Summary of Instream Flow Requirements for Riparian Forest Maintenance at Judith River and Arrow Creek, Central Montana

Summary of Instream Flow Requirements for Riparian BLM Forest Maintenance at Judith River and Arrow Creek, Central Montana and Upper Missouri River Breaks National Monument Federal Reserved Water Rights Proposal

July 2011 Upper Missouri River National Breaks Monument

U.S. Department of the Interior

Bureau of Land Management Summary of Instream Flow Requirements for Riparian Forest Maintenance at Judith River and Arrow Creek, Central Montana

and Upper Missouri River Breaks National Monument Federal Reserved Water Rights Proposal

July 2011

Chad R. Krause1

BLM/MT/ST-11/009+7250P

Copies may be obtained from:

United States Department of Interior Bureau of Land Management Upper Missouri River Breaks National Monument 920 NE Main Street Lewistown, MT 59457-1160

http://www.blm.gov/mt/st/en/fo/lewistown_field_office.html

1 Hydrologist, U.S. Bureau of Land Management, Upper Missouri River Breaks National Monument

Cover Photo: Judith River TABLE OF CONTENTS

BACKGROUND ...... 1

WATERSHED DESCRIPTIONS AND RESOURCE SETTINGS ...... 3

Judith River ...... 3

Arrow Creek ...... 5

APPROACH AND METHODS ...... 8

ANALYSIS OF FLOW REQUIREMENTS...... 15

PROPOSAL SUMMARY...... 27

Judith River ...... 27

Arrow Creek ...... 28

Proposal Requests Across Both Basins ...... 29

ACKNOWLEDGEMENTS ...... 30

REFERENCES ...... 31

APPENDICES ...... 33

A-1 Judith River Mean Daily Discharge ...... 33

A-2 Arrow Creek Daily Mean Discharge (2007, 2008, 2009) ...... 34

A-3 Arrow Creek and Judith River Digital Terrain Models ...... 35

A-4 Arrow Creek and Judith River Watersheds ...... 37

A-5 Arrow Creek Watershed ...... 38

A-6 Judith River Watershed ...... 39

LIST OF FIGURES

Figure 1. Upper Missouri River Breaks National Monument ...... 1

Figure 2. Judith River Watershed ...... 3

Figure 3. Judith River Valley near PN Ranch ...... 5

Figure 4. Arrow Creek Watershed...... 6

Figure 5. Lower Arrow Creek during Flood Conditions ...... 7

Figure 6. Judith River Study Reach ...... 9

Figure 7. Arrow Creek Study Reach...... 10

Figure 8. Judith River DTM and Valley Bottom Cross Section Cut Lines ...... 11

Figure 9. Judith River at Approximately 2,000 cfs on May 27, 2008 ...... 12

Figure 10. Example of Vegetation Sampling Along Cross Section 5 on Judith River ...... 12

Figure 11. Pole Cottonwood Tree Excavated to the Establishment Surface ...... 13

Figure 12. Judith River and Arrow Creek Piezometer Locations ...... 14

Figure 13. Judith River Historical Photos ...... 16

Figure 14. Full Valley Bottom Flooding on Arrow Creek in June 2011, Approximate Flow Range – 4,000 to 5,000 cfs (Estimate by C. Krause) ...... 17

Figure 15. Channel Change and Channel Abandonment on Judith River in 2011 ...... 17

Figure 16. Judith River DTM with Inundation Depths at 3,000 cfs ...... 18

Figure 17. Inundation Mapping and Relative Water Depths of 3,000 cfs on the Judith River and HEC-geoRAS Model Water Surface for 3,000 cfs Profile at Judith River Cross Section J5 ...... 19

Figure 18. Inundating Discharges of Cross Sections 5 and 6 at Judith River ...... 20

Figure 19. Judith River Cross Section 5 Rating Curve ...... 21

Figure 20. 2008 Species Presence, in Fraction of Bottomland Plots Relative to Inundating Discharge ..... 22

Figure 21. Channel Movement at Judith River Cross Section 5 ...... 23

Figure 22. Daily Mean Discharge (cfs), Real Groundwater Elevations (ft), and Real River Stages at Arrow Creek (A12A) and Judith River (J8C) ...... 24

Figure 23. Judith River Recession Rate Analysis ...... 25

LIST OF TABLES

Table 1. Estimated Largest Annual Peak Flows for Water years 1950-2009 (Sando, written communication, provisional – in review and subject to revision) ...... 15

BACKGROUND

The Upper Missouri River Breaks National Monument (UMRBNM) was established on January 17, 2001 when President Clinton issued a Proclamation under the provisions of the Antiquities Act of 1906. The Monument includes about 375,000 acres of Bureau of Land Management (BLM) land in north central Montana in Blaine, Chouteau, Fergus, and Phillips counties and generally corresponds with the Upper Missouri National Wild and Scenic River (UMNWSR) from Fort Benton to Arrow Creek. Below Arrow Creek, the Monument widens downstream to the Charles M. Russell National Wildlife Refuge. Private and state lands are intermingled with the Monument.

Figure 1. Upper Missouri River Breaks National Monument

The Monument proclamation reserves water in two tributaries to the Missouri River, Arrow Creek and the Judith River. The proclamation states “These tributaries contain outstanding objects of biological interest that are dependent on water, such as a fully functioning cottonwood gallery forest ecosystem that is rare in the Northern Plains. Therefore, there is hereby reserved, as of the date of this proclamation and subject to valid existing rights, a quantity of water in the Judith River and Arrow Creek sufficient to fulfill the purposes for which this monument is established.”

Federal reserved water rights may be created when federal lands are withdrawn from the public domain (e.g. national parks, wildlife refuges, national forests). Federal reserved water rights are different from state appropriated water rights. They: may apply to both instream and out-of-stream water uses; may be created without actual diversion or beneficial use (as defined by State law); are not lost by non-use; have priority dates established as the date the land was withdrawn; and are for the minimum amount of water reasonably necessary to satisfy both existing and foreseeable future uses of water for the primary

1 purposes for which the land is withdrawn (Modified from Alaska Department of Natural Resources, 2000).

Federal reserved water rights must be resolved as part of Montana’s statewide adjudication process. In 1979, the Reserved Water Rights Compact Commission (RWRCC) was created by the Montana legislature to: “conclude compacts for the equitable division and apportionment of waters between the State and its people and the several Indian Tribes claiming reserved water rights within the state (Montana Code Annotated (MCA) 85-2-701) and between the State and its people and the federal government claiming non-Indian reserved waters within the State (MCA 85-2-703).”

A vast majority, but not all, of cottonwood forest growing in the two drainages are located on either private or state land. Although BLM managed lands do account for less than ten percent of the river bottom ownership, riparian forest resources are present on public land as well. We emphasize that BLM has no management authority over private or state lands and minerals. However, the BLM has been given the responsibility under presidential proclamation to protect a flow regime in the Judith River and Arrow Creek that supports the water dependent objects for which the Monument was established. Although the BLM will not pursue the acquisition of private land unless approached by a landowner or their representative, if those lands were to come into title of the United States, they would be withdrawn from the public domain and be reserved as part of the Monument.

Instream flow studies were recommended by the Upper Missouri River Breaks National Monument Record of Decision and Approved Resource Management Plan before beginning negotiations with the RWRCC (BLM, 2008). The BLM needed to know the magnitude, timing, and frequency of flows necessary to support the outstanding water-dependent biological resources, and cottonwood galleries that were the basis for the reserved water rights.

