The Role of Benthic Invertebrate Species in Freshwater Ecosystems

February 1999 119

Small invertebrates are function-ally important in many terres-trial and aquatic ecosystems

(Wilson 1992 Freckman et al 1997Palmer et al 1997 Postel and Car-penter 1997) In freshwater sedi-ments benthic invertebrates are di-verse and abundant but they areoften patchily distributed and rela-tively difficult to sample especiallywhen they live in deep subsurfacesediments Thus the species richnessand functional importance of fresh-water benthic invertebrates gener-ally go unnoticed until unexpectedchanges occur in ecosystems Unan-ticipated changes in freshwater eco-systems are often due to alterationsin the complex connections amongsediment-dwelling species and asso-ciated food webs (eg Goedkoopand Johnson 1996 Lodge et al1998b Stockley et al 1998) or todisturbances such as floods ordrought (eg Covich 1993 Power1995 Johnson et al 1998) that alterthe species composition of thebenthos In addition benthic speciescan themselves constitute a distur-

bance such as when they transmitdiseases For example certain benthicinvertebrate species (eg Tubifex tu-bifex) serve as parasite-transmittingvectors if these invertebrates increasein abundance in stream sedimentsthey may spread a lethal disease totrout causing trout populations todecline (Brinkhurst 1997) Fish killsmay also occur because of increasedaccumulation of nutrients whichcause formation of toxic algalblooms deoxygenation of deeperdensity-stratified waters and high con-centrations of ammonia or hydrogensulfide (Covich 1993)

The bottom muds of lakes andstreams may at first glance appear tobe uniform and therefore unlikelyhabitats for high biodiversity How-ever physical chemical and bio-logical processes create significanthorizontal and vertical heterogene-ities in the substrata (Figure 1) thatprovide a physical template for dis-tinct niches (Hutchinson 1993)These sedimentary processes includechanges in direction and rates of

flows differential deposition of sedi-ment grain sizes and dead organ-isms growth and death of rootsburrowing and sediment reworkingand fecal production by benthic con-sumers Microhabitats are also cre-ated by chemical gradients andmicrozonation in concentrations ofdissolved oxygen hydrogen sulfideammonia phosphorus and othercritical chemicals (Groffman andBohlen 1999) Colwell (1998) em-phasizes that such ldquobiocomplexityrdquoof habitats and biological relation-ships is an important aspect ofbiodiversity Bioturbation and otherbiotic interactions create extensivebiocomplexity in freshwater sedi-ments (Charbonneau and Hare1998) These biocomplexities mustbe better understood if clean drink-ing water and recreational uses offresh waters are to be maintainedScience-based policies require an eco-system perspective on the multiple rolesof many diverse benthic species

Previous studies have often dealtwith the ldquogoodsrdquo produced bybenthic species such as the quantityof prey items consumed by fish Thesegoods are clearly important compo-nents of food webs but how theirfunctional relationships respond tochanges in species composition arealso important In this article wehighlight examples of how some spe-cies have a disproportionately largeimpact on food-web dynamics andhow particular species provide es-sential ecosystem services These eco-system functions include sedimentmixing nutrient cycling and energyflow through food webs

The Role of Benthic InvertebrateSpecies in Freshwater EcosystemsZoobenthic species influence energy flows and nutrient cycling

Alan P Covich Margaret A Palmer and Todd A Crowl

The integrity of thefreshwater supply

depends on how variousbenthic species make

their living andcontribute to complex

food webs

Alan P Covich (e-mail alanccnrcolostateedu) is a professor in the Depart-ment of Fishery and Wildlife Biology atColorado State University Fort CollinsCO 80523 Margaret A Palmer (e-mailmp3 umailumdedu) is a professor inthe Department of Biology at the Univer-sity of Maryland College Park MD 20742Todd A Crowl (e-mail facrowlccusuedu) is an associate professor in the De-partment of Fisheries and Wildlife and theEcology Center at Utah State UniversityLogan UT 84322 copy 1999 American In-stitute of Biological Sciences

120 BioScience Vol 49 No 2

Diversity of benthicfreshwater communities

Freshwater benthic species evolvedfrom many phyla over millions ofyears and represent a rich fauna Inthe fourth and last volume of A Trea-tise on Limnology G EvelynHutchinson (1993) reevaluated thequestion he first posed 40 years agomdashldquoWhy are there so many kinds ofanimalsrdquomdashbut in the context of thezoobenthos Hutchinson (1993) con-cluded that ldquothe Diptera are by farthe most diverse order of insects infresh water they are in fact the mostdiversified of any major taxon offreshwater organismsrdquo He estimatedthat more than 20000 Dipteran spe-cies breed in fresh water worldwideapproximately four times the num-ber of Coleoptera Others estimatethat there are large numbers ofbenthic species of protozoa crusta-cea and other groups (Palmer et al1997) Moreover systematists esti-mate that only a small percentage ofcertain taxa (eg freshwater nema-todes) have been described Diverseforms are continuously discoveredespecially in deep groundwaters inwhich regional endemics reflect iso-lation and evolutionary adaptations

to specific conditions (eg Holsinger1993) Many species still remainundescribed both taxonomically andecologically (Hutchinson 1993Palmer et al 1997) Protecting di-verse benthic communities will re-quire more thorough understandingof long-term functional relationshipsamong these species in an ecosystemcontext

Importance of individualspecies in ecosystem processesIt is evident from studies of terres-trial species that the number of spe-cies per se is not necessarily relatedto rates of ecosystem production(eg Chapin et al 1997 Tilman etal 1997) Instead each species isadapted to function under variableconditions with different speciesbeing of different relative importanceto particular ecological processesChanges in distributions and abun-dances of one species can result indisproportionate and unexpectedresponses by other species as theyattempt to compensate functionallyfor changes in the associated species(Frost et al 1995 Naeem 1998)

