encyclopedia of inland waters || lakes as ecosystems

Lakes as EcosystemsW M Lewis, University of Colorado, Boulder, CO, USA

ã 2009 Elsevier Inc. All rights reserved.

Introduction of the Ecosystem Concept

Scientific studies of lakes began as early as the seven-teenth century, but at first were descriptive ratherthan analytical. Toward the end of the nineteenth cen-tury, measurements and observations on lakes becamemore directed. For example, the thermal layering oflakes was attributed to specific physical causes, andsuch phenomena as the movement of plankton in thewater columnwere the subject of hypotheses that weretested with specific kinds of data.Comprehensive studies of lakes began with the

work of Alphonse Forel (1841–1912) on Lac Leman(Lake Geneva), Switzerland, as well as other Swisslakes. In a three-volume monograph (Le Leman:1892, 1895, 1904), Forel presented data on a widevariety of subjects including sediments and bottom-dwelling organisms, fishes and fisheries, water move-ment, transparency and color, temperature, andothers. Thus Forel, who introduced the term ‘limnol-ogy’ (originally the study of lakes, but later expandedto include other inland waters), demonstrated theholistic approach for understanding a lake as an envi-ronmental entity, but without application of anexplicit ecosystem concept.The conceptual basis for studying lakes as eco-

systems was first clearly given by Stephen Forbes(1844–1930; Figure 1) through a short essay, TheLake as a Microcosm (1887). Forbes was professorof biology at the University of Illinois in Champaign,IL, USA, and director of the Illinois Natural HistoryLaboratory (subsequently the Illinois Natural HistorySurvey), which was charged with describing and ana-lyzing the flora and fauna of Illinois. Forbes realizedthat it was not possible to achieve a full understandingof a lake or, by implication, of any other environmen-tal system such as a stream or forest, simply fromknowledge of the resident species. Forbes proposedthat the species in a particular environment, wheninteracting with each other and with the nonlivingcomponents of the environment, show collective (sys-tem) properties. Themicrocosm that Forbes describedtoday would be called an ecosystem, although thisterm did not come into use until 48 years later throughthe work of the British botanist A. G. Tansley. Bycurrent usage, an ecosystem is any portion of theEarth’s surface that shows strong and constant inter-actions among its resident organisms and betweenthese organisms and the abotic environment.

Forbes not only described accurately the modernecosystem concept, but also identified ways in whichcritical properties of ecosystems could be measuredand analyzed. He named four common propertiesof lakes, each of which provides a cornerstone forthe study of lakes and other ecosystems, as shownin Table 1. Although Forbes’s concepts have beenrenamed, they are easily visible in the modern studyof lakes as ecosystems.

Although the essay by Forbes now is considered aclassic in limnology and in ecology generally, itcaused no immediate change in the practices of lim-nologists or ecologists. Like many important discov-eries in science, it was a seed that requiredconsiderable time to germinate.

In studies of Cedar Bog Lake, Minnesota, for hisPh.D. at the University of Minnesota, RaymondLindeman (1915–1942; Figure 2) gave limnologiststhe clearest early modern example of the study oflakes as ecosystems. Rather than focusing on a par-ticular type of organism or group of organisms, whichwould have been quite typical for his era, Lindemandecided that he would attempt to analyze all of thefeeding relationships (‘trophic’ relationships) amongorganisms in Cedar Bog Lake. Thus, his Ph.D. workextended from algae and aquatic vascular plants toherbivorous invertebrates, and then to carnivores,and conceptually even to bacteria, although therewere no methods for quantifying bacterial abundanceat that time. Lindeman’s descriptions of feeding rela-tionships were voluminous but straightforward towrite up and publish, but he sought some more gen-eral conclusions for which he needed a new concept.

Lindeman took a postdoctoral position withG. Evelyn Hutchinson at Yale University in 1942.Hutchinson had become a limnologist of notethrough his quantitatively oriented studies of plank-ton and biogeochemical processes in the small kettlelakes near New Haven. He would in subsequentdecades become the world’s most influential lim-nologist, and part of his reputation grew out of hiscontributions to the field of biogeochemistry, animportant tool of ecosystem science.