Much of the vegetation field work and retrospective analyses were conducted in collaboration with USGS as part of a larger, long-term collaborative study of relations between streamflow, grazing, and cottonwod riparian vegetation in and around the Upper Missouri Breaks National Monument (Auble and others 2010, 2005; Auble and Scott 1998, Bovee and Scott 2002, Scott and Auble 2002, Scott and others 1997).

This summary has the format of a scientific report and a fairly in depth discussion of study setup, flow modeling, and interpretation of results. However, it is emphatically noted that the purpose and intent of this report is to provide the thought process, rationale, and justification that the BLM followed in development of the UMRBNM federal reserved water right proposal. The science, study setup, and the interpretation of results are intended as documentation of that thought process. Although both Judith River and Arrow Creek are summarized in this report, the analysis, results, and interpretation are often very similar between the two. So, for the purpose of keeping this report simpler and more manageable, often only one or the other is discussed where appropriate.

WATERSHED DESCRIPTIONS AND RESOURCE SETTINGS

Judith River

The Judith River originates in the Little Belt, Big Snowy, and Judith Mountain ranges. Draining significant portions of Fergus and Judith Basin counties, the river flows northward into the Missouri River. The Judith River watershed is approximately 2,762 square miles (1,767,810 acres) and includes the communities of Lewistown, Hobson, Moore, Stanford, and Denton. Numerous other towns and small settlements are also present. Major tributaries include Big Spring Creek, Warm Spring Creek, Wolf Creek, and Ross Fork Creek. While the Judith River and its tributaries do provide water-based recreation, wildlife and fisheries resources, and some municipal needs, it is a valuable source of water for agricultural purposes, and most water rights claims are agriculture related. Ackley Lake is an important storage structure for irrigation within the watershed. East Fork Reservoir and Casino Creek Reservoir were originally built for flood control above the city of Lewistown. There are many water rights, contracts, laws and agreements that dictate the management of flows in the Judith River watershed. Still, the Judith River does maintain the semblance of a natural hydrograph although the base flows are severely over appropriated. Above the confluence of Big Spring Creek and Warm Spring Creek, the Judith is often dewatered during summer months.

Figure 2. Judith River Watershed

Climatic characteristics of the Judith River watershed are continental. Winters are generally cold with average January maximum temperatures ranging from 32 to 34 degrees Fahrenheit (F) at Lewistown,

Denton, and Utica (WRCC, 2011). Between cold waves, there are periods of mild and windy weather known as “chinook” weather (WRCC, 2011). These “chinook” winds often reach speeds of 25 to 50 mph and can last for several days. Summer temperatures are warm, but not hot, through most of the watershed. July average maximum temperatures range from 78 to 83 degrees F. Precipitation is very dependent upon elevation, and annual precipitation varies from 13 to 15 inches in the lower end of the watershed, while the top of the Snowy and Little Belt Mountains receive around 45 inches of precipitation with most falling as snow. Eighty five percent of the watershed receives 22 inches or less of precipitation (Montana NRIS, 2011). Elevations range from 2,400 feet at the mouth of the Judith to 9,127 feet at the highest point of the watershed in the Little Belts. Because of the climatic characteristics of the basin, the Judith River hydrograph is a snow-melt dominated hydrograph with peaks occurring in late May and June and flows receding throughout the remainder of the summer and fall. Prairie snow melt in March may create double peaked hydrographs, and late summer thunderstorms may cause short duration high flow events. One infrequent anomaly in the Judith River hydrograph is that occasionally release of water from ice jams may cause river stages and flows that are greater than the summer high flows.

The UMRBNM boundary encompasses approximately the last 12 ½ river miles of the Judith River. Much of the river bottomland is privately owned, while a good deal of the surrounding hills and badlands are managed by the BLM Central Montana District and State of Montana trust lands. The lower end of the watershed within the UMRBNM is decidedly warmer and drier than much of the basin. Elevations range from 2,400 feet at the mouth of the Judith to 3,500 feet on the canyon rims. Annual precipitation is around 13 to 15 inches, and average maximum temperature in July for Winifred, which is immediately adjacent to the lower end of the Judith watershed, is 86 degrees F (WRCC, 2011).

The reach of Judith River within the UMRBNM boundary flows through a wide valley bottom of Quaternary alluvium (Wilde and Porter, 2001). These alluvial deposits of modern day channels and floodplains are set within a deep canyon of Cretaceous Judith River Formation sandstone and Claggett Shale (Wilde and Porter, 2001). Numerous faults result in exposures of Cretaceous Eagle Formation sandstone. The geomorphology of the landscape today has also been affected by glaciations during the Pleistocene (Wayne et al., 1991). Adjustments to the landscape from a complex series of continental glaciations, pre and postglacial river valleys, glacial lakes, and outwash channels has resulted in features such as stream channels that are under fit to their valley size (i.e. The Big Sag, a side tributary to the Judith River).

The river morphology of the Judith River within the UMRBNM boundary is broadly characterized as a Rosgen C- channel type as described in Rosgen (1996). According to Rosgen (1996), Rosgen C-channel types are often found in broad valleys with terraces, in association with floodplains and alluvial soils. Rosgen C channel types are low gradient, meandering, point-bar with riffle/pool morphology (Rosgen, 1996). A wide range of substrate particle sizes are present from fines in low shear stress zones to gravel and cobble in the active channel.

Bare, moist sites that are safe from future disturbances summarize the conditions required for cottonwood recruitment, and these restrictive requirements are met most frequently by flow-induced

4 channel change (Auble et al., 2005). Because of the Judith River’s stream channel type, which laterally migrates across the valley bottom, and the river’s ability to flood, which has resulted in substantial channel change from 1938 to 2006, broad, multi-aged, cottonwood forests are present on the lower Judith. These cottonwood forests, which are the basis for the reserved water rights, are dominated by plains cottonwood (Populus deltoides). Minor amounts of narrowleaf cottonwood (Populus angustifolia) or narrowleaf/plains cottonwood hybrids are present but represent a very small component of the cottonwood forest. Other riparian tree and shrub species present include peach-leaf willow (Salix amygdaloides), yellow willow (Salix lutea), and sandbar willow (Salix exigua). Non-native and invasive Russian olive (Elaeagnus angustifolia) is very common on the lower Judith River.

Figure 3. Judith River Valley near PN Ranch

Arrow Creek

Arrow Creek originates in the Little Belt and Highwood mountain ranges and drains portions of Judith Basin, Fergus, and Chouteau Counties. The Arrow Creek watershed is approximately 1,224 square miles (783,352 acres) and sits immediately west of the Judith River watershed. Arrow Creek also flows northward into the Missouri River and encompasses the communities of Geraldine, Geyser, Coffee Creek and Square Butte. Major tributaries include Flat Creek, Lone Tree Creek, and Cottonwood Creek. Most water right claims in the watershed are agriculture related and are associated with irrigation or livestock watering facilities. No major storage structures are present within the watershed, and irrigation normally occurs when water is available. Most irrigation takes place near the edge of the mountains where a more perennial source of water is likely. As Arrow Creek drops into the badlands of the Missouri Breaks, little to no irrigation development exists. Numerous small reservoirs and stock water ponds are located throughout the watershed. Arrow Creek also has a relatively natural hydrograph and may go dry in late summer and early fall during dry years. Whether that is natural or due to water developments is unknown.

