A global meta-analysis of the ecological impacts ofnonnative crayfish

Laura A. Twardochleb1AND Julian D. Olden2

School of Aquatic and Fishery Sciences, University of Washington, Seattle, Washington 98195 USA

Eric R. Larson3

Department of Ecology and Evolutionary Biology, University of Tennessee,Knoxville, Tennessee 37996 USA

Abstract. Nonnative crayfish have been widely introduced and are a major threat to freshwaterbiodiversity and ecosystem functioning. Despite documentation of the ecological effects of nonnativecrayfish from .3 decades of case studies, no comprehensive synthesis has been done to test quantitativelyfor their general or species-specific effects on recipient ecosystems. We provide the first global meta-analysis of the ecological effects of nonnative crayfish under experimental settings to compare effectsamong species and across levels of ecological organization. Our meta-analysis revealed strong, butvariable, negative ecological impacts of nonnative crayfish with strikingly consistent effects amongintroduced species. In experimental settings, nonnative crayfish generally affect all levels of freshwaterfood webs. Nonnative crayfish reduce the abundance of basal resources like aquatic macrophytes, prey oninvertebrates like snails and mayflies, and reduce abundances and growth of amphibians and fish, but theydo not consistently increase algal biomass. Nonnative crayfish tend to have larger positive effects ongrowth of algae and larger negative effects on invertebrates and fish than native crayfish, but effect sizesvary considerably. Our study supports the assessment of crayfish as strong interactors in food webs thathave significant effects across native taxa via polytrophic, generalist feeding habits. Nonnative crayfishspecies identity may be less important than extrinsic attributes of the recipient ecosystems in determiningeffects of nonnative crayfish. We identify some understudied and emerging nonnative crayfish that shouldbe studied further and suggest expanding research to encompass more comparisons of native vs nonnativecrayfish and different geographic regions. The consistent and general negative effects of nonnative crayfishwarrant efforts to discourage their introduction beyond native ranges.

Key words: Orconectes rusticus, Orconectes virilis, Pacifastacus leniusculus, Procambarus clarkii, taxonomiceffect, manipulative experiment.

Humans have a penchant for introducing species toareas beyond their native geographic distributions,thereby providing the opportunity for these nonna-tive species to become invaders (Elton 1958). Theecological consequences of nonnative species canrange from beneficial (Schlaepfer et al. 2011) todetrimental, and in the latter case, often lead tosignificant ecological damage ranging from theextinction of native species to alteration of ecosystemprocesses (e.g., Vitousek et al. 1996, Mack et al. 2000,Vila et al. 2009). Ecologists have long been challenged

to identify, quantify, and predict the ecologicalimpacts of invasive species (Parker et al. 1999), andmeta-analytical approaches have emerged as a pow-erful tool to shed insight into the general ecologicaleffects of and traits possessed by invasive species. Inrecent applications of meta-analyses in invasionecology, investigators have developed frameworksto organize invader impacts based on attributes ofinvasive species (Thomsen et al. 2011). Meta-analysesalso have addressed such fundamental topics ininvasion ecology as the influence of taxonomicidentity on invasion impact (Ricciardi and Atkinson2004), understanding the environmental determinantsof establishment success (Cassey et al. 2005), andassessing the differential vulnerability of native vs

1 E-mail addresses: ltwardoc@uw.edu2 olden@uw.edu3 elarson8@utk.edu

Freshwater Science, 2013, 32(4):1367–1382’ 2013 by The Society for Freshwater ScienceDOI: 10.1899/12-203.1Published online: 5 November 2013

1367

nonnative species to climate change (Sorte et al. 2012).Meta-analyses are limited only by the taxonomy andgeography of available data on nonnative species(Pysek et al. 2008) and will continue to be animportant tool in invasion biology.

Crayfish have become one of the most widelyintroduced freshwater taxa through a variety ofpathways, including bait-bucket releases, intentionalintroduction to support fisheries, and release after usefor education (Hobbs et al. 1989, Gherardi 2010). Theecological effects of nonnative crayfish are welldocumented from .3 decades of case studies (Hobbset al. 1989, Snyder and Evans 2006, Lodge et al. 2012),but have yet to be synthesized quantitatively to test forgeneralized effects on recipient ecosystems. Populationdeclines, extirpations, and extinctions of native cray-fishes are among the most alarming effects of wide-spread crayfish introductions (Lodge et al. 2000, Perryet al. 2001). However, the ecological effects are muchmore far-reaching. Crayfish are generalist omnivores,and thus, have large effects on both primary andsecondary producers (Lodge et al. 1994, Perry et al.2000). Crayfish are ecosystem engineers that increaserates of leaf-litter breakdown and nutrient cycling instreams (Charlebois and Lamberti 1996, Bobeldyk andLamberti 2008), and their grazing and burrowing canreduce benthic algae and macrophyte cover, producinga state change in lakes and wetlands from clear- tophytoplankton-dominated turbid-water systems (Fem-inella and Resh 1989, Matsuzaki et al. 2009). Coupledwith habitat modification, crayfish predation drivesdeclines in diversity and abundances of native inver-tebrates (McCarthy et al. 2006, Correia and Anastacio2008), and reduces amphibian populations throughpredation on eggs and larvae (Gamradt and Kats 1996,Gamradt et al. 1997). Last, crayfish invasions haveresulted in fish declines through predation, sheltercompetition, and indirect competition for prey (re-viewed in Reynolds 2011). Because of their potentialeffects across all levels of ecological organization,nonnative crayfish are a major threat to freshwaterbiodiversity and ecosystem functioning.

Difficulties in evaluating and predicting effects ofinvasive species are particularly relevant to crayfishbecause past research suggests that the ecologicaleffects of crayfish may differ according to the identityof the introduced crayfish and resident species in theenvironment (Larson and Olden 2010, Lodge et al.2012). Case studies of red swamp crayfish Procambarusclarkii invasions have documented dramatic declinesin macrophytes (Matsuzaki et al. 2009), whereasstudies of rusty crayfish Orconectes rusticus invasionsoften reveal the largest effects on benthic invertebratecommunities (McCarthy et al. 2006). These examples

suggest that the species of crayfish, by way ofpreferential feeding habits, may influence the typeand magnitude of their ecological effects on residentnative species. Factors extrinsic to the crayfish species,including competitors and prey in the receivingenvironment, also can influence the effects of crayfishinvasions.

Caution should always be practiced when makinggeneralizations about effects of invasive species, andprevious reviews on the ecological impacts of nonna-tive crayfish have been largely narrative and lackedquantitative synthesis across species. Case studiesexamining the ecological effects of specific crayfishspecies on recipient ecosystems are beneficial, butindividually, they provide little insight into general-ized patterns of crayfish impacts. However, takentogether, the rich body of existing literature lendsitself to a systematic overview of the ecological effectsof nonnative crayfish. A literature synthesis thatcompares effects across nonnative crayfish species(and where possible in relation to native crayfisheffects) will help identify generalities in impacts thatcannot be addressed by individual studies alone.

Here, we provide the first systematic review andglobal meta-analysis of the ecological effects ofnonnative crayfish. Our aim is to test current evidencefor the ecological effects of nonnative crayfish acrossdifferent levels of ecological organization by address-ing the following questions: What is the empiricalevidence for the effects of nonnative crayfish? Dononnative crayfish differ from native crayfish in theirecological effects? Does species identity of nonnativecrayfish determine the magnitude and direction oftheir ecological effects? We expect that nonnativecrayfish will have larger effects than native crayfish ata given location. We also predict that all nonnativecrayfish species will have direct and indirect negativeeffects on most levels of ecological organizationincluding macrophytes, benthic invertebrates, suchas insects and snails, crayfish, amphibians, and fishand positive effects on algal biomass. Last, we predictthat nonnative crayfish will be more successful thannative crayfish in agonistic interactions for territoryand shelter. By comparing and contrasting theecological effects of different nonnative and nativecrayfish species across levels of ecological organiza-tion, we aim to inform future risk assessments andcontrol strategies for nonnative crayfish.

