University of Nevada, Reno
Plant Community Invasibility in Riparian Landscapes: Role of Disturbance,
Geomorphology, and Life History Traits.
A dissertation submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy in Ecology, Evolution, and Conservation Biology
by
Susan G. Mortenson
Dr. Peter J. Weisberg / Dissertation Advisor
December 2009
THE GRADUATE SCHOOL
We recommend that the dissertation prepared under our supervision by
SUSAN GRACE MORTENSON
entitled
Plant Community Invasibility in Riparian Landscapes: Role of Disturbance, Geomorphology, and Life History Traits
be accepted in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
Peter J. Weisberg, Ph.D., Advisor
Jeanne C. Chambers, Ph.D., Committee Member
Elizabeth A. Leger, Ph.D., Committee Member
Ashley D. Sparrow, Ph.D., Committee Member
Thomas F. Bullard, Ph.D., Graduate School Representative
Marsha H. Read, Ph. D., Associate Dean, Graduate School
December, 2009
i
ABSTRACT
Riparian landscapes are highly susceptible to invasion by non-native plant species. High productivity and frequent disturbances associated with flooding contribute to invasibility of riverbanks and floodplains. The hydrogeomorphology of riparian landscapes is intricately tied to plant community composition and structure. Plant invasions often coincide with the period directly following regulation because novel riparian habitat is created. Additionally, river regulation in the form of dams, diversions, and inter-basin water transfers alter disturbance regimes (flood frequency, magnitude, timing, and duration). Native species lag behind introduced species in colonization of new riparian habitat due to life history constraints imposed by adaptation to the previous disturbance regime. Plant invasions often coincide with the period directly following river regulation, but non-native plant species also spread along rivers that have not been hydrologically altered.
Tamarix spp. are invasive shrubs that have spread prolifically throughout riparian
landscapes of the southwestern US. Tamarix has a long period of seed release and high
salinity and drought tolerance relative to native pioneer shrubs and trees. These
characteristics combined with the ability to form a thick litter layer contribute to
competitive exclusion of native shrubs. The presence of large, dense stands of Tamarix along rivers also resists scour from floods, stabilizes the river channel, causes channel incision, and reduces overbank flooding which is required for establishment of native pioneers and overall ecosystem function (e.g., nutrient availability, decomposition).
Tamarix control is currently conducted with a variety of methods (herbicide application, mechanical removal, biocontrol, burning, flooding). I researched the potential for ii prevention of Tamarix establishment using controlled flooding along one of the most
regulated rivers in the US, the Colorado River through the Grand Canyon National Park.
The high geomorphologic diversity of the Grand Canyon and long period of flow
records make this an ideal study site for understanding the influence of
hydrogeomorphology and other environmental factors on Tamarix recruitment. Although
Tamarix has been present in the Grand Canyon since at least the 1930s, this species expanded its range extensively following post-dam flooding in 1965 and 1973. Through a tree-ring analysis I investigated influences of past hydrologic and climatic conditions on
Tamarix establishment and persistence at different flow stage elevations from 1984 through 2008. Historic records and scarcity of pre-1983 Tamarix revealed high Tamarix mortality from the 1983 through 1985 floods, but these floods also initiated a large establishment event. Tamarix establishment was positively correlated with years of high
summer flows and low precipitation the following year. The importance of precipitation
for large recruitment episodes and evidence of Tamarix establishment in every year after
1983 suggests that controlled floods are unlikely to prevent future Tamarix establishment.
Elevated levels of establishment caused by the 1983 through 1985 floods led us to
conclude that controlled floods should not be conducted during Tamarix seed release
(April through September).
In addition to hydrologic and climatic regimes, bi-trophic interactions (e.g.,
herbivory, pathogens) also influence riparian plant community composition and structure.
I investigated the potential for selective foraging by beavers to affect coarse-scale spatial
patterns of riparian vegetation along the Colorado River in the Grand Canyon. Spatial
associations of beaver occurrence and Salix and Tamarix cover were analyzed using iii multiple linear regression models after accounting for relationships with geomorphic variables (geomorphic reach, sinuosity, and rock resistivity). Beaver presence had a strong positive association with Tamarix cover and a slight, positive association with
Salix cover. This suggests that Tamarix and beavers occupy similar habitats even at spatial scales less than 4.5 ha, beavers prefer habitats with high Tamarix cover, or that
beavers promote Tamarix dominance through selective foraging of Salix .
To explore the hypothesis that river regulation increases dominance of non-native
species I conducted a survey of riparian shrubs and trees along 20 river segments across
the southwestern US. I created an index of flow alteration based on hydrologic
conditions prior to and after dam completion or prior to and after 1962 for non-dammed
rivers. Regressions of dominance of native ( Salix exigua and Populus spp.) and non-
native ( Tamarix spp. and Elaeagnus angustifolia ) woody species with the degree of flow
alteration revealed a positive relationship between Tamarix and flow alteration, a
negative relationship between Populus and flow alteration, and no significant relationship
between S. exigua and E. angustifolia and flow alteration. Native and non-native species
respond to hydrologic characteristics based on their life history strategies, not based on
their native status. Therefore, river regulation does not necessarily increase cover of non-
native woody plant species.
I organized seven guidelines for river restorationists based on a comprehensive
review of riparian research. These guidelines focus on restoration principles relevant to
woody riparian vegetation and are particularly applicable for regulated rivers where flow
regime can be altered for restoration purposes. I advocate formulation of alternative
flood regimes based on knowledge of natural variability, consideration of effects of iv increased fire frequency in regulated systems and potential opportunities to use fire and floods to reinvigorate plant establishment and geomorphic processes, and use of functional groups and consideration of multi-trophic species interactions to predict effects of management scenarios. Watershed-scale and site-scale restoration approaches are often required for restoration of connectivity, variability, and geomorphic processes. The most promising strategy for riparian restoration planning, implementation, and monitoring remains adaptive management. However, the temporal scale of adaptive management must extend to incorporate climate change scenarios.
v
OVERVIEW
The majority of this dissertation focuses on the ecology of non-native, invasive
riparian plants, river regulation, and the potential for restoration. For chapters one and
two, I researched different aspects of the Tamarix invasion along the Colorado River through the Grand Canyon. Tree-ring analyses and surveys for Tamarix seedling
establishment and adult density were used to understand the influences of hydrologic,
climatic, and geomorphic factors on Tamarix establishment and persistence. The second
chapter examines the spatial association of beavers, Tamarix , and Salix and the hypothesis that beavers may contribute to Tamarix dominance through selective foraging
of Salix . The third chapter describes a regional survey of river segments in the
southwestern US. Through this survey I addressed relationships between flow regime,
climatic factors, and changes in flow regime with dominance of native and non-native
woody species. Chapter four provides seven principles for riparian restoration for
restoration practitioners. This chapter is based on a comprehensive literature review of
riparian research.
Important questions addressed by this study include: 1) How has the regulated
flow regime along the Colorado River in the Grand Canyon influenced Tamarix
establishment and persistence? 2) Do bi-trophic interactions contribute to Tamarix dominance? 3) How does the degree of flow alteration affect dominance of non-native, invasive woody plants?
vi
ACKNOWLEDGEMENTS
I am grateful to my future husband, Chris Kratt, who made many sacrifices so that
I could pursue this degree. Chris spent weeks in the field with me digging up Tamarix , driving all over the southwestern US, and cutting Tamarix slabs. He also provided unending encouragement when I was frustrated with the magnitude of my Ph.D. project.
Without Chris’s support, I do not know how I would have accomplished the tasks necessary to receive this degree. My mother, Julia Mortenson, also assisted with field work and has always encouraged me to pursue my goals. I thank her for a lifetime of support and excellent advice.
The exciting nature of this Ph.D. project and constant enthusiasm of my advisor,
Peter Weisberg, were integral to completion of this research. Peter played an active role in my research and professional development. I especially appreciate all of the time that he spent with me in the field. Thanks also for the speedy return of drafts which allowed me to expedite the Ph.D. process. I enjoyed our weekly meetings that often produced new and exciting ideas or research directions.
My graduate committee (Jeanne Chambers, Beth Leger, Ashley Sparrow, and
Tom Bullard) actively participated throughout the Ph.D. process. My research and manuscripts have benefited tremendously from their involvement. Thanks also to Larry
Stevens who shared his passion for the Grand Canyon. Steve Jenkins was also actively involved in my Ph.D. research, and the ideas for my favorite dissertation chapter were conceived in his research design course. Past and present members of the Great Basin
Landscape Ecology Lab all assisted in various ways. I am grateful for the friendships of
Amy Leist, Blake Engelhardt, & Kristen Schmidt. Thanks for cheering me on ladies! vii
TABLE OF CONTENTS
ABSTRACT……………………………………………………………………………….i
OVERVIEW……………………………………………………………………………....v
ACKNOWLEDGEMENTS………………………………………………………………vi
LIST OF TABLES……………………………………………………………………….xii
LIST OF FIGURES…………………………………………………………………...... xv
CHAPTER 1: FLOODS, PRECIPITATION, AND GEOMORPHOLOGY INFLUENCE
TAMARIX ESTABLISHMENT AND DISTRIBUTION IN THE GRAND
CANYON…………………………………………………………………………1
ABSTRACT ……………………………………………………………………....1
INTRODUCTION………………………………………………………………...2
Controls of riparian woody plant establishment…………………………..2
The Tamarix invasion in riparian landscapes……………………………..3
STUDY AREA……………………………………………………………………5
QUESTIONS……………………………………………………………………...8
METHODS………………………………………………………………………11
Retrospective study of Tamarix establishment…………………………..11
Correlation of seedling establishment and adult abundance with
geomorphic variables…………………………………………….13
Statistical analyses……………………………………………………….14
RESULTS………………………………………………………………………..15
Influence of hydrology, climate, and flow stage on Tamarix……………15
Spatial patters of recent seedling establishment…………………………17 viii
Current distribution of Tamarix in Grand Canyon………………………17
Comparison among river systems………………………………………..18
DISCUSSION……………………………………………………………………19
Flow regimes and Tamarix establishment in Grand Canyon…………….19
Hydrogeomorphic influences on Tamarix and Salix establishment……..22
Flow regimes and Tamarix establishment among river systems………...24
Management implications………………………………………………..25
ACKNOWLEDGEMENTS……………………………………………………...28
REFERENCES…………………………………………………………………..29
TABLES…………………………………………………………………………35
FIGURE CAPTIONS……………………………………………………………40
FIGURES………………………………………………………………………..41
CHAPTER 2: DO BEAVERS PROMOTE INVASION OF NON-NATIVE TAMARIX
IN THE GRAND CANYON RIPARIAN ZONE?...... 46
ABSTRACT ……………………………………………………………………..46
INTRODUCTION……………………………………………………………….47
Study Area……………………………………………………………….50
METHODS………………………………………………………………………51
GIS data layers…………………………………………………………...51
Statistical analysis………………………………………………………..54
RESULTS………………………………………………………………………..56
Abundance and distribution of Tamarix and Salix ………………………56
Distribution of beavers along the Colorado River……………………….57 ix
Associations among beavers, Tamarix , and Salix ………………………..57
DISCUSSION……………………………………………………………………59
Spatial relationships among beavers, Tamarix , and Salix ………………..59
Recommendations for further research and management………………..61
ACKNOWLEDGEMENTS……………………………………………………...62
REFERENCES…………………………………………………………………..63
TABLES…………………………………………………………………………66
FIGURE CAPTIONS……………………………………………………………68
FIGURES………………………………………………………………………...69
CHAPTER 3: DOES RIVER REGULATION INCREASE DOMINANCE OF
INVASIVE WOODY SPECIES IN RIPARIAN LANDSCAPES?...... 73
ABSTRACT ……………………………………………………………………..73
INTRODUCTION……………………………………………………………….74
Study Area……………………………………………………………….78
METHODS………………………………………………………………………80
Dams and flow regimes………………………………………………….80
Hydrology and woody vegetation………………………………………..81
RESULTS……………………………………………………………………….84
Effects of dams on flow regime………………………………………….84
Vegetation patterns across river segments……………………………….85
Effects of hydrology on woody vegetation………………………………86
DISCUSSION……………………………………………………………………88
Management implications………………………………………………..92 x
ACKNOWLEDGEMENTS……………………………………………………...93
REFERENCES…………………………………………………………………..94
TABLES…………………………………………………………………………99
FIGURE CAPTIONS…………………………………………………………..106
FIGURES………………………………………………………………………107
CHAPTER 4: GUIDELINES ECOSYSTEM RESTORATION OF REGULATED
RIVERS…..…………………………………………………………………….111
ABSTRACT ……………………………………………………………………111
INTRODUCTION……………………………………………………………...112
GUIDELINES FOR RIVER RESTORATION………………………………...115
CONCLUSIONS……………………………………………………………….128
REFERENCES…………………………………………………………………129
TABLES………………………………………………………………………..136
FIGURE………………………………………………………………………...138
APPENDIX A: APPENDIX FOR CHAPTER 1……………………………………….139
Soil characteristics……………………………………………………………...139
Years Tamarix established according to geomorphic reach……………………140
Life history characteristics of riparian shrub and tree species………………….141
APPENDIX B: APPENDIX FOR CHAPTER 3……………………………………….143
Species sampled………………………………………………………………...143
Constancy of focal taxa………….……………………………………………..144
Relative cover of all species…………..………………………………………..145
Hydrologic indicators for period of gage records…..…………………………..151 xi
Hydrologic indicators prior to dam construction or 1962…………………...... 152
Hydrologic indicators after dam construction or 1962…………….…………...153
Eigenvectors for hydrologic indicators from principal components analysis…..154
xii
LIST OF TABLES
Chapter 1
Table 1. Characteristics of geomorphic reaches of the Grand Canyon……………35
Table 2. Explanatory variables used in analyses…………………………………..36
Table 3. Relative importance of each explanatory variable for the annual frequency
of Tamarix establishment ………………………………………………..37
Table 4. Comparison of most plausible generalized linear models for Tamarix
establishment……………………………………………………………..38
Table 5. Relative importance of each explanatory variable for adult Tamarix
density, Tamarix seedling presence, and Salix seedling presence……….38
Table 6. Predicted responses of common riparian shrubs and trees of the
southwestern US to flood scenarios……………………………………...39
Chapter 2.
Table 1. Spatial variables used in the nested multiple linear regression models…66
Table 2. A priori candidate models explaining variation in Salix cover according to
various combinations of geomorphic and vegetation variables…………67
Chapter 3.
Table 1. Characteristics of river segments sampled………………………………99
Table 2. Descriptions of the Indicators of Hydrologic Alteration (IHA)………..101
Table 3. Species characteristics and number of river segments at which the species
was found……………………………………………………………….102
Table 4. Pearson correlation coefficients and significance for hydrologic and
climatic variables with vegetation cover………………………………..103 xiii
Table 5. Mean differences in the Indicators of Hydrologic Alteration (IHA)
parameter medians before and after dam construction or 1962 for non-
dammed rivers…………………………………………………………..104
Table 6. Variables included in the three best generalized linear models of relative
vegetation cover according to AIC c model comparisons………………105
Chapter 4.
Table 1. Seven guidelines for riparian restoration……………………………….136
Table 2. Framework for grouping plant species according to life history
characteristics…………………………………………………………...137
Appendix A
Table 1. Average soil characteristics in hydroriparian zone of Green and Colorado
Rivers in Canyonlands National Park and the Colorado River in Grand
Canyon National Park…………………………………………………..139
Table 2. Relative importance of each soil variable across models for presence /
absence of S. exigua and Tamarix seedlings in Canyonlands and Grand
Canyon………………………………………………………………….139
Table 3. Years Tamarix establishment was documented in each geomorphic
reach…………………………………………………………………….140
Table 4. Life history traits of common riparian shrubs and trees of the southwestern
US………………………………………………………………………141
Appendix B
Table 1. Species sampled in river survey with abbreviations and native status…143
Table 2. Constancy of focal taxa among river segments………………………...144 xiv
Table 3. Relative area-weighted cover (%) of all species sampled along twenty
river segments…………………………………………………………..145
Table 4. Hydrologic indicators for all un-dammed and dammed river sections from
the entire period of gage records………………………………………..151
Table 5. Hydrologic indicators for dammed river segments prior to dam
construction and un-dammed river segments prior to 1962…………….152
Table 6. Hydrologic indicators for dammed river segments following dam
construction and un-dammed river segments after 1962……………….153
Table 7. Eigenvectors for hydrologic indicators from principal components analysis
of alteration in flow characteristics……………………………………..154
xv
LIST OF FIGURES
Chapter 1.
Figure 1. Annual hydrographs of the Colorado River near Grand Canyon, AZ and
Green River at Green River, UT…………………………………………41
Figure 2. Map of study area………………………………………………………...42
Figure 3. Number of Tamarix established by year in the Grand Canyon………….43
Figure 4. The number of sites that experienced Tamarix establishment in each year
from 1984 to 2006 according to a) July – September peak flows and b)
annual peak flows………………………………………………………..44
Figure 5. Elevations of accurately-aged Tamarix samples above most recent high
water line according to year of establishment……………………………44
Figure 6. Dendrochronology results of Tamarix establishment along the Colorado
River in Grand Canyon and Green River below Flaming Gorge Dam…..45
Chapter 2.
Figure 1. The Colorado River through the Grand Canyon…………………………69
Figure 2. Occurrence of beaver and Salix , and cover of Salix and Tamarix , for a
representative section of the Colorado River…………………………….70
Figure 3. Proportion of riparian surfaces covered by Tamarix and Salix and number
of beaver occurrences divided by length of each geomorphic reach…….71
Figure 4. Ripley’s K-function for beaver and Salix across lag distances from 100 –
500 meters……………………………………………………………….72
Chapter 3.
Figure 1. Map of study sites………………………………………………………107 xvi
Figure 2. Relative area-weighted canopy cover of the most common non-native
(Tamarix and Elaeagnus ) and native ( Populus and S. exigua ) taxa……108
Figure 3. PCA plot of the first two principal component axes for hydrologic
variables “pre” and “post” dam construction or “pre” and “post” 1962 for
non-dammed river segments……………………………………………109
Figure 4. Regressions of vegetation cover on the Euclidean distance of Indicators of
Hydrologic Alteration parameters………………………………………110
Chapter 4.
Figure 1. Conceptual model for riparian restoration with an emphasis on riparian
plants……………………………………………………………………138
1
CHAPTER 1. FLOODS, PRECIPITATION, AND GEOMORPHOLOGY
INFLUENCE TAMARIX ESTABLISHMENT AND DISTRIBUTION IN THE
GRAND CANYON
ABSTRACT
Along highly regulated rivers, the decoupling of climate and hydrology combined
with introduction of non-native species create novel abiotic and biotic conditions.
Tamarix , a non-native shrub, dominates plant communities along many waterways in the southwestern US including the Colorado River through Grand Canyon. We combined a retrospective tree-ring study with surveys of seedling and adult distribution to describe the relative influences of climate, hydrology, and geomorphology on Tamarix establishment and persistence in Grand Canyon. Floods in 1983 through 1986 were associated with high mortality but also initiated a large establishment event. Tamarix establishment has been low but relatively continuous in subsequent years, excluding a peak in 2000 caused by a planned April flood and steady summer flows. Geomorphic diversity likely contributes to years of consecutive Tamarix establishment despite variable flow conditions.
From 1984 to 2006 Tamarix establishment was greatest when years of high, late- summer flows were followed by years of low precipitation. This combination provided moist surfaces for establishment of Tamarix during the late summer peak in seed dispersal and likely reduced erosion and scour of newly-germinated seedlings the following year. The annual peak flow of major tributaries, the primary source of fine sediment in this highly regulated river, positively influenced establishment at certain flow stages. Restoration floods for ecosystem restoration in March 1996 and November 2004 2 were associated with low levels of establishment. To decrease future Tamarix
establishment, water managers should avoid floods during the period of seed release
(April through September) which encompasses the timing of historic floods. However,
high relative importance of precipitation and geomorphology for Tamarix establishment
suggests that restoration floods alone cannot prevent Tamarix spread.
INTRODUCTION
Controls of riparian woody plant establishment
Complex interactions among hydrology, geomorphology, climate, and competition influence riparian plant community composition, structure, and species abundance. The reproductive phenologies of disturbance-adapted woody plants such as
Tamarix , Populus , and Salix are intricately tied to flow regime characteristics, particularly the timing and magnitude of floods, because their seeds are short-lived and require open, moist areas for germination. Germination sites must be available during the short period of seed release (approximately two to three months for Populus and Salix and six months for Tamarix ) while seeds are viable which is often less than four weeks
(Shafroth et al ., 1998; Guilloy-Froget et al ., 2002; Karrenberg et al ., 2002). Floods result in scouring forces, sediment deposition, increased moisture, and elevated water tables.
Woody plant establishment occurs when rare, timely flood events are of sufficient magnitude to deposit sediments at stage elevations where subsequent floods will not remove seedlings and saplings. Riparian shrubs often establish on fine-textured substrates with a high water-holding capacity at higher elevations, or on coarser 3 substrates (e.g., cobble bars) that are not subject to the scouring that occurs nearer to the river channel (Scott et al ., 1997).
Even if germination occurs, seeds often establish in locations that are ultimately unfit for long-term survival and recruitment. Seedling mortality in the Salicaceae is very high, often approaching 100% in riparian habitats (Cooper et al. , 1999; Johnson, 2000;
Karrenberg et al ., 2002; Polzin & Rood, 2006). Agents of riparian seedling mortality include inundation, scour, burial, and drought as well as competition for light, water, and nutrients (Hupp & Osterkamp, 1996). The rate of drawdown following floods strongly influences the probability of seedling survival in semi-arid regions. Root growth must equal the rate of groundwater decline (i.e., recession rate) which is influenced by sediment texture (Mahoney & Rood, 1998). Precipitation also can interact with flow regime to provide necessary moisture for seedling establishment (Baker, 1990; Johnson,
2000). Riparian plants become more resistant to environmental stresses with age
(Johnson, 2000; Levine & Stromberg, 2001). When plants grow larger they trap sediments, stabilize geomorphic surfaces, and prevent future scour (Graf, 1978; Edwards et al., 1999). Frequent disturbances prevent competitive exclusion from regulating diversity, particularly in productive environments (Poff & Ward, 1989; Huston, 1994;
Bornette et al ., 2008). However, if disturbance frequency or intensity decreases,
competition and other biotic processes become more influential (Stanford et al. , 1996).
The Tamarix invasion in riparian landscapes
In the southwestern US, the composition and abundance of riparian plants has
changed as a result of flow regime alteration and the invasion of non-native invasive 4 plant species, especially Tamarix . Tamarix are arborescent shrub species native to
Eurasia that have spread prolifically near springs, lakes, rivers, reservoir deltas, and other moist habitats in western North America, Mexico, and Australia. Many species and hybrids of Tamarix grow in the introduced range; the most common plants in the
American Southwest are Tamarix ramosissima X Tamarix chinensis hybrids (Gaskin &
Schall, 2002). Although the ecological effects of Tamarix invasion are currently disputed
(Stromberg et al ., 2009), the dominance of this shrub in riparian habitats necessitates a thorough understanding of its autecology and interactions with local environmental factors. Research associated with Tamarix invasion may also provide insight into management of other riparian invasions in which life history strategies are intricately tied with disturbance regimes.
The invasion of Tamarix spp. is attributed to the broad ecological amplitude of the species and changes in disturbance regimes (e.g., construction of dams) coincident with naturalization. Tamarix spp. are disturbance-adapted but also have high drought and salinity tolerance (Glenn & Nagler, 2005). Tamarix has a longer period of seed release than many native shrubs that are adapted to historical spring floods (Howe & Knopf,
1991; Shafroth et al., 1998; Roelle & Gladwin, 1999). Seed availability throughout the growing season may confer a selective advantage to Tamarix along regulated rivers that no longer undergo spring floods and have midsummer peak flows when hydroelectric energy needs are higher. Young Tamarix are inferior competitors when compared with native, riparian trees of the southwestern US (Stevens, 1989; Sher & Marshall, 2003;
Bhattacharjee et al ., 2008; DeWine & Cooper, 2009), but the reproductive phenology of
Tamarix is well-suited to altered hydrologic regimes that hinder or preclude native shrub 5 establishment (Stromberg et al., 2007). Adult Tamarix dominance is highest along rivers
with a moderate degree of hydrologic alteration, whereas Populus dominance is highest along rivers with minimal alteration (Merritt & Poff, 2010; Mortenson & Weisberg,
2010). However, Tamarix recruitment is limited along highly altered rivers due to lack of floods that create establishment sites (Merritt & Poff, 2010).
Previous dendrochronological investigations identified aspects of the hydrologic regime and climate that influenced Tamarix establishment on the Green and Yampa
Rivers (Cooper et al., 2003; Birken & Cooper, 2006). On various segments of the regulated Green River, the discharge of maximum flows during, the year prior to, and the year following Tamarix germination explained the presence of Tamarix recruits in constrained reaches. In unconstrained reaches, precipitation during July through August also influenced Tamarix presence (Cooper et al., 2003). Further downstream along the
Green River, high magnitude peak flows followed by low peak flows the next year initiated Tamarix establishment (Birken & Cooper, 2006). These conditions correspond with the requirement of Tamarix and other common pioneer shrubs for bare, moist sites for germination (provided by high flows) and safety from scour and burial (provided by subsequent low flows).
STUDY AREA
The study area encompasses the riparian habitat along the Colorado River in the
Grand Canyon National Park from Lees Ferry to 384 km downriver (Figure 1). Grand
Canyon boasts a long period of flow records and high geomorphologic diversity, making it an ideal study site for understanding the influence of hydrogeomorphology and other 6 environmental factors on Tamarix recruitment and survival. The channel morphology of the Colorado River is characterized by alternating pools and rapids (Leopold, 1969).
Abundant debris from side-valley fan deposits constrict the river and create areas of sediment deposition upstream and downstream of rapids. Three types of fluvial deposits
(tributary fans, cobble bars, and fine-grained sediments) (Howard & Dolan, 1981) are available for plant establishment. Schmidt & Graf (1990) identified six geomorphic units based on type and origin of fluvial deposits: terrace, sandbar, return-current channel, channel margin, debris fan, and cobble bars. The grain-size composition and elevation above the river channel of these units influences plant composition. At a larger scale, geomorphic reaches are differentiated by bedrock composition, channel width, gradient, and major tributaries (Schmidt & Graf, 1990). Wider reaches have lower gradients, greater area of deposition, and more tributaries.
Prior to completion of the Glen Canyon Dam just upriver of Grand Canyon in
1963, the flow regime of the Colorado River through Grand Canyon was dominated by spring, snowmelt floods and fall sediment inputs from monsoons (Topping et al. , 1999).
Spring floods caused erosion and mobilized sediments in the riverbed, while summer flash floods added fine-grained sediments (Howard & Dolan, 1981). The pre-dam
Colorado River transported a large suspended sediment load (~143 million tons per year
[Leopold, 1969]). Suspended sediment load decreased by 90% after dam construction, and fine sediments are now provided primarily by tributaries (Schmidt & Graf, 1990).
Regulation of the Colorado River has dramatically decreased the magnitude of high flows, increased the daily flow fluctuation, changed the season in which peak flow occurs
(Figure 1), and trapped fine sediments. The current flow regime consists of low 7 magnitude flows with seasons of increased magnitude dependent on hydropower needs
(e.g., Figure 1d).
There is a tight relationship between the Colorado River hydrograph and vegetation mortality and establishment. The high magnitude floods on the pre-dam
Colorado River resulted in annual scouring of riparian vegetation (Johnson, 1991). The completion of Glen Canyon Dam decreased flooding frequency and magnitude and caused an increase in riparian vegetation cover. Tamarix invasion of lower riparian zones along the Colorado River began after post-dam floods in 1965 and 1973 (Martin, P. and
B. Hayden, written communications), and the plant was widespread in Grand Canyon by the 1980s (Turner & Karpiscak, 1980). Tamarix is the dominant species in riparian communities of Grand Canyon, often growing in mixed patches with native shrubs.
Tamarix reproductive phenology along the Colorado River corridor varies according to elevation and height above river. In Grand Canyon, seed dispersal peaks in late May and early June. Plants on lower surfaces release seed throughout the growing season with a short late-summer peak (Warren & Turner, 1975; Stevens & Siemion, in prep ).
The Colorado River through Grand Canyon is unlike many riparian areas in southwestern US due to the nearly complete absence of riparian trees. Pre-dam records of Grand Canyon flora note Populus fremontii presence at tributary confluences and several sites along the mainstream where it is infrequent today (Turner & Karpiscak,
1980). Salix gooddingii was formerly somewhat more common, but the few remaining stands are threatened by beaver foraging, lack of springtime recruitment floods, and post- dam coarsened sand substrata (Stevens, 1989; Mast & Waring, 1997; Mortenson et al ., 8
2008). Other common native riparian shrubs in Grand Canyon include Baccharis spp.,
Salix exigua , Pluchea sericea , Prosopis glandulosa, and Celtis laevigata .
In recent decades four restoration floods have been implemented by the Bureau of
Reclamation in association with an adaptive management working group. These floods varied in timing, magnitude, and duration. High magnitude, short duration floods in
March 1996, November 2004, and March 2008 were intended to rebuild sandbars by redistributing fine sediments from the bottom of the river channel. Managers also hoped that floods would provide soil moisture to high-elevation, pre-dam vegetation (Stevens et al ., 2001). The timing, low magnitude (1274 m 3s-1), and brief duration (March 26 to
April 2) of the 1996 flood temporarily raised the elevation of sandbars (Stevens et al .,
2001; Figure 1e). Restoration floods in 2004 and 2008 were of similar duration and magnitude and had similar results. A low magnitude (915 m 3s-1) flow pulse in April, low, steady flows (226 m3s-1) from June – September and a four-day high flow (850 m 3s-1) in
late September (Figure 1f) were implemented in 2000. This release was designed to
enhance native fish populations through preservation of backwater nurseries (Stevens &
Gold, 2003). We were particularly interested in how these restoration floods affected the
abundance of flood-adapted Tamarix .
