S Flathvad 3asln 333.9115For*st Practices, f=17fr water Quality and 1991 Fisheries Cooperative Pr ogra* Final report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program STATE DOCUMENTS eOLLKTION AUG 2 8 1991 MONTANA STATE LIBRARY 1515 E. 6th AVE. HELENA, MONTANA 59620 Final Report June 1991 About Tms Report This is the Final Report summarizing ten individual studies conducted for the Flathead Basin Forest PracticesAVater Quality and Fisheries Cooperative Program. The Cooperative Program was administered by a Coordinating Team representing the Montana Department of State Lands Forestry Division, the Flathead National Forest, Plum Creek Timber Company, L.P., the Montana Depart- ment of Fish, Wildlife and Parks, the Montana Department of Health and Environmental Sciences' Water Quality Bureau, the University of Montana, and the Rathead Basin Commission. The Cooperative Program's specific objectives were (1) to document, evaluate, and monitor whether forest practices affect water quality and fisheries within the Flathead Basin, and (2) if detrimental impacts exist, to establish a process to utilize this information to develop criteria and administrative procedures for protecting water quality and fisheries. The ten individual studies included the evaluation of: (1) specific practices at the site level, (2) accumulation of practices at the watershed level, (3) general stream conditions, (4) water quality variables relative to levels of management activity in small watersheds, (5) fish habitat and abundance relative to stream variables influenced by forest practices at the watershed level, (6) long- term changes in large-stream dynamics related to historical records of natural and man-related disturbances, and (7) changes in lake sediments relative to historical records of natural and man- related disturbances. Contributors U.S. Forest Service — ^Flathead National Forest Plum Creek Timber Company, L.P. Montana Department of State Lands Forestry Division Water Quality Bureau of the Montana Department of Health and Environmental Sciences Montana Department of Natural Resources and Conservation University of Montana Flathead Lake Biological Station School of Forestry Montana Forest and Conservation Experiment Station U.S. Department of Agriculture — Mclntire-Stennis Program Montana Department of Fish, Wildlife and Parks Flathead Basin Commission Montana Environmental Quality Council Montana Chapter of the American Fisheries Society Governor's Office, State of Montana Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Final Report June 1991 pubushed by Flathead Basin Commission 723 Fifth Avenue East Kalispell, Montana 59901 Abstract of the Final Report OF THE Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program The Flathead Basin Forest Practices/Water Quality and Fisheries Cooperative Program was initiated in 1988. It's goal was to evaluate the effects of forest practices on water quality and fisheries in the Flathead Basin. The Cooperative Program was administered by a Coordinating Team representing the Mon- tana Department of State Lands Forestry Divi- sion, the Flathead National Forest, Plum Creek Timber Company, L.P., the Montana Depart- ment of Fish, Wildlife and Parks, the Montana Department of Health and Environmental Sci- ences' Water Quality Bureau, the University of Montana, and the Flathead Basin Commission. The Cooperative Program's specific objec- tives were ( 1 ) to document, evaluate, and moni- tor whether forest practices affect water quality and fisheries within the Flathead Basin, and (2) if detrimental impacts exist, to establish a proc- ess to utilize this information to develop criteria and administrative procedures for protecting water quality and fisheries. Eight study leaders developed ten individ- ual studies (modules). They conducted these studies during 1989 and 1990. They held a coordination workshop early in 1989 to de- velop linkages among the studies and to obtain external review from three watershed and fish- eries scientists from Idaho and Washington. A set of study watersheds was selected at this workshop to help integrate as many of the field studies as possible. The ten individual studies included the evalu- ation of: (1) specific practices at the site level, (2) accumulation of practices at the watershed level, (3) general stream conditions, (4) water quality variables relative to levels of manage- ment activity in small watersheds, (5) fish habi- tat and abundance relative to stream variables influenced by forest practices at the watershed level, (6) long-term changes in large-stream dynamics related to historical records of natural and man-related disturbances, and (7) changes in lake sediments relative to historical records of natural and man-related disturbances. Each of the individual studies were docu- mented in "stand-alone" reports available through the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). The study leaders summarized their reports for inclusion in this Final Report.. Then the study leaders individually drafted conclusions and recommendations as a basis for discussion and interaction. The study team leaders held workshops during the spring of 1991 to review results and to develop a consensus set of summary conclu- sions and summary recommendations for con- sideration by the cooperating land management organizations. The land management organiza- tions then developed a formal response to the recommendations for inclusion in this Final Report. Following final editing and printing, this Final Report will be presented at a summer workshop for public information and response. Flathead Basin Cooperative Program Final Report Contents Introduction 1 Cooperative Study Program 3 Purpose and Specific Objectives 4 Nature of Project Results 4 Funding 5 Implementation 5 Structure of the Program 5 Information about this Report g Summaries of Individual Study Module Reports 9 Historical Perspective A. An Analysis of