ISSN: 1683-1489
Mekong River Commission
The Stationary Trawl (Dai) Fishery
of the Tonle Sap-Great Lake System,
Cambodia
MRC Technical Paper
No. 32
August 2013
Ca m b o d i a . L a o P D R . Th a i l a n d . Vi e t N a m
For sustainable development
Page 1
Mekong River Commission
The Stationary Trawl (Dai) Fishery
of the Tonle Sap-Great Lake System,
Cambodia
MRC Technical Paper
No. 32
August 2013
Ca m b o d i a . L a o P D R . Th a i l a n d . Vi e t N a m
For sustainable development
Published in Phnom Penh, Cambodia in August 2013 by the Mekong River Commission
Cite this document as:
Halls, A.S.; Paxton, B.R.; Hall, N.; Peng Bun, N.; Lieng, S.; Pengby, N.; and So, N (2013).
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake, Cambodia. MRC Technical Paper
No. 32, Mekong River Commission, Phnom Penh, Cambodia, 142pp. ISSN: 1683-1489.
The opinions and interpretations expressed within are those of the authors and do not necessarily
reflect the views of the Mekong River Commission.
Cover Photo: J. Garrison
Editors: K.G. Hortle, T. Hacker, T.R. Meadley and P. Degen
Graphic design and layout: C. Chhut
Office of the Secretariat in Phnom Penh (OSP)
576 National Road, #2, Chak Angre Krom,
P.O. Box 623,
Phnom Penh, Cambodia
Tel. (855-23) 425 353
Fax. (855-23) 425 363
Office of the Secretariat in Vientiane (OSV)
Office of the Chief Executive Officer
184 Fa Ngoum Road, P.O. Box 6101,
Vientiane, Lao PDR
Tel. (856-21) 263 263
Fax. (856-21) 263 264
© Mekong River Commission
E-mail: [email protected]
Website: www.mrcmekong.org
Table of contents
List of tables ........................................................................................................................................... vi
List of figures.......................................................................................................................................... ix
List of Plates......................................................................................................................................... xiv
Acknowledgements . ............................................................................................................................. xv
Abbreviation and acronyms.................................................................................................................. xvi
Glossary............................................................................................................................................... xvii
Glossary of parameters (from formulas used in the paper)................................................................... xxi
Abstract ............................................................................................................................................... xvii
1. Introduction .........................................................................................................................................1
1.1. The Tonle Sap-Great Lake (TS-GL) System.............................................................................1
1.1.1. Location, origins and physiography..............................................................................1
1.1.2. Hydrology and hydrodynamics.....................................................................................3
1.1.3. Ecology of the Tonle Sap..............................................................................................5
Biodiversity and endemism...........................................................................................5
The flood pulse..............................................................................................................6
1.1.4. Fisheries of the TS-GL System.....................................................................................8
Threats to fishery resources.........................................................................................12
1.2. The Tonle Sap dai fishery . .....................................................................................................12
1.3. The purpose and scope of the paper........................................................................................13
1.4. The structure of the report.......................................................................................................14
2. The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System . ..........................................................15
2.1. The Location of the fishery.....................................................................................................15
2.2. Target species..........................................................................................................................18
2.3. Fishing gear and operation......................................................................................................24
2.4. Fish disposal............................................................................................................................30
2.4.1. Prahok.........................................................................................................................30
iii
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2.5. Economics ............................................................................................................................31
2.6. Management............................................................................................................................35
2.6.1. Legislation...................................................................................................................35
2.6.2. Closed season..............................................................................................................35
2.7. Effort control/Licensing..........................................................................................................36
2.8. Gear restrictions......................................................................................................................37
3. Routine monitoring activities and survey methodologies.................................................................39
3.1. Purpose of monitoring.............................................................................................................39
3.2. A history of survey methodologies and sampling regimes......................................................39
3.3. Spatial and temporal variation in sampling effort...................................................................43
3.4. Current survey methodology...................................................................................................49
3.4.1. Sample stratification....................................................................................................49
3.4.2. Data collectors and data collection process................................................................50
3.4.3. Variables enumerated .................................................................................................50
3.5. Catch estimation methodology................................................................................................52
3.5.1. Aggregated total catch................................................................................................52
3.5.2. Species-wise CPUE and catch....................................................................................54
4. The Dai Fishery database: storage and processing............................................................................55
4.1. Database evolution..................................................................................................................55
4.2. Description of the current database.........................................................................................58
4.2.1. Data tables...................................................................................................................58
4.2.2. Lookup Tables.............................................................................................................59
4.2.3. Length Frequency tables.............................................................................................60
4.2.4. Additional tables.........................................................................................................60
4.3. Current Database Query descriptions......................................................................................60
5. The Ecology of the Fishery................................................................................................................61
5.1. Longitudinal (upstream/downstream) variation......................................................................61
5.1.1. Aggregated catch, effort and CPUE............................................................................61
5.1.2. Catch diversity and composition . ..............................................................................64
5.1.3. Fish size (weight)........................................................................................................70
5.2. Variation in catch rates among individual dais........................................................................71
5.3. Intra-annual variation..............................................................................................................77
5.3.1. Abundance (CPUE).....................................................................................................77
5.3.2. Species diversity and similarity .................................................................................83
iv
Table of contents
5.4.
5.5.
5.3.3. Size (weight) ..............................................................................................................91
Inter-annual variation..............................................................................................................93
5.4.1. Catch, effort and CPUE...............................................................................................93
5.4.2. Catch diversity and species composition....................................................................93
5.4.3. Size (weight).............................................................................................................101
Hydrological effects on fish biomass....................................................................................103
6. Summary and Conclusions.............................................................................................................. 113
6.1. Management Implications..................................................................................................... 117
6.2. Recommendations................................................................................................................. 119
6.2.1. Monitoring fisheries resources in the TS-GL and beyond........................................ 119
6.2.2. Recommendations to improve the existing dai fishery monitoring programme . ....120
6.2.3. Further research.........................................................................................................121
7. References .......................................................................................................................................123
8. Annex ..............................................................................................................................................131
8.1. Dai Fishery principal data tables (Cans and Ngor, 2006).....................................................133
8.2. 2009 Dai Fishery database principal data tables...................................................................137
Page v
List of tables
Table 1 Dai locations in the Tonle Sap during the 2007–08 season . ................................................. 16
Table 2 Mean (and range) dimensions, mesh sizes, and periods of operation of each dai net type . . 24
Table 3 Reasons reported by dai operators to mechanise hauling operations . .................................. 29
Table 4 Dai Fishery profit estimates 1999–2000................................................................................ 32
Table 5 Dai unit licence allocation by type for the 2007–08 season .................................................. 36
Table 6 Maximum gear dimensions and minimum mesh sizes for the three types of dai nets . ........ 37
Table 7 Estimates of catches prior to and including the 1994–95 seasons ........................................ 41
Table 8 Monthly dai sampling regime planned for the 1997–98 season............................................ 42
Table 9 Monthly dai sampling regime proposed for the 2000–01 fishing season ............................. 44
Table 10 Monthly dai sampling regime planned for the 2002–03 fishing season ............................... 45
Table 11 Summary of changes made to the dai fishery survey design (1994–2008)
and numbers of dais sampled each season............................................................................. 46
Table 12 The relative locations of High (shaded cells) and Low Yield (unshaded cells) dais ............. 49
Table 13 Description of the main tables in the dai fishery database developed ................................... 57
Table 14 Description of the main tables in the dai fishery database 2009............................................ 59
Table 15 Estimates of aggregated catch and effort by season and municipality................................... 61
Table 16 Estimates of fishing mortality, (F) for the dai fishery for a range of fishing efforts.............. 64
Table 17 Species contributing up to 80% of mean cumulative percentage (Cum%) dissimilarity
between Phnom Penh Municipality and Kandal Province in each season............................. 69
Page vi
List of tables
Table 18 ANOVA test results for the null hypothesis that the average weight of fish
(irrespective of species) landed by the fishery does not change with dai row number
after accounting for the effects of month, season and mesh size .......................................... 71
Table 19 Estimates of water velocity and water depth below each dai recorded in February 2008 .... 72
Table 20 The seasonal frequency of upward (↑) and downward (↓) changes in the abundance
of common species between consecutive months, 1997–98 to 2008–09............................... 88
Table 21 Estimated catches of the most abundant species reported for the years 1938–39.................. 94
Table 22 Principal families and species making up 99% of the dai catches listed in descending
order of their total percentage contribution............................................................................ 98
Table 23 Summary of the indices used to describe the various attributes of the hydrograph
hypothesised to affect fish biomass...................................................................................... 106
Table 24 ANOVA table for the GLM model....................................................................................... 107
Table 25 Summary from the existing dai database of the number of distinct dais sampled
in each of the different strata between 1994 and 2008 ........................................................ 131
Table 26 tbl_MainWi&Dos................................................................................................................. 133
Table 27 tbl_SpeciesWin&Dos........................................................................................................... 133
Table 28 tbl_Effort.............................................................................................................................. 134
Table 29 tlkp_SeasonYear................................................................................................................... 134
Table 30 tlkp_GearCode...................................................................................................................... 134
Table 31 tlkp_Species.......................................................................................................................... 134
Table 32 tlkp_SpeciesStandard........................................................................................................... 135
Table 33 tbl _Phnom Penh Port Water level........................................................................................ 136
Table 34 tbl _MoonFace...................................................................................................................... 136
Page vii
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 35 Alphanumeric gear codes in the Access database incorporating to lunar phase
and dai yield (B0001–B004)................................................................................................ 136
Table 36 tbl_MainWi&Dos................................................................................................................. 137
Table 37 tbl_SpeciesWin&Dos........................................................................................................... 138
Table 38 tbl_Effort.............................................................................................................................. 138
Table 39 tlkp_GearCode...................................................................................................................... 138
Table 40 tlkp_Species.......................................................................................................................... 139
Table 41 tlkp_SpeciesStandard........................................................................................................... 139
Table 42 Leng_tbllengthFreq (Asterisks indicate name changes or new fields)................................. 140
Table 43 LengtblSpecies (Asterisks indicate name changes or new fields)........................................ 141
Table 44 Leng_tblLocation (Asterisks indicate name changes or new fields).................................... 141
Table 45 tbl_LunarAge&Phase* (Asterisks indicate name changes or new fields)............................ 141
Table 46 tblOtherInfo* (new table) (Asterisks indicate name changes or new fields)........................ 141
Table 47 Annual Hydrological Indices* (Asterisks indicate name changes or new fields)................ 142
Table 48 tbl_LunarAge&Phase* (Asterisks indicate name changes or new fields)............................ 142
Page viii
List of figures
Figure 1 The Tonle Sap–Great Lake System.........................................................................................2
Figure 2 Mean, maximum and minimum inflow (negative values) and outflow
(positive values) discharges from the Tonle Sap measured in m3 s-1 at Prek Kdam
Between 2004 and 2008..........................................................................................................3
Figure 3 Monthly mean water level (amsl) and inundated lake area . ..................................................4
Figure 4 Mean monthly TSS flux, concentration and water levels in the Lake for
the seasons 1997–2003 . .........................................................................................................7
Figure 5 Fishing Lots in Cambodia ....................................................................................................10
Figure 6 The location of dai fishery in the Tonle Sap..........................................................................17
Figure 7 The species composition of the dai catch in 2009-10...........................................................18
Figure 8 Main structures of the dai fishery in the Tonle Sap . ............................................................26
Figure 9 Reported mean minimum and maximum mesh sizes of each net type by dai row.............. 27
Figure 10 Cumulative number of dai operators using diesel engines to close and haul their nets (top)
and (bottom) cumulative total engine horsepower (HP) employed in the dai fishery
since 1999..............................................................................................................................29
Figure 11 Fish disposal pathways for low value fish (LVF) or small size fish species.........................30
Figure 12 Actual (upper line) and inflation adjusted (lower line) unit price of trey riel ..................... 33
Figure 13 Mean values of key economic variables for the Cambodian dai fishery plotted
as a function of dai row number ...........................................................................................34
Figure 14 License costs for 25 dais plotted as a function of their annual profit . .................................35
Figure 15 Number of licensed dais . .....................................................................................................37
Page ix
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 16 Changes in sampling effort between the 1994–95 and 2008–09 expressed as the total
number of hauls sampled in each season...............................................................................46
Figure 17 Total number of hauls sampled each (a) month and (b) lunar phase
(Quarters 1–4) each season................................................................................................... 47
Figure 18 (a) Total number and (b) mean number of hauls per dai (±SD)
sampled by row each season.................................................................................................48
Figure 19 Outline of the stratification of the dai fishery sampling regime . .........................................53
Figure 20 Structure of the database files held in the DOS version ARTFISH showing
folders tiered by Administrative Province, season and month .............................................55
Figure 21 Structure of the database files held in the Windows version ARTFISH showing
folders tiered by Season, Administrative province, month and file type ..............................56
Figure 22 Entity-relationship diagram for the dai fishery database developed ....................................57
Figure 23 Entity relationship diagram for the Dai fishery database in use from 2009..........................60
Figure 24 Mean sampled catch rates (1997–2009) plotted by row .......................................................62
Figure 25 The Delury depletion model fitted to mean sampled catch rates (1997–2009)
expressed as the number of fish caught per dai unit per day, and cumulative fishing
effort measured in dai units from the most upstream row 15................................................63
Figure 26 Estimates of (a) species richness (S) (Mean±SE) and (b) Shannon Diversity Index (H)
(Mean±SE) for each row from Row 1 (downstream) to 15 (upstream)
for the seasons 1997–98 to 2008–09 averaged across months and lunar phases..................66
Figure 27 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root
transformed mean CPUE for individual dais in January during the peak lunar phase
(second quarter), 1997–98 to 2002–03 seasons.....................................................................67
Figure 28 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root
transformed mean CPUE for individual dais in January during the peak lunar phase
(second quarter), 2003–04 to 2008–09..................................................................................68
Figure 29 Estimates of mean weight of the fish assemblage sampled from the dai fishery
by row and season................................................................................................................ 70
Page x
List of figures
Figure 30 Estimates of mean depth below each dai row.......................................................................73
Figure 31 Estimates of mean water velocity at each dai row................................................................74
Figure 32 The relationship between water velocity and depth .............................................................74
Figure 33 Mean loge-transformed catch rates during each lunar quarter (1–4) of the survey month
(February 2008) plotted as a function of the estimated water depth below the dai unit.......75
Figure 34 Mean loge-transformed catch rates during each lunar quarter (1–4) of the survey month
(February 2008) plotted as a function of the estimated water velocity at each dai unit.......76
Figure 35 Mean dai CPUE (kg) calculated for each day of the fishing season during
(a) 1997–98, (b) 1998–99 and (c) 1999–2000 . ....................................................................78
Figure 36 Mean dai CPUE (kg) calculated for each day of the fishing season during (a) 2000–01,
(b) 2001–02 and (c) 2002–03................................................................................................79
Figure 37 Mean dai CPUE (kg) calculated for each day of the fishing season during
(a) 2003–04, (b) 2004–05 and (c) 2005–06 . ........................................................................80
Figure 38 Mean dai CPUE (kg) calculated for each day of the fishing season during
(a) 2006-07, (b) 2007–08 and (c) 2008–09 ..........................................................................81
Figure 39 Loge-transformed average daily catch rates (kg/dai/day) by (a) month and
(b) lunar phase for the seasons 1997–98 to 2008–09 ...........................................................82
Figure 40 Estimates of mean (±95% CI) (a) species richness (S) and (b) Shannon Diversity
Index (H) (Mean ±95% CI) of the assemblage sampled during phase 2 of the lunar
cycle of each month and season ...........................................................................................84
Figure 41 Estimates of mean (±95% CI) (a) species richness and (b) the Shannon Diversity
Index (H) of the assemblage sampled during each quarter (phase) of the lunar cycle
in January of each season .....................................................................................................85
Figure 42 MDS ordinations illustrating rank similarities in the species assemblage sampled
each month during lunar phase 2, 1997–98 to 2002–03 ......................................................86
Figure 43 MDS ordinations illustrating rank similarities in the species assemblage sampled
each month during lunar phase 2, 2003–04 to 2008–09 ......................................................87
Page xi
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 44 MDS ordinations illustrating rank similarities between the species assemblage
sampled during each lunar phase of January, 1997–98 to 2002–03 .....................................89
Figure 45 MDS ordinations illustrating rank similarities between the species assemblage sampled
during each lunar phase of January, 2003–04 to 2008–09....................................................90
Figure 46 Changes in mean fish weight (all species combined) through the fishing season
(1997–08 to 2008–09) ..........................................................................................................91
Figure 47 Mean weight changes by month for the six most abundant species
(1997–08 to 2008–09)...........................................................................................................92
Figure 48 (a) Total Catch, (b) effort and (c) CPUE (1997–8 to 2008–9)..............................................93
Figure 49 The species composition of the dai fishery catch, 1997–98 to 2008–09...............................95
Figure 50 (a) The Species Richness (S) (Mean ±95% CI) and (b) Shannon Diversity Indices (H)
(Mean ±95% CI) of fish species caught by a dai on a day for the seasons
1997–98 to 2008–09 averaged across all sampling months in each season..........................97
Figure 51 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root
transformed mean CPUE for individual dais for the 1997–98 to 2008–09 seasons
for (a) October and (b) November.........................................................................................99
Figure 52 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root
transformed mean CPUE for individual dais for the 1997–98 to 2008–09 seasons
for (a) December and (b) January........................................................................................100
Figure 53 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root
transformed mean CPUE for individual dais for the 1997–98 to 2008–09 seasons
for (a) February and (b) March............................................................................................101
Figure 54 Mean weight of fish landed (all species combined) 1997–2009.........................................102
Figure 55 Trends in the mean weight of the six most abundant species in the fishery
estimated for December each year (1995–06 to 2008–09)..................................................102
Figure 56 The location of the Kampong Luong gauging station in the TS-GL...................................104
Figure 57 Daily water level (m) measured at Kampong Luong gauging station and estimates
of the flooded area of the TS-GL System, 1997–2009........................................................105
Page xii
List of figures
Figure 58 Illustration of the estimation of hydrological indices. . ......................................................105
Figure 59 The GLM observed and predicted mean catch rates 1997–2009 .......................................108
Figure 60 Model Residuals..................................................................................................................108
Figure 61 The relationship between the mean predicted daily catch rate of a dai unit
during the fishing season and the flood index (FI) for the TS-GL System.........................109
Figure 62 The relationship between mean sampled fish weight (all species combined)
and the flood index with fitted exponential mode ..............................................................109
Figure 63 Mean weight estimates for the six most abundant species in the fishery
in December each year........................................................................................................110
Figure 64 Changes in mean fish weight and the flood index from 1997–98 to 2008–09....................110
Figure 65 Fish abundance plotted as a function of the flood index.....................................................111
.
Page xiii
List of plates
Plate 1 Arrow fence traps constructed in the flooded forest margins of the Lake .............................. 11
Plate 2 Rows of dai nets anchored across the Tonle Sap ...................................................................15
Plate 3 Cirrhinus lobatus (trey angkam) ............................................................................................19
Plate 4 Henicorhynchus cryptopogon (trey riel angkam)....................................................................19
Plate 5 Paralaubuca barroni (trey slak russey) . ................................................................................20
Plate 6 Labiobarbus lineatus (trey khnawng veng) ............................................................................20
Plate 7 Henicorhynchus siamensis (trey riel tob) ...............................................................................21
Plate 8 Labiobarbus siamensis (trey ach kuk) ....................................................................................21
Plate 9 Labeo chrysophekadion (trey kaek) . ......................................................................................22
Plate 10 Pangasius pleurotaenia (try chhwiet) ....................................................................................22
Plate 11 Puntioplites proctozysron (trey chrakeng) .............................................................................23
Plate 12 Thynnichthys thynnoides (trey linh) .......................................................................................23
Plate 13 Emptying the codend into sorting compartments . .................................................................28
Plate 14 The codend being emptied directly into traders boats . ..........................................................28
Plate 15 Diesel engines used to haul the dai net ..................................................................................28
Page xiv
Acknowledgements
This paper is an output of the Fisheries Ecology Valuation and Mitigation (FEVM) Component of the
MRC’s Fisheries Programme. The paper aims to improve stakeholder capacity to monitor and evaluate
the status and trends of fisheries resources in the Lower Mekong Basin (LMB) in the context of
fisheries management and basin development activities (FEVM Logframe Outputs 1, 3 & 5).
The paper also forms an output of the ACIAR-funded project: Analyses of three databases of
fisheries data from the Mekong River (FIS/2006/137) - a collaborative research project between the
MRC, IFReDI, LARReC and Murdoch University, funded by ACIAR.
The routine fishery monitoring activities described in Section 3.4 were initiated by Niek van
Zalinge and later managed by Kent G. Hortle, Niklas Mattson and then Ashley Halls. The field work
was supervised by Ngor Peng Bun, Deap Loueng, Yim Chea, Heng Kong, Chhoun Chamnan, Souen
Sotthia and Lieng Sopha. Ngor Pengby and Tan Phalla have managed and improved the database for
the past five years. Their outstanding efforts, including those of the many field enumerators, and the
cooperation of the dai operators, are gratefully acknowledged.
We are grateful to Kent G. Hortle and Nao Thuok for their reviews and comments which
significantly improved an earlier draft of the paper.
The preparation of this paper was facilitated by the MRC Fisheries Programme with funding
from DANIDA, SIDA and ACIAR.
Page xv
Abbreviations and acronyms
TAMCF
AMSL
ANOVA
ARTFISH
CPUE
DoF
DOS
FEVM
FI
FiA
IFReDI
LMB
LVF
MDS
MFCF
MFD
MRC
MRCS
PRIMER
SD
SE
SIMPER
TSBR
TS-GL
TSS
UNESCO
Page xvi
Assessment of the Mekong Capture Fisheries (Programme)
Above Mean Sea Level
Analysis of Variance
Approaches, Rules and Techniques for Fisheries Statistical Monitoring (FAO Database)
Catch Per Unit of Effort
Department of Fisheries
Disk Operating System
Fisheries Ecology Valuation and Mitigation
Flood Index
Fisheries Administration
Inland Fisheries Research and Development Institute
Lower Mekong Basin
Low Value Fish
Multi-dimensional Scaling
Management of the Freshwater Capture Fisheries (Programme)
Mekong Fish Database
Mekong River Commission
Mekong River Commission Secretariat
Plymouth Routines in Multivariate Ecological Research (software)
Standard Deviation
Standard Error
Similarity Percentage (sub-routine in PRIMER)
Tonle Sap Biosphere Reserve
Tonle Sap-Great Lake
Total Suspended Solids
United Nations Educational Scientific and Cultural Organisation
Glossary
Analysis of variance
Analysis of variance (ANOVA) is a collection of statistical
models, and their associated procedures, in which the observed
variance in a particular variable is partitioned into components
attributable to different sources of variation. In its simplest
form ANOVA provides a statistical test of whether or not
the means of several groups are all equal, and therefore
generalizes t-test to more than two groups. Doing multiple
two-sample t-tests would result in an increased chance of
committing a type I error. For this reason, ANOVAs are useful
in comparing two, three or more means.*
ARTFISH
A standardized tool adaptable to most fisheries in the
developing countries. Its design was driven by the need
to provide users with robust, user-friendly and errorfree approaches with computer software, and achieve the
implementation of cost-effective fishery statistical systems
with minimal external assistance.**
Blackfish
Species that possess morphological and physiological
adaptations to extreme environmental conditions including low
dissolved oxygen concentrations, and desiccation.
Bray-Curtis (dis)similarity Index
The Bray–Curtis dissimilarity is a statistic used to quantify
the compositional dissimilarity between two different sites. It
is equivalent to the total number of species that are unique to
any one of the two sites divided by the total number of species
over the two sites. In other words, it is the ratio between the
turnover of species between the two sites and the total species
richness over the two sites.*
Catchability coefficient
The proportion of the population removed by one unit of
effort.
Coefficient
A multiplicative factor in some term of an expression.
Page xvii
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Coefficient of determination
The coefficient of determination R2 is used in the context of
statistical models whose main purpose is the prediction of
future outcomes on the basis of other related information. It
is the proportion of variability in a data set that is accounted
for by the statistical model. It provides a measure of how well
future outcomes are likely to be predicted by the model.*
Correlation coefficient
A measure of the strength of the linear relationship between
two variables.*
Delury Depletion Model
A method to estimate animal abundance by monitoring how
indices of abundance (e.g. catch rates) decline in response to
cumulative fishing effort.*****
Flood Index
A quantitative description of the extent and duration of
flooding corresponding to the area beneath and the areaduration curve above mean flood levels.
General Linear Model (GLM)
The general linear model incorporates a number of different
statistical models. * In this document GLM provides a general
version of multiple linear regression where explanatory
variables take the form of factors and covariates.
Lunar phase or (lunar quarter)
Lunar quarters relate to four consecutive seven day periods
starting from the new (dark phase) moon. Quarter 2, when
catch rates in the dai fishery are observed to peak, corresponds
to the period of approximately 7–14 days after the new moon
when between approximately 50–100 % of the moon is visible.
This period between what are commonly termed the first
quarter and full moon phases is also known as the ‘Waxing
Gibbous’ phase.
Multi-dimensional scaling (MDS)
Multidimensional scaling (MDS) is a set of related
statistical techniques often used in information
visualization for exploring similarities or dissimilarities
in data. MDS is a special case of ordination. An MDS
algorithm starts with a matrix of item–item similarities,
then assigns a location to each item in N-dimensional
space, where N is specified a priori. For sufficiently small
N, the resulting locations may be displayed in a graph or
3D visualisation.*
Page xviii
Glossary
Population dynamics
Population dynamics is the branch of life sciences that
studies short-term and long-term changes in the size
and age composition of populations, and the biological
and environmental processes influencing those changes.
Population dynamics deals with the way populations are
affected by birth and death rates, and by immigration and
emigration, and studies topics such as ageing populations
or population decline.*
Primary production
Primary production is the production of organic
compounds from atmospheric or aquatic carbon dioxide,
principally through the process of photosynthesis, with
chemosynthesis being much less important. Almost
all life on earth is directly or indirectly reliant on
primary production. The organisms responsible for
primary production are known as primary producers or
autotrophs, and form the base of the food chain.*
PRIMER
(Plymouth Routines In Multivariate Ecological Research
Version 6): PRIMER
6 is a collection of specialist routines for analyzing
species or sample abundance (biomass). It is primarily
used for ecological and environmental studies.
Multivariate routines include:
• grouping (CLUSTER);
• sorting (MDS);
• principal component identification (PCA);
• hypothesis testing (ANOSIM);
• sample discrimination (SIMPER);
• trend correlation (BEST);
• comparisons (RELATE); and
• diversity, dominance, and distribution calculating. ***
Recruitment
The number of fish (recruits) added to the exploitable
stock, in the fishing area, each year, through a process
of growth (i.e. the fish grows to a size where it becomes
catchable) or migration (i.e. the fish moves into the
fishing area).**
Page xix
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Shannon Diversity Index
The Shannon Diversity index, sometimes referred to as
the Shannon-Wiener Index is one of several diversity
indices that can be used to measure species diversity.
The advantage of this index is that it takes into account
the number of species and the evenness of the species.
The index is increased either by having additional unique
species, or by having a greater species evenness.*
SIMPER
Identifies the species primarily providing the
discrimination between two observed sample clusters.***
(See PRIMER)
Species richness
Species richness is the number of different species in
a given area. It is represented in equation form as S.
Typically, species richness is used in conservation studies
to determine the sensitivity of ecosystems and their
resident species. The actual number of species calculated
alone is largely an arbitrary number.*
Standard deviation
Standard deviation shows how much variation or
"dispersion" exists from the average. A low standard
deviation indicates that the data points tend to be very
close to the mean, whereas high standard deviation
indicates that the data points are spread out over a large
range of values. The standard deviation of a statistical
population, data set, or probability distribution is the
square root of its variance.*
Standard error
An estimate of that standard deviation, derived from a
particular sample used to compute the estimate*
Survey stratification
The process of dividing members of the population into
homogeneous subgroups (stratum) before sampling to
reduce sample variance.
Type I Error
Falsely rejecting the null hypothesis when it is true.
Whitefish
Migratory species intolerant of low dissolved oxygen
conditions and typically inhabit lotic (flowing water)
environments.
*http://en.wikipedia.org
** http://www.fao.org
*** http://www.primer-e.com/index.htm
***** Hilborn & Walters (1992)
Page xx
Glossary of parameters (from formulas used in the paper)
Α, a
Β, b
AD
AG
C
CPUE
D
d
DSI
e
E
e
ε
F
FA
FC
FE
FS
h
H
KAN
lp
m
N
p
p
pi
PP
q
r
r
R2
S
TL
TVC
VC
WL
wt
y
Constant
Coefficient
Active Days (fishing days)
Active Gears (dai units)
Catch
Catch Per Unit of Effort
Flood duration
A given day of the year
Dry Season Index
Effort (dai units)
(Cumulative) fishing effort
Exponential
Error (residual)
Instantaneous fishing mortality rate
Flooded Area
Fixed Costs
Flood End
Flood Start
Haul
Shannon Diversity Index
Kandal
Lunar Phase (1–4)
Calendar month
Fish abundance (number of fish)
Administrative zone (Province)
Probability of committing a Type I Error
Relative abundance of species i
Phnom Penh
Catchability coefficient
Dai row
Correlation coefficient
Coefficient of determination
Species richness
Total Length
Total Variable Costs
Variable Costs
Water Level
Sampled haul weight
Yield or year
Page xxi
Abstract
The Tonle Sap-Great Lake (TS-GL) system is an integral part of the history, culture, ecology and
economics of the Mekong region. Species of blackfish and whitefish are the target of industrial,
artisanal and subsistence fisheries operating in the TS-GL System. Strong competition exists among
the fisheries to land in excess of 200,000 tonnes of fish each year equivalent to approximately 10%
of the total weight of fish consumed in the entire Lower Mekong Basin (LMB) each year. The dai
fishery on the Tonle Sap, established almost 140 years ago, is an important component of the industrial
fishery, landing approximately 14% of the annual catch taken from the TS-GL System. It targets the
refuge migrations of a multi-species assemblage of fish as they migrate from the Great Lake to the
Mekong main channel via the Tonle Sap with the receding floodwaters each year. In addition to its
significant socio-economic value both locally and nationally, the dai fishery provides a valuable source
of data and information to monitor trends in migratory fish populations which seasonally utilise the
TS-GL System and beyond. This paper represents the first attempt to compile and analyse the
available data and information about the Cambodian dai fishery in a single document. It therefore
serves as an important reference document for present and future workers involved in the
management, monitoring and administration of the fishery. It also contains new insights into the
ecology and dynamics of target fish populations important for their management. The fishery exhibits
considerable spatial and temporal variation in catch rate indicators of fish biomass and abundance.