QUESTIONS
We investigated patterns of seedling establishment and persistence of Tamarix at
a variety of spatial and temporal scales to understand abiotic and biotic controls on
Tamarix distribution and abundance along a 363-km stretch of the Colorado River in 9
Grand Canyon (Figure 2). A landscape-level dendrochronological analysis of Tamarix was used to address the following questions:
1) How have the specific flow characteristics of the regulated Colorado River, including
restoration floods, influenced the probability of Tamarix establishment and persistence in
the past? 1b ) How do precipitation, temperature, tributary input, or river stage further influence the likelihood of Tamarix establishment in Grand Canyon National Park? We
expected hydrologic variables to exert the greatest influence over Tamarix establishment.
Unfortunately, a tree-ring analysis of Salix exigua establishment, a native riparian shrub
that grows in similar habitats as Tamarix , was infeasible because of its ability to spread
clonally.
2) How are Tamarix and Salix distribution patterns (seedlings and adults) related
to geomorphic variables? The history of long-term survival and mortality potentially
confounds establishment studies that are based on the ages of adult trees (Harper, 1977).
Studies of seedling establishment should be linked with investigations of age structure
(e.g. dendrochronology) to truly understand the spatial and temporal patterns of plant
recruitment. Therefore, we investigated the spatial distribution of seedlings and adults
through surveys for current Tamarix seedlings (0 to 5 years) and transects of adult
Tamarix density. We also assessed seedling establishment patterns of Salix exigua . We
were interested in the importance of geomorphic characteristics for Tamarix and Salix
seedling distribution and adult Tamarix density. Of particular interest was the relative
importance of water and sediment inputs from unregulated tributaries, over which river
managers have no influence. 10
Because Tamarix and Salix require fine sediments and seeds to establish, we
predicted that tributaries would have positive, localized effects on Tamarix , and that
Tamarix density and frequency of Tamarix establishment would be greater at sites farther downstream. We considered tributary density within five km upstream because this distance corresponds approximately with the scale of pool and riffle alternation (Leopold
1969) but is coarse enough to allow for variation in tributary density. High debris fan densities and sediment limitation from the main channel led us to hypothesize that hillslope processes would affect Tamarix density and Tamarix and Salix seedling
presence in Grand Canyon. The geomorphic reach of a site exhibits many coarse-scale
controls on vegetation through river channel dynamics (Richards et al ., 2002). Tamarix
establishment should be prevalent in unconstrained reaches that contain more potential
establishment sites. Other, fine-scale geomorphic variables that vary within reaches were
also considered (channel width, gradient, and sinuosity).
3) How does seedling establishment and adult abundance of Tamarix in the
highly regulated Colorado River downstream of Glen Canyon Dam compare with other
less regulated river sections in the Colorado River Basin? We expected riparian
vegetation to be influenced by the highly regulated flows of the Colorado River below
Glen Canyon Dam in different ways than in other, less regulated sections that continue to
experience annual, spring floods. We compared patterns of Tamarix and S. exigua
seedling establishment along the Colorado and Green Rivers in Canyonlands National
Park with those in Grand Canyon. We also considered differences in establishment and
persistence of Tamarix in Grand Canyon (this study) and the Green River (Birken &
Cooper, 2006). The comparison of Tamarix establishment patterns in the Green River 11 and the Colorado River permitted generalizations about drivers of Tamarix establishment
across river systems.
METHODS
Retrospective study of Tamarix establishment
Sampling sites for the tree-ring study were randomly selected by river mile across
226 miles with the constraint that recreation areas and historic southwestern willow
flycatcher breeding sites were avoided. Then, we stratified the sites to ensure that all
geomorphic units (terrace, sandbar, return-current channel, channel margin, debris fan,
and cobble bars) and geomorphic reaches (Table 1) were representatively sampled. Four
field expeditions were conducted in spring and fall of 2006 and 2007, and 43 sites and
409 Tamarix individuals were sampled. Sites varied in extent because sites varied in the
area needed to include all geomorphic units present. Three representative trees from all
Tamarix size classes on each geomorphic unit were selected for tree-ring sampling and excavated with hand tools. Precise measures of Tamarix age required collection of cross- sections from below the germination point (i.e., root crown) to the soil surface. The elevation of the root crown above the high water line was determined with surveying equipment, and the GPS location (precision from 2-59 m) was recorded.
Processing of tree-ring samples involved sectioning the cross section into approximately 5-cm thick slabs with a band saw. For most samples, this process yielded four to five root crown sections and two surface slabs for cross-dating within each individual. Slabs were sanded with progressively finer sand paper (80 to 600 grit) in preparation for ring counting. The original position of each slab on the plant was 12 recorded, and rings were counted by two observers with the aid of a stereo binocular microscope. A decrease in the number of rings and absence of pith towards the lower portion of the sample indicated presence of root crown. The maximum number of rings at the root crown revealed the precise age of the sample. We cross-dated within samples of the same individual but were unable to cross-date among most Tamarix individuals.
Many samples were impossible to age due to rot, insect damage, compressed rings from
burial, and inability to collect the root crown; these samples were not used in statistical
analyses.
The response variables for the statistical analyses were designed to elucidate how
hydrologic conditions and flow stage elevation affected Tamarix establishment and
persistence. Stevens and Waring (1985) estimated that Tamarix sustained 45% mortality
from the 1983 – 1985 floods. Because of high mortality of pre-1983 Tamarix , we
analyzed the number of sites that experienced Tamarix establishment in each year from
1984 to 2006. We used a GIS layer derived from an USGS flow stage model to divide
the samples among flow stage elevations based on GPS locations. These elevations
corresponded with pre- and post-dam discharges, post-dam flow discharges, and post-
dam restoration flood discharges respectively (see Figure 1) and served as a proxy for
flooding frequency, inundation duration, and water availability.
We considered climate variables that influence moisture availability, including
summer maximum temperature and precipitation the year of and following germination
(Table 2). Hydrologic variables were based on a January to December water year, and
some variables were calculated using Indicators of Hydrologic Alteration software
(Smythe Scientific Software, Boulder, CO). We included annual peak flow the year prior 13 to, during, and following germination. We expected high flows before and during the year of germination to positively influence establishment through the creation of bare surfaces for seed germination. High peaks the year following germination were expected to have a negative effect. Higher magnitudes of annual minimum flow were predicted to favor seedling establishment through increased water availability. Timing of maximum and minimum flows and the magnitude of maximum flow occurring during spring (April
– June) and summer (July – September) were included in the models to incorporate the importance of flow seasonality with respect to reproductive phenology. The duration, frequency, and rate of flows also were considered. We expected major tributaries (Paria and Little Colorado Rivers) to influence Tamarix establishment by providing fine sediment which is limited in the post-dam Colorado River.
Correlation of seedling establishment and adult abundance with geomorphic variables
Surveys for seedlings and adult abundance were non-destructive and, therefore, were not restricted to the site locations sampled for the tree-ring study. Instead, field sampling sites were randomly located throughout the river. We searched for Tamarix and
Salix seedlings (0 to 5 years) or vegetative recruits at 217 sites throughout the study area
on a randomly-selected side of the river. Sites extended 20 m parallel to the river channel
and varied in width according to the width of riparian habitat.
During fall 2006 we sampled adult Tamarix density at 316 “floating transects” in
Grand Canyon. Transects extended approximately 100 m on both sides of riparian shoreline. We counted the number of Tamarix individuals to quantify adult Tamarix 14 density while floating down the river. Individuals were differentiated by phenological state and variations in foliage or flower color. The clonal nature of Salix precluded
similar density estimates. GPS locations of all sites were recorded with an average
precision of 24 m; greater spatial precision was not possible because of the deep canyon
environment. We also conducted seedling surveys and floating transects along the Green
and Colorado Rivers in Canyonlands National Park during fall 2008.
We determined geomorphic characteristics of seedling survey and floating
transect sites (Table 2). The geomorphic reach and distance downstream from Lees Ferry
were interpreted for each site with 2002 1-m resolution orthophotographs and GIS layers
provided by the U.S. Geological Survey Grand Canyon Monitoring and Research Center
using ArcGIS software (ESRI, Redlands, CA). Channel width was calculated with a
combination of ET GeoTools (ET Spatial Techniques, Pretoria, South Africa) and
ArcToolbox applications. A 5-m DEM was used to determine channel gradient at the
scale of 5.8 km (30 times the maximum channel width). Sinuosity was calculated as the
distance along the river centerline 1.65 km upstream divided by the Euclidean distance.
The distance from nearest major tributary (drainage area > 30 km 2) and density of moderately-sized tributaries (drainage area > 10km 2) 5-km upstream was determined
using tributary information from Webb et al . (2000).
Statistical analyses
Generalized linear regression analyses were used to discern the influence of hydrograph properties, climate factors, and flow stage on Tamarix establishment and the influence of geomorphic variables on Tamarix and Salix distribution. Separate 15 generalized linear models were created to define environmental conditions that allowed
Tamarix to establish and persist at different elevations above the river channel (all elevations, 226.5 – 708 m 3s-1, and 708 – 1,274 m 3s-1 flow stages). Generalized linear models were created for possible combinations of explanatory variables (Table 2), and
Akaike’s information criterion (AIC) scores were determined for each model using R software. AICc, an index recommended for small sample sizes (Burnham & Anderson,
2002), and Poisson distributions, which are appropriate for right-skewed count data
(Ramsey & Schafer, 2002), were used for the tree-ring analyses. We used logistic regression for analyses where the response variable was presence or absence of seedlings.
We calculated the relative importance of explanatory variables using the sum of AIC weights for possible models as demonstrated by Burnham & Anderson (2002).
Nagelkerke pseudo- R 2 values (see Nagelkerke, 1991) were calculated for the most
plausible models with the lowest AIC scores using the “Design” package in R.
RESULTS
Influence of hydrology, climate, and flow stage on Tamarix establishment and
persistence
Of 409 Tamarix individuals collected, 149 were accurately aged based on
unambiguous evidence of root crown in the sample. An additional 89 lacked such
evidence but were considered to be approximately aged. We found two Tamarix
individuals that had resprouted from buried stems and, therefore, did not contain root
crowns. We were unable to accurately age pre-dam Tamarix , and few trees were sampled
that predated the 1980s floods. At some sites, the root crowns of older Tamarix were 16 difficult to excavate because of deep sand deposits associated with 1980s and recent planned flooding. Tamarix sustained high establishment during 1983 through 1986 and
1999 through 2000 and had continuous, low levels of establishment in other years from
1987 to 2004 (Figure 3). A positive relationship between summer peak flow and
Tamarix establishment was primarily driven by high establishment from 1984 through
1986 and 1999 through 2000 (Figure 4a). The association between establishment and annual peak flow was also positive, but the R 2 was lower (Figure 4b).
Regression analyses of post-1983 Tamarix establishment (1984 – 2006)
demonstrated the overwhelming influence of summer precipitation in the year following
germination (Table 3). Tamarix establishment was greater in years that had lower
summer precipitation the following year. This variable was only influential in additive
models that contained measures of flow magnitude or, for lower flow stages, recession
rate or Paria River flow (Table 4). High magnitude maximum flows were positively
related to sites with Tamarix establishment in all flow stages and high flow stages but
were negatively related to sites with Tamarix establishment in low flow stages (Table 3).
In high flow stages, peak flow of the Little Colorado River was the strongest positive
predictor of Tamarix establishment. High establishment in 1993, when the Little
Colorado River flooded at a magnitude of 509 m 3s-1, influenced this response.
All surviving pre-dam Tamarix persisted on upper riparian zone terraces.
However, during the post-dam period, Tamarix establishment and persistence were
documented on every type of geomorphic unit (Figure 5). Only nine of the accurately-
aged Tamarix originated prior to 1983, and these individuals grew on high-elevation
terraces, sandbars, and channel margins. Establishment on channel margins occurred 17 continuously but was more frequent from 1983 through 1986 and 1999 through 2000.
Tamarix establishment on return current channels was infrequent, and only one Tamarix established on a sandbar from 1987 through 1999. Individuals growing on cobble bars were poorly represented in the tree-ring samples because the tree rings were often unreadable. However, a large cohort of Tamarix saplings was observed on cobble bars during all research trips. The elevation of Tamarix establishment above current river level was high from 1983 through 1986 and gradually decreased in recent years (Figure
5). Recent low-elevation establishment has occurred primarily on sandbars, cobble bars, and debris fans.
Spatial patterns of recent seedling establishment
Tamarix and Salix seedlings (0 to 5 years) were present at 51% and 20% of survey sites respectively. Tamarix seedlings were more likely to be present in spring
(65%) than fall (38%). Generalized linear models of Tamarix seedling presence based on geomorphic variables explained less variance than the intercept model. The sampling season did not play a role in Salix seedling presence, but geomorphic reach identity was influential (Table 5). Salix seedlings were more likely to occur in wide reaches, with the exception of Muav Gorge (Table 1). Tributaries did not influence the presence of Salix seedlings (Table 5).
Current distribution of Tamarix in Grand Canyon
Currently, Tamarix grows along 98% of the 363-km stretch of the Colorado River through Grand Canyon National Park and occurs at an average density of 34 and a 18 maximum density of 165 individuals per 100 m. According to AIC-derived variable importance values, Tamarix density is most strongly influenced by geomorphic reach
(Table 5). When geomorphic reach was excluded from the analysis, channel width was the most important explanatory variable (positive effect). The three widest reaches
(Marble Canyon, Furnace Flats, and Canyon) had the highest Tamarix density (Table 1).
Comparison among river systems
In most floating transects in Canyonlands National Park, Tamarix stands were too dense to differentiate individuals. Tamarix seedlings were similarly abundant at survey sites in Grand Canyon (51%) and the less regulated reaches of the Colorado and Green
Rivers in Canyonlands National Park (62%). However, Salix exigua seedlings were present in 20% of the survey sites in Grand Canyon compared with 87% of sites in
Canyonlands.
The oldest Tamarix that we sampled established in 1937 or earlier near Cardenas
Creek; we were unable to determine the accurate age due to rot in the root crown section.
This is consistent with the oldest Tamarix sampled in the Green River (1938) by Birken
& Cooper (2006). However, Tamarix established and persisted in every year from 1983 to 2006 in Grand Canyon but in only eight years from 1983 to 2004 along the Green
River (Birken & Cooper, 2006; Figure 6). To decipher potential hydrologic causes of these differences we compared annual hydrographs of the Green River and Colorado
River through Grand Canyon (Figure 1). Both rivers were dammed in the early 1960s.
Prior to dam construction, the Green and Colorado Rivers exhibited nearly identical hydrographs with distinct spring snowmelt floods in May to June (Figure1a). Following 19 dam construction, the Green River continued to peak, albeit at lower magnitudes, but the
Colorado River hydrographs were much more erratic (Figure 1c, d). Both rivers experienced floods reminiscent of pre-dam floods (similar timing, magnitude, duration) in 1983-1985 (Figure 1b). This was the most recent time period when the Colorado River through Grand Canyon experienced a natural flow regime.
DISCUSSION
Flow regimes and Tamarix establishment and persistence in Grand Canyon
We found a low number of surviving Tamarix that originated prior to 1983
(Figure 3), and previous documentation of abundant Tamarix prior to this time period
(Turner & Karpiscak, 1980; Webb, 1996) provides evidence of high Tamarix mortality during the 1983 through 1985 floods (Stevens & Waring, 1985). However, the same floods that caused mortality also created habitat for subsequent Tamarix establishment.
The largest remaining cohort of Tamarix established in 1985. Floods in the mid-1980s were associated with Salix gooddingii establishment in Grand Canyon (Mast & Waring,
1997), Tamarix and Acer negundo establishment along the Green and Yampa Rivers
(Cooper et al., 2003; Birken & Cooper, 2006; DeWine & Cooper, 2007), and Tamarix establishment throughout riparian corridors in the arid and semi-arid southwestern US
(Merritt & Poff, 2010). Floods are necessary for establishment pulses of native and non- native, disturbance-adapted riparian species and often instigate consecutive years of establishment (Scott et al., 1997; Edwards et al., 1999; Cooper et al., 2003; Polzin &
Rood, 2006). 20
In constrained channels, such as Grand Canyon, plant establishment is controlled primarily by flood deposition of fine sediment (Scott et al., 1996). However, the particular characteristics of floods (magnitude, timing, duration, recession rate) and their relationships to plant life history traits ultimately determine the probability of establishment. Restoration floods during March 1996 and November 2004 were timed to prevent high levels of Tamarix establishment. These floods failed to initiate a large pulse in establishment despite having similar magnitude as in 1984 through 1986 (Figure 4b).
The 1996 and 2004 floods occurred outside of the period of seed release, and bare sediments were most likely too dry for germination once seeds were available (Stevens et al., 2001). The results of these restoration floods exhibit the importance of the interaction between timing of peak flows and Tamarix reproductive phenology.
In Grand Canyon, high flows during July through September were a better indicator of the probability of Tamarix establishment than annual peak flow magnitude
(Table 3; Figure 4). Tamarix exhibits a second seed release peak in late summer (Warren
& Turner, 1975; Stevens & Siemion, in prep ), and high flows during summer allow
Tamarix to colonize surfaces unavailable to early-dispersing native riparian shrubs. Seed viability of Tamarix also may be highest during this time (Merkel & Hopkins, 1957).
High flows during late summer also maintain high water tables and increase water availability for seedlings. In regional surveys of woody vegetation in the southwestern
US, Mortenson & Weisberg (2010) and Merritt & Poff (2010) also observed greater dominance of Tamarix along rivers with high magnitude flows in late summer. Even though Tamarix has a long period of seed release, the seasonal characteristics of flows 21 interact with fluctuations in seed release magnitude and viability to determine the likelihood of establishment.
Many factors including erosion, drought, and inundation cause seedling mortality.
This study revealed the overwhelmingly negative influence of summer precipitation the year following germination on Tamarix establishment in Grand Canyon (Table 3). We predicted that precipitation during the second growing season would favor Tamarix establishment through increased water availability, but these results suggest that heavy precipitation may cause collapse of river banks and scour and burial of the prior year’s seedlings or pathogen-induced mortality. Late summer precipitation along the Upper
Green and Yampa Rivers also hindered Tamarix establishment in some reaches (Cooper et al., 2003). Tamarix may be vulnerable to erosion during the second growing season due to its ability to establish late in the summer. For seedlings that establish in fall, the first year of growth is abbreviated, and seedling roots are less likely to resist subsequent erosion. We did not find evidence for drought mortality driven by climatic factors, but fast recession of water levels had a strong negative effect in models of Tamarix establishment at low stage elevations (Tables 3, 4).
We observed Tamarix establishment on progressively lower-elevation surfaces similar to observations of Populus deltoides establishment along the Missouri River
(Figure 5; Scott et al., 1997). Scott et al . (1997) suspected that recent establishment on
lower surfaces was temporary and that seedlings and saplings would be removed during
future floods in this minimally-regulated river segment. Populus establishment along the regulated reach of the Green River below Flaming Gorge Dam was limited to low- elevation islands and cutbanks where mortality reached 100% (Cooper et al., 1999). 22
Cooper et al . (2003) also observed similar patterns in Tamarix establishment elevation, but these patterns were more apparent in constrained reaches. In Grand Canyon, recent
Tamarix establishment on low-elevation sand and cobble bars combined with reduced flooding may result in narrowing of the river channel in less-constrained reaches despite sediment limitation. We observed many Tamarix saplings on cobble bars that established during the 2000 steady flows (LES, pers.obs.). These seedlings grew to sufficient size and persisted through the 2004 test flood. The 2000 cohort currently stabilizes cobble bars. A large flood of similar magnitude to the mid-1980s floods would most likely remove these saplings; however, such floods are not presently planned by the river management team.
Hydrogeomorphic influences on Tamarix and Salix establishment and adult distribution
In reaches where hillslope processes are important, predictions of establishment based solely on stream flow characteristics are infeasible (Friedman et al., 2006). Reach- scale effects on Tamarix density and Salix seedling presence are apparent (Table 5).
Wider reaches have higher Tamarix density and are more likely to have Salix seedlings
(Table 1). Generally, wide reaches have lower channel gradients which encourage deposition of fine sediments (Hupp & Bornette, 2003) on which disturbance-adapted shrubs rely for establishment. The high percentage of sites with Salix seedlings in Muav
Gorge, a constrained reach, may be driven by sediment or propagule input from Kanab and Havasu Creeks, major tributaries that intersect the Colorado River in Muav Gorge.
The Muav Gorge also contains many reattachment and channel margin deposits (Schmidt 23
& Graf, 1990). The low frequency of S. exigua seedlings in Grand Canyon when compared to riparian habitats in Canyonlands National Park likely is caused by a combination of constrained reaches, reduction of fine sediments from Glen Canyon Dam, lack of spring flooding, and high salinity in Grand Canyon (Stevens, 1989; Mortenson &
Weisberg, 2009).
Geomorphic variables alone are poor predictors of Tamarix seedling presence
(Table 5). Tamarix is a generalist and is more drought and salt tolerant than native
shrubs (Busch & Smith, 1995; Shafroth et al., 1998). Tamarix seedlings can establish in
cracks of bedrock (pers. obs.). Moisture is the only requirement for seed germination,
however many requirements must be met for the subsequent survival of Tamarix
seedlings. Geomorphology influences Tamarix establishment but not spatial patterns of
establishment. Under the current flow regime, Tamarix seedlings are inundated by
seasonally high flows from May to August, resulting in significant seedling mortality.
Therefore, we were less likely to find seedlings during fall surveys. However,
germination can occur following subsidence of the seasonal increase (end of August).
This also corresponds with the final days of the North American Monsoon that transports
fine sediments from tributaries and hillslopes.
Tamarix seed availability does not limit establishment along the Green and
Yampa Rivers (Cooper et al ., 1999). Similarly, continuous distribution of adult Tamarix
throughout the Colorado River corridor through Grand Canyon excludes the mechanism
of seed limitation remediated by tributary input. However, the high relative importance
of annual peak flows of the Paria and Little Colorado Rivers (Table 3) and negligible
influence of tributaries on Salix seedling presence and Tamarix density (Table 5) may 24 indicate that tributaries are only influential during flood years. Additionally, fine sediments from tributaries may be transported in suspension for long distances (Topping et al., 1999) and negate potential effects of tributary proximity.
Flow regimes and Tamarix establishment among river systems
In Grand Canyon the largest number of Tamarix established during 1985 which was a year of high flow followed by years of low flow (Figure 3). These results are consistent with patterns of Tamarix establishment along the Green River (Birken &
Cooper, 2006). However, the recent continuous establishment in Grand Canyon is contrary to results of Birken & Cooper (2006) who documented years of establishment dictated by inter-annual hydrologic patterns and lack of establishment in surrounding years (Figure 6). The pre-dam flow regime in Grand Canyon, the period of flooding in the mid-1980s, and the current flow regime of the Green River at Green River, Utah are characterized by floods extending from May to July. These hydrologic conditions are ideal for Tamarix establishment. However, the annual occurrence of floods causes
mortality of the previous year’s seedlings. For example, an entire Populus cohort
established in 1993 was killed during restoration floods in 1994 along the Rio Grande
River (Taylor et al ., 1999). The current flow regime of Grand Canyon does not consist of
annual floods. Instead, low magnitude, daily fluctuating flows with seasons of increased
magnitude dependent on hydropower needs are status quo (Figure 1d). Years of
consecutive Tamarix establishment in Grand Canyon are more likely than along the
Green River due to lack of annual spring floods in Grand Canyon. 25
Hydrology alone does not explain Tamarix establishment patterns. Tamarix
establishment along the Yampa and Green River immediately below Flaming Gorge Dam
occurred in many years, and Cooper et al . (2003) attributed this pattern to geomorphic diversity. Conditions are suitable for establishment on some geomorphic units in most years (Figure 5). Tamarix establishment was more frequent in Gray Canyon than in other
segments sampled by Birken and Cooper (2006) along the Green River. Debris fan
complexes dominate these river segments as well as the Colorado River through Grand
Canyon. Geomorphic variability, such as occurs at debris fan complexes, increases the
probability that suitable microsites are available. Tamarix establishment patterns are
influenced by the interaction of hydrology and geomorphology, which modulate local
effects of a particular hydrologic event.
Management implications
In the highly regulated Colorado River through Grand Canyon, knowledge of the
hydrologic conditions that allowed colonization of Tamarix in the past may be applied to
minimize Tamarix establishment in the future through flow manipulation. Similarly,
knowledge of native species biology may allow predictions of species-specific responses
to flow (Table 6). Tamarix germination occurs in every year, but large cohorts of
Tamarix establish in years with high flows during the period of seed release. For
example, the restoration flows in 2000, which included increased flow in April (Figure
1f), allowed elevated levels of establishment (Stevens & Gold, 2003). The short-
duration, high flow in April corresponded with Tamarix seed release. To prevent large
cohorts of Tamarix from establishing in the future, floods during the time of Tamarix 26 seed release (April through September) should be avoided. However, lack of floods during this time may also prohibit establishment of native pioneers ( Populus , Salix )
(Table 6). We recommend “unnatural” floods (flooding outside of the growing season
[November through March]) that benefit clonal expansion of native shrubs (e.g., Salix exigua , Pluchea sericea ) in Grand Canyon (see Barsoum, 2002). If these floods are of similar magnitude and duration as the 1996 restoration flood, they should not negatively affect fish populations, including the endangered humpback chub (Valdez et al ., 2001).
Moreover, winter floods may become more frequent with global change due to increased precipitation falling as rain and earlier melting of snowpack (Nijssen et al ., 2001).
The most natural flow regime that has occurred in Grand Canyon since construction of the Glen Canyon Dam (mid-1980s floods) initiated high levels of
Tamarix establishment. Similarly, reinstatement of the natural flow regime in riparian areas of southwestern US may encourage Tamarix establishment. However, similar hydrologic conditions (floods) allow Populus , Salix gooddingii , Acer negundo , and non-
native Elaeagnus angustifolia to establish (Friedman & Lee, 2002; DeWine & Cooper,
2007). Recent studies demonstrate the inferior competitive abilities of Tamarix seedlings
and adults when compared with native riparian tree species (Stromberg , 1997; Sher &
Marshall, 2003; Bhattacharjee et al., 2008; DeWine & Cooper, 2009). In systems where
these species are abundant and currently recruiting, well-timed floods may ultimately
benefit native species. However, native riparian trees are rare in Grand Canyon and,
therefore, spring “natural” floods are detrimental to Tamarix control in this system.
Differences in species composition and climate confound the relationship between flow regimes and Tamarix establishment. Temperature and elevation influence seed 27 availability of many riparian woody species, and, in unregulated to minimally regulated systems, these factors relate closely with natural flow regimes (Stella et al., 2006;
Stevens & Siemion, in prep ). Along highly regulated rivers, the decoupling of climate
and hydrology combined with introduction of non-native species create novel abiotic and
biotic conditions. Predictions of flow treatment effects must take into account all of these
factors and must be site specific. The differences in our results compared to those of
similar studies along the Green and Yampa Rivers attest to the complicated nature of
these relationships. Annual fall releases of 25-day duration that inundate and may cause
mortality of Tamarix seedlings as suggested by Roelle & Gladwin (1999) would be an
effective management strategy for Tamarix in Grand Canyon. However, recent drought,
increased water extraction, and energy needs make this strategy infeasible. Although
annual floods similar to recent restoration floods would benefit Tamarix control efforts,
frequent floods would also degrade marsh habitat and likely cause mortality of
endangered species (Stevens et al., 1995).
The overwhelming importance of subsequent precipitation on Tamarix
establishment in Grand Canyon and Tamarix establishment in every year after 1983
(Figure 3) suggests that flow regime treatments are not a panacea for Tamarix control,
even along highly-regulated rivers. Merritt & Poff (2010) also concluded that restoration
floods would not significantly reduce Tamarix establishment. Along rivers where
Tamarix dominance is a concern, repeated monitoring of native and non-native woody
seedlings should be conducted similar to Johnson (2000). By following the fate of
seedlings on all types of geomorphic units, researchers can understand hydrologic,
geomorphic, and climatic conditions that allow establishment and persistence. In this 28 way land managers can focus localized restoration efforts on sites that foster long-term survival of Tamarix and other problematic species. Likewise, planting of natives can be targeted to suitable sites. Although watershed-scale restoration strategies are ideal, our results suggest that a combination of site-scale and watershed-scale restoration efforts are necessary for Tamarix control.
ACKNOWLEDGEMENTS
We are grateful for the many, enthusiastic volunteers that assisted with field work.
We also thank Cole Crocker-Bedford (retired NPS) for helping us obtain permits. Jeri
Ledbetter and Monte Tillinghast were instrumental in planning and executing river expeditions. Chris Kratt assisted with tree slab processing. We thank Gibby Siemion for sharing her knowledge of the research area and Tamarix autecology. Kelly Burke (Grand
Canyon Wildlands), Justin Salamon, and Kerrie Medeiros (UNR) assisted with river trip logistics. Tom Gushue and Barb Ralston (USGS) provided GIS data layers. Julisa
Edwards and Stephanie Kilburn assisted with tree-ring counting. Jeanne Chambers,
Ashley Sparrow, and the EECB peer review group provided helpful comments on early drafts. Funding was provided by the USDA NRI “Biology of Weedy and Invasive
Plants” program, grant # 2005-35320-16327, and by the National Park Service through the Great Basin Cooperative Ecosystem Studies Unit, Task Agreement # J8R07070014.
29
REFERENCES
Baker WL. 1990. Climatic and hydrologic effects on the regeneration of Populus angustifolia James along the Animas River, Colorado. Journal of Biogeography 17 : 59-75.
