RESTORATION AND MANAGEMENT OF LAKES AND RESERVOIRS - CHAPTER 9 pot

For distri-example, Schindler 1978 found a highly significant correlation r = 0.69 between pelagic pro-ductivity and steady state lake P concentrations for 66 P-limited lakes shallow an

9 Biomanipulation

9.1 INTRODUCTION

Intensive research over many decades has developed an understanding of factors regulating bution, abundance, productivity, and species composition of phytoplankton, especially in deep lakes(reviews by Pick and Lean, 1987; Hecky and Kilham, 1988; Kilham and Hecky, 1988; Seip, 1994,among many) The common approach for long-term control of nuisance algae is to lower nutrientconcentrations, an approach supported by controlled experimental laboratory, field enclosure, andwhole lake investigations demonstrating that phosphorus (P) and sometimes nitrogen (N) concen-trations are causally linked, especially on a long-term basis, to algal production (e.g., Schindler,1977; Smith and Bennett, 1999) The link between P concentration and algal biomass is frequentlyillustrated with a log-log TP-chlorophyll regression indicating that most long-term changes in algalbiomass are explained by changes in P concentration (Figure 9.1) However, when these data areplotted on a linear scale (Figure 9.2), especially on a short-term basis, variances are apparent,suggesting other factors in addition to nutrients can be important in determining algal biomass For

distri-example, Schindler (1978) found a highly significant correlation (r = 0.69) between pelagic

pro-ductivity and steady state lake P concentrations for 66 P-limited lakes (shallow and deep), ranging

in latitude from 38° S to 75° N The relationship, however, explained only about half of the variance,meaning that P concentration and chlorophyll are highly correlated, but that the relationship might

be weak or non-existent for some lakes in some years or parts of years Grazing, mixing, and/orallelopathic materials, might influence and/or control algal biomass in these lakes

An example is Square Lake, Minnesota (Osgood, 1984), a lake with a greater Secchi Disc (SD)transparency than expected from its TP concentration of about 20 μg P/L (mesotrophic) SD wasmore than 7 m during summer months, a depth typical of oligotrophic lakes, and apparently due

to Daphnia grazing.

One purpose of this chapter is to examine factors other than resources (e.g., nutrients, light)that can control algal biomass in deep lakes, and to discuss how lake managers might use thisknowledge to address phytoplankton problems Since most lakes are shallow and may be dominated

by either phytoplankton or macrophytes, this chapter also examines factors determining whichproducer type dominates, and how this knowledge can be used to manage shallow lakes

9.2 TROPHIC CASCADE

As illustrated by the Square Lake, Minnesota case history, zooplankton grazing is a source of algalmortality, sometimes leading to lower algal biomass in the water column than expected for a givennutrient level This might occur when planktivores are suppressed by piscivores But in some lakes,zooplankton herbivory may be low during some periods of the summer, perhaps due to fish orinsect planktivory, allowing phytoplankton to bloom

A set of hypotheses was developed to explain the roles of resources (nutrients, light) and trophiclevel interactions The original organization of these ideas was directed at terrestrial communities(Hairston et al., 1960), and was later (Smith, 1969) proposed for lakes (see Hairston and Hairston,1993) Hairston et al (1960) predicted that in systems with three dominant trophic levels (producers,herbivores, primary carnivores), producers would be resource-controlled, whereas in four trophic

level systems (top predator also), producers would be consumer (herbivore)-controlled and the firstcarnivore level would be controlled by predation.

Observations of ponds by Hrbacek et al (1961) supported the above ideas Fish planktivory(no piscivory) reduced zooplankton grazing, leading to algal blooms Further early evidence, from

Brooks and Dodson (1965), found that planktivory by the alewife (Alosa pseudoharengus) in New England lakes led to elimination of the most efficient herbivores (large-bodied Daphnia) and to selection for smaller-sized zooplankton (e.g., Bosmina) that have lower grazing rates and choose

smaller food (algae) particles They proposed the “size–efficiency” hypothesis to explain grazingimpacts of smaller and larger-bodied zooplankton on phytoplankton

