Geomorphology 77 (2006) 249 – 269

www.elsevier.com/locate/geomorph

Flow and form in rehabilitation of large-river ecosystems:

An example from the Lower Missouri River
Robert B. Jacobson a,⁎, David L. Galat b,1
a
U.S. Geological Survey, Columbia, Missouri, USA
b
U.S. Geological Survey, Cooperative Research Units, University of Missouri, Columbia, Missouri, USA
Received 11 June 2004; received in revised form 11 March 2005; accepted 5 January 2006
Available online 28 February 2006

Abstract

On large, intensively engineered rivers like the Lower Missouri, the template of the physical habitat is determined by the nearly
independent interaction of channel form and flow regime. We evaluated the interaction between flow and form by modeling four
combinations of modern and historical channel form and modern and historical flow regimes. The analysis used shallow, slow
water (shallow-water habitat, SWH, defined as depths between 0 and 1.5 m, and current velocities between 0 and 0.75 m/s) as an
indicator of habitat that has been lost on many intensively engineered rivers and one that is thought to be especially important in
rearing of young fishes. Two-dimensional hydrodynamic models for modern and historical channels of the Lower Missouri River at
Hermann, Missouri, indicate substantial differences between the two channels in total availability and spatial characteristics of
SWH. In the modern channel, SWH is maximized at extremely low flows and in overbank flows, whereas the historical channel
had substantially more SWH at all discharges and SWH increased with increasing discharge. The historical channel form produced
3–7 times the SWH area of the modern channel regardless of flow regime. The effect of flow regime is evident in increased within-
year SWH variability with the natural flow regime, including significant seasonal peaks of SWH associated with spring flooding.
Comparison with other reaches along the Lower Missouri River indicates that a) channel form is the dominant control of the
availability of habitat even in reaches where the hydrograph is more intensively altered, and b) rehabilitation projects that move
toward the historical condition can be successful in increasing topographic diversity and thereby decreasing sensitivity of the
availability of habitat to flow regime. The relative efficacy of managing flow and form in creating SWH is useful information
toward achieving socially acceptable rehabilitation of the ecosystem in large river systems.
Published by Elsevier B.V.

Keywords: Physical habitat; Flow regime; Missouri River; 2-dimensional modeling

1. Introduction Consequently, restoration of the flow regime has been

cited as a necessary, and often sufficient, condition for
Flow regime is generally considered the primary vari- restoration of the ecosystem (National Research
able driving processes in the river ecosystem (Richter Council, 1992). The biological basis for the primacy
et al., 1997; Poff et al., 1997; Richter et al., 2003). of flow regime is based on well-documented arguments
that life stages of many species and many ecosystem
⁎ Corresponding author. Fax: +1 573 876 1904.
functions depend on a dynamic flow regime (Bayley,
E-mail addresses: [email protected] (R.B. Jacobson),
1995; Sparks, 1995; Poff et al., 1997; Galat et al.,
[email protected] (D.L. Galat). 1998; Middleton, 2002). Periodic variation of flow is
1
Fax: +1 573 884 5070. considered important in renewal of riparian vegetation
0169-555X/$ - see front matter. Published by Elsevier B.V.
doi:10.1016/j.geomorph.2006.01.014
250 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

communities, episodically connecting fish spawning, flow regime and channel form can be (and often must be)
rearing, and foraging habitats to the channel, and for manipulated independently. Separate adjustment of flow
transporting energy and nutrients between the channel and form gives managers additional flexibility to achieve
and flood plain (Sparks, 1995). ecological, social, and economic objectives, but this abi-
The complementary geomorphic argument is based on lity also implies the need to develop a detailed understan-
the idea that rivers adjust morphology — hence available ding of the interplay between flow and form in setting the
physical habitat — to periodic geomorphically effective physical template for ecosystem processes.
flows (Wolman and Miller, 1960; Leopold et al., 1964). Social and economic factors may provide hard con-
Flood flows have been identified as necessary for straints on achievement of some rehabilitation goals. For
creation of sandbars (Webb et al., 1999), creation of new example, rehabilitation activities on rivers used for
flood plain surfaces associated with channel migration navigation may achieve greater habitat diversity within
(Friedman et al., 1997), and rejuvenation of spawning the engineered channel, but it is typically unacceptable to
gravels (Kondolf and Wilcock, 1996). allow the channel to migrate freely. Although the full
Large rivers, those with drainage areas in excess of natural range of lateral erosion and deposition processes
250,000 km2, are typically used by society for a variety of may not be possible under engineered conditions, some
economic services including hydropower, navigation, and increased geomorphic dynamism may be achievable
water supply. Flood plains of large rivers are valued for the within the banks.
extensive flat land and fertile soil and are often intensively
exploited for agriculture and urban development; maxi- 2. Physical habitat in river rehabilitation
mum economic value of these lands is achieved when they
are protected from flooding by flow regulation and levees. In this paper we explore the interplay between channel
Because of the large economic benefits that accrue from form and flow regime in determining the spatial and
engineering and active management of large rivers and temporal distribution of physical habitat in large,
flood plains, restoration to pre-managed conditions is intensively engineered rivers. We use the Lower Missouri
seldom realistic (Gore and Shields, 1995; Stanford et al., River (LMOR) as an example (Fig. 1). The LMOR is
1996). Hence, attempts to restore ecosystem integrity of defined as the Missouri River downstream of Gavins
large rivers tend to focus on specific functions or goals of Point dam near Yankton, South Dakota, to its confluence
species populations that can be accomplished without with the Mississippi River near St. Louis, Missouri. The
compromising traditional economic benefits, rather than Missouri River is the longest river in the United States
holistic restoration (Gore and Shields, 1995). The process (more than 4000 km long) and drains more than
of restoring components of the ecosystem has been called 1,300,000 km2. Like many large rivers, the LMOR is
rehabilitation or naturalization to distinguish it from ho- intensively managed and subject to conflicting manage-
listic restoration (National Research Council, 1992; ment objectives, including flood control, navigation,
Sparks et al., 1998; Rhoads et al., 1999). hydropower production, irrigation, water supply, recrea-
Many large rivers have flow regimes that have been tion, and support of threatened and endangered species.
altered by dams, diversions, and hydrologic changes in the The history and environmental context of management
watershed, and they have channel forms that have been conflict of the Missouri River are summarized by the
altered by bank stabilization, channel training structures National Research Council (2002).
(wing dikes), and levees. Whereas the natural flow para- Our emphasis is on understanding the template of
digm (Poff et al., 1997) assumes that channel form is physical habitat that results from interactions of channel
substantially adjusted to flow, morphology of an engi- form and flow regime. We use the definition of physical
neered river can be practically independent of flow. This habitat as the three-dimensional structure in which
limits the efficacy of restoring the flow regime alone. On riverine organisms live, modified to emphasize that time
intensively engineered rivers, restoring the hydrograph (frequency, duration, sequence, rate of change) adds a
may restore important flow-related factors like timing of critical fourth dimension (Gordon et al., 1992). Aquatic
floods, water temperature, and turbidity. Without a natu- habitat typically includes physical and chemical char-
ralized morphology, or flow capable of maintaining acteristics of the space occupied by organisms, however
channel-forming processes, however, the hydrologic this paper is confined to physical characteristics, in-
pulses will not be realized in habitat availability. Con- cluding water depth, flow velocity, and substrate. Water
versely, completely natural channel form without tempo- temperature and turbidity are typically also strongly
ral hydrologic cues would also not be expected to support associated with depth and flow velocity. Physical aquatic
natural ecosystem functions. In large, engineered rivers, habitat results from interaction of water with the
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 251

Fig. 1. Missouri River basin, the Lower Missouri River, and locations of modeling reaches and streamflow gaging stations addressed in this study. The
Lower Missouri River is the part of the river downstream of Gavins Point Dam near Yankton, South Dakota, and extending to the junction with the
Mississippi River near St. Louis.

