Responses of obligate versus faculative riparian shrubs following river damming

Abstract

Riparian or streamside woodlands include obligate riparian trees and shrubs (obligates) that are restricted to streamside zones, and facultative riparian species that are abundant in, but not restricted to the riparian areas. Due to their distinctive life history requirements, it may be predicted that the ecological specialist obligates would be more vulnerable than the facultative generalists to impacts from river damming and flow regulation. We tested this along the Snake River through Hells Canyon, USA, where two native riparian shrubs dominate: the obligate sandbar willow (Salix exigua), and the facultative, netleaf hackberry (Celtis reticulata). We assessed riparian conditions over the past century by comparing ground-level and aerial photographs taken after 1907 and in the 1950s in advance of three dams, versus recent conditions. These comparisons revealed three changes downstream from the dams: (1) the depletion of surface sands and sandbars and (2) reductions in sandbar willow versus (3) the proliferation of hackberry in dense bands above the typical high-water line. The willow decline probably resulted from the depletion of sand following sediment trapping by the reservoirs, combined with changes in the seasonal water flow pattern. The increase in hackberry may have resulted from a beneficial ‘irrigation effect’ of daily water releases for power generation during the summer. The opposing responses reflect the plants' differing life histories and may partially resolve impacts of river regulation on alluvial sediments versus the instream flow pattern. We consider other riparian studies that suggest that obligates such as cottonwoods (Populus angustifolia, P. deltoides and P. fremontii) are highly vulnerable to river regulation, while facultative trees and shrubs such as trembling aspen (Populus tremuloides), wolf-willow (Elaeagnus commutata) and velvet mesquite (Prosopis velutina) are more resilient. These results suggest that conservation of riparian woodlands should emphasize the ecological specialist obligates, while facultative species may be less vulnerable to river regulation. Copyright © 2009 John Wiley & Sons, Ltd.

Full text

RESPONSES OF OBLIGATE VERSUS FACULTATIVE RIPARIAN SHRUBS
FOLLOWING RIVER DAMMING
STEWART B. ROOD,
a
*
JEFFREY H. BRAATNE
b
and LORI A. GOATER
a
a
Department of Biological Sciences, University of Lethbridge, AB, Canada T1K 3M4
b
Department of Fish, Wildlife and Range Resources, University of Idaho, Moscow, ID 83844, Russia
ABSTRACT
Riparian or streamside woodlands include obligate riparian trees and shrubs (obligates) that are restricted to streamside zones,
and facultative riparian species that are abundant in, but not restricted to the riparian areas. Due to their distinctive life history
requirements, it may be predicted that the ecological specialist obligates would be more vulnerable than the facultative
generalists to impacts from river damming and flow regulation. We tested this along the Snake River through Hells Canyon,
USA, where two native riparian shrubs dominate: the obligate sandbar willow (Salix exigua), and the facultative, netleaf
hackberry (Celtis reticulata). We assessed riparian conditions over the past century by comparing ground-level and aerial
photographs taken after 1907 and in the 1950s in advance of three dams, versus recent conditions. These comparisons revealed
three changes downstream from the dams: (1) the depletion of surface sands and sandbars and (2) reductions in sandbar willow
versus (3) the proliferation of hackberry in dense bands above the typical high-water line. The willow decline probably resulted
from the depletion of sand following sediment trapping by the reservoirs, combined with changes in the seasonal water flow
pattern. The increase in hackberry may have resulted from a beneficial ‘irrigation effect’ of daily water releases for power
generation during the summer. The opposing responses reflect the plants’ differing life histories and may partially resolve
impacts of river regulation on alluvial sediments versus the instream flow pattern. We consider other riparian studies that suggest
that obligates such as cottonwoods (Populus angustifolia , P. deltoides and P. fremontii) are highly vulnerable to river regulation,
while facultative trees and shrubs such as trembling aspen (Populus tremuloides), wolf-willow (Elaeagnus commutata) and
velvet mesquite (Prosopis velutina) are more resilient. These results suggest that conservation of riparian woodlands should
emphasize the ecological specialist obligates, while facultative species may be less vulnerable to river regulation. Copyright #
2009 John Wiley & Sons, Ltd.
key words: Celtis reticulata; hackberry; riparian ecology; instream flow needs; river damming; Salix exigua; Snake River; willow
Received 10 October 2008; Revised 22 December 2008; Accepted 15 January 2009
INTRODUCTION
Riparian or streamside woodlands represent biologically rich interfaces between rivers and terrestrial ecosystems
(Naiman et al., 2005). Riparian woodlands such as floodpl ain forests are general ly vulnerable to river regulation but
there are consi derable differences in sensitivities of different trees and shrubs (Auble et al., 1994; Johnson et al.,
1995; Jansson et al., 2000). Ecological studies of riparian plant species have been undertaken in an attempt to
recognize patterns of vulnerability and this understanding could assist in managing river flows for environmental
conservation (Shafroth et al., 1998; Dixon and Johnson, 1999; Nilsson and Svedmark, 2002; Shafroth et al., 2002;
Auble et al., 2005).
A primary categorization of riparian plants considers their distribution, and obligate riparian plants or ‘obligates’
are ecological specialists that are restricted to the streamside zones. Ecological specialists typically display
distinctive adaptations which provide a high degree of benefit but over a narrow range of conditions (Feinsinger
et al., 1981; Futuyma and Moreno, 1988; Wilson and Yoshimura, 1994), and riparian obligates may have traits such
as restricted intervals of seed release and viability, narrow substrate requirements for successful colonization, and
particular flood and drought tolerances (Krasny et al., 1988; Karrenberg et al., 2002; Rood et al., 2003a). In
RIVER RESEARCH AND APPLICATIONS
River. Res. Applic. 26: 102–117 (2010)
Published online 22 February 2009 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/rra.1246
*Correspondence to: Stewart B. Rood, Department of Biological Sciences University of Lethbridge, AB, Canada T1K 3M4.
E-mail: rood@uleth.ca
Copyright # 2009 John Wiley & Sons, Ltd.

