|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 HABITAT RELATIONSHIPS OF SMALL BODIED FISH IN THE GRAND CANYON REACH OF THE COLORADO RIVER, ARIZONA: EMPHASIS ON NATIVE FISH AND EVALUATION OF BACKWATER HABITATS By MICHAEL JAMES DODRILL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012
2 2012 Michael James Dodrill
3 This thesis is dedicated to my loving and sup portive parents.
4 ACKNOWLEDGMENTS I would like to thank my graduate advisor, Dr. William E. Pine, III for support and guidance during my time at the University of Florida. I would also like to thank my committee, Dr s Mike Allen and Joann Mossa. I thank my fellow graduate students who worked on the Near Shore Ecology project, Brandon Gerig and Colton Finch for tremendous help in the field and enlightening discussions in the office. Many people were involved in the field sampling effort and deserv e special thanks, Jessie Pierson, Jake Hall, Nick Bene, Max Kn ight and numerous others I would also like to thank the boatmen Brian Dierker, Steve Jones, Brian Smith, and others for safe trips down the Colorado River The USGS, Grand Canyon Monitoring and Research Center staff provi ded tremendous support and I thank Mike Yard, Bill Persons, Glen Bennet t Carol Fritzinger, Dave Foster, and Seth Felder. Most of all, I thank my parents (Jim and Julie) for their support and encouragement throughout the years.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTI ON ................................ ................................ ................................ .... 13 Altered Riverine Ecosystems and the Decline of Native Fish Populations .............. 13 River Restoration ................................ ................................ ................................ .... 14 Adaptive Management of Native fish in the Colorado River, Grand Canyon ........... 16 Habitat Requirements and Physical Habitat Restoration ................................ ........ 17 Research Needs and Study Objectives ................................ ................................ .. 18 2 ABUNDANCE, DENSITY, AND HABITAT SELECTION PATTERNS OF SMALL BODIED FISH WITH EMPHASIS ON NATIVE FISH CONSERVATION AND BACKWATER HABITATS ................................ ................................ ....................... 20 Methods ................................ ................................ ................................ .................. 24 Study Site ................................ ................................ ................................ ......... 24 Fish Sampling in Backwater Habitats ................................ ............................... 25 Fish Sampling in Other Habitat Types ................................ .............................. 26 Abundance Estimation ................................ ................................ ..................... 27 Available Nearshore Habitat ................................ ................................ ............. 29 Density Estimation and Habitat Selection ................................ ......................... 30 Results ................................ ................................ ................................ .................... 31 Catch ................................ ................................ ................................ ................ 31 Capture Probability ................................ ................................ ........................... 32 Habitat Availability ................................ ................................ ............................ 32 Abundance by Habitat Type ................................ ................................ ............. 32 Density by Habitat Type ................................ ................................ ................... 34 Habitat Selection ................................ ................................ .............................. 35 Discussion ................................ ................................ ................................ .............. 36 3 DIFFERENTIAL PRE DATION RISK: ASSESSING THE ROLE OF HABITAT AND TURBIDITY IN A LARGE RIVER ECOSYSTEM ................................ ............ 71 Methods ................................ ................................ ................................ .................. 76 Study Site ................................ ................................ ................................ ......... 76 Tethering Trials ................................ ................................ ................................ 76
6 Analysis ................................ ................................ ................................ ............ 78 Results ................................ ................................ ................................ .................... 79 Discussion ................................ ................................ ................................ .............. 80 4 CONCLUSIONS ................................ ................................ ................................ ..... 89 LIST OF REFERENCES ................................ ................................ ............................... 93 BIOGRAPHIC AL SKETCH ................................ ................................ .......................... 106
7 LIST OF TABLE S Table page 2 1 Definitions of habitat types, modified from Converse et al. (1998). ..................... 43 2 2 Common and scientific name for each taxon. ................................ ..................... 44 2 3 Fish species and the number captured with seines in backwater habitats each month d uring 2009 depletion sampling. ................................ .................... 45 2 4 Fish species and the number captured with seines in backwater habitats each month during 2010 depletion sa mpling ................................ ...................... 46 2 5 Fish species and the number captured with electrofishing in talus, debris fan, cliff, and sand ha bitats during 2009 ................................ ................................ .... 47 2 6 Fish species and the number captured with electrofishing in talus, debris fan, cliff, and sand habitats during 2010 ................................ ................................ .... 48 2 7 Total available shoreline habitat sampled during 2009 and 2010 ....................... 49 2 8 estimates ( ) for speckled dace during 2009 ................................ ..................... 50 2 9 estimates ( ) f or speckled dace during 2010 ................................ ..................... 51 2 10 estimates ( ) for rainbow trout < 100 mm TL during 2009 ................................ 52 2 11 estimates ( ) for rainbow trout < 100 mm TL durin g 2010 ................................ 53 2 12 estimates ( ) for fathead minnow du ring 2009 ................................ ................... 54 2 13 estimates ( ) for fathead minnow durin g 2010 ................................ ................... 55 2 14 estimates ( ) for plains killifish during 2009 ................................ ....................... 56 2 15 estimates ( ) for plains killifish during 2 010 ................................ ....................... 57 3 1 Turbidity shown in nephelometric turbidity units (NTU) during each of the experimental trials. ................................ ................................ ............................. 85
8 3 2 Akaike Information Criteria (AIC) model weighting of habitat and turbidity heterogeneity of tethered fish mortality ................................ ............................... 86 3 3 Pairwise comparisons of Wilcox rank sum tests with Bonferroni correction applied (significance level = 0.05) ................................ ................................ ...... 87
9 LIST OF FIGURES Figure page 2 1 Map showing the Colorado River in Marble and Grand Canyons, Arizona. The study area is located just downstream of the confluence of the Colorado and Little Co lorado River ................................ ................................ .................... 58 2 2 Figure showing the three sampling sites used for the electrofishing mark recapture. Each site is approximately 1500 m of shoreline on each site of the r i ver ................................ ................................ ................................ .............. 59 2 3 Overhead aerial photo of the Colorado River within the sampling area. The photo is overlain with a GIS layer demarcati ng the 50 m ................................ ................................ ................................ ...... 60 2 4 Capture probability of small bodied fish (40 99 mm TL, all species) sampled using boat electrofishing st ratified by habitat type ................................ 61 2 5 Juvenile humpback chub (40 99 mm TL) abundance estimates by habitat type during July, August, September, and October of 2009 (A ) and 2010 (B) .... 62 2 6 Juvenile bluehead sucker (40 149 mm TL) abundance estimates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B) .... 63 2 7 Juvenile fannelmouth sucker (40 149 mm TL) abundance estimates by habitat type during July, August, September, and October of 20 09 (A) and 2010 (B) ................................ ................................ ................................ .............. 64 2 8 Juvenile humpback chub (40 99 mm TL) density estimates by habitat type during July, August, Septembe r, and October of 2009 (A) and 2010 (B) ............ 65 2 9 Juvenile bluehead sucker (40 149 mm TL) density estimates by habitat type during July, August, September, and Oc tober of 2009 (A) and 2010 (B) .... 66 2 10 Juvenile flannelmouth sucker (40 149 mm TL) density estim ates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B) .... 67 2 11 n estimates ( ) for juvenile humpback chub (40 99 mm TL) during 2009 (A) a nd 2010 (B) ................................ ........................ 68 2 12 estimates ( ) for juvenile bluehead sucker (40 149 mm TL) d uring 2009 (A) and 2010 (B) ................................ ............... 69 2 13 ) for juvenile flannelmouth sucker (40 149 mm TL) during 2009 (A) a nd 2010 (B) ................................ .... 70
10 3 1 Predation risk expressed as percent mortality of tethered fish in each experimental treat ment ................................ ................................ ....................... 88
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for th e Degree of Master of Science HABITAT RELATIONSHIPS OF SMALL BODIE D FISH IN THE GRAND CANYON REACH OF THE COLORADO RIVER, ARIZONA: EMPHASIS ON NATIVE FISH AND EVALUATION OF BACKWATER HABITATS By Michael James Dodrill August 2012 Chair: William Pine Major: Fisheries and Aquatic Sciences Many management actions in aquatic ecosystems are directed at restoring or improving specific habitats to improve fish growth and survival. In the Grand Canyon reach of the Colorado River, experimental flow operations as part of the Glen Canyon Dam Adapti ve Management Program often consider the creation of sandbars and associated backwater habitats as critical for juveni le native fish conservation. T hese habitats are thought to be warmer and more s imilar to the pre dam environment thus providing areas for improved growth and survival for native fish Previous research has demonstrated that both native and non native fishes occur in backwater habitats including the federally endangered humpback chub Gila cypha yet little is known about the distribution of fish among other habitat types. This study compared abundance, density and habitat selection patterns between shoreline habitats (cliff, talus, debris fan, sand, backwater) and quantified relative predation risk to compare backwaters to other availabl e h abitats. Within the study area, abundance of juvenile humpback chub was consistently highest in talus habitats and lowest in backwater habitats. Backwater habitat contained low total abundance of many fish species in this study, but densities of
12 many native and n on native fish were higher in backwaters compared to other habitats. Many species, including juvenile humpback chub showed po sitive selection for backwaters and other patterns of habitat selection were largely specific to individual species D espite the high densities and selection for backwater habitats by native fish species, the spatial extent of backwaters was small and supports a small proportion of the overall fish abundance. T hus management actions directed at manipulating this habitat type will a ffect a small proportion of the overall abundance of native fish in this river reach Concerns over predation in backwater habitats were evaluated to test whether backwater and debris fan habitat has different relative predat ion risks for juvenile fish across a range in turbidity Under low and intermediate turbidity conditions, debris fan habitat showed higher relative predation risk compared to backwater habitat. These results suggest turbidity is an important factor mediating predation risk and should be considered when managing for certa in habitat types.
13 CHAPTER 1 INTRODUCTION Altered Riverine Ecosystems and the Decline of Native Fish Populations Riverine ecosystems are highly modified throughout the world to provide human populations with numerous socioeconomic goods and services. Societal needs for water, energy and transportation are commonly met through engineered solutions such as dams, navigation channels, and flood control st ructures (Gore and Shields 1995; Gleick 2003). These solutions, which modify riverine flows to meet human needs, are common with great er than half of large river systems impacted by dams globally and 77% impacted in the northern hemisphere (Dynesius and Nilsson 1994; Gore and Shields 1995; Nilsson et al 2005). Common physical modifications to riverine environments include altered flow r egimes, loss of lateral and longitudinal connectivity, and channel manipulations includ ing dredging and engineered structures aimed at improving navigation. A dampening in the timing, duration, and magnitude of flood events has widely occurred via dam and reservoir management aimed at providing more consistent and reliable water supplies (Moyle and Mount 2007; FitzH ugh and Vogel 20 10). Alteration of flow regimes has led to loss of lateral and longitudinal connectivity essential to riverine species that ha ve adapted life history strategies that exploit resources dictated by natural flo ws (Bunn and Arthington 2002). Concurrent with these changes to river flow regimes and the physical structure of riverine habitats are often large scale biological changes. C hanges in aquatic biodiversity (Bunn and Arthington 2002), food web structure (Power 1996), and fish community structure (Bain et al. 1988) have been attributed i n
14 part to the altered physical template of riverine eco systems Native fish species are often most adversely affected (Moyle and Mount 2007), culminating in local declines in abundance or extirpation, a trend observed across the American West ( Minckley et al. 2003; Olden and Poff 2005 ). Altered conditions often faci litate the establishment of non native species (Lozon and MacIsaac 1997), which have largely contributed to declines in native species via a set of both direct and indirect mec hanisms (Allan and Flecker 1993; Richter et al 1997). Furthermore, nonnative species are often implicated as a leading factor in the decline of native fish species (Olden and Poff 2005; Coggins et al. 2011; Cucherousset a nd Olden 2011; Yard et al. 2011 ). River Restoration River restoration activities have increased exponentially over the past decade in North Amer ica with annual spending topping 1 billion dollars in the continental U.S. (Bernhardt et al. 2005). A common tenant of these restoration activities in regulated rivers is that they often seek to restore key hydrologic features of the unmodified flow regime often with the expectation that depressed native species populations wil l respond positive ly (Stanford et al. 1996; Bednarek and Hart 2005). Flow restoration activiti es are commonly aimed at benefiting native fish species by restoring the naturally variable flows that create and maintain a diversity of habitats types (Poff et al 1997). Succes sful restoration of native riverine fishes may need to consider more than just restoring elements of the physical environment. Non native biota often proliferates in modified lotic ecosystems and can influence patterns of native biodiversity (Townsend and Crowl 199 1; Pine et al. 2007). Non native control or active management of these species may also be needed in order to successfully restore native fish communities (Lessard et al. 2005 ).
15 Although river restoration and habitat manipulations are often succ essful in small streams (Riley a nd Fausch 1995; Gowan and Fausch 1996), these techniques are often not applied to larger systems (Gore and Shields 1995). As an example, i n stream structures such as boulder weirs or deflectors, and large woody debris are o ften used to restore or create habitat in streams and small rivers ( Roni et al. 2008 ) yet the scale of large rivers precludes the application of common site based approaches ( Gore and Shields 1995 ). Thus managers of large river systems are often limited in habitat resto ration or enhancement techniques and must primarily rely on flow based app roaches Despite these efforts, large river restoration has been viewed as largely unsuccessful in restoring native fauna, increasing biodiversity and improving ec ological function (Bernhardt et al. 2005). Numerous reasons are cited for this failure including a lack of studies demonstrating population responses to habitat manipulations, or inappropriate scale of restoration efforts (Bernhardt et al. 2005; Wohl 2005; Bernhardt and Palmer 2011). Large river s are inherently complex mosaics of environments, aquatic species, and human users all with diverse needs and expectations of the riverine ecosystem. This complexity creates extensive management challenges req uiring innovative solutions from resource managers tasked with protecti ng or restoring riverine ecosystems In regulated rivers, biotic and abiotic responses to flow operations are often not known with certainty P re scribing water releases from dams in o rder to illicit specific fish population responses downstream is an intuitively appealing solution to mitigating environmental impacts to flow modifications. However, ecosystem responses to these types of actions are often highly uncertain and examples o f counterintuitive
16 response s exist (Pine et al. 2009). One approach to managing riverine and other ecosystems is not to prescribe management actions but instead adaptively manage an ecosystem in ways to explicitly learn about the ecosystem by assessing re sponses to management actions. Adaptive management is an approach to confronting uncertainty in ecosystem responses to natural resource policy decisions that treats policy choices as explicit experiments (Walters and Hilborn 1978; Walters 2007) Adaptiv e management thu s provides a potential framework for addressing the complex issues managers face in large river ecosystems. Adaptive Management of Native fish in the Colorado River, Grand Canyon The Glen Canyon Dam Adaptive Management Program (GCDAMP) was formed in 1996 and is charged with managing the Colorado River ecosystem within Grand Canyon. Under this framework multiple management goals for the Colorado River ecosystem have been established including meeting water and power production needs, protecting recreational and cultural resources, and conservation of native species including humpback chub. As an Endangered species under the US Endangered Species Act (BO; USDOI 2008) protection and ultimately population recovery of humpback chub is a focal aspect of the GCDAMP. T he 2008 Biological Opinion ( USDOI 2008) on the operation of Glen Canyon Dam (GCD) identifies conservation measures necessary to conserve and protect native fi sh. The BO outlines several experimental actions designed to benef it native fish by manipulating downstream habitats including backwaters Backwater habitats have been the focus of several experimental habitat manipulations and have received intense study in Grand Canyon for about 20 years ( AZGFD 1996; Ho ffnagle 2000; G rams et al. 2010 ) B ackwater habitats are areas of low velocity water that are partially isolated from
17 the main channel, often in the lee of an emergent reattachment sandbar or tributary debris fan (Behn et al. 2010). These areas are perceived to be impo rtant to juvenile native fish due to warmer w ater temperatures than the main stem river that more closely approach the pre dam environmental conditions ( AZGFD 1996; Trammell 2 002). P revious research has extensively sampled backwater habitats throughout Gra nd Canyon (AZGFD 1996; Hoffnagle 2000; Grams et al. 2010) and has suggested that one possible reason for the observed decline of native fish populations in Grand Canyon may be related to the reduction in availability of backwater habitats (Valdez and Ryel 1995). Habitat Requirements and Physical Habitat Restoration Physical habitat restoration or modification i s one of the main tools used in fish and wildlife management to benefit degraded animal populations. In the U. S. and Canada, habitat protection a nd re storation is mandated by legislation (U.S. Endangered Species Act; Canadian Species at Risk Act). In 1996, reauthorization of the Magnuson Stevens Fishery Conservation and Management Act through the Susta inable Fisheries Act took place. This legislat ion contained a provision that Essential Fish Habitat (EFH) must be identified for marine and estuarine species of economic or cultural importance managed by eight Regional Fishery Management Councils that oversee t he U.S. Exclusive Economic Zone (EEZ) This legislation demonstrates the continued view of the importance of habitat characteristics to fish populations. In order for habitat restoration efforts to benefit target populations, it is essential to understand species habitat associations and ident ify the approp riate habitats for restoration Habitat preference, selection and requirement are often poorly defined ter ms (Rosenfeld and Hatfield 2006 ) and my definitions of these terms follow definitions
18 provided by Rosenfeld and Boss (2000) H abitat preference is determined in choice experiments where environmental or b iotic factors can be controlled H abitat selection occurs when a species shows differential use of a particular habitat type (Rosenfeld and Boss 2000). H abitat requirement refers to a feature of the environment that is necessary for the growth and persistence of individuals and populations (Rosenfeld and Boss 2000) Typically habitats which have high abundance of target species (Bond and Lake 2003) and for which species are positively selecting are identified for possible restoration efforts (Rosenfeld and Hatfield 2006). Determining habitat use and selection patterns is the first step to identifying critical habitats required by species (Rosenfeld and Boss 2001 ). Research Needs and S tudy Objectives In order for effective restoration of native fish species in riverine ecosystems to occur, habitat requirements must be established to inform restoration efforts In the Grand Canyon reach of the Colorado River, backwater habitats have been the targets of habitat restoration, yet it is uncertain whether these habitats are required for native fish. Comparing small bodied native and nonnative fish abundance, density, and habitat selection patterns between backwaters and other availa ble habitats will aid in managing the physical resources downstream of GCD to benefit native fish conservation goals. The overall goal of this thesis is to further our knowledge about habitat relationships of both native and nonnative small bodied fish using shoreline habitats in Grand Canyon More specifically this thesis addresses backwater habitats by combining inferences on habitat selection with an experiment to assess predation risk in backwaters. The objectives of chapter two are to (1) compare abundance of small bodied fish between habitats with emphasis on humpback chub, (2) compare densities
19 of juvenile and small bodied native and nonnative fish between habitats and (3) determine habitat selection patterns of small bodied fish. The objective of chapter three is to assess the relative predation risk between a backwater habitat, and a debris fan habitat across a range of turbidities.
