Beaver dams shift desert fish assemblages towarddominance by non-native species (Verde River,Arizona, USA)Polly P. Gibson1, Julian D. Olden1, Matthew W. O’Neill21School of Aquatic and Fishery Sciences, University of Washington, 1122 NE Boat St., Seattle, WA 98105, USA2Bubbling Ponds Native Fish Conservation Facility, Arizona Game and Fish Department, 1600 N. Page Springs Rd., Cornville, AZ 86325, USA

Accepted for publication April 13, 2014

Abstract – The reintroduction of beaver (Castor canadensis) into arid and semi-arid rivers is receiving increasingmanagement and conservation attention in recent years, yet very little is known about native versus non-native fishoccupancy in beaver pond habitats. Streams of the American Southwest support a highly endemic, highlyendangered native fish fauna and abundant non-native fishes, and here we investigated the hypothesis that beaverponds in this region may lead to fish assemblages dominated by non-native species that favour slower-water habitat.We sampled fish assemblages within beaver ponds and within unimpounded lotic stream reaches in the mainstemand in tributaries of the free-flowing upper Verde River, Arizona, USA. Non-native fishes consistently outnumberednative species, and this dominance was greater in pond than in lotic assemblages. Few native species were recordedwithin ponds. Multivariate analysis indicated that fish assemblages in beaver ponds were distinct from those in loticreaches, in both mainstem and tributary locations. Individual species driving this distinction included abundant non-native green sunfish (Lepomis cyanellus) and western mosquitofish (Gambusia affinis) in pond sites, and nativedesert sucker (Catostomus clarkii) in lotic sites. Overall, this study provides the first evidence that, relative tounimpounded lotic habitat, beaver ponds in arid and semi-arid rivers support abundant non-native fishes; theseponds could thus serve as important non-native source areas and negatively impact co-occurring native fishpopulations.

Key words: Castor canadensis; desert fishes; assemblage structure; impoundments; microhabitat use; dryland stream

Introduction

Management of freshwater ecosystems takes place atthe intersection of numerous ecological and socialtrends, which can create novel challenges for naturalresource conservation and policy. Recent decadeshave witnessed growing interest in employing ormimicking the remarkable engineering abilities ofbeaver (Castor canadensis and C. fiber) in streamrestoration efforts (Pollock et al. 2007; DeVries et al.2012). After widespread extirpations of North Ameri-can beaver during several centuries of intense furtrade, hunting restrictions combined with deliberatereintroductions contributed to successful beaverrecovery (Pollock et al. 2003). However, restoration

of beaver populations also raises new questions forecosystem management, especially within arid andsemi-arid environments (Gibson & Olden 2014). Forexample, the introduction and continuing spread ofinvasive fishes pose a substantial threat to freshwaterecosystems (Cucherousset & Olden 2011), and it isunclear how native and non-native fish species areusing beaver pond habitat, and thus how contempo-rary fish assemblages might respond to changes inthe abundance of beaver ponds as beaver return tomore of their historical range.Beavers are widely recognised as ecosystem engi-

neers with major influence on the structure of aquaticecosystems. Beaver dams create distinct lentic habitatpatches within the lotic stream corridor; these beaver

Correspondence: P. P. Gibson, School of Aquatic and Fishery Sciences, University of Washington, 1122 NE Boat St., Seattle, WA 98105, USA.E-mail: ppgibson@uw.edu

doi: 10.1111/eff.12150 1

Ecology of Freshwater Fish 2014 � 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

ECOLOGY OFFRESHWATER FISH

ponds promote landscape heterogeneity and alterstream hydrology, sediment dynamics, nutrientcycling, and aquatic and riparian biotic communities(Rosell et al. 2005). Previous research has shown thatbeaver pond habitat can have substantial conse-quences for stream fish populations: individual spe-cies, particularly salmonids, have shown highergrowth, survival or abundance within beaver pondsrelative to unimpounded stream habitat (Murphyet al. 1989), and some species preferentially selectbeaver pond habitats during certain life stages orenvironmental conditions (e.g., Schlosser 1995). Inaddition to upstream pond habitat, high-velocity habi-tat immediately downstream of beaver dams mayfavour fluvial specialists (Smith & Mather 2013).Beaver ponds can influence composition of fishassemblages at scales ranging from individual ponds(Snodgrass & Meffe 1999) to entire drainages (Sch-losser & Kallemeyn 2000).Results demonstrating benefits of beaver ponds for

salmonid fishes (e.g., Pollock et al. 2004) tend tolead to the assumption that beaver dams will improvehabitat for native fish. However, this remains anuntested hypothesis for many systems (Kemp et al.2012), especially in the context of invasion by non-native fish. In southern Chile, where both beaver andrainbow trout (Oncorhynchus mykiss) are non-native,the density of native puye (Galaxias maculatus) wasfound to be substantially higher in stream reacheswith beaver dams, regardless of whether trout werepresent (Moorman et al. 2009). Habitat heterogeneitycreated by beaver dams and backwater habitat alongthe Provo River (Utah) supported several species ofnative fish in a system otherwise dominated by non-native brown trout (Salmo trutta) (Billman et al.2012). In another Utah stream, native cutthroat trout(Oncorhynchus clarkii utah) moved across beaverdams more frequently than non-native brown trout orbrook trout (Salvelinus fontinalis) (Lokteff et al.2013). These studies reveal that beaver dams canhave complex effects on native and non-native fish,but their influence on the overall structure of contem-porary fish assemblages containing both native andnon-native species is unknown.Community dynamics between native and non-

native fishes are particularly relevant in arid andsemi-arid rivers of the American Southwest. Thenative fish fauna of the lower Colorado River Basin(LCRB) is both highly endemic and highly endan-gered (Olden & Poff 2005): two-thirds of the extantnative fish species in Arizona, for example, are feder-ally listed under the US Endangered Species Act(Turner & List 2007). Non-native fishes pose a pri-mary threat to the persistence of the native fish fauna.Numerous species have been introduced to theLCRB, and spread of non-native fish has been

accompanied by dramatic declines of native fish pop-ulations (Olden & Poff 2005; Rinne & Miller 2006).Predation by non-native fish on native species,accompanied by competition and indirect effects ofpredation risk, is considered the primary mechanismfor the widespread replacement of native fish (Minck-ley et al. 2003). Numerous studies have documenteddirect predation, agonistic interactions and shifts inresource use by native fishes in response to non-natives, including both large-bodied predators andabundant small insectivores (Cucherousset & Olden2011). Non-native crayfishes (primarily Orconectesvirilis and Procambarus clarkii) have also been intro-duced and spread throughout the LCRB, althoughhistorically crayfish were absent from the entire Col-orado River Basin (Martinez 2012; Moody & Taylor2012). Effective management for conservation ofnative fish requires understanding how habitat andenvironmental features can influence their coexis-tence with non-native fish and crayfish.Very little is known of the influence of beaver

dams on fish assemblages in the LCRB. Historically,beaver dams were abundant in parts of the basin(Gibson & Olden 2014), and presumably beaverponds would have been occupied by native fishes.Current reintroductions of extirpated beaver popula-tions are usually motivated by growing recognition ofthe potential value of beaver engineering for achiev-ing conservation goals (e.g., Fredlake 1997). How-ever, within the contemporary landscape, promotionof beaver dams may also have unexpected conse-quences: natural history and previous studies suggestthat the habitat conditions created by beaver pondsare likely to benefit some non-native fishes in theLCRB. Many of the common non-native species areassociated with pool habitat or slow current velocities(Olden et al. 2006; Rinne & Miller 2006). In a long-term study of fish assemblages in the Gila River sys-tem (New Mexico), higher densities of non-nativefish were associated with low flows and greater abun-dance of pools (Propst et al. 2008). Rinne and Miller(2006) conclude that deep (>2 m) pools provide opti-mum habitat for large non-native predatory fishessuch as bass and catfishes, while lack of poolsreduces their abundance. Additionally, beaver pondsmay increase success of non-native fishes by mitigat-ing the effects of extreme high and low flows. Deeppools behind beaver dams could promote survival –especially of large-bodied species – through harshdrought conditions, while the physical beaver damstructure could provide immediate refuge fromfloods. Although these hydrological effects of beaverdams would apply to native and non-native fishesalike, non-native fishes lack an evolutionary historywith the region’s highly variable streamflow (Oldenet al. 2006), and thus non-native species might bene-

2

Gibson et al.

fit disproportionately from mitigation of these hydro-logical disturbances.The strong impacts of non-native fishes on native

populations and the potential for beaver ponds to pro-mote non-native species suggest the general hypothe-sis that beaver dams in the LCRB may lead to fishassemblages dominated by non-native species, withpotential negative effects on native fishes. In thisstudy, we investigate how beaver ponds influence thestructure of mixed native and non-native fish assem-blages in the Verde River system in Arizona. Specifi-cally, we address the following questions: (i) Doesfish assemblage structure differ between beaver pondand lotic habitat types? (ii) How do native and non-native species differentially occupy these habitats?(iii) Do fish assemblage structure and the influenceof beaver ponds differ between the mainstem riverand its tributary streams?