contrast, ecological generalists display less specific adaptations that are beneficial over a broader niche range.
Facultative riparian plants are ecological generalists that are often abundant in riparian areas, but are also typically
sparsely distributed in adjacent upland zones and can thus grow and reproduce in either riparian or upland
environments .
Of the riparian plants, the woody shrubs and trees are especially important with respect to ecological services
including sediment and bank stabilization, the assimilation of nutrients and other pollutants and the provision of
wildlife habitats (Naiman et al., 2005). Woody plants are consequently of high priority for riparian conservation
(Richter and Richter, 2000; Rood et al., 2005a). Both obligate and facultative species occur in most areas and this
provides an opportunity to compare consequences of ecological specialization. In comparing patterns across
diverse taxa, McKinney and Lockwood (1999) conclude that many ecological generalists, such as some invasive
weeds, are often more resilient than specialists and may even thrive following human alteration. From this pattern it
would be predicted that riparian obligates would be more vulnerable to river damming and flow regulation than
facultative plants.
We recognized an ideal study opportunity with the Snake River through Hells Canyon (Schmidt et al., 1995;
Braatne et al., 2008). This is a physically dramatic riverscape in a narrow, deep V-valley cut through basalt bedrock
(Vallier, 1998). With very hot and dry summers (hence, ‘Hells Canyon’), relatively few plant species occur and this
simplifies ecological comparison. Within the narrow riparian bands, two native woody plants predominate: the
obligate shrub, sandbar willow (Salix exigua), and the facultative shrub or small tree, netleaf hackberry (Celtis
reticulata).
Since native riverine organisms are generally adapted to the natural flow and distur bance regimes (Poff et al.,
1997; Shafroth et al., 2002; Rood et al., 2007), we predicted that there would be negative impacts of river regulation
on most native woody riparian plants in Hells Canyon. We further anticipated that the specialist, sandbar willow,
would be particularly vulnerable to human alteration, possibly due to more specific hydrogeomorphic
requirements, the demands for particular patterns of water and sediment substrate (Karrenberg et al., 2002; Rood
et al., 2003a; Dixon and Turner, 2006). In contrast, due to a broader niche range, we predicted that the generalist,
netleaf hackberry, would be more robust relative to the environmental alteration.
We further sought to understand the underlying basis for the responses of these two shrubs and analysed river
hydrology and considered alluvial sediments (Kondolf, 1997). We thus considered a general prediction relative to
the ecological specialization and vulnerability of riparian shrubs to river regulation, and then considered possible
explanations for differing responses of the two shrubs.
METHODS
Study system
The Snake River through Hells Canyon defines the border between Oregon and Washington, and Idaho, USA
(Figure 1, Braatne et al., 2008). Here, the Snake River has a fairly steep gradient and was developed for
hydroelectric power generation with the implementation of three large dams, Brownlee, Oxbow and Hells Canyon,
in 1958, 1961 and 1967, respectively. These are collectively managed as the Hells Canyon Complex (HCC). Below
the HCC, Hells Canyon provides a relatively inaccessible landscape and this has limited other human impacts.
Instream flows are almost entirely determined by water release from the HCC since contri butions from local
streams are minor above the Imnaha and Salmon River inflows (Figure 1).
Through Hells Canyon, riparian vegetation is limited to narrow bands flanking the river and the two native woody
plants that predominate are very different with respect to ecological specialization. The obligate sandbar willow,
S. exigua Nuttall (also ‘ coyote’ or ‘narrowleaf’ willow; syn. S. interior Rowlee), is common along streams
throughout western North America and is restricted to low elevation streamside zones (Krasny et al., 1988;
Ottenbreit and Staniforth, 1992; Friedman et al., 2006; Dixon and Turner, 2006). In contrast, the generalist netleaf
hackberry (C. reticulata Torr.; syn. Celtis laevigata Willd.) is abundant in higher riparian zones and also occurs
extensively but sparsely in the upland (Debolt and McCune, 1995; Salzer et al., 1996). The designations of sandbar
willow as an obligate and of hackberry as a facultative species are consistent with regional designations relative to
wetland occurrence (US-FWS, 1996).
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
103

Photograph comparisons
To investigate changes in these shrubs following damming, we obtained the General Land Office (GLO) survey
records from the 1870s to 1935 for the Idaho portion through Hells Canyon. We also searched regional archives for
ground-level photographs taken in Hells Canyon prior to the HCC. Photographs were sought that had sharp clarity,
and we avoided locations with apparent human impacts such near homes or ranches. Suitable photographs provided
comparison points that were relocated and re-photographed during river floats in the summers of 1998 through
2000. This provided compara tive photograph pairs with the original photograp h representing the pre-HCC
conditions and the contemporary photograph revealing conditions after four decades of operation of the HCC. This
research strategy was similar to that undertaken for the Colorado River through Grand Canyon by Webb (1996) and
represents a temporal comparison of pre- versus post-project conditions (Underwood, 1994, Braatne et al., 2008).
The black and white photographs were digitally scann ed at 300 dots per inch to provide resolutions similar to
those of the contemporary digital photo graphs, which were viewed as grey-scale images for consistency. The pairs
Figure 1. Map of the reaches and dams along the Snake River through Hells Canyon. Modified from Braatne et al. 2008.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
104 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

