Figure 5. Average annual differences in shape among all stations on Pole Creek in Pole Creek Pasture. Positive values indicate channel is getting narrower and deeper. Bars with different letters differ significantly.
By B.J. Rhodes, C.B. Marlow, and H.W. Sherwood*
Stream channel form and bank stability are important both for stream function and the biotic community structure of riparian ecosystems (11, 14). Stream form and structure not only dictates habitat for fish and other aquatic organisms (1, 13), but also regulates erosion, flow regime and groundwater recharge (17, 5). Consequently, stream channel stability has become a fundamental component of the Forest Service-Bureau of Land Management preferred management alternative described in the Rangeland Reform '94 Final Environmental Impact Statement (20).
The public land management agencies intend to promulgate national guidelines and standards to achieve a 27 percent improvement in riparian function on Bureau of Land Management lands, and a 7 percent improvement on Forest Service lands (20). It is intended these guidelines would become the framework for state and local standards (19).
However, one recent federal riparian management manual (2), and an earlier scientific journal article (10) both arrive at the same general conclusion: "that each watershed, stream, stream reach and riparian area has unique characteristics that must be accounted for in developing a grazing strategy." This poses a significant challenge to the development of national standards, because land managers must assume that all rangeland and riparian sites within the same region and forest will respond in a similar manner. If streams, or even stream reaches react differently to grazing, then the application of national standards and guidelines may accomplish little more than prolonging the controversy over livestock use of riparian areas.
Results from a long-term riparian and rangeland monitoring project conducted under a cooperative agreement between the US Forest Service, Matador Cattle Company and the Montana Agricultural Experiment Station (7) can provide insight into the question of whether streams under the same grazing management strategy respond in a similar manner, or if each responds differently. While the objective of the cooperative project was to describe the response of streambanks, willow populations and rangeland to a particular grazing strategy, systematic collection of stream channel morphological data allowed us to compare the response of two separate stream systems managed under the same allotment management plan.
Study Site Description. The study site lies within the Long Creek Forest Service allotment, Beaverhead National Forest, about 44 miles southeast of Dillon, Montana. The Long Creek allotment occupies the Ruby Valley-Centennial Valley divide and ranges in elevation from 7,000 to 7,960 feet. Upland vegetation is dominated by the mountain big sagebrush (Artemisia tridentata Nutt.)/Idaho fescue (Festuca idahoensis Elmer.) habitat type (8). Riparian vegetation is dominated by Geyer willow (Salix geyeriana Anderss.), Booth willow (Salix boothii Dorn.), sedges (Carex spp.) and Kentucky bluegrass (Poa pratensis L.). Annual precipitation ranges from 14.5 inches to 30 inches with most occurring during the winter and spring months (21).
The allotment is located on an ancient erosional surface that has been preserved along the crest of the adjacent Gravelly Range. The surface is gently rolling with the deeply weathered underlying rocks being prone to landslides, while ridges are underlain by more resistant sandstone. The ancient surface is overlain in places by gravels, which in turn are overlain by a thick preglacial soil that is preserved in only a few small areas (15). Drainage is strongly rectilinear, and controlled by fractures in the bedrock. Streams are fed primarily by groundwater, which is also fracture controlled (15).
Grazing History
Unrestricted livestock grazing was curtailed in 1935 when the US Forest Service established the present Long Creek Allotment boundary. From 1935 to 1962, livestock on the allotment varied from 535 cows and 75 horses to as many as 1,238 cows and 115 horses or 1,351 sheep and 120 horses. Season of use was June 1 to October 31 for horses and August 1 to October 31 for cattle and sheep.
In 1963 and 1964, when the Forest Service aerially applied 2,4-dichlorophenoxyacetic acid (2,4-D) to control sagebrush, the allotment was not grazed by livestock. In 1965, the allotment was fenced into Pole Creek Pasture, Jones Creek Pasture, Long Creek Pasture and Lone Butte Pasture. Long Creek also runs through Lone Butte Pasture.
Two of these pastures were used in the study reported here: Pole Creek Pasture and Lone Butte Pasture and are part of the system (Figure 1). Season of use was set at 90 days, from July 16 to October 15, for 800 cow/calf pairs.