Figure 4. Arrow Creek Watershed

Not surprising since the Arrow Creek watershed borders the Judith and they share headwaters in the same mountain ranges, the watershed and climatic characteristics of the catchments are similar. The climatic characteristics of the Arrow Creek watershed are also continental with cold winters. The average January maximum temperature in Geraldine is 34 degrees F (WRCC, 2011). “Chinook” weather is also common in the Arrow Creek basin, and cold waves are often interrupted by periods of mild, windy weather. Summer temperatures are warm, and can be hot particularly in the lower reaches of the drainage. The average July maximum temperature is 85 degrees F in Geraldine (WRCC, 2011). Precipitation is also very dependent upon elevation, and annual precipitation varies from 13 to 15 inches in the lower end of the watershed while the portions of the watershed within the Little Belt and Highwood mountain ranges receive around 30 inches. Arrow Creek is somewhat more arid than the Judith Basin with 85 percent of the watershed receiving 19 inches or less of precipitation (Montana NRIS, 2011). Elevations range from 2,440 at the mouth of Arrow Creek to 7,952 in the highest portion of the watershed in the Little Belts. Arrow Creek generally peaks in May and June as spring rains supplement any remaining snow melt. Arrow Creek is more arid than the Judith and the headwaters hold less snow for a shorter period of time. Prairie snow melt in March may create double peaked hydrographs, and late summer thunderstorms may cause short duration high flow events. Arrow Creek is often dry in August and September of drought years.

Arrow Creek is very sinuous, and the UMRBNM boundary encompasses approximately the last 27 ½ river miles. As with the Judith, much of the bottomland is privately owned although State of Montana trust

6 lands and BLM managed lands do compose a greater percentage than the Judith. Most of the surrounding hill sides are BLM lands or State of Montana trust lands. The reach of Arrow Creek within the Missouri Breaks is arid and warmer than much of the basin. Elevations range from 2,440 at the mouth to 3,500 feet on the canyon rims. Annual precipitation is 13 to 15 inches, and the average maximum temperature in July for Winifred and Iliad is 86 degrees F (WRCC, 2011).

Figure 5. Lower Arrow Creek during flood conditions

Arrow Creek flows through a wide valley bottom of Quaternary alluvium (Wilde and Porter, 2001), and these deposits of modern day channels are set within a canyon of sedimentary layers. Since the sedimentary layers in this area dip slightly eastward, and Arrow Creek sits west of the Judith, the canyon sides show older geologic formations. The formations still include the Cretaceous Claggett shale and Eagle sandstone formations, but the older, Cretaceous Telegraph Creek Formation is also present (Wilde and Porter, 2001). Because of the wide alluvial valley and floodplain, Rosgen C-channel type characterizes most of lower Arrow Creek. Rosgen C channels are low in gradient, meander, and have point-bar with riffle/pool morphology (Rosgen, 1996). These characteristics are created as the stream migrates laterally across the floodplain. Channel substrate materials contain large percentages of fines, but gravel and cobble are also present.

Similar to the Judith River, Arrow Creek’s ability to flood and active disturbance processes that create and abandon river channel have met the requirements for cottonwood recruitment, which are bare, moist sites that are safe from subsequent disturbance (Auble et al., 2005; Scott et al. 1996). Bare, moist sites are created with point-bar development and channel widening during flood events. Those sites become safer from future scour from flow and ice over time as the active channel moves away from them and vertical accretion of the floodplain moves them higher. The cottonwood forest on Arrow Creek is dominated by plains cottonwood (Populus deltoides). Other riparian tree and shrub species include peach-leaf willow (Salix amygdaloides), yellow willow (Salix lutea), and sandbar willow (Salix exigua). Russian olive (Elaeagnus angustifolia) is present on Arrow Creek but not in large amounts currently.

APPROACH AND METHODS

The requirements for the establishment and recruitment of riparian, pioneer species such as plains cottonwood (Populus deltoides) are relatively well known. Plains cottonwood is a pioneer, flood- dependent species. For establishment to be successful, cottonwoods need bare, moist surfaces that are protected from future disturbances (Scott et al. 1993, 1996; Scott and Auble, 2002; Auble et al., 2005). Plains cottonwood produces very large numbers of seeds that are dispersed by wind and water. In the plains of central Montana, they are dispersed roughly six weeks from early to mid June through mid to late July. Typically, germination and establishment of seedlings occur on bare, moist alluvium following medium to large floods (Scott et al., 1993). Seedlings are positioned low enough that they receive adequate moisture and avoid desiccation; however, they must be high enough to avoid subsequent scour by flow and ice. Often the recruitment band is between 0.6 and 2 meters in elevation above the late summer stream stage (Mahoney and Rood, 1998).

Root growth of new seedlings must keep pace with declining water tables through the summer (Mahoney and Rood, 1998; Scott et al., 1993), and the rate of stream stage decline should not exceed 2.5 cm (1 inch)/day (Mahoney and Rood, 1998) although manipulative lab experiments have shown some seedling can keep pace with declines up to 8 cm/day (survival less than 25%). Base summer flows are important for maintaining adequate moisture when evapotranspiration is highest, and historic alluvial water tables must be maintained. Declines in historic, shallow groundwater elevation have been shown to cause mortality even in mature, existing trees. Scott et al. (1999) found that in medium alluvial sands, sustained groundwater declines greater than one meter produced leaf desiccation and branch dieback within three weeks, and 88% mortality of trees over a three-year period.

Perhaps most important for long-term maintenance of cottonwood forest is the creation of new disturbed sites suitable for regeneration from seed. These sites are generally dependent on channel movement, sediment deposition and erosion associated with streamflow variation and infrequent high flow events. However, if the conditions that created the disturbance are repeated in the same locations, those sites are unlikely to be safe enough in the future to allow for recruitment of trees to larger size classes. These processes are important not only for creating new, suitable, bare, moist substrates, but also for moving the active channel away from established sites and for vertical accretion of the floodplain, which moves established cottonwoods into safer positions from water and ice scour.

Flow variability in terms of magnitude and timing is crucial for maintaining riverine disturbance and riparian forest. A decrease in channel migration rate has been associated with a decrease in reproduction of riparian pioneer species (Friedman et al., 1998). Without flooding, zones of dense vegetation encroach on the water’s edge competing with cottonwood seedlings, erosion and deposition processes are reduced; thereby decreasing the amount of suitable, recruitment sites and the seasonal pattern of spring flooding and gradual flow decline for seedlings is lacking (Rood and Mahoney, 1995).

Based upon the above mentioned requirements, the approach on the Judith River and Arrow Creek was to measure and quantify the flow regimes on those streams which meet the conditions suitable for the establishment and recruitment of cottonwood. A United States Geological Survey (USGS) stream gage has been in operation on the Judith River since 2000 (Station # 06114700), and the BLM began

8 measuring stream flow on Arrow Creek in 2007. Actually, the BLM began evaluating flow conditions in 2005 and 2006, but it wasn’t until 2007 that continuous record through the year was collected. However, because of the short period of record of these stations, we also initiated some additional retrospective studies. These include estimates of likely historical streamflow using regional analyses and analysis of vegetation and channel change from historical aerial photography. The objectives of this work were to better understand what variability in river flows have occurred in the past, what time periods were associated with changes in riparian forest and river channel, and what types of river flows occurred during those times.

We established intensive study reaches on both the Judith River and Arrow Creek which were representative of those stream’s vegetative and morphologic characteristics within the UMRBNM boundary. The study reaches were placed immediately below the stream gages in order to facilitate the calibration of a subcritical flow hydraulic model and floodplain modeling.

1000 feet N

Judith River Stream Gage

Judith River Study Reach Aerial View

Figure 6. Judith River Study Reach

Channel cross section locations were chosen based upon locations that would be conducive for recruitment of cottonwood (i.e. point bar formations), locations that would provide hydraulic, section control for floodplain modeling (i.e. top of a riffle), and in intermediate locations to decrease the channel distances between cross sections.