morphology of the stream channel and adjacent flood tems (Gorman and Karr, 1978; Schlosser, 1987; Power
plain. River hydrologic characteristics determine how et al., 1988; Jeffries and Mills, 1990). Factors other than
much water is in the channel, when, and for how long. physical habitat (for example, nutrients, energy, compe-
Hydrologic characteristics can be assessed using the five tition, predation, and contaminants) are also important in
aspects of flow regime listed by Poff et al. (1997): mag- determining ecosystem functions. Most management
nitude, frequency, duration, timing, and rate of change. emphasis on rivers, including the LMOR, however, has
The form of the river channel determines how the water is been on physical habitat because the magnitude of human-
distributed across the channel, thereby creating the spatial induced change to physical habitat is large, and because of
distribution of depth, velocity, and substrate. the direct connection of physical habitat to management
Ecologists generally accept the concept that physical actions (U.S. Fish and Wildlife Service, 2000, 2003;
habitat is the organizational template for aquatic ecosys- Clarke et al., 2003; U.S. Army Corps of Engineers, 2004).
252 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

In addition, we have chosen to simplify this analysis 3. History and management of the Lower Missouri
by concentrating on one measure of physical habitat: River
areas of shallow and slow velocity water, known as
shallow-water habitat (SWH). We focus on SWH Flow regulation began on the Missouri River in the late
because survival and growth of young fishes are 1930s with the construction of Fort Peck Dam in Montana,
associated with the availability of shallow, low velocity but regulation achieved significance with the closure of the
water (Scheidegger and Bain, 1995; Bowen et al., 1998; Missouri River Reservoir System in 1954. The Missouri
Freeman et al., 2001). The bounds of depth and velocity River Reservoir System, consisting of six mainstem dams
vary among studies. For example, Bowen et al. (2003a) (Fig. 1), is now the largest water management system in
use water less than 1-m deep and current velocities less the nation, with nearly 92,500 km3 (73.4 million acre feet)
than 0.25 m/s. Recent habitat analyses on the Missouri of water storage (U.S. Army Corps of Engineers, 2004).
River have used a rather arbitrary definition of SWH as The system is managed for multiple purposes including:
0–5 ft (0–1.5 m) and 0–2 ft/s (0–0.6 m/s) (U.S. Fish and maintenance of navigation flows, flood control, hydro-
Wildlife Service, 2000, 2003; U.S. Army Corps of power, public water supply, recreation, and fish and
Engineers, 2004). Although the complex interactions wildlife resources. A historical perspective on hydrologic
among water properties, substrate, topography, and biota changes is discussed in Galat and Lipkin (2000) and
cannot be fully described by this one class or the hydrologic effects of alternative dam management
dimensions assigned to it, we adopt it for the purpose of scenarios are illustrated in Jacobson and Heuser (2001).
analysis as an operational index of a habitat type that is in Pegg and Pierce (2002) and Pegg et al. (2003) classified
short supply in the river system. the river into hydrologically distinct reaches. These

Fig. 2. Duration hydrographs illustrating effects of flow regulation, Lower Missouri River. Daily flow duration for 100 years of daily data; note
differences in discharge scales. A. Missouri River at Sioux City, Iowa. B. Missouri River at Hermann, Missouri. ROR, run-of-the-river simulation
model; CWCP, current water control plan simulation model, operational control plan 1967–April 2004.
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 253

analyses document substantial alteration to the annual attaining higher stages than they did before extensive
hydrograph below the reservoirs, with generally decreased river engineering, and the reach around Kansas City has
spring pulses and increased summer low flows (Fig. 2A). degraded (Fig. 3; U.S. Army Corps of Engineers, 1996).
The intensity of hydrologic alteration diminishes down- Increased stages for larger discharges have been
stream from the dams as minimally regulated tributaries attributed to channel constriction by revetments and
enter the Missouri River. The lower 590 km downstream levees, whereas channel degradation has been attributed
of the confluence with the Kansas River has a notably to sediment deficits and constriction by wing dikes
altered hydrograph, particularly with respect to low flows, (Pinter et al., 2002). The downstream most 400 km is
but has regained substantial variability (Fig. 2B). characterized by modest channel incision, resulting in
Morphological alterations to the Missouri River decreased stages for low discharges. Channel degrada-
began much earlier than hydrologic alteration. Clearing, tion downstream from Gavins Point dam is a continuing
snagging, and stabilization of the Missouri River began process that limits present-day rehabilitation strategies
in the early 1800s to improve conditions for steamboat and may affect future habitat availability in the LMOR.
navigation (Chittenden, 1903). Most of the rock dikes Since 1989, management agencies for the Missouri
and revetments in the river, however, are the direct result River basin have been involved in the contentious process
of the Missouri River Bank Stabilization and Navigation of revising the operating rules for the mainstem reservoirs,
Project, part of the Pick-Sloan act of 1944 (Ferrell, the Missouri River Master Manual. While a revised
1996). Wing dikes and revetments have stabilized the Master Manual was released in 2004 (U.S. Army Corps of
riverbanks, and narrowed and focused the thalweg Engineers, 2004), management of the river continues to
to maintain a self-dredging navigation channel from be debated. Proposals to restore elements of natural
St. Louis, Missouri, 1200 km upstream to Sioux City, variability to the hydrograph have been particularly
Iowa. The result has been to create a narrow, swift, and controversial. The U.S. Fish and Wildlife Service
deep channel from what was historically a shallow, proposed a revised annual hydrograph with a spring rise
shifting, braided river. Loss of riverine habitat on the and a summer low-flow period (U.S. Fish and Wildlife
LMOR has been estimated as much as 400 km2 (Funk Service, 2000, 2003); the need for a naturalized
and Robinson, 1974). Substantial declines in integrity of hydrograph was supported by a national science review
the ecosystem have been associated with this habitat loss (National Research Council, 2002). In these publications,
(Hesse and Sheets, 1993). Recognition of the scope of specific habitat functions attributed to a spring rise were:
habitat loss has increased interest in rehabilitating parts rejuvenation of emergent sandbar habitats, seasonal
of the Missouri River (Latka et al., 1993). connection of the channel with low-lying flood-plain
In addition to flow alteration and direct effects of the lands, and a spawning cue for native fishes, especially the
Bank Stabilization and Navigation Project, the LMOR endangered pallid sturgeon (Scaphirhynchus albus).
has been adjusting to greatly diminished sediment load Specific habitat functions attributable to the summer
and effects of channel constriction. The channel has low flow included exposure of sandbars for nesting by the
incised 3–5 m in the first 100 km downstream from federally threatened piping plover (Charadrius melodus)
Gavins Point Dam. Further downstream, floods are and federally endangered lest tern (Sterna albifrons) and

Fig. 3. General degradation trends, Lower Missouri River, shown as stage changes 1954–mid-1990's, for given discharges. Data from U.S. Army
Corps of Engineers (1996).
254 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Table 1
Combinations of scenarios explored by two-dimensional hydrodynamic modeling of physical habitat, Lower Missouri River at Hermann, Missouri
Form
1894 Channel form 2000 Channel form
Flow Run-of-the-river flow regime Historical/Historical Historical/Modern
Current water control plan flow regime a Modern/Historical Modern/Modern
a
The current water control plan operated approximately 1967–2004; in 2004 it was replaced with a slightly altered water control plan.

increased area of SWH for rearing of young fishes. As an The ROR model treats the reservoirs as flow-through
alternative, engineered rehabilitation of channel morphol- water bodies and, therefore, produces an estimate of the
ogy has been proposed as an effective means to manage natural hydrograph, with small biases during summer
habitat availability without requiring flow changes (U.S. months when evapotranspiration in the modeled reser-
Army Corps of Engineers, 2003). Hence, the dominant voirs produces somewhat lower discharges than actual
debate on LMOR management has focused on the (U.S. Army Corps of Engineers, 1998a). The data were
tradeoffs between flow and form in balancing ecosystem analyzed to determine duration of flow for every day of
and traditional economic values and services. the year.

