Brochures

Abrupt shift from clear to turbid state in a shallow eutrophic, biomanipulated lake

Description
Abrupt shift from clear to turbid state in a shallow eutrophic, biomanipulated lake
Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  PRIMARY RESEARCH PAPER Abrupt shift from clear to turbid state in a shalloweutrophic, biomanipulated lake Istva´n Ta´trai   Gergely Boros   A´gnes I. Gyo ¨rgy   Ka´lma´n Ma´tya´s   Ja´nos Korponai   Piroska Pomogyi   Ma´te´ Havasi   Tama´s Kucserka Received: 23 May 2008/Revised: 24 September 2008/Accepted: 27 September 2008/Published online: 15 October 2008   Springer Science+Business Media B.V. 2008 Abstract  Monitoringdatawereusedtoassesscausesbehind a recent shift from a clear-water to a turbid-water state in Lake Major, a 10 ha shallow lake inHungary. In 1999–2000, fish manipulation was con-ducted in this hypertrophic lake. Reduced fish stock resulted in clearing water and the development of adense ( [ 80% coverage) submerged vegetation in2005. During the recent abrupt shift, which occurredin 2007, submerged vegetation subsequently declinedafter a two-year period of clear water and abundantvegetation. An intense decay of macrophytes withinthe lake produced a rapid transition between the clear-and turbid-water states. During the clear-water state in2005–2006, the most important variables predomi-nantly correlating with macrophyte cover were Secchitransparency, temperature and TN, while TN, temper-ature, Secchi depth and chlorophyll-a were the mostsignificant variables during the turbid-water state in2007. Nitrogen may play a significant role in the coverof submerged macrophytes when TP is moderate. Weargue that several factors in concert are necessary toinitiate a shift. Water temperature likely has contrib-uted to triggering shift through inter-year-dependentchangesincoverofmacrophytes,withfishrecruitmenthaving key roles in the dynamics of shallow lakes. Keywords  Fish manipulation    Alternative states   Macrophytes    Nutrients Introduction Excessive nutrient loading has caused many shallowlakes to shift to a turbid state. For some time, it hasgenerally been accepted that one of the solutions tothis problem is to cut off the external nutrient load(Moss et al., 2005; Phillips et al., 2005; Jeppesen et al., 2005). Upon reduction of such nutrient loading,many shallow lakes may shift back to a clear state ata lower critical nutrient level than the nutrient level atwhich they switched to the eutrophic state (Schefferet al., 2001; Jeppesen et al., 2005). Most shallow lake ecosystems are thus either in a relatively turbidmacrophyte-free state, or in a clear state with highvegetation coverage (Scheffer et al., 1993; Jeppesenet al., 1999).Attempts have frequently been made by lakemanagers to apply various measures with the aim of  Handling editor: Luigi Naselli-FloresI. Ta´trai ( & )    G. Boros    A´. I. Gyo¨rgyBalaton Limnological Research Institute, 8237 Tihany,Hungarye-mail: tatrai@tres.blki.huA´. I. Gyo¨rgy    M. Havasi    T. KucserkaLimnology Department, Pannon University,8200 Veszpre´m, HungaryK. Ma´tya´s    J. Korponai    P. PomogyiKis-Balaton Laboratory, Transdanubian Water Authority,8360 Keszthely, Hungary  1 3 Hydrobiologia (2009) 620:149–161DOI 10.1007/s10750-008-9625-4  slowing down or reversing eutrophication. Numerousfactors may influence the realised stable state of lakesand may vary from lake to lake (Scheffer & van Nes,2007). The central idea behind the srcinal alternativestable state theory is that shallow lakes tend to shiftrather suddenly between clear and turbid states(Scheffer et al., 1993). Nevertheless, the question arises whether an unexpected shift in state representsfew years or even more months. Shifts between thesealternative states may have catastrophic effects suchas a severely reduced plant abundance (Jeppesenet al., 2000). The presence of alternative stable statesimplies that if a lake has gone through a state shift ittends to remain in the new state until the responsiblefactor (e.g. nutrients) is changed back to a muchlower level (van Nes & Scheffer, 2005). It may bethat a strategy combining initial intensive measures(fish removal/stocking) with subsequent follow-upactions is the way forward to restore clear water and amacrophyte-dominated state.Several studies have revealed that both statescould be stabilised by a number of mechanisms(Moss, 1990; Jeppesen et al., 1998; Scheffer, 1998). The stability of the two states varies with nutrientlevel, and the critical nutrient level for lakes tobecome turbid is higher for smaller lakes. Moreover,there is a higher chance for small water bodies to bein a vegetated state (Jeppesen et al., 1990; Schefferet al., 1993; Scheffer & van Nes, 2007). This