The term “trophic cascade” was introduced by Paine (1980) to describe the roles of speciesthat he termed “strong links” or “strong interactors” in intertidal communities These are specieswhose removal (or introduction) produced dramatic changes in prey biomass If the prey was acompetitively superior species, the effects could “cascade” from predator to trophic levels one ortwo links away Pace et al (1999) defined “trophic cascades” as (p 483): “reciprocal predator–preyeffects that alter the abundance, biomass or productivity of a population, community or trophiclevel across more than one link of the food web.” The pelagic trophic cascade, with and without adominant (“strong link”) top carnivore level, is illustrated in Figure 9.3

Carpenter et al (1985) argued that trophic cascades could explain the large variances (Figure9.2) in algal biomass or productivity between lakes with similar nutrient concentrations Theyproposed that nutrient levels determined the long-term productivity or trophic state of a lake, butthe year-to-year variances from expected trophic state were set by trophic-level interactions Strong

FIGURE 9.1 Relationship between summer chlorophyll and total phosphorus in a number of lakes (From

Shapiro, J 1979 U.S Environmental Protection Agency National Conference on Lake Restoration USEPA

piscivory suppressed planktivores, allowing zooplankton grazing to reduce algal biomass, a trophiccascade extending from piscivores to phytoplankton.

DeMelo et al (1992) suggested that evidence for trophic cascades in lakes is weak, except forstrongly manipulated lakes, a conclusion contradicted by some experiments, but supported by others.Jeppesen et al (2000) described trophic cascades in shallow Danish lakes Brett and Goldman(1996) noted that there have been few whole-lake studies, but in 54 pond and enclosure experimentsthere was evidence of trophic cascades Trophic cascades continued over a multi- year period infertilized, dimictic lakes (Carpenter et al., 2001) In contrast, Drenner and Hambright (2002) foundthat 10 of 17 experiments, not confounded by other manipulations, failed to support the trophiccascade hypothesis However, in most lakes dominated by planktivores (three trophic levels) theslope of the chlorophyll: TP regression (Figure 9.1) was three times that of four level (piscivore-dominated) lakes, indicating that piscivore control of planktivory may lead to enhanced zooplanktongrazing and lower algal biomass than expected for a given nutrient level There also may be a

“behavioral cascade.” Planktivorous fish seek refuge in macrophyte beds in the presence of

pisci-vores caged in open water, allowing longer open water feeding by Daphnia than when piscipisci-vores

are absent (Romare and Hansson, 2003)

FIGURE 9.2 Replotting of part of the data from Figure 9.1 (From Shapiro, J 1979 U.S Environmental

Protection Agency National Conference on Lake Restoration USEPA 440/5-79-001, pp 161–167.)

The trophic cascade hypothesis of Carpenter et al (1985) is one of the most significant inmodern limnology It stimulated new research and led to a paradigm about control of lake produc-tivity that included both biotic interactions and the role of resources Readers are urged to examinethe many books and review articles describing this concept (e.g., Kerfoot and Sih, 1987; Carpenter,1988; Gulati et al., 1990; Elser and Goldman, 1991; Carpenter and Kitchell, 1992, 1993; Hansson,1992; McQueen, 1998).

FIGURE 9.3 Hypothetical scheme showing the connections involved in food-chain biomanipulation in lakes.

Shaded area represents tentative connections (From Benndorf, J et al 1984 Int Rev ges Hydrobiol 69:

407–428 With permission.)

Low Possibly High

(small species)

(Large) (Colonies) Effects on water quality

High secchi depth Normal or high pH Normal or extreme

Low secchi depth

(Small body size)

(large body size) Zooplankton

Predators

plankton feeders

Zoo-Phytoplankton

Role of the carnivorous zooplankton

Role of the nutrient load and of the hydrophysical conditions

?

?