20 CHAPTER 2 A BUNDANCE, DENSITY, A ND HABITAT SELECTION PATTERNS O F SMALL BODIED FISH WITH EMP HASIS ON NATIVE FISH CONSERVATION AND BACKWATER HABITATS A large proportion of the fish and wildlife management efforts in the U.S. and Canada are concerned with restoration of degraded populations, especia lly threatened or endangered species. A central tenet of this effort is the preservation and restoration of habitats assumed to be critical to the species persistence (U.S. Endangered Species Act; Canadian Species at Risk Act). This emphasis on physical habitat management arose from observations that certain habitats are often disproportionally important to species (Fausch et al. 2002) and that species declines often correspond with habitat degradation or loss of habitat heterogeneity (Rosen z weig 1995). Riverine ecosystems are often highly modified in terms of species and habitats (Dynesius and Nilsson 1994; Nilsson et al. 2005) and as with other freshwater ecosystems, rivers have experienced losses in biodiversity (Ricciardi and Rasmussen 1999; Strayer a nd Dudgeon 2010) and declines of native species (Minckley et al. 2003; Olden and Poff 2005) and hence have received much interest in restoration (Bernhardt et al. 2005). Since 1990, river restoration efforts have increased in the continental U.S. with an nual spending exceeding 1 billion dollars annually (Bernhardt et al. 2005). In arid climates, sources of physical and biological modification are often the result of river regulation from dams or diversions and the corresponding approaches to restoring ha bitat has focused on reestablishing historical flow elements, or timing releases to reconnect or create historical riparian and aquatic habitats (Stromberg 2001; Bednarek and Hart 2005; Bernhardt and Palmer 2011). These efforts are motivated by both ecolo gical theory and empirical work demonstrating a positive relationship between
21 habitat heterogeneity and species diversity (Ricklefs and Schluter 1993; Tews et al. 2004; Palmer et al. 2010). Building on this foundation is the belief that restoring a divers ity of habitat types in river ecosystems will result in increased biodiversity (Palmer 2010). In order for habitat restoration efforts to benefit target populations, it is essential to understand spec ies habitat relationships by identifying habitats that species are positively selecting Typically, habitats have been targeted for restoration by associating high abundance of target species and inferring high habitat quality (Bond and Lake 2003). Although this approach is often unavoidable, considering th e availability of habitats and determining selection patterns is preferred (Rosenfeld and Hatfield 2006). Determining habitat selection patterns and coupling fitness measures within selected habitats provides stronger inference in identifying critical hab itats (Rosenfeld and Boss 2001). Identifying habitat selection and critical habitats is necessary because this allows targeted habitats to be restored and increases the chances of restoration success (Bond and Lake 2003 ). If such habitats are not identif ied then efforts such as habitat protection or enhancement might not have the intended consequences of benefiting the target species (Rosenfeld and Hatfield 2006) and undesirable or unforeseen consequences of restoration may e merge (Ebersole et al. 1997). The Colorado River has been the site of extensive development of water infrastructure since the 1930s creating one of the most regulated and highly controlled rivers on E arth (Fradkin 1981). Concurrently, populations of native fish have declined across the basin, with seven of t he 14 native fish species listed as imperiled (Jelks et al.
22 2008) and four species listed as federall y endangered : humpback chub Gila cypha bonytail Gila elegas Colorado pikeminnow Ptychocheilus l ucius and razorback sucker Xyra uchen texanu s (Minckley et al 2003). D ecades of research by federal and state agencies in the Colorado Basin have focused on restoring native fish populations. In the Grand Canyon reach of the Colorado River three high flow experiments designed to mimi c elements of pre dam floods have occurred in 1996, 2004, and 2008 to inform restoration activities as part of the GCDAMP. These floods represent an effort to create hydrologic and habitat conditions that mimic the pre dam environment (Patten et al. 2001 ) Restoring high flow events is one strategy to aid in restoring the physical and ecological resources of Grand Canyon an d reverse the observed declines in the physical and biological resources, including fish One aspect of the pre dam environment that these floods sought to enhance was the redistribution of sand and sediment from the river bed to shoreline areas. Historically 60 million tons of sediment annually were conveyed through Marble and Grand Canyon (Topping et al. 2000) However, 90 % (or more) of this sediment is now retained in Lake Powell following the closure of Glen Canyon Dam in 1963 (Topping et al. 2000). One consequence of this sediment loss has been a likely decrease in the extent and complexity of shoreline sand deposits wh ich may provide habitat for juvenile native fish ( Schmidt and Graf 1990 ; Converse et al. 1998 ) While backwater habitats are generally thought by managers to have declined in Grand Canyon, Goeking et al. (2003) assessed a time series of aerial photography images from 1935 2000 and found high inter annual variation in backwaters but no progressive decline over time.
23 Another change from the pre dam conditions is the shift in temperatures from high seasonal variation to much colder with little variation. This is due to epilimnetic releases from GCD which likely reduce humpback chub growth rates and may affect survival of larvae ( Robinson and Childs 2001; Coggins and Pine 2010 ). Low rates of water exchange between backwater habitats (Behn et al. 2010 ) and the mainstem river an d generally shallow depths results in warmer temperatures in backwater habitat than the mainstem river (AZGFD 1996). This warm water is another reason backwaters are considered important to juvenile native fish because the warmer tem peratures more closely mimic the pre dam conditions ( Hoffnagle 1996; Trammell et al. 2002 ). In Grand Canyon, r esear ch has documented the presence of juvenile humpback chub in backwater habitats throughout Grand Canyon (AZGFD 1996; Hoffnagle 2000; Grams et al. 2010) and the physical processes associated with the creation and persistence of sandbars has been extensi vely studied (Rubin et al. 1990; Topping et al. 2005 ). The creation and maintenance of backwater habitats to benefit native fish is one of two fu as a flow policy from GCD. Despite years of research and the implementation of experimental flow policies based on the assumption that backwater habitats are critical for native fish, f undamental questions remain related to the role of backwaters and native fish ecology in Grand Canyon Key uncertainties of management importance include whether or not backwaters are sele cted over other habitat types and whether backwater habitats are required for native fish populations to persist This is important because backwaters are a habitat type that can be managed for through experimental
24 floods from GCD coupled with sediment inputs from unregulated tributaries such as the Paria and Little Colorado rivers. Previous research into habitat relationshi ps of juvenile humpback chub has identified talus and debris fan habitats as having higher relative densities of humpback chub than bedrock or sand habitats (Conver se et al. 1998) Although earlier work did not calculate habitat selection results from Converse et al. (1998) suggest that juvenile humpback chub select talus habitats over other, more homogenous habitat types as a predation refuge. Additionally, facto rs such as differential food availability or energetic costs associated with higher flow conditions could influence habitat selection patte rns of juvenile humpback chub. To date, no habitat studies have quantified habitat selection for juvenile humpback c hub or other small bodied fish in Grand Canyon. This information is critical to native fish conservation efforts because management actions such as annual flow levels or prescribed experimental floods can be done to create or protect specific habitat t ypes The objectives of this chapter were to (1) compare abundance of small bodied fish between habitats with emphasis on humpback chub, (2) compare densities of juvenile and small bodied native and nonnative fish between habitats and (3) determine habita t selection patterns of small bodied fish. This study investigates habitat relationships of seven species of both native and nonnative fish commonly captured in nearshore areas. This information is vital to guide management efforts designed to benefit n a tive fish in the Grand Canyon Methods Study Site This study was conducted in the Little Colorado River (LCR) inflow reach of the mainstem Colorado River in Grand Canyon, Arizona (Figure 2 1). Sampling covered an
25 area from below Heart Island (102 km ) t o Lava Chuar rapid (about 106 km as measured downstr eam from Lees Ferry, Arizona). Sampling occurred on four trips annually in July, August, September, and October, during 2009 and 2010. Th is area of river was chosen because the largest aggregation of humpback chub occur s in this river reach (Gorman and Stone 1999) Additionally this reach has a high density of debris fan complexes that potentially form backwater habitats (Schmidt and Graf 1990) a s well as other common habitat types in which to compare habitat use of juvenile humpback chub. Fish Sampling in Backwater Habitats During each sampling trip, backwaters in the study area were sampled with seines (9 25 m in length depending on backwa ter size x 1.2 m height x 3 mm mesh) during nighttime hours at the beginning and end of each sampling trip (generally 12 days between sampling events) Sampling was conducted at night because dam operations during July and August produce daily fluctuation s in river stage such that backwater habitats are only available during the nighttime low flow period. During most of the daylight hours, backwater habitats were of ten unavailable due to increased stage height that overtops the sandbar which isolates the backwater from the main channel of the river. Prior to seining a backwater, a large seine was deployed across the mouth of the backwater to prevent fish from escaping the backwater prior to sampling. After closing the backwater, multiple seine pass remov al s were completed until the backwater was depleted of fish (minimum of five passes). After each pass, individual fish were removed from the net and placed in aerated buckets for processing. Individual fish were identified to species lengths were record ed (total length TL and fork length FL), and identifying visual imp lant elastomer (VIE) marks or passive integrated transponder
26 (PIT) tags given. The VIE marks were given to humpback chub between 40 99 mm TL and humpback chub over 99 mm TL were given a PI T tag. The VIE marks were given to all other species greater than 40 mm TL and native fish 150 mm TL also received a PIT tag. A coded combination of color and body location of the VIE mark was used to designate the sampling gear (seines) sampling trip, and sampling habitat (backwater habitat) If fish were subsequently collected outside of the backwater habitat by cooperative fish sampling efforts, the capture history information (describing habitat use) for this fish could be decoded. After processin g fish were released into the backwater habitat. Fish Sampling in Other Habitat Types Three sampling sites (sites 1, 2 and 3) were established within the study area of approximately 1500 m each depending on hydrologic features (Figure 2 2). Each Site wa s then subdivided further into individual spatially referenced 50 at Sampling Unit s s Figure 2 3). All electrofishing was conducted after dark between 1900 and 0300 hours. Each site was repeatedly sampled three times over five day periods in 2009 for sites one and two (48h between passes), over successive nights in site 3 (2009, 24 h between pas ses) and over successive nights in all of 2010 (24 h between passes). Experienced boat operators used slow speed boat mounted electrofishing (5 10 seconds per meter of shoreline) to methodically sample nearshore areas for small bodied fish (See Korman et al. 2009 for additional description). Aluminum hulled boats (4.9 m length) with 5000 W generators and Coffelt Mark XXII control units were used. Fish captured in each HSU were placed in a numbered bucket corresponding to the HSU to track catch by habita t. Each HSU was classified with one of five discrete habitat types (backwater, sand, debris fan, talus, cliff, see Table 2 1)
27 based on information from Geographic Information Systems (GIS) habitat maps and verified in field. Fish captured during electro fishing were processed similar to fish captured by seining with the exception that fish received a VIE mark corresponding to sampling trip, gear, and site. All fish were returned to nearshore areas within the HSU where they were caught. Abundance Estimati on For each of the species considered, r emoval estimators were used to estimate species specific abundanc e ( N ) and capture probability ( p ) at each backwater site using a maximum likelihood multinomial estimation method (Gould and Pollock 1997). The full unconditional likelihood is given as: (2 1 ) s ( N p ), and n i is a sequ ence of removals. The values for is commonly denoted with Q and the multinomial likelihood can be written as the products of two likelihoods (Gould and Pollock 1997): (2 2) (2 3) (2 4) The estimation is accomplished in a two step process where the capture probability is estimated first then the estimation of abundance is conditional on the first likelihood.
28 This model assumes that all individuals have the same probability of capture for one unit of effort, that each unit of effort is independent, and all removals of individuals are k nown (Gould and Pollock 1997). The model also assumes that the removal samples come from a closed population, which was ensured by the use of a block seine p laced across the mouth of the backwater sites prior to seining. Estimation of 95% confidence intervals (CI) of abundance was done using a parametric bootstra p technique with the FISHMETHODS package in program R ( R Development Core Team 2007 ) following the methods of Gould and Pollock (1997). Some confidence intervals could not be estimated because the inversion of the hessian matrix failed due to increase s in catch on successive passes. A small number of abundance estimates could also not be made for som e species because of failure to deplete catch In this case the cumulative catch of all seine passes within a sampling event was used in place of abundance. Total abundance of fish in backwater habitats within the study area was calculated by summing abu ndance estimates from individual backwater sites for a given trip Capture probability estimates ( p i ) for electrofishing were calculated for each habitat type, excluding backwaters, for each site and trip by dividing the VIE recaptures ( R i ) from passes two and three by marks ( M i ) given on pass one and two for each habitat type. Marks and recaptures were pooled across species for the capture probability estimates. (2 5) A nonparametric bootstrap procedure using Poptools add in in Micro soft Excel was used to estimate habitat and year specific mean p i The bootstrap procedure re sampled (1,000 iterations) with replacement the estimates of p i for each habitat type,
29 site and each year. The 95% CI of the means were approximate from the 2.5 th and 97.5 th percenti les of the bootstrap samples. Multiple HSUs of one continuous habitat type were assumed closed over the three (or five) day mark recapture (see below) The standard assumptions of no additions (births or immigrants) or deletions (de aths or emigrants) from the population, equal capture probability for each animal, and marks are not lost or overlooked, applied to each experiment (Pollock et al. 199 0). Species specific a bundance estimates were calculated by taking the mean pooled unmar ked fish for each habitat type ( U i ) each trip and dividing by habitat and year specific capture probability estimates. (2 6) The upper and lower 95% confidence intervals for abundance in each habitat type were calculated in the same manner as the abundance estimates but using the 95% upper and lower confidence intervals for the habitat and year specific capture probability estimates. Available Nearshore Habitat Nearshore aquatic habitats in this reach of the Colorado River can be temporally dynamic and subject to changes resulting from the dominant sediment or flow conditions. To account for shifts in habitat availability due to changes in stage or sediment aggradation/degrada tion, shorelines were surveye d and recla ssified during each sampling trip. Habitat maps derived from GIS were used to classify each individual HSU which was then verified in the field. The linear shoreline length of each habitat type (excluding backwater) was determined by summation of the individual HSU lengths The shoreline lengths of backwaters were estimate d during each sampling
30 occasion and summed across the entire study area. Generally shoreline habitats consisted of multiple HSUs of the same habitat type forming conti nuous stretches of one habitat This is particularly true for river right (looking downstream) which contains large distances of continuous talus and cliff. The assumption of closure for the electrofishing abundance estimation was across these distances of con tinuous habitat. Density Estimation and Habitat Selection Both habitat specific densities and selection ratios were estimated for seven species to evaluate both abundance and the availability of each habitat type. Density ( D i ) was calculated by dividing the abundance in each habitat by the linear shoreline length ( SL i ) s caled to 100 m of each habitat. The 95% CI for density were calculated in the same manner using the 95% upper and lower bounds of the abundance estimates. Densities were conside red significantly different if 95% CI did not overlap. (2 7) In order to test if habitat selection was occurring, first a chi squared test was used to examine if habitat use was different than what would be expected from random (Byers et al. 1984). This was followed by constructing species specific selection ratios for each habitat during each sampling trip following the type I methods described in Manly et al. (2002). The proportion of used resource units ( o i ) was calculated by taking the abundance in a given habitat and divided by the total abundance calculated for the study reach. The proportion of available resource units ( i ) was calculated by taking the shoreline length of each habitat and dividing by the total shoreline l ength of the study reach.