the Effect of Timber Harvest on Streamflow Quantity and Regime: An Examination of Historical Records (Hauer) 1 1 B. Evaluation of Historical Sediment Deposition Related to Land Use through Analysis of Lake Sediments (Spencer) 17 Water Quality and Fisheries C. The Effect of Timber Management on Stream Water Quality (Hauer and Blum) 41 D. Fisheries Habitat and Fish Populations (Weaver and Fraley) 51 E. Application of the Montana Nonpoint Source Stream Reach Assessment in the Flathead Basin (Tralles) 69 Evaluation of Forest Practices F. Assessments of Best Management Practices (Potts) 81 G. Management Guidelines for Riparian Forests (Pfister and Sherwood) 89 Flathead Basin Cooperative Program Final Report Pagei Contents Evaluation of Watersheds H. Application of the Sequoia Method for Determining Cumulative Watershed Effects in the Flathead Basin (Potts) 103 I. A Forest Management Nonpoint Source Risk Assessment Geographic Information Systems Application (Potts) 113 J. Linear Correlation/Regression Analysis of Forestry Models, Risk Assessment, and Water Quality and Fisheries Data (Sirucek, Hill, Hauer, Fraley, Weaver, Potts, and Tralles) 125 General Discussion 139 Summary of Conclusions 143 Summary of Recommendations 151 References 169 Appendixes 177 A. Memorandum of Understanding 179 B. Participants in the Flathead Basin Forest Practices/ Water Quality and Fisheries Program 185 C. Abies lasiocarpa/Oplopanax horridum h.t. (Subalpine Fir/Devil's Club h.t.) 1 89 D. Montana Forest Habitat Types 195 E. Data Set for Module J (Linear Correlation/Regression Analysis of Forestry Models, Risk Assessment, and Water Quality and Fisheries Data 199 F. Nonpoint Source Stream Reach Assessment Montana Water Quality Bureau Field Report Form 205 G. List of Flathead Basin Streams with Major Bull Trout Spawning ans/or Rearing Areas or Concentrated Use by Spawning Westslope Cutthroat Trout 2 1 3 Page ii Flathead Basin Cooperative Program Final Report Contents Figures and Tables Introduction Figure 1 : Linkages of study modules in Flathead Basin Study 8 B. Evaluation of Historical Sediment Deposition Related to Land Use through Analysis of Lake Sediments (Spencer) Figure B-1: Mean sediment accumulation rates in Whitefish Lake over the last 125 years, and timber harvest activity over the time period (since 1929) when harvest records could be assembled. 22 Figure B-2: Mean sediment accumulation rates in Swan Lake over the last 125 years, and timber harvest activity over the time period (since 1948) when harvest records could be assembled 30 Figure B-2: Mean sediment accumulation rates in Lake McDonald over the last 135 years 33 C. The Effect of Timber Management on Stream Water Quality (Hauer and Blum) Table C-1: Wilcoxson's signed-rank comparison of mean total suspended solids (TSS), three forms of nitrogen (NH,, NO,^, and total perslfate nitrogen — TPN), and two forms of phosphorus (soluble reactive phosphorus — SRP and total phosphorus — TP) between no to low activity watershed sites paired to moderate to high activity watershed sites 45 Figure C- 1 : Annual maximum total phosphotur (TP) and total persulfate nitrogen (TPN) at each of ten paired watershed sites. 46 Figure C-2: Summer maximum Chlorophyll a and ash-free-dry-masss (AFDM) at each of ten paired watershed sites. 48 Figure C-3: Autumn taxa richness and Simpson's index of riffle habitat macroinvertebrates at each of ten paired watershed sites 49 D. Fisheries Habitat and Fish Populations (Weaver and Fraley) Table D-1 : List of study streams showing activities scheduled for each stream included as part of the Flathead Basin Commission Cooperative Forest Practice, Water Quality, and Fisheries Study 54 Figure D- 1 : Watershed Study Sites 55 Figure D-2: Relationship between numbers of westslope cutthroat trout fry successfully emerging from replicates of six gravel mixtures and the percentage of material smaller than 6.35 mm in each mixture 57 Flathead Basin Cooperative Program Final Report Pageiii Contents Figure D-3: Relationship between number of bull trout fry successfully emerging from replicates of six gravel mixtures and the percentage of material smaller than 6.35 mm in each mixture 57 Figure D-4: Relationship between transformed substrate scores and juvenile bull trout densities (number of trout > 75 mm/100 m*) for 15 tributary reaches in the Flathead River Basin during 1989 60 Table D-2: Summary of 1989 rainbow trout spawning site inventories 60 Table D-3: Summary of westslope cutthroat trout spawning site inventories from 1986-1987 61 Table D-4: Summary of annual bull trout spawning site inventories between 1979 and 1989 62 Figure D-5: Summary of annual bull trout redd counts in the North and Middle Forks of the Flathead River Drainage from 1979 through 1989 63 Figure D-6: Summary of annual bull trout redd counts in the Swan River Drainage from 1982 through 1989 63 Table D-5: Comparison of total fish density, juvenile bull trout density, and juvenile cutthroat trout density (for those > 75 mm/l(X)m^ calculated from electrofishing estimates at 28 sites around the Flathead Basin during 1989 65 Figure D-7: Relationship between the arcsine transformations of the Sequoia index and transformed McNeil Core results for 28 watersheds in the Flathead Basin during 1989 66 Figure D-8: Relationship between the arcsine transformations of HjOY model output and transformed McNeil core results for 28 watersheds in the Flathead Basin during 1989 66 Figure D-9: Relationship between the arcsine transformations of the Sequoia index and transformed substrate scores for 28 watersheds in the Flathead Basin during 1989 '67 Figure D-10: Relationship between the arcsine transformations of HjOY model output and transformed substrate scores for 28 watersheds in the Flathead Basin during 1989 67 Application of the Montana Nonpoint Source Stream Reach Assessment in the Flathead Basin (Tralles) Figure E-1: Map: Completed NSPS Stream Reach Assessments Upper