In 1986 concerns of increasing sagebrush caused the permittee to voluntarily reduce the permit by 200 cow/calf pairs (6). Further reductions in livestock grazing were discussed in 1989 due to concerns over riparian condition. This prompted the development of a double rest rotation grazing strategy which would be administered under a Memorandum of Understanding between the permittee, Forest Service and Montana Agricultural Experiment Station (7).
The current grazing plan was developed by Resource Concepts Inc. (1991), and implemented with the 1991 grazing season (Table 1).
Year Pole Creek Lone Butte Long Creek Jones Creek
1991 Grazed early Rested Grazed late Grazed mid
1992 Grazed early Rested Grazed mid Grazed late
1993 Rested Grazed early Grazed late Grazed mid
1994 Rested Grazed early Grazed mid Grazed late
Under the grazing plan cattle numbers were to be increased to about 775 cow-calf pairs. This produced a stocking rate of 2.5 to 3.7 acres per animal unit month from mid-July to mid-October. Grazing in the Jones Creek and Long Creek pastures rotated yearly between mid and late season use. The Pole Creek and Lone Butte pastures were grazed for two years, followed by two years of rest (double rest-rotation). Livestock grazed each pasture for three to four weeks and were moved when the stubble height of riparian grasses had been reduced to about three to four inches as mandated in the Beaverhead Forest Plan. Another component of the plan was to haze livestock from the riparian areas to improve distribution and ensure compliance with the Allotment Management Plan.
Field Techniques. Eight stations were established, four in each of the two pastures. At each station, we established three transects across the creek, and one set of transects in a fenced "exclosure" that excluded both cattle and large game. We closed off one additional area in the lower corner of Lone Butte Pasture with a fence that excluded only cattle. The exclosures were an attempt to account for natural variation in stream morphology. The cattle and game exclosures have not been grazed since 1990, and the cattle exclosure has not been grazed since 1989.
Permanent transects were marked at about 50-foot intervals and placed across the channel perpendicular to water flow (11). Stakes were placed at the endpoints, about four inches inland from high water edge, and driven to a depth sufficient to ensure permanency. A line level was used to set the horizontal line from which depth was measured, and reset to these measurements each time the transect was measured. The distance from the transect tape to the channel bottom was measured every four inches along the tape, always beginning on the right side of the channel while facing downstream.
Data Analysis. Gini-coefficient values were used to determine changes in channel morphology over time (11). Gini coefficients describe the profile of the stream channel and indicate whether the channel is becoming wider and flatter or narrower and deeper. Positive percent change [("after grazing" minus "before grazing" divided by / after grazing) times 100] in Gini values indicate the channel is becoming deeper and narrower. Negative values that it is becoming wider and shallower.
Four observations were made for each grazed station every year. Stream channel transects were measured:
Gini values were evaluated by analysis of variance (ANOVA) to determine differences in Gini-coefficient values between pastures, stations, seasons and years. All statistical tests were performed using the General Linear Model procedures of SAS (16), and considered significant if there was less than a five percent chance of the result occurring by chance (alpha = 0.05).
The results indicate that stream channel response to land management practices can be irregular. For example, when we looked at the stream channel response to similar management, stations varied (P<0.01) even within the same pasture. The mean values of the Gini coefficient for the Lone Butte stations and the cattle exclosure were all similar (P>0.05) during the study, with the exception of those from Lone Butte Station 1 on Long Creek (Figure 2). All stations in Pole Creek pasture were different (P<0.05) from each other (Figure 3).
No difference in Gini coefficient values between pastures (P>0.26), seasons (P>0.07), or years (P>0.70) was detected in the analysis. Similarly, the interactions of season by year (P>0.17) and station by year (P>0.06) were not significant. However, stations varied significantly (P<0.01) within pastures.
The mean values of the Gini-coefficients for Lone Butte Pasture's cattle exclosure, cattle and game exclosure, and grazed transects 2 and 3 were all similar (P>0.05) during the course of the study (Figure 2). The mean value for Lone Butte Station 1 was different than Lone Butte Stations 2, 3, or cattle and game exclosure at a highly significant level (P<0.01).