31 (2 8) Habitat is used in proportion to availability if = 1, if >1 this indicates positive habitat selection, and if <1this indicates negative selection or avoidance. Bonferroni corrected 95% CI were calcu lated around each selection ratio estimate in order to determine if selection or avoidance was occurring for a particular habitat type during a particular trip. Independence was assumed between animals (i.e. no territoriality) and all animals were assumed to have equal access to all available habitats (Thomas and Taylor 1990; Manly et al. 2002 ). The analysis wa s completed using the ADEHABITAT package in program R. Results Catch During this study, a number of both native an d nonnative fish were collecte d A total of 5 133 and 3 960 fish were captured by seines in 200 9 and 2010 respectively (Table 2 2, Table 2 3 and Table 2 4). The majority of fish captured with seines were nonnative fathead minnows Pimephales promelas comprising 52% (20 09) and 57% (2010) of the catch. Bluehead sucker Catostomus discobolus flannelmouth sucker Catostomus latipinnis and larval sucker spp. were the next most numerous in the seine catch In 2010 the proportion of plains killifish Fundulus zebrinus in the catch increa sed relative to 2009 (Table 2 3 and Table 2 4) Seining captured 149 (3% total catch) humpback chub in 2009 and 203 (5% total catch) in 2010. The electrofishing effort captured 9 611 total fish in 2009 and 9 925 in 2010, with fathead minnows being the ma jority of th e catch (70% and 65%, Table 2 5 and Table 2 6).
32 Capture Probability Estimation of capture probability for each of the habitats and gear types allowed the possibility of differences in detection by habitat to be addressed. Electrofishing capt ure probability ranged from 0 .11 to 0 .18 (Figure 2 4). The highest capture probability for 2009 was in sand ( 0 .16) and the lowest capture probability was in talus habitat ( 0 .11). During 2010, the highest capture probability was observed in cliff habitat ( 0 .18). Capture probability estimates between habitats types did not exhibit s ignificant differences (Figure 2 4). Seine capture probability estimates were highly variable across all species Habitat Availability Backwater habitat made up a small prop ortion (~1%) of the total available shoreline habita t in the study area (Table 2 7). Talus habitat (~36%) was the most common followed by debris fan (~24%), sand (~20%), and cliff (~18%) habitats. Although individual backwater sites where ephemeral, the overall availability of backwater habitat was consistently around 1% throughout the study. Temporal changes in habitat availability for ea ch habitat type were very small (Table 2 7) Abundance by Habitat Type Juvenile humpback chub showed consistent trends in abundance across habitat types during 2009 and 2010. Abundance of humpback chub was significantly higher in talus habitat compared to other available habitats and this trend was consistent across sampling trips and years (Figure 2 5 ). Generally, debris fan habitat was second in abundance followed by sand and cliff habitats. Humpback chub abundance in backwater habitats was significantly lower than the other available shoreline habitats during all of the 2009 and 2010 sampling trips
33 These patterns in abundance did not persist with other native species. As an example, juvenile bluehead sucker showed large variation in abundance both temporally and between habitat s during 2009 a nd 2010 (Figure 2 6 ). Blue head sucker abundance was low est in July 2009 across all habitats compared to other sampling trips and significant differences between habitat types were generally not observed. Abundance of juvenile bluehead sucker was higher in backwater habitat during August 2009, September 2009 (significant), and October 2010 than other habitat types Sand habitats had the next highest abundance of juvenile bluehead sucker in Jul y, August and September 2010. Patterns in j uvenile f lannelmouth sucker abund ance were mos t simi lar to juvenile bluehead sucker with occasionally higher abundance in backwaters (July 2009) and sand habitat containing higher abundance compared to other habitats (Figure 2 7 ). S peckl ed dace Rhinichthys osculus abundance was highest in debris fan and talus habitats (Table 2 8 and Table 2 9 ). This trend was consistent ac ross sampling trips and years. Abundance of speckled dace in backwater habitats was generally low, especially in 2010, with some exceptions (July and September 2009). Both small bodied and juvenile nonnative fish showed species specific trends in abundance across habitats. Juvenile r ainbow trout Oncorhynchus mykiss abundance was generally two to three times higher in talus than other habitats followed by debris fans wi th low abundance found elsewhere ( Table 2 10 and Table 2 11 ) Fathead minnow occurred in highest abundance in talus habitat, although usua lly not significantly and showed variable abundance in backwater habitats (Table 2 12 and Table 2 13 ). Refer to T ab le 2 14 and 2 15 for trends in p lains killifish abundance.
34 Density by Habitat Type Humpback chub density in backwaters was generally high compared to other habi tats but this pattern exhibited considerable temporal heterogeneity acr oss all sampling trip s (Figure 2 8 ). Backwater habitats represented a small (and ephemeral) habitat type overall in contrast to talus which had high availability, high abundance, and generally the first or second highest juvenile humpback chub density. There were consistentl y no significant dif ferences in density of juvenile humpback chub between sand, debris fa n, and cliff habitats Particularly during July, August and September of 2010, densities of humpback chub were very similar between sand, debris fan, talus, and cliff habitats (Figure 2 8). Other native species also occurred in relatively high dens ities within backwaters including juvenile bluehead sucker which consistently exhibited higher densities in backwater habitats than other habitat types (Figure 2 9 ). In con trast bluehead sucker also occurred in relatively hi gh densities in sand habitats with densities in other habitat types showing co nsiderable temporal variation Juvenile flannelmouth sucker exhibited patterns similar to juvenile bluehead sucker, with hig h density in backw ater and sand ha bitats (Figure 2 10 ). Speckled dace density was also high in backwater habitats, during 2009, but similar to what was observed in other ha bitat types during 2010 (Table 2 8 and Table 2 9 ). Speckled dace densities were hi gher in debris fan (especially 2009) and talus habitats with lower densities generally in cliff habitats. Nonnative fish densities were variable among habitat types. Rainbow trout densities were highest in talus and debris fan habitats, but also showed high variation b etween s ampling trips (Table 2 10 and Table 2 11 ). Densities of fathead minnow were highest in backwater habitats compared to other habitat types (Table 2 12 and Table 2
35 13 ). Plains killifish were infrequently sampled in 2009 prohibitin g the estimation of reliab le density estimates (Table 2 14 ). However, this species occurred in high er densities within backwater habitats relative to other available habitats during 2010 (Table 2 15 ). Habitat Selection In general, n ative fish showed positive selection for backwater habitat s and largely species specific patterns of selection for other habitat types Juvenile humpback chub demonstrated positive selection for backwater habitats, except during October 2009 ( Figure 2 11 ). Humpback chub a lso positively selected for talus habitats. Sand, debris fan, and cliff habitats were generally avoided although these trends were not as consistent. Hab itat selection trends for juvenile bluehead and flannelmouth sucker diff ered from humpback chub. W hile both species strongly selected for backwater ha bitats (Figure 2 12 and Figure 2 13 ), positive selection for sand habitats and against talus contrasted to the pattern s observed with humpback chub. Speckled dace generally showed positive selection for backwater habitats and debris fan habitats while avoiding cliff in both years (Table 2 8 and Table 2 9 ). Habitat selection patterns for nonnative species were variable and species specific. In 2009 juvenile rainbow trout habitat selection was highly va riable except for positive select ion of talus habitats (Table 2 10 ). Rainbow trout selected for debris fan and talus habitats in 2010 and demonstrated negative selection (avoidance) of sand and cliff habitats (Table 2 11 ) Fathead minnow showed positive selection for backwater habitats in 2009, and variable selection and avoidance of backwater habitats in 2010 (Table 2 12 and Table 2 13 ). Fathead minnow generally demonstrated no
36 trends in habitat selection and used habitats in proportion to availability. Plains killifish demonstrated positive selection for backwater habitats but largely no other patterns in ha bitat selection (Table 2 1 4 and Table 2 15 ). Discussion This study represents the first attempt at comparing juvenile native fish abundance betwee n backwater habitat s and other available habitat types in Grand Canyon and builds upon previous research identifying native fish habitat use of backwater s (AZGFD 1996; Hoffnagle 2000; Grams et al. 2010). Previous studies of shoreline habitat use by juveni le and subadult humpback chub demonstrated positive associations with talus and debris fan habitats (Valdez and Ryel 1995; Converse et al. 1998). Refining these earlier approaches by estimating the true abundance by habitat, considering habitat availability, and determining selections patterns, represents the next step in understanding habitat associations of native fish in the mainstem Colorado The results presented here suggest that backwater habitats are selected for by juvenile humpback chu b, but backwater habitats are small in ar ea compared to other available habitats Low abundance of juvenile humpback chub found in backwaters the limited area of this habitat and the ephemeral nature of backwaters in this reach suggests that backwaters may not be required by humpback chub for species persistence in the LCR reach of the Colorado River. Juvenile humpback chub abundance was highest in structurally complex habitats including talus which represent s around 36% of available shoreline habitat t o juven ile fish within the study area. Abundance of juvenile humpback chub in backwater habitat in comparison was generally low. This result is due to the small amount of available backwater habitat. Backwaters represented around 1% of the available sho reline
37 habitat within the study reach B ackwater habitats are ephemeral and the amount of avai lable habitat is small throughout Grand Canyon (Grams et al. 2010 ). Although it has been demonstrated that experimental high flows can increase the extent of b ackw ater habitat (Grams et al. 2010 ), it would be very unlikely to create and maintain the necessary amount of backwater habitat in order to approach the high er abundance of humpback chub in other habitats within the LCR reach. Although the tota l area of backwater habitats is low, and the contribution of humpback chub in backwater to the overall humpback chub population may be small, h igh densities of juvenile humpback chub and other native fish were observed in backwater habitats. High densities of nati ve fish suggest that backwater habitat is high quality, yet t here is much debate among ecologists as to the value of using density to infer habitat quality ( Bock and Jones 2004; Perot and Villard 2009 ). Advocates argue that research supports the correlati on between higher densities and measures of fitn ess (Bock and Jones 2004 ), often density is a reliable indicator of quality (Bock and Jones 2004; Perot and Villard 2009), and the prevalence of ecological traps (where high density leads ultimately to low fi tness) is largely unknown and with little empirical support (Battin 2004 ; Robertson and Hutto 2006). The seminal work of Van Horne (1983) elucidated possible breakdowns of the density habitat quality assumptions and cautions against using density as the sole metric of inferring habitat quality. Decoupling of the density habitat quality relationship has been attributed to social interac tions or temporally dynamic habitats (Van Horne 1983). Density dependent growth from interspecific competition is als o common and especially important for juvenile fish ( Lorenzen and Enberg 2001 ). Therefor caution should be used in interpreting high
38 densities of native fish found in backwaters as a sign of high quality habitat and other lines of inference should be considered. While juvenile humpback chub densities were highest in backwater habitats, similar densities were largely estimated across o t her available habitat types. This eve nness of density is inconsistent with previous studies in this area of the Colorado River that found the highest relative densities of humpback chub in talus and debris fan shorelines and associated with vegetative cov er ( Valdez and Ryel 1995 ; Converse et al. 1998). Increased relative densities associated with cover was hypothesized to be in response to biotic (predation) and abiotic (velocity refuges or decreased light) factors (Converse et al. 1998). Vegetative cov er was not considered in the present study due to the limited extent of such habitat in the study area (personal observation). The evenness of juvenile humpback chub densities between habitats and the differences in patterns observed in this study and pre vious research could be due to several factors. First, the previous habitat studies occurred in the early adult (> age 4) hu mpback chub populations were relatively high ( Coggins and Walters 2009 ), yet the proportion of juvenile fish (<100 mm T L) i n the mainstem Colorado River relative to the overall population is unknown, but likely lower th an during the present study because mainstem river temperatures were lower (Voicheck and Wright 2007) The possible differences in overall juvenile abundan ce between the study periods could result in differing patterns of fish density across habitats. Second, differences could be attributed to differential capture probability between habitats biasing the relative densities of Converse et al. ( 1998 ) Third, size differences between the fish considered in the present study and the earl ier investigations could result in differences with respect
39 to habitat use due to ontogenetic habitat shifts (Stone and Gorman 2006). These factors alone or in concert could pa rtially explain the differences between the present study and previously observed habitat use patterns of juvenile humpback chub. The habitat selection patterns suggest that juvenile humpback chub are not randomly distributed across shoreline habitats bu t consistently selecting for backwaters and to a lesser degree talus habitat. Juvenile chub also showed negative selection (avoidance) for cliff and sand habitats, but these patterns were not as strong or consistent. Selection for backwater habitats is p robability due to increased water temperatures in backwaters (unpublished data) Other factors such as prey resources for juvenile fish are comparable between backwaters and other habitats (Brouder et al. 1999; Behn et al. 2010 ). Differences in predation risk between habitats could also influence selection patterns (See Chapter 3). Humpback chub showed some avoidance of sand and cliff habitats. Contrasting the patterns observed for humpback chub, sucker species showed higher densities and selection for sand habitats. This could be due to species specific associations with the benthos or the availability of benthic food resources for these species. Positive selection for talus habitats and higher densities of humpback chub in talus relative to other c ommon habitat types could be explained by several factors. Heterogeneous habitats are thou ght to be important in rivers and provide a variety of benefits especially to juvenile fish (Pretty et al. 2003). Increased habitat complexity provides low velocity refuges that decrease energy loss es from swimming (Crook and Robertson 1999). For drift feeding fish hydraulic cover may also provide efficient foraging sites where faster nearby current provides food which fish da rt out and capture
40 (Fausch 1984; Allouche 2002). Structure may also serve as refuge from predation (Allouche 2002) and generally reduces foraging efficiency of p redators (Savaino and Stein 1982 ). Habitat selection is generally thought to incorporate a variety of ecological processes t hat effect habitat use decisions (Morris 2003). Individual animals are expected to seek habitats that maximize their fitness (Krebs and Kacelnik 1991) and riverine fish are thought to occupy positions relative to currents that provide a net energy gain (F ausch 1984). Habitat selection studies have been used to identify important habitat features for the conservation of native fishes in large river systems (Koehn 2009). High selection does not explicitly infer that these habitats are required however, and evaluation of the associated fitness consequences is still necessary to (Rosenfeld and Hatfield 2006). This s tudy aids in establishing habitat selection patterns, so th at future research may quantify t he habitat specific fitness measures (gr owth or survi val) of using different nearshore habitats This information would greatly benefit the management of native fish species in Grand Canyon, by providing further inference in defining habitat requirement s Removal methods are commonly applied to estimate ab undance in fisheries and wildlife studies. These methods can produce bias results if the assumptions are violated (Rosenberger and Dunham 2005) Studies from small streams, where application of multiple pass removal methods are common, have examined the bias associated with violating the assumption of constant capture probability across passes ( Peterson et al. 2004 ; Rosenberger and Dunham 2005 ). These studies have demonstrated that declining capture probability over removal passes results in an
41 overestim ation of capture probability and an underestimation of abundance ( Rosenberger and Dunham 2005 ). Increasing the number of removal passes greater than three has been suggested as a method to decrease this bias ( Riley and Fausch 1992 ). This study used a min imum of five passes in order to address this potential bias. There is the potential that some of the removal estimates of abundance in backwaters are biased low, however due to the large differences often observed in abundance between backwaters and other habitats, the underestimation would generally not affect the conclusions of this study. How do these findings help to inform policy choices related to managing for specific habitat types? The results suggesting that backwater habitats may not be requir ed for humpback chub persistence is dependent on the absence of fitness advantages to juvenile humpback chub from backwater (or other) habitat types that could lead to improved growth or survival for juveniles in these habitats. These fitness advantages c ould ultimately lead to higher contribution to the adult population and higher future recruits from a differentially smaller portion of the juveniles that used specific habitat types. Whether there are fitness benefits to juvenile humpback or other nativ e fish living in some habitats types requires further investigation, but preliminary evidence suggests that at least growth and survival is similar among habitat types based on the following. First, the study was conducted during two months (July and Augu st) each year of a fluctuating flow regime in which many backwaters within the study area were dewatered daily, subjecting fish to the same growth and survival condition as other shoreline habitats. Second, individuals marked with VIE that identified thes e fish as backwater
42 captures were often re captured in other habitats suggesting at least a moderate degree of mixing on timescales of weeks to months between backwaters and other habitats. While preliminary, these results suggest that differential fitnes s benefits did not exist among habitat types within this reach of the Colorado River and that conservation efforts designed to benefit humpback chub that target backwater habitats will only influence a small proportion of the humpback chub abundance in the LCR reach. Whether these same patterns of hab itat use and potential fitness consequences persist in areas outside of the LCR reach, is a key arena for future work in Grand Canyon.
43 Ta ble 2 1. Definitions of habitat types, modified from Converse et al. ( 1998 ) Habitat Type Definition Backwater Areas of low velocity water that is partially isolated from the main channel, often in the lee of an emergent reattachment sandbar. Sand Areas of predominantly exposed sand. Beaches can have very steep banks or be very flat. Debris Fan Debris, predominantly boulder, transported from a tributary during a flooding event. It is characterized by boulders with some degree of embeddedness, intermittent sand beaches, and a small percentage of gravel. The angle of repose is generally flatter than that of talus. The boulders are more rounded as inferred by the process of transportation. Talus Colluvium, predominantly boulder, deposited by rockfall or rockslide activity on the canyon walls. It is characteristically n ot embedded and has a steeper angle of repose than a debris fan. Debris is more angular as inferred by its process of transportation. Cliff Any rock that is in its original location and has not been transported or broken up by any means. This includes she ar walls and laterally or vertically emerging ledges.
44 Table 2 2. Common and scientific name for each taxon. Common Name Scientific Name Black Bullhead Ameiurus melas Bluehead Sucker Catostomus discobolus Black Crappie Pomoxis nigromaculatus Brown Trout Salmo trutta Channel Catfish Ictalurus punctatus Common Carp Cyprinus carpio Fathead Minnow Pimephales promelas Flannelmouth Sucker Catostomus latipinnis Green Sunfish Lepomis cyanellus Humpback Chub Gila cypha P lains Killifish Fundulus zebrinus R ainbow Trout Oncorhynchus mykiss R ed Shiner Cyprinella lutrensis S peckled Dace Rhinichthys osculus S triped Bass Morone saxatilis
45 Table 2 3 Fish species and the number captured with seines in backwater habitats each month during 2009 deple tion sampling. Proportions of total catch in each sam pling trip shown in parentheses Species July August September October Total Black Bullhead 1 (0.001) 1 (< 0 .001) Bluehead Sucker 174 (0.260) 455 (0.182) 48 (0.043) 677 ( 0 .132) Fathead Minnow 107 (0.125) 262 (0.392) 1331 (0.534) 977 (0.875) 2677 ( 0 .522) Flannelmouth Sucker 389 (0.456) 50 (0.075) 209 (0.084) 5 (0.004) 653 ( 0 .127) Humpback Chub 59 (0.069) 45 (0.067) 38 (0.015) 7 (0.006) 149 ( 0 .030) P lains Killifish 10 (0.012) 5 (0.007) 11 (0.004) 32 (0.029) 58 ( 0 .011) R ainbow Trout 2 (0.001) 2 (< 0 .001) R ed Shiner 5 (0.007) 4 (0.002) 9 ( 0 .002) S peckled Dace 58 (0.068) 48 (0.072) 192 (0.077) 24 (0.021) 322 ( 0 .062) Unknown Catostomus Spp. 229 (0.268) 80 (0.120) 250 (0.100) 21 (0.019) 580 ( 0 .113) Unknown 2 (0.001) 3 (0.003) 5 ( 0 .001) Total 853 ( 0 .166) 669 ( 0 .130) 2494 ( 0 .486) 1117 ( 0 .218) 5133
46 Table 2 4 Fish species and the number captured with seines in backwater habitats each month during 2010 depletio n sampling. Proportions of total catch in each sa mpling trip shown in parenthes es. Species July August September October Total Black Bullhead 3 (0.005) 3 (< 0 .001) Bluehead Sucker 8 (0.006) 63 (0.107) 17 (0.022) 114 (0.086) 202 ( 0 .051) Channel Catfish 1 (0.002) 1 (< 0 .001) Common Carp 2 (0.003) 2 (< 0 .001) Fathead Minnow 266 (0.206) 349 (0.592) 580 (0.764) 1046 (0.792) 2241 ( 0 .566) Flannelmouth Sucker 18 (0.014) 12 (0.020) 21 (0.028) 31 (0.023) 82 ( 0 .021) Humpback Chub 23 (0.018) 80 (0.136) 84 (0.111) 16 (0.012) 203 ( 0 .051) P lains Killifish 36 (0.028) 20 (0.034) 27 (0.036) 100 (0.076) 183 ( 0 .046) R ainbow Trout 4 (0.003) 7 (0.012) 2 (0.003) 13 ( 0 .003) R ed Shiner 1 (0.001) 2 (0.003) 1 (0.001) 4 ( 0 .001) S peckled Dace 59 (0.046) 19 (0.032) 25 (0.033) 14 (0.011) 117 ( 0 .030) S triped Bass 1 (0.001) 1 (< 0 .001) Unknown Catostomus Spp. 875 (0.678) 26 (0.044) 1 (0.001) 902 ( 0 .228) Unknown 6 (0.010) 6 ( 0 .002) Total 1290 ( 0 .326) 590 ( 0 .149) 759 ( 0 .192) 1321 ( 0 .334) 3960
47 Table 2 5 Fish species and the number captured w ith electrofishing in talus, deb ris fan, cliff, and sand habitats dur ing 2009. Proportions of total catch in each sa mpling trip shown in parenthes es Species July August September October Total Black Bullhead 16 (0.011) 4 (0.002) 5 (0.002) 2 (<0.001) 27 ( 0 .003) Bluehead Sucker 48 (0.034) 139 (0.082) 137 (0.063) 56 (0.013) 380 ( 0 .040) Black Crappie 1 (<0.001) 1 (< 0 .001) Brown Trout 1 (0.001) 3 (0.002) 4 (0.002) 3 (0.001) 11 ( 0 .001) Channel Catfish 1 (0.001) 1 (<0.001) 2 (< 0 .001) Common Carp 2 (0.001) 10 (0.006) 2 (0.001) 14 ( 0 .001) Fathead Minnow 695 (0.495) 672 (0.396) 1460 (0.668) 3939 (0.910) 6766 ( 0 .704) Flannelmouth Sucker 83 (0.059) 182 (0.107) 79 (0.036) 46 (0.011) 390 ( 0 .041) Humpback Chub 312 (0.222) 355 (0.209) 302 (0.138) 135 (0.031) 1104 ( 0 .115) P lains Killifish 3 (0.001) 3 (< 0 .001) R ainbow Trout 72 (0.051) 90 (0.053) 55 (0.025) 46 (0.011) 263 ( 0 .027) R ed Shiner 18 (0.011) 7 (0.003) 9 (0.002) 34 ( 0 .004) S peckled Dace 106 (0.076) 131 (0.077) 102 (0.047) 81 (0.019) 420 ( 0 .044) Unknown Catostomus Spp. 67 (0.048) 93 (0.055) 29 (0.013) 7 (0.002) 196 ( 0 .020) Total 1403 ( 0 .146) 1697 ( 0 .177) 2184 ( 0 .227) 4327 ( 0 .450) 9611
48 Table 2 6 Fish species and the number captured with electrofishing in talus, deb ris fan, cliff, and sand habitats dur ing 2010. Proportions of total catch in each sa mpling trip shown in parenthes es. Species July August September October Total Black Bullhead 3 (0.002) 1 (<0.001) 1 (<0.001) 2 (0.001) 7 ( 0 .001) Bluehead Sucker 22 (0.012) 125 (0.043) 135 (0.053) 69 (0.027) 351 ( 0 .035) Brown Trout 6 (0.003) 1 (<0.001) 1 (<0.001) 1 (<0.001) 9 ( 0 .001) Channel Catfish 3 (0.002) 2 (0.001) 1 (<0.001) 6 ( 0 .001) Common Carp 13 (0.007) 5 (0.002) 4 (0.002) 2 (0.001) 24 ( 0 .002) Fathead Minnow 1031 (0.552) 2032 (0.696) 1579 (0.620) 1808 (0.698) 6450 ( 0 .650) Flannelmouth Sucker 47 (0.025) 29 (0.010) 107 (0.042) 125 (0.048) 308 ( 0 .031) Green Sunfish 1 (<0.001) 1 (< 0 .001) Humpback Chub 185 (0.099) 320 (0.110) 310 (0.122) 204 (0.079) 1019 ( 0 .103) P lains Killifish 9 (0.005) 65 (0.022) 34 (0.013) 28 (0.011) 136 ( 0 .041) R ainbow Trout 221 (0.118) 98 (0.034) 110 (0.043) 131 (0.051) 560 ( 0 .056) R ed Shiner 10 (0.005) 9 (0.003) 20 (0.008) 4 (0.002) 43 ( 0 .004) S peckled Dace 309 (0.166) 177 (0.061) 243 (0.095) 216 (0.083) 945 ( 0 .095) S triped Bass 1 (0.001) 1 (< 0 .001) Unknown Catostomus Spp. 7 (0.004) 55 (0.019) 3 (0.001) 65 ( 0 .007) Total 1867 ( 0 .188) 2919 ( 0 .294) 2548 ( 0 .257) 2591 ( 0 .261) 9925
49 Table 2 7. Total available shoreline habitat sampled during 2009 and 2010. The percentage of the overall shoreline in each habitat classification is shown in parentheses. Year Sampling Month Habitat Availability (m) Backwater Cliff Debris Fan Sand Talus 2009 July 137 (1.7 %) 1493 (18.4 %) 1959 (24.1 %) 1623 (20.0 %) 2923 (35.9 %) 2009 August 106 (1.3 %) 1496 (18.4 %) 1916 (23.5 %) 1661 (20.4 %) 2967 (36.4 %) 2009 September 138 (1.7 %) 1486 (18.3 %) 1868 (23.0 %) 1677 (20.6 %) 2967 (36.5 %) 2009 October 136 (1.7 %) 1486 (18.2 %) 1813 (22.3 %) 1746 (21.4 %) 2967 (36.4 %) 2010 July 121 (1.5 %) 1492 (18.2 %) 1972 (24.0 %) 1617 (19.7 %) 3014 (36.7 %) 2010 August 106 (1.3 %) 1492 (18.1 %) 1978 (24.0 %) 1663 (20.2 %) 2994 (36.4 %) 2010 September 74 (0.9 %) 1496 (18.4 %) 1834 (22.5 %) 1754 (21.5 %) 2994 (36.7 %) 2010 October 86 (1.1 %) 1496 (18.3 %) 1834 (22.5 %) 1764 (21.6 %) 2974 (36.5 %)
50 Table 2 8 Summary of a bundance, density, and abitat selection estimates ( ) for speckled dace during 2009 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Year Month Habitat Abundance Density (fish per 100 m) Selection Ratio ( ) Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2009 July Backwater 58 57 59 42.34 41.76 42.91 7.13 5.41 8.85 Sand 44 26 54 2.69 1.6 0 3.3 0 0.45 0.32 0.57 Debris Fan 280 168 342 14.29 8.56 17.44 2.15 1.97 2.34 Talus 154 118 186 5.27 4.03 6.35 0.67 0.56 0.77 Cliff 21 7 28 1.42 0.47 1.85 0.19 0.1 0 0.28 August Backwater 56 56 56 52.83 52.83 52.83 6.87 5.16 8.59 Sand 125 74 153 7.52 4.47 9.2 0 0.96 0.81 1.11 Debris Fan 280 168 342 14.61 8.75 17.83 1.7 0 1.54 1.86 Talus 145 111 175 4.89 3.74 5.89 0.48 0.39 0.56 Cliff 113 37 147 7.57 2.51 9.82 0.77 0.62 0.92 September Backwater 113 108 118 81.88 78.05 85.72 11.12 9.27 12.96 Sand 125 74 153 7.45 4.43 9.12 0.99 0.84 1.15 Debris Fan 210 126 256 11.24 6.73 13.72 1.37 1.2 0 1.53 Talus 190 146 229 6.42 4.91 7.73 0.65 0.56 0.75 Cliff 42 14 55 2.86 0.95 3.71 0.3 0 0.2 0 0.4 0 October Backwater 12 12 12 8.82 8.82 8.82 1.61 0.71 2.51 Sand 106 63 130 6.08 3.61 7.44 1.09 0.91 1.27 Debris Fan 171 102 209 9.44 5.65 11.52 1.54 1.34 1.74 Talus 82 62 98 2.75 2.1 0 3.31 0.38 0.29 0.46 Cliff 149 49 193 10 .00 3.31 12.98 1.42 1.2 0 1.65
51 Table 2 9 Summary of a bundance, density, and abitat selection estimates ( ) f or speckled dace during 2010 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Year Month Habitat Abundance Den sity (fish per 100 m) Selection Ratio ( ) Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2010 July Backwater 47 47 47 38.84 38.52 39.17 2.93 2.11 3.75 Sand 147 99 180 9.12 6.1 0 11.12 0.69 0.58 0.79 Debris Fan 341 202 418 17.28 10.25 21.19 1.31 1.19 1.42 Talus 448 323 531 14.87 10.73 17.61 1.12 1.04 1.2 0 Cliff 105 77 124 7.03 5.18 8.33 0.53 0.43 0.63 August Backwater 17 14 20 16.04 12.89 19.18 1.74 0.92 2.56 Sand 160 107 195 9.6 0 6.43 11.71 1.04 0.9 0 1.18 Debris Fan 243 144 298 12.31 7.3 0 15.09 1.33 1.19 1.47 Talus 278 201 330 9.3 0 6.71 11.01 1.01 0.91 1.1 0 Cliff 61 45 72 4.07 3 .00 4.82 0.44 0.34 0.55 September Backwater 18 18 18 14.88 14.88 14.88 1.98 1.08 2.89 Sand 147 99 180 9.12 6.1 0 11.12 0.69 0.58 0.79 Debris Fan 285 169 350 14.46 8.58 17.73 1.27 1.14 1.39 Talus 455 328 539 15.1 0 10.89 17.87 1.24 1.16 1.32 Cliff 94 69 111 6.29 4.64 7.45 0.51 0.41 0.61 October Backwater 11 11 11 9.09 9.09 9.09 1.13 0.46 1.79 Sand 74 49 90 4.56 3.05 5.56 0.37 0.29 0.45 Debris Fan 285 169 350 14.46 8.58 17.73 1.37 1.23 1.5 0 Talus 414 299 490 13.75 9.91 16.27 1.22 1.14 1.31 Cliff 143 106 170 9.62 7.09 11.39 0.84 0.72 0.97
52 Ta ble 2 10 Summary of a bundance, density, and abitat selection estimates ( ) for rainbow trout < 100 mm TL during 2009 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Abundance Density (fish per 100 m) Selection Ratio ( ) Year Month Habitat Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2009 July Backwater 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Sand 19 11 23 1.15 0.69 1.41 0.39 0.22 0.56 Debris Fan 23 14 28 1.19 0.71 1.45 0.41 0.25 0.56 Talus 154 118 186 5.27 4.03 6.35 1.8 0 1.63 1.97 Cliff 42 14 55 2.84 0.94 3.69 0.97 0.71 1.23 August Backwater 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Sand 62 37 76 3.76 2.23 4.6 0 1 .00 0.78 1.22 Debris Fan 93 56 114 4.87 2.92 5.94 1.3 0 1.08 1.52 Talus 136 104 164 4.58 3.5 0 5.52 1.22 1.07 1.37 Cliff 14 5 18 0.95 0.31 1.23 0.25 0.12 0.38 September Backwater 4 4 4 4 .00 0.17 7.83 1.78 0.06 3.5 0 Sand 31 19 38 1.86 1.11 2.28 1.15 0.79 1.5 0 Debris Fan 16 9 19 0.83 0.5 0 1.02 0.51 0.27 0.75 Talus 82 62 98 2.75 2.1 0 3.31 1.69 1.46 1.92 Cliff 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 October Backwater 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Sand 6 4 8 0.36 0.21 0.44 1.01 0.31 1.71 Debris Fan 16 9 19 0.86 0.51 1.05 2.42 1.6 0 3.24 Talus 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Cliff 7 2 9 0.48 0.16 0.62 1.34 0.48 2.2 0
53 Table 2 11 Summary of a bundance, density, and abitat selection estimates ( ) for rainbow trout < 100 mm TL during 2010 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoi dance of a particular habitat type. Year Month Habitat Abundance Density (fish per 100 m) Selection Ratio ( ) Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2010 July Backwater 2 2 2 1.65 1.65 1.65 0.2 0 0 .00 0.48 Sand 129 86 157 7.98 5.34 9.73 0.98 0.83 1.13 Debris Fan 111 66 136 5.64 3.35 6.92 0.69 0.58 0.81 Talus 333 240 394 11.04 7.96 13.07 1.36 1.25 1.46 Cliff 94 69 111 6.29 4.64 7.45 0.77 0.63 0.92 August Backwater 7 7 7 6.6 0 6.6 0 6.6 0 2.76 0.75 4.77 Sand 18 12 22 1.11 0.74 1.35 0.46 0.26 0.66 Debris Fan 63 37 77 3.16 1.88 3.88 1.32 1.05 1.59 Talus 81 59 96 2.72 1.96 3.22 1.14 0.95 1.33 Cliff 28 20 33 1.85 1.36 2.19 0.77 0.51 1.04 September Backwater 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Sand 6 4 7 0.38 0.25 0.46 0.18 0.04 0.31 Debris Fan 42 25 51 2.12 1.26 2.59 1.15 0.85 1.44 Talus 109 78 129 3.61 2.6 0 4.27 1.83 1.63 2.02 Cliff 6 4 7 0.37 0.27 0.44 0.19 0.03 0.34 October Backwater 0 0 0 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Sand 12 8 15 0.76 0.51 0.93 0.24 0.11 0.38 Debris Fan 83 49 102 4.23 2.51 5.19 1.59 1.32 1.86 Talus 115 83 137 3.83 2.76 4.54 1.36 1.18 1.53 Cliff 22 16 26 1.48 1.09 1.75 0.52 0.31 0.72
54 Table 2 12 Summary of a bundan abitat selection estimates ( ) for fat head minnow during 2009 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Abundance Density (fish per 100 m) Selection Ratio ( ) Year Month Habitat Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2009 July Backwater 107 106 108 78.1 0 77.67 78.53 4.11 3.35 4.86 Sand 294 174 359 18.09 10.74 22.13 0.93 0.84 1.03 Debris Fan 397 237 484 20.24 12.12 24.71 0.95 0.86 1.04 Talus 571 437 688 19.54 14.94 23.54 0.77 0.71 0.83 Cliff 474 157 615 31.76 10.52 41.22 1.3 0 1.19 1.42 August Backwater 288 252 324 271.7 237.31 306.09 14.25 12.77 15.74 Sand 337 200 413 20.31 12.06 24.85 1.05 0.95 1.15 Debris Fan 342 205 418 17.85 10.69 21.8 0.84 0.75 0.92 Talus 499 381 601 16.81 12.85 20.25 0.66 0.6 0 0.72 Cliff 325 108 422 21.75 7.21 28.24 0.89 0.79 0.99 September Backwater 1023 1012 1034 741.3 733.35 749.26 21.45 20.4 22.5 Sand 375 223 459 22.36 13.28 27.35 0.64 0.58 0.7 0 Debris Fan 435 261 532 23.31 13.96 28.46 0.6 0 0.55 0.66 Talus 1015 776 1223 34.23 26.17 41.23 0.74 0.7 0 0.79 Cliff 347 115 450 23.33 7.73 30.28 0.53 0.47 0.58 October Backwater 451 433 469 331.62 318.5 0 344.73 9.36 8.57 10.16 Sand 500 297 612 28.63 17 .00 35.02 0.79 0.73 0.86 Debris Fan 614 368 750 33.89 20.29 41.37 0.86 0.79 0.92 Talus 1242 950 1496 41.87 32.01 50.44 0.89 0.84 0.93 Cliff 594 197 771 39.99 13.24 51.9 0 0.88 0.81 0.95
55 Table 2 13 Summary of a bundance, density, and abitat selection estimates ( ) for fat head minnow during 2010 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Year Month Habitat Abundance Density (fish per 100 m) Selection Ratio ( ) Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2010 July Backwater 8 8 8 6.61 6.61 6.61 0.31 0.1 0 0.52 Sand 1130 757 1378 22.79 15.26 27.8 0 1.06 0.96 1.15 Debris Fan 1606 953 1970 22.57 13.39 27.68 1.05 0.96 1.13 Talus 1915 1381 2267 21.41 15.44 25.34 0.99 0.93 1.05 Cliff 1076 794 1275 20.34 15.01 24.1 0 0.94 0.85 1.04 August Backwater 49 41 57 46.23 39.02 53.44 1.78 1.29 2.28 Sand 504 337 614 30.29 20.28 36.94 1.17 1.08 1.26 Debris Fan 508 301 623 25.67 15.22 31.48 0.99 0.92 1.06 Talus 727 524 860 24.27 17.5 0 28.73 0.94 0.88 0.99 Cliff 348 256 412 23.3 0 17.19 27.6 0 0.9 0 0.81 0.99 September Backwater 14 14 14 11.57 11.41 11.73 0.76 0.36 1.16 Sand 491 329 599 30.39 20.35 37.06 1.13 1.04 1.21 Debris Fan 452 268 554 22.92 13.6 0 28.11 0.99 0.91 1.07 Talus 801 578 949 26.59 19.18 31.48 1.08 1.02 1.13 Cliff 270 199 320 18.12 13.37 21.47 0.73 0.65 0.81 October Backwater 111 111 111 91.74 91.74 91.74 4.51 3.69 5.32 Sand 448 300 547 27.73 18.57 33.82 0.89 0.81 0.96 Debris Fan 487 289 597 24.69 14.64 30.27 0.93 0.85 1 .00 Talus 903 651 1069 29.97 21.62 35.48 1.06 1.01 1.11 Cliff 386 285 458 25.89 19.1 0 30.67 0.9 0 0.82 0.98
56 Table 2 14 Summary of a bundance, dens abitat selection estimates ( ) for plains killifish du ring 2009 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Abundance Density (fish per 100 m) Selection Ratio ( ) Year Month Habitat Value Lower 95% Upper 95% Value Lower 95% Upper 95% Value Lower 95% Upper 95% 2009 July Backwater 10 9 11 7.3 0 6.73 7.87 76.85 76.85 76.85 Sand 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Debris Fan 0 0 0 0 0.0 0.00 0.00 0.00 0.00 0.00 Talus 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Cliff 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 August Backwater 4 4 4 3.77 3.59 3.96 76.85 76.85 76.85 Sand 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Debris Fan 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Talus 0 0 0 0.00 0.00 0.00 0 0.0 0.00 0.00 Cliff 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 September Backwater 13 13 13 9.42 9.42 9.42 58.95 58.95 58.95 Sand 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Debris Fan 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Talus 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Cliff 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 October Backwater 8 8 8 5.88 5.74 6.03 22.9 0 10.43 35.37 Sand 6 4 8 0.36 0.21 0.44 1.37 0.46 2.28 Debris Fan 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 Talus 9 7 11 0.31 0.23 0.37 0.89 0.34 1.44 Cliff 0 0 0 0.00 0.00 0.00 0.00 0.00 0 0.0
57 Table 2 15 Summary of a bundance, density, and abitat selection estimates ( ) for plains killifish du ring 2010 Selection ratios greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type. Year Month Habitat Abundance Density (fish per 100 m) Selection Ratio ( ) Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI Value Lower 95% CI Upper 95% CI 2010 July Backwater 22 22 22 18.18 18.02 18.34 27.09 18.31 35.86 Sand 12 8 15 0.76 0.51 0.93 1.13 0.57 1.69 Debris Fan 21 12 26 1.06 0.63 1.3 0 1.58 1.04 2.11 Talus 0 0 0 0.00 0 0.0 0.00 0.00 0.00 0.00 Cliff 0 0 0 0.00 0.00 0.00 0.00 0.00 0.00 August Backwater 3 3 3 2.83 2.83 2.83 0.89 0.00 1.9 0 Sand 98 66 120 5.91 3.96 7.21 1.87 1.57 2.16 Debris Fan 42 25 51 2.11 1.25 2.59 0.67 0.48 0.85 Talus 68 49 80 2.27 1.64 2.69 0.72 0.57 0.86 Cliff 50 37 59 3.33 2.46 3.94 1.05 0.79 1.31 September Backwater 11 11 11 0.68 0.68 0.68 6.44 2.75 10.13 Sand 92 62 112 5.7 0 3.82 6.95 2.28 1.94 2.61 Debris Fan 7 4 9 0.35 0.21 0.43 0.16 0.04 0.28 Talus 34 24 40 1.13 0.81 1.33 0.49 0.34 0.64 Cliff 44 33 52 2.96 2.18 3.51 1.28 0.95 1.61 October Backwater 65 65 65 53.72 53.4 54.04 31.45 25.19 37.7 Sand 12 8 15 0.76 0.51 0.93 0.29 0.13 0.45 Debris Fan 49 29 60 2.47 1.46 3.03 1.1 0 0.84 1.37 Talus 20 15 24 0.68 0.49 0.8 0 0.29 0.17 0.4 0 Cliff 50 37 59 3.33 2.46 3.94 1.38 1.05 1.71
58 Figure 2 1. Map showing the Colorado River in Marble and Grand Canyons, Arizona. The study area is located just downstream of the confluence of the Colorado and Little Colorado River. Numbers indicate km downstream of Glen Canyon Dam
59 Figure 2 2. Figure showing the three sampling sites used for the electrofishing mark recapture. Each site is approximately 1500 m of shoreline on each site of the river. Note the proximity of the sampling sites to the Little Colorado River.
60 Figure 2 3. Overhead aerial photo of the Colorado River within the sampling area. The photo is overlain with a GIS l ayer demarcating the 50 Unit s s ). Each of these HSUs was discreetly classified as talus, debris fan, sand, cliff or backwater habitat and this habitat classification was verified in the field. Note the river is flowing from the t op of the figure to the bottom.
61 Figure 2 4. Capture probability of small bodied fish (40 99 mm TL, all species) sampled using boat electrofish ing stratified by habitat type. Electrofishing effectively sampled sand, debris fan, talus, and cl iff habitats. See Table 2 1 for definitions of the habitat types. The bars represent approximate 95% CI for the capture probability estimates calculated using a bootstrapping technique. Capture probability from electrofishing sampling during both 2009 and 2010 is shown.
62 Figure 2 5 Juven ile humpback chub (40 99 mm TL) abundance estimates by habitat type during July August, September, and October of 2009 (A) and 2010 ( B). Bars represent approximate 95% confidence intervals.
63 Figure 2 6 J uvenile bluehead sucker (40 149 mm TL) abundance estimates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B). Bars represent approximate 95% confidence intervals.
64 Figure 2 7 Juvenile fannelmouth sucker (40 149 mm TL) abundance estimates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B). Bars represent approximate 95% confidence intervals.
65 Figure 2 8 Juvenile humpback chub (40 99 mm TL) density estimates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B). Bars represent approximate 95% confidence intervals.
66 Figure 2 9 Juvenile bluehead sucker (40 149 mm TL) density estimates by habitat type during July, August, September, a nd October of 2009 (A) and 2010 (B). Bars represent approximate 95% confidence intervals. The secondary y axis on the right of each panel corresponds to the density of fish in backwater habitats.
67 Figure 2 10 Juvenile flannelmouth sucker (40 149 mm TL) density estimates by habitat type during July, August, September, and October of 2009 (A) and 2010 (B). Bars represent approximate 95% confidence intervals. The secondary y axis on the right of each panel corresponds to the density of fish in bac kwater habitats.
68 Figure 2 11 abitat selection estimates ( ) for juvenile humpback chub (40 99 mm TL) during 2009 (A) and 2010 (B) The legend refers to the sampling trip and the x axis refers to the habitat type. Values greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type.
69 Figure 2 12 type I h abitat selection estimates ( ) for juvenile bluehead sucker (40 149 mm TL) during 2009 (A) and 2010 (B) The legend refers to the sampling trip and the x axis refers to the habitat type. Values greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type.
70 Fig ure 2 13 abitat selection estimates ( ) for juvenile flannelmouth sucker (40 149 mm TL) during 2009 (A) and 2010 (B) The leg end refers to the sampling trip and the x axis refers to the habitat type. Values greater than 1 represent positive selection for a habitat type and values less than 1 represent avoidance of a particular habitat type
71 CHAPTER 3 DIFFERENTIAL PREDATI ON RISK: ASSESSING THE ROLE OF HABITAT AND TURBIDITY IN A LARGE RIVER ECOSYSTEM The importance of predator prey relationships in structuring biotic communities was identified by early ecologists, such as Aldo Leopold in Game Management (1933). During the past seven decades, predation has been recognized as an important element structuring aquatic communities (Sih et al. 1985; Estes 1996; Carpenter and Kitchell 1993 ) Predation can have direct and indirect impacts on prey populations by alteri ng distributions or habitat use (Townsend and Crowl 1991), and causing behavioral responses (Lima 1998) that frequently result in decreased growth and survival. Altered or degraded environmental conditions can facilitate the establishment of nonnative spe cies (Lozon and MacIsaac 1997; Meador et al. 2003 ; Johnson et al. 2008 ). Introductions of predatory species are of particular concern because the effects on novel prey can be disproportionally large (Townsend and Crowl 1991; Blinn et al. 1993). Along wit h habitat degradation (Alla n and Flecker 1993; Dudgeon et al. 2006), nonnative species are commonly implicated as a leading factor in the decline of native species (Olden and Poff 2005 ; Coggins et al. 2011 ; Cucherousset and Olden 2011 ; Yard et al. 2011 ). Concerns over the decline and extirpation of native fishes have prompted research and management efforts aimed at recovering these populations and reducing extinc tion risks in the United States. T hese actions are often in response to legislative requirem ents under the Endangered Species Act (ESA) in the U.S. by restoring habitat deemed essential or mitigating the negative impacts of nonnative species. Restoration strategies designed to benefit native riverine species include restoring habitat (Roni et
72 al 2008), historical flow conditions ( Poff et al. 1 997; Bunn and Arthingt on 2002), and removals of nonnative species (Tyus and Saunders 2000; Coggins et al. 2011 ). These efforts are often implemented with little monitoring of responses (Bernhardt et al. 20 05) which makes attributing the biological responses to management or environmental factors difficult. Few studies have compared management alternatives (e.g., flow restoration, nonnative removal some joint combination ) to determine whic h approaches are most effective. However, recently Marks et al. ( 2010) has shown larger increases in native fish abundance to nonnative control efforts than to flow restoration and other long term research suggests that flow management singularly may not illicit desired native fish population responses ( Bradford et al. 2011) Thus identifying what factors ( i.e., biotic interactions, habitat limitation, temperature tolerance) limit population recovery is of primary interest to management agencies. The Grand Canyon reac h of the Colorado River downs tream from GCD has been highly modified as a result of river regulation and since the early 1970s the physical and biological resources of the Colorado River have been studied to ascertain effects on the ecosystem from dam oper ations. Currently the GCDAMP seeks to understand how dam operations affect downstream physical, cultural, and biological resources using large scale ecosystem experiments. The altered physical conditions downstream of GCD include a modified flow regime, cold stenothermic temperature regime (9.4 o C mean from 1988 2005) and decreased sediment supply (Topping et al. 2005; Voicheck and Wright 2007 ). These changes to the physical environment facilitated the establishment of abundant nonnative species, parti cularly rainbow trout that have in recent years numerically dominated the fish assemblage (Gloss and Coggins 2005 ).
73 Predation b y both rainbow and brown trout Salmo trutta has been documented in Grand Canyon (Valdez and Ryel 1995; Marsh and Douglas 1997) with differential predation on native fish spe cies revealed in a study of piscivory (Yard et al. 2011). Nonnative fish have been implicated in the decline of native fishes across the Southwest ern U.S. (Minckley 1991; Tyus and Saunders 2000). Across th e C olorado River basin declines in native fishes has been observed ( Minckley 1991) with this decline often attributed to changes in the physical environment resulting from the development of water infrastructure ( Van Steeter and Pitlick 1998 ; Osmundson et al. 2002 ), or interactions with introduced species (Tyus and Saunders 2000; Olden et al. 2006). This decline is exempli fied in the Grand Canyon reach of the Colorado River whe fishes have been extirpated with the exception of the federally endan gered humpback chub (Minckley et al. 2003). The largest remaining population of humpback chub is found near the confluence of an unregulated tributary, the Little Colorado River (Gorman and Stone 1999). This population is thought to be limited by recruitment of 1 3 yr old fish to the adult population (Coggins et al. 2006). Juvenile humpback chub are known to use nearshore areas including backwater habitats (Valdez and Ryel 1995; Robinson et al. 1998; Stone and Gorman 2006). These habita ts are hypothesized to be important to juvenile native fish due to warmer water temperatures ( Hoffnagle 1996 Trammell e t al. 2002) than the mainstem Colorado River (Endangered under the ESA) and ESA mandated protecti on of critical habitats backwater habitats are routinely sampled in the mainstem Colorado River during research and monitoring expeditions (Valdez a nd Ryel 1995 ; AZGFD 1996 ; Grams et al
74 2010 ) to assess the status and trends of native fish populations ( Hoffnagle 2000 ; Grams et al. 2010). The physical processes associated with sandbars and the creation of these habitats has been extensively studied (R ubin et al. 1990; Topping et al. 2005) and it is well demonstrated that experimental high flow events can increase the area of backwater habitats (Grams et al. 2010). However the relationship between backwater habitats and humpback chub population respons es in not understood. Glen Canyon Dam has decreased downstream sediment supply significantly from pre dam conditions ( Topping et al. 2000 ), ye t two tributaries, the Paria (0.32 k m) and the Little Colorado River (98.5 km as measured downstream from Lees Fe rry, Arizona ) contribute sediment to the mainstem river in the Glen and Grand Canyon reaches and these tributaries can significantly affect turbidity levels in the mainstem river (Topp ing et al. 2005). Turbidity is thought to reduce foraging success of pr edators and thus decrease the direct and indirect effects of predation ( Barrett et al. 1992; Carter et al. 2010) A variety of predator and prey responses to increased turbidity have been observed, particularly for piscivorous species ( Gregory and Levings 1996 ; Yard et al 2011). These results are likely related to species specific responses of both predator and prey to turbidity (Gregory 1993; Abrah ams and Kattenfeld 1997). T urbidity as a form of cover has been suggested by several authors ( Gregory 1993; Allouche 2002 ) as a mechanism for how predation risk for juvenile fish may change under varying turbidity levels A ltered sediment and flow regimes have also likely changed shoreline habitat conditions in the post dam Grand Canyon Post dam sand d eposits are less extensive and structurally complex than historical conditions (Schmidt and Rubin 1995). This
75 suggests decreased availability of habitats such as backwaters that are created by sand deposits ( Converse et al. 1998). Backwaters often contai n homogeneous sand substrates and compared to other more structurally complex habitats would be expected to have higher predation risk (Warfe and Barmuta 2004; Camp et al. 2011 ). More complex habitat types, such as debris fans, with heterogeneous substrat es are thought to provide refugia from predation and would be expected to have lower predation risk (Ostrand et al. 2004; Gadomski and Parsley 2005). Despite the presence of humpback chub in backwater habitats and the perceived importance of this habi tat type, there is not clear indication that expansion or contraction of backwater habitats through time has led to a measurable chang e in humpback chub populations Key questions remain regarding the biological importance of backwaters as juvenile fish h abitat and the negative impacts nonnative fish have in these and other nearshore areas. Structurally simple habitats such as backwaters often have high predation risk (Ostrand et al. 2004; Gadomski and Parsley 2005). Also, p iscivorous fish are often foun d to co occur in backwater ha bitats with juvenile native fish in Grand Canyon (Grams et al. 2010). Understanding predation risk will inform managers of potential biotic interactions structuring habitat use. This type of investigation will aid in defining the continued efficacy of high flow experiments to benefit native fish via the creation of backwater habitat T urbidity influences predator prey interactions in aquatic ecosystems and because of the variability in turbidity conditions within this reach of the C olorado River, it is necessary to examine predation risk across a range of turbidity conditions. The objective of this study was to assess the relative predation risk between a backwater habitat, and a debris fan habitat across a range of turbidit ies.