Methods

Study area

The Verde River, a semi-arid tributary of the LCRB,drains over 17,000 km2 of central Arizona (Fig. 1).The perennial mainstem river runs approximately300 km from its headwaters in the Big Chino Wash(elevation = 1325 m) to its confluence with the SaltRiver north of Phoenix, AZ (elevation = 402 m),while numerous perennial tributaries contribute run-off from the southwestern edge of the Colorado Pla-teau (Rinne 2005; Blasch et al. 2006). Our studyregion included stream reaches in the upper VerdeRiver mainstem and in five major tributaries(Fig. 1). The upper Verde River is largely free-flow-ing, a rarity for perennial rivers in the Desert South-west. However, several diversion dams and smallimpoundments are present in the watershed, primar-ily on tributaries. Development is limited within theVerde system, and primary anthropogenic distur-bances are livestock grazing and reduction in baseflows as a result of groundwater withdrawals.Hydrology in the upper Verde River mainstem ischaracterised by relatively steady, spring-fed baseflow, with high-flow events that vary in magnitudeand timing among years in response to storm run-off. Tributary hydrology, relative to the mainstem, ismore variable and includes more snowmelt run-off.Mean annual streamflow in the upper mainstem was1.2 m3�s�1 for the period 1963–2004; streamflow inthe four perennial and one ephemeral (Dry BeaverCreek) tributaries examined in this study rangedfrom 0.9 to 2.3 m3�s�1 during a similar period ofrecord (1961–2013 and 1981–2013, respectively)(Blasch et al. 2006). Beavers are common but bea-ver dams are spatially clustered throughout the

system (J. Agyagos, pers. comm.; P. Gibson,unpublished data).The Verde River system, especially the headwaters

of the mainstem, is an important site for conservationof native fish in Arizona (Turner & List 2007; Poolet al. 2013). At least twelve fish species were nativeto the system, but the fish assemblage is changingrapidly, and only five native species have beenobserved since 1997 (Rinne 2005). Many of the ori-ginal native species have been absent from the VerdeRiver system for decades, but federally endangeredspikedace (Meda fulgida) was last documented in theVerde River in 1997, and its current status in theVerde system is uncertain. Two other small-bodiednative species, longfin dace (Agosia chrysogaster)and speckled dace (Rhinichthys osculus), havebecome rare in the Verde River mainstem, and nativeroundtail chub (Gila robusta) has been proposed forfederal listing. Numerous non-native fishes are pres-ent in the Verde system, including several species ofcentrarchids, ictalurid catfishes and minnows(Table 1; Rinne 2005). Fish community compositionof the Verde River varies widely among years, butnon-native fish have generally outnumbered nativesfor the past two decades. Management goals for fishassemblages of the upper Verde River include main-tenance of current native populations and reintroduc-tion of extirpated species, including spikedace (Rinne2005, 2012).

Site selection

We surveyed fish assemblages in two different habi-tat types within the Verde River mainstem and fivetributaries: (i) beaver ponds formed by dams acrossthe main stream channel (representing the most com-mon position of dams in this system); and (ii) streamreaches without beaver dams (lotic sites). Focusingon stream segments with a high probability of beaverdam presence (as suggested by regional managers),continuous longitudinal census of 64 km of streamsin the Verde River system located 37 functional (i.e.,impounding water) beaver dams in April 2012 (dataavailable at http://dx.doi.org/10.6084/m9.figshare.867703). The highest density of beaver dams andthe largest and most complex ponds were locatedon the mainstem (23 dams in 24.0 km of stream),while dams on the tributaries were less common (14dams in 40.4 km) and usually smaller than mainstemdams.Pond sampling sites (mainstem n = 6; each tribu-

tary n = 0–3) were designated proportionally to pondabundance within a given stream (Fig. 1); fish sam-pling occurred in the first 100 m upstream of eachdam. With one exception, all ponds extended beyondthe sampled 100 m: on average, ponds were 180 m

3

Beaver dams shift desert fish assemblages

long (range = 90–250 m) with a maximum depth of1.5 m (range = 1.1–2.2 m). Lotic sampling sites(mainstem n = 10; each tributary n = 3–4) were dis-tributed throughout the stream segments that hadbeen searched for beaver dams, at distances rangingfrom 40 to 2650 m to the nearest dam (when damswere present). At each lotic site, we designated asampling reach approximately 100 m long(range = 80–115 m), which typically contained 1–2pool-riffle-run sequences.

Sampling methods

We sampled 26 lotic and 12 pond sites during latespring (May–June) 2012. We isolated the ~100 mlotic reaches with block nets (4 mm bar mesh) andsampled with double-pass removal backpack electro-fishing (Smith-Root Model L-24, 180–320 V,30 Hz), where 90% of fish and 100% of species werecaught after two passes relative to three-pass removalof selected sites (data not shown). A standard effortlevel of 1200 total seconds (st. dev. = 162 s) was

adjusted as needed to account for length and variablehabitat complexity within the reach. All fish capturedwere identified, enumerated and released alive afterelectrofishing was complete. In rare instances, fishobserved but not captured (i.e., misses or largeschools of small fish) were included in counts whenthe fish could be confidently identified. Althoughelectrofishing is somewhat size-selective for largerindividuals, this sampling approach is one of the leastselective of all active fish capture methods (Barbouret al. 1999). Additionally, in our study, groups ofsmall fish were often readily visible during sampling,increasing our confidence in the effectiveness of thismethod for characterising lotic fish communities.Pond sites were typically too large, deep and com-

plex to be effectively sampled with backpacking elec-trofishing, and too remote for use of boatelectrofishing; instead, we followed previous studies(e.g., Smith & Mather 2013) in using a standardisedcombination of baited minnow traps (12 minnowtraps and 12 crayfish traps set for 12 h at 0.2–1.5 mdepth), trammel nets (3 nets, 9.1 9 1.8 m, 0.3 m

MAINSTEM SITES

TRIBUTARY SITES

Verde River

Sycam

ore C

reek

Oak Creek

Dry Beaver Creek

Wet Beaver Creek

West Clear Creek

Lotic sites

Pond sites

Stream length searched for beaver dams

Fig. 1. Map of the study area in the Verde River system and fish sampling sites, distributed among the mainstem and five tributaries. Grayshading indicates stream segments that were searched for beaver dams.

4

Gibson et al.

wall size, 51 mm mesh size, set for 3 h), seine hauls(4 mm mesh, 5 hauls sampling approximately 60 m2

total) and snorkel survey (one person snorkelled fullpond perimeter plus one transect down pond centreline; visibility was generally excellent in tributarysites but limited in mainstem sites) to achieve anaccurate representation of species occurrence and rel-ative abundance. Most of these sampling methodswould have been ineffective in our lotic sites: waterdepth was often too shallow for snorkelling or foreffective capture using trammel nets. Therefore, weselected sampling methods that maximised efficiencyand minimised error in the characterisation of fishassemblages within each habitat type.Crayfish were also sampled at all sites. In pond

sites, all crayfish captured in seine hauls, minnowtraps and crayfish traps were counted; we calculatedcrayfish catch per unit effort (CPUE) for each geartype, as well as the sum total of crayfish caughtacross all gears at each site. In lotic sites, we usedtimed D-net (0.5 mm mesh) sweeps to sample cray-fish density along with other benthic macroinverte-brates. The crayfish sample for each lotic siteconsisted of 10 point samples, covering a total sur-face area of 4 m2 and total sampling effort time of150 s, in standardised microhabitats from both theupstream and downstream ends of the samplingreach. All crayfish with carapace length ≥3 mm and≤30 mm (larger crayfish were not sampled effectivelywith this methodology) were sorted from the sampleand counted.

Statistical analysis

Sampling effort was constant among all pond sitesand among all lotic sites; thus, species abundance foreach site was calculated as the total number of indi-viduals of that species captured at the site across allgear types. Raw abundances were log-transformed toreduce the influence of highly abundant species andthen standardised to relative abundance at each site.Using relative rather than raw abundances allowed usto make comparisons between pond and lotic sam-pling sites despite different sampling methods. Toexamine differences in relative abundance of individ-ual species between pond and lotic sites, we calcu-lated the mean of pond-lotic pairwise differences inrelative abundance for each species (separately formainstem and tributary sites). Within pond sites andwithin lotic sites, we also tested for correlationsbetween abundance of crayfish and abundance ofsmallmouth bass (Micropterus dolomieu) and ofall non-native fishes combined (abundanceslog-transformed).Patterns in multivariate fish assemblage structure

were examined using principal coordinate analysis(PCoA; Legendre & Legendre 1998) of species rela-tive abundances at each site. Notably, ordinationsusing presence–absence data produced qualitativelysimilar results. PCoA is similar to principal compo-nent analysis (PCA), but generalised to allow the useof any ecologically relevant dissimilarity measure.For this analysis, we used Bray–Curtis distance

Table 1. Our study recorded four native fish species, nine non-native fish species and one non-native crayfish species in the upper Verde River mainstem andtributaries.

Family Species† Common name Code‡

Proportion of sites where species was present

All sitesn = 38

Pond sitesn = 12

Lotic sitesn = 26

NativeCatostomidae Catostomus clarkii Desert sucker CACL 0.45 0.00 0.65

Catostomus insignis Sonora sucker CAIN 0.42 0.58 0.35Cyprinidae Gila robusta Roundtail chub GIRO 0.29 0.25 0.31

Rhinichthys osculus Speckled dace RHOS 0.05 0.00 0.08Non-nativeCentrarchidae Micropterus dolomieu Smallmouth bass MIDO 0.89 0.92 0.88

Lepomis cyanellus Green sunfish LECY 0.61 0.83 0.50Ambloplites rupestris Rock bass AMRU 0.05 0.00 0.08

Cyprinidae Cyprinella lutrensis Red shiner CYLU 0.37 0.42 0.35Cyprinus carpio Common carp CYCA 0.16 0.33 0.08Pimephales promelas Fathead minnow PIPR 0.08 0.25 0.00

Ictaluridae Ameiurus natalis Yellow bullhead AMNA 0.29 0.25 0.31Poeciliidae Gambusia affinis Western mosquitofish GAAF 0.37 0.50 0.31Salmonidae Oncorhynchus mykiss Rainbow trout ONMY 0.11 0.08 0.12Non-native crayfishCambaridae Orconectes virilis Northern crayfish ORVI 1.00 1.00 1.00

†Within each family, species are listed in order of commonness.‡Combination of first two letters of genus and first two letters of species. Used to indicate species on ordination plots.

