Stream Geomorphology and Fish Community Structure in Channelized and Meandering Reaches of an Agricultural Stream

Kelly M. Frothingham

Department ofGeography and Planning. Bufilo State College, Buffalo. New York

Bruce L. Rhoads

Department o J C e o p p h y , Universify oJlllino~s, Urbana, Illinois

Edwin E. Herricks

Department of Civil and Environmental Engineering, Universi* of Illinois, Urbana, Illinois

In environments dominated by human activity, such as the agricultural Mid- west, channel morphology is strongly influenced by anthropogenic factors. Past research has shown that human-induced channel modifications, including stream channelization, affects both the abiotic and biotic components of a stream. How- ever, connections between changes in geomorphological form and function and ecological conditions are still poorly understood. Knowledge of these connec- tions is critical for successful approaches to stream naturalization. The objective of this research is to explore the linkage between geomorphological variability and ecological conditions in adjacent channelized and unchannelized reaches of an agricultural stream in Illinois. Geomorphological and fisheries data were col- lected in the headwaters of the Embarras River where channel maintenance for agricultural drainage produced a straight channelized reach adjacent to a mean- dering unchannelized reach. Site characterization involved topographic mapping of each reach and establishing permanent cross sections to document the full range of morphologic variability at the sites. Fish sampling was conducted ap- proximately monthly to provide information on community structure in the me- andering and channelized reaches. The meandering reach has greater morpho- logical variability, both over time and space, than the channelized reach. The me- andering reach also contains more fish and larger individual fish than the chan- nelized reach, suggesting that increased geomorphological complexity results in increased fish abundance and total biomass. Fisheries analysis also indicates that fish-community structure varies temporally in both reaches with the greatest con- trast occurring seasonally.

INTRODUCTION

Geomorpbic Processes and Riverine Habitat Human modification of watersheds has produced pro- Water Science and Application Volume 4, pages 105.117 found changes in streams throughout the agricultural Mid- Copyright 2CQ1 by the American Geophysical Union west. A growing emphasis on environmental quality in this

105

106 STREAM GEOMORPHOLOGY AND FISH COMMUNITY STRUCTURE

region of the United States has generated concern about the effects of humans on stream ecosystems. The concept of stream naturalization has been proposed as a management concept to address human and biophysical dimensions of environmental-quality issues in this human-dominated en- vironment [Rhoads and Herricks, 1996; Rhoads and Monahan, 1997; Rhoads el a l , 19991. Naturalization aims to establish morphologically and hydraulically varied, yet dynamically stable [Rhoads, 19951 stream systems that are capable of supporting biologically diverse aquatic ecosys- tems. Successful implementation of naturalization strate- gies requires an improved understanding of the geomor- phological and ecological dynamics of human-modified streams [Frothingham et a/. , in press], as well as the ef- fects of social interactions between scientists and non- scientists in community-based environmental decision- making [Rhoads et al., 19991.

Fluvial geomorphologists and aquatic ecologists both are interested in stream dynamics; however, the focus of these two disciplines differs in important ways. Geomorpholo- gists primarily are interested in the interaction among flow properties, sediment transport, and channel morphology [Dietrich, 1987; Rhoads and Werford, 19911, whereas ecologists typically are concerned about the structure and function of aquatic communities [Osborne and Wiley, 1992; Aadland, 19931. The potential linkage between geo- morphology and ecology centers on a common interest in the physical properties of streams where geomorphological conditions defne the amount, diversity and structure of physical habitat. Ecologists recognize that spatial variabil- ity of physical properties, such as mean velocity, flow depth and substrate characteristics, has a strong influence on relationships between habitat and community structure and function [Aadland, 1993; Beebe, 1996; Gore and Ham- ilton, 1996; Fisher, 19971. Channel morphology also has been acknowledged as an important factor in habitat diver- sity and species richness in streams [Southwood, 1977; FrisseN et 01.. 1986; Gelwick, 1990; Osborne and Wiley, 1992; Aadland, 1993; Fisher, 1997; Kemp et a/. , 1999; Montgomery et a/ , , 19991. Fluvial geomorphologists, on the other hand, have developed theories and techniques related to the dynamic interaction between fluvial forms and processes [Hook and Hamey, 1983; Dietrich, 1987; Rhoadr and Werford, 19911. These theories provide a foundation for an improved understanding of physical habitat and time-related changes in habitat conditions, but integration of geomorphological and ecological analysis has yet to be achieved for low-gradient streams in the agri- cultural Midwest.

Human-induced channel modification, referred to as stream channelization, affects both the abiotic and biotic components of a stream. Stream channelization results in decreased spatial variation of channel morphology and 3-D

flow structure [Diehich, 1987; Rhoads and Werord, 1991; Rhoads and Urban, 19971. Physical processes impacted by channelization include pool-riffle development and alter- nate bar formation [Rhoadr and Werord, 1991; Gregory et a/. , 19941. Channelization also changes aquatic communi- ties, reducing species richness and altering local commu- nity structure and function [Brooks, 1988; Swales, 19881. The reduced species richness in channelized streams has been attributed mainly to decreased physical habitat diver- sity associated with uniform channel morphology [Frissell et a/., 1986; Gelwick, 1990; Osborne and Wiley, 1992; Aadland, 19931.

In watersheds of East Central Illinois, stream channeliza- tion typically occurs in a piecemeal fashion, whereby only a few kilometers of a stream are modified in any particular project. As a result, streams are impacted most severely at the planform scale, i.e. the scale at which the planform characteristics of streams can be clearly discerned rho ad^ and Herricks, 1996; Frothingham et al., in press]. Differ- ences in the spatial extent and timing of channelization has produced a diverse array of channel morphologies [Rhoads and Herricks, 19961, hut the exact relation between geo- morphological form and function and ecological conditions is still poorly understood. Such knowledge is critical for naturalization of these stream systems, which depends on an integrated eco-geomorphological approach to watershed management [Frothingham el al., in press].

The objective of this research is to explore how channel morphology is related to stream ecology using fish- community characteristics as an indicator of stream ecosystem condition or state. To achieve this objective, geomorphological and fish-community data are compared for an adjacent straight channelized reach and an unmodified meandering reach of an agricultural stream in East Central Illinois. The central hypothesis of this study is that increased spatial and temporal variability in stream geomorphology results in increased fish species abundance, biomass and richness.

STUDY AREA

Geomorphological and fisheries data were collected from adjacent meandering and channelized reaches in the headwaters of the Embarras River in Champaign County, IL (Figure I). The Embarras River originates on the Ur- bana moraine at the southern edge of the Champaign- Urbana metropolitan area and flows south between the West Ridge and Philo moraines. The two study sites are located approximately 10 km south of Champaign-Urbana.

Channelization of the Embanas River began in the late 19" century and over the last 100 years many sections have been subject to repeated maintenance. The meandering reach is situated within an 800-111 section of the Embarras

FROTHINGHAM ET AL. 107

Figure 1. The Embarras River showing the location of the meandering and channelized reaches (.).

River that has not been channelized since at least the 1930s. This 800-111 section represents unmodified channel conditions in the sense that channel dimensions and plan- form characteristics have fully adjusted to any past chan- nelization and the river channel has been dynamically sta- ble over the past several decades. The study reach within this unmodified section consists of a high-amplitude me- ander bend approximately 70 m in length (Figure 2). Bank- full channel width ranges &om 13.0 to 18.0 m and bankfull depth is between 1.0 and 2.0 m. The channelized reach is located immediately downstream of the 800-m long un- channelized section. It is located at the upstream end of a 275-111 section of the river that was straightened between 1973 and 1982 [Rhoacls and Herricks, 1996; Rhoads and Urban, 19971. This 275-111 channelized section subse- quently remained straight, but in 1996 channel mainte- nance was conducted to deepen the channel by 0.65 m and to regrade the channel banks to 3:l side slopes. The chan- nelized reach is 50 m long and has a trapezoidal channel cross section. Bankfull width varies between 13.0 and 16.0 m and bankfull depth is approximately 2.0 m (Figure 3).

METHODOLOGY

The field measurement program for each reach included: (1) mapping in detail the channel morphology, (2) docu- menting changes in channel form, and (3) electrofishing to assess fish community structure. Channel morphology was mapped in June 1997 using a total station. Two local benchmarks were established at each site for horizontal and vertical control; survey data were reduced to Cartesian (X,Y,Z) coordinates. Contour maps for each reach were generated from the coordinate information using the com- puter software package Surfer, which defines contours for a uniform grid of elevation data interpolated via kriging.

Changes in channel morphology were determined by conducting repeat surveys at monumented cross sections oriented orthogonally to the planform alignment of the bankfull channel (Figures 2 and 3). The cross sections en- compass the full range of morphologic variability in the two reaches. Repeat surveys of the cross sections were conducted three times between October 1997 and June 1998.

108 STREAM GEOMORPHOLOGY AND FISH COMMUNITY STRUCTURE

0.1 m mniour interval i 4

Fieure 2. IlcJ tupufraphy of the memdcrlng rcnch md locat~ons of pcmmml cross-secl~ons Cross sections are ortented or- rl~ogonally lo the locdl channel direct~on 3s defined by a n arc of cunal~re along the ccnterl~ne of thr hmkfull channel.

Spatial variability of channel morphology within each reach was evaluated qualitatively by visually inspecting variations in bed elevation on contour maps. Temporal variability in channel morphology was defined by changes in the profiles of surveyed cross sections. Sediment dynamics were quantified in terms of net sediment accumulation or removal by volume within each reach between channel surveys. Calculations of net volumetric sediment flux were performed in Surfer by interpolating a

uniform grid of elevation data from the survey information using kriging, and then differencing the interpolated elevation values for each grid node for successive surveys. Negative flux volumes indicate net deposition and positive flux volumes indicate net erosion.

Fish community sampling was conducted on a quasi- monthly basis from July 1997 through December 1998 at the straight and meandering reaches. Ten collections were obtained in the meandering and straight reaches [Tompkim,

FROTHMGHAM ET AL. 109

direction

'" i

5 rn 0.1 m mntour interval

Figure 3. Bed topography of the channelized reach and locations of permanent cross-sections. Cross sections are oriented or- thogonally to the local channel direction as defined by the linear path of the channel centerline.

1998; Hauser, 19991. Samples were collected using a Cof- felt Mark 10 backpack electro-shocker and block nets were installed at the upstream and downstream ends of a reach to ensure that fish could not escape. Sampling campaigns

were conducted only during sub-bankfull conditions and all fish collections were obtained between 8:00 a.m. and 5:00 p.m. Fish were identified, weighed and measured. Fisheries data &om both reaches were analvzed throueh general assessment of species number and length.

RESULTS

Spatial Variability in Channel Morphology

Topographic mapping indicates that the meandering reach has greater spatial variability in channel morphology, and thus physical habitat, than the chamelized reach (Fig- ures 2 and 3). Topographic highs and lows occur througb- out the meandering reach (Figure 2). Point bars are present along the inner bank of the upstream (cross-sections 5-8) and downstream (cross-sections 12-14) parts of the reach. Pools exist at the bend entrance (cross-section 1-3) and opposite the point bars at bend apices (cross-sections 6-8 and 12-14). The point bar-pool morphology produces pro- nounced asymmetry of channel cross sections, which are characterized by a nearly vertical cut bank and a gently sloping point-bar face (Figure 2). Cross-stream relief of the channel bed is as much as 0.75 m at the bend apices. Rif- fles between the pools have fairly symmetrical channel geometry (cross-sections 3-5 and 9-11). Longitudinal variation in the thalweg elevation between the riffles and pools is about 0.5 m.

The geometry of the channel in the channelized reach is essentially trapezoidal, reflecting recent channelization (Figure 3). Channel banks have uniform side slopes that extend to the edge of a relatively flat channel bottom. The elevation of the channel bed varies less than 0.1 m, except at the upstream end of the reach where a bar along the west bank produces local cross-stream relief of about 0.45 m. Longitudinal variation in bed elevation along the thalweg is less than 0.25 m throughout the reach.

Spatial Variability in the Fish Community

The fish collections were analyzed to assess differences in community structure between the meandering and the channelized reaches (Table 1). Species richness is similar in the channelized and meandering reaches, but fish abun- dance and biomass is greater in the meandering reach than in the channelized reach (Figures 4 and 5, Table 1). The fish community in these two reaches is composed mainly of mixed cyprinids, including various minnows and carp, and some additional benthic species like yellow bullhead and golden redhorse (Figures 4 and 5). Average fish abun- dance for the meandering reach exceeds that of the chan- nelized reach by about 25%. Fish in the meandering reach also are, on average, twice as large as those in the channel- ized reach (Table I).