This reflects (pulsed) migratory behaviour associated with the lunar cycle, the hydrological cycle
(drawdown effects), depletion effects as fish migrate through the fishery, and inter-annual hydrological
effects on fish growth and biomass. Above average levels of recruitment were probably responsible
for the very high catches observed during 2004–05 and 2005–06. Factors responsible for these high
levels of recruitment remain uncertain. There is evidence that the timing of migrations is species
and size-dependent with larger species and larger individuals of the same species migrating earlier
than smaller fish. These responses are consistent with earlier studies on the system and in other
tropical river systems. Whilst the dai fishery is the focus of most fisheries monitoring and evaluation
efforts in Cambodia, it is not the only, nor most significant, component of the entire TS-GL fishery.
Other components with which it interacts and competes with, particularly the other lot, artisanal
and subsistence fisheries, are also significant and therefore must be given greater consideration in
the future. In spite of the present restricted focus, the monitoring efforts directed at the dai fishery
have generated the only continuous long-term data set for an inland fishery in Cambodia. Analyses
of indicators estimated from this data set described here have been informative for policy and
management evaluation, revealing little or no compelling evidence of changes in the abundance,
biomass, size or diversity of migratory fish populations that seasonally utilise the TS-GL System and
beyond often over distances of more than 600 km. Furthermore, time series of these indicators have
equipped managers with an important baseline against which to monitor any impacts of management
and basin development activities. A key finding of this research is that inter-annual variation in the
biomass of the multispecies assemblage targeted by the fishery (and hence landings) can be largely
explained by flood duration and extent effects on fish growth. Fish growth, indicated by mean fish
Page xxii
Abstract
weight, increases exponentially with the flood extent and duration presumably reflecting changes in
feeding opportunities or competition. This response has been modelled, allowing predictions to be
made of how the relative biomass of the multispecies assemblage targeted by the fishery (and hence
catches) are likely to vary under different flooding conditions, whether natural or modified as a
consequence of climate change and/or water management projects in the Basin. Owing to the highly
migratory nature of the target fish species, these predicted hydrological responses may be observed
over large distances, affecting fisheries and piscivorous fish populations beyond the immediate vicinity
of the system. The unexplained variation in the model may well reflect variation in recruitment to the
system each year in addition to variation in fishing effort (mortality) applied by the other important
fisheries within the TS-GL System or over the migratory range of the target species, reinforcing
the need for two more comprehensive monitoring programmes. These results also urge caution
when monitoring mean fish size as a proxy for rates of exploitation in the TS-GL System and other
highly fluctuating environments. By applying depletion model theory, this research has provided the
first estimates of the proportion of fish removed over the range of the fishery, dai gear catchability
(efficiency), and dai fishing mortality rates subject to a number of assumptions. These results are
an important first step towards understanding, and thereby controlling if necessary, the relative
sources of fishing effort (mortality) over the migratory range of populations of important species of
fish. Additional studies and monitoring programmes will be necessary to determine the validity of
the assumptions underlying these estimates and to quantify the spatial distribution of the remaining
sources of fishing mortality in the TS-GL System and beyond. Recommendations are made for these
studies and programmes, as well as to improve the existing dai fishery monitoring programme.
xxiii
1. Introduction
1.1. The Tonle Sap-Great Lake (TS-GL) System
Rising in the Himalayan highlands of China, the Mekong River flows through Myanmar, Lao PDR,
Thailand and Cambodia before discharging into the South China Sea in Viet Nam. At 4,880 km, it is
the 12th longest river in the world and 10th largest in terms of water volume (Gupta and Liew 2007;
Kummu and Sarkkula, 2008). In Cambodia it empties onto a vast alluvial plain where its channel
becomes less well-defined and where more hydraulically complex floodplain processes prevail
(MRC, 2005). Roughly 4,300 km from its source at the Chaktomuk junction near the city of Phnom
Penh it is met on its right bank by the Tonle Sap. The Great Lake–the largest wetland in Southeast
Asia (Kummu et al., 2008)–is situated at the apex of the Tonle Sap around 130 km to the northwest of
this confluence. These two water bodies–the Tonle Sap coupled to the Great Lake–form the Tonle SapGreat Lake (TS-GL) system.
The Tonle Sap-Great Lake system is an integral part of the history, culture, ecology and economics
of Southeast Asia. It plays a crucial role in mitigating floods in Cambodia and Viet Nam, particularly
in the case of extreme flood events (Hai et al., 2008), provides habitats for a diversity of aquatic and
terrestrial plant and animal species (Campbell et al., 2006) and an important source of raw materials,
nutrition, income and livelihoods for upwards of one million people living in and around it
(Keskinen, 2003; Sarkkula et al., 2003 and Lamberts, 2006). Together with rice cultivation, fishing is
amongst the most important of these livelihood activities (Keskinen, 2003; Ahmed et al., 1996) and
areal yields from the Lake have been estimated to be 230 kg ha-1 year-1 (Baran et al., 2001).
1.1.1. Location, origins and physiography
The TS-GL System is one of the most distinctive hydro-geomorphic features within the larger Mekong
River basin. The Great Lake depression that today comprises the Lake itself was believed to have
formed through the subsidence of the Cambodian platform around 5,700 years ago (Carbonel, 1963
cited in Campbell et al., 2006). Higher-than-present rainfall and sea level transgression suggest that, at
this time, the depression was inundated and that there was less seasonal variation in depth and possibly
tidal influence in the vicinity of the modern lake (Penny, 2006). From the mid-Holocene onwards,
however, a weakening of the southwest (wet) monsoon led to drier, more seasonal climate. These
factors, combined with sea level regression would have contributed to hydrological and hydrodynamic
conditions in the Lake becoming similar to what they are today (Penny 2006; Penny 2008).
The catchment of the Lake basin covers an area of around 67,000 km2 (Ahmed et al., 1996),
bounded to the north by the Dongrek mountain range and to the southwest by the Cardamom and
Page 1
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Elephant ranges. It is separated into a northern and a southern basin with a deltaic region at its
southern end formed by the deposition of sediments brought in from the Mekong by the Tonle Sap
(Figure 1).
Figure 1
Page 2
The Tonle Sap–Great Lake System.
Introduction
1.1.2. Hydrology and hydrodynamics
During the dry season (October–May) water drains from the TS-GL System into the Mekong River
via the Tonle Sap. As the wet season advances from June/July onwards, however, flooding in the delta
downstream of Phnom Penh causes water levels in the Mekong River to rise higher than those in the
Great Lake. This causes flow in the Tonle Sap to reverse, and instead of draining into the Mekong, the
waters are pushed back upstream towards the Great Lake, inundating its floodplains. At the peak of the
flood the aerial extent of the Lake will increase between three and six times (van Zalinge et al., 2004).
Towards the end of the flood, backed-up waters in the Lake and concurrently subsiding water levels in
the Mekong, cause the flow in the Tonle Sap to reverse once more. The waters are then carried out of
the Lake, back into the Mekong River and towards the delta.
By the height of the flood season in August, discharges through the Tonle Sap into the Great Lake
peak at approximately 10,000 m3 s-1 or roughly 25% of the mean mainstream flows at this time of year
(MRC, 2005). After the peak of the flood, in November, discharges once more decline and the backed
up waters in the Great Lake begin to drain back into the Mekong at similar rates (10,000 m3 s-1) to the
inflow (MRC, 2005) (Figure 2).
Figure 2
Mean, maximum and minimum inflow (negative values) and outflow (positive values) discharges
from the Tonle Sap measured in m3 s-1 at Prek Kdam Between 2004 and 2008.
Page 3
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
At the height of the flood, the inundated area of the Great Lake increases from approximately
3,500 km2 during the dry season to 14,500 km2 as water levels rise over the wet season (Figure 3,
Kummu et al., 2008). Over this same period the Lake volume will increase from 1.5 km3 to 60–70 km3
(MRC, 2005).
14000
area
12000
water level
Area [km2 ]
10000
12
10
8000
8
6000
6
4000
4
2000
2
Rising flood
Water level [m], AMSL in Hatien
14
Receding flood
0
Apr
Mar
Feb
Jan
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
0
Figure 3
Monthly mean water level (amsl) and inundated lake area.
Error bars represent minimum and maximum mean monthly water levels (Kummu et al., 2008).
The mean annual inflow to the Tonle Sap is 79 km3–roughly comparable with its outflow (78.6
km3) and it exhibits high inter-annual variability (44.1 km3 in 1998 and 106.5 km3 in 2000) (Kummu
and Sarkkula, 2008). Minimum depths of 0.5 m in April increase to depths of around 6–9 m at the
height of the flood between late September and early October (MRC, 2005). Campbell et al. (2006)
noted a statistically significant decline in the maximum and minimum water levels of 0.52 m between
the years 1925–35 and 1996–2002. This translates to reduction in inundated area of 623 km2. They rule
out the possibility of gauging error to explain this decline and suggest instead that a reduced rainfall
and a minor reduction in flows from the Tonle Sap may be responsible.
Of the flows entering the Lake, 52% originate from the Mekong River via the Tonle Sap and 5%
from overland flows from the mainstream over the flood season. The remainder of the flows originates
from the Lake’s tributaries (30%) or directly from rainfall (13%) (Baran et al., 2007). The left
Page 4
Introduction
bank tributaries of Lao PDR together with the Se Kong, Se San and Sre Pok Rivers, that combined,
contribute 25% of the mean annual flow volume of the Mekong River at Kratie, are important
components of the flow regime in the Lower Mekong Basin and key to driving the flow reversal of
the Tonle Sap (MRC, 2005). Ecological processes in the TS-GL System are therefore vulnerable
to hydrological modification in upstream locations caused by water management structures and
abstractions.
1.1.3. Ecology of the Tonle Sap
Biodiversity and endemism
The status of the TS-GL System as a wetland of international ecological and conservation importance
was reinforced when it was approved by UNESCO as the Tonle Sap Biosphere Reserve (TSBR)
in 1997 (Davidson, 2006). Many of the plant and animal communities in the Lake itself and the
surrounding floodplains are adapted to the fluctuating water levels which is both a key driver of
ecosystem processes and a major factor affecting the species composition in different areas of the
Lake.
Davidson (2006) estimated that roughly 200 vascular plants occur on the surrounding floodplains.
One of the most distinctive plant communities on these floodplains are the flooded or swamp forests
that occupy about 10% of land on the floodplain area and are situated around the edge of the central
portions of the Lake that are permanently inundated (Campbell et al., 2006). They comprise woody
tree species that vary in height between 7–15 m, the dominant species being Barringtonia acutangula
(Lecythidaceae). Other species that are also found in this community include Diospyros cambodiana,
Terminalia cambodiana, a local endemic and Samandura harmandii, a narrow endemic
(Campbell et al., 2006; Davidson, 2006). The swamp forests become submerged for up to
6 months of the year between August and January during which time the majority of the deciduous
tree species loose their leaves. Flowering and fruiting occurs in the late dry/early wet season after
which the trees drop their seeds to be carried away in the flood (Davidson, 2006). Aside from the
swamp forests, the swamp scrubland vegetation community that occupies a much greater proportion
of the floodplain (up to 80%) also consists of woody species, but these do not exceed more than 4 m
in height. The scrubland vegetation community is dominated by Euphorbiaceae, Fabaceae and
Combretaceae (Campbell et al., 2006). These flooded forests are critical for ecological processes in
the basin, providing spawning, nursery and feeding habitats for fish and wildlife (Ahmed et al., 1996).
Although mammal diversity is not high around the Great Lake, distinctive primates such as
the slow loris Nycticebus coucang, long-tailed macaque Macaca fascicularis and silvered langur
Semnopithicus cristatus have all been recorded in the Prek Toal Core Area of the TSBR (Campbell
et al., 2006). The Prek Toal is also recognised as important site for breeding colonies of waterbirds,
supporting about 100 species of which 13 are of global significance (Clements et al., 2007). Nearendemic species include giant ibis Pseudibis gigantea–the largest of the world’s ibises and the only
Page 5
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Indochinese population of the critically endangered Bengal florican bustard Houbaropsis bengalensis
is restricted to the grasslands surrounding the Tonle Sap (Gray et al., 2009) and Prek Toal supports
the only known Southeast Asian breeding colonies of the globally threatened spot-billed pelican
Pelecanus philipensis and milky stork Mycteria cinerea (Clements et al., 2007).
The TS-GL System also supports a diversity of reptiles including the Siamese crocodile Crocodylus
siamensis, although numbers are low (Cambell et al., 2006). Seven species of homalospine watersnake
also inhabit the system including the endemic Tonle Sap watersnake Enhydris longicauda
(Stuart et al., 2000).
Occupying the permanently inundated waters and the floodplain during the wet season are upwards
of 149 fish species in 35 families, although the exact number is difficult to determine due to absence of
extensive surveys, survey objectives and taxonomic uncertainties (Campbell et al., 2006;
Junk et al., 2006). Although the Mekong River has a relatively high level of fish endemism (24%),
none of these species are restricted to the TS-GL System itself (Campbell et al., 2006). Five
of the fish species that occur in the TS-GL System are globally threatened, one of which is
the Giant Mekong catfish Pangasianodon gigas (Davidson, 2006).
The flood pulse
As water levels in the main channel begin to rise each year and flow into the Tonle Sap they transform
the previously dry terrestrial floodplain habitat surrounding the Lake into an ephemeral aquatic
habitat. Together with the increase in the amount of habitat available to aquatic organisms, nutrients
and carbon are exchanged between the land and water giving rise to a peak in primary productivity.
Invertebrates, fish and plants adapted to living in river-floodplains systems exploit this increased
availability of food and habitat for feeding and sheltering their young and have therefore evolved
to synchronise their reproduction and recruitment to coincide with this flooding period. The central
importance of the flood to these processes led (Junk et al., 1989) to conclude that this ‘flood pulse’ was
“the principal driving force responsible for the existence, productivity and interactions of the major
biota in river-floodplains systems”. It therefore follows that any natural or anthropogenic changes to
either the magnitude, timing, duration of the flood, or speed with which the floodwaters rise (rate of
change), will affect the overall productivity of the system.
Floodwaters carry sediments and nutrients essential to supporting ecosystem processes in the Lake
and on its floodplain. The combined mean annual suspended sediment flux into the lake from the
Mekong River (72%) and tributaries (28%) is estimated to be 7–9 million t y-1. Most sediments are
deposited in the Lake or are trapped by the floodplain vegetation around the Lake margins
(Sarkkula et al., 2003; Kummu et al., 2008). However, deposition and erosion of sediments are
considered to be in balance and net sedimentation is therefore considered to be at or close to zero
(Sarkkula and Koponen, 2003). There is considerable intra-annual variation in total suspended solid
(TSS) concentrations that vary between 3.5 × 109 kg to 9 × 109 kg. TSS increases on the rising flood
Page 6
Introduction
2500
Rising flood
150
Receding flood
2000
120
Flux in
TSS conc.
1500
water level
90
1000
60
500
30
0
0
-500
-30
May
Figure 4
TSS concentration [mg/l] /
Water level [dm], AMSL
TSS flux [10^6 kg]
Flux out
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
Mean monthly TSS flux, concentration and water levels in the Lake for the seasons
1997-2003 (Kummu et al. 2008).
and reaches its peak in July and August (Figure 4). It begins to decline thereafter on the receding flood,
reaching its lowest levels at the beginning of the dry season (Kummu et al., 2008).
Inter-annual variability in river flows are strongly reflected in the sedimentation patterns. During
a dry year, very little sediment reaches the Lake, whereas during a wet year, the sediments become
widely distributed through the Lake system (Kummu et al., 2005). Long-term sedimentation rates have
been reported at less than 1 mm y-1. Although this is not considered a significant threat to the Lake
itself (Penny et al., 2005) caution that the channel morphology at the mouth of the Tonle Sap where it
enters the Lake may be more sensitive to altered sedimentation regimes.
The TS-GL System is mesotrophic and most likely phosphorous limited, particularly over the high
water period. Most of this phosphorous is sediment-bound and believed to be brought by the flood
(Sarkkula and Koponen, 2003; Sarkkula et al., 2003). This sediment-bound phosphorous is made
available to phytoplankton through the growth and subsequent decomposition of vascular plants
growing in the aquatic-terrestrial transition zone (Kummu et al., 2008).
Oxygen concentrations in the Lake differ spatially (between the Lake and its floodplain), as well as
through the seasons. In the permanently inundated central portion of the Lake, wind and wave mixing
Page 7
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
ensures well-oxygenated water is distributed throughout the water column, even during the flood. In
parts of the floodplain during the high water period, particularly in the North where water velocities
are lower and there is sheltering by vegetation and little wind, anoxic conditions predominate in all but
the uppermost layers of the water column due to oxygen consumption by decomposing organic matter
of terrestrial origin inundated by the flood (Sarkkula et al., 2003 and Sarkkula et al., 2004). Oxygen
concentrations measured at the floodplain-lake interface exhibit high variability possibly because of
adventive mixing. The general trend in this region is towards an increase in concentration from dry
season (July) to the peak of the flood (November) (Sarkkula and Koponen, 2003).
Sediment-bound phosphorous is made available to the aquatic food web through assimilation by
higher plants growing in the aquatic-terrestrial transition zone (Sarkulla et al., 2003). Phytoplankton
growth occurs during the phosphorous-limited high water period (October–December) when
Aulacoseira diatoms and copepods dominate the pelagic trophic structure. During the low water
period when turbidity is high, phytoplankton growth is concentrated near the surface layers where
the predominant algae are positively buoyant Anabaena species (Sarkulla et al., 2003; Sarkulla and
Kopenen, 2003). Algal blooms are not common due to high zooplankton biomass suggesting that
nutrients are taken up and processed in the ecosystem very efficiently (Sarkulla and Koponen, 2003).
1.1.4. Fisheries of the TS-GL System
The fish fauna exploited by the fisheries of the TS-GL System can be separated into two principle
ecological guilds: (1) Blackfish that are tolerant of anoxic conditions and may undertake lateral
migrations between the Lake and floodplain and (2) Whitefish that undertake longitudinal migrations
between the Lake and the Mekong River via the Tonle Sap. Amongst the latter, the Cirrhinus and
Henicorhynchus genera are amongst the most important for river fisheries, whereas the former include
genera such as Channa, Trichogaster, Anabas, Oxyeleotris and Mystus (Lim et al., 1999). Both groups
depend on the advance and retreat of flood waters across the floodplains. An estimated 7 million
snakes or 777 tonnes are removed from the Lake annually (Brooks et al., 2007). Snakes provide an
inexpensive source of protein but current rates of exploitation are believed to be unsustainable
(Brooks et al., 2007).
Under Cambodian Fishery Law, the fisheries exploiting these resources are divided into two broad
categories: limited and open-access fisheries. Fishing takes place in two seasons: Open (October–May)
and closed (June–September).
The limited access fisheries are managed by means of ‘fishing lots’ and may operate only during
the open season. Fishing lots vary from a simple anchoring position for dais in the Tonle Sap to large
areas (up to 500 km2) of floodplain around the Great Lake, Tonle Sap, Mekong, Bassac and other
rivers and tributaries (Figure 5). Lots in the TS-GL System contain mostly natural habitats including
flooded forests, shrub forest and grasslands, but rice fields and villages are sometimes within their
boundaries. Approximately 35 fishing lots exist in the System fished using large scale gear
Page 8
Introduction
(van Zalinge, 2002). Many lots operate large-scale seines and bamboo fence traps to target blackfish
species such as Channa (see Plate 1). Barrage traps are also set in the delta region of the TS-GL
System to target longitudinally migrating whitefish (Lamberts, 2001). Exploitation rights for these
large-scale (industrial) fisheries are issued by the Fisheries Administration on a licence basis. Largescale fisheries licences are issued for two years. Licence revenue exceeds millions of dollars annually
(van Zalinge, 2002).
Open access fisheries comprise two main components: Middle-scale (artisanal) fisheries that can
operate only during the open season and family (subsistence) fisheries that can operate year-round and
in flooded rice fields (Sam et al., 2002).
Artisanal and subsistence fisheries may operate in open access areas. These fisheries employ a
wide variety of gear types including gillnets, seines, bamboo fence traps, cast nets and longlines.
Illegal fishing gear include the use of explosives, electro-fishing and poisoning (Lamberts, 2001).
Watersnakes are gillnetted, caught in bamboo traps, noosed or speared (Stuart et al., 2000).
Competition for fish resources among sectors is intense as illustrated by artisanal and subsistence
fishers operating from small boats between rows of dais in the Tonle Sap. During the 2004–05 open
season, reports of high catch rates also attracted thousands of additional artisanal and subsistence
fishers from throughout Cambodia’s lowlands. In January 2005, an estimated 7,000 boats were fishing
on the Tonle Sap including at times between rows of dai nets and around the Mekong Junction. Most
fishers were using drifting gill nets, with some using larger gear such as seines and trawls. The total
catch from these boats was estimated at about 4,000–5,000 tonnes in February 2005, a similar amount
to the dai catch that month (Hortle et al., 2005). Fish not caught by dai nets or other industrial gear are
vulnerable to capture by these competing fisheries as demonstrated by a tag-release study
(Hortle et al., 2004a). Many thousands of people also fish along the banks of the Tonle Sap using
cast-nets and small gear. Floodplains along the Tonle Sap are also fished where they drain into the
Tonle Sap by small-scale fishers and formerly by some commercial lot fishers (Hortle pers. Comms;
Valbo-Jorgensen et al., 2001 and Dubeau et al., 2001).
With the exception of the dai fishery, most of fisheries of the TS-GL System have not been subject
to detailed investigation or systematic routine monitoring, and are generally poorly documented. The
implications of this limited focus for interpreting data generated by the dai monitoring programme,
and in the wider context of fisheries policy and management activities are discussed in Section 6.
It has been estimated that the livelihoods of more than a million people depend on the fish
resources of the TS-GL System (van Zalinge, 2002). Most of these fishers are engaged in the openaccess fisheries, living at the edges of the floodplain or in floating villages or houses built on tall stilts
(van Zalinge, 2002).
Hap et al. (2006) reported that the total annual fish yield landed by all fisheries operating in the
TS-GL System to be in the order of between 200,000 and 218,000 tonnes with a landed value of
US$ 150–250 million. Van Zalinge (2002) estimated the yield to be approximately 235,000 tonnes.
Page 9
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 5
Page 10
Fishing Lots in Cambodia
Introduction
Plate 1
Arrow fence traps constructed in the flooded forest margins of the Lake
Page 11
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
These estimates are equivalent to approximately 50% of total annual yield of inland fish for Cambodia
(480,000 tonnes) estimated by (Hortle, 2007). Lamberts (2006) cautions that yield estimates for
TS-GL System may be unreliable due to the absence of quantitative surveys.
Van Zalinge et al. (2004) hypothesized that three factors affect fish production in the system:
(1) nutrient bearing sediment that influences rates of primary production; (2) oxygen conditions on the
floodplain that effectively determines habitat availability and (3) the transport of fish eggs and larvae
to the system.
Threats to fishery resources
Campbell et al. (2006); Keskinen (2003) cite human population growth and concomitant
over-harvesting of resources together with potential future hydrological changes as amongst the
principle threats to fisheries and wildlife in the TS-GL System. To this, Hortle et al. (2004a) add
habitat destruction and fragmentation by dams. It has been estimated that up to a third of flooded
forests was lost between 1973 and 1993 in response to increasing demand for rice growing land and
firewood, and subsequent soil errosion (Nao and van Zalinge, 2001). Lim et al. (2004) report that the
construction of rural water irrigation infrastructure has also impeded fish migrations in the system.
According to anecdotal reports, large and medium sized fish have become scarcer in response
(van Zalinge et al., 1998 and van Zalinge, 2002). Water management projects in upstream locations
may pose the greatest threat since they are likely to alter the magnitude and timing of flows, as well as
trap sediments and nutrients (Ahmed et al., 1996; Lamberts, 2008).
Conservation of this valuable resource and predicting how it will respond to future anthropogenic
interventions requires understanding of the physical and biological processes that support its
productivity.
1.2. The Tonle Sap dai fishery
The dai fishery on the Tonle Sap, established almost 140 years ago, is an important component of the
large scale (industrial) fisheries of the TS-GL System described above, landing up to an estimated 14%
(33,000 tonnes) of the total catch from the system (235,000 tonnes) and up to 7% of the estimated
annual inland fish landings in Cambodia.
The dai fishery targets the refuge migrations of a multi-species assemblage of fish as they migrate
from the TS-GL System to the Mekong main channel via the Tonle Sap with the receding floodwaters
each year. Fish also enter the Tonle Sap from its adjacent floodplains via numerous tributaries and
channels (Hortle pers comms). Quantitative estimates of the relative importance of these migratory
routes are not available. Judging by the distribution of fishing lots in the Lake and neck of the Tonle
Page 12
Introduction
Sap, and the absence of dais in the tributaries and channels of the Tonle Sap, it would therefore not
appear unreasonable to assume that most fish caught in the dai fishery migrate from the Lake itself.
Further studies are required to confirm or reject this assumption.
The fishery provides important seasonal employment opportunities for more than 2,000 rural
people and is a major supplier of the essential ingredient for prahok, a fermented paste which is an
important protein source for many, particularly towards the end of the dry season when fish is scarce
(Halls et al., 2007).
In addition to its significant socio-economic value both locally and nationally, the dai fishery of the
Tonle Sap also provides a valuable source of data and information to monitor trends in the migratory
fish populations that seasonally utilize the TS-GL System and beyond and forming the basis of
important fisheries.
Understanding the ‘drivers’ of these trends is important for effective resource management but
also to gain insights into how fish populations respond to important environmental variation such as
flooding patterns. Improving our understanding of the latter is likely to become increasingly important
in the face of climate change and basin development which both have the potential to significantly
modify flooding patterns throughout the Mekong basin.
1.3. The purpose and scope of the paper
This paper aims to provide a comprehensive and detailed description of the Tonle Sap dai fishery
including target species, management and monitoring activities (past and present), and the dynamics of
the exploited fish populations. The paper includes the results of analyses of the most recently available
survey data. The paper discusses the implications of these findings in the context of fisheries policy
and management, and basin development. Recommendations are made to improve existing monitoring
activities and priority research needs are identified. The paper therefore provides an important
reference for those engaged in the management of the migratory fish stocks that seasonally inhabit the
TS-GL System.
The paper contributes to Logframe Outputs 1, 3 and 5 of the Fisheries Ecology Valuation and
Mitigation (FEVM) Component of the MRC’s Fisheries Programme: (1) Improved information on the
status and trends of the fisheries in the LMB is available to riparian governments, national Mekong
Committees and the MRCS; (3) Improved information on the ecology of the fisheries of the LMB and
models for basin planning purposes are available to basin planners and development agencies, and (5)
Improved stakeholder and institutional capacity to (a) monitor and evaluate the status, value and trends
of fisheries in the LMB, (b) assess and mitigate environmental impacts on fisheries and (c) initiate and
sustain research and development activities.
Page 13
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
1.4. The structure of the report
The context and importance of the Cambodian dai fishery has been described in Section 1. Section 2
gives an account of the history of the fishery, its gears and operation, target species and the disposal of
its outputs. Section 2 also provides information relating to the economics of the fishery and details of
management activities undertaken by the Government of Cambodia Fisheries Administration (FiA).
Ad hoc and routine survey activities are described in Section 3. Details of the content and structure
of databases used to store and process the data are provided in Section 4. More detailed descriptions
of queries to generate raw data for analyses are available in a companion working document. Section
5 contains a description of the spatial and temporal dynamics of the fishery based upon a detailed
analysis of fish abundance, biomass, fish size, and species diversity and similarity. A sub-section
is devoted to examining the effect of hydrology on fish biomass. The findings are discussed in the
context of fisheries policy and management in Section 6 which includes recommendations for future
monitoring and research activities.
Page 14
2. The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
The stationary trawl or Loh Dai fishery was introduced into Cambodia by the French colonial
authorities between 1873 and 1889 for harvesting small-size fish primarily for fish oil production
(Touch, 1998). The fish oil was used to replace engine oil during World War I. The importance of
the fishery for fish oil production has diminished and most of the landed catch is now for human
consumption.
2.1. The Location of the fishery
The Dai fishery is located in the lower section of the Tonle Sap spanning more than 30 km across
the municipality of Phnom Penh and Kandal Province (Figure 6). Dai nets are arranged in up to 15
separate rows of between one and seven nets anchored perpendicularly to the channel, with the net
mouths facing upstream (Table 1 D; Plate 2). The most upstream Row (15) is located approximately
35 km from Phnom Penh. Individual dais within a row are allocated an identification letter from ‘A’ to
‘H’. This letter combined with the row number provides a unique alpha-numeric identification code for
each dai (e.g. 10A).
In 2001, dai row #1 with 3 dai units, situated nearby Chhroy Changvar bridge (presently known
as the Cambodian-Japanese Friendship Bridge), was decommissioned as a result of a fishery policy
reform in 2000. Some additional dai units (8B, 11A’, 12G and 15E) are reported to be in operation
in Kandal Province every year. According to the Department of Fisheries Affairs of the Fisheries
Administration (FiA), dais 8B, 11A’ and 15E are operated with licenses issued by the Ministry of
Agriculture, Forestry and Fisheries whereas 12G is operated without a license.
Plate 2
Rows of dai nets anchored across the Tonle Sap
Page 15
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 1
Province
Kandal
Province
Sub-Total
Dai locations in the Tonle Sap during the 2007–08 season
104º48.580’
East ends
A
A
A
B
C
B
C
D
C
D
E
F
E
D
E
F
G
G
H
5
7
Coordinates
North ends
104º47.266’
A
B
E
Total number of dai
units forming each
row
Approximate
cumulative distance
between rows (km)
11º52.110’
11º53.585’
104º48.111’
104º47.675’
A
A’
D
D
Relative transversal positions of dai nets in the
Tonle Sap channel
Row
No.
37.50
11º51.618’
104º49.383’
C
C
D
G
3
7
5
6
1
3
5
33.00
11º50.349’
104º50.515’
B
C
F
Row 15
31.92
11º47.447’
104º51.026’
B
E
Row 14
28.93
11º42.257’
104º51.360’
D
Row 13
23.07
11º40.963’
104º51.969’
C
Row 12
13.17
11º40.477’
C
Row 11
10.77
11º39.685’
B
Row 10
4.87
B
Row 9
4.28
G
Row 8
F
Row 7
4
5
4
42
D
4
G
D
F
104º52.581’
C
D
E
11º38.867’
C
D
3.77
C
C
Row 6
B
9 rows
104º53.328’
B
5
104º53.809’
A
F
11º38.363’
A
B
11º38.295’
104º55.116’
104º54.705’
E
3.28
11º37.640’
A
2.75
11º37.068’
22
D
Row 5
1.40
5 rows
64
C
Row 4
0.00
B
Row 3
Sub-total
15 rows
Phnom Penh
Municipality
Row 2
Grand total
Page 16
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Figure 6
The location of dai fishery in the Tonle Sap
Page 17
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
The transversal position of individual dai rows within the channel varies along the upstreamdownstream axis. Rows 2–4 and row 7 are positioned towards the right bank of the channel (facing
northwards), whilst row 13 and 14 are positioned towards the left bank of the channel. Other rows are
positioned more centrally in the channel (Table 1). These positions have remained largely unchanged
for more than a century and may have been chosen to maximize catch rates determined by river
morphology and hydrology. The locations of the dai rows as well as the dai units are now controlled
under Fisheries Law (see Section 2.6).