Recently Palmer et al (1997) pro-posed that particular benthic species

are especially important for deter-mining how organic matter is pro-cessed in freshwater ecosystemsThey described the diversity and eco-logical roles of freshwater benthicspecies and the major processes thatthese species influence in freshwaterecosystems In this article we sum-marize several recent studies thatshow how specific zoobenthic spe-cies alter ecosystem processes Weemphasize that zoobenthic speciesespecially crustaceans influence bothenergy flow through freshwater foodwebs and nutrient cycling (Figure 2)We suggest that in some cases thepresence or absence of a single spe-cies can dramatically alter ecologicalprocesses such as rates of grazingand decomposition

In benthic communities evenclosely related species may obtaintheir food resources differently Con-sequently species are anticipated todiffer in the ways or rates at whichthey perform a distinct ecosystemservice (eg acting as primary pro-ducers herbivores predators ordetritivores) Although we highlightonly a few examples there are nu-merous food-web linkages in whichone species interacts positively ornegatively with others or in whichthe addition or loss of a single spe-cies alters food-web dynamics Basedon current information about theseparation of niches among benthicspecies we conclude that differentspecies of sediment-dwelling macro-invertebrates are unlikely to be in-terchangeable components in manycomplex ecosystem processes

Diversity and speciesredundancy in ecosystemsLinkage of niche theory to trophicdynamics led to the ldquorivet hypoth-esisrdquo (Ehrlich and Ehrlich 1981)which postulates that each specieshas the potential to perform an es-sential role in the persistence of thecommunity and the ecosystem andthat some species may remain as thesole representatives of a particularfunctional group (Ehrlich and Walker1998)

Although it is clear that at somelevel each species is unique overlapin resource use among species is notunusual especially in freshwater foodwebs For decades ecologists have

Figure 1 Benthic macroinvertebrates burrow deeply into layered sediments andaccelerate nutrient cycling Burrowing bivalves crayfish tubificid worms and aquaticinsect larvae mix the sediments aerate deeper layers of sediments and increase rates ofrecycling of macronutrients (nitrogen phosphorus and organic carbon) and micronu-trients (trace elements) by bioturbation and fecal production Mysid shrimp amphi-pods and gastropods enhance microbial growth and nutrient cycling through theirmixing of surface sediments and breakdown of organic detritus

February 1999 121

questioned how much overlap in re-source use can persist over timeamong competing species Recentlythis question has been rephrased toask if and under what conditionsthe functional roles of each speciesare necessary for ecosystem processesto persist This ldquoredundancy hypoth-esisrdquo predicts that not all species areequally necessary at any one time forecosystem processes to continue(Lawton and Brown 1994 Ehrlichand Walker 1998) If some specieswere ldquoredundantrdquo in terms of theirfunctional relations then their losswould not result in observablechanges in energy flow or nutrientcycling The concept of ldquoparallel re-dundancyrdquo used in engineering analy-sis for system reliability is likely to beapplicable for comparing speciesrsquo rolesin ecosystem studies (Naeem 1998)

The redundancy hypothesis canbe broken down into three subhypo-theses The ldquofunctional group hy-pothesisrdquo predicts that as long as onespecies from each functional groupis present ecosystem processes willcontinue The ldquotrophic-level hypoth-esisrdquo predicts that as long as thebiomass or turnover of organisms ateach trophic level remains relativelyuniform and is independent of spe-cies composition energy flow andecosystem processes will persist Fi-nally the ldquokeystone species hypoth-esisrdquo predicts that not all species areof equal functional importancerather only a few are truly necessaryfor ecosystem processes eventhrough these species may not beabundant

Species redundancy infreshwater sedimentsAs mentioned above essential infor-mation on the unique contributionsmade by individual benthic species islacking Moreover there is insuffi-cient information about how indi-vidual zoobenthic species interactwith one another under the dynamicrange of natural conditions in fresh-water sediments Nevertheless fromdetailed field observations it appearsthat redundancy in many freshwaterbenthic ecosystems is low For ex-ample numerous zoobenthic speciesoccupy particular microhabitatsalong stream channels or at variousdepths in lakes (eg Hutchinson

1993) and at various times of year(eg Cummins et al 1989) Thesespatial and temporal distributionssuggest that benthic species have dif-ferent preferences for particularranges of temperature pH currentvelocity and types of substrata Colo-nization studies of streams and riv-ers also suggest that there are impor-tant differences in preferred use ofmicrohabitats (Milner 1987 Malm-qvist et al 1991) These differencesin the ability of species to disperse toand live in certain microhabitats be-come especially important after ma-jor disturbances when species abun-dances and community structure mayshift From these observations weinfer that rates of ecosystem process-ing may change after a major distur-bance because species compositionoften changes

Different spatial patterns of dis-tribution have formed a basis forgeneralizations about functional re-lationships of zoobenthic species indifferent freshwater ecological pro-cesses One example is the RiverContinuum Concept (Vannote et al1980) which relates sources of en-ergy and the dominant ecologicalmode by which energy is obtained in

headwater and tributary streams tothe types of consumers distributedalong a drainage network (Cumminset al 1995) According to this con-cept particular groups of benthicconsumers use different sources ofenergy such as riparian leaf litter orin-stream plant productivity fromalgae or macrophytes (Wallace andWebster 1996 Parkyn et al 1997Wallace et al 1997) Certain speciesof aquatic insects that live in smallheadwater streams use specializedmouthparts or feeding appendagesto break up large pieces of organicdetritus into smaller fragments Inthe process of feeding some shred-ded and suspended fragments aretransported downstream (along withfecal pellets) Other species are spe-cialized to filter out variously sizedparticles and are typically locateddownstream from the shredders Suchspecializations suggest that the lossof some pivotal species such as shred-ders would alter food availabilityfor suspension feeders and therebyalter ecosystem processing of detri-tal carbon However experimentalstudies on the roles of single speciesare generally lacking for stream eco-systems (Heard and Richardson