4. Approach and methods 4.2. Modern channel model

We address the interaction between flow and form in To model flow discharges through the modern
river rehabilitation by assessing the spatial and temporal channel, we used River2D (Steffler and Blackburn,
distribution of SWH under current conditions and under 2001; version 0.90, July 23, 2002) and its supporting
a historical reference condition. Using flow and form we programs, R2D_Bed and R2D_Mesh. This model code
explore four scenarios (Table 1) that are intended to solves the shallow-water, depth-averaged equations to
provide insight into the interaction between flow and balance mass and momentum on a finite element mesh.
morphology; they are not intended to model realistic River2D handles wetting and drying by converting to
rehabilitation alternatives. We use two-dimensional ground-water flow equations for subsurface flow, and it
hydrodynamic models to create the inventory of SWH explicitly handles transitions between sub- and super-
in the modern (2000) and pre-engineered (1894) critical flow (Steffler and Blackburn, 2001).
channel. Understanding the limitations of models for We modeled steady discharges from 566 to 6790
instream flow (Hudson et al., 2003), we use hydrody- m3 s− 1 (in increments of 283 and 566 m3 s− 1), a range
namic models to indicate trend and sensitivity rather than corresponding to 97 — 2% duration of flow under the
to predict specific biotic responses. CWCP and 96 — 4% flow duration under the ROR.
This range includes low flows to just over bank (with
4.1. Hydrologic model data analysis present-day channel morphology).
Channel bathymetry was obtained by the U.S. Army
Hydrologic data used in the analyses came from the U. Corps of Engineers in 1998 and 1999 using an
S. Army Corps of Engineers Daily Routing Model for the echosounder to collect depth data and was supplemented
Missouri River (U.S. Army Corps of Engineers, 1998a). with high-resolution depth data that we collected in
The modeled flows were synthesized from historical data 2000–2001 at specific locations. A combination of cross-
on tributary inflows, calculations of streamflow deple- section and feature-based design was used. Cross-sections
tions because of evapotranspiration and consumptive use spaced approximately every 20 m (5% of channel width)
of water, and modifications of outflows according to achieved an average data density of 6 points per 100 m2.
water-control rules that comprise flow-management We used a high-resolution depth sounder and a
alternatives. The model reproduces how reservoirs differentially corrected global positioning system to
would be managed under a set of water control rules, map bathymetry that resulted in positional errors
given the actual range of variability of historical inflow estimated at ±0.75 m and depths ±0.07 m (Jacobson et
data 1898–1998. al., 2002). Depths were converted to elevations using
Modeled flows were used for the streamgage station surveyed elevations of the water surface, and depth points
at Hermann, Missouri for two management scenarios, were merged with U.S. Army Corps of Engineers channel
the current water control plan (CWCP) that was in place bathymetry and a 5-m-cell flood-plain digital elevation
until April 2004 and the run-of-the-river model (ROR). model (DEM, U.S. Army Corps of Engineers, 2000).
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 255

Additional elevation points were surveyed on wing- tional mesh was derived from the 5-m grid and edited
dike crests by total station. Elevation data were gridded in River2D. The substrate of the reach was mapped
to 5-m cells using a kriging algorithm. The computa- using hydroacoustic methods (Jacobson et al., 2002),

Fig. 4. A. Modern (2000) and historical (1894) channel maps, Lower Missouri River at Hermann, Missouri. Modern channel is extent of modern
2-dimensional hydrodynamic model. B. Schema for developing channel cross-sections as constrained by historical map. Extent of cross-sections
indicates extent of historical 2-dimensional model.
256 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Fig. 5. Modeled and measured water-surface elevations. Measured elevations are validation dataset.

and was used to initialize roughness heights in the measured and modeled elevations of the water surface
model. for comparison. Calibration of the model was considered
Boundary conditions, stage–discharge curves at adequate. In addition to close agreement between
upstream and downstream ends of the reach, were measured and modeled elevations of the water surface,
developed by surveying water surfaces corresponding models were considered successful if they achieved a
with discharges obtained from the U.S. Geological low net outflow (less than 5% of the flow was
Survey streamflow gaging station 06934500, Missouri unaccounted by the model) over run times sufficient to
River at Hermann, Missouri, in the middle of the achieve a steady state (usually greater than 10,000 time
modeled reach (Fig. 4). Other parameters required by steps). Model results were also evaluated for whether
River2D include, 1 and 2 (coefficients used to calculate they realistically reproduced known flow patterns —
Boussinesq eddy viscosity), an upwinding coefficient to such as eddies downstream of wing dikes. In some cases,
parameterize the finite element solving scheme, and modeled local instabilities in the flow field were
ground-water transmissivity and minimum depth coeffi- accepted if they affected only small areas and did not
cients for aiding in wetting and drying calculations at the substantially affect habitat area calculations.
wetted boundary. Eddy viscosity and upwinding para-
meters were kept as default values (0, 0.5, and 0.5). 4.3. 1894 Channel model
Sensitivity analyses documented that modeled velocities
in the main channel and recirculating eddies down- Bathymetric and discharge data do not exist to
stream of wing dikes were generally insensitive to the construct a similarly calibrated and validated hydrody-
choice of viscosity parameters for the eddy. Ground- namic model for the 1894 river. A variety of historical
water transmissivity was set at 0.01 to minimize ground- information is available, however, to construct a sta-
water discharge. tistical model of the historical river. Combined with
Elevation profiles of the water surface were used as calibrated parameters of flow from the modern channel,
the primary means of calibration. Elevations of the water and a few reasonable assumptions, the statistical model
surface for discharges 1245–4560 m3 s− 1 were surveyed allows construction of a reference hydrodynamic model.
at two benchmarks, and stage was available from the The value of the reference model, albeit based on a series
streamgage in the center of the reach. Of 14 measured of assumptions, is that it provides a way to explore the
elevations of the water surface, seven were used for historical reference condition using the same quantities
calibration (by adjusting roughness height) and seven as the modern condition. Comparison to the historical
were used for validation of model results. Fig. 5 shows reference condition can provide insight into how
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 257

Fig. 6. Historical channel cross-section synthesis. A. Fitting of cross-sections to Weibull 4-parameter distribution, parameterized by channel width,
width : depth ratio, thalweg position, and shape coefficient. B. Schema for construction of compound channel sections with maximum of three
channels and two islands.

ecosystem drivers, in this case, channel form and flow (depth in the thalweg), the width of the channel, the
regime, provide a template for ecosystem functions. position of the maximum depth along the cross-section
Planform maps from the late 1800s (Missouri River (thalweg position) and a shape parameter.
Commission, 1894) are the foundation of the channel
model (Fig. 4). The 1894 planform map was used to
1−d 
1 d−1
d−1 d x−b d−1 d
identify and digitize key linear features: banks, thalweg, Depth ¼ −a þ
d c d
and island axes. This is the most subjective and inter- "

1 d #
pretive step in the process. Cross-sections were laid out x−b d−1 d d−1

e − þ þ
at 50 m spacing perpendicular to the main thalweg. c d d
Coordinates of the points of intersection of cross-
sections with banks, thalweg, and bar medial axes were
attributed with a geomorphic significance (Fig. 4B) and Where:
extracted from the geospatial database. Channel mor-
phology was generalized so the most complicated case a maximum amplitude, or depth of thalweg,
consisted of three thalwegs and two islands (Fig. 6); b distance from bank to thalweg,
channel sections could also consist of two thalwegs and c channel width,
one island, or one thalweg with no islands. d shape coefficient, and
The 1894 maps contain a great deal of detail on x distance along cross-section
channel width and form, but do not have any depth data.
Another series of maps from 1920 (War Department, The sample of 1920 channel cross-sections provided
1920) present sparse sounding data (Fig. 7). Cross- an estimate of pre-engineered width : maximum depth
sections from non-engineered reaches were selected and ratio (74.2) and the shape coefficient (1.5). From 1894
compiled to develop a statistical model for channel bankfull channel widths, we calculated maximum depth
cross-section shape. The typical channel shape can be and from the interpreted position of the thalweg, we
modeled as an upside-down, 4-parameter Weibull calculated distance from the bank to the thalweg.
distribution (Fig. 6A, Eq. (1); Weibull, 1951). The Computer code automated the calculations to generate
parameters of the distribution relate to maximum depth depths at 120 points along each cross-section. The
258 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Fig. 7. Channel morphology, Lower Missouri River in 1920. A. Example of 1920 planimetric map of part of the Lower Missouri River with sparse
sounding data. B. Examples of channel cross-sections assembled from 1920 sounding data, multiple locations along LMOR.