were viewed on computer monitors, allowing adjustment of image brightness, contrast and size. After initial
qualitative comparison, we rated the changes in the extents of: (1) exposed surface sands that occurred as sand bars
and as interstitial sands between cobbles, (2) sandbar willow and (3) netleaf hackberry. For each photograph pair
where each feature existed, we three authors independently rated the apparent post-dam change as decrease ()or
increase (þ). If our initial interpretations were not unanimous, we reassessed the pair, until unanimous assessments
were reached. This provided the resp onse for each photograph pair, and the probabilities of patterns across the
multiple pairs were investigated with binomial probabilities (i.e. the probability of 5 of 5 comparisons showing a
decrease ¼ (½)
5
).
For each photograph point, we obtained aerial photograph pairs from the United States Department of
Agriculture Aerial Photographic Records office, Salt Lake, UT, with one in the pre-dam period (typically
August 1955, black and white) and a contemporary match (typically August 1997, true colour). Aerial photograph
scales ranged from 1:6000 to 1:20 000.
Hydrology
River discharges (Q, flow rate) from Snake River hydrometric gauges were obtained from the United States
Geological Survey (USGS) water website (http://water.usgs.gov/) and from Idaho Power Company for flow
releases from the Hells Canyon Dam, the final dam in the HCC. We analysed mean annual (Q
a
), annual maximum
mean daily (Q
max
), monthly (Q
Month
) and mean daily (Q
d
) values and annual hydrographs for the periods of verified
record, generally from 1911 to 2005. Q
d
values were averaged over three, 11-year intervals to compare flow regimes
from 1911 to 1921 (‘natural’ ¼ prior to the major dams upstream of the HCC), 1946 to 1956 (‘pre-HCC’ ¼ after
upstream damming but prior to the HCC) and 1990 to 2000 (‘post-HCC’, after implementation of the HCC). To
reveal the summer diurnal pattern, 15-min interval Q data (Q
i
¼ Q ‘instantaneous’) were accessed from the USGS
‘near real-time ’ data for 2006. Statistical analyses with Microsoft Excel (Redmond, WA) and JMP (SAS, Cary, NC)
investigated historic trends through regression (Rood et al., 2005b).
RESULTS
Comparative photograph pairs
We located 240 photographs that displayed the Snake River above and through Hells Canyon, prior to the HCC.
Photographs were particularly located at the Washington State University, the Idaho Historical Society, Lewis and
Clark State College, the Nez Perce County Historical Society and Idaho Power Company (Blair et al., 2002). Many
photographs focussed on people and some were situated at home-sites or other altered locations, and riparian zones
were visible in the backgrounds, providing supplemental information about the pre-HCC conditions. Thirty-one
historic photographs provided views that were suitable for the photograph comparison points, with sufficient focus
and contrast to reveal the riparian substrate and shrubs, and limited human impact.
Twelve of the 31 photograph points occurred along the unimpounded Hells Canyon reach of the Snake River,
between the Hells Canyon Dam and the Salmon River inflow (Figure 1). Three of these photograph pairs are
presented in Figure 2 and interpretations for all 12 pairs are provided in Table I, along with information from the
land survey records. For reference, all 31 photographs and pairs are presented in Blair et al. (2002).
Six, five and six of the photograph points showed sites that were inundated or impacted by Brownlee, Oxbow and
Hells Canyon dams and reservoirs, respectively. One pre-dam photograph is provided for each reservoir in Figure 3
and information for all reservoir sites is provided in Table II. The river valley becomes progressively more confined
downstream and the landscape inundated by the Oxbow and Hells Canyon reservoirs closely resembled the reach
downstream from the HCC (Figure 3).
Three photo graph points represented sites between Weiser and Brownlee Reservoir, the u pper HCC reservoir.
These revealed an alluvial reach with a broad valley and physical landscape that was very different from that
through Hells Canyon (Braatne et al., 2008). The archival photographs were from 1902 and 1908 and displayed
sand-covered banks and abundant sandbar willow and hackberry. In contrast, by 1999, these sites were dominated
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
105

by exotic woody plants including salt-cedar (Tamarix spp.), Russian olive (Elaeagnus angustifolia) and European
willows (Salix alba and other species).
Physical conditions
The photograph pairs consistently indicated minimal change in the physical landscape (Figure 2). Channel and
bank positions and conditions were apparently constant, reflecting the dominant erosion-re sistant bedro ck. Large
boulders were often conspicuous and some recognizable boulders were relocated.
In the pre-HCC photographs for sites along the downstream reach and between the reservoirs, surface sands were
abundant (Tables I and II, Figures 2 and 3). In contrast, these same positions were often covered by cobbles in 1999
through 2003 (Figure 2). In the pre-dam condition, surface sands were conspicuous in bars such as along the inside
of meanders and in protected sites, and also in relatively exposed sites (such as in Figure 2E, right-side).
The depletion of surface sands along the downstream Hells Canyon reach provided the most conspicuous
difference betwee n the historic ground-level photographs and recent conditions (Table I). This depletion was also
apparent in the comparisons of pre- versus post-HCC aerial photographs (Figure 4) that reinforced the prior
analyses by Schmidt et al. (1995) and thus is not further presented here.
Figure 2. Three ground-level photograph comparisons for sites SR-3 (top, a and b, sites described in Table I), SR-4 (middle, c and d) and SR-11
(bottom, e and f) along the Hells Canyon reach of the Snake River, showing pre-dam versus post-dam conditions. The ‘s’ indicates sand and the
‘h’ indicates hackberry. Photos a and c—1955 by F.R. McCormick; b, d and f—1999, J. Braatne; e—1903, John Miller Collections, Nez Perce
Historical Society.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
106 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