Barsoum N. 2002. Relative contributions of sexual and asexual regeneration strategies in Populus nigra and Salix alba during the first years of establishment on a braided gravel bed river. Evolutionary Ecology 15 : 255-279.
Bhattacharjee J, Taylor JP Jr, Smith LM, Haukos DA. 2008. Seedling competition between native cottonwood and exotic saltcedar: implications for restoration. Biological Invasions DOI: 10.1007/s10530-008-9357-4.
Birken AS, Cooper DJ. 2006. Processes of Tamarix invasion and floodplain development along the lower Green River, Utah. Ecological Applications 16 : 1103-1120.
Bornette G, Tabacchi E, Hupp C, Puijalon S, Rostan JC. 2008. A model of plant strategies in fluvial hydrosystems. Freshwater Biology 53 : 1692-1705.
Burnham KP, Anderson DR. 2002. Model selection and multimodel inference: a practical information-theoretic approach . Springer: New York.
Busch DE, Smith SD. 1995. Mechanisms associated with decline of woody species in riparian ecosystems of the southwestern U.S. Ecological Monographs 65 : 347- 370.
Cooper DJ, Merritt DM, Andersen DC, Chimner RA. 1999. Factors controlling the establishment of Fremont cottonwood seedlings on the Upper Green River, USA. Regulated Rivers: Research & Management 15 : 419-440.
Cooper DJ, Andersen DC, Chimner RA. 2003. Multiple pathways for woody plant establishment on floodplains at local to regional scales. Journal of Ecology 91 : 182-196.
DeWine JM, Cooper DJ. 2007. Effects of river regulation on riparian box elder ( Acer negundo ) forests in canyons of the Upper Colorado River Basin, USA. Wetlands 27 : 278–289.
DeWine JM, Cooper DJ. 2009. Habitat overlap and facilitation in tamarisk and box elder stands: implications for tamarisk control using native plants. Restoration Ecology doi: 10.1111/j.1526-100X.2008.00494.x .
30
Edwards P, Kollmann J, Gurnell AM, Petts GE, Tockner K, Ward JV. 1999. A conceptual model of vegetation dynamics on gravel bars of a large Alpine river. Wetlands Ecology and Management 7: 141-153.
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 : 78-91.
Friedman JM, Lee VJ. 2002. Extreme floods, channel change, and riparian forests along ephemeral streams. Ecological Monographs 72 : 409-425.
Gaskin JF, Schaal BA. 2002. Hybrid Tamarix widespread in U.S. invasion and undetected in native Asian range. Proceedings of the National Academy of Sciences of the United States of America 99 :11256-11259.
Glenn EP, Nagler PL. 2005. Comparative ecophysiology of saltcedar ( Tamarix ramosissima ) and native trees. Journal of Arid Environments 61 : 419–446.
Graf WL. 1978. Fluvial adjustments to the spread of tamarisk in the Colorado Plateau region. Geological Society of America Bulletin 89 : 1491-1501.
Guilloy-Froget H, Muller E, Barsoum N, Hughes FMR. 2002. Dispersal, germination, and survival of Populus nigra L. (Salicaceae) in changing hydrologic conditions. Wetlands 22 : 478-488.
Harper JL. Í977. Population Biology of Plants . Academic Press: London.
Howard A, Dolan R. 1981. Geomorphology of the Colorado River in the Grand Canyon. The Journal of Geology 89 : 269-298.
Howe WH, Knopf FL. 1991. On the imminent decline of Rio Grande cottonwoods in central New Mexico. The Southwestern Naturalist 36 : 218-224.
Hupp CR, Bornette G. 2003. Vegetation as a tool in the interpretation of fluvial geomorphic processes and landforms in humid temperate areas. pp 269-288 In Tools in Fluvial Geomorphology , Kondolf GM, Piégay, H, (eds) John Wiley & Sons: West Sussex, UK.
Hupp CR, Osterkamp WR. 1996. Riparian vegetation and fluvial geomorphic processes. Geomorphology 14 : 277-295.
Huston MA. 1994. Biological Diversity: the coexistence of species on changing landscapes. Cambridge University Press: Cambridge, UK.
31
Johnson RR. 1991. Historic changes in vegetation along the Colorado River in the Grand Canyon. pp 178-205 In Colorado River Ecology and Dam Management . Marzolf, GR. (ed) National Academy Press: Washington D.C.
Johnson WC. 2000. Tree recruitment and survival in rivers: influence of hydrological processes. Hydrological Processes 14 : 3051-3074.
Karrenberg S, Edwards PJ, Kollman J. 2002. The life history of Salicaceae living in the active zone of floodplains. Freshwater Biology 47 : 733-748.
Leopold LB. 1969. The rapids and the pools—Grand Canyon. Geological Survey Professional Paper 669-D.
Levine CM, Stromberg JC. 2001. Effects of flooding on native and exotic plant seedlings: implications for restoring south-western riparian forests by manipulating water and sediment flows. Journal of Arid Environments 49 : 111-131.
Mahoney JM, Rood SB. 1998. Streamflow requirements for cottonwood seedling recruitment- an integrative model. Wetlands 18 : 634-645.
Mast JN, Waring G. 1997. Dendrochronological analysis of Goodding willows in Grand Canyon National Park. Proceedings of the Third Biennial Conference of Research on the Colorado Plateau National Parks , The National Park Service, Flagstaff, AZ.
Merkel DL, Hopkins HH. 1957. Life History of Salt Cedar ( Tamarix gallica L.). Transactions of the Kansas Academy of Science 60 : 360-369.
Merritt DM, Poff NL. 2010. Shifting dominance of riparian Populus and Tamarix along gradients of flow alteration in western North American rivers. Ecological Applications , in press.
Mortenson SG, Weisberg PJ, Ralston BE. 2008. Do beavers promote the invasion of non- native Tamarix in the Grand Canyon riparian zone? Wetlands 28 : 666-675.
Mortenson SG, Weisberg PJ. 2009. Plant community response to tamarisk invasion and hydrologic regime in the Cataract Canyon, Canyonlands National Park: a preliminary investigation. Final Report, National Park Service: Moab, UT.
Mortenson SG, Weisberg PJ. 2010. Does river regulation increase dominance of invasive woody species in riparian landscapes? Global Ecology and Biogeography , accepted.
32
Mortenson SG. 2009. Plant community invasibility in riparian landscapes: role of disturbance, geomorphology, and life history traits. University of Nevada Reno, Ph.D. Dissertation, Reno, NV.
Nagelkerke, N. J. D. 1991. A note on a general definition of the coefficient of determination. Biometrika 78 : 691-692.
Nijssen B, O'Donnell G, Hamlet A, Lettenmaier D. 2001. Hydrologic sensitivity of global rivers to climate change. Climatic Change 50 :143-175.
Poff NL, Ward JV. 1989. Implications of streamflow variability and predictability for lotic community structure: a regional analysis of streamflow patterns. Canadian Journal of Fisheries and Aquatic Science 46 : 1805-1818.
Polzin L, 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.
Ramsey FL, Schafer DW. 2001. The Statistical Sleuth. Duxbury/Thomson Learning: Pacific Grove, CA.
Richards K, Brasingon J, Hughes F. 2002. Geomorphic dynamics of floodplains: ecological implications and a potential modelling strategy. Freshwater Biology 47 : 559-579.
Roelle JE, Gladwin DN. 1999. Establishment of woody riparian species from natural seedfall at a former gravel pit. Restoration Ecology 7: 183-192.
Schmidt JC, Graf JB. 1990. Aggradation and degradation of alluvial sand deposits, 1965 to 1986, Colorado River, Grand Canyon National Park, Arizona. U.S. Geological Survey Professional Paper 1493 , Washington DC.
Scott ML, Friedman JM, Auble GT. 1996. Fluvial process and the establishment of bottomland trees. Geomorphology 14 : 327-339.
Scott ML, Auble GT, Friedman JM. 1997. Flood dependency of cottonwood establishment along the Missouri River, Montana, USA. Ecological Applications 7: 677-690.
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.
33
Sher A, Marshall DL. 2003. Seedling competition between native Populus deltoides (Salicaceae) and exotic Tamarix ramosissima (Tamaricaceae) across water regimes and substrate types. American Journal of Botany 90 : 413-422.
Stanford JA, Ward JV, Liss WJ, Frissell CA, Williams RN, Lichatowich JA, Coutant CC. 1996. A general protocol for restoration of regulated rivers. Regulated Rivers: Research and Management 12 : 391-413.
Stella JC, Battles JJ, Orr BK, McBride JR. 2006. Synchrony of seed dispersal, hydrology and local climate in a semi-arid river reach in California. Ecosystems 9: 1200- 1214.
Stevens LE. 1989. Mechanisms of riparian plant community organization and succession in the Grand Canyon, Arizona. Ph.D. Dissertation, Northern Arizona University, Flagstaff, AZ.
Stevens LE, Ayers TJ, Bennet JB, Christensen K, Kearsley MJ, Meretsky VJ, Phillips III AM, Parnell RA, Spence J, Sogge MK, Springer AE, Wegner DL. 2001. Planned flooding and Colorado River riparian trade-offs downstream from Glen Canyon Dam, Arizona. Ecological Applications 11 : 701-710.
Stevens LE, Gold BD. 2003 Monitoring for adaptive management of the Colorado River ecosystem in Glen and Grand Canyons. In Monitoring Ecosystems: Interdisciplinary Approaches for Evaluating Ecoregional Initiatives , Busch DE, Trexler JC. (eds.) pp. 101-134. Island Press: Washington DC.
Stevens LE, Schmidt JC, Ayers TJ, Brown BT. 1995. Flow regulation, geomorphology, and Colorado River marsh development in the Grand Canyon, Arizona. Ecological Applications 5: 1025-1039.
Stevens LE, Waring GL. 1985. The effects of prolonged flooding on the riparian plant community in Grand Canyon. In: USDA General Technical Report RM-120 , pp. 81-86. Tucson, AZ.
Stromberg JC. 1997. Growth and survivorship of Fremont cottonwood, Goodding willow, and salt cedar seedlings after large floods in central Arizona. Great Basin Naturalist 57 : 198–208.
Stromberg JC, Chew MK, Nagler PL, Glenn EP. 2009. Changing perceptions of change: the role of scientists in Tamarix and river management. Restoration Ecology 17 : 177-186.
Stromberg JC, Lite SJ, Marler R, Paradzick C, Shafroth PB, Shorrock D, White JM, White MS. 2007. Altered stream-flow regimes and invasive plant species: the Tamarix case. Global Ecology and Biogeography 16 : 381-393. 34
Taylor J P, Wester DB, Smith LM. 1999. Soil disturbance, flood management, and riparian woody plant establishment in the Rio Grande floodplain. Wetlands 19 : 372–382.
Topping DJ, Rubin DM, Nelson JM, Kinzel III PJ, Bennett JP. 1999. Linkage between grain-size evolution and sediment depletion during Colorado River floods. In The Controlled Flood in Grand Canyon. Geophysical Monograph 110 , Webb RH, Schmidt JC, Marzolf GR, Valdez RA. (eds.) pp. 71-98. American Geophysical Union: Washington DC.
Turner RM. Karpiscak MM. 1980. Recent vegetation changes along the Colorado River between Glen Canyon Dam and Lake Mead, Arizona. Geological Survey Professional Paper 1132. US Dept. of Interior: Washington DC.
Valdez RA, Hoffnagle TL, McIvor CC, McKinney T, Leibfried WC. 2001. Effects of a test flood on fishes of the Colorado River in Grand Canyon, Arizona. Ecological Applications 11 : 686–700.
Warren DK, Turner RM. 1975. Saltcedar ( Tamarix chinensis ) seed production, seedling establishment, and response to inundation. Journal of the Arizona Academy of Science 10 : 135-144.
Webb RH. 1996. Grand Canyon, a Century of Change . University of Arizona Press: Tucson, AZ.
Webb RH, Griffiths PG, Melis TS, Hartley DR. 2000. Sediment delivery by ungaged tributaries of the Colorado River in Grand Canyon, AZ. USGS Water-Resources Investigations Report 00-4055 . Tuscon, AZ.
TABLES
Table 1. Characteristics of geomorphic reaches of Grand Canyon according to Schmidt & Graf (1990) including length, mean channel width, and gradient. Channel width refers to width at 708 m 3s-1 discharge. Mean Tamarix density and standard deviation from floating transects and percent of sites with Tamarix and Salix seedlings are given. The number of sampled transects or sites is noted parenthetically. Salix exigua distribution is limited to upstream reaches (not Canyon and Lower Granite).
Length (km) Channel width (m), gradient Tamarix density Tamarix seedlings % Salix seedlings % Permian 17.7 85, 0.0010 0.37 ± 0.11 (6) 40 (10) 20 (10) Supai 18.5 64, 0.0014 0.34 ± 0.13 (8) 67 (3) 0 (3) Redwall 27.8 67, 0.0015 0.27 ± 0.18 (40) 57 (7) 0 (7) Marble Canyon 36 107, 0.0010 0.58 ± 0.24 (35) 52 (23) 48 (23) Furnace Flats 24.5 119, 0.0021 0.44 ± 0.18 (25) 43 (21) 29 (21) Upper Granite 63.5 58, 0.0023 0.21 ± 0.14 (43) 52 (58) 5 (58) The Aisles 14 70, 0.0017 0.42 ± 0.25 (16) 70 (10) 20 (10) Middle Granite 23.2 64, 0.0020 0.16 ± 0.12 (15) 63 (8) 0 (8) Muav Gorge 32.1 55, 0.0012 0.15 ± 0.09 (32) 52 (27) 37 (27) Canyon 86.7 94, 0.0013 0.44 ± 0.29 (77) 52 (44) (0) Lower Granite 42 73, 0.0016 0.29 ± 0.18 (19) 17 (6) (0)
35
36
Table 2. Explanatory variables used in analyses. * indicates variables that were calculated with Indicators of Hydrologic Alteration software (Smythe Scientific Software, Boulder, CO). USGS gage and climate data obtained from http://waterdata.usgs.gov and http://www.wrcc.dri.edu. Tree-ring analyses: Climate (Phantom Ranch, AZ; station #026471) - Mean summer (June – August) maximum monthly temperature - Total precipitation from May through September - Mean summer (June – August) maximum monthly temperature year following germination - Total precipitation from May through September following year Flow Magnitude - Annual peak flow magnitude Colorado River near Grand Canyon, AZ (USGS gage #09402500) - Annual peak flow magnitude of previous year - Annual peak flow magnitude following year - Minimum flow magnitude of 30-day duration* Flow Timing - Date of maximum flow: number days after initiation of Tamarix seed dispersal (May 1) - Julian date of annual minimum flow* - Magnitude of maximum flow, April through June - Magnitude of maximum flow, July through September Tributaries - Annual peak flow magnitude of Paria River (USGS gage #09382000) - Annual peak flow magnitude of Little Colorado River (USGS gage # 09402000) Rate of Change - Recession rate* : Median of negative daily differences (fall rate) Flow Frequency - Number of high flow pulses* : defined as periods where discharge 25% above median - Number of high flow pulses two years following germination* Flow Duration - Duration of high flow & flood pulses* : number of days in which high flow or flood pulses occur Geomorphic Analyses: - Geomorphic reach : as defined by Schmidt & Graf (1990) See characteristics in Table 1. - Distance downstream from Glen Canyon Dam Channel characteristics - Channel width at 8,000 ft 3s-1 discharge - Channel gradient : slope of thalweg 5.8 km (30 X maximum channel width) upstream - Sinuosity : channel distance / Euclidean distance of channel 1.65 km upstream Tributary influence (tributaries as presented by Webb et al ., 2000) - Density of moderately-sized tributaries (drainage area > 10 km 2) 5 km upstream - Distance from nearest upstream major tributary (drainage area > 30 km 2)
37
Table 3. Relative importance of each explanatory variable or sum of AIC weights ( wi) across models for the annual frequency of Tamarix establishment across sites in all, low (226 - 708 m 3s-1), and high (708 – 1,274 m 3s-1) flow stages. The direction of effect (positive or negative) indicates the sign of a bivariate correlation coefficient. Bold font indicates the two most important variables. See Table 2 for variable descriptions.
All stages Low High Climate precipitation (following May – Sept) 0.736 - 0.774 - 0.487 - max temperature (following year) 0.021 + 0.079 + 0.052 + maximum summer temperature 0.011 - 0.004 - 0.027 + precipitation (May – Sept) 0.009 + 0.002 + 0.023 + Flow Magnitude annual peak flow magnitude 0.212 + 0.002 - 0.206 + 30-day minimum flow 0.024 + 0.006 - 0.070 + annual peak flow (previous year) 0.022 + 0.018 - 0.058 + annual peak flow (following year) 0.010 + 0.019 - 0.046 + Flow Timing maximum flow (July – Sept) 0.466 + 0.002 - 0.132 + maximum flow (April – June) 0.299 + 0.003 - 0.151 + date of maximum flow 0.014 - 0.002 - 0.035 - date of minimum flow 0.011 - 0.002 - 0.163 + Tributaries Little Colorado River annual peak 0.033 + 0.002 + 0.257 + Paria River annual peak 0.017 + 0.438 + 0.028 + Rate of Change recession rate 0.008 - 0.532 - 0.023 + Flow Frequency # high pulses 0.024 - 0.007 - 0.025 - # high pulses (2 following years) 0.009 + 0.036 - 0.023 + Flow Duration Duration high pulses & floods 0.019 + 0.050 - 0.047 +
Table 4. Comparison of most plausible generalized linear models for Tamarix establishment in all, low (226 - 708 m 3s-1), and high (708 – 1,274 m 3s-1) stages. Models with ∆ AICc no greater than 2 of the best model are shown. AIC weights are relative to all possible models. Nagelkerke pseudo-R2 values are also given.
Variables AICc ∆ AICc AIC weight R2 All flow stages precip (following May – Sept) + maximum flow (July – Sept) 96.096 0 0.281 0.56 precip (following May – Sept) + maximum flow (April – June) 96.228 0.132 0.263 0.50 precip (following May – Sept) + annual peak flow magnitude 96.908 0.812 0.187 0.48 Low flow stages precip (following May – Sept) + recession rate 70.165 0 0.517 0.56 precip (following May – Sept) + Paria River peak flow 71.691 1.526 0.241 0.53 High flow stages precip (following May – Sept) + annual peak flow magnitude 68.815 0 0.151 0.34 precip (following May – Sept) + maximum flow (April – June) 69.934 1.119 0.086 0.33
Table 5. Relative importance of each explanatory variable or sum of AIC weights ( wi) across models for adult Tamarix density and Salix seedling presence. Bold font indicates the most important variable. The Nagelkerke pseudo-R2 for the best univariate model is given. No significant model was found for Tamarix seedling presence. See Table 1. for adult Tamarix densities and percent of sites with Tamarix and Salix seedlings in different reaches.
Tamarix density (n= 316 ) Salix seedling (n= 167 ) geomorphic reach 1.00 (R 2 = 0.34) 0.997 (R 2 = 0.28) distance downstream 0.111 - 0.107 - channel width 0.132 + 0.095 + channel gradient 0.135 - 0.105 - sinuosity 0.153 + 0.159 + distance from tributary 0.088 - 0.182 + density of tributaries 0.138 - 0.102 - 38
Table 6. Predicted responses of common riparian shrubs and trees of the southwestern US to flood scenarios. Pluses and minuses indicate positive or negative response of seedling establishment / adult persistence. The responses are based on life history characteristics summarized by Mortenson (2009). The duration of all flood scenarios is 10 days excluding the fall flood which continues for 30 days. Wetland indicator status is in bold.
Fall flood, Early spring flood Summer flood long duration Winter flood No floods Obligate wetland Salix exigua + /+ + / + - / + - / + - / + Salix gooddingii + / + - / + - / + - / + - / + Facultative wetland Populus fremontii + / + - / + - / + - / + - / + Baccharis emoryi - / + + / + - / - - / + - / - Baccharis salicifolia - / + + / + - / - - / + - / - Pluchea sericea - / + + / - - / - - / + - / + Tamarix ramosissima - / + + / + - / - - / + - / + Acer negundo - / + + / - - / - - / + + / + Celtis laevigata - / + - / + - / - - / - + / + Elaeagnus angustifolia - / + + / - - / - + / + + / + Facultative Baccharis sarothroides - / - - / - - / - - / - + / + Prosopis glandulosa - / - - / - - / - - / - + / + 39
40
FIGURE CAPTIONS
Figure 1. Annual hydrographs of the Colorado River near Grand Canyon, AZ (USGS gage #09402500) and Green River at Green River, UT (gage #09315000). Note changes in y-axis scale. These hydrographs represent flow regimes characteristics of: a) pre-dam (prior to 1963), b) post-dam floods (1983 – 1986), c) high-fluctuating flows (1965 – 1982; 1987 – 1991), d) low-fluctuating flows (1992 – present excluding restoration flood years), e) restoration floods (1996, 2004, 2008), and f) steady flows (2000).
Figure 2 . Map of study area. Sites sampled for the Tamarix tree-ring study and geomorphic reaches as defined by Schmidt & Graf (1990) are shown.
Figure 3. Number of approximately-aged and accurately-aged Tamarix established by year in Grand Canyon. The peak annual flow recorded by USGS gage #09402500 is also given.
Figure 4. The number of sites that experienced Tamarix establishment in each year from 1984 to 2006 according to a) July – September peak flows and b) annual peak flows. Note differences in x-axis scale. Restoration floods were conducted in 1996 and 2004.
Figure 5. Elevations of accurately-aged Tamarix samples above most recent high water line according to year of establishment. Geomorphic unit is indicated by symbol shape and fill. Water levels were comparable during all sampling trips. (Note: n=131 because one field notebook with elevations was lost.)
Figure 6. Comparison of dendrochronology results of Tamarix establishment along the Colorado River in Grand Canyon and Green River below Flaming Gorge Dam (modified from Birken & Cooper, 2006).
41
FIGURES
Figure 1.
42
Figure 2.
43
Figure 3.
44
Figure 4.
Figure 5.
45
Figure 6.
46
CHAPTER 2. DO BEAVERS PROMOTE INVASION OF NON-NATIVE
TAMARIX IN THE GRAND CANYON RIPARIAN ZONE?
Published September 2008 in Wetlands
ABSTRACT
Beavers ( Castor canadensis Kuhl) can influence the competitive dynamics of plant species through selective foraging, collection of materials for dam creation, and alteration of hydrologic conditions. In the Grand Canyon National Park, the native Salix gooddingii C.R.Ball (Goodding’s willow) and Salix exigua Nutt. (coyote willow) are a staple food of beavers. Because Salix competes with the invasive Tamarix ramosissima
Ledeb., land mangers are concerned that beavers may cause an increase in Tamarix
through selective foraging of Salix . A spatial analysis was conducted to assess whether
the presence of beavers correlates with the relative abundance of Salix and Tamarix .
These methods were designed to detect a system-wide effect of selective beaver foraging in this large study area (367 linear km of riparian habitat). Beavers, Salix , and Tamarix co-occurred at the broadest scales because they occupied similar riparian habitat, particularly geomorphic reaches of low and moderate resistivity. Once the affinity of
Salix for particular reach types was accounted for, the presence of Salix was independent of beaver distribution. However, there was a weak positive association between beaver presence and Salix cover. Salix was limited to geomorphic settings with greater sinuosity and distinct terraces, while Tamarix occurred in sinuous and straighter sections of river channel (cliffs, channel margins) where it dominated the woody species composition.
47
After accounting for covariates representing river geomorphology, the proportion of riparian surfaces covered by Tamarix was significantly greater for sites where beavers were present. This indicates that either Tamarix and beavers co-occur in similar habitats, beavers prefer habitats that have high Tamarix cover, or beavers contribute to Tamarix dominance through selective use of its native woody competitors. The hypothesis that beaver herbivory contributes to Tamarix dominance should be considered further through more mechanistic studies of beaver foraging processes and long-term plant community response.
INTRODUCTION
Selective foraging and utilization of woody materials by North American beavers
(Castor canadensis ) can exert profound effects on the species composition and structure of plant communities (Rosell et al. 2005) and may shift community structure toward non- preferred species (Johnston and Naiman 1990). For instance, selective felling of Populus deltoides Bartram ex. Marsh (cottonwood) by beavers has facilitated the growth of
Tamarix (tamarisk) and Elaeagnus angustifolia L. (Russian olive), two non-native invasive shrubs, along several rivers in eastern Montana (Lesica and Miles 2004). Along the Marias River in north-central Montana, beaver browsing was three times more likely on P. deltoides than on E. angustifolia , and the damage to Populus trees was more severe, with beaver damage to the main stem on Populus as opposed to the basal branches on E. angustifolia (Lesica and Miles 1999). Pearce and Smith (2001) also found that Populus was more susceptible to beaver harvesting than E. angustifolia , which led them to predict that E. angustifolia will become more abundant at the expense of Populus along the Milk
48
River, Montana. Beavers have also greatly reduced the abundance of Populus along the
Green and Yampa Rivers of the upper Colorado River Basin, and sapling predation by beavers threatens continued recruitment of Populus (Breck et al. 2003a). Beavers are capable of altering the competitive relationships between preferred and non-preferred forage species.
Bank-dwelling beavers colonize dens on the banks of large-order rivers and influence vegetation by forage selection for preferred species. Salix (S. exigua and Salix gooddingii ) is a staple food for beavers (Johnson 1991, Baker and Hill 2003), leading to the concern that abundant beaver populations in the Grand Canyon National Park
(GCNP) may lead to an increase in invasive Tamarix over native riparian plant species.
Beaver utilization is apparent in the majority of Salix stands and threatens the persistence of Salix gooddingii (Goodding’s willow) stands. The potential effect of beavers on the native S. gooddingii is of more imminent concern because few stands of S. gooddingii remain. Beaver foraging contributes to the decline of S. gooddingii populations along the main channel of the Colorado River in the Grand Canyon (Johnson 1991).
Plants respond differently to herbivory depending on a variety of factors including timing, nutrient availability, and intensity of grazing (Maschinski and Whitham 1989).
Plant competition can cause a more negative reaction (i.e., lower fitness) to biomass removal (Harper 1977). Foraging of Salix by beavers may yield areas for future establishment of Tamarix , particularly when foraging occurs in late summer when annual growth has subsided and compensatory growth is not possible (Kindschy 1989).
Alternatively, when foraging occurs in early spring, beaver damage may increase the cover of Salix , as has been observed for Populus fremontii S. Wats. (McGinley and
49
Whitham 1985). In Utah, P. fremontii changes growth forms from a tree to a shrub due to beaver herbivory; repeated branch removal by beavers causes production of many branches below the original Populus branch (McGinley and Whitham 1985). Similar to
Populus , this type of growth pattern could increase canopy cover and, possibly, increase the competitive ability of Salix if browsing occurs early in the growing season.
We investigated the potential for selective foraging by bank-dwelling beavers to correlate with coarse-scale spatial patterns of riparian vegetation along the Colorado
River in the Grand Canyon National Park. Our analyses were designed to assess: 1) correlations in the distribution of beavers, Salix , and Tamarix , 2) the spatial association of species occurrence after accounting for the distribution in the form of large-scale geomorphic variables, and 3) the relatedness of beaver presence to canopy cover of
Tamarix and Salix . Because beavers and Salix co-occur in similar habitats in GCNP
(Ruffner 1983), a negative association between beavers and Salix after accounting for habitat variability may suggest that beavers have had a significant effect on riparian plant community composition. A positive association may not necessarily indicate a causal relationship, but only that beavers and Salix exist in similar habitats. In contrast, because beavers rarely feed on Tamarix (Lesica and Miles 2004, unpublished data), a positive association would imply that beavers prefer habitats high in Tamarix or suggest a potential effect of beavers for facilitating Tamarix at the expense of native plant species.
It is unknown whether beaver herbivory along the lower Colorado River has contributed to the invasion of the non-native Tamarix . This question is important for managers, as evidenced by a recent proposal by the Hualapai Tribe to conserve native riparian vegetation through a 50% reduction of the beaver population in the Lower Grand
50
Canyon (E. Leslie, National Park Service, pers. comm., 2005). There are also important ecological implications of a system-wide effect of beaver foraging on shrub species composition. If beavers strongly influence the species diversity of riparian ecosystems, the current bottom-up view of riparian systems with hydrology and geomorphology as the key structuring processes (Tickner et al. 2001) should be reconfigured to better incorporate biotic interactions as recommended by Naiman and Rogers (1997).
Study Area
The study area encompasses the riparian habitat along the Colorado River in the
Grand Canyon National Park from Lees Ferry to 383.8 km downriver (Figure 1). From
Lees Ferry to Diamond Creek (river km 387) the Colorado River drops 542 meters, and the majority of elevation loss occurs in short rapids created by debris flows at tributary mouths (Schmidt and Graf 1990). Eleven geomorphic reaches from Lees Ferry to
Diamond Creek have been proposed by Schmidt and Graf (1990). These reaches differ fundamentally in their bedrock composition which causes variation in channel width, slope, and other landform characteristics. In 1963 the Glen Canyon Dam was completed and, thereafter, has drastically reduced the annual streamflow discharge, increased daily water-level fluctuations, and altered the season of high flows of the Colorado River. Flow stabilization has created more riparian habitat, and many shrub species have colonized this newly-available habitat including Salix exigua (coyote willow), Pluchea sericea
Nutt., Tamarix ramosissima , and Baccharis spp. Beavers are also thought to have become more common because of this increase in available habitat (Turner and Karpiscak 1980).
Tamarix , an Eurasian native, was distributed as isolated individuals prior to the Glen
51
Canyon Dam, but spread prolifically with native riparian vegetation after 1963. No major flood has occurred in Grand Canyon since 1983. Currently, Salix exigua and Tamarix are the two of the most dominant riparian shrubs in the Grand Canyon (Turner and Karpiscak
1980).