110 STREAM GEOMORPHOLOGY AND FISH COMMUNITY STRUCTURE

Table 1. Fish collection summary results

Site Meandering reach Channelized reach 28

-- Total number of species 27 Range of number of fish col- 192-850 151-495 - lected per sampling campaign

Average number of fish col- 434 lected per sampling campaign

Average weight of fish weigh- I5 grams ing greater than one gram

6 grams

Percent of fish weighing less 4% I I%

Further analysis of fish collection data provides addi- tional insight into the effects of spatial variability in geo- morphological conditions on fish community structure. This detailed analysis examines the average number of fish, average length and difference ratios for number and length for the six most common species in the two reaches (Table 2). A positive value for the difference ratio indicates a greater value in the meandering reach; correspondingly, a negative value indicates a greater value in the channelized reach. For all species evaluated, with the exception of the Creek chub, more specimens were collected in the mean- dering reach and the average length is larger in the mean- dering reach than in the channelized reach. Although there is little difference in the annual community composition in morphologically varied and morphologically uniform reaches, there are clear differences in fish size/hiomass characteristics. Fish in the meandering reach are larger, and generally occur in greater numbers than fish in the channelized reach.

Temporal Variability in Channel Morphology

The repeat surveys of monumented cross sections show that the meandering reach exhibits greater temporal change in channel morphology than the channelized reach (Figures 6 and 7). At the entrance to the meandering reach (cross- section I), erosion of a bar platform along the inner bank and deposition on the channel bed has transformed the ge- ometry of the pool at this location (Figure 6). The most prominent channel change in the meandering reach in- volves systematic erosion of the outer cut bank, especially at bend apices. Bank erosion at the two apices (cross- sections 6 and 12) exceeds one meter per year (Figure 6). Maximum erosion has occurred at the downstream end of the reach (cross-section 14) where the outer bank has moved by about 2 m between 1997 and 1998. The cut bank erodes mainly through slab failures that introduce large blocks of bank material into the stream. These blocks alter local flow conditions near the bank and enhance physical-

habitat diversity for fish. Net deposition is evident on point bars along the inner banks of the upstream (cross-section 6) and downstream (cross-sections 12) bends.

Changes in channel cross sections in the channelized reach are small and localized compared to those in the me- andering reach (Figure 7). Minor systematic erosion is evi- dent at the toe of the right (south) bank where flow is de- flected laterally toward this side of the channel by the bar at the upstream end of the reach (cross-section 5). Overall, however, channel form in this reach is highly stable.

Temporal Variability in Sediment Dynamics

Calculations of net sediment flux within each reach con- firm that the meandering reach is more dynamic over time than the channelized reach (Table 3). Two time intervals were used to illustrate temporal variability in sediment flux: October 1997 to early June 1998 (June 1 and June 5 for the straight and highly sinuous reach, respectively) and early June to late June 1998 (June 25). The first eight- month period provided information for the entire win- terlspring period, which included a variety of hydrological events ranging from baseflow to overbank conditions. The second, twenty-day interval was characterized by sustained near-bankfull to bankfull flow in the two reaches.

Erosion was the dominant process in both reaches; how- ever, the volume of net erosion is greater in the meander- ing reach than in the channelized reach (Table 3). During the first interval net erosion in the meandering reach was an order of magnitude greater than in the channelized reach, whereas during the second interval it was about two times greater than in the channelized reach. The positive sediment flux for the meandering reach suggests that, over the period of record, erosion of cut banks exceeded deposi- tion on point bars. Net sediment flux in the meandering reach for the twenty-day period in June 1998 is only slightly less than the net flux for the eight-month period from October 1997 until June 1998. This similarity in net flux illustrates the important influence of sustained bank-

~- - - -

. A i ~-

7nB7 Collection Date

CA = z

Figure 4. Fish community composition of the meandering reach.

full flow on the rate of channel-shaping processes. The net flux is nearly zero for the eight-month period at the chan- nelized reach, suggesting that erosion and deposition ap- proximately were balanced during this period of variable flow conditions. The channelized reach apparently was able to transfer downstream sediment evacuated from the meandering reach upstream. The comparatively large posi- tive net flux during the twenty-day period confirms that bankfull events most strongly influence the rate of sedi- ment dynamics in the channelized reach. The difference in net sediment flux between the channelized and meandering reaches indicates that the greater spatial variability of channel morphology in the meandering reach is associated with higher temporal rates of erosional and depositional processes.