Table I. Descriptions of sites along the Hells Canyon reach of Snake River (SR) from General Land Office (GLO) Surveys and comparisons of historic and recent
photographs
Photo point Location
(year of
historic photo)
GLO Records Surface features Vegetation Change
Historic photo Current photo Historic photo Current photo Sand Willow Hackberry
SR-1 Deep Ck
RM247 (1951)
1
Willow, buckbrush
scattered pine/fir (1935)
Fines, bedrock,
sandbars
Cobble and bedrock
2
Scattered pine
and hackberry
Scattered pine
and hackberry
oo o
SR-2 Saddle creek
RM 236 (1934)
Willow, maple
3
with
scattered pine/fir (1910)
Alluvial fan with
fine to cobble
Alluvial fan with
fine to cobble
Hackberry with sparse
cover along Saddle Ck
Hackberry with
diverse riparian
vegetation along
Saddle Ck
 o þ
SR-3 Johnson bar
RM230 (1953)
Fines to bedrock Cobble to bedrock Scattered pine
and hackberry
Extension of hackberry
with increased foliage
 þ
SR-4 Sheep creek
RM229.5 (1953)
Rocky rapids with
sands to cobble
Rocky rapids
with cobble
Scattered pine and
hackberry
Expansion and growth
of hackberry
 þ
SR-5 Pine bar
RM 227.5 (1930s)
Large sandbar
deposit
Highly-eroded
sandbar
Scattered pine with
extensive willow
Scattered pine, some
hackberry, no willow
 o
SR-6 Pittsburg landing
RM 215 (1928)
Willow, maple
2
with
scattered pine/fir
(1901-04)
Bedrock to cobble Bedrock to cobble Scattered hackberry Scattered hackberry  o þ
SR-7 Pittsburg landing
RM 214.5 (1953)
Fines to cobble banks Cobble banks Fringe of hackberry Expanded fringe
of hackberry
 þ
SR-8 Getty creek
RM 206 (1928)
Bedrock to
cobble slopes
Bedrock to
cobble slopes
Sparse upland vegetation Sparse upland
vegetation
ooþ
SR-9 Ragtown bar
RM 205 (1928)
Cobble slopes cobble slopes Scattered hackberry scattered hackberry – o þ
SR-10 Dug bar
RM 196.5 (1928)
Cliffs, ravines, basalt
ledges, no
timber (1906)
Basalt
cliffs/talus slopes
Basalt cliffs/talus
slopes
Scattered hackberry Scattered hackberry – o þ
SR-11 Eureka bar
RM 191.5 (1903)
No timber/ no
other notes
Rocky slopes
with large
and small sandbars
Remnant sandbar Willow along
sandbar fringe
Altered willow patch  o
SR-12 Above Salmon R.
RM188.5 (1928)
Steep rocky slopes Steep rocky slopes Minimal riparian
vegetation
Minimal riparian
vegetation
oo o
Total (binomial p) 9/9–( p ¼ 0.002) 5/5–(P ¼ 0.039) 8/8 þ ( p ¼ 0.004)
Three photograph pairs are presented in Figure 2 and all pairs are displayed in Blair et al. (2002). Apparent changes in surface sand, sandbar willow and netleaf hackberry were assessed in
photograph pairs as decrease () or increase (þ). These features were absent or insufficient for assessment (o) at some sites.
1
RM ¼ river mile.
2
There was some construction alteration at this site.
3
Maple’ in the GLO record very probably refers to hackberry, which has a similar leaf shape.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
107

Willow versus hackberry
Two changes in woody vegetation were consistently indicated in the comparative pairs and in comparisons of the
survey records and observations of the 240 total historic photographs, versus field observations and photographs
along the river valley in 1999 through 2003 (Table I and Figure 2). Sandbar willow was consistently the first-listed
plant in the GLO surveys, indicating prominent occurrence from 1901 through 1935 (Table I). Due to its ligh t,
pubescent, narrow leaves, the shrub was less conspicuous in some of the pre-dam photo graphs but in photographs
with sharper resolution and closer views, the prevalence of sandbar willow prior to damming was confirmed
(Tables I and II and Figures 2 and 3). In each of the five cases in which formal comparison was possible, there was a
considerable reduc tion in sandbar willow in the post-HCC photograph (Table I). Con firming the early abundance of
sandbar willow through Hells Canyon, it was present in 15 of the 16 photographs of the riparian zones upstream of,
or inund ated by the HCC reservoirs (Table II).
The comparisons also indicated a second change in woody riparian vegetation after damming, an apparent
increase of netleaf hackberry (Table I and Figure 2). In both the pre- and post-dam photographs, hackberry occurred
sparsely in the upland zones (Figures 2 and 3). Hackberry was consistently more abundant in the riparian zones, and
Figure 3. Historic photographs of zones inundated by the Hells Canyon Complex of dams and reservoirs showing typical conditions with
common sandbars and sandbar willow: (a) Brownlee Dam site, 1953 (BR-6, Table II); (b) Oxbow Dam site, 1953 (OX-1); (c) Kinney Creek
Rapids, 1953 (HC-5) (a and b by F.R. McCormick, c by P. Basche).
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
108 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

Table II. Riparian conditions in historic photographs of sites along the Snake River upstream from Hells Canyon (Weiser, WR)
and through Hells Canyon in sites that were inundated by Brownlee (BR), Oxbow (OX) or Hells Canyon (HC) reservoir
Site Location Year Sand Willow Hackberry
WR-1 Weiser bridge 1908 þþ þ
WR-2 Westlake island ferry 1908 þþ þ
WR-3 Farewell bend 1902 þþ þ
BR-1 Burnt river bridge 1899 þþ þ 
BR-2 Morgan creek 1952 þþþ þ
BR-3 Hibbard creek 1950’s þ 
BR-4 Soda creek 1952 þþ 
BR-5 Powder river 1956 þþþ þ þ
BR-6 Brownlee dam 1953 þþ þþ 
OX-1 Oxbow dam 1953 þþ þ þ
OX-2a Scorpion creek 1953 þþ ??
OX-2b Scorpion creek 1953 þþ - þ
OX-3a Oxbow tip 1953 þ ? þ
OX-3b Oxbow tip 1953 þ ? þ
HC-1 Hells Canyon park 1950 þ 
HC-2 Ballard bridge 1953 þþ -
HC-3 McGraw creek 1953 þþ þ
HC-4 Spring creek 1953 þ þ
HC-5 Kinney creek rapids 1953 þþ þ
HC-6 Eagle bar landing 1953  ??
Total 14/19þ 15/16þ 12/18þ
Three photographs are presented in Figure 3 and all photographs are displayed in Blair et al. (2002). From the photographs, surface sand, sandbar
willow and netleaf hackberry were assessed as absent (), present (þ), abundant (þþ) or profuse (þþþ). A few photographs were taken from
points high above the river producing ambiguity relative to shrub occurrence (?) and these were excluded from Totals.
Figure 4. Pairs of aerial photographs from 1955 (left) versus 1997 (right) for the Snake River through Hells Canyon along Dug Bar (SR-10)
showing the depletion of sand bars along a gradual meander lobe (1) and in a protected cove (2).
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
109