The mean values for all stations in the Pole Creek Pasture were different (P<0.05) from each other (Figure 3). The Pole Creek cattle and game exclosure transects were not statistically different than those of Lone Butte Station 1, which were grazed. The rest of the stations in the Pole Creek pasture were different (P>0.05) from those in Lone Butte.
There was a pasture by year interaction (P>0.01) in the Gini-coefficient values for all stations. Long Creek's channel became 7.4 percent narrower and deeper in Lone Butte Pasture between 1992 and 1993, significant at the (P>0.05) level (Figure 4). Although the values for 1992 and 1993 were different from each other, they were not statistically different from 1991 or 1994. A negative 4.7 Gini percent change was observed between 1993 and 1994.
For the Pole Creek Pasture, Gini values for 1992 and 1994 were different (P>0.05) from each other, but were not statistically different than 1991 or 1993 (Figure 5). A 7.6 percent change was noted between 1991 and 1992. A negative 5.6 percent change occurred between 1992 and 1993, and another negative 3.6 percent change occurred between 1993 and 1994.
Review of changes in stream channel morphology between 1991 and 1994 indicates that the stream reach within the Lone Butte pasture did not appear to respond to the grazing management imposed upon it. Little or no change can be attributable to livestock use according to the information from Lone Butte Pasture's cattle and big game exclosure and stations 2 and 3. The same pattern existed between Pole Creek Pasture's cattle and game exclosure and its stations 2 and 3 (Figures 2 and 3).
The difference between Lone Butte Pasture's Station 1 and the rest of the sample stations located there (Figure 2) was due to its use by cattle as a stock crossing (18). Similarly, Pole Creek Pasture's Station 1 is located directly on a stock crossing (Figure 3). Based on statistical comparisons, these stations are not representative of the rest of the stream in either pasture. If these sites had been the only comparisons made against the protected reaches to determine if Forest Service standards were being met, one could easily have concluded that the grazing strategy was insufficient to protect bank stability.
Therefore, areas of abnormal impact should be avoided unless the objective is to study such areas (17, 4). Intense beaver and moose activity in the Long Creek area (6) create confounding factors in monitoring efforts and can add to the differences between sites, even within streams. This is evident in the Pole Creek Pasture (Figure 3) where all four stations responded differently.
Because Forest Service and national standards assume that all riparian areas respond to grazing or other disturbances in a similar manner, the variation noted among streambank monitoring stations along Pole Creek (Figure 3) as well as the variation between Pole Creek channel response and the response in Long Creek (Figure 2) cannot be overlooked.
Random selection of sampling locations can cause problems in monitoring, because streams are not random in variation but tend to cluster around variables based on physical and chemical laws (4). Similarly, abnormal impacts are probably not random in their variability, but are concentrated around different spatial variables. The variable response of streambanks to grazing may have resulted from the spatial variability of existing streambank vegetation conditions (3).
Monitoring stream channel systems is difficult because of the large amount
of variability within a stream system. Consequently, monitoring and
management efforts of stream channels need to be stream-type and reach specific due to
the large variation between sampling locations (9). Additional variation in
monitoring results may occur because many measure
ment techniques have not been adequately tested (12).
Riparian grazing standards could potentially work well in the Lone Butte pasture with the exception of "ramps" or stock crossings, because all stations responded in a similar fashion. However, Pole Creek, less than one mile away, showed a great deal of variability among the reaches. This makes national standards of questionable utility in development of grazing management plans.
Stream channel monitoring and management efforts should be directed toward specific areas of concern, rather than the grazing unit (allotment) as a whole. To work well in this context, riparian grazing standards must be developed for the specific reach or stream conditions. However, care should be taken because management at one critical location could lead to development of another problem at a different location, or create different problems at a later date (5). Widespread use of the same set of grazing standards over large areas and entire grazing allotments can not be considered a reliable basis for monitoring efforts because of the natural variation occurring within and between streams in the same watershed. The use of forest-wide or national standards may not be beneficial to either the natural resource base or local economic uses.
The authors wish to thank Matador Cattle Company, and the Sheridan Ranger District for their support in this research project. (References are after the following notes.)
Go to Current and Classic MSU-Bozeman Feature Stories Organized by Topic
You are the 12060th person to access this page.