points provided a dense mesh from which a depth grid It was impossible to calibrate or validate the 1894
(depths as negative numbers) was generated. An model; the validity of the model depends on the 2000
elevation grid was calculated by adding the depth grid calibration of roughness height and the many assump-
to a grid representing the mean bankfull elevation of the tions that went into constructing the topographic mesh.
present flood plain. This last calculation was based on Early 19th century paintings (e.g., Karl Bodmer; Josyln
the assumption that valley slope and bankfull elevation Art Museum, 1984) and the long history of steamboat
have not changed significantly since 1894. wrecks on the Missouri River (Chittenden, 1903)
A computational mesh was generated from the document that large woody debris (LWD) was common
synthetic elevation grid, and the same series of discharges in the channel. LWD certainly created hydraulic rough-
were modeled for the synthetic 1894 condition (historical) ness that is not captured in the model, and it was
as for the 2000 condition (modern). The calibrated probably responsible for fine-scale topographic features
roughness height (similar to medium sand) was retained in the bed that also are not represented in the
from the 2000 model runs. Stage–discharge relations topography.
were generated from the measurement data for discharge The degree to which the synthetic morphology of the
at the Hermann streamgage using 1928–1935, prior to channel agrees with the 1894 reality depends on several
extensive channel engineering in the Hermann reach. The interpretations, statistical models, and assumptions. As
relations were extended to the upstream and downstream such, it is not expected to be an accurate depiction of the
margins using a streamwise slope. 1894 condition, but it is expected to be realistic. Its
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 259

value lies in its ability to show general conditions of the the SWH class: 0–1.5 m, depth and 0–0.75 m/s
pre-modern reference condition. Because no other current velocity.
quantifiable reference condition exists for large rivers Grids with 5-m cell size were constructed from
like the Lower Missouri, we believe these assumptions 2-dimensional model outputs of depth and velocity.
are justified. Cells meeting depth and velocity criteria were classified
as SWH in a third grid. The SWH grid was analyzed for
4.4. Habitat assessment total area, total edge, and mean patch area to char-
acterize some basic spatial statistics (McGarigal and
Simulation models of hydrodynamic habitats can Marks, 1995).
produce a continuum of depths and depth-averaged
velocities for a modeled area at each modeled discharge. 5. Results
Many different combinations of depth and velocity may
be important habitat characteristics for some species or Results of hydrodynamic modeling were evaluated in
some life stages, and a complete assessment of habitat terms of the relations between discharge and habitat area
should probably consider a myriad of measures of in the historical and modern channel, and by looking at
spatial relations of habitat patches as well as temporal the distribution of SWH during the year under historical
measures of sequence, duration, and timing. For the and modern flow regimes. Results from the Hermann
purposes of this analysis, however, we look at only at reach are compared to upstream reaches to assess affects

Fig. 8. Two-dimensional hydrodynamic modeling results for selected discharges, modern channel of the Lower Missouri River at Hermann, Missouri,
showing the mapped distributions of depth, current velocity, and shallow-water habitat.
260 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

of the alteration of more severe flow and spatial Lewis and Clark Corps of Discovery commented on
variability within the LMOR. high velocities in the Missouri River in 1804: “…the
Boat run on Logs three times to day, owing [to] her
5.1. Shallow-water habitat in the historical and modern being too heavily loaded a Sturn… I saw a number of
channel Goslings to day on the Shore, the water excessively rapid,
and Banks falling in.” May 15, 1804, William Clark
Hydraulic conditions in the modern channel are gen- (Moulton, 2002).
erally more uniform than in the 1894 channel (Figs. 8 Total SWH is much greater over all discharges under
and 9), with the exception of deep, slow areas associated historical conditions (Fig. 10A, Table 2). Unlike the
with wing-dike scours. Maximum current velocities are modern channel in which SWH increases with decreas-
greatest in the 1894 channel in the discrete area where ing discharge (for flows less than bankfull), SWH in the
multiple channels converge in a relatively narrow reach historical channel increases with increasing discharge.
(Fig. 9). Whereas it is commonly accepted that chan- The continuous and gradually varying distribution of
nelization of the Missouri River increased mean velocity topographic surfaces from the thalweg to the flood plain
(Latka et al., 1993), spatial variability was clearly much supported abundant SWH at all discharges. Because of
greater before channel engineering and discrete areas of the broadly convex bars, more of the surface is in-
very high velocity existed in the historical river. The undated with shallow depths as discharge increases.

Fig. 9. Two-dimensional hydrodynamic modeling results for selected discharges, historical reconstructed channel of the Lower Missouri River at
Hermann, Missouri, showing the mapped distributions of depth, current velocity, and shallow-water habitat.
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 261

two flow regimes and two channel forms (Table 1). The
median area of 100 years of daily values for each day of
the year was plotted over the year (Fig. 11A). With the
modern channel form, SWH is scarce during all times
of the year for the modern and historical flow regimes
(Fig. 11A). SWH area increases by a factor of 3–7 for
the historical channel form for both flow regimes,
demonstrating the strong affect of channel form on the
availability of habitat. For historical and modern flow
regimes, within-year variability is greater for the
historical channel form (within-year ranges of 7.4 and
11.4 ha) compared to the modern (within-year ranges of
1.7 and 2.0 ha). The greatest within-year variability
results from the historical channel form and historical
flow regime (11.4 ha). The historical form/historical
flow combination also contains a discrete peak in habitat
availability during late June that results from the larger,
second mode of the annual flow pulses (Fig. 2). This
habitat availability spike would have been a persistent
event (occurring at least 50% of the time) during which
high bar and low flood-plain surfaces were connected to
the main channel.

5.2. Comparison to other segments of the Lower

Missouri River

Because the Hermann, Missouri, segment has been

highly affected by channel engineering, but has less affect
of flow regulation compared to segments closer to the
Fig. 10. Relations between discharge and selected spatial variables for mainstem reservoirs, we also considered the situation
shallow-water habitat, 1894 and 2000 channels, Lower Missouri River where the hydrograph was more intensively altered. Daily
at Hermann, Missouri. A. Total shallow-water habitat (SWH) area and discharges at the Sioux City, Iowa, streamgaging station
discharge. B. Average size of SWH area. C. Total edge length for
SWH.
(Figs. 1, 2) were multiplied by the ratio of mean annual
flow at Hermann to the mean annual flow at Sioux City to
synthesize a flow regime for Hermann that is as
Average size of SWH patches is also greater at all intensively regulated as that at Sioux City. Similar to
discharges in the historical channel than the modern the previous analysis, the distribution of SWH availability
channel. Patches of SWH in the modern channel are during the year was calculated from the time series of
fragmented whereas those in the historical channel are SWH for the synthetic flow (Fig. 11B). As in the previous
large and continuous (Figs. 8–10B). Historical patches analysis, the historical form provides much more SWH
are elongate compared to the modern patches, and have than the modern form over the entire year. The
greater total edge length compared to the modern con- combination of modern form/historical flow presents a
dition (Fig. 10C; Table 2). These conditions may have peak of habitat availability associated with the late June
favored species with affinities for habitat edges. Edge flood peak as the overbank area just begins to be inundate.
density (total edge length divided by total patch area) is The regime for the modern form/modern flow has a small
substantially higher for the historical condition for flows peak of availability related to flow reductions in March.
up to 5660 m3 s− 1, indicating that the historical channel Although regimes for modern and historical flows
had large quantities of productive edges even for the provide abundant SWH during the entire year with the
large area of SWH present. historical channel form, the within-year variability of
SWH abundance over the season was assessed by SWH under the historical flow regime is substantially
interpolating total SWH area from the discharge–SWH greater than that under the modern flow regime (within-
relation, for one hundred years of daily discharges for year range of 21.4 ha compared to 7.7 ha).
 262
Table 2
Summary patch statistics for shallow-water habitat (SWH) in the modeled modern and historical channel, Lower Missouri River, at Hermann, Missouri

