NAT. NAT. HIST. BULL. SIAM Soc. 51 (1): 5-36 ，2003

RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS OF SMALL CARPS (CYPRINIDAE) IN THE MEKONG RIVER

Ian ιBa かI1 'd 1ヘMark S. Flahe 同'1， and Bounpheng Phylavanh 1

ABSTRA Cf

τ'hro ughout history ，many differ 耳目 tcultures have associa 胞d lunar cycles with changes in variety a variety of human and animal behaviors. In the southem-most part of La os ，血血 .e area known 鼠“Siphandone" or 血.e 4，0∞islands ，rur 百 1 fishers living on islands 泊 the middle of the mains 悦 am Mekong River are especially conscious of the influence of lunar cycles on aquatic life. life. They associate upriver migrations of large quantities of small cyprinid fishes from

Cambodia Cambodia to La os at the beginning of each year with lunar ph 舗 es. 百 is article examines the fishery fishery for small cyprinids in 血e Kh one Falls area ，Kh ong District ， Champasak Pr ovince ， southem southem La o PDR ，飢da five-year time series of catch -e ffort fisheries da 旬 for a single fence- fJl ter 釘ap are presented. 百lese da 筒 are then compared with catch da 組合om the bag-net fishery fishery in the Tonle Sap River 泊 C 釘 nbodia. It is shown 白紙 the migrations of small cyprinids ， particul 釘'i y Henicorhynchus lobatus and Paralaubuca 砂'P us ，眠 highly correlated with new moon periods at 血e Kh one Falls. Many small cyprinids migrate hundr 哲也 of km up the Mekong River River to Kh one Falls 台。 m 血eTo 叫巴 Sap River and probably 血.e Great Lak e in Cam bodia. Th e evolutionary evolutionary conditions 白紙 have led to the behavior of these fish are discussed ， and manage-

ment ment implications are conside 陀 d.

Key words: Cambodia ，La os ，Mekong River ，副 isanal cap 加 re fisheries ，Cyprinidae ， lunar cycles ，migrations

別 TRODUCTION

As the moon follows its natural phases ， orbiting the e紅白 once per mon 出， our view of it dramatically changes. It begins as a new moon ，hidden from E 紅白's view ，and over 血e course of 29.5 days becomes fully visible and brigh t. Thr oughout history ，many di 丘'erent cultures cultures have associated the cyclic lunar transformations wi 白 predictable changes in a variety variety of human and animal behaviors. 百 e scientific evidence for the existence of these relationships relationships is mixed. In vestigations 加to the behavioral e釘ects of lunar phases on fish 釘 e often often inconclusive (e.g. ，ROBERTSON EJ AL. ，1990; ROOKER & DENNIS ， 1991). Ne 河川leless ， 抑 ong associations between lunar ph 酪 es and some behaviours of p制 icul 紅 fish and crustacean species have been documented; In particul 紅， spawning cycles (e.g. ，CRABτ 'RE E ， 1995;JERL 町 GANDW ∞LDRIDGE ，1992; JOHANNES ， 1981). Fishers around the world have long believed that 血e most successful fishing times are associated with particul 紅 looar

Ip.O. Ip.O. Box 860 ，Pakse ，La o PDR ，[email protected] 2Ge ography De partment ，University of Victoria ，Victoria ，B.C ・， Canada Received Received 30 May 2002; ぉ cepted 8 November 2002.

5 6 IAN G. BAIRO ，MARK S. FLAHERTY ANO BOUNPHENG PHYLAVANH periods. periods. Wh ile this body of knowledge based on personal experiences m ak: es up fishing folklore ，scientific evidence also supports the association between fish catch and lunar cycle cycle (e.g. ，0 叩 BUS 町， 1990). In In the Lao People's Democratic Republic (L ao PDR or Laos) rural ethnic Lao people still still predominantly use a lunar calendar. Th ey are well aw 紅 'e of how lunar cycles affect natural natural and human processes. In th 巴 southem-most part of Laos ， in an area known as “Siphandone" ，or 出e4 ，000 islands ，rural fishers living on islands in the middle of the mainstream mainstream Mekong River are especially conscious of the lunar cycles influence have on aquatic aquatic life. They associate upriver migrations of large quantities of small cyprinid fishes from Cambodia to Laos at the beginning of each year with lunar phases (BAIRD ， 1998; SINGHANOUVONG ET AL. ， 1996; ROBERTS & BAIRD ， 1995; ROBERTS & WARREN ， 1994; ROBERTS ， 1993). Highly Highly migratory populations of small cyprinids are extremely important to 血e people of of the lower Mekong River basin in several ways (BAIRD ET AL. ， 1998; ROBERTS ， 1997; JENSEN ， 1996; ROBERTS & BAIRD ， 1995; LIENG ET AL. ， 1995; ROBERTS ， 1993; BARDACH ， 1959). 1959). Ecologically ， they are important primary consumers of phytoplankton in the mainstream mainstream Mekong River during the dry season (SINGHANOUVONG ET AL. ， 1996; ROBERTS & BAIRD ， 1995; ROBERTS & WARREN ， 1994). Th ey produce the 1釘 gest proportion of animal animal protein consumed by people in the lower Mekong River basin (BAIRD ET AL. ， 1998; ROBERTS ， 1997; LIENG ET AL. ， 1995). Socially and culωrally ， they support ways of life based based around the m 叩 y artisanal fisheries 血at tぽ get them over their long joumey. In In Cambodia ， the bulk of the bag-net fishery (，ぬ i in Kh mer) catch in the Tonle Sap River ，one of the largest inland commercial fisheries in the world ，is dominated by small cyprinids. cyprinids. Th e most import 加 t are the labeoin cyprinids Henicorhynchus lobatus 佃 d， to a lesser lesser extent ， Henicorhynchus siamensis (LIENG ET AL. ， 1995). Th ese species and many others ， migrate out of the Great L ak: e and other seasonally inundated areas in central Cambodia at the end of the monsoon rainy season between November and March of each year ，and move into and up the Mekong River via the Tonle Sap River (VAN ZAL 町 GEET L.， AL.， 1999; BAIRD ， 1998; DIEP ET AL.， 1998; ROBERTS ， 1997; LIENG ET AL. ， 1995). ROBERTS & BAIRD (1 995) note that H. lobatus (referred to by them as Cirrhinus lobatus) is an ecologically ecologically significant species in the lower Mekong River in southem Laos due to its large large biomass ，and it may be the single most important species in the Mekong River below 出e Kh one Falls (ROBERTS ， 1997). Important Important fisheries artisanal in the Kh one Falls area 紅 e also based on migrations of Henicorhynchus spp. ，Paralaubuca typus ，Labiobarbus leptocheilus ， Lobocheilus melanotaenia ， Botia modesta and a number of other species (BAIRD ， 1998; ROBERTS &

BAIRD ， 1995; ROBERTS & WARREN ， 1994; ROBERTS ， 1993). Th ese fish migrations 訂 e similarly similarly significant for fishers living in other parts of Siphandone above the Kh one Falls. BAIRD ET AL. (1998) found that approximately 40 percent of the families in Kh ong District ， Champas ak: Pr ovince (Siphandone 紅 ea) reported 血at Henicorhynchus spp. were 白e most abundant abundant fish species caugh t. Despite the small size of individual fish ， the to ta1 catch su 甲assed all other groups in terms of weight. BAIRD (1 998) proposed that the small cyprinid fence-filter trap (tone in Lao) fishery at at the Kh one Falls targets largely the s創 ne populations of fish that 訂 e caught in 血e bag-net bag-net fishery in central Cambodia. Moreover ，it is believed 出at some of these small cyprinids cyprinids migrate past 出e Khone Falls to the border between Laos and 百四land ，組d RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRA TIONS 7 possibly possibly as far upriver as Vientiane or beyond. The migration of important fish stocks 100s or or possibly over 1，000 km between three countries ，highlights the need for the nations in the the Mekong basin to cooperate in improving their understanding of the biology and socio- economic significance of Henicorhynchus lobatus and other migratory small cypri 凶 d fishes. Joint Joint research and management efforts are warranted to ensure the long-term sustainability of of the fisheries based on such species. Th is paper examines the fishery for small cyprinids in the Khone Falls area ，Khong District ，Champasak Province ，southem Lao PDR. The nature of the fishery is reviewed and a time series of catch-effort fisheries data for a single fence-filter trap are presented. These data are then compared with catch data from the bag-net fishery in the Tonle Sap River River in Cambodia.

STUDY AREA

Th e Mekong River basin supports one of the most diverse fish faunas in the world ， and probably the most fish species for a single river basin in Asia. Approximately 1，200 species species of fish occur in the Mekong basin ，including brackish water areas ， although many have not yet been taxonomically described (VAN ZALINGE ET AL. ， 2000; RAINBOTH ， 1996). As of 2000 ，201 fish species in 109 genera and 39 families had been identified from the Mekong River and adjacent tributaries below the Khone Falls 泊Kh ong District (BAIRD ， 200 1). Many species seasonally migrate long distances up the Mekong River from as f;釘 away as the Great Lake in Cambodia and the South China Sea in Vietnam (BAIRD ET AL. ，

1999; 1999; BA ，IRD ， 1998; VAN ZALINGE ET AL. ， 2000; LIENG ET AL. ， 1995; ROBERTS & BAIRD ， 1995). 1995). Other species are relatively sedentary (BAIRD ET A L.， 2001b; BAIRD ET AL. ， 1999). The Siphandone Wetland 訂 ea is a complex ecosystem found in the mainstream of the Mekong River in the extreme south of Laos. is It made up of large and small inhabited and uninhabited uninhabited islands ，channels ，seasonally inundated forests ， deep-water pools ， rapids and waterfalls waterfalls (DACONTO ， 2001; ALTOBELLI ET A L.， 1998; CLARIDGE ， 1996). Th e Siphandone Wetland area is largely situated in Khong District ，which is in the southem-most part of Champasak Pr ovince (Fig. 1). The aquatic environment of the area is characterized by high biodiversity biodiversity and productivity (BAIRD ， 2001). 百lere are approximately 65 ，000 people in Khong ， the vast majority of whom are ethnic ethnic Lao peasants. For the most p制， these people 紅 e semi-subsistence paddy rice farmers ，who have a long history of habitation in the area. Approximately 94% of families in in the district are involved in subsistence artisanal fisheries and many market part of their catch. catch. The estimated total fish catch for the district in 1996 was 4，000 metric tons ，and US$ 1 million worth of fish was reportedly exported from Kh ong during that year (BAIRD ET AL. ，1998). τ'h e wild-capture fisheries of Kh ong may be more important to local people than than in any other district in Laos. Of the 136 villages in Khong ，86 are situated on islands ， and most of 出e rest lie along the eastem bank of the Mekong River (BAIRD ET AL. ， 1998). The Khone Falls ， in the southem-most part of Khong District ， is the most important area area in Siphandone for wild-capture fisheries ，especially for those targeting migratory species species (Fig. 2) (BAIRD ， 2001; BAIRD ET AL. ， 1998; S町 GHANOUVONG ET AL. ， 1996; ROBERTS & BAIRD ， 1995; ROBERTS ， 1993). 8 lAN G. BAIRD ，MARK S. FL A H ERTY AND B OUNPHENG PH YLAVAN H

.-、 、_ -、‘ 司~

zawJF 1・J 。， 、、 China k'ar ー、' Vietnam 、‘ -，人ー-~

，、

J 、，ヘ r 、- Laos l Gulfof t Tongken Tongken 、，f 〆 、 、 j f. ~ 〆ノ‘

』

玄 n fo Thailand Thailand 、- h

• Bangkok

Gu ぴof Thailand Thailand

。 100

Figure Figure 1. The study area - Khon g Di str ict ，C h ampasak Provin c巴， So ulh ern Lao PDR NZJ日 ヘ叶 ZYAMO

明↓出回刷用

NH 〈開刃「

CZ 〉刃沼山〉的関白〉

~ Seasona い un 制 ed ZU 玄』。

Villages mN 1 Ban Don Tho lathi 10 Ban Don トSom 〉叶 2 Ban Don Sang 11 Ban Don Sahong -OZM 3 Ban Don Det Tok 12 Ban Houa Sadam 4 Ban Don Det Oke 13 Ban Don Phapheng CAMBODIA 5 Ban Don En 14 Ban Hang Khone 6 Ban Don Tan Tok 15 Ban Hang Sadam 7 Ban Don Tan Oke 16 Ban Veun Gnang 8 Ban Khone Tai 17 Ban Don Langa o 3 km 9 Ban Khone Neua 18 Ban Veun Kham