5

Beaver dams shift desert fish assemblages

(which excludes double absences and which is widelyused for community abundance data) and correctedfor negative eigenvalues. Rare species (present in ≤2sites), which can have a disproportionate effect onmultivariate results (Gauch 1982), were removedfrom this analysis. Statistical significance of thePCoA axes was determined according to the brokenstick rule (P < 0.05; Legendre & Legendre 1998).Raw species correlations with the axes were overlaidas vectors on the ordination plots to examine the con-tribution of individual species to observed patterns incommunity data.We formally assessed differences between pond

and lotic fish assemblage using permutational multi-variate analysis of variance (perMANOVA; Anderson2001). A significant difference identified by thisapproach may indicate a difference in group mean(i.e., location in PCoA ordination space) or in groupvariability (i.e., spread in ordination space) or a com-bination of the two. Therefore, we also tested forhomogeneity of multivariate dispersions betweengroups (Anderson 2006), a multivariate analogue tothe univariate Levene’s test, to locate the source ofany significant differences identified by the perMA-NOVA. This dispersion test computes an F-statisticto compare the average distance between an individ-ual sample and the group centroid defined in thePCoA space of a chosen dissimilarity measure (in ourcase, Bray–Curtis dissimilarity). Both of these multi-variate tests estimate statistical significance (P value)by permuting the appropriate least-squares residuals.All analyses were performed in R version 2.13.1(R Development Core Team 2011), using the veganpackage 2.0–2 for multivariate community analysis(Oksanen et al. 2011).

Results

Species occurrence and abundance

During our study, we recorded >12,500 individualfish from four native and nine non-native species(Table 1). Smallmouth bass was the most frequentnon-native fish, followed by green sunfish (Lepomiscyanellus), western mosquitofish (Gambusia affinis)and red shiner (Cyprinella lutrensis). Desert sucker(Catostomus clarkii) was the most frequent nativespecies, found in 45% of sites (Table 1). Speciesrichness per site varied from 1 to 8 species, with aconsiderably larger species pool in the Verde Rivermainstem (mean spp. richness � SE = 6.1 � 0.35,range 4–8) than in tributaries (2.7 � 0.20, range 1–4). Several species were found only in the mainstem:native roundtail chub and non-native western mosqui-tofish, common carp (Cyprinus carpio) and fatheadminnow (Pimephales promelas). Additionally, native

Sonora sucker (Catostomus insignis) and non-nativered shiner were found primarily in the mainstem butalso in one tributary. Desert sucker, speckled daceand rock bass (Ambloplites rupestris) were presentonly in lotic samples, while fathead minnow wasonly found in pond sites. Sonora sucker was the onlynative species frequently found in ponds. Speckleddace and rock bass were found only in two lotic sites;these rare species were excluded from multivariateanalyses.Almost without exception, non-native fishes

outnumbered natives in samples (mean = 85% non-native; range = 14–100%), and the non-native frac-tion of the assemblage was consistently greater inpond than in lotic sites, in both mainstem and tribu-tary locations (Fig. 2). This difference primarilyreflect large numbers of small non-native fish in mostpond samples. The greater dominance of non-nativefish in ponds was also apparent across individual spe-cies, with the non-native species displaying generallyhigher relative abundances in ponds, whereas nativespecies were relatively more abundant in lotic sites(Fig. 3). The difference in relative abundancebetween habitat types was especially strong for desertsucker and green sunfish in tributaries. Smallmouthbass and, to a lesser extent, yellow bullhead(Ameiurus natalis) were the only non-native specieswhose mean relative abundance was greater in loticsites than in pond sites at both mainstem and tribu-tary locations.Northern crayfish (Orconectes virilis) were present

at all sites, although density varied widely amongboth lotic sites (range = 0.3–28.3 individuals per m2)and pond sites (range = 0–37 individuals per minnowtrap) (Table 2). Within pond sites, there was a mar-ginally significant positive relationship betweenabundance of crayfish and abundance of smallmouthbass (r = 0.55, P = 0.07), but not between crayfishand all non-native fish combined (r = 0.14,P = 0.66). By contrast, in lotic sites, there was noevidence of a relationship between abundance ofcrayfish and either smallmouth bass (r = �0.01,P = 0.97) or all non-native fish combined(r = �0.24, P = 0.23).

Fish assemblages

Multivariate ordination revealed strong differences infish assemblage structure between mainstem and trib-utary sites (Fig. 4a), reflecting the different speciespools in the two locations. The first three principalcomponent axes were statistically significant, and col-lectively explained 52% the total variation in fishassemblage. A two-way perMANOVA analysis indi-cated that fish assemblage differed significantly byboth location (mainstem vs. tributary) and habitat

6

Gibson et al.

type (pond vs. lotic), although location contributedthe greater component of variation (Table 3). Therewas no evidence of an interaction effect, nor of anydifferences in within-group variability. To focus ondifferences between habitat types, subsequent analy-ses were conducted separately for mainstem and fortributary sites.

An ordination of mainstem sites showed modestdifferentiation between pond and lotic habitat types(Fig. 4b); perMANOVA confirmed a significant dif-ference in fish assemblage by habitat type, althoughthere was no difference in variability (Table 3). Inthe ordination, the first two principal coordinate axesexplained 57% of the total variation; only the first

Fig. 3. Mean difference in relative abundance between lotic and pond sites (averaged for all pond site-lotic site pairwise comparisons,within mainstem sites and within tributary sites), for each species recorded in this study (with the exception of speckled dace and rockbass). Positive values (i.e., bars above the midline) indicate higher relative abundance in lotic sites than in pond sites, while negative valuesindicate higher relative abundance in pond sites. Error bars show standard error of the mean.

Fig. 2. Boxplots showing the distributions of the proportion native fish (lighter shades) and its converse, the proportion non-native fish(darker shades), across lotic sites (left panel) and across pond sites (right panel). Distributions are calculated separately for mainstem sites(solid fill) and for tributary sites (hatched fill).

7

Beaver dams shift desert fish assemblages

two axes were statistically significant. Axis 1appeared to describe a separation of fish assemblagesalong a gradient from pond sites on the left-handside of the plot to lotic sites on the right. Westernmosquitofish (GAAF) and green sunfish (LECY)were strongly correlated with each other and withAxis 1, indicating that pond sites on the left side ofthe plot were associated with abundant mosquitofishand sunfish, while lotic sites to the right were moreassociated with relatively abundant desert sucker(CACL). By contrast, other species, including com-mon smallmouth bass (MIDO) showed a stronger cor-relation with Axis 2, and no strong tendency towardseither pond or lotic sites. Axis 2 described a generalgradient from sites dominated by small-bodied redshiner (CYLU) and mosquitofish (GAAF) on the bot-tom to sites dominated by larger-bodied smallmouthbass (MIDO) and Sonora sucker (CAIN) on the top, agradient present in both pond and lotic sites.Within tributary sites, fish assemblage again dif-

fered significantly between pond and lotic habitattypes (Table 3). An ordination of tributary sites(Fig. 4c) showed, again, some clustering of pondsites, although pond and lotic sites appear less differ-entiated than in the mainstem ordination. Addition-ally, patterns in habitat associations of individualspecies were similar to the patterns seen for mainstemsites. The first two principal coordinates explained62% of the total variation; only the first two axeswere statistically significant. Pond sites were clus-tered together in the lower-right quadrant of the plot,

but within ordination space also occupied by loticsites. Although lotic sites occupied a greater area inthis ordination space than did pond sites, there wasno evidence for a difference in group variability(Table 3). Green sunfish (LECY) again showed astrong correlation with Axis 1, and again the ordina-tion plot suggested a gradient between pond siteswith abundant sunfish to the lower-right and loticsites with relatively abundant desert sucker (CACL)to the upper left. Finally, the smallmouth bass(MIDO) vector was again perpendicular to the sun-fish-desert sucker gradient, and again lacked a strongtendency towards either pond or lotic sites. Axis 2described a general gradient from sites almostentirely dominated by smallmouth bass at the bottom,moving to sites with more desert sucker (for loticsites) and relatively fewer bass at the top. This tribu-tary ordination plot was strongly influenced by thetwo pond and three lotic sites (mostly clustered onthe far right-hand side of the plot) from one tributarystream (Dry Beaver Creek), which was smaller,ephemeral and supported a different fish species poolthan the other sampled tributaries. Red shiner(CYLU) were found only in this tributary, and thereonly within pond sites, resulting in the largely hori-zontal vector for this species in the ordination plot.

Discussion

We assessed the influence of beaver pond habitats instructuring the composition of mixed native and non-

Table 2. Density and catch per unit effort (CPUE) of non-native northern crayfish in lotic and pond habitats.

Habitat type Abundance metric

Mainstem sites Tributary sitesEstimated CL† (mm)

n Mean Range n Mean Range Range

Lotic sites, density Individuals per m2 10 sites 9.0 0.3–28.3 16 sites 7.3 1.5–22.3 3–30 (mean = 8.4)Pond sites, CPUE Crayfish traps (individuals per trap) 72 traps 3.0 0–9 71 traps 1.8 0–10 20–60

Minnow traps (individuals per trap) 72 traps 5.0 0–37 71 traps 1.1 0–12 20–50Seine hauls (individuals per m2 seined) 33 hauls 4.1 0–21.0 32 hauls 0.5 0–4.3 10–70

†Size range of crayfish caught in each gear type. Sizes indicate carapace length [CL] in mm. Crayfish caught in lotic sites (top row) were measured; crayfishcaught in pond sites (bottom three rows) were not measured, and these size ranges are estimates only, to indicate the approximate size of individuals sampledby each gear type.

Table 3. Results of perMANOVA and multivariate dispersion tests examining effects of location (mainstem vs. tributary) and habitat type (pond vs. lotic) onfish assemblages within all sites and within mainstem sites or within tributary sites only.

Sites Factor n

perMANOVAHomogeneity ofdispersions

pseudo F P pseudo F P

All sites Habitat type 9 Location 38Location 18.91 <0.001 1.31 0.227Habitat type 5.65 0.003 0.62 0.450Location 9 Habitat type 1.60 0.158 NA NAMainstem sites Habitat type 16 3.61 0.009 0.38 0.550Tributary sites Habitat type 22 3.85 0.014 0.37 0.551

8

Gibson et al.