Flathead River Basin 73 Figure E-2: Nonpoint Source Stream Reach Assessment Overview 74 Table E- 1 : Streams Ordered by Impairment Value 75 Figure E-3: Impairment Values by Percentage of Total Streams (30) 76 Figure E-4: Impairment Values by Percentage of Total Reaches (95) 77 Page iv Flathead Basin Cooperative Program Final Report Contents Figure E-5: Assessment Categories Rated as Impaired by Percentage of Total Reaches (95) 77 Figure E-6: Impairment Values by Percentage of Managed Reaches (75) 78 Figure E-7: Impairment Values by Percentage of Unmanaged Reaches (20) 78 Table E-2: Summary of Management Activities as Actual or Potential Sources of Stream Problems 79 F. Assessments of Best Management Practices (Potts) Figure F-1: Percentage of timber sales with at least one major impact 84 Figure F-2: Percentage of timber sales with at least one minor impact 84 Figure F-3: Percentage of impacts by BMP category 85 G. Management Guidelines for Riparian Forests (Pfister and Sherwood) Table G- 1 : Habitat types observed on two or more of the BMP audit sites 93 Table G-2: Streamside Management Zone widths for various soil and slope classes 96 Table G-3: Soil erodibility based on parent material type (Cline and others 1981) 97 Table G-4: Evaluation of proposed guides (Table G-3) on 48 sites 97 Figure G- 1 : Field guide for locating the edges of the SMZ 98 Figure G-2: Montana "Risk" Matrix (3/91) 1(X) H. Application of the Sequoia Method for Determining Cumulative Watershed Effects in the Flathead Basin (Potts) Table H- 1 : Runoff Coefficients and Recovery Rates 1 06 Table H-2: Extent of Activities — Equivalent Acres 107 Figure H- 1 : S wan Watershed Map 1 08 Table H-3: Comparison of model results 1 10 I. A Forest Management Nonpoint Source Risk Assessment Geographic Information Systems Application (Potts) Figure I- 1 : Montana "Risk" Matrix (3/91) 117 Figure 1-2: The Howard Creek Watershed, Montana 118 Figure 1-3: Geology Map for Howard Creek 1 19 Figure 1-4: Checkerboaid Ownership in the Howard Creek Watershed 120 Flathead Basin Cooperative Program Final Report Page v Contents Figure 1-5: Timber Harvest Activity in the Howard Creek Watershed 121 Table I- 1 : "Risk" Calculations for Howard Creek, Lolo National Forest, Montana 123 J. Linear Correlation/Regression Analysis of Forestry Models, Risk Assessment, and Water Quality and Fisheries Data (Sirucek, Hill, Hauer, Fraley, Weaver, Potts, and Tralles) Table J-1: List of study sites by watetrshed and the application of water yield increase and/or sediment yield models (H^OY and WATSED), the Sequoia risk analysis, and the study sites that were investigated in the water quality module (C), the fisheries module (D), and the Water Quality Bureau's qualitative research assessment module (E) 129 Figure J- 1 : Watershed Study Sites 130 Table J-2: Independent and dependent variables applied to the linear correlation/regression analysis 131 Table J-3: Statistically significant regressions of independent and dependent variables 132 Figure J-2: Sequoia CRA to Transformed McNeil Results 134 Figure J-3: Sequoia CRA to Substrate Score Transformed 134 Figure J-4: Sequoia CRA to Mean TPN 134 Figure J-5: Sequoia CRA to Max Chlorophyll a 134 Figure J-6: H^OY WYl % to Transformed McNeil Core Value 135 Figure J-7: H^OY WYl % to Transformed Substrate Score 135 Figure J-8: H,OY WYl % to Mean TPN 135 Figure J-9: H,OY WYl % to Max Chlorophyll a 135 Figure J- 10: WATSED Annual WYl % to Mean TPN 135 Figure J- 1 1 : H,OY WYl % (75% Peak Duration) to Mean TPN 135 Figure J-12: WATSED Annual WYl % to Max ChlorophyU a 136 Figure J-13: H,OY WYl % (75% Peak Duration) to Max Chlorophyll a 136 Figure J-14: % of Basin Harvested to Mean TP 136 Figure J- 1 5 : % of Basin Harvested to Mean TPN 1 36 Figure J- 16: % of Basin Harvested to Max Chlorophyll a 136 Figure J-17: H,OY WYl % to Channel Stability Rating 136 Figure J- 18: Sequoia CRA to Channel Stability Rating 137 Figure J- 19: WATSED Annual WYl % to Change in Channel Stability Rating 137 Figure J-20: WATSED WYl % (75% Peak Duration) to Change in Channel Stability Rating 137 Figure J-21 : WATSED Predicted Sediment to Suspended Sediment 137 Page vi Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Introduction Flathead Basin Cooperative Program Final Report Page 1 Page 2 Flathead Basin Cooperative Program Final Report Introduction o, 'ver the past few decades, the make-up of the economy of the Flathead Basin has stead- ily changed. Once almost dominated by timber processing and agriculture, the area has experi- enced steady growth in recreation and tourism, retirement population, and more diversified manufacturing. People have been attracted by the Basin's considerable amenity values, un- crowded landscapes, a relatively mild temper- ate climate, and a clean environment. As popu- lations have grown, the concern to keep Flat- head Basin waters pristine has also increased, and forest management activities have come under more intensive public scrutiny. Many people have challenged logging and road con- struction activities as a perceived threat to water quality and fishery habitat. The concern over what may happen to state forested watersheds has been expressed through the passage of House Joint Resolution 49 by the Montana Legislature in 1987, appeals of na- tional forest plans, and increased public scru- tiny of timber management programs of State- owned and privately-owned lands. Timber man- agers and purchasers are in turn affected by the uncertainty, as planned operations are delayed and adverse economic impacts result. In response to these concerns the principal forest land managers in the Flathead River Basin (Flathead National Forest, Plum Creek Timber Company, L.P., and Montana Depart- ment of State Lands) proposed a cooperative effort to learn how forest practices arc affecting water quality and fisheries within the Basin. Cooperative Study Program The Flathead Basin Forest Practices/Water Quality and