Figure 2 The roles of benthic macroinvertebrates in cycling nutrients and controllingnutrient outflows from ecosystems The benthos transforms organic detritus fromsedimentary storage into dissolved nutrients that can be mixed into overlying watersand used by rooted plants (macrophytes) and algae (phytoplankton) to enhance primaryproductivity Some benthic species are omnivores and feed on macrophytes algae andzooplankton Many benthic species are consumed by fishes Through their mixing ofsediments and consumption of diverse resources benthic invertebrates can directly andindirectly influence microbial production and release of greenhouse gases (CO2 andCH4) toxic gases (H2S and NH4) and nitrogen (N2)

122 BioScience Vol 49 No 2

1995) An integration of empiricaland theoretical studies is essential ifthe linkages of benthic speciesrsquo rolesto freshwater ecosystem processesare to be better understood

Roles of benthic speciesin ecological processesBenthic species perform a variety offunctions in freshwater food websFirst as already described benthicinvertebrates provide essential eco-system services by accelerating detri-tal decomposition (van de Bund etal 1994 Wallace and Webster 1996)Dead organic matter is one of themain sources of energy for benthicspecies in shallow-water habitats(Covich 1988 Hutchinson 1993Wallace and Webster 1996) Benthicinvertebrates are estimated to pro-cess 20ndash73 of riparian leaf-litterinputs to headwater streams Sec-ond benthic invertebrates releasebound nutrients into solution by theirfeeding activities excretion andburrowing into sediments (Figures 1and 2) Bacteria fungi algae andaquatic angiosperms can quickly takeup these dissolved nutrients acceler-ating microbial and plant growth(van de Bund et al 1994 Cumminset al 1995 Pelegri and Blackburn1996 Wallace et al 1997) This in-creased growth of benthic microbesalgae and rooted macrophytes is inturn consumed by herbivorous andomnivorous benthic invertebrates(Creed 1994 Lodge et al 1994Nystrom et al 1996 Cronin 1998)Third many benthic invertebratesare predators that control the num-bers locations and sizes of theirprey (Crowl and Covich 1990 1994)Fourth benthic invertebrates supplyfood for both aquatic and terrestrialvertebrate consumers (eg fishesturtles and birds) Finally benthicorganisms accelerate nutrient trans-fer to overlying open waters of lakes(eg Lindegaard 1994 Threlkeld1994 Blumenshine et al 1997Clarke et al 1997) as well as toadjacent riparian zones of streams(eg Covich et al 1996 Johnsonand Covich 1997 Naiman andDeacutecamps 1997 Wallace et al 1997)

The extent of understanding ofthe effects of benthic organisms onfreshwater ecosystem processes var-ies with the type of freshwater sys-

tem (ie streams lakes and wet-lands) For example much more isknown about how benthic species ofaquatic insects and other consumersinfluence detrital processing instreams than about how they do soin lakes or wetlands (Hutchinson1993 Wallace and Webster 1996Rosemond et al 1998) Species-spe-cific linkages are known to enhancealgal growth and productivity(Dodds 1991 Pringle et al 1993)and field experiments are beginningto show that benthic macroinverte-brates have species-specific roles inprocessing organic matter For ex-ample one species of freshwatershrimp can process leaf litter fasterthan another shrimp species in aninsular tropical headwater stream(see discussion below) Althoughboth shrimp are detritivores they donot substitute completely for oneanother in leaf-detrital processingand nutrient cycling In streamreaches where both of these shrimpspecies co-occur (Covich andMcDowell 1996) their interactionsand different modes and rates ofleaf-litter processing may enhanceeach otherrsquos effectiveness

Different benthic species alterrates of decompositionIf sufficient dissolved oxygen andappropriate substrata are availablethen many species of benthic organ-isms especially insects and crusta-ceans can accelerate microbial pro-cessing of dead organic materialBecause many species process or-ganic detritus most freshwater ecolo-gists have generally categorized theseconsumers into functional feedinggroups (Cummins et al 1995) or feed-ing guilds (Hawkins and MacMahon1989) To simplify data collection andanalyses most investigators ldquolumprdquospecies making the assumption forexample that those with similar feed-ing appendages or mandibular mor-phology generally perform similarroles in processing leaf litter (Merrittand Cummins 1996)

Although ecologists still disagreeabout how to best categorize differ-ent species (eg Wallace and Webster1996) it is widely agreed that shred-ders feed by tearing up large piecesof microbially conditioned leaf de-tritus with specialized mouth parts

whereas scrapers feed on attachedalgae or ldquobiofilmsrdquo of bacteria andalgae However some species ofscrapers also consume bacteria andfungi from fresh and decomposingleaf surfaces (Kornijow et al 1995)As species scrape and shred coarseplant litter in the process of obtain-ing their food they convert coarselitter into fine particulates Collec-tors filter suspended organic par-ticulates from flowing waters or fromsmall water-filled spaces within thesediments Although these functionalclassifications are useful for somestudies they can obscure importantfood-web dynamics that result fromdifferences among individual speciesand changes in feeding behavior un-der specific conditions

Additions of benthic species to foodwebs Given the general lack of em-pirical and theoretical work on theroles of single species in freshwaterecosystems one way to glean con-ceptual insights may be to considerrange extensions of benthic speciesinto additional habitats as a sourceof information on the role of par-ticular species in ecological processesAlthough we do not advocate mov-ing species around it is essential tomonitor the ecosystem-level conse-quences of any movements that dooccur The arrival of an additionalspecies is often associated with theloss of one or more resident species(Lodge et al 1998b) However it isalso important to ask about the ef-fects on ecosystem processes Becausenative species are generally welladapted to local conditions move-ments of additional species into fresh-water assemblages can sometimesalter energy flow and change nutri-ent cycling

Although many range extensionsand introductions are transient thespread of some benthic species ispersistent Successful invaders oftenhave biotic attributes that predis-pose them to have major impacts onhighly variable ecosystems Identify-ing these invasive characteristics maybe useful for better understandinghow resident native benthic speciesfunction in an ecosystem contextThese attributes include aggressiveuse of food resources rapid repro-duction larvae that are well adaptedfor dispersal or resistant resting