may be because higher plants are less affected by turbiditysince they will disappear more gradually fromshallow parts of the lake, or they are more protectedfrom the effects of wind (van Nes et al., 2002; van deHaterd & Ter Heerdt, 2007). In-lake restorationmeasures are frequently observed to have poor long-term effects in small (0.5–5 ha) and shallow eutro-phic lakes (Jeppesen et al., 2007). Nevertheless, fish manipulation may not necessarily work in largerlakes where waves and poorer light conditions mightprevent macrophyte establishment irrespective of otherwise conducive conditions (van de Haterd &Ter Heerdt, 2007). However, a nutrient-mediatedincrease in periphyton is often described as beingresponsible for loss of higher plants from shallowlakes, yet this violates the stochastic assumptions of alternative equilibrium theory (Jones & Sayer, 2003). Our previous studies in a small and shalloweutrophic lake showed the potential of making adistinction between states with higher chlorophylllevels and states with lower chlorophyll levels as aneffect of food web manipulation (Ta´trai et al., 2003,2005). The development of submerged macrophytesand the structure and biomass of phytoplankton andcrustacean plankton responded rapidly to a removalof 60% of omnivorous cyprinid fish. Following ashort turbid period, an increase in transparency and adecrease in the concentrations of chlorophyll-a,phytoplankton and phosphorus occurred simulta-neously with the increased presence of submergedmacrophytes. The latter covered [ 40% of the lakearea compared with \ 10% during the premanipula-tion period. The success of this fish manipulationdemonstrated the potential of this measure as a short-term management strategy conducted in a small,shallow and eutrophic wetland-type lake (Ta´traiet al., 2005).Although the objective of re-establishing sub-merged macrophytes seemed to have beensuccessfully attained by 2005, the resulting clear-water state proved to be unstable in the long term.From 1999 to 2002, the lake displayed turbid water,followed from 2005 to 2006 by clear water and anabundant growth of aquatic plants. Following three tofour years of fish manipulation, we experienced alarge and abrupt variation in macrophyte coverage in2007. Since the yearly external nutrient load ispractically zero in Lake Major, due to an intensedecay of macrophytes, turn out between statesoccurred like an emergency.Here we examine probable causes behind a recentshift from clear to turbid conditions in Lake Major.Our data set does not allow us to examine whether thetwo states are true stable states in a dynamic sense.Instead, our aim is to assess probable causes behindthe shift by comparing conditions during the shiftwith conditions during the previous period of clearwater and abundant vegetation, which prevailed forfew years.The question of how important environmentalfactors are for the creation and subsequent mainte-nance of clear water in a small, fish-manipulatedshallow lake has yet to be addressed. The primaryobjective of this study was to determine the mostimportant factors associated with clear- vs. turbid-water states. We hypothesised that macrophyte cover,nutrientlevels,crustacean abundanceandfishbiomasswould be the most important factors differentiatinglow and high chlorophyll conditions. 150 Hydrobiologia (2009) 620:149–161  1 3  Materials and methods Study siteThe study site, Lake Major, is part of a wetland area,the so-called Kis-Balaton Water Protection System(KBWPS), which traps nutrients transported by themain inflow, the River Zala, to a large and shallowlake, Lake Balaton in Hungary (Fig. 1). KBWPRconsists of two parts (KBWPR-I and KBWPR-II) andLakeMajorissituated inthecentral partofKBWPR-I.KBWPR-I has a surface area of 18.5 km 2 and a meandepth of 1.1 m and was created in 1985 (Pomogyi,1996).Since1991,wetlandreconstructionhasresultedin improved water quality and decreasing chlorophyll- a  values in Lake Balaton (Ta´trai et al., 2000). In 1990, however, KBWPR-I became hypertrophicwith mean annual chlorophyll-a concentrations of 150–250  l g l - 1 at the outflow. Cyanobacteria algalblooms began to appear with increased frequencyduring the summer. A recent survey showed that thetotal standing stock of fish in KBWPS-I was over200 kg ha - 1 . Omnivorous cyprinid fish, mainly cru-cian carp,  Carassius auratus  (L.), common carp, Cyprinus carpio  L., roach,  Rutilus rutilus  (L.),common bream,  Abramisbrama (L.)andwhitebream,  Abramis bjoerkna  (L), dominated the fish community(Ta´trai et al., 2005). The experimental area, LakeMajor, is a shallow and eutrophic lake with a meandepth of 1.1 m and an area of 10 ha. The lake iscircular in shape with moderately steep sides, hasneither inlets