Fisheries Cooperative Program rep- resents a concerted and coordinated effort by state, federal, and private interests to work together to learn how forest practices are affect- ing water quality and fisheries, and to develop methods to utilize the findings in the manage- ment of Flathead Basin forests. The Cooperative Program was administered by a Coordinating Team representing the Mon- tana Department of State Lands Forestry Divi- sion, the Flathead National Forest, Plum Creek Timber Company, L.P., the Montana Depart- ment of Fish, Wildlife and Parks, the Montana Department of Health and Environmental Sci- ences' Water Quality Bureau, the University of Montana, and the Flathead Basin Commission. A representative of the Environmental Quality Council served as a liaison to the Cooperative. The Flathead Basin Commission was the nominal sponsor of the effort, providing logis- tical and staff support and serving as an "um- brella" organization under which the Coopera- tive Program operated. Each specific research project was under the direction of a scientific study leader, while resource specialists from various organizations provided technical assis- tance as needed. Citizen participation in the Cooperative Program was coordinated through the Flathead Basin Commission, with both for- mal and informal opportunities for the inter- ested public to review and comment on the Flathead Basin Cooperative Program Final Report Page 3 Introduction research design and findings. A formal Memo- randum of Understanding was signed by all the cooperating members in August of 1988. (See Appendix A.) Purpose and Specific Objectives Purpose: To improve the management of Flat- head area forested watersheds through the development and application of state-of-the-art information to prevent or mitigate the potential adverse ef- fects of specific forest practices on water quality and fisheries. Specific Objectives: • To document, evaluate, and moni- tor whether forest practices affect water quality and fisheries within the Rathead Basin; and, • If detrimental impacts exist, to establish a process to utilize this information todevelop criteria and administrative procedures for pro- tecting water quality and fisher- ies. Nature of Project Results The initial product of the research, monitor- ing, and analysis was a series of reports and scientific publications that define how certain forest practices affect water quality and fisher- ies. The second and most important product of the Cooperative Program will be the use of the research results in the management of Montana forests. This cooperative program included a vari- ety of modules. Some studies were conducted to obtain a better understanding of the overall interactions among forest practices, water qual- ity, and fisheries. Other studies were direct approaches to evaluate and provide new infor- mation on management practices. A variety of research methods were used, including histori- cal analysis of existing data, collection of new field data, field audits of management practices by teams of experts, formal summaries of ex- pert opinion, field assessment of environmental conditions, experimental work, and evaluation of models. Workshops were held during the program in an attempt to communicate and explore all possible opportunities to link studies together where possible and appropriate. The studies were conducted in cooperation with the land managers and resource manage- ment agencies. However, every attempt was made to have the researchers maintain inde- pendence of analysis and interpretation of re- sults. The organization of reports illustrates this process. The complete individual module re- ports stand alone as scientific documentation of findings of the individual studies. This report provides a summary of module results with the addition of specific conclusions. The administrative structure of the Coop- erative Program itself was conducive to trans- lating the results of the research into on-the- ground management. The participants included the major land managers and natural resource agencies in the upper Flathead Basin. These participants worked cooperatively throughout the study. The public has also been informed and involved in the Cooperative Program through the Flathead Basin Commission. The consequence of this structure was a shared "ownership" of the research results by all par- ties, and thus a substantial momentum toward understanding and improving forest manage- ment practices based on the research results. Page 4 Flathead Basin Cooperative Program Final Report Introduction Funding Structure of the Program The Coordinating Team worked closely with the Rathead Basin Commission to coordinate the logistics of funding. All participants in the Cooperative Program provided financial, tech- nical, and in-kind contributions. Each study component was funded sepa- rately and from a variety of sources. The Coop- erative Program approach facilitated funding assistance. Some study proposals incorporated ongoing projects already funded through Mcln- tire-Stennis Research Program funds and the Montana Riparian Association. A grant request was also received from the Renewable Re- source Development program administered by the Department of Natural Resource and Con- servation for partial support of the program. Implementation The scientist study leaders implemented the individual projects following thorough devel- opment of a study plan, peer review, approval of the project by the Coordinating Team, and allocation of research funds. Each scientist study leader conducted independent research accord- ing to the study plan. The Coordinating Team reviewed annual work plans, offered sugges- tions, encouraged scientific team efforts, and facilitated appropriate technical assistance from resource specialists working with the various organizations involved in timber management and oversight in the Flathead Basin. During the first year of the program, the