METHODS
GIS Data Layers
Beaver occurrence data at the spatial resolution of 0.1 river mile units along the longitudinal axis of the river were compared with Tamarix and Salix occurrence data, derived from a remote sensing classification and aggregated to the same spatial resolution. There were two primary spatial data layers used in this study, a point coverage of beaver observations obtained from the Science Center at Grand Canyon National Park and a vegetation map developed by the Grand Canyon Monitoring and Research Center
(GCMRC). From these layers and others, several GIS variables were derived and used for analysis (Table 1). All variables were sampled at the level of 0.1 Stevens river mile (RM) units, the common distance system used in GCNP (Stevens 1990). This measurement system refers to Lees Ferry as RM 0, and downstream locations are referred to by their distance downstream from RM 0 along the river centerline. The 2274, 0.1 RM units between Lees Ferry (RM 0) and RM 238.5, excluding the section from 85.8 RM to 96.0
RM, formed the set of observations used in this analysis; these comprised the extent of river length common to beavers and the vegetation data sets.
Continuous sampling for beaver activity was conducted by National Park Service employees from rafts. Indications of beaver presence and location (i.e., RM) were noted.
52
The beaver database included 929 beaver observations from 1999 – 2003, distributed among years as follows: 2 in 1999, 494 in 2000, 123 in 2001, 142 in 2002, and 168 in
2003. The vast majority (97%) of these observations was taken from April – June. This data set encompassed the river corridor from Lees Ferry to Lake Mead, and included the following variables: observer, date, river mile (to the nearest one-tenth), bank (left or right), whether the observation was active or inactive, density, and comments which sometimes indicated the type of observation (i.e., slide, burrow, lodge, den, tracks, holes, cuttings, animals). In order to achieve consistency in the analysis, we simplified the beaver data set by considering only signs of beaver presence or absence at each 0.1 river mile (Table 1; Figure 2a). Because beaver sampling was ad hoc and inconsistent between sample periods, levels of sampling effort differed among years. Therefore, this data set cannot be used to assess patch occupancy dynamics of beaver populations.
The vegetation classification was derived from airborne digital imagery collected between May 25 and June 3, 2002, when dam release was approximately 8,000 cfs. The resulting four-band image mosaics (blue, green, red, and near-infrared) were aggregated to a resolution of 44 cm. The vegetation base map was constructed through a combination of ground surveys, image processing, and automated supervised classification procedures to identify vegetation classes (Ralston, unpublished data). The accuracy of the classified database was estimated by sampling approximately 10% of the river corridor and comparing areas at least 100 m² in size. The overall accuracy of the vegetation map was
80%. We used the single-species level vegetation classification, which differentiated
Salix and Tamarix from other species. Salix exigua and Salix gooddingii were combined into a common class for Salix spp. following a river reconnaissance trip which suggested
53 that the classification did not accurately distinguish the two species. The fuzzy accuracy assessment (Gopal and Woodcock 1994) revealed 92% accuracy in omission and 50% accuracy in commission for the Salix class values, while Tamarix had values of 80% accuracy in omission and 78% accuracy in commission.
The vegetation classification was used to determine the aerial cover of each species, and each observation describes the vegetation of a circular plot of 120-m radius, comprising 4.52 ha, centered on each 0.1-RM point. Circles of different radii were compared, and the 120-m distance was chosen because it minimized overlap with neighboring circular buffers, while capturing all vegetation within 60-m from water, the maximum foraging distance reported for beavers (Rosell et al. 2005). The proportion of riparian surfaces covered by Tamarix and Salix were used as response variables.
Therefore, the area covered by water at a flow of 8,000 cfs and the area covered by desert vegetation were not considered in the analysis. The variables derived from the vegetation classification and used in this analysis are described in Table 1.
Three geomorphic variables were developed for use as covariates in the statistical analyses: geomorphic reach, rock resistivity, and sinuosity index (Table 1). Eleven geomorphic reaches were adopted from Schmidt and Graf (1990) and include the
Permian Gorge, the Supai Gorge, the Redwall Gorge, the Marble Canyon, Furnace Flats,
Upper Granite Gorge, The Aisles, Middle Granite Gorge, Muav Gorge, Lower Canyon
Reach, and Lower Granite Gorge (Figure 1). In addition, we considered three levels of rock resistivity (low, moderate, high) as interpreted from major geologic formations at river level between Lees Ferry and Phantom Ranch (Ruffner 1983). High resistivity reaches tend to have more constrained river channels, less riparian vegetation, and rocky
54 banks because the bedrock is not easily eroded. A river sinuosity index was calculated as the ratio of channel length to Euclidean distance for each circular plot, where distances were calculated from 0.5 RM above to 0.5 RM below each focal 0.1 RM point. All spatial variables were extracted from the 120-m buffer polygons in ArcGIS and exported to S-
Plus software for statistical analysis.
Statistical Analysis
The standard chi-square test, ANOVA analysis, and Bonferroni multiple comparisons tests were used to test whether occurrence or mean cover of the three taxa varied according to geomorphic river reach or rock resistivity class. Contingency table analyses were used to test for association among beaver occurrence, Salix occurrence
(Figure 2b), and the two reach type variables (geomorphic reach and rock resistivity).
Tamarix occurrence (presence-absence) was not analyzed further because Tamarix was nearly ubiquitous along the river corridor. The spatial association between beavers and
Salix was assessed using contingency table analysis, for the entire river corridor as well as individually within rock resistivity classes. A chi-square test was used for bivariate contingency tables, and the Mantel-Haenszel chi-square test (Legendre and Legendre
1998) was used to test for bivariate associations (e.g., Salix and beaver occurrence) after
accounting for a third categorical covariate (e.g., rock resistivity class).
Nested multiple linear regression models were used to test whether beaver
occurrence significantly predicted the proportion of riparian surface area covered by Salix and Tamarix (Figures 2c, d), after accounting for the environmental covariates of geomorphic reach, river sinuosity, riparian surface area, and proportion of riparian shrub
55 cover (Table 1). This analysis allowed us to determine the relative importance of beaver presence to Salix and Tamarix distribution and abundance after removing the correlation attributed to their shared environments. Salix and Tamarix cover values were first transformed using the natural logarithm to achieve more normal distributions.
Multicollinearity among predictor variables was low. Rock resistivity was not included in regression models because this variable was not available for the entire river corridor.
Akaike’s Information Criterion (AIC) was used to rank the effectiveness of alternative covariate models for fitting the data, given the number of parameters included. Models were compared on the basis of the ∆AIC, or the difference in AIC between the model
with smallest AIC value and the current model, and Akaike weights (w i), where the weights of all models considered are constrained to sum to unity. Thus, for AIC differing by at least 2, the most parsimonious model was that with the lower AIC value (Burnham and Anderson 1998). A drop-in-deviance test (Ramsey and Schafer 2002) was then used to test whether addition of the “beaver” variable to the more parsimonious covariate model significantly improved model fit.
A second-order spatial point pattern analysis technique (Ripley’s K) was applied to the distributions of Salix and beaver observations (Getis and Boots 1978). Each
presence of either species at a 0.1 RM plot (with the 120-m radius applied for Salix ) was
extracted as a point observation, and then analyzed by plotting Ripley’s K functions with
a 95% confidence envelope indicating the k-function for completely spatially random
(CSR) data. In this manner, distributions can be described as clustered, random, or
uniform over a range of spatial scales. The Ripley’s K analysis allowed us to further
investigate the spatial patterns of beaver and willow occurrence.
56
RESULTS
Abundance and Distribution of Tamarix and Salix on the Colorado River
Tamarix occurred at 2,265 of 2,274 (99.6%) plots, and Salix occupied 1,796 of
2,274 (79%) of plots. These frequency rates were high because occupancy was assessed at a coarse spatial scale. Greater differences between the two species were apparent for the proportion of riparian surface area covered (Figure 3). Tamarix dominated an average of 7.8% (95% CI: 7.4–8.1%) of riparian area (including rock and unvegetated surfaces) across all plots, while the cover of Salix was an order of magnitude lower, with a mean of only 0.52% (95% CI: 0.44%–0.60%).
Salix occurrence was not independent of geomorphic river reach ( χ2 = 183.16, df
= 11, p < 0.001), with Salix species occurring more commonly in some reaches than in others. Salix presence/absence varied according to resistivity of the rock formation at
river level ( χ2 = 71.90, df = 2, p < 0.001), with a higher proportion of Salix occurrences
than expected in reaches of low and moderate resistivity. Similarly, mean Salix cover
differed across geomorphic river reach types (F 11, 2262 = 54.16, p < 0.001) and rock resistivity classes (F 2, 855 = 97.50, p < 0.001). Salix cover was greatest in reaches of moderate resistivity (mean = 1.42%), intermediate in reaches of low resistivity (mean =
1.07%), and lowest in reaches of high resistivity (mean = 0.29%). Tamarix cover also
varied according to the eleven geomorphic reaches considered (F 11, 2262 = 49.70, p <
0.001) (Figure 3), as well as by rock resistivity class (F 2, 855 = 9.02, p < 0.001). Tamarix
cover was greatest in reaches of moderate resistivity (mean = 12.84%), but did not differ
57 significantly between reaches of low and high resistivity (mean = 9.84% and 8.98%, respectively).
Distribution of Beavers along the Colorado River
Between 1999 and 2003, observations of beavers or their sign were reported for
444 of 2,274 (19.4%) plots. Of the 444 observations, 278 plots (63%) had recorded beaver observations for only a single year, 117 (26%) had beavers recorded for two years,
36 (8%) for three years, 11 (2.5%) for four years, and only two (0.5%) for all five years.
Beaver presence/absence was not independent of geomorphic river reach ( χ2 = 228.38, df
= 11, p < 0.001) (Figure 3). Beaver occurrence also varied according to rock formation resistivity ( χ2 = 61.72, df = 2, p < 0.001), with a higher proportion of beaver occurrences
than expected in reaches of low and moderate resistivity.
Associations among Beaver, Tamarix , and Salix Occurrence and Abundance
Considering all 2,274 plots in the study, beaver and Salix occurrence were not
independent ( χ2 = 27.18, df = 1, p < 0.001). Salix was present in 88% of plots where
beavers occurred, even though beavers occurred only in approximately 20% of plots. The
association between Salix and beaver occurrence was significant even when the effect of geomorphic river reach was accounted for (Mantel-Haenszel χ2 = 7.186, df = 1, p <
0.001). However, there was no significant association between Salix and beaver occurrence after accounting for the effect of rock resistivity (Mantel-Haenszel χ2 =
0.1244, df = 1, p = 0.724). Individual chi-square analyses conducted for each of the three
58 rock resistivity classes also found that Salix and beaver occurrences were independently distributed within each rock resistivity unit.
The most parsimonious covariate model for Salix cover included the following
variables: GeoReach (effect varied depending upon reach), VegArea (negative effect),
and Sinuosity (positive effect) (Table 2). It appears from the model comparisons that, of
all the variables, GeoReach has the greatest explanatory power for predicting Salix cover
(Table 2). The presence/absence of beavers slightly improved this model (drop-in-
deviance = 2.74, df = 1, p = 0.13). After holding the three covariates constant, Salix cover
increased by 1.1% on sites where beavers were present (95% CI = 1.05–1.14). The only
parsimonious covariate model for Tamarix (∆AIC ≤ 10) included the following variables:
GeoReach (effect varied depending upon reach), VegArea (negative effect), and
Sinuosity (negative effect). The presence/absence of beavers significantly improved this model (drop-in-deviance = 12.34, df = 1, p < 0.001). On plots where beavers were present, after holding all covariates constant, Tamarix cover was on average 1.34 % greater (95% CI = 1.24–1.46). This is an increase of 18% of the mean tamarisk cover value (7.8%)
The dispersion pattern of beaver and Salix occurrences was aggregated across all spatial scales (Figure 4). The Ripley’s k-functions rapidly rise above the confidence envelope for complete spatial randomness, indicating clustered spatial patterns. The beaver k-function shows more intense aggregation over shorter distances, as is expected for the species which is rarer and less diffuse in its distribution.
59
DISCUSSION
Spatial Relationships among Beavers, Tamarix , and Salix
Beavers, Salix , and Tamarix tend to co-occur in similar riparian habitat at coarse spatial scales. As also observed by Ruffner (1983), beavers and Salix occur more frequently in reaches of low and moderate resistivity. Such reaches are more likely to have greater channel width, greater sinuosity (more meandering channel), and more extensive terraces for riparian vegetation to establish than high-resistivity reaches
(Howard and Dolan 1981). Tamarix is the dominant woody species along the Grand
Canyon section of the Colorado River. Although its presence at coarse spatial scales is nearly ubiquitous, there is considerable variation in cover of Tamarix at finer spatial scales. Relative Tamarix cover is greatest on straighter sections of river channel (cliffs and narrow channel margins) where other riparian shrubs cannot establish. Salix and many other riparian shrub species ( Baccharis , Pluchea ) are more likely to dominate where channels are meandering and geomorphic structures are complex, allowing for formation of marsh habitats and multiple terraces with a mixture of coarse and fine substrates. Tamarix absolute cover is also high in such settings. However, this species is dominant on rocky cliffs in narrow gorges likely because of its high drought tolerance relative to native riparian shrubs (Smith et al. 1998).
After the variation associated with geomorphic reach was accounted for, the positive association of beavers and Salix was still significant, suggesting that both taxa selected for similar habitat features at spatial scales finer than the geomorphic reach.
However, the ecological amplitude for beavers along the river corridor appears to be more limited than that for Salix , resulting in a more clustered distribution (Figure 4).
60
Once the occurrence of the two species in low resistivity rock types was accounted for,
Salix presence was independent of beaver distribution. However, beaver presence was slightly positively associated with Salix cover. Beavers and Salix may occupy similar habitats at spatial scales finer than 4.52 ha, or beavers may prefer habitats with greater
Salix cover. These results are consistent with the idea that Salix is a staple food for beavers. This association among the two species does not support the hypothesis that beavers negatively influence Salix cover along the Grand Canyon river corridor.
The cover of Tamarix was also greatest in 4.52-ha plots where beavers were present, after accounting for geomorphic covariates. This effect was highly significant and greater than the effect of beavers on Salix cover. This indicates either that Tamarix
and beavers occupy similar habitats even at spatial scales finer than 4.52 ha, that beavers
prefer habitats with high Tamarix cover, or that beavers promote Tamarix dominance
through selective herbivory and use of competing riparian woody species. Beavers have
likely occupied the same sites over the past few decades, given that most optimal beaver
sites have been occupied continuously since approximately fifteen years following
construction of the Glen Canyon Dam (Ruffner 1983). Riparian plant communities can
shift rapidly under pressure from intense, selective herbivory because plant community
composition depends largely upon initial successional response to frequent, episodic
disturbance. For instance, cattle grazing of Populus and Salix in unregulated riparian
habitat indirectly increases Tamarix abundance (Stromberg et al. 2007). The hypothesis
that beaver herbivory contributes to Tamarix dominance is consistent with our results.
Beaver occurrence may facilitate Tamarix dominance at the scale of 4.52-ha plots.
61
Recommendations for Further Research and Management
The effects of beaver activities on riparian vegetation composition should be investigated further using more mechanistic and detailed studies. More specifically, we recommend exploration of the relationship between beavers and Tamarix at fine scales. It would be advisable to initiate exclosure studies, where beaver-proof fences are constructed around patches of riparian vegetation in areas occupied by beavers as in
Andersen and Cooper (2000). Beaver observations from exclosure studies do not provide a reliable indicator of population size, which requires more robust, intrusive methods involving trapping and mark-recapture statistics, or radio telemetry as in Breck et al.
(2001). Accurate population estimates and spatial models of beaver distribution are essential for understanding effects of beaver herbivory. Finally, it is critical to consider beaver effects with respect to life history traits of riparian woody species and the interaction of these with managed flow regimes. Flow regulation may increase availability of favored species to beavers (Breck et al. 2003b), as well as dramatically alter riparian community composition apart from beaver effects (Stevens et al. 1995,
Stromberg et al. 2007).
It is important to differentiate between bank-dwelling beaver and dam-building beaver strategies when considering the effects that beavers have on wetland communities.
Dam-building beavers are considered ecosystem engineers because they increase landscape heterogeneity through creation of impoundments. Beaver ponds increase beta diversity of herbaceous species in the Adirondacks (Wright et al. 2002) and may shift competitive dynamics in favor of native shrubs. Beaver dams may also inhibit Tamarix
62 growth due to the relative intolerance of Tamarix for inundation when compared with native Salix and Populus (Albert and Trimble 2000). The use of Tamarix for beaver dam construction in northwestern Colorado may have indirectly enhanced Salix exigua distribution and abundance (Baker and Hill 2003).
Dam-building beavers are often introduced to benefit native plants and wildlife in ecological restoration projects (Albert and Trimble 2000), but beaver populations are also controlled (e.g. removed to another habitat, fatally trapped) to reduce their negative influence on rare, native shrubs and trees (Longcore et al. 2007). These treatments obviously contradict each other. Perhaps much of the confusion concerning the effects of beavers on wetland health arises due to the different activities of dam-building and bank- dwelling beavers. Beaver dams may benefit restoration projects that aim for greater diversity, while selective foraging of beavers may favor non-native plant species. The potential for divergent ecological effects of beavers that construct ponds and bank- dwelling beavers should be taken into account in planning for wetland management.
ACKNOWLEDGMENTS
Bert Frost, Elaine Leslie, and Lori Makarick, all of the U.S. National Park
Service, were instrumental in initiating, supporting, and funding this research. Thomas
Gushue (GCMRC) made available numerous other, essential GIS layers. Rekha Pillai of
UNR assisted with technical GIS support. Larry Stevens (Grand Canyon Wildlands
Council) contributed to overall understanding through ongoing discussions of Grand
Canyon vegetation dynamics. We thank Steve Jenkins, Larry Stevens, Chris Lowry,
63
Mark Miller, and three anonymous reviewers for helpful comments on the manuscript.
Funding was provided by the USDA NRI “Biology of Weedy and Invasive Plants” program, grant # 2005-35320-16327, and by the National Park Service through the Great
Basin Cooperative Ecosystem Studies Unit (CESU), Task Agreement # JBR07040010.
REFERENCES
Albert, S. and T. Trimble. 2000. Beavers are partners in riparian restoration on the Zuni Indian Reservation. Ecological Restoration 18:87–92.
Andersen, D. C. and D. J. Cooper. 2000. Plant-herbivore-hydroperiod interactions: Effects of native mammals on floodplain tree recruitment. Ecological Applications 10:1384–1399.
Baker, B. W. and E. P. Hill. 2003. Beaver ( Castor canadensis ). p. 288–310 In G.A. Feldhamer, B. C. Thompson, and J. A. Chapman. (eds.). Wild Mammals of North America: Biology, Management, and Conservation , second edition. The John Hopkins University Press, Baltimore, MD, USA.
Breck, S. W., K. R. Wilson, and D. C. Andersen. 2001. The demographic response of bank-dwelling beavers to flow regulation: a comparison on the Green and Yampa Rivers. Canadian Journal of Zoology 79:1957–1964.
Breck, S. W., K. R. Wilson, and D. C. Andersen. 2003a. Beaver herbivory and its effect on Populus trees: influence of flooding along matched regulated and unregulated rivers. River Research and Applications 19:43–58.
Breck, S. W., K. R. Wilson, and D. C. Andersen. 2003b. Beaver herbivory of Salix under two flow regimes: a comparative study on the Green and Yampa rivers. Western North American Naturalist 63:463–471.
Burnham, K. P. and D. R. Anderson. 1998. Model Selection and Inference: a Practical Information-Theoretic Approach . Springer-Verlag, New York, NY, USA.
Getis, A. and B. Boots. 1978. Models of Spatial Processes: An Approach to the Study of Point, Line, and Area Patterns . Cambridge University Press, Cambridge, MA, USA.
Gopal, S. and C. Woodcock. 1994. Theory and methods of accuracy assessment using fuzzy sets. Photogrammetric Engineering and Remote Sensing 60:181–188.
64
Harper, J. L. 1977. Population Biology of Plants . Academic Press, London, UK.
Howard, A. and R. Dolan. 1981. Geomorphology of the Colorado River in the Grand Canyon. Journal of Geology 89:269–298.
Johnson, R. R. 1991. Historic changes in vegetation along the Colorado River in the Grand Canyon. p. 178–206. In G. R. Marzolf (ed.) Colorado River Ecology and Dam Management . National Academy Press, Washington DC, USA.
Johnston, C. A. and R. J. Naiman. 1990. Browse selection by beaver: effects on riparian forest composition. Canadian Journal of Forest Research 20:1036–1043.
Kindschy, R. R. 1989. Regrowth of willow following simulated beaver cutting . Wildlife Society Bulletin 17:290–294.
Legendre, P. and L. Legendre. 1998. Numerical Ecology, second edition. Elsevier, Amsterdam, The Netherlands.
Lesica, P. and S. Miles. 1999. Russian olive invasion into cottonwood forests along a regulated river in north-central Montana. Canadian Journal of Botany 77:1077– 1083.
Lesica, P. and S. Miles. 2004. Beavers indirectly enhance the growth of Russian olive and Tamarix along eastern Montana rivers. Western North American Naturalist 64:93–100.
Longcore, T., C. Rich, and D. Müller-Schwarze. 2007. Management by assertion: beavers and songbirds at Lake Skinner (Riverside County, California). Environmental Management 39:460–471.
Maschinski, J. and T. G. Whitham. 1989. The continuum of plant responses to herbivory: the influence of plant association, nutrient availability, and timing. The American Naturalist 134:1–19.
McGinley, M. A. and T. G. Whitham. 1985. Central place foraging by beavers ( Castor canadensis ): a test of foraging predictions and the impact of selective feeding on the growth form of cottonwoods ( Populus fremontii ). Oecologia 66:558–562.
Naiman, R. J. and K. H. Rogers. 1997. Large animals and system-level characteristics in river corridors . BioScience 47:521–529.
Pearce, C. M. and D. G. Smith. 2001. Saltcedar: distribution, abundance, and dispersal mechanisms, Northern Montana, USA. Wetlands 23:215–228.
65
Ramsey, F. L. and D. W. Schafer. 2002. The Statistical Sleuth: A Course in Methods of Data Analysis , second edition. Duxbury, Pacific Grove, CA, USA.
Rosell, F., O. Bozser, P. Collen and H. Parker. 2005. Ecological impact of beavers Castor fiber and Castor canadensis and their ability to modify ecosystems. Mammal Review 35:248–276.
Ruffner, G. A. 1983. Abundance and distribution of beaver and coyote willow in Grand Canyon, Arizona. Unpublished report to the Grand Canyon National Park, National Park Service, Grand Canyon, AZ, USA.
Schmidt, J. C. and J. B. Graf. 1990. Aggradation and degradation of alluvial-sand deposits, 1965 to 1986, Colorado River, Grand Canyon National Park, Arizona. Professional Paper 1493.,U.S. Geological Survey, Washington DC, USA.
Smith, S. D., D. A. Devitt, A. Sala, J. R. Cleverly, and D. E. Busch. 1998. Water relations of riparian plants from water desert regions. Wetlands 18:687–696.
Stevens, L. E. 1990. The Colorado River in Grand Canyon , third edition. Red Lake Books, Flagstaff, AZ, USA.
Stevens, L. E., J. C. Schmidt, T. J. Ayers, and B. T. Brown. 1995. Flow regulation, geomorphology, and Colorado River marsh development in the Grand Canyon, Arizona. Ecological Applications 5:1025–1039.
Stromberg, J. C., S. J. Lite, R. Marler, C. Paradzick, P. B. Shafroth, D. Shorrock, J. M. White, and M. S. White. 2007. Altered stream-flow regimes and invasive plant species: the Tamarix case. Global Ecology and Biogeography 16:381–393.
Tickner, D. P., P. G. Angold, A. M. Gurnell, and J. O. Moutford. 2001. Riparian plant invasions: hydrogeomorphological control and ecological impacts. Processes in Physical Geography 25:22–52.
Turner, R. M. and M. M. Karpiscak. 1980. Recent vegetation changes along the Colorado River between Glen Canyon Dam and Lake Mead, Arizona. Professional Paper 1132, U.S. Geological Survey, Washington DC, USA.
Wright, J. P., C. G. Jones, and A. S. Flecker. 2002. An ecosystem engineer, the beaver, increases species richness at the landscape scale. Ecosystems Ecology 132:96– 101.
66
TABLES
Table 1. Spatial variables used in the nested multiple linear regression models. All vegetation variables are derived from 2002 aerial photography according to the Grand
Canyon Monitoring and Research Center vegetation classification. “RM” = Stevens river mile.
Ecological Variable Abbreviation Description Beaver Occurrence BEAV Presence/Absence at each 0.1 RM, pooled for 1999 – 2003 Salix Occurrence paSALIX Presence/Absence of Salix spp. in a 120-m radius circle centered on each 0.1 RM point, as classified from 2002 aerial photography. Tamarix Occurrence paTARA Presence/Absence of Tamarix spp. in a 120-m radius circle centered on each 0.1 RM point, as classified from 2002 aerial photography. Salix Cover cvSALIX Proportion of riparian surfaces in 2002 covered by Salix Tamarix Cover cvTARA Proportion of riparian surfaces in 2002 covered by Tamarix Riparian Surface Area VegArea Area within the 4.52 ha plot with riparian surfaces where vegetation is or could be established (m2)
Geomorphic Variable Abbreviation Description Geomorphic Reach GeoReach The 11 bedrock- and landform-defined geomorphic reaches of the Grand Canyon derived from Schmidt and Graf (1990) Rock Resistivity Resistivity Classifications of rock resistivity at river level, interpreted from major geologic formations between Lees Ferry and Phantom Ranch. Three levels: Low, Moderate, High. Sinuosity Sinuosity River sinuosity index, calculated as the quotient of channel length and straight-line distance, for 1.0 RM channel lengths centered on each 0.1 RM point.
67
Table 2. A priori candidate models explaining variation in Salix cover according to various combinations of geomorphic and vegetation variables (abbreviations defined in
Table 1). K indicates the number of parameters plus one. Only candidate models with
∆AIC ≤ 10 are shown.
VARIABLES K AIC ∆AIC AIC WEIGHT
GeoReach VegArea Sinuosity 15 1908.4 0.0 0.86
GeoReach VegArea 14 1912.2 3.8 0.14
68
FIGURE CAPTIONS
Figure 1. The Colorado River through the Grand Canyon. The twelve geomorphic reaches identified by Schmidt and Graf (1990) are shown. The study area encompasses
Lees Ferry to river km 408 but excludes the section from 162 km to 178 km.
Figure 2. Occurrence of A) beaver and B) Salix and relative cover of C) Tamarix and D)
Salix for a representative section of the Colorado River from Stevens River Mile 194 –
201. For (a) and (b) a circle indicates presence of beaver or Salix , respectively, in the
120-m radius plots centered on each 0.1 River Mile. For C) and D) larger sizes reflect greater values for the proportion of riparian surfaces covered by Tamarix and Salix .
Figure 3 . The proportion of riparian surfaces covered by Tamarix and Salix and the number of beaver occurrences divided by the length of each geomorphic reach in river miles. One occurrence indicates that there was some sign of beaver activity during one of the years (1999-2003) of beaver surveys. The reach abbreviations are: (A) Permian
Gorge, (B) Supai Gorge, (C) Redwall Gorge, (D) Marble Canyon, (E) Furnace Flats, (F)
Upper Granite Gorge, (G) The Aisles, (H) Middle Granite Gorge, (I) Muav Gorge, (J)
Lower Canyon, and (K) Lower Granite Gorge.
Figure 4 . Ripley’s K-function for beaver (thick solid line) and Salix (thick dashed line) across lag distances from 100 – 500 meters. The x-axis starts at a lag distance of 100 m because the closest observations are approximately that far apart. The thin dashed lines at the bottom of the graph show the confidence envelope expected for a spatially random distribution. The lines are consistently above the confidence envelope, indicating a clumped distribution.
69
FIGURES
Figure 1.
70
Figure 2.
71
Figure 3.
72
Figure 4.
73
CHAPTER 3. DOES RIVER REGULATION INCREASE DOMINANCE OF
INVASIVE WOODY SPECIES IN RIPARIAN LANDSCAPES?
Accepted for publication in Global Ecology and Biogeography (2010)
ABSTRACT
A regional analysis was used to explore the influence of river regulation on dominance of non-native, invasive shrubs and trees. We addressed the following questions. 1) How do large dams affect hydrologic parameters that influence riparian vegetation? 2) How do flow regimes affect dominance of non-native woody species? 3)
How do changes in flow regimes affect dominance of non-native woody species? We sampled the canopy cover of woody species on 179 point bars along seven non-dammed and thirteen dammed river segments. Wilcoxon rank sum tests were used to determine differences between flow parameters in dammed and non-dammed rivers. We used correlation analyses and generalized linear model comparisons to examine associations of flow parameters and canopy cover of native (Populus and Salix ) and non-native (Tamarix and Elaeagnus ) taxa. An index of flow alteration that was created using principal
components analysis was regressed with vegetation cover.
Tamarix cover was positively related to drainage area, flow constancy, August and May median flow, and flow recession rate, but Elaeagnus cover was unrelated to
flow variables. River segments with peak flows in late summer or high constancy had the
highest Tamarix cover. Populus cover was positively influenced by low maximum
temperatures and frequent high pulses. Flow alteration was negatively related to Populus cover and positively related to Tamarix cover. Total non-native, Elaeagnus , and Salix
74 covers were not correlated with flow alteration. Rivers with large drainage area and low flow variability are inherently more vulnerable to invasions. River regulation does not necessarily increase cover of non-native, invasive species. Instead, changes in flow allow proliferation of species that have life history traits suited to modified flow regimes. River restoration projects that aim to reinstate natural flow regimes should be designed with knowledge of native and non-native species’ life history strategies.