nor outlets, and is isolated from otherparts of the reservoir by dams. A food web manipu-lation experiment and subsequent monitoring werecarried out from 2005 to 2007, during which data onwater chemistry and biota were collected.Fishery and restockingThe whole-lake fish manipulation experiment in LakeMajor was initiated in May 1999 with sampling forphytoplankton, zooplankton, fish and nutrients (Ta´traiet al., 2005). In April 2000, the lake was drainedwithin three days by pumping out the water, aided bythe digging of canals, to enable the removal of fish.The total removed biomass was 308 kg ha - 1 of which 92% were cyprinids (carp, crucian carp, breamand roach). Recapture estimates made in May–June2000 yielded a total fish biomass including YOY(young of the year) populations of ca 170 kg ha - 1 Fig. 1  The Kis-BalatonWater Protection System(KBWPS) with its two partsand the experimental LakeMajorHydrobiologia (2009) 620:149–161 151  1 3  (Ta´trai et al., 2003). The fish stock of Lake Majorwas reduced by approximately 60%, and the biomassratio of predators to cyprinids based on recaptureestimates increased more than twofold (from 3 to 7%)by 2002 (Ta´trai et al., 2005).The species composition, number and the biomassof the fish stock were determined between April andOctober in the period 2005–2007 as catch per uniteffort (CPUE) using multiple mesh gill nets (7 meshsizes ranging from 11 to 50 mm, total net length35 m, height 1.2 m). In July–August 2005–2007, theratio of the most abundant YOY fish (roach, breamand crucian carp) in total CPUE number wasdetermined. Total length of YOY fishes consideredranged between 5 and 9 cm. Catches in gill nets wererelated to the weight of the population of interest by acatchability coefficient (Hansson & Rudstam, 1994)and thus provided data on catch per unit effort (CPUEas g - 1 standard net - 1 h - 1 ). Gill net samples werestandardised for gear size (standard unit is a 10 mlength of each gillnet panel) and fishing time(standard setting time was 1 h) in PASGEAR(Kolding, 1997).Sampling and analyses procedureIntegrated water samples for analysis of solublereactive phosphorus (SRP), total phosphorus (TP),total nitrogen (TN) and chlorophyll- a  were taken witha surface mud tube sampler (diameter 5.6 cm andlength 2 m) at 6 randomly selected stations on eachoccasion, monthly from April to September from2005 to 2007. Samples were taken at 6 differentstations and pooled on each occasion. From eachpooled sample, a sub-sample of 100 ml was pre-served with acid Lugol’s solution for subsequentalgal cell counting immediately after sampling. Cellswere counted under an inverted microscope and sizedto derive biovolume from appropriate geometricshapes. Phytoplankton biomass was measured photo-metrically as chlorophyll- a  following methanolextraction (Iwamura et al., 1970). Transparency wasmeasured as Secchi depth at the same 6 stations.Turbidity was measured as Nephelometric TurbidityUnits (NTU) in situ using Water Checker (U-10,Horiba Co, Japan) at two depths at the 6 stations.Crustacean zooplankton samples were taken at thesame 6 stations with a Schindler-Patalas sampler andpooled. Samples were immediately preserved in 70%ethanol. On each occasion, 15–20 l of water werefiltered through a 65  l m mesh to concentrate theorganisms. Organisms were counted with an invertedmicroscope after sedimentation in chambers for atleast 2 h. Minima of 100 organisms of the mostnumerous groups were counted. Biomass was calcu-lated with length–weight regressions according toBottrell et al. (1976). The crustacean potential filter-ing rate and the grazing pressure on phytoplanktonwere calculated according to Jeppesen et al. (1994),assuming that cladoceran ingest 100% and copepods50% of their biomass per day.Macroinvertebrates were collected using aHargrave sampler at monthly intervals from Marchto October each year. Three samples were collected ateach station (six in total) and pooled into one sample.The samples were sieved (mesh size 0.25 mm) andmacroinvertebrates retained before being identified,counted and weighed.Macrophyte cover in the lake was mapped byaerial colour infrared digital photographs (Do¨mo¨to¨rfyet al., 2003) to estimate distributions of submergedmacrophytes during July and/or August 2005–2007.A digital elevation model was built from a digitalcontour map of the study area and point data from aGlobal Positioning System (GPS). Species wereidentified by mapping from a boat all over the lake.StatisticsThe srcinal data sets from May to September wereused in the statistical analyses.  