Fi g ur 巴 2. Th e Khon 巴 Fall s ar ea，Khong Di strict ，Champa sak Province ，Southern Lao PDR 、。 10 IAN G. BAIRD ，MARK S. fLAI四 rv 刷 D BOUNPHENG PHYLAVANH

PR O，厄 CT BACKGROUND ，AND THE FENC E- FILTERτ 'RAP FISHERY

Th e Lao Community Fisheries and Dolphin Pro tection Pr oject (LIαDP 吋W 出 established 田 a small ，non-governmen ta1 organization (N ∞) supported ，government proje 心t in Khong Dis 位ict in Janu 紅 Y 1993. 百le LCFDPP'S m 副n objectives were to work with local officials from Kh ong Di蜘 ict' s Agriculture and Fores 町 Of 目白 (AFO) and villagers to facili 飽te the establishment establishment of community-b 部 ed living aquatic resources co-m 組 agement systems. In April April 1993 ，出 .e LCFDPP also established a systematic fisheries monitoring program just below the Kh one Falls in Ban Hang Kh one ，a rural fishing situated village on an island 泊 the the middle of the Mekong Ri ver adjacent to 白e border with Cambodia (F ig. 2). 百le 命y season season fence-filter trap fishery is of great importanceωvillagers living ne 紅白.e Kh one Falls ，including inhabitants of Ban Hang Kh one. It supplies them with considerable amounts of animal protein as well as income. 百le southem end of Nok Kas oum Island ，which is about about 2 km northe 部 t of Ban Hang Kh one ，is 血 .e site of important fence-filter 住aps belonging to to villagers from Ban H 佃g Kh one ，B 佃Kh one Neua and Ban Don Som (see ROBERTS & BAIRD ， 1995). In In July 1997 ， the Environment Pr otection and Community Development in Siphandone Wetland Proj 倒 (E 陀 DSWP) took over 血e responsibilities of the LCFDPP ， including work rela 旬 d to the living aquatic resource co ・management program and the fisheries monitoring program based in Ban Hang Kh one. 百le new project ，however ，con 白lO ed to monitor the fence-filter fence-filter trap fishery at Nok Kasoum Island 泊 1998 and 1999. Th e fence-filter 位'ap fishery has long been ex 位emely important for villagers living 加 Ban Hang Kh one and 0 血ers 泊由.e Kh one Falls 紅 'ea. Fence-filter 回 IPS are commonly used at at the edges of rice paddies 佃 d 泊 small s悦創ns throughout Laos and other p紅 ts of mainland Southeast Asia. 百le tone 回.p fishery dealt with here is ，however ， fundamentally di 首erent and unique ，and is found nowhere else in 血eMekongb 部恒outside of the Kh one Falls. Falls. Th e tone traps used at 血.e edges of paddy fields and 泊 streams 紅 'e gene i" ally smaller 血姐 those situated at 白.e Kh one Falls. Th ey target different species of fish ，and 紅 e generally generally used right at the end of the rainy season rather 出組 at the height of dry season ， 部 is 血e case for the tone fishery at 出e Kh one Falls. 百ley also generally catch as fish they 創芭 moving downstream out of seasonally inundated areas ，rather 白血 when they are moving upriver 泊 the Mekong ，and 血ey 紅 'e often designed differently than tone traps in 血e Mekong (see CLARIDGE EI' AL. ， 1997; ROBERTS & BAIRD ， 1995). Each ye 釘 at the beginning of 血.e dry season ，villagers cons 釘uct fence-fil 加「佐'aps of various various sizes and designs using locally obtained wood ，bamboo ，rattan and vines. Each 回 p is is slightly different ， as it is built 旬白血ep 紅白ul 釘 geographical and hydrological conditions found at the site it is designed for. Th ese traps 釘宮 constructed to target the large m 部 ses of of small cyprinids 白紙 migrate up 白e Mekong River 企om Cambodia each ye 紅 .τ 'h e fish sw 加 up 血e many small channels and rapids 白紙 make up the Kh one Falls. However ， some of the channels cannot be easily ascended due to the presence of imp 部 sible rapids and waterfalls so 血ey must move up and down the channels until 白ey fmd a channel 也ey can get up (ROBERTS ， 1997; ROBERTS & BAIRD ， 1995). Only a few channels ， the most notable notable being Hou Sahong and ， to a lesser ex 低 nt ，Hou Sadam 釘 'e easily passable year round (BAIRD EI' AL. 2001a; ROBERTS ， 1997; BAIRD ， 1996; ROB 眠時& BAIRD ， 1995). Villagers Villagers understand the local movements of these fish very well. Th ey cons 凶 ct 血eir fence-filter fence-filter traps so as to catch the migrating fish not when they 釘 'e moving ups 位eam ， as RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRA TIONS 11 might might be expected ， but when they move downstream after unsuccessfully attempting to pass pass a particular channel with a high waterfall or steep rapid at its upper end. Retuming fish fish are sometimes caught in large quantities over a few days each year during peak periods of of the highly seasonal fishing period (BAIRD ， 1998; ROBERTS & BAIRD ， 1995). One of the most important fishing areas for the dry season fence-filter trap fishery at Khone Falls is the southem tip of Nok Kasoum Island (Fig. 2). Located in the geographical center center of the Khone Falls ，local fishers who hold traditional tenure over particular trap sites generally generally catch large qu 加 tities of fish during peak catch periods (ROBERTS & BAIRD ， 1995). 1995). Owing to the strategic location of Nok Kasoum Island ， the relatively large catches from from the area ， and its close proximity to project's base at Ban Hang Khone ，it was decided that that this would be an ideal place to study and monitor the fence-filter trap fishery. Several Several fence filter traps are operated in the Khone Falls area. It was decided ，however ， to to concentrate data collection at one key fence-filter fishing operation at Nok Kasoum Island Island (Figs. 3-10). Accurately monitoring more sites would have required much more project project labor than was available. We were fortunate in that the third author is the son-in- law law of one of the co-owners of a good fence-filter trap site on Nok Kasoum Island. His in-laws in-laws are from Ban Hang Khone while the other owner lives in Ban Khone Neua. Th e two communities are situated at either end of the 5・km long Khone Island ，which straddles the the Khone Falls at its center. The fence-filter trap fishery in the Khone Falls area is subject to a long-practiced tenure tenure system that dictates the nature of the fishery. Because there are only a limited number of high quality tone trap sites in the Khone Falls area ， and since some sites are more effective than others ， high quality f，巴 nce-filter trap sites have long been in high demand. demand. As a result ， an indigenous tone trap tenure system has developed over generations. It It gives individual families ， or groups of two or more families ， exclusive rights for constructing constructing tone traps at particular sites. Tenure is initially established through building a tone trap at a particular location. If the site has never been used or claimed by another family family for fish trapping ， and does not negatively impact on the ability of previously established established traps nearby to catch fish ， the first claimants become de facto owners of the site. site. These ownership rights are not unlike other private prope 口y rights in that fishing site rights rights can be freely traded or rented to others ， and can also be inherited through family lines. lines. For example ，if a family owns a particular tone site but for some reason decides not to to build a trap there during a particular year ，others who do not hold traditional tenure over the the site must ask permission of the owner to build a trap there. They may also be obliged to to pay rent to the owner for the right to fish there for a single season. After the end of the fishing fishing season ， tenure reverts to the original owner (BAIRD ET A L.， 2001a; ROBERTS & BAIRD ， 1995). In In many cases tone fishing sites are owned by single families. These families may or may not have sufficient labor to run a particular tone fishing operation on their own. If family family labor is insufficient ， the owner may decide to bring in one or more partners to assist in in building the trap and subsequently running the fishing operation. However ， these agreements agreements rarely take the form of normal employer/employee relationships. Although only only one family owns the trap site ，all trap construction costs and labor expenditures associated associated with the fishery are divided equally between the partners. Moreover ， the fish catch catch is divided equally amongst all those participating in the fishery ，regardless of their ownership ownership status (ROBERTS & BAIRD ， 1995). 12 I AN G. B AIRD, M ARK S. FLAHERT Y AND B OUNPHENG PH YLAVANH

Fi gure 3. Fence-fi lter trap at Nok Kasoum Island, Khong Di stri ct Champasak Province, southern Lao PDR

Figure 4. (same as above) RHYT I-I MS OF T I-I E R1 VE R: LU NAR PH ASES AN D M JGRAT IONS 13

Figllt 官 7. Pαralaubu ca typU .I' (pa te p)

Figllr e 5. (sallle as above)

Figllr e 6. Relllo vin g th e ca tch ba sk巴I fr Olll th 巴 fe nce- fill 巴r Ir ap at Nok Ka SO lllll I slancl 14 lAN G . s A IRD ，MA RK S. FL A H ERT Y AN D s O UN PH ENG P H Y LA VAN H

F igur e 8. T he fir sl author w 巴ighsc atch fro l11 -fil lerl rap F igure 9. S un -dri ed P ara lallbll ca l yp ll s (pa lep ) fO I 日t Nok K asou l11 Island in the pr ese nce o f sal巴 in lh e Pak s巴 l11 arket，C hal11 pasak fishers Prov ince ，s outh巴rn La o PDR

F igur e 10. S l11 all cypri nids al巴 laid out on ro cks on No k K asou J1l lsland 10 bes un -dri ed RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 15

Considering Considering the egalitarian nature of such 担 Tangements ， which clearly indicates the value value of labor in the production proc 巴ss ， one might wonder what ，if any ， advantages trap owners owners receive over their non-owning pぽ tners. In fact ，there appe 釘 s to be no immediate material material advantage for the owners. Owners do ，however ， have several important long-term advantages. advantages. Th e frrst is that they have the exclusive right to sell or rent out the tone trap site site to others ，and g釘 ner all the associated economic benefits. Secondly ， owners have the right right to decide whether they w 叩 t partners and ， if so ，how m 加 y. If after one ye 釘 of cooperating cooperating with one family the owner decides 白at he would rather work with someone else ， he is under no obligation to renew the old agreement for the coming year. Th is provides provides the owners with more occupational security 白an their non-owning p訂 tners. 官官dly ， it it gives owners the opportunity to gain social influence in their villages ，especially when they they own a particularly good fishing site (see ROBERTS & BAIRD ， 1995). Historically ， people in the Khone Falls area who do not have tone trap sites ， or who were were unable to catch enough fish to feed themselves ， can obtain fish by asking the owners of of good trap sites to give them small to moderate amounts of fish for free. Even today ， villagers villagers sometimes make requests when they arrive at trap sites during peak fishing periods. periods. Although the growing commercialization of 白e 白shery is making it increasingly di 妊icu 1t for people to ask for fish for 仕ee ， the tradition of giving fish away still persists in in the Kh one Falls area (ROBERTS & BAIRD ， 1995). However ， those who get fish for free 釘 e，inωrn ， expected to give fish away when 血ey have a lot of fish and others 釘 'e in greater greater need. 官邸has lead to the saying ，“ kho pheng koua seu ，" which roughly translates into ，“ It is more expensive to ask for something for free 由加 to pay for it."

METHODS

Th e tone tenure system in the Nok Kasoum Island 訂 ea is well established ，and in recent recent years ，all of the fence-filter 位aps in 出e 釘 ea have been bui 1t in essentially the same places places and in the same ways. Th e same people have also operated the traps ，and there have been been no new en 凶es into the fishery in 血e Nok Kasoum Island area for at least the last ten years. years. However ， the fishery undoubtedly experienced a period of exp 佃 sion a few decades ago. ago. Th ere are a dozen or so traps of various shapes ，sizes and efficiency levels in the Nok Kasoum Island area ，and probably more than 200 throughout the Khone Falls (ROBERTS & BAIRD ， 1995). Data Data collection at the tone site at Nok Kasoum Island was quite straightforward. One single single immovable trap built in the same manner ， and situated at 出e same location in the river river each year was monitored. Each ye 訂 about five or six families operate it ， although there there are only two owners. Th is facilitated obtaining yearly data for the fishery that are directly directly comparable. Since the vast majority of fish caught in 出is fishery 訂 e taken during daylight daylight hours ，project officers were stationed at the designated trap with notebooks and scales. scales. Th ey recorded the species and weights of all fish caught for entire seasons ， which generally generally run 仕'O m between late J佃 U 訂 y and March or early April ， depending on hydrological cycles. cycles. On several occasions ， the small size of individual fish and the large quantities caught over over short periods necessitated lumping large numbers of fish together for weighing. Wh en there there was insufficient time to sep 訂 ate large quantities of fish by species ， l-kg random 16 16 IAN G. BAIRD ，MARK S. F'LA HERTY AND BOUNPHENG PHYLAVANH

samples samples were taken 合om large batches and were treated as representative of the total catch. Because Because large quantities of fish 釘 e caught over short periods of time during pe ak: seasons ， it it would have excessively inconvenienced fishers if we had insisted on weighing each fish.