module study leaders and technical resource specialists conducted a coordination workshop. A team of scientists provided a formal review of the individual cooperative program and offered suggestions for improvements. The participants identified linkages among modules and new modules were developed to provide missing information. The funding agencies agreed on nine sepa- rate "study modules." (A) An Analysis of the Effect of Timber Harvest on Streamflow Quantity and Regime: An Examination of Histori- cal Records. This study by the Flat- head Lake Biological Station statisti- cally examined the relationships be- tween water flow, weather, logging, and fire data for the basin dating back to the turn of the century. The final study report discusses the implications of past fires and timber harvests on stream flow in the Swan River and the North and Middle Forks of the Flat- head River. (B) Evaluation of Historical Sediment Deposition Related to Land Use through Analysis of Lake Sediments. To evaluate the impact of land use activities, floods, and fires on sediment deposition in several lakes in the Flat- head Basin over the last 150 years. The first lake (Whitefish Lake) is located in a watershed that has had extensive log- ging activity during the past century. The second lake (Swan Lake) is lo- cated in an area that has had a number of natural and human-related distur- bance activities. The third lake (Lake McDonald) is located in an area with no logging — although a major road was constructed during the middle part of this century. (C) The Effect ofTimber Management on Stream Water Quality. This was an in- stream study to measure any specific changes in aquatic ecology (biologi- cal, chemical, and physical character- istics) due to forestry practices and to Flathead Basin Cooperative Program Final Report Pages Introduction describe the implications of those changes. The streams evaluated in- cluded "control" streams in undevel- oped watersheds and test streams in watersheds from which timber has been harvested. The Flathead Lake Biologi- cal Station conducted this study. (D) Fisheries Habitat and Fish Popula- tions. This study, conducted by the Montana Department of Fish, Wildlife and Parks, examined cutthroat and bull trout habitat and how changes in stream- bottom sediment conditions are impor- tant to populations of these species. (E) Application of the Montana Nonpoint Source Stream Reach Assessment in theFlatheadBasin. This module evalu- ated impairment to beneficial uses in the Flathead Basin and evaluated the accuracy of the assessment procedure by comparing its results with the re- sults of quantitative studies performed by other Cooperative Program mem- bers. The Montana Department of Health and Environmental Sciences completed this assessment. (F) Assessments of Best Management Practices. The University of Montana School of Forestry conducted this inter- disciplinary team review of completed timber sales. The review evaluated the success of implementing forestry Best Management Practices (BMPs) forpre- venting soil erosion and protecting water quality. They examined fifty- three field locations on a mix of land ownership (federal, state, and private) within the Flathead Basin. (G) Management Guidelines for Ripar- ian Forests. This study module devel- oped management guidelines or "rec- ommended management practices" that are specifically tailored to the wetland vegetation types found along stream courses in the Flathead Basin. It also determined field procedures for con- sistent field identification of Stream- side Management Zones and a soil ero- sion risk matrix. (H) Application of the Sequoia Method for Determining Cumulative Water- shed Effects in the Flathead Basin. This study assessed possible cumula- tive effects in a watershed due to tim- ber harvesting and road building ac- tivities. The University of Montana School of Forestry conducted this study. (I) A Forest Management Nonpoint Source Risk Assessment Geographic Information Systems Application. This module developed a computer geo- graphic display capability for a water- shed risk assessment model based on a variety of practices with risk assess- ment for specific slopes and soils. This methodology will be used to help evalu- ate future management strategies for individual watersheds. The University of Montana School of Forestry con- ducted this study. In addition, the study participants decided that they needed a module to determine direct linkages between timber harvest activities and the presently used modeling and monitoring methods. (J) LinearCorrelationlRegressionAnaly- sis of Forestry Models, Risk Assess- ment, and Water Quality and Fisher- ies Data. Those working on this mod- ule provided the current modeling and monitoring data for testing relation- ships with data from other modules in the study, evaluated correlations of variables currently available fi-om the Page 6 Flathead Basin Cooperative Program Final Report Introduction A -HISTORICAL STREAM RECORDS V J D - FISHERIES HABITAT E - STREAM REACH ASSESSMENT F- SITE AUDITS OF BEST MANAGEMENT PRACTICES G-MANAGEMENT GUIDELINES, SMZ'S & RISK MATRIX H - SEQUOIA ACTIVITY ASSESSMB^ I - WATERSHED PRACTICES/RISK G.I.S. H/ETHOOaOGY J -CORRELATION OF MODELS, WATER QUALrrY DATA AND RSHERIES DATA Figure 1- Direct(^ — ►) and reference( ) linkages among study modules. Flathead Basin Cooperative Program Final Report Page? Introduction Rathead National Forest database, con- ducted correlation analyses for all se- lected variables, and jointly interpreted the results of the correlation analyses. The Flathead National Forest led this effort with the cooperation of several study leaders. Although the Flathead Basin Forest Prac- tices/Water Quality and Fisheries Study Coop- erative reports were designed as individual, independent modules, the study team made several efforts to develop coordination as the program progressed. Figure 1 demonstrates