Z  -tests were used toexamine significant differences between turbid- andclear-waterstatesformeasuredphysical,chemicalandbiological variables for the summer (June–August)months. Canfield & Jones (1984) suggested that about50% coverage ofa lake’s surface area mustbe coveredby macrophytes to produce a significant impact onnutrient cycling, and accordingly, the data sets weredivided into two periods: the lower chlorophyll periodfrom2005to2006andthehigherchlorophyllperiodin2007. We chose to treat the samples collected in eachmonth as individual cases. Samples taken within ayear are potentially autocorrelated, though shallowlakes can show fast shifts. Consequently, our correl-ative analyses must be interpreted with care and theresults are only indicative. To determine which factorsassociate with turbid water conditions compared withclear-water conditions in Lake Major, the abiotic and 152 Hydrobiologia (2009) 620:149–161  1 3  biotic environment characteristics of each type werecompared using a combination of univariate andmultivariate analyses. The criterion for significancewas set to  P \ 0.05. Correlations between totalphosphorus and phytoplankton biomass measured aschlorophyll- a  for the two separate periods wereobtained by linear regression analyses. All statisticalanalyses were carried out in SYSTAT 11.0 forWindows (SYSTAT Inc., Germany). Results Several factors were associated with the developmentof the clear-water state (CWS) in 2005–2006 com-pared with the turbid-water state (TWS) of 2007(Table 1). TWS had significantly higher water tem-perature (  z  = - 1.817,  P  =  0.034), lower Secchitransparency (  z -test,  z  =  3.941,  P  =  0.000) andhigher turbidity (  z  = - 4.353,  P  =  0.000), chloro-phyll- a  (  z  =  2.313,  P  =  0.00), soluble reactivephosphorus (  z  = - 2.848,  P  =  0.002), total phospho-rus (  z  = - 3.602,  P  =  0.001), Cladocera biomass(  z  = - 3.000,  P  =  0.001), Cladocera/Copepodabiomass ratio (  z  =  4.213,  P  =  0.002), macroinverte-brates biomass (  z  =  2.443,  P  =  0.011), and fishCPUE biomass (  z  =  5.091,  P  =  0.000).The results of stepwise multivariate linear regres-sion showed that a model with macrophyte cover asthe dependent variable and Secchi transparency,chlorophyll- a , temperature, total phosphorus (TP)and total nitrogen (TN) as predictor variables duringCWS was significant (df   =  11,  F   =  9.238, P  =  0.012), explaining 69% ( r  2 =  0.693) of the totalvariance in macrophyte cover. Regression analysisindicated that the most important factor differentiat-ing TWS from CWS was macrophyte coverage. Thepercentage cover of  [ 50% proved to be suitable fordifferentiation between the two states. During CWSin 2005–2006, the most important variables predom-inantly correlating with macrophyte cover wereSecchi transparency (df   =  11,  t   = - 5. 038, P  =  0.001), temperature (df   =  11,  t   = - 3.397, P  =  0.008) and total nitrogen (TN) (df   =  11,  t   =- 2.642,  P  =  0.027). However, there was no signif-icant correlation between macrophyte cover and totalphosphorus (TP) (df   =  11,  t   = - 0.908,  P  =  0.387).Stepwise multivariate linear regression showedthat, in particular, a model with macrophyte cover asdependent variable and Secchi transparency, chloro-phyll- a , temperature, total phosphorus (TP) and totalnitrogen (TN) as predictor variables during TWS washighly significant (df   =  5,  F   =  13.019,  P  =  0.001)and explained 93% ( r  2 =  0.928) of the total variancein macrophyte cover. During TWS in 2007, themost important variables predominantly correlatingwith macrophyte cover were TN (df   =  5,  t   =- 6.147,  P  =  0.001), Secchi transparency (df   =  5, t   = - 3.898,  P  =  0.012), chlorophyll- a  (df   =  5, t   = - 2.979,  P  =  0. 017) and temperature (df   =  5, t   = - 2.879,  P  =  0.021). In contrast to CWS, thecorrelation between macrophyte cover and TP in Table 1  Mean summervalues (June–August) forenvironmental variables inturbid—(TWS) and clearwater states (CWS)(mean  ±  95% CI)* Denotes significantdifference at 95%confidencePeriod: May–SeptemberVariable 2005–2006 2007CWS TWSAverage depth, m 1.21 1.09Water temperature* 19.9  ±  6.9 25.2  ±  0.5Secchi transparency, cm* 44.5  ±  14.9 20.4.6  ±  1.8Turbidity, NTU* 16.4  ±  13.7 73.8.4  ±  16.9Soluble reactive phosphorus,  l g l - 1 * 33.5  ±  15.5 78.3  ±  20.3Total phosphorus,  l g l - 1 * 147.5  ±  51.0 291.4  ±  43.5Total nitrogen,  l g l - 1 * 1,728  ±  501 4,580  ±  481Chlorophyll-a,  l g l - 1 * 75.5  ±  43.4 255.2  ±  16.7Coverage of macrophytes, % 81 28Cladocera biomass, mg l - 1 * 5.4  ±  0.5 8.5  ±  2.1Cladocera/copepoda ratio* 0.52  ±  0.18 0.24  ±  0.12Macroinvertebrates biomass, g ww m - 2 * 9.3  ±  2.9 3.7  ±  3.2CPUE-biomass, g h - 1 st. net - 1 * 1,612  ±  373 2,960  ±  362Hydrobiologia (2009) 620:149–161 153  1 3
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks