I. Introduction
A. The importance of the long view (space and time) in river restoration
Rivers can be studied in their present state and at a specific location and this will yield only an instantaneous snapshot of the river at a point in space and time. If we are using the river in its present state to define what we think it should look like we may be in for some trouble. Even looking at a 10 year data set may not be sufficient to give us much of an idea of how a natural system functions or what it is responding to.
“ Perhaps the present is too short to be a key to the past or future.” (Schumm, 1991)
River restoration is a growing industry in many parts of the world. People are becoming increasingly aware that river systems are more than just water conveyance mechanisms and that river environments are essential to ecological function and improved environmental conditions. However, if we are thinking about restoration what are we trying to restore a system to?
B. What is fluvial geomorphology and how does it help with the long view?
1. Understanding the trends in system response
Geomorphology is the study of landforms on Earth’s surface. Dynamic equilibrium: the forms within a landscape maintain their character as long as controls on landscape form do not change dramatically. However, what time scale are we talking about? The system is responding to events that have occurred throughout the watershed at different time scales. For example, the system could be responding to the addition of sediment from mining impacts on the decadal scale (James, 1999), but on longer time scales may still be adjusting to the end of the last glacial period (Schumm, 1991). Geomorphologists tend to think of different types of equilibrium relating to different time scales. Static equilibrium represents instantaneous time while steady state equilibrium represents a collection of periods of steady time on the order of 100s to thousands of years
Time is a means of measuring change. It can be viewed as a proxy for the rate of energy expenditure, work done or change in entropy just as drainage basin area can be used to determine discharge from ungaged basins (Schumm, 1991).
2. The uniqueness of an individual system
Geomorphology also provides us with the tools to understand how processes may differ between different landscapes. For example, if a river is flowing through shale it will quickly downcut and erode away any relief in the landscape whereas gneissic bedrock landscape may create a very different channel network, even if the climatic variables are constant between the two. As practitioners of river restoration, we need to understand the natural range of variability within watersheds as we think about their long-term function (Wohl et al., 2005). Regional differences in climate, geology and topography define the system and therefore are essential to consider if we are to pursue a restoration project in the area.
Legacy effects can manifest themselves in physical and biological differences in the system – geomorphology can help us to better constrain past events that may be affecting the current system. Each local system may have differences in legacy effects that may alter the restoration approach at a given site. Understanding the geomorphology of a watershed may aid in untangling these legacy effects.
II. Reading the Landscape (the results of Fluvial Processes at work)
A. Channel Form
Two sets of variables must be considered to understand channel form, those that drive change and those that resist it (Lane, 1955). Resisting variables include sediment size, hydraulic roughness, bank cohesiveness and variables that drive change include discharge (Q), which is the volume of water per unit time, slope (S), sediment supply (Qs), sediment size. Understanding the interaction between these variables is complex because not only is the fluid in the channel constantly deforming and adjusting, but the channel shape itself is adjusting to changes in flow. Neither fluid nor container remains constant. Table 1 summarizes channel response to changes in discharge (Q) and sediment load (Qs). The work conducted by Schumm (1977) continues to be refined as additional research in both natural channels and flumes sheds light on how channels respond to adjustments in discharge and sediment supply (Gaueman et al., 2005; Madej et al., 2009).