Temporal Variabilify in Fish Community Structure

of temporal variability of fish-community composition was performed for the meandering reach only. In this analysis, the fisheries collections fiom the meandering reach were used to cluster different collection dates based on the simi- larity in community composition (Jaccard coefficient of similarity for presencelabsence data) and fish abundance (Euclidean distance coefficient for presencelabundance data). Both dendrograms indicate that the fish community changes through time (Figwes 8 and 9). Adjacent months often are the most similar, whereas the largest dissimilari- ties generally are associated with different seasons. There- fore, the changes observed in community structure in the meandering reach are likely related to short term phe- nomenon such as changing flow stage and water tempera- ture and seasonal migration or movement of fish [Cham- bers, 19941.

DISCUSSION

Given the similarity in community composition between The fmdings of this study generally are consistent with the meandering and channelized reaches (Table I), analysis those from other investigations comparing stream mor-

112 STREAM GEOMORPHOLOGY AND FISH COMMUNITY S T R U C m

- L7 1 1 I24197 7/7/97 Collection Date

Figure 5. Fish community composition of the channelized reach.

Table 2. Fish collection and difference analysis comparison for adults

Average Average Average # Average # Average # length un- length Average collected collected

Species unmodified straight modified straight length ratio

difference meandering trapezoidal (S:N) meandering trapezoidal ratio (S:N)

(mm) (mm) Bluntnose 158 103 0.53 57 45 0.79 Minnow

Central 56 45 0.20 Stoneroller

66

Creek Chub

39

Greenside Daner

4

Longear Sunfish

9

Striped 30 3 0.90 100 70 0.70 Shiner

7.0 7.0

00 5.0 10.0 15.0 a 0 I " " l " " l " " I L k h ~ ? (m) 00 5.0 100 15.0 200

Distance (m)

Figure 6. Changes in channel cross sections at the meandering teach for the period October 1997 through June 1998

phologic variability and fish-community characteristics andering reach of the Emharras River, which has greater [Gorman and Karr, 1978; Schlosser, 1982; Swales, 1988; spatial variability in channel morphology than the channel- Gelwick, 1990; Lobb and Orth, 19911. Results support the ized reach, contained more and larger fish than the chan- hypothesis that increased spatial variability in channel nelized reach. This association suggests that availability of morphology results in improved habitat for fish. The me- physical habitat, defined on the basis of stream morphol-

114 STREAM GEOMORPHOLOGY AND FISH COMMUNITY STRUCTURE

Cross-section 5 11.0

Figure 7. Changes in channel cross sections at the channelized reach for the period October 1997 through June 1998.

ogy, has a detectable influence on ecological conditions in Embarras River has remained relatively stable over the last human-modified agricultural streams. No substantial dif- 100 years [Hauser, 19991. The similarity in community ference in species richness was observed between the composition in this study reflects the mobility of fish and, channelized and meandering sections of the E m b m considering that the sampled reaches are adjacent to one River. Other work has shown that species richness in the another, is not surprising. While the fish community in the

FROTHINGHAM ET AL. 115

Table 3. Sediment flux results

Site Date Volume of erosion Volume of deposition Net volume change (m3) (m') (m3m-I)

Meandering reach I011 0197-615198 83.7 36.7 0.67 615198-6125198 55.8 17.3 0.55

Channelized reach 1013197-6/15/98 16.8 17.5 0.01 611198-6/25/98 26.1 11.0 0.29

meandering reach changed over time, the relationship of this temporal variability to channel morphology is uncer- tain. Fish community conditions were affected by short- term factors (i.e., weather-related hydrologic variability) and longer-term phenomenon (i.e., seasonal hydrological variability and temperature regimes). Although channel morphologyhabitat changed during the period of fish sam- pling, the effects of these changes on fish-community structure could not be discerned from the type of informa- tion collected in this study.

The linkage between physical-habitat and fish commu- nity structure documented in this investigation represents a first step in the development of a knowledge base to guide integrated eco-geomorphological approaches to stream naturalization in East Central Illinois [Frothingham el al., in press]. Because naturalization may involve myriad man- agement strategies, ranging from passive recovery to ecol- ogically sensitive channel maintenance, a critical need ex- ists for adequate knowledge to defme the ecological bene- fits that will accrue from a particular style of naturaliza- tion. Without such knowledge, both the formulation of ecological goals and assessments of the capacity to attain these goals will be uninformed, increasing the risk of pro- ject failure. Concern about failure is especially acute in East Central Illinois where most naturalization projects are likely to he undertaken by drainage-district commissioners, who are concerned fmt and foremost about drainage, hut who also have an interest in being environmental stewards [Rhoads et al., 19991. The money to support drainage- district projects is raised through taxes imposed on land- owners within a district. The commissioners feel obligated

to use their neighbors tax dollars wisely and to demonstrate that expenditures produce tangible benefits.