INTRODUCTION
The increasing human population has spurred numerous efforts to harness flowing water for domestic use, irrigation, hydropower, aquaculture, navigation, and flood control worldwide, but the negative effects of river regulation on riparian communities threaten their sustainability and capacity to provide ecosystem services (Nilsson & Berggren,
2000). Concurrently, invasions of non-native plants in riparian communities have altered the structure and function of riparian communities and threaten biodiversity (Richardson et al., 2007). However, the relationships between river regulation and plant invasions are ambiguous. Elton (1958) recognized the ability of “water control” to spur biological invasions into natural communities, and human-caused flow alteration is presently thought to foster invasion of non-native species in aquatic habitats (Bunn & Arthington,
2002; Nilsson & Svedmark, 2002). Spread of invasive plants and animals in riparian landscapes is often coincident with flow alteration (Nilsson & Berggren, 2000;
Richardson et al., 2007). There are also examples of invasions without flow alteration
(see D’Antonio et al., 1999). Understanding the consequences of river regulation is necessary to design suitable management strategies and is made more important by
75 numerous efforts to restore natural flow regimes (see Poff et al., 1997). The hypothesis that regulated rivers are more susceptible to non-native plant invasions needs to be explored systematically across multiple river systems and broad extents.
We considered particular hydrologic components associated with woody plant invasion across a broad region using consistent sampling methods. We addressed the following questions. 1) How do large dams affect hydrologic parameters that influence riparian vegetation in the southwestern US? This preliminary question allowed us to identify dam-induced flow regime changes and establish sensitivity of selected indices to extreme flow regulation. The effects of dams on flow regimes are well-researched yet variable across regions (Walker & Thoms, 1993; Magilligan & Nislow, 2005; Graf, 2006;
Poff et al., 2007). We hypothesized that rivers in the southwestern US would experience similar changes in flow regime as other arid and semi-arid regions. Dams commonly reduce flow variability and decrease the magnitude of peak flows (Walker et al., 1995;
Graf, 2006). Large dams in the US generally increase the magnitude of minimum flows and alter timing of annual maximum and minimum flows downstream from impoundments (Graf, 2006).
2) How do flow regimes affect dominance of non-native woody species? We focused on interactions between specific hydrologic characteristics and plant community structure. The dominance of common pioneer species (e.g., Salix , Populus , Baccharis , and Tamarix ) is closely tied to flow regime characteristics, particularly flood magnitude, timing, duration, and frequency (Karrenberg et al., 2002; Beauchamp & Stromberg,
2007). Floods create bare areas for germination and establishment. The germination site of seedlings must remain moist and protect the seedling from subsequent burial, scour,
76 and prolonged inundation (Scott et al., 1997). Rapid recession of flows can cause mortality of Populus seedlings if root growth does not equal rates of groundwater recession (Mahoney & Rood, 1998). Riparian shrubs and trees that do not require floods for establishment benefit from occasional high flows that raise groundwater levels. We hypothesized that native and non-native species would respond to hydrologic characteristics based on their life history strategies.
3) How do changes in flow regimes affect dominance of non-native woody species? The primary objective of this study was to compare the prevalence of non- native, invasive woody species on rivers that span a continuum of river regulation to determine if and to what extent non-native cover is influenced by altered flow regimes.
The proposed analysis differs from previous studies by quantifying river regulation instead of merely comparing regulated with unregulated rivers; most rivers in the world have experienced some degree of regulation (Nilsson & Berggren, 2000).
Dam construction and other forms of river regulation alter hydrologic and geomorphic processes and, therefore, influence riparian vegetation (Friedman et al.,
1998; Jansson et al., 2000; Merritt & Cooper, 2000). Although floods of lower magnitude are less likely to cause mortality of riparian vegetation, cessation of floods and associated sediment reworking limits bare areas available for pioneer tree and shrub establishment (Scott et al., 1996). In addition, a change in timing of flood events may benefit species that disperse seeds during the modified flood season, have a longer period of seed dispersal or seed longevity, or reproduce vegetatively (Jansson et al., 2000).
Riparian species composition is driven by the combination of flow and geomorphic alteration that influences dominance of species with different life history traits
77
(Richardson et al., 2007). Therefore, we hypothesized that native and non-native species would respond to changes in hydrologic characteristics based on their life history strategies.
Along regulated rivers in the US, the recruitment of native pioneer trees such as
Populus spp. (cottonwood) and Salix gooddingii (Goodding willow) is limited because of
lack of spring floods (Fenner et al., 1985). Salix and Populus release their short-lived
seeds during the approximate time of historic floods (Shafroth et al., 1998). Changes in
the timing of high and low flows disadvantage Populus and Salix because their seed production, dispersal, and longevity are synchronized with hydrologic events (Fenner et al., 1985; Shafroth et al., 1995; Karrenberg et al., 2002). However, populations of native shrubs that reproduce clonally such as Baccharis spp., Pluchea sericea , and Salix exigua
can increase with regulation because reduced disturbance frequency and intensity favor
clonal reproduction (Jansson et al., 2000).
Tamarix spp. and Elaeagnus angustifolia (Russian olive) are the dominant, non-
native woody invaders in riparian habitats in the southwestern US. Tamarix is well-
suited to natural flow regimes of the southwestern US and benefits from a longer period
of seed release than native shrubs (Harris, 1966). Tamarix also has greater drought
tolerance (Glenn & Nagler, 2005) and, therefore, may colonize areas that are unsuitable
for native phreatophytes. Tamarix requires disturbance to establish although seedlings
also are susceptible to frequent disturbances (D’Antonio et al., 1999). In contrast,
Elaeagnus does not require disturbance for establishment and is relatively shade-tolerant.
The reduced magnitude and frequency of floods caused by dams may favor the spread of
78
Elaeagnus but this has not yet been documented (Lesica & Miles, 1999; Katz et al.,
2005).
Several studies have shown Tamarix to be more prominent along regulated vs. unregulated rivers (Merritt & Cooper, 2000; Shafroth et al., 2002; Stromberg et al.,
2007). However, other studies have shown non-native woody species to be equally or more abundant along unregulated rivers (Katz et al., 2005; Beauchamp & Stromberg,
2007) indicating that altered flow regimes do not consistently increase non-native shrub cover. In this study, we hypothesized that relationships among the relative abundance of
Tamarix and Elaeagnus and flow regimes are consistent with species-specific life history strategies. We expected to find positive relationships between the degree of flow alteration and Tamarix cover relative to native riparian vegetation because common native trees and shrubs such as Salix and Populus have difficulty establishing in altered flow regimes (Scott et al., 1997). This assumes that native shrubs compete directly with
Tamarix for resources such as bare areas for establishment. In contrast, Elaeagnus cover was not expected to differ with increasing flow alteration because Elaeagnus does not rely on bare areas for establishment (Shafroth et al., 1995; Katz et al., 2001).
Study Area
The study area includes rivers in the Colorado Plateau and the Basin and Range physiographic provinces of the US, encompassing 1,237,576 km 2 (Fig. 1). In this region floods originate in early spring (May – June) following snowmelt and in late summer
(August – September) during the North American Monsoon. Floods associated with monsoons are more common in the southern part of these provinces, while snowmelt
79 floods dominate hydrographs in the northern part (Webb et al., 2007). The southwestern
US is an ideal study area because the riparian habitat contrasts strongly with adjacent semi-arid habitat, and many rivers have a long period of gage records.
The construction of dams and diversions has disrupted the river continuum such that sections within a river become functionally independent with regard to flows of energy, organisms, and matter (Ward & Stanford, 1983). Therefore, the river segment, or section of a river between a gage station and features that alter river flow (e.g., dams, tributaries, diversions, intensive agriculture or development) is the study unit. We conducted field surveys of twenty river segments located on the Carson, Colorado,
Green, Gunnison, Humboldt, Rio Grande, Salt, San Juan, San Pedro, Truckee, Verde,
Virgin, Walker, and White Rivers (Fig. 1; Table 1). These river segments have extensive flow records, relatively unconstrained channels, and perennial flow (excluding the
Carson). River segments were chosen to represent a full continuum of regulation, from the free-flowing San Pedro river segment to the highly regulated Colorado River in
Grand Canyon (Colorado 3). Individual segments varied in length from eight to twenty- eight km. Regional surveys such as this, particularly studies that rely on a long period of gage records, have inherently small sample sizes (see Jansson et al., 2000; Stromberg et al., 2007), but are necessary to assess regional trends in ecosystem conditions and coordinate control of widespread, noxious invaders (Mack, 2005).
80
METHODS
Dams and Flow Regimes
Statistics that describe parameters relevant to hydrologic change were derived from US Geological Survey (USGS) mean daily streamflow records and calculated with
Indicators of Hydrologic Alteration (IHA) software (Smythe Scientific Software,
Boulder, CO). We identified a subset of these parameters that change consistently with dam construction (Magilligan & Nislow, 2005; Graf, 2006), the major agent of river alteration in this region (Table 2), and influence riparian vegetation composition and structure. We used parameters that represent all components of flow regime (magnitude, frequency, duration, timing, and rate of change of river flow [Poff et al. 1997]) and interact with plant life histories to influence the likelihood of germination, establishment, and survival. We considered multiple indicators of magnitude and timing to address the effects of different temporal scales.
The predictability of timing and magnitude of flows structures riparian
communities (Poff & Ward, 1989). Two IHA parameters that represent elements of
predictability deserve further explanation. Constancy measures variation in daily flow
magnitudes, and contingency measures variation in timing of similar magnitude flows
among years. Poff and Ward (1989) outline the calculation of constancy and contingency
for river flow which are based on equations by Colwell (1974). Because predictability is
the sum of constancy and contingency (Colwell, 1974) we multiplied the two parameters
provided by the IHA software, “predictability” and “constancy / predictability” to obtain
constancy, and constancy was subtracted from “predictability” to calculate contingency.
81
The IHA parameters were calculated using non-parametric statistics according to water year (Oct 1 - Sept 30). We assessed changes in IHA parameters using records from at least 20 years before and after dam construction. For unregulated rivers, an equal period “before” and “after” 1962, the mean year of regional alteration, was analyzed.
The dams considered in this analysis are impoundments on the main channel documented by the National Inventory of Dams (USACE, 1992). Segments that did not have flow data 20 years prior to dam construction were excluded from the flow alteration analysis
(Colorado 1, Green 1, Rio Grande 3, Salt, Truckee 1, Truckee 2). The White and
Colorado 2 River segments were included in the non-dammed category because their flow regimes are considered unaltered despite small, upstream dams (Webb et al., 2007).
To assess the effect of dams on hydrologic parameters, changes in IHA parameters from dammed and non-dammed rivers were compared using separate Wilcoxon rank sum tests.
Hydrology and Woody Vegetation
To test whether flow regime parameters and river regulation have influenced non- native shrub abundance in the southwestern US, we mapped woody vegetation on four to fourteen (mean = 9) point bars or islands within each of the twenty river segments for a total of 179 point bars or islands. Point bars (areas of sediment deposition that collect on the inside of meander bends) and islands were chosen as the sampling unit because they are young geomorphic surfaces and occur within all river segments. We expected woody vegetation age classes found on these surfaces to fall within the period of hydrologic record. Point bars and islands within river segments were identified prior to field sampling. All selected point bars were less than two km from a road unless the sites were
82 accessed by boat. Vegetation was not sampled within 25 m of development, more than
100 m from the low flow channel, or in areas where restoration activities were evident.
Sites varied in sampled area (0.11 ha – 10.21 ha) but did not extend more than 500 m
parallel to the axis of the river.
Vegetation patches were delineated in the field using 1-m resolution, National
Agricultural Imagery Program (NAIP) truecolor orthophotography. The percent canopy
cover of riparian shrub and tree seedlings, saplings, and adults was ocularly estimated
within all patches of each point bar or island by the same observer (SGM). Cover was
used as the measure of abundance because the negative effects of invasive woody species
are proportional to their relative cover (Lundholm & Larson, 2004). Populus age cohorts
were mapped in decadal categories to document time of establishment based on a
subsample of trees that were cored to allow comparison between recent and past Populus establishment. It is difficult to estimate the age of other common taxa due to lack of a straight, upright growth form.
The analyses included three Tamarix species (T. chinensis , T. parviflora , and T. ramosissima ) and their hybrids, Elaeagnus angustifolia , two Populus species ( P.
deltoides subsp. wislizenii and P. fremontii ), and Salix exigua . Taxonomic groups
occurring at fewer than 10 (50%) river segments were not included in statistical analyses
(Table 3). The area of each digitized patch was calculated using ArcGIS software, and
the percent cover of each species was multiplied by the area of the associated patch to
obtain an area-weighted cover estimate. The area-weighted cover of total non-native
species, the most common non-native ( Tamarix and Elaeagnus ), and native ( Populus , young Populus cohorts estimated to be less than 30-years old, and Salix exigua ) species
83 was divided by the total area of woody vegetation for all point bars within a river segment to determine the relative area-weighted percent cover of each taxon.
We analyzed the relationship between species dominance and IHA parameters
(Table 2). We also considered the effects of drainage area which influences hydrologic indicators, land use patterns, and climate. River reaches with large drainage areas tend to occur at low elevations and have higher temperatures, more anthropogenic development, and greater rates of discharge. Although climate is intricately tied to hydrology, in regulated systems these influences are often decoupled. Therefore, we also considered the influence of minimum January temperature, maximum July temperature, mean annual precipitation, and mean annual snowfall.
We calculated IHA metrics from the earliest recorded water year (Table 1) to
2007 for all river segments. The influence of each IHA parameter and climate variable was analyzed with correlation analyses to explore its relationship to total non-native,
Tamarix , Elaeagnus, Populus , young Populus , and S. exigua . Transformations were
performed if necessary. The geographic distribution of Elaeagnus is limited by temperature (Friedman et al., 2005). Therefore, absences associated with Elaeagnus cover in southern Arizona were excluded from analyses.
We used Akaike’s Information Criterion (AIC) to address the influence of univariate, additive, and interaction effects of environmental variables on vegetation cover. Generalized linear models for each taxon were compared on the basis of AIC C, an
AIC adjusted for small sample sizes (Burnham & Anderson, 2002). The lowest AIC C value indicates the most plausible model given the data, and differences in AIC C ( ∆i) of >
2 units reveal models with considerably less support. Akaike weights ( wi) allow
84 comparison within a set of models based on each model’s relative likelihood (Burnham &
Anderson, 2002). These analyses allowed us to investigate the combined effects of drainage area, climatic variables, and flow regime on the relative cover of common taxa.
The relationships between changes in flow regime and vegetation cover were assessed with an index of flow alteration. Linear combinations of the thirteen flow indicators (Table 2) were created using principal components analysis (PCA). PCA avoided problems introduced by high multicollinearity of hydrologic parameters by producing uncorrelated axes representing the variation of the original indicators (McCune
& Grace, 2002). The Euclidean distance between “before” (before 1962 or prior to dam construction) and “after” time periods for each river segment was calculated from the first five PCA axes that explained more than 90% of the variance. This created an index of flow alteration that was regressed against the cover of the taxon of interest. In this way we investigated associations between vegetation cover and flow alterations due to cumulative effects of many forms of river regulation.
RESULTS
Effects of dams on flow regime
Changes in many of the IHA parameters before and after dam construction or before and after 1962 differed significantly for dammed and non-dammed river segments
(Table 5). The number of high pulses, constancy, August flow, minimum flow, base flow, date of maximum flow, and recession rate increased compared to the pre-regulation period in dammed river segments, but did not significantly change in non-dammed river segments. Contingency and maximum flow decreased at a greater magnitude in dammed
85 segments. Change in high pulse duration, mean annual flow, timing of minimum flow, and May median flow were not significantly different among dammed and non-dammed rivers.
Vegetation patterns across river segments
Non-native cover on point bars varied among river segments from <1 % to 89 %
(Fig. 2). Riparian areas in the Great Basin had lower cover of non-native taxa than in other physiographic provinces. Tamarix was present along all segments and composed a large majority of non-native plant cover except along the San Juan, White, and Rio
Grande 2 where Elaeagnus was dominant (Table 3). River segments with higher maximum temperatures had greater Tamarix cover (Table 4). In contrast, river segments with lower minimum temperatures had greater Elaeagnus cover (r = -0.49, p = 0.03), which justified exclusion of southern Arizona river segments in our hydrologic analyses.
The highest Tamarix cover was sampled along the Virgin River, while Elaeagnus cover was greatest along the San Juan River segment.
The distribution of native species generally followed climate zones. Native
Populus spp. were present along every river segment excluding the Colorado 3 and
Humboldt. Salix spp. were ubiquitous. Salix exigua was present along all segments excluding the San Pedro, Verde, Salt, and Rio Grande 3 which were the most southerly latitudes sampled. Salix exigua cover was greater along river segments with higher mean
annual snowfall (Table 4). Salix gooddingii was common in deserts along the Rio
Grande 2, Rio Grande 1, Virgin, Colorado 3, Verde, San Pedro, and Salt River segments.
Populus estimated to be less than 30 years-old comprised greater than fifty percent of
86
Populus sampled along the San Pedro, Verde, Carson, and Colorado 2 segments (Fig. 2).
Baccharis spp. (B. emoryi , B. salicifolia , and B. sarothroides ) were present along thirteen segments in arid steppe and desert environments. The Colorado 3 segment had high cover of Baccharis emoryi and Pluchea sericea . Acer negundo (box elder) was abundant in colder environments along the Colorado 1 and Green 1, and Shepherdia argentea was common along the Walker River segment in the Great Basin.
Effects of hydrology on woody vegetation
Pearson correlation coefficients revealed positive associations between total non- native cover and constancy, drainage area, August flow, mean annual flow, May flow, and recession rate (Table 4). These relationships were influenced by combined effects of flow regime on Tamarix and Elaeagnus . River segments with large drainage area and
high constancy and August flow had greater Tamarix cover. Elaeagnus cover was not significantly correlated with the hydrologic variables analyzed, although river segments with short duration, high pulses and low minimum temperatures generally had higher
Elaeagnus cover. Populus cover was negatively correlated with drainage area and
August flow. The influence of environmental parameters on young Populus (Table 4) was largely driven by two river segments with high cover of young Populus cohorts
(Carson & San Pedro). Salix exigua cover was negatively related to constancy and high pulse count and positively related to high pulse duration (Table 4).
Model comparisons with AIC c revealed the influence of combinations of explanatory variables, some of which did not manifest their effects in the bivariate correlation analyses (Table 6). Multiple plausible models existed for total non-native
87 cover, and the interaction and additive effects of constancy and drainage area were included in the top two models. The interaction of constancy and date of maximum flow provided the best single model for Tamarix cover with an AIC weight of 0.71, and date of maximum flow was included in the top three models. River segments that had high constancy or moderate constancy but a late peak flow (July or August) had higher
Tamarix cover. The interaction and additive effects of minimum January temperature and duration of high pulses comprised the most plausible models for Elaeagnus . The
best model for Populus cover included maximum July temperature and the number of
high pulses, although neither of these explanatory variables was significant in the
univariate analysis. River segments with high maximum temperatures had low Populus
cover, unless they experienced frequent high pulses over the growing season.
Interactions between constancy or snow and the duration and frequency of high pulses
were important in the models for Salix exigua cover (Table 6).
The first two PCA axes of hydrologic variables explained 45 and 21% of the
variation in flow parameters, respectively. Axis one described high flow magnitudes
(high mean annual, May, August, and seven-day maximum and minimum flows),
recession rate, and constancy (Fig. 3). Axis two expressed high contingency, high pulse
duration, and few high pulses. The Colorado 3 River segment, which is controlled by
Glen Canyon Dam, experienced the greatest overall change in flow along these two PCA-
derived dimensions, while the San Pedro experienced the least flow alteration (Fig. 3).
The first five PCA axes explained 92% of the variation in flow regime parameters, and
the Euclidean distance of the river segments in the “pre” and “post” periods was
calculated for these axes. The ordering of river segments according to flow regime
88 alteration did not closely correspond to the presence of a major dam upstream; the non- dammed Colorado 2 and Humboldt had high alteration in flow regime (Figs. 3 & 4).
Total non-native and Tamarix covers were not significantly related to flow regime
alteration (Figs. 4a, b). However, if high cover of Tamarix along the Virgin River
segment was excluded from the analysis, Tamarix cover was significantly positively
related to alteration in flow regime (Fig. 4b). Exclusion of the Virgin River segment did
not affect the relationship of total non-native cover and flow alteration (R 2 = 0.09, p =
0.3). Elaeagnus cover was unrelated to flow alteration (Fig. 4c). Populus cover was negatively related to flow regime alteration, and the negative relationship between young
Populus and flow alteration was marginally significant (p = 0.06) (Figs. 4d, e). A non- linear effect of flow regulation on young Populus cover, where abundant Populus establishment has recently occurred only for the two river segments with the least flow regime alteration, was suggested by this relationship (Fig. 4d). Salix exigua cover was unrelated to flow alteration (Fig. 4f).
DISCUSSION
We found differences in flow alteration of dammed and non-dammed rivers similar to those described by Magilligan and Nislow (2005) who documented a significant increase in seven-day minimum flow, August flows, and high pulse count and a decrease in seven-day maximum flows along 21 dammed rivers across the US (Table
5). This established sensitivity of these hydrologic parameters to dam construction.
The non-linear relationship between young Populus and flow alteration (Fig. 4d) suggests that beyond a certain degree of alteration, establishment of large cohorts of
89
Populus is less likely. In a similar study Merritt and Poff (2010) also observed
diminished Populus recruitment along rivers with minimal levels of flow modification.
This and the negative relationship between total Populus cover and flow alteration (Fig.
4c) justifies concerns of Populus decline caused by river regulation (Fenner et al., 1985;
Rood & Mahoney, 1990) but could also be a product of episodic recruitment patterns of
Populus (Scott et al., 1997). If conditions needed for seedling establishment are
consistently prevented by flow alteration, Salix exigua populations are also expected to
decline (Douhovnikoff et al., 2005). In the short term, clonal growth allows S. exigua to
cope with high flow variability (low constancy) and flow alteration. Salix exigua is
highly tolerant of inundation and increases root and shoot growth in response to rises in
the water table (Amlin & Rood, 2001). These characteristics allow S. exigua to flourish
despite long periods of high pulses.
The response of Populus to flow regime and climate characteristics is driven by
water availability (Table 6). Populus cover was lower along river segments with high
maximum temperature unless frequent flood pulses prevented desiccation of seedlings.
Populus cover is also greater on perennial vs. intermittent river reaches in Arizona
(Stromberg et al., 2007). However, Populus cover was lower and Tamarix cover was
greater along segments with high August flows (Table 4). Floods late in the growing
season can cause mortality of Populus seedlings and also overlap with the prolonged seed
dispersal period of Tamarix (Warren & Turner, 1975). Additionally, Merkel & Hopkins
(1957) observed higher viability of Tamarix seeds during August (51%) when compared
with June (19%). These factors likely contribute to high Tamarix cover along rivers with
moderate constancy but late summer peak flows (Table 6).
90
River segments with high constancy have greater total non-native and Tamarix cover on point bars (Table 4). This relationship is not always a function of river regulation; in this study, river segments with high constancy (White, Virgin) had low levels of river regulation. The highest Tamarix cover sampled was along the Virgin
River segment which is not regulated by a large dam. According to Christensen (1962) one of the earliest documents of Tamarix naturalization (1925) was based on observations along the Virgin River. The early naturalization of Tamarix may have led to greater dominance of this species along the Virgin River segment. Webb et al. (2007) attribute
Tamarix dominance along this portion of the Virgin River to an artificial flood caused by a tributary dam failure in December, 1989. We observed a significant trend in Tamarix dominance with increasing flow alteration after removing the Virgin River segment from the analysis (Fig. 4b). Interestingly, Tamarix dominance was highest at moderate levels of flow alteration. This trend was also documented by Merritt and Poff (2010).
Total non-native and Tamarix cover values were higher along river segments with larger drainage areas (Table 4). Increases in non-native plants in downstream river reaches have been attributed to higher temperatures and human use in France and
Australia (Tabacchi et al., 1998). However, Elaeagnus distribution and cover were positively related to low January temperatures, indicating differential climatic responses of these two non-natives. River segments with high minimum temperature but short duration of high pulses also had high Elaeagnus cover on point bars (Table 6) which suggests superior drought tolerance or intolerance to inundation. More studies are needed to determine ecological tolerances of this species (Katz & Shafroth, 2003). Katz et al. (2005) documented the establishment of many recent cohorts of Elaeagnus on
91 dammed and non-dammed rivers in Colorado. Our results also indicate that Elaeagnus can have high cover regardless of the degree of river regulation.
River regulation is not directly associated with total non-native cover (Fig. 4a).
This is caused by the differential influence (or lack of influence) of hydrologic factors on native and non-native species with different life history traits. Tamarix increased and
Populus decreased with flow alteration. These species depend upon scouring floods to
create suitable germination surfaces. Elaeagnus is less sensitive than Tamarix to specific
aspects of flow regime (Katz & Shafroth, 2003). Its seeds are primarily animal-dispersed
and capable of germinating under riparian forest canopies (Shafroth et al., 1995). Salix
exigua can reproduce clonally. Therefore, Elaeagnus and S. exigua are not as dependent
as Populus and Tamarix on seasonal flooding to disperse seeds and create new
germination surfaces.
Dispersal characteristics and the conditions needed for establishment and
persistence determine the effect of disturbance regimes on native and non-native species.
Vegetation response to river regulation depends upon the interactions of species-specific
life history strategies with climate, flow regime characteristics, and flow alteration, along
with stage of invasion. The common dichotomy of native vs. non-native has limited
ecological significance in regards to the effects of flow or climatic regimes on vegetation.
Future studies should focus on identifying species-specific responses of focal species to
certain aspects of the flow regime.
92
Management Implications
River regulation does not always result in an increase in non-native, invasive species. The response of invasive species to flow alteration is more complicated than previous studies (Bunn & Arthington, 2002; Nilsson & Svedmark, 2002) have suggested.
Intrinsic watershed characteristics such as large drainage areas and high hydrologic constancy (i.e., low flow variability) make some river segments more susceptible to invasion. The natural flow regime paradigm encourages the management of rivers based on the hydrologic conditions that existed prior to human alteration of the river system
(Poff et al., 1997). For example, the return of high spring flows with a low recession rate on the Truckee River in Nevada encouraged establishment of Populus (Rood et al.,
2003). Tamarix is not yet abundant along this stretch of the river but would have likely increased under these flow conditions if seeds were available. Dominance of non-native, invasive species complicates process-based restoration strategies because invasive species may be well-suited to certain natural flow regimes. Once an invasive species has established along a river, returning to natural flow conditions alone will not ensure the success of native plant species.
The outcomes of implementing natural flows may be unpredictable without considering the relationships of the life history strategies of native and non-native species to specific aspects of the flow regime. For example, along the Colorado River in the
Grand Canyon, high steady flows were released from Glen Canyon Dam from June through August, 2000 to benefit native fish populations. Unfortunately, these releases also spurred establishment of a large cohort of Tamarix (Stevens & Gold, 2003).
Restoration prescriptions must be designed from an ecosystem perspective that
93 encompasses diverse taxa. Closer investigation of the ecological amplitude of potential non-native and native species will enable managers to focus on species that will increase or decrease due to certain forms of river regulation.
ACKNOWLEDGEMENTS
We are grateful to Julia Mortenson and Christopher Kratt for logistical and field assistance. Jeanne Chambers, Ashley Sparrow, members of the Great Basin Landscape
Ecology Lab, and the EECB peer review group provided valuable insights throughout manuscript preparation. John Peterson, Ondrea Hummel (US Army Corps of Engineers), and Elizabeth Milford (Natural Heritage New Mexico) provided imagery and vegetation data. We appreciate the constructive comments of Julian Olden and four anonymous reviewers. Funding was provided by the USDA NRI “Biology of Weedy and Invasive
Plants” program, grant # 2005-35320-16327, and by the National Park Service through the Great Basin Cooperative Ecosystem Studies Unit, Task Agreement # J8R07070014.
Thanks also to The Nature Conservancy for access to the McCarran Ranch Preserve.
94
REFERENCES
Amlin, N.A. & Rood, S.B. (2001) Inundation tolerances of riparian willows and cottonwoods. Journal of the American Water Resources Association, 37 , 1709- 1720.
Beauchamp, V.B. & Stromberg, J.C. (2007) Flow regulation of the Verde River, Arizona encourages Tamarix recruitment but has minimal effect on Populus and Salix stand density. Wetlands , 27 , 381-389.
Bunn, S.E. & Arthington, A.H. (2002) Basic principles and ecological consequences of altered flow regimes for aquatic biodiversity. Environmental Management , 30, 492-507.
Burnham, K.P. & Anderson, D.R. (1998) Model Selection and Multimodel Inference , Second Edition. Springer, New York, NY.
Christensen, E.M. (1962) The rate of naturalization of Tamarix in Utah. The American Midland Naturalist , 68, 51-57.
Colwell, R.K. (1974) Predictability, constancy, and contingency of periodic phenomena. Ecology , 55 : 1148-1153.
D'Antonio, C.M., Dudley, T.L. & Mack, M. (1999) Disturbance and biological invasions: direct effects and feedbacks. Ecosystems of Disturbed Ground . (ed. by L.R. Walker) pp. 413-452. Elsevier, New York, NY.
Douhovnikoff, V., McBride, J.R. & Dodd, R.S. (2005) Salix exigua clonal growth and population dynamics in relation to disturbance regime variation. Ecology , 86 , 446-452.
Elton, C.S. (1958) The Ecology of Invasions by Animals and Plants . The University of Chicago Press, Chicago, IL.
Fenner, P., Brady, W.W. & Patton, D.R. (1985) Effects of regulated water flows on regeneration of Fremont cottonwood. Journal of Range Management , 38 , 135- 138.