Table 1. Table adapted from Schumm (1977) showing adjustments to channel attributes with changes in discharge (Q) and sediment supply(Qs). (+) indicates increase, (-) decrease, (±) could increase or decrease (c) variable remains constant, (na) no response or unknown response
The majority of these variables are responding to changes in stream power Ω=γQS where γ is the specific weight of water, Q is discharge and S is slope. A slight increase or decrease in stream power has the potential to change sinuosity ( = channel length / straight-line valley length) in a stream reach (Schumm, 1977) or impact the rate of bank erosion (Lawler et al., 1999). The type of channel form that will develop in response to a given stream power depends on the forces that are resisting change, in particular, bank material and basal stream material (Simon and Rinaldi, 2006). This concept is important in thinking about restoration, because incision is often one of the reasons that restoration projects are initiated (Zaimes et al., 2006; Florsheim et al., 2008) . A primarily cobble bedded stream will respond to moderate increases in discharges by eroding their banks rather than incising (Simon and Rinaldi, 2006). Large discharge increases, however, may cause cobbles to move and even a cobble bedded stream will respond by incising (Madej et al., 2009). Bank resistance is affected by the percentage of silt versus clay in the banks and the presence of vegetation. Resistant bank material will decrease rates of river migration (Hudson and Kesel, 2000; Brooks et al., 2003), while lack of resistant material will inhibit meander development (Hooke, 2003). Whether a channel responds to an increase in discharge by incising or widening depends on the ratio of resistance between bed and bank material (Simon and Rinaldi, 2006).
B. Landforms
Fluvial response to external forcing is primarily investigated by looking at terraces and terrace sequences (e.g. Bull, 1991) with the understanding that much of the sediment stored in terraces reflects sediment deliveries from tributaries and tributary fans (Meyer et al., 1995). Terraces, at the very least, are landforms that represent periods of aggradation and incision by any number of processes. They can also represent periods of vertical stability which is the case with bedrock strath terraces and fill-strath terraces where formation is primarily by lateral planation (Bull, 1990). These straths are often covered by thin deposits, where deposits that are less than or equal to ~ 2 times the average bank height do not constitute aggradation, but represent a vertically stable stream (Pierce et al., 2011).
Additional features that can aid in interpreting fluvial processes are oxbow lakes and paleochannels. Since rivers devoid of human impact are extremely rare (Wohl, 2006), finding references to compare pre and post-disturbance is difficult. The use of the long term fluvial record (paleochannels) is one way to get around this investigative problem (e.g. Starkel, 1995).
Figure 1. Paleochannels on Hell Roaring fan filled with snow. Photo is taken from the top of Nemesis Mountain, just east of Hell Roaring Canyon, Centannial Valley, MT. The modern channel is on the bottom right in this photo and is flowing north, away from the viewer (photo courtesy of Hayes Buxton, USGS).
III. What controls the Landforms we discussed above?
The landscapes that we see represent a balance between the forces that drive change and those that resist them. Resisting forces include bedrock, regolith, soil and rooted vegetation while driving forces include climate, tectonics and gravity.
Since we have described landscapes above as being in equilibrium than there must be periods when they are in disequilibrium. A shift into disequilibrium occurs when a threshold is crossed. Thresholds may be extrinsic (earthquake) or intrinsic (freeze thaw action over hunderds of years), but once a threshold has been crossed the landscape, or landform, is now in disequilibrium and will respond to a change in conditions. In the case of a river, it may be downcutting in response to glacial climate a threshold was crossed and the river is still trying to reach equilibrium.
The location of a given reach within the channel network whether it be headwaters, midsection or lower basin, may help dictate how a stream should function in a given reach (Schumm, 1977). Energy gradient changes downstream and is directly responding to hillslopes, channel gradient, discharge and sediment supply. This gives reaches distinct disturbance regimes (Montgomery, 1999; Wohl et al., 2005).
A. Climate
The first direct effect of climate in terms of streams is that there is a change in the volume of precipitation reaching the ground. This has a direct effect on rivers in terms of discharge down the channel. The amount of water in the channel directly affects its power to transport sediment and erode the channel bed and banks. With a lower discharge the stream may not be in equilibrium and may loose the ability to transport sediment. This can happen on short time scales (low end of summer flows) or be a more permanent condition as the climate moves toward drier conditions. The intensity and duration of precipitation events can also vary under different climate regimes and have a dramatic effect on channel discharge.
The climate that exists at a locale also has an impact on the vegetation types present. Vegetation can inhibit erosion in areas with consistent moisture and thick vegetation.