Methods

Systematic search

Our protocols for search and selection followedthose outlined by Pullin and Stewart (2006) for

1368 L. A. TWARDOCHLEB ET AL. [Volume 32

systematic review, which included formation ofsearch protocol and data inclusion, data extraction,and analysis. We searched the Institute of ScientificInformation (ISI; Thomson Reuters) Web of Scienceonline database and identified peer-reviewed paperspublished through the end of 2012 that containedresults of experimental manipulations used to quan-tify the effects of nonnative crayfish on recipientecosystems. We also included studies referencedwithin articles obtained from this search. We includedonly manipulative experimental studies to avoidunstandardized treatment levels (crayfish densities)and the confounding effects of other predators onresponse organisms. Our search terms included key-word combinations: ‘experiment* or manipulation*’and ‘non-native* or nonnative * or invasive* or alien*or exotic* or non-indigenous* or nonindigenous* orintroduced*’ and ‘crayfish’ or ‘crawfish’ or ‘Procam-barus clarkii* or Pacifastacus leniusculus* or Orconectesrusticus* or Orconectes virilis* or Orconectes propinquus*’in the article. The species listed have the longest andmost widespread history of invasion studies (Hobbset al. 1989, Lodge et al. 2012), and additional searchingrevealed only limited sample sizes for other nonnativecrayfish (e.g., Orconectes neglectus; Rabalais andMagoulick 2006). We included O. propinquus in oursearch terms, but we did not include this species inour meta-analysis because of the continued uncer-tainty regarding whether O. propinquus is nonnative tothe regions where it has been studied in Wisconsin(Hobbs and Jass 1988). We further filtered searchresults to include journals that publish articles inEnglish within the categories of ecology, marine orfreshwater biology, fisheries, limnology, zoology,behavioral sciences, environmental sciences, or biodi-versity conservation.

Study selection

We read the abstract for each article returned fromthe Web of Science search and further filtered resultsto include studies that fell into 1 of 3 categories: thosethat tested responses in biomass or abundance ofnative taxa to 1) native and nonnative crayfish or 2)nonnative crayfish only, or 3) those that paired anative and nonnative crayfish in agonistic or compet-itive interactions. When comparing the ecologicaleffects of native and nonnative crayfish, we includedonly those studies that simultaneously examined theresponse of groups exposed to a native crayfish vsnonnative crayfish vs a control with no crayfish.

We identified effects of nonnative crayfish acrosslevels of ecological organization by selecting studiesthat examined the effects of exposure to nonnative

crayfish. We included studies in these analyses if theycompared groups of organisms exposed to nonnativecrayfish to groups without crayfish. In all studiesincluded in our analyses, investigators manipulatedspecies in the laboratory, outdoor ponds, or in situwith crayfish enclosures or exclusions.

We assessed whether nonnative crayfish win moreagonistic interactions for shelter or territorial domi-nance than native crayfish. Authors of all studiesincluded in this analysis observed agonistic behaviorsin laboratory tanks. Authors of studies of shelter-dominance interactions defined the winner as thecrayfish occupying the shelter, and reported thenumber of wins for each species. Authors of studiesof territorial-dominance interactions identified thewinner from a set of behaviors, including approaches,threats, and strikes, and each author reported thenumber of each behavior displayed by each crayfishspecies or reported the number of wins as determinedby the experimenters. Size is often a determinant ofthe success of an aggressor, so we used replicates withlarge nonnative crayfish and small native crayfish asour treatment group and replicates with large nativecrayfish and small nonnative crayfish as our controlswhen calculating the effect size for each study.

Data extraction

We categorized articles according to the focalnonnative crayfish species examined and groupedresponse organisms taxonomically as benthic algae,macrophytes, benthic invertebrates (further dividedby family or order), crayfish, fish, or amphibians(Table 1). We refer to these as ecological responses.We coded each study for experimental venue, i.e.,whether the experimenters used laboratory tanks,outdoor mesocosms, or cages in situ to permitanalyses of how experimental design potentiallyinfluenced the responses.

We extracted statistics for control and treatmentgroups, including sample sizes, means, proportions,and standard deviations or standard errors, fromtables and results in the articles. For articles that didnot report those statistics, we requested data fromauthors and, when necessary, we used the data-extraction software, DataThief III (http://datathief.org) to extract data from figures. Authors of somearticles reported data in a way that did not permitextraction of sample size, standard deviation, orstandard error measurements, and we excluded thesestudies. We were able to obtain the required datafrom 52 of the 96 articles that fit our search criteria.These articles presented 61 experiments with 6nonnative crayfish species (Appendix S1; available

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1369

online from: http://dx.doi.org/10.1899/12-203.1.s1)from 8 countries (Fig. 1A).

Analysis

We measured directional effect sizes for ecologicalresponses to nonnative crayfish relative to nativecrayfish to examine whether nonnative crayfish havelarger ecological effects than native crayfish. We testedfor effects of crayfish on algal biomass and biomassand abundances of macrophytes, invertebrates, andfish, and we tested whether mean effect sizes ofecological responses differed among nonnative cray-fish species (Table 1). Responses examined for benthicalgae were direct consumptive and indirect (viatrophic cascades) changes in biomass or concentrationsof chlorophyll a. For floating, submerged, and emer-gent macrophytes, we tested the effects of consump-tion by nonnative crayfish on abundances and biomassof seeds, seedlings, and plants. We also tested thedirect effects of predation by nonnative crayfish onabundances and biomass of benthic invertebrates. Weexamined changes in rates of native crayfish survivalas direct (predation) and indirect (behavioral) effects ofexposure to nonnative crayfish or as number ofvictories in agonistic interactions. Furthermore, weexamined predatory effects of nonnative crayfish onabundances of fish and amphibians and changes ingrowth mediated by behavioral responses to predationrisk. Last, we had adequate sample sizes to test fordifferences in mean effect sizes among experimentalvenues for macrophytes and invertebrates, but noother ecological response.

All meta-analytic and statistical calculations wereimplemented using the software MetaWin, version 2.0

(Rosenberg et al. 1999). We used Hedges’ d (Hedges 1981)as the metric of effect size for ecological responses.Hedges’ d is a commonly used measure of effect size inecological studies (Møller and Jennions 2002) and isappropriate for use in traditional meta-analyses because ithas a low type-I error rate (Lajeunesse and Forbes 2003).Hedges’ d computes the effect size as the standardizedmean difference between treatment and control groups,and includes a weighting factor to correct for small samplesizes (Rosenberg et al. 1999). We calculated Hedges’ d as:

d=XT{XC

� �S

J ½1�

where XT is the mean of the treatment group, XC is themean for the control group, and S is the pooled standarddeviation. We calculated S as:

S=

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiNT{1ð Þ STð Þ2z NC{1ð Þ SCð Þ2

NTzNC{2

s½2�

where ST and SC are the standard deviations for thetreatment and control groups, respectively. J is theweighting factor based on the sample sizes for the

treatment (NT) and control (NC) groups, respectively.We calculated J as:

J=1{3

4 NCzNT{2ð Þ{1: ½3�

We computed the variance of Hedges’ d as:

Vd=NCzNT

NCNTz

d2

2(NCzNT): ½4�

TABLE 1. Effects of nonnative crayfish on different taxonomic groups. Evidence column indicates whether our meta-analysisprovided empirical support for positive or negative effects. Orconectes hylas and Orconectes neglectus are not included in this table,but these crayfish were included in sums for crayfish–crayfish studies.

Taxonomic groupNumber of

experimentsOrconectes

rusticusOrconectes

virilisProcambarus

clarkiiPacifastacusleniusculus Evidence

Algae 9 5 0 1 3 +/2Macrophytes 20 8 0 9 3 –Invertebrates 23 6 1 7 9 –

Amphipoda 5 1 1 1 2 +/2Chironomidae 6 2 1 1 2 +/2Ephemeroptera 3 2 0 1 0 –Gastropoda 13 4 1 4 4 –Trichoptera 5 3 0 1 1 +/2

Crayfish 10 0 0 4 4 –

Agonism 5 0 0 1 3 –Survival 5 0 0 3 1 +/2

Fish 7 0 2 2 3 –Amphibians 9 0 0 8 1 –

1370 L. A. TWARDOCHLEB ET AL. [Volume 32

Negative values for Hedges’ d indicate a negativeeffect of nonnative crayfish on the measured res-ponse variable (biomass or abundance) compared tocontrols of native crayfish or no crayfish.

We calculated the log-odds ratio (OR) for the effect-size estimate of agonistic interactions between non-native and native crayfish. The OR computes themean effect size for studies where a 2 3 2 contingencytable with categorical responses is appropriate. Wecompared the probability of the native specieswinning an agonistic interaction in control groups,in which the native species is larger than thenonnative, to treatment groups with nonnative spe-cies larger than native. Following Rosenberg et al.(1999), we computed the OR as:

lnOR=

XOi{OOiX

Vi

½5�

where Oi equals the observed responses from the

treatment group (number of times the native wins), OOi

is the expected number of responses assuming notreatment effect, and vi is the variance. A negativeeffect size for the OR indicates that the native specieswins agonistic interactions fewer times than expectedby chance alone.