Friedman, J.H., Osterkamp, W.R., Scott, M.L. & Auble, G.T. (1998) Downstream effects of dams on channel geometry and bottomland vegetation: regional patterns in the Great Plains. Wetlands , 18 , 619-633.
Friedman, J.H., Auble, G.T., Shafroth, P.B., Scott, M.L., Merigliano, M.F., Freehling, M.D. & Griffin, E.R. (2005) Dominance of non-native riparian trees in western USA. Biological Invasions , 7, 747-751.
95
Glenn, E.P. & Nagler, P.L. (2005). Comparative ecophysiology of Tamarix ramosissima and native trees in western U.S. riparian zones. Journal of Arid Environments , 61 : 419-446.
Graf, W.L. (2006). Downstream hydrologic and geomorphic effects of large dams on American rivers. Geomorphology , 79 , 336-360.
Harris, D.R. (1966) Recent plant invasions in the arid and semi-arid southwest of the United States. Annals of the Association of American Geographers , 56 , 408-422.
Jansson, R., Nilsson, C., Dynesius, M. & Andersson, E. (2000) Effects of river regulation on river-margin vegetation: a comparison of eight boreal rivers. Ecological Applications , 10 , 203-224.
Karrenberg, S., Edwards, P.J. & Kollman, J. (2002) The life history of Salicaceae living in the active zone of floodplains. Freshwater Biology , 47 , 733-748.
Katz, G.L., Friedman, J.M., & Beatty, S.W. (2001) Effects of physical disturbance and granivory on establishment of native and alien riparian trees in Colorado, U.S.A. Diversity and Distributions 7, 1-14.
Katz, G.L. & Shafroth, P.B. (2003). Biology, ecology, and management of Elaeagnus angustifolia . Wetlands , 23 , 763-777.
Katz, G.L., Friedman, J.M. & Beatty, S.W. (2005) Delayed effects of flood control on a flood-dependent riparian forest. Ecological Applications , 15 , 1019-1035.
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.
Lundholm, J.T. & Larson, D.W. (2004) Dominance as an overlooked measure of invader success. Biological Invasions , 6, 505-510.
Mack, R.N. (2005) Assessing biotic invasion in time and space: the second imperative. Invasive Alien Species: A New Synthesis (ed. by H.A. Mooney, R.N. Mack, J.A. McNeely, L.E. Neville, P.J. Schei, and J.K. Waage) pp. 179-208. Island Press, Washington DC.
Magilligan, F.J. & Nislow, K.H. (2005) Changes in hydrologic regime by dams. Geomorphology , 71 , 61-78.
Mahoney, J.M. & Rood, S.B. (1998) Streamflow requirements for cottonwood seedling recruitment- an integrative model. Wetlands 18 , 634-645.
96
McCune, B. & Grace, J.B. (2002) Analysis of Ecological Communities . MJM Software Design, Gleneden Beach, OR.
Merkel, D.L. & Hopkins, H.H. (1957) Life History of Salt Cedar ( Tamarix gallica L.). Transactions of the Kansas Academy of Science , 60 , 360-369.
Merritt, D.M. & Cooper, D.J. (2000) Riparian vegetation and channel change in response to river regulation: a comparative study of regulated and unregulated streams in the Green River Basin, USA. Regulated Rivers: Research and Management , 16 , 543-564.
Merritt, D.M & Poff, N.L. (2010) Shifting dominance of riparian Populus and Tamarix along gradients of flow alteration in western North American rivers. Ecological Applications , in press.
Nilsson, C. & Berggren, K. (2000) Alteration of riparian ecosystems caused by river regulation. BioScience , 50 , 783-792.
Nilsson, C. & Svedmark, M. (2002) Basic principles and ecological consequences of changing water regimes: riparian plant communities. Environmental Management , 30, 468-480.
Poff, N.L. & Ward, J.V. (1989) Implications of streamflow variability and predictability for lotic community structure: a regional analysis of streamflow patterns. Canadian Journal of Fisheries Aquatic Science , 46, 1805-1818.
Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D., Sparks, R.E. & Stromberg, J.C. (1997) The natural flow regime: a paradigm for river conservation and restoration. BioScience , 47 , 769-784.
Poff, N.L., Olden, J.D., Merritt, D.M. & Pepin, D.M. (2007) Homogenization of regional river dynamics by dams and global biodiversity implications. Proceedings of the National Academy of Sciences , 104 , 5732-5737.
Richardson, D.M., Holmes, P.M., Esler, K.J. Galatowitsch, S.M., Stromberg, J.C., Kirkman, S.P., Pysek, P. & Hobbs, R.J. (2007) Riparian vegetation: degradation, alien plant invasions, and restoration prospects. Diversity and Distributions , 13 , 126-139.
Rood, S.B. & Mahoney, J.M. (1990) Collapse of riparian poplar forests downstream from dams in western prairies: probable causes and prospects for mitigation. Environmental Management , 14 , 451-464.
97
Rood, S.B., Gourley, C.R., Ammon, E.M., Hiki, L.G., Klotz, J.R., Morrison, M.L., Mosley, D., Scoppettone, G.G., Swanson, S. & Wagner, P.L. (2003) Flows for floodplain forests: a successful riparian restoration. BioScience , 53 , 647-656.
Scott, M.L., Friedman, J.M. & Auble, G.T. (1996) Fluvial process and the establishment of bottomland trees. Geomorphology 14 , 327-339.
Scott, M.L., Auble, G.T. & Friedman, J.M. (1997) Flood dependency of cottonwood establishment along the Missouri River, Montana, USA. Ecological Applications , 7, 677- 690.
Shafroth, P.B., Auble, G.T. & Scott, M.L. (1995) Germination and establishment of the native plains cottonwood ( Populus deltoides Marshall subsp. monilifera ) and the exotic russian- olive ( Elaeagnus angustifolia L.). Conservation Biology , 9, 1169- 1175.
Shafroth, P.B., Auble, G.T., Stromberg, J.C. & Patten, D.T. (1998) Establishment of woody riparian vegetation in relation to annual patterns of streamflow, Bill Williams River, Arizona. Wetlands , 18 , 577-590.
Shafroth, P.B., Stromberg, J.C. & Patten, D.T. (2002) Riparian vegetation response to altered disturbance and stress regimes. Ecological Applications , 12 , 107-123.
Stevens, L.E. & Gold, B.D. (2003) Monitoring for adaptive management of the Colorado River ecosystem in Glen and Grand Canyons. Monitoring Ecosystems: Interdisciplinary Approaches for Evaluating Ecoregional Initiatives (ed. by D.E. Busch and J.C. Trexler) pp. 101-134. Island Press, Washington, DC.
Stromberg, J.C., Lite, S.J., Marler, R., Paradzick, C., Shafroth, P.B., Shorrock, D., White, J.M. & White, M.S. (2007) Altered stream-flow regimes and invasive plant species: the Tamarix case . Global Ecology and Biogeography , 16 , 381-393.
Tabacchi, E., Correll, D.L., Hauer, R., Pinay, G., Planty-Tabacchi, A. & Wissmar, R.C. (1998) Development, maintenance and role of riparian vegetation in the river landscape. Freshwater Biology , 40 , 497-516.
US Army Corps of Engineers (USACE). (1992) National Inventory of Dams . CD-ROM. Venture Computing Group Inc., Troy, NY.
Walker, K.F. & Thoms, M.C. (1993) Environmental effects of flow regulation on the lower river Murray, Australia. Regulated Rivers: Research and Management , 8, 103-119.
Walker, K.F., Sheldon, F. & Puckridge, J.T. (1995) A perspective on dryland river ecosystems. Regulated Rivers: Research and Management , 11 , 85-104.
98
Ward, J.V. & Stanford, J.A. (1983) Serial discontinuity concept of lotic ecosystems. Dynamics of Lotic Systems (ed. by T.D. Fontaine and S.M. Bartell) pp. 29- 42. Ann Arbor Science, Ann Arbor, MI.
Warren, D.K. & Turner, R.M. (1975) Saltcedar ( Tamarix chinensis ) seed production, seedling establishment, and response to inundation. Journal of the Arizona Academy of Science , 10 , 135-144.
Webb, R.H., Leake, S.A. & Turner, R.M. (2007) The Ribbon of Green: Change in Riparian Vegetation in the Southwestern United States . The University of Arizona Press, Tucson, AZ.
99
TABLES
Table 1. Characteristics of river segments sampled. The number of point bars sampled is indicated in parentheses. The coordinates of the most upstream point bar are given.
USGS gage data obtained from: http://waterdata.usgs.gov. Dam information from
National Inventory of Dams (US Army Corps of Engineers, 1994).
Drainage Mean annual Dam name & Gage number & Location (dms) River segment (n) Area (km 2) flow (m 3s-1) completion year earliest record N W Great Basin : Carson (9) 3,372 10.6 None 10312000, 1911 39°16’57” 119°23’17” Humboldt(10) 11,163 10.6 None 10321000, 1943 40°46’19” 115°54’40” Truckee 1 (10) 3,706 22.9 Lake Tahoe 1913 10350000, 1899 39°30’31” 119°39’13” Truckee 2 (5) 4,341 11.7 Derby 1905 10351600, 1918 39°35’09” 119°26’30” Walker(7) 2,497 5.8 None 10300000, 1914 38°50’20” 119°16’54” Upper Colorado River Basin : Colorado 1 (10) 20,850 108.2 Granby 1941 09095500, 1933 39°16’07” 108°15’16” Shadow Mtn. 1946 Colorado 2 (10) 62,419 203.1 same as upper CO 09180500, 1914 38°45’44” 109°19’24” Green 1 (10) 76,819 118.5 Flaming Gorge 1964 09261000, 1946 40°25’29” 109°14’41” Green 2 (10) 116,162 171.9 Flaming Gorge 1964 09315000, 1894 39°04’41” 110°08’31” Gunnison(14) 20,534 72.1 Morrow Pt. 1963 09152500, 1896 38°59’10” 108°27’06” Crystal 1973 White(8) 10,412 19.3 Taylor Draw 1984 09306500, 1923 39°58’24” 109°10’04” Lower Colorado River Basin : Colorado 3 (7) 366,744 433.3 Glen Canyon 1963 09402500, 1922 36°06’14” 111°49’46” Salt(8) 16,141 27.5 T. Roosevelt 1911 09502000, 1934 33°33’22” 111°32’22” Stewart Mtn. 1930 San Pedro(6) 3,196 1.5 None 09471000, 1904 31°33’57” 110°08’36” San Juan(9) 33,411 57.7 Navajo 1962 09368000, 1934 36°46’01” 108°39’57” Verde(4) 15,957 18.5 Bartlett 1939 09510000, 1904 33°45’24” 111°40’22” Horseshoe 1944 Virgin(9) 13,183 6.9 None 09415000, 1929 36°52’35” 113°56’06” Rio Grande Basin : Rio Grande 1 (10) 41,699 39.4 Cochiti 1975 08319000, 1927 35°28’34” 106°24’12” Rio Grande 2 (12) 45,170 32.7 Cochiti 1975 08330000, 1942 35°06’11” 106°41’32” Rio Grande 3 (11) 76,276 28.2 Elephant Butte 1916 08361000, 1916 33°06’06” 107°17’33” 100
Table 2. Descriptions of the Indicators of Hydrologic Alteration (IHA), predicted direction of change following dam construction, and plant life history aspect that is influenced by the indicator. ↑ = increase, ↓ = decrease, and 0 = no change expected.
Underlined life history stages were hypothesized to be strongly influenced by the indicator. (1) Magilligan and Nislow (2005),
(2) Graf (2006), (3)Mahoney & Rood (1998), (4)Scott et al. (1997), (5)Shafroth et al. (1998), (6)Shafroth et al. (2002).
Indicator Description Predicted change Life history stage Magnitude Mean annual flow (MAF) Mean daily discharge for analysis years ? establishment (4), survival (2) 7-day maximum flow (Max) Maximum median discharge of 7-day duration ↓ (1) germination (4), establishment (4), survival 7-day minimum flow (Min) Minimum median discharge of 7-day duration ↑ (1) establishment, survival Base flow 7-day minimum / mean annual flow ↑ (2) establishment, survival Constancy Lack of variance in daily discharge among years ↑ survival Frequency High pulse count (HPC) Number periods where discharge 25% above median ↑ (1) establishment, survival Duration High pulse duration (HPD) Median duration of high pulses ↓ (1) establishment (2) , survival Timing Date maximum flow Julian date of maximum median discharge 0 (2) germination (6), establishment (5,6) Date minimum flow Julian date of minimum median discharge 0 (2) establishment May median flow (May) Median of May median discharge ↓ (1) establishment , survival August median flow (August) Median of August median discharge ↑ (1) establishment (6), survival Contingency Similarity of discharge on certain days among years ↓ establishment, survival Rate of Change Recession rate Median of negative daily differences (fall rate) ↓ (2) germination (3,5), establishment (3) 101
102
Table 3. Species characteristics and number of river segments at which the species was found. Twenty additional species had low frequency (< 1 river segment) and were excluded from this list.
Scientific Name Family Native status Frequency Acer negundo L. Aceraceae native 2 Artemisia sp. L. Asteraceae native 7 Atriplex sp. L. Chenopodiaceae native 2 Baccharis sp. L. Asteraceae native 5 Baccharis salicifolia Asteraceae native 7 (Ruiz & Pav.) Pers. Baccharis sarothroides A. Gray Asteraceae native 2 Chrysothamnus nauseosus Asteraceae native 2 Pall. ex Pursh Elaeagnus angustifolia L. Elaeagnaceae non-native 14 Juniperus sp. L. Cupressaceae native 3 Pluchea sericea (Nutt.) Coville Asteraceae native 3 Populus deltoides Salicaceae native 5 Bartram ex Marsh. Populus fremontii S. Watson Salicaceae native 12 Populus tremuloides Michx. Salicaceae native 2 Prosopis glandulosa Torr. Fabaceae native 6 Rhus trilobata Nutt. Anacardiaceae native 5 Ribes sp. L. Grossulariaceae native 2 Rosa sp. L. Rosaceae native 3 Salix exigua Nutt. Salicaceae native 16 Salix gooddingii C.R. Ball Salicaceae native 7 Salix sp. L. Salicaceae native 8 Sarcobatus vermiculatus Chenopodiaceae native 3 (Hook.) Torr. Shepherdia argentea (Pursh) Nutt. Elaeagnaceae native 5 Tamarix spp. L. Tamaricaceae non-native 19 Ulmus pumila L. Ulmaceae non-native 9
Table 4. Pearson correlation coefficients and significance for hydrologic and climatic variables with vegetation cover. Climatic data obtained from: http://www.wrcc.dri.edu. * = p ≤ 0.05 and ** = p ≤ 0.01. All others are not significant (p ≥ 0.05). N = 20 for all except Elaeagnus where n = 17. See table 2 for parameter abbreviations.
Parameter Non-native Tamarix Elaeagnus young Populus All Populus Salix exigua Hydrologic : Constancy 0.62** 0.57** 0.14 -0.36 -0.29 -0.46* ln Drainage area 0.61** 0.56** 0.15 -0.51* -0.57** -0.19 ln August 0.53* 0.49* 0.16 -0.54* -0.50* -0.30 ln Mean annual flow 0.48* 0.43 0.10 -0.56** -0.40 -0.16 ln May 0.47* 0.37 0.18 -0.62** -0.35 -0.01 ln recession rate 0.45* 0.36 0.20 -0.56** -0.34 -0.17 ln max 0.44 0.37 0.13 -0.42 -0.34 -0.14 arc-sin Min 0.43 0.32 0.21 -0.48* -0.37 -0.15 Date Max 0.32 0.41 -0.14 0.16 -0.09 -0.13 ln base flow 0.26 0.16 0.22 -0.36 -0.35 0.01 Date Min 0.19 0.11 0.18 -0.39 -0.17 -0.03 ln High pulse duration -0.10 0.04 -0.42 -0.26 -0.27 0.49* Contingency -0.05 0.02 -0.22 0.08 -0.03 0.42 High pulse count 0.04 0.07 0.14 0.41 0.19 -0.45* Climatic : Max temperature 0.33 0.46* -0.11 -0.26 -0.44 -0.37 Min temperature. -0.12 0.10 -0.43 0.04 -0.08 -0.35 Snowfall -0.33 0.06 -0.07 0.05 0.18 0.55* Precipitation -0.20 -0.38 0.12 0.22 0.04 -0.19 103
104
Table 5. Mean differences ([post-regulation value] – [pre-regulation value] / [pre- regulation value]) in the Indicators of Hydrologic Alteration (IHA) parameter medians before and after dam construction or 1962 for non-dammed rivers with standard errors.
Positive numbers indicate an increase in the parameter, and negative numbers indicate a decrease. The p-values from exact Wilcoxon rank sum tests reveal results from a test of equal distributions of the change in IHA parameters in dammed and non-dammed rivers.
Parameter Dammed (n = 7) Non-dammed (n = 7) Difference p-value
High pulse count 0.48 ± 0.22 -0.16 ± 0.06 0.64 0.01 Constancy 0.21 ± 0.12 0.01 ± 0.03 0.20 0.03 August median 0.46 ± 0.15 -0.08 ± 0.11 0.54 0.03 Seven-day minimum 0.72 ± 0.28 0.01 ± 0.10 0.71 0.03 Base flow 0.76 ± 0.29 0.11 ± 0.09 0.65 0.04 Date maximum 0.13 ± 0.09 -0.02 ± 0.02 0.15 0.04 Recession rate 0.30 ± 0.29 -0.15 ± 0.10 0.18 0.06 High pulse duration -0.10 ± 0.15 0.23 ± 0.26 0.24 0.50 Mean annual flow -0.03 ± 0.09 0.01 ± 0.09 0.04 0.62 Date minimum -0.11 ± 0.14 -0.05 ± 0.05 0.06 0.62 May median -0.02 ± 0.19 -0.09 ± 0.09 0.07 1.00 Contingency -0.46 ± 0.08 -0.14 ± 0.06 0.32 0.01 Seven-day maximum -0.35 ± 0.14 -0.13 ± 0.10 0.22 0.02
Table 6. Variables included in the three best generalized linear models of relative vegetation cover according to AIC c model comparisons. wi = Akaike weight or normalized likelihood relative to set of three models. ∆i = difference in AIC c score from best model. N = 20 for all except Elaeagnus where n = 17. See table 2 for other variable abbreviations.
Response Variable Best model ( wi) Second-best model ( wi ; ∆i) Third-best model ( wi ; ∆i)
Non-native Constancy (0.40) Constancy (0.34 ; 0.33) Date Max (0.25 ; 0.93) * ln Drainage area + ln Drainage area + ln Drainage area
Tamarix Date Max (0.71) Date Max (0.17 ; 2.81) Date Max (0.12 ; 3.53) * Constancy * ln Drainage area + ln Drainage area
Elaeagnus Min temperature (0.54) Min temperature (0.45 ; 0.36) Date Max (0.01 ; 7.22) * ln High pulse duration + ln High pulse duration + ln Drainage area young Populus ln Mean annual flow (0.43) Max temperature (0.37 ; 0.32) ln Mean annual flow (0.20 ; 1.5) + ln Max * High pulse count * ln Max
Populus Max temperature (0.69) Max temperature (0.19 ; 2.62) Max temperature (0.12 ; 3.5) * High pulse count + High pulse count + ln High pulse duration
Salix exigua Constancy (0.36) Snow (0.36 ; 0.03) Snow (0.28 ; 0.49) * High pulse duration * High pulse duration * High pulse count 105
106
FIGURE CAPTIONS
Figure 1. Map of study sites. The twenty river segments are shown with symbols that
indicate total non-native cover (%). Segments with large dams upstream are noted with
bold text, and shading represents physiographic provinces.
Figure 2. Area-weighted canopy cover of the most common non-native ( Tamarix and
Elaeagnus ) and native ( Populus [including Populus < 30 years-old] and S. exigua ) taxa
relative to the total vegetation cover of all point bars sampled in river segments. Bars do
not total 100% because of the presence of other, less common species (see Table 3).
Figure 3. PCA plot of the first two principal component axes for hydrologic variables
“pre” and “post” dam construction or “pre” and “post” 1962 for non-dammed river
segments. Influential loadings of flow variables are noted in the axis labels. CO =
Colorado and Rio = Rio Grande
Figure 4. Regressions of vegetation cover on the Euclidean distance of Indicators of
Hydrologic Alteration parameters “before” and “after” dam construction or “before” and
“after” 1962 for non-dammed river segments calculated from PCA axes one through five.
The x-axis is unitless because it is derived from the PCA. An outlier (hollow circle) is
excluded from the trend line in 4b. The relative order of river segments on the x-axis is
shown at the top of the graphs. * denotes significant regression models (p ≤ 0.05).
107
FIGURES
Figure 1.
108
Figure 2.
109
Figure 3.
110
Figure 4.
111
CHAPTER 4. GUIDELINES FOR ECOSYSTEM RESTORATION
OF REGULATED RIVERS
ABSTRACT
The negative effects of river regulation have stimulated much research on river restoration. Restoration of hydrologic connectivity, flow variability, and geomorphic processes are ultimate goals of restoration ecologists. Presently, ecologists recognize the need for watershed-scale, process-based approaches that may or may not emulate historic reference conditions. The restoration of riparian vegetation poses a challenge to managers because of the interaction between vegetation, fluvial geomorphic processes, and non-native plant invasions. We formulated seven ecological restoration principles that bring together research in riparian systems and focus on riparian vegetation. These principles are designed to link the science of restoration with restoration practices.
We advocate formulation of alternative flow regimes based on knowledge of natural variability. The repercussions of increased fire frequency in regulated systems must be considered along with opportunities to use fires and floods to reinvigorate plant establishment and geomorphic processes. Functional groups can then be used to predict the effects of management scenarios on different plant life stages. The influence of multi-trophic species interactions which are often cryptic must also be considered. Site and watershed-based restoration approaches are often required for restoration of connectivity, variability, and geomorphic processes. Long-term restoration trajectories are necessary to prepare for future climate change. Instead of discussing how we should
112
restore riparian ecosystems, we must now ask how we can preserve or enhance the adaptation strategies of species and rivers. The most promising strategy for riparian restoration planning, implementation, and monitoring remains adaptive management.
However, the temporal scale of adaptive management must extend to incorporate climate change scenarios.
INTRODUCTION
The indisputable value of rivers and obvious negative effects of river regulation have stimulated much research on river restoration (Goodwin et al. 1997; Arthington &
Pusey 2003; Hughes & Rood 2003). Dams, diversions, and groundwater withdrawal reduce water and sediment discharge (i.e., longitudinal connectivity), floodplain inundation (i.e., lateral connectivity), and groundwater upwelling and downwelling (i.e., vertical connectivity). Channel incision can further decrease floodplain inundation because the magnitude of flows required for floodplain inundation increases (Kondolf et al. 2006). Hydrologic connectivity, flow variability, and fluvial geomorphic processes determine riparian ecosystem structure and function (Stanford et al. 1996; Ward et al.
2001; Kondolf et al. 2006). The science of restoration ecology is based on understanding these organizing factors, their influence on riparian biota, and ways to restore ecosystem function through alteration of connectivity, variability, and geomorphic processes.
Floods promote longitudinal, latitudinal, and vertical connectivity. Because most
forms of river regulation suppress floods, controlled floods are released along dammed
rivers to revive ecological processes that rely on flooding. Flood and sediment releases
113
are examples of watershed-scale restoration that influence both reach- and site-scale processes (Hughes et al. 2001). Restoration planning at the watershed scale is most appropriate, but political consensus at this scale is often impossible (Hughes et al. 2001;
Wohl et al. 2005; Richardson et al. 2007). Therefore, managers may need to focus efforts on watersheds that are relatively undeveloped, particular river reaches, or target sites
(Richardson et al. 2007). Regardless of the scale of restoration, projects should not be implemented without first considering interactions of watershed-scale ecosystem processes with management activities.
The use of a particular end state, form, or reference condition as a restoration goal
has been discouraged. Instead, ecologists recommend restoring function or process to
riparian systems (Goodwin et al. 1997; Hughes et al. 2005; Wohl et al., 2005; Kondolf et
al. 2006). This requires restoring hydrologic connectivity, flow variability, and fluvial
geomorphic processes (Richter et al. 1998; Kondolf et al. 2006). Along dammed rivers,
water release timing and amount can be altered to meet these process-oriented goals.
Much research has focused on designing flow prescriptions (Arthington & Pusey 2003;
Hughes & Rood 2003; Richter & Thomas 2007). Stanford et al. (1996) suggest
restoration of floods, sediment addition, and base flow stabilization. Other methods of
river restoration such as river channel reconstruction (meander and riffle construction)
and removal of in-stream structures also restore connectivity (Kondolf et al. 2006). In
some cases, increased connectivity may have negative effects (e.g., the spread of invasive
species via inter-basin water transfers).
114
Riparian plants interact with fluvial geomorphology by influencing spatial and temporal patterns of erosion and deposition (Gurnell 1995). In turn, fluvial geomorphology, in the form of floods and sediment reworking, interacts with plant life history traits to determine the probability of establishment and persistence (Hupp &
Osterkamp 1996). Native plant populations that are adapted to specific flood characteristics often decline following river regulation. Conversely, populations of non- native species often expand following these habitat changes (Stanford et al. 1996). For instance, the decline of Populus spp. in North America is associated with river regulation
(Howe & Knopf 1991) and was coincident with the Tamarix spp. invasion (Merritt &
Poff 2010). Similarly, Eucalyptus spp. have declined in Australia due to reduced flood
frequency (Bren 1988), whereas non-native Salix spp. have spread prolifically (Stokes &
Cunningham 2006). The presence of non-native plants per se does not necessitate
control. However, if non-native plants act as transformer species that alter ecosystem
function, removal may be required to reestablish native plant communities and restore
fluvial geomorphic processes (Richardson et al. 2007).
We define riparian restoration as, “…assisting the recovery of ecological integrity
in a degraded watershed system by reestablishing the processes necessary to support the
natural ecosystem within a watershed” (Wohl et al. 2005). Because of the ability of
vegetation to influence fluvial geomorphic processes, we focus on restoration principles
relevant to woody riparian vegetation. Restoration of processes that maintain woody
vegetation addresses three of the five most common restoration goals identified by
115
Bernhardt et al. (2005): 1) to manage riparian zones, 2) enhance water quality, and 3) improve in-stream habitat.
Riparian restoration provides the ultimate challenge for understanding riparian ecosystem structure and function. The opportunity to readily change ecosystem processes through flow and sediment alterations along regulated rivers is unique. In the future, river managers will need to formulate restoration prescriptions in light of climate change predictions. This will require long-term, sustainable, and adaptable approaches for riparian restoration. We have organized seven guidelines for river restorationists based on a comprehensive review of riparian research (Table 1). The majority of these guidelines are most applicable to regulated rivers where flow regimes can be altered for restoration purposes. It is our hope that these guidelines, arranged along a continuum of temporal scale and integrating fundamental theory of natural disturbance, ecological succession and patch dynamics (Figure 1), will more closely link the science and implementation of riparian restoration.
GUIDELINES FOR RIVER RESTORATIONISTS:
1) Use the natural flow regime concept as a guide to understand how the ecosystem
functions. Do not attempt to emulate the natural flow regime without
considering effects of climate change, non-native invasive species, and prior
management strategies on ecosystem structure and function.
The natural flow regime concept acknowledges the connection between components of the flow regime (magnitude, timing, frequency, rate of change, duration) and riparian
116
ecosystem function (Poff et al. 1997). Variability in flow components maintains species and landscape diversity. Broad acceptance of this concept has spurred interest in managing toward a natural flow regime. Attempts to restore the natural flow regime to rivers are a vast improvement in restoration strategies as compared with merely maintaining minimum flows (Poff et al. 1997). However, many factors have changed with river regulation, and restoration of historic flow regimes on regulated rivers may not result in desired outcomes (Stromberg 2001). For example, Tamarix populations would likely increase along many regulated rivers in the southwestern US if historic spring floods were reintroduced (Birken & Cooper 2006; Mortenson et al. in prep ).
Knowledge of the natural variability of a landscape promotes understanding of the way in which the system functioned in the past (Landres et al. 1999). However, use of natural variability concepts (e.g., natural flow regime) for formulation of management strategies can be sociopolitically infeasible if historical disturbances were of a large magnitude, size, or intensity. We also do not understand the cumulative effects of disturbances including the interactions of disturbance events. For instance, because the effects of one flood can persist for 100 years, the consequences of future floods will depend on prior flood characteristics. Past and future changes in climate and the legacy of previous management activities may alter patterns and processes so that past variability has no application to current ecosystem function (Landres et al. 1999).
Therefore, although we need to understand the hydrologic characteristics of natural flow regimes, it is not always advisable to attempt to reconstruct historical or pre-dam conditions.
117
Controlled floods should be designed with knowledge of the effects of river regulation on historic flood characteristics, plant and animal populations, and river channel morphology (Richter & Thomas 2007). Changes in hydrologic conductivity, variability, and patterns of erosion and deposition should be assessed. If little change has occurred and the ecosystem appears minimally altered (i.e., relict ecosystem), natural flood characteristics may represent appropriate management strategies. For novel riparian ecosystems ( sensu Johnson 2002), alternative flow regime scenarios should be considered that are socially and economically feasible and have the potential to restore conductivity, variability, and geomorphic processes. Controlled floods may need to be implemented as experiments and their lasting effects monitored (i.e., adaptive management) (Walters 1997). In this way we can monitor the response of novel riparian ecosystems to alternative flow regimes.
2) Reduced flood frequency and magnitude associated with river regulation will
likely increase fire occurrence in riparian corridors. Managers must consider
potential consequences of fire for riparian ecosystems. In addition, fires may be
used in congruence with small floods to trigger establishment of riparian trees.