76 Methods Study Site The study site was at Colorado River kilometer 104.1 measured downstream from is below the LCR and was chosen because the largest aggregation of humpback chub is found in proximity of the LCR (Gorman and Stone 1999). Juvenile humpback chub disperse out of the LCR ( Valdez and Ryel 1995, Robinson et al. 1998 ) into the ma instem Colorado River where non native predators are more abundant, making this area important to study preda tion risk This ar ea is key to many management actions directed at improving humpback chub growth and survival including a large nonnative fish removal prog ram that took place between 2003 and 2006 that removed close to 20,000 rainbow trout (Coggins et al. 2011 ) Tethering Trials Tethering experiments have been widely used to test hypotheses about predation risk in marine ecosystems (McIvor and Odum 1988, Adams et al. 2004), riverine ecosystems (White and Harvey 2001 ; Camp et al. 2011 ), and to examine the effects of turbidit y ( Laplante Albert et al. 2010) on predation risk. Tethering trials were conducted to assess predation risk in different habitats (backwater and debris fan) during September and October 2010. These trials were conducted over the wide range of turbidity l evels typical of this section of river. Turbidity was monitored continuously using acoustic Doppler profilers at U. S. Geological Sur vey gaging station 09402500 (141 km (Topping et al. 2004). These data were adju sted to account for water travel time by back calculation using a flow model for Grand Canyon ( Wiele and Griffin 1997 ).
77 Nonnative fathead minnows were used as surrogates for native fish throughout all trials and were collected either by seining nearshor e areas or boat electrofishing and held in a net pen until sufficient numbers were captured. Fathead minnows between 40 70 mm total length (TL) were used for the experiment. Fathead minnows were tethered with 0 .5 m of braided fishing line (4.5 kg test, 0 .15 mm diameter) passed through the ca ud al peduncle dorsa l of the spine and secured ventrally with a uni knot (see Camp et al. 2011) Tethered fish were attached to 6.4 mm diameter lead core line using small metal clip s spaced 1 m apart Ea ch ~12 m lead line was anchored with mesh rock bags and attached to shore. Tethering trials consisted of 10 tethered fathead minnows placed in each habitat (backwater or debris fan) and left undisturbed for 1 2 hours. The experiment al design consisted of a habitat trea tment with two levels (backwater and debris fan) and a turbidity treatment with three levels (low, intermediate, high). Each treatment and level (i.e., backwater and low turbidity) received 23 to 26 replicate trials. The debris fan site was characterized by mixed substrate of embedded small to large boulders with intermittent sand, slow surface velocity and a 75% bank slope. The backwater site had a smooth sand bottom substrate, near zero surface velocity, and a very shallow slope (<5%). The tethered f ish were placed at about 75 cm and 90 cm average depths in debris fan and backwater sites respectively. At the conclusion of each trial, fish were ascribed to one of two cat e gories; predation event occurred ( fi sh missing from tether or evidence of predati on) or no predation event occurred ( fish found on tether) The use of tethering experiments in ecology has received criticism (Peterson and Black 1994) a nd much subsequen t debate (Aronson and Heck 1995; Aronson et al.
78 2001) over the potential for method ological artifacts to bias possible inferences. Tethering inhibits prey from escaping and when applied in the same manner across experimental treatments yields a relative, not absolute measure of predation. The potential for bias arises when experimental treatments interact differently with methodological artifacts (Peterson and Black 1994). For example, if prey species tethered in seagrass beds tangle more readily compared with prey in bare substrates and produce erroneously high (or low) mortality rate s due to tangling. In order to address this potential bias, an exclosure was used as a control in both habitats to estimate the attrition rate of tethered fish absent predation. The tethering methods used were identical to the methods in the other experi mental treatments. Analysis The binomial likelihood function and the numerical optimization tool Solver (Microsoft Excel) were used to calculate maximum likelihood estimates of tethered fish mortality. Model selection using A kaike information criteria (AIC) was calculated to compare alternative models of tethered fish mortality (predation risk) by site and t urbidity. AIC scores were calculated by maximum likelihood estimation of the bionomial likelihood function ( n = number of trials, y = number of mo rtalities, m = probability of mortality). (3 1) The most parsimonious model was chosen and m solved for using maximum likelihood estimation. A Monte Carlo analysis (1,000 iterations) was conducted to estimate 95 percent upper (95% UL) and lower (95% LL) confi dence limits. Pairwise
79 Wilcox rank sum tests with a Bonferroni correction were used to test for significant differences (significance level = 0 .05) between m estimates. Results A total of 145 trials were conducted during the experiment. The average turbidity for each sampling period is summarized in Table 3 1 The likelihood framework allowed AIC comparisons of alternative models allowing the probability of a predation event to vary between habitats and turbidity. The highest weighted model allowed the probability of a predation event to vary by both turbidity and habitat (Table 3 2). The percent mortality of tethered fish estimated for each treatment ranged from 5% (backwater, low turbidity) to 73% (debris fan, intermediate turbidity) (Figure 3 1). The highest per cent mortality occurred during intermediate levels of turbidity with 73 .0% at the debris fan site and 45.7% at and backwater site During low turbidity conditio ns the mortality of tethered fish decreased to 5% at the backwater site and 35% at the debris fan site Under high turbidity conditions m ortality was 11.7% at the ba ckwater site and 7% at the debris fan sit e Significant differences were observed b etween habitat types during low and intermediate turbidity conditions with debris fan habitat having higher percent mortality (Figure 3 1). The pairwise Wilcox rank sum te sts confirmed significant differences (significance level = 0 .05) between several of the experimental treatments (Table 3 3). Based on the results from the predator exclosures, it is unlikely that any potential bias was introduced from methodological art ifacts. T he incidence of false positives (mortality events) was zero at the backwater site (0% attrition in 50 trials) and very low the debris fan site (0.0086% attrition in 23 trials) These levels of attrition were very low and no further analyses for this data were considered.
80 Discussion Predation risk was significantly lower in backwater than debris fan habitats during low and intermediate turbidity conditions. This result was unexpected because generally predation risk decreases with increasing habitat complexity (Ostrand et al. 2004; Warfe and Barmuta 2004; Gadomski and Parsley 2005; Camp et al. 2011 ). Loss of visual detection is often cited as a mechanism responsible for the behavioral changes seen in both predators and prey ( Crowde r and Cooper 1982 ). Shifts in foraging strategies (Savino and Stein 1982) and prey selectivity (Carter et al. 2010) are common responses of predators to increased complexity. Behavioral responses of prey species include reduced antipredator behavior in s tructurally complex habitats (Miner and Stein 1996). The higher predation risk in debris fan habitats could be explained by the higher densities and moderate selection of these areas by rainbow trout (See chapter 2). The structure that provides potential refugia for vulnerable sizes of fish may also provide preferred foraging habitat for larger fishes. This study indicates that predation risk in both habitats is strongly influenced by turbidity with the highest risk at intermediate levels of turbidity. Increased turbidity affects fish foraging success by modifying the reaction distance and encounter rate (Abrahams and Kattenfeld 1997; Turesson and Bronmark 2007). Most studies show reaction distance decreasing nonlinearly with increasing turbidity (Mine r and Stein 1996; Sweka and Hartman 2001; Sweka and Hartman 2003). Therefore, foraging success (and in turn predation risk) generally decreases with increasing turbidity for many species (Carter et al. 2010), although small or no change in foraging succes s has also been observed in experiments (Gregory and Levings 1996). These results agree with a
81 recent study that observed higher trout piscivory in this section of the Colorado River under increased turbidity (Yard et al. 2011). So who are the most likely predators? Adult rainbow trout and humpback chub are likely the species responsible for the majority of pi scivory in this river reach Both species exhibit piscivo ry as adults (Stone and Gorman 2006; Yard et al. 2011 ) and are present in this study reach B rown trout, which can be highly piscivorous (Crowl et al. 1992; McDowall 2006) and channel catfish Ictalurus punctatus are also present in this reach, but these species are not locally common compared to adult humpback chub and rainbow trout. Addition ally, s patial segregation of rainbow and brown trout has also been proposed based on disparities between interspecific electofishing catch, suggesting rainbow trout generally oc cupy nearshore areas under turbid conditions ( Speas et al. 2004 ). The overl ap of effective feeding breadth between rainbow trout and humpback chub is one explanation for the increase in predation risk at intermediate turbidities. Rainbow trout may shift feeding strategies during times of increased turbidity from drift feeding to active searching behavior ( Al Shaw and Richardson 2001). This response to increased turbidity has also been shown in brook trout (Sweka and Hartman 2001). Humpback chub evolved in a highly turbid but seasonally variable riverine environment that histori cally transported 60 million tons of sediment yearly past Lees Ferry, Arizona (Topping et al. 2000) To survive in this environment humpback chub have evolved the capacity to effectively forage during pe riods of high turbidity. Adult humpback chub have been documented to be piscivorous and move inshore at night to feed (Stone and Gorman 2006). Additionally, adults are thought to be negatively phototaxic (Stone and
82 Gorman 2006). In 2010, a shift to nearshore and backwater areas during increased tur b idity was observed during a sonic telemetry study of sub adult humpback chub ( Gerig 2012 ). Prey responses such as decreased antipredator behavior during increased turbidity could also help to explain the higher predation risk ( Gr egory 1993 ; Abrahams and Kattenfeld 1997). The patterns observed in predation risk in Grand Cayon between habitats and turbidities would likely change if the nonnative predator community changed. One species of particular concern which is found in the upper Colorado River basin and is thought to have negative interactions with native Colorado River species is s mallmouth bass Micropterus dolomieu (Johnson et al 2008) Smallmouth bass and other centrarchids could possibly be attracted to backwater habitats because of abundant pre y lower velocities, and warmer water temperatures. Monitoring of nonnative species in the LCR reach of the Colorado River and invasion risks to this reach from tributary and mainstem inputs should be considered in long term conservation planning for hump back chub What insights does this work provide to helping recover humpback chub and inform management of the Colorado River in Grand Canyon? Currently, management of sand and sediment resources is a high priority within the GCDAMP given the importance o f sand to protect archeological sites and provide camping sites for visitors. Sand laden sediment is managed through experimental floods to create sandbars and backwater habitats and since 1996 three experimental high flow releases from GCD have been impl emented to restore elements of the historical flow regime and study sediment dynamics ( Grams et al. 2010; Melis 2011 ) It is unclear whether managing the
83 Colorado River to protect or enhance backwater habitats would have a positive benefit on humpback chu b populations. By choosing to manage for backwater habitat this would imply that these areas are required for humpback chub persistence and are areas that provide positive growth and favorable survival rates to small bodied fish. This assumption still re quires further investigation to validate. As with many river restoration efforts, much of the impetus for flow restoration is directed at the creation of historic habitat conditions that are thought to illicit positive biotic responses, particularly from native species. Implicit to the justification of such habitat manipulations is a positive relationship between habitat and population size (Rosenfeld and Hatfield 2006), which to date has not been demonstrated in Grand Canyon for backwater habitats Com mon approaches to identifying habitat abundance relationships includes id entifying habitat selection p atterns (see Chapter 2 ) and then quantifying the associated fitness consequences (growth, survival, fecundity) of the habitat use ( Railsback et al. 2003 ; Rosenfeld and Hatfield 2006). Unfortunately these metrics are often difficult to assess and surrogates such as presence or abundance metrics (density, catch per unit effort) are used, which can be misleading (Van Horne 1983 ; Garshelis 2000) The pattern s of predation risk observed in this study, while not providing direct met rics of habitat quality, can inform future management decisions or experiments. Until recently, there has been a failure to critically evaluate many river restoration projects ( Be rnhardt et al. 2005 ) and in some circumstances the hypothesis (P almer et al. 1997 ) continues to guide restoration. Riverine ecosystems are complex and examples of counterintuitive responses to management
84 actions are abundant ( P ine et al. 2009 ) often because of unpredicted responses in juvenile survival due to competition or predation. Although investigating predation risk between habitats and across environmental conditions does not yield demographic rates, inferences can be dr awn in evaluating management actions (i.e. high flows to increase backwater area) and ascertaining mechanisms responsible for counterintuitive responses. The results from tethering experiments such as this can provide managers with informatio n for making informed decisions concerning habitat manipulations or other management actions Despite the limited scope of the study, this information can be useful in understanding biotic interactions in habitats that are the focus of management actions. Although the type of fine scale interactions examined in this study are not commonly a component of evaluating biotic responses to management actions, this learning can be used to formulate hypotheses about the mechanisms responsible for biotic responses to managem environmental flow restoration grows, evaluation of policy actions such as experimental high flows is essential. A unique opp ortunity exists in areas of natural resource management tha t have committed to active adaptive management such as Grand Canyon where learning from studies such as this can be directly considered in guiding future management.
85 Table 3 1. Turbidity shown in nephelometric turbidity units (NTU) during each of the experimental trials. Category Dates Mean SD Minimum Maximum Low 9/8 9/10/2010 16 6 7 34 Intermediate 10/22 10/24/2010 146 207 27 818 High 9/11 9/13/2010 1426 978 509 4830
86 Table 3 2. Akaike Information Criter ia (AIC) model weighting of habitat and turbidity heterogeneit y of tethered fish mortality. M(habitat*turbidity) = mortality differed by habitat and turbidity M(turbidity) = mor tali ty differed by turbidity, M(habitat) = mortality differed by habitat, M(.) = constant mortality across habitat and turbidity Model LL K AIC W i M(habitat *turbidity) 291 6 0 1 M(turbidity) 350 3 113 0 M(habitat ) 479 2 369 0 M(.) 508 1 425 0
87 Table 3 3. Pairwise comparisons of Wilcox rank sum tests with Bonferroni correction applied (significance level = 0 .05). Values in the table can be interpreted like p values. Experimental Treatment Backwater High Backwater Low Backwater Intermediate Debris Fan High Debris Fan Low Backwater Low 1 Backwater Intermediate <0.001 <0.001 Debris Fan High 1 1 <0.001 Debris Fan Low 0.011 <0.001 1 <0.001 Debris Fan Intermediate <0.001 <0.001 0.003 <0.001 <0.001
88 Figure 3 1 Predation risk expressed as percent mortality of tethered fish in each experimental treatment Upper and lower 95% confidence intervals are shown.