Once large quantities of fish 釘 e caught ，v i1l agers 創 e generally anxious to lay the fish out on rocks to 命yin 白e sun so 血at they do not spoi l. τ'h e tone fishery data collected have been entered into a relational database management sy 取 m (RDMS) using 出e Microsoft programs EXCEL 組 d ACCESS. 百le variables recorded were weights of individual fish species in grams ， the fishing ge 紅 used to catch the fish ， the the number of ge 紅 s used ， the n創 nes of the fishers ， the time periods fishing effort took place ，and the number of hours of fishing.

RESULTS

Table Table 1 presents the total catch summaries by species and weight for each of the five fence- ft1 ter trap or tone fish 泊gse 部 ons monitored between 1995 佃 d 1999. Table 2 combines the the results of all five se 邸 ons. 官 le results c1 ear 行 indicate 出e dominance of 血e small cyprinids cyprinids Henicorhynchus lobatus (pa soi houa lem) and Paralaubuca typ 凶{抑 tep) in 血e fishery. fishery. H. lobatus made up 45% of the combined catch for the five ye 紅 period ，wh i1 e P. 砂'P us constituted 33%. Together they made up almost 79% of the to ta1 combined harvest for for all ye 釘 s. Al though P. typus catches exceeded those of H. lobatus in 1995 and 1996 ， between between 1997 and 1999 H. lobatus was more abundan t. Ann ual H. lobatus catches ranged from between 35 and 77% ，wh i1 e P. typus catches were between 11 and 48%. 官 le 白ird most important species in 出e tone fishery was Henicorynchus siamensis

(pa (pa soi houa po) ，which made up 釦 average of 5.4% of the catch over 出e five ye 紅 S (創 mual range 2. 6- 11%). 百le fourth most abundant species was 血e cyprinid La biobarbus leptocheilus leptocheilus (pa lang khon) ， with 4.8% (armual range O.ι12%). Th e cobitid loach Botia modesta modesta (pa mou man) was fi 伯 most abundant with 4.4% (range 0. 4- 14%). 百le small cyprinid cyprinid Crossocheilus reticulatus (p a tok thoi) made up 1.2% (r ，如ge 0.08- 1. 7%). 百le remaining remaining 84 species recorded in catches all constituted less 白血 1% of the total catch for the the combined five-year period. However ，it is notable 血at Cirrhinus microlepis (pa phone mak koke) made up 4.1 % of the catch 泊 1995 ， but were very rare or absent in catches for all all 0 出er years. SINGHANOUVONG ET AL. (1 996) also recorded relatively high catches of Cirrhinus Cirrhinus microlepis in the Si 抑制done Wetland area upriver from the Kh one Falls in early 1995. 1995. Tables Tables 1 and 2 in c1 ude the minimum ，maximum and mean individual weights ， along with with their standard deviations ， for each of the species recorded 泊 the tone fishery monitored. However ， only p紅 t of the total catch could be used to derive these statistics ， as it was not possible possible to in c1 ude fish weighed in bulk. From the five ye 釘's of catch data available ，it h部 not been possible to detect any significant significant trends 泊 catch de c1 ines or increases for the tone fishery. However ， since only one fishing site was monitored ，albeit one of the most important single 住aps 泊 the Falls ， it it is possible 白at the fish catch data collected may not have been representative of the overall overall fishery. However ，we believe ，from having observed many tone 佐ap catches in other other charmels of the Kh one Falls ，出鉱山 e relative differences between 創官邸 are generally not not great ， although there 創ち exceptions. RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 17

Table Table l. Summary of 1995-1999 fence-filter trap (tone) fish catches at Nok Kasoum Island ，Kh ong District ，Champasak Pr ovince ，Lao PD R. Th e top 8 species are listed ， plus others having at least 0.5% of the catch for the years.

# Latin name T_ weight (g) % catch Min (g) Max (g) Mean (g) StD ev (g) 1995 1995 Paralaubuca Paralaubuca typus 1，648 ，137 47.63 3 25 7 4 Henicorhynchus 2 IHenicorhynchus lobatus 1，163 ，126 33.62 2 11 5 2 Cirrhinus 3 Cirrhinus microlepis 142 ，780 4.13 60 200 124 27 Henicorhynchus 4 IHenicorhynchus siamensis 130 ，992 3.7 9 5 60 29 15 Labiobarbus 5 ILabiobarbus leptocheilus 99 ，617 2.88 5 27 13 7

6I Botia m刀desta 62 ，938 1. 82 4 20 10 4 Crossocheilus 7I Crossocheilus reticulatus 38 ，668 1.1 2 3 13 5 3 8I Barbodes altus 17 ，320 0.5 0 25 150 74 31 9 IOthers (68 species +) 156 ，469 4.52 Totals Totals 3，460 ，047 100.00 1996 1996 Paralaubuca η 'P us 2，053 ，534 48 .3 9 3 5 4 Henicorhynchus 2 IHenicorhynchus lobatus 1，792 ，600 42.24 6 7 7 Labiobarbus 3 ILabiobarbus leptochei/us 123 ，390 2.91 10 17 13 5 Henicorhynchus 4 Henicorhynchus siamensis 111 ，819 2.63 8 8 8 Crossocheilus 5 Crossocheilus reticulatus 55 ，055 1.3 0 3 5 4 6I Botia modesta 41 ，093 0.97 5 10 7 3 Lobocheilos 7I Lobocheilos melanotaenia 34 ，317 0.81 5 5 5 8 I Garra fasciacauda 7 ，882 0.1 9 3 5 4 9 I Others (37 species +) 24 ，100 0.57 Totals Totals 4，243 ，790 100.00 1997 1997 Henicorhynchus Henicorhynchus /obatus 1，745 ，148 44.95 19 6 2 Paralaubuca 2 IParalaubuca typus 806 ，310 20.77 3 21 6 2 3I Botia modesta 544 ，308 14.02 3 20 8 3 Labiobarbus 4 Labiobarbus /eptocheilus 303 ，768 7.82 3 40 10 5 Henicorhynchus 5 IHenicorhynchus siamensis 145 ，187 3.7 4 5 56 14 8 Thynnichthys 6 I Thynnichthys thynnoides 95 ，580 2.46 4 20 13 4 Crossocheilus 7I Crossocheilus reticulatus 65 ，772 1. 69 2 10 4 8 Garra fasciacauda 34 ，703 0.89 2 7 4 18 18 lA N G. BAIRD ，MARK S. FLAHERTY AND BOUNPHENG PHYLAVANH

Table Table 1 (continued).

# Latio oame T_ weight (g) %ωItch Mio (g) M 鉱 (g) Mω 圃 (g) StD ev (g)

Lobocheilos 9 Lobocheilos melanotaenia 31 ，204 0.80 2 33 10 9 Gyrinocheilus 10 Gyrinocheilus pennocki 28 ，150 0.7 3 3 100 22 25 11 11 Others (52 species +) 82 ，494 2.1 3 Totals Totals 3，882 ，624 l∞.∞ 1998 1998 Henicorhynchus Henicorhynchus lobatus 615 ，121 35.08 3 45 7 5 21 Paralaubuca か1fJ us 451 ，406 25.74 2 37 6 5 3 I La biobarbus leptocheilus 214 ，225 12.22 5 110 21 18 Henicorhynchus 4 IHenicorhynchus siamensis 196 ，396 11. 20 7 220 39 43

5 IScaphogn 四thops bandanensis 39 ，351 2.24 10 250 45 29 61 Botia modesta 30 ，197 1. 72 3 150 18 26 71 Barbodes altus 24 ，175 1. 38 20 310 89 48 Crossocheilus 8 Crossocheilus reticulatus 24 ，021 1.3 7 2 15 7 3 Thynnichthys 9 I Thynnichthys thynnoides 20 ，070 1.1 4 12 33 20 10 10 10 Sikukia gudgeri 17 ，433 0.99 4 70 35 14

11 11 Lobocheilos melan ゅtaenia 15 ，726 0.90 5 60 20 16 I 12 12 Mekongina erythrospila 12 ，582 0.72 5 402 128 101 13 13 Garra fasciacauda 12 ，555 0.72 4 160 11 28 14 14 Gyrinocheilus pennocki 8 ，937 0.51 10 138 37 37 15 15 Others (67 species +) 71 ，336 4.07 Totals Totals 1，753 ，531 100.00 1999 1999 Henicorhynchus Henicorhynchus lobatus 1，818 ，947 76.93 Henicorhynchus 2 IHenicorhynchus siamensis 257 ，7∞ 10.90 31 Paralaubuca 砂1fJ us 252 ，369 10.67 Labiobarbus 4 Labiobarbus leptocheilus 13 ，065 0.55 51 Botia modesta 9 ，170 0.3 9 Lobocheilos 61 Lobocheilos melanotaenia 5 ，115 0.22 7 I Garra fasciacauda 3 ，292 0.1 4 Crossocheilus 81 Crossocheilus reticulatus 1 ，962 0.08 910 由ers (6 species) 2 ，650 0.1 1 Totals Totals 2，364 ，270 l∞ .00 RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 19

Table Table 2. Summary of tot a1 combined 1995-1999 fence-filter trap (tone) 自由 catches at Nok Kasoum Island ，Khong Dis 甘ict ，Champasak Province ，Lao PDR

T_weight T_weight %catch Min Max Mean StD ev # La tin name Lao 皿ame 匂) (g) ω ω ω Henicorhynchus Henicorhynchus lobatus pa soi houa lem 7，134 ，942 45.44 45 6 2I Paralaubuca η'pus pa tep 5，211 ，756 3 3. 19 2 37 6 4 Henicorhynchus 3 I Henicorhynchus siamensis pa soi houa po 842 ，0ヲ4 5.3 6 5 220 27 32 4 ILa biobarbus leptocheilus pa lang khon 754 ，065 4.80 3 110 13 12

5I Botia modesta pa mou 刷 n 687 ，706 4.3 8 3 150 11 14 Crossocheilus 6I Crossocheilus reticulatus pa toke thoi 185 ，478 1.1 8 2 15 5 2

7I Cirrh 印刷 microlepis pa phone 脚 kkoke 148 ，135 0.94 20 210 106 39

Thynnichthys 8 I Thynnichthys t伽noides pa ko 附 117 ，820 0.75 4 33 14 5

9μ bocheil ω脚 lanotaenia pa khiang 101 ，996 0.65 2 ω 13 12

10 10 Scaphognathops bandanensis pa plO n 65 ，354 0.42 10 250 42 25 II II Garra fasciacauda pa khiko 61 ，841 0.3 9 2 160 7 17 12 12 Barbodes altus pa vianfai 45 ，015 0.29 20 310 82 40 13 13 Gyrinocheilus pennocki pa ko 43 ，620 0.28 3 138 26 27

14 14 Sikukia gudgeri pa khao 加 35 ，727 0.23 4 2ω 35 19 15 15 Puntioplites falc 俳r pa sakang 30 ，047 0.19 3 165 64 30

16 16 Hypsibarbus 脚 Icolmi pa pak kom 28 ，894 0.18 5 3ω 92 pa pa sa. “ 17 17 Mekongina eryth 叩'spila ， 23 ，196 0.15 5 402 81 77 18 18 Cy c/ ocheilichthys enop ω pa choke. 仰 chok 19 ，670 0.13 5 1，530 320 327 19 19 Himantura chaophraya pa fa lailpa fa hang 14 ，8∞ 0.09 1，3∞ 1，3∞ 1，3∞

20 20 Coius undecimradiatus pa se 削 13 ，076 0.08 6 1，2∞ 2ω 202 21 21 Scaphognathops st り，negen pa plO n 11 ，ω5 0.07 12 3ω 93 56 22 22 H. 10batusl L. leptocheilus pa soilpa lang khon 11 ，040 0.07 23 23 Cosmochilus harmandi pa mak ban 9 ，854 0.06 3 950 “ 159 24 24 Botia ωudipunc 加talspp pa mou man 8.1 99 0.05 3 15 7 3 25 25 Bagarius yarre l/ ilspp. pa khe 7，885 0.05 10 8ω 254 240 26 26 Bangana behri pa va na no 6，765 0.04 10 3∞ 73 63

27 27 Mystacoleuc ω附 rgmatus pa lang ko 5，867 0.04 5 180 17 25 28 28 Henicorhynchus lineatus pa soi lai 5，555 0.04 2 6 5 Belodontichthys 29 Belodontichthys dinema pa khop 4，ωo 0.03 100 1，100 356 312 30 30 Cirrhinus molitore l/ a pa kellg 4 ，600 0.03 5 4∞ 46 74 31 31 Botia helodes pa kheo kai 4，451 0.03 25 12 5 32 32 Pangasius pleurotaenia pa gnone thong khom 3 ，760 0.02 22 150 77 41 33 33 Crossocheilus siamensis pa khang lai noi 3，652 0.02 2 15 5 2 34 34 Hemibagrus nemurus pa kot leuang 3，545 0.02 20 判。 135 90 35 35 Rasbora spp pa s，eu ao 3，451 0.02 3 15 6 3 36 36 Cirrhinus mrigala pa nang chan 3，395 0.02 60 280 139 68 20 20 IAN G. BAJRD ，MARK S. FLAHERTY AND BOUNPHENG PHYLAVANH

Table Table 2 (con 血lU ed).

