SPECIES COMPOSITION AND POPULATION DYNAMICS OF THE PHYTOPLANKTON COMMUNITY IN KOTMALE RESERVOIR

: The present worlr was carried out in the Kotmale reservoirfrom April 1994 to April 1096 with tlle objectiveof studying thespeciescomposition and the seasorlal, spatial and cliurnill distributionofpliytoplankton. Thephytoplankton of the t.eservoir was composed of four major families: Chlorophyceae, Cyanopliyceae, H~~cillnriophyceile and Dinophyceae. Chlorophyceae dominated thc phytoplanliton, with Stnrrrnsh.~~ni species forming the mqjor component. ,, Ihe results show seasonal v;i~.i;ition in the major phytoplanlcton types of ttie ~~eservoirthatcorrelates with the water level. Iluringthe wetmonths,Stn~~.rostrum species dominated the phytoplankton while in the dry months &ficl.~sy.~tis nerz~gir~osa became prominent especially in the upstream region of the reservoir. This periotlic occurrence of M. aeruginosn bloom was a prominent feature of the impo~uidment. Although the filamentous diatom, iVie1osir.a species and the dinoflagellate, Peridil~i~~lli. C ~ N C ~ L L I ~ formed an important part of the plankton community, no contrasting periodicity was observed as in M. aer~~ginosa.


INTRODUCTION
The EZotnialc reservoir is the uppermost impoundment in Sri Laillia and was constructed it1 1985, under the Accelerated Mahaweli programme for the generation of'hyd~~oelcctsicity. A 87 m high rock filled dam was built across the Kotmalc Oya, which is a main tributary of the river Mahaweli, to create the 174 million cubic meter capacity reservoir.' According to thc information gathered since 1987, Kotmale is a thermally stratified waterbody indicating a clinogradc type of dissolved oxygen ~urve~~%nd the reservoir is very sensitive to eutrophication." The intensity of eutrophication was indicated by the thick Cort.esponcl~ng author bloom of Microcystis aerz~g.ilzosa in 1991, that occurred despite high flow through rates.
The extensive enrichment of the reservoir is mainly due to the delivery of nutrients through agricultural and urban sources of the cat~hment.","~ A study was commenced in April 1994, with the objective of investigating types of fish plankton interactions in the reservoir and the possibility of using fish for controlling the bloom during extreme situations. The present paper deals with the observations on species composition, seasonality and diurnal migration patterns of the phytoplankton community of the waterbody, related to the above study.  Phytoplankton samples werc collected once a month using a closing type net (mesh slzc, 30ym) which was attached to a winch and operated by a boat. Samples at lorn intelvals were taken from bottom to surface at the Major statlo11 1, dur~ng the pcriod of study Horizontal samples were collected fi-om thl ce malo1 statlolls (N11. M2 and M3) and two sub stations (I? and G) by dragging the net lor. one minute while thc boat was moving at a very slow speed. These pla~lkton samples wc1.c collcctcd between 1000 and 1300h. Three diurilal samj)li12g:'" \ver.e cr'u.ric(l out at six hour'ly inte~.vals to find out the diurnal migration pnttelns ol thc planlrtoii These samples werc immediately transferred into pre-labelled bottles and preserved in 5% formalill. I11 the laboratory, the samples were diluted to a known volumc and shaken tl~oroughly to achieve a uniform distribution before lml was pipetted out on to a Sedgwick-Rafter cell. The phytoplankton were identified with the aid of Abcywickrama,' Abcywickrama &Abeywlckrama.$ and Whitford & Schumacl~er," using a llght lnicl oscopc and the lncan of'the three replicates was taken as the final value Thc total numbei.s of individual species collected at each station cand at each depth were quarltlfjed and tlie phytoplankton density was expressed as nil m ber of' cc I I s pet rn ' of' the reservoir water.

METHODS AND MATERIALS
Temperature, p1-l and ~oliductivity of the surface water sarnplcs at each statlon were measured usiilg a digital thermometer, pH meter (Jenway, model 3070) and a conductivity meter (Jenway, model 4070) respectively. Secchi depth (SD) transparency was determined as the averaged reading derived from lowcring and raising a standard 25cm diameter, black and white disk (Secchi disk) from the shady side of'the boat. From these readings, the euphotic limit (Zri,) ol' the reservoir was cnlci~lated by multiplyii~g with a factor of 2.5.l" T11c data were analysed using Microsoft Excel 5.1 (1985Excel 5.1 ( -1992 version and correlation coefficient valilcs ( r ) we1.c calculated to find out the siguificance of the ~.OSLII 1s.