2.2. Target species
The dai fishery primarily targets small cyprinids including Cirrhinus lobatus, Henicorhynchus
cryptopogon, Paralaubuca barroni, Labiobarbus lineatus, Henicorhynchus siamensis and
Labiobarbus siamensis, collectively known as trey riel in Khmer, as they migrate from the Tonle
Sap Lake to the Mekong River with receding floodwaters between October and March each year.
Other species making at important contribution to landings are Labeo chrysophekadion, Pangasius
pleurotaenia, Puntioplites proctozystron and Thynnichthys thynnoides (Figure 7 and Plate 3 – Plate
12). During the 2009–2010 fishing season, a total of 123 individual species were reported to have been
landed. Small-sized fish are often used to produce Prahok, fish meal and fish sauce (see below).
Ngor (2000) notes that species composition varies during the six months of the fishing season. He
reports that large and medium size fish species such as Pangasianodon gigas, Catlocarpio siamensis,
Probarbus jullieni, Cirrhinus microlepis, Pangasius spp., Cyclocheilichthys enoplos etc., are caught
in larger numbers at the start of the season (October and November) compared to later in the season.
Inter and intra-specific differences in the timing of migrations conditioned primarily by fish size has
been described in the Mekong since the mid-1950s (Welcomme, 1985). Inter-specific differences in
migration timing are described in more detail in Section 5.3.2.
1.5% 1% 1%
Cirrhinus lobatus
1% 1% 1%
Lobocheilos cryptopogon
2%
2%
2%
Paralaubuca barroni
Labiobarbus lineatus
4%
27%
Henicorhynchus siamensis
Labiobarbus siamensis
Labeo chrysophekadion
11%
Pangasius pleurotaenia
Puntioplites proctozystron
Thynnichthys thynnoides
12%
21%
12%
Clupeichthys aesarnensis
Syncrossus helodes
Pangasius larnaudii
Yasuhikotakia modesta
Cirrhinus microlepis
Figure 7
Page 18
The species composition of the dai catch in 2009–10
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Plate 3
Cirrhinus lobatus (trey riel angkam). A migratory herbivorous species, which seems to play a key
role in the food chain; It is a protogynous hermaphrodite, which spawns in June and July in the
main channel and in floodplains (MFD, 2003).
Plate 4 Henicorhynchus cryptopogon. Found at midwater to bottom depths in canals, ditches and small
streams of floodplains, and more commonly in larger rivers as the temporary water bodies dry up.
It undertakes lateral migrations on to seasonally inundated land during the rainy season, where it
feeds on algae, periphyton, and phytoplankton (Rainboth, 1996).
Page 19
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Plate 5
Plate 6
Page 20
Paralaubuca barroni (trey slak russey) Feeds on zooplankton; Found in slow flowing or standing
water in the Middle Mekong Basin; Processed to Prahok (MFD, 2003).
Labiobarbus lineatus (trey khnawng veng). Omnivorous species found basin-wide in rivers and
streams. Undertakes seasonal migrations onto floodplains to spawn. (MFD, 2003).
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Plate 7
Plate 8
Henicorhynchus siamensis (trey riel tob). Abundant herbivorous species occurring basin-wide in
large and small rivers. It is highly migratory and spawns at the beginning of the flood. The species
is extremely important in the dai fisheries and is also caught with other gear. Most of the fish is
processed to Prahok. (MFD, 2003).
Labiobarbus siamensis (trey ach kuk). An important whitefish species that feeds on aquatic
animals, small water plants and algae. It is believed to migrate from Cambodia to Lao PDR
in January–February to spawn on upstream floodplains in June–July. After spawning, the fish
migrates down the Mekong River again (MFD, 2003).
Page 21
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Plate 9 Plate 10 Page 22
Labeo chrysophekadion (trey kaek). Predominantly herbivorous occurring in flowing and standing
water including reservoirs throughout the Mekong Basin. It spawns mainly in the early flood
season in a variety of habitats. Commercially important marketed fresh or dried and salted.
(MFD, 2003).
Pangasius pleurotaenia (Chhwiet). An omnivorous catfish with a basin wide occurrence although
apparently more abundant in the Lower Mekong. It is believed to undertake significant spawning
migrations during the beginning of the flood season. After spawning in the mainstream, eggs and
larvae drift to the nursery areas. The drifting larvae are caught with special gear and cultured. A
very important food fish in the Lower Mekong and also used in the aquarium trade (MFD, 2003).
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Plate 11
Puntioplites proctozysron (trey chrakeng). Omnivorous species occurring in slowly moving and
standing water including reservoirs. Moves laterally into the flooded forest during the high water
season. Spawning occurs during the flood. (MFD, 2003).
Plate 12
Thynnichthys thynnoides (trey linh). Pelagic feeding in a variety of habitats basin-wide including
the mainstream. It migrates in the mainstream during the dry season. Enters floodplains during the
high water season and spawns in the flooded littoral zone possibly throughout the flood season.
Caught with a variety of small to large scale gear; Sold fresh and processed’ (MFD, 2003).
Page 23
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2.3. Fishing gear and operation
A dai is effectively a stationary trawl net anchored within the river channel to intercept migrating fish.
A dai can be operated singly or joined with up to six other dais in a row. It comprises a long
(100–120 m) cone-shaped net suspended from two anchored bamboo rafts or (sampans) and a floating
work platform (or a movable floating house) positioned at the codend of the net (Figure 8 a). A small
boat or floating platform is secured between the two anchored rafts by bamboo poles that serve as a
gangway and help to keep the net mouth open. The distance between the two anchored rafts
(25–27 m) and the depth of the water determines the net mouth area. In all, up to 30 anchors attached
to steel cables may be required to secure the dai in position. Another raft is positioned above the
entrance of the net mouth. This holds a winch for raising and lowering the chain ‘footrope’ thereby
opening and closing the net mouth (Figure 8 b and c). The bag net is kept open by the force of the
current and with help of anchors.
Three types of nets varying according to mesh size and shape are used by dai operators during the
course of the fishing season according to the prevailing hydrological conditions and the mean size of
migrating fish (Table 2). All three nets are conical in shape but the Yore net has a much smaller toppanel and longer U-shaped top-rope to maintain net buoyancy during low flow conditions towards the
end of the season (January–March) when lift forces exerted on the net generated by the flow are much
lower. With a much reduced top-panel area, fishers must first drive fish towards the codend before
hauling it.
Table 2
Net type
Mean (and range) dimensions, mesh sizes, and periods of operation of each dai net type reported
by all 64 dai unit operators during the 2007–2008 season.
Length (m)
Width (m)
Depth (m)
Min mesh (m)
Max mesh (mm)
Dai Chieu
157 (110 – 180)
29 (30 – 27)
21 (32 – 10)
24 (50 – 13)
182 (250 – 100)
Dai Nheuk
157 (110 – 180)
29 (30 – 27)
17 (28 – 8)
16 (25 – 15)
118 (220 – 80)
Dai Yor
158 (110 – 180)
29 (30 – 27)
14 (26 – 6)
15 (20 – 15)
70 (200 – 20)
Period of operation
Net type
Oct
Nov
Dec
Jan
Feb
March
Dai Chieu
Dai Nheuk
Dai Yor
During the first two months of the season (October–November), the dai chieu net is used to target
medium and large size fish species which fishers believe migrate from the Great Lake first followed by
smaller sized species and individuals later in the season. These fisher observations appear consistent
with those derived from the analysis of monitoring programme data (see Sections 5.3.2 & 5.3.3).
Strong water currents during this period apply significant drag forces to the dai net which limit the
Page 24
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
minimum net mesh sizes that can be used. During the first two months when haul weights are low
but fish size is relatively large, an open weave basket made of bamboo and rattan is often attached to
the codend (Figure 8 d). The Dai nheuk net is used between December and February to target smaller
species including Cirrhinus lobatus, Henicorhynchus siamensis, Paralaubuca spp. (trey slak russey)
and Labiobarbus lineatus (trey khnawng veng). The mesh sizes of all three nets decrease from the net
mouth to the codend. Note that gear dimensions and net mesh sizes used by dai operators often exceed
the legal maximum and minimum sizes respectively as stated in the Burden Book (see Section 2.8).
(a) dai nheuk
(b) mouth of a dai
Page 25
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
(c) Winches used to lift the codend of the net and to close the dai mouth.
1.8–2 m
30–50 cm
for small dai only
bamboo codend basket
(d) Open weave-baskets attaching to the codend of the dai, used especially in October and November
Figure 8
Main structures of the dai fishery in the Tonle Sap
(source: Deap et al., 2003, pp. 228–232)
Page 26
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Upstream dai operators positioned closest to the Lake tend to use Dai chieu nets with a smaller
maximum and minimum mesh size compared to those operators downstream with differences of up to
100 mm (Figure 9). Conversely, Yor nets used upstream tend to have a larger maximum mesh size than
those used downstream. Minimum and maximum mesh sizes of nheuk nets show no obvious trend
with row number.
mesh size (mm)
Min. mesh
45
40
Dai
DaiChieu
Chieu
Dai
DaiNheuk
Nheuk
Dai
DaiYor
Yor
35
30
25
20
15
10
5
0
0
0
2
2
44
6
6
8
8
Row
Row
10
10
12
12
16
Dai
DaiChieu
Chieu
250
Max.
Min. mesh
mesh size
size (mm)
(mm))
14
14
Dai
DaiNheuk
Nheuk
Dai
DaiYor
Yor
200
150
100
50
0
0
0
Figure 9
2
2
44
6
6
88
Ro
Row
10
10
12
12
14
14
16
16
Reported mean minimum and maximum mesh sizes of each net type by dai row
To empty the net of fish, the net mouth is first closed by raising the foot rope. The codend of the net
is then winched aboard the working platform and emptied into sorting compartments (Plate 13) or
directly into boats belonging to fish buyers during peak landing periods (Plate 14).
Page 27
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Plate 13 Emptying the codend into sorting
compartments
Plate 14 Plate 15
The codend being emptied
directly into traders boats
Diesel engines used to haul the dai net
A survey undertaken in 2007 revealed that after 1999, diesel engines (Plate 15) began replacing the
traditional hand-powered wooden winches (Figure 8 c) used to raise the foot rope to close the net
mouth and to raise the codend during net hauling. Since 2006 all dai operators except 6G are now
mechanized using a total of almost 2500 HP (Figure 10).
Page 28
Dais using engines
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
100
90
80
70
60
50
40
30
20
10
0
1998
2000
2002
2004
2006
2008
Year
3500
Total engine HP
3000
2500
2000
1500
1000
500
0
1998
2000
2002
2004
2006
2008
Year
Figure 10 Cumulative number of dai operators using diesel engines to close and haul their nets (top) and (bottom) cumulative total engine horsepower (HP) employed in the dai fishery since 1999.
Ninety-percent of the dai operators reported
that they mechanised their operations to
increase the number of hauls they could make
and the individual size of each haul during
peak catch periods. Almost all reported their
prime motive was to reduce hauling time,
increasing the effective fishing time (net soak
time) each day. Only 10% of operators reported
that they employed engines to reduce labour
costs (Table 3).
Table 3
Reasons reported by dai operators to
mechanise hauling operations
Reason
% of operators (63)
Increase hauls/day?
81
Lift more fish each haul?
71
Both of the above?
90
Reduce time hauling?
98
Reduce labour ?
10
Page 29
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2.4. Fish disposal
Small-sized low value fish (LVF) fish species that form the great majority of the annual catch by
weight (98% in 2007–2008) are typically landed and sold to fish traders either on the floating dai
platform or on the nearest riverbank. During periods of high catch rates, the entire haul may be landed
directly into traders’ boats. Typical pathways of disposal are illustrated in Figure 11.
a
LVF sauce manufacturers
LVF exporters
a
Markets
a
a
LVF wholesalers
Viet Nam
Provinces/cities
a
Tonle Sap dai fishery
operators
LVF traders
Fish/animal raisers
b
a
Rice farmers
a
Phnom Penh & Kandal
market retailers
Consumers
Occasional (less common) pathway.
Pathway recommended by the MoFL. for research dais
(rice farmers given priority to landings over traders).
a
b
Figure 11
Fish disposal pathways for low value fish (LVF) or small size fish species.
Source: So et al. (2007), p20.
2.4.1. Prahok
Prahok is a protein-rich paste made of salted and fermented low-value fish typically trey riel, trey slak
russey, trey khnawng veng, and trey linh. Prahok is a traditional condiment in most Cambodian dishes,
a symbol of national cuisine and an important source of protein, calcium and vitamin A for the rural
poor in Cambodia, particularly towards the end of dry season (April–May) when fish is scarce. Prahok
is produced within the vicinity of the dai fishery for both subsistence and commerce trade. Subsistence
production is mainly by poor farmers or households from rural areas where fish is scarce or where
farming activities prevail. They establish makeshift processing camps along the riverbanks each year
during the peak catch period (December–January). The peak catch period is often communicated by
word of mouth or via announcements by the FiA on national television. Each household camp may
process between 80 and 200 kg of fresh fish to produce prahok. The fish is dressed, often beheaded
Page 30
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
and partly or wholly eviscerated and mixed with salt. After some days it may receive further treatment
including filleting, cleaning and packaging in airtight containers. Commercial producers use machines
to remove the head and de-scale the fish. Salt and either rice, grains or fruit may be added to lower the
pH to increase storage life and add to the taste (Sverdrup-Jensen, 2002; Halls et al., 2007).
Commercial producers supply wholesalers, retailers and exporters in markets countrywide.
Commercial producers also export prahok especially to Thailand. A large-scale producer typically
processes between 60 and 150 tonnes of fresh fish to produce between 20 to 50 tonnes of prahok
(So et al., 2007).
Small-size fresh fish from the dai fishery are also processed into other forms of fermented fish
(pra-ork), smoked fish (trey cha-eur), sweet fish (mam), and fish sauce for human consumption
(Lieng et al., 1995; Ngor, 2000 and So et al., 2007). Following the growth of aquaculture along the
Tonle Sap during the last two decades, small-size fish are also used as feed for fish cultured in cages or
ponds including Channa micropeltes, Pangasianodon hypophthalmus and Clarias spp., as well as
feed for animals including ducks, pigs and chickens (So and Nao, 1999; Hav and Ngor 2005 and
So et al., 2005). For the latter, fish may be first sun-dried. So et al. (2007) report that each drying
operator may buy up to 150 tonnes of small-sized fish landed by the dai fishery. Sun-dried fish is fed
to ducks, chickens and pigs, and for producing fish meal for export to neighbouring countries.
Recent reports (So et al., 2007) suggest that small-size fresh fish landed by the dai fishery are also
exported to Viet Nam by boat or road for both human consumption, fish and other animal feed. This
practice was prohibited by the government in 1990 to control prices and ensure adequate supply for
rural farmers and households (Touch, 1993).
More valuable medium and large-sized fish species including Osteochilus melanoplerus,
Pangasius larnaudii, Cyclocheilichthys enoplos and Pangasianodon hypophthalmus are often kept
alive in bamboo cages suspended below the working platform of the dai. A single cage can contain up
to 20 tonnes of live fish (Lieng et al., 1995). These fish are sometimes sold during the closed season
(March–September) when fish supply is low and prices are high. High value species are also sold
during the fishing season to fish traders to supply nearby markets or for export.
Ngor (2000); So et al. (2007) report that the supply and mean size of these medium and large-sized
fish species have declined since the 1990s. Factors postulated to be responsible for these declines
include increasing rates of exploitation and illegal gear use, and habitat destruction and modification.
Changes in the species composition of the dai fishery landings are examined in detail in Section 5.4.
2.5. Economics
From a sample of 16 dais, Hap and Ngor (2001) estimated that the mean net profit of a dai unit is
approximately US$14,000 per annum corresponding to a mean annual rate of return to investment
Page 31
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
of 60% (Table 4). A more recent study (So et al., 2007) estimated returns to be approximately 50%
but with significant variation ranging from 3%–74%. This lower return may reflect the exclusion of
revenue generated from the sale of medium and large high value species from the study.
Table 4 Dai Fishery profit estimates 1999-2000. Hap and Ngor (2001), p. 224
Items
I. Gross Return
Total values ('000 Riels)
USD
183,857
47,755
Variable Costs (VC)
Salary and wages of labours
10,426
Fuel/oil
8,306
Food expenditure
3,405
Other expenses (medical, cigarette, fruit etc)
2,100
A. Total VC
24,237
Fixed Cost (FC)
Lease fees
Government tax (5% on lease fee)
8,389
443
Depreciation of:
Gear/equipment
Boat (motorized and non-motorized)
Engine boat
Engine for electricity & dynamo
Floating-house and cage for stocking fish
Total depreciation
Repairs/maintenance
Interest rate on borrowed funds
17,381
1,694
818
170
1,632
21,694
26,007
50,321
B. Total FC
106,855
II. Total Costs
131,092
III. Gross Profit (I - A)
159,620
IV. Net Profit (I-II)
Opportunity Cost of Management
52,765
V. Rate of Return to Capital
59.75
VI. Operating Profit Margin Ratio
56.03
Note:
Page 32
13,705
521
Exchange rate is 3,800 Riels per one US$ (January, 2000). Opportunity cost of management = 5% of salary/wage of labour
(374,839 Riels). Total value of assets = Total values (costs) of fixed asset (i.e. Gear/equipment, boat, engine for boat and
electricity and floating house) and TVC; Adjusted Net Income = Net Profit + Interest Rate + Taxes. Rate of Return to Capital =
(Return to Capital/Total values of Assets)* 100 Return to Assets/Return to Capital = Adjusted Net Income - Opportunity Cost of
Unpaid Labour - Opportunity Cost of Management.
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
The price of trey riel which forms approximately 40% of the total annual landings of the dai fishery
has increased by almost a factor of 10 from 1995–6 to 2007–8 (Figure 12). So et al. (2007) report a
similar trend in price for both fresh and preserved fish at dai landing sites.
2000
14000
Average weighted price (Riel/kg)
Price (riel/kg)
12000
Adusted price (Riel/kg)
1600
Annual catch (tonnes)
10000
1400
1200
8000
1000
6000
800
600
4000
400
2000
2007-8
2006-7
2005-6
2004-5
2003-4
2002-3
2001-2
2000-1
1999-0
1998-9
1997-8
1996-7
1995-6
200
0
Annual catch (tonnes)
1800
0
Season
Figure 12
Actual (upper line) and inflation adjusted (lower line) unit price of trey riel. In 2004, 4,000 Riel
was equivalent to about US$1.
Source: Updated from Hortle et al. (2005). Correlation between annual catch and average adjusted price (r = 0.84)
unexpectedly indicating that price increases with supply.
Using data on costs and revenue reported by So et al. (2007), the annual profit of dai units
tends to decline from the most upstream dai row 15 to row 2 (Figure 13). Equipment, labour,
and fuel costs are all higher upstream and unit fish prices are lower, but these differences are
more than compensated for by disproportionately larger landings made by the upstream dai
units. Noteworthy, is that licence fees (both official and unofficial combined) paid by dai
operators appears independent of their reported operating profit (Figure 14).
The inclusion of high-value medium and large size fish which form only 3% of the catch
by weight but 11% by value in 2007–08 might alter these conclusions.
Page 33
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
P < 0.01
b
c
10000
P < 0.001
P = 0.02

Labour costs ($)
a
7500
5000








2500





0
15 14 13 12 11 10 9 8 7 6
5 4 3 2
Row
d
P = 0.09
g
Page 34
P < 0.001
h
P < 0.001
Figure 13 e
f
P = 0.04
i
P < 0.01
P = 0.02
Mean values of key economic variables for the Cambodian dai fishery plotted as a function of dai
row number. Error bars show mean values +/- 1 S.D. P-values < 0.05 indicate a significant trend in
the variable value with row number. Data from So et al. (2007).
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
Figure 14 License costs for 25 dais plotted as a function of their annual profit (R2 = 0.002; p = 0.85).
Data from So et al. (2007).
2.6. Management
2.6.1. Legislation
The dai fishery is legislated under Cambodian Fisheries Law (2006) (article 39). The ‘burden book’
describes management legislation including operating seasons, the position of the dais in the river,
gear size restrictions, payment rules and harvest rules. The Fisheries Administration (FiA) Inspectorate
and Cantonment are responsible for monitoring and enforcing the rules and regulations described in
the burden book.
2.6.2. Closed season
Fishing with dais in the Tonle Sap is permitted only between 1st October and 30th of March in the
following year after which all dai structures have to be completely removed from the River.
Re-installation of the dai gear is permitted to begin in September.
Page 35
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2.7. Effort control / Licensing
Under Cambodian Fisheries Law, a dai unit is classified as a large scale fishing gear or ‘fishing lot’
(No. 028 KOR SOR KOR Table A III). Rights to these lots in the form of a fishing licence are either
auctioned by the Government to the highest bidder for exclusive exploitation for a two-year period
(Deap et al., 2003) or are allocated by the FiA for research purposes (Table 5).
Table 5
Dai unit licence allocation by type for the 2007–08 season
Provinces
Phnom Penh
Kandal 2
Total
1
Dai fishery licences
Auctioned dai (units)
Research dai (units)
Total
18
26
44
4
12
16
22
38
60
Source: FiA (2007)
The processes and procedures for the auction of the dai fishery are stipulated in a sub-decree of
Cambodian fisheries law. This sub-decree provides rules on renting inland and marine fishery domains
for exploitation comprising 29 articles (CoM, 1989). The auction is open to all Cambodians except for
government officials (Article 6) and conducted at provincial levels. Articles regarding the dai fishery
or fishing lot auction written in the sub-decree contain rules on the auction process, payment of fishing
fees by the auction winners, rules on fish harvesting and rules on enforcement responsibilities.
Since 2001, 60 dai units in 14 rows (row 2–15) have been granted licenses by the FiA to operate
each season (van Zalinge et al., 2003). Unlicensed fishing occurs. During 2007–08 one dai unit in
Kandal Province operated without a licence and other three units were operating under additional
licenses issued by MAFF and FiA.
In 1938–39, 108 dai units were permitted to fish in 23 rows (Chevey and Le Poulain, 1940) but by
1962–63 the total was just 61 units in 15 rows (Fily and d’Aubenton, 1965). The mid-eighties saw an
increase to 86 units followed by a decline to present day numbers (Figure 15).
1
Research dais in Phnom Penh include all dai units in row #2 (2A, 2B, 2C and 2D). This title of research dais in Phnom has
no timeframe limitation.
Research dais in Kandal Province include row #14 (14A, 14B and 14C), row #13 (13A) #12 (12A, 12B, 12C, 12D, 12E) and row #9
(9B, 9C and 9D). All research dais in Kandal have a limited timeframe between 2003–09.
2
Page 36
The Dai Fishery of Tonle Sap-Great Lake (TS-GL) System
120
Dais Licensed
110
100
90
80
70
60
50
40
1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
Year
Figure 15 Number of licensed dais (1938–2007) reported by Chevey and Le Poulain (1940); Nguyen and
Nguyen (1991); Lieng et al. (1995); Ngor (2000) and Ngor and van Zalinge (2001).
2.8. Gear restrictions
The ‘Burden Book’ states the maximum gear dimensions and minimum mesh sizes for the three types
of nets that can be used in the fishery during the open season (Table 6).
Table 6
Maximum gear dimensions and minimum mesh sizes for the three types of dai nets
Types and size of dai gears
Period of operation
Start
End
Dai Chieu
Net dimension:
- Size: 120 m (Length) x 27 m (width)
- Mesh size: 20 mm – 220 mm
01 October
01 January
Dai Nheuk
Net dimension:
- Size: 120 m (Length) x 27 m (width)
- Mesh size: 15 mm – 220 mm
02 January
28 February
Dai Yor
Net dimension:
- Size: 120 m (Length) x 27 m (width)
- Mesh size: 15 mm – 30 mm
01 March
15 March
Source: Dai fishery’s Burden Book 2005–07
Page 37
Page 38
3. Routine monitoring activities and survey methodologies
3.1. Purpose of monitoring
The primary purpose of monitoring has been to generate estimates of total aggregated monthly and
seasonal estimates of landings and their monetary value. These figures are reported to the Fisheries
Administration (FiA), Ministry of Agriculture, Forestry and Fisheries, in Phnom Penh. Total catch
and more recently, catch rate estimates have also been used to monitor resource trends in response to
environmental variability and management initiatives.
3.2. A history of survey methodologies and sampling regimes
Routine monitoring of dai fishery has been undertaken since 1994 by the Management of the
Freshwater Capture Fisheries programme, the Assessment of Mekong Capture Fisheries Component
of the MRC Fisheries Programme (AMCF, 2003–2006) and since 2007 by the Fisheries, Ecology
Valuation and Mitigation (FEVM) component of the MRC Fisheries Programme in cooperation with
Inland Fisheries Research and Development Institute (IFReDI)–the research component of the FiA.
Combined, these programmes have generated the only continuous long-term data set for an inland
fishery in Cambodia–and one of only two in the Mekong Basin–the other being the lee trap fishery in
southern Lao PDR (MRC, 2010).
It is likely that both the dai and lee trap fisheries became foci for intensive monitoring because
they target fish migrations through ‘bottlenecks’: the Tonle Sap and the few channels passable
by fish at the Khone falls in southern Lao PDR. Fish densities and therefore catch rates are high
at these locations offering opportunities to sample relatively large proportions of migrating fish
populations during short periods of time giving rise to high sampling efficiency and the prospect to
accurate population estimates. The ‘Fishing Lots’ (barrage and floodplain) described in Section 1.1.4
also offer opportunities to sample significant proportions of exploited fish populations efficiently
owing to the often large-scale nature of the fishing gears operating within them. Attempts by the
FiA were made in the past to monitor these lots, but neither lot operators nor officials were prepared
to cooperate (van Zalinge, 2002). Whilst related ad hoc studies and anecdotes have been published
(e.g. Hortle et al., 2005; Valbo-Jorgensen et al., 2001 and Dubeau et al., 2001) attempts to routinely
monitor the other, more dispersed, sectors of the fishery (i.e. middle-scale artisanal and family
fisheries) have also failed in the past owing to a lack of capacity (van Zalinge, 2002).
Therefore whilst the dai fishery continues to provide valuable information about the variability and
long-term trends of migratory fish populations that seasonally utilise the TS-GL System, no equivalent
programmes exist to monitor populations of valuable Blackfish species that inhabit the system
year-round (van Zalinge, 2002).
Page 39
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Moreover, without information about the country’s other fishery sectors, it is difficult to interpret
variability or long-term trends in the dai (or any other) fishery without assuming that effort has
remained relatively static over the period of interest. This might not be unreasonable if the large-scale
barrier and fence traps common in TS-GL System and in other bottlenecks over the migratory range
of stocks exert the greatest overall effort (fishing mortality), but the artisanal and subsistence fisheries
may also be significant in this respect. This subject requires further investigation.
However, a focus on only one of several components of the fishery cannot provide all the necessary
data for the estimation of national landings by species–information that is important for economic
valuation and for environmental impact assessment purposes. It also makes recommendations for the
control of fishing effort (mortality) to ensure the sustainability of stocks difficult.
Recommendations to complement data and information generated by the dai fishery monitoring
programme for management and policy planning and evaluation purposes are made in Section 6.
The earliest ad hoc surveys of the dai fishery were documented by Chevey and Le Poulain (1940)
who reported catches over the 1938–39 season. They described the location of the fishery, the
structure and operation of dai fishing gear, recorded the relative abundances of individual species
and estimated the total catch to be around 13,500 tonnes (Table 7). Fily and d’Aubenton (1965)
monitored the dai fishery between 1962–63. They provided a general description of the fishery and
monitored fish catches and species composition on selected dai units. They also tested the effects of
current speed on species composition and catch volume. After the 1960s, there appears to have been
no more monitoring of the fishery until 1986–88 when Nguyen and Nguyen (1991) conducted a study
(written in Vietnamese) on freshwater fish productivity in Cambodia that included the dai fishery.
Fisheries statistics for the Department of Fisheries (DoF; the present-day FiA) were collected
more consistently from 1980 onwards. A lack of human resources and technical facilities during this
early period, however, meant that rigorous sampling protocols could not be followed. Prior to the
mid-1990s, data were collected using logbooks which dai operators were required to complete under
their licence agreement. However, most operators under-reported their landings fearing that reporting
their true landings might raise the cost of their operating licence. These logbook-generated data are
therefore regarded as unreliable.
In 1994, as part of the Mekong River Commission (MRC) Fisheries Programme and in cooperation
with the Department of Fisheries, the MFCF was established to improve fisheries monitoring
activities, the main focus being on mobile gear and large-scale fisheries. The software programmes
ARTFISH (Stamatopoulos, 1994) and LenFreq were introduced to aid data analysis. In the same year,
Lieng et al. (1995) applied a new data collection scheme–the Catch Assessment Survey (CAS)–to
estimate the total catch by species for the dai fishery.
Over the 1994–95 fishing season, Lieng et al. (1995) reported 73 dais distributed among 15 rows.
For their study, Lieng et al. (1995) stratified the dais into three groups of rows (rows 1–5; 6–10 and
11–15) according to their downstream location. Because peak catches were strongly associated with
Page 40
Routine monitoring activities and survey methodologies
the lunar phase, they also introduced a ‘peak’ and ‘low’ fishing period as an additional sampling
stratum. The Peak Period was defined as occurring during the waxing moon phase (khnaet) 4–6 days
before the full moon and the Low Period as those days falling outside of this period during the waning
moon phase (ronouch).
Dais were randomly selected for sampling within each combination of these strata and sampling
was conducted twice a month in December 1994 and in January and February 1995. Catch per haul
was sampled from selected dais and the number of hauls made during 24 hours estimated from the
observed hauls per hour. Lieng et al. (1995) estimated that total catch for the 1994–95 season was
18,410 tonnes. Their estimates were considerably higher than those of Fily and d’Aubenton’s (1965),
comparable with DoF and Chevey and Le Poulain’s (1940) estimates, and roughly equivalent to those
of Nguyen and Nguyen’s (1991) (Table 7).
Table 7 Estimates of catches prior to and including the 1994-95 seasons reported by Lieng et al. (1995)
No. of dais
Season
Total catch (tonne)
1938–39
13,568
106
128
1962–63
2,135
61
35
1981–93
5,000–12,839
97–35
–
1983–88
7,413–18,026
86
86–209
1994–95
10,755
73
161
1994–95
18,410
73
252
1995–96
Catch/dai (tonne)
63
Reference
Chevey and Le Poulain (1940)
Fily and d'Aubenton (1965)
DOF statistics
Nguyen and Nguyen (1991)
DOF (pers. comm.) (1995)
Project estimates (1995)
Auction results (1995)
Only 18 dais were sampled between December 1994 to February 1995 (Annex Table 25) hindering
comparisons with data collected under subsequent surveys.