T_weight T_weight %伺 tch Min Max M 個目 StD ev # La tin name Lao name (g) (g) (g) (ω ω 匂)

37 37 Lucioson 昭 bleekeri pa 剛 kvai 3 ，2 58 0.02 20 180 47 39 38 38 Acantopsis sp. or spp. pa hak kouay 2，866 o.ω 10 7 3

39 39 Hypsihar 伽slagleri papak 問 2，605 0.02 20 435 165 120

40 40 Osteochilus microcephaluslwaandersii pa 幼ang lai gnai 2，354 0.01 2 22 8 6 41 41 Opsarius pulchellus pa lai kho ω ng 2，183 0.01 10 4 2 42 42 Glyptothorax spp. pa kon 1，711 0.01 2 10 5 2 43 43 Morulius chrysophekodionlspp. pa phia 1，685 0.01 5 1，0∞ 2ω 333 44 44 Chi 伯， la blanci pa tong kai 1，620 0訓 80 9∞ 540 419 45 45 Osteochill 郎脚lanopleu. 附 pa khang lai gnai 1， 'ω5 0.01 10 1∞ 44 28 46 46 Boesemania microlepis pa kouang 1，600 0.01 1，'ωo 1，6∞ 1 ，6ω 47 47 Hemisilurus mekongensis pa nang deng 1，530 0.01 10 480 121 162

48 48 HYl 悶ibarb. 凶 we. 仰 orei pa pak thong leuanglpa pak 蜘 m 1，.制。 0.01 1，480 1，480 1" 綿0 Cyclocheilichthys 49 Cyclocheilichthys armatus pa doke ngieu l刈8 0.01 5 230 72 61 50 50 Tenualosa thibaudeaui pa mak phang 1，245 0.01 10 90 39 17 51 51 Kryptopterus spp. pa pik kai 1，230 0.01 5 150 32 29 52 52 Wallago attu pa 蜘。 1，2∞ 0.01 1，2∞ 1 ，2ω 1，2∞

53 53 Pangasius 初印刷i pa 蜘耐醐 1，150 0.01 267 350 308 59 54 54 Pangasius conchophil ω， pa phalpa ke 995 0.01 l∞ 3ω 199 89 55 55 Toxotes microlepis pa mong 905 0.01 20 l∞ 50 24 56 56 Helicophagus waandersi pa no uJ pa hoi 8お 0.01 15 240 123 91 57 57 Pangasius macronema pa gnone tha. 附幼 815 0.01 10 l∞ 41 28 HenicorhynchuslParalaubuca 58 HenicorhynchuslParalaubuca spp. pa soilpa tψ， 790 0.01 59 59 Notopterus notopterus pa ωIng na 7ω o. ∞ 20 200 117 70 ω Micronema apogonlmicronema pa nang khaolpa sangoua 575 o. ∞ 5 240 73 93

61 61 Hemibagr 郎町'ckioides pa 幼eung 511 o.ω 10 501 256 347 62 62 Lycothrissa crocodill ω pa mak chan 5ω o.ω 38 38 38

ω Labeo 町'thropte~凶 pa va souang 420 o.ω 10 100 49 30 64 64 La ides hexane maJ spp. pa gnone thong 405 o.ω 5 50 19 14 65 65 Parambassis wo lffiJ spp. pa khap khong 紛4 o. ∞ 54 28 37

66 66 Ompok bimac uJ atus pa seu ，間 390 o.ω 20 120 78 41 67 67 Hampala macrolepidota pa sout 3ω o. ∞ 3∞ 3ω 3∞ 68 68 Mystus singaringanlspp. pa kha gneng 3∞ aω 10 55 33 16 69 69 Onychostoma cf. elongat 酬， pa khiang fai 295 o. ∞ 10 40 20 10 70 70 Probar 加， labeamajor pa eun khao 250 aω 250 250 250

71 71 Mastacemblus am 招tuslspp. pa lat 214 aω 2 40 24 15 72 72 Tetraodon leiur ω，:/spp. papao 205 a∞ 10 65 24 19 RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRA TIONS 21

Table Table 2 (continued).

T_weigbt T_weigbt %catch Min Max Mean StD ev 事 Latin name Lao name (g) (g) (g) (g) (g) (g) 73 73 Rhinogobius spp pa bou 183 。∞ 3 5 4 74 74 Osphronemus exodon pa men 150 O. ∞ 20 50 35 21 Amblyrhynchichthys 75 Amblyrhynchichthys truncatus pa ta po 150 O. ∞ 25 100 63 53 76 76 Raiamas guttatus pa sanak 120 O. ∞ 50 70 ω 14 77 77 Xenentodon canc i/ a pa kathong l飢} O. ∞ 5 20 13 6 78 78 Opsarius koratensis pa Sleu 100 0.00 79 79 Poropuntius deauratus pa chat l∞ O. ∞ 5 30 14 II Pristolepis 80 Pristolepis fascia ω pa ka 80 O. ∞ 20 20 20 。 81 81 Epalzeorh y. 励。'sfrenatum pa dout hin 80 O. ∞ 5 5 5 。 82 82 Systomus orphoides pa pok 65 O. ∞ 5 50 22 25 83 83 Glossogobius koragensis pa bou khao 45 。∞ 5 10 6 2 84 84 Bagrichthys macracanthus pa kouay souk 40 O. ∞ 10 30 20 14 85 85 Euryglossa panoides pa pan gn Ul 40 O. ∞ 20 20 20 86 86 Hemimyzon sp pa tit hin 35 O. ∞ 5 5 。 87 87 Pseudomystus siamensis pa hia khi 30 。∞ 10 20 15 7

88 88 Macrognathus siamen 四islspp pa lot 30 。ω 10 20 15 7 89 89 Hypsibarbus malcolmi (j uv.) pa khao lan 30 0.00 30 30 30 90 90 Chonerhinus nφ'stus pa pao louang 25 O. ∞ 5 15 10 7 91 91 Kη 'ptopterus cryptopterus pa pik kai 20 。ω 92 92 Achiroides spp. pa pan 15 。ω 5 10 8 4 93 93 Mystacoleucus atridorsalis pa lang ko 15 。ω 5 。 94 94 Oxyeleotris marmorata pa bou 10 O. ∞ 5 5 5 。 Totals Totals 15 ，703 ，5 25 l削J. ∞ 22 lAN G. BAIRD ，MARK S. FLAHERTY AND BOUNPHENG PHVLAVANH

Table 3. Fish species believed to migrate f加m the Tonle Sap Ri ver to Kh one Falls at the begiru 由19 of each dry season. Fish species listed as being caught in 血e tone fishery fishery by ROBERTS & BAIRD (1995) but not included here 紅 'e presently not believed believed to migrate from the Tonle Sap River to 血e Kh one Falls.

書 Latin name Listed by Roberts & 8aird (199 5) Listed by Lieng et al. (1995)

Henicorhynchus Henicorhynchus lobatus yes ，箇 Cirrhinus loba 似S yes ，飴 Henicorhynchus spp. 2 Henicorhynchus siamensis y凶， as Cirrhinus siamensis yes ，飴 Henicorhynchus spp. 3 Henicorhynchus lineatus yes ， as Cirrhinus lineatus yes? ，鎚 Henicorhynchus spp. 4 Cirrhinus microlepis yes yes 5 Paralaubuca 砂'P us yes yes 6 La biobarbus leptocheilus yes yes ，鎚 Dangila spp. Thynnichthys 7 Thynnichthys thynnoi ゐ， yes yes

8 Lo bocheilus melanotaenia yes 00 ，possibly mixed wi 曲 other spp.?

9 Garra fasciacauda yes 00 ，small size -ov 町 Iooked? 10 10 Barbodes altus yes yes

11 11 Gyrinocheilus pennocki* yes 00

12 12 Sikukia gudgeri* yes ，mcon 事ctly 俗 S. stejnegeri 00

13 13 Puntioplites falcifer yes ，incorrectly ぉ P. proctozysron yes ，incon 切 tly 箇 P. proctozysron

Cyclocheilichthys 14 Cyclocheilichthys enaplos yes ye 唱

Cyclocheilichthys 15 Cyclocheilichthys sp .l spp. 00 ye 民俗 C. apogo n/ spp.

16 16 Cosmochilus harmandi* 00 ye 唱

17 17 Epalzeorhychus frenatum yes ，泊 correctly 鎚 E.frenatus 00

18 18 Crossocheilus reticulatus yes 00 ，small size -overlooked?

19 19 Crossocheilus siamensis yes ，鎚 Epalzeorhychus siamensis 00 ，small size -overl ∞ ked? 20 20 Osteochilus melanopleurus 目。 yes ，iocorrectly ぉ O. melanapleura 21 21 Osteochilus microcephalus yes 目。

22 22 Amblyrhychichthys truncatus 00 yes

23 23 Lu ciosoma bleekeri yes 00

24 24 Leptobarbus hoeveni 00 ，large size ，oot caught 泊 tone yes

25 25 Rasbora sp.* 00 目。 26 26 Tenualosa thibaudeaui yω yes

27 27 Schistura sp. 00 00 ，small size -overlooklω? 28 28 Acantopsis sp. yes 目。 29 29 Botia madesta yes yes ，鎚 Botia spp. 30 30 Botia helodes yes yes ， Ii sted as Botia spp.

31 31 Botia caudipunctata 00 yes? ，as Botia spp. 32 32 Parambassis wo !ffi 目。 yes ...... Origins of migrations uoclear RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRA TIONS 23