Species composition and abundance
Fol-ty three species of pliytoplanliton were identified during the two year study pcriod and were catcgorisccl into four taxonomic groups as indicated in Table 1, where the major pliytoplanlctoil types were grouped accol-ding to thcir abundance. Of these, Chlorophyceae was the most abundant with 26 species while the group Dinophyceae contained olily one species,,Pe~-idil-~i~,tr~ citzctujn.
Cyanophyceac was the seco~id abu1ida11.t group wit11 10 species svllile six species were recorded from tlie group Bac~llariophyceae. Of'the Chlorophyceae, the genus Stai~t.astl-z~in was the most dominant, with more than seven species. As indicated in Fig. 2 and 3. during most months, S~UZLI-astrzsm species represented over 80 percent of the total phytoplankton population of the reservoir and could be coilsidered as the only genus belonging to the category of 'very abundant'.   Othcr than Star~/.nstl-zcl)i specics, Cosn~ul-iz/m spccics was the oiily othcr chlorophyte, which was prcsent throughout tl~c year in coi~siderable amounts.

Spatial distribution of phytoplankton
Al.1, the phytoplankton types were distributed throughout the reservoir water, but some phytoplanliton were more common in certain areas of the impouud-~nent (Fig. 3). I'nl-titularly, the M .ae~-r~g-il~osn and P. cl~~ctz~nz sl~owed highest dcnsitics in the up stream region. In May 1995, over 55 percent of' the pliyt,oplankton population in up st re an^ region, was composed ofM. ac?~-rlgt~~osu. whilc Major station 1 a~id 2 h21d 39.3 percent and 52.3 pc13cent ~.cspecti\wly. When compat.cd to the above mcntioiiecl species, the distribution ofStccr~rustrr~~~~~ spccics was more homogeneous while Melosira species was morc abundant closcr to the dam than in the upstream region. When c0nsiderin.g the vertical, clistribu tion of the phytoplankton, a considerable amount ofindividuals were present throu.ghout the water column (Fig.4).  (3) 0.05

Temporal distribtioli of phytoplankton
As illustrated in Fig. 5    Physicocliemical properties and phytoplankton dyilamics Fig. G(a ) ~llustt-ates tl,c \water lcvcl fluct~iatioii at the thrcc M:\jor statio~is Although during the ra111y season, the reservoir water level was high, the correlation coefficient value between tkc two parameters sl~owed a negative value ( Table 2). This could be due to the combined efIhcts oflag period betwceo the rainfall of the catchement and water received by the impoundment and duc to high flow through rates during power generation Sccchi clepthvalues varied between 2.7 to 0.33 m and. was high.er.inMajol: station 1 thanin otller stations. Fig G(b) indicates low secchi dcptli \~al.ues during low water levels or during drought. This may bc corl.elatcd eitllcl with lligli phytoplankton growth or. with high turbidity of' the water due to sed.i.mcnt particles ( As shownin Fig 5 the vertical d~stribut.10~1 oftotal phytopla~llrtonin Malor stat~on 1 var~ed with the euphotic limlt. When the eupllotic 11rnlt was h~gh, the phytopl anktondensity was also high and the plankton was found throughout the water column. In April to May 1995, the euphotic zone was low and tlic plankton dlstributlon was 11mltcd to the upper water layel 5 Thc temperature variation of the reservoil \water was minimum, with temperaturcs around 26°C for most part of'tlie year,. The 131-1 ot'thc water val:it!tl bc1;ween 6.2 to 9.3 (Fig 6). Urhcn tile M~:croc;ystds density was higli, the pH and temperature of the rcscl.voir water was also high and the water lcvel was low before the density of Mici-ocystis increased. After the lleavy rains in June, the density dropped to low levels i13 J~lly. Also P. CIILC~LLJZ showed high densities when the water level, of the reservoir was low (Fig, 2,3 and 6).

Species conlposition and abundance
Tl~c present sti~dy agrees with the fact, that thc phytoplaultton conirnunities o(' the newly built upland ~-esc~:voi~.s of'the river Mahaweli are doniinntcd by tlic family Chlorophyccae, and Stazl.rastr~i~n species. Howcver, P. CLI~.C~ZL/~~ which significantly contributed to the total phytoplankto~l of'theKotmale rescrvoil-was not recorded from Victoria and Rantan~be.*l.'~ 117 the lowland reservoirs of Sri Lanka, (e.g. Parakrama Samudra), Chloi-ophyceae were represented by a large numbel. of species, but in low densities and the do.mina17t phytopl.anl<ton groi~p was the cyanopl~yta.'" It is correlated with thc fjct that thcy arc limi~ologically different fiom the upland reservoirs in their morphology and flow through regimes l4 Lake George of Uganda, which is a shallow tropical fresh water body was dominated by Microsyslis species'5vhile in water bodies in Kenya, tlic prom inent phytoplanlcton recorded belonged to the family Chl~rophyceae.'~