Due to river regulation, non-native plant invasions, and fire suppression, fuels have accumulated in riparian corridors (Ellis 2001). Continued river regulation will increase fire frequency and intensity in riparian ecosystems. Will the effects of fires be similar to the effects of floods? The life history traits that allow riparian plants to respond to floods are similar to those of fire-adapted species (resprouting, clonal reproduction, production
118
of many seeds) (Dwire & Kauffman 2003). However, pioneer shrubs and trees require reworking of sediment to create recruitment sites for seedlings (Hughes & Rood 2001).
Sediment reworking does not occur with fire alone. Fires may also cause mortality of adult, riparian trees (Ellis 2001; White & Stromberg 2007). Populus and Salix gooddingii , two riparian tree species that are obligate seeders with only limited resprouting ability, may not be capable of returning to previous densities via seedling establishment following a fire. In the absence of floods, fires may favor the dominance of resprouting or clonal shrubs (e.g., Tamarix , Salix exigua ) or herbaceous vegetation
(White & Stromberg 2007).
Large floods or other forms of intense disturbance (e.g., mechanical tree removal) are necessary to shift dominance to native riparian trees in the southwestern US (Stromberg
& Chew 2002). However, in areas where development exists along the floodplain of a river, high-magnitude floods are infeasible. Moreover, if floods are excluded from a river for 10 – 15 years, establishment of woody vegetation near the river channel may constrain the channel so that future floods are ineffective at removing vegetation
(Friedman & Auble 1999). In these situations, controlled burns combined with small floods may be an appropriate restoration option. Ellis (2001) observed vegetation changes in sites that had experienced recent floods and fire. Her results indicated that fire and subsequent flooding in early spring may be necessary to promote establishment of native trees in the Rio Grande bosque. A combination of fire and small floods may result in sediment reworking and subsequent seedling establishment (Dwire & Kauffman 2003).
119
In riparian corridors fires and small floods can restore fluvial geomorphic processes.
Large rainfall events that occur immediately after a fire may initiate erosion. Fires also increase available water resources because burned areas are often converted from shrubland to grassland (White & Stromberg 2007). Therefore, lateral and vertical hydrologic conductivity can increase through the combination of disturbance types. Bank erosion and downstream transport of sediments also promote longitudinal conductivity.
3) Functional groups defined by phenology, life history traits, and tolerance to
disturbance and stress are useful for predicting community response to flow
regime scenarios. Artificial groupings such as “native” and “non-native” are
counterproductive.
The species-specific phenology, life history traits, and flood adaptations of riparian species are indicators of their potential response to flow regimes and fires. Merritt et al.
(2009) provide a review of the existing models that link these characteristics of riparian vegetation to flow regime response. Functional groups or guilds based on these characteristics allow prediction of potential results of management treatments (Roberts
2002; Merritt et al. 2009). Determination of the functional groups present in an ecosystem might allow generalized predictions to be made across ecosystems (Merritt et al. 2009). In this way, restoration strategies for target species assemblages could be designed with prior knowledge obtained from other riparian ecosystems.
For example, riparian trees and shrubs in the southwestern US can be grouped according to their establishment requirements and tolerance to floods, drought, and shade
120
(Table 2). The ability to reproduce clonally is a defining property that allows plants to spread in the absence of disturbance and recolonize areas quickly following severe disturbance. Shade tolerant species do not require disturbance prior to establishment, and therefore may become more dominant in the absence of moderate to high severity floods
(Katz et al. 2005; Dewine & Cooper 2008). Seed dispersal timing and persistence also need to be taken into account to determine the probability of establishment after specific management treatments. This general, functional framework can aid restorationists in prescribing floods or controlled burns based on information from other shrub and tree species.
Functional groups clarify the similarities and differences between native and non- native focal species. However, the search for certain flow regime treatments that increase native species richness and decrease non-native species richness (e.g., Howell & Benson
2000) is counterproductive. The response of species to flow regime changes is independent of their native status. Instead, the functional traits of native and non-native species determine their relative abundance under specific environmental conditions
(Mortenson & Weisberg 2010). A life-history perspective should be used to compare non-native species that alter ecosystem function and warrant control with native species included in the target assemblage (also see guideline #5). Restoration of native species may only be achievable if the functional traits of non-native transformers do not significantly overlap with native target species.
121
4) Incorporate into management plans knowledge of the entire plant life cycle
including the establishment niche, patterns of growth, persistence, and
mortality. S uccessional processes may need to be managed through mechanical
removal or planting.
The recruitment box model (see Mahoney & Rood 1998) is a valuable conceptual
model that focuses on flow conditions required for Populus establishment. The model
incorporates the timing of Populus seed release with flow stage elevation and recession
rate to identify the shape of the annual hydrograph that should result in seedling
establishment. However, juveniles and adults of the same species usually have different
water availability requirements for survival (Roberts 2002). Therefore, management
strategies should not be limited to one life stage. Riparian plant persistence is controlled
by future disturbances and water and light availability in congruence with physiological
adaptations that allow tolerance to inundation, scour, burial (e.g., resprouting, bendable
shoots and branches, clonality), drought, and shade. Differential survival may ultimately
determine the fate of restoration treatments. Although establishment is a necessary
condition for species persistence, it is not a sufficient condition and there should not be
sole emphasis on the establishment phase.
A comprehensive understanding of successional trajectories is becoming increasingly
important due to reduction in disturbance magnitude and frequency along regulated
rivers. Reduction in flooding causes competition to become more influential in
determining community biodiversity (Huston 1994). For example, an entirely different
riparian community has persisted following construction of the Glen Canyon Dam that
122
drastically reduced floods along the Colorado River through the Grand Canyon (Johnson
1991). Pre-dam vegetation communities occurred at high flow stage elevations which were protected from annual scouring floods and provided with water. Vegetation now extends to low elevations where the invasive shrub Tamarix is dominant (Johnson 1991).
Management of these communities is hampered by uncertainty of successional trajectories especially in relation to Tamarix . In some areas willows may outcompete
Tamarix (Stevens 1989), but without full understanding of this process, other forms of
Tamarix control must be considered. Because Tamarix is shade intolerant, Dewine &
Cooper (2008) advocate planting Acer negundo (box elder) or other shade tolerant riparian trees to replace Tamarix . Galatowitsch & Richardson (2005) also recommend planting shade-tolerant riparian trees to reduce non-native woody plants along rivers in
South Africa.
The majority of restoration treatments are designed to interrupt succession through mechanical interventions (e.g., invasive plant removal, meander construction) and managed floods. In areas where managed floods are not an option, managing succession through competitive interactions may be a more sustainable restoration method than mechanical removal of vegetation. Because prior successional trajectories were historically interrupted by flooding and long-term data about plant succession following dam construction are lacking, the predictability of this approach is limited (Nilsson &
Berggren 2000). However, knowledge of plant life history strategies can aid in selection of competitive plant species for directed succession. In the absence of flooding or adequate seed supply, planting species that can establish following removal of non-native
123
stands of shrubs or trees and quickly grow large canopies to exclude reestablishment of non-native species is a valid restoration option . In time, more shade-tolerant species may
establish or require planting (Galatowitsch & Richardson 2005).
5) Manage at the ecosystem level and consider species interactions. Managing for
single species or single taxa can result in unpredictable consequences.
Riparian restoration may be mandated by endangered species legislation, but restoration plans focused on one species are likely to fail to meet holistic restoration goals
(Ellis et al. 2002). For example, a spring flood and high, steady summer flows implemented along the Colorado River in Grand Canyon were designed to encourage spawning of native fish but also resulted in increased establishment of Tamarix , an
invasive shrub (Stevens & Gold 2003; Mortenson et al. in prep ). Instead of designing
restoration treatments for particular species, target species assemblages can be
established and their abundance managed (Hughes et al. 2005). Ideally these
assemblages would include species from multiple trophic levels.
Multi-trophic interactions can covertly influence riparian ecosystems. There are
many examples of multi-trophic interactions in relation to Tamarix in North America.
Dominance of Tamarix in the Grand Canyon may be maintained by leaf hoppers
(Siemion & Stevens, in review ) and beavers (Mortenson et al. 2008). Leaf hoppers
(Opsius stactogalus Fieber ) are abundant herbivores on Tamarix . Leaf hopper exudates support populations of fungi that can reduce seed viability and cause seedling mortality of potential competitors (Siemion & Stevens, in review ). Beavers may contribute to
124
Tamarix dominance through selective foraging of Salix spp. (Mortenson et al. 2008).
Competition experiments with Populus and Tamarix seedlings and mycorrhizae additions revealed that Populus may benefit from mycorrhizal associations (Beauchamp et al.
2005). In areas where Tamarix has grown in monoculture for extended periods of time,
populations of mycorrhizae may be depleted. These areas may not be suitable for
Populus if mycorrhizae are necessary for persistence (Beauchamp et al. 2005).
One well-known example of multi-tropic interactions in riparian systems involves three trophic levels. The decline of wolf populations in Yellowstone National Park in the early 20 th Century allowed elk populations to increase. Elk browse heavily on willows in riparian areas. Elk herbivory shortens willow stems and reduces the effectiveness of stems for beaver dam construction (Wolf et al. 2007). Therefore, many beavers abandoned their beaver ponds. The absence of beaver ponds caused channel incision.
Declines in the water table resulting from channel incision reduced willow stands.
Restoration of willows in Yellowstone riparian areas will now require construction of ponds and exclusion of elk (Wolf et al. 2007).
Multi-trophic interactions are often ignored in the context of riparian restoration.
However, restoration treatments that are not designed with multi-trophic species interactions in mind may fail. For example, in a restoration project in Maryland, survival of planted seedlings was four times greater if shelters were installed to prevent herbivory
(Sweeney et al. 2002). Knowledge of the interactions of target species assemblages with other components of the ecosystem including geomorphology, flow regime, and other species are necessary.
125
6) In most cases site- and watershed-scale restoration approaches are necessary to
achieve restoration goals.
In riparian systems site-based restoration measures include selective species removal
or addition, river channel reconstruction, and road restoration while watershed-based
approaches include altering the flow or sediment regime and dam removal (Hughes et al.
2001). Watershed-based approaches are preferable over site-based approaches for
riparian restoration because of ability to restore ecological function and increase
sustainability (Wohl et al. 2005; Arthington et al. 2006). However, dominance of new
species may create an abiotic or biotic threshold (Whisenant 1999) or an alternative state
(Suding et al. 2004) that can prevent restoration via abiotic or biotic processes alone.
Biotic or abiotic processes may need to be restored at the site scale. Ultimately, the
crossing of biotic and abiotic thresholds must be assessed to determine the likelihood of
restoration success (Chambers et al. 2004; Suding et al. 2004) and the scale at which
restoration activities should be conducted.
Removal of non-native plants and reintroduction of native plants is often necessary to
achieve restoration goals (Taylor et al. 1999; Galatowitsch & Richardson 2005; Taylor et
al. 2006). Combinations of prescribed floods and mechanical removal of invasive plants
has been a successful strategy for restoration of Populus forests along the Rio Grande. In
1993 and 1994 the Rio Grande was flooded during the historic time of pre-dam floods.
Populus establishment was aided by the floods and mechanical removal of adult Tamarix ,
but Tamarix seedlings were abundant in all treatments (Taylor et al. 1999). A more
126
recent survey of the treatment area revealed increased abundance of Populus and decrease or no change in Tamarix (Taylor et al. 2006). Site-based approaches can restore ecosystem function. In South Africa, non-native invasive trees are mechanically removed from fynbos, grassland, and savannah communities to increase streamflow (Van
Wilgen et al. 1998).
7) When designing flow prescriptions, consider the long-term flow regime,
including frequency and predictability, not just potential effects of one flood
treatment. Similarly, restoration projects should encompass longer time scales
(50 years or more) to align with the temporal scale of patch dynamics in which
the interaction between geomorphology and biotic communities are strongest.
In this way, restoration plans can encompass climate change predictions.
The design of river restoration strategies has traditionally considered time scales of ten years or less. However, the patch dynamics of riparian landscapes are maintained by processes occurring at a time scale of 100 years or greater (Steiger et al. 2005).
Moreover, the sequence of floods and droughts (i.e., cumulative interaction of floods) determines the response of a river channel to certain magnitude floods (Hooke 1996).
Therefore, a controlled flood with the same characteristics may generate different results depending on the characteristics of prior floods. Plants evolve with disturbance regimes so that the effect of floods on vegetation is also constantly evolving (Hughes 1994). We must design flow prescriptions that incorporate the complex multi-scale structure of river systems and focus on frequency and predictability of the flow regime.
127
Restoration of flow regimes should incorporate knowledge of long-term
hydrogeomorphic processes and interactions with life history traits to create a sustainable
system. Hughes et al. (2005) recommend avoiding assessment of restoration projects on
an annual scale which may be misleading due to intrinsic inter-annual variability.
Instead, proposed restoration trajectories that incorporate potential effects of multiple
restoration treatments over decades should reveal the ability of restoration practices to
meet goals. With this framework, external forces such as climate change and non-native
species introductions might change the restoration trajectory (Hughes et al. 2005).
Restoration results will only be sustainable if river ecosystems are able to adjust to future
changes. Instead of discussing how we should restore riparian ecosystems, we must now
ask how we can preserve or enhance the adaptation strategies of species and rivers.
Nijssen et al. (2001) used climate models to make climate change predictions for nine large river basins across the globe. Temperatures are expected to warm within all of these river basins, especially those at high latitudes. Precipitation predictions are variable within watersheds as are predictions of river discharge (Palmer et al. 2008). Tropical river basins are expected to experience an increase in precipitation but a decrease in streamflow due to increased evaporation resulting from higher temperatures.
The most dramatic streamflow changes are expected in snowmelt-driven river basins.
Due to overall warming trends, less precipitation is expected to fall as snow. Therefore, more winter flooding and earlier, lower magnitude spring runoff is predicted (Nijssen et al. 2001). In the western US, the amount of moisture stored in snowpack decreased in the
20 th Century, and temperature predictions suggest that these trends will continue (Mote et
128
al. 2005). Streamflows in late summer may be drastically reduced without continued snowpack input. Severe droughts may cause mortality of riparian plants, particularly along regulated rivers in arid or semi-arid habitats. Droughts may also reduce sediment transport and deposition which maintain riparian habitat.
Although regulated rivers are seemingly decoupled from climate change, changing human needs for water and power associated with warming will indirectly influence these systems (Meyer et al. 2001). Dammed rivers are expected to require more restoration actions designed to mitigate floods and droughts, while non-dammed rivers are expected to be more resilient to predicted disturbances (Palmer et al. 2008). Dam failure associated with winter flooding will be an increasing concern. Changes in variability and seasonality of streamflow due to changing climate are expected to most influence riparian ecosystems (Meyer et al. 2001). Increased climate variability in the future adds to the difficulty of predicting restoration trajectories. Investigations of past responses of landform and vegetation dynamics to Holocene climate change can serve as valuable information for prediction of the effects of future climate change (Hughes 1994; Miller et al. 2004; Tausch et al. 2004).
CONCLUSIONS
Adaptive management (sensu Holling 1978) requires treating management activities as large-scale experiments. This approach is limited in relation to riparian restoration projects because there are no opportunities for true controls or replication at the scale of river reaches or watersheds. Assessment of the long-term influences of
129
restoration treatments is difficult when using a before-after statistical design due to environmental changes not related to the treatment (Smith 2002). However, the adaptive management paradigm promotes the essential involvement of scientists and managers during restoration implementation and monitoring (Walters et al. 1997).
The incorporation of adaptive management strategies with climate change predictions can function as a viable framework for future restoration (Lawler 2009). The lack of knowledge and uncertainty of complex ecological system functions combined with uncertainties regarding climate change produce a need for adaptive management.
As new knowledge is obtained through management treatments (e.g., controlled floods) and improved climate predictions, future management plans can be adjusted. However, the appropriate temporal scale in which adaptive management plans should be formulated has not been established (Wohl et al. 2005). The idea of adaptive management must integrate with long-term restoration plans (see restoration principle #7).
Riparian restoration projects are costly (see Bernhardt et al. 2005), and prioritization must be a cornerstone of long-term restoration planning. Some watersheds may be extensively developed and sediment dynamics so altered that restoration is unlikely to succeed. Which rivers will be most affected by climate change? Increased flooding or droughts will affect the fluvial geomorphology of alluvial rivers and have minimal effects on bedrock-constrained rivers (Goudie 2006). Which species may be capable of adapting to new flow regimes? Species with general reproductive phenologies are likely to be resilient to changes in flow. The necessary complexity and long-term
130
commitments of riparian restoration strategies may necessitate concentration of efforts on minimally altered or critically important river ecosystems.
Because human demands for fresh water are increasing, the water requirements of
river ecosystems must be included in political water allocations (Arthington & Pusey
2003). The ultimate goal of river restoration is to recreate sustainable ecosystems that
can adapt to changing climate and disturbance conditions. Restoration of hydrologic
variability and connectivity to dammed rivers will require a mixture of watershed
(alternative flow prescriptions) and site-scale (planting of shade trees) approaches. The
strategies should be based on the expected response of all life stages of riparian plants
and species from other trophic levels to management scenarios. This will require a long-
term adaptive management strategy and long-term commitments from society.
Continued funding of restoration projects and water allocations for riparian ecosystems
are necessary to achieve these goals.
REFERENCES
Arthington, A. H., S. E. Bunn, N. L. Poff, and R. J. Naiman. 2006. The challenge of providing environmental flow rules to sustain river ecosystems. Ecological Applications 16 : 1311-1318.
Arthington, A., and B. Pusey. 2003. Flow restoration and protection in Australian rivers. River Research and Applications 19 : 377-395.
Beauchamp, V. B., J. C. Stromberg, and J. C. Stutz. 2005. Interactions between Tamarix ramosissima (saltcedar), Populus fremontii (cottonwood), and mycorrhizal fungi: effects on seedling growth and plant species coexistence. Plant and Soil 275 : 221- 231.
131
Bernhardt, E. S., M. A. Palmer, J. D. Allan, G. Alexander, K. Barnas, S. Brooks, J. Carr, S. Clayton, C. Dahm, J. Follstad-Shah, D. Galat, S. Gloss, P. Goodwin, D. Hart, B. Hassett, R. Jenkinson, S. Katz, G. M. Kondolf, P. S. Lake, R. Lave, J. L. Meyer, T. K. O'Donnell, L. Pagano, B. Powell, and E. Sudduth. 2005. Synthesizing U.S. River Restoration Efforts. Science 308 : 636-637.
Birken, A. S., and D. J. Cooper. 2006. Processes of Tamarix invasion and floodplain development along the lower Green River, Utah. Ecological Applications 16 : 1103-1120.
Bren, L. J. 1988. Effects of river regulation on flooding of a riparian red gum forest on the River Murray, Australia. Regulated Rivers: Research and Management 2: 65- 77.
Chambers, J. C., J. R. Miller, D. Germanoski, and D. A. Weixelman. 2004. Process-based approaches for managing and restoring riparian ecosystems. Pages 262-292 In J. C. Chambers and J. A. MacMahon, editors. Great Basin Riparian Ecosystems. Island Press, Washington D.C.
Dewine, J. M., and D. J. Cooper. 2008. Canopy shade and the successional replacement of tamarisk by native box elder. Journal of Applied Ecology 45 : 505-514.
Dwire, K. A., and J. B. Kauffman. 2003. Fire and riparian ecosystems in landscapes of the western USA. Forest Ecology and Management 178 : 61-74.
Ellis, L. M. 2001. Short-term response of woody plants to fire in a Rio Grande riparian forest, Central New Mexico, USA. Biological Conservation 94 : 159-170.
Ellis, L. M., C. S. Crawford, and M. C. Jr. Molles. 2002. The role of the flood pulse in ecosystem-level processes in southwestern riparian forests: a case study from the middle Rio Grande. Pages 51-107 In B. A. Middleton, editor. Flood Pulsing in Wetlands: Restoring the Natural Hydrological Balance. John Wiley & Sons, New York, NY.
Friedman, J. M., and G. T. Auble. 1999. Mortality of riparian box elder from sediment mobilization and extended inundation. Regulated Rivers: Research & Management 15 : 463-476.
Galatowitsch, S., and D. M. Richardson. 2005. Riparian scrub recovery after clearing of invasive alien trees in headwater streams of the Western Cape, South Africa. Biological Conservation 122 : 509-521.
132
Goodwin, C. N., C. P. Hawkins, and J. L. Kershner. 1997. Riparian restoration in the western United States: overview and perspective. Restoration Ecology 5(4S): 4- 14.
Goudie, A. S. 2006. Global warming and fluvial geomorphology. Geomorphology 79 : 384–394.
Gurnell, A. M. 1995. Vegetation along river corridors: hydrogeomorphological interactions. Pages 237-260 in A. Gurnell and G. Petts, editors. Changing River Channels. John Wiley & Sons, New York, NY.
Holling, C. S. 1978. Adaptive Environmental Assessment and Management. John Wiley & Sons, New York, NY.
Hooke, J. M. 1996. River responses to decadal-scale changes in discharge regime: the Gila River, SE Arizona. Pages 191-204 in J. Branson, A. G. Brown, and K. J. Gregory, editors. Global Continental Changes: the Context of Palaeohydrology. Geological Society Special Publication 115. Geological Society, London, UK.
Howe, W. H., and F. L. Knopf. 1991. On the imminent decline of Rio Grande cottonwoods in central New Mexico. The Southwestern Naturalist 36 : 218-224.
Howell, J., and D. Benson. 2000. Predicting potential impacts of environmental flows on weedy riparian vegetation of the Hawkesbury-Nepean River, south-eastern Australia. Austral Ecology 25 : 463-475.
Hughes, F. M. R. 1994. Environmental change, disturbance and regeneration in semi-arid floodplain forests. Pages 321-345 in A. Millington and K. Pye, editors. Environmental Change in Drylands: Biogeographical and Geomorphological Perspectives. John Wiley & Sons, West Sussex, UK.
Hughes, F. M. R., W. M. Adams, E. Muller, C. Nilsson, K. S. Richards, N. Barsoum, H. Decamps, R. Foussadier, J. Girel, H. Guilloy, A. Hayes, M. Johansson, L. Lambs, G. Pautou, J.-L. Peiry, M. Perrow, F. Vautier, and M. Winfield. 2001. The importance of different scale processes for the restoration of floodplain woodlands. Regulated Rivers: Research and Management 17 : 325-345.
Hughes, F. M. R., A. Colston, and J. O. Mountford. 2005. Restoring riparian ecosystems: the challenge of accommodating variability and designing restoration trajectories. Ecology and Society 10 : 12.
Hughes, F. M. R., and S. B. Rood. 2001. Floodplains. Pages 105-1221 in A. Warren and J. R. French, editors. Habitat Conservation: Managing the Physical Environment. John Wiley & Sons, New York, NY.
133
Hughes, F. M. R., and S. B. Rood. 2003. Allocation of river flows for restoration of floodplain forest ecosystems: a review of approaches and their applicability in Europe. Environmental Management 32 : 12-33.
Hupp, C. R., and W. R. Osterkamp. 1996. Riparian vegetation and fluvial geomorphic processes. Geomorphology 14 : 277-295.
Huston M. A. 1994. Biological Diversity: the coexistence of species on changing landscapes. Cambridge University Press, Cambridge, UK.
Johnson, R. R. 1991. Historic changes in vegetation along the Colorado River in the Grand Canyon. Pages 178-205 in G. R. Marzolf, editor. Colorado River Ecology and Dam Management. National Academy Press, Washington D.C.
Johnson, W. C. 2002. Riparian vegetation diversity along regulated rivers: contribution of novel and relict habitats. Freshwater Biology 47 : 749-759.
Katz, G. L., J. M. Friedman, and S.W. Beatty. 2005. Delayed effects of flood control on a flood-dependent riparian forest. Ecological Applications 15 : 1019-1035.
Kondolf, G. M., A. J. Boulton, S. O’Daniel, G. C. Poole, F. J. Rahel, E. H. Stanley, E. Wohl, A. Bång, J. Carlstrom, C. Cristoni, H. Huber, S. Kolijonen, P. Louhi, and Landres, P. B., P. Morgan, and F. J. Swanson. 1999. Overview of the use of natural variability concepts in managing ecological systems. Ecological Applications 9: 1179-1188.
Lawler, J. 2009. Climate change adaptation strategies for resource management and conservation planning. Year In Ecology and Conservation 1162 :79-98.
Mahoney, J. M. and S. B. Rood. 1998. Streamflow requirements for cottonwood seedling recruitment- an integrative model. Wetlands 18 : 634-645. Merritt, D. M., M. L. Scott, N. L. Poff, G. T. Auble, and D. A. Lytle. 2009. Theory, methods and tools for determining environmental flows for riparian vegetation: riparian vegetation-flow response guilds. Freshwater Biology, doi:10.1111/j.1365- 2427.2009.02206.x.
Meyer, J. L., M. J. Sale, P. J. Mulholland, and N. L. Poff. 1999. Impacts of climate change on aquatic ecosystem functioning and health. Journal of the American Water Resources Association 35 : 1373-1386.
134
Miller, J. R., K. House, D. Germanoski, R. J. Tausch, and J. C. Chambers. 2004. Fluvial geomorphic responses to Holocene climate change. Pages 49-77 In J. C. Chambers and J. A. MacMahon, editors. Great Basin Riparian Ecosystems. Island Press, Washington D.C.
Mortenson, S. G., P. J. Weisberg, B. E. Ralston. 2008. Do beavers promote the invasion of non-native Tamarix in the Grand Canyon riparian zone? Wetlands 28 : 666-675.
Mortenson, S. G., and P. J. Weisberg. 2010. Does river regulation increase dominance of invasive woody species in riparian landscapes? Global Ecology and Biogeography, accepted.
Mote, P. W., A. F. Hamlet, M. P. Clark, and D. P. Lettenmaier. 2005. Declining mountain snowpack in western North America. American Meteorological Society, January.
Nakamura, K. 2006. Process-based ecological river restoration: visualizing three- dimensional connectivity and dynamic vectors to recover lost linkages. Ecology and Society 11 : 5.
Nijssen, B., G. O'Donnell, A. Hamlet, and D. Lettenmaier. 2001. Hydrologic sensitivity of global rivers to climate change. Climatic Change 50 :143-175.
Nilsson, C. and K. Berggren. 2000. Alterations of riparian ecosystems caused by river regulation. Bioscience 50 : 783-792.
Palmer, M. A., C. A. Reidy Liermann, C. Nilsson, M. Flörke, J. Alcamo, P. S. Lake, and N. Bond. 2008. Climate change and the world's river basins: anticipating management options. Frontiers in Ecology and the Environment 6: 81-89.
Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks, and J. C. Stromberg. 1997. The natural flow regime: a paradigm for river conservation and restoration. BioScience, 47 : 769-784.
Richardson, D. M., P. M. Holmes, K. J. Esler, S. M. Galatowitsch, J. C. Stromberg, S. P. Kirkman, P. Pyšek, and R. J. Hobbs. 2007. Riparian vegetation: degradation, alien plant invasions, and restoration prospects. Diversity and Distributions 13 : 126- 139.
Richter, B. D., J. V. Baumgartner, D. P. Braun, and J. Powell. 1998. A spatial assessment of hydrologic alteration within a river network. Regulated Rivers: Research and Management 14 : 329-340.
135
Richter, B. D., and G. A. Thomas. 2007. Restoring environmental flows by modifying dam operations. Ecology and Society 12 : 12.
Roberts, J. 2002. Species-level knowledge of riverine and riparian plants: a constraint for determining flow requirements in the future. Australian Journal of Water Resources 5: 21-31.
Smith, E. P. 2002. BACI design. Pages 141-148 in A. H. El-Shaarawi and W. W. Piegorsch (editors). Encyclopedia of Environmetrics. John Wiley & Sons, Chichester, UK.
Stanford, J. A., J. V. Ward, W. J. Liss, C. A. Frissell, R. N. Williams, J. A. Lichatowich, and C. C. Coutant. 1996. A general protocol for restoration of regulated rivers. Regulated Rivers: Research and Management 12 : 391-413.
Steiger, J., E. Tabacchi, S. Dufour, D. Corenblit, and J.L. Peiry. 2005. Hydrogeomorphic processes affecting riparian habitat within alluvial channel-floodplain river systems: a review for the temperate zone. River Research and Applications 21 : 719-737.
Stevens, L. E. 1989. Mechanisms of riparian plant community organization and succession in the Grand Canyon, Arizona. PhD Dissertation, Northern Arizona University, Flagstaff, AZ.
Stevens L. E., and B. D. Gold. 2003. Monitoring for adaptive management of the Colorado River ecosystem in Glen and Grand Canyons. Pages 101-134 in D. E. Busch and J. C. Trexler, editors. Monitoring Ecosystems: Interdisciplinary Approaches for Evaluating Ecoregional Initiatives. Island Press, Washington DC.
Stokes, K. E., and S. A. Cunningham. 2006. Predictors of recruitment for willows invading riparian environments in south-east Australia: implications for weed management. Journal of Applied Ecology 43 : 909-921.
Stromberg, J. C. 2001. Restoration of riparian vegetation in the south-western United States: importance of flow regimes and fluvial dynamism. Journal of Arid Environments 49 : 17-34. Stromberg, J. C., and M. K. Chew. 2002. Flood pulses and restoration of riparian vegetation in the American Southwest. Pages 11-49 in B. A. Middleton, editor. Flood Pulsing in Wetlands: Restoring the Natural Hydrological Balance. John Wiley & Sons, New York, NY.
Suding, K. N., K. L. Gross, and G. R. Houseman. 2004. Alternative states and positive feedbacks in restoration ecology. Trends in Ecology & Evolution 19 : 46-53.
136
Sweeney B. W., S. J. Czapka, and T. Yerkes. 2002. Riparian forest restoration: increasing success by reducing plant competition and herbivory. Restoration Ecology 10 : 392–400.