Figures Figures 11 t 'O 15 c'O mp 釘 e daily t 'O tal fish catches f'O r each ye 釘 against lunar phases (the (the t'O ps 'O f the darkened triangles represent full m 'O'O n peri 'O ds while the b'O tt 'O ms represent new m 'O'O n peri 'O ds). Figure 16 includes summaries 'O f daily t 'O tal catches f'O r all five seas 'O ns. It It is evident 血at fish catches 紅 'e cl 'O sely ass 'O ciated with lunar cycles. Peak catches always 'O ccur ar 'O und the new m 'O'O n，when nights 創泡 the darkes t. τ'h e main peak catches f'O r each ye 釘'O ccu 汀 'edd 町 ing a new moon phase in late January 'O r early t 'O mid-February ，f'O ll 'O wed by a sec 'O nd smaller peak alm 'O st exac t1 y 'O ne lunar m 'O nth later ， again ar 'O und the time 'O f the the new m 'O'O n. H 'O wever ，泊 1997 there w 槌 a deviati 'O n 合om 血is pattern in 血at a small peak peak in catches 'O ccurred ar 'O und the time 'O f the full m 'O'O n，ab 'O ut half a lunar m 'O nth after 出 e frrst main peak in catches. It is unclear why 血is 'O ccurred ， but the fish may have been delayed delayed f'O rs 'O me reas 'O n，'O r they may have l 'O st a number 'O f days 甘ying t 'O ascend 'O ther channels channels in the Kh'O ne Falls area bef 'O re reaching N 'O k Kas 'O um Island. An 'O ther small peak 'O ccuπed d町加g 白e new m 'O'O na m 'O nth after the frrst main peak ， half am 'O nth after the unexpected unexpected small peak ar 'O und the full m 'O'O n. In 1999 ，'O nly a frrst peak in catches w 槌 rec 'O rded. Th e sec 'O nd smaller peak was probably n'O t rec 'O rded because water levels dr 'O pped m 'O re rapidly 泊 1999 血佃泊 previ 'O us ye 紅 s. By the time a fulllunar m 'O nth had p槌 sed since since the main peak in catches ， water levels had already dr 'O pped t'O'O much f'O r tone fishing t 'O take place. In 'O ther w 'O rds ，fish may have been migrating at the time ， but the tone trap being being m 'O nit 'O ring at N 'O k Kas 'O um Island c'O uld n'O t catch them. Figure Figure 17 depicts m 'O nthly t 'Ota1 fish catches f'O r the bag-net fishery 泊血eT 'O nle Sap River River in cen 甘al Camb 'O dia f'O r each 'O f 白e five years between 1995 and 1999. It als 'O includes includes the m 'O nthly t 'O tal catches f'O r the fence-filter 佐'ap at N 'O k Kas 'O um Island 'O ver 血e S釘 ne five-year peri 'O d. 官le weights f'O rb 'O th fisheries 狙'e n'O t direc t1 yc 'O mparable ，h'O wever ， because because the bag-net fishery catches 紅 e 関心'O rded 泊 me 住ic t'O ns ，and have been ex 凶 p 'O lated t 'O represent the t 'Ota1 catch f'O r all bag-net 'O perati 'O ns (LIENG ET AL. ，1995). 百 le fence-filter trap trap catches 紅 e rec 'O rded in grams and are 'O nly available f'O ras 泊gle 位ap. Due t 'O the wide r組 ge 'O f efficiency f'O r different traps ，n 'O attempt has been made t 'O extrap 'O late results f'O r the the wh 'O le fishery. F 'O r example ，s'O me small traps 泊 min 'O r channels catch less 血an 100 kg a seas 'O n， while the 位aps at N 'O k Kas 'O um Island gen 釘 ally catch 'O ver 3，000 kg per seas 'O n. Even if a large sample 'O f traps in different channels c'O uld have been m 'O nit 'O red with with a large am 'O unt 'O f lab 'O r ， it still w 'O uld be di 妊icult t 'O ex 佐ap 'O late the results because even even traps in 出e same channels 'O ften have quite di 鉦'erent catches. Peak bag-net fish catches generally 'O ccur ab 'O ut 白ree-quarters 'O fa lunar m 'O nth bef 'O re the the peak fence-filter trap se 槌'O ns. It appe 釘 s that ye 釘 s 'O f particularly high bag-net catches ， such such as 1995 and 1998 (1 8，410 and 14 ， 671 t'O ns ，respectively (VAN ZAL 町 GE ET AL. ， 1999; LIENG ET AL. ，1995) ，釘 'e ass 'O ciated with l'O w t 'O m 'O derate fence-filter trap catches (3.5 and 1. 8t 'O ns ，respectively). In 1996 佃 d 1997 when tone catches were highest (4.2 and 3.9 t'O ns) ， dai catches were m 'O derate (1 4，4 29 and 15 ，4 88 t'O ns) (VAN Z 札別GE ET AL. ， 1999). In In 1999 ， dai catches were 恥 l'O west in five years (8 ，894 t'O ns) (VAN Z 札町GEET AL. ，1999) ， and while tone catches were als 'O l'O w ，出 ey were nevertheless higher th 加泊1998 (2.4 t'O ns in in 1999 c'O mpared t 'O 1. 8ωns 泊 1998). Th e ab 'O ve data 加古 n'O tstr 'O ng en 'O ugh t 'O make any defmitive defmitive c'O nclusi 'O ns ab 'O ut the relati 'O nship between the dai fishery and 血e tone fishery. H 'O wever ， the p 'O ssible implicati 'O n is 白紙出.esu ∞ess 'O f the tone fishery is str 'O ngly influenced by h'O w m 組 y fish are caught in 出e dai fishery. The 1993-1997 water discharges in the mainstream Mek 'O ng River at Pakse (appr 'O ximately 130 km upriver fr 'O m the Kh'O ne Falls; Fig. 18) indicate 血at peak small 24 IAN G . B A IR D ，恥1A RK S. FL A H E RTY AN D BOUNP H E N G P HYLAVA N H

15 匝刃 .OCIJ 2963 Lu na r I! end ・TolalCa tch W 珂 hl 日:xJ.0CIJ "hm-u m )

』傷 4∞.田沼吾、 Cコ a J E 200. 慌泊

nυ nu

国間・ 川両 的 明 出伊邑〈口円凶品別』 Z 国 可 品出口山品。 mmwcEfmmc出 山田 mmaω規制 同 mmvh明白局主 mm wmaUωmwuωmuω mc 町宮 whm- mzad 山 e 海草 ・己《・凶 ωω C amw aA he伺主 喝主的内 a 旦 伺「 伺「国内 也弘』国- 主。円 〈 凸 口・∞凸岬戸 口 fm hhehh hhA AF司 ι - ー L ー h 一 刊刊1 品別 円 mp N m 一伊 叫戸 口問 円内 v 内

Pea l< Ofda r同 ned Ina ngleS ::; fu Umoon

Figur e 11. Tota l 1995 fish catc h es fo r th 巴 fe 口氏 一fil t巴r trap (t on e) fishe ry at Nok Ka soum Island ，K hong Di stri ct， C hamp as ak Provi nce，L ao PDR agai nst lun ar phas es

15 15 一一ー一一一一一一一ーーー一一一一一一ー一一一一一一一一一一一 Lunar I一一一ーren d T OOJ 庄町 12 12 .T0 161 CalchWeigh l ~ 600. 1)羽 の nv hzu 田 ) hmE4 a∞.ぼ旧吾、 r む ω 」 ;;: E 沼田:XJ 3

nu

mquωmmuυωmmum間剛山 市町田 国 凶餌と畑町づ門戸回目 回目。 由 回目 曲目 田市山 岨 由自活主品曲師』帽凶肋昆豆内向曲目』曲目・岨 曲目昆〈品戸曲目己《・持 m-C m m m ・ 山元 U@ φ ecamc 2ωsaLhS aω ha邑〈・NF 曲口 与の m ω eEeR 口・凸品 口出戸口出向 『 「 弘，何 比凶 L E E 〈凶 ・ 一明戸ー i ・由 emN ・一前 mN 咽何 m・ F F

Peak Peak 01 darkened Iriangles = {ull moon

Figu re 12.Tota l 1996 fish ca tch es f or the fence-fi ltert r ap (t one) fishe ry at Nok Ka soum lsland，K hong Di str ict， Cham pas ak Prov in ce，L ao PDR agai nstl una r ph as巴S

15 15 国:XJ αm Luna r lI end 12 12 • Tota! Catch Welg 刷 、 区沼周珂 h GJpb { 伊 曹司繍 hmWC a∞.聞吾、 コ 由 」 主 2 ∞.OCIJ

nU nu 田町 8 姐山 国 唱曲ω@ 九町，h 九回 hm hmA-wkm-hm-hm hmahmwthmhm h hmh hm hawhm mZ 目と帽勺向戸 肋，局室内村 目 aU@白・ U ce aCEと旬、 宮 一宮弘aωL， 渇室町 ぉ 品 己〈・由・ 己《円戸・ 己呼内向- 』門崎〈・ ua伺口 口・内向口 『， 比内 E ER ・ ，・-罰戸 hhmw 今内 e E 前 ・ e ・明 i由 山戸 RE 潟 岬- 由 h N 円 F

Peak Peak o(oa 山ene dI ria n gles ;:: fu Jl打 、00 0

Figu re 13.To tal 1997 fi sh ca tch es fo r th e fence-fi lt er 廿ap (I one) fi she ry at Nok Ka soum Island，K hong Di strict， C hamp asak Prov ince ，Lao PDR aga in st lun ar ph ases RHYTHMS OF THE RIVER : LUNAR PHA SES AND MIGRA Tl ONS 25

I} 800 民10 L凶、 ultend 12 12 .Tol a.l C 山 hW 創酔 t 刷叫岨地間醐脚

・he96 官 V 旬 司戸 、ecg 益田 ω J 註

。 日 E12EZ bh EZ EZ 宮 宮 "。EdziR宮号弘円"。宮ふよ宝 宮号比門何宮 ESEa宮，ョ宮 E-EAHEE 自 宜白置・ ， U ER-P cd， 8 HdEA， -MdEAN ・昆〈門-昆 己 幽凸・ sz ・ 口 凸 口問 白角 h E 〈 〈 〈お ， 「白。-向。 旬 4hH 2・ mv -R

eak Peak o(da rk e ned 1，la l¥ g l&<; ::; full moon

Figur e 14. Total 1 998 fi sh ca tch es fo r th e fence -filt er trap (ton e) fi she ry at Nok Ka so um Island ，K hong District ， Champa s ak Province ，Lao PDR aga in st lun ar pha ses

回 3.137 15 日Xl.度目

Lun artU~lid 12 Tolal C剖chWelgh l • 日刃 .成田 瓜whハヲ 。司』 '" mmC 4∞脚色 pb E- ~ - E Z 泊四刃

nu nu 田由 田 館。国酌・ 田町 忠告8 m 明白定づ凶内申町内副忠告主FmbawhhgAmm酢 酢酌』幡圭mm酢町』伺主，口問冊目 明白』身内-m 肋問 町 品と 酌』偲芸品 m ωω口auω ωωIω C4・ 』伺主・円 』何主 49q ヰ《。』 ω ω 4 ωHh £円円 巳《 口田 口・山 口町内 口町刊 3 ・ ・ i ・曲 2 2 町内 申 由- -R R hN 戸 F

Peak Peak 0' da ，t. ..'e ned In angles::; fufl moon

Figu re 15. Tot al 1999 fi sh catche s for th 巴 fence-f ilt er trap (ton e) fish巴ry at Nok K aso um !sland ，Khon g D istrict ， Champa s ak Pr ovinc e，Lao PDR aga in st lun ar pha s巴s

RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 27

Bag .n elft s hery

1200000 掴聞聞DaiWelgh {ωE日目日日日日nunununU。。日日日日日目nununU ， 百 ，.，e hIl--1 Eub ・1 コ h ・ U}ωpmwzuω-E4 ell 、 1 .11 0.i!i ， ・'・・ U内 a---1 ーl l ・ 口 i .2 l---11 ，， 20 日日日日 、 2 日日日 0 ，1 nU

Fence Fence f，lter Ir ap ftshery

1 ，2日目。， 000 ， 骨骨 ・ . 一、 ~，日日日日 00 -・・・ Tone Welghl 。ischarge 3 ，500 ，目。。 1，000 .0 00 ， ω 、日・nunununununNnuNnu日日日目nu ω J弓 』 '" {回 E 800 ，000 .'" " 50 内 )E925Z ， 4q Unu u ". 五 . 門 ，" ‘，，Ununυ B 600 ，0日o . ・ ・ ・ .， .， 封筒 0> 5 門 nMnuの 」 E unuunu ." " ， ・ ~ ~OO ，O∞ ; ." υ U .，nN D il l ' 戸 δ" 2日0.000 1 1& " ， ‘ 、 50 日.日 00 、・~ "，. 『可.' 。

1，200 ，0目。 ~ .0 00.0 日日

OaiWelghl OaiWelghl I = ~Dl 側scharge 3 ，500 ，0日日 1，000 .0 00 2 -‘、 .・ " 3 ，000 ，000 " ::・ .ー‘ .ー‘ 自00 .0 00 ' : 1齢、 ，司 、，‘E3nupb口口nununuDOnununUnunu : ' ， ， . 豆町一 旬間 .、 . 』 ω OZu--FUDnU nυ'nunu ， ， 、 主吾首 4 - ・白 ー， ‘' U

D nN内 : ( ， 2・ u ， ' 4日日 00 日

' φ- 、. 1.00 日日日日

， ‘ 2日0 .0日日 . -‘ ‘ 500 .0 00 ‘. ‘， ， 。、-- '

Figur e 17.Monthly total fish catches for the bag-net fis h巴ry in the Tonle Sap R ive r in Ce ntr al Cambod ia an d th e fence- fiI ter tra p fi she ry at th e Khone Fa ll s in Southern Laos between 1995- 1999

RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 29

ROBERTS & BAIRD (1 995) recorded 27 species of fish as being caught regularly 泊出e tone tone fishery 加出e Khone Falls. We believe that 32 fish species migrate from the Tonle Sap River to 白e Kh one Falls each dry season (Table 3). Species listings in ROBERTS &