Hutchinson made suggestions that no doubt werecritical to Lindeman’s groundbreaking paper, TheTrophic Dynamic Aspect of Ecology (1942, publishedin the journal Ecology), which now is recognized asa landmark in limnology and in ecology generally.

431

432 Biological Integration _ Lakes as Ecosystems

Lindeman proposed a way of converting the tremen-dous mass of highly specific information for CedarBog Lake into a format that would allow comparisonswith any other lake or even with other kinds of eco-systems. Building on the work of the German limnol-ogist August Thienemann and the British ecologistCharles Elton, Lindeman organized the feeding rela-tionships as a feeding hierarchywithinwhich each kindof organism was assigned to a specific feeding level(trophic level). He then proposed that the feeding rela-tionship represented by any given link in the food webbe quantified as an energy flow. Thus, the total energyflow from level 1 (plants) to level 2 (herbivores) couldbe quantified as the summation of energy flowsbetween all pairs of plants and herbivores; a similarestimate could be made for all other pairs of trophiclevels. In thisway, the flowof energy across the levels ofthe food web could be expressed in quantitative terms.The first important conclusion from Lindeman’s

energy-based approach was that each transfer of

Figure 1 Stephen A. Forbes. Reproduced from the Illinois

Natural History Survey website, with permission from the IllinoisNatural History Survey.

Table 1 Four key properties of ecosystems identified by S. A. Forbe

examples

Forbes concept Modern nomen

Web of interactions Food-web dyna

Building up and breaking down of organic matter Ecosystem met

Circulation of matter Biogeochemistr

Distribution of organisms along gradients Community org

energy between trophic levels is governed by thesecond law of thermodynamics, which requires thatsignificant energy loss must occur each time energyis transferred. Thus, Lindeman demonstrated whyfood webs have relatively few trophic levels: pro-gressive dissipation of energy as it passes throughthe food web from the bottom (plants) to the upperlevels (upper-level carnivores) ultimately providesinsufficient energy for expansion of the food webto further levels. Also, analysis of a food web inthis way sets the stage for calculating efficiencies ofenergy transfer, comparison of efficiencies acrossdifferent ecosystem types, and the identificationand analysis of bottlenecks restricting the flow ofenergy within the food web.

The contributions of G. Evelyn Hutchinson in the1940s on the biogeochemistry of carbon in lakes alsomust be counted as landmarks in the development ofecosystem science. Even so, ecosystem science wasscarcely represented in the research agenda of ecolo-gists or in the academic curriculum as late as 1950.Penetration of the ecosystem thinking into research,teaching, and public awareness occurred first throughthe publication of a textbook, Fundamentals of Ecol-ogy (1953), written by Eugene P. Odum, and espe-cially through the second edition of the same book(1959), written by E. P. Odum and Howard T. Odum.The Odums visualized ecology as best viewed fromthe top down, with ecosystems as a point of departureand studies of ecosystem components as infrastruc-ture for the understanding of ecosystems.

The ecosystem perspective has not displacedthe more specialized branches of ecology that dealwith particular kinds of organisms or specific kindsof physical phenomena, such as studies of water move-ment, optics, or heat exchange. Rather, ecosystem sci-ence has had a unifying effect on studies of ecosystemcomponents (Figure 3). Study of a specific ecosystemcomponent produces not only a better understandingof that component, but also a better understanding ofthe ecosystem, which is a final objective for the scienceof an ecosystem type, such as lakes.

s, along with their modern nomenclatural counterparts and some

clature Modern studies

mics Food-web complexity and efficiency

abolism Total ecosystem photosynthesis and respiration

y Dynamics and cycling of carbon, nitrogen, andphosphorus

anization Vertical and horizontal patterns in the distribution of

fishes, invertebrates, and algae

Atmosphere: deposition,

gas exchange

Food web

BenthosPlankton

Lake outflowWatershedinflow

Massinput and output:

water, organic matter,nutrients, etc.