Lieng et al. (1995) recommended that future monitoring programmes should introduce a sampling
stratification scheme based upon dai licence cost (value) to replace the existing row stratification on
the assumption that catch rates should be correlated with licence value. Whilst this assumption appears
difficult to reconcile with the results presented in Section 2.5, this stratification scheme was introduced
during the following fishing season. Unfortunately, this survey did not include the dai identification
code (see Section 2.1), or sample haul frequency also hindering comparisons with data collected under
subsequent surveys.
During the next fishing season (1996–97), the Department of Fisheries attempted a census of dai
catch rates (Ngor, 2000; Ngor and van Zalinge, 2001). However, Ngor (2000) reported that only 49%
of dai units were included in the survey because of insufficient resources. Dais were ranked according
to their percentage contribution to the total catch recording during the survey. Those dais that landed
more than 50% of the cumulative total catch were classified as ‘High Yield’ dais and the remainder as
‘Low Yield’ dais. It remains unclear how dais that were not sampled were classified.
Page 41
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Although the design of the 1997–98 survey was intended to incorporate the new High and Low
Yield stratum into the sampling regime (Table 8), the database records show no evidence of having
been stratified by yield before the 1998–99 season (Annex Table 25). From the 1997–98 season
however, the dai yield stratum, based on the 1996–97 census, has been consistently applied to stratify
the sampling programme. Dais continued to be selected randomly within each stratum (Table 8).
Table 8 Monthly dai sampling regime planned for the 1997–98 season (Deap, 1999, Ngor and van Zalinge,
2001). Dais within each combination of strata were selected randomly. Although the sampling
regime reported here appears to have been stratified by dai yield in this season, examination of
records in the database suggest that dai yield stratum was only introduced from the 1998–99
season onwards.
Stratum
No. of dais
High Yield
8
Low Yield
60
Peak Period
Low Period
Dais sampled
8
Dais sampled
8
Hauls/dai/day
10
Hauls/dai/day
10
Total
80
Total
80
Dais sampled
14
Dais sampled
7
Hauls/dai/day
10
Hauls/dai/day
10
Total
140
Total
70
Total hauls
220
Total hauls
150
In 1998–99, the Department of Fisheries introduced another major stratum based on the location
of the rows in two administrative zones: Kandal Province and Phnom Penh Municipality (Ngor and
van Zalinge, 2001). Further stratification was included during the 2000–01 fishing season prompted
by MFCF interest in the species composition of the landings by dais in row 2. During this season only,
Row 2 and ‘all the other dai units’ formed an additional stratum (Table 9).
In the 2001–02 season, dai row 1 (three individual units in Phnom Penh Municipality) was
Decommissioned reducing the total number of dais sampled to 65. The survey design has remained
unchanged since 2002–03 with three main sampling strata: (1) Administrative Zone (Kandal Province
and Phnom Penh); (2) Dai Yield (High Yield and Low Yield); and (3) Lunar Period (Peak Period and
Low Period) (Table 10).
The number of dai units sampled each season has remained relatively constant since 1999–2000
(Table 11). Prior to this season, the sampling effort tended to grow each year. Since 1998–99 almost all
dais appear to have been sampled at least once each season.
Prior to 2004, dais were randomly selected within each stratum. After 2004, recommendations were
made to sample the same dais each year as those selected for the 2003–04 season to maintain temporal
continuity (Sopha Lieng, pers. comms.). However, judging from the temporal and spatial variation in
sampling effort (see Annex Table 25), these recommendations appear not to have been adopted.
Page 42
Routine monitoring activities and survey methodologies
3.3. Spatial and temporal variation in sampling effort
In addition to changes to the survey design, there has been considerable inter-seasonal variation in
the sampling effort as measured by the total number of hauls sampled by the data collectors over the
survey period as indicated by the counts of data entry forms, or ‘DOCs’ in the database (Figure 16).
At the commencement of surveys in 1994, no more than 18 hauls were sampled from 18 separate
dais. Thereafter, the total number of hauls sampled increased steadily to a maximum of 2,426 over the
2001–02 season. From 2003–04 onwards, sampling effort declined to around half this value (1,261)
and aside from a peak in effort during 2007–08, sampling has remained consistent at around
1,200–1,500 samples per season until present.
From 1997–98, the total number of hauls sampled in each month has varied between four and
approximately 650 (Figure 17 a). Most of the sampling effort was focused on the peak fishing months
of November, December and January. Within each month, sampling effort also varied with lunar phase
and was most intense during the Peak Period–approximately 6 days before the full moon or second
lunar phase (Figure 17 b). This is consistent with the sampling protocol which prescribes an increase
in sampling effort (i.e. sampling over a 24-hour as opposed to a 12-hour period) and personnel (i.e. the
employment of four as opposed to two data collectors in each administrative zone).
Sampling effort can also be seen to vary longitudinally (upstream/downstream) as reflected in the
counts of the total number of hauls sampled in each row and season (Figure 18 a). The general trend is
for a decline in sampling effort with distance downstream. In most seasons, but particularly 2000–01
and 2001–02, Row 2 has been sampled most frequently. Rows 9 and 13 tend to be sampled less
frequently in Kandal Province, particularly in the later seasons.
In contrast to the total number of hauls sampled by row and season, the mean number of hauls
sampled per dai in Kandal Province (Rows 7–15) were found to be higher than those in Phnom Penh
(Rows 1–6) (Figure 18 b). This is true for at least five of the twelve seasons shown (1998–99,
1999–2000, 2001–02, 2002–03 and 2004–05), but is most pronounced in 2001–02 and 2002–03.
For the former case, the mean number of hauls sampled per dai in Kandal Province was twice that of
Phnom Penh.
Page 43
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 9 Kandal Province (43 units)
No. of dais
10 (2.9) Hauls/dai/day
7 (11) Dais sampled
11 No. of dais
110 (25.1)
10 (3.5)
11 (11)
Low Period
Dais sampled
70 (19.2) Samples/month
Peak Period
Hauls/dai/day
11
Samples/month
32 No. of dais
- (3.4)
28 (27)
32
No. of dais
10 (3) Hauls/dai/day
15 (18) Dais sampled
Row 2 dais
4* No. of dais
Low Period
4
Monthly dai sampling regime proposed for the 2000–01 fishing season (Ngor and Zalinge, 2001). In this season, the Row 2 inclusion/exclusion
stratum was added. For each dai yield and lunar period stratum combination the following was reported: No. of dais: the total number of dais
in each stratum combination; Dais sampled: the planned number of dais to be sampled each month (actual counts of the number of dais sampled
each month from the database are reported in brackets); Hauls/dai/day: the proposed number of hauls/dai/day to be sampled (the mean number
of hauls/dai/day in each month estimated from the database are reported in brackets), Samples/month: the proposed total number of samples/
month (the mean number of samples collected in each month estimated from the database are reported in brackets). Within each combination of
strata dais were selected randomly.
Minor Stratum:
High Yield
Low Yield
Dais sampled
280 (55)
Peak Period
Phnom Penh Municipality (25 dai units)
150 (21.2) Samples/month
Hauls/dai/day
Samples/month
All dais excluding Row 2
21* No. of dais
40 (4)
Low Period
10 (7.9)
21 No. of dais
14* (4) Dais sampled
Peak Period
10 (11.3) Hauls/dai/day
No. of dais
10 (5.3) Hauls/dai/day
40* (24) Dais sampled
400 (28.6)
10 (4) Hauls/dai/day
140 (47.1) Samples/month
20 (23) Dais sampled
400 (56.6) Samples/month
Dais sampled
200 (60.6) Samples/month
Hauls/dai/day
Samples/month
Note: the same dai may be sampled on several occasions in the same season.
*There are inconsistencies between the total number of dais in each stratum combination (No. of dais) and the proposed number of dais to be sampled (Dais sampled) since it was expected that the
same dai would be sampled on several occasions on the same day (Mr. Pengbun Ngor, pers comm.).
Page 44
5 (5.2) Hauls/dai/day
145 (66.6) Samples/month
10 (5.4) Hauls/dai/day
250 (69.6) Samples/month
Hauls/dai/day
Samples/month
Dais sampled
32 No. of dais
29 (29) Dais sampled
32 No. of dais
25 (27) Dais sampled
No. of dais
55 (34.3) Samples/month
5 (5.7) Hauls/dai/day
100 (46.5) Samples/month
10 (6.4) Hauls/dai/day
Samples/month
Hauls/dai/day
11 No. of dais
11 (11) Dais sampled
11 No. of dais
10 (11) Dais sampled
6 No. of dais
180 (55) Samples/month
11.3 (5.4) Hauls/dai/day
16 (16) Dais sampled
16 No. of dais
100 (43.1) Samples/month
16.7 (7.8) Hauls/dai/day
6 (6) Dais sampled
Low Period
Phnom Penh Municipality (22 units)
Peak Period
Dais sampled
Low Period
No. of dais
Note: the same dai may be sampled on several occasions in the same season.
Low Yield
High Yield
Peak Period
Kandal province (43 units)
140 (43)
8.7 (4)
16 (17)
16
60 (23.1)
10 (4.9)
6 (7)
6
Monthly dai sampling regime planned for the 2002–03 fishing season. The Row 2 stratum was dropped from the sampling stratification and
the stratification in this season reflects the stratification at the time of writing. Within each stratum dais were randomly selected for sampling.
For each dai yield and lunar period stratum combination the following was reported: No. of dais: the total number of dais in each stratum
combination; Dais sampled: the planned number of dais to be sampled each month (actual counts of the number of dais sampled each month
recorded in the database are reported in brackets); Hauls/dai/day: the proposed number of hauls/dai/day to be sampled (the mean number of
hauls/dai/day in each month calculated from the database are reported in brackets), Samples/month: the proposed total number of samples/month
(the mean number of samples collected in each month calculated from the database are reported in brackets). Within each combination of strata
dais were selected randomly.
Minor Stratum:
Table 10 Routine monitoring activities and survey methodologies
Page 45
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 11
Summary of changes made to the dai fishery survey design (1994–2008) and numbers of dais
sampled each season.
Season
Dais sampled
Database structure and features
1994–95
Stratified by lunar period
Low sample effort
16
1995–96
Stratified by lunar period
No dai identification reported
1996–97
Stratified by lunar period
Census conducted to determine dai yield
48
1997–98
Stratified by lunar period
Dai yield stratum not yet introduced
59
1998–99
Stratified by lunar period
Administrative zone stratum introduced
Dai yield stratum introduced (20 October 1998)
60
1999–00
Unchanged from above
64
2000–01
Stratified by lunar period, administrative zone and dai yield
Row 2 dais in Phnom Penh stratum sampled separately
Row 1 discontinued
67
2001–02
Stratified by lunar period, administrative zone and dai yield
Row 2 dai stratification not applied
64
2002–03
Unchanged from above
68
2003–04
Unchanged from above
66
2004–05
Unchanged from above
64
2005–06
Unchanged from above
62
2006–07
Unchanged from above
64
2007–08
Unchanged from above, but separation of the catch into ‘Large Fish’ and ‘Small
Fish’
65
no records
3000
2000
1500
1000
08-09
07-08
06-07
05-06
04-05
03-04
02-03
01-02
00-01
99-00
98-99
97-98
0
96-97
500
94-95
No. of hauls sampled
2500
Season
Figure 16 Page 46
Changes in sampling effort between the 1994–95 and 2008–09 expressed as the total number of
hauls sampled in each season.
Routine monitoring activities and survey methodologies
(a)
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
Total no. hauls sampled
600
400
200
0
Total no. hauls sampled
600
400
200
0
Total no. hauls sampled
600
400
200
0
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Month
Month
Month
Month
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
Feb
Mar
(b)
Total no. hauls sampled
1250
1000
750
500
250
Total no. hauls sampled
1250
1000
750
500
250
Total no. hauls sampled
1250
1000
750
500
250
1
2
3
Lunar Phase
Figure 17 4
1
2
3
Lunar Phase
4
1
2
3
Lunar Phase
4
1
2
3
4
Lunar Phase
Total number of hauls sampled each (a) month and (b) lunar phase (Quarters 1–4) each season.
Page 47
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
(a)
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
Total no. hauls sampled
500
400
300
200
100
Total no. hauls sampled
500
400
300
200
100
Total no. hauls sampled
500
400
300
200
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row number
(b)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row number
97-98
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row number
98-99
Row number
99-00
00-01
No. of samples/dai
8
6
4


2




























































0
01-02
02-03
03-04
04-05
No. of samples/dai
8
6









4


2











































0
05-06
06-07
07-08
08-09
No. of samples/dai
8
6
4




























2




























0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row
Figure 18 Page 48
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Row
(a) Total number and (b) mean number of hauls per dai (±SD) sampled by row each season. The
broken vertical line indicates the boundary between Phnom Penh Municipality (Rows 1–6) and
Kandal Province (7–15).
Routine monitoring activities and survey methodologies
3.4. Current survey methodology
The following sections give further details of the current survey design and enumeration methods
employed since 2005–06 designed to generate estimates of total catch weight and value by species
landed by the fishery each season.
3.4.1. Sample stratification
Nine rows of dai operate in Kandal Province comprising 11 High Yield units: 10A, 10B, 10D, 10E,
11A, 12A, 12B, 12C, 13A, 14A and 14B (Table 12). The remaining 27 units are Low Yield units
including three additional units which are often seen operating in this stratum. Within the Phnom
Penh Municipality, there are five dai rows and six High Yield units: 2A, 2B, 2C, 2D 3C and 3D. The
remainder (16) are Low Yield units (Table 12). The Peak Period commences seven days before the full
moon (second quarter) and ends one day before full moon. All days that fall outside of this period are
classified as the Low Period. Sampling occurs on approximately 17 days each month. During the Peak
Period, dais are sampled every day and during the Low Period, every second or third day.
Table 12
The relative locations of High (shaded cells) and Low Yield (unshaded cells) dais.
A’ = additional dai that came into operation subsequent to the rows having been labelled.
Province
Relative transversal positions of dai nets
in the Tonle Sap channel
Row No.
Row 15
Kandal Province
B
C
B
C
Row 12
A
B
C
D
E
Row 11
A’
A
B
C
D
Row 10
A
B
C
D
E
F
G
Row 9
B
C
D
Row 8
B
C
D
E
F
G
H
C
D
E
F
G
Row 14
A
Row 13
A
F
5
3
G
Row 5
6
5
7
3
9 rows
7
5
42
Row 6
Phnom Penh
Municipality
E
1
Row 7
Sub-Total
D
Total number
dai units
B
C
D
E
F
C
D
G
5
E
F
Row 4
A
B
C
D
5
4
Row 3
A
B
C
D
4
Row 2
A
B
C
D
4
Sub-total
5 rows
22
Grand total
15 rows
64
Page 49
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
3.4.2. Data collectors and data collection process
Survey enumerators are trained and supervised by IFReDI. Typically, two enumerators are assigned
to each administrative zone each sampling two dais per day. For each dai, they sample 3–4 hauls
over a 12-hour period. During the seven days corresponding to the Peak Period each month, sampling
occurs over a 24-hour period. Additional enumerators are employed during this period–three in Kandal
Province and two in Phnom Penh Municipality. During this period, eight, instead of four dais are also
sampled daily in each zone. However, day and night samples are not distinguished in the database.
The duration of the ‘Peak Period’ does not appear to be fixed each month, rather it appears to be
subjectively identified by survey supervisors on the basis of changes to catch rates.
3.4.3. Variables enumerated
Catch per haul
The catch per haul is the total weight (kg) of all species combined in each haul, or lift of the dai net.
During the Low Period the total catch per haul is relatively small and can be weighed using a five
kilogram scale. All individual fish and each species can be enumerated during this period. During Peak
Periods, however, catches per haul are usually too high to be weighed on a scale. In these instances,
a visual estimate of the fish catch per haul is made by the enumerator with guidance from the dai
operator. Three alternative methods for visually estimating the weight of the haul are applied:
1) Visual estimation per codend: Enumerators estimate the proportion of the codend that is
filled by the haul. The weight of a full codend is known (100–200 kg depending on the
design) and the total weight of the hauls is reported as a fraction of this;
2) Visual estimation per basket: If the quantity of fish in a single haul exceeds the volume of
the codend, the data collectors count the number of baskets sold directly to buyers waiting
in boats beside the dais. The weight of a single basket is known and the total catch per
haul can then be estimated as the product of the unit weight and the number of baskets
landed; and
3) Visual estimation per boat: Very large hauls are loaded directly into the boats of the
buyers. In these cases, enumerators report landed weight estimated by the dai operator on
the basis of the capacity of the boat.
Species composition and sub-sampling
If the haul weight is less than 5 kg, the total haul is sampled. Individual fish in the haul are separated
into species and the number and total weight of fish belonging to each species recorded. For haul
weights greater than 5 kg, sub-samples totalling between 5 kg and 15 kg are first taken from the haul.
Page 50
Routine monitoring activities and survey methodologies
Total length (TL) of individual fish
Within each sample or sub-sample, ecologically and economically important species (large fish, high
value or high demand) are selected for length and weight measurement. The total length (TL) of each
fish is measured with a measuring board accurate to one millimetre. Large fish are measured to the
nearest centimetre and small fish are measured to the nearest 0.5 cm. It is left to the discretion of the
enumerators to decide which species are important.
Fish size
This variable was introduced in the 2007–08 fishing season to determine the relative proportion
of large and small fish. Large fish comprise a smaller proportion of the total weight of the catch,
but command a higher price. They are therefore usually kept alive by the dai operators in a cage
suspended under the floating platform of the dai. Fish of the same species are first sorted into large and
small categories before recording their number, weight and unit value.
The weight categories as they are entered into the data entry form are therefore:
A = Total Weight of all fish per haul
B = Total Weight small fish
C = Sample Weight of small fish (sub-sample of B)
D = Total Weight big fish
Fishing effort
Effort is measured as the number of hauls per dai per day. This is estimated by the enumerator from
the average observed soak time or interval between each successive haul over the 12 or 24 hours
monitoring period. Enumerators correct the estimates for time spent repairing or cleaning nets during
the unobserved period following the advice of the dai operator.
To estimate the total landings made by the fishery each month from the sampled catch rates, the
following additional measures of fishing effort are reported:
1) Active fishing gear: The number of dais operating during the survey month. Enumerators
record the number of dais observed operating on sampling days to provide a mean
estimate for the month; and
2) Active fishing days: The number of days corresponding to the Peak and Low Period each
month subjectively estimated from field observations and daily catch rates.
Initially these raising factors were informally reported but they are now required to be recorded on
data collection forms.
Page 51
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Fish Prices
Unit prices are recorded for each species. Fish prices typically change each year according to supply
(Sopha Lieng pers comm.) and are used to estimate landing values.
3.5. Catch estimation methodology
3.5.1. Aggregated total catch
The estimation of total catch for the dai fishery is based on the principles outlined in
Stamatopoulos (2002). The dai unit forms the primary sampling unit. The daily catch rate
(CPUE) of a dai unit is estimated as the product of the sampled weight for haul, i and the
estimated number of hauls in a day:
CPUEd,m,p,l,y,Dai,i =wtm,p,l,y,Dai,i .hd,m,p,l,y,Dai
Where d = day, m = month, p = administrative zone (province), Kandal (KAN); Phnom Penh (PP),
l = lunar period, y = dai yield category, Dai = individual dai unit, wt = weight of haul, and
h = estimated number of hauls in a day. The mean daily CPUE (mean total weight dai-1 day-1) for any
given stratum combination within a month is then obtained by averaging across all dai unit samples
in each stratum combination.
The monthly estimate of the total catch ( C ) is given by:
Cm,p,l,y = CPUEm,p,l,y x Effortm,p,l,y
The total monthly fishing effort (Effort) is:
Effortm,p,l,y =ADm,p,l,y x AGm,p,l,y
Where AD is the estimate of active days and AG is the estimate of active gears. The total catch for a
month is therefore given by:
p=KAN l=2 y=h
TotalCatchm= ∑
∑∑
p=PP l=1 y=l
The total catch for the season (TotalCatch) is then obtained by summing the monthly catches:
m=Mar
TotalCatch = ∑
m=Oct
Page 52
Routine monitoring activities and survey methodologies
SEASON (TOTAL CATCH)
Month (Total Catch)
Phnom Penh Municipality
Kandal Province
High Period
Low Period
High Yield
Low Yield
Dai 1
Big fish sample
Small fish sub-sample
Total haul
Figure 19 Dai 2
Dai 3
Species name
No. of fish
Body weight
Length (selected species)
Mean daily CPUE per stratum
Mean daily Effort per stratum
Outline of the stratification of the dai fishery sampling regime. The mean catch weight per haul
(CPUE) is calculated for a dai on a day (large shaded area) within each stratum. The total catch is
obtained by multiplying the stratum-specific estimate of the mean daily CPUE by the two stratumspecific raising factors: the active dais and active days.
Page 53
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
3.5.2. Species-wise CPUE and catch
A sub-sample of small fish is taken from each haul that is too large for a total sample, whereas all
the big fish are removed from the haul as a total sample. These two samples, i.e. the total sample of
large fish and sub-sample of small fish, are then separated into species, enumerated, weighed and
selected species are measured to obtain Total Lengths (TL) (Figure 19 and Section 3.3.3). The species
composition of the total catch in terms of weight and abundance can therefore be estimated from the
proportion of species present in the small fish sub-sample to which the big fish total sample is later
added.
Page 54
4. The Dai Fishery database: storage and processing
4.1. Database evolution
Between 1994 and 2006, the data recorded by the data collectors was stored in ARTFISH software
(Stamatopoulos, 1994; 1995). Over this period two successive versions of ARTFISH were used: the
DOS version which stored data files in Dbase IV (*.dbf) format (1994–2001) and the Windows version
(2001–2006) which was stored in text file (*.txt) format. Individual data files in ARTFISH were stored
in nested folders tiered by administrative province, fishing season and month (Figure 20). The data
files themselves were separated into fishing effort (EF), fishes landed (LN) and estimates of total
catches (ES). In total, ARTFISH generated 500 files analysable only by the software itself. Additional
limitations of the earlier DOS version of ARTFISH included the fact that: (1) it only captured 38
of the numerically most dominant species (Cans and Ngor, 2006) and grouped the remainder as
‘X-OTHERS’, and (2) that a maximum of only 20 species could be reported for any given landing
(Baran et al., 2001b).
Figure 20
Structure of the database files held in the DOS version ARTFISH showing folders tiered by
Administrative Province, season and month. Alphanumeric file labels are coded by (1) the type of
file: fishing effort (EF), fishes landed (LN) and estimates of total catches (ES) and (2) landing site
(2 digits), month (2 digits) and year (2 digits) (Cans and Ngor, 2006).
Page 55
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Baran et al. (2001b) identified the need to transform the database into a form that would enable
analyses by means of a range of different software packages. They developed an access function that
merged the individual ARTFISH files into a single database and restructured the species table using the
‘restructure’ function in SPSS Version 13. A detailed description of the steps taken to merge the DOS
Version files and the Access source code used to do this can be found in Appendix A of Cans and Ngor
(2006).
Cans and Ngor (2006) identified several problems with the storage of effort data in the original
DOS version of ARTFISH. These included: (1) the fact that mean number of hauls per day was
calculated manually prior to the data being entered into the system, (2) the active days raising factor
was only available in printed reports and (3) mean number of hauls per day, active days and active
gears were missing in several instances in the DOS version. To resolve this, Cans and Ngor (2006)
estimated the mean number of hauls per day for a gear type and period on the basis of prior records.
The basic structure of the files and folders storage system of the Windows version of ARTFISH was
different from the DOS version, with the primary tier being the fishing season rather than province and
different file types being stored in separate folders rather than being differentiated by file name (Figure
21). The merging into an Access database appears to have been more straightforward than it was in the
case of the DOS version.
Problems identified in the Windows version of ARTFISH
included: (1) species lists and species codes appeared to be
mismatched from one month to the other; (2) species lists
and codes were occasionally duplicated and (3) there were
inconsistencies in the spelling of Khmer common names of
fish species in the two lists. The species lists were therefore
corrected and linked to a standard species table used by the
AMCF project which incorporates the scientific name and
general information on the species which was obtained from the
Mekong Fish Database (2000).
The database developed by Cans and Ngor (2006) contains
three types of data tables. These contain the sampling
information and primary data collected, four lookup tables that
contain gear, species and sampling season information and
two additional tables that contain data on water level and lunar
phase (Table 13). Descriptions of the fields in each of these
tables appear in Annex, Table 26 – Table 34.
Figure 21 Page 56
Structure of the database files held in the Windows version ARTFISH showing folders tiered by
Season, Administrative province, month and file type (effort, landings, results and look up tables).
The Dai Fishery database: storage and processing
Table 13
Description of the main tables in the dai fishery database developed by Cans and Ngor (2006) and
in use between 2006–2009.
Tables
Description
Data Tables
tbl_MainWin&Dos
Main table: Stores all the sample information: date, location, gear used,
total weight of the haul, weight, value, price and fish number of the sample,
number of hauls per day.
tbl_SpeciesWin&Dos
Sample species table: Stores species information caught in the sample,
species name, code, weight, number, value and price.
tlb_Effort
Effort table: information on the effort, date, stratum, gear code,
active gears and active days.
Lookup Tables
tlkp_GearCode
List and code of the fishing gear used
tlkp_Species
Species list in Khmer with the MFD, 2000 corresponding code
tlkp_SeasonYear
This table provides the season year according to the date
tlkp_SpeciesStandard
List of fish names both English & Khmer (translation), general species
information issued from MFD, 2000
Additional Tables
tbl_PhnomPenh Port Water Level
Water level at Phnom Penh Port from 1994 to present.
tbl_moonFace
Daily moon face percentage table added for further analysis
Figure 22 Entity-relationship diagram for the dai fishery database developed by Cans and Ngor (2006) and in
use between 2006–2008. Bold text indicates primary keys.
Page 57
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
The primary keys in the main database tables (bold text in Figure 22) linked via one-to-one or
one-to-many relationships include:
(1) DOC: the document number of the paper record;
(2) Date: date the sample was collected;
(3) StratumNum: administrative zone (Kandal or Phnom Penh);
(4) GearCode: an alphanumeric gear code that incorporates the lunar phase and dai yield
(B0001–B004) as well as the stratifications applied in the earlier sampling years
(B0005–B0010) i.e. (a) no stratification (All DAI) applied prior to the 1997–98 season
and (b) the stratification scheme that included or excluded Row 2 (Row2/Excl2) that
was applied between the 1998–99 and 1999–00 seasons (Appendix II, Table 10–11);
and
(5) KhmerNameCode: Khmer species name linking the samples of individual species with
the standard species table.
4.2. Description of the current database
Since Cans and Ngor (2006) developed the Access version of the dai fishery database, it has
undergone several further modifications resulting from minor changes to the sampling regime as well
as the inclusion of length frequency tables (Table 14).
4.2.1. Data tables
As in the original Access database, the Main Table (tbl_MainWin&Dos), describes sample information
relating to the date of the sample, the gear used (DaiID, dai size and mesh size) and gear code (stratum
combination) (Table 14). It also contains aggregated catch values including: the measured or estimated
total weight of each haul (TotalWeight) and the total weight of the samples taken from the haul. Each
row in the database relates to a single sampled haul which can be traced to the original data entry form
via a document number recorded in the DOC field. The four primary keys that link the hauls to other
tables in the database are the same as those in the original database: DOC, date and GearCode and
StratumNum (Administrative zone).
In addition to the way in which information about the catch is recorded, other minor changes include
the manner in which dais are identified and described. In the original database, the size of the dai was
described by a single figure. This, ‘SizeDai’ field in the original database is now split into three fields
that describe the depth, width and length of the dai and the minimum mesh size as measured at the
codend of the bagnet. The ‘Skipper’ field has been replaced by ‘DaiID’ that contains the alphanumeric
code for the row number (1–15) and the transversal position across the channel (A–G). Many of the
fields remain unchanged (Annex Table 36).
Page 58
The Dai Fishery database: storage and processing
Table 14 Description of the main tables in the dai fishery database 2009. Asterisks indicates tables added
subsequent to Cans and Ngor (2006).
Tables
Description
Data Tables
tbl_MainWin&Dos
Main table: Stores all the sample information: date, location, gear used, total
weight of the haul, weight, value, price and fish number of the sample, number of
hauls per day
tbl_SpeciesWin&Dos
Sample species table: Stores species information caught in the sample, species
name, code, weight, number, value and price
tlb_Effort
Effort table: information on the effort, date, stratum, gear code, active gears and
active days
Lookup Tables
tlkp_GearCode
List and code of the fishing gear used
tlkp_SeasonYear
This table provides the season year according to the date
tlkp_Species
Species list in Khmer with the MFD, 2000 corresponding code
tlkp_SpeciesStandard
List of fish names both English & Khmer (translation), general species
information issued from MFD, 2000
Length Frequency Tables
Leng_tbllengthFreq*
Species, total length and number of fish in the length interval
Leng_tblSpecies*
Species, total weight of the haul from which the sample was taken, weight of the
sample, weight of the sample used to obtain length information
Leng_tblLocation*
Administrative zone from which the length data was obtained
Key additional Tables
Annual Hydrological Indices
Flood Indices, water levels from 1994 to present, flood start and end,
flood duration
tbl_LunarAge&Phase
Daily moon age in days from the new moon and phase
tblOtherInfo*
Beginning and end of the bagnet soak time, total catch per haul, number of hauls
The second principal data table in the database, referred to as tbl_SpeciesWin&Dos, is linked
to the Main table by the four primary keys and contains information pertaining to the abundance
(FishNumber) and weight (Catch) of individual species in the sample (Annex Table 37). Also in this
table are fields containing information on the value and price of the individual species. The 2009
format provides for the separation of fish samples into ‘big’ and ‘small’–a stratification which was
introduced in 2007 and therefore incorporates three additional fields namely: sample weight of the
small fish (S_Sampled weight), total weight of the small fish (S_TotalWeight), and total weight of
the big fish (B_TotalWeight). The third principal data table (tbl_Effort) in the 2009 database contains
information on the active days and active gears and remains unchanged from the original database
(Annex Table 38).
4.2.2. Lookup Tables
The data tables are linked to four look-up tables that contain information linking the season, date and
year (tlkp_SeasonDateYear), gearcodes (tlkp_GearCode) and species information (tlkp_Species and
tlk_SpeciesStandard) (Annex Table 39–Table 41).
Page 59
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
4.2.3. Length Frequency tables
The three length frequency tables were incorporated into the database in 2007. These tables detail the
length and number of individuals of each species sampled (Leng_tbllengthFreq), the weight of the
species sample selected for length measurements (LengtblSpecies), and location (Leng_tblLocation)
(Annex Table 42–Table 44).