1000 feet

Figure 7. Arrow Creek Study Reach

Two additional cross sections on Arrow Creek were actually present during initial setup, A1 and A2, which were located on the big meander bend to the north of A3. However, they proved to be too problematic for subsequent flow modeling, particularly at high flows. When river flows become very large (i.e. full valley bottom flooding), the water surface profile is no longer section or channel controlled. In fact, the water surface profile (actually energy gradient) begins to approach the valley bottom slope. This effect was difficult to model, especially where the river channel runs perpendicular to the valley bottom. So, the study reach was ended at cross section A3.

The channel cross sections on the Judith River and Arrow Creek were surveyed with survey-grade global positioning system (GPS) (Trimble R-8 base and rovers). A Pentax PTS-III total station connected to the Trimble data logger was used to fill in locations where the canopy was interfering with good satellite reception. While the timeframe of the study was too short to establish any long-term trends in changes in river morphology at the cross sections, the cross sections were surveyed during three different field visits in order to measure channel morphology change.

Although the surveyed channel cross sections were able to encompass flows within the active channel and smaller overbank flows, there have clearly been flows larger than the surveyed cross sections. Furthermore, the extent of bottomland forest is obviously wider than the intensely surveyed cross section. So each section was extended to the width of the valley bottom with a digital terrain model (DTM) with one foot contour intervals. This was accomplished with a photogrammetry flight of the study reaches, and the DTM was created in the form of a triangulated irregular network (TIN) compatible with ArcGIS software.

Figure 8. Judith River DTM and Valley Bottom Cross Section Cut Lines

A low flow HEC-RAS hydraulic model and a high flow HEC-geoRAS hydraulic model were completed for both the Judith River and Arrow Creek. These models were calibrated to known stage-discharge relations at the stream gages on the upstream end of the reach and to surveyed water surface elevations for varying flows at the cross sections. High-water marks from dates and times of known discharges were surveyed at least at the uppermost four cross sections for calibration. At times, actual water surface elevations were measured at each cross section for a given discharge. The models were calibrated by adjusting Manning’s roughness coefficients until given discharges matched known water surface elevations at the cross sections. The Judith River models were fairly strong up to approximately 3,000 cubic feet per second (cfs) (the highest discharge to occur between 2005 and 2009), and the Arrow Creek models are fairly strong up to approximately 825 cfs. Above those values, their accuracy is less certain, but no errors occurred when running steady flow analysis. The purpose of the hydraulic modeling was to combine the low and high flow models in order to create as accurate as possible stage- discharge ratings for each individual cross section.

Figure 9. Judith River at Approximately 2,000 cfs on May 27, 2008

Vegetative transects were sampled in three late-summer field trips (2007, 2008, and 2009) using methods similar to those described for permanent monitoring sites on the mainstem Missouri River in the UMRBNM (Auble and Scott 1998). The transects coincided with the channel cross sections that were representative of features common on the river but had greater potential for riparian woody species recruitment. Cross sections across laterally active areas such as point bars were chosen as opposed to cross sections across more static transition reaches. Cross sections used for the collection of vegetative data were J5 and J6 on the Judith River and A3, A7, and A8 on Arrow Creek. The transects were ten meter wide belt transects (20 meters for mature trees), centered on the cross section line, and ran the extent of the valley bottom width. The presence and absence of every woody species and age class (seedling, sapling, pole, mature) were recorded at 1 meter intervals along the transect.

Figure 10. Example of Vegetation Sampling Along Cross Section 5 on Judith River.

Because the establishment surface of existing cottonwood trees is commonly buried by sediment deposition, some trees were excavated to their establishment surface and that elevation was surveyed. Trees from different stands were also cored and aged for comparison to the flow histories of the basins. Because of time and equipment constraints, the excavations did not go deep enough to determine an absolute establishment point. However, they are useful for determining boundary conditions: establishment elevations were generally no higher than the current ground elevation in the absence of evidence of exposed aboveground roots; establishment elevations were no lower than the establishment surface determined by excavation; and trees are at least as old as the largest ring count.

Figure 11. Pole Cottonwood Tree Excavated to the Establishment Surface

The belt transects and 1 meter presence/absence zones essentially create thousands of virtual plots on the vegetation transects. Because the horizontal locations of the plots on the transects are known and the vegetation transects are centered on the channel cross section lines, the zones can be plotted relative to their channel elevation and inundating discharge class. The elevations of the edges of each plot were determined by linear interpolation between surveyed elevations and the edge elevations were averaged for a single representative plot elevation. Water surface elevations with specific discharges were used to give each plot an inundating discharge class. These were derived from rating curves for each cross section developed from the hydraulic models.

In order to measure and monitor the shallow, alluvial groundwater, twelve shallow piezometers or well points were installed, five on the Judith and seven on Arrow Creek. The well points were installed with a geoprobe and consisted of 2 inch polyvinyl chloride (PVC) with the bottom five feet being well screen. Depths ranged from 10 to 25 feet. The well points were installed along the cross sections used for the vegetation transects and at the upper most cross sections, which coincide with the stream gages. The purpose for measuring groundwater at the stream gages was to have a location where there was a concurrent measurement of both river stage and groundwater elevation. As previously mentioned, the rest of the piezometers were installed along the cross sections used for vegetation sampling. The purpose was to evaluate the groundwater/surface water interaction at those sections.

N N

Judith River Piezometers Arrow Creek Piezometers

Figure 12. Judith River and Arrow Creek Piezometer Locations

The well points were surveyed in relative to the study reach’s datum, and pressure transducers were used to record groundwater elevations over time to find summer low flow levels, compare recession rates of the surface and groundwater elevations, and see how changes in surface stage corresponds to groundwater elevation.

Because the recession rate of groundwater decline is a crucial component of cottonwood seedling survival, a recession rate analysis was completed for the Judith River and Arrow Creek. Using the rating curves for the channel cross sections, the amount of discharge associated with 1 inch (2.5 cm), 2 inch (5 cm), and 4 inch (10 cm) stage declines for varying levels of flow was determined for each cross section. The values from each cross section were averaged to calculate a mean change in discharge for each stream associated with the aforementioned stage declines at different flow levels.

ANALYSIS OF FLOW REQUIREMENTS

As expected for the Judith River and Arrow Creek, based upon the air photo intervals of river channel and riparian forest change, historic flow reconstruction, tree aging, inundating discharge classes of vegetative plots, and hydraulic modeling of floodplain features, successful recruitment of cottonwood trees is associated with infrequent flood events and geomorphic processes such as lateral channel migration and channel narrowing following widening due to flood events. Sando (written communication, provisional – in review and subject to revision) found that there were two relatively short time periods during which there were several years when high-streamflow conditions in the Arrow Creek and Judith River watersheds probably were much above average: 1962-1970, and 1975-1981. During water years 1950-2009, three years probably were among the 5 largest annual peak flows for both Arrow Creek and Judith River at or near the mouths: 1953, 1975, and 1981 (Sando, written communication, provisional – in review and subject to revision).

Table 1. Estimated Largest Annual Peak Flows for Water years 1950-2009 (Sando, written communication, provisional – in review and subject to revision)

Rank Arrow Creek at Arrow Creek Judith River at the Water Year the mouth upstream from mouth the mouth of Flat Creek 1 6,420 5,070 10,700 1953 2 6,210 4,910 10,400 1981 3 5,170 4,090 8,650 1975 4 4,850 3,830 8,110 1997 5 4,670 3,690 7,810 1962 6 3,920 3,090 6,550 1978 7 3,730 2,950 6,240 1964 8 3,180 2,510 5,320 1970 9 3,170 2,500 5,300 1967 10 3,090 2,440 5,160 1974 11 3,050 2,410 5,100 1996 12 3,000 2,370 5,020 1993 13 2,950 2,330 4,940 1969 14 2,790 2,200 4,670 1979 15 2,760 2,180 4,610 1965

The majority of new riparian forest established in abandoned channels. Sixty percent of the new forest at Judith River and 86 percent of the new forest at Arrow Creek established in channels abandoned within the preceding 50 years (Auble, provisional – in review and subject to revision). With any retrospective analysis there is uncertainty about processes and mechanisms, but both Judith River and Arrow Creek have exhibited important patterns. There has been substantial flow variability between 1950 and 2009, with periods of channel widening and migration during intervals of higher than average flows and channel narrowing during intervals of relatively milder flow years. As measured by thalweg

15 location, new area occupied by channel, and channel width, substantial channel change has occurred from 1938 to 2006 (Auble, provisional – in review and subject to revision).