Processes(metabolism,

nutrientcycling)

Figure 3 Ecosystem diagram of a lake.

Figure 2 R. L. Lindeman and his famous study site, Cedar Bog Lake, MN. Reproduced from People of Cedar Creek website, with

permission from Cedar Creek Natural History Area.

Biological Integration _ Lakes as Ecosystems 433

Metabolism in Lakes

The dominant anabolic component of metabolismin lakes is photosynthesis based on carbon dioxideand water plus solar irradiance as an energy source.The dominant catabolic component is aerobic respira-tion based on the oxidation of organic matter withoxygen as an electron acceptor. Conversion of solarenergy to stored chemical energy in the form of

biomass seldom reaches 1% efficiency in lakes becauseof losses inherent in the wavelengths that can beintercepted by photosynthetic pigments, inefficiencyin the interception process, and thermodynamic lossesin the conversions leading to the production of bio-mass. Respiration also involves thermodynamic losses.Therefore, the solar energy source greatly exceedsphotosynthetic output, and cellular capture of the


energy released from organic matter by respirationgreatly exceeds the energy stored in the organic matter.Aerobic photosynthesis is characteristic of the

aquatic vascular plants, attached algae, and phyto-plankton of lakes and is universal wherever light andoxygen are present. Aerobic respiration is character-istic of aerobic autotrophs as well as consumers andmost bacteria; it occurs wherever oxygen is present(Table 2). Other categories of metabolism occur undereither of twomore restrictive conditions: (1)where lightpenetrates into an anaerobic zone, and (2) where thereis an interface or mixing between oxidizing and reduc-ing conditions, as is common near a sediment–waterinterface.A study of whole-ecosystem metabolism would

require consideration of all of the metabolic cate-gories listed in Table 2. The dominance of aerobicphotosynthesis and aerobic respiration, however,often allows ecosystem studies to focus on these twometabolic components. In open water, photosynthesisand respiration are often measured with a verticalseries of paired transparent and darkened incubationbottles filled with lake water; rate of decline of oxy-gen in the dark bottles indicates respiration rate, andrate increase of oxygen in the transparent bottle indi-cates net photosynthesis. In unproductive lakes,uptake of 14C-labeled CO2 can be used as an evenmore sensitive indicator of photosynthesis. Wheremacrophytes or attached algae are important, as isoften the case in small lakes, separate measurementsmust be made of their photosynthesis, typically by theuse of enclosures. The metabolic rates of microbes indeepwater sediments can be inferred from the rate ofoxygen loss from the hypolimnion of a stratified lake,or can be measured with enclosures.Annual rate of photosynthesis and respiration per

unit area (typically given as mg C m�2 year�1) is themost commonly used metabolic statistic for lakes.A complete metabolic accounting would also include

Table 2 Summary of metabolic processes in lake ecosystems

Metabolic process Energy source Capabilities

Eukaryoticplants

Aerobic

photosynthesis

Solar Yes

Anaerobicphotosynthesis

Solar No

Aerobic

chemosynthesis

Inorganic oxidation No

Aerobic respiration Organic oxidation YesAnaerobic respiration Inorganic or organic

reduction

No

processing of organic matter entering a lake from thesurrounding terrestrial environment, primarilythrough stream flow.

Total annual net production of autotrophs (pro-duction in excess of respiration) is a measure of thecapacity for a lake to generate biomass at highertrophic levels (Figure 4).

Also, the time-course of production, which showsseasonal and nonseasonal variation, is often a usefulpoint of departure for the analysis of mechanisms thatcontrol biotic functions in a lake. Factors that maysuppress production include deep mixing of the watercolumn, exhaustion of key nutrients, grazing, andhydraulic removal of biomass.

Photosynthesis and respiration often respond dif-ferently to a specific physical or chemical change.

Food-Web Analysis

Modern food-web analysis follows the example givenby Lindeman. The sophistication of the analysis ismuch greater today than it was in 1942, however,and the uses of food-web analysis have diversified aswell.