R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Modeled discharge Modern channel Historical channel
(m3/s)
Area Linear density Mean patch Total patch Total edge Area Linear density Mean patch Total patch edge Total edge density
(ha) (ha/km of river) size (ha) edge (m) density (m/ha) (ha) (ha/km of river) size (ha) (m) (m/ha)
850 28.6 5.0 0.2 36,100 156.8 63.4 7.9 1.44 45,600 155.3
1140 16.6 2.9 0.09 31,100 132.3 77.8 9.7 2.43 52,200 157.2
1420 12.0 2.1 0.06 26,400 110.6 88.9 11.1 2.78 56,300 152.0
1700 9.4 1.6 0.05 22,900 94.9 100.1 12.5 2.86 62,400 152.3
1990 8.3 1.4 0.04 20,900 85.8 118.1 14.8 2.62 70,700 154.5
2160 7.8 1.4 0.04 20,400 83.3 121.3 15.2 2.64 72,400 150.6
2270 7.3 1.3 0.04 19,400 79.3 124.8 15.6 2.35 73,500 148.0
2550 7.0 1.2 0.03 18,700 75.9 125.4 15.7 2.37 73,300 140.0
2830 6.7 1.2 0.03 18,200 73.2 129.5 16.2 2.64 75,600 137.2
3400 6.1 1.1 0.03 17,100 68.3 130.3 16.3 2.41 77,100 134.0
3970 5.8 1.0 0.03 16,600 65.6 133.7 16.7 2.62 78,500 130.6
4530 5.2 0.9 0.02 15,200 59.7 149.3 18.7 3.83 79,100 118.8
5100 4.9 0.9 0.02 14,300 55.7 165.7 20.7 4.6 81,000 112.0
5660 6.9 1.2 0.03 15,900 61.3 193.0 24.1 5.36 80,300 101.4
5950 14.3 2.5 0.06 21,300 79.5 223.0 27.9 4.65 73,000 83.8
6510 33.5 5.8 0.16 30,700 101.2 231.7 29.0 4.07 61,800 63.9
7080 31.2 5.4 0.16 27,600 91.0 217.7 27.2 4.03 61,000 61.2
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 263

Fig. 11. Median daily distributions of shallow-water habitat during the year using modern and historical channel discharge-habitat relations,
combined with modern and historical flow regimes. A. Shallow-water habitat at Hermann, Missouri using historical and modern flow regimes. B.
Shallow-water habitat at Hermann, Missouri using synthetic modern and historical flow regimes reflecting flow regulation similar to that affecting the
Missouri River at Sioux City, Iowa.

We also addressed how the Hermann reach compared 1-dimensional model case, only the depth criterion for
to others along the LMOR by synthesizing the results SWH was used.
of hydraulic modeling from seven additional reaches Four distinct channel forms result in four distinct
(Figs. 1, 12). This analysis establishes that the Hermann discharge–SWH area curves for the reaches (Fig. 12).
reach is representative of much of the altered LMOR, The Blair (Nebraska), Nebraska City (Nebraska),
but that systematic deviations exist because of res- Doniphan (Kansas), and Baker (Missouri) reaches
toration activities or other exogenous influences on have discharge–SWH relations similar to the modern
channel morphology. For six of these reaches we Hermann reach because they are intensively engineered
used 2-dimensional results of the hydrodynamic model reaches with little diversity in elevation. In these
developed by the Army Corps of Engineers (U.S. reaches, SWH becomes abundant only at extremely
Army Corps of Engineers, 1998b); the seventh was a low flows. The Rocheport (2-d model) and Lisbon-
1-dimensional hydraulic model reach (Jacobson et al., Jameson (1-d model) reaches provide greater abun-
2004). In the 2-dimensional models, the depth and dance of SWH over a similar range of discharges
velocity classes of the U.S. Army Corps of Engineers because they have greater diversity of elevation. The
were combined to match the SWH definition; in the Rocheport reach is located on a bend with an extensive
264 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Fig. 12. Relations between normalized discharge and shallow-water habitat (SWH) hectares per kilometer of river in reaches along the Lower
Missouri River. Discharge was normalized by dividing by mean annual discharge at the nearest gaging station. Closed symbols denote the engineered
channel reaches; grey symbols denote rehabilitated or abnormally wide channel reaches; open symbols denote reconstructed historical reach or
minimally engineered reaches.

pointbar, sandbar complex. The Lisbon-Jameson reach abundant information from which ecological functions
has been extensively rehabilitated with substantial can be inferred (Power et al., 1988). Strictly for the
widening of the channel and a flow-through a side- purpose of this analysis, we have adopted the definition
channel chute (Jacobson et al., 2004). and assumed ecological importance of SWH as an
The Vermillion, South Dakota reach is in the non- indicator of ecological function (for example, Schei-
channelized segment of the Missouri River just degger and Bain, 1995; Bowen et al., 1998). Historical
downstream of Gavins Point dam and upstream of descriptions of the pre-engineered river and the effects of
the navigation channel at Sioux City, Iowa. While bank stabilization and navigation structures clearly
maintaining a braided-anastomosing morphology sim- demonstrate that shallow-water areas have been greatly
ilar to the historical Missouri River, the channel has diminished (Funk and Robinson, 1974; Hesse and
degraded as much as 5 m as a result of a diminished Sheets, 1993). Moreover, shallow, slow water has been
sediment load (U.S. Army Corps of Engineers, 1996). shown to be important as rearing and nursery habitat for
Over the modeled range of discharges, SWH peaks at many young fishes (Gozlan et al., 1998; Robinson et al.,
about 0.75 times the mean annual discharge as flow 1998; Schiemer et al., 2000). Shallow, slow water areas
covers abundant sandbars, and then diminishes as with associated fine sediment also might be expected to
increasing flow encounters the near-vertical banks. trap and retain organic matter in an engineered system
Modeled discharges did not include discharges in excess designed to transport material efficiently downstream.
of 2.0 times the mean annual discharge, which Empirical evidence for the ecological functioning of
presumably reach the former flood plain and produce SWH on the LMOR relative to other areas is limited
extensive SWH area. The Vermillion discharge–SWH because of little comparative evaluation. Limnological
relation differs from the reconstructed Hermann histor- studies on the Missouri River documented that turbidity
ical curve because the historical channel form at generally limits light penetration and consequently rates
Hermann was not incised, and included a gradual of algal growth; as mean depths decrease with addition
continuum of elevations from the thalweg to bankfull. of SWH, more in-channel primary productivity could be
expected (Knowlton and Jones, 2000). Large numbers
6. Discussion: ecological and restoration and high species richness of fishes, particularly young-
implications of-year (age-0), have been reported from Missouri River
channel and floodplain SHW (Pflieger and Grace, 1987;
Although an incomplete descriptor of complex Brown and Coon, 1994; Tibbs and Galat, 1997; Grady
aquatic ecosystems, physical habitat models contain and Milligan, 1998). The one systematic sampling of
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 265