9.3 BASIC TROPHIC CASCADE RESEARCH

Research on the trophic cascade hypothesis (Carpenter et al., 1985) generated a greater ing of lakes, and many new questions about lake ecology Trophic cascades may be more common

understand-in mesotrophic lakes than understand-in oligotrophic or hypereutrophic lakes, leadunderstand-ing to an “understand-intermediatetrophic state” hypothesis (Carney, 1990; Figure 9.4) In mesotrophic lakes, edible, nutritious algaedominate the plankton, and large-bodied zooplankton are abundant In eutrophic lakes, especiallyhypereutrophic lakes, the phytoplankton community may be dominated by cyanobacteria that may

be toxic, inedible, non-nutritious, and/or produce clogging of the filter-feeding mouthparts ofzooplankton (Gliwicz, 1990; Lampert, 1982) These factors may explain the rarity of large-bodiedzooplankton in eutrophic lakes, rather than planktivory (deBernardi and Giussani, 1990; Gliwicz,1990; DeMott et al., 2001)

The “intermediate trophic state” hypothesis was examined by comparing grazing in oligotrophic Lake Tahoe (California/Nevada), mesotrophic Castle Lake, California, and eutrophicClear Lake, California, using enclosures with ambient or enhanced ambient zooplankton, and

ultra-enclosures with added Daphnia pulex (Elser and Goldman, 1991) In the eutrophic lake, ambient zooplankton (even at eight times in-lake density) had no impact on phytoplankton biomass D pulex had a weak effect It was believed that Anabaena circinalis, (42% of the phytoplankton) was

either inedible or interfered with filter-feeding In the mesotrophic lake there were large declines

in phytoplankton but increases in primary productivity, suggesting grazing and nutrient recycling

by zooplankton In oligotrophic Lake Tahoe enclosures, copepods dominated with no grazingimpact The trophic cascade was strongest in the mesotrophic lake

McQueen et al (1986, 1989, 1992) proposed a “top-down bottom-up” model to explain trophiclevel interactions, based on enclosure experiments and a multi-year study of eutrophic Lake St.George, Ontario Regressions between TP, chlorophyll, and fish biomass (piscivore and planktivore)indicated that bottom-up forces (nutrients) were strongest at trophic levels nearest resources, whiletop-down forces were strongest near the piscivore level They predicted that bottom-up control

FIGURE 9.4 Schematic diagram of the strength of herbivore grazing and nutrient regeneraton in relation to

trophic state (From Carney, H.J 1990 Verh Int Verein Limnol 24: 487–492 With permission.)

Lakes, Ponds Cladocerans, esp Daphnia,

with higher filtering rates, nutrient regeneration

Copepods dominate, lower grazer concen- trations, less nutrient regeneration

Algae with greater defenses, less palatable and nutritious Trophic state

becomes increasingly important relative to top-down control in enriched lakes Therefore, piscivoresare less likely to have a cascading effect on algal biomass in the most productive lakes.

A 1982 fish winterkill in Lake George eliminated 72% of largemouth bass, allowing a test ofthe model, using a 6-year database In 1983–1984, planktivore biomass increased rapidly, and thenbegan to decrease in 1985 as the bass population recovered As predicted, a strong top-down cascadefrom piscivores through planktivores to zooplankton was observed, though negative correlations

between planktivore biomass and either total zooplankton or Daphnia biomass were not significant Correlations of zooplankton or Daphnia biomass with chlorophyll and transparency also were not

significant, but correlations between log chlorophyll and log TP were positive and significant.McQueen et al (1989) argued that long-term processes determining trophic level biomassdepend on resources and energy flow (“bottom-up”) In accordance with Carpenter et al (1985),McQueen et al (1989) concluded that short-term disturbances or cascades set “realized biomass”limits Lakes are strongly influenced by year to year changes in precipitation and associated waterand nutrient loading and by climate (mixing events, fish winterkill), and these stochastic events inturn affect both top and bottom trophic levels, including effects on fish biomass, reproduction, andmortality (Carpenter et al., 1985) These events may not be instantaneous effects of stochasticchanges, but instead there will be lags or inertias that produce responses at other times “Algalproduction today may depend on yesterday’s zooplankton, which depended on zooplanktivoresduring the past month, which depended on piscivore recruitment the previous year” (Carpenter etal., 1985, p 637)

The controls of algal biomass in shallow and deep lakes and at various levels of enrichment

remain controversial Empirical data from a large sample (n = 446) of shallow and deep lakes, fromarctic to temperate zones and ranging from oligotrophic to hypereutrophic, did not support thehypothesis of McQueen et al (1989) that the cascading effect might be greatest in oligotrophiclakes Instead, the survey gave partial support to the “intermediate state hypothesis” (Elser andGoldman, 1991), and indicated that at high TP, the most probable condition (even with removal ofplanktivores) is high algal biomass and turbidity (Jeppesen et al., 2003a)