At the base of the food web is organic mattergenerated within a lake or coming to a lake from itswatershed and atmosphere above (Figure 5). All ofthese sources of organic matter can be quantified, asexplained in the preceding section, but require theapplication of several methods and must take intoaccount both spatial and temporal variability in thesynthesis or transport of organic matter.

Quantification of linkages in the food web beginswith feeding relationships between autotrophs (pri-mary producers) and herbivores (primary consu-mers). Analysis of gut contents is one method forestablishing the linkage between a specific kind ofconsumer and one or more plant foods. A linkage

Occurrence conditions inlakes

Cyanobacteria Otherbacteria

Yes No Universal, photic, oxic

No Specialists Occasional, photic, anoxic

No Specialists Common, oxic/anoxic

interfaces

Yes Yes Universal, oxicNo Yes Common, oxic/anoxic

interfaces

Sunlight

Photo-synthesis

P

Imports

Respira-torymachinery of plants

Herbi- vores

Decom- posers

Carni- vores

Top c

Exports

Heat

R

Figure 4 Example of an early metabolic (energy-flow) diagram for an aquatic ecosystem. Reproduced from Odum H T (1956)

Limnology and Oceanography 1: 102–118, with permission from the American Society of Limnology and Oceanography.


drawn in this way has two disadvantages, however:(1) it is not quantitative because food items oftencannot be identified fully, and (2) it is subject to errorsof interpretation caused by the ingestion of foods thatare not assimilated or only partially assimilatedthrough the gut wall. The first problem can be over-come either by the use of feeding experiments or bythe quantification of growth rate of the consumer,with some empirically-based assumptions about thegrowth efficiency of the consumer. The second prob-lem can be resolved by experimental use of tissuelabels (typically isotopes) or, more efficiently, by theuse of stable isotopes as passive tracers (i.e., rely-ing on the natural abundance of stable isotopes toinfer food sources). Because primary producers ofdifferent categories may differ substantially in theirconcentration of the stable carbon isotope 13C, analy-sis of 13C content of protoplasm from the consumermay allow a quantitative estimation of the relativeimportance of several possible foods contributing tothe synthesis of biomass by the consumer.Above the level of primary consumers are carni-

vores, which may be secondary, tertiary, or even qua-ternary consumers, depending on their food source.

Measurements of growth rate, along with gut-contentanalysis and use of passive or active tracers, can beapplied to carnivores just as they are to herbivores.Within the carnivore trophic levels, however, theassignment of consumers to a specific trophic levelmay be difficult because carnivores often consumefoods belonging to more than one trophic level. Thestable nitrogen isotope 15N is useful in assigning aspecies to a fractional position on the food webbecause 15N shows increasing tissue enrichmentfrom one trophic level to the next.

When completed, a food-web analysis shows theefficiency of energy transfer from one level to thenext. Because of the thermodynamic limits on effi-ciency and the typical observed efficiency, which are amatter of record, a food-web analysis based on energyshows whether particular linkages are unusually weakor strongby comparisonwith expectations. Suchobser-vations in turn lead to hypotheses about mechanismsof control for energy transfer within the food web.For example, modern ecosystem theory includes theconcept of ‘trophic cascades’ involving ‘top-down’and ‘bottom-up’ effects on trophic dynamics, which iseasily applicable to lakes. Change in one trophic level

Autochthonous and allochthonous DOM

Viruses

Autotrophicpicoplankton

ProtozoaHeteronano-flagellates

Bacteria

BacteriaMixotrophic flagellatesCiliatesHeterotrophic dinoflagellatesAmoebaeHeliozoa

CrustaceaRotifers

CrustaceaRotifers

Net phytoplankton

Nano phytoplankton

AlgaeZoobenthos

Zooplankton

Top carnivores

Carnivorouszooplankton

Herbivorouszooplankton

Classic grazer food chain in the plankton

Benthic/littoral food chain Birds Fish

Microbial food web

Figure 5 A modern diagrammatic view of a lacustrine food web, including both microbial and nonmicrobial components