89 CHAPTER 4 CONCLUSIONS Examining abundance, density and habitat selection patterns of small bodied fish using nearshore habitats was the focus of chapter two. This chapter established that abundance of native fish was generally low in backwater habitats compared of other available habitats within the study area. This chapter also demonstrated high densities of both nativ e and nonnative species occur in backwater habitats. Native species including humpback chub demonstrated positive selection for backwater habitats, although other patterns of habitat selection were often specie s specific. Predation risk in backwaters and debris fan habitat across a range in turbidity was examined in chapter three. Predation risk was higher in debris fan compared to backwater habitats under low and intermediate levels of turbidity. During high turbidity co nditions the predation risk in both habitats decreased and was not significantly different. These results suggest that management for a certain habitat type such as backwaters or debris fans should consider the role of turbidity in mediating predation r isk. The inferences from both chapters will help to inform management of habitats in the Grand Canyon reach of the Colorado River. The c omparisons of abund ance across habitats should in particular have bearing on the direction of subsequent high flow ex periments Previous inquiry into backwater habitats prompted experimental high flows designed to increase the proportion of backwater habitat (Melis 2011). In light of the abundance trends in chapter two experimental prescriptions that seek to increase the growth and survival of juvenile humpback chub should target the habitats with the largest abundances of fish, primarily talus. Flow experiments options are limited in
90 these habitats, considering that the competence of high flows to alter the colluvium derived substrates is outside the operational regime of GCD. Additionally, juvenile fish in high angle habitats may not be affected by changes in stage as much as in low angle habitats (Converse 1998; Korman and Campana 2009). The distinct geomorphic st ructure of Grand Canyon ( Schmidt and Graf 1990 ) means that fish responses to flows may be dependent on geomorphic reach (Korman et al. 2004). The examination of abundance, density and selection patterns in other areas of the Colorado River may help to str engthen the inferences presented here in relation to management actions that effect the entire length of Grand Canyon. Combining gears and sampling methodologies to study small bodied fish in large river systems has been advo cated The information pres ented in chapter two demonstrates that by combining gears (seines, electrofishing) and sampling methodologies (depletion and capture recapture) for juvenile and small bodied fish, metrics such as density and selection ratios can be applied. Designing smal l scale targeted experiments, such as the tethering experiment in chapter three, also provides inference into specific questions. Over time these techniques within a habitat stratified sampling regime will provide essential information for identifying res ponses to management actions that may differ between habitats or physical condition s (turbidities). This will be critical information especially in evaluating future management actions such as experimental high flows or the efficacy of turbidity measures designed to disadvantage nonnative species. As an example, e valuation of the physical (backwater area, volume) responses to a 2008 high flow experiment (HFE) was very quantitative, demonstrating short term
91 increases, but catch of fish occupying backwater s was insufficient to demonstrate any response or compare backwaters to other habitat features (Grams et al. 2010). The biological results related to native fish in backwaters from the 2008 HFE demonstrated that native fish occur in backwaters (Grams et al. 2010), a pattern that has been in informing policy choices related to managing for specific habitat types Such a study would have benefited greatly from addressing capture probabili ty combining sampling methodologies and estimating metrics such as abundance or density especially pre and post flood in backwaters and other nearshore areas. This thesis prese nts a sampling and analytical framework f or examining the response of small b odied fish to management actions or changes in environmental conditions. Management efforts focused on the preservation and restoration of altered habitats is a common theme in modern fish eries management. If using habitat management approaches to illicit positive responses in target species is going to be successful in river resto ration, approaches to estimate abundance or density by habitat type and accurately measure responses of the target species to habitat manipulations is essential. Juvenile fish are most likely more sensitive than adults to alterations in biotic and abiotic conditions and thus monitoring juvenile fish responses to management actions would provide important information to managers. The approaches presented here should benefi t not only managers of the Colorado River in Grand Canyon but other restoration efforts targeting juvenile fish in large riverine ecosystems. Considering the dynamic nature of riverine species and river systems, caut ion should be used with temporal or spatial extrapolation of the density and selection results
92 presented in chapter two Juvenile fish habitat use is often mediated by predation risk (Werner and Hall 1988; Harvey and Stewart 1991; Walters and Juanes 1993) and changes in use and thus selection could result from changes in predator populations. This study was conducted during a time of relatively low rainbow trout abundance (GCMRC, unpublished data), a species which has the largest cumulative predation effe ct s on native fish in this area of the Colorado River (Yard et al. 2011). The study period also coincided with warm water releases from GCD (GCMRC, unpublished data). Warm water species, such as humpback chub, may respond to colder temperatures by incre asing selection for backwaters, especially if the gradient of backwater warming is increased. Habitat selection patterns may change across a range of predator abundances and abiotic conditions, therefore management actions such as experimental high flows may lead to different responses under these conditions. Continued monitoring will determine how robust habitat selection and density patterns are to a range of physical and biological conditions. Currently rainbow trout abundance is increasing and decreas ed water temperatures are forecasted for GCD releases, making 2012 an important year to test such hypotheses. These efforts coupled with evaluating habitat specific growth or survival rates would greatly improve native fish conservation efforts within the Colorado River basin.
93 LIST OF REFERENCES Abrahams, M. V., and M. G. Kattenfeld. 1997. The role of turbidity as a constraint on predator prey interactions in aquatic environments. Behavioral Ecology and Sociobiology 40(3):169 174. Adams, A. J., J. V. L ocascio, and B. D. Robbins. 2004. Microhabitat use by a post settlement stage estuarine fish: Evidence from relative abundance and predation among habitats. Journal of Experimental Marine Biology and Ecology 299(1):17 33. Al Shaw, E., and J. S. Richardson 2001. Direct and indirect effects of sediment pulse duration on stream invertebrate assemblages and rainbow trout ( O ncorhynchus mykiss ) growth and survival. Canadian Journal of Fisheries and Aquatic Sciences 58(11):2213 2221. Allan, J. D., and A. S. Fle cker. 1993. Biodiversity conservation in running waters. Bioscience 43(1):32 43. Allouche, S. 2002. Nature and functions of cover for riverine fish. Bulletin Franais De La Pche Et De La Pisciculture (365 366):297 324. Aronson, R. B., and K. L. Heck Jr. 1995. Tethering experiments and hypothesis testing in ecology. Marine Ecology Progress Series 121(1):307 309. Aronson, R. B., K. L. Heck Jr, and J. F. Valentine. 2001. Measuring predation with tethering experiments. Marine Ecology Progress Series 214:311 312. AZGFD (Arizona Game and Fish Department). 1996. Ecology of Grand Canyon Backwaters, Final Report to U.S. Bureau of Reclamation, Glen Canyon Environmental Studies, Cooperative Agreement 9 FC 40 07940, Phoenix, Arizona Available: http://www.gcmrc.gov / library/reports/biological/Fish_studies/AZGame&Fish/1996/AZGame&Fish1996.pdf (January 2012). Bain, M. B., J. T. Finn, and H. E. Booke. 1988. Streamflow regulati on and fish community structure. Ecology 69(2) :382 392. Barrett, J. C., G. D. Grossman, and J Rosenfeld. 1992. Turbidity induced changes in reactive distance of rainbow trout. Transactions of the American Fisheries Society 121(4):437 443. Battin, J. 2004. When good animals love bad habitats: Ecological traps and the conservation of animal popula tions. Conservation Biology 18(6):1482 1491. Bednarek, A. T., and D. D. Hart. 2005. Modifying dam operations to restore r ivers: Ecological responses to Tennessee R iver dam mitigation. Ecological Applications 15(3):997 1008.
9 4 Behn, K.E., T.A. Kennedy and R. O. Hall. 2010. Basal resources in backwaters of the Color ado River below Glen Canyon Dam effects of discharge regimes and comparison with mainstem depositional environments: U.S. Geological Survey Ope n File Report 2010 1075 Available: http://pubs.usg s.gov/of/2010/1075/. (January 2012). Bernhardt, E. S., and M. A. Palmer. 2011. River restoration: The fuzzy logic of repairing reaches to reverse catchment scale degradation. Ecological Applications 21(6):1926 1931. Bernhardt, E. S., M. Palmer, J. Allan, G. Alexander, K. Barnas, S. Brooks, J. Carr, S. Clayton, C. Dahm, and J. Follstad Shah. 2005. Synthesizing U S river restoration efforts. Science 308(5722):636. Blinn, D. W., C. Runck, D. A. Clark and J. N. Rinne. 1993. Effects of rainbow trout predation on Little C olorado spinedace. Transactions of the American Fisheries Society 122(1):139 143. Bock, C. E., and Z. F. Jones. 2004. Avian habitat evaluation: Should counting birds count? Frontiers in Ecology and the E nvironment 2(8):403 410. Bond, N., and P. Lake. 2003. Characterizing fish habitat associations in streams as the first step in ecological restoration. Austral Ecology 28(6):611 621. Bradford, M. J., P. S. Higgins, J. Korman, and J. Sneep. 2011. Test of an e nvironmental flow release in a British C olumbia river: Does more water mean more fish? Freshwater Biology 56(10):2119 2134. Brouder, M.J., D. W. Speas, and T. L. Hoffnagle. 1999. Changes in number, sediment composition, and benthic invertebrates of ba ckwaters, in Webb, R.H., Schmidt, J.C., Marzolf, G.R., and Valdez, R.A., eds., The controlled flood in Grand Canyon: Washington, D.C., American Geophysical Union, Geophysical Mono graph Series Bunn, S. E., and A. H. Arthington. 2002. Basic principles and e cological consequences of altered flow regimes for aquatic biodiversity. Environmental Management 30(4):492 507. Byers, C. R., R. K. Steinhorst, and P. R. Krausman. 1984. Clarification of a technique for analysis of utilization availability data. The Jour nal of Wildlife Management 48(3):1050 1053. Camp, E., D. Gwinn, W. Pine III, and T. Frazer. 2011. Changes in submersed aquatic vegetation affect predat ion risk of a common prey fish L ucania parva (cyprinodontiformes: Fundulidae) in a spring fed coastal ri ver. Fis heries Management and Ecology Carpenter, S. R., and J. F. Kitchell. 1993. The trophic casca de in lakes. Cambridge University Press
95 Carter, M. W., D. E. Shoup, J. M. Dettmers, and D. H. Wahl. 2010. Effects of turbidity and cover on prey selectivi ty of adult smallmouth bass. Transactions of the American Fisheries Society 139(2):353 361. Coggins, L. G., W. E. Pine, C. J. Walters, D. R. Van Haverbeke, D. Ward, and H. C. Johnstone. 2006. Abun dance trends and status of the Little Colorado R iver popula tion of humpback chub. North American Journal of Fisheries Management 26(1):233 245. Coggins L. G. and C. Walters. 2009. Abun dance trends and status of the L it tle Colorado R iver population of humpback chub; an update considering data from 1989 2008. U S Geological Survey Open File Report 1075:18. Coggins, L. G., and W. E. Pine. 2010. Development of a temperature dependent growth model for the endangered humpback chub using capture recapture data. The Open Fish Science Journal 3:122 131. Coggins, L. G., M. D. Yard, and W. E. Pine. 2011. Nonnative fish control in the Colorado River in Grand Canyon, A rizona: An effective program or serendipitous timing? Transactions of the American Fisheries Society 140(2):456 470. Converse, Y. K., C. P. Hawkins, and R. A. Valdez. 1998. Habitat relationships of subadult humpback chub in the Colorado River through Grand C anyon: Spatial variability and implications of fl ow regulation. Regulated Rivers Research and Management 14(3):267 284. Crook, D. and Robertson, A. 1999. R elationships between riverine fish and woody debris: Implications for lowland rivers. Marine and Freshwater Research 50(8):941 953. Crowder, L. B., and W. E. Cooper. 1982. Habitat structural complexity and the interaction between bluegills and their prey. Ecology 63(6) :1802 1813. Crowl, T. A., C. R. Townsend, and A. R. Mcintosh. 1992. The impact of introduced brown and rainbow tro ut on native fish: The case of A ustralasia. Reviews in Fish Biology and Fisheries 2(3):217 241. Cucherousset, J., and J. D. Ol den. 2011. Ecological impacts of nonnative freshwater fishes. Fisheries 36(5):215 230. Dudgeon, D., A. H. Arthington, M. O. Gessner, Z. I. Kawabata, D. J. Knowler, C. Lvque, R. J. Naiman, A. H. Prieur Richard, D. Soto, and M. L. J. Stiassny. 2006. Fresh water biodiversity: Importance, threats, status and conservation challenges. Biological Reviews of the Cambridge Philosophical Society 81(2):163 182. Dynesius, M., and C. Nilsson. 1994. Fragmentation and flow regulation of river systems in the northern th ird of the world. Science 266(5186):753.
96 Ebersole, J. L., W. J. Liss, a nd C. A. Frissell. 1997. Restoration of stream habitats in the western United S tates: Restoration as re expression of habitat capacity. Environmental Management 21(1):1 14. Estes, J. A. 1996. Predators and ecosystem management. Wildlife Society Bulletin 24(3):390 396. Fausch, K. D. 1984. Profitable stream positions for salmonids: Relating specific growth rate to net energy gain. Canadian Journal of Zoology 62(3):441 451. Fausch, K. D ., C. E. Torgersen, C. V. Baxter, and H. W. Li. 2002. Landscapes to riverscapes: Bridging the gap between research and conservation of stream fishes. Bioscience 52(6):483 498. FitzHugh, T. W., and R. M. Vogel. 2010. The impact of dams on flood flows in th e United S tates. River Research and Applications 27(10):1192 1215. Fradkin, P. L. 1981. A river no m ore: The Colorado River and the West. Tucson, University of Arizona Press. Gadomski, D. M., and M. J. Parsley. 2005. Effects of turbidity, light level, and cover on predation of white sturgeon larvae by prickly sculpins. Transactions of the American Fisheries Society 134(2):369 374. Garshelis, D. L. 2000. Delusions in habitat evaluation: Measuring use, selection, and importance. Pages 111 164 in Resear ch Techniques in Animal Ecology. Columbia University Press, N ew York, New York. Gleick, P. H. 2003. Global freshwater resources: Soft path solutions for the 21st century. Science 302(5650):1524. Gloss, S. O., and L G. Coggins. 2005. Fishes of Grand Canyon. U.S. Geological Survey. Circular 1282:33 56 Goeking, S. A., J. C. Schmidt, and M. K. Webb. 2003. Spatial and temporal trends in the size and number of backwaters between 1935 and 2000, Marble and Grand Canyons, A rizona. Report Prepared for US Geological Survey, Grand Canyon Monitoring and Rese arch Center, Flagstaff, Arizona Gore, J. A., and F. D. Shields. 1995. Can large rivers be restored? Bioscience 45(3):142 152. Gorman, O. T., and D. M. Stone. 1999. Ecolog y of spawning hum pback chub, G ila cypha in the Little Colorado River near Grand Canyon, A rizona. Environmental Biology of Fishes 55(1):115 133.
97 Gould, W., and K. Pollock. 1997. Catch effort maximum likelihood estimation of important population paramete rs. Canadian Journal of Fisheries and Aquatic Sciences 54(4):890 897. Gowan, C., and K. D. Fausch. 1996. Long term demographic responses of trout populations to habitat manipulation in six C olorado streams. Ecological Applications 6(3) :931 946. Grams, P. E., J.C. Schmidt, and M. E. Andersen. 2010. 2008 high flo w experiment at Glen Canyon Dam morphologic response of eddy deposited sandbars and associated aquatic backwater habitats along the Colorado River in Grand Canyon National Park: U.S. Geological S urvey Open File Report 2010 1032. Gregory, R. S. 1993. Effect of turbidity on the predator avoidance behavio ur of juvenile chinook salmon ( O ncorhynchus tshawytscha ). Canadian Journal of Fisheries and Aquatic Sciences 50(2):241 246. Gregory, R. S., and C. D. Levings. 1996. The effects of turbidity and vegetation on t he risk of juvenile salmonids, O ncorhynchus spp., to predation by adult cutthroat trout, O. clarkii Environmental Biology of Fishes 47(3):279 288. Harvey, B., and A. Ste wart. 1991. Fish size and habitat depth relationships in headwater streams. Oecologia 87(3):336 342. Hoffnagle, T. L. 1996. Changes in water quality parameters and fish usage of backwaters during fluctuating vs. short trem steady flows in the Colorado R iver, Grand Canyon. Available: http://www.gcmrc.gov/library/reports/physical/hydrology/Hoffnagle1996.pdf (January 2012). Hoffnagle, T.L. 2000. Backwater fish communities in th e Colorado River Grand Canyon. F inal r eport prepared for the U.S. Fish and Wildl ife Service, Grand Canyon fishery resource o ffice. Flagstaff, Arizona. Available: http://www.gcmrc.gov/library/reports/biological/Fish_studies/ AZGame&Fish/2000/Hoffnagle2000.pdf (January 2012). Jelks, H. L., S. J. Walsh, N. M. Burkhead, S. Contreras Balde ras, E. Diaz Pardo, D. A. Hendrickson, J. Lyons, N. E. Mandrak, F. McCormick, and J. S. Nelson. 2008. Conservation sta tus of imperiled North A merican freshwater and diadromous fishes. Fisheries 33(8):372 407. Johnson, B. M., P. J. Martinez, J. A. Hawkins, and K. R. Bestgen. 2008. Ranking predatory thre ats by nonnative fishes in the Yampa River, C olorado, via bioenergetics modeling. North American Journal of Fisheries Management 28(6):1941 1953.
98 Johnson, P. T. J., J. D. Olden, and M. J. Vander Zanden. 2008. Dam invaders: Impoundments facilitate biological invasions into freshwaters. Frontiers in Ecology and the Environment 6(7):357 363. Koehn, J. 2009. Multi scale habitat selection by murray cod M accullochella peelii peelii in two lowland rivers. Journal of Fish Biology 75(1):113 129. Korman, J., S. M. Wiele, and M. Torizzo. 2004. Modelling effects of discharge on habitat quality and dispersal of juvenile hum pback chub ( G ila cypha ) in the Colorado River, Grand C anyon. River Research and Applications 20(4):3 79 400. Korman, J., and S. E. Campana. 2009. Effects of hydropeaking on nearshor e habitat use and growth of age 0 rainbow trout in a large regulated river. Transactions of the American Fisheries Society 138(1):76 87. Korman, J., M. Yard, C. Walters, an d L. G. Coggins. 2009. Effects of fish size, habitat, flow, and density on capture probabilities of age 0 rainbow trout estimated from electrofishing at discrete sites in a large river. Transactions of the American Fisheries Society 138(1):58 75. Krebs, J.R. and A. Kacelnik. 1 991. Decision making. Pages 105 136 in Krebs, J. D. and N. B. Davies, editors. Behavioral ecology: an evolutionary a pproach. Oxford, Blackwell Scientific Publications. Laplante Albert, K. A., M. A. Rodrguez, and P. Magnan. 2010. Quantifying habitat dependent mortality risk in lacustrine fishes by means of tethering trials and survival analyses. Environmental Biology of Fishes 87(3):263 273. Leopold, A. 1933. Game Management. Charles Scribner's Sons. Reprinted in 1986 by Un iversity of Wisconsin Press, Madison. Lessard, R. B., S. J. D. Martell, C. J. Walters, T. E. Essington, and J. F. Kitchell. 2005. Should ecosystem management involve active control of species abundances. Ecology and Society 10(2):1. Lima, S. L. 1998. Nonlethal effects in the ecology of predator prey interactions. Bioscience 48(1):25 34. Lorenzen, K., and K. Enberg. 2002. Density dependent growth as a key mechanism in the regulation of fish populations: Evidence from among population comparisons. P roceedings of the Royal Society of London. Series B: Biological Sciences 269(1486):49. Lozon, J. D., and H. J. MacIsaac. 1997. Biological invasions: Are they dependent on disturbance? Environmental Reviews 5(2):131 144.