Figure 13. Judith River Historical Photos

Because plains cottonwood is a disturbance, flood-dependent species, flow variability is important to the river processes and maintaining the riparian forest on Judith River and Arrow Creek. When disturbance patches (bare, moist sites) occur in the same location year after year, the likelihood of any germinating cottonwood seedlings becoming mature trees is small because they are constantly subjected to continual disturbance. So, on Judith River and Arrow Creek, the key is getting bare, moist sites in different locations through the years.

Figure 14. Full Valley Bottom Flooding on Arrow Creek in June 2011, Approximate Flow Range – 4,000 to 5,000 cfs (Estimate by C. Krause).

During infrequent large flow events and periods of above average flow conditions, the river channel widens and in some cases could be described as blown out. Most regeneration of cottonwood at Judith River and Arrow Creek is likely associated with these conditions (i.e. 6,000 to 12,000 cfs at Judith River) as seedlings establish on the declining limb on the high-flow hydrographs or in years following as the channel narrows back to dimensions associated with more normal flows. Maintaining the ability of these river systems to flood and do so within a relative range of the frequency at which the flow events occurred historically is a crucial component of maintaining the riparian forest on these streams.

Judith River on July 11, 2011, Judith River on May 27, 2008 following a flood event of greater (approximately 2,000 cfs) than 10,000 cfs in June 2011.

Figure 15. Channel Change and Channel Abandonment on Judith River in 2011.

The results of the HEC-RAS and HEC-geoRAS hydraulic model and floodplain modeling were consistent with the magnitudes of flows that have occurred historically and what surfaces they may have inundated. The models were very well calibrated up to 3,000 cfs on the Judith and approximately 825 cfs on Arrow Creek, which were the highest flows to occur between 2005 and 2009. Flows above that were not calibrated to known water surface elevations, but no errors occurred when running a steady flow analysis and the flow values are within a reasonable range of values that have occurred historically.

Figure 16. Judith River DTM with Inundation Depths at 3,000 cfs

Because the Judith River and Arrow Creek have been laterally very dynamic, an assumption regarding channel morphology needs to be made that the existing channel morphology is similar to the past even though the channel may be in a different lateral position. One, there is no way to get around that assumption in terms of floodplain modeling with the existing morphology, and two, even in terms of cottonwood establishment surfaces that may be under sediment deposition, flow was there at one point in time that deposited that sediment. Furthermore, inundating discharges may be conservative estimates for periods of very high flow because during those times, scour on the rising limb of the hydrograph would result in channel enlargement. So, just looking at a flow necessary to inundate a given elevation would give smaller estimates with current channel morphology. An example of the results of the HEC-geoRAS model for a 3,000 cfs flow is shown in Figure 16.

Cross Section J5

Figure 17. Inundation Mapping and Relative Water Depths of 3,000 cfs on the Judith River and HEC-geoRAS Model Water Surface for 3,000 cfs Profile at Judith River Cross Section J5

Various profiles were run as steady flow analysis in both the HEC-RAS and HEC-geoRAS flow models in order to pick rating points to create a stage-discharge relation for each cross section. As mentioned previously, a low flow HEC-RAS model was used for flows less than 3,000 cfs on the Judith River and 825 cfs on Arrow Creek because the channel was tightly surveyed with survey grade GPS. The HEC-geoRAS model and DTM were used to pick rating points for higher flows although both models were calibrated to known water surface elevations for given discharges at the cross sections. The roughness coefficients were somewhat lower than anticipated, particularly for the overbank areas (.037-.040), but that is what was necessary to calibrate the model to known stage-discharge relations.

As mentioned above, flows above 3,000 cfs at Judith River and 825 cfs at Arrow Creek were not calibrated to the gages. However, comparison of the results of the hydraulic modeling with the results of the reconstruction of the flow histories matches very well. Using Judith River as an example, 2,000 cfs (approximately a 2-year flow), occurs around bankfull of the active channel for channel morphology associated with recent flow conditions. Five to 10,000 cfs inundates a good portion of the valley bottom and extent of bottomland forest, which is in line with the more infrequent, high flow events between 1950 and 2009.

Figure 18. Inundating Discharges of Cross Sections 5 and 6 at Judith River

The cross section rating curves for HEC-RAS and HEC-geoRAS use a linear relation between profiles that were run through the model. Because the purpose of the rating curves was to relate stream flow to riparian vegetation, it was necessary to produce more precise rating curves. The HEC-RAS and HEC- geoRAS models were used to pick the stages and discharges for rating points. The rating points for each cross section were imported into the rating development tool box in the Aquarius software package. Aquarius allows the user to fit a rating curve to the rating points. The result is a rating curve that gives a discharge for every 0.01 feet of elevation. The results were exported to Microsoft Excel in order to assign an inundating discharge class to the vegetation plots.

J5 Cross Section Rating

2497

2496

2495

2494

2493 J5 Cross Section Rating 2492

2491 Stage Stage (feet) J5 Rating Points 2490

2489

2488

2487

2486 0 5000 10000 15000 Discharge (cfs)

Figure 19. Judith River Cross Section 5 Rating Curve

The Judith River is larger than Arrow Creek, so mature cottonwood trees are accordingly associated with larger discharges and higher bank positions at Judith River than Arrow Creek. New seedling establishment of both cottonwood and willow occurs in zones that were wetted by the recent year’s flows. New seedling establishment associated with flows between 2007 and 2009 are, as expected, in low bank positions because of the low flow conditions that occurred during this period. However, even the estimated establishment surfaces of mature trees indicate a relatively low inundating discharge. This pattern further emphasizes the importance of the creation of new channel, abandonment of old channel, and the deposition of sediment and floodplain creation in old channel, rather than simple inundation.

Figure 20. 2008 Species Presence, in Fraction of Bottomland Plots Relative to Inundating Discharge

There has been some limited channel movement and riparian forest regeneration associated with moderate high flow events (i.e. 2,000 to 3,000 cfs at Judith River). For example, river flows of approximately 3,000 cfs at the Judith River in 2008 resulted in about 20 feet of channel movement between 2007 and 2008. No channel movement occurred in 2009. Relative to the average bottomland width at Judith River, which is 2,268 feet, and the lifespan of a plains cottonwood tree (100 to 200 years), this rate of channel movement (20 feet/every few years) would not have resulted in the amounts of riparian forest seen today.

Figure 21. Channel Movement at Judith River Cross Section 5

Analysis of the shallow groundwater data and river stage and flow indicated that the alluvial groundwater is very dependent upon river stage. Furthermore, the recession rate of stage decline in the river and the recession rate of groundwater elevation are closely connected, with little to no lag time between the two. This is as expected for an alluvial system with substantial amount of coarser material. Although fines are prevalent within the system, channel substrates are gravel and some cobble.

The figure on the next page plots daily mean discharge, observed groundwater elevations, and observed river stage during 2009. A12A is the piezometer located at the Arrow Creek gage, and J8C is the piezometer at the Judith River gage. The other ten piezometers show similar relations although the river stages are estimated with a model rather than measured in the field.