4.2.4. Additional tables
The key additional tables in the current database contain information on flood magnitude, timing and
duration (Annual Hydrological Indices), lunar age (in days) and phase (tbl _LunarAge&Phase). Other
tables include daily records of water levels at Prek Kdam (Pkdam), Phnom Penh port (Ppport), and
Kampong Luong (Kluong). The
table ‘Dai units by row & Et’
lists the number of dai units by
season and row (Annex
Table 45–Table 48).
Figure 23 Entity relationship diagram for the Dai fishery database in
use from 2009.
4.3. Current Database Query descriptions
The queries in the database have been designed to extract information on individual dai and daily
catch rates, to estimate the total catch for each stratum (e.g. month and season), as well as to calculate
these statistics for individual species. A number of alternative approaches (queries) are available
according to different assumptions regarding the structure of the data itself and/or the reliability of
certain variables such as information on fishing effort. Thus variation from the application of the
different methods represents “model” or “structural” uncertainty–an additional form of uncertainty in
catch estimation that exists alongside the statistical uncertainty associated with estimates of variables
of interest derived from the data. Details of these queries are described in a working document held at
IFReDI.
Page 60
5. The Ecology of the Fishery
5.1. Longitudinal (upstream/downstream) variation
5.1.1. Aggregated catch, effort and CPUE
On average, most (76%) of the catch is landed in Kandal province, upstream of Phnom Penh, where
the majority (64%) of the dai units are located (Table 15). The resulting catch rates in Phnom Penh are
on average less than half the rates estimated for Kandal, indicating a substantial difference (decline) in
fish abundance between Kandal and Phnom Penh.
Table 15 Estimates of aggregated catch and effort by season and municipality.
Catch
Season
Phnom
Penh
(kg)
98–99
5,951,708
99–00
00–01
Effort
Kandal
(kg)
Total (kg)
Phnom Kandal
Phnom
Kandal
Total
Phnom Kandal
Penh
(%)
Penh
(Dai units) (Dai units) Penh
(%)
(%)
(Dai units)
(%)
5,835,913
11,787,621
50%
50%
25
38
63
40%
60%
3,389,174
8,892,105
12,281,279
28%
72%
25
38
63
40%
60%
5,723,004
23,417,799
29,140,803
20%
80%
25
38
63
40%
60%
01–02
6,249,828
12,327,838
18,577,667
34%
66%
22
38
60
37%
63%
02–03
5,043,345
10,675,150
15,718,495
32%
68%
22
39
61
36%
64%
03–04
2,529,287
8,067,296
10,596,583
24%
76%
22
39
61
36%
64%
04–05
4,368,836
21,526,992
25,895,828
17%
83%
22
40
62
35%
65%
05–06
5,924,983
32,801,968
38,726,952
15%
85%
22
42
64
34%
66%
06–07
2,947,799
19,444,630
22,392,429
13%
87%
22
42
64
34%
66%
07–08
1,939,297
11,584,989
13,524,286
14%
86%
22
42
64
34%
66%
08–09
1,792,376
11,217,913
13,010,289
14%
86%
22
42
64
34%
66%
Mean
4,169,058
15,072,054
19,241,112
24%
76%
23
40
63
36%
64%
A detailed examination of mean sampled catch rates plotted as a function of row number or cumulative
effort measured from the most upstream row of dai units closest to the Lake is indicative of a depletion
of fish (Figure 24). The depletion response is most pronounced between rows 13 and 6 reflected in the
significant linear decline in catch rates.
Removals of fish by other small-scale gear such as gillnets and seines operating between the dai rows
may have contributed to this apparent depletion. However, information about this component of the
fishery is sparse and poorly documented (see Section 1.1.4).
Page 61
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
The depletion response is non-linear over the range of the fishery. Fish may become more vulnerable
to capture (i.e. gear catchability increases) where catch rates rise between rows 15 and 13 and again
through rows 6 to row 2, possibly as the result of changing hydrological and morphological conditions
which favour fish capture. Differences in small-scale fishing effort and/or migrations of fish from
adjacent floodplains between dai rows might also be responsible for this non-linear response.
Figure 24 Mean ln transformed catch rates (1997–2009) plotted by row.
Assuming that both: (i) migrations of fish from adjacent floodplains to between-dai row locations and
(ii) removals of fish by any small-scale gears operating between the dai rows, are negligiable, then the
observed depletion response can be described by the depletion model of Delury (Hilborn and Walters,
1992). This model can provide approximate estimates of the mean numbers of fish arriving at the dai
fishery during each season, the catchability coefficient of a dai unit, and the rates of exploitation by
the fishery. The catchability coefficient, (q) is a measure of gear efficiency and can be defined as the
proportion of the population removed by one unit of effort. Since we are considering changes in fish
abundance indicated by catch rates through rows of dais, (r) instead of through time (t), the Delury
model is expressed as:
CPUEr = qN15e – qEr
Where N15 is the number of fish arriving at row 15 and Er is the cumulative fishing effort measured in
dai units from the most upstream row 15. Thus, E15 = 0.
Page 62
The Ecology of the Fishery
Taking logarithms, we then get:
ln CPUEr = ln[qN15] – qEr
This form of the model can be fitted with linear regression methods where ln CPUEr is the
dependent variable and Er is the independent variable. The regression coefficient (slope) gives the
estimate of (q) (Figure 25). The average number of fish arriving at the dai fishery during a fishing
season is estimated from the regression intercept a and the slope, b: N15 = ea/-b.
Figure 25 The Delury depletion model fitted to mean ln transformed catch rates (1997–2009) expressed as
the number of fish caught per dai unit per day, and cumulative fishing effort measured in dai units
from the most upstream row 15.
The regression model was found to be highly significant (p < 0.001) with slope, b = - 0.028 and
intercept, a = 10.745. The average number of fish arriving at the dai fishery each day of each fishing
season, N15 is estimated to be 1,657,052.
The model predicts that each dai removes approximately 2.8% of fish migrating through the
fishery equivalent to an instantaneous fishing mortality rate, F (F = Eq) of 1.79 for 64 dai units. In
other words approximately 83% of fish arriving at the dai fishery are caught. This compares with
95% estimated for the 1938–39 fishing when 106 dais were licensed to fish. Estimates of (F) and the
proportions of migrating fish escaping and removed for a range of effort are given in Table 16.
Page 63
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 16 Estimates of fishing mortality, (F) for the dai fishery for a range of fishing efforts.
Dai Units
F (per season)
Proportion Escaped
Proportion Captured
0
0
1.00
0.00
10
0.28
0.76
0.24
20
0.56
0.57
0.43
30
0.84
0.43
0.57
40
1.12
0.33
0.67
50
1.4
0.25
0.75
60
1.68
0.19
0.81
70
1.96
0.14
0.86
80
2.24
0.11
0.89
90
2.52
0.08
0.92
100
2.8
0.06
0.94
110
3.08
0.05
0.95
5.1.2. Catch diversity and composition
The results of the Delury depletion analysis described above demonstrated a strong depletion response
of fish abundance and biomass through the fishery. If some species are selectively removed by the
gear, e.g. if mesh sizes select only larger species, then downstream changes in the species composition
of the catch would be expected.
PRIMER software (Clarke and Warwick, 2001) was used to test for significant differences in species
diversity and assemblage composition among dai rows and also between administrative zone (Phnom
Penh Municipality and Kandal Province). Mean daily dai catch rates (N dai-1 day-1) were employed
as the index of species abundance and dai units were treated as sample replicates. Only those species
forming more than 0.1% of the total catch between 1997–98 and 2008–09 were included in the
analyses.
Species diversity was indicated by the total number of species (Richness, S) and the Shannon Diversity
Index (H) that accounts for both the abundance and evenness of the species present:
k
H = –∑Piln Pi
i=1
where (pi) is the proportion of species i relative to the total number of species present, (k). High values
of (H) are indicative of high evenness or richness. Low values are indicative of low species rich or
evenness i.e. the assemblage is dominated by a one or a few species. Both the counts of the number of
species (S) and calculation of the Shannon Diversity Index were performed.
Non-parametric multi-dimensional scaling (MDS) was used to explore assemblage
(dis)similarity among dai rows and between administrative zones on fourth-root transformed species
abundance estimates. Assemblage similarity was described using the Bray-Curtis similarity
Page 64
The Ecology of the Fishery
index. The similarity percent (SIMPER) routine was used to identify taxonomic groups that most
contributed to sample(dis) similarity. To eliminate the confounding effects of month and lunar phase
(see Section 5.3.2), only catch rate samples corresponding to the peak month (January) and lunar
phase (second quarter) were selected in each season.
The number of species (S) in each season varied between a minimum of two and a maximum of 80
species (mean = 21.2). The Shannon Diversity Index (H) ranged between 1.1 and 2.5. Both (S) and (H)
tended to decrease almost stepwise downstream between Kandal Province (Rows 7–15) and Phnom
Penh municipality (Rows 1–6) (Figure 26). This pattern is particularly pronounced after 2001–02.
In 2001–02, 2002–03 and 2006–07, approximately twice as many species were recorded in Kandal
Province compared to Phnom Penh.
The MDS ordinations confirmed the existence of significant dissimilarities in species assemblages
reported in the two administrative zones (Figure 27 and Figure 28). However, consistent patterns of
dissimilarity among dai rows were not apparent in the ordinations.
In six of the 12 seasons, C. lobatus contributed the highest percentage dissimilarity between
administrative zones except in 1997–98 and 2004–05 (Table 17). P. barroni and
L. lineatus also contributed significantly to assemblage dissimilarity between the administrative zones
in most years. However, these patterns were not consistently evident. For example, during the 1997–98
season, Parambassis apogonoides contributed the most (33.6%) to the assemblage dissimilarity but in
2000–01 and 2005–06, the greatest contribution to the dissimilarity was made by H. cryptopogon (21.9
and 29.2% respectively).
The distance between the most upstream and downstream dais does not exceed 30 km (Section
2.2) and the distance between rows 6 and 7 is no more than 500 m. Differences in (S) and (H) at the
Phnom Penh/Kandal boundary may simply be an artefact of sampling rather than the existence of real
longitudinal changes in species diversity.
Selective depletion of certain species downstream or species-specific catchability differences
between dais was not detected by the multivariate analyses. The observed longitudinal differences
in the assemblage may simply be an artefact of the capacity of sampling teams to identify and report
species present in landings. Different dai sampling teams are responsible for sampling dais in the
two administrative zones and these were never interchanged, thereby increasing the likelihood
inconsistencies in species identification and/or enumeration. That these differences are not consistent
among seasons may be due to changes in sampling personnel over the period of observations.
Page 65
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
(a)
(b)
Figure 26 Page 66
Estimates of (a) species richness (S) (Mean±SE) and (b) Shannon Diversity Index (H) (Mean±SE) for each
row from Row 1 (downstream) to 15 (upstream) for the seasons 1997–98 to 2008–09 averaged across months
and lunar phases. The broken line indicates the boundary between Phnom Penh Municipality (Rows 1–6) and
Kandal Province (Rows 7–15).
The Ecology of the Fishery
1997-98
2D Stress: 0.13
11
12
1998-99
2D Stress: 0.15
5
3
5
14
13
9
7
7
8
9
6 10
10
4
2
10
10
4
2
11
8
2
14
2
117
8
8
12
8
11
8
15
6 5
12
1310
9
14
3
12
9
11
11
4
5
15
1
10
60%
55%
1999-00
2D Stress: 0.13
8
15
15
2000-01
8
4
9
3
12
8
10
6
5
12
2
2
5
2
10
13
10
9
11
8
12
14
14
5
1
6
5
3
10
14
2
4
2 2
2
10
6
6
55%
55%
2D Stress: 0.11
2001-02
4
14
7
14
2
9
3
3
12
7
4
5
3
108
5
11
6
8
2
8
8
8
10
14
2
4
2
7 4
6
9
10 11
3
3
5
2D Stress: 0.11
8
4
12
9
2002-03
4
2
712
3
1
4
3
2
14
15
1
1
5
3
3
1
11
3
5
4
4
14 15
12
2D Stress: 0.12
5
66
6
3
2
5
7
2
5
9
5
3
4
2
9
12
9
14
1111
14
10
3 6
5
8
10
13
12
77
55%
Figure 27 55%
MDS ordinations based on Bray-Curtis index of similarity derived from fourth-root transformed
mean CPUE for individual dais in January during the peak lunar phase (second quarter), 1997–98
to 2002–03 seasons. Black triangles represent Phnom Penh Municipality, circles represent Kandal
Province and replicates are individual dais labelled by Row number. Stress values are reported at
top right and vary from 0.11 to 0.16. Clusters are grouped by percentage similarity and reported at
bottom left. The broken line indicates the split between administrative zones.
Page 67
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2D Stress: 0.15
2003-04
8
9
2D Stress: 0.16
2004-05
10
2
6
9
11
8
15
10
10
12
10
12
4
14
13
4
2
3
10
14
12
6
5
2
15
9
3
11
11
15
12
14
12
7
7 7 13
2
9
10 12
6
3
3
10
8
4
5
2
3
5
4
15
4
6
2
5
2
10
6
3
6
5
11 8
7
15
12
3
4
7
8
8
14
14
6
12
12
2
55%
60%
2005-06
2D Stress: 0.18
10
3 2
5
4
6
6
2
5
52
6
3
5
36
4
3
12
12
10
4
13
4
7
14
118
14
8
14
15
13
8
2
5
15
15
5
6
6
10
12
3
9
15
3
5
4
5
63
3
4
6
2
12
10
4
5
2
15
8
10
7
11
12 11
9
14
10
9
11
14
15 12
7 12
10
7 8
10
7
12
11
7
8
8
10
8
9
2
2D Stress: 0.15
2006-07
8
10
2
11
14
11
9
60%
60%
2007-08
2D Stress: 0.21
14
7
7
14
15
15
14
15
8
12
9
12
2
5
15
4
15
6
13
24
10
2
12
7
11
9 8
10
7
10
11
12
12
8
3
5 6 23
6 10
3
5
8
7
4
8
12
10
Page 68
12
1211
77
12 11
14
15 7
87
8
913
12
14
65
3
9
15
12
6
2
15
7
2
4
10
10
10
3
2
10
11
9
11
11
8
Figure 28 5
6
5 6
4
2D Stress: 0.17
5
11
3
3
8
65%
2008-09
3
5
3
4
5
4
11
60%
MDS ordinations based on Bray-Curtis index of similarity derived from fourth-root transformed
mean CPUE for individual dais in January during the peak lunar phase (second quarter), 2003–04
to 2008–09. Black triangles represent Phnom Penh Municipality, circles represent Kandal Province
and replicates are individual dais labelled by Row number. Stress values are reported at top right
and vary from 0.1 to 0.17. Clusters are grouped by percentage similarity and reported at bottom
left. The broken line indicates the split between administrative zones.
-
-
-
-
18.2
-
33.6
Labiobarbus lineatus
Henicorhynchus cryptopogon
Henicorhynchus siamensis
Puntioplites proctozysron
Clupeichthys aesarnensis
Labeo chrysophekadion
Parambassis apogonoides
-
6.6
-
-
Thynnichthys thynnoides
Osteochilus lini
Parachela siamensis
Yasuhikotakia morleti
-
9.9
Paralaubuca barroni
Labiobarbus siamensis
9.6
4.3
-
3.9
3.9
-
-
-
-
-
-
-
14.9
15.3
45.9
1997–98 1998-99
6.3
1.1
-
3.1
1.2
-
1.8
1.8
-
-
-
6.9
8.2
33.4
1999-00
4.1
4.2
3.7
6.3
6.2
-
-
4.3
-
2.5
21.9
7.5
14.0
9.2
2000-01
4.1
1.8
0.9
1.3
2.8
-
0.7
0.8
0.7
6.8
12.3
12.8
7.3
22.2
2001-02
-
8.1
2.2
2.2
4.4
0.7
0.7
5.6
-
5.7
6.6
1.9
8.3
15.4
2002-03
-
1.5
1.4
3.1
1.2
-
1.7
3.9
-
3.9
13.9
1.9
12.0
27.4
2003-04
-
-
-
-
-
-
-
2.9
-
11.3
11.2
9.1
1.6
4.2
2004-05
-
-
0.8
-
-
-
-
-
-
7.6
29.2
2.9
33.3
9.3
2005-06
-
-
-
2.4
-
-
16.6
-
8.5
1.5
3.6
26.5
4.6
1.6
2006-07
-
-
-
1.6
-
-
3.1
1.9
9.6
6.2
16.8
1.2
14.3
18.9
2007-08
-
2.2
-
-
-
-
11.8
-
32.2
19.8
-
8.8
1.3
7.3
2008-09
1.5
1.6
1.6
2.0
2.1
2.9
3.0
3.2
4.2
6.2
9.6
10.0
11.6
20.0
Mean
%
79.4
77.9
76.3
74.7
72.7
70.7
67.8
64.8
61.7
57.4
51.2
41.6
31.6
20.0
Cum
%
Species contributing up to 80% of mean cumulative percentage (Cum%) dissimilarity between Phnom Penh Municipality and Kandal Province in
each season.
Cirrhinus lobatus
Species
Table 17
The Ecology of the Fishery
Page 69
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 29 Estimates of mean weight of the fish assemblage sampled from the dai fishery by row and season.
Estimates are weighted by numbers of fish sampled.
5.1.3. Fish size (weight)
An analysis of variance (ANOVA) indicated that the mean weight of fish caught using nets with
the most common mesh size (i.e. mesh range from 100 to 15 mm) recorded in the database for both
municipalities, does not vary significantly with dai row (Table 18). Mean weights were however found
to vary significantly with both month and season (see Sections 5.3.3 and 5.4.3). This suggests that the
trends illustrated in Figure 29 are likely to reflect the use of nets with larger meshes in the downstream
rows of the fishery, particularly in the Phnom Penh Municipality (rows 1–6), rather than species or
fish size-specific changes in gear catchability with dai row. However, this could not be confirmed
because of the generally poor quality and paucity of mesh size records in the database but the available
evidence supports this interpretation at least for chieu nets (see Section 2.3). Effort should be directed
towards improving the quality and quantity of these records in the database.
Page 70
The Ecology of the Fishery
Table 18 ANOVA test results for the null hypothesis that the average weight of fish (irrespective of species)
landed by the fishery does not change with dai row number after accounting for the effects of
month, season and mesh size. Mean weight estimates by dai row (1–15), month (October–March)
and season (1998–99 to 2007–08) were weighted by the numbers of individual fish sampled.
Source
Type of
III Sum of
Squarres
df
Mean Square
F
Sig.
Corrected Model
.003a
32
8.839E-05
7.131
.000
Intercept
.002
1
.002
130.777
.000
1.024E-05
1
1.024E-05
.826
.366
SEASON
.002
8
.000
16.149
.000
MONTH
.000
4
6.107E-05
4.927
.001
SEASON *MONTH
.001
18
7.662E-05
6.181
.000
Error
.001
100
1.239E-05
Total
.019
133
Corrected Total
.004
132
ROW
Test of Between-Subjects Effects, Dependent Variable: Weight (kg)
a
R Squared = .695 (Adjusted = .598)
5.2. Variation in catch rates among individual dais
In February 2008, IFReDI undertook a survey over a four day period (9–13 February 2008) to estimate
the depth of water and water velocity below each dai unit as potentially important factors responsible
for the observed variation in catch rates among individual dai units. Depths ranged from 8 to 28 m
and velocity from 0.2 to 0.57 m/s (Table 19). Most dais had fish depths of between 10 and 14 m, but
much deeper water is found below rows 2, 3, 4 and 14 (Figure 30). Water velocities of between 0.3 and
0.5 m/s are common throughout the fishery but below this range only around row 2 (Figure 31). As
expected, water velocity declines with depth (Figure 32). Neither water depth, or velocity was found to
have any detectable effect on catch rates during the survey period (Figure 33 and Figure 34)
Page 71
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 19 Page 72
Estimates of water velocity and water depth below each dai recorded in February 2008.
Dai
Row
Water Velocity (m/s)
Depth (m)
2A
2
0.22
24.0
2B
2
0.20
24.5
2C
2
0.30
26.0
2D
2
0.32
28.0
3A
3
0.21
17.0
3B
3
0.29
17.0
3C
3
0.39
18.5
3D
3
0.31
20.0
4A
4
0.34
15.0
4B
4
0.31
17.0
4C
4
0.34
19.0
4D
4
0.35
19.0
5B
5
0.37
10.0
5C
5
0.43
11.0
5D
5
0.42
12.0
5E
5
0.42
14.0
5F
5
0.41
15.0
6C
6
0.54
16.0
6D
6
0.45
15.0
6E
6
0.43
13.0
6F
6
0.44
10.0
6G
6
0.41
9.5
7C
7
0.36
13.0
7D
7
0.37
14.0
7E
7
0.37
16.0
7F
7
0.45
16.0
7G
7
0.37
13.5
8C
8
0.39
13.0
8D
8
0.39
13.5
8E
8
0.46
15.0
8F
8
0.39
12.5
8G
8
0.32
13.0
8H
8
0.28
10.0
9A
9
0.41
16.5
9B
9
0.32
15.0
9C
9
0.29
13.0
10A
10
0.37
13.0
10B
10
0.41
11.0
10C
10
0.57
10.0
10D
10
0.49
9.5
10E
10
0.44
9.5
10F
10
0.42
8.0
10G
10
0.34
8.0
The Ecology of the Fishery
Figure 30 11A1
11
0.33
12.5
11A
11
0.36
12.0
11B
11
0.36
14.0
11C
11
0.32
13.0
11D
11
0.34
12.0
12A
12
0.35
10.0
12B
12
0.33
11.0
12C
12
0.36
11.5
12D
12
0.35
13.0
12E
12
0.35
12.5
12F
12
0.35
11.0
13A
13
0.28
13.0
14A
14
0.34
22.0
14B
14
0.40
18.0
14C
14
0.34
16.0
Estimates of mean depth below each dai row.
Page 73
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 31
Estimates of mean water velocity at each dai row.
0.60
0.55
0.50
Water velocity (m/s)
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
5
10
15
20
25
Depth (m)
Figure 32
Page 74
The relationship between water velocity and depth (R2 = 0.18; p < 0.001).
30
The Ecology of the Fishery
Figure 33 Mean loge-transformed catch rates during each lunar quarter (1–4) of the survey month (February
2008) plotted as a function of the estimated water depth below the dai unit.
Page 75
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 34 Page 76
Mean loge-transformed catch rates during each lunar quarter (1–4) of the survey month (February
2008) plotted as a function of the estimated water velocity at each dai unit.
The Ecology of the Fishery
5.3. Intra-annual variation
5.3.1. Abundance (CPUE)
The dai fishery operates during the falling water period targeting the migrations of fish as they
move from the Lake to the Mekong mainstream with the receding flood waters. During this period,
fish abundance indicated by daily catch rates of sampled dai units exhibits significant variation
both between and within the six months comprising the fishing season. Catch rates can vary from
approximately 1 kg to 80,000 kg.dai-1.day-1 (Figure 35–Figure 38).
Typically catch rates peak (mean = 10,818 kg.dai-1.day-1; ± SD 18,708) during January
(nine out of the 12 seasons examined) and are lowest (mean = 86.9 kg.dai-1.day-1; ± SD 438.7) in
October (eight out of 12 seasons) (Figure 39 a).
In all the seasons examined, mean catch rates peaked during the second quarter of the lunar cycle
and were lowest during the 4th quarter (Figure 39 b). Here the lunar quarters relate to four consecutive
seven day periods starting from the new (dark phase) moon. Quarter 2 therefore relates to the period of
approximately 7–14 days after the new moon when between approximately 50–100% of the moon is
visible. This period, between what are commonly termed the first quarter and full moon phases, is also
known as the ‘Waxing Gibbous’ phase.
Overall, mean fish abundance indicated by loge-transformed daily catch rates of sampled dais
(1997–98 to 2008–09) varied significantly (p < 0.001) between months and between the four
phases (or quarters) of the lunar cycle. In addition to significant inter-annual variation in mean fish
abundance, fish migrations therefore appear to be strongly influenced by the lunar cycle, and possibly
also water levels as peak migrations typically occur around January or December coinciding with the
end of the flood season.
Page 77
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 35
Page 78
Mean dai CPUE (kg) calculated for each day of the fishing season during (a) 1997–98,
(b) 1998–99 and (c) 1999–2000.
The Ecology of the Fishery
Figure 36 Mean dai CPUE (kg) calculated for each day of the fishing season during (a) 2000–01,
(b) 2001–02 and (c) 2002–03.
Page 79
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 37 Page 80
Mean dai CPUE (kg) calculated for each day of the fishing season during (a) 2003–04,
(b) 2004–05 and (c) 2005–06.
The Ecology of the Fishery
Figure 38 Mean dai CPUE (kg) calculated for each day of the fishing season during (a) 2006–07,
(b) 2007–08 and (c) 2008–09.
Page 81
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
(a)
(b)
Figure 39 Page 82
loge-transformed average daily catch rates (kg/dai/day) by (a) month and (b) lunar phase for the
seasons 1997–98 to 2008–09.
The Ecology of the Fishery
5.3.2 Species diversity and similarity
Sections 6.3.1 and 6.3.2 illustrated how fish migrations tend to peak with each ‘Waxing Gibbous’
phase of the lunar cycle and towards the end of the flood period (December–January). Species richness
in catch samples was found to peak either during November (2002–03; 2004–05; 2006–07 and 2007–
08) or December (five seasons: 1997–98; 1999–00; 2000–01; 2005–06 and 2008–09) (Figure 40 a)
consistent with anecdotes (Chhun Haing Tong, pers comms). The Shannon-Weiner index of diversity
peaked more frequently during November (eight of the 12 seasons) (Figure 40 b). These differences in
the timing of peak values reflect the dominance of one or more species in the catch early in the season.
The species richness of the assemblage caught by the fishery does not appear to vary significantly
during the lunar cycle, although there is some evidence that species richness is marginally higher
during phase 1 and 2 between 2002–03 and 2008–09 (Figure 41 a). The diversity index did not display
any consistent trends or patterns associated with lunar phase (Figure 41 b), suggesting that either the
lunar phase has little or no species-specific effect on migration or it may be that a quarterly (lunar
phase) resolution is insufficient to detect such effects.
However, ordinations illustrating the rank similarities in the assemblage composition between
months in each season during the second quarter of the lunar cycle showed evidence of a change in
composition during the season (Figure 42 and Figure 43). Although these trends were not consistent
among seasons, there was some evidence of assemblages dissimilarity at the start (October/November)
and end (January/February/March) of the season. This apparent shift in the composition of the
assemblage was attributable to: (i) a decline in relative abundance of Clupeichthys aesarnensis at the
start of the season (8 out of the 12 seasons); (ii) an increase in the relative abundance of most other
species, particularly the Cirrhinus, Labiobarbus and Henicorhynchus genera from October to January;
and (iii) a decrease in the relative abundance of these three genera thereafter (Table 20). These patterns
are consistent with the results for the univariate diversity indicators described above.
Ordinations of the rank similarity of the assemblage composition among the four lunar phases in
January each season revealed some evidence of assemblage dissimilarity between the first (Phases 1
and 2) and second half (Phases 3 and 4) of the lunar cycle, but not consistently among seasons (Figure
44 and Figure 45). Nonetheless, these signals do suggest that the intensity of migrations among species
varies during the lunar cycle.
Page 83
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
(a)
(b)
Figure 40 Page 84
Estimates of mean (±95% CI) (a) species richness (S) and (b) Shannon Diversity Index (H) (Mean
±95% CI) of the assemblage sampled during phase 2 of the lunar cycle of each month and season.
The Ecology of the Fishery
(a)
(b)
Figure 41 Estimates of mean (±95% CI) (a) species richness and (b) the Shannon Diversity Index (H) of the
assemblage sampled during each quarter (phase) of the lunar cycle in January of each season.
Page 85
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
1997-98
1998-99
40%
30%
1999-00
2000-01
30%
40%
2001-02
2002-03
30%
40%
Figure 42 Page 86
MDS ordinations illustrating rank similarities in the species assemblage sampled each month
during lunar phase 2, 1997–98 to 2002–03. Samples from phase 1 were used in those cases when
no samples were available for phase 2. Similarities between samples were described using the
Bray-Curtis index using fourth-root transformed mean catch rates sampled from individual dais
(O = October, N = November, D = December, J = January and F = February). Stress values range
from 0.12 to 0.16. Clusters samples are grouped by percentage similarity and these values are
reported at bottom left.
The Ecology of the Fishery
2003-04
2004-05
40%
40%
2005-06
2006-07
40%
40%
2007-08
2008-09
40%
40%
Figure 43 MDS ordinations illustrating rank similarities in the species assemblage sampled each month
during lunar phase 2, 2003–04 to 2008–09. Samples from phase 1 were used in those cases when
no samples were available for phase 2. Similarities between samples were described using the
Bray-Curtis index using fourth-root transformed mean catch rates sampled from individual dais
(O = October, N = November, D = December, J = January and F = February). Stress values range
from 0.1 to 0.14. Clusters samples are grouped by percentage similarity and these values are
reported at bottom left.
Page 87
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 20
November–December
↑
↓
%
December–January
↑
↓
%
January–February
↑
↓
%
February–March
The seasonal frequency of upward (↑) and downward (↓) changes in the abundance of common species between consecutive months, 1997–98 to
2008–09. The average percentage contribution made by each species to the overall sample dissimilarity for each comparison, generated using
the SIMPER routine of the the PRIMER software, is also given.
October–November
%
0.5
↓
1.4
↑
0.4
1
%
0.7
3
1.2
↓
0.3
2
2.6
↑
1
1.5
5
0.6
2.2
4
2
2
4.6
0.6
1
2
5.8
2
0.3
3
3
Acantopsis sp.