99 Manly, B. F. J., L. L. McDonald, D L. Thomas, T. L. McDonald, and W. P. Erickson. 2002. Resource selection by animals. 2nd ed. Kluwer Academic Publishers, Dordrecht, the Netherlands. exotic species r emoval on recovery of native fish: Lessons from a dam decom missioning. Restoration Ecology 18(6):934 943. Marsh, P. C., and M. E. Douglas. 1997. Predation by introduced fishes on endangered humpback chub a nd other native species in the Little Colorado Rive r, A rizona. Transactions of the American Fisheries Society 126(2):343 346. McDowall, R. 2006. Crying wolf, crying foul, or crying shame: Alien salmonids and a biodiversity crisis in the southern cool temperate galaxioid fishes? Reviews in Fish Biology a nd Fisheries 16(3):233 422. McIvor, C. C., and W. E. Odum. 1988. Food, predation risk, and microhabitat selection in a marsh fish assemblage. Ecology :1341 1351. Meador, M. R., L. R. Brown, and T. Short. 2003. Relations between introduced fish and enviro nmental conditions at large geographic scales. Ecological Indicators 3(2):81 92. Melis, T. S., editor. 2011. Effects of three high flow experiments on the Colorado River ecosystem downstream from Glen Canyon Dam, Arizona: U.S. Geological Survey Circular 1 366. Minckley, W. 1991. Nat ive fishes of the Grand C anyon region: An obituary. Colorado River Ecology and Dam Management. National Academy Press, Washington, DC 124 177. Minckley, W., P. C. Marsh, J. E. Deacon, T. E. Dowling, P. W. Hedrick, W. J. Matthews and G. Mueller. 2003. A conservation plan for native fishes of the lower Colorado R iver. Bioscience 53(3):219 234. Miner, J. G., and R. A. Stein. 1996. Detection of predators and habitat choice by small bluegills: Effects of turbidity and alternative pr ey. Transactions of the American Fisheries Society 125(1):97 103. Morris, D. W. 2003. Toward an ecological synthesis: A case for habitat selection. Oecologia 136(1):1 13. Moyle, P. B., and J. F. Mount. 2007. Homogenous rivers, homogenous faunas. Proceedi ngs of the National Academy of Sciences 104(14):5711. Nilsson, C., C. A. Reidy, M. Dynesius, and C. Revenga. 2005. Fragmentation and flow regulation of the world's large river systems. Science 308(5720):405.
100 Olden, J. D., N. L. R. Poff, and K. R. Bestgen 2006. Life history strategies predict fish inv asions and extirpations in the Colorado R iver basin. Ecological Monographs 76(1):25 40. Olden, J., and N. Poff. 2005. Lon g term trends of native and non native fish faunas in the American S outhwest. Animal Bi odiversity and Conservation 28(1):75 89. Osmundson, D., R. J. Ryel, V. Lamarra, and J. Pitlick. 2002. Flow sediment biota relations: Implications for river regulation effects on native fish abundance. Ecological Applications 12(6):1719 1739. Ostrand, K. G., B. J. Braeutigam, and D. H. Wahl. 2004. Consequences of vegetation density and prey species on spotted gar foraging. Transactions of the American Fisheries Society 133(3):794 800. Palmer, M. A., R. F. Ambrose, and N. L. R. Poff. 1997. Ecological t heory and community restoration ecology. Restoration Ecology 5(4):291 300. Palmer, M. A., H. L. Menninger, and E. Bernhardt. 2010. River restoration, habitat heterogeneity and biodiversity: A failure of theory or practice? Freshwater Biology 55:205 222. Patten, D. T., D. A. Harpman, M. I. Voita, and T. J. Randle. 2001. A managed flood on the Colorado River: Background, objectives, design, and implementation. Ecological Applications 11(3):635 643. Perot, A., and M. A. Villard. 2009. Putting density back in to the h abitat q ual ity equation: Case study of an o pen n esting forest bird. Conservation Biology 23(6):1550 1557. Peterson, C. H., and R. Black. 1994. An experimentalist's challenge: When artifacts of intervention interact with treatments. Marine Ecology Progress Series 111(3):289 297. Peterson, J. T., R. F. Thurow, and J. W. Guzevich. 2004. An evaluation of multipass electrofishing for estimating the abundance of stream dwelling salmonids. Transactions of the American Fisheries Society 1 33(2):462 475. Pine, W.III. E., T. J. Kwak, and J. A. Rice. 2007. Modeling management scenarios and the effects of an introduced apex predator on a coastal riverine fish community. Transactions of the American Fisheries Society 136(1):105 120. Pine, W., S. Martell, C. Walters, and J. Kitchell. 2009. Counterintuitive responses of fish populations to management actions: Some common causes and implications for predictions based on ecosystem modeling. Fisheries 34(4):165 180.
101 Poff, N. L. R., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks, and J. C. Stromberg. 1997. The natural flow regime. Bioscience 47(11):769 784. Pollock, K. H., J. D. Nichols, C. Brownie, and J. E. Hines. 1990. Statistical inference for captu re recapture experiments. Wildlife Monographs 107 :3 97. Power, M. E., W. E. Dietrich, and J. C. Finlay. 1996. Dams and downstream aquatic biodiversity: Potential food web consequences of hydrologic and geomorphic change. Environmental Management 20(6):8 87 895. Pretty, J., S. Harrison, D. Shepherd, C. Smith, A. Hildrew, and R. Hey. 2003. River rehabilitation and fish populations: Assessing the benefit of instream structures. Journal of Applied Ecology 40(2):251 265. R Development Core Team. 2007. R: A l anguage and environment for statistical computing reference index version 2.13.1 R Foundation for Statistical Computing, Vienna, Austria. ISBN 3 900051 07 0. Available: http://www.Rproject.org. Railsback, S. F., H. B. Stauffer, and B. C. Harvey. 2003. Wh at can habi tat preference models tell us? T ests using a virtual trout population. Ecological Applications 13(6):1580 1594. Ricciardi, A., and J. B. Rasmus sen. 1999. Extinction rates of North A merican freshwater fauna. Conservation Biology 13(5):1220 1222. Richter, B. D., D. P. Braun, M. A. Mendelson, and L. L. Master. 1997. Threats to imperiled freshwater fauna. Conservation Biology :1081 1093. Ricklefs, R. E., and D. Schluter. 1993. Species diversity in ecological communities: historical and geographica l perspectives. University of Chicago Press, Chicago, Illinois. Riley, S. C., and K. D. Fausch. 1992. Underestimation of trout population size by maximum likelihood removal estimates in small streams. North American Journal of Fisheries Management 12(4):7 68 776. Riley, S. C., and K. D. Fausch. 1995. Trout population response to habit at enhancement in six northern C olorado streams. Canadian Journal of Fisheries and Aquatic Sciences 52(1):34 53. Robertson, B. A., and R. L. Hutto. 2006. A framework for under standing ecological traps and an evaluation of existing evidence. Ecology 87(5):1075 1085. Robinson, A. T., R. W. Clarkson, and R. E. Forrest. 1998. Dispersal of larval fishes in a regulated river tributary. Transactions of the American Fisheries Society 127(5):772 786.
102 Robinson A. T., and M. R. Childs. 2001. Juvenile growth of native fishes in the Little Colorado River and in a thermally modified portion of the Colorado River. North American Journal of Fisheries Management 21(4):809 815. Roni, P., K. Ha nson, and T. Beechie. 2008. Global review of the physical and biological effectiveness of stream habitat rehabilitation techniques. North American Journal of Fisheries Management 28(3):856 890. Rosenberger, A. E., and J. B. Dunham. 2005. Validation of abu ndance estimates from mark recapture and removal techniques for rainbow trout captured by electrofishing in small streams. North American Journal of Fisheries Management 25(4):1395 1410. Rosenfeld, J. S., and S. Boss. 2001. Fitness consequences of habita t use for juvenile cutthroat trout: Energetic costs and benefits in pools and riffles. Canadian Journal of Fisheries and Aquatic Sciences 58(3):585 593. Rosenfeld, J. S., and T. Hatfield. 2006. Information needs for assessing critical habitat of freshwate r fish. Canadian Journal of Fisheries and Aquatic Sciences 63(3):683 698. Rosenzweig, M. L. 1995 Species diversity in space and time. Cambridge University Press. Cambridge. Rubin, D. M., J. C. Schmidt, and J. N. Moore. 1990. Origin, structure, and ev olution of a reattachment bar, Colorado River, Grand Canyon, A rizona. Journal of Sedimentary Research 60(6):982 991. Savino, J. F., and R. A. Stein. 1982. Predator prey interaction between largemouth bass and bluegills as influenced by simulated, submer sed vegetation. Transactions of the American Fisheries Society 111(3):255 266. Schmidt, J.C., and J. B. Graf. 1990. Aggradation and degradation of alluvial sand deposits, 1965 to 1986, Colorado River, Grand Canyon National Park, Arizona: U.S. Geological S urvey Professional Paper 1493. Available: http://pubs.er.usgs.gov/usgspubs/citfor/pp/pp1493. (January 2012). Schmidt, J. C., and D. Rubin. 1995. Regulated streamflow, fine grained deposits, and effective discharge in canyons with a bundant debris fans. Geop hysical Monograph American Geophysical Union 89:177 177. Sih, A., P. Crowley, M. McPeek, J. Petranka, and K. Strohmeier. 1985. Predation, competition, and prey communities: A review of field experiments. Annual Review of Ecology and Systematics 16:269 311
103 Speas, D. W., C. J. Walters, D. L. Ward, and R. S. Rogers. 2004. Effects of intraspecific density and environmental variables on electrofishing catchability of brown and rainbow trout in the Colorado R iver. North American Journal of Fisheries Management 24(2):586 596. Stanford, J. A., J. Ward, W. J. Liss, C. A. Frissell, R. N. Williams, J. A. Lichatowich, and C. C. Coutant. 1996. A general protocol for restoration of regulated rivers. U S Department o f Energy Publications 43. Stone, D. M., and O. T. Gorman. 2006. Ontogenesis of endangered hum pback chub ( G ila cypha ) in the Little Colorado River, A rizona. The American Midland Naturalist 155(1):123 135. Strayer, D. L., and D. Dudgeon. 2010. Freshwater b iodiversity conservation: Recent progress and future challenges. Journal of the North American Benthological Society 29(1):344 358. Stromberg, J. C. 2001. Restoration of riparian v egetation in the south western United S tates: Importance of flow regimes an d fluvial dynamism. Journal of Arid Environments 49(1):17 34. Sweka, J. A., and K. J. Hartman. 2003. Reduction of reactive distance and foragi ng success in smallmouth bass, M icropterus dolomieu exposed to elevated turbidity levels. Environmental Biology of Fishes 67(4):341 347. Sweka, J. A., and K. J. Hartman. 2001. Effects of turbidity on prey consumption and growth in brook trout and implications for bioenergetics modeling. Canadian Journal of Fisheries and Aquatic Sciences 58(2):386 393. Tews, J., U. Brose, V. Grimm, K. Tielbrger, M. Wichmann, M. Schwager, and F. Jeltsch. 2004. Animal species diversity driven by habitat heterogeneity/diversity: The importance of keystone structures. Journal of Biogeography 31(1):79 92. Thomas, D. L., and E. J. Taylo r. 1990. Study designs and tests for comparing resource use and availability. The Journal of Wildlife Management 54(2) :322 330. Topping, D. J., Melis, T. S., Rubin, D. M., and Wright, S. A. 2004. High resolution monitoring of suspended sediment concentr ation and grain size in the Colorado River in Grand C anyon using a laser acoustic system. Proceedings of the Proceedings of the Ninth International Symposium o n River Sedimentation, October 18 21. Topping, D. J., D. M. Rubin, and J. C. Schmidt. 2005. Regu lation of sand transport in the Colorado R iver by changes in the surface grain size of eddy sandbars over multi year timescales. Sedimentology 52(5):1133 1153.
104 Topping, D. J., D. M. Rubin, an d L. Vierra Jr. 2000. Colorado River sediment transport 1. N atur al sediment supply l imitation and the influence of Glen Canyon D am. Water Resources Research 36(2):515 542. Townsend, C. R., and T. A. Crowl. 1991. Fragmented po pulation structure in a native New Z ealand fish: An effect of introduced brown trout? Oikos 61 (3) :347 354. Trammell, M., R. Valdez, S. Carothers, and R. Ryel. 2002. Effects of a low steady summer flow experime nt on native fishes of the Colorado River in Grand Canyon, Arizona. Report to U.S. Geological Survey 4:2 1. Turesson, H., and C. Brnmark. 2 007. Predator prey encounter rates in freshwater piscivores: Effects of prey density and water transparency. Oecologia 153(2):281 290. Tyus, H. M., and J. F. Saunders. 2000. Nonnative fish control and endangered f ish recovery: Lessons from the Colorado R iver. Fisheries 25(9):17 24. U.S. Department of the Interior. 2008. Final biological opinion for the operation of G len C anyon D am. U.S. Fish and Wildlife Service, AESO/SE 22410 1993 F 167R1. Available: http://www.usbr.gov/uc/envdocs/bo/FinalGCDBO2 26 08 .pdf (January 2012). Valdez, R., and R. Ryel. 1995. Life history and ecology of the hum pback chub ( G ila cypha ) in the Colorado River, Grand Canyon, Arizona. Final report to Bureau of Reclamation, Salt Lake City, U tah. Contract no. 0 CS 40 091 10. Bio West R eport no.TR 250 08 Van Horne, B. 1983. Density as a misleading indicator of habitat quality. The Journal of Wildlife Management 47(4) :893 901. Van Steeter, M. M., and J. Pitlick. 1998. Geomorphology and endange red fish habitats of the upper Colorado Riv er 1. H istoric changes in streamflow, sediment load, and channel morphology. Water Resources Research 34(2):287 302. Voichick, N., and S. A. Wr ight. 2007. Water temperature data for the Colorado River and tributaries between Glen Canyon Dam and Spencer Canyon, northern Arizona, 1988 2005: U.S. Geological Survey Data Series 251. Walters, C. J. 2007. Is adaptive management helping to solve fisheries problems? AMBIO: A Journal of the Human Environment 36(4):304 307. Walters, C. J., and F. Juanes. 1993. Recruitment limitation as a consequence of natural selection for use of restricted feeding habitats and predation risk taking by juvenile fishes. Canadian Journal of Fisheries and Aquatic Sciences 50(10):2058 2070. Walters, C. J., and R. Hilborn. 1978. Ec ological optimization and adaptive management. Annual Review of Ecology and Systematics 9:157 188.
105 Warfe, D. M., and L. A. Barmuta. 2004. Habitat structural complexity mediates the foraging success of multiple predator species. Oecologia 141(1):171 178. Werner, E. E., and D. J. Hall. 1988. Ontogenetic habitat shifts in bluegill: The foraging rate predation risk trade off. Ecology 69(5) :1352 1366. White, J. L., and B. C. Harvey. 2001. Effects of an introduced piscivorous fish on native benthic fishes in a coastal river. Freshwater Biology 46(7):987 995. Wiele, S. M., and E. R. Griffi n. 1997. Modifications to a one dimensional model of unsteady flow in the Colorado River through Grand Canyon, Arizon a: U.S. Geological Survey Water Resources Investigations Report 97 4046. Wohl, E., P. L. Angermeier, B. Bledsoe, G. M. Kondolf, L. MacDonnell, D. M. Merritt, M. A. Palmer, N. L. R. Poff, and D. Tarboton. 2005. River restoration. Water Resources Research 41(10):W10301. Yard, M. D., L. G. Coggins C. V. Baxter, G. E. Bennett, and J. Korman 2011. Trout piscivory in the Colorado River, Grand C anyon: Effects of turbidity, temperature, and fish prey availability. Transactions of the American Fisheries Society 140(2):471 486.
106 BIOGRAPHICAL SKE TCH Michael (Mike) James Dodrill was born in Denver, Colorado. He spent his childhood in Lakewood, a suburb of Denver, until leaving to attend college in Fort Collins, at Colorado State University ( CSU ). Mike took ti me to explore the rivers of northern Colorado fly rod in hand, and graduated from CSU in 2 degree in Fisheries Biology. After graduation he went to work for the Colorado Division of Wildlife as a technician in the southern mountains of the state. After a short winter spent enjoying the snow at Crested Butte he worked for the USGS studying riparian community structure along the Missouri River in Montana. In 2009, Mike started ecology in the Grand Canyon reach of the Colorado River.