The findings of this study also support the contention of drainage-district commissioners that a management trade- off exists between the static stability of a channelized reach and the dynamic stability of unmodified stream channels. Streams in East Central Illinois have low values of stream power [Rhoads and Herricks, 19961 and generally do not exhibit dramatic post-channelization adjustment, such as downcutting and channel widening [e.g., Harvey and Wat- son, 1986; Simon and Hupp, 19901 or rapid re- establishment of meanders [e.g. Brookes, 19881. Many channelized streams in this region are statically stable [Rhoads and Urban, 1997; Urban, 20001--a consequence that is viewed as desirable from a land-drainage perspec- tive, but as undesirable ecologically. The lack of geomor- phological recovery from hnman-induced disturbance pre- vents the development of physical habitat within morpho- logically homogenous channelized streams.

The fmdings here show that the dynamic stability of the unmodified meandering reach provides diverse physical habitat that locally enhances the abundance and biomass of fish communities. One approach to naturalization in East Central Illinois is to preserve meandering segments as "habitat islands" within extensively modified agricultural stream systems [Frothingham et al., in press]. This ap- proach was used to protect the 800-m meandering segment from channelization in 1996 [Rhoads and Herricks, 19961. However, drainagdistrict commissioners still perceive erosion rates in this section of the river as excessive [Rhoads et al., 19991. This st1 ns that the mean-

Figure 8. Meandering reach fisheries presencelabsence dendrogram (laccard Dissimilarity)

Figure 9. Meandering reach fisheries presence

dering reach is migrating laterally; however, the rates of migration documented in this short-term field investigation are consistent with erosion rates in unmodified meandering segments of the Emharras River over the last 60 years [Rhoads and Urban, 1997; Urban, 20001 and with rates of erosion for other channels of similar size and drainage area [ H o o k , 1980; H o o k and Redmond, 19921. Those respon- sible for stream management in East Central Illinois, if they endeavor to improve environmental quality, should understand the dynamic aspects of river channels and the linkage of these dynamics, via processes that maintain spa- tial heterogeneity of physical habitat, to ecological condi- tions. Such understanding depends not only on the avail- ability of accurate scientific information on eco- geomorphological linkages, an issue this study has aimed to address, but also on meaningful social interaction be- tween scientists and local people [Rhoads et al . , 19991.

CONCLUSION

This study has provided detailed data on the linkages between ecological conditions and morphological variabil- ity in a straight channelized reach and an unmodified me- andering reach of an agricultural stream in East Central Illinois. The meandering reach has greater morphological variability over time and space, resulting in greater habitat diversity, fish abundance and fish biomass than the chan- nelized reach. This finding is consistent with previous re- search that has emphasized availability of physical habitat as a key factor governing fish community characteristics in human-modified streams in the Midwest [Schlosser, 1982; Terhaar and Herricks, 19891. Ongoing work is examining how the difference in morphological complexity between the two reaches influences three-dimensional flow charac- teristics, thereby creating a difference in the diversity of hydraulic habitat [Frothingham, 20011.

This study is part of a comprehensive research program that seeks to develop a knowledge base to guide stream- naturalization activities in East Central Illinois. The com- prehensive program is based on a multiscale eco- geomorphological perspective [Frothingham el al. , in

labundance dendrogram (Euclidean Distance).

press]. It combines historical analysis of fisheries [Cham- bers, 1994; Hauser, 19991 and stream-dynamics [Urban, 20001 with field-based analysis of current geomorphologi- cal and ecological conditions [Frothingham, 20011 to bet- ter understand eco-geomorphological linkages in agricul- tural stream systems. The knowledge emerging from this research will underpin efforts to develop stream- management alternatives that promote dynamically stable stream systems capable of supporting biologically diverse aquatic ecosystems.