References
1. Bauer, S.B., and T.A. Burton. 1993.
Monitoring Protocols to Evaluate Water Quality Effects
of Grazing Management on Western Rangeland Streams. USEPA 910/R-93-017.
2. Chaney, E., W. Elmore, and W.S. Platts.
1993. Managing Change: Livestock Grazing on Western Riparian Areas. Environ.
Protect. Agency, Region 8, Denver, CO.
3. Clifton, C. 1989. Effects of Vegetation and
Land Use on Channel Morphology. In:
Practical Approaches to Riparian Resource
Management: An Educational Workshop. Billings, Montana.
4. Harrelson, C.C., C.L. Rawlins, and
J.P. Potyondy. 1994. Stream Channel Reference Sites: An Illustrated Guide to Field
Techniques. USDA General Technical Report. RM-245.
5. Heede, B.H. 1980. Stream Dynamics:
An Overview for Land Managers. USDA Forest Service General Technical Report RM-72
6. Manoukian, M.E. 1994. Evaluation of
Tall Willows Within a Livestock Grazing Allotment in Southwest Montana. M.S. Thesis.
Montana State University, Bozeman, Montana.
7. Memorandum of Understanding. 1991.
Between the Matador Cattle Company, Montana Fish Wildlife and Parks, Montana State
University, and Forest Service, United States Dept.
of Agriculture.
8. Mueggler, W.F., and W.L. Stewart.
1980. Grassland and shrubland habitat types of
western Montana. Dept. of Ag. Forest Service, Int.
Mtn. Forest and Range Expt. Stn. Int-66
9. Myers, T.J., and S. Swanson. 1992. Variation
of stream stability with stream type and
livestock bank damage in northern Nevada. Water Resources Bulletin 28:743-754
10. Myers, T.J., and S. Swanson. 1991.
Aquatic habitat condition index, stream type, and livestock bank damage in northern
Nevada. Water Resources Bulletin 27(4):667-677
11. Olson-Rutz, K.M., and C.B. Marlow.
1992. Analysis and interpretation of stream
channel cross-sectional data. North American J.
Fisheries Manage. 12:55
12. Platts, W.S., C. Armour, G.D. Booth, M.
Bryant, J.L. Bufford, P. Cuplin, S. Jensen, G.W. Lienkaemper, G.W. Minshall, S.B. Monsen,
R.L. Nelson, J.R. Sedell, and J.S. Tuhy. 1987. Methods for Evaluating Riparian Habitats
With Applications to Managements. USFS General Technical Report INT-221
13. Platts, W.S. 1979. Relationships Among
Stream Order, Fish Populations, and Aquatic Geomor 14. Prichard, D., H. Barrett, J. Cagney, R. Clark,
J. Fogg, K. Gebhardt, P.L. Hanson, B. Mitchell, and D. Tippy. 1993. Process for
Assessing Proper Functioning Condition. USDI-BLM Riparian Area Management TR1737-9
15. Rupple, E.T. 1993. Geologic Map of the
Long Creek Grazing Allotment Beaverhead County, Montana. Montana Bureau of Mines
and Geology. Montana College of Mineral Science and Technology. Butte, Montana.
16. SAS Institute. 1988. SAS/STAT Guide
for Personal Computers. SAS Institute Inc. Cary, NC.
17. Skinner, Q.D. 1994. A Primer to Provide Help
in Understanding How Riparian Zones Function and Thoughts about Monitoring Natural
and Induced Change. University of Wyoming Cooperative Extension Service.
18. Trimble, S.W. 1994. Erosional Effects of
Cattle on Streambanks in Tennessee, U.S.A. Earth Surface Processes and Landforms. 19:451-464.
19. United States Department of Interior. Bureau
of Land Management. 1994. Rangeland Reform `94. Draft Environmental Impact
Statement. USDI Bureau of Land Management. Washington, D.C.
20. United States Department of Interior. Bureau
of Land Management. 1994. Rangeland Reform `94. Final Environmental Impact
Statement. USDI Bureau of Land Management. Washington, D.C.
21. United States Department of Agriculture.
Soil Conservation Service. 1973. Soil and
Vegetation Inventory of Near-Pristine Sites,
Montana. USDA Soil Conservation Service. Bozeman, Montana.
phology in an Idaho River Drainage. Fisheries 4(2):5-9