(from Weisse and Stockner, 1993 as modified by Kalff, 2002; with permission).


may be visible in other trophic level, in the manner of acascade (Figure 6). Top-down effects pass from anytrophic level to the next trophic level below. For exam-ple, an unusual abundance of algal biomass in a lakecould be traced to unusually efficient removal of herbi-vores through predation at such a rate as to leave algaemostly uneaten (a top-down effect).Similarly, bottom-up control passes from any tro-

phic level to the next level above. The inability of anoligotrophic (nutrient-poor) lake to grow substantialamounts of plant biomass, e.g., exerts a bottom-upeffect on all higher trophic levels by restricting theamount of energy that is available at the base of thefood chain.Trophic-dynamic analysis also leads to other fun-

damental questions involving the structure of bioticcommunities. For example, food webs might or mightnot be more productive or more efficient with a highdiversity of herbivores than with strong dominancefrom a very successful, specific type of herbivore.Such questions bear not only on the analysis of natu-ral ecosystems, but also on ecosystem management.

Biogeochemistry

Many of the functional attributes of lake ecosystemscan be analyzed through biogeochemical studies.While metabolism and trophic dynamics are viewedin terms of energy flux, biogeochemistry typically isviewed in terms of mass flux. As with energy analysis,the foundation is basic physics: mass is neither creatednor destroyed (under conditions that are compatiblewith the presence of life). The conservation of massleads to the mass-balance equation, which can be givenas follows for an ecosystem: I � O ¼ �S, where I isinput of mass of a particular type (carbon or phospho-rus, for example), O is output of mass, and DS ischange in storage of mass. Input and output pathwaysfor lakes are either hydrologic (flow) or atmospheric(deposition or gas exchange). Over the short term,change in storage may involve changes in concentra-tion of mass in the water column, but over the longterm, change in storage mostly reflects accumulationof mass in sediments. Any element can be a target forbiogeochemical studies, but the most frequently

Nutrient input rate

Res

pons

e of

pro

duce

rs

P, N

P, N

More responsiveto nutrients

Less responsiveto nutrients

Figure 6 Contrasts in nutrient response for lakes that have

strong control by top carnivores (lower line) and lakes that have

weak control by top carnivores (upper line). The top carnivore

exercises top-down control by suppressing smaller fish that eatlarge grazing zooplankton. Survival of more large zooplankton

holds phytoplankton in check, thus weakening response to plant

nutrients (P, N). Reproduced from Carpenter S R (2003) Regime

Shifts in Lake Ecosystems, with permission from the InternationalEcology Institute, Oldendorf/Luhe, Germany.

Inorganic Organic

Dissolved

Particulate

CO2, H2CO3, HCO3−

CaCO3

Nonliving debris,organisms

Humic and fulvicacids from soils

Protein, carbohydrate,and nucleic acid

breakdown productsfrom organisms

Figure 7 Compartment matrix for analysis of carbon cycling in

the water column of a lake.


studied biogeochemical processes in aquatic ecosys-tems involve carbon, phosphorus, and nitrogen.The mass-balance equation applies not only to the

entire ecosystem, but also to compartments withinthe ecosystem (Figure 7). For example, carbon in thewater column could be partitioned into compart-ments, each ofwhichwould have its ownmass-balanceequation. Some of these compartments would havea high turnover rate, while others would not. Thedynamics of compartments explain restrictions on spe-cific biological processes such as photosynthesis, andaccount for differences in individual lakes or categoriesof lakes, such as those that differ greatly in nutrientsupply.Studies of the carbon cycle are ideal complements

to studies of lake metabolism and food webs. CO2,which is the feedstock for photosynthesis, entersaquatic autotrophs either in the form of free CO2

(CO2þH2CO3) or bicarbonate (HCO3�). CO2 is con-

verted to organic matter by the reduction process ofphotosynthesis. Either within an autotroph orthrough consumption of autotroph or other biomassby consumers or decomposers, reduced carbon isreconverted to its oxidized form as CO2 throughaerobic respiration.