Pond mainstem

Pond tributary

Lotic mainstem

Lotic tributary

−1.0

−0.5

0.0

0.5

1.0

Axis 1 (24%)

Axis

2(1

8%

)

−1.0

−0.5

0.0

0.5

1.0

Axis 1 (35%)

Axis

2(2

2%

)

−1.0

−0.5

0.0

0.5

1.0

−1.0 −0.5 0.0 0.5 1.0

Axis 1 (40%)

Axis

2(2

2%

)

MIDO CAIN

CACL

AMNA

CYLU

LECY

GAAF

LECY

CYLU

CACL

MIDO

(a)

(b)

(c)

Fig. 4. PCoA ordination plots of fish assemblage data for (a) all sites; (b) mainstem sites only; and (c) tributary sites only. Hulls are drawnaround tributary versus mainstem sites in panel (a) and around pond versus lotic sites in panels (b) and (c). Only statistically significant(P < 0.05) vectors are displayed.

9

Beaver dams shift desert fish assemblages

native fish assemblages in a free-flowing drylandriver. To our knowledge, this is the first study toaddress the effect of beaver ponds on non-nativefishes at the community level. Our analyses showedthat in both the upper Verde River mainstem and inits tributaries, there were significant differences infish assemblage composition between lentic beaverpond and lotic stream habitat. Non-native speciesdominated the fish assemblage to a greater extentwithin ponds than within lotic sites: small non-nativefish (both small-bodied fishes and juveniles of largerspecies) were often highly abundant in pond samples,while native species were rarely documented withinponds. These results are generally consistent with thehypothesis that beaver ponds could favour fishassemblages dominated by non-native species.Our results are also consistent with studies of fish

assemblage in relation to artificial dams. Impound-ments behind small, low head dams are the closestanalogue to beaver pond habitat: Beatty et al. (2009)found abundant non-native fish (primarily fatheadminnow and white sucker, Catostomus commersonii)within and downstream of an artificial wetland cre-ated by a small dam on an upper Colorado RiverBasin stream (Wyoming), while native fishes wererestricted to upstream of the wetland and to tributar-ies without impoundments. Dominance of non-nativefishes within large reservoirs of the LCRB is wellestablished (Mueller & Marsh 2002), and reservoirscan promote proliferation of non-native speciesdownstream, even without major changes to down-stream flow or thermal regimes (Martinez et al.1994). However, beaver dams are typically smaller,more permeable and much less permanent than artifi-cial analogues, which may affect the direction ormechanism of influences on fish.

Fish assemblage responses

We found that responses to beaver pond habitat var-ied by species. Here we focus on the native fishesand the four most common non-native species in ourstudy: smallmouth bass, green sunfish, western mos-quitofish and northern crayfish.Smallmouth bass were the most common fish in

our study (present in 87% of sites) and often the mostabundant large-bodied fish, in both lotic and pondhabitat types. In lotic samples, especially, bass oftendominated the total fish assemblage. The high num-bers of bass found in both habitat types in our studyconfirm the habitat flexibility of this species. From aconservation perspective, smallmouth bass are of par-ticular interest because this widespread species isbelieved to pose a substantial threat to native fishesin the LCRB. A bioenergetics model for the upperVerde River identified smallmouth bass as the great-

est source of potential piscivory and suggested thatnon-native control efforts should target this species(Bonar et al. 2004). Similarly, a bioenergetics modelof non-native fishes in the Yampa River (upper Colo-rado River Basin) indicated that the total potential pi-scivory of abundant smallmouth bass exceeded thatof two larger but less numerous non-native predators(northern pike, Esox lucius, and channel catfish, Ict-alurus punctatus) (Johnson et al. 2008).Green sunfish were frequently present in both lotic

and pond habitats, but in lotic sites they were usuallypresent only at low numbers, whereas in ponds theywere often highly abundant. This pattern was true forboth mainstem and tributary sites. The association ofgreen sunfish with pond habitats is consistent withother studies which report that species of the genusLepomis were common in beaver pond fish assem-blages (Snodgrass & Meffe 1999; Pollock et al.2003), and with the species’ known habitat prefer-ence for slow current velocity (Dudley & Matter2000; Olden et al. 2006). Presence of green sunfishhas been implicated in steep declines of native fishpopulations in the LCRB (e.g., Clarkson et al. 2010).Although gut content analysis of green sunfish fromthe upper Verde River found a very low incidence ofpiscivory (mean 0.61% of stomach content by vol-ume, n = 754; Bonar et al. 2004), laboratory andfield evidence shows that this species can be aneffective predator of larval and juvenile native fishesin the LCRB (Dudley & Matter 2000; Carpenter &Mueller 2008). Increased d15N isotope signaturesfrom green sunfish relative to native fishes in anotherArizona river system, the Gila River, also demon-strate a strong propensity towards piscivory (Pilgeret al. 2010). Additionally, highly aggressive behav-iour by green sunfish indicates a probable advantageover native fishes in competition for space or food(Karp & Tyus 1990). These lines of evidence suggestthat there is potential for green sunfish to have sub-stantial impacts on native fishes in the Verde Riversystem, especially given the high densities of greensunfish that we observed in beaver ponds.Western mosquitofish showed a strong association

with beaver pond habitat, although high mosquitofishabundances were also observed in some mainstemlotic sites. Consistent with our results, preferred mos-quitofish habitat features include shallow water, slowcurrent and dense vegetation (Pyke 2008), all typicalfeatures of beaver ponds (Rosell et al. 2005). Mos-quitofish have been widely introduced worldwide(Pyke 2008), and, in the LCRB, spread of mosquito-fish is closely associated with declines of the ecologi-cally similar native Gila topminnow (Poeciliopsisoccidentalis) (Meffe 1985). Laboratory and fieldstudies have demonstrated that agonistic interactionswith aggressive mosquitofish can lead to injury,

10

Gibson et al.

changes in behaviour and reduced growth for nativefishes (particularly small-bodied species), and mos-quitofish also prey directly on eggs and larvae ofnative fish (Meffe 1985; Mills et al. 2004; Ayalaet al. 2007). In isolated backwater habitats otherwisefree of non-native fish, presence of mosquitofish isbelieved to be responsible for failed recruitment bystocked native razorback suckers (Xyrauchen tex-anus) (Ley et al. 2012); we suggest that abundantmosquitofish may similarly impede attempts torestore populations of spikedace in the Verde River.In addition to non-native fishes, we observed high

densities of non-native northern crayfish within bothlotic and beaver pond habitats of the Verde Riversystem. This study is one of the first to report abun-dance metrics for crayfish in Arizona. In pools of aGila River system tributary stream (Arizona), Carpen-ter (2005) reported crayfish densities between 3 and11 large individuals (>25 mm carapace length [CL])per m2, comparable to our mean densities of 9.3 and7.4 small individuals (mean CL = 8.4 mm) per m2 inmainstem and tributary sites, respectively. Martinez(2012) reported somewhat higher densities (10.9 indi-viduals per m2 in 2005; mean CL = 16.1 mm) in theYampa River (upper Colorado River Basin), wherethe total estimated river-wide biomass of crayfishexceeded that of all fish and other invertebrates com-bined. Invasive crayfish can have strong impacts onmultiple levels of aquatic food webs (Twardochlebet al. 2013), and laboratory and field evidence indi-cate that crayfish can compete with and prey onnative LCRB fishes directly (Carpenter 2005). Manynon-native fishes consume crayfish (Johnson et al.2008), while the native fishes sampled in this studyare more exclusively invertivores/herbivores (Oldenet al. 2006; Pilger et al. 2010), and therefore morelikely to compete with crayfish than to prey on them(Carpenter 2005; Arena et al. 2012). The greatestimpact of crayfish on native fishes may be indirect,via apparent competition: abundant crayfish may pro-vide an alternative prey source when small fish havebeen depleted, thus stabilising large populations ofnon-native predators. Our finding of a positive rela-tionship between crayfish and smallmouth bass abun-dances in ponds is consistent with this hypothesis. Inthe Yampa River, Martinez (2012) documentedsimultaneous large increases in northern crayfish andsmallmouth bass populations (along with steepdeclines in small-bodied native fish). By creating rel-atively stable hydrology and deep pools, beaverponds may provide favourable habitat for crayfish(e.g., Light 2003); the influence of beaver ponds oncrayfish populations in the LCRB requires additionalinvestigation.Native fishes in general showed a tendency

towards higher abundance in lotic habitats: all four of

the native species recorded in this study were presentin lotic sites, but only Sonora sucker and roundtailchub were found in ponds. The two native suckers inour species pool differ in modes of feeding and useof habitat. Desert suckers are herbivorous and inhabitpredominantly swift-flowing streams with hard-bot-tom substrate, whereas Sonora suckers feed on inver-tebrates and detritus and are abundant in deep poolswith restricted flow and fine substrate (Minckley &Marsh 2009). These differences were reflected infindings of our study. Desert suckers showed a strongassociation with lotic habitat: they were frequentlypresent and often relatively abundant in lotic sites,but never found within ponds. Sonora suckers, bycontrast, were present and occasionally abundant inboth habitat types. Native roundtail chub, much likeSonora suckers, are omnivorous feeders associatedwith deep pool habitat. However, we rarely foundthis species in ponds. It is possible that roundtailchub presence or abundance was underestimated inponds due to size selectivity of sampling methodol-ogy: size frequency data from pond samples (data notshown) suggest that mid-sized fish (approximately150–300 mm total length) may have been consis-tently under-sampled in ponds. We believe that thisis the most important influence of a gear effect in ourresults. Species most likely to have been underesti-mated in this way are roundtail chub, desert sucker,and, to a lesser extent, smallmouth bass. Alterna-tively, the rarity of roundtail chub in beaver pondscould reflect unfavourable habitat conditions for thesefish, whether due to unique abiotic properties of bea-ver ponds or to the abundant non-native fish.The three small-bodied fish present in the upper