Figure 22. Daily Mean Discharge (cfs), Real Groundwater Elevations (ft), and Real River Stages at Arrow Creek (A12A) and Judith River (J8C)

Because of the connectivity between river flow and the shallow groundwater, maintaining some flow in the channel is important for maintaining historic groundwater levels. Declines in historic, shallow groundwater elevation have been shown to cause mortality, even in mature, existing trees. Scott et al. (1999) found that in medium alluvial sands, sustained groundwater declines greater than one meter produced leaf desiccation and branch dieback within three weeks, and 88% mortality of trees over a three-year period. Cottonwood roots grow and spread at groundwater elevations that the river system has historically provided. For example, once the roots hit the historic water table, they do not keep growing indefinitely. So, even though cottonwood trees in general may be capable of surviving in different groundwater conditions, it is important to consider the river specific conditions that they have adapted to.

During 2009, the time period where both river stage and groundwater elevation were measured, Arrow Creek maintained base flow throughout the late summer months (July, August, and September). This is often not the case as Arrow Creek frequently goes dry during these months. Conditions at Arrow Creek are probably at or near a point where low flows in late summer may be affecting seedling survival. The base flow at Judith River is generally maintained at or above 160 cfs from contributions due to Big Spring Creek and Warm Spring Creek.

Large drops in the receding limb of the hydrograph of these streams can cause groundwater declines that cottonwood seedling root growth is unable to keep up with. Whether the drops occur near the peak of the hydrograph or near base flow, the large, sudden stage declines can desiccate new seedlings. Root growth of new seedlings must keep pace with declining water tables through the summer (Mahoney and Rood, 1998; Scott et al., 1993), and the rate of stream stage decline should not exceed 2.5 cm (1 inch)/day (Mahoney and Rood, 1998) although manipulative lab experiments have shown some seedling can keep pace with declines up to 8 cm/day (survival less than 25%). Using the Judith River as an example, the percent change in discharge associated with a 1 inch (2.5 cm), 2 inch (5 cm), and 4 inch (10 cm) stage decline is relatively consistent at 5, 10, and 15 percent of the given discharge respectively. The notable change occurs as discharges approach low flow levels where small changes in river stage are large percentages of the discharge, further emphasizing the importance of maintaining base flow conditions.

Judith River Recession Rate Analysis

30%

25% Percent Change in Discharge Associated with 1" (2.5 cm) 20% Stage Decline

15% Percent Change in Discharge Associated with 2" (5 cm) Stage 10% Decline Percent Change in Discharge 5% Associated with 4" (10 cm) Stage Decline 0% 10000 8000 6000 4000 2000 0 Discharge (cfs) PercentChange Discharge in (percent)

Figure 23. Judith River Recession Rate Analysis

Without a doubt, at Judith River and Arrow Creek, the different components of the hydrograph (rising limb, peak, receding limb, and base flow) all play a role in sustaining the riparian forest. Although different, each component is important because if any of the pieces of the puzzle are missing, it would preclude recruitment of riparian forest. The key message is that maintaining the cottonwood forest at Judith River and Arrow Creek in the future will require protection of the magnitude of peak flows and

25 the frequency and timing at which they occur within a relative range of what they have occurred historically. Maintenance of the shallow, alluvial aquifers is dependent on at least some water being in the channel, particularly in later summer, and the rate of flow decline during the germination window is also important. The BLM proposal to the State of Montana RWRCC was intended to follow this thought process. The proposal is summarized below along with the rationale and supporting justification for each request.

PROPOSAL SUMMARY

The following is a summary of the BLM proposal for the negotiation of the UMRBNM water rights and the rational for each request.

Judith River

The BLM requests an instream flow for those flows remaining in the river at the time of the UMRBNM proclamation, subject to an amount for future development by the State of Montana. A base flow consistent and concurrent with the Montana Department of Fish, Wildlife and Parks (MT FWP) state reservation of 160 cubic feet per second (cfs) as measured at the stream gage on the Judith River (USGS station number 06114700) is requested as the basis for the United States to place a call on junior water rights and as an additional limit to new appropriations. The purpose of maintaining instream flow within the channel is for maintaining the shallow, alluvial groundwater. The instream flow value of 160 cfs was chosen for consistency with the State of Montana instream flow reservation already in place. Moreover, late summer low flows often drop to around 160 cfs, particularly during periods of drought. Existing, mature cottonwoods have survived based upon groundwater levels maintained by these types of flows.

Second, the BLM requests no new mainstem storage structures will be permitted on the Judith River. As mentioned above, maintaining the ability of the Judith River to flood within a relative range of historic frequencies is critical for maintenance of the cottonwood forest. As compared to off-site storage structures or storage structures on smaller order side streams, a mainstem storage structure has the ability to significantly attenuate the magnitude of flow entering the structure and considerably decrease the frequency at which certain magnitude flows occur. Just as important, mainstem reservoirs can disrupt the sediment regime, which is vital for the geomorphic processes that create depositional features suitable for cottonwood establishment.

Third, the BLM requests a cap on new development at 1,990 cfs, which represents the difference between 160 cfs and the median peak discharge for the Judith River Basin (2,150 cfs). This Judith River Available Water Supply (JRAWS) for new appropriations would be based upon flow rate as opposed to volume. In theory, much of the water the BLM hopes to protect is not realistically available for appropriation because it is not available on a reasonable basis. However, the BLM believes that it is a reasonable to make the flow rate between base flow and the median peak discharge available for development. The median peak discharge was chosen as the upper limit for the JRAWS because these flows would be available on a somewhat regular basis (every couple years) and allowing for development within these limits would be consistent with Montana water law. Furthermore, although allowing for development within the basins will decrease the flooding frequency and potentially lead to a flatter hydrograph (approximately half the years), the attenuation of much larger flow events (i.e. 6,000 to 12,000 cfs) that are most effective in creating conditions for extensive cottonwood recruitment would have a much smaller effect on riparian processes than the attenuation of the median, or 50th percentile flows. The regression equation estimates for the median peak discharge might be considered near the lower range of appropriate estimates (Sando, written communication, provisional – in review and subject to revision). However, we used them for the proposal because the flood-frequency relations

27 for the regression equations provide the most appropriate estimates for all recurrence intervals because they are associated with detailed and reliable error estimates.

Terms and conditions on direct from source diversions such as pumps, ditches, and canals would include only appropriating water when the base instream flow requirements are being satisfied (160 cfs). In addition, the BLM requests that permits for direct from source diversions greater than 10 cfs be conditioned to operate under a “ramped diversion” regime that prevents an increase in diversion of more than 10 cfs per day or 20 percent of the total allowed diversion, whichever is greater, in any 24- hour period to prevent sudden, drastic drops in the natural recession rate of the river. The purpose of the ramped diversion regime for large diversions is to protect the hydrograph from water developments that can individually create drastic changes in flow rate. So for example, a new 100 cfs diversion would have to ramp up to 100 cfs at 20 cfs intervals over a 5-day period.

Of additional concern to the BLM is the development of storage structures, which cumulatively may not take a significant percentage of the basins annual volume yield-but which take it all during runoff events. The BLM requests that applications for storage reservoirs larger than 15 acre-feet capacity be required to include hydrologic analysis showing the expected 2-year recurrence interval peak flow for the location. This flow number will be subtracted from the JRAWS. This would eventually cap the total amount of storage development in the watersheds, but it would allow for the development to occur flexibly (i.e. many smaller reservoirs or fewer larger ones) as needed by the users within the basin.