Seasonal variation
According to Horne & Goldman,15 phytoplankton is subjected to strong seasollal influences. In thc temperate and polar zoncs, there is a great contl-ast between the summe]. and thc wintcr. and in thc tropics between the rainy and the clry seasons. Algac respond to this contrast rearrangcmcnt of' the physical and chemical s-tructure of their en.vironment w.ith characteristic popl-lat ti on fl.uctuations.
The seasonal variation of' phytoplankton commun~ty in thc Kotmalc reservoir secms to correlate with water lcvel fluctuations, caused by the rainfjll and the out flow due to powcr generation, seepage and evaporation. The depth from which watcr is drawn out for power generation depends on Ilic water Jcvcl as the bvutcr cntcring the tunnel rnay orlglnate from ncar the s~n.t'ace or at must from a dcpth 0135 m when the reservo~ris at fill L supply Ievcl.'" This 1s ~mportant slnce water drawn from the surface 1s rlcli ul pbytoplanlcton and poor in nutl.icnts co~i~pa~wl to water drawn out fivm lower. lcvcls '" Th.e South West monsoon fi-om May to September brings the highest rajnfall to the catchment and February is the driest month. Inter monsoonal periods give more rainfall than the North East monsoon." Hence for most part of the year, the reservoir has high water levels and during such periods Staurastri~.m species remain the dominant group. T11e driest period of the reservoir catchment was generally from December to March and the water level of' the reservoir drops especially in the upstream region. During this pcriod clue to reduced power generation, the flow through rates were low and the temperature and pI-I ranges were high. These conditions favoured the dominance of M. uerz~gitzosc~..~'~~"'O Chandrananda" has ~.ecorded thc highest fWcrocystis densities duri~lg 1 , --September to Novemberin 1991, and this was due to the fact that the catchment of the reservoir did not experience the soutll west monsoon adequately. Due to Accordi~lg to Ryding & Rust," algae can bc accumulated to nuisance Jcvcls in a lake or a reservoir, only when the algac growth rate is f~~stcr. than thc rcnewal or the rate of' flushing of thc water body Duv~ng thc. present investigation, in 1994 and 1995 normal mo~lsoonal ralns were cxp~r~~~icccl T11c bloom condit~ons werc observcd only during May to June, due to low rainf'all from Dccember to March.
During the Microcystzs bloom period, the domlnant phytoplankton, Stazrrastl-un~ species droppcd to low levels. Tl~c dlnoflagellatc, P cllztz~nz appears to be the only other phytoplankton whichrcachcd near to the populatioi~ dcnsitics of'Stuurastrunz species and as recorded by Chandranancla' it appcars to increase at times when the water lcvcl is very low.

Spatial diurnal variations
Phytoplankton was not evenly distributcd in the Kotmale reservoir. Biomass was usually greater ill mid stream to upstream regi0n.s but lowcr at greatcr depths, wlzich ngrces with Rott,':: who states that tl~ere was a distinct diffci.euce bctwecn thc phytoplai~kton of Southc111, Middle and No~~thcrn parts of' the Parakrama salnudra iluriagl~is study period. The sub stati.o~xs, 17 and G appear to have the highest population densities ol'Adic~.ocys.ti,s and P. CLII.(:~~I,IIL clurjng certain months. This indicates that the intense growth of'tl.lcse spccics were ;it the upstrcam regions. These stations rcceive nutrient rich watcr dil.ectly from the tributaries (the Kotmala Oya and the Pundalu Oyaj. Thcy bring the surface runoff from the dense tea cstates of the up-country, where intensive use of agrochemicals including hrtilizers, pesticides and fungicides is being The studies carried out in dif'fBrcnt water bodies in B1:azil have shown, that the diurnal cycles are as irnpulta~lt as the seasonal cycles and they establish patterns of' spatial heterogeneity, that will produce density g~-eclients interfering with thc sirlking rates of pl~ytoplankton and tlic nutrient distribution of the watcr col~~mn.'~ccordiilg to Westlakc ct nl.,':' diurnal. alternations in phytoplankton are only possible if their migratory velocities exceed the rate of vertical displacement of water by turbulence. Mjgration by changing buoyancy and active migration by flagellates are the common mechanisms involved.
In the Kotmalc reservoir both types of migrations were observed. Microcystis species having gas vesicles within their cel.1~ showed migration by changing buoyancy and according to Humphries & Lyne'j a reason for the dominance of cyanophytes is due to this vertical migration which allows them to exploit any spatial. separation between the nutrients in the hypolymnion and light i nteusi ty in the epilimni on. Unlilre the Mlc7-ocystls, P. ciltctr~7,z showed active migration and as they are photoactive they nor.mally swim up to the surface in the morning for phot~synthesis.~"