Verde River within recent decades (longfin dace,speckled dace and spikedace) were not recorded inour study, with the exception of two tributary siteswhere speckled dace were present. These three spe-cies are generally associated with riffle habitat(Minckley & Marsh 2009); therefore, it is unlikelythat adults would be found in beaver pond habitat,although they might occupy fast tailwaters immedi-ately downstream of dams (Smith & Mather 2013).Extensive conversion of lotic habitat to lentic beaverponds, as has been documented on the regulated BillWilliams River, Arizona (Shafroth et al. 2010), couldlimit the potential for restoration of these native rifflespecialists. This situation occurred on Bonita Creek,AZ, where extensive beaver dam activity has resultedin the cessation of stocking efforts for spikedace andloach minnow (Rhinichthys cobitis), due in part to alack of lotic habitat and an increase in abundance ofgreen sunfish (H. Blasius, pers. comm.).However, the greatest importance of beaver ponds

for native and non-native fish communities in theVerde River system may be their function as spawn-

11

Beaver dams shift desert fish assemblages

ing or rearing habitat for juvenile fish. Previous stud-ies have found ontogenetic shifts in fish use of beaverpond habitat, with some species moving into pondsto spawn (Schlosser 1998; Snodgrass & Meffe 1999).Schlosser (1995, 1998) further suggests that beaverponds may function as reproductive ‘sources’ forsome species, while adjacent stream reaches are‘sinks’ with little successful recruitment. Beaverponds often include extensive shallow, vegetatedareas with slow current, many of the same featuresthat characterise typical rearing habitat for juvenileLCRB native fishes (Childs et al. 1998). Backwatersand other off-channel rearing habitats have been thetarget of restoration activities (Minckley et al. 2003),and given the importance of backwaters for success-ful recruitment of native fish, the possibility that bea-ver ponds could provide equivalent rearing habitatconditions is worthy of investigation. However, theabundant non-native fish that we documented withinmargins and backwaters of beaver ponds may makethese areas unsuitable for juvenile native fish (Minck-ley et al. 2003; Carpenter & Mueller 2008). Sizeselectivity of different sampling gears prevents formalanalysis of differences in habitat use by life stage inour data. However, we did record young-of-the-yearindividuals of several native and non-native species(native Sonora sucker and non-native smallmouthbass, green sunfish, common carp and yellow bull-head) within ponds, suggesting that these speciesmay sometimes utilise beaver ponds for rearinghabitat.

Variation over time

This study provides a snapshot view of fish assem-blage at a single point in time, focusing on the dis-tinction between pond and lotic habitat types. Long-term studies of fish assemblage in the Verde Riverand similar systems document substantial interannualvariation in absolute and relative abundance of nativefishes, especially in response to stream discharge(e.g., Rinne 2005; Propst et al. 2008). Similarly, theresponse of fish assemblage to beaver pond habitatsis likely to vary with time and discharge; therefore,caution should be used in extrapolating results of thepresent study to different environmental conditions.Sampling for this study occurred during May andJune, typically the driest months of the year (Blaschet al. 2006), and during a drought year; the upperVerde River mainstem had not experienced a signifi-cant high-flow event since 2010 (USGS 2013). Lowflow conditions are of particular importance from theperspective of native fish conservation. Numerousstudies have found strong positive correlationsbetween discharge and LCRB native fish abundance(Rinne & Miller 2006; Stefferud et al. 2011). Native

fishes as a group tend to increase in absolute and rel-ative abundance following large peak flows or yearswith high discharge; conversely, during dry years,non-native species are more likely to become estab-lished and to dominate assemblages, and droughtconditions are associated with extirpations of nativespecies (Propst et al. 2008; Gido et al. 2013).Drought and low discharge likely represent criticalperiods for persistence of native fishes and thereforethe most important time to understand the influenceof beaver dams and other habitat features on nativeand non-native assemblage structure.

Contrasting mainstem and tributary habitats

Our finding of a strong distinction between fishassemblages of the upper Verde River mainstem andits tributaries is consistent with other observed differ-ences between the two locations, including differ-ences in species pool, habitat features anddistribution of beaver dams. We found that beaverdams were generally larger and more abundant in theupper mainstem than in the tributaries, and mainstemponds appeared older and more complex. This mayreflect the more stable hydrology in the upper main-stem relative to the tributaries: the number of beaverdams in the upper mainstem tends to increase steadilyover several years during periods of moderate dis-charge, until occasional large floods destroy most orall dams (D. Campbell, pers. comm.). In Verde Rivertributaries, by contrast, dams are more likely to washout annually in snowmelt or monsoon floods(K. Schonek, pers. comm.), thus preventing enlarge-ment and succession of the beaver pond habitat overmultiple years. It seems likely that longevity of bea-ver ponds may affect their ecological function, andseveral studies have demonstrated that fish assem-blage in beaver pond habitats varies with pond age(Snodgrass & Meffe 1998; Schlosser & Kallemeyn2000). However, despite differences in hydrologyand pond stability, we observed some consistent pat-terns in individual species’ response to pond habitatin both the mainstem and tributaries.

Net effects of beaver ponds on fish communities

In this study, we documented patterns of presenceand abundance for native and non-native fish withinbeaver ponds as compared to unimpounded loticstream reaches. Our results point to further questionsabout the consequences of these patterns (also seeGibson & Olden 2014): what will be the net effect ofbeaver dams on fish populations in the Verde Riversystem and similar desert rivers? Beaver dams andponds are complex landscape features that alternumerous different aspects of aquatic ecosystems at

12

Gibson et al.

Table4.

Hypothesisedconsequences

ofvarioustypesof

ecosystem

change

resulting

from

constructionof

beaver

damsfornativeandnon-nativefishesin

theCRB,as

wellas

possibleoutcom

esof

each

change

fornative

fishin

amixed

native–non-nativeassemblage.

Arrow

sindicate

whether

theeffect

onfishisexpected

tobe

positive(↑)or

negative(↓).Beaverdamsandpondsarehighlyvariable,andnotalleffectswillapplyto

allbeaver

ponds.

Category

Effect

ofbeaver

dam

†Consequencesfornativefishes‡

Consequencesfornon-nativefishes§

Potentialoutcom

es¶

Directhabitat

effects

A.Pooledwater,reduced

currentvelocity.

↑Deeppoolsprovidehabitatforlarger

orpool-

dwellingfish,

potentially

including‘big

river’

nativefish.

↑Shallowlenticwater

inpond

margins

and

backwatersprovides

habitatforlentic/

marsh

habitatspecialists.

↓Decreased

habitatavailabilityto

fluvial

specialists

dueto

conversion

oflotic

tolentichabitatin

beaver

ponds.

↑Deeppoolsprovidehabitatforlarger

orpool-

dwellingfish,

includingcarp,catfishesand

somesunfishes.

↑Shallowlentichabitatprovides

habitatfor

lentic/m

arsh

habitatspecialists.

↑Relativestability

ofbeaver

pond

habitat

resembles

typicalhabitatin

thenativerange

ofmanynon-nativespecies.

↑Reduced

intensity

ofbioticinteractions

dueto

habitatsegregation;

or↓Increasedcompetitionandpredationdueto

high

densities

ofnon-nativefishesin

ponds.

B.Increasedhabitatcomplexity,

includingboth

channel

morphology(side

channels,backwaters)

andhabitatstructure(large

wood,

macrophytes).

↑Spawning/rearing

habitatfornumerous

speciesin

shallow,vegetated,

backwater-

type

habitats.

↑Habitatstructureprovides

coverandrefuge

from

predation.

↑Spawning/rearing

habitatfornumerous

speciesin

shallow,vegetated,

backwater-

type

habitats.

↑Habitatstructureprovides

coverfor

predators,increasespredationsuccess;

or↓Habitatstructureprovides

coverforprey

andreducespredationsuccess.

↑Reduced

predationdueto

availabilityof

cover

andrefuge

habitats.

↑Com

plexity

prom

otes

habitatsegregationand

reducesintensity

ofbioticinteractions;or

↓Abundantsm

allnon-nativefishpresentin

potentialnativefishrearinghabitatincrease

ratesof

competitionandpredationon

larvalnativefish.

C.Retentionof

sediment,

reducedturbidity.

↑Reducesedimentationdownstream

ofponds.

↓Fine

sedimentwithin

pondsunfavourablefor

speciesthat

prefer

coarsersubstrate.

↓Loss

ofvisualcoverdueto

decreased

turbidity.

↑Reducesedimentationdownstream

ofponds.

↓Fine

sedimentwithin

pondsunfavourablefor

speciesthat

prefer

coarsersubstrate.

↑Increase

successof

visualpredatorsdueto

decreasedturbidity.

↑Im

proved

downstream

habitatquality

for

somespeciesdueto

alteredsediment

flows.

↓Increasedpredationdueto

greatervisibility.

D.Increasedfluvialhabitat(fast

water,coarse

substrate)

immediatelydownstream

ofbeaver

dams.

↑Provide

habitatforsm

allfluvialspecialists.

↑Provide

habitatforsm

allfluvialspecialists.

↑↓Changein

quality

orquantityof

habitat

availableto

fluvialspecialists.

E.Alteredwater

quality,

includingtemperature

regime

anddissolvedoxygen

concentration.

↑Reducethermalstress

dueto

temperature-

bufferingeffect

ofponds;or

↓Increase

thermalstress

dueto

higher

temperaturesin

ponds.

↑Reducethermalstress

dueto

temperature-

bufferingeffect

ofponds;or

↓Increase

thermalstress

dueto

higher

temperaturesin

ponds.

↓Increasedcompetitiveadvantagefornon-

nativefishesdueto

thermalstress.