link- ages of the modules. Cooperative Program results will be re- ported in several ways. (1) The nine official study reports (Mod- ules A through I) have been printed for distribution to interested parties. (2) The study leaders have worked together to develop this summary report to syn- thesize individual study results and rec- ommendations into a consolidated document for public distribution. (3) Land owners have prepared a written response to the recommendations (which is incorporated in this Final Report). (4) The study results and management rec- ommendations will be presented to the public in a workshop hosted by the Flathead Basin Commission. This will include a discussion by the study lead- ers of their results and recommenda- tions as well as an interactive panel discussion between the three principal land managers and the public. (5) A videotape has been prepared for pub- lic television and organizational view- ing. The video details the Coopera- tive's goals, the study modules, the study results, and the management rec- ommendations. Information about Tras Report This report is organized into eight major sections: (1) Introduction, (2) Summaries of Individual Study Module Reports, (3) General Discussion, (4) Summary of Conclusions, (5) Summary of Recommendations, (6) Response of Major Forest Land Managers to the "Sum- mary of Recommendations," (7) References, and (8) Appendixes When reading the summaries of individual study module reports, be aware that the re- searchers used many different kinds of scien- tific data analyses. For those studies involving statistical analysis, the term "significant" means that the data show a relationship that can be stated with a degree of confidence. A probabil- ity level is usually shown in parentheses or with asterisks. For example: * = 90% level of confidence ** = 95% level of confidence *** = 99% level of confidence. For regression or correlation analyses, the probability value is commonly used to indicate that a relationship exists between the variables (for example, p < 0.01 means at the 99 percent level, p < 0.05 means at the 95 percent level, and p < 0.10 means at the 90 percent level). Often the r on* value is also included as a measure of variation of points about the regression line. In any statistical test, the sample size is important in determining "statistical signifi- cance." Furthermore, statistical interpretations relate to a population. Therefore use caution when evaluating a specific site or population. Pages Flathead Basin Cooperative Program Final Report I Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Summaries of Individual Study Module Reports This section of the Final Report summarizes studies published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^^^ Page 1 0 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module A: An Analysis of the Effect of Timber Harvest on Streamflow Quantity and Regime: An Examination of Historical Records By F. Richard Hauer This section of the Final Report summarizes a study of the same name published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report S® P^Se 1 2 Flathead Basin Cooperative Program Final Report An Analysis of the Effect of Timber Harvest ON Streamflow Quantity and Regime: An Examination of Historical Records By F. Richard Hauer' Introduction The piupose of this module was to deter- mine the effect, if any, of timber management practices on streamflow quantity or regime. The analyses were based on Flathead Basin historical records of: (1) streamflow (United States Geological Survey) (2) precipitation, temperature, and snow pack (National Weather Service), (3) timber harvest quantity, location, and type (Flathead National Forest, De- partment of State Lands, and Plum Creek Timber Co.), and (4) historical fire records (Flathead Na- tional Forest and Glacier National Park). I analyzed streamflow quantity and regime from 22 gauging sites located throughout the Flathead Basin for changes that may be attrib- uted to timber harvest. These analyses were conducted in light of natural interannual vari- abilities of climate, past fire history, and current logging practices. I focused the study on deter- mining possible changes in four components of annual streamflow: (1) discharge response to specific rainfall events that are distinct and separate from the influence of water stored as snow pack, (2) spring runoff quantity or regime (for example, height, breadth, and timing), (3) annual discharge to annual precipita- tion relationships, and (4) the relationship of annual maximum and minimum discharge. This report does not attempt to detail every analysis that was conducted. Many of the com- parative statistics that were run, particularly on untransformed data (that is, data which did not account for interannual variability in tempera- ture, precipitation, or snow pack), were insig- nificant due to high interannual variability Fur- thermore, some of the databases, particularly streamflow on very small creeks, spanned rela- tively short time periods (typically three years), which did not permit long term comparative analyses. Results Results of the analyses revealed that: (1) There was no significant correlative relationship between recorded rainfall events during the summer and fall (that is, not affected by snow pack) and increased river discharge during the 'Dr. F. Richard Hauer is a Research Associate Professor with the Hathead Lake Biological Station in Poison, Montana. Flathead Basin Cooperative Program Final Report Page 13 An Analysis of the Effect of Timber Harvest on Streamflow Quantity and Regime same time period. This was attributed to insufficient weather recording sta- tions rather than to an uncorrelated relationship. Mountain precipitation is currently unmeasured outside of accu- mulated snow pack during winter. Thus high precipitation events are underesti- mated by valley precipitation record- ings. (2) When the climatic factors that