1.4
3
2
1
2
0.5
0.8
0.7
Barbichthys nitidus
2
2.3
0.4
4
0.7
0.6
0.3
2
2.8
5.5
1
6.2
0.4
0.6
7
2
1
11
1
Belodontichthys truncatus
Yasuhikotaki helodes
Yasuhikotaki modesta
Yasuhikotaki morleti
Clupeichthys aesarnensis
1
1.5
Cyclocheilichthys armatus
Cyclocheilichthys enoplos
4.4
1.2
Gyrinocheilus aymonieri
2
5
7
1.2
5
2
2.8
3
7
Henicorhynchus cryptopogon
2.4
Cirrhinus lobatus
1
0.3
1.4
6
1
3
1.8
9
0.3
3
4.6
1
3
5.8
Albulichthys albuloides
9
3
1
10
1.6
Amblyrhynchichthys micracanthus
1
3
1
1
6
1
2.6
1
0.5
8
6
3.2
1.9
1
2.8
2.5
2.7
1
5
0.5
2
1
1.3
1.3
3
2.3
0.6
4
0.7
4.3
1
0.5
0.5
1
1
1
0.5
12
1
1.0
7
1
2
6
1.0
0.3
0.7
1
0.6
0.7
1
1.0
15
2
1
2
1
2
11
6.3
1.5
1.4
2
6
10
1
0.4
1.2
4.8
3
0.3
15
2
9
1
1
2.6
0.7
0.3
0.6
3
2
1
Henicorhynchus siamensis
7
2
1
1
1
0.5
0.3
1
1
17
1
Labeo chrysophekadion
Labiobarbus lineatus
Labiobarbus siamensis
Hemibagrus nemurus
Osteochilus lini
Pangasius pleurotaenia
Paralaubuca barroni
Parachela siamensis
Parambassis apogonoides
1
10
1
Parambassis ranga
Puntioplites proctozysron
Rasbora tornieri
Thynnichthys thynnoides
Count of number of species
Page 88
The Ecology of the Fishery
1997-98
1998-99
40%
40%
1999-00
2000-01
40%
40%
2001-02
2002-03
40%
40%
Figure 44 MDS ordinations illustrating rank similarities between the species assemblage sampled during
each lunar phase of January, 1997–98 to January, 2002–03. Similarities between samples were
described using the Bray-Curtis index using fourth-root transformed mean catch rates sampled
from individual dais. Stress values range from 0.1 to 0.14. Clusters are grouped by percentage
similarity reported in the bottom left corner of each ordination.
Page 89
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
2003-04
2004-05
2D Stress: 0.11
3
3
3
3 1
3
4
4
4
4
3
3
3
4
4
44
4
4
4
1
4
40%
40%
2005-06
2006-07
40%
40%
2007-08
2008-09
40%
40%
Figure 45 Page 90
4
1
1
4
4
1
14
1
2
2 2
2
1 2
2
1
22
3
1 2
3
12 2 2 2
3
22 2
3
32 2 2 2
22
22
1
3 3
2 2 12 2 2
22 23 2 2 2
2
2
2 2 12
4
2
2
1
4
4
MDS ordinations illustrating rank similarities between the species assemblage sampled during
each lunar phase of January, 2003–04 to January, 2008–09. Similarities between samples were
described using the Bray-Curtis index using fourth-root transformed mean catch rates sampled
from individual dais. Stress values range from 0.08 to 0.12. Clusters are grouped by percentage
similarity reported in the bottom left corner of each ordination.
The Ecology of the Fishery
5.3.3. Size (weight)
In spite of significant inter-annual variation, changes in the mean weight of fish caught during the
fishing season follow a largely consistent pattern with a rapid rise in fish size at the start of the season
which peaks in November or December and is then is followed by a decline (Figure 46). This is
consistent with the monthly changes in species diversity reported in Section 5.3.2.
Figure 46 Changes in mean fish weight (all species combined) through the fishing season (1997–98 to
2008–09). Estimates of the mean are weighted by numbers of fish sampled.
Page 91
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
These results suggested that the mean weight changes observed during the season reflect changes
in the timing of migration of species of different sizes with a greater abundance of larger species
migrating during the flood recession in November and December compared to other months.
Monthly changes in mean weight for the most abundant species are illustrated in Figure 47.
Lobocheilos crytopogon exhibits a monotonically increasing weight through time with some slowing
of growth following the end of flooding typical of the pattern of growth of species inhabiting tropical
rivers (Welcomme, 1985). The remaining species, except O. lini, also exhibit evidence of declining
growth rates through the season, but often with deviations above and then below the expected
trajectory. This may be indicative of larger individuals migrating first during the falling water period
(November and December) followed by smaller juvenile individuals.
Figure 47 Mean weight changes by month for the six most abundant species (1997–98 to 2008–09).
Estimates of the mean are weighted by numbers of fish sampled.
The tendency for both larger species and larger individuals of each species to leave the Great Lake
earlier than smaller fish was reported by workers during the 1950’s as well as in other tropical river
systems (see Welcomme 1985 for review). Dai fishers also report this behaviour (Section 2.3). Given
that large fish also often fail to leave deeper floodplain pools (Welcomme, 1985) suggests that depth is
a major factor affecting the timing of refuge-seeking fish migrations along with dissolved oxygen and
temperature.
Page 92
The Ecology of the Fishery
5.4. Inter-annual variation
5.4.1. Catch, effort and CPUE
75
70
65
60
55
-9
99 9
-0
00 0
-0
01 1
-0
02 2
-0
03 3
-0
04 4
-0
05 5
-0
06 6
-0
07 7
-0
08 8
-0
9
98
97
-9
8
50
600
500
400
300
200
100
-9
98 8
-9
99 9
-0
00 0
-0
01 1
-0
02 2
-0
03 3
-0
04 4
-0
05 5
-0
06 6
-0
07 7
-0
08 8
-0
9
0
97
Historical records of the species
composition of the dai catches date from
surveys undertaken in the 1930s (Chevey
and Le Poulain, 1940) and 1960s (Fily
and D’Aubenton, 1965), but as pointed
out in Section 3.2, consistent and regular
sampling of the fishery dates from the
1994–95 fishing season. Comparing catches
from the early surveys with those of the
1994–95 period, Lieng et al. (1995) found
that the dominant species had remained
largely unchanged with catches being
dominated by the Cirrhinus, Lobochelius
and Henicorhynchus genera (collectively
classified as Henicorhynchus species) that
had the highest relative abundance in the
1938–39 season and comprised 25.4% and
CPUE (tonnes/dai)
5.4.2. Catch diversity and species
composition
Effort (dai units)
97
-9
98 8
-9
9
99
-0
00 0
-0
01 1
-0
02 2
-0
03 3
-0
04 4
-0
05 5
-0
06 6
-0
07 7
-0
08 8
-0
9
Catch (tonnes)
During the 12 year period for which effort data are believed to be reliable, estimates of total annual
catch for the fishery (aggregated across species) has varied from between approximately 8,000 and
33,000 tonnes, with a mean of approximately 15,000 tonnes but with no obvious trend (Figure 48 a).
Very high catches were observed in
2004–05 and 2005–06. Effort has ranged
40,000
from 60 to 68 dai units, again with no
35,000
obvious trend (Figure 48 b). Catch per
30,000
dai per season–an index of fish biomass
25,000
(Figure 48 c) exhibits a similar coefficient of
20,000
15,000
variation as catch per season (approximately
10,000
46%) and there is also no obvious trend
5,000
through time. Catch rates ranged from
0
approximately 120 to 530 tonnes dai-1
season-1 with a mean of approximately 240
80
tonnes dai-1 season-1.
Season
Figure 48 (a) Total Catch, (b) effort and (c) CPUE
(1997–98 to 2008–09).
Page 93
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
67.5% of the catch in the 1962–63 and 1994–95 seasons respectively (Table 21). The 1962–63 season
differed from the other two seasons in that Cirrhinus microlepis (18.6%) and Thynnichthys thynnoides
(14.9%) were second and third most dominant respectively. Paralaubuca barroni (listed as typus)
comprised only 0.3% of the catch in 1962–63 whereas it comprised 13.4% in the 1994–95 season.
Table 21 Estimated catches of the most abundant species reported for the years 1938–39
(Chevey and Le Poulain, 1940), 1962–63 (Fily and D’Aubenton, 1965), 1994–95
(Lieng et al., 1995) – the latter estimated as a sum of the catches for December, January and
February 1994 (adapted from Lieng et al., 1995).
Species
1938–39
1962–63
Rel.
abundance
% Total
Catch
1994–95
Total Catch
(tonne)
% Total
Catch
Henicorhynchus spp.
25.4
12,432
67.5
Paralaubuca typus 1
0.30
2,460
13.4
-
979
5.3
14.9
515
2.8
2.0
471
2.6
18.6
398
2.2
0.2
270
1.5
Osteochilus hasseltii Dangila spp.
2
Thynnichthys thynnoides
Morulius chrysophekadion4
Cirrhinus microlepis
Botia spp.5
3
Paralaubuca typus (Bleeker, 1864) (Trey slak russey) is currently listed in the Dai Database as Paralaubuca barroni (Fowler, 1934).
1
Lieng et al. (1995) and Baran et al. (2001) lists Osteochilus hasseltii (Valenciennes, 1842) as (Trey kros). In the current database,
(Trey kros) is listed as Osteochilus lini.
2
Dangila sp. is a synonym for a Labiobarbus sp. Lieng et al. (1995) list this species as (Trey khnawng veng). In the current database
(Trey khnawng veng) is listed as L. lineata (Sauvage, 1881). The correct spelling for this species, however is lineatus.
3
Morulius chrysophekadion (Bleeker, 1850) (Trey kaek) is a synonym for Labeo chrysophekadion (Bleeker, 1850) as it is currently listed in
the database.
4
For the purposes of comparison, the three Botia (now named Yasuhikotakia) spp. were combined in the 2008–09 catch estimation.
5
Baran et al. (2001b) analysed the species composition of the dai catches between 1995 and
2000 by means of Principle Components Analysis, focussing on the 35 most significant species.
They concluded that although the relative abundance of some species changed between years, there
appeared to be no significant inter-annual variability in catch composition for the bulk of the species.
Subsequent to Baran et al. (2001b) report, there have been no further analyses of species composition.
In this section, data from the 1997–98 season (the first reliable catch estimates for reasons outlined
in Section 3.3) to present (2008–09) are examined for evidence of changes in species composition
following the methods of Clark and Warwick (2001).
Consistent with the historical record, the Cirrhinus, Lobochelius and Henicorhynchus genera
(previously referred to as the Henicorhynchus group) were found to be the most important genera
in the catch, contributing between 2,715 to 12,712 tonnes, or 20% and 50% of the annual catch,
respectively. The other dominant species consistent with the historical record include: P. typus,
L. lineatus, T. thynnoides and L. chrysophekadion (formerly listed as M. chrysophekadion). There
is evidence of an increasing contribution to the total catch from the Cirrhinus, Lobochelius,
Page 94
The Ecology of the Fishery
Henicorhynchus and Paralaubuca genera from 1997–98. There is also considerable inter-annual
variation in the contribution made by other species (Other spp.), particularly during the earlier seasons
(1997–2003) (Figure 49). However, as with the diversity indices, preliminary investigations appear to
support the conclusion that these trends may be an artefact of changes to the survey techniques, i.e.
sample stratification and/or species identification, and therefore may not be biologically significant.
97- 98
98 - 99
12,624 t
4,270
4,874
9,373 t
99 - 00
2,759
3,333
3,824
775
222
369
535
7,155
400
592 671
00 - 01
11,954 t
1,345
140
207
336
1,179
21,346 t
01- 02
18,505 t
02- 03
2,863
3,642
6,184
14,890 t
605
9,423
814
1,486
290
597
11,886
2,741
1,090
9,230
475
417
440
807
937
03 - 04
04 - 05
5,869 t
5,397
1,695
177
594
06 - 07
769
455
1,814
2,508
462
07- 08
21,969 t
5,481
6,754
10,823 t
08 - 09
2,707
9,321
774
670
33,640 t
12,712
9,386
400
224
859
181
05 - 06
10,663
2,715
10,190 t
2,174
4,844
293
271
235
246
5,126
1,327
1,279
2,336
3,386
Figure 49 18,733 t
1,426
1,081
Heicoryhnchus spp.
Paralaubuca barroni
Labiobarbus lineatus
Thynnichthys thynnoides
Labeo chrysophekadion
Other spp.
The species composition of the dai fishery catch, 1997–98 to 2008–09.
Page 95
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
The greatest number of species reported caught by a single dai on a day was 80 by dai 12B
on 22 January 2002. The second and third highest number of species reported to have been caught
on a dai on a day (75 and 77 species respectively) occurred on the two days prior to this date
(20–21 January 2002). The mean number of species caught by a dai on a day across all seasons was
22.8. The highest diversity (H) was recorded on these same three days: 20, 21 and 22 January 2002:
4.01, 3.99 and 3.97 respectively.
The mean species richness and diversity of the dai catches across all months and lunar periods
show evidence of considerable interannual variability (Figure 50 a and b). After a decline from
1997–98, estimates of the mean number and diversity of species caught by a dai on a day show a
steady increase between 1998–99 and 2002–03. Thereafter, aside from a peak in 2006–07, the values
decline and then stabilise to a mean of approximately 26 species/dai/day. Thus the trends suggest
an overall increase in the number and diversity of species has occurred over the twelve year period.
However, strong parallels exist with the number of hauls sampled in each season (Section 3.3,
Figure 17) suggesting that these trends merely reflect changes in sampling effort and/or the skill of
enumerators rather than changes in the fish community itself. Sampling effort in terms of the total
number of hauls sampled each season has varied considerably since the start of the monitoring period.
A total of 435 hauls were sampled in 1997–98, with this number increasing to 2,426 hauls in 2001–02.
Sampling effort then declined somewhat after 2003–04 and this is paralleled by the sharp drop in
species numbers and diversity over this season. The only inconsistencies in the parallels between
species diversity and effort is that the former peaked in 2006–07, whereas the latter peaked in
2007–08.
Multivariate analyses revealed significant differences (p < 0.01) in relative species abundance
among fishing seasons (Figure 51–Figure 53). Whilst these differences were not significant for every
pair of seasons tested using the ANOSIM routine, differences in the assemblage before and after the
2000–01 season were evident in the MDS plots.
Catch records from the earliest surveys (1930s), suggest that the catch composition has remained
relatively unchanged through the seasons with the smaller, short-lived annual cyprinids belonging to
the Cirrhinus, Lobocheilos, Henicorhynchus, Paralaubuca and Labiobarbus genera remaining the
most dominant in terms of weight and number. Other important non-cyprinid fish important in the
catch include the larger Pangasiid catfish such as Pangasius larnaudii as well as the Yasuhikotakia
genus (e.g. Y. modesta) belonging the Cobitidae family (Table 22). Whilst there is some evidence
that these species have increased in relative abundance through time, this trend may simply reflect
sampling error. Efforts to minimise sampling error or species mis-identification will be required to
determine if the assemblage utilising the TS-GL System is changing with time.
Page 96
The Ecology of the Fishery
(a)
(b)
Figure 50 (a) The Species Richness (S) (Mean ±95% CI) and (b) Shannon Diversity Indices (H) (Mean ±95%
CI) of fish species caught by a dai on a day for the seasons 1997–98 to 2008–09 averaged across
all sampling months in each season.
Page 97
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 22
Principal families and species making up 99% of the dai catches listed in descending order of
their total percentage contribution. The number of species in each family is reported together with
the three dominant species by weight in each family, also in descending order of their percentage
contribution.
Family
Cyprinidae
No. of species
53
Species
Cirrhinus lobatus
Paralaubuca typus
Henicorhynchus cryptopogon
Pangasiidae
9
Pangasius larnaudii
Pangasius pleurotaenia
Pangasianodon hypophthalmus
Cobitidae
5
Yasuhikotaki modesta
Yasuhikotaki morleti
Yasuhikotaki helodes
Siluridae
8
Belodontichthys truncatus
Phalacronotus micronemus
Wallago attu
Gyrinocheilidae
1
Gyrinocheilus aymonieri
Soleidae
1
Synaptura marginata
Bagridae
8
Hemibagrus nemurus
Mystus singaringan
Hemibagrus wyckii
Clupeidae
4
Clupeichthys aesarnensis
Tenualosa thibaudeaui
Anodontostoma chacunda
Engraulidae
3
Coilia lindmani
Setipinna melanochir
Lycothrissa crocodilus
Cynoglossidae
1
Cynoglossus feldmanni
Sciaenidae
1
Boesemania microlepis
Belonidae
1
Xenentodon cancila
Notopteridae
2
Chitala ornata
Notopterus notopterus
Polynemidae
3
Polynemus multifilis
Polynemus dubius
Polynemus borneensis
Bagriichthidae
Page 98
1
Bagrichthys majusculus
The Ecology of the Fishery
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
2D Stress: 0.25
2D Stress: 0.22
Figure 51 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root transformed
mean CPUE for individual dais for the 1997–98 to 2008–09 seasons for (a) October and (b)
November.
Page 99
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
2D Stress: 0.19
2D Stress: 0.17
Figure 52 Page 100
MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root transformed
mean CPUE for individual dais for the 1997–98 to 2008–09 seasons for (a) December and (b)
January.
The Ecology of the Fishery
97-98
98-99
99-00
00-01
01-02
02-03
03-04
04-05
05-06
06-07
07-08
08-09
2D Stress: 0.21
2D Stress: 0.19
Figure 53 MDS ordinations based on Bray-Curtis indices of similarity derived from fourth-root transformed
mean CPUE for individual dais for the 1997–98 to 2008–09 seasons for (a) February and (b)
March.
5.4.3. Size (weight)
Assuming that recruitment is constant, a decline in the mean weight of individuals in a population
through time is indicative of increasing rates of exploitation as fewer large (older) individuals survive
with time (Sparre and Venema, 1992).
Page 101
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Estimates of mean weight for all species combined as well as those species that contribute to the
majority of the catch by weight have shown considerable variation during the fifteen year monitoring
period, but with no evidence of a continuous monotonic decline (Figure 54 and Figure 55).
Figure 54 Mean weight of fish landed (all species combined) 1997 – 2009. Estimates of the mean are
weighted by numbers of fish sampled.
Figure 55
Trends in the mean weight of the six most abundant species in the fishery estimated for December
each year (1995–96 to 2008–09). Estimates of the mean are weighted by numbers of fish sampled.
Page 102
The Ecology of the Fishery
5.5. Hydrological effects on fish biomass
The hydrology of the TS-GL System and the basin as a whole may change in the future as a
consequence of basin development activities such as hydropower dam construction and water
abstractions or diversions for irrigation, as well as climate change. Improving our understanding of
hydrological influences on the fisheries resources of the TS-GL System potentially offers a means to
predict how the fisheries resources in this system and in other parts of the LMB are likely to respond
to future hydrological conditions under different scenarios of basin development and climate change.
Hydrological effects on fish populations have been described for several tropical and European
river systems from as early as 1910. Typically, these have been described in the form of correlations or
linear regressions between indices of flood extent and duration and annual catch or some other index
of fish biomass such as catch per unit of effort, CPUE. Other workers (e.g. Dudley 1972; Kapetsky,
1974) have demonstrated correlations between hydrological indices and annual growth increments of
fish (see Welcomme, 1985 for review).
Several workers including Lieng et al. (1995); Baran et al. (2001 b) and van Zalinge et al. (2004)
have described similar effects for the TS-GL System using regressions between annual catch estimates
for the dai fishery and maximum annual water level measured at Phnom Penh as a proxy of flood
extent. However, these earlier models took no account of changes to fishing effort in the fishery and
the selected flood indices accounted only for the extent of flooding over a relatively short period of
time 1–31 days and therefore poorly described the inter-annual changes in flood duration.
Below, we describe a model that employs the mean daily catch rate of a dai unit as the biomass
index. A flood index (FI) is used to quantify both the extent and duration of the flood each year, (y):
FIy = ∑
d
FAy,d
Where (FAy,d) is the flooded area of the TS-GL System in year (y) on day (d), measured above the mean
flooded area for the model period 1 January 1997 to 31 March 2009.
Estimates of (FAy,d) were derived from daily observations of water level (WLy,d) at Kampong Luong
gauging station in the Great Lake (Figure 56 and Figure 57) and the following second-order
polynomial (John Forsius pers comms):
FAy,d = 716.64 + 1094.19WLy,d + 30.05WL2y,d
Page 103
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 56 Page 104
The location of the Kampong Luong gauging station in the TS-GL.
The Ecology of the Fishery
Figure 57 Daily water level (m) measured at Kampong Luong gauging station and estimates of the flooded
area of the TS-GL System, 1997 to 2009.
Water Level /
Flooded Area
1.5
Flood
start
Flood
end Flood
amplitude
Max
WL / FA
Flood rise
1
0
-0.5
1
Flood
index
Drawdown
0.5
Dry
season
index
Rising water
duration
Dry season duration
Figure 58 Time
(Days)
Min
WL / FA
-1
-1.5
Mean
WL / FA
falling water
duration
Flood duration
Illustration of the estimation of hydrological indices. WL–Water level; FA–Flooded area.
Page 105
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 23
Summary of the indices used to describe the various attributes of the hydrograph hypothesised to
affect fish biomass.
Index
Description
Calculation
Hypothesised effect
References
Flood
Index, FI
Measure of the
extent and duration of flooding.
Sum of flooded
area for each day
of the flood.
Potentially influences growth,
survival and recruitment and
therefore biomass.
Dry Season
Index, DSI
Measure of the
severity and
duration of the
dry season.
Sum of dry area
for each day of
the dry season.
Potentially influences fish
survival and therefore
spawning stock biomass and
recruitment.
Welcomme (1979; 1985),
Welcomme and Halls. (2001);
Kapetsky, 1974 and
Dudley (1972).
Annual
Flood Index
Measure of dry
FSI - DSI
season and flood
season habitat
availability.
Combined effect of FSI and
previous DSI.
Flood Start,
FS
Relative start
time of flood.
Number of days
since January, 01
to start of flood
Potentially influences survival
of drifting larvae and feeding
opportunities.
Flood End,
FE
Relative end
time of flood.
Number of days
before or after
January, 01
corresponding to
end of flood.
Determines flood duration and
therefore feeding opportunities.
Number of days
between flood
start and flood
end date.
Potentially influences feeding and growth opportunities.
Growth effects may also influence size-dependent mortality
and subsequent recruitment.
Welcomme and Hagborg (1977);
Halls and Welcomme (2004).
Flood Dura- Duration of the
tion, D
flood.
Flood Rise
Rate, FRR
Rate of flooding. Flood rise / rising
water duration
Influences recruitment success
and colonisable area of
floodplain.
Welcomme and Hagborg (1977);
Halls et al. (2001); Halls and
Welcomme (2004).
Drawdown
Rate, DDR
Rate at which
waters recede.
Influences survival rates and
therefore subsequent
recruitment.
Welcomme and Hagborg (1977);
Halls et al. (2001); Halls and
Welcomme (2004).
Drawdown /
falling water
duration
A general linear model (GLM) was used to test whether the flood index had a significant effect on fish
biomass indicated by sampled loge-transformed daily dai catch rates accounting for the effects of interannual differences in fishing effort (Section 5.4.1), and the spatial (dai row) and intra-annual (month
and lunar phase) variation in catch rates associated with fish migrations described in Sections 5.1.1
and 5.3.1.
The flood index, (fi) and the fishing effort, e (number of dais) were therefore treated as covariates in
the model and lunar phase, (lp) (1–4); month, (m) (October–March); and dai row, (r) (1–15) were
treated as fixed factors:
ln CPUEijkl = α + βfii + βei + lpj + mk + rl + εijkl
i = 1,2,..n; j = 1,2,3,4; k = 10,11,12,1,2,3; l = 1,2,…15.
Page 106
The Ecology of the Fishery
Where εijkl is the residual (unexplained) component and (n) is the number of observations.
Dai row, (r) was treated as a fixed factor in the model rather than a covariate because of the apparent
non-linear decline in catch rates with dai row or cumulative effort, (Er) illustrated in Section 5.1.1.
Other hydrological indices were also considered during model fitting including indices to account for
variation arising from potential dry season (survival) related, and rate of flood rise and fall related
effects (Figure 58 and Table 23). However, the (FI) alone was found to provide the best-fitting and
parsimonious model (Table 24). Catch rates were predicted to increase with the (FI) and, as expected,
catch rates were predicted to decline with increasing fishing effort. Overall the model explains almost
70% of the variation in the observed catch rates (Figure 59) and the model residuals are reasonably
well behaved (Figure 60). However, the model predictions overestimate some of the very low catch
rates and underestimate the very high ones (Figure 59).
Table 24
ANOVA table for the GLM model
Source
Type III Sum of Squares
Corrected Model
df
Mean Square
F
Sig.
25427.520(a)
349
72.858
38.867
.000
Intercept
901.877
1
901.877
481.121
.000
Effort
320.501
1
320.501
170.976
.000
FI
740.492
1
740.492
395.028
.000
MONTH
3039.725
5
607.945
324.319
.000
Lunar Phase
3700.012
3
1233.337
657.945
.000
Row
462.154
14
33.011
17.610
.000
1914.482
15
127.632
68.088
.000
Month * Row
248.902
70
3.556
1.897
.000
Lunar phase * Row
105.415
42
2.510
1.339
.071
1.597
.000
Month * Lunar phase
Month * Lunar phase * Row
592.624
198
2.993
Error
11307.169
6032
1.875
Total
223390.183
6382
36734.689
6381
Corrected Total
Dependent Variable: ln CPUE (kg/dai/day)
R Squared = .692 (Adjusted R Squared = .674)
Page 107
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
9
Ln CPUE (kg/dai/day)
8
7
6
5
4
3
2
Observed
1
Predicted
20
08
20
07
20
06
20
06
20
05
20
04
20
03
20
03
20
02
20
01
20
01
20
00
19
99
19
99
19
98
19
97
0
Year
Figure 59
The GLM observed and predicted mean catch rates 1997–2009. For the purposes of illustration,
mean monthly, instead of daily, catch rates are illustrated.
Figure 60
Model Residuals.
Page 108
The Ecology of the Fishery
The model predicts that fish biomass, indicated by the mean daily catch rate of a dai unit during the
fishing season (October–March), increases exponentially with the (FI) (Figure 61) as follows:
Mean CPUE (kg/dai/day)
CPUE = 83.88.e1.6063E–06FI
450
400
350
300
250
200
150
100
50
0
0
200000 400000 600000 800000 1000000 1200000
Flood Index (km2 days)
Figure 61
The relationship between the mean predicted daily catch rate of a dai unit during the fishing season
and the flood index (FI) for the TS-GL System.
This response appears to be mediated through the effects of the (FI) on fish growth. Growth variation,
indicated by mean fish weight responds in the same way having a similar exponent value and therefore
the same four-fold increase in predicted values from the minimum to the maximum observed flood
index (Figure 62). This same type of growth response is also evident for the six most abundant species
(Figure 63).
0.025
Weight (kg)
0.020
0.015
0.010
0.005
0.000
0
200000 400000 600000 800000 1000000 1200000
2
Flood Index (km days)
Figure 62 The relationship between mean sampled fish weight (all species combined) and the flood index
with fitted exponential model. Weight = 0.0054e1.231E–06FI . R2 = 0.59.
Page 109
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Figure 63 Mean weight estimates for the six most abundant species in the fishery in December each year
(corresponding to the end of the flood season) plotted as a function of the flood index.
1200000
Weight
FI
0.020
1000000
800000
0.015
600000
0.010
400000
-0
9
-0
8
08
-0
7
07
-0
6
06
-0
5
05
-0
4
04
-0
3
03
02
01
00
99
98
97
-0
2
0
-0
1
0.000
-0
0
200000
-9
9
0.005
-9
8
Weight (kg)
0.025
Flood Index (km2 days)
ANOVA demonstrated that when variation due to the flood index is accounted for, there is no
significant change in mean weight through time for any of the six species examined. The relatively
low mean weights observed during the last six years are therefore likely to reflect growth responses to
below average flood conditions, rather than the effects of increasing rates of exploitation as suggested
by some workers (Figure 64).
Season
Figure 64 Page 110
Changes in mean fish weight and the flood index from 1997–98 to 2008–09.
No such relationship was evident between fish abundance indicated by mean CPUE for the season,
expressed as number of fish caught per dai per day, and the (FI) (Figure 65).
Figure 65 Fish abundance plotted as a function of the flood index.
Inter-annual fish biomass variation in the TS-GL System indicated by catch rates sampled from the
dai fishery appear to be growth mediated in response to feeding opportunities dictated by flood extent
and duration. Mean fish weights for 2004–05 and 2005–06 were consistent with the flood index.
Therefore, above average levels of recruitment were probably responsible for the very high catches
observed during these two seasons described in Section 5.4.1. Factors responsible for these high levels
of recruitment remain uncertain but a campaign by the FiA to confiscate illegal gear, particularly
during 2003 and 2004, may have been influential (Hortle et al., 2005). The Mekong River Commission
(MRC) report that sediment inputs, which can influence primary production in the TS-GL System,
peaked during these two seasons (MRC, pers. comms.). These conditions may have also favoured
recruitment.
Page 111
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Page 112
6. Summary and Conclusions
The Tonle Sap-Great Lake (TS-GL) system is an integral part of the history, culture, ecology and
economics of the Mekong region. As well as mitigating floods in Cambodia and Viet Nam, it provides
a source of raw materials, nutrition, income and livelihoods for upwards of one million people living
in and around it. The system is of international ecological and conservation importance supporting a
diverse flora and fauna including 149 species of fish, five of which are globally threatened. The annual
flood pulse is believed, by some, to be the principle driving force responsible for this productivity and
diversity, transporting and recycling biolimiting nurtrients and providing diverse ephemeral critical
habitats.
Blackfish and whitefish species are the target of industrial, artisanal and subsistence fisheries
operating in the TS-GL System. The large-scale industrial fisheries are managed through a system of
fishing lots - demarcated areas of productive fish habitat of up to 500 km2, or positions along migration
corridors such as the Tonle Sap and other tributaries forming the system. These limited access fisheries
operate only during the open season (October–May) and typically employ large barrier, bagnet or
fence type gears designed to divert and/or intercept migrating fish. Artisanal (middle-scale) and
subsistence (family) fisheries operate in open access areas with a variety of gear types including
gillnets, seines, bamboo fence traps, cast nets and longlines. The former may operate only during the
open season, whereas subsistence fisheries operate year-round. Strong competition exists among the
fisheries to land in excess of 200,000 tonnes of fish each year equivalent to approximately 10% of the
total quantity of fish consumed in the entire LMB each year.
The dai fishery on the Tonle Sap, established almost 140 years ago, is an important component of
the industrial fishery landing approximately 14% of the annual catch taken from the TS-GL System.
It targets the refuge migrations of a multi-species assemblage of fish as they migrate from the Great
Lake to the Mekong main channel via the Tonle Sap with the receding floodwaters each year. Up to
15 rows of stationary trawl nets are set over a 35 km stretch of the Tonle Sap in Kandal and Phnom
Penh Municipalities, with 1–7 trawl nets or dai units forming each row. Diesel engines have recently
replaced the traditional hand-powered wooden winches for hauling the nets as frequently as 4 times
per hour during peak catch rate periods.
In addition to its significant socio-economic value both locally and nationally, the dai fishery
provides a valuable source of data and information to monitor trends in migratory fish populations
which seasonally utilise the TS-GL System and beyond. Monitoring these trends provides an important
means to evaluate the performance of fisheries management measures as well as potential impacts
arising from basin development activities.
Page 113
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Small cyprinids belonging to the Cirrhinus, Lobocheilos, Labiobarbus and Henicorhynchus genera
form the bulk of the catch landed by the dai fishery. Other species making an important contribution
to landings are Labeo chrysophekadion, Pangasius pleurotaenia, Puntioplites proctozystron and
Thynnichthys thynnoides.