Organic carbon resides not only within organisms,but also in water or sediments in dissolved or nonliv-ing particulate form. Dissolved organic carbonderives from the watershed, atmosphere, or organ-isms within the lake. Watershed contributions to dis-solved organic carbon in lakes are composed to alarge extent of humic and fulvic acids, which are thebyproduct of the degradation of organicmatter withinsoils. Humic and fulvic acids are refractory (resistantto breakdown), although they are slowly decomposedby microbes and can be broken down by ultravioletlight in a water column. Humic and fulvic acids aregenerally present in concentrations of 1�10mg l�1 inlakes and, when present at concentrations above5mg l�1, generally impart a brown or orange colorto the water.

Organic compounds that are released to the watercolumn of lakes by the resident organisms varygreatly in composition. All soluble organic com-pounds present in organisms can be found in thewater column at measurable concentrations. Thus, acareful analysis of lake water would show a widevariety of amino acids, carbohydrates of varyingcomplexity, and other metabolites of organisms.These compounds enter the water column throughleakage (excretion), death, or the production of fecalmatter. As might be expected, the turnover rate formetabolites is generally high because most of thesecompounds are labile (easily used), in contrast tohumic and fulvic acids.

Particulate organic matter present in a water col-umn may be either living or nonliving. Frequently,these two compartments are physically joined, as allnonliving particulate organic matter in water is colo-nized by bacteria and, under some conditions, fungi.


Most of the carbon attributable to living organisms isaccounted for by phytoplankton and zooplankton;fish and bacteria make smaller contributions in thesense of mass but have important ecosystem effects.Carbon is continually stored in sediments, which accu-mulate at rates often about 1mm per year, much ofwhich is organic. As sediments become buried by addi-tional sediments, their decomposition slows because ofthe lack of oxygen and chemically hostile conditionsfor active metabolism of bacteria. Thus, while someof the organic matter in sediments is remobilized intothe water column, other organic matter in the sedi-ments becomes long-term storage.A study of mass flux and storage, when viewed in

terms of ecosystem compartments and subcompart-ments, tells a story about the mechanisms by which alake functions. For example, lakes may differ greatlyin the terrestrial contribution to carbon processingand carbon storage, and also may differ greatly inspeed of carbon turnover in specific living and non-living compartments.The cycling of phosphorus in lakes has been studied

intensively, because phosphorus is one of the two ele-ments (nitrogen is the other) most likely to be criticallydepleted by aquatic autotrophs. Thus, phosphorus atthe ecosystem level is commonly viewed as an ecosys-tem regulator; lakes in which it is abundant (eutrophiclakes) have potential to produce high amounts of plantbiomass (phytoplankton, attached algae, or aquaticvascular plants) and have water-quality characteristicsthat are often viewed as undesirable (low transparency,green color, potential for odor production, and severeloss of oxygen in deep water). In contrast, lakes havinglow supplies of phosphorus (oligotrophic lakes) mayshow constant suppression of plant growth throughphosphorus scarcity. Such lakes have much loweramounts of carbon in the water, higher transparency,often appear blue, and retain oxygen in deep waterconsistent with the requirements of fish and othereukaryotes.Because the phosphorus requirement of plants is

only approximately 1% of dry biomass, scarcity ofphosphorus can be offset by relatively modestincreases in the phosphorus additions to a lake. Addi-tion of 1 kg of phosphorus per unit volume or area,e.g., could easily generate 100 kg of dry mass or500 kg of wet mass of autotrophs. Thus, mobilizationof phosphorus by humans has the potential to changelakes and has done so throughout the world whereverhuman populations liberate phosphorus throughwaste disposal, agriculture, and disturbance of soil.For this reason, the study of trophic state (nutrientstatus) has received more attention than any otherecosystem feature of lakes. The practical applicationof modeling or analysis of trophic status in lakes arises

through the desire to prevent changes in trophic stateor to reverse changes in trophic state that have alreadyoccurred, which requires ecosystem-level understand-ing of the lake and particularly of its nutrient budgets.