1. Changes in discharge
Tree ring records, drought records, fire records tied to what we observe in landforms?
2. Changes in vegetation
a) Riparian corridor (bank stability, nutrient exchange, floodplain development)
b) Hillslopes (potential for landslides, forest fires, snow hydrology)
B. Tectonics
The other issue is that fluvial landforms could also form in response to a change in base level caused by tectonic forces, so that the tectonic signal dominates over any signal resulting from a changing climate. Wegmann and Pazzaglia (2002) specifically address this issue by investigating a stream network in the tectonically active Olympic Mountains of western Washington State. By comparing Holocene and Pleistocene rates of fluvial incision along with Neogene exhumation rates in concert with the type and form of terrace deposits, the authors hoped to get a sense of the dominant process controlling fluvial form and response. The study determined that rock uplift rates have remained more constant than climate through this time period and that rates of incision in the Holocene are much more rapid than during the Pleistocene. Wegmann and Pazzaglia (2002) suggest that the Pleistocene river was over capacity with respect to sediment load and so the channel was well above the strath level and not actively eroding the uplifting bedrock. Once sediment load declined following the glacial period, the river has sought to return to equilibrium and only now is beginning to actively erode the uplifting bedrock. Uplift is part of what is driving incision, but terrace have formed primarily in response to addition of material from hillslopes either from climate induced changes or earthquake events. The lack of convincing evidence of upstream convergence of terraces with a knickpoint adds additional rationale against the tectonic influence on terrace formation in the Holocene.
1. Changes in sediment supply
2. Changes in base level
C. Human impacts (this is a quick look – I will focus on the natural processes in more detail)
1. Dams and diversions
2. Channel Alterations
3. Other channel structures/mining
(James, 1999)
D. Geomorphologist at work… How do we determine why the system looks as it does now?
1. Longer time scales:
Although, each study has its own issues that it is trying to address, fluvial response studies generally begin by mapping terraces based on field relationships (e.g. Wegmann and Pazzaglia, 2002). Field relationships of terraces are usually first determined by elevation above the channel at some given flow (e.g. bankfull) and then more detailed analysis is made of the degree of soil development on the terrace tread, which may indicate the relative stability of that surface, and thus, age (Bull, 1990). Other distinctive characteristics of the sediment, such as grain size and sorting are also considered when making correlations across, or down, the channel. There may also be a distinctive layer, such as a tephra unit, to aid in distinguishing a given terrace level (Meyer et al., 1995). Once these field relationships are established, and a relative chronology developed, absolute dating can be used to interpret terrace units and make it possible to place them in context with other fluvial systems and global events.
2. Shorter time scales
Surveying , xsec analysis of channel form – mapping landforms… where will water go
Maybe stick the biogeomorphic bit in here (Osterkamp and Hupp, 2010).
IV. What are we dealing with now? How can we use what we know about process and the system to aid in solving current issues on river systems?
A. Uses of geomorphology in restoration
1. Practice
Understanding that a landscape has a history (Schumm, 1991) and that it may still be responding to past events. In work done in Nevada it was found that most rivers in the region were downcutting in response to ??? (Germanoski and Miller, 2002). If restoration had proceeded and tried to deal with the downcuttig problem it would have been unsuccessful. Too many times river restoration projects are undertaken without understanding the landscape controls that are at work. Restoration needs to be holistic in time and space.
Use of air photos to aid in historical assessment of function and aid in geomorphic mapping.
2. Problems
Focusing on a fixed endpoint rather than understanding the variability in a system will lead to addressing specific systems rather than working on addressing process linkages that have broken down within a watershed (Wohl et al., 2005)
Holistic watershed level work rather than reach scale – otherwise we are fixing symptoms rather than process based problems since headwater issues are very likely to have ramifications downstream (e.g. sediment inputs from a road building project) (Wohl et al., 2005).