Small species are sold to fish traders either on the floating dai platform or on the nearest riverbank
for marketing, processing or export. The remaining more valuable medium-sized and large-sized fish
species including Osteochilus melanoplerus, Pangasius larnaudii, Cyclocheilichthys enoplos and
Pangasianodon hypophthalmus are often kept alive in bamboo cages suspended below the working
platform of the dai and sold during the closed season (March–September) when the fish supply is low
and prices are high.
Typical annual operating profit per dai unit is in the region of US$14,000. Annual profit tends to
decline from the most upstream dai row 15 to row 2 with declining catch rates. Licence fees (both
official and unofficial combined) paid by dai operators appears independent of their reported operating
profit. Unexpectedly, the price of trey riel appears positively correlated with supply (landings).
Under Cambodian Law, the fishery is managed using a closed season together with effort and gear
size restrictions. Ad hoc surveys of the fishery began in the late 1930s but comprehensive routine
monitoring programmes did not begin until 1994–95 undertaken by the FiA/IFReDI with support from
the MRC and the FAO. The survey design has evolved over the 15-year monitoring period mainly
with changes to survey stratification to reduce sample variance. In addition to changes to the structure
and stratification of the sampling regime, there has been considerable inter-seasonal variation in the
sampling effort as measured by the total number of hauls sampled by the data collectors over the
survey period.
The database used to store and process the data collected from the fishery has also evolved and
been supported by different software platforms. This paper has provided a detailed description of
the changes that have been made including the full details of the tables providing a useful reference
document for both IFReDI and others working with the database. Database queries used to extract
data from the tables have been described in a companion working document (Cambodian Dai Fishery
Database–Query Reference Manual) held at IFReDI. This Manual contains a range of alternative
queries that were developed to analyse the data. The alternative queries described in the manual reflect
different assumptions regarding the structure of the data itself and/or the reliability of certain variables
such as information on fishing effort. The conclusions drawn from the data generated by the queries
may differ slightly depending on these underlying assumptions. This variation effectively represents
"model" or "structural" uncertainty–an additional form of uncertainty in catch estimation that exists
alongside the statistical uncertainty associated with estimates of parameters using the data. Due to
ambiguity surrounding the interpretation of the data no single method has been recommended.
Catch rates sampled from the dai fishery exhibit considerable spatial and temporal variation.
Catch rates in Phnom Penh (downstream) are on average less than half the rates estimated for Kandal
Page 114
Summary and Conclusions
indicative of a depletion effect on the migrating fish. A Delury depletion model was used to describe
this depletion response. It predicted that each dai unit removes approximately 2.8% of fish migrating
through the fishery equivalent to an instantaneous fishing mortality rate, (F) of 1.79 for 64 dai units. In
other words approximately 80% of fish arriving at the dai fishery are estimated to be caught.
However, the non-monotonic decline in catch rates with cumulative effort evident in the depletion
model suggests that the required assumption of constant catchability may have been violated. This
might reflect the use of different net mesh size (see Section 5.1.3) and/or non-random sampling effort
through space and time (Section 3.3). The use of larger mesh sizes would be expected to result in lower
catch rates and vice-versa.
Furthermore, the corresponding estimates of fishing mortality generated by this analysis will be
biased if removals of fish between dai rows by small-scale (artisanal and subsistence) gear were
significant. Significant lateral migrations of fish from or to adjacent floodplains between dai rows
would also bias the estimates. Current knowledge of these small-scale fisheries and fish migrations
routes in the TS-GL System is sparse and poorly documented. This will need to be addressed in order
to reliably interpret these results.
If lateral migrations can be ignored, then the estimate of the proportion of the fish removed over the
range of the dai fishery (regardless of gear type present) would remain unchanged i.e. approximately
80%. However, the catchability coefficient, (q) for the dai gear and thereby the estimate of the
instantaneous fishing mortality rate, (F) for the dai fishery would be upwardly biased if removals by
small-scale gear were significant.
Most importantly, the estimate of the proportion of fish removed by the dai fishery, and the
corresponding estimate of the average instantaneous fishing mortality rate (F) during the dai fishing
season, relate only to the population of fish remaining after exploitation (removals) by other gear
operating upstream in the Lake including those used by other lot, artisanal and subsistence fisheries.
They are no estimates for the entire population of fish inhabiting the TS-GL System during a period of
one year.
Water depth and velocity effects on individual dai catch rates could not be detected. In addition to
depletion effects on the population, the observed spatial variability in dai catch rates may also reflect
differences in catchability influenced by factors such as bathymetry or other hydrological variables not
considered here.
Whilst no strong evidence was found to suggest that the species assemblage landed by the dai
fishery varies significantly through the dai rows, some evidence was found to suggest that the mean
weight of fish landed increases downstream. This may reflect the use of nets with larger mesh size.
Fish abundance, indicated by the mean daily catch rate of sampled dai units, exhibits significant
variation both between and within the six months comprising the fishing season. Catch rates typically
Page 115
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
peak in December or January and are lowest in October. Catch rates also peak during the second
quarter of the lunar cycle known as the ‘Waxing Gibbous’ phase. These results indicate that fish
migrations are strongly influenced by the lunar cycle, and possibly water levels as peak migrations
typically occur around January or December coinciding with the end of the flood season as also
suggested by Baird et al. (2003). This pulsed, rather than continuous, form of migratory behaviour
from the Lake peaking during the second quarter of the lunar cycle was reported as early as the 1950’s
(Welcomme, 1985).
However, it remains uncertain why migrations from the Lake should peak during this period
particularly when levels of illumination are high potentially increasing the threat of predation by
sight-feeding predators. For this reason, migrations of European eels (Anguilla anguilla) and lampreys
(Petromyzonidae) tend to be weaker driving full-moon or moonlit conditions (Wootton, 1990).
Seaward migrations of coho salmon (Onchorhynchus kisutch) smolts also tend to peak during the
(dark) new moon phase. In the Tonle Sap however, illumination may not matter as the water is very
turbid, and predators would be swamped (Hortle, pers comms).
Lunar phase-related migrations of amphidromous or anadromous species returning to rivers are
generally interpreted in the context of variation in tidal height than as a consequence of variable light
intensity (Lucas and Baras, 2001). Historically, tidal influences on the TS-GL System may have been
significant (see Section 1.1.1). Baird et al. (2003) therefore proposed that stenohaline species including
small cyprininds may have evolved migratory behaviour to avoid encountering saline intrusions in the
Mekong mainstream associated with the highest (spring) tides. However, this is hard to reconcile with
the fact that spring tides occur every 14 days (twice per month), whilst migrations clearly peak only
once per month.
There was also evidence that different species migrate from the Lake at different times with a
general trend toward increasing species richness and diversity from the beginning of the season
(October), to the middle (November and December), before tailing off toward the end of the season
in March. Corresponding changes in mean fish weight were also observed peaking in November or
December and followed by a decline. These patterns may be indicative of larger species migrating
during the flood recession in November and December compared to other months. There is also
some evidence that larger individuals of the same species migrate first during the falling water period
(November and December) followed by smaller juvenile individuals.
The tendency for both larger species and larger individuals of the same species to leave the Great
Lake earlier than smaller fish was reported by workers during the 1950s as well as in other tropical
river systems (see Welcomme 1985 for review). This behaviour is also reported by the dai operators.
Given that large fish also often fail to leave deeper floodplain pools, Welcomme (1985) suggests that
depth is a major factor affecting the timing of refuge-seeking fish migrations along with dissolved
oxygen and temperature.
The fishery experiences significant spatial, intra- and inter-annual variation in catch rates and
landings in catch rates and therefore total landings. During the 12 year period for which effort data
Page 116
Summary and Conclusions
are believed to be reliable, total catch (aggregated across species) by season has varied from between
approximately 8,000 and 33,000 tonnes, with a mean of approximately 15,000 tonnes but with no
obvious trend. Very high catches were observed in 2004–05 and 2005–06. Effort has ranged from
60 to 68 dai units, again with no obvious trend. The mean weight of fish has shown considerable
variation during the fifteen year monitoring period, but with no evidence of a continuous monotonic
decline. Multivariate analyses provided no compelling evidence to indicate that species composition
has changed significantly through time. Rather changes in species richness appear to reflect changes in
sampling effects/error.
The extent and duration of flooding in the TS-GL System each year has a significant effect on fish
biomass migrating from the Lake indicated by dai catch rates after accounting for variation in fishing
effort. Relative fish biomass increases exponentially with the index of flood extent and duration
exhibiting a four-fold increase across the observed range. This response appears to be mediated
through the effects of the flood index on fish growth, rather than on recruitment.
Observed mean fish weight for 2004–05 and 2005–06 were consistent with those predicted
(expected) for the flooding conditions. Therefore, above average levels of recruitment were probably
responsible for the very high catches observed during these two seasons.
6.1. Management Implications
This paper represents the first attempt to compile and analyse the available data and information
about the Cambodian dai fishery in a single document. It therefore serves as an important reference
document for present and future workers involved in the management, monitoring and administration
of the fishery. It also contains new insights into the ecology and dynamics of target fish populations
important for their management.
The paper has attempted to place the dai fishery in the wider context of the fisheries of the
TS-GL System, Cambodia and regionally. This has served to illustrate that whilst the dai fishery is
the focus of most fisheries monitoring and evaluation efforts in Cambodia, it is not the only, nor most
significant, component of the entire TS-GL fishery. Other components with which it interacts and
competes with, particularly the other lot, artisanal and subsistence fisheries, are also significant and
therefore should be considered.
Like the lee trap fishery in southern Lao PDR, it is likely that the dai fishery became a focus for
intensive monitoring because it targets fish migrations through a ‘bottleneck’ giving rise to high
sampling efficiency and the prospect of accurate population estimates owing to the relatively large
proportion of the population that can be sampled over short periods. Adding to this, attempts to
monitor other lot fisheries have failed in the past because neither lot operators nor officials were
prepared to cooperate. Attempts to monitor the other, more dispersed, sectors of the fishery
(i.e. middle-scale artisanal and family fisheries) also failed owing to a lack of institutional capacity.
Page 117
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
In spite of this restricted focus, the monitoring efforts directed at the dai fishery have generated the
only continuous long-term data set for an inland fishery in Cambodia–and one of the only two in the
Mekong Basin–the other being for the Lee trap fishery.
Analyses of indicators estimated from this data set have been informative for policy and
management evaluation, revealing little or no compelling evidence of changes in the abundance,
biomass, size or diversity of migratory fish populations that seasonally utilise the TS-GL System
and beyond often over distances of more than 600 km (Baird and Flaherty, 2004; Poulsen et al.,
2004 and Adamson et al., 2009). Furthermore, the time series of these indicators have also equipped
managers with an important baseline against which to monitor any impacts of management and basin
development activities.
A key finding of this research is that inter-annual variation in the biomass of the multispecies
assemblage targeted by the fishery (and hence landings) can be largely explained by flood duration
and extent effects on fish growth. Fish growth, indicated by mean fish weight, increases exponentially
with the flood extent and duration described here by the ‘flood index’. Presumably food resources and
available feeding time increases with the flood index, or competition for food resources is less intense
during larger, longer floods (Halls et al., 2008).
This response has been modelled, allowing predictions to be made of how the relative biomass
of the multispecies assemblage targeted by the fishery (and hence catches) are likely to vary under
different flooding conditions whether natural or modified as a consequence of climate change and/
or water management projects in the Basin. Owing to the highly migratory nature of the target fish
species, these predicted hydrological responses may be observed over large distances, affecting
fisheries and piscivorous fish populations beyond the immediate vicinity of the system.
The unexplained variation in the model may well reflect variation in recruitment to the system each
year in addition to variation in fishing effort (mortality) applied by the other important fisheries within
the TS-GL System or over the migratory range of the target species, reinforcing the need for more
comprehensive monitoring programmes (see below). These results also urge caution when monitoring
mean fish size as a proxy for rates of exploitation in the TS-GL System and other highly fluctuating
environments.
By applying depletion model theory, this research has provided the first estimates of the proportion
of fish removed over the range of the fishery, dai gear catchability (efficiency), and dai fishing
mortality rates subject to a number of assumptions described above. These results are an important first
step towards understanding, and thereby controlling if necessary, the relative sources of fishing effort
(mortality) over the migratory range of populations of important species of fish. Additional studies
and monitoring programmes will be necessary to determine the validity of the assumptions underlying
these estimates and to quantify the spatial distribution of the remaining sources of fishing mortality in
the TS-GL System and beyond (see below).
Page 118
Summary and Conclusions
6.2. Recommendations
6.2.1. Monitoring fisheries resources in the TS-GL and beyond
The dai fishery monitoring programme has provided a relatively efficient means of obtaining valuable
data and information about the ecology, variability and long-term trends in migratory fish populations
that seasonally utilise the TS-GL System for fisheries policy and management purposes. However, it
is not a panacea for all information needs because of the characteristics of the dai fishery including its
interaction with other important fisheries.
Characteristically, the dai fishery targets only migratory whitefish species and therefore does not
provide any information about valuable blackfish species that inhabit the TS-GL System year-round.
Moreover, whilst located in a migration corridor or bottleneck, the dai fishery effectively operates
only over an area spanning approximately 35 km of the Tonle Sap. Without detailed knowledge of
the migratory range of the species targeted by the fishery and their population structure it is uncertain
over what geographical scale information generated by monitoring the dai fishery applies. Monitoring
other fisheries that target these stocks is also necessary to understand, and thereby control, if necessary,
sources of fishing effort (mortality) over the migratory range of populations of important species of
fish, as well as to provide other important catch-related data.
Its interaction with other fisheries exploiting the TS-GL System also makes it difficult to interpret
variability or long-term trends in the dai (or any other) fishery without assuming that effort has
remained relatively static over the period of interest. This might not be unreasonable if the large-scale
barrier and fence traps common in TS-GL System and in other bottlenecks over the migratory range of
stocks exert the greatest overall effort (fishing mortality) since their effort remains largely unchanging.
However, effort exerted by the artisanal and subsistence fisheries may be significant and variable in
response to fish abundance or may have increased with population growth. At the same time, many
fish populations will never enter the TS-GL System and therefore never be vulnerable to capture by the
dai fishery or others operating in the system.
These characteristics combine to make it impossible to rely solely on the dai fishery monitoring
programme to meet all the information needs of managers and policy makers. Consideration should
be therefore given to establishing additional monitoring programmes to supplement the data and
information currently generated by the dai fishery monitoring programme.
Detailed recommendations for these additional monitoring programmes including the statistical
aspects of their design lie outside the scope of this paper and will depend largely on the information
needs of the FiA and other major stakeholders reflecting the country’s and regional fisheries policy
and development priorities. A consultative process with these stakeholders will therefore be necessary
to develop programmes that meet their data and information needs. Useful guidelines to support this
process are described among others by FAO (1999) and Halls et al. (2005).
Page 119
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
However, on the basis of the findings described in this paper, future programmes should attempt to
sample basic statistics (catch by species and effort by gear type) from the major sectors of the fishery
(lot fisheries, artisanal and subsistence fisheries) stratified by major habitat type. Estimates of the total
number and size of gear within each sampling stratum could be used to raise catch rate samples and
effort census to total estimates, and to standardise effort across gears to account for differences in their
catchability (Gulland, 1983). Mapping the results would help to identify where management efforts
to control fishing mortality might best be targeted. Data generated by this type of survey would also
help to improve understanding of the migratory behaviour of fish in the TS-GL System and elsewhere
and provide estimates of total catch weight and value by species for economic valuation and for
environmental impact assessment purposes.
These efforts might begin in the TS-GL to complement the existing data and models for the dai
fishery and to test the validity of the assumptions described above before extending to other locations
in the country depending on available resources and capacity.
6.2.2. Recommendations to improve the existing dai fishery monitoring programme
In addition to these supplementary monitoring activities, the following recommendations are made to
improve the existing dai fishery monitoring programme.
There is evidence of considerable variation in sampling effort between the dai rows. This is
significant from a survey design (and total catch estimation) perspective because of the apparent
decline in catch rates from the most upstream row 15 to row 2 which is believed to reflect the
depletion of fish as they pass through the fishery. Disproportionate sampling effort directed towards
the most upstream dai rows in any given year could therefore generate unrepresentatively high sample
catch rates for the entire fishery and corresponding total catch estimates. The current programme
should therefore be reviewed by a qualified statistician specialising in survey design to maximise
the accuracy and precision of estimates of catch and effort, and to allow for valid inter-annual
comparisons of estimates of total catch, effort, and fish biomass indicated by dai catch rates, given the
available resources.
Reasons for apparent longitudinal differences in species size, diversity and composition between
upstream and downstream locations remain uncertain but differences in dai net mesh size and the
capacity of survey teams to correctly identify species between upstream and downstream locations
may be important. These differences might account for the unexplained variation in the depletion
model and observations of species composition through space and time. It is therefore recommended
that greater effort be given to accurately recording net mesh size for each sampled haul. The ability
of enumerators to correctly identify species should also be checked and training provided where
necessary.
Page 120
Summary and Conclusions
Whilst numerous database queries have been written for the purposes of the analyses described in
this paper, it is recommended that ‘reports’ be programmed using the existing software platform to
automatically generate basic information in the format that is routinely required by the FiA and the
MRC for monitoring and evaluation purposes.
6.2.3. Further research
The relative contribution of the upstream versus adjacent floodplain sources of fish to dai catches
needs to be investigated to determine the reliability of the estimates of the proportion of migrating fish
removed by the dai fishery and the corresponding fishing mortality rate estimates. An expansion of
survey activities to gear targeting fish migrations through channels connecting adjacent floodplains to
the Tonle Sap (downstream of Kampong Tralach) during the open season would provide useful data
for this purpose.
Efforts to establish the migratory range of the target species targeted by dai fishery would
determine the geographic extent over which any management measures or modification to
hydrological conditions in the TS-GL System are likely to propagate to add to recent work (see Halls
et al. in press).
Analyses of the length frequency data sampled from the fishery might also provide further insights
into the population dynamics of the species targeted by the fishery including growth, mortality and
recruitment.
Page 121
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Page 122
7. References
Adamson, E.A.S.; Hurwood, D.A.; Baker, A.M and Mather, P.B (2009) Population subdivision in
the Siamese mud carp Henicorhynchus siamensis in the Mekong River basin: implications for
management. Journal of Fish Biology, 75: 1371-1392.
Ahmed, M.; Navy, H.; Vuthy, L. & Tiongco, M (1998) Socio economic assessment of freshwater
capture fisheries of Cambodia. Report on a household survey, MRC/DANIDA/DoF.
Management of the freshwater capture fisheries of Cambodia, 186pp.
Baird, I.G and Flaherty, M.S (2004) Beyond National Borders: Important Mekong River Medium
Sized Migratory Carps (Cyprinidae) and Fisheries in Laos and Cambodia. Asian Science
Fisheries, 17: 279-298.
Baird, I.G.; Flaherty, M.S and Phylavanh, B (2003) Rhythms Of The River: Lunar Phases and
Migrations of Small Carps (Cyprinidae) In The Mekong River. Nat.Hist.Bul.Siam Soc, 51(1):
5-36.
Baran, E.; Starr, P and Y. Kura (2007) Influence of built structures on the Tonle Sap fisheries.
Cambodia National Mekong Committee and Worldfish Center. Phnom Penh, Cambodia, 44pp.
Baran, E.; Van Zalinge, N and Ngor, P. B (2001a) Analysis of the Cambodian Bagnet ("Dai") fishery
data. ICLARM, Penang, Malaysia, Mekong River Commission Secretariat and Department of
Fisheries, Phnom Penh, Cambodia, Penang, Malaysia, 50pp.
Baran, E.; van Zalinge, N.; Ngor, P. B.; Baird, I. G and D. Coates (2001 b) Fish resource and
hydrobiological modelling approaches in the Mekong basin. ICLARM, Penang, Malaysia and
the Mekong River Commission Secretariat, Phnom Penh, Cambodia, 60pp.
Brooks, S. E.; Allison, E. H and J.D. Reynolds (2007) Vulnerability of Cambodian water snakes: initial
assessment of the impact of hunting at Tonle Sap Lake. Biological Conservation, 139: 401-414.
Campbell, I. C.; Poole, C.; Giesen, W and J. Valbo-Jorgensen (2006) Species diversity and ecology of
Tonle Sap Great Lake, Cambodia. Aquatic Science, 68: 355-373.
Cans, G and Ngor, P. B (2006) Cambodian Bagnet ("Dai") fishery database: overview of the
conversion from ARTFISH to MS Acces database description. Mekong River Commission,
Fisheries Ecology, Valuation and Mitigation Component, Phnom Penh, Cambodia, 26pp.
Page 123
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Chevey, P and F. Le Poulain (1940) La pêche dans les eaux douces du Cambodge 5e mémoire.
Travaux de L'Institut Océanographique de L'Indochine, 195pp.
Clarke, K.R and R.M. Warwick (2001) Change in marine communities: an approach to statistical
analysis and interpretation, 2nd edition. PRIMER-e Ltd. Plymouth.
Clements, T.; O'Kelly, H and Visal, S (2007) Monitoring of large waterbirds at Prek Toal, Tonle Sap
Great Lake 2001–2007. UNDP/GEF Tonle Sap Conservation Project, 60pp.
CoM (1989) Sub-decree on renting inland and marine fishery domains for exploitation (in Khmer
Version), Council of Ministers, Phnom Penh.
Davidson, P (2006) The biodiversity of the Tonle Sap Biosphere Reserve: 2005 status review. Phnom
Penh: UNDP/Wildlife Conservation Society.
Deap, L (1999) The bagnet [dai] fishery in the Tonle Sap River. In Annual Meeting of the Department
of Fisheries of the Ministry of Agriculture, Forestry and Fisheries, eds. N. Van Zalinge and T.
Nao. Mekong River Commission and Department of Fisheries, Phnom Penh.
Deap, L.; Degen, P and van Zalinge, N (2003) Fishing Gears of the Cambodian Mekong. Phnom Penh,
Cambodia Fisheries Technical Paper Series IV, 269pp.
Dubeau, P.; Ouch, P and Sjorslev, J. G (2001) Estimating fish and aquatic animal productivity/yield per
area in Kampong Tralach: an integrated approach. Cambodia Fisheries Technical Paper Series,
3: 143-164.
Dudley, R.G (1972). Growth of Tilapia of the Kafue floodplain, Zambia: predicted effects of the kafue
Gorge Dam. Ph.D. Dissertation, University of Idaho, Moscow, USA, 50pp.
FAO (1999) Guidelines for the routine collection of capture fishery data. FAO Fish.Tech. Pap. 382,
FAO, Rome, 113pp.
Fily, M and F. d'Aubenton (1965) Cambodge - Grand Lac Tonle Sap - Technologie des Pêches
1962–1963. République Française, Minisţère des Affaires Etrangères, service de coopération
technique.
Gray, T. N. E.; Collar, N. J.; Davidson, P. J. A.; Dolman, P. M.; Evans, T. D.; Fox, H. N.; Chamnan,
H.; Borey, R.; Kim Hout, S. and R.N. van Zalinge. (2009) Distribution, status and conservation
of the Bengal Florican Houbaropsis bengalensis in Cambodia. Bird Conservation International,
19: 1-14.
Gulland, J.A (1983) Fish Stock Assessment: A Manual of Basic Methods, Chichester: John Wiley and
Sons, 223pp.
Page 124
References
Gupta, A and S.C. Liew (2007) The Mekong from satellite imagery: a quick look at a large river.
Geomorphology, 85: 259-274.
Hai, N.T., Sunada, Y.K., Oishi, S., and Ikejima, K. Tonle Sap Ecosystem Fish Species Biological
Groups and Hydroecological Index. International Journal of Ecological Economics and
Statistics, 8: 6-26.
Halls, A.S.; Paxton, B.R.; Hall, N.; Hortle, K.G.; So, N.; Chea, T.; Chheng, P.; Putrea, S.; Lieng, S.;
Peng Bun, N.; Pengby, N.; Chan, S.; Vu, V.A.; Nguyen Nguyen, D.; Doan, V.T., Sinthavong,
V.; Douangkham, S.; Vannaxay, S.; Renu, S.; Suntornratana, U.; Tiwarat, T. and Boonsong,
S. (2013). Integrated Analysis of Data from MRC Fisheries Monitoring Programmes in the
Lower Mekong Basin. MRC Technical Paper No. 33, Mekong River Commission, Phnom Penh,
Cambodia, 130pp.
Halls, A.; Lieng, S.; Ngor, P. B and Tun, P (2008) New research reveals ecological insights into Dai
fishery. Catch and Culture, 14(1): 8-12.
Halls, A. S., Arthur, R., Bartley, D., Felsing, M., Grainger, R., Hartmann, W., Lamberts, D., Purvis, J;
Sultana, P., Thompson, P and Walmsley, S (2005) Guidelines for Designing Data Collection and
Sharing Systems for Co-Managed Fisheries. Part 1 & 2: FAO Fisheries Technical Paper. No.
494/1&2. Rome, FAO. 2005, Part 1, 42pp and Part 2, 108pp.
Halls, A.S and Welcomme, R.L (2004) Dynamics of river fish populations in response to hydrological
conditions: A simulation study. River Research and Applications, 20: 985-1000.
Halls, A.S.; Kirkwood, G.P and Payne, A.I (2001) A dynamic pool model for floodplain-river fisheries.
Ecohydrology and Hydrobiology, 1(3): 323-339.
Halls, A.S., Sopha, L and Ngor, P (2007) Landings from Tonle Sap dai fishery in 2006–07 above the
12-year average. Catch & Culture, 13(1): 7-11.
Hap, N and Ngor, P (2001) ‘An economic analysis of fish production in the Dai Fisheries in Phnom
Penh/Kandal Province, Cambodia’, in N. van Zalinge, S. Nuov, R. Ounsted and L. Sopha (eds.),
Cambodia fisheries technical paper series, Volume III, The Management of the Freshwater
Capture Fisheries of Cambodia Component of the Mekong River Commission’s Program for
Fisheries Management and Development and the Department of Fisheries of Cambodia, Phnom
Penh.
Hap, N., Chuenpagdee, R and J. Kurien (2006) Livelihood importance and values of the Tonle Sap
Lake fisheries. Inland Fisheries Research and Development Institute, Phnom Penh, Cambodia,
8pp.
Page 125
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Hav, V and Ngor, P (2005) An overview of aquaculture in Cambodia, Working Paper 2, An output
from the DFID-funded Post-Harvest Fisheries Livelihood Project, Department of Fisheries,
Phnom Penh.
Hilborn, R and Walters, C.J (1992) Quantitative Fisheries Stock Assessment. Choice, Dynamics and
Uncertainty. London, Chapman Hall, 570pp.
Hortle, K.G (2007) Consumption and the yield of fish and other aquatic animals from the Lower
Mekong Basin. Mekong River Commission, Vientiane. MRC Technical Paper No. 16, 87pp.
Hortle, K.G.; Lieng, S and Valbo-Jorgensen, J (2004) An introduction to Cambodia's inland fisheries.
Mekong Development Series No. 4. Mekong River Commission, Phnom Penh, Cambodia, 41pp.
Hortle, K.G.; Ngor, P., Hem, R and Lieng, S (2004) Trends in the Cambodian dai fishery: floods and
fishing pressure. Catch & Culture, 10(1): 7-9.
Hortle, K.G.; Ngor, P.; Hem, R and Lieng, S (2005) Tonle Sap yields record haul. Catch & Culture,
11(1): 3-7.
Junk, W J; Bayley, P B and Sparks, R E (1989) The flood pulse concept in river-floodplain systems. In.
Dodge, D P (ed). Proceedings of the International Large River Symposium. Canadian Special
Publication of Fisheries and Aquatic Science, 106: 110-127.
Junk, W. J.; Brown, M.; Cambell, I. C.; Finlayson, M.; Gopal, B.; Ramberg, L. and B.G, Warner
(2006) The comparative biodiversity of seven globally important wetlands: a synthesis. Aquatic
Science, 68: 400-414.
Kapetsky, J.M (1974) Growth Mortality and Production of Five Fish Species of the Kafue River
Floodplain, Zambia. PhD Thesis. University of Michigan, 194pp
Keskinen, M (2003) Socio-economic survey of the Tonle Sap, Cambodia. A Master of science thesis
submitted for inspection in Espoo, 2003.
Kummu, M and J. Sarkkula (2008) Impact of the Mekong River flow alteration on the Tonle Sap flood
pulse. Ambio, 37: 185-192.
Kummu, M.; Koponen, J and J. Sarkkulan (2005) Modelling sediment transportation in Tonle Sap
Lake for impact assessment. In: SIMMOD '05 International Conference on Simulation and
Modelling 2005. Simulation and Modelling: Integrating Sciences and Technology for Effective
Resource Management. Asian Institute of Technology, Bangkok, Thailand. (http://www.mssanz.
org.au/simmod05/papers/C3-02.pdf).
Page 126
References
Kummu, M.; Penny, D.; Sarkkula, J and J. Koponen (2008) Sediment: curse or blessing for Tonle Sap
Lake? Ambio, 37(7): 158-163.
Lamberts, D (2008) Little impact, much damage: the consequences of Mekong River flow alterations
for the Tonle Sap ecosystem. In: Modern Myths of the Mekong, eds. M. Kummu, M. Keskinen
& O. Varis: 3-18.
Lamberts, D (2001) Tonle Sap fisheries: a case study on floodplain gillnet fisheries in Siem Reap,
Cambodia. RAP Publication. FAO Regional Office for Asia and the Pacific, Bangkok, Thailand,
133pp.
Lamberts, D (2006) The Tonle Sap Lake as a productive ecosystem. Water Resources Development,
22: 481-495.
Lieng, S.; Yim, C and van Zalinge, N. P (1995) Freshwater fisheries of Cambodia, I: the bagnet (Dai)
fishery in the Tonle Sap River. Asian Fisheries Science, 8: 255-262.
Lim Song, S.; Lieng, S.; Ing, T and S Heng (2004) The unsustainable exploitation of inland fisheries
resources in Cambodia. Discussion Paper 14 In: Overcoming factors of unsustainablility and
overexploitation in fisheries: selected papers on issues and approaches. International workshop
on the Implementation of International Fisheries Instruments and Factors of Unsustainability
and Overexploitation in Fisheries , Siem Reap, Cambodia, 19-16 September 2004. FAO/Japan
Government Cooperative Programme. (available google book).
Lim, P.; Lek, S.; Touch, S.T.; Mao, S.-O and Chhouk, B (1999) Diversity and spatial distribution of
freshwater fish in Great Lake and Tonle Sap river (Cambodia, Southeast Asia). Aquatic Living
Resources, 12: 379-386.
Lucas, M.C and Baras, E (2001) Migration of Freshwater Fishes. Oxford, Blackwell Science Ltd,
420pp.
MRC (2005) Overview of the hydrology of the Mekong Basin. Mekong River Commission, Vientiane,
73pp.
MFD (2003) Mekong Fish Database. Mekong River Commission 2003.
Nao, T and N, van Zalinge (2001) Challenges in managing Cambodia fisheries. How we can meet
them. Mekong River Commission and Department of Fisheries, Phnom Penh.