All control and treatment groups used in ouranalyses were exposed to experimenter-defined cray-fish densities. Therefore, we excluded in situ exper-iments that considered natural crayfish densities (e.g.,

FIG. 1. Geographic distribution of publications across countries and states/provinces of the USA and Canada, respectively (A)and photographs of Orconectes rusticus (B), Orconectes virilis (C), Pacifastacus leniusculus (D), and Procambarus clarkii (E), the 4 moststudied nonnative crayfish. Photographs by E. R. Larson (B), J. D. Olden (C), C. Capinha (D), and F. Tomasinelli (E).

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1371

uncaged stream plots). When calculating effect sizesfor ecological responses to native vs nonnativecrayfish, we used groups exposed to nonnativecrayfish as treatments and groups exposed to nativecrayfish as controls. In this way, we compared theeffects of nonnative crayfish directly to effects ofnative crayfish. In our analyses of noncrayfishecological responses to nonnative crayfish, groupsexposed to nonnative crayfish served as treatmentsand were compared to groups without crayfish(controls). When different crayfish densities wereused in a study, we selected only the highest-densitytreatment group for our analysis. Furthermore, inmany experiments, the effects of crayfish on treatmentgroups were not manifested immediately after theexperiment began. To capture the delayed response tocrayfish treatments, we used only data from the lastsampling date when multiple post-treatment sampleswere collected.

Some authors reported .1 response measure forthe final sampling date, e.g., multiple biomass orabundance measures on the same individuals ortaxonomic groups in an experiment. We calculatedseparate effect sizes and variances for each responsemeasure in each crayfish–response pairing. To avoidpseudoreplication in our analysis, we pooled effectsizes and variances from repeated-response measuresto obtain a mean effect size for every nonnativecrayfish–response pairing in each experiment andused these effect sizes in our final analysis (see vanKleunen et al. 2010 for further description and anexample of this method). To retain as much informa-tion as possible from each study, we did not pooleffect sizes across different experiments from thesame publication (Gurevitch et al. 1992).

For all analyses, we used random-effects andmixed-effects models (random-effects models with agrouping variable). Random-effects models assumethat random sources of variation in effect sizes existbetween studies and that sampling error accounts forheterogeneity within studies. Mixed-effects modelsincorporate 2 sources of variation that are importantto ecological studies, the study-specific samplingerror and between-study differences in true effect-sizes (Gurevitch and Hedges 1999). Grouping vari-ables in mixed-effects models allows tests of signifi-cance on the cumulative effect size (e.g., acrosscrayfish species) and between effect sizes (e.g.,between crayfish species) for groups of studies(Rosenberg et al. 1999). For comparisons of effectsamong nonnative crayfish, we used crayfish speciesas the grouping variable, and for tests for effects ofexperimental venue, we used venue (laboratory,outdoor mesocosm, or cage) as the grouping variable.

We calculated bias-corrected 95% bootstrap confi-dence intervals (CIs; 9999 permutations) to testwhether mean effect sizes were significantly differentfrom 0 (Adams et al. 1997). A CI that does not overlapwith 0 indicates a statistically significant effect size(Gurevitch et al. 1992). We used Pearson’s x2 tests toexamine whether variance among effect sizes for eachexperiment (QTotal) was significantly larger than wouldbe expected from sampling error alone. For mixed-effects models, we also used a x2 test to evaluatedifferences in variance between study groups (Qb) andwithin groups (Qw). A significant QTotal indicates thateffect sizes are not equal across studies, whereas asignificant Qb or Qw indicates that effect sizes varysignificantly between (e.g., between nonnative crayfishspecies) or within study groups, respectively.

For the overall ecological response to nonnativecrayfish and the effects of experimental venue, wepresent data from mixed-effects models. Otherwise, wepresent data from random-effects models for 2 reasons:1) low sample sizes did not permit inclusion of allcrayfish species and experiment types in mixed-effectsmodels, resulting in reduced statistical power for someanalyses and 2) results from mixed-effects modelsindicated that total and between-species heterogeneitywere not statistically significant for any response(Appendices S2, S3; available online from: http://dx.doi.org/10.1899/12-203.1.s2, http://dx.doi.org/10.1899/12-203.1.s3). Therefore, random-effects modelsare appropriate for our analyses (Rosenberg et al.1999). We ran a separate random-effects model for eachspecies of crayfish and experimental venue with asample size of n § 2 to incorporate more experimentsinto our estimates of the cumulative effects on eachecological response. We also obtained an estimate ofthe cumulative effect-size variance and CI for crayfishon each ecological response.

Meta-analyses can be influenced by the publicationbias associated with a tendency of journals to publishstudies showing significant results (also termed ‘‘the filedrawer problem’’; Rosenthal 1979). To address potentialpublication bias, we calculated a fail-safe number foreach mixed-effects and random-effects model (Appen-dices S2, S3). The fail-safe number reflects the numberof additional nonsignificant studies that, if included inthe analysis, could change the model output fromsignificant to nonsignificant (Rosenberg 2005). Thus, thelarger the fail-safe number, the more confidence wehave in the validity of a significant result.

Results

Nonnative crayfish exhibit larger positive effects onalgae and somewhat larger negative effects on aquatic

1372 L. A. TWARDOCHLEB ET AL. [Volume 32

insects, snails, and fish than do native crayfish(Fig. 2). While both native and nonnative crayfishhave negative effects on macrophytes, nonnativecrayfish exhibit slightly smaller effects, and thus, apositive effect size compared to native crayfish(Fig. 2). However, these patterns were quite variable,and differences were large for only 4 studies thatexamined responses in algal biomass and snails. Adirect comparison of the ecological effects of nativeand nonnative crayfish is a challenge because of thelimited number of studies. Therefore, differencesobserved here may not exist or may vary as a functionof species identity and type of ecological response.

Our meta-analyses supported predicted effects ofnonnative crayfish on the suite of ecological responsesconsidered, with the exception of effects on algae(Table 1). The negative effects of nonnative crayfishacross the food web are substantial (Fig. 3). With theexception of O. virilis, the mean effect sizes for allspecies were statistically significant and were &|0.8|,which is a strong effect according to Gurevitch et al.(1992). Mean effect sizes among crayfish species didnot differ significantly (Qb = 5.25, p = 0.15).Procambarus clarkii and P. leniusculus had the largestoverall effects across ecological responses. However,heterogeneity in effect size was high among studieswithin each group (i.e., crayfish species; Qw = 81.4,

p = 0.02), indicating that differences in effect sizes foreach species are, in part, explained by other indepen-dent variables not considered in this analysis.According to Rosenthal’s fail-safe number an addi-tional 1127 studies with nonsignificant results wouldbe required to reverse our findings (i.e., support thenull hypothesis that crayfish have no ecologicaleffects) (Appendix S2).

Effects on primary producers

Nonnative crayfish had little effect on algae underexperimental conditions (Fig. 4A). Effect sizes for P.leniusculus and O. rusticus were slightly positive, butnot statistically significant (Fig. 4A). Total heteroge-neity was significantly larger than the expectedsampling error (QTotal = 22.6, p , 0.01; AppendixS3), suggesting that factors, such as experimentaldesign, geographical, or ecological factors other thancrayfish identity, may explain the heterogeneityamong effect sizes. In contrast, effects of nonnativecrayfish on macrophytes were consistently negativeacross all species individually and combined(Fig. 4B), and the total heterogeneity in effect sizesfor the random effects model was not significant(QTotal = 26.0, p = 0.17; Appendix S3), indicatingbroad similarities across species in their effects onmacrophytes. Rosenthal’s fail-safe number indicatedthat adding 278 additional studies with nonsignificantresults to the analysis could reverse our finding thatcrayfish have significant negative effects on macro-phytes.

FIG. 2. Mean (6 mean variance) estimated effect sizes(Hedges’ d, symbol scaled to precision of estimate) incomparisons of native and nonnative crayfish (Astacusastacus, Astacus italicus, Cambaroides japonicus, Orconectespropinquus, Orconectes rusticus, Orconectes virilis, Procambarusclarkii, Pacifastacus leniusculus). Nonnative species used ineach study are listed on the left, and native species are listedto the right. Nonnative crayfish demonstrate slightlygreater, but highly variable, ecological effects compared tonative crayfish. 1 = Luttenton et al. 1998, 2 = Nystrom andStrand 1996, 3 = Arce et al. 2006, 4 = Hazlett et al. 1992, 5 =

Olsen et al. 1991, 6 = Olden et al. 2009, 7 = Usio et al. 2006,8 = Ellrott et al. 2007.