Tausch, R. J., C. L. Nowak, and S. A. Mensing. 2004. Climate change and associated vegetation dynamics during the Holocene: the paleoecological record. Pages 24- 48 In J. C. Chambers and J. A. MacMahon, editors. Great Basin Riparian Ecosystems. Island Press, Washington D.C.
Taylor, J. P., D. B. Wester, and L. M. Smith. 1999. Soil disturbance, flood management, and riparian woody plant establishment in the Rio Grande floodplain. Wetlands 19 : 372-382.
Taylor, J. P., L. M. Smith, and D. A. Haukos. 2006. Evaluation of woody plant restoration in the middle Rio Grande: ten years after. Wetlands 26 : 1151-1160.
Van Wilgen, B. W., R. M. Cowling, and D. C. Le Maitre. 1998. Ecosystem services, efficiency, sustainability and equity: South Africa’s Working for Water programme. Trends in Ecology & Evolution 13 : 378.
Walters, C. 1997. Challenges in adaptive management of riparian and coastal ecosystems. Conservation Ecology 1:1.
Ward, J. V., K. Tockner, U. Uehlinger, and F. Malard. 2001. Understanding natural patterns and processes in river corridors as the basis for effective river restoration. Regulated Rivers: Research and Management 17 : 311-323.
Whisenant, S. G. 1999. Repairing Damaged Wildlands. Cambridge University Press, Cambridge, UK.
White, J., and J. C. Stromberg. 2009. Resilience, restoration, and riparian ecosystems: case study of a dryland, Urban River. Restoration Ecology 17 : 1-11.
Wohl, E., P. L. Angermeier, B. Bledsoe, G. M. Kondolf, L. MacDonnell, D. M. Merritt, M. A. Palmer, N. L. Poff, and D. Tarboton. 2005. River restoration. Water Resources Research 41 : W10301.
Wolf, E. C., D. J. Cooper, and N. T. Hobbs. 2007. Hydrologic regime and herbivory stabilize an alternative state in Yellowstone National Park. Ecological Applications 17 : 1572-1587.
137
TABLES
Table 1. Seven guidelines for riparian restoration.
1) Use the natural flow regime concept as a guide to understand how the ecosystem functions. Do not attempt to emulate the natural flow regime without considering effects of climate change, non-native invasive species, and prior management strategies on ecosystem structure and function. 2) Reduced flood frequency and magnitude associated with river regulation will likely increase fire occurrence in riparian corridors. Managers must consider potential consequences of fire for riparian ecosystems. In addition, fires may be used in congruence with small floods to trigger establishment of riparian trees. 3) Functional groups defined by phenology, life history traits, and tolerance to disturbance and stress are useful for predicting community response to flow regime scenarios. Artificial groupings such as “native” and “non-native” are counterproductive. 4) Incorporate knowledge of the entire plant life cycle including the establishment niche, patterns of growth, persistence, and mortality into management plans. Successional processes may need to be managed through mechanical removal or planting. 5) Manage at the ecosystem level and consider species interactions. Do not manage for singe species or single taxa. This can result in unpredictable consequences. 6) In most cases site- and watershed-scale restoration approaches are necessary to achieve restoration goals. 7) When designing flow prescriptions, consider the long-term flow regime, including frequency and predictability, not just potential effects of one flood treatment. Similarly, restoration projects should encompass longer time scales (50 years or more) to align with the temporal scale of patch dynamics in which the interaction between geomorphology and biotic communities are strongest. In this way, restoration plans can encompass climate change predictions.
Table 2. Framework for grouping plant species according to life history characteristics and tolerance to floods, shade, and drought for shrubs and trees of the southwestern US. Functional groups are indicated in the second column. Tolerance to floods includes tolerance to scour, burial, and inundation. See Mortenson et al. (2009) for detailed information used to determine characteristics.
Drought Species Functional group Clonal Flood tolerant Shade tolerant tolerant Salix gooddingii Drought intolerant X Populus fremontii Drought intolerant X Salix exigua Clonal X X Pluchea sericea Clonal X X X Acer negundo Shade tolerant X X X Elaeagnus angustifolia Shade tolerant X X Baccharis sarothroides Flood intolerant X Prosopis glandulosa Flood intolerant X Tamarix ramosissima Drought tolerant X X Ulmus pumila Drought tolerant X X
138
139
FIGURE
Figure 1. Conceptual model for riparian restoration with an emphasis on riparian plants. The numbers indicate relationships of components of the conceptual model with the guidelines for river restoration (see Table 1).
Single Disturbance Event Plant Life Cycle Disturbance Regime
Mycorrhizae Establishment Flood regulation Climate Flood 4) Growth 1) Alternative flow regime Magnitude Herbivores Timing Duration Mortality 5) Disturbance Rate of change up nal gro Frequency Functio Parasites, 7) + pathogens Predictability 3) 2) Fire Plant Community Alternative fire regime Succession Magnitude 4) 6) Timing Patch dynamics
Manage Event Manage Sites and Species Manage Process
1 YR. TEMPORAL SCALE 100 YR.
140
APPENDIX A: APPENDIX FOR CHAPTER 1
Table 1. Average soil characteristics in hydroriparian zone of Green and Colorado Rivers in Canyonlands National Park and the Colorado River in Grand Canyon National Park. Asterices indicate statistical differences (p ≤ 0.05). Statistical differences are based on Welch-modified t-tests which are appropriate for samples with unequal variance. The sample sizes are parenthetically noted.
-1 River NH 4-N (µg/g) NO 3-N (µg/g) pH EC (dSm ) Fines (%) Grand Canyon 0.31±0.4 1.35±5.8 8.01±0.5* 4.77±7.9* 14.07±12.8* (136) Canyonlands 0.40±0.4 1.31±3.2 8.94±0.3 2.24±1.5 20.73±17.5 (90)
Table 2. Relative importance of each soil variable or sum of AIC weights ( wi) across models for presence / absence of S. exigua and Tamarix seedlings in Canyonlands and Grand Canyon. The direction of effect is also given.
GRAND CANYON (n=136) Variable Tamarix wi Salix exigua wi EC - 0.519 - 0.301 nitrate - 0.411 - 0.641 pH + 0.244 + 0.302 Fines - 0.209 - 0.186 ammonium + 0.145 - 0.165
CANYONLANDS (n=90) Variable Tamarix wi Salix exigua wi nitrate + 0.807 - 0.579 ammonium - 0.456 - 0.137 EC - 0.343 - 0.142 Fines - 0.099 + 0.171 pH + 0.086 - 0.730
Table 3. Years Tamarix establishment was documented in each geomorphic reach. UG, MG, and LG = upper, mid, and lower Granite. The number of sites sampled in each reach is noted parenthetically. Reaches are in upstream to downstream order. Supai (3) Redwall (3) Marble (4) Furnace (4) UG (4) Aisles (4) MG (4) Muav (2) Canyon (5) LG (2) 1966 X 1968 X X 1969 X X 1976 X 1980 X X 1982 X 1983 X X X X 1984 X X X X X X X 1985 X X X X X X 1986 X X X X X X 1987 X X 1988 X X X 1989 X 1990 X X X 1992 X X X 1993 X X X X 1994 X X X 1995 X X 1996 X 1997 X X X 1998 X X 1999 X X X X X X 2000 X X X X X X X 2001 X X 2002 X X X X 2004 X X 2005 X X X X 2006 X X X 141
Table 4. Life history traits of common riparian shrubs and trees of the southwestern US. Wetland indicator status is in bold. Citations are noted parenthetically. References cited follow:
(1) USDA, NRCS. 2009. The PLANTS Database (http://plants.usda.gov, 2 November 2009). National Plant Data Center, Baton Rouge, LA. (2) Stevens, L.E. & G. Siemion, in prep (3) USDA, Rocky Mountain Research Station. 2009. Forest Service Fire Effects Information System (http://www.fs.fed.us/database/feis/, 11 October 2009). Fire Sciences Laboratory, Missoula, MT. (4) Stevens, L.E. & G.L. Waring. 1986. The effects of post-dam flooding on riparian substrates, vegetation, and invertebrate populations in the Colorado River corridor in Grand Canyon, Arizona. National Technical Information Service Report, US Department of Commerce, Springfield, VA. Available online at: http://www.gcmrc.gov. (5) Katz, G.L. & P.B. Shafroth. 2003. Biology, ecology, and management of Elaeagnus angustifolia . Wetlands 23: 763-777. (6) Cleverly, J. R., S.D. Smith, A. Sala, & D.A. Devitt. 1997. Invasive capacity of Tamarix ramosissima in a Mojave Desert floodplain: the role of drought. Oecologia 111: 12-18. (7) Glenn, E.P. & P.L. Nagler. 2005. Comparative ecophysiology of Tamarix ramosissima and native trees in western U.S. riparian zones. Journal of Arid Environments 61: 419-446. (8) Levine, C.M. & J.C. Stromberg. 2001. Effects of flooding on native and exotic plant seedlings: implications for restoring south-western riparian forests by manipulating water and sediment flows. Journal of Arid Environments 49: 111-131. (9) Turner, R.M. & M.M. Karpiscak. 1980. Recent vegetation changes along the Colorado River between Glen Canyon Dam and Lake Mead, Arizona. Geological Survey Professional Paper 1132. U.S. Dept. of Interior, Washington D.C. (10)Stevens, L.E. & G.L. Waring. 1985. The effects of prolonged flooding on the riparian plant community in Grand Canyon. USDA General Technical Report RM-120. USDA. Tuscon, AZ. Available online at: http://www.gcmrc.gov. (11)Friedman, J.M. & G.T. Auble. 1999. Mortality of riparian box elder from sediment mobilization and extended inundation. Regulated Rivers: Research & Management 15: 463-476. (12)Stromberg, J.C. 1993. Riparian mesquite forests: a review of their ecology, threats, and recovery potential." Journal of the Arizona-Nevada Academy of Science 27: 111-124.
142
Conditions Needed for Establishment Vegetative reproduction Bare surface Seed dispersal timing Seed persistence Clonal Resprouter? Obligate wetland (1) Salix exigua yes Spring - Summer (1) no (1) yes (1,3) yes (1) Salix gooddingii yes (3) Spring (1) no (1) no (1) yes (1) Facultative wetland (1) Baccharis emoryi yes Summer - Fall (9) no no yes Baccharis salicifolia yes Summer - Fall (1) no (1) no (1) yes (1) Pluchea sericea yes Summer no yes yes Populus fremontii yes Spring (1,3) no (1,3) no (1) yes (1) Tamarix ramosissima yes (3) Spring - Fall (2) no (3) no yes (7) Acer negundo no (3) Summer - Fall (1) no (1,3) no (1) yes (1,3) Celtis laevigata no Summer - Fall (1) yes (1) no (1) yes (1) Elaeagnus angustifolia no (3, 5) Summer - Winter (1,5) yes (1,5) no yes (1) Facultative (1) Baccharis sarothroides Spring - Summer (1) yes (1) no (1) no (1) Prosopis glandulosa no (12) Summer - Winter (12) yes (12) no yes (3) Persistence with disturbance Persistence without disturbance Tolerance to: Scour Burial Inundation Drought Shade Salix exigua genet:high (3,4) high (3,10) high (1) moderate (1) moderate (1) Salix gooddingii high (4) moderate (8) high (1,4) moderate (1) intolerant (1) Baccharis emoryi low (9) moderate low (4) low (9) intolerant Baccharis salicifolia low (1,4) moderate (8) low (1) low (1) intolerant (1) Pluchea sericea genet:moderate (4) moderate moderate (1,4) high (6) intolerant Populus fremontii high high (8) moderate (1) moderate (1) intolerant (1,3) Tamarix ramosissima moderate (4) moderate-high (8,10) moderate (4) high (6) intolerant Acer negundo high moderate (11) moderate (1) high (1) tolerant (1) Celtis laevigata moderate (1) low (1) tolerant (1) Elaeagnus angustifolia moderate moderate low (1) moderate - high (1,5) moderate (3) Baccharis sarothroides moderate (4) low (10) low (1) high (1) intolerant (1) Prosopis glandulosa low low low (12) high (3) intolerant (3) 143
144
APPENDIX B: APPENDIX FOR CHAPTER 3
Table 1. Species sampled in river survey with abbreviations and native status. Scientific Name Abbreviation Common Name Native status Acacia greggii ACAGRE catclaw acacia native Acer negundo ACENEG box elder native Ailanthus altissima AILALT tree of heaven non-native Alhagi maurorum ALHMAU camelthorn non-native Amorpha fruticosa AMOFRU desert false indigo native Artemisia sp. ARTSP sagebrush native Atriplex sp. ATRSP saltbush native Baccharis emoryi BACEMO seep-willow native Baccharis salicifolia BACSAL baccharis native Baccharis sarothroides BACSAR desert broom native Baccharis sp. BACSP baccharis native Brickellia longifolia BRILON brickelbush native Chrysothamnus nauseosus CHRNAU rabbitbrush native Elaeagnus angustifolia ELAANG Russian olive non-native Flaveria macdougallii FLAMAC yellowtops native Fraxinus sp. FRASP ash native Fraxinus velutina FRAVEL velvet ash native Gutierrezia sarothrae GUTSAR broom snakeweed native Isocoma acradenia ISOACR goldenbush native Juglans major JUGMAJ Arizona walnut native Juniperus sp. JUNSP juniper native Lycium andersonii LYCAND wolfberry native Morus alba MORALB white mulberry non-native Parthenocissus quinquefolia PARQUI Virginia creeper native Pluchea sericea PLUSER arrowweed native Populus angustifolia POPANG narrowleaf cottonwood native Populus deltoides wislizeni POPDEL Rio Grande cottonwood native Populus fremontii POPFRE Fremont cottonwood native Populus tremuloides POPTRE trembling aspen native Prosopis glandulosa PROGLA honey mesquite native Prosopis pubescens PROPUB screwbean mesquite native Rhus trilobata RHUTRI skunkbush sumac native Ribes sp. RIBSP current or gooseberry native Robinia neomexicana ROBNEO New Mexico locust native Rosa sp. ROSSP rose native Salix exigua SALEXI coyote willow native Salix gooddingii SALGOO Goodding's willow native Salix laevigata SALLAE red willow native Salix sp. SALSP willow native Sarcobatus vermiculatus SARVER greasewood native
145
Table 1 continued.
Shepherdia argentea SHEARG buffaloberry native Tamarix spp. TAMSP tamarisk or saltcedar non-native Toxicodendron rydbergii TOXRYD western poison ivy native Ulmus pumila ULMPUM Siberian elm non-native
Table 2. Constancy of focal taxa (percent of point bars and islands in which taxa occurred) among river segments.
Total River non-native Tamarix Elaeagnus Populus Salix exigua
Carson 89 67 89 89 56 upper CO 100 100 40 60 90 middle CO 100 100 10 40 100 lower CO 100 100 0 0 86 upper Green 100 100 80 80 10 lower Green 100 100 100 70 70 Gunnison 100 100 86 86 93 Humboldt 60 10 60 0 100 upper Rio 100 100 100 100 70 middle Rio 100 93 92 100 100 lower Rio 100 100 18 0 0 Salt 100 100 0 50 0 San Juan 100 100 100 33 56 San Pedro 33 33 0 100 0 upper Truckee 30 10 0 90 50 lower Truckee 100 0 100 100 100 Verde 100 100 0 50 0 Virgin 100 100 0 22 89 Walker 29 14 14 29 86 White 100 100 100 100 40
Table 3. Relative area-weighted cover (%) of all species sampled along twenty river segments. See table 1 for abbreviations.
ACAGRE ACENEG AILALT ALHMAU AMOFRU ARTSP ATRSP Carson 0 0 0 0 0 0.37 0 upper CO 0 10.57 0 0 0 2.18 0 middle CO 0 0 0 0 0 0 0.09 lower CO 1.3 0 0 0.66 0 0 trace upper Green 0 1.12 0 0 0 1.82 0 lower Green 0 0 0 0 0 0 0 Gunnison 0 0 0 0 0 0.55 0 Humboldt 0 0 0 0 0 0 0 upper Rio 0 0 1.02 0 0 0 0 middle Rio 0 0 0 0 0.25 0 0 lower Rio 0 0 0 0 0 0 0 Salt 0 0 0 0 0 trace 0 San Juan 0 0 0 0 0 0 0 San Pedro 0 0 0 0 0 0 0 upper Truckee 0 0 0 0 0 0.32 0 lower Truckee 0 0 0 0 0 10 0 Verde 0 0 0 0 0 0 0 Virgin 0 0 0 0 0 0 0 Walker 0 0 0 0 0 0 0 White 0 0 0 0 0 0 0
146
Table 3 continued.
BACEMO BACSAL BACSAR BACSP BRILON CHRNAU ELAANG FLAMAC Carson 0 0 0 0 0 0 0.37 0 upper CO 0 0 0 3.69 0 0 1.19 0 middle CO 0 0.28 0 0 0 0 0.16 0 lower CO 15.01 0 0 0 0.51 0 0 0.17 upper Green 0 0 0 0.02 0 0 10.14 0 lower Green 0 0.4 0 0 0 0 23.91 0 Gunnison 0 0 0 3.57 0 0 7.02 0 Humboldt 0 0 0 0 0 0.52 1.53 0 upper Rio 0 0 0 0 0 0 24.85 0 middle Rio 0 0 0 trace 0 0 17.56 0 lower Rio 0 9.5 0 0 0 0 0.03 0 Salt 0 3.06 0.19 0 0 0 0 0 San Juan 0 0 0 0 0 0 39.02 0 San Pedro 0 5.4 2.6 0 0 0 0 0 upper Truckee 0 0 0 0 0 0 0 0 lower Truckee 0 0 0 0 0 1.57 8.68 0 Verde 0 17.28 0 0 0 0 0 0 Virgin 0 0.05 0 0 0 0 0 0 Walker 0 0 0 0 0 0 0.78 0 White 0 0 0 0.94 0 0 37.79 0
147
Table 3 continued.
FRAVEL FRASP GUTSAR ISOACR JUGMAJ JUNSP LYCAND MORALB Carson 0 0 0 0 0 0 0 0 upper CO 0 0 0 0 0 0 0 0 middle CO 0 0 0 0 0 0.09 0 0 lower CO 0 0 0.06 0.09 0 0 0 0 upper Green 0 0 0 0 0 0 0 0 lower Green 0 0 0 0 0 0 0 0 Gunnison 0 0 0 0 0 trace 0 0 Humboldt 0 0 0 0 0 0 0 0 upper Rio 0 0 0 0 0 0.16 0 0 middle Rio 0 0 0 0 0 0 0 0.36 lower Rio 0.04 0 0 0 0 0 trace 0 Salt 0 0 0 0 0 0 0 0 San Juan 0 0 0 0 0 0 0 0 San Pedro 0 0 0 0 0.01 0 0 0 upper Truckee 0 0.39 0 0 0 0 0 0 lower Truckee 0 0 0 0 0 0 0 0 Verde 0 0 0 0 0 0 0 0 Virgin 0 0 0 0 0 0 0 0 Walker 0 0 0 0 0 0 0 0 White 0 0 0 0 0 0 0 0
148
Table 3 continued.
PARQUI PLUSER POPANG POPFRE POPDEL POPTRE PROGLA PROPUB Carson 0 0 0 73.53 0 0 0 0 upper CO 0 0 0 0 20.68 0 0 0 middle CO 0 0 0 2.63 0 0 0 0 lower CO 0 18.62 0 0 0 0 23.6 0 upper Green 0 0 0 19.23 0 0 0 0 lower Green 0 0 0 15.28 0 0 0 0 Gunnison 0 0 0 0 25 0.99 0 0 Humboldt 0 0 0 0 0 0 0 0 upper Rio 0.02 0 0 0 48.48 0 0 0 middle Rio 0 0 0 0 31.82 0 0 0 lower Rio 0 1.25 0 0 0 0.53 4.74 9.04 Salt 0 0 0 21.05 0 0 13.49 0 San Juan 0 0 0 0 10.92 0 0 0 San Pedro 0 0 0 73.36 0 0 0.1 0 upper Truckee 0 0 0 80.44 0 0 0 0 lower Truckee 0 0 0 52.85 0 0 0 0 Verde 0 0 0.33 10.33 0 0 0.1 0 Virgin 0 6.41 0 0.42 0 0 1.27 0 Walker 0 0 0 4.82 0 0 0 0 White 0 0 0 37.34 0 0 0 0
149
Table 3 continued.
RHUTRI RIBSP ROBNEO ROSSP SALEXI SALGOO SALLAE SALSP Carson 0 0 0 0 7.47 0 0 5.94 upper CO 6.02 0 0.32 0 19.65 0 0 1.56 middle CO 3.29 0 0 0 26.11 0 0 0 lower CO 0 0 0 0 3.53 3 0 0 upper Green 0.8 0 0 0 1.82 0 0 0 lower Green 1.57 0 0 0 11.06 0 0 0.04 Gunnison 6.54 0 0 0 14.52 0 0 0.05 Humboldt 0 0.45 0 trace 97.37 0 0 0 upper Rio 0 0 0 0 7.18 0.03 0 0 middle Rio 0 0 0 0 39.12 trace 0 0 lower Rio 0 0 0 0 0 0 0 0.10 Salt 0 0 0 0 0 13.56 0 0 San Juan 0 0 0 0 2.21 0 0 0 San Pedro 0 0 0 0 0 18.07 0 0 upper Truckee 0 0 0 0.24 7 0 0.59 9.16 lower Truckee 0 0 0 0 26.97 0 0 trace Verde 0 0 0 0 0 56.83 0 0 Virgin 0 0 0 0 3.18 1.34 0 0 Walker 0 1.54 0 1.33 45.31 0 0 0.75 White 0 0 0 0 3.23 0 0 0
150
Table 3 continued.
SARVER SHEARG TAMSP TOXRYD ULMPUM Carson 0 1.24 7.72 0 0 upper CO 0.14 0 33.46 0 1.22 middle CO 4.2 0 63.13 0 0.02 lower CO 0 0 33.73 0 0 upper Green 0 0 64.1 0 0.15 lower Green 7.79 0 38.87 0 0.2 Gunnison 0 0 41.09 0.18 0.53 Humboldt 0 0.06 0.08 0 0 upper Rio 0 0 16.54 0 0.45 middle Rio 0 0 7.21 0 3.66 lower Rio 0 0 74.74 0 0 Salt 0 0 49.18 0 0 San Juan 0 0 47.67 0 0.47 San Pedro 0 0 0.35 0 0 upper Truckee 0 0.43 trace 0 1.43 lower Truckee 0 0 0 0 0 Verde 0 0 14.68 0 0 Virgin 0 trace 88.63 0 0 Walker 0 45.44 0.02 0 0 White 0 0 18.84 0 0
151
Table 4. Hydrologic indicators for all un-dammed and dammed river sections from the entire period of gage records.
River HPC Constancy Aug. Min BF DateMax FR HPD DateMin May Cont. Max Un-dammed: Carson 4 0.17 0.1 0.0 0.0 138 15 7 273 28.0 0.23 49.4 middle CO 3 0.42 95.5 52.6 0.3 151 200 10 237 62.0 0.16 775.6 Humboldt 2 0.12 0.6 0.3 0.0 154 8 25 258 20.0 0.22 45.9 San Pedro 9 0.23 1.1 0.1 0.1 222 1 4 181 0.2 0.15 15.6 Virgin 10 0.38 2.2 1.6 0.3 224 9 3 194 4.9 0.14 21.3 Walker 3 0.30 4.5 1.0 0.2 155 6 9 273 9.0 0.15 20.5 White 5 0.46 11.0 6.6 0.3 154 25 3 342 40.3 0.15 73.2
Dammed: upper CO 2 0.46 68.6 36.1 0.4 156 95 31 371 203.8 0.20 454.7 lower CO 4 0.42 383.6 123.9 0.3 158 500 4 338 504.0 0.09 989.5 upper Green 3 0.35 58.8 31.5 0.3 147 120 15 305 276.4 0.16 454.7 lower Green 3 0.36 88.3 32.9 0.2 152 190 10 347 378.0 0.17 716.8 Gunnison 3 0.38 29.1 17.2 0.3 144 80 6 235 177.8 0.13 284.2 upper Rio 5 0.36 22.4 7.6 0.2 144 54 4 260 78.1 0.10 121.8 middle Rio 6 0.24 13.2 1.2 0.0 146 50 3 272 72.2 0.13 108.3 lower Rio 4 0.36 38.9 0.1 0.0 185 10 12 304 45.1 0.14 61.7 Salt 6 0.31 36.4 0.1 0.0 203 40 4 356 32.5 0.16 54.8 San Juan 5 0.31 22.1 8.7 0.2 159 70 3 235 123.8 0.11 203.8 upper Truckee 5 0.34 8.1 6.5 0.3 128 23 5 239 37.9 0.09 61.1 lower Truckee 4 0.14 0.7 0.2 0.0 131 16 6 280 18.9 0.09 40.0 Verde 8 0.22 8.4 1.2 0.1 72 24 5 189 5.8 0.06 70.8
152
Table 5. Hydrologic indicators for dammed river segments prior to dam construction and un-dammed river segments prior to 1962. Only dammed segments with at least 20 years of gage records prior to dam construction are included.
HPC Constancy Aug. Min BF DateMax FR HPD DateMin May Cont. Max MAF Un-dammed: Carson 4 0.17 0.0 0.0 0.0 139 16 9 275 28.6 0.27 48.2 349.6 Humboldt 2 0.15 0.8 0.4 0.1 151 8 47 256 20.1 0.24 45.5 335.8 middle CO 3 0.40 98.0 49.0 0.2 150 245 7 253 679.0 0.25 1208.5 8118.0 San Pedro 8 0.27 2.4 0.1 0.1 220 2 4 178 0.2 0.17 21.5 62.0 Virgin 12 0.40 2.0 1.7 0.3 225 9 2 193 4.2 0.18 23.4 232.5 Walker 4 0.29 4.1 0.9 0.2 159 8 12 377 9.4 0.19 23.5 187.7 White 6 0.47 12.3 7.1 0.4 161 28 3 349 42.8 0.16 72.2 716.4
Dammed: Gunnison 4 0.39 23.1 13.8 0.2 144 73 6 236 191.4 0.21 355.3 2446.0 lower CO 3 0.37 261.4 95.4 0.2 155 383 9 254 1162.0 0.21 1927.2 16430 lower Green 3 0.34 78.7 29.7 0.2 153 165 9 349 408.8 0.23 749.8 5932.0 middle Rio 6 0.19 7.3 0.2 0.0 149 50 2 268 50.4 0.17 106.7 993.6 San Juan 5 0.26 22.1 5.0 0.1 155 84 5 249 138.6 0.21 247.4 2272.0 upper Rio 4 0.33 18.2 5.0 0.2 138 54 3 254 58.8 0.14 107.5 1240.0 Verde 7 0.34 9.9 2.8 0.1 54 20 5 189 5.1 0.18 264.3 829.4
153
Table 6. Hydrologic indicators for dammed river segments following dam construction and un-dammed river segments after 1962. Only dammed segments with at least 20 years of gage records prior to dam construction are included.
HPC Constancy August Min BF DateMax FR HPD DateMin May Cont. Max MAF Un-dammed: Carson 4 0.18 0.1 0.0 0.0 135 14 7 254 26.6 0.22 53.6 395.5 Humboldt 2 0.16 1.0 0.7 0.1 149 10 33 260 21.0 0.22 56.2 458.1 middle CO 2 0.44 82.9 51.3 0.3 156 200 20 232 364.0 0.15 732.8 6723 San Pedro 9 0.23 0.7 0.0 0.0 225 1 3 187 0.2 0.20 12.0 40.5 Virgin 9 0.38 2.0 1.5 0.3 216 9 3 206 5.6 0.14 20.6 237.7 Walker 3 0.30 4.9 1.1 0.2 154 5 12 288 8.9 0.16 14.4 210.8 White 4 0.46 11.0 6.3 0.3 146 23 4 345 37.4 0.15 76.9 681.4 Dammed: Gunnison 7 0.44 40.3 25.5 0.4 143 70 5 210 130.6 0.08 180.7 2459 lower CO 8 0.55 445.2 194.2 0.5 180 1148 2 308 378.0 0.04 562.8 14290 lower Green 3 0.42 84.7 50.6 0.3 148 170 9 342 313.6 0.14 586.9 5543 middle Rio 8 0.30 14.5 5.1 0.2 145 50 3 275 87.1 0.14 123.9 1347 San Juan 6 0.34 33.0 10.0 0.2 168 70 3 231 90.4 0.09 144.3 2013 upper Rio 5 0.41 24.8 9.0 0.2 151 50 3 280 79.5 0.10 110.5 1399 Verde 8 0.19 8.5 0.5 0.0 88 27 5 12 6.9 0.10 44.2 501.3
154
Table 7. Eigenvectors for hydrologic indicators from principal components analysis of alteration in flow characteristics.
Axis 1 2 3 4 5 MAF 0.389 0.152 0.157 0.040 -0.017 Max 0.310 0.294 0.196 0.089 -0.340 Min 0.392 -0.011 0.068 -0.028 0.270 Base flow 0.290 -0.272 -0.366 -0.078 -0.060 Constancy 0.293 -0.265 -0.300 -0.064 -0.309 High pulse count -0.037 -0.500 0.332 0.228 0.005 High pulse duration -0.071 0.390 -0.165 -0.038 0.598 Date max 0.035 -0.203 -0.105 0.859 0.192 Date min 0.116 0.141 -0.711 0.117 -0.070 May 0.322 0.277 0.183 0.131 -0.299 August 0.389 0.004 0.117 -0.009 0.256 Contingency -0.136 0.452 0.005 0.395 -0.166 Recession rate 0.368 -0.050 0.080 -0.039 0.360
155