Ngor, P and N, van Zalinge (2001). Dai (Bagnet) fishery: 1994/95-2000/01: catch assessment
methodology and results. Project for Management of the Freshwater Capture Fisheries of
Cambodia. Mekong River Commission/DoF/DANIDA, 47pp.
Page 127
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Ngor, P. B (2000) Dai fisheries in the Tonle Sap River of Phnom Penh and Kandal Province (including
a review of the census data of 1996-97). In: van Zalinge, N.P.; Nao, T and S. Lieng (eds.)
Management aspects of Cambodia’s Freshwater Capture Fisheries. Eleven presentations given
at the annual meeting of the Department of Fisheries of the Ministry of Agriculture, Forestry
and Fisheries. Pp 30-47. Mekong River Commission and Department of Fisheries Phnom Penh,
Cambodia, 27-28 January 2000, 169pp.
Nguyen, X.T and V.H, Nguyen (1991) Investigation of the freshwater fish productivity of Cambodia
(1986-88) Fishery Service of Viet Nam, Hanoi, 126pp.
Penny, D (2006) The holocene history and development of the Tonle Sap, Cambodia. Quarternary
Science Reviews, 25: 310-322.
Penny, D (2008) The Mekong at climatic crossroads: lessons from the geological past. Ambio, 37(3):
164-169.
Penny, D.; Cook, G and S.S, Im (2005) Long-term rates of sediment accumulation in the Tonle Sap,
Cambodia: a threat to ecosystem health? Journal of Paleolimnology, 33: 95-103.
Poulsen, A.F.; Hortle, K.G.; Valbo-Jorgensen, J.; Chan, S.; Chhuoun, C.K.; Viravong, S.;
Bouakhamvongsa, K.; Suntornratana, U.; Yoorang, N.; Nguyen, T.T and Tran, B.Q (2004)
Distribution and Ecology of Some Important Riverine Fish Species of the Mekong River Basin.
MRC Technical Paper, 10: 116pp.
Rainboth, W.J (1996) Fishes of the Cambodian Mekong. FAO Species Identification Field Guide for
Fishery Purposes. FAO, Rome, 265pp.
Sam, N.; Lieng, S and Thor, S (2002) Improving Inland Capture Fishery Statistics in Cambodia. In
New Approaches for the Improvement of Inland Capture Fishery Statistics in the Mekong Basin.
RAP Publication 2003/01, FAO, Rome, 145pp.
Sarkkula, J and J, Koponen (2003) Modelling Tonle Sap for environmental impact assessment and
management support. Water Utilization Program - modelling of the flow regime and water
quality of the Tonle Sap MRCS/WUP-FIN Project, 110pp.
Sarkkula, J.; Baran, E.; Chheng, P.; Keskinen, M.; Koponen, J and Kummu, M (2004) Tonle Sap
pulsing system and fisheries productivity. In XXIX International Congress of Limnology
(SIL 2004). Lahti, Finland.
Sarkkula, J.; Kiirikki, M.; Koponen, J. and M. Kummu (2003) Ecosystem processes of the Tonle Sap
Lake. In 1st Workshop of Ecotone Phase II. Phnom Penh, Cambodia.
Page 128
References
So, N.; Sy Vann, L and Yumiko, Y (2007) Study of the catch and market chain of low value fish along
Tonle Sap River, Cambodia and implications for management of their fisheries - A preliminary
study. Consultancy report for the WorldFish Center. Inland Fisheries Research and Development
Institute and WorldFish Center’s Greater Mekong Region, Phnom Penh, Cambodia, 56pp.
So, N.; Eng, T.; Souen, N and Hortle, K.G (2005). Use of freshwater low value fish for aquaculture
development in the Cambodia's Mekong basin. Inland Fisheries Research and Development
Institute, Phnom Penh and Mekong River Commission Fisheries Program, Vientiane, Lao PDR.
So, N and Nao T (1999) Aquaculture Sector Review, Cambodia. Constancy report for APIP/World
Bank- Fisheries Component, Department of Fisheries, Phnom Penh, 50pp.
Sparre, P and Venema, S.C (1992) Introduction to tropical stock assessment. FAO Fisheries Technical
Paper 306/1. Rome, FAO, 376pp.
Stamatopoulos, C (1994) ARTFISH: a microcomputer system for the statistical monitoring of artisanal
fisheries. Version 1. FAO Rome.
Stamatopoulos, C (1995) Statistical monitoring of the freshwater capture fisheries in Cambodia.
Mission report, field document No. 3. DoF/MRC/DANIDA. Management of the freshwater
capture fisheries of Cambodia project, Phnom Penh, 6pp.
Stamatopoulos, C (2002) Sample-based fishery surveys. FAO Fisheries Technical Paper 425. FAO
Rome, 132pp.
Stuart, B.L.; Smith, J.; Davey, K.; Din, P and S.G. Platt (2000) Homalospine watersnakes: the harvest
and trade from the Tonle Sap, Cambodia. TRAFFIC Bulletin, 18: 115-124.
Sverdrup-Jensen, S (2002) Fisheries in the Lower Mekong Basin: status and perspectives. Mekong
River Commission, Phnom Penh. MRC Technical Paper.
Touch, S.T (1998) Inland fisheries in historic prospective, an afterthought of commercialization. MRC
Symposium, 7-8 December 1998, 17pp.
Touch, S.T (1993) Fish supply and demand in rural Svay Rieng Province, Cambodia, A thesis
submitted as a partial fulfillment of the degree of master, Asian Institute of Technology (AIT),
Bangkok.
Valbo-Jorgensen, J.; Chan, S and Chhuon, K. C (2001) Lateral fish migrations between the Tonle Sap
River and its floodplain. Cambodia Fisheries Technical Paper Series, 3: 120-128.
Page 129
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
van Zalinge, N (2002) Update on the status of the Cambodian inland capture fisheries sector with
special reference to the Tonle Sap Great Lake. Catch and Culture, 8(2): 1-5.
van Zalinge, N.; Sarkkula, J.; Koponen, J.; Loeung, D and Ngor, P.B (2004) Mekong flood levels and
Tonle Sap fish catches. In: Welcomme, R.L., Petr, T. (Eds.), Second International Symposium on
the Management of Large Rivers for Fisheries. FAO, Bangkok, Phnom Penh, Cambodia.
van Zalinge, N.; Degen, P.; Pongsri, C.; Nuov, S.; Jensen, J. G.; van Hao, N. and X, Choulamany
(2003) The Mekong River system. In Second International Symposium on the Management of
Large Rivers for Fisheries. Phnom Penh.
van Zalinge, N.; Thouk, N and Touch, S.T (1998) Where there is water, there is fish? Fisheries issues
in the Lower Mekong Basin from a Cambodian perspective. Project for Management of the
Freshwater Capture Fisheries of Cambobia. Mekong River commission. Seventh Common
Propoerrty Conference of the International Association for the Study of Common Property,
Vancouver, Canada, 10-14 June 1998, 14pp.
Welcomme, R.L and Hagborg, D (1977) Towards a model of a floodplain fish population and its
fishery. Environmental Biology of Fishes, 2: 7-24.
Welcomme, R.L (1979) Fisheries Ecology of Floodplain Rivers, London: Longman, 317pp.
Welcomme, R.L and Halls, A.S. (2001) Some considerations of the effects of differences in flood
patterns on fish populations. Ecohydrology and Hydrobiology, 1(3): 313-321.
Welcomme, RL (1985) River Fisheries. FAO Fisheries Technical Paper 262: 330pp.
Wootton, R.J (1990) Ecology of Teleost Fishes. Chapman & Hall, London, 404pp.
Page 130
Page 131
4
5
8
6
7
30
4
5
5
5
7
3
29
7
7
7
6
27
6
6
7
5
7
7
38
13
10
10
3
1
37
3
12
11
12
11
8
57
14
14
13
11
52
7
13
19
16
13
10
78
10
7
4
7
7
35
4
5
7
9
5
6
36
2
6
17
7
9
41
7
8
9
13
10
10
57
10
12
7
11
6
46
7
9
16
6
8
7
53
7
11
17
11
46
6
17
27
19
22
13
104
47
8
11
11
6
4
40
6
14
16
11
3
24
14
8
13
62
16
36
30
31
21
134
Phnom Penh Municipality
Kandal Province
Both
B0001 B0002 B0003 B0004 B0007 B0008 B0009 B00010 B0001 B0002 B0003 B0004 B0007 B0008 B0009 B00010 B0001 B0002 B0003 B0004 B0005 B0006
2
7
3
2
4
5
13
No data on effort or dai identity
9
38
30
19
13
109
24
47
41
37
25
174
39
68
78
71
58
314
44
75
101
85
83
64
452
9
5
4
18
Total
Summary from the existing dai database of the number of distinct dais sampled in each of the different strata between 1994 and 2008 (count of
unique DaiID in the Dai Fishery database). Dai Yield and Lunar Period strata: B0001 High Yield- Peak Period; B0002 High Yield- Low Period;
B0003 Low Yield- Peak Period; B0004 Low Yield-Low Period; B0005 All DAI- Low Period; B0006 All DAI- Peak Period; B0007 Row2- Low
Period; B0008 Row2- Peak Period; B0009 Excl2- Low Period; B00010 Excl2- Peak Period. No effort information in terms of the number of
hauls or dai ID was recorded for the 1996–96 season.
1994–2000
Row Labels Mon
94–95
Dec
Jan
Feb
94–95 Total
95–96
Nov
Dec
Jan
Feb
Mar
95–96 Total
96–97
Nov
Dec
Jan
Feb
Mar
96–97 Total
97–98
Nov
Dec
Jan
Feb
Mar
97–98 Total
98–99
Oct
Nov
Dec
Jan
Feb
98–99 Total
99–00
Oct
Nov
Dec
Jan
Feb
Mar
99–00 Total
Table 25 8. Annex
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Oct
Nov
Dec
Jan
Feb
Oct
Nov
Dec
Jan
Feb
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Feb
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Feb
Mar
Oct
Nov
Dec
Jan
Feb
Mar
2000–2008
Row Labels Mon
00–01
Oct
Nov
Dec
Jan
Feb
Mar
00–01 Total
01–02
01–02 Total
02–03
02–03 Total
03–04
03–-04 Total
04–05
04–05 Total
05–06
05–06 Total
06–07
06–07 Total
07–08
07–08 Total
Grand Total
21
6
6
6
6
6
30
6
2
2
6
6
6
6
5
4
14
5
7
6
6
6
30
21
2
5
5
4
1
17
3
2
2
5
2
4
4
4
4
4
2
22
3
6
6
5
5
3
28
6
5
5
4
1
63
8
9
11
14
12
54
42
12
13
12
12
16
65
10
12
13
15
13
12
12
11
7
54
6
12
13
9
6
46
53
8
13
10
12
9
52
12
12
13
12
5
13
18
16
16
11
6
80
7
15
15
10
7
10
64
12
12
13
10
6
45
5
9
8
10
4
36
34
6
10
8
10
9
43
7
7
9
12
10
8
11
8
7
31
7
9
9
7
6
38
4
8
10
4
4
14
18
14
11
5
62
20
18
22
19
16
95
21
20
21
18
17
5
102
11
13
12
22
11
69
22
22
30
16
90
658
64
5
14
16
13
13
61
11
15
16
11
7
5
65
13
15
23
13
13
77
5
22
16
16
16
75
726
3
1
3
2
92
209
Phnom Penh Municipality
Kandal Province
Both
B0001 B0002 B0003 B0004 B0007 B0008 B0009 B00010 B0001 B0002 B0003 B0004 B0007 B0008 B0009 B00010 B0001 B0002 B0003 B0004 B0005 B0006
4
17
7
17
1
4
2
14
15
6
8
7
18
1
5
4
18
13
3
10
8
22
2
1
1
4
4
19
17
9
6
7
16
1
1
4
4
20
19
8
8
6
15
1
4
4
7
13
7
4
7
9
25
18
95
77
33
43
35
97
3
1
3
2
4
13
9
7
4
9
4
14
9
21
3
14
4
18
4
17
6
9
4
17
10
13
11
15
4
20
8
14
5
7
12
7
20
82
40
84
29
61
6
13
6
4
16
10
6
10
8
10
15
15
6
9
8
7
12
20
6
11
10
4
13
14
6
13
9
6
14
9
3
4
2
5
7
8
33
60
43
36
77
76
8
19
7
13
8
10
6
12
2
10
22
3
5
4
6
6
24
12
14
18
7
51
396
30
3
8
8
5
6
30
1
11
9
14
5
40
430
77
12
13
12
8
45
524
95
9
14
16
11
50
510
18
7
5
6
4
22
219
25
5
6
6
7
24
233
Total
46
75
87
84
85
55
432
63
87
78
90
73
27
418
65
85
83
73
69
42
417
45
77
83
70
48
5
328
66
88
88
83
76
401
74
78
86
83
64
10
395
54
78
85
85
64
366
6
100
99
118
74
397
4,221
Page 132
Annex
8.1. Dai Fishery principal data tables (Cans and Ngor 2006)
Table 26
tbl_MainWi&Dos
Name
Type
Description
DOC
Text
Document number of the original paper record , number of the data page (per month)
DAY
Number
Number of the day of the month
MONTH
Number
Number of the month
YEAR
Number
Year
Date
Date/Time
Concatenation of Day, Month and Year
GearCode
Text
Gear Code, linked with the Gear list table
Number of haul
Number
Number of haul per day
Sampled weight
Number
Total weight of the sample
TotalWeight
Number
Total weight of the haul
Recorder
Text
Recorder name
StratumNum
Number
Identification of the strata: 1 Phnom Penh, 2 Kandal, and 3 both of them.
SKIPPER
Text
Name of the skipper or No of the Dai unit/ no of row ( No of row (=dai)/Letter of
unit in a row, ex: 12E = dai number 12, 5th bagnet.
SizeDai
Number
Size of the dai in meters
Meshed
Text
Meshed use
NOSP
Number
Number of people working at the dai station
TOTV
Number
Total value of the sample ( SampledWeight*TOTP) ( * 1,000 riels )
TOTP
Number
Weighted average of the sample price ( * 1,000 riels)
TOTN
Number
Total number of individuals in the sample
Table 27 tbl_SpeciesWin&Dos
Name
Type
Description
DOC
Text
Document number , number of the data page ( per month) foreign key from Main table
Date
Date/Time
Date foreign key from main table
GearCode
Text
Gear code key from main table
StratumNum
Number
Stratum number, foreign key from main table
Khmer name
Text
Khmer species name
Catch
Number
Weight of this species (kg)
Value
Number
value of the catch = catch* price
Price
Number
value ( *1,000 riels) per kg
FishNumber
Number
Number of individuals of that species
Page 133
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 28 tbl_Effort
Name
Type
Description
StratumNum
Number
Stratum number
GearCode
Text
Gear code
Recorder
Text
Recorder name
Date
Date/Time
Date
month
Number
Month
year
Number
Year
Active Gear
Number
Number of active gear
Active Day
Number
Number of active days
Table 29 tlkp_SeasonYear
Name
Type
Description
MONTH
Number
Month
YEAR
Number
Year
Season
Text
Fishing season
Table 30 tlkp_GearCode
Name
Type
Description
Gear Code
Text
Code of the gear
Gear name
Text
Name of the gear
Table 31 tlkp_Species
Name
Type
Description
Code in form
Number
Code of the species in the field data entry form
Merge for form
Text
Concatenation of the Khmer name and the code
MFD CODE
Number
MFD, 2000 corresponding code
Page 134
Annex
Table 32 lkp_SpeciesStandard
Name
SpeciesNo
Type
Number
SpeciesCode
lngSpeciesCode
txtSpeciesName
TAXON
Family
Fish or OAA
txt Habitat
txt Indigenous
BlackWhite
Feeding Simple
CommEnglish
KhmerName
Vietnamese
Lao
LocalName Thai
boolInclude
GenusMnS
SpeciesMnS
MFDGenus
MFDSpecies
Logbook
Feeding
Comment
memFeeding
txtSpeciesStatus
Cambodia
LaoPDR
Thailand
VietNam
Yunnan
Rainboth
OldGenusRainboth
OldSpeciesRainboth
memGlobalOcc
memMekongOcc
memHabitat
txtDryHabitat
txtFloodHabitat
MaxLenFemales
txtMaxTypeFem
Reservoirs
Swamps
Lakes
Floodplain
Brackish
Estuaries
Marine
Anadromous
Catadromous
Potamodromous
Limnodromous
Oceanodromous
Amphidromous
Number
Number
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Yes/No
Text
Text
Text
Text
Yes/No
Text
Text
Memo
Text
Text
Text
Text
Text
Text
Text
Text
Text
Memo
Memo
Memo
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Text
Description
LEK number 1–193 was missing 65,176,192 (190 species), plus new nos. added for new
AMCF databases
from Dai data entry form
MFD species code. 2–1323, total 924, numbers > r 1323 added arbritarily
Species Name in MFD
Taxa used for analysis, can combine species by user depending on the data
Family
Fish or other aquatic animal
Indigenous, Exotic
Black, White, Estuarine, Marine
Simplified feeding category
English Name
Khmer Name
Vietnamese Name
Lao Name
Thai Name
Included in LEK survey
Migration survey genus name
Migration survey species name
MFD genus name
MFD species name
In drop-down list for data entry
Feeding category
Comment on the names used
Notes on feeding from MFD
Indicate the status of the species (Questionable, Exclude, Expected or Confirmed)
Indicates if the species occurs in Cambodia, Source (Rainboth, 1996)
Indicates if the species occurs in Lao PDR, Source (Kottelat)
Indicates if the species occurs in Thailand, Source (Chavalit)
Indicates if the species occurs in Viet Nam, Source (Kottelat and Yen)
Indicates if the species occurs in Yunnan (China) (Dr Chen)
Indicates if this species included in Rainboths field-guide
Generic name of the species as used in Rainboth field-guide
Specific name of the species as used in the Raimboth field-guide
Remarks on the global distribution of this species
Remarks on the occurrence within the Mekong basin
Information on habitat
Typical habitat where the species can be found during the dry season
Typical habitat where the species can be found during the flood season
Maximum length for females; note this was always > for males
Type of length measurement
Recorded in reservoirs, man-made
Recorded in swamps
Found in lakes
Recorded on floodplains
Recorded in brackish water
Recorded in estuaries
Recorded in the sea
Known anadromous - ascends rivers from the sea to breed
Known catadromous - descends rivers to estuaries or the sea to breed
Known potamodromous - migrates within rivers to breed
Known limnodromous - migrates within lakes to breed
Known oceanodromous - migrates in the sea to breed
Known amphidromous - moderates up or downstream to breed
Page 135
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 33 tbl _Phnom Penh Port Water level
Name
Type
Description
Date
Date/Time
Date
Level (m)
Number
Water level at Phnom Penh port
Dai Fishery additional Tables (Cans and Ngor, 2006)
Table 34 tbl _MoonFace
Name
Type
Description
Date
Date/Time
Date
Moon Illumination Percent
Number
Percentage of moon face illumination
Table 35 Alphanumeric gear codes in the Access database incorporating
to lunar phase and dai yield (B0001–B004), as well as the
alternative stratifications applied in the earlier sampling
years, i.e. no stratification (All DAI) applied prior to the
1997-98 season and the scheme that included or excluded Row
2 (Row2/Excl2) applied between the 1998–99 and 1999–00
season.e
Gear code
Gear name
B0001
High Yield- Peak Period
B0002
High Yield- Low Period
B0003
Low Yield- Peak Period
B0004
Low Yield- Low Period
B0005
All DAI- Low Period
B0006
All DAI- Peak Period
B0007
Row2- Low Period
B0008
Row2- Peak Period
B0009
Excl2- Low Period
B00010
Excl2- Peak Period
Page 136
Annex
8.2. 2009 Dai Fishery database principal data tables
Table 36 tbl_MainWi&Dos
Name
Type
Description
DOC
Text
Document number of the original paper record , number of the data page (per month)
DaiID
Text
Name of the skipper or number of the Dai unit/row ( number of row(=dai)/ Letter of
unit in a row, ex: 12E = dai number 12, 5th bagnet.
DAY
Number
Number of the day of the month
MONTH
Number
Number of the month
YEAR
Number
Year
Date
Date/Time
Concatenation of Day ,Month and Year
GearCode
Text
Gear Code, linked with the Gear list table
Number of haul
Number
Number of hauls per day
S_Sampled weight
Number
Total weight of the sample taken of the small species
S_TotalWeight
Number
Total weight of all the small species in the haul (TotalWeight– B_TotalWeight)
B_TotalWeight
Number
Total weight of big species in the haul (not sub-sampled)
TotalWeight
Number
Total weight of haul the haul (small and big species combined)
Recorder
Text
Recorder name
StratumNum
Number
Identification of the administrative zone: 1 Phnom Penh, 2 Kandal and 3 both zones
Fisher
Text
Name of the Dai owner
SizeDai_Depth
Number
Depth of the Dai (m)
SizeDai_Width
Number
Width of the Dai (m)
Size Dai_Length
Number
Length of the Dai (m)
Meshed_Max
Number
Maximum mesh size of the bagnet (mm)
Meshed_Min
Number
Minimum mesh size of the bagnet (mm)
Mesh Size_ARTFISH ONLY
Text
Mesh size from ARTFISH database NB Not classified as Max or Min (Max and Min
data combined)
NOSP
Number
Number of people working at the dai station
TOTV
Number
Total value of the sample ( SampledWeight*TOTP) ( * 1,000 riels )
TOTP
Number
Weighted average of the sample price ( * 1,000 riels)
TOTN
Number
Total number of individuals in the sample
Page 137
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 37 tbl_SpeciesWin&Dos
Name
Type
Description
DOC
Text
Document number , number of the data page ( per month) foreign key from Main table
Date
Date/Time
Date foreign key from main table
GearCode
Text
Gear code key from main table
StratumNum
Number
Stratum number, foreign key from main table
Khmer new name
Text
Khmer species name
S_Catch
Number
Weight of the small species (kg)
S_Value
Number
Value of the small catch=catch* price
S_Price
Number
Value of the small catch (/1,000 riels) per kg - ie all prices have been divided by 1,000
S_FishNumber
Number
Number of individuals of small species
B_Catch
Number
Weight of the big species (kg)
B_Value
Number
value of the big catch= catch* price
B_Price
Number
value of the big catch ( /1,000 riels) per kg - ie all prices have been divided by 1,000
B_FishNumber
Number
Number of individuals of big species
Table 38 tbl_Effort
Name
Type
Description
StratumNum
Number
Stratum number (Administrative Zone)
GearCode
Text
Gear code
Recorder
Text
Recorder name
Date
Date/Time
Date
month
Number
Month
year
Number
Year
Active Gear
Number
Number of active gears
Active Day
Number
Number of active days
Table 39 tlkp_GearCode
Name
Type
Description
Gear Code
Text
Code of the gear
Gear name
Text
Name of the gear
Period_numeric
Number
Numeric code for the period (Peak = 1; Low = 2)
Yield_numeric
Number
Numeric code for the Dai Yield classification
(Both = 2; High = 1; Low = 0)
2009 Dai Fishery database lookup tables.
Page 138
Annex
Table 40
tlkp_Species
Name
Type
Description
NewCode
Number
Code of the species in the field data entry form
KhmerName
Text
Khmer species name
KhmerNameCode
Text
Concatenation of the khmer name and the code
MFD CODE
Number
Mekong Fisheries Database 2000 corresponding code
Table 41 tlkp_SpeciesStandard
Name
Type
Description
SpeciesNo
Number
LEK number 1–193 was missing 65,176,192 (190 species), plus new nos. added for
new AMCF databases
SpeciesCode
Number
From Dai data entry form
lngSpeciesCode
Number
MFD species code. 2–1323, total 924, numbers > r 1323 added arbritarily
txtSpeciesName
Text
Species Name in MFD
TAXON
Text
Taxa used for analysis, can combine species by user depending on the data
Family
Text
Family
Fish or OAA
Text
Fish or other aquatic animal
txtHabitat
Text
Habitat
txtIndigenous
Text
Indigenous, Exotic
BlackWhite
Text
Black, White, Estuarine, Marine
Feeding Simple
Text
Simplified feeding category
CommEnglish
Text
English Name
KhmerName
Text
Khmer Name
Vietnamese
Text
Vietnamese Name
Lao
Text
Lao Name
LocalName Thai
Text
Thai Name
boolInclude
Yes/No
Included in LEK survey
GenusMnS
Text
Migration survey genus name
SpeciesMnS
Text
Migration survey species name
MFDGenus
Text
MFD genus name
MFDSpecies
Text
MFD species name
Logbook
Yes/No
In drop-down list for data entry
Feeding
Text
Feeding category
Comment
Text
Comment on the names used
memFeeding
Memo
Notes on feeding from MFD
txtSpeciesStatus
Text
Indicate the status of the species (Questionable, Exclude, Expected or Confirmed)
Cambodia
Text
Indicates if the species occurs in Cambodia, Source (Rainboth, 1996)
LaoPDR
Text
Indicates if the species occurs in Lao PDR, Source (Kottelat)
Thailand
Text
Indicates if the species occurs in Thailand, Source (Chavalit)
VietNam
Text
Indicates if the species occurs in Viet Nam, Source (Kottelat and Yen)
Yunnan
Text
Indicates if the species occurs in Yunnan (China) (Dr. Chen)
Page 139
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Rainboth
Text
Indicates if this species included in Rainboths field-guide
OldGenusRainboth
Text
Generic name of the species as used in Rainboth field-guide
OldSpeciesRainboth
Text
Specific name of the species as used in the Raimboth field-guide
memGlobalOcc
Memo
Remarks on the global distribution of this species
memMekongOcc
Memo
Remarks on the occurrence within the Mekong basin
memHabitat
Memo
Information on habitat
txtDryHabitat
Text
Typical habitat where the species can be found during the dry season
txtFloodHabitat
Text
Typical habitat where the species can be found during the flood season
MaxLenFemales
Text
Maximum length for females; note this was always > for males
txtMaxTypeFem
Text
Type of length measurement
Reservoirs
Text
Recorded in reservoirs, man-made
Swamps
Text
Recorded in swamps
Lakes
Text
Found in lakes
Floodplain
Text
Recorded on floodplains
Brackish
Text
Recorded in brackish water
Estuaries
Text
Recorded in estuaries
Marine
Text
Recorded in the sea
Anadromous
Text
Known anadromous - ascends rivers from the sea to breed
Catadromous
Text
Known catadromous - descends rivers to estuaries or the sea to breed
Potamodromous
Text
Known potamodromous - migrates within rivers to breed
Limnodromous
Text
Known limnodromous - migrates within lakes to breed
Oceanodromous
Text
Known oceanodromous - migrates in the sea to breed
Amphidromous
Text
Known amphidromous - moderates up or downstream to breed
Table 42 Leng_tbllengthFreq (Asterisks indicate name changes or new fields)
Name
Type
Description
DOC
Text
Document number and number of the data page ( per month)
foreign key from Main table
DaiID
Text
Dai identification number
Date
Date/Time
Date foreign key from main table
SpeciesName
Text
Species Name
Length
Number
Total Length (cm)
Num_Fish
Number
Number of fish of length (Length Field)
2009 Dai Fishery database Length Frequency Tables (new)
Page 140
Annex
Table 43 LengtblSpecies (Asterisks indicate name changes or new fields)
Name
Type
Description
DOC
Text
Document number , number of the data page ( per month) foreign key from Main table
DaiID
Text
Dai identification number
Date
Date/Time
Date foreign key from main table
SpeciesName
Text
Gear code key from main table
Total
Number
Total weight of the haul from which the sample was taken
Total_Sample
Number
Total weight of the sample
Subtotal
Number
Weight of the sample by species used to obtain the length information
Table 44 Leng_tblLocation (Asterisks indicate name changes or new fields)
Name
Type
Description
DOC
Number
Document number , number of the data page ( per month) foreign key from Main table
DaiID
Text
Dai identification number
Date
Date/Time
Date foreign key from main table
Recorder
Text
Name of the recorder
Location
Text
Administrative zone (Phnom Penh and Kandal)
Table 45 tbl_LunarAge&Phase* (Asterisks indicate name changes or new fields)
Name
Type
Description
Date
Date/Time
Date
MoonAge
Number
Number of days following the New Moon
Quarter
Number
Moon phase (1–First Quarter, 2–Second Quarter, 3–Third Quarter and 4–Full Moon)
2009 Dai Fishery database Additional Tables
Table 46 tblOtherInfo* (new table) (Asterisks indicate name changes or new fields)
Name
Type
Description
DOC
Text
Document number , number of the data page ( per month) foreign key from Main table
Date
Date/Time
Date foreign key from main table
GearCode
Text
Gear code key from main table
StratumNum
Number
Stratum number, foreign key from main table
Time_Start*
Date/Time
Beginning bagnet soak time
Time_End*
Date/Time
End bagnet soak time
TotalCatch/haul*
Number
Total catch per haul (kg)
No_of_haul*
Number
Number of hauls
Page 141
The Stationary Trawl (Dai) Fishery of the Tonle Sap-Great Lake System, Cambodia
Table 47 Annual Hydrological Indices* (Asterisks indicate name changes or new fields)
Name
Type
Description
Season
Text
Fishing season
FI (km days)
Number
Flood Index
Fiy-1 (km2 days)
Number
Flood Index–1 year
DSI (km days)
Number
Dry Season Index
AFI (km days)
Number
Annual Flood Index
Flood start
Number
Serial days since 1st January
Flood end
Number
End of flood (days relative to January, 1)
Flood duration 1
Number
Days between start and end of flood
FRR (m/day)
Number
Flood Rise rate
DDRy-1 (m/day)
Number
Drawdown rate (previous year)
2
2
2
Table 48 tbl_LunarAge&Phase* (Asterisks indicate name changes or new fields)
Name
Type
Description
Date
Date/Time
Date
MoonAge
Number
Serial days since new moon
Quarter
Number
Lunar quarter (moon age / 7)
Page 142
Mekong River Commission
Office of the Secretariat in Phnom Penh (OSP)
Office of the Secretariat in Vientiane (OSV)
576 National Road, #2, Chak Angre Krom,
Office of the Chief Executive Officer
P.O. Box 623,
184 Fa Ngoum Road, P.O. Box 6101,
Phnom Penh, Cambodia
Vientiane, Lao PDR
Tel. (855-23) 425 353
Tel. (856-21) 263 263
Fax. (855-23) 425E-mail:
363
[email protected] Fax. (856-21) 263 264
Website:
www.mrcmekong.org
© Mekong
River Commission
E-mail: [email protected]
Website: www.mrcmekong.org
Office of the Secretariat in Phnom Penh (OSP)
Office of the Secretariat in Vientiane (OSV),
576 National Road, #2, Chak Angre Krom,
Office of the Chief Executive Officer
P.O. Box 623,
184 Fa Ngoum Road, P.O. Box 6101,
Phnom Penh, Cambodia
Vientiane, Lao PDR
Tel. (855-23) 425 353
Tel. (856-21) 263 263
Fax. (855-23) 425 363
Fax. (856-21) 263 264