FIG. 3. Mean (695% bias-corrected, bootstrap confidenceintervals) estimated effect sizes (Hedges’ d, symbol scaled toprecision of estimate) from mixed-effects models fornonnative crayfish (Orconectes rusticus, Orconectes virilis,Pacifastacus leniusculus, Procambarus clarkii). Number ofexperiments is indicated in parentheses. Nonnative crayfishhave strong negative effects on all ecological responses ofrecipient ecosystems.

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1373

Effects on invertebrates

Nonnative crayfish were responsible for largechanges in the biomass and abundance of all benthicinvertebrates (Qtotal = 27.8, p = 0.15; Fig. 5A) andsnails specifically (Qtotal = 13.7 p = 0.32; Fig. 5B)(Appendix S3), and total heterogeneity across crayfishspecies was not significant. Amphipoda, Chirono-midae, and Ephemeroptera responded negatively innonnative crayfish treatments, but the effects weresignificant only for Ephemeroptera (Fig. 5C, Appen-dix S3). In contrast, the response of Trichoptera wasslightly positive, but this effect was highly variableand not significant (Fig. 5C). Rosenthal’s fail-safenumber lends substantial support to our results.Two hundred eighty-four and 110 additional studieswith nonsignificant results could reverse our findingsof significant negative effects on all benthic inverte-brates and snails, respectively.

Effects on crayfish

Nonnative crayfish were more aggressive (i.e., morevictorious) in agonistic interactions compared tonative crayfish, but results were highly variable(Fig. 6A). Overall, nonnative crayfish did not signif-icantly influence survival of native crayfish (Fig. 6B).Total heterogeneity was not significantly differentfrom expected sampling error in random-effectsanalyses of agonistic interactions (QTotal = 2.78,

FIG. 4. Mean (695% bias-corrected, bootstrap confidenceintervals) estimated effect sizes (Hedges’ d, symbol scaledto precision of estimate) from random-effects models fornonnative crayfish (Orconectes rusticus, Pacifastacus lenius-culus, Procambarus clarkii) on algal biomass or chlorophylla (A) and macrophyte abundance or biomass (B). Numberof experiments is indicated in parentheses. Nonnativecrayfish do not affect algae, but negatively affect macro-phytes.

FIG. 5. Mean (695% bias-corrected, bootstrap confidenceintervals) estimated effect sizes (Hedges’ d, symbol scaled toprecision of estimate) from random-effects models fornonnative crayfish (Orconectes rusticus, Pacifastacus leniuscu-lus, Procambarus clarkii) on abundance or biomass of benthicmacroinvertebrates (A), gastropods (B), and selected inver-tebrate taxa (C). Number of experiments is indicated inparentheses. Nonnative crayfish have significant negativeeffects on total benthic invertebrates, gastropods, andEphemeroptera, but not Amphipoda, Chironomidae,or Trichoptera.

1374 L. A. TWARDOCHLEB ET AL. [Volume 32

p = 0.60) or survival (QTotal = 3.96, p = 0.41)(Appendix S3).

Effects on vertebrates

Nonnative crayfish had dramatic effects on thegrowth and abundances of fish (Fig. 7A) and am-phibians (Fig. 7B) in experimental settings. Experi-ments with P. clarkii were responsible for most of theoverall crayfish effect on amphibians (Fig. 7B). Totalheterogeneity was not significant for total effects onfish (QTotal = 6.48, p = 0.37) or amphibians (QTotal =

8.09, p = 0.42), indicating that sampling error explainsmost of the heterogeneity among effect sizes (Appen-dix S3). Rosenthal’s fail-safe number indicated that 42additional studies with nonsignificant effects arerequired to change our results for the effects ofcrayfish on fish and amphibians.

Effects of experimental venue

The magnitude of ecological effects was similarregardless of experimental venue. We found nosignificant differences among experiments testing foreffects on invertebrates (QTotal = 28.2, p = 0.13; Qb =

4.14, p = 0.13; Qw = 24.1, p = 0.19; Appendix S2,Table 2) and macrophytes (QTotal = 25.1, p = 0.16; Qb

= 2.33, p = 0.31; Qw = 22.7, p = 0.16; Appendix S2,Table 2).

Discussion

Crayfish often play a pivotal role in food webs byfeeding across trophic levels. They have been regard-ed as ecosystem engineers because of their burrowingactivities that increase sediment transport in loticsystems, their rapid detrital processing, and theirability to shift ecosystems from macrophyte to algaedominated (Creed and Reed 2004, Matsuzaki et al.2009, Statzner 2012). Our meta-analysis supports tovarying degrees the assessment of crayfish as ecosys-tem engineers and strong interactors that affect alllevels of food webs, thereby contributing to theecological effects of crayfish as invasive species.

FIG. 6. Mean (695% bias-corrected, bootstrap confidenceintervals) estimated effect sizes (Hedge’s d, symbol scaledto precision of estimate) from random-effects models forthe probability of winning agonistic interactions (effect sizebased on log odds ratio) (A) and surviving competitiveinteractions (Hedge’s d) (B) between nonnative (Orconecteshylas, Orconectes neglectus, Pacifastacus leniusculus, Procam-barus clarkii) and native crayfish. Number of experiments isindicated in parentheses. Nonnative crayfish win moreagonistic interactions than native crayfish, but do notreduce survival of native crayfish in competitive interac-tions.

FIG. 7. Mean (695% bias-corrected, bootstrap confidenceintervals) estimated effect sizes (Hedges’ d, symbol scaled toprecision of estimate) from random-effects models fornonnative crayfish (Orconectes virilis, Pacifastacus leniusculus,Procambarus clarkii) on fish (A) and amphibian (B) growthand abundance. Number of experiments in indicated inparentheses. Nonnative crayfish have strong negativeeffects on fish and amphibians.

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1375

Ecological effects of native and nonnative crayfish

In our comparison of the effects of native andnonnative crayfish, nonnative crayfish often hadlarger, but variable, effects on ecological responsesthan did native crayfish. Nonnative crayfish had morepositive effects than native crayfish on algae andweakly negative effects on other organisms. Too fewstudies were published in which effects of both nativeand nonnative crayfish were tested for us to calculatean overall effect size. Therefore, we are unable tojudge whether nonnative crayfish are inherentlystronger interactors in ecosystems than native cray-fish. However, in another recent meta-analysis thatcompared global effects of native and nonnativeconsumers, Paolucci et al. (2013) found that nonnativespecies have significantly larger ecological effectsthan native consumers, and nonnative invertebrateconsumers have particularly strong effects. Giventhese results and our analysis, we suggest that thepotential is high for nonnative crayfish to exhibitlarger effects than native crayfish in some ecosystems.

Experiments with O. rusticus made up most of thecomparative literature and 3 of the 4 large effect sizesin our analysis. Thus, we do not know whether theslight trends toward larger effects of nonnativespecies result from intrinsic behavioral traits or otherecological functions of O. rusticus specifically or fromgeneral traits of nonnative crayfish. Invader body orchelae size, susceptibility to predation, and behavioraltraits (e.g., burrowing activity) all may contribute todifferences in ecological effects of nonnative andnative crayfish species (reviewed by Gherardi 2006 forP. clarkii). Furthermore, nonnative crayfish mayoccupy a wider range of habitats than native crayfishand, thus, affect a greater variety of food resources(Olsson et al. 2009). We suggest that future investiga-tors explore differences in ecological traits betweenco-occurring native and nonnative crayfish thatexplain the potential absence of or heightenedecological effects of nonnative crayfish.

Our study highlights the dearth of direct compar-isons of nonnative to native crayfish and the limitedor equivocal support for greater ecological effects ofnonnative than native crayfish. It is likely that thegreatest effects of nonnative crayfish will manifest inecosystems that historically lacked crayfish or func-tionally similar organisms (e.g., freshwater shrimp,crabs) altogether. More studies of greater taxonomicbreadth and geographic scope are needed to investi-gate the potentially more-subtle ecological effects thatmay manifest when nonnative crayfish are introducedinto communities with native crayfish.