following damming the shrub occurred in dense and relatively continuous bands near the typical high water line
(Figures 2 and 5). Bands were often particularly dense on steep banks such as along those sloping down from the
extensive terraces formed by the massive Bonneville floods (Vallier, 1998, Figure 5). During the 1999–2003 field
observations, the dense hackberry bands were conspicuous along the length of the Snake River through the Hells
Canyon reach and the increase was even apparent in some of the aerial photographs.
Historic hydrology
Despite the considerable upstream water diversions for agricultural irrigation (Parkinson et al ., 2003), total
annual discharge of the Snake River through Hells Canyon did not dramatically decline through the past century
(r ¼ 0.058, ns; Figure 6). After about 1960 the annual discharge has been highly variable, and there was an increase
in the frequency of low-flow years in recent decades (Q
a
< 350 m
3
s
1
(the low flow in the pre-dam interval); 4 years
from 1911 to 1960 vs. 12 years from 1961 to 2005, x
2
¼ 4.9, p < 0.05).
The spring peak is important for seedling recruitment of riparian Salicaceae (Scot t et al., 1996, Mahoney and
Rood, 1998; Karrenberg et al., 2002; Dixon and Turner, 2006) and while daily maximum flows did not change
consistently over the past century (Figure 7), the broad spring peak was progressively attenuated (Figure 8). Major
Figure 5. The typical, dense band of hackberry (h) above the mean annual high water line along the Snake River in Hells Canyon
(September 1999, near SR-6, S.B. Rood).
Figure 6. Mean annual discharge (Q
a
) of the Snake River at Weiser, Idaho (1910–2004) before (pre-HCC) and after (post-HCC) the
implementation of the Hells Canyon Complex (HCC) of three hydroelectric dams. To simplify comparison, a dashed line represents
350 m
3
s
1
, the approximate pre-dam low flow.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
110 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

changes occurred after the upstream dams were implemented (Figure 8B) and further attenuation has followed
(Figure 8C). The attenuation was largely due to increased regulation upstream since the daily and spring peaks were
generally similar upstream and downstream of the HCC (Figures 7 and 8).
While the spring peak has been attenuated, the subsequent post-peak recession has persisted (prior to Julian Day
200, Figure 8). Low flows continued to occur in late summer, and flows often increased with autumn rains. The
combination of rain and snowmelt increased flows through the winter to the next spring peak, and the HCC has
slightly altered this seasonal component (Figure 8).
In contrast to the reduction of spring flows, late summer flows were elevated following the early damming of the
Snake River (Figure 9). Prior to the implementation of the major Snake River dams, August discharge averaged
about 200 m
3
s
1
(Figure 9). From 1925 to 1960, during the period of damming and increasing regulation upstream
of the HCC, the August discharge progressively increased to about 350 m
3
s
1
(Figure 9). After 1960, the
hydrologic record reflects extensive variation in the late sum mer flows (Figures 8 and 9).
Although river flow data are often represented and archived as daily mean discharges, hydroelectric dams are
generally operated with diurnal flow pulsing as has occurred for the HCC (Fig ure 10). The pulsing shown for
August 2006 was typical for the low flow period of mid- to late summer (Figure 8). As shown, the daily flow varied
2.2-fold, from about 250–550 m
3
s
1
(Figure 10). This daily flow pulsing would elevate the river stage for a few
hours in most days.
DISCUSSION
This study provided a pre- versus post-damming temporal comparison, one of the most obvious approaches to
assess ecological impacts from river damming (Braatne et al., 2008). This study approach is often hindered by the
lack of pre-dam data and the comparison of photograph pairs is somewhat qualitative and may thus reveal the
direction of change , but not detailed aspects relating to the magnitude or timing.
Our observation of the depletion of surface sands is consistent with the well-established sediment-trapping by
large reservoirs (Schmidt et al., 1995, Kondolf, 1997). The depletion of sand bars through Hells Canyon was
previously describ ed by Schmidt et al., (1995), following analyses of aerial photographs. With respect to this
depletion of alluvial sediments, 48 upstream dams along the Snake River and its tributaries would have severed the
Figure 7. Annual maximum mean daily discharges (Q
max
) of the Snake River at Weiser (o), Idaho and in Hells Canyon (*).
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
111

supply of sands from those upstream watersheds and contributed to the depletion, in additi on to the local impacts
from the HCC (Parkinson et al., 2003).
The second apparent change was the depletion of sandbar willow and prior studies have revealed reductions in
this obligate riparian shrub following damming of other rivers (Lesica and Miles, 1999; Roo d et al., 2003b). This
shrub is often associated with sandbars and is commonly very abundant around the sandbar margins, while the
central portions of sandba rs are generally barren of vegetation. This distribution may reflect the additional stability
around the margins due to cobbles and boulders that provide a secure substrate and prevent uprooting of the willow
with flood flows that suspend sands.
The decline of sandbar willow through Hells Canyon probably results at least partly from the depletion of sands .
The loss of sands would reduce the moisture-retention capacity and capillarity of the riparian substrate (Mahoney
and Rood, 1998; Amlin and Rood, 2002). Subsequent drought stress would have been increased during the more
frequent, low-flow years (Figure 6). Changes in the seasonal streamflow pattern could degrade conditions relative to
Figure 8. Seasonal daily mean discharge ( Q
d
) for Snake River at Weiser, Idaho, for intervals prior to major dams (‘natural’, 1911–1921), and
before and after the implementation of the Hells Canyon Complex (HCC). Flows through Hells Canyon are also indicated in the post-HCC
interval.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
112 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