BAIRD (1 995) and LIENG ET AL. (1 995) 紅 'e also included. Fish species added to the list of long long distance migrators since ROBERTS & BAIRD (1 995) include: (1) Botia caudipunctata (a (a newly described cobitid loach); (2) Schistura sp. (a mysterious smallloach 出at shows up in peak tone catche トーbut due to 出e fact 血at Schisturas 紅 'e not k:n own to migrate in other other places ，this species is only tentatively included); (3) Leptobarbus hoeveni (a carp too large large to be caught in tone fisheries but believed by villagers to migrate 企om the Gr eat Lake in in Cambodia); (4) the minnow Rasbora sp. (a species found 泊 peak tone catches and of mysterious mysterious origin); 組 d (5) Osteochilus melanopleurus (a carp believed by villagers to be highly highly migratory). Future research should be dir ，田 ted towards confmning the migration pattems pattems of these species (Table 3). 百le Great Lake and other large wetlands in Central Cambodia shrink vastly in size du 血19 伽句season (v AN Z 札削GE ET AL. ， 1998; RAlN BO 百， 1996; LIENG ET AL. ， 1995). 百lI s necessitates many of 血e fish inhabiting these 釘 'eas during the monsoon se 踊 on to look look for other places to survive during the harsh low-water season. Peak dai catch periods 泊 the Tonle Sap River occur between one and six days before the full moon (LIENG ET AL. 1995; 1995; N. van Zalinge ， pers. comm. ，1999) ， while peak tone catches at 白e Kh one Falls occur occur approximately at 白e new moon ， or slightly before or 泊 .er it. To determine 血e approximate swimming speed of the small cyprinids ，we have assumed that that peak dai fishing periods occurred each month fo 町 days before the full moon. We have counted counted the days between those peak catches and peak catch days for the tone fishery at 血e Kh one Falls ， based on our daily catch data from Nok Kasoum Island. It took between 15 阻 d 25 days for the small cyprinids to migrate between the dai fishery and 血e tone fishery. 官 le mean 住aveling time w 出 19.8 days (SD = 2.99). Nine lunar month periods over over five years were used to estimate 血e 位aveling time for peak runs of small cyprinids. Th at translates to average swimming speeds of between 16 佃 d 27 km per day ， with a mean of of about 20 km. 官 le Project for Management of the Freshwater Capture Fisheries of Cambodia conducted a ye 釘・long sampling program at six landing sites along the Mekong River River in Cambodia and found a swimming speed for the small cyprinids of 17 km/ day between between the ぬi traps and Stung Treng Pr ovince (N. van Zalinge ，pers. comm. ， 1999). Experienced Experienced La o fishers from the Kh one Falls area believe that tlle main peak fish 泊g period period for the tone fishery occ 町 's each year around the 3 rd day of the rising moon during 出e 3 rd Lao lunar month ，which coincides with Chinese New Year (S 町 GHANOUVONG ET AL. ， 1996; ROBERTS & BAIRD ， 1995; ROBERTS & WAR 臨 N ， 1994; ROBERTS ， 1993). Data collected collected at Nok Kasoum Island indicate 白紙 villagers are correct 泊 assuming 出is to be tlle tlle case ， although tlle time of peak catches can v紅 y by a few days either way (Figs. 11 to to 15). In In 1996 出 e frrst main peak tone catches were recorded during the new moon of the fourth fourth Lao lunar montll ，rather 出an the third. An other smaller peak occurred d町泊g the new moon at the beginning of the fifth lunar month during the s創 ne ye 也官lese data suggest suggest 白紙 the fish arrlved at 出.e Kh one Falls one lunar month late. However ， after comp 釘 ing tlle peak dai fishing periods with the peak tone fishing periods ，it became apparent apparent tllat the fish were not a month late 油 .er al l. On around January 1，1996 (four days before before the full moon) ， large dai catches were recorded (these catches were only slightly 30 IAN G. B AlR D. MARK S. FLAHERTY AND BOUNPHENG PHYLAVANH

間中assed the oext lunar mooth-5 ，ω4 compared to 5，282 toos) (N. v佃 Zalioge ，pers. comm. ， 1999). However ， the frrst peak period for the tone fishery w 部 oot uotil around February February 23 ，1996 ，54 days a負.er the frrst dai peak. Coosidering what we already know about about swimming speed ，ぬ is much looger than the average 20-day time period appareotly requII 泡d for the fish to reach the Kh ooe Falls 企om the Toole Sap River. However ， if a 29.5-day 29.5-day luoar mooth is subtracted 企om 血.e 54 days ， an approximately 25 day 回 veling tIIO e remaios ，which sounds much more plausible. 百lerefore ，o町 cooclusioo is 白紙 a run of of fish probably did arrive at 血e Kh ooe Falls at the eod of January ，1996 ， but because wa 回 levels dropped particularly slowly 白紙 ye 紅，出.e water was too high forω ne fishing to to occur ，組d 00 fish were caugh t.百 le fish probably passed the Kh ooe Falls relatively unhnpeded ，and 由is may explain why 1996 was coosidered a good ye 訂 for small cyprinid fishing fishing upriver from 出.e Kh ooe Falls (BAIRD ET AL. ， 1998). Particularly high water levels 泊 1996 may also explain why it took 25 days for the fish to reach the Kh ooe Falls ， compared to 血e 20 days 00 average that the joumey takes. It is possible 白紙偽esw 泊四1泊g speed speed of 血e fish was reduced because the fish had to swim against strooger curre 附 than are are oormal. Cooclusioos Cooclusioos based 00 our es tIIO ate for the swimmiog speed of the small cypri 凶ds should should be viewed with some cautioo. To begin wi 血， the assumptioo that peak dai catches took took place fo 町 days before the full mooo may oot hold 泊 all years. Moreover ，it is difficult to to pin-po 泊t clear pe 紘 s 泊白e tone catch data. Some tIIO es fish reach the Kh ooe Falls and 凶 vel up ooe or more impassible channels before turr 血 g back and finally 釘ying to go up 出e widest channelleading to Nok Kasoum Island (Fig. 2) (R OBERTS & BAIRD ， 1995). Nok Kasoum Island is ideally sited 泊伽largest channel 泊血e Kh ooe Falls area ，just upriver from Hou Sahoog ， the main migratioo path for fishes moving above the Kh ooe Falls (BAIRD ET AL. ， 2001a; BAIRD ， 1996; ROB 臨時&BAIRD ， 1995) (Fig. 2). Th erefore ， there is is less risk of theω ne fishery at Nok Kasoum Island 凶 ss 泊gm 吋or migratioos of small cyprinids cyprinids comp 釘ed to other tone fisheries io 白e Kh ooe Falls 白紙釘e si 似 ated in smaller and less ceotrally located channels. Nevertheless ，a few days may be lost io some mooths

by fish 血at unsuccessfully 町 to asceod impassible waterfalls. Th ey have to double 幽 back before before moving up 泊to 血.e Hou Sahoog channel and Nok Kasoum Island area (see ROBERTS & BAIRD ，1995 ， for a detailed explanatioo of the channels of the Kh ooe Falls). An other factor 血at influeoces wheo migrating small cyprinids 釘 'e caught 加 tone 凶 ps is is the lapse of tIIO e betweeo wheo they pass 白e trap moviog ups 釘'eam ，and wheo 白ey retum retum after eocountering an hnpassible obstacle. 百ley geoeral 取movep 部 tthe 回 ps without being being caught io significant oumbers ， and theo hours later 出ey are caught 泊 l紅 ge oumbers 部出eyre 旬m. Th ese hours can ，however ，sep 釘 ate days with a night and influeoce sw IInmin g speed speed calculatioos. Coosidering all the above ，it is probable 血at sw 泊四由19 speeds will exhibit exhibit coosiderable variatioo around the me 加. An appareotly small ruo of small cyprinids passes 出.e Kh ooe Falls arouod the beginning of 血.e 2nd Lao luoar mooth each ye 低官le tone fishery at Nok Kasoum Island does oot target target these migrators because water levels ar 芭 too high at 白紙 time of year. However ，we have have recorded small ruos of Henicorhynchus lobatus caught 泊 cylin 合 ical cu 町eot 住aps (ch 似 oe 紅 B 佃 Hang Kh ooe 泊出e Kh ooe Falls 紅白around the beg 恒国og of the 2nd La o lunar lunar mooth during various years. 百lese migrators are called "pa soi ho ω pi" 泊 Lao ， which translates 部“出 e first Henicorhynchus of the year" (ROBERTS & BAIRD ， 1995). Th ere is app 間 otly at least a weak link betweeo relatively large dai catches and low RH Y1百 MS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 31

tone tone catches ，alth 'O ugh '0町 data are far fr 'O m c 'O nclusive. One 'O f 出 ep 'O ssible implicati 'O ns 'O f these results 紅 e 白紙 neither dai n'O r tone fish catches al 'O ne are ag 'O'O d indicat 'O r 'O f the size size 'O f migrating small cyprinid p'O pulati 'O ns in the Mek 'O ng basin during particular ye 但"8. In stead it may be necess 紅 Y t 'O m 'O nit 'O rm 'O re 由加'O ne fishery 泊'O rder t 'O estimate p'O pulati 'O n sizes. sizes. It is p 'O ssible that 加 s'O me ye 紅's large numbers 'O f small cyprinids are able t 'O escape cap 伽re in the T 'O nle Sap River dai fishery due t'O hydr 'O l 'O gical 'O r 'O ther reas 'O ns. H 'O wever ， small small cyprinids als 'O certainly migrate 'O ut 'O f 'O ther wetlands in central Camb 'O dia. Yet it is still still unclear h'O w much these fish c 'O ntribute t 'O small cyprinid migrati 'O ns in 白 e Mek 'O ng basin. basin. F 'O rn 'O w we d 'O n'O t have en 'O ugh inf 'O rmati 'O nt 'O answer these questi 'O ns ，which have imp 'O民組 t management implicati 'O ns. Whil e it is quite clear 血 at small cyprinids arrive at the Kh'O ne Falls each year at 出 e beginning beginning 'O f the rising m 'O'O n during the 3 rd La'O lunar m 'O nth ， as villagers claim ，it is n'O t clear clear whether individual species m 'O ve at different speeds. Nevertheless ，s 'O far we have n 'O data data t 'O indicate 白紙出ey d 'O. In fact ，it is c'O mm 'O nt 'O see large quantities 'O f many species in in catches during peak tone fishing peri 'O ds. It is als 'O unclear whether the small cyprinids travel travel at different speeds thr 'O ugh 'O ut 血 eir l'O ng j 'O urneys. Th ey apparently 'O nly migrate d町泊g daylight h'O urs ， as small cyprinids 訂 'e never caught in tone traps in large quantities during during the nigh t. Th eir daytime swimming pattems ，h 'O wever ， are n'O t well underst 'O'O d. Henicorhynchus Henicorhynchus lobatus appear t 'O migrate in lines ne 紅 the edge 'O f channel while Paralaubuca 砂'P us frequently travel cl 'O ser t'O血 e surface and 泊 deeper water (pers 'O nal 'O bservati 'O ns. ，1. G. B 剖rd). On e questi 'O nc 'O mes t 'O mind. Is it by design 白紙 masses 'O f small cyprinids pass the Kh'O ne Falls during the darkest p紅 t 'O f 血 em 'O n血，加d is 白 is why they time their dep 佐知re fr 'O m the Great Lake as 血 ey d 'O? On the 'O ne hand ，it is plausible 白紙 it is c'O incidental 白紙 the the small cyprinids reach Kh'O ne Falls during new m 'O'O n peri 'O ds ，and 由 at the time 血 ey 倒 rive is simply a functi 'O n 'O f the number 'O f days it takes f'O r migrating masses 'O f fish t 'O swim 合om the T 'O nle Sap River 'O r 'O ther wetlands up t'O出 e Kh'O ne Falls. Nevertheless ， the Kh'O ne Falls represents the 'O nly maj 'O r ‘ b'O ttleneck' f'O r the migrat 泊， g fish 'O ver 血 ec 'O urse 'O f血 eir l'O ng j'O urney (it is 恥'O nly maj 'O r waterfalls 'O n 恥Lo wer Mek 'O ng Ri ver) 侭 A 町 BOTH ， 1996). 1996). It s田 ms p'O ssible 白紙 it c'O uld be advantage 'O us f'O r the fish t 'O pass the Kh'O ne Falls during during dark peri 'O ds ，泊'O rder t'O reduce the risk 'O f predati 'O n during 血 e nights when 出 ey 創'e n'O t migrating and re 釘eat 加t'O small refuges ar 'O und the Falls. Duri ng peak migrati 'O n