Trophic cascades are an important determinant of biomass at planktivore, herbivore, andproducer trophic levels, and explain some of the variance observed in regressions between resourcesand biomass Exciting controversies remain about forces that organize ecosystems, and about theadaptations that species populations make to counter those forces But lake managers want to knowwhether manipulations of trophic levels can produce clearer lakes, and if so, how long will theeffect last?

9.4 BIOMANIPULATION

Caird (1945) was the first to publish observations about phytoplankton responses to increasedpiscivorous fish biomass Caird suspected that largemouth bass addition to a 15 ha lake in Con-necticut was associated, through food chain effects, with 4 years of reduced phytoplankton blooms,resulting in a termination of copper sulfate applications

Shapiro et al (1975) proposed the term “biomanipulation”, which he defined (Shapiro, 1990)

as “a series of manipulations of the biota of lakes and of their habitats to facilitate certain interactionsand results that we as lake users consider beneficial — namely reduction of algal biomass and, inparticular, of blue-greens” (p 13) Shapiro et al (1975) included effects on algal biomass from

“top-down” control of zooplanktivores by piscivores, and “bottom-up” effects on algae such asnutrient cycling by benthivorous fish Many lake managers apply the term only to top-down control

of planktivorous fish (see Drenner and Hambright, 2002) More recently, the term has referred tonearly all ecological manipulations to manage algae and aquatic plants

There have been many review articles and books about biomanipulation, including Shapiro(1979), Gulati et al (1990), DeMelo et al (1992), Carpenter and Kitchell (1992), Moss et al

(1996a), Hosper (1997), Jeppesen (1998), McQueen (1998), Bergman et al (1999), and Drennerand Hambright (2002) Lazzaro (1997) is a particularly useful comparative summary.

The following examines biomanipulation case histories, particularly their effectiveness andlongevity Several investigators (e.g., Jeppesen et al., 1997) indicated that resource control of algalbiomass is weaker in shallow lakes, suggesting biomanipulation may be more successful in them.Therefore, the examination of shallow and deep lake case histories is separated

9.5 SHALLOW LAKES

Characteristics of deep and shallow lakes were described in Chapter 2 Briefly, shallow lakes have

a mean depth less than 3 m, are usually polymictic, and often have significant nutrient recyclingaffecting the entire water column Compared with deep lakes, fish biomass per volume is higher,the impacts of fish on turbidity and sediment nutrient release are greater, and the area colonized

by macrophytes may be close to 100% (Cooke et al., 2001) These and other characteristics aresummarized in Moss et al (1996a) and Scheffer (1998)

Shallow lakes are more common than deep ones Unlike deep lakes, shallow lakes, at moderate(30–100 μg P/L) nutrient levels, appear to exist in alternative states: Either they are clear withrooted plant dominance, or turbid with algae dominance (Scheffer, 1998) At low nutrient concen-trations the clear-water vegetated state is most likely, whereas at higher (perhaps > 100 μg P/L)concentrations, the turbid state is more likely (Hosper, 1997) At concentrations between theextremes, either the clear or turbid state can occur It is the forces determining lake state that aresubject to manipulation, leading to the possibility of “switching” from one state to the other Goodexamples of clear and turbid shallow lakes existing within a limited geographic area (Alberta) werepresented by Jackson (2003)

Macrophyte-dominated, clear-water lakes are resistant to development of algal dominance fromincreased external nutrient loading because plants reduce wind and boat-generated resuspension of

sediments, provide daytime refuge to algae-grazing Daphnia, their periphyton may take up

signif-icant amounts of nutrients, and some macrophytes release compounds inhibitory to algae Piscivoresmay thrive in macrophyte-dominated lakes, controlling fish that prey on zooplankton and on

periphyton-consuming snails (Bronmark and Weisner, 1996) This last effect of fish is important

because, as discussed later, abundant periphyton may reduce rooted plant growth

Resistance of the clear water macrophyte-dominated lake to change is reduced by some plantmanagement activities (e.g., harvesting, grass carp; Chapters 14 and 17), by increased fish produc-tion (young of the year (YOY) of most fish species are zooplanktivorous), and by introduction of

toxins (e.g., copper sulfate, herbicides, insecticides) lethal to Daphnia and to plants There is a

nutrient-based stability threshold for the clear water, macrophyte-dominated state of about 50–100