benthic fishes from multiple habitat types throughout tion on habitat requirements of key species and at
the entire riverine Missouri showed that catches of most multiple life stages.
benthic fishes were highest in inside bends and The critical and persistent policy and scientific
connected secondary channel mesohabitats (Berry et question on the LMOR is the extent to which SWH
al., 2004), both of which contain large areas of SWH rehabilitation can substitute for, or will interact with, a
(Galat et al., 2001). Ninety percent of shorebird use naturalized hydrograph. Two attributes of the natural-
along LMOR floodplain wetlands occurs on exposed ized hydrograph have been identified as particularly
saturated mudflats and water less than 5 m3 deep important: 1) a spring rise and 2) a summer low flow
(McColpin, 2002). SWH in a constructed side-channel (U.S. Fish and Wildlife Service, 2000). Channel func-
chute near Overton, Missouri, was shown to be ex- tions attributed to the spring rise include: a) a spawning
tremely efficient in recruiting and retaining large woody cue for native fishes, including the pallid sturgeon, b)
debris (Jacobson et al., 2004). sandbar formation or alteration for shorebirds (including
Timing of the appearance of maximum area of SWH the listed interior least tern and piping plover), and c)
may be as important to recruitment of riverine biota as seasonal connection with flood plain or other low-lying
the total amount available. Maximum amount of SWH surfaces to allow for exchange of nutrients, energy, and
in the historical flow-historical channel form scenarios biota. The functions attributed to the summer low flow
at Hermann using both Hermann and Sioux City flow are: a) sandbar availability for nesting and migrating
regimes was in late June (Fig. 11). Galat et al. (1998) shorebirds, and b) SWH availability in mid-late summer
reported that this was near the end of the annual for fish nursery. Galat et al. (1998) illustrated the relation
spawning season for many Missouri River fishes and, of many seasonally important bioevents to river stage
thus, ample in-channel SWH would have been available and precipitation for LMOR floodplain wetlands.
as nursery for recently hatched larvae of obligate fluvial Our conclusion that total SWH availability is highly
species (e.g., Scaphirhynchus sturgeons, Macrhybopsis sensitive to channel form addresses only one of these
chubs, Hybognathus minnows). Additionally, late June functions and should not be seen as a conclusion
was also at the end of seed dispersal by pioneer sandbar diminishing the value of a naturalized flow regime. This
colonizing cottonwoods (Populus deltoides) and wil- analysis also documents that timing of events of habitat
lows (Salix spp.) (Mazourek, 1998). availability has to be controlled by flow regime. A
In a modeling study of habitat availability in comprehensive perspective on ecosystem integrity
modified and unmodified reaches of the Upper needs to include habitat types and availability over
Yellowstone River, Bowen et al. (2003b) documented many life stages of many species and on functional
similar results: bank-stabilized reaches of the river attributes operating at intra- and inter-annual temporal
produced smaller areas of shallow, slow water at high scales. Rehabilitation activities to create SWH have
discharges compared to unmodified reaches. The been proposed to reduce the need for mid-summer low
authors inferred that lack of rearing habitat during flow, but would probably increase the efficacy of spring
spring flows would likely diminish juvenile fish rises in seasonally connecting low-lying flood plain,
abundance. similar to the historical case illustrated at Hermann.
A functioning aquatic ecosystem is much more The challenge in rehabilitating large, intensively
complex than the presence or extent of one habitat engineered river systems like the Lower Missouri is to
unit. SWH, resulting from channel rehabilitation, exists design a combination of channel form and flow regime
as discrete zones along an intensively engineered river that results in socially acceptable improvements to
and fragmentation of the physical habitat template may ecosystem functions. Most of the LMOR is intensively
diminish its ecological value. Whereas SWH may engineered and provides discharge–SWH relations
provide some of the functions necessary for some life similar to Hermann for which extensive SWH can
stages of many aquatic species, most biota require a only be achieved through very low flows or overbank
variety of physical habitats for different life stages flows. Some stakeholders consider very low flows to be
(Fausch et al., 2002). For example, young (3–6-year incompatible with uses of the river for navigation,
old) pallid sturgeon have been associated with relatively industrial and municipal water supply, and cooling of
deep and swift habitats during summer and early fall water from power plants. Discharge–SWH relations at
(Elliott et al., 2004) and deep, slow areas for over Rocheport and Lisbon-Jameson indicate that channel
wintering (Grady et al., 2001; Jacobson and Laustrup, forms with abundant SWH over a broad range of
2002). A more complete understanding of abiotic–biotic discharge can exist in the LMOR. Rehabilitated reaches
relations in the LMOR awaits more complete informa- like Lisbon-Jameson lie between the intensively
266 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

engineered condition and the historical reference increasing discharge. The modern channel SWH occurs
condition, and illustrate a practical endpoint of physical in smaller, more fragmented patches than the historical
rehabilitation wherein SWH is abundant over a range of channel. Combining discharge–SWH area relations
discharges, thereby decreasing reliance on specific flow with modern and natural-hydrograph flow-regimes
targets to achieve physical habitat goals. It is not known indicates the relative efficacy of flow and form regime
at this time, however, if achieving habitat rehabilitation in achieving SWH during the year. The historical
goals in the absence of flow naturalization will yield channel form produces three to seven times the SWH
ecological benefits acceptable to stakeholders. area of the modern channel regardless of flow regime.
The value of this analysis lies in identifying the The effect of flow regime is evident in increased within-
large influence of rehabilitation of form on SWH year SWH variability with the natural flow regime,
availability in rivers where channel form is intensively including significant seasonal peaks of SWH associated
engineered. If rehabilitation of channel form can with spring flooding. A synthetic hydrologic record
supply sufficient SWH for essential ecosystem func- representing hydrology similar to that just downstream
tions during mid to late summer months, managers of the Missouri River mainstem dams illustrates that
would have additional flexibility in meeting multiple channel form persists as the dominant control on within-
stakeholders' desires. channel SWH availability even where hydrograph
regulation by dams is more severe.
7. Summary Comparisons of discharge–SWH habitat relations
along the LMOR indicate three conditions of the
On large, intensively engineered rivers, like the modern channel that contrast with the historical refe-
Lower Missouri, the physical habitat template is rence condition. Most of the LMOR has discharge–
largely determined by the interaction of channel SWH relations similar to the modeled reach at
form and flow regime. In contrast to natural rivers Hermann, Missouri, in which maximum SWH area is
where form is equilibrated to a range of flows, achieved by extremely low flow or overbank flows;
engineering structures decouple form and flow for the intermediate flows within normal river regulation
lifetime of the engineering structures, thereby making parameters achieve little SWH in these reaches be-
them substantially independent and amenable to cause of the lack of diversity of elevation in the
separable management strategies. We evaluated the engineered channel. Two reaches show substantially
interaction between flow and form by modeling the greater amounts of SWH over a wide range of flows
modern channel morphology and reconstructed histor- because of either existence of large pointbar sandbar
ical channel morphology using the modern and complexes or intensive channel rehabilitation. SWH
historical flow regimes. The four combinations of area is less sensitive to discharge in these reaches,
flow and form cover the range of possible conditions indicating a condition where substantial SWH can be
for river rehabilitation from the present condition to achieved with a variety of flow regimes. The up-
the historical reference condition. stream-most reach of the LMOR is unchannelized and
This analysis focused on one very important habitat contains many of the complex channel features of the
class: shallow, slow velocity water (SWH) defined by historical channel, but it is incised substantially below
natural resource agencies as depths between 0 and 1.5 the pre-dam flood plain. As a result, the discharge–
m, and current velocities between 0 and 0.75 m/s. This SWH relation peaks at moderate flow and then de-
habitat class is an incomplete measure of ecosystem creases substantially as water over mid-channel bars
function, but serves as an index of a critical habitat unit deepens but does not extend to overbank flow. The
nearly lost because of river regulation and engineering. reconstructed historical channel at Hermann has a
SWH is especially important as nursery area for young continuously and gradually varying distribution of
fishes. channel elevations from the thalweg to the top of bank.
Two-dimensional hydrodynamic models for the mo- In contrast to the modern and rehabilitated channels,
dern and reconstructed historical channels at Hermann, this form results in a marked increase in SWH with
Missouri, indicate substantial differences between the increasing discharge.
two channels in total availability and spatio-temporal Examples like the Rocheport and Lisbon-Jameson
characteristics. In the modern channel SWH is maxi- reaches indicate that practical rehabilitation of channel
mized at extremely low flows and in overbank flows, form on the LMOR can produce habitat conditions
whereas the historical channel had substantially more intermediate between the modern, engineered channel
SWH at all discharges and SWH increased with and the historical reference channel. The engineering
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 267