February 1999 123

stages that survive in muds for manyyears Intensive monitoring studiesare just beginning to reveal the eco-logical traits that account for somebenthic speciesrsquo abilities to extendtheir ranges and to alter ecosystemprocesses Well-designed field experi-ments are needed to identify uniqueroles of both invasive and non-invasive species before the conse-quences of losing resident nativespecies can be fully understood

Losses of benthic species in foodwebs Loss of some species will likelyalter or degrade critical ecosystemprocesses because of the unavailabil-ity of replacement species Althoughthe exact consequences of each spe-ciesrsquo loss cannot be predicted Wil-son (1992) noted that if one speciesafter another were lost from an eco-system then at some point the eco-system would likely change drasti-cally Once species are lost the costsfor maintaining natural ecosystemswith engineering processes would beprohibitively expensive (Postel andCarpenter 1997) If at least one spe-cies were to remain in each func-tional group and the rate of process-ing by that species were sufficientlyhigh then in theory at least ecosys-tem processes should continue How-ever because environmental condi-tions change over time populationsof some of these remaining specieswould most likely become locally ex-tinct disrupting ecosystem process-ing Consequently ecosystems com-posed of a bare minimum of species ina fluctuating environment probablycould not continue to function overtime merely by compensating for thelosses of some species with increaseddensities biomass or processingrates of the few remaining species

Freshwater ecosystemprocessing by crustaceansSeveral studies have shown that crus-taceans play important roles instream and lake food webs We out-line these studies as examples to stimu-late additional field studies and toemphasize the need for more effectiveconceptualization of cross-linkagesamong different benthic species

Detritivorous shrimp An exampleof how individual species can alter

detrital processing comes from a re-cent experimental manipulation ofdecapods in a tropical headwaterstream in the Luquillo ExperimentalForest in Puerto Rico This rain for-est is one of the sites in the NationalScience Foundationrsquos Long TermEcological Research (LTER) Pro-gram The Luquillo LTER is focusedon the effects of disturbances such ashurricanes and drought on forest andstream ecosystems (Covich et al1996) Leaf litter was manipulatedin a series of pools to identify theeffects of two shrimp species on leaf-detrital processing The stream poolsat this elevation (550 m) are locatedabove several high waterfalls whereno fish predators occur (Todd Crowland Alan Covich unpublished data)the food web is relatively simple andis dominated by several decapod spe-cies (Covich and McDowell 1996)By clearing pools of all naturallyoccurring leaf litter and macroin-vertebrates it was possible to mea-sure detrital processing by each de-capod species separately in responseto additions of leaf litter from asingle riparian tree species Leaf lit-ter from Cecropia schreveriana (anearly successional tree that oftencolonizes disturbed riparian habi-tats after mud slides and hurricanes)was added back to the cleared poolsalong with either of two naturallyco-occurring species of detritivorousshrimp (Atya lanipes and Xiphocariselongata) Predatory shrimp (Macro-brachium carcinus Macrobrachiumcrenulatum) were excluded from thepools with in-stream fencing to fur-ther reduce the number of speciesinteractions that could affect rates ofleaf decomposition

Although both shrimp species in-fluenced the rates of leaf-litter de-composition their effects were dis-tinctly different (Todd Crowl andAlan Covich unpublished data)Over the 23 days of the experimentXiphocaris shredded the leaf litter asthey ingested leaf fragments and thebacteria and fungi that colonized thedecomposing leaf As a result of thisshredding Xiphocaris increased theconcentration and rate of down-stream transport of suspended fineparticulate organic matter as well asthe concentrations of both total dis-solved nitrogen and dissolved or-ganic carbon Atya also increased

the rate of leaf breakdown relative tocontrols but because they both shredand scrape leaf surfaces as well asfilter out suspended detritus theirprocessing resulted in less down-stream transport of suspended fineorganic particulates Thus a singlefunctional classification for Atya isnot as effective as for XiphocarisMoreover these species are not com-plete substitutes for one another interms of food-web dynamics and eco-system processing

Because of their functional differ-ences these two shrimp species maycomplement one another whereverthey co-occur resulting in a lessldquoleakyrdquo headwater ecosystem thanone containing just a single shrimpspecies That is few leaves are washeddownstream out of pools containingboth species because Xiphocarisbreaks leaf material into small sizefractions that are then available tofilter-feeding Atya Thus the rela-tive spatial location of these twospecies within or between pools couldalter the effectiveness of overall de-trital processing For example whenAtya occur downstream of Xipho-caris growth of the former could beenhanced by the increased availabil-ity of suspended fine organic par-ticulates In addition the degree ofcomplementarity depends on streamflow and water depth In a shallowpool or riffle that has sufficient ve-locity to suspend organic particu-lates leaf shredding by Xiphocariscould increase concentrations of de-trital particulates for filter-feedingAtya

The ldquoprocessing chainrdquo that re-sults from different species of shrimpinteracting as detritivores within andbetween pools is similar to that hy-pothesized in the River ContinuumConcept in that aquatic insect shred-ders occur primarily in upstreamreaches and are thought to increasethe availability of fine organic par-ticulates for downstream collectorsand suspension filter feeders (egVannote et al 1980 Heard andRichardson 1995) More field ma-nipulations are needed to determinehow various species of aquatic in-sects crustaceans gastropods andother benthic invertebrates differ intheir individual species effects onrates of detrital processing and nu-trient cycling

124 BioScience Vol 49 No 2

Grazing crayfish Herbivorousbenthic species have distinct func-tional morphologies feeding prefer-ences and behaviors resulting inmajor differences in their grazingrates (Cronin 1998 Lodge et al1998a) Grazing crayfish generallydo not consume all types of rootedmacrophytes as widely as they con-sume most algae in stream food webs(Nystrom et al 1996) Rooted an-giosperms were derived from terres-trial ancestors and contain indigest-ible cellulose and secondarycompounds such as glucosinolates(Newman 1991) Such chemical de-fenses against herbivores appear tobe restricted to particular species ofgrazers The size and structure ofmacrophytes are also known to in-fluence consumption by differentgrazers (Nystrom et al 1996 Cronin1998) This combination of chemi-cal and structural defenses may re-sult in uniformly low rates of her-bivory among particular assemblagesof producer and consumer species