Small stock reservoirs (less than 15 acre-feet capacity) and domestic and stockwater wells less than 35 gallons per minute to 10 acre-feet per year would be exempted from the development cap based on the MCA 85-2-306 permit exceptions. Should the MCA 85-2-306 exceptions be changed to a less restrictive standard, or should these exceptions to permit requirements be removed, these developments would become a part of the State’s available development cap.

Arrow Creek

The proposal for Arrow Creek is structured similarly to the Judith River proposal; however, there are some differences including the instream flow values. The BLM requests an instream flow water right for those flows remaining in the river at the time of the UMRBNM proclamation, subject to an amount for future development by the State of Montana. The BLM is also requesting a base flow of 5 cubic feet per second from March 1 to July 31 of each year that would be the basis for the BLM to place a call on junior water rights. As with the Judith, the purpose of maintaining instream flow within the channel is for maintaining the shallow, alluvial groundwater. Although Arrow Creek flows year round some years, many years it goes dry by late summer, particularly during drought years. Arrow Creek typically flows from at least March 1 to July 31 of each year.

The BLM also requests no new mainstem storage structures will be permitted on Arrow Creek. As mentioned above, maintaining the ability of Arrow Creek to flood within a relative range of historic frequencies is critical for maintenance of the cottonwood forest. As compared to off-site storage structures or storage structures on smaller order side streams, a mainstem storage structure has the ability to significantly attenuate the magnitude of flow entering the structure and considerably decrease

28 the frequency at which certain magnitude flows occur. Just as important, mainstem reservoirs can disrupt the sediment regime, which is vital for the geomorphic processes that create depositional features suitable for cottonwood establishment.

The BLM requests a cap on new development at 457 cfs, which represents the difference between 5 cfs base flow and the median peak discharge for Arrow Creek upstream of the mouth of Flat Creek (462 cfs). The Arrow Creek Available Water Supply (ACAWS) for new appropriations would be based upon flow rate as opposed to volume. All junior, new developments within the cap would be subject to the same terms and conditions as described for the Judith River basin. These include direct from source appropriations only being used when the base instream flow requirements are satisfied; direct from source diversions greater than 10 cfs be required to operate under a “ramped diversion” regime that prevents an increase in diversion of more than 10 cfs per day or 20 percent of the total allowed diversion, whichever is greater, in any 24-hour period; and applications for storage reservoirs larger than 15 acre-feet capacity must include hydrologic analysis showing the expected 2-year recurrence interval peak flow. The rational for these requirements is the same as that explained for the Judith River.

Proposal Requests across Both Basins

The BLM requests that a State water right enforcement project be established for the Judith River and Arrow Creek basins, including the appointment of a water commissioner to enforce the provisions of this compact. The BLM believes that this compact proposal is reasonable and allows for a considerable amount of development; and that, if enforced, it would adequately protect the BLM’s water right for the UMRBNM. The BLM believes that the enforcement of terms and conditions for existing permits has historically been extremely difficult or non-existent. Failure to enforce the compact has the ability to adversely affect the BLM’s water right, particularly the critical base instream flows; thus the request for an enforcement project in these basins.

Groundwater appropriations not exempted from permit requirements under the MCA 85-2-306, that directly influence surface water will be subtracted from the available water supply.

Instream flow applications for non-consumptive uses, pursuant to the MCA 85-2-316, will not be subtracted from the available water supply.

For both Judith River and Arrow Creek basins, once the available water supply is exhausted, the basin will be closed to new applications for appropriation.

ACKNOWLEDGEMENTS

Assessment of the instream flow requirements for riparian forest maintenance at Judith River and Arrow Creek would not have been possible without the contributions from numerous individuals. In particular, Jim Fogg, retired BLM surface water specialist (BLM National Operations Center) assisted with study design; and more importantly, provided mentorship and exhibited extreme patience for the collection of hydrologic data and flow modeling. Special thanks is given to Greg Auble, USGS riparian ecologist (Fort Collins Science Center) for collaborating with BLM on studies of riparian vegetation in the Upper Missouri River Breaks National Monument. Steve Sando, USGS hydrologist (Montana Water Science Center) has been conducting reconstructions of the flow histories of the basins and completed the flow variability and flood frequency analysis. Paul Summers, BLM groundwater specialist (BLM National Operations Center) helped with choosing locations for the piezometers and provided suggestions for evaluating the shallow groundwater. Mike Scott, USGS riparian ecologist (Fort Collins Science Center) and Mike Merigliano, riparian ecologist (University of Montana) lent a hand with field work, and more importantly provided numerous insights, guidance and suggestions. None of the assessment work would have been completed without Sheila Tesarek, former BLM hydrologic technician (BLM Lewistown). Sheila always ensured all cadastral surveying, discharge measurements, and field work were completed in a professional, timely manner. Frances Rieman, BLM water rights specialist (BLM Montana State Office) provided the cross walk between the on-the-ground science and an actual water right proposal. Many others assisted with field work, completion of the digital terrain models, and successful completion of the project, but their names are too numerous to mention.

REFERENCES

Alaska Department of Natural Resources. 2000. Federal Reserved Water Rights. www.dnr.state.ak.us/mlw/factsht/wtr_fs/fed_rsv.pdf.

Auble, G., M. Scott, M. Merigliano, and C. Krause. 2010. Monitoring of cottonwood seedling demography in the Missouri Breaks National Monument. Poster presentation at Decade of Discovery, National Landscape Conservation System Science Symposium, May 26-28, 2010, Albuquerque, NM.

Auble, G., M. Scott, J. Frazier, C. Krause, and M. Merigliano. 2005. Cottonwood in the Missouri Breaks National Monument. U.S. Geological Survey, Biological Resources Discipline, Fact Sheet 2005-3132, 4 p.

Auble, G.T., and M.L. Scott. 1998. Fluvial disturbance patches and cottonwood recruitment along the upper Missouri River, MT. Wetlands 18: 546-556.

Bovee, K. and M.L. Scott. 2002. Implications of flood pulse restoration for Populus regeneration on the upper Missouri River. River Research and Applications 18:287-298.

Bureau of Land Management (BLM). 2008. Lewistown Field Office. Upper Missouri River Breaks National Monument Record of Decision and Approved Resource Management Plan.

Mahoney, John M. and Stewart B. Rood. 1998. Streamflow requirements for cottonwood seedling recruitment – an integrative model. Wetlands 18: 634-645.

Montana Natural Resource Information System (NRIS). 2011. Geographic Information: climate, land information, water. http://nris.mt.gov/

Rood, Stewart B. and John M. Mahoney. 1995 River Damming and Riparian Cottonwoods along the Marias River, Montana. Rivers 5: 195-207

Rosgen, D. 1996. Applied river morphology. Wildland Hydrology, Pagosa Springs, CO. 352 pp.

Friedman, J.M., W.R. Osterkamp, M.L. Scott, and G.T. Auble. 1998. Downstream effects of dams: regional patterns in the Great Plains. Wetlands 18: 619-633.

Scott, M.L., and G.T. Auble. 2002. Conservation and restoration of semi-arid riparian forests: a case study from the upper Missouri River, Montana, USA. Pages 145-190 in Flood Pulsing and Wetland Restoration in North America, B. Middleton, (ed.), John Wiley and Sons, Inc.

Scott, M.L., G.T. Auble, and J.M. Friedman. 1997. Flood dependency of cottonwood establishment along the Missouri River, Montana, USA. Ecological Applications 7:677_690.

Scott, M.L., J.M. Friedman, and G. T. Auble. 1996. Fluvial process and the establishment of bottomland trees. Geomorphology 14: 327-339.

Scott, M.L., P.B. Shafroth, and G.T. Auble. 1999. Responses of riparian cottonwoods to alluvial water table declines. Environmental Management Vol. 23(3): 347-358.