Ecological effects across trophic levels

Our meta-analytical review of the nonnative cray-fish literature revealed large and negative generalecological effects of nonnative crayfish on recipientecosystems with consistent effects among crayfishspecies. Crayfish significantly and negatively affectedbiomass and abundances of macrophytes, inverte-brates (all benthic invertebrates and snails specifical-ly), fish, and amphibians in experimental settings.These findings are consistent with results of previousmeta-analyses that showed negative effects of crayfishon invertebrates and macrophytes for selected cray-fish species and regions (McCarthy et al. 2006,Matsuzaki et al. 2009). Taken together, our resultsprovide quantitative evidence that nonnative crayfishinvasions are associated with substantial, but variable,effects across multiple levels of freshwater food webs.

Relationships between crayfish and primary pro-ducers can be complex. Nonnative crayfish havedirect negative effects on macrophyte abundance(Nystrom et al. 1999, Matsuzaki et al. 2009) andindirect positive effects on algal biomass via con-sumption of grazers and competing macrophytes(e.g., Lodge et al. 1994, Nystrom et al. 1999). However,we found little evidence for directional effects ofnonnative crayfish on algal biomass. Only O. rusticusand P. leniusculus had positive (albeit weak) effects on

TABLE 2. Results of mixed-effects model analysis comparing mean effect sizes of nonnative crayfish on invertebrates andmacrophytes in different experimental venues.

Taxonomic group Experimental venue Number of experiments Effect size 95% confidence interval

Invertebrate All 22 21.49 22.09 to 21.04Cage 10 21.33 21.99 to 20.89Laboratory 7 22.23 23.92 to 21.39Outdoor mesocosm 5 20.771 22.09 to 0.163

Macrophyte All 20 22.02 22.84 to 21.29Cage 12 21.59 22.60 to 20.748Laboratory 3 22.58 26.74 to 20.819Outdoor mesocosm 5 22.96 24.25 to 21.94

1376 L. A. TWARDOCHLEB ET AL. [Volume 32

algae. This result might be expected given that short-term experiments, which are common in the litera-ture, are not likely to capture trophic cascades thatmanifest over longer time scales. On the other hand,crayfish effects on algae may be confounded by taxon-specific responses. For example, Creed (1994) foundthat grazing crayfish reduced abundance of thefilamentous algae Cladophora by 103, releasing dia-toms and increasing their abundance 203.

Nonnative crayfish have strong negative effects onmacrophyte biomass and abundances. Coupled withthe large negative effects of crayfish on snails, theseresults suggest that the activities of nonnative crayfishmay ultimately lead to trophic cascades and shiftsfrom macrophyte- to algae-dominated systems (e.g.,Lodge et al. 1994, Smart et al. 2002). Many studieshave focused on the effects of P. clarkii on macro-phytes (reviewed in Matsuzaki et al. 2009), but ouranalysis suggests that O. rusticus and P. leniusculus aresimilarly capable of eliminating macrophyte stands(e.g., Peters et al. 2008, Usio et al. 2009).

We also found strong negative, yet variable, effectsof nonnative crayfish on benthic invertebrates. Acrossnonnative crayfish, the strongest negative effects wereon snails, suggesting that snails may be particularlysusceptible prey. Previous studies have shown dra-matic changes in benthic invertebrate communities(e.g., Nystrom et al. 1999, McCarthy et al. 2006),especially thin-shelled species of snails, in response tocrayfish over both short and long time scales (Krepset al. 2012). However, authors of few studies in ouranalysis examined effects on other invertebrate taxa,including Amphipoda, Chironomidae, Ephemerop-tera, and Trichoptera. Thus, more research is neededto assess crayfish effects on common benthic inverte-brates, which are important prey items in crayfishdiets (Momot 1995).

Crayfish have been implicated in declines of fishand amphibians throughout the world, and theseeffects are mediated by predation and indirectcompetition for habitat and prey (Ilheu et al. 2007,Reynolds 2011). According to our review, crayfish hadsignificant negative effects on benthic invertebrates,which when combined with evidence that crayfishmodify habitat conditions by reducing macrophytebiomass and abundances, supports the notion thatcrayfish decrease the abundance of shared preyresources and available habitat for fish and amphib-ians (Dorn and Mittelbach 1999, Ilheu et al. 2007).Crayfish had significantly negative effects on juvenileand adult fish and amphibian abundances andgrowth, results suggesting that crayfish have substan-tial effects on vertebrates during early and mature lifestages. Our analyses of effects on vertebrates were

based on small sample sizes, but we obtained largeRosenthal’s fail-safe numbers. Thus, our resultsappear to be relatively unbiased, and nonnativecrayfish should be expected to have negative effectson aquatic vertebrates. Amphibians and freshwaterfish are globally imperiled (Dudgeon et al. 2006), andresearch is needed to link data on nonnative crayfishinvasions and management actions aimed at mini-mizing their ecological impacts.

Interactions among crayfish

Nonnative crayfish are widely considered thelargest threat to native crayfish in regions includingJapan, North America (Taylor et al. 2007), and Europe(Lodge et al. 2000), yet little consensus existsregarding the relative importance of aggression indisplacement of native crayfish (Hobbs et al. 1989,Snyder and Evans 2006). Our analysis did not indicatethat nonnative crayfish influenced survival of nativecrayfish. However, nonnative crayfish won moreagonistic interactions than native crayfish. Nonnativecrayfish are frequently cited as more aggressive thantheir native counterparts (Gherardi and Cioni 2004),and we suggest that enhanced competitive abilitymay be a behavioral trait common to nonnativecrayfish (Pintor and Sih 2009). Increased aggressionduring encounters is hypothesized to give nonnativecrayfish an edge over native species in predatoravoidance and the pursuit of shelter and foodresources, and this edge may explain their successin displacing native crayfishes in the invasive range(Hill and Lodge 1994, Lodge et al. 2000, Olden et al.2011). However, in studies of aggression betweennative and nonnative crayfish, investigators mostoften use naıve individuals with no experience withinvaders (e.g., Larson and Magoulick 2009). Hayes etal. (2009) suggested that native crayfish with experi-ence with nonnatives fare better in aggressiveencounters than their naıve counterparts. Only asmall number of studies in our analysis examinedaggression and survival of native crayfish in thepresence of nonnative crayfish, and this area couldbenefit from further exploration to inform efforts tomitigate effects of nonnative crayfish on nativecongeners. Agonistic encounters with nonnativecrayfish may be an important component of risk tonative crayfish and should be considered in manage-ment plans for threatened or endangered species.

Comparisons of ecological effects among crayfish species

Invasion success and ecological effects vary greatlyamong nonnative crayfish (Larson and Olden 2010,Lodge et al. 2012), but the lack of significant

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1377

heterogeneity in effect sizes among crayfish species inour study suggests that species identity may be lessimportant than extrinsic factors in determining effectsof crayfish. Crayfish are opportunistic feeders that eatwhat is available in the environment. For example, P.clarkii consumes macroinvertebrates in proportion totheir seasonal availability in rice fields of Portugal(Correia 2002) and supplements this diet withvegetative material when macroinvertebrates arescarce (Correia 2003). Stable-isotope analyses, tradi-tional feeding studies, and diet analyses show thatP. leniusculus and O. rusticus prey preferentially onmacroinvertebrates but also consume large volumesof detritus and vegetative material as they areavailable (Bondar et al. 2005, Roth et al. 2006). Ourresults are consistent with these feeding studies inillustrating that the polytrophic generalist feedinghabits of nonnative crayfish affect native taxa rangingfrom macrophytes to vertebrates.

Effects of experimental venue on conclusions aboutecological effects

The design of an experiment can affect its outcomeby influencing how species interact (Skelly 2002).Specifically, small experimental enclosures may over-emphasize the importance of interactions (Gurevitchet al. 1992). Nevertheless, we did not find evidencethat venue influenced the strength of ecological effectsof nonnative crayfish. This result suggests that thetypes of venues used in experiments included in ourstudy (laboratory tanks, outdoor mesocosms, and insitu cages) can be used to provide unbiased tests ofthe effects of crayfish.