Because benthic species differ intheir abilities to consume rootedmacrophytes the rate of removal ofsubmerged vegetation can be greatlyaltered if additional herbivorous spe-cies extend their range into shallow-water ecosystems An example of theaddition of a new herbivore to alittoral food web occurred when theldquorustyrdquo crayfish Orconectes rusti-cus moved into northern Wisconsinlakes from Indiana and Ohio Thesenorthern shallow lakes had previ-ously been dominated by anotherspecies of crayfish Orconectes viriliswhose abundance had already beendecreased by the earlier immigrationof another non-native crayfish Orco-nectes propinquus (Lodge et al 19941998b) The native crayfish Ovirilis was less aggressive in its diur-nal feeding on submerged macro-phytes than O rusticus O rusticusremoved entire macrophyte beds inthe littoral zones of lakes from whichit displaced O virilis by clippingplant stems The removal of the mac-rophyte beds had a major effect onassemblages of other species ofbenthic invertebrates such as gas-tropods (Lodge et al 1998b) Fur-thermore because larval fish requirethe protective cover of submergedvegetation to avoid predation bylarger fishes recruitment of juvenile

fish declined and within a few yearsfishermen caught fewer large fish

Omnivorous crustaceans Crayfishconsume a wide range of both plantand animal foods and the spread ofO rusticus into northern habitatstherefore also illustrates how spe-cies-specific differences in feeding byomnivores can change energy flowsthrough benthic food webs In addi-tion to altering the structure of mac-rophyte beds O rusticus also al-tered the gastropod communityassociated with submerged plantsBefore the arrival of O rusticus andthe consequent removal of macro-phytes gastropods had access toabundant periphyton growing on themacrophytes and found protectionfrom fish and crayfish predators amongthe leaf cover After the macrophyteswere removed the gastropod speciesrsquoshell thickness and their ability toevade shell-breaking predators (suchas crayfish) were important predic-tors of which species of gastropodspersisted (Covich et al 1994 Lodgeet al 1994) For example some spe-cies of gastropods avoid day-activepredators by burrowing into the sedi-ments if macrophytes are unavail-able for structural cover (Alan Covichunpublished data) While finding mi-crobial food in the sediments thesegastropod species actively recycle nu-trients and continue their ecologicalroles as consumers

There are other examples of om-nivorous crustaceans that illustratesome unexpected results because ofunanticipated foraging behaviors bynon-native benthic species enteringfresh waters For instance lake man-agers in the western United Statesdid not anticipate the decline in fish-eries caused by intentional introduc-tions of a species of crustacean preytransferred from some deep north-ern lakes to other lakes to increasefish production (Martinez andBergersen 1989) These benthic crus-taceans ldquoopossum shrimprdquo (Mysisrelicta) consume microorganismsduring the day when they remain inor on the sediments (eg Johannsson1992) At night however they moveoff the bottom and switch to feedingon phytoplankton and zooplankton(which also migrate vertically in theevening) As a result of its variedfeeding locations Mysis can be fed

on by lake trout and kokanee salmononly under specific light conditions

In one case Mysis spread down-stream (from an initial introductionto a relatively shallow lake) throughan outflowing river and into Flat-head Lake Montana Althoughmovement of Mysis along rivers hadnot been expected (because thesecrustaceans are usually not found inrivers) it soon disrupted this largerdeeper lakersquos open-water communityof planktonic crustaceans by feedingon juvenile zooplankton Adult cla-doceran zooplankton had previouslybeen prey for lake trout and kokaneesalmon which had also been intro-duced earlier in this deep lake Theloss of open-water zooplankton preyled to declines in sport fishes in Flat-head Lake and the Mysis remainedin deeper darker waters where theyavoided fish predation After thekokanee declined fewer bald eaglesand bears were observed feeding inthe inflowing river because there werefewer fish left to swim upstream tospawn (Spencer et al 1991)

Mysis were deliberately intro-duced into many other western lakesand reservoirs on a trial-and-errorbasis Often the results were not asexpected fisheries declined insteadof flourishing (Martinez and Berger-sen 1989) Dispersal of benthic spe-cies from one lake to another as amanagement tool is now generallyrecognized as inappropriate but un-intentional introductions of manydifferent species are increasing asrecreational and commercial boattraffic expands

Conclusions andpolicy implicationsThese examples illustrate that benthicinvertebrate species function in dif-ferent ways that are important tomaintaining ecosystem functionssuch as energy flow in food websMany benthic species convert liveplant and dead organic material intoprey items for larger consumers incomplex food webs In the process ofmaintaining energy flow thesebenthic species simultaneously pro-vide essential ecosystem servicessuch as nutrient cycling and aerationof sediments Different species com-prise distinct functional groups thatprovide ecological integrity In some

February 1999 125

cases these functional groups maybe represented by only a few speciesso that any loss of species diversitycould be detrimental to continuedecosystem functioning Thus it isincreasingly important to protect thebiodiversity of benthic communitiesto lower the risk of unexpected andunwanted consequences