seedling recruitment while consequences on clonal expansion could be different, and even beneficial (Krasny et al.,
1988; Ottenbreit and Staniforth, 1992; Shafroth et al., 1998; Amlin and Rood, 2002; Dixon and Turner, 2006).
Sandbar willow is probably tolerant of, and dependent upon, the natural flow and disturbance regimes (Ottenbreit
and Staniforth, 1992; Karrenberg et al., 2002; Dixon and Turner, 2006; Rood et al., 2007) and changes due to river
regulation would degrade conditions for this riparian obligate.
Our third observation was unexpected and most novel, the apparent proliferation of hackberry in dense bands
above the typical high-water line. Like sandbar willow, hackberry can establish through seedlings or expand
through clonal root suckers but it is adapted to a broader environmental range than sandbar willow (Debolt and
Figure 9. Mean discharge in August of the Snake River at Weiser, Idaho and in Hells Canyon after the implementation of the Hells Canyon
Complex (HCC) of three hydroelectric dams. A linear regression is fit to the intermediate interval since these data appeared to display a linear
trend.
Figure 10. Discharge averaged over 15-min intervals of the Snake River in Hells Canyon, from 14 August to 1 September 2006, showing diurnal
variation associated with hydroelectric power generation.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
113

McCune, 1995; Salzer et al., 1996). As a facultative riparian plant, hackberry is more drought-tolerant and less
dependent upon the natural instream flow and fluvial disturbance regimes. We had anticipated that due to alteration
away from the natural conditions, both the obligate and facultative riparian shrub s would be diminished following
river regulation. In contrast, hackberry appar ently increased following damming.
The increase in hackberry may result partly from reduced competition due to the decline of sandbar willow and
also possibly as a resu lt of release from prior sheep grazing pressure (Wissmar et al., 1994; Parkinson et al., 2003).
We also believe that a substantial benefit originated from the supplemental moisture provided by the diurnal flow
pulsing associated with hydroelectric power generation. Due to this diurnal pulsing, the adjacent riparian fringe
would be saturated and following some groundwater infiltration, the root zone would remain moist for some
interval thereafter. We thus believe that the daily peak of 550 m
3
s
1
rather the daily mean discharge would
probably be more relevant for hackberry growth (Fig ure 8, Stromberg et al., 1993; Springer et al., 1999; Rood et al.,
2003a).
With diurnal flow pulsing, the root-zone below the hackberry band would receive supplemental moisture through
the stressful, hot and dry summer interval. This could act somewhat like the periodic sub-irrigation in a hydroponic
greenhouse operation, in which a coarse substrate is saturated by elevating the water level for a daily interval. We
characterize this as a beneficial ‘irrigation effect’ that probably contributes to the proliferation of hackberry.
Consistent with this interpretation, there were generally dense but narrow bands of hackberry above the typical
high-water line (Figure 5), a distribution that differed somewhat from the more dispersed natural pattern (Figure 2,
for example 2C vs. 2D).
It might be expected that the irrigation effect would also benefit sandbar willow but we observed that the dense
bands were consistently dominated by hackberry (Figure 5). Sandbar willow is an ecological pioneer that requires
barren sites for colonization and expansion, and is probably a poor competitor with established vegetation (Krasny
et al., 1988; Ottenbreit and Staniforth, 1992; Dixon and Turner, 2006).
The proliferation of hackberry is notable as an increase in the abundance of a native, woody riparian plant,
following river regulation. There are prior reports of increases in riparian trees and shrubs following damming and
these cases have often reflected expansion into zones in which natural flooding would have previously excluded
colonization (Johnson, 1994; Friedman et al., 1998; Shafroth et al., 2002). With the increase in hackberry through
Hells Canyon there would be a corresponding increase in wildlife habitat. We observed that spiders and webs were
very abundant in the hackberry band, and this might suggest an abundant invertebrate community. In Hells Canyon,
hackberry also provides dense nesting habitat for many birds and especially neotropical migrants , and these and
other wildlife species are likely to benefit from the hackberry band (Blair et al., 2002).
Our results from the photograph comparisons also provide direction for further field study. We would thus
encourage quantitative assessments of riparian surface sediments and vegetation, and particularly sandbar willow
and hackberry, above, through and below Hells Canyon. We would also recommend similar analyses along the
adjacent lower gorge of the free-flowing Salmon River, to provide a complementary spatial comparison (Braatne
et al., 2008).
Ecological specialization and vulnerability of other riparian plants
Following the opposing responses of sandbar willow versus hackberry, a further question follows, ‘Is the
association between ecological specialization and vulnerability typical for other riparian shrubs and trees?’
Possibly the most extensively studied riparian plants are cottonwoods, Populus species that often comprise
dominant trees in riparian woodlands around the Northern Hemisphere (Scott et al., 1996; Shafroth et al., 1998 ;
Rood et al., 2003b; Braatne et al., 2007). The section Aigeiros prairie cottonwood, Populus deltoides, and Fremont
cottonwood, Populus fremontii, occur as riparian obligates in the drier ecoregions of western North America, and
are particularly sensitive to river damming. In contrast, the section Tacamahaca, ‘balsam poplars’, balsam poplar,
Populus balsamifera, and black cottonwood, Populus trichocarpa , are facultative riparian trees that may be less
vulnerable (Krasny et al., 1988; Polzin and Rood, 2006). The narrowleaf cottonwood (Populus angustifolia)is
somewhat intermediate between those two groups, with respect to ecological specialization and environmental
vulnerability (Rood et al., 2003a). Further extending the pattern, the trembling aspen, Populus tremuloides,is
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
114 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