μg P/L Continued loss of stability as nutrient loading increases and/or plant removal occurs canproduce an abrupt switch to the alternative, turbid, macrophyte-free state Moss et al (1996a) calledthese changes leading to the turbid state “forward switches.” Figure 9.5 illustrates this model ofalternative states and the forces that promote a switch from one state to the other Note that eitherthe clear or turbid water state can occur without change in overall nutrient concentrations Switching a turbid lake to a clear water lake may not occur, even when external loading issignificantly reduced This means that the common advice given to shallow lake owners to reduceexternal loading as a method to clear up the water may not produce expected results Sediment re-suspension by wind, boat, and fish activity will attenuate light, preventing re-establishment ofmacrophytes Extensive internal nutrient recycling may continue to sustain phytoplankton andreduce transparency, preventing re-establishment of clear water and macrophytes Fish removals,followed by piscivore stocking and enclosures to protect plants from birds, are among the bioma-nipulation procedures that trigger the switch to a clear water state Reduction of nutrient concen-trations through diversion (Chapter 4) or P inactivation (Chapter 8) increases the probability of the

FIGURE 9.5 The alternative stable states model for dominance by aquatic plants or phytoplankton in shallow lakes, over the gradient of total phosphorus concentrations

that includes both pristine values and those encountered in polluted conditions (From Moss, B et al 1996 A Guide to the Restoration of Nutrient-Enriched Lakes.

Broads Authority, Norwich, Norfolk, UK With permission.)

PLANT

DOMINANCE

(may operate at any nutrient concentration in the overlap range) Mechanical or boat damage, herbicides, exotic vertebrate grazers, pesticides, increased salinity, differential kills of piscivores.

PHYTOPLANKTON DOMINANCE, TURBID WATER

Tall plants System increasingly stabilized by Cladoceran (Daphnia) grazing to maintain clear water.

Sparser plants but clear water retained by grazing.

PLANT DOMINANCE, CLEAR WATER

System often with abundant blue-green algae, some of which may be poorly edible, but main buffer is the lack of zooplankton grazing due

to lack of refuges for Cladocera.

Green algae become more abundant at high nutrient levels, and are generally grazeable.

System buffered by lack of grazers and heavy periphyton growth on any developing plants;

possibly also by inhibiting conditions in sediments.

REVERSE SWITCHES

Copyright © 2005 by Taylor & Francis

switch Figure 9.6 illustrates the pattern of resistance of clear and turbid shallow lakes to increasing

or decreasing external nutrient loading

Biomanipulation, especially top-down procedures, is more likely to be successful in shallowlakes, and in turn, shallow lakes are easier to biomanipulate because nearly all fish can be removed.The following case histories illustrate the outcomes of some efforts to restore turbid lakes to theclear, macrophyte-dominated state Most examples are European because they prefer rooted plant-dominated, clear water conditions in their shallow lakes In North America, with a high density ofshallow lakes, ponds, and reservoirs, lake users appear to want an algae-free, non-turbid, macro-phyte-free lake, regardless of factors preventing this condition from being stable or even possible(e.g., high internal nutrient recycling, high external loading, and/or stocking of exotic herbivorousfish) This unrealistic goal is possible only with continual reliance on expensive mechanical and/orchemical controls

More realistic expectations about the trophic state of shallow lakes might be of value to NorthAmerican lake users For example, a common tactic is to attempt to manage a macrophyte-dominatedlake toward an intermediate biomass of plants that may satisfy the lake users who want a macrophyte-free lake This condition is unlikely where external and internal nutrient loads are high or increasing,and where plant removal (e.g., harvesting, stocking of grass carp) is extensive These conditionsmay drive the lake to the turbid state It could be more realistic, in some cases, to manage someshallow lakes in an area toward macrophyte-free water that may be compatible with boating andswimming, while managing other nearby lakes toward the clear water condition (Van Nes et al.,