challenge of rehabilitation will be to achieve the benefits Calow, P., Petts, G.E., 1992. The Rivers Handbook. Blackwell
of increased diversity of habitat while maintaining Scientific Publications, Oxford.
Chittenden, H.M., 1903. History of Early Steamboat Navigation on the
sediment transport in the navigation channel. Missouri River. Francis P. Harper, New York.
Focus on a single class of habitat cannot adequately Clarke, S.J., Bruce-Burgess, L., Wharton, G., 2003. Linking form and
describe the many relations that determine ecosystem function: towards an eco-hydromorphic approach to sustainable
functions. The spatial and temporal distributions of river restoration. Aquatic Conservation: Marine and Freshwater
Ecosystems 13, 439–450.
physical habitat determined by flow and form,
Elliott, C.M., Jacobson, R.B., DeLonay, A.J., 2004. Physical
however, determine the template upon which river aquatic habitat assessment, Fort Randall segment of the
ecosystem processes play out (Calow and Petts, 1992). Missouri river, Nebraska and south Dakota. U.S. Geological
Among the continuum of physical conditions in the Survey Open-File Report 2004-1060. 80 p. On CD-ROM or
habitat, SWH is a useful indicator of the diversity of online at: http://infolink.cr.usgs.gov/RSB/USGS_OFR_2004-
habitat in multipurpose rivers where SWH has been 1060/index.htm.
Fausch, K.D., Torgersen, C.E., Baxter, C.V., Li, H.W., 2002.
diminished by flow regulation and engineering. Hence, Landscapes to riverscapes: bridging the gap between research
recognition of the importance of channel form relative and conservation of stream fishes. Bioscience 52, 483–497.
to flow regime in determining SWH should contribute Ferrell, J., 1996. Soundings — 100 Years of the Missouri River
increased flexibility toward achieving socially accept- Navigation Project. U.S. Army Corps of Engineers, Kansas City.
able rehabilitation of ecosystems in large river systems. Freeman, M.C., Bowen, Z.H., Bovee, K.D., Irwin, E.R., 2001.
Flow and habitat effects on juvenile fish abundance in natural
and altered flow regimes. Ecological Applications 11,
179–190.
Acknowledgements Friedman, J.M., Scott, M.L., Auble, G.T., 1997. Water management
and cottonwood forest dynamics along prairie streams. Ecological
This is a contribution from the Missouri Cooperative Studies 125, 49–71.
Funk, J.L., Robinson, J.W., 1974. Changes in the Channel of the
Fish and Wildlife Research Unit (U.S. Geological Lower Missouri River and Effects on Fish and Wildlife. Missouri
Survey, Missouri Department of Conservation, Univer- Department of Conservation, Jefferson City, Missouri.
sity of Missouri, and Wildlife Management Institute Galat, D.L., Lipkin, R., 2000. Restoring the ecological integrity of
cooperating). We thank Joanna Reuter, Mark Laustrup, great rivers: historical hydrographs aid in defining reference
and Gary D'Urso for extensive help with field work and conditions for the Missouri River hydrosystem. Hydrobiologia
422/423, 29–48.
data reduction. The manuscript benefited substantially Galat, D.L., Fredrickson, L.H., Humburg, D.D., Bataille, K.J., Bodie,
from comments from Doug Shields, Michael Urban, and J.R., Dohrenwend, J., Gelwicks, G.T., Havel, J.E., Helmers, D.L.,
an anonymous reviewer. Hooker, J.B., Jones, J.R., Knowlton, M.F., Kubisiak, J.,
Mazourek, J., McColpin, A.C., Renken, R.B., Semlitsch, R.D.,
1998. Flooding to restore connectivity of regulated, large-river
References wetlands — natural and controlled flooding as complimentary
processes along the Lower Missouri River. Bioscience 48,
Bayley, P.B., 1995. Understanding large river-floodplain ecosystems. 721–733.
Bioscience 45, 153–158. Galat, D.L., Wildhaber, M.L., Dieterman, D.J., 2001. Spatial Patterns
Berry, C.R., Wildhaber, M.L., Galat, D.L., 2004. Fish Distribution and of Physical Habitat, vol. 2. Population Structure and Habitat Use of
Abundance, vol. 3. Population Structure and Habitat Use of Benthic Benthic Fishes Along the Missouri and Lower Yellowstone Rivers.
Fishes Along the Missouri and Lower Yellowstone Rivers. U. S. U. S. Geological Survey, Cooperative Research Units, University
Geological Survey, Cooperative Research Units, South Dakota State of Missouri, Columbia. Online at: http://www.nwo.usace.army.
University, Brookings. Online at: http://www.nwo.usace.army.mil/ mil/html/pd-e/benthic_fish/benthic_fish.htm.
html/pd-e/benthic_fish/benthic_fish.htm. Gordon, N.D., McMahon, T.A., Finlayson, B.L., 1992. Stream
Bowen, Z.H., Freeman, M.C., Bovee, K.D., 1998. Evaluation of Hydrology: an Introduction for Ecologists. John Wiley and Sons,
generalized habitat criteria for assessing impacts of altered flow Chichester, England.
regimes on warmwater fishes. Transactions of the American Gore, J.A., Shields Jr., F.D., 1995. Can large rivers be restored?
Fisheries Society 127, 455–468. Bioscience 45, 142–152.
Bowen, Z.H., Bovee, K.D., Waddle, T.J., 2003a. Effects of flow Gorman, O.T., Karr, J.R., 1978. Habitat structure and stream fish
regulation on shallow-water habitat dynamics and floodplain communities. Ecology 59, 507–515.
connectivity. Transactions of the American Fisheries Society 132, Gozlan, R.E., Mastrorillo, S., Dauba, F., Tourenq, J.-N., Copp, G.H.,
809–823. 1998. Multi-scale analysis of habitat use during late summer
Bowen, Z.H., Bovee, K.D., Waddle, T.J., 2003b. Effects of channel for 0+ fishes in the River Garonne (France). Aquatic Sciences
modification on fish habitat in the Upper Yellowstone River: U.S. 60, 99–117.
Geological Survey Open-File Report 03-476, 30 p. + appendices. Grady, J.M., Milligan, J., 1998. Status of Selected Cyprinid Species at
Brown, D.J., Coon, T.G., 1994. Abundance and assemblage structure Historic Lower Missouri River Sampling Sites. Columbia
of fish larvae in the lower Missouri River and its tributaries. Fisheries Resources Office, U.S. Fish and Wildlife Service,
Transactions of the American Fisheries Society 123, 718–732. Columbia, Missouri.
268 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269