peri 'O也， the small cyprinids ar 芭 s'O metimes f'O rced t 'O c 'O ngregate in large numbers 泊 relatively small small shall 'O w p 'O'O ls in the Kh 'O ne Falls 創芭a where they c'O uld be easy prey if night visibility visibility w 回 g'O'O d at 白 at time ， which it is n'Ot.官邸 c'O uld be the ev 'O luti 'O nary reas 'O n why they they have c'O me t 'O pass the Kh'O ne Falls when they d 'O. Thi s sc 岨 ari 'O makes even m 'O re sense when the ge 'O l 'O gical hist 'O ry 'O f the Mek 'O ng basin is is c 'O nsidered. Th e Great Lake was '0凶 y created during 白 em 'O st recent subsidence event 'O f 恥 Camb 'O dian pla ぜorm between 5，590 組 d5 ，850 years ag 'O侭 A 町 BOTH ，1996) ，and d町加 g relatively recent hist 'O ry ，much 'O f 血 el 'O wer Mek 'O ng basin in s'O uthem Camb 'O dia and and Vie 旬 am was under the sea. 百leref 'O re ，it is likely that the small cyprinids had a much g白紙er peri 'O d 'O f time t 'O ev 'O lve their migrati 'O n timing f'O r passing the Kh'O ne Falls th 佃 f'O r adap 曲 19 t 'O c'O nditi 'O ns in 血， e Great Lak e 'O r 'O ther parts 'O fs 'O uthem and central Camb 'O dia and s'O uthem Vietnam. Furtherm 'O re ，bef 'O re the Great Lake and 'O ther seas 'O nally inundated wetlands wetlands in s'O uthem and central Camb 'O dia and s'O uthem Vietnam were f'O rmed ， it is p'O ssible 白紙 p'O pulati 'O ns 'O f small cyprinids were c 'O nsiderably smaller 出組曲ey 紅 en 'O w 32 lA N G. BAIRD ，MARK S. FLAHERTY AND BOUNPHENG PHYLAVANH

due to not having such vast wetlands available to reproduce and nurse. It is also 町田 that when the small cyprinids reach the Kh one Falls they are generally split up into smaller groups groups that try to ascend different channels. Th erefore ，their collective s回 tegy of reducing 出 e potential impact of predation by traveling in large groups is compromised more at 血e Kh one Falls than anywhere else along their migration route. All of the above helps to explain explain why it would be advantageous for the small cyprinids to pass the Kh one Falls at the the darkest time of the lunar month. Th ere is still another reasonable explanation of why the small cyprinids time their migrations migrations as 血ey do. At first ，it is not easy to recognize why they pass the Tonle Sap River River from the Great Lake one to six days before the full moon. Wh y would they want to move a number of days before the full moon? However ， if the recent geological history is is again considered in a different light ，a possible explanation becomes app 紅 'ent. Whi le the Great Great Lake and Tonle Sap Rivers are no longer directly affectβd by the tides ， the spawning and nursing grounds of the small cyprinids were possibly signific 叩 tly affected by tidal influences influences before the Great Lake was formed ， and during the period when much of southern Cambodia and Vietnam were under the sea. The largest tides of the month ，called spring tides ，generally occur shortly after full moons (Wayne Gum and Bernard O'Callaghan ， pers. pers. comm. ，1999).τ 'h erefore ，salt-water intrusion into the Mekong River is likely to be greatest greatest during that period ，回pecially at the end of the rainy season when Mekong River downstream discharge declines. Assuming 白紙 the small cyprinids evolved in 合'eshwater environments environments and would want to avoid particularly saline tidal influenced salt-water intrusions intrusions from the sea ，it makes sense that they would have evolved to leave the wetlands before before the high tides and during low neap tide periods ，which occur before the full moon. An other factor 血at adds suppo 託 to the second hypothesis is 出at migrations of other medium-sized medium-sized cyprinids 白紙 annually migrate 合om 血e Sekong ， Sesan and Srepok basins in in Cambodia up to the Khone Falls at around the same time of year (including

Scaphognathops Scaphognathops bandanensis ，Mekongina erythrospila ，Labeo erythropterus ，etc.) 紅 e not correlated correlated with lunar cycles (BAIRD & FLAHERTY ， 2001). If the Kh one Falls had been the major major influence on the lunar migratory patterns of the small cyprinids ，one might expect 出e same to be the case for other fishes migrat 泊g past the Khone Falls at about the same time time of year. In any case ，it appe 紅 s that the timing of the small cyprinid migrations can be evolutionarily explained by bo 血 factors related to conditions at 血e Kh one Falls ，and by historical tidal influences in lower p訂 ts of the basin. For now ，we are not able to suggest suggest which of the above two scenarios is likely to have influenced the evolution of migratory migratory patterns of the small cyprinids more. Perhaps a combination of both resulted in the the present circumstances. It It is likely 血at fishers catch many of the small cyprinids that pass the Tonle Sap River before before they reach Kh one Falls. Non-human predators ，including fish ， birds and mammals probably probably also feed heavily on 出 e cyprinids. Dry season cyprinid migrations 訂 'e not associated with with spawning behaviour (most small cyprinid species spawn during the monsoon season) (WARREN ET AL. ， 1998; SINGANOUVONG ET AL. ， 1996; ROBERTS & BAIRD ， 1995; ROBERTS & WARREN ， 1994; ROBERTS ，1993). 百le migrations appe 紅白be conducted by the small cyprinids cyprinids in order to use the available feeding areas along all the large lowland rivers of the the basin. Thi s enables them to feed on algae that grow as the Mekong becomes less turbulent turbulent at the beginning of the 句 season (see W. 組 REN ET AL. ， 1998; S町 GANOUVONG ET A L.， 1996; ROBERTS & BAIRD ， 1995). Some fish probably distribute themselves for RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRATIONS 33

feeding feeding pu 中 oses in lower sections of the Mekong ，below the Kh one Falls. We do not know whether whether the dis 凶bution pattern changes from ye 紅 to year ， depending on environmen ta1 or other other conditions. Many small cyprinids migrate up the Sekong ，Sesan 佃 d Srepok Rivers in in northeastern Cambodia ，出 us not reaching the Kh one Falls. Villagers from Kh ong Dis 肘ct have have long believed 白at the health of the cyprinid fishery at the Kh one Falls is 1釘 gely de 戸 ndent on what happens downriver in Cambodia. For example ，villagers claim that migr ョ.ting fish at 血e Kh one Falls were ex 回 ordinarily abundant between 1975 and 1978 ， when the Kh mer Rouge banned all commercial fishing operations and also severely lir 凶ted subsistence subsistence fishing (ROBERTS ， 1993). Cambodian fishers seemly have the advantage in terms of having the first crack at 出e small small cyprinids. In terms of other 1紅 ge-river m 姐 agement issues ， such as those related to hydroelectric hydroelectric dam construction and water pollution emissions ，however ，be 泊g downs 住.eam is is not necessarily advantageous. Wh at is needed is joint-cooperation between all countries in 血e region so 血at each coun 町 does not negatively impact on fisheries of other coun 凶 es. Th e goal should be to maximize benefits to the environment and ultimately local people. Cambodia needs to make sure 白紙 Laos and Th ailand get their fair sh 但芭 of migrating small cyprinids. cyprinids. On the other hand ， the Lao ， and the Th ais need to avoid impacting on hydrological cycles 白at influence natural processes downriver ， or extracting too much water ， or emitting too too much pollution into 出e river system. If all the coun 甘ies in the basin do not cooperate to to P問 vent and solve these critical problems ， everyone is likely to suffer. Sustainable fisheries fisheries based on highly migratory species depend on not over-harvesting fish throughout their their migration routes as well as maintaining riverine conditions suitable for the successful reproduction reproduction and survival of fish populations.

CONCLUSIONS

Several Several conclusions c佃 be made based on 出 e analysis of the tone fishery data from Nok Kasoum Island in the Kh one Falls area in southem Laos ， and the fish catch data from 出e dai fishery 泊 the Tonle Sap River in Cen tra1 Cambodia. Th e frrst is 白紙 the small cyprinid cyprinid migrations up 血e Kh one Falls and the fence-futer trap fishery are highly correlated with with lunar cycles ， with peak catches always occurring during the new moon ， followed by a second smaller peak in catches one lunar month later.τ 'h is makes sense considering 出at peak peak dai fishery catches 釘 .e already known to be associated with lunar phases (LIENG Ef AL. ，1995). 百lere is considerable evidence to show that fish caught 泊 the dai fishery are from the same populations as those targeted by the ω ne fishery (BAIRD ， 1998). τ'h e small cyprinid migrations from the Great Lake to the Mekong River 釘 e 凶 ggered by lunar cycles. However ，it is evident from catch data 企om 血e dai fishery 白紙 hydrological conditions conditions play a significant role in determining when fish migrate. For example ， in 1996 ， when water levels drop slowly ，significant amounts of fish were caught in early January ， but but many more were caught in February and March (N. van Zal 泊.ge ，pers. comm. ， 1999). In In contrast ， in 1999 when water levels dropped their quickest ， the bulk of the small cyprinids cyprinids had migrated past the Tonle Sap River by early January (N. van Zal inge ，pers. comm. ，1999). 官 lerefore ，it is reasonable to believe that the combined effects of lunar and hydrological hydrological cycles determine the nature of small cyprinid migrations from the Great Lake. Th e most significant findings of our investigations 紅 e that small cypri 凶ds ，which 釘 e 34 lA N G. BAIRD. MARK S. FLAHERTY AND BOUNPHENG PHYLAVANH of of great social and economic importance to local people in southem Laos and other parts of of the Mekong River basin ， pass at least 400 km ， and probably over 1，000 km ， between Cambodia ，Laos and Th ailand. Al though lunar phases greatly in fI uence their movements ， 叩 y significant alterations of the hydrological regime are likely to greatly impact their migration migration pattems. Moreover ，if even one of ten the large dams envisioned for the mainstream Mekong River between the Great Lake in Cambodia and the upper reaches of the Mekong River River in Laos and Th ailand is built (Mekong Secretariat ，1994) ， small cyprinid migrations could could be totally blocked. This could result in ex 仕emely serious impacts to fish populations ， the the overall ecology of the Mekong River basin ， and the millions of small-scale fishers and farmers farmers who depend on fisheries for subsistence and income. Because of the great importance of of the Hou Sahong channel for migrating fish (BAIRD EI AL. ， 2001a; ROBERTS ， 1997; BAIRO ， 1996; ROBERTS & BAIRD ，1995) ，it would be a grave mistake to 出 sume that d釘 nming 白紙 channel would have little impact on migrating fish.τoat is because other channels channels in 由e Khone Falls area are generally impassible by fish ，especially in the low- water water season (BAIRD ET AL. ， 2001a; BAIRO ， 1996; ROBERTS & BAIRO ， 1995). It It is critical that a regional approach to research and fisheries management be developed so so as to encourage the long-term maintenance of local fisheries throughout the lower Mekong basin. We need to think regionally while at the same time being sensitive to local conditions conditions along the Mekong River and her tributaries. Striking a suitable balance between regional regional objectives and local priorities is critical ， and remains one of the major future challenges challenges for the peoples of the Mekong River basin. However ，govemments have often made development decisions 白紙訂 'e deleterious to bo 血 fish populations and the people who depend on them (IRN ， 1999; Fisheries Depar 佃 lent ，Ratanakiri Province 叩 d Nl 下P Project ， 2000). It is our hope that this destructive pattem will not continue to be replicated in in the Mekong River basin. We firmly believe that sustainable fisheries ，which 釘 e irreplaceable irreplaceable resources ， must be the ultimate goal ， not the sacrifice of fisheries and livelilioods to to achieve short-term replaceable resources that do not adequately take the overall picture and and welfare of the region and its people into accoun t.