1999, 2002) For clear water lakes with high resilience (low probability of switching to the turbidstate because nutrient loading and benthivorous fish biomass are low), some macrophyte removalcould occur in high use areas Management goals should be consistent with reality (Welch, 1992a).The following case histories were chosen from situations where the lake was intentionallymanipulated rather than from instances of unplanned and drastic biological changes such as a winterfish kill or a drought

9.6 BIOMANIPULATION: SHALLOW LAKES

Cockshoot Broad (3.3 ha, mean depth 1.0 m) is one of several small, riverine, shallow lakes ineastern England Originally they were macrophyte-dominated, but recently dense phytoplanktonreplaced macrophytes (Moss et al., 1996b) Aquatic plants in the UK are considered an asset,

FIGURE 9.6 Eutrophication and oligotrophication in relation to algal biomass, showing a typical hysteresis

curve (From Hosper, H 1997 Clearing Lakes An Ecosystem Approach to the Restoration and Management

of Shallow Lakes in The Netherlands RIZA, Lelystad, The Netherlands With permission.)

Nutrient load

providing high biodiversity, and efforts were made to rehabilitate some of these lakes, includingCockshoot Broad.

Consistent with conventional wisdom of the time (early 1980s), P removal at municipal water facilities was believed sufficient to rehabilitate Cockshoot Broad However, high internal Precycling was a major P source, and the Broad was therefore isolated from the adjacent nutrient-rich river, and about one meter of P-rich sediment removed TP fell, and aquatic plants returned.However, by 1984–1985, submersed plants declined and the phytoplankton-dominated state

waste-returned because Daphnia pulex, abundant following isolation and dredging, declined to small

numbers by 1984 as planktivorous fish populations recovered (Moss et al., 1986b)

Biomanipulation via nearly complete fish removal occurred in winter 1989 and 1990 Fishremoval in winter is easier because fish tend to aggregate at this time, making electro-fishing and

seining easier Maintenance fish removal continued in subsequent winters Daphnia returned,

chlorophyll concentration declined, and submersed macrophytes recolonized the broad High ent concentrations occurred in the macrophyte and phytoplankton-dominated conditions, indicating

nutri-that these alternative states were influenced by biological interactions When Daphnia were absent,

chlorophyll concentrations were highly correlated with TP There was no correlation in years of

high Daphnia–low planktivore densities.

Grazer control of phytoplankton appeared to be linked to macrophytes that served as physicalrefuges from planktivory (Timms and Moss, 1984; Moss et al., 1986b, 1994) Unlike deep lakes,where vertical migration can provide a daytime refuge for zooplankton from fish predation (Gliwicz,1986), shallow lake zooplankton may employ diel horizontal migration (DHM) to and from the

littoral zone to provide daytime refuge from planktivory While Daphnia appear to be chemically repelled by some macrophytes (e.g., Myriophyllum exalbescens), they use other macrophytes to

avoid fish (Lauridsen and Lodge, 1996; Burks et al., 2001) However, if littoral zones are dominated

by planktivores (including YOY piscivores), Daphnia mortality may be high (Perrow et al., 1999).

These authors suggested that if macrophytes comprise 30–40% of lake volume that is sufficientrefuge from fish for zooplankton to maintain clear water The clear-water stabilizing effect of DHMappears to be high when macrophytes are abundant and littoral-associated piscivores control plank-tivory (Burks et al., 2002) The clear water state may be possible, even at elevated nutrient

concentrations, when Daphnia grazing is extensive Additional studies on DHM are needed.