Aclo~owledgments Michael Urban, Melinda Newell, Marta Graves, Matthew Ladewig, Perry Cabot, Alexander Sukhodolov, and Kathleen Gregory assisted in the collection of geomor- phological data. Timothy Hauser, Mark Tompkins, John Schwartz, Todd Looby, Monica Perez, and Beth Wentzel assisted in the collection of fisheries data. David R. Montgomery, E. Eric Knudsen, and an anonymous reviewer helped the manuscript considerably by providing many constructive comments. This work was supported by grants from the National Science Founda- tion (SBR-9811322), the U.S. Environmental Protection Agency (R82-5306-OIO), the University of Illinois Research Board, and the Department of Geography at the University of Illinois.

REFERENCES

Aadland, L. P., Stream habitat types: their fish assemblages and relationship to flow, North American Journal of Fisheries Management, 13, pp. 790-806, 1993.

Beebe, I. T., Fluid speed variability and the importance to manag- ing fish habitat in rivers, Regulated Rivers: Research and Management, 12, pp. 63-79, 1996.

Brookes, A,, Physical effects, in Channelized Rivers, pp. 81-1 10, John Wiley and Sons, Chichester, 1988.

Chambers, M. D., Spatial and temporal variation offi.~h commu- nilies in the headwaters ofthe Kaskaskia River, Master of Sci- ence thesis, University of Illinois, Urbana, 1994.

Dietrich, W. E., Mechanics of flow and sediment transport in river bends, in River Channels & Environment and Process, edited by K . Richards, pp. 179-224, Basil Blackwell, Oxford, 1987.

Fisher, S. G., Creativity, idea generation, and the functional mor- phology of streams, Journal of the North American Ben- thological Sociely, 16, pp. 305-3 18, 1997.

FROTHINGHAM ET AL. 1 17

Frissell, C. R., W. J. Liss, C. E. Warren, and M. D. Hurley, A Impact, Monitoring, andAssessmen1, edited by E. E. Herricks, hierarchical framework for stream habitat classification: view- pp. 65-75, Lewis, Boca Raton, 1995. ing streams in a watershed context, Environmental Manage- Rhoads, B. L., and E. E. Herricks, Naturalization of headwater menl, 10, pp. 199-214, 1986. streams in Illinois, in River Channel Restoration, edited by A.

~ r o t h i n ~ h h , . ~ . M., GeomorphologicalProcesses in Meandering Brookes, and D. F. Shields Jr, pp. 33 1-367, John Wiley and Sons, and Straiaht Reaches o f an Apricultural Stream in East Chichester. 1996.

Central 11li&is: Relatio& to ~ q u a l i c Habitat, Ph.D. thesis, Uni- versity of Illinois, Urbana, 2001.

Frothingham K. M., B. L. Rhoads, and E. E. Herricks, A multis- cale conceptual framework for integrated eco-geomorphic re- search to support stream naturalization in the agricultural Mid- west, Environmenlol Management, in press.

Gelwick, F. P., Longitudinal and temporal comparisons of rime and pool fish assemblages in a northeastern Oklahoma Ozark stream, Copeia, 4, pp. 1072-1082, 1990.

Gore, J. A,, and S. W. Hamilton, Comparison of flow-related habitat evaluations downstream of low-head weirs on small and large fluvial ecosystems, Regulated Riven: Research and Management, 12. pp. 459-469, 1996.

Gorman, 0 . T., and J . R. K m , Habitat structure and stream fish communities, Ecology, 59, pp. 507-515, 1978.

Gregory, K. J., A. M. Gurnell, C. T. I-lill, and S. Tooth, Stability of the pool-riffle sequence in changing river channels, Regu- lated Rivers: Research and Management, 9, pp. 35-43, 1994.

Harvey, M. D., and C. C. Watson, Fluvial processes and morpho- logical thresholds in incised channel restoration, Water Re- sources Bulletin, 22, pp. 359-368, 1986.

Hauser, T. A., Descriptive comparison of resident and transient firh communities in headwater streams in East Central Illinois, Master of Science thesis, University of Illinois, Urbana, 1999.

Hooke, J. M., Magnitude and distribution of rates of river bank erosion, Earth Surface Processes, 5, pp. 143-157, 1980.

Hooke, J . M., and A. M. Harvey, Meander changes in relation to bend morphology and secondary flows, Special Publications of the International A~sociation of Sedimenlologists. 6 , pp. 121- 132,1983.