drive spring runoff are accounted for by com- paring years of similar temperature re- gimes and the data are transformed based on available snow pack, the ac- cumulated volume of the spring runoff of Hathead Basin Rivers occurs earlier in the runoff period in years since tim- ber management compared to years prior to such management. Simply stated, spring runoff waters are coming into the Flathead Rivers earlier in the year today than prior to extensive tim- ber harvests. (3) Comparison of long term trends of an- nual maximum and minimum discharge relationships (Qmax:Qmin) suggest that the Middle Fork of the Flathead River, which has experienced relatively little logging, is becoming increasingly stable in its QmaxrQmin relationship while the North Fork, which has had significantly more timber harvest, is becoming increasingly variable in its Qmax:Qmin relationship. In other words, the North Fork is trending to- ward a discharge pattern of higher maximum flows compressed over a shorter time period while the Middle Fork is tending toward a discharge pat- tern in which peak flows are lower and the higher discharge of spring runoff extends over a longer time period. This is particularly important in light of the decrease in fires in both drainages, which presumably would result in a longer runoff period. Thus, the Middle Fork, which had similar area to that of the North Fork involved in forest fires prior to the advent of broad scale fire prevention (that is, since the 1930s), has been progressing to increased dis- charge stability, while the North Fork has decreased in discharge stability since the advent of extensive logging. Conclusions The analyses conducted within the auspices of this study provide some insight into the effects of timber harvest on streamflow. How- ever, they were limited by the lack of long-term data collection among the smaller watersheds. All of the long-term databases of streamflow in the Flathead Basin are restricted to the major streams and rivers (for example, North Fork, Middle Fork, Swan River, etc.). This is a par- ticularly important consideration since it is the lower order watersheds that are likely to re- spond most dramatically and quickly to altera- tion of the landscape. Nonetheless, important conclusions can be drawn from this study. (1) The North Fork and Middle Fork drain- ages experienced several large fires between 1 880 and 1930. Subsequently, fire suppression efforts have greatly reduced forest fire frequency and size. (2) The North Fork and Swan River drain- ages have experienced a substantial quantity of timber harvest in compari- son to that of the Middle Fork drainage. Most of this timber harvest has oc- curred since 1950. (3) The accumulation of spring runoff waters has occmred earlier in the North Page 14 Flathead Basin Cooperative Program Final Report An Analysis of the Effect of Timber Harvest on Streamflow Quantity and Regime Fork and Swan River drainages during (4) The maximum discharge of the spring the past two decades compared to pre- runoff has demonstrated a general trend 1950 runoff patterns and when com- toward increasing in the North Fork pared to trends that have occurred in and decreasing in the Middle Fork, the Middle Fork during the same time periods and the same comparison years. Flathead Basin Cooperative Program Final Report Page 1 5 Page 16 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quauty and Fisheries Cooperative Program Study Module B: Evaluation of Historical Sediment Deposition Related TO Land Use THROUGH Analysis of Lake Sediments By Craig N. Spencer This section of the Final Report summarizes a study of the same name published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^^^ Page 1 8 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition Related to Land Use Through Analysis of Lake Sediments By Craig N. Spencer' Introduction One of the biggest environmental concerns with regard to timber harvest is the potential for enhanced erosion and transport of sediment to surface waters. Increased sediment loadings are considered undesirable as they may degrade gravel spawning habitat used by stream fish (see Weaver and Fraley, Module D). Further- more, since sediments represent a major source of water-borne nutrients (Mortimer 1941; Perry and Stanford 1982) increased erosion and sedi- ment transport may accelerate the eutrophica- tion process in surface waters, especially in lakes and reservoirs. Current debate over the impact of logging activities on water quality in the Flathead Basin is hampered by a scarcity of quantitative data on conditions prior to the commencement of tim- ber harvest. Early logging activities began 50 or more years ago in many parts of the Basin. However, most water quality monitoring ef- forts in the Basin were initiated only within the last 10 to 20 years or less. Without pre-logging water quality data, it is difficult to assess the impact of harvest activity on water quality in the Flathead Basin. Previous studies in other areas have docu- mented increased sediment loadings to surface waters resulting from timber harvest activities (Likens et al. 1970, Lowe et al. 1986). Never- theless, it may not be appropriate to extrapolate findings from other areas to the Flathead Basin. A number of streams in the B asin are flanked by steep, naturally occurring, unstable stream banks composed of sand, silt and clay. The presence of these erosive deposits may result in naturally elevated sedimentation rates in the Basin. Thus it is possible that erosion of sediments associ- ated with logging activities in the Flathead Basin may be minor in comparison to natural sediment loadings. The vast majority of suspended stream sediments carried into large deep lakes are deposited within the quiescent lake environ- ment. Significant changes in surface erosion and sediment transport within the catchment of a lake should be reflected in changes in the sediment character and the