In the past few decades freshwa-ter habitats have received significantremediation (Norton and Davis1997) as a result of the Clean WaterActrsquos call for greater ecological in-tegrity in particular their biodi-versity has increased For examplethe Cuyahoga River in Ohio is nolonger so contaminated with petro-leum wastes that it catches fire (Ol-ive et al 1988) Lake Eriersquos levels ofdissolved oxygen are increasing inits bottom waters and mayflies arebeginning to return to sediments inthe shallow western basin (eg Kolaret al 1997) that was once thought tobe ldquodeadrdquo The return of mayfliesmeans that nutrients are again rap-idly converted from long-term stor-age in lake sediments into prey thatare available to many species of fishesand other consumers rather than ac-cumulating in the muds Withoutthese benthic insects many nutrientswould reach high concentrations inthe sediments and not be available toconsumer species However the eco-system is being modified now by thespread of invasive non-native spe-cies such as the zebra mussels(Dreissena polymorpha and Dreis-sena bugensis) that have altered theflow of energy within the benthiccommunity (Stewart et al 1998)These invasive mussels alter foodwebs in several ways First filterfeeding by these mussels removessuspended organic materials in over-lying waters and can enrich sedi-ments for use by other benthic spe-cies Second zebra mussel shellsprovide hard surfaces for coloniza-tion by various benthic invertebrates(Botts et al 1996) Finally theseshells also serve as structural refugiafor prey so that many types of benthicinvertebrates may avoid fish preda-tors From available information itis not possible to predict preciselyhow the addition and persistence ofthese invasive mussels will ultimatelychange energy flow and influence wa-ter quality However their addition to

North American fresh waters clearlydemonstrates that particular speciescan alter ecosystem productivity

There is now widespread agree-ment that the global ecosystem can-not function without an adequatesupply of inland waters As the de-mand for fresh water increases inresponse to population growth cli-mate shifts and economic develop-ment additional conflicts amongcompeting users (such as drinkingwater supply for cities or irrigation)will create new challenges to eco-logical processes in natural and man-aged inland waters (Naiman et al1995 Meyer 1997) In the UnitedStates discussions about reauthoriz-ing the Endangered Species Act andthe Clean Water Act will requireecologists to inform policymakers ofalternative solutions to complex is-sues involving large drainage areasand their connections to wetlandslakes and rivers Besides communi-cating the intrinsic value of indi-vidual species it will be important toexplain the functions of diversebenthic communities under differentecological conditions before any ad-ditional species are lost

Freshwater ecologists understandthat benthic species provide impor-tant ecosystem services but an ad-equate model for evaluating theseservices is lacking The public gener-ally understands that water is ldquore-usedrdquo in the hydrologic cycle evapo-ration from surface waters andtranspiration by plants provides wa-ter vapor for cloud formation andprecipitation back to the earthrsquos sur-face But there is less public under-standing of how water enters thebelowground water table and re-charges aquifers Few individualsappreciate that much of flowingwater eventually passes through thesubsurface zones where a rich fau-nal diversity contributes to a widerange of ecological services Thefreshwater benthic biota (microbesto macrofauna) mediate biogeo-chemical transformations and actdirectly to prevent the buildup ofcarbon in the sediments and thedeoxygenation of bottom watersThey also sequester and move con-taminants and excess nutrients fromgroundwaters and sediments whileinfluencing the flux of greenhousegases (carbon dioxide and methane)

Although these diverse species maybe hidden and ldquoinvisiblerdquo becausethey live below the surface the in-tegrity of the freshwater supply de-pends on how various benthic spe-cies make their living and contributeto complex food webs

AcknowledgmentsWe appreciate helpful reviews by WK Dodds D M Lodge and RRudman The ideas reviewed herewere stimulated by a recent sympo-sium sponsored by the American As-sociation for the Advancement ofScience This work was supportedby grants from the National ScienceFoundationrsquos Division of Environmen-tal Biology This article is a project ofthe Committee on Soil and SedimentBiodiversity and Ecosystem Function-ing a component of DIVERSITAScoordinated by SCOPE

References citedBlumenshine SC Vadeboncoeur Y Lodge

DM Cottingham KL Knight SE 1997Benthicndashpelagic links Responses ofbenthos to water-column nutrient enrich-ment Journal of the North AmericanBenthological Society 16 466ndash479

Botts PS Patterson BA Schloesser DW 1996Zebra mussel effects on benthic inverte-brates Physical or biotic Journal of theNorth American Benthological Society 15179ndash184

Brinkhurst RO 1997 On the role of tubificidoligochaetes in relation to fish diseasewith special reference to the MyxozoaAnnual Review of Fish Disease 6 29ndash40

Chapin FS Walker BH Hobbs RJ HooperDU Lawton JH Sala OS Tilman D 1997Biotic control over the functioning of eco-systems Science 277 500ndash504

Charbonneau P Hare L 1998 Burrowingbehavior and biogenic structures of mud-dwelling insects Journal of the NorthAmerican Benthological Society 17 239ndash249

Clarke KD Knoechel R Ryan PM 1997Influence of trophic role and life-cycleduration on timing and magnitude ofbenthic macroinvertebrate response towhole-lake enrichment Canadian Jour-nal of Fisheries and Aquatic Sciences 5489ndash95

Colwell R 1998 Balancing the biocomplexityof the planetrsquos life systems A twenty-firstcentury task for science BioScience 48786ndash787

Covich AP 1988 Geographical and histori-cal comparisons of Neotropical streamsBiotic diversity and detrital processing inhighly variable habitats Journal of theNorth American Benthological Society 7361ndash386

______ 1993 Water and ecosystems Pages40ndash55 in Gleick PH ed Water in Crisis

126 BioScience Vol 49 No 2

Oxford Oxford University PressCovich AP McDowell WH 1996 The stream

community Pages 433ndash459 in Reagan DPWaide RB eds The Food Web of a Tropi-cal Rain Forest Chicago University ofChicago Press

Covich AP Crowl TA Alexander JE Jr VaughnCC 1994 Predator-avoidance responses infreshwater decapodndashgastropod interactionsmediated by chemical stimuli Journal ofthe North American Benthological Society13 283ndash290

Covich AP Crowl TA Johnson SL Pyron M1996 Distribution and abundance of tropi-cal freshwater shrimp along a stream corri-dor Response to disturbance Biotropica28 484ndash492

Creed RP 1994 Direct and indirect effects ofcrayfish grazing in a stream communityEcology 75 2091ndash2103