Indirect

effects

F.Increasedrate

ofprimary

productivity

andstanding

stockof

organicmatter.

↑Increase

food

availabilityforherbivores

and

detritivores.

↑Increase

food

availabilityforherbivores

and

detritivores.

↑Reduced

competitiondueto

greater

availabilityof

resources;

or↓Increasedcompetitionandpredationdueto

higher

densities

ofnon-nativefish.

G.Increasedbiom

assand

alteredcommunity

structure

ofbenthicinvertebrates.

↑Increase

food

availabilityforinvertivores.

↓Reducerelativeabundanceof

lotic

invertebrate

taxa

preferredby

some

species.

↑Increase

food

availabilityforcrayfish-

consum

ingspecies.

↓Reducerelativeabundanceof

lotic

invertebrate

taxa

preferredby

some

species.

↑Reduced

competitiondueto

greater

availabilityof

resources;

or↓Increasedcompetitionandpredationdueto

higher

densities

ofnon-nativefish.

13

Beaver dams shift desert fish assemblages

Table

4(con

tinued)

Category

Effect

ofbeaver

dam

†Consequencesfornativefishes‡

Consequencesfornon-nativefishes§

Potentialoutcom

es¶

Drainage-level

effects

H.Flow

stabilisation:

dampened

peak

flows.

↑Provide

slackwater

refuge

during

peak

flows,

particularlyforvulnerablelarvalor

juvenile

fish.

↓Reducerecruitm

entsuccessdependenton

peak

flowevents.

↓Reducepower

ofpeak

flowsto

move

sedimentandrejuvenate

spaw

ning

and

rearinghabitats.

↑Provide

slackwater

refuge

during

peak

flows:

may

beespeciallyimportantfornon-native

speciesthat

lack

adaptations

forsurviving

flood

conditions

↓Reductionin

theability

ofnativefishesto

benefit

from

peak

flowevents.

↓Increasedsurvivalof

non-nativefishthrough

peak

flowevents

and/or

increasedspeedof

recolonisationfollowingdisturbance.

I.Flow

stabilisation:

maintenance

ofsurface

water

within

beaver

pond

and/or

downstream

ofdam.

↑Provide

refuge

habitatduring

droughtand

lowflows.

↑Provide

refuge

habitatduring

droughtand

lowflows:

may

beparticularlyimportantfor

non-nativespeciesthat

lack

adaptations

for

survivingdroughtconditions.

↑Increasedability

ofnativefishesto

persist

throughcombinedstressorsof

droughtand

non-nativefish;

or↓Increasednegativebioticinteractions

dueto

concentrationof

nativeandnon-native

specieswithin

smallrefuge

habitats.

↓Increasedsurvivalof

non-nativefishthrough

droughtevents

and/or

increasedspeedof

recolonisationfollowingdisturbance.

J.Barriersto

fishmovem

ent.

↓Reducehabitatconnectivity,fishdispersal

betweenpopulations.

↓Reducehabitatconnectivity,fishdispersal

betweenpopulations.

↓Slowrate

ofspread

upnewdrainagesand

tributarysystem

s.

↑Potentialfordamsto

create

habitatpatches

that

aretemporarilyisolated

from

non-native

fish.

↓Dam

slim

itability

tomovebetweenhabitats

inresponse

todroughtor

othervarying

environm

entalconditions.

K.Uniquelentichabitatpatches

within

lotic

habitat

corridors.

↑Provide

reproductivesource

habitatforsome

species.

↑Promotepersistenceof

lentic/m

arsh

habitat

specialists

within

adrainage.

↑Potentially

increase

theupstream

extent

ofsuitablehabitatfor‘big

river’fishesin

smallerriveror

tributarysystem

s.

↑Provide

reproductivesource

habitatforsome

species.

↑Promotepersistenceof

lentic/m

arsh

habitat

specialists

within

adrainage.

↑Potentially

increase

theupstream

extent

ofsuitablehabitatforlargepiscivores

such

ascatfishesin

smallerriveror

tributary

system

s.

↑Higherspeciesrichness

ofnativefisheswithin

adrainage.

↓Higherspeciesrichness

ofnon-nativefishes

within

adrainage.

↓Spilloverfrom

pond

habitatincreases

abundanceof

non-nativefishthroughout

adrainage.

L.Reduced

channelisationand

increasedabundanceof

marshyandcienega

habitats.

↑Diversity

ofaquatic

habitats

prom

otes

regionalspeciesrichness.

↑Diversity

ofaquatic

habitats

prom

otes

regionalspeciesrichness.

↑↓Differenthabitatsupports

adifferent

assemblageof

nativeandnon-nativefishes.

†Well-establishedeffectsof

beaver

damsarein

plaintext,whilehypothesised

oruncertaineffectsareindicatedin

italics.

‡Responseof

nativefish,

independentof

orin

theabsenceof

non-nativefish.

§Responseof

non-nativefish,

with

orwithoutnativefishpresent.

¶Hypothesisednetoutcom

esfornativefishpopulations,whennon-nativefisharepresent,as

aresultof

thebeaver

dam

effect.In

somecases,

mutually

exclusivealternatives

arepresented.

Sources

inform

ingdevelopm

entof

this

tableinclude:

Schlosser

(1995,

1998);Snodgrass

&Meffe

(1998,

1999);Pollock

etal.(2003);Rosellet

al.(2005);Propstet

al.(2008);Minckley&

Marsh

(2009);Stefferud

etal.

(2009);Kem

pet

al.(2012);Smith

&Mather(2013);Gibson&Olden

(2014).

14

Gibson et al.

multiple scales (Pollock et al. 2003; Rosell et al.2005). Some of these beaver dam effects may be ben-eficial to fishes while others are detrimental; someeffects may influence native fishes differently thanthey do non-native fishes; and the overall effect ofdams on native fishes may vary depending onwhether non-natives are present. This complexitymakes it difficult to distinguish the mechanisms bywhich beaver dams produce observed responses infish communities. In Table 4, we address this com-plexity by systematically formulating separatehypotheses about how each type of beaver dam effectis likely to influence native and non-native fish popu-lations in the LCRB. Our intention is that this chartwill help to guide and organise thinking about howand why beaver dams influence fish communities;additionally, these hypotheses may suggest profitabledirections for future research. In the following discus-sion, we expand on some of the particularly impor-tant or interesting hypotheses.At the scale of individual beaver ponds (Table 4,

‘Direct habitat effects’), a better understanding isneeded of how fish use of pond habitat varies withseason, environmental conditions and life stage; iden-tifying which species may be employing beaverponds for spawning or rearing habitat would be par-ticularly valuable. Fish use of beaver pond habitatlikely also varies with physical habitat features ofponds, such as pond size or location within the riversystem (e.g., Snodgrass & Meffe 1998). However,the net effect of utilising beaver pond habitat willdepend on the degree of biotic interactions occurringin these habitats, and whether beaver ponds tend toincrease or decrease negative interactions betweennative and non-native fishes. Beaver ponds generallyincrease habitat complexity relative to unimpoundedstream, in terms of both channel morphology (e.g.,secondary channels, overbank flooding; Kemp et al.2012; Smith & Mather 2013) and habitat structure(e.g., aquatic vegetation, large wood; Pollock et al.2003). Habitat complexity may provide refuge frompredation and increase spatial segregation betweennative and non-native fishes, thus reducing negativeinteractions (Crowder & Cooper 1982; Meffe 1985;Billman et al. 2012). However, if beaver pond habi-tats also support significantly higher densities of non-native fish than do unimpounded stream reaches, thenthis could override any advantages of habitat struc-ture for native fish (Table 4, point B). In particular,use of beaver pond margins by native fishes as rear-ing habitat could increase predation on vulnerablenative fish larvae when, as in our results, these areasalso support abundant small non-native fishes.Some habitat features of beaver ponds are also

found in natural stream pools: fish preferring deepwater and reduced current velocity will likely make

use of both habitat types (Table 4, point A). How-ever, the complexity and habitat structure of beaverponds are not associated with natural pools, andflooding behind beaver dams increases channel widthand often includes abundant shallow grassy area(Table 4, point B; Rosell et al. 2005). Beaver damscan greatly increase retention of sediment and organicmatter (Table 4, point C). This retentiveness, togetherwith the more open canopy and addition of woodydebris due to beaver foraging, affects nutrient cyclingand other ecosystem processes within ponds, whichin turn influences the biotic communities (Table 4,points F-G; Naiman et al. 1986; Rosell et al. 2005).The beaver dam itself provides a potential barrier tomovement of fish and other organisms, creating amore closed system than a natural pool (Table 4,point J). Finally, although an individual beaver pondmay be no bigger than an average stream pool, oftenmultiple beaver dams are constructed within a smallarea, forming a very large complex of continuousbeaver pond habitat (Rosell et al. 2005). Futureresearch should seek to clarify which, if any, of theseunique habitat features are most influential for fishspecies of interest.The influence of beaver dams on fish assemblages

can also extend beyond the pond itself to the rest ofthe stream network (Table 4, ‘Drainage-leveleffects’). Fish from beaver pond assemblages mayspill over into neighbouring stream reaches (Schlosser1998; Snodgrass & Meffe 1999). Beaver ponds canprovide source habitat for some species, as discussedearlier. In general, beaver ponds promote fish speciesrichness at the drainage level by supporting speciesthat otherwise might not persist in unmodified habitat(e.g., Hanson & Campbell 1963; Snodgrass & Meffe1998; Smith & Mather 2013), but this effect may beundesirable if it is the richness of non-native speciesthat is increased (Table 4, point K). Beatty et al.(2009), who documented abundant non-native fishwithin and downstream of an artificial impoundmenton a Wyoming stream, suggest that this impoundmentprovided source habitat for the non-native fish, notingthat tributary streams lacking any impoundments wereinstead dominated by native fishes. The abundantgreen sunfish and western mosquitofish recordedwithin ponds in our study suggest that beaver pondscould similarly provide source habitat for these smalllentic fishes in the Verde River system. However, wefound no evidence for a correlation between distancefrom a pond and abundance of these or any other spe-cies (unpublished results); additionally, thriving popu-lations of non-native species in similar streamswithout beaver activity (e.g., Fossil Creek, Arizona;M. O’Neill, pers. obs.; Marks et al. 2010) indicate thatbeaver pond habitats are not essential to their life his-tory. Nonetheless, beaver ponds could promote higher