Recommendations for future research

Nonnative crayfish have been introduced to everycontinent except Antarctica (Lodge et al. 2012), but theexperimental studies considered here focused on theeffects of nonnative crayfish in only 3: North America,Europe, and Japan (Fig. 1A). In invasion ecology,geographical and taxonomic bias are common prob-lems that limit our understanding of the effects ofinvaders because data are often missing from less-studied regions like Africa and Asia and for recentlyintroduced species (Pysek et al. 2008). This bias isprevalent in the crayfish literature and limits ourunderstanding of global and regionally specific effectsof nonnative crayfish. Experimental work on nonnativecrayfish is needed from heavily invaded regions likeChina (Liu et al. 2011), and from regions likeMadagascar where invasions threaten rare and imper-iled native crayfish species (Jones et al. 2009). We foundalmost no experimental studies on nonnative crayfish

in Australia, a hotspot of global crayfish diversity andconservation need (Crandall and Buhay 2008). Austra-lian crayfish species Cherax quadricarinatus and Cheraxdestructor have been introduced within the continentwhere they threaten other Australian endemic cray-fishes (Elvey et al. 1996) and are increasingly foundglobally in locations ranging from southeast Asia(Ahyong and Yeo 2007) to Europe (Scalici et al. 2009).Experimental studies are needed to gauge whethereffects of these emerging invaders resemble those ofmore widespread and commonly studied nonnativecrayfish originating from North America.

Most experiments used in our study were done totest effects of 4 nonnative crayfish (Fig. 1B–E; colorversion available online from: http://dx.doi.org/10.1899/12-203.1.s4), all originating from North Americaand members of the Astacidae and Cambaridae, andsome responses were dominated by effects of a singlespecies. For example, our study revealed substantialecological impacts of nonnative crayfish on amphib-ians, but almost all available experiments were donewith P. clarkii (e.g., Cruz and Rebelo 2005, Cruz et al.2006, Gamradt and Kats 1996). Our meta-analysisshows that competition and predation by nonnativecrayfish can have major consequences for freshwatervertebrates, and research is needed to assess andprevent effects of other nonnative crayfish species onamphibian and fish populations. Nonnative crayfishin need of study include the emerging Australianinvaders (above), the parthenogenetic Marmorkrebs(Jimenez and Faulkes 2011), and the many extralimitalspecies that have been introduced in regions of highcrayfish diversity like the southeastern USA (Larsonand Olden 2010). The many studies on nonnativecrayfish like P. clarkii (5 studies in Portugal) and O.rusticus (14 studies in the USA) are a product ofresearcher- and region-specific programs in whichthese widespread species are used as model organ-isms in invasion biology, an invaluable contributionto our understanding of effects of invasive crayfishthat would benefit from more research in otherregions and on overlooked or emerging invasivespecies.

Conclusions

Our quantitative meta-analysis supports the con-clusions of past qualitative reviews that invasion bynonnative crayfish can have major effects throughoutfreshwater communities and ecosystems (Lodge et al.2000, 2012, Snyder and Evans 2006, Gherardi 2010).The effects of nonnative crayfish range from reducingbasal resources like aquatic macrophytes to preyingdirectly on invertebrates to negatively affecting

1378 L. A. TWARDOCHLEB ET AL. [Volume 32

amphibians and fish. Nonnative crayfish did notconsistently increase algae, a result supporting conclu-sions by Usio (2000) that omnivorous crayfish candecouple trophic cascades by consuming both fresh-water invertebrates and their preferred basal resources.We caution that crayfish invasions should be expectedto have complex and context-dependent indirecteffects in freshwater ecosystems (e.g., stable stateshifts; Matsuzaki et al. 2009) because crayfish havepotentially strong interactions at all trophic levels.

Our study revealed geographic and taxonomicbiases in studies of the effects of nonnative crayfish,but results were generally, albeit not universally,consistent across crayfish species. We have identifieda number of directions for future research on theecological effects of nonnative crayfish, but emphasizethat ample evidence already exists to support pre-vention and discouragement of introduction of theseorganisms beyond their native ranges.

Acknowledgements

We thank the researchers who shared their data forthis analysis, including E. Almeida, F. Banha, V.Brenneis, S. Hudina, P. Johnson, T. Jonsson, K. Klose,K. Mueller, J. Peters, L. Pintor, and N. Usio.Comments from B. Helms and 2 anonymous refereesgreatly improved the final manuscript. Financialsupport was provided by a National Science Founda-tion Graduate Research Fellowship (LAT), the Uni-versity of Washington H. Mason Keeler EndowedProfessorship (JDO), and the US EnvironmentalProtection Agency Science to Achieve Results (STAR)Program (grant 833834) to LAT and JDO. An openaccess database that includes publications and sum-mary data used in our analyses can be found onlineat: http://dx.doi.org/10.6084/m9.figshare.746992.

Literature Cited

ADAMS, D. C., J. GUREVITCH, AND M. S. ROSENBERG. 1997.Resampling tests for meta-analysis of ecological data.Ecology 78:1277–1283.

AHYONG, S. T., AND D. C. J. YEO. 2007. Feral populations of theAustralian red-claw crayfish (Cherax quadricarinatus vonMartens) in water supply catchments of Singapore.Biological Invasions 9:943–946.

ARCE, J. A., F. ALONSO, E. RICO, AND A. CAMACHO. 2006. Astudy on the possible effect of two crayfish species onepilithic algae in a mountain stream from central Spain.Bulletin Francais de la Peche et de la Pisciculture 380–381:1133–1144.

BOBELDYK, A. M., AND G. A. LAMBERTI. 2008. A decade afterinvasion: evaluating the continuing effects of rustycrayfish on a Michigan river. Journal of Great LakesResearch 34:265–275.

BONDAR, C. A., K. BOTTRIELL, K. ZERON, AND J. S. RICHARDSON.2005. Does trophic position of the omnivorous signalcrayfish (Pacifastacus leniusculus) in a stream food webvary with life history stage or density? Canadian Journalof Fisheries and Aquatic Sciences 62:2632–2639.

CASSEY, P., T. M. BLACKBURN, R. P. DUNCAN, AND J. L.LOCKWOOD. 2005. Lessons from the establishment ofexotic species: a meta-analytical case study using birds.Journal of Animal Ecology 74:250–258.

CHARLEBOIS, P. M., AND G. A. LAMBERTI. 1996. Invadingcrayfish in a Michigan stream: direct and indirect effectson periphyton and macroinvertebrates. Journal of theNorth American Benthological Society 15:551–563.

CORREIA, A. M. 2002. Niche breadth and trophic diversity:feeding behaviour of the red swamp crayfish (Procam-barus clarkii) towards environmental availability ofaquatic macroinvertebrates in a rice field (Portugal).Acta Oecologica 23:421–429.

CORREIA, A. M. 2003. Food choice by the introduced crayfishProcambarus clarkii. Annales Zoologici Fennici 40:517–528.

CORREIA, A. M., AND P. M. ANASTACIO. 2008. Shifts in aquaticmacroinvertebrate biodiversity associated with thepresence and size of an alien crayfish. EcologicalResearch 23:729–734.

CRANDALL, K. A., AND J. E. BUHAY. 2008. Global diversity ofcrayfish (Astacidae, Cambaridae, and Parastacidae -Decapoda) in freshwater. Hydrobiologia 595:295–301.

CREED, R. P. 1994. Direct and indirect effects of crayfishgrazing in a stream community. Ecology 75:2091–2103.

CREED, R. P., AND J. M. REED. 2004. Ecosystem engineering bycrayfish in a headwater stream community. Journal ofthe North American Benthological Society 23:224–236.

CRUZ, M. J., S. PASCOAL, M. TEJEDO, AND R. REBELO. 2006.Predation by an exotic crayfish, Procambarus clarkii, onnatterjack toad, Bufo calamita, embryos: its role on theexclusion of this amphibian from its breeding ponds.Copeia 2006:274–280.

CRUZ, M. J., AND R. REBELO. 2005. Vulnerability of southwestIberian amphibians to an introduced crayfish, Procam-barus clarkii. Amphibia-Reptilia 26:293–303.

DORN, N. J., AND G. G. MITTELBACH. 1999. More than predatorand prey: a review of interactions between fish andcrayfish. Vie et Milieu 49:229–237.

DUDGEON, D., A. H. ARTHINGTON, M. O. GESSNER, Z.-I.KAWABATA, D. J. KNOWLER, C. LEVEQUE, R. J. NAIMAN,A.-H. PRIEUR-RICHARD, D. SOTO, M. L. J. STIASSNY, AND C.A. SULLIVAN. 2006. Freshwater biodiversity: importance,threats, status and conservation challenges. BiologicalReviews 81:163–182.

ELLROTT, B. J., J. E. MARSDEN, J. D. FITZSIMONS, J. L. JONAS, AND

R. M. CLARAMUNT. 2007. Effects of temperature anddensity on consumption of trout eggs by Orconectespropinquus and O. rusticus. Journal of Great LakesResearch 33:7–14.