ACKNOWLEDGMENTS

Th anks to Mr. Phongsavath Kisouvannalath ， Mr. Vixay In thaphaisy ，民11". Bou 凶lO ng

Mounsouphom ， Mr. Peter Cunningham ， Mr. Kh amsouk Xaiyamanivong 組 d 民11". Bounthong Vongsenesouk for their con 仕ibutions in collecting and compiling fisheries data from Ban Hang Khone between 1995 and 1999. Mr. To Mounsouphom entered most of the data into the the computer database with the assistance of the Vientiane-based NGO ， CIDSE. Mr. Pierre Dubeau ，a former CUSO co-operant ， helped develop the Ban Hang Kh one fisheries database. Ms. Sarah Adams extracted data from the database to create the graphs and tables included in in this article Dr. Tyson R. Roberts provided valuable input regarding the data collection program at Ban Hang Kh one during the early stages of research design. He helped identify m 叩 Y of the fish species collected at Ban Hang Kh one. Mr. Wayne Gum and Mr. Bemard O'Callaghan O'Callaghan provided valuable insights regarding possible tidal in fI uences. Research between between 1995 and 1997 was conducted under the auspices of the Lao Community Fisheries and Dolphin Protection Pr oject (LCFDPP). Work completed in 1998 and 1999 was done through through the Environmental Protection and Community Development in Siphandone Wetland RHYTHMS OF THE RIVER: LUNAR PHASES AND MIGRA TIONS 35

Project (EPCDSWP) ，which was implemented by the NGO ，CESVI Cooperation and Development and funded by the European Commission. Mr. Ole Heggen 合'O m 血eGeography Dep 紅加lent of the University of Victoria ，Canada prepared 血e maps. 1組 G. Baird took all photos ，except for one 白紙 was taken by Peter Cunningham.τ 'h anks 紅 'e also extended to the Agriculture and Forestry Division of Champasak Pr ovince ，the Agriculture and Fores 町 Office of Khong District ，and the Center for Protected Areas and Watershed Management (CPAWM) at the Department of Forestry ，Vientiane ，for their support. Special thanks to Mr. Mr. Vilay A 出n 紘 hone ， Mr. Sounthone Athn 北 hone ， Mr. Nouphai Athn 紘 hone ，Mrs. Le n A 白n 紘 hone ， Mr. Say ， Mr. Khamsouk ，and Mr. Souan for allowing us to monitor their fence-filter fence-filter trap fish catches at the southem end of Nok Kasoum Island ，Kh ong District ， Champasak Pr ovince. Without their patience and understanding this research would not have been possible. Finally ，we would like to th 創Ik Mr. Nicolaas van Zalinge and Mr. Nao Th uok from the pr 吋ect for M 叩 agement of the Freshwater Capture Fisheries of Cambodia ，which is being implemented by the Cambodia' s Department of Fisheries and the Mekong River Commission (JI.偲 c) with funding suppo 此 from DANIDA. They graciously provided the five years of bag-net fisheries data from the Tonle Sap River used in this paper.

REFERENCES

ALTOBELLI ，A ， G. DACONTO ， E. FEO Ll， J. F. MAXW 乱 L ，c. MILE 凪 AND E. RUI. 1998. Environmental Assessment 。f Siphandone Wetland. Unpublished interim 陀 po 此 prep 紅 ed for the Environmental Pro tection and Com- munity munity De velopment in Siphandone Wetland Pro ject ，CESVI ，Pakse ，La o PDR ，43 pp.+ BAIRD ，1. G. 1996. Kh one Falls Fishers. Mekong Fish Catch and Culture. Mekong Fisheries Network Newsletter ， Mekong River Commission ，2(2): 1 -4. BAIRD ，1. G. 1998. pr 芭Ii minary Fishery Stock Assessment Results from Ban Hang Khone ，Khong District ， Champasak Pro vince ， Southern Lao PD R. Unpublished technical repo 目 prep 紅 ed for the Environmental

Pro tection 如 d Community Development in Siphandone Wetland P吋ect ，CESVI ，Pakse ，La o PDR ， 112 pp. pp. B Al RD ，1. G. 200 1. Aquatic biodiversity i目白e Siphandone wetlands. Pages 61-74 in G. Daconto (ed.) ， Siphandone

Wetlands. Wetlands. Environmental Protection and Communi ηDe 阿 lopment in Sip 加 ndone Wetland Project ，CESVI ，

Berg 創 no ，1飽ly. BAIRD ，1. G. ， V. INTHA 聞 AISY ， B. PHYLAIVANH ，AND P. KISOUVANNALAT. 1998. A Rapid Fisheries Survey in Khong District ，Champas 紘Pro vince ， Southern La o PDR. Unpublished technical report prepar 吋 for the Environ- mental 針。 tection and Comrnunity De velopment in Siphandone Wetl 釦 d 針。~ect ， CESVI ，Pakse ，La o PDR ， 31 pp. BAIRD ，1. G. ， V. 1NT HAPHAYSI ， P. KISOUVANNALAT ， B. PHYLAVANH ，AND B. MOUNSOUPHOM. 1999. Fishes 01 Southern Southern La os (In La o). Lao Comrnunity Fisheries and Dolphin Pr o臨 tion P吋ect ，Minis 町 of Agricul- 加問釦d Forestry ，Lao PDR ， 161 pp. BAIRD ，1. G. ， Z. H ∞AN ， B. PHYLA V ANH ，AND P. Moy 臼. 200la. A Communal Fishery for 白巴 Migratory Catfish Pangasius Pangasius macronema 泊 the Mekong River. Asian Fisheries Science ， 14 (2001): 25 -41. BAIRD ，1. G. ， B. PHYLAVANH ， B. VONGSENESOUK ，AND K. XAIYAM ANlV ONG. 200lb. 百le Ec ology and Conservation of of the Smallscale Croaker Boesemania microlepis (Bleeker 1858-59) 白血e Mains 回 am Mekong River ， Southern Southern La os. Nat. Bull. Hist. Siam Soc. ， 49: 161-176. BAIRD ，1. G. AND M.S. FLAHERTY 200 1. Mekong River Artisinal Fisheries: G i1l Netting for Medium Sized Migratory Migratory Carps Below the Khone Falls in Southern Laos. Technical report prepared the for Environmen-

tal tal Pr otection and Comrnunity Development in Siph 佃 done Wetland Pr oject ，CESVI ，Pakse ，La o PD R. BARDACH ，J. E. 1959. Report on Fisheries 泊 Cambodia. USOM ，Cambodia ，Phn om Penh ， 55 pp. CLA 悶悶E， G. (compiler). 1996. An Inventory 01 Wetlands 01 the La o PDR. Wetlands Pr ogram ，IUCN - Th e World Conservation Union ，Vientiane ，Lao PDR. 36 lA N O. BAIRD ，MARK S. FLAHERTY AND BOUNPHENG I'H YLAVANH

CLA 悶悶E， O. F. ， T. SORANGKHOUN ，AND 1. O. BAIRD. 1997. Community Fisheries in La o PDR: A Survey of Techniques Techniques and Issues. IUCN ー百le World Conservation Union ，Vientiane ，Lao PD R. CRABτ "REE， R. 1995. Relationship between lunar phase and spawr 血 g activity of Tarpon ，Megalops Atlanticus ， wi 白 notes 0目白.e dis 住ibution of larvae. Bul. Mar. Sci. 56 (3): 895-899. DA' ∞悶0 ，0.2 凹 1. Siphandone Wetlands. Environmental Protection and Communi 砂Development in Siphandone Wetlands. Wetlands. CESVI ，Bergamo ，ltaly ， 192 pp. DIEP ，L. ， S. LY ，AND N. P. VAN ZAL 別 GE (eds ふ 1998. Catch Statistics of the Cambodian Freshwa 包r Fisheries. MRC /D oF /D anida Pro ject for the Managem 巴nt of the Freshwa 町 Capture Fisheries of Cambodia. Mekong River River Commission ，Phnom Penh ， Cambodia. Fisheries Fisheries Office ，Ratanakiri Province ， and N 宵 P Project 2000. A Study of the Downstream li 叩 acts l!打 'he Yali Falls Falls Dam in the Se San River Basin in Ratanakiri Province ，Northeast Cambodia. Ban Lung ，Ratanakiri Province ，Cambodia ， 66 pp. temational 1o temational Rivers Network. 1999. Power Struggle; ・The Impacts of Hydro-Development in La os. 10 町 national Rivers Rivers Network ，Berkeley ，CA ，USA ， 68 pp. JENSEN ，J. 1996. 1，000 ，000 tonnes of fish from the Mekong? Mekong Fish Catch and Culture ，Mekong Fisheries Network Newsletter. Mekong River Commission 2 (1): 1， 12. J限L1 NG ， H. AND T. WOOLDRIDGE. 1992. Lunar influence on distribution of a calanoid copepod 泊 the water column column of a shallow ，tempera 旬 es 旬ary. Marine Biology 112: 309- 312. Jo 臥 N間活， R. E. 198 1. Word of the La goon: Fishing and Marine La re in the Palau District Q川licronesia. University University of Califomia Pre ss ，Berkeley ，CA ，USA. LIENG ，S. ， C. Y 削， AND N. P. VAN Z札町GE. 1995. Fisheries of the Tonle Sap River Cambodia ，1: Th e Bagnet (Dai) Fishery. Fishery. Asian Fisheries Science 8: 258-265.

Mekong Secretariat 1994. Mekong Mainstream Run-of-River Hydropower. Bangkok ，百泊 l組 d， 20 pp. 0 町 BU 捌， S. 1990. E伽包of lunar periods and some 0 伽 parame 蜘 on fish catch in Lake K嗣~i ， Nigeria. Fisheries Fisheries Research. 8 (3): 233-245. PANTULU ， V. R. 1986. Th e Mekong River sy 臨 m. Pages 695 ー719 in B. R. Davies and K. F. Walker (eds.) ，The Ecology Ecology of River Syste. 附. W. Junk ，Do由民 ht ，百 e Ne 血erlands. RAINBOTH ，W. J. 1996. Field Guide to Fishes ofthe Cambodian Mekong. Food and Agriculture Organization of the the United Nations. Rome ，ltaly ， 1-265 + plates. ROBERTS ， T. R. 1993. Art isinal fisheries and fish ecology below the gr 伺 t waterfalls of the Mekong River in Southem La os. Nat. Bull. Hist. Siam Soc. 41 (1): 31 -6 2. ROB 限時， T. R. 1997. Systematic revision of 白巴回pical Asian labeoin cyprinid fish genus Cirrhinus ， with descrip 凶ons of new species and biological observations on C. lobatus. Nat. Bull. Hist. Siam Soc. 45: 171-203. 171-203. ROBERTS ， T. R. AND T. J. WARREN. 1994. Observations on fishes and fisheries in Southem Laos and Northeastem Cambodia ，Oc tober 1993 - February 1994. Nat. Hist. Bull. Siam Soc. 42: 87-115. ROBERTS ， T. R. AND 1. 0. BAIRD 1995. Traditional fisheries and fish ecology on the Mekong River at Kh one Waterfalls Waterfalls in Southem Laos. Nat. Bull. Hist. Siam Soc. 43: 219-262. ROBERTSON ，D. ， C. PETERSEN ，AND J. BROWN. 1990. Lunar 悶 :productive cycles of benthic-brooding reef fishes: Reflections Reflections of larval biology or adult biology. Ecological Monographs 60 (3): 311-329. R∞KER ，J. AND O. DENNIS. 199 1. Diel ，lunar and seωonal changes in a mangrove fish assemblance 0 釘 sou 血ー westem pt 随時o Rico. Bull. Mar. Sci. 49 (3): 684-6 98. SINGHANOUVONG ，D. ， C. SOU Ll GNAVONG ，K. VONGHACHAK ， B. SAAD 喧Y，AND T. J. WARREN. 1996a. Th e Main Dry- Season Season Fish Migrations of the Mekong Mainstream at Hat Village ，Muang Kh ong Dis 位ict ，Hee Village ， Muang Mouan (Sic) Dis 凶ct and Ban Hatsalao Village ，Paxse. 1o digenous Fishery Development Project ， Fisheries Fisheries Ec ology Technical Report ，Vientiane ，La o PDR ，3: 1-13 1. VAN ZA Ll NGE ， N. P. ，NAO THUOK ，AND TOUCH SEANG TANA. 2∞ O. Wh ere there is wa 旬 :r ，there is fish? Cambodian and and Mekong River Basin fisheries issues 泊 perspective. Pages 10 9- 139 in M. Ahnled and P. Hirsch (eds) ， Common Property in the Mekong: Issues of Sustainabili 砂 and Subsistence. ICLARM Studies and Reviews ，Manila ，Philippines. VAN Z 札附GE ， N. P. ，NAO THUOK ，AND D 臥 P Lo EUNG (eds.). 1999. 針esent S阻tus of Cambodia's Freshwa 飽r Capture Capture Fisheries and Management Im plications. Nine presentations given at the Ann ual Meeting of the De阿武 ment of Fisheries of 血.e Min is 町 of Agriculture ，Fores 町 and Fisheries ，1 9- 21 Janu 町 1999. Mekong River Commission and De p紅 tment of Fisheries ，Phnom Penh ，Cambodia ， 149 pp.

WARREN ， T. J. ， O. C. CHAPMAN ，AND D. SINGHANOUVONG. 1998. 百 e Ups 田町助y-S 側 on Migrations of Some Impo 同組t Fish Species in the Lo wer Mekong River in La os. Asian Fisheries Science 11 (1998): 239-25 1.