primarily an upland tree but is also a facultative riparian species and appears to be less responsive to river damming
than cottonwoods.
The comparisons can also be made acro ss genera. In many semi-arid regions of western North America, the
facultative riparian shrub, wolf-willow, Elaeagnus commutata, displays distributions similar to hackberry,
occurring sparsely in upland areas and more densely in the moister riparian zones. We have observed abundant
wolf-willow in riparian zones downstream from dams, in areas where cottonwoods have declined, and quantitative
analyses of this shrub would be worthwhile. Wolf-willow is related to Russian olive, E. angustifolia, an invasive,
exotic tree that thrives along some regulated river reaches (Lesica and Miles, 1999; Pearce and Smith, 2000; Katz
and Shafroth, 2003). Wolf-wil low and Russian olive are facultative riparian species and their apparent abundance
downstream of dams supports the pattern of reduced vulnerability of ecological generalists.
South of Hells Canyon, mesquites (Prosopis sp.) are abundant facultative riparian shrubs and trees. Different
species occur, with velvet mesquite (Prosopis velutina) being well studied in riparian zones (Stromberg et al.,
1993). The velvet mesquite life history is partly coordinated with the natural flow regime, but it is probably less
sensitive to flow alteration than the obligate riparian Fremont cottonwood that overlaps in distribution. This
provides another comparison that supports the greater vulnerability of obligates.
With this association between ecologi cal specialization and vulnerability to river regulation, we recommend that
obligate riparian trees and shrubs would provide the appropriate focus for analyses of the ecological health of
riparian woodlands. Due to their higher sensitivity, these species should be more diagnostic for scientific studies
and for monitoring management practices intended for the conservation and restoration of floodplain forests
(Richter and Richter, 2000; Rood et al., 2005a). We further recommend that the niche breadth and degree of
ecological specialization should be characterized across other riparian shrubs and trees, to identify candidate
species for monitoring and conservation efforts in other ecoregions worldwide. For such characterization,
hydrogeomorphic requireme nts would be defined, along with life history characteristics relative to seed and
seedling development, and aspects of environmental adaptation such as flood and drought tolerance and nutrient
requirements and adaptations.
ACKNOWLEDGEMENTS
This research was supported by funding from Idaho Power Company (IPC) and NSERC Canada. The authors
extend thanks to C. Blair of CH2M, Idaho, R. Simons of Simons & Assoc., Colorado and G. Holmstead, T.
Holthuijzen, F. Edelman and A. Ansell of IPC for their insightful discussions.
REFERENCE
Amlin NM, Rood SB. 2002. Comparative tolerances of riparian willows and cottonwoods to water-table decline. Wetlands 22: 338–346.
Auble GT, Friedman JM, Scott ML. 1994. Relating riparian vegetation to present and future streamflows. Ecological Applications 4: 544–554.
Auble GT, Scott ML, Freidman JM. 2005. Use of individualistic streamflow-vegetation relations along the Fremont River, Utah, USA to assess
impacts of flow alteration on wetland and riparian areas. Wetlands 25: 143–154.
Blair C, Braatne JH, Simons RK, Rood. SB. 2002. Effects of construction and operating the Hells Canyon hydroelectric complex on wildlife
habitat. Hells Canyon hydroelectric project, Idaho Power Company, Boise, Idaho. Technical Report E.3.2.
Braatne JH, Jamieson R, Gill KM, Rood. SB. 2007. Instream flows and the decline of riparian cottonwoods along the Yakima River, Washington,
USA. River Research and Applications 23: 247–267.
Braatne JH, Rood SB, Goater LA, Blair CL. 2008. Analyzing the impacts of dams on riparian ecosystems: a review of research strategies and
their relevance to the Snake River through Hells Canyon. Environmental Management 41: 267–281.
Debolt AM, McCune B. 1995. Ecology of Celtis reticulata in Idaho. Great Basin Naturalist 55: 237–248.
Dixon MD, Johnson WC. 1999. Riparian vegetation along the middle Snake River, Idaho: zonation, geographical trends, and historical changes.
Great Basin Naturalist 59: 18–34.
Dixon MD, Turner MG. 2006. Simulated recruitment of riparian trees and shrubs under natural and regulated flow regimes on the Wisconsin
River, USA. River Research and Applications 22: 1057–1083.
Feinsinger P, Spears EE, Poole RW. 1981. A simple measure of niche breadth. Ecology 62: 27–32.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
115