ELTON, C. S. 1958. The ecology of invasions by animals andplants. University of Chicago Press, Chicago, Illinois.

ELVEY, W., A. M. M. RICHARDSON, AND L. BERMUTA. 1996.Interactions between the introduced yabby, Cherax

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1379

destructor, and the endemic crayfish, Astacopsis franklinii,in Tasmanian streams. Freshwater Crayfish 11:349–363.

FEMINELLA, J. W., AND V. H. RESH. 1989. Submersedmacrophytes and grazing crayfish: an experimentalstudy of herbivory in a California freshwater marsh.Holarctic Ecology 12:1–8.

GAMRADT, S. C., AND L. B. KATS. 1996. Effect of introducedcrayfish and mosquitofish on California newts. Conser-vation Biology 10:1155–1162.

GAMRADT, S. C., L. B. KATS, AND C. B. ANZALONE. 1997.Aggression by non-native crayfish deters breeding inCalifornia newts. Conservation Biology 11:793–796.

GHERARDI, F. 2006. Crayfish invading Europe: the case studyof Procambarus clarkii. Marine and Freshwater Behaviourand Physiology 39:175–191.

GHERARDI, F. 2010. Invasive crayfish and freshwater fishes ofthe world. Revue Scientifique et Technique–OfficeInternational des Epizooties 29:241–254.

GHERARDI, F., AND A. CIONI. 2004. Agonism and interferencecompetition in freshwater decapods. Behaviour 141:1297–1324.

GUREVITCH, J., AND L. V. HEDGES. 1999. Statistical issues inecological meta-analyses. Ecology 80:1142–1149.

GUREVITCH, J., K. L. MORROW, A. WALLACE, AND J. S. WALSH.1992. A meta-analysis of competition in field experi-ments. American Naturalist 140:539–572.

HAYES, N. M., K. J. BUTKAS, J. D. OLDEN, AND M. J. VANDER

ZANDEN. 2009. Behavioural and growth differencesbetween experienced and naıve populations of a nativecrayfish in the presence of invasive rusty crayfish.Freshwater Biology 54:1876–1887.

HAZLETT, B. A., F. E. ANDERSON, L. A. ESMAN, C. STAFFORD, AND

E. MUNRO. 1992. Interspecific behavioral ecology of thecrayfish Orconectes rusticus. Journal of FreshwaterEcology 7:69–76.

HEDGES, L. V. 1981. Distribution theory for Glass’s estimatorof effect size and related estimators. Journal of Educa-tional Statistics 6:107–128.

HILL, A. M., AND D. M. LODGE. 1994. Diel changes in resourcedemand: competition and predation in species replace-ment among crayfishes. Ecology 75:2118–2126.

HOBBS, H. H., AND J. P. JASS. 1988. The crayfishes and shrimpof Wisconsin. (Decapoda: Palaemonidae, Cambaridae).Special Publications in Biology and Geology, No. 5.Milwaukee Public Museum, Milwaukee, Wisconsin.

HOBBS, H. H., J. P. JASS, AND J. V. HUNER. 1989. A review ofglobal crayfish introductions with particular emphasison two North-American species (Decapoda, Cambar-idae). Crustaceana 56:299–316.

ILHEU, M., J. M. BERNARDO, AND S. FERNANDES. 2007. Predationof invasive crayfish on aquatic vertebrates: the effect ofProcambarus clarkii on fish assemblages in Mediterra-nean temporary streams. Pages 543–558 in F. Gherardi(editor). Biological invaders in inland waters: profiles,distribution, and threats. Springer, Dordrecht, TheNetherlands.

JIMENEZ, S. A., AND Z. FAULKES. 2011. Can the parthenogeneticmarbled crayfish Marmorkrebs compete with other

crayfish species in fights? Journal of Ethology 29:115–120.

JONES, J. P. G., J. R. RASAMY, A. HARVEY, A. TOON, B. OIDTMANN,M. H. RANDRIANARISON, N. RAMINOSOA, AND O. R.RAVOAHANGIMALALA. 2009. The perfect invader: a parthe-nogenic crayfish poses a new threat to Madagascar’sfreshwater diversity. Biological Invasions 11:1475–1482.

KREPS, T. A., A. K. BALDRIDGE, AND D. M. LODGE. 2012. Theimpact of an invasive predator (Orconectes rusticus) onfreshwater snail communities: insights into habitat-specific effects from a multilake long-term study.Canadian Journal of Fisheries and Aquatic Sciences 69:1164–1173.

LAJEUNESSE, M. J., AND M. R. FORBES. 2003. Variable reportingand quantitative reviews: a comparison of three meta-analytical techniques. Ecology Letters 6:448–454.

LARSON, E. R., AND D. D. MAGOULICK. 2009. Does juvenilecompetition explain displacement of a native crayfish byan introduced crayfish? Biological Invasions 11:725–735.

LARSON, E. R., AND J. D. OLDEN. 2010. Latent extinction andinvasion risk of crayfishes in the southeastern UnitedStates. Conservation Biology 24:1099–1110.

LIU, X., Z. GUO, Z. KE, S. WANG, AND Y. LI. 2011. Increasingpotential risk of a global aquatic invader in Europe incontrast to other continents under future climatechange. PloS One 6:e18429.

LODGE, D. M., A. DEINES, F. GHERARDI, D. C. J. YEO, T. ARCELLA,A. K. BALDRIDGE, M. A. BARNES, W. L. CHADDERTON, J. L.FEDER, C. A. GANTZ, G. W. HOWARD, C. L. JERDE, B. W.PETERS, J. A. PETERS, L. W. SARGENT, C. R. TURNER, M. E.WITTMANN, AND Y. ZENG. 2012. Global introductions ofcrayfishes: evaluating the impact of species invasions onecosystem services. Annual Review of Ecology, Evolu-tion, and Systematics 43:449–472.

LODGE, D. M., M. W. KERSHNER, J. E. ALOI, AND A. P. COVICH.1994. Effects of an omnivorous crayfish (Orconectesrusticus) on a freshwater littoral food web. Ecology 75:1265–1281.

LODGE, D. M., C. TAYLOR, D. HOLDICH, AND J. SKURDAL. 2000.Nonindigenous crayfishes threaten North Americanfreshwater biodiversity: lessons from Europe. Fisheries25(8):7–20.

LUTTENTON, M. R., M. J. HORGAN, AND D. M. LODGE. 1998.Effects of three Orconectes crayfishes on epilithic micro-algae: a laboratory experiment. Crustaceana 71:845–855.

MACK, R. N., D. SIMBERLOFF, W. M. LONSDALE, H. EVANS, M.CLOUT, AND F. A. BAZZAZ. 2000. Biotic invasions, globalconsequences, and control. Ecological Applications 10:689–710.

MATSUZAKI, S. S., N. USIO, N. TAKAMURA, AND I. WASHITANI.2009. Contrasting impacts of invasive engineers onfreshwater ecosystems: an experiment and meta-analy-sis. Oecologia (Berlin) 158:673–686.

MCCARTHY, J. M., C. L. HEIN, J. D. OLDEN, AND M. J. VANDER

ZANDEN. 2006. Coupling long-term studies with meta-analysis to investigate impacts of non-native crayfishon zoobenthic communities. Freshwater Biology 51:224–235.

1380 L. A. TWARDOCHLEB ET AL. [Volume 32

MØLLER, A. P., AND M. D. JENNIONS. 2002. How much variancecan be explained by ecologists and evolutionarybiologists? Oecologia (Berlin) 132:492–500.

MOMOT, W. T. 1995. Redefining the role of crayfish in aquaticecosystems. Reviews in Fisheries Science 3:33–63.

NYSTROM, P., C. BRONMARK, AND W. GRANELI. 1999. Influenceof an exotic and a native crayfish species on a littoralbenthic community. Oikos 85:545–553.

NYSTROM, P., AND J. A. STRAND. 1996. Grazing by a native andan exotic crayfish on aquatic macrophytes. FreshwaterBiology 36:673–682.

OLSEN, M. T., D. M. LODGE, G. M. CAPELLI, AND R. J. HOULIHAN.1991. Mechanisms of impact of an introduced crayfish(Orconectes rusticus) on littoral congeners, snails, andmacrophytes. Canadian Journal of Fisheries and Aquat-ic Sciences 48:1853–1861.