Cronin G 1998 Influence of macrophyte struc-ture nutritive value and chemistry on thefeeding choices of a generalist crayfish Pages307ndash317 in Jeppesen E Sondergaard MSondergaard M Christoffersen K eds TheStructuring Role of Submerged Macrophytesin Lakes New York Springer

Crowl TA Covich AP 1990 Predator-inducedlife history shifts in a freshwater snail Sci-ence 247 949ndash951

______ 1994 Responses of a freshwater shrimpto chemical and tactile stimuli from a largedecapod predator Journal of the NorthAmerican Benthological Society 13 291ndash298

Cummins KW Wilzbach MA Gates DM PerryJB Taliaferro WB 1989 Shredders andriparian vegetation BioScience 39 24ndash30

Cummins KW Cushing CE Minshall GW1995 Introduction An overview of streamecosystems Pages 1ndash8 in Cushing CECummins KW Minshall GW eds Riverand Stream Ecosystems AmsterdamElsevier

Dodds WK 1991 Community interactionsbetween the filamentous alga Cladophoraglomerata (L) Kuetzing its epiphytes andepiphyte grazers Oecologia 85 572ndash580

Ehrlich PR Ehrlich AH 1981 Extinction TheCauses and Consequences of the Disap-pearance of Species New York RandomHouse

Ehrlich PR Walker B 1998 Rivets and redun-dancy BioScience 48 387

Freckman DW Blackburn TH Brussaard LHutchings P Palmer MA Snelgrove PVR1997 Linking biodiversity and ecosystemfunctioning of soils and sediments Ambio26 556ndash562

Frost TM Carpenter SR Ives AR Kratz TK1995 Species compensation and comple-mentarity in ecosystem function Pages 224ndash239 in Jones CG Lawton JH eds LinkingSpecies amp Ecosystem New York Chapmanand Hall

Goedkoop W Johnson RK 1996 Pelagicndashbenthic coupling Profundal benthic com-munity response to spring diatom deposi-tion in mesotrophic Lake Erken Limnologyand Oceanography 41 636ndash647

Groffman PM Bohlen PJ 1999 Soil and sedi-ment biodiversity BioScience 49 139ndash148

Hawkins CP MacMahon JP 1989 GuildsThe multiple meanings of a concept AnnualReview of Entomology 34 423ndash451

Heard SB Richardson JS 1995 Shredderndash

collector facilitation in stream detrital foodwebs Is there enough evidence Oikos 72359ndash366

Holsinger JR 1993 Biodiversity of subterra-nean amphipod crustaceans Global pat-terns and zoogeographic implications Jour-nal of Natural History 27 821ndash835

Hutchinson GE 1993 A Treatise on Limnol-ogy Vol 4 The Zoobenthos New YorkJohn Wiley amp Sons

Johannsson O 1992 Life-history and produc-tivity of Mysis relicta in Lake Ontario Jour-nal of Great Lakes Research 18 154ndash168

Johnson SL Covich AP 1997 Scales of obser-vation of riparian forests and distributionsof suspended detritus in a prairie riverFreshwater Biology 37 163ndash175

Johnson SL Covich AP Crowl TA Estrada ABithorn J Wurtsbaugh WA 1998 Do sea-sonality and disturbance influence repro-duction in freshwater atyid shrimp in head-water streams Puerto Rico Proceedings ofthe International Association of Theoreti-cal and Applied Limnology 26 2076ndash2081

Kolar CS Hudson PL Savino JF 1997 Condi-tions for the return and simulation of therecovery of burrowing mayflies in westernLake Erie Ecological Applications 7 665ndash676

Kornijow R Gulati RD Ozimek T 1995 Foodpreference of freshwater invertebrates Com-paring fresh and decomposed angiospermsand a filamentous alga Freshwater Biology33 205ndash212

Lawton JH Brown VK 1994 Redundancy inecosystems Pages 255ndash270 in Schulze EDMooney HA eds Biodiversity and Ecosys-tem Function New York Springer-Verlag

Lindegaard C 1994 The role of zoobenthos inenergy flow of two shallow lakesHydrobiologia 275276 313ndash322

Lodge DM Kershner MW Aloi JE Covich AP1994 Effects of an omnivorous crayfish(Orconectes rusticus) on a freshwater lit-toral food web Ecology 75 1265ndash1281

Lodge DM Cronin G van Donk E Froelich AJ1998a Impact of herbivory on plant stand-ing crop Comparisons among biomes be-tween vascular and nonvascular plants andamong freshwater herbivore taxa Pages149ndash174 in Jeppesen E Sondergaard MSondergaard M Christoffersen K eds TheStructuring Role of Submerged Macrophytesin Lakes New York Springer

Lodge DM Stein RA Brown KM Covich APBronmark C Garvey JE Klosiewski SP1998b Predicting impact of freshwater ex-otic species on native biodiversity Chal-lenges in spatial and temporal scaling Aus-tralian Journal of Ecology 23 53ndash67

Malmqvist B Rundle S Bronmark CErlandsson A 1991 Invertebrate coloniza-tion of a new man-made stream in southernSweden Freshwater Biology 26 307ndash324

Martinez PJ Bergersen EP 1989 Proposedbiological management of Mysis relicta inColorado lakes and reservoirs North Ameri-can Journal of Fisheries Management 9 1ndash11

Merritt RW Cummins KW 1996 Trophicrelations of macroinvertebrates Pages 453ndash474 in Hauer FR Lamberti GA eds Meth-ods in Stream Ecology New York Aca-demic Press

Meyer JL 1997 Conserving ecosystem func-tion Pages 136ndash145 in Pickett STA Ostfeld

RS Shachak MS Likens GE eds The Eco-logical Basis of Conservation New YorkChapman and Hall

Milner AM 1987 Colonization and ecologicaldevelopment of new streams in Glacier BayNational Park Alaska Freshwater Biology18 53ndash70

Naeem S 1998 Species redundancy and eco-system reliability Conservation Biology 1239ndash45