Hooke, I. M., and C. E. Redmond, Causes and nature of river planform change, in Dynamics of Gravel-bed Rivers, edited by P. Billi, R. D. Hey, C. R. Thorne, and P. Tacconi, pp. 557-571, John Wiley & Sons, Chichester, 1992.

Kemp, J., D. M. Harper, and G. A. Crosa, Use of "functional habitats" to link ecology with morphology and hydrology in river rehabilitation, Aquatic Conservotion Marine Freshwater Ecosystems, 9, pp. 159-178, 1999.

Lobb 111, M. D., and D. I. Orth, Habitat use by an assemblage of fish in a large warmwater stream, Transactions of the American Fisheries Society, 120, pp. 65-78, 1991.

Montgomery, D. R., E. M. Beamer, G. R. Pess, and T. P. Quinn, Channel type and salmonid spawning distribution and abun- dance, Canadian Journal ofFisheries and Aquatic Science, 56, pp. 377-387, 1999.

Osborne, L. L., and M. 1. Wiley, Influence of tributary spatial position on the structure of warmwater fish communities, Ca- nadian Journal of Fisheries and Aquatic Science, 49, pp. 671- 681, 1992.

Rhoads, B. L., and M. R. Welford, Initiation of river meandering, Progress in Physical Geography, IS, pp. 127-156, 1991.

Rhoads, B. L., Stream power: a unifying theme for urban fluvial geomorphology, in Stormwater Runoff and Receiving Systems:

Rhoads, B. L , and K. M. Monahan, Geomorphological principles for 'naturalizing' streams and rivers in Illinois, in Governor's Conference on the Management of the Illinois River System, edited by A. M. Strawn, pp. 79-86, Water Resources Center, University of Illinois, Urbana, 1997.

Rhoads, B. L., and M. A. Urban, Human-induced geomorphic changes in low-energy agricultural streams: an example from east central Illinois, in Management of Landrcapes Disturbed bv Channel Incision. edited by S . S. Y. Wanp. E. J. Lanzen- - - doen, and

F. D. Shields. Jr.. oo. 968-973. The Center for Comoutational . . . . Hydroscience and Engineering, The University of Mississippi, Oxford, 1997.

Rhoads, B. L., D. Wilson, M. A. Urban, and E. E. Herricks, Inter- action between scientists and nonscientist in community-based watershed management: emergence of the concept of stream naturalizafion, Environmental Management, 24, pp. 297-308, 1999.

Schlosser, I. J., Fish community structure and function along two habitat gradients in a headwater stream, Ecological Mono- gmphr, 52, pp. 395-414, 1982.

Simon, A., and C. R. Hupp, 'The recovery of alluvial systems to imposed channel modifications, West Tennessee, USA, in Vegetation and Erosion, edited by J. B. Thomes, pp. 145-160, John Wiley and Sons, Chichester, 1990.

Southwood, T. R. E., Habitat, the templet for ecological strate- gies?, Journal ofAnimo1 Ecology, 46, pp. 337-365, 1977.

Swales, S., Fish populations of a small lowland channelized river in England subject to long-term river maintenance and man- agement works, Regulated Rivers: Research and Management, 2, pp. 493-506, 1988.

TerHaar, M. J., and E. E. Herricks, Management and Develop- ment of Aquatic Habitat in Agricullural Drainape Systems. UILU-WRC-89-212, 145 pp., Water Resources Center, Uni- versity of Illinois, Urbana, 1989.

Tompkins, M. R., Analysis of thejisherier in maintainedstreams using the PHABSIM Model and a habitat driven population model, Masters of Science thesis, University of Illinois, Ur- bana, 1998.

Urban, M. A., Conceptualizing anthropogenic change in fluvial system: drainage development on the upper Embarras River, Illinois, Ph.D, thesis, University of Illinois, Urbana, 2000.

Kelly M. Frothingham, Buffalo State College, Department of Geography and Planning, 1300 Elmwood Avenue, Buffalo, NY 14222

Bruce L. Rhoads, University of Illinois, Department of Geog- raphy, 607 South Mathews Avenue, 220 Davenport Hall, Urbana IL61801

Edwin E. Herricks, University of Illinois, Department of Civil and Environmental Engineering, 1105 Newmark Civil Engineer- ing Laboratory, 205 North Mathews Avenue, Urbana, IL 61801