OLDEN, J. D., E. R. LARSON, AND M. C. MIMS. 2009. Home-fieldadvantage: native signal crayfish (Pacifastacus leniuscu-lus) out consume newly introduced crayfishes forinvasive Chinese mystery snail (Bellamya chinensis).Aquatic Ecology 43:1073–1084.

OLDEN, J. D., M. J. VANDER ZANDEN, AND P. T. J. JOHNSON. 2011.Assessing ecosystem vulnerability to invasive rustycrayfish (Orconectes rusticus). Ecological Applications21:2587–2599.

OLSSON, K., P. STENROTH, P. NYSTROM, AND W. GRANELI. 2009.Invasions and niche width: does niche width of anintroduced crayfish differ from a native crayfish?Freshwater Biology 54:1731–1740.

PAOLUCCI, E. M., H. J. MACISAAC, AND A. RICCIARDI. 2013.Origin matters: alien consumers inflict greater damageon prey populations than do native consumers. Diver-sity and Distributions (in press). doi:10.1111/ddi.12073

PARKER, I. M., D. SIMBERLOFF, W. M. LONSDALE, K. GOODELL, M.WONHAM, P. M. KAREIVA, M. H. WILLIAMSON, B. VON HOLLE,P. B. MOYLE, J. E. BYERS, AND L. GOLDWASSER. 1999. Impact:toward a framework for understanding the ecologicaleffects of invaders. Biological Invasions 1:3–19.

PERRY, W. L., J. L. FEDER, G. DWYER, AND D. M. LODGE. 2001.Hybrid zone dynamics and species replacement be-tween Orconectes crayfishes in a northern Wisconsinlake. Evolution 55:1153–1166.

PERRY, W. L., D. M. LODGE, AND G. A. LAMBERTI. 2000. Crayfish(Orconectes rusticus) impacts on zebra mussel (Dreissenapolymorpha) recruitment, other macroinvertebrates andalgal biomass in a lake-outlet stream. American Mid-land Naturalist 144:308–316.

PETERS, J. A., T. KREPS, AND D. M. LODGE. 2008. Assessing theimpacts of rusty crayfish (Orconectes rusticus) on sub-mergent macrophytes in a north-temperate U.S. lakeusing electric fences. American Midland Naturalist 159:287–297.

PINTOR, L. M., AND A. E. SIH. 2009. Differences in growth andforaging behavior of native and introduced populationsof an invasive crayfish. Biological Invasions 11:1895–1902.

PULLIN, A. S., AND G. B. STEWART. 2006. Guidelines forsystematic review in conservation and environmentalmanagement. Conservation Biology 20:1647–1656.

PYSEK, P., D. M. RICHARDSON, J. PERGL, V. JAROSIK, Z. SIXTOVA,AND E. WEBER. 2008. Geographical and taxonomic biasesin invasion ecology. Trends in Ecology and Evolution23:237–244.

RABALAIS, M. R., AND D. D. MAGOULICK. 2006. Is competitionwith the invasive crayfish Orconectes neglectus chaeno-dactylus responsible for the displacement of the nativecrayfish Orconectes eupunctus? Biological Invasions 8:1039–1048.

REYNOLDS, J. D. 2011. A review of ecological interactionsbetween crayfish and fish, indigenous and introduc-ed. Knowledge and Management of Aquatic Ecosys-tems 401:article 10. doi: http://dx.doi.org/10.1051/kmae/2011024

RICCIARDI, A., AND S. K. ATKINSON. 2004. Distinctivenessmagnifies the impact of biological invaders in aquaticecosystems. Ecology Letters 7:781–784.

ROSENBERG, M. S. 2005. The file-drawer problem revisited: ageneral weighted method for calculating fail-safenumbers in meta-analysis. Evolution 59:464–468.

ROSENBERG, M. S., D. C. ADAMS, AND J. GUREVITCH. 1999.MetaWin statistical software for meta-analysis. Version2.0. Stony Brook, New York.

ROSENTHAL, R. 1979. The ‘‘file drawer problem’’ andtolerance for null results. Psychological Bulletin 86:638–641.

ROTH, B. M., C. L. HEIN, AND M. J. VANDER ZANDEN. 2006.Using bioenergetics and stable isotopes to assess thetrophic role of rusty crayfish (Orconectes rusticus) in lakelittoral zones. Canadian Journal of Fisheries andAquatic Sciences 63:335–344.

SCALICI, M., S. CHIESA, F. GHERARDI, M. RUFFINI, G. GIBERTINI,AND F. N. MARZANO. 2009. The new threat to Italianinland waters from the alien crayfish ‘‘gang’’: theAustralian Cherax destructor Clark, 1936. Hydrobiologia632:341–345.

SCHLAEPFER, M. A., D. F. SAX, AND J. D. OLDEN. 2011. Thepotential conservation value of non-native species.Conservation Biology 25:428–437.

SKELLY, D. K. 2002. Experimental venue and estimation ofinteraction strength. Ecology 83:2097–2101.

SMART, A. C., D. M. HARPER, F. MALAISSE, S. SCHMITZ, S. COLEY,AND A. C. G. DE BEAUREGARD. 2002. Feeding of the exoticLouisiana red swamp crayfish, Procambarus clarkii(Crustacea, Decapoda), in an African tropical lake: LakeNaivasha, Kenya. Hydrobiologia 488:129–142.

SNYDER, W. E., AND E. W. EVANS. 2006. Ecological effects ofinvasive arthropod generalist predators. Annual Reviewof Ecology, Evolution, and Systematics 37:95–122.

SORTE, C. J. B., I. IBANEZ, D. M. BLUMENTHAL, N. MOLINARI, L. P.MILLER, E. D. GROSHOLZ, J. M. DIEZ, C. M. D’ANTONIO, J. D.OLDEN, S. J. JONES, AND J. S. DUKES. 2012. Poised toprosper: a cross-system comparison of climate changeimpacts on native and non-native species performance.Ecology Letters 16:261–270.

STATZNER, B. 2012. Geomorphological implications of engi-neering bed sediments by lotic animals. Geomorpholo-gy 157–158:49–65.

2013] ECOLOGICAL EFFECTS OF NONNATIVE CRAYFISH 1381

TAYLOR, C. A., G. A. SCHUSTER, J. E. COOPER, R. J. DISTEFANO,A. G. EVERSOLE, P. HAMR, H. H. HOBBS, H. W. ROBISON,C. E. SKELTON, AND R. F. THOMA. 2007. A reassessment ofthe conservation status of crayfishes of the United Statesand Canada after 10+ years of increased awareness.Fisheries 32:372–389.

THOMSEN, M. S., J. D. OLDEN, T. WERNBERG, J. N. GRIFFIN, AND

B. R. SILLIMAN. 2011. A broad framework to organize andcompare ecological invasion impacts. EnvironmentalResearch 111:899–908.

USIO, N. 2000. Effects of crayfish on leaf processing andinvertebrate colonisation of leaves in a headwaterstream: decoupling of a trophic cascade. Oecologia(Berlin) 124:608–614.

USIO, N., R. KAMIYAMA, A. SAJI, AND N. TAKAMURA. 2009. Size-dependent impacts of invasive alien crayfish on a littoralmarsh community. Biological Conservation 142:1480–1490.

USIO, N., K. SUZUKI, M. KONISHI, AND S. NAKANO. 2006. Alienvs. endemic crayfish: roles of species identity in

ecosystem functioning. Archiv fur Hydrobiologie 166:1–21.

VAN KLEUNEN, M., E. WEBER, AND M. FISCHER. 2010. A meta-analysis of trait differences between invasive and non-invasive plant species. Ecology Letters 13:235–245.

VILA, M., C. BASNOU, P. PYSEK, M. JOSEFSSON, P. GENOVESI, S.GOLLASCH, W. NENTWIG, S. OLENIN, A. ROQUES, D. ROY,P. E. HULME, DELIVERING ALIEN INVASIVE SPECIES INVENTORIES

FOR EUROPE (DAISIE) PARTNERS. 2009. How well do weunderstand the impacts of alien species on ecosystemservices? A pan-European, cross-taxa assessment. Fron-tiers in Ecology and the Environment 8:135–144.

VITOUSEK, P. M., C. M. D’ANTONIO, L. L. LOOPE, AND R.WESTBROOKS. 1996. Biological invasions as global envi-ronmental change. American Scientist 84:468–478.

Received: 11 December 2012Accepted: 19 August 2013

1382 L. A. TWARDOCHLEB ET AL. [Volume 32