Naiman RJ Deacutecamps H 1997 The ecology ofinterfaces Riparian zones Annual Reviewof Ecology and Systematics 28 621ndash658

Naiman RJ Magnuson JJ McKnight DMStanford JA 1995 The Freshwater Impera-tive A Research Agenda Washington (DC)Island Press

Newman RM 1991 Herbivory and detritivoryon freshwater macrophytes by invertebratesA review Journal of the North AmericanBenthological Society 10 89ndash114

Norton DJ Davis DG 1997 Policies for pro-tecting aquatic diversity Pages 276ndash300 inBoyce MS Haney A eds Ecosystem Man-agement Applications for Sustainable For-est and Wildlife Resources New Haven(CT) Yale University Press

Nystrom P Bronmark C Graneli W 1996Patterns in benthic food websmdasha role foromnivorous crayfish Freshwater Biology36 631ndash646

Olive JH Jackson JL Bass J Holland L SaviskyT 1988 Benthic macroinvertebrates as in-dexes of water quality in the upper CuyahogaRiver Ohio Journal of Science 88 91ndash98

Palmer MA et al 1997 Biodiversity and eco-system processes in freshwater sedimentsAmbio 26 571ndash577

Parkyn SM Rabeni CF Collier KJ 1997 Ef-fects of crayfish (Paranephrops planfironsParastacidae) on in-stream processes andbenthic fauna A density manipulation ex-periment New Zealand Journal of Marineand Freshwater Research 31 685ndash692

Pelegri SP Blackburn TH 1996 Nitrogen cy-cling in lake sediments bioturbated byChironomus plumosus larvae under differ-ent degrees of oxygenation Hydrobiologia325 231ndash238

Postel S Carpenter S 1997 Freshwater ecosys-tem services Pages 195ndash214 in Daily GCed Naturersquos Services Societal Dependenceon Natural Ecosystems Washington (DC)Island Press

Power ME 1995 Floods food chains andecosystem processes in rivers Pages 52ndash60in Jones CG Lawton JH eds Linking Spe-cies and Ecosystems New York Chapmanand Hall

Pringle CM Blake GA Covich AP Buzby KMFinley AM 1993 Effects of omnivorousshrimp in a montane tropical stream Sedi-ment removal disturbance of sessile inver-tebrates and enhancement of understoryalgal biomass Oecologia 93 1ndash11

Rosemond AD Pringle CM Ramirez A 1998Macroconsumer effects on insect detritivoresand detritus processing in a tropical streamFreshwater Biology 39 515ndash523

Spencer CN McClelland BR Stanford JA 1991Shrimp stocking salmon collapse and eagledisplacement BioScience 41 14ndash21

Stewart TW Miner JG Lowe RL 1998 Anexperimental analysis of crayfish (Orco-nestes rusticus) effects on a Dreissena-domi-nated benthic macroinvertebrate commu-

February 1999 127

nity in western Lake Erie Canadian Journalof Fisheries and Aquatic Science 55 1ndash7

Stockley RA Oxford GS Ormond RFG 1998Do invertebrates matter Detrital process-ing in the River Swale-Ouse Science of theTotal Environment 210 427ndash435

Threlkeld ST 1994 Benthicndashpelagic interac-tions in shallow water columns Anexperimentalistrsquos perspective Hydrobio-logia 275276 293ndash300

Tilman D Knops J Wedin D Reich P RitchieR Siemann E 1997 The influence of func-tional diversity and composition on ecosys-tem processes Science 277 1300ndash 1305

van de Bund WJ Goedkoop W Johnson RK1994 Effects of deposit-feeder activity onbacterial production and abundance inprofundal lake sediment Journal of theNorth American Benthological Society 13532ndash539

Vannote RL Minshall GW Cummins KWSedell JR Cushing CE 1980 The rivercontinuum concept Canadian Journal ofFisheries and Aquatic Sciences 37 130ndash137

Wallace JB Webster JR 1996 The role ofmacroinvertebrates in stream ecosystemfunction Annual Review of Entomology41 115ndash139

Wallace JB Eggerton SL Meyer JL WebsterJR 1997 Multiple trophic levels of a foreststream linked to terrestrial litter inputsScience 277 102ndash104

Wilson EO 1992 The Diversity of Life NewYork WW Norton

Biology Reporting Awards

The Awards

The American Institute of Biological Sciences Media Award wasestablished in 1995 to recognize outstanding reporting on research

in biology This yearrsquos winners will each receive $1000 and expenses toattend the annual meeting of the American Institute of BiologicalSciences 11ndash14 November 1999 where the awards will be presented

The awards are designed to encourage the communication of biologyto the public One award is for print journalism specifically the otherfor broadcast journalism AIBS intends to promote public understand-ing of how biologists approach their research collect and interprettheir data and reach conclusions as well as how the research and itsconclusions are relevant to society

Rules

The awards will be limited to nontechnical journalism Articlespublished in newspapers and magazines are elegible for the print

award and stories broadcast on radio and television are eligible for thebroadcast award Both freelancers and staff writers are eligible Profes-sional scientists writing in their area of research are not eligible Booksand articles in technical journals will not be considered Articlesappearing in BioScience the publication of the American Institute ofBiological Sciences are not eligible

Biological research is broadly defined to include laboratory and fieldwork as well as theoretical advances For the purposes of this awardit does not include testing of medical or veterinary treatments

Entries will be judged on the basis of clarity reporting and writingskills originality and appeal to the general public

Applicants may submit a single contribution or a series Stories musthave been published or broadcast betwen 1 January 1998 and 31December 1998 A series will be accepted if more than half of itappeared between those dates Applications may be submitted by thejournalist or on his or her behalf

For information and entry form

Send a self-addressed envelope to AIBS Media Award 1444 Eye StNW Suite 200 Washington DC 20005

All applications and submissions must be received by 1 April 1999Submissions will not be returned

Judges

The award will be judged by a panel of science journalists andscientists chosen by the American Institute of Biological Sciences

The winner will be notified by 1 July 1999