An assessment of zooplankton grazing in increasing transparency may be difficult to determineusing traditional sampling methods Only nighttime sampling reveals the actual density of zooplankton

in macrophyte-dominated shallow lakes Daytime open water sampling fails to capture zooplankton

in refuges (Meijer et al., 1999) Artificial refugia for large-bodied zooplankton in the English broads,including bundles of brush, strands of polypropylene rope, and mesh cages did not enhancezooplankton survival (Moss, 1990; Irvine et al., 1990)

9.6.2 LAKE ZWEMLUST (AND OTHER DUTCH LAKES)

The Lake Zwemlust (1.5 ha, mean depth 1.5 m) case history is instructive because of its long-termdata set, and because of problems in maintaining the clear water state after biomanipulation In

1968, a broad-spectrum herbicide (diuron) was applied, eliminating macrophytes There was a rapidshift to the turbid, algae-dominated state A minimum transparency of 1.0 m in swimming lakes

is required in The Netherlands, but blooms of Microcystis aeruginosa reduced transparency below

this criterion In winter, 1987, Lake Zwemlust was seined, electro-fished, and drained to eliminate

planktivorous and benthivorous fish It was then stocked with pike (Esox lucius) and rudd dinius erythrophthalmus), willow twigs were added as shelter for pike fingerlings, yellow water lily (Nluphar lutea) and Chara were planted, and Daphnia magna and D hyalina (1 kg wet weight)

(Scar-were introduced (Gulati, 1990; van Donk et al., 1990)

Though external nutrient loading remained high (2.4 g P/m2 per year; van Donk et al., 1993),

the water became clear In 1988–1989 Elodea nuttalli dominated, and phytoplankton became

N-limited Summer chlorophyll concentrations were low in 1987, the first summer of the treatment,

partly from intense Daphnia grazing By 1991, small-bodied Daphnia became dominant after planktivory resumed, and grazing on algae declined Ceratophyllum demersum was the dominant

macrophyte in 1990–1991, but was nearly absent in 1992–1994 Late summer algal blooms resumed

in 1992–1994 Potamogeton berchtholdii appeared in the spring of 1992–1994, but became covered

with epiphytes and was eliminated (van Donk and Gulati, 1995) Another planktivore removal tookplace in 1999, followed by a clear water period and a slow return of macrophytes

What caused the shift back to the turbid water condition? Herbivorous birds (coot, Fulica atra) invaded the lake in 1989, removing Elodea at rates up to 7 kg dry weight/day during late autumn.

Rudd are also herbivorous, and the two grazers (fish and birds) shifted plant dominance to coontail

(Ceratophyllum demersum) and Potamogeton Coot then grazed on Ceratophyllum in 1991–1992 Because Daphnia grazing was effective only in 1987, it may have been macrophytes that maintained

the clear water state from 1988–1991 (perhaps through alleleopathy, by inducing N-limitation ofphytoplankton, and/or by preventing sediment resuspension) After macrophyte elimination by fish,birds, and epiphytes, algal blooms returned to Lake Zwemlust (van Donk et al., 1994; van Donkand Gulati, 1995; Figure 9.7) Ultimately high nutrient loading prevented establishment of a perma-nent clear water state Only the turbid state appeared to be stable (van de Bund and van Donk, 2002).One factor causing loss of macrophytes is light limitation due to epiphyte coverage (Sand-Jensen and Søndergaard, 1981) This apparently occurred in Lake Zwemlust (van de Bund and vanDonk, 2002) Bronmark and Weisner (1992) proposed that biomanipulation of the littoral zonefood web to eliminate molluscivores (often Centrarchidae or sunfish), thereby maintaining biomass

of epiphyton-grazing snails, could play a role in stabilizing the clear water state This hypothesis

has gained support In nutrient-enriched aquaria, C demersum exhibited extensive production of new biomass at high snail (Physa, Helisoma) densities (coontail obtains its nutrients directly from

the water) High nutrient conditions with low or no snails permitted heavy epiphyton growth and

apparent light limitation of C demersum (Lombardo, 2001) In shallow U.K lakes, macrophyte

biomass was negatively correlated with periphyton biomass, which, in turn, was negatively

corre-lated with invertebrate biomass (r2= 0.714) Invertebrate density on plant surfaces was lower in

FIGURE 9.7 Schematic presentation of mechanisms buffering the stability (horizontal arrows) and inducing

the transition (vertical arrows) of the clear water state and turbid state in Lake Zwemlust (From van Donk,

E and R.D Gulati 1995 Water Sci Technol 32: 197–206 With permission.)

C.d.

P.b.

N – lim.

Zooplankton Macrophytes

+

+

−

−