15

Beaver dams shift desert fish assemblages

abundances of these fish or increase their resistance orresilience to disturbance. Finally, the degree to whichbeaver ponds influence fish populations will alsodepend on the abundance and distribution of pondsand other pool habitat within a river system. Pondhabitat is likely to be more influential in systemswhere beaver ponds are abundant or where ponds pro-vide a unique habitat type (Table 4, point K), espe-cially when other pool habitat is scarce.Within desert river systems like the Verde River,

the influence of beaver ponds on fish assemblageresponse to large-scale disturbance by drought andfloods is of particular interest. Previous studies havesuggested that beaver ponds can provide refuge habi-tat for fish during drought and seasonal low flows(Table 4, point I; Hanson & Campbell 1963; Magou-lick & Kobza 2003; White & Rahel 2008). This func-tion may be particularly important in intermittenttributary streams, especially with projected decreasesin perennial surface flow due to water withdrawalsand climate change (Marshall et al. 2010). In ephem-eral Dry Beaver Creek in our study, we observedlarge native Sonora suckers within deep pools thathad been expanded by beaver dams. However, con-centrating fish within small refuge habitats typicallyincreases the intensity of biotic interactions (Magou-lick & Kobza 2003): enforced habitat overlap betweennon-native predators and vulnerable native fisheswithin ponds could lead to extensive predation andother negative interactions, such that native fisheswould derive no net benefit from the refuge habitat.Beaver dams may also provide fish some protectionfrom floods or peak flows (Kemp et al. 2012), poten-tially speeding recolonisation of a drainage by non-native species following the flood (Table 4, point H;Pool & Olden 2014). Future research should addressthe ability of beaver ponds to provide refuge fromthese disturbances, and the relative impact of that ref-uge on native and non-native fishes.Ultimately, the consequences of beaver dam-building

activity for native fish populations will depend on theextent to which a dam’s influence extends beyondthe pond: if beaver ponds support a high density ofnon-native fish only within the pond itself, then thenet impact on native fish populations at the drainagelevel may be negligible. However, if abundant non-native fish spill over into adjacent stream reaches, orif beaver ponds promote reproduction and maintainsource populations of non-native species, then beaverdams could have far-reaching consequences for theoverall composition of the fish assemblage.

Acknowledgements

We thank L. Kuehne and J. Walters for essential field assis-tance. We also thank K. Schonek of The Nature Conservancy

for providing useful information and insight into the ecologyof the Verde River system. J. Sorensen, the Arizona Gameand Fish Department, and The Nature Conservancy of Arizonaprovided substantial logistical support. D. Beauchamp, M.Pollock and two anonymous reviewers provided valuablecomments that improved the manuscript. Funding for PPGwas provided by a University of Washington Top ScholarGraduate Fellowship and by the H. Mason Keeler Endowmentfor Excellence through the School of Aquatic and Fishery Sci-ences. JDO was supported in part by an H. Mason KeelerEndowed Professorship and by the Department of Defense –Strategic Environmental Research and Development Program(RC-1724).

References

Anderson, M.J. 2001. A new method for non-parametric mul-tivariate analysis of variance. Austral Ecology 26: 32–46.

Anderson, M.J. 2006. Distance-based tests for homogeneity ofmultivariate dispersions. Biometrics 62: 245–253.

Arena, A., Ferry, L.A. & Gibb, A.C. 2012. Prey capturebehavior of native vs. nonnative fishes: a case study fromthe Colorado River drainage basin (USA). Journal of Experi-mental Zoology 317: 103–116.

Ayala, J.R., Rader, R.B., Belk, M.C. & Schaalje, G.B. 2007.Ground-truthing the impact of invasive species: spatio-tem-poral overlap between native least chub and introduced wes-tern mosquitofish. Biological Invasions 9: 857–869.

Barbour, M.T., Gerritsen, J., Snyder, B.D. & Stribling, J.B.1999. Rapid bioassessment protocols for use in streams andwadeable rivers: periphyton, benthic macroinvertebrates, andfish. Washington, DC: U.S. Environmental ProtectionAgency, Office of Water, EPA 841-B-99-002.

Beatty, R.J., Rahel, F.J. & Hubert, W.A. 2009. Complex influ-ences of low-head dams and artificial wetlands on fishes ina Colorado River tributary system. Fisheries Managementand Ecology 16: 457–467.

Billman, E.J., Kreitzer, J.D., Creighton, J.C., Habit, E.,McMillan, B. & Belk, M.C. 2012. Habitat enhancement andnative fish conservation: can enhancement of channel com-plexity promote the coexistence of native and introducedfishes? Environmental Biology of Fishes 96: 555–566.

Blasch, K.W., Hoffman, J.P., Grasner, L.F., Bryson, J.R. &Flint, A.L. 2006. Hydrogeology of the upper and middleVerde River watersheds, central Arizona. U.S. GeologicalSurvey Scientific Investigations Report 2005-5198.

Bonar, S.A., Leslie, L.L. & Velez, C.E. 2004. Influence ofspecies, size class, environment, and season on introducedfish predation on native fishes in the Verde River System,Arizona. Arizona Cooperative Fish and Wildlife ResearchUnit, Fisheries Research Report 01-04.

Carpenter, J. 2005. Competition for food between an intro-duced crayfish and two fishes endemic to the ColoradoRiver basin. Environmental Biology of Fishes 72: 335–342.

Carpenter, J. & Mueller, G.A. 2008. Small nonnative fishes aspredators of larval razorback suckers. The SouthwesternNaturalist 53: 236–242.

Childs, M.R., Clarkson, R.W. & Robinson, A.T. 1998.Resource use by larval and early juvenile native fishes inthe Little Colorado River, Grand Canyon, Arizona. Transac-tions of the American Fisheries Society 127: 620–629.

16

Gibson et al.

Clarkson, R.W., Marsh, P.C., Melis, T.S., Hamill, J.F., Coggins,L.G. Jr, Grams, P.E., Kennedy, T.A., Kubly, D.M. & Ralston,B.E. 2010. Effectiveness of the barrier-and-renovate approachto recovery of warmwater native fishes in the Gila Riverbasin. In: Melis, T.S., Hamill, J.F., Bennet, G.E., Coggins,L.G. Jr, Grams, P.E., Kennedy, T.A., Kubly, D.M. & Ralston,B.E., eds. Proceedings of the Colorado River basin scienceand resource management symposium. U.S. Geological Sur-vey Scientific Investigations Report 2010-5135: pp. 209–217.

Crowder, L.B. & Cooper, W.E. 1982. Habitat structural com-plexity and the interaction between bluegills and their prey.Ecology 63: 1802–1813.

Cucherousset, J. & Olden, J.D. 2011. Ecological impacts ofnonnative freshwater fishes. Fisheries 36: 215–230.

DeVries, P., Fetherston, K.L., Vitale, A. & Madsen, S. 2012.Emulating riverine landscape controls of beaver in streamrestoration. Fisheries 37: 246–255.

Dudley, R.K. & Matter, W.J. 2000. Effects of small greensunfish (Lepomis cyanellus) on recruitment of Gila chub(Gila intermedia) in Sabino Creek, Arizona. The Southwest-ern Naturalist 45: 24–29.

Fredlake, M. 1997. Environmental assessment: re-establish-ment of North American beaver (Castor canadensis) intothe San Pedro Riparian National Conservation Area. Tucson,AZ: U.S. Bureau of Land Management, EnvironmentalAssessment no. AZ-060–97-004.

Gauch, H.G. 1982. Multivariate analysis in community ecol-ogy. Cambridge: Cambridge University Press. 312 pp.

Gibson, P.P. & Olden, J.D. 2014. Ecology, management, andconservation implications of North American beaver (Castorcanadensis) in dryland streams. Aquatic Conservation:Marine and Freshwater Ecosystems 24: 391–409.

Gido, K.B., Propst, D.L., Olden, J.D., Bestgen, K.R. & Rosen-feld, J. 2013. Multidecadal responses of native and intro-duced fishes to natural and altered flow regimes in theAmerican Southwest. Canadian Journal of Fisheries andAquatic Sciences 70: 554–564.

Hanson, W.D. & Campbell, R.S. 1963. The effects of poolsize and beaver activity on distribution and abundance ofwarm-water fishes in a north Missouri stream. AmericanMidland Naturalist 69: 136–149.

Johnson, B.M., Martinez, P.J., Hawkins, J.A. & Bestgen, K.R.2008. Ranking predatory threats by nonnative fishes in theYampa River, Colorado, via bioenergetics modeling. NorthAmerican Journal of Fisheries Management 28: 1941–1953.

Karp, C.A. & Tyus, H.M. 1990. Behavioral interactionsbetween young Colorado squawfish and six fish species. Co-peia 1990: 25–34.

Kemp, P.S., Worthington, T.A., Langford, T.E.L., Tree,A.R.J. & Gaywood, M.J. 2012. Qualitative and quantitativeeffects of reintroduced beavers on stream fish. Fish andFisheries 13: 158–181.

Legendre, P. & Legendre, L. 1998. Numerical ecology.Amsterdam, New York: Elsevier. xvi + 990 pp.

Ley, G., Fencl, J.S., Kesner, B.R. & Marsh, P.C. 2012. Impe-rial Ponds native fish off-channel habitat progress and sum-mer telemetry study. Laughlin, NV: Presented at annualColorado River Aquatic Biologists meeting.