Scott, M.L., M.A. Wondzell, and G.T. Auble. 1993. Hydrograph characteristics relevant to the establishment and growth of western riparian vegetation. Pages 237-246 in H.J. Morel-Seytoux, ed., Proceedings of the Thirteenth Annual American Geophysical Union Hydrology Days. Hydrology Days Publications, Atherton, CA.

Wayne, W.J., Aber, J.S., Agard, S.S., Bergantino, R.N., Bluemle, J.P., Coates, D.A., Cooley, M.E., Madole, R.F., Martin, J.E., Mears, Jr., B., Morrison, R.B., and Sutherland, W.M., 1991, Quaternary geology of the northern Great Plains, in Morrison, R.B., ed.: Geology of North America, volume K-2, The Geological Society of America, Boulder, CO, USA, p. 441-476.

Western Regional Climate Center. 2011. Historical Climate Information. www.wrcc.dri.edu/

Wilde, Edith M., and Karen W. Porter. 2001. Geologic Map of the Winifred 30’ by 60’ Quadrangle Central Montana. Montana Bureau of Mines and Geology Open File No. 437.

APPENDICES

A-1 Judith River Mean Daily Discharge (2000-2010)

Judith River Mean Daily Discharge (cfs) (2000- 2010) 1200

1000

800 Judith Mean Daily 600 Discharge (cfs) Discharge (cfs) 400

200

Time

A-2 Arrow Creek Daily Mean Discharge (2007, 2008, 2009)

Arrow Creek 2007 Daily Mean Discharge (cfs) 450 400

350 Arrow 300 Creek 2007 250 Daily Mean 200 Discharge 150 (cfs) Discharge (cfs) 100 50 0 Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Time

Arrow Creek 2008 Daily Mean Discharge (cfs) 800 700 Arrow

600 Creek 2008 500 Daily Mean 400 Discharge 300 (cfs)

Discharge (cfs) 200 100 0 Mar-08 Apr-08 May-08 Jun-08 Jul-08 Aug-08 Sep-08 Oct-08 Time

Arrow Creek 2009 Daily Mean Discharge (cfs) 140 120

100 Arrow Creek 2009 Daily 80 Mean 60 Discharge (cfs)

Discharge (cfs) 40 20 0 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Time

A-3 Arrow Creek and Judith River Digital Terrain Models

Arrow Creek Digital Terrain Model

Judith River Digital Terrain Model

UPPER MISSOURI RIVER BREAKS NATIONAL MONUMENT

ARROW CREEK and JUDITH RIVER WATERSHEDS

UPPER MISSOURI RIVER BREAKS Ë NATL MONUMENT

Montague !(

Arrow Creek USGS GAGE: Judith River, JudithRiver Geraldine near mouth; near Winifred (!

Flat Creek

Square Butte ARROW CREEK !( WATERSHED Arrow Creek (1271 Sq. Mi )

Arrow Creek Judith River

Hay Creek Coffee Creek Arrow Creek !( JUDITH RIVER Denton WATERSHED (! (2762 Sq. Mi.) Geyser !( Surprise Creek Deerfield Colony !( Hilger Wolf Creek !( Coyote Creek Sage Creek Danvers !(

Warm Spring Creek Stanford Maiden (! !( Big Spring Creek

Lone Tree Creek Willow Creek

Kolin !( Dry Wolf Creek Windham Benchland !( !( Louse Creek Ross Fork Judith River!( Moccasin Lewistown !( (!

Cottonwood Creek

Running Wolf Creek Hobson (! Judith River Olsen Creek Utica Heath Moore !( !( (!

Rock Creek Hansen Creek Sapphire Village

!( Antelope Creek Middle Fork Judith River

Buffalo !(

East Fork Big Spring Creek

South Fork Judith River Garneill !( Ross Fork Creek

Legend LAND MANAGEMENT BLM Loc Govt MT State Other Federal Private USFS 02.5 5 10 Miles Legend UPPER MISSOURI RIVER ARROW CREEK BASIN BREAKS NATIONAL MONUMENT BOUNDARY JUDITH RIVER BASIN

09-09-2011 No warranty is made by the Bureau of Land Management as to the accuracy, reliability, or completeness of these data for individual or aggregate use with other data. Original data were compiled from various sources. This information may not meet National Map Accuracy Standards. This product was developed through digital means and may be updated without notification. UPPER MISSOURI RIVER BREAKS NATIONAL MONUMENT ARROW CREEK WATERSHED

UPPER MISSOURI RIVER BREAKS Ë NATIONAL MONUMENT

Unnamed1

Unnamed

Flat Creek Pass Coulee

Mud Spring Coulee Unnamed

Unnamed

Telephone Coulee Flat Creek

Windy Coulee Unnamed Deadman Coulee Cut Bank Coulee Nance Coulee G e r a l d i n e

Fahlgren Coulee

Phantom Coulee Wilson Coulee

Lepleys Creek Slide Coulee UnnamedAlder Creek Woodcock Coulee Spring Coulee Big Coulee Steele Lake Coulee ARROW CREEK Kelley Creek Melton Coulee Soda Spring Coulee Cowboy Creek WATERSHED Warren Creek

Boyd Creek Butte Creek

(1271 Sq. Mi ) Little Battle Creek Gerard Creek

Ole Coulee

Possum Run Creek Martin Creek Fall Creek Cottonwood Creek Pine Creek Rose Creek Davis Creek

Black Coulee Byrne Creek Braun Creek Big Coulee

Hay Creek Arrow Creek Coffee Creek Legend

BLM

Lone Tree Creek Surprise Creek Leiberg Coulee Local Govt

Dipping CreekVat MT State Othr Fed

Old Geyser CrowCreek Coulee Private Mccarty Creek USFS Fox Coulee Upper Missouri

Shannon Creek River Breaks North Fork Surprise Creek National Monument

Wolf Butte Coulee

Arch Coulee Sun Creek Bower Gulch

Peterson Gulch

Legend 0 3.5 7 14 Miles ARROW CREEK BASIN JUDITH RIVER BASIN

09-07-2011 No warranty is made by the Bureau of Land Management as to the accuracy, reliability, or completeness of these data for individual or aggregate use with other data. Original data were compiled from various sources. This information may not meet National Map Accuracy Standards. This product was developed through digital means and may be updated without notification. UPPER MISSOURI RIVER BREAKS NATIONAL MONUMENT JUDITH RIVER WATERSHED

UPPER MISSOURI RIVER BREAKS NATIONAL MONUMENT Ë

JUDITH RIVER WATERSHED

(2762 Sq. Miles)

ARROW CREEK WATERSHED

(1271 Sq. Mi )

D e n t o n

Wolf Creek Coyote Creek

Warm Spring Creek S t a n f o r d Sage Creek Big Spring Creek Willow Creek

Dry Wolf Creek

Louse Creek Cottonwood Creek L e w i s t o w n

Rock Creek Running Wolf Creek H o b s o n Judith River Olsen Creek Ross Fork Creek

M o o r e

Hansen Creek

Antelope Creek

Middle Fork Judith River

East Fork Big Spring Creek South Fork Judith River

Legend

BLM Local Govt MT State Othr Fed Private USFS Upper Missouri River Breaks National Monument Legend USGS GAGING 0 5 10 20 Miles ARROW CREEK BASIN $+ STATION JUDITH RIVER BASIN

No warranty is made by the Bureau of Land Management as to the accuracy, reliability, or completeness of these data for individual or aggregate use with other data. Original data were compiled from various sources. This information may not meet National Map Accuracy Standards. This product 09-08-2011 was developed through digital means and may be updated without notification.