Lakes may also be limited by low concentrations ofthe forms of inorganic nitrogen that are readily avail-able to aquatic autotrophs (ammonium, nitrate). Inthis case, nitrogen limitation rather than phosphoruslimitation may control plant growth. In fact, the twotypes of nutrient limitation may occur sequentiallyacross seasons or across years in a single lake.

Nitrogen cycling in lakes is much more compli-cated than is phosphorus cycling because nitrogenhas a gaseous atmospheric component that phospho-rus lacks, and because nitrogen exists in seven stableoxidation states within a lake, which sets the stagefor the use of nitrogen as a substrate for oxidationreduction reactions supporting microbial metabo-lism (Table 2). Like phosphorus, nitrogen enterslakes through the watershed and, in small amounts,with precipitation. Unlike phosphorus, nitrogen alsoenters lakes as a gas by diffusion at the air–waterinterface in the form of N2. N2 is so inert chemicallythat it cannot be used as a nitrogen source by mostorganisms. Certain prokaryotes have the ability touse N2 by converting it to ammonium, which thencan be used in organic synthesis, provided that asubstantial energy supply is available. This processis called nitrogen fixation, in that it converts thegaseous nitrogen to a solid that is soluble in water(ammonium). The dominant nitrogen fixers in lakesare the cyanobacteria. Not all cyanobacteria fix nitro-gen, but certain taxa that have a specialized nitrogen-fixing cell (heterocyst) grow commonly in lakes thatshow nitrogen depletion.

Nitrogen fixers escape the limitation of growth asso-ciated with nitrogen depletion, thereby gaining anadvantage over other autotrophs. Thus, lakes thatare enriched with phosphorus but showing nitrogendepletion often have nuisance growths of nitrogen-fixing cyanobacteria, which produce all of the expectedsymptoms of nutrient enrichment, and sometimes alsoproduce toxins. There is currently much interest inpredicting and preventing the development of circum-stances that lead to large and persistent blooms ofnitrogen-fixing cyanobacteria.

Nitrogen is a multiplier element, just as phospho-rus is. It constitutes approximately 5% of dry mass inplants. Therefore, when it is added to lakes, providedthat phosphorus is added at the same time, it supportsa 20-fold multiplication of dry plant biomass.

A detailed documentation of the carbon, nitrogen,and phosphorus cycles for a lake produces an under-standing of factors that regulate the metabolic rates oflakes and the accumulation of biomass of various


kinds in lakes. These ecosystem phenomena havenumerous practical applications, ranging from inter-est in biomass production (e.g., fish) to interest inconstraining water quality within boundaries thatare either natural or that favor human purposes.

Community Organization

As foretold by Forbes, ecosystems have a strongdegree of spatial organization involving the arrange-ment of organisms according to abotic constraintsbased on factors such as amount of irradiance, con-centration of dissolved oxygen, or substrate charac-teristics. The requirement for irradiance is especiallyimportant because it dictates the distribution andgrowth potential for aquatic autotrophs. Phytoplank-ton grow strongly only within the euphotic zone,which corresponds approximately to water receivingat least 1% of the solar irradiance available at thewater surface. Thus, vigorously growing phytoplank-tons often are confined to the upper part of the watercolumn (epilimnion and sometimes metaliminon).The same principle applies to periphyton and rootedaquatic plants, which can occupy substrates only overportions of the sediment that are exposed to at least1% surface irradiance. The gradient of irradiance iscontrolled by the transparency of the water.Spatial organization of animal communities and

microbial communities is dictated partly by physicalconditions and partly by the distribution of autotrophs.Crustacean zooplankton, e.g., may migrate large dis-tances vertically in the water, hiding in deep waterduring the daytime but rising to feed within surfacewaters at night. The zooplankton of the littoral zonediffers from that of open water, in that the littoralzone offers food types (periphyton especially) that arenot available in open water. Distribution of fishes maybe dictated in part by the presence of structure asso-ciated with littoral zone. The effect of structure mayeven be related to life history in that immature or smallfish may seek the structure of the littoral zone forshelter from predation.Because organisms conduct the business of an

ecosystem, an understanding of their habitat require-ments, reflected as spatial organization, leads toexplanations of the total abundances of certain cate-gories of organisms, the failure of certain organismsto fare well in one lake but not in another, and theconsequences of habitat disturbances of natural orhuman origin.