Light, T. 2003. Success and failure in a lotic crayfish inva-sion: the roles of hydrologic variability and habitat alter-ation. Freshwater Biology 48: 1886–1897.

Lokteff, R.L., Roper, B.B. & Wheaton, J.M. 2013. Do beaverdams impede the movement of trout? Transactions of theAmerican Fisheries Society 142: 1114–1125.

Magoulick, D.D. & Kobza, R.M. 2003. The role of refugia forfishes during drought: a review and synthesis. FreshwaterBiology 48: 1186–1198.

Marks, J.C., Haden, G.A., O’Neill, M. & Pace, C. 2010.Effects of flow restoration and exotic species removal onrecovery of native fish: lessons from a dam decommission-ing. Restoration Ecology 18: 934–943.

Marshall, R.M., Robles, M.D., Majka, D.R. & Haney, J.A.2010. Sustainable water management in the southwesternUnited States: reality or rhetoric? PLoS ONE 5: e11687.

Martinez, P. 2012. Invasive crayfish in a high desert river:implications of concurrent invaders and climate change.Aquatic Invasions 7: 219–234.

Martinez, P., Chart, T., Trammell, M., Wullschleger, J. &Bergersen, E. 1994. Fish species composition before and afterconstruction of a main stem reservoir on the WhiteRiver, Colorado. Environmental Biology of Fishes 40: 227–239.

Meffe, G.K. 1985. Predation and species replacement inAmerican southwestern fishes: a case study. The Southwest-ern Naturalist 30: 173–187.

Mills, M.D., Rader, R.B. & Belk, M.C. 2004. Complex inter-actions between native and invasive fish: the simultaneouseffects of multiple negative interactions. Oecologia 141:713–721.

Minckley, W.L. & Marsh, P. 2009. Inland fishes of the greaterSouthwest: chronicle of a vanishing biota. Tucson: Univer-sity of Arizona Press. xxxiv + 426 pp.

Minckley, W.L., Marsh, P.C., Deacon, J.E., Dowling, T.E.,Hedrick, P.W., Matthews, W.J. & Mueller, G. 2003. A con-servation plan for native fishes of the lower Colorado River.BioScience 53: 219–234.

Moody, E.K. & Taylor, C.A. 2012. Red swamp crawfish(Procambarus clarkii) discovered in the San Pedro River,Arizona: a new invader in a threatened ecosystem. TheSouthwestern Naturalist 57: 339–340.

Moorman, M.C., Eggleston, D.B., Anderson, C.B., Mansilla,A. & Szejner, P. 2009. Implications of beaver (Castorcanadensis) and trout introductions on native fish in theCape Horn Biosphere Reserve, Chile. Transactions of theAmerican Fisheries Society 138: 306–313.

Mueller, G.A. & Marsh, P.C. 2002. Lost, a desert river and itsnative fishes: a historical perspective of the Lower ColoradoRiver. U.S. Geological Survey Information and TechnologyReport USGS/BRD/ITR-2002-0010.

Murphy, M.L., Heifetz, J., Thedinga, J.F., Johnson, S.W. &Koski, K.V. 1989. Habitat utilization by juvenile Pacific sal-mon (Onchorynchus) in the glacial Taku River, SoutheastAlaska. Canadian Journal of Fisheries and Aquatic Sciences46: 1677–1685.

Naiman, R.J., Melillo, J.M. & Hobbie, J.E. 1986. Ecosystemalteration of boreal forest streams by beaver (Castor canad-ensis). Ecology 67: 1254–1269.

Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin,P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Henry, M.,Stevens, H. & Wagner, H. 2011. Vegan: community ecologypackage. R package version 2.0-9. Available at: http://CRAN.R-project.org/package=vegan (accessed 25 September 2012)

17

Beaver dams shift desert fish assemblages

Olden, J.D. & Poff, N.L. 2005. Long-term trends of nativeand non-native fish faunas in the American Southwest. Ani-mal Biodiversity and Conservation 28: 75–89.

Olden, J.D., Poff, N.L. & Bestgen, K.R. 2006. Life-historystrategies predict fish invasions and extirpations in the Colo-rado River basin. Ecological Monographs 76: 25–40.

Pilger, T.J., Gido, K.B. & Propst, D.L. 2010. Diet and trophicniche overlap of native and nonnative fishes in the GilaRiver, USA: implications for native fish conservation. Ecol-ogy of Freshwater Fish 19: 300–321.

Pollock, M.M., Heim, M. & Werner, D. 2003. Hydrologic andgeomorphic effects of beaver dams and their influence onfishes. In: Gregory, S.V., Boyer, K. & Gurnell, A., eds. Theecology and management of wood in world rivers. Bethesda,MD: American Fisheries Society, pp. 213–233.

Pollock, M.M., Pess, G.R., Beechie, T.J. & Montgomery,D.R. 2004. The importance of beaver ponds to coho salmonproduction in the Stillaguamish River basin, Washington,USA. North American Journal of Fisheries Management 24:749–760.

Pollock, M.M., Beechie, T.J. & Jordan, C.E. 2007. Geomorphicchanges upstream of beaver dams in Bridge Creek, an incisedstream channel in the interior Columbia River basin, easternOregon. Earth Surface Processes and Landforms 32: 1174–1185.

Pool, T.K. & Olden, J.D. 2014. Assessing long-term fishresponses and short-term solutions to flow regulation in adryland river basin. Ecology of Freshwater Fish. DOI:10.1111/eff.12125.

Pool, T.K., Strecker, A.L. & Olden, J.D. 2013. Identifyingpreservation and restoration priority areas for desert fishes inan increasingly invaded world. Environmental Management3: 631–641.

Propst, D.L., Gido, K.B. & Stefferud, J.A. 2008. Natural flowregimes, nonnative fishes, and native fish persistence in arid-land river systems. Ecological Applications 18: 1236–1252.

Pyke, G.H. 2008. Plague minnow or mosquito fish? A reviewof the biology and impacts of introduced Gambusia species.Annual Review of Ecology, Evolution, and Systematics 39:171–191.

R Development Core Team. 2011. R: a language and environ-ment for statistical computing. Vienna, Austria: R Founda-tion for Statistical Computing.

Rinne, J.N. 2005. Changes in fish assemblages, Verde River,Arizona, 1974-2003. In: Rinne, J.N., Hughes, R.M. & Cala-musso, B., eds. Historical changes in fish assemblages oflarge rivers in the Americas. Bethesda, MD: American Fish-eries Society, pp. 115–126.

Rinne, J.N. 2012. Fish and aquatic organisms. In: Neary,D.G., Medina, A.L. & Rinne, J.N., eds. Synthesis of upperVerde River research and monitoring 1993–2008. Fort Col-lins: USDA Forest Service, Gen. Tech. Rep. RMRS-GTR-291, pp. 189–232.

Rinne, J.N. & Miller, D. 2006. Hydrology, geomorphologyand management: implications for sustainability of native

southwestern fishes. Reviews in Fisheries Science 14:91–110.

Rosell, F., Bozser, O., Collen, P. & Parker, H. 2005. Ecologi-cal impact of beavers Castor fiber and Castor canadensisand their ability to modify ecosystems. Mammal Review 35:248–276.

Schlosser, I.J. 1995. Dispersal, boundary processes, and tro-phic-level interactions in streams adjacent to beaver ponds.Ecology 76: 908–925.

Schlosser, I.J. 1998. Fish recruitment, dispersal, and trophicinteractions in a heterogeneous lotic environment. Oecologia113: 260–268.

Schlosser, I.J. & Kallemeyn, L.W. 2000. Spatial variation infish assemblages across a beaver-influenced successionallandscape. Ecology 81: 1371–1382.

Shafroth, P.B., Wilcox, A.C., Lytle, D.A., Hickey, J.T.,Andersen, D.C., Beauchamp, V.B., Hautzinger, A., McMul-len, L.E. & Warner, A. 2010. Ecosystem effects of environ-mental flows: modelling and experimental floods in adryland river. Freshwater Biology 55: 68–85.

Smith, J.M. & Mather, M.E. 2013. Beaver dams maintain fishbiodiversity by increasing habitat heterogeneity throughout alow-gradient stream network. Freshwater Biology 58: 1523–1538.

Snodgrass, J.W. & Meffe, G.K. 1998. Influence of beavers onstream fish assemblages: effects of pond age and watershedposition. Ecology 79: 928–942.

Snodgrass, J.W. & Meffe, G.K. 1999. Habitat use and tempo-ral dynamics of blackwater stream fishes in and adjacent tobeaver ponds. Copeia 1999: 628–639.

Stefferud, J.A., Stefferud, S.E. & Clarkson, R.W. 2009.Fishes: historical changes and an imperiled native fauna. In:Stromberg, J.C. & Tellman, B., eds. Ecology and conserva-tion of the San Pedro River. Tucson: University of ArizonaPress, pp. 192–216.

Stefferud, J.A., Gido, K.B. & Propst, D.L. 2011. Spatially var-iable response of native fish assemblages to discharge, pre-dators and habitat characteristics in an arid-land river.Freshwater Biology 56: 1403–1416.

Turner, D.S. & List, M.D. 2007. Habitat mapping and conser-vation analysis to identify critical streams for Arizona’snative fish. Aquatic Conservation: Marine and FreshwaterEcosystems 17: 737–748.

Twardochleb, L.A., Olden, J.D. & Larson, E.R. 2013. A glo-bal meta-analysis of the ecological impacts of nonnativecrayfish. Freshwater Science 32: 1367–1382.

USGS (U.S. Geological Survey). 2013. National Water Infor-mation System. Available at: http://waterdata.usgs.gov/nwis/(accessed 6 September 2013).

White, S.M. & Rahel, F.J. 2008. Complementation of habitatsfor Bonneville cutthroat trout in watersheds influenced bybeavers, livestock, and drought. Transactions of the Ameri-can Fisheries Society 137: 881–894.

18

Gibson et al.