Grady, J.M., Mauldin, L., Davison, B., Milligan, J., 2001. Missouri National Research Council, 2002. The Missouri River Ecosystem —
River Pallid Sturgeon Survey Route 19 Bridge Replacement Exploring the Prospects for Recovery. National Academies Press,
Project, Hermann, Missouri. Columbia Fisheries Resource Office, Washington, DC.
U.S. Fish and Wildlife Service, Columbia, Missouri. Pegg, M.A., Pierce, C.L., 2002. Fish community structure in the
Hesse, L.W., Sheets, W., 1993. The Missouri River hydrosystem. Missouri and lower Yellowstone rivers in relation to flow
Fisheries 18, 5–14. characteristics. Hydrobiologia 479, 155–167.
Hudson, H.R., Byrom, A.E., Chadderton, W.L., 2003. A critique of Pegg, M.A., Pierce, C.L., Roy, A., 2003. Hydrological alteration along
IFIM—instream habitat simulation in the New Zealand context. the Missouri River basin: a time series approach. Aquatic Sciences
Science for Conservation 231. 65, 63–72.
Jacobson, R.B., Heuser, J.L., 2001. Visualization of Flow Alternatives, Pflieger, W.L., Grace, T.B., 1987. Changes in the fish fauna of the
Lower Missouri River. U.S. Geological Survey Open-File lower Missouri River, 1940–1983. In: Matthews, W.J., Heins, D.C.
Report 02-122. Online at: http://www.cerc.usgs.gov/rss/visualize/ (Eds.), Community and Evolutionary Ecology of North American
visualize.htm. Stream Fishes. Univ. Oklahoma Press, Norman, pp. 166–177.
Jacobson, R.B., Laustrup, M.L., 2002. Habitat Assessment for Pallid Pinter, N., Wlosinski, J.H., Heine, R., 2002. The case for utilization of
Sturgeon Overwintering Surveys, Lower Missouri River. U.S. stage data in flood-frequency analysis: preliminary results from the
Geological Survey. Online at: http://www.cerc.usgs.gov/pubs/ Middle Mississippi and Lower Missouri River. Hydrologic Science
center/pdfDocs/micra_whole.pdf. and Technology Journal 18, 173–185.
Jacobson, R.B., Laustrup, M.L., Reuter, J.M., 2002. Habitat Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L.,
assessment, Missouri River at Hermann, Missouri. U.S. Geological Richter, B.D., Sparks, R.E., Stromberg, J.C., 1997. The natural
Survey Open-File Report 02-32, On CD-ROM or online at: http:// flow regime. Bioscience 47, 769–784.
www.cerc.usgs.gov/pubs/Moriver/herm_hab.htm. Power, M.E., Stout, R.J., Cushing, C.E., Harper, P.P., Hauer, F.R.,
Jacobson, R.B., Johnson, H.E., Laustrup, M.S., D'Urso, G.J., Matthews, W.J., Moyle, P.B., Statzner, B., Wais De Badgen, I.R.,
Reuter, J.M., 2004. Physical habitat dynamics in four side-channel 1988. Biotic and abiotic controls in river and stream commu-
chutes, Lower Missouri River. U.S. Geological Survey Open-File nities. Journal of the North American Benthological Society 7,
Report 2004-1071, On CD-ROM or online at: http://infolink.cr. 456–479.
usgs.gov/RSB/USGS_OFR_2004-1071/index.htm. Rhoads, B.L., Wilson, D., Urban, M.L., Herricks, E.E., 1999.
Jeffries, M., Mills, D., 1990. Freshwater Ecology: Principles and Interaction between scientists and nonscientists in community-
Applications. Belhaven Press, N.Y. based watershed management: emergence of the concept of stream
Josyln Art Museum, 1984. Karl Bodmer's America. University of naturalization. Environmental Management. 24, 297–308.
Nebraska Press, Lincoln, Nebraska. Richter, B.D., Baumgartner, J.V., Wigington, R., Braun, D.P., 1997.
Knowlton, M.F., Jones, J.R., 2000. Seston, light, nutrients and How much water does a river need? Freshwater Biology 37,
chlorophyll in the Lower Missouri River, 1994–1998. Journal of 231–249.
Freshwater Ecology 15, 283–297. Richter, B.D., Mathews, R., Harrison, D.L., Wigington, R., 2003.
Kondolf, G.M., Wilcock, P.R., 1996. The flushing flow problem: Ecologically sustainable water management: managing river flows
defining and evaluating objectives. Water Resources Research 32, for ecological integrity. Ecological Applications 13, 206–224.
2589–2599. Robinson, A.T., Clarkson, R.W., Forrest, R.E., 1998. Dispersal of
Latka, D.C., Nesteler, J., Hesse, L.W., 1993. Restoring physical habitat larval fishes in a regulated river tributary. Transactions of the
in the Missouri River: a historical perspective. In: Hesse, L.W., American Fisheries Society 127, 772–786.
Stalnaker, C.B., Benson, N.G., Zuboy, J.R. (Eds.), Proceedings of Schlosser, I.J., 1987. A conceptual framework for fish communities in
the Symposium on Restoration Planning for the Rivers of the small warmwater streams. In: Matthews, W.J., Heins, D.C. (Eds.),
Mississippi River Ecosystem. National Biological Survey Report, Community and Evolutionary Ecology of North American Stream
vol. 19, pp. 350–359. Fishes. Univ. Oklahoma Press, Norman, pp. 269–299.
Leopold, L.B., Wolman, M.G., Miller, J.P., 1964. Fluvial Processes Scheidegger, K.J., Bain, M.B., 1995. Larval fish distribution and
in Geomorphology. W.H. Freeman and Company, San Francisco. microhabitat use in free-flowing and regulated rivers. Copeia 1995,
522 p. 125–135.
Mazourek, J.C., 1998. Dynamics of seed bank and vegetation Schiemer, F., Keckeis, H., Reckendorfer, W., Winkler, G., 2000. The
communities in existing and potential emergent marshes in the “inshore retention concept” and its significance for large rivers.
Missouri River flood plain. Master's thesis. University of Archiv für Hydrobiologie. Supplement 135, 509–516.
Missouri, Columbia. Sparks, R.E., 1995. Need for ecosystem management of large rivers
McColpin, A., 2002. Waterbird use of lower Missouri River floodplain and their flood plains. Bioscience 14, 168–182.
wetlands. Master's thesis. University of Missouri, Columbia. Sparks, R.E., Nelson, J.C., Yin, Y., 1998. Naturalization of the flood
McGarigal, K., Marks, B.J., 1995. FRAGSTATS: spatial pattern anal- regime in regulated rivers. Bioscience 48, 706–720.
ysis program for quantifying landscape structure. U.S. Department Stanford, J.A., Ward, J.V., Liss, W.J., Frissell, C.A., Williams, R.N.,
of Agriculture Forest Service Gen. Tech. Rep. PNW-351. Lichatowich, J.A., Coutant, C.C., 1996. A general protocol for
Middleton, B.A., 2002. Flood Pulsing in Wetlands: Restoring the restoration of regulated rivers. Regulated Rivers: Research and
Natural Hydrological Balance. Wiley and Son, New York. Management 12, 391–413.
Missouri River Commission, 1894. Missouri River Commission Survey Steffler, P., Blackburn, J., 2001. River2D — Two-Dimensional Depth
Maps — 1879. Missouri River Commission, 1:63,360 scale. Averaged Model of River Hydrodynamics and Fish Habitat,
Moulton, G.E., 2002. The Definitive Journals of Lewis and Clark. Introduction to Depth Averaged Modeling and User's Manual.
University of Nebraska Press Lincoln, Nebraska. University of Alberta, Edmonton.
National Research Council, 1992. Restoration of Aquatic Ecosystems. Tibbs, J.E., Galat, D.E., 1997. Larval, Juvenile and Adult Small Fish
National Academy Press, Washington, DC. Use of Scour Basins Connected to the Lower Missouri River.
 R.B. Jacobson, D.L. Galat / Geomorphology 77 (2006) 249–269 269

Cooperative Fish and Wildlife Research Unit, University of Review and Update. Northwestern Division, U.S. Army Corps of
Missouri, Columbia. Engineers, Portland, Oregon.
U.S. Army Corps of Engineers, 1996. Missouri River Gavins Point U.S. Fish and Wildlife Service, 2000. Biological Opinion on the
Dam Degradation Trends Study. Omaha District, Omaha, Operation of the Missouri River Main Stem Reservoir System,
Nebraska. Operation and Maintenance of the Missouri River Bank Stabili-
U.S. Army Corps of Engineers, 1998a. Reservoir regulation studies- zation and Navigation Project, and Operation of the Kansas River
daily routing model studies. U.S. Army Corps of Engineers, Reservoir System. U.S. Fish and Wildlife Service, Bismarck,
Northwest Division Missouri River Region, Master Water Control North Dakota.
Manual Missouri River Review and Update Study, v. 2A, Omaha, U.S. Fish and Wildlife Service, 2003. Amendment to the 2000
Nebraska. Biological Opinion on the Operation of the Missouri River Main
U.S. Army Corps of Engineers, 1998b. Environmental Studies- Stem Reservoir System, Operation and Maintenance of the
Riverine Fisheries (supplement), Appendix C-Physical habitat Missouri River Bank Stabilization and Navigation Project, and
analysis. U.S. Army Corps of Engineers, Northwest Division Operation of the Kansas River Reservoir System. U.S. Fish and
Missouri River Region, Master Water Control Manual Missouri Wildlife Service, Minneapolis, Minnesota.
River Review and Update Study, v. 7d-s1, Omaha, Nebraska. War Department, 1920. Missouri River from the Mouth to Quindaro
U.S. Army Corps of Engineers, 2000. Digital Elevation Data. U.S. Bend. War Department, scale 1:12,000, Washington, D.C.
Army Corps of Engineers, Kansas City, Missouri. Webb, R.H., Schmidt, J.C., Marzolf, G.R., Valdez, R.A., 1999. The
U.S. Army Corps of Engineers, 2003. Supplemental Biological controlled flood in Grand Canyon. American Geophysical Union
Assessment for the Current Water Control Plan — Appendix B Monograph, vol. 110. Washington, D.C.
— Detailed description of the Corps' alternative to the 2000 BiOP Weibull, W., 1951. A statistical distribution function of wide
Gavins Point RPA. U.S. Army Corps of Engineers, Northwestern applicability. Journal of Applied Mechanics 18, 293–297.
Division, Portland, Oregon. Wolman, M.G., Miller, J.P., 1960. Magnitude and frequency of forces
U.S. Army Corps of Engineers, 2004. Summary Missouri River Final in geomorphic processes. Journal of Geology 68, 54–74.
Environmental Impact Statement — Master Water Control Manual