Synthetic Analysis of Ecosystems

Synthetic analysis of ecosystems often involves thestatistical study of empirical information, particularly

if the objective is to test a particular hypothesis.A quantitative overview of multiple ecosystem func-tions or of the intricate detail for an ecosystem com-ponent typically involves modeling as a supplementto other types of quantitative analysis. When compu-ters first became available, it was anticipated thatecosystem science would ultimately be driven almostentirely by mechanistic models that would predictecosystem responses to natural or anthropogenic con-ditions. These expectations were not realized. Ecosys-tems, like other complex systems such as climate oreconomics, aremoderately predictable frommodelingthat combines a modest number of well-quantifiedvariables addressed to a specific question, but typi-cally are unpredictable on the basis of large number ofvariables addressing more general questions. Even so,modeling of numerous variables in an ecosystem con-text can be useful in setting limits on expected out-comes or showing possible outcomes of multipleinteractions that cannot be easily discerned from thestudy of individual variables. Therefore, modeling isuseful in promoting the understanding of ecosystems.

Conclusion

Lakes first inspired the ecosystem concept, and havebeen a constant source of ideas about ecosystemstructure and function. Ecosystem science as appliedto lakes, is supported by and is consistent with, otherkinds of ecological studies that are directed towardspecific organisms, groups of organisms, or specificcategories of abiotic phenomena in lakes. The strengthof ecosystem science lies in its relevance to the under-standing of all subordinate components in ecosys-tems, and to its often direct connection to humanconcerns in understanding and managing lakes.

Glossary

Biogeochemistry – Scientific study of the mass flux ofany element, compound, or group of compoundswithin or across the boundaries of an ecosystem orany other spatial component of the environment.Mass flux within an ecosystem is often designatedas nutrient cycling or element cycling.

Cycling – A biogeochemical term referring to themovement of elements or compounds within an eco-system or any other bounded environmental system.

Ecosystem – Any portion of the earth’s surface thatshows strong and constant interaction among itsresident organisms and between those organismsand the abotic environment.


Lake trophic state – Fertility of a lake, as measuredeither by its concentrations of key plant nutrients(especially phosphorus and nitrogen) or the annualproduction of plant biomass (aquatic vascularplants and algae). Trophic-state categories includeoligotrophic (weakly nourished), mesotrophic(nutrition of intermediate status), and eutrophic(richly nourished).

Mass flux – Movement of mass per unit time, often inecosystem studies expressed as mass/area/time.

Stable isotope – Any isotope of an element that doesnot decay spontaneously.

Trophic dynamics – Fluxes of energy or mass causedby feeding relationships, including rates of grazingby herbivores on plant matter, rates of predation bycarnivores on other animals, or rates of decomposi-tion of organic matter by microbes.

See also: Carbon, Unifying Currency; Microbial FoodWebs; Modeling of Lake Ecosystems; Nitrogen;Phosphorus; Regulators of Biotic Processes in Streamand River Ecosystems; Trophic Dynamics in AquaticEcosystems.

Further Reading

Belgrano A, Scharler UM, Dunne J, and Ulanowicz RE (2006)

Aquatic Food Webs: An Ecosystem Approach. New York:Oxford University Presss.

Carpenter SR (2003) Regime Shifts in Lake Ecosystems: Patternand Variation.Oldendorf/Luhe, Germany: International Ecology

Institute.Golley FB (1993) A History of the Ecosystem Concept in Ecology:

More than the Sum of the Parts. New Haven/London: Yale

University Press.Kalff J (2002) Limnology: Inland Water Ecosystems. Upper Saddle

River, NJ: Prentice Hall.

encyclopedia of inland waters || lakes as ecosystems

Documents