Friedman JM, Osterkamp WR, Scott ML, Auble GT. 1998. Downstream effects of dams on channel geometry and bottomland vegetation:
regional patterns in the Great Plains. Wetlands 18: 619–633.
Friedman JM, Auble GT, Andrews ED, Kittel G, Madole RF, Griffin ER, Allred TM. 2006. Transverse and longitudinal variation in woody
riparian vegetation along a montane river. Western North American Naturalist 66: 79–91.
Futuyma DJ, Moreno G. 1988. The evolution of ecological specialization. Annual Review of Ecology and Systematics 19: 207–233.
Jansson R, Nilsson C, Renofalt B. 2000. Fragmentation of riparian floras in rivers with multiple dams. Ecology 81: 899–903.
Johnson WC. 1994. Woodland expansion in the Platte River, Nebraska: patterns and causes. Ecological Monographs 64: 45–84.
Johnson WC, Dixon MD, Simons R, Jenson S, Larson K. 1995. Mapping the response of riparian vegetation to possible flow reductions in the
Snake River, Idaho. Geomorphology 13: 159–173.
Karrenberg S, Edwards PJ, Kollman J. 2002. The life history of Salicaceae living in the active zone of floodplains. Freshwater Biology 47: 733–
748.
Katz GL, Shafroth PB. 2003. Biology, ecology and management of Elaeagnus angustifolia L. (Russian olive) in western North America.
Wetlands 23: 763–777.
Kondolf GM. 1997. Hungry water: effects of dams and gravel mining on river channels. Environmental Management 21: 533–551.
Krasny ME, Vogt KA, Zasada JC. 1988. Establishment of four Salicaceae species on river bars in interior Alaska. Ecography 11: 210–
219.
Lesica P, Miles S. 1999. Russian olive invasion into cottonwood forests along a regulated river in north-central Montana. Canadian Journal of
Botany 77: 1077–1083.
Mahoney JM, Rood SB. 1998. Stream flow requirements for cottonwood seedling recruitment—an integrative model. Wetlands 18: 634–645.
McKinney ML, Lockwood JL. 1999. Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends in Ecology
and Evolution 14: 450–453.
Naiman RJ, De
´
camps H, McClain ME. 2005. Riparia: Ecology, Conservation and Management of Streamside Communities. Elsevier Academic
Press: Amsterdam.
Nilsson C, Svedmark M. 2002. Basic principles and ecological consequences of changing water regimes: riparian plant communities.
Environmental Management 30: 468–480.
Ottenbreit KA, Staniforth RJ. 1992. Life-cycle and age structure of ramets in an expanding population of Salix exigua (sandbar willow).
Canadian Journal of Botany 70: 1141–1146.
Parkinson S, Anderson K, Conner J, Milligan J. 2003. Sediment transport, supply, and stability in the Hells Canyon reach of the Snake River.
Hells Canyon hydroelectric project, Idaho Power Company, Boise, ID. Technical Report E. 1-1.
Pearce CM, Smith DG. 2000. Plains cottonwood’s last stand: can it survive invasion of Russian olive onto the Milk River, Montana floodplain?
Environmental Management 28: 623–637.
Poff NL, Allan D, Bain MB, Karr JR, Prestegaard KL, Richter BD, Sparks RE, Stromberg JC. 1997. The natural flow regime: a paradigm for river
conservation and restoration. BioScience 47: 769–784.
Polzin ML, Rood SB. 2006. Effective disturbance: seedling safe sites and patch recruitment of riparian cottonwoods after a major flood of a
mountain river. Wetlands 26: 965–980.
Richter BD, Richter HE. 2000. Prescribing flood regimes to sustain riparian ecosystems along meandering rivers. Conservation Biology 14:
1467–1478.
Rood SB, Braatne JH, Hughes FMR. 2003a. Ecophysiology of riparian cottonwoods: stream flow dependency, water relations and restoration.
Tree Physiology 23: 1113–1124.
Rood SB, Gourley C, Ammon EM, Heki LG, Klotz JR, Morrison ML, Mosley D, Scoppettone GG, Swanson S, Wagner PL. 2003b. Flows for
floodplain forests: a successful riparian restoration. BioScience 53: 647–656.
Rood SB, Samuelson GM, Braatne JH, Gourley CR, Hughes FMR, Mahoney JM. 2005a. Managing river flows to restore floodplain forests.
Frontiers in Ecology and Environment 3: 193–201.
Rood SB, Samuelson GM, Weber JK, Wyrot KA. 2005b. Twentieth-century decline in streamflows from the hydrographic apex of North
America. Journal of Hydrology 306: 215–253.
Rood SB, Goater LA, Mahoney JM, Pearce CM, Smith DG. 2007. Floods, fire, and ice: disturbance ecology of riparian cottonwoods. Canadian
Journal of Botany 85: 1019–1032.
Salzer MW, McCord VAS, Stevens LE, Webb RH. 1996. The dendrochronology of Celtis reticulata in the Grand Canyon: assessing the impact of
regulated flow on tree growth. In Environment and Humanity, Dean JS, Meko DM, Swetnam TW (eds). RadioCarbon, University of Arizona:
Tucson; 272–281.
Schmidt JC, Grams PE, Webb RH. 1995. Comparison of the magnitude of erosion along two large regulated rivers. Water Resources Bulletin 31:
617–631.
Scott ML, Auble GT, Friedman JM. 1996. Fluvial process and the establishment of bottomland trees. Geomorphology 14: 327–339.
Shafroth PB, Auble GT, Stromberg JC, Patten DT. 1998. Establishment of woody riparian vegetation in relation to annual patterns of streamflow,
Bill Williams River, Arizona. Wetlands 18: 577–590.
Shafroth PB, Stromberg JC, Patten DT. 2002. Riparian vegetation response to altered disturbance and stress regimes. Ecological Applications 12:
107–123.
Springer AE, Wright JM, Shafroth PB, Stromberg JC, Patten DT. 1999. Coupling groundwater and riparian vegetation models to assess effects of
reservoir releases. Water Resources Research 35: 3621–3630.
Stromberg JC, Wilkins SD, Tress JA. 1993. Vegetation-hydrology models: implications for management of Prosopis velutina (velvet mesquite)
riparian ecosystems. Ecological Applications 3: 307–314.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
116 S. B. ROOD, J. H. BRAATNE AND L. A. GOATER

Underwood AJ. 1994. On beyond BACI —sampling designs that might reliably detect environmental disturbances. Ecological Applications 4:
3–15.
United States Fish & Wildlife Service (US-FWS). 1996. 1996 National list of vascular plant species that occur in wetlands. http://www.fws.gov/
nwi/Plants/list96.html (accessed on November 2008).
Vallier T. 1998. Islands & Rapids, The Geologic Story of Hells Canyon. Confluence Press: Lewiston, ID.
Webb RH. 1996. Grand Canyon, a century of change. University of Arizona Press: Tuscon, AZ.
Wilson DS, Yoshimura J. 1994. On the coexistence of specialists and generalists. American Naturalist 144: 692–707.
Wissmar RC, Smith JE, McIntosh BA, Li HW, Reeves GH, Sedell JR. 1994. A history of resource use and disturbance in riverine basins of eastern
Oregon and Washington (early 1800s–1900s). Northwest Science 68: 1–35.
Copyright # 2009 John Wiley & Sons, Ltd. River. Res. Applic. 26: 102–117 (2010)
DOI: 10.1002/rra
RIVER DAMMING AND RIPARIAN SHRUBS
117