rate of sediment accumulation in the lake (Berglund 1 986). Thus, undisturbed sediments deposited on the bottom of lakes contain a record of the past history of sedimentation from their respectivecatchments. Modem paleolimnological techniques are avail- able which allow estimation of past sedimenta- tion rates through detailed analysis of the "paleo" record preserved in the lake sediments (Ber- glund 1986). A record of past changes is most evident in sediments deposited in the deep- water (profundal) region of the lake. This envi- ronment is much more stable than near-shore, shallow lake, or stream environments; deep lake sediments may remain largely undisturbed 'Dr. Craig N. Spencer is a Research Assistant Professor with the University of Montana, Flathead Lake Biological Station in Poison, Montana. V J Flathead Basin Cooperative Program Final Report Page 19 Evaluation of Historical Sediment Deposition for thousands of years. One of the more notable paleolimnological studies, by Hutchinson and others ( 1 970), docu- mented increased sedimentation rates in a lake in Italy (Lago di Monterosi) over 2000 years ago, coinciding with construction of the Roman Road, the Via Cassia, in about 171 B.C. Numer- ous studies report changes in sedimentation rates in other lakes which correlate with various land disturbance activities including timber harvest, plowing of fields, and road building (Davis 1975, Batterbee and others 1985, see reviews in Berglund, 1986). Although there have been previous paleolimnological studies in the Flathead Basin (Moore et al. 1982), none have quantified recent sedimentation rates. The present research was initiated to study the historical record of sediment deposition over the last 100-150 years in three lakes in the Flathead Basin. Whitefish Lake and Swan Lake are located in watersheds that have a history of land disturbance activity (primarily road build- ing and timber harvest) throughout much of this century. Lake McDonald is located in Glacier National Park, and its watershed has not been logged. However, a road was constructed from the bottom to the top of the catchment during the middle part of this century. Methods Sediment cores were collected from the deep central basin of each lake using a freeze- coring technique modified from Shapiro (1958). Sampling locations were located on flat areas of the lake bottom away from steeper regions which may be subject to underwater landslides or slumps which would distort the sediment record. The sediments deposited in the deep central regions of the study lakes consist of fine sediments composed largely of silt and clay sized particles. The cores were sectioned in horizontal slices (approximately 1 cm thick), and the date of deposition was estimated for each section. Dates were established using a naturally occurring radioisotope (^'°Pb) which decays at a known rate (half life=23 years). By measuring the activity of ^'°Pb in each section, it was possible to establish a time line of sedi- ment deposition dates along the length of each core (Appleby and Oldfield 1983). In undis- turbed sediments, this technique has been shown to be useful in dating sediments deposited up to 150 years ago. An independent technique was used for an alternate estimate of the location of the year 1963, along the sediment core. Atmospheric testing of atomic weapons peaked in 1963, and previous studies have documented a peak in '"Cs activity in sediments deposited in 1963, due to global fallout of the radioactive decay particles (Pennington and others 1973). Com- parison of dates indicated close agreement be- tween the two dating methods in the study cores. Once chronological ages were determined, sediment accumulation rates (mg dry wt/cmV yr) were estimated. These rates were calculated by multiplying the width of the core section (cm) by the dry weight density of the section (mg/cm') and dividing by the length of time (years) spanned by the section. These rates represent mean sedimentation values which occurred over the time span of each 1 cm thick section of sediment. One core from each lake was chosen for de- tailed analysis. Ideally, several cores would have been analyzed from each lake; however, the available budget was not sufficient for multiple core analysis. Although only one core was analyzed in detail from each lake, the selected cores appeared to be representative of whole-lake conditions in Whitefish and Lake McDonald. For example, distinct patterns of horizontal banding noted in a particular lake Page 20 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition core also were visible in the same relative location in other cores collected from the same lake. Since no banding was observed in cores collected from Swan Lake, visual cross-com- parisons between cores was not possible. It is important to note that the present analy- ses were not designed for quantitative estab- lishment of whole-lake sediment budgets over the last 150 years. Thus the study results should not be used to estimate the total sediment load- ing to the study lakes over the period of record. Such analyses would require analyses of a number of cores collected from various parts of each lake. Rather, the present analysis simply utilized the stable deep-lake sediments as a continuous monitor of relative changes in depo- sition of fine sediment from the watershed to the lake environment. Thus, the absolute sedimen- tation rates are of less interest that the relative change in these rates over time. Results and Discussion Whitefish Lake The Whitefish Lake catchment has been subject to a number of land disturbance activi- ties (both man-induced and natural) which may have influenced the sedimentation rate in this lake. I assembled a history of major land distur- bance activities in the catchment, including natural and human-related activities. These various activities will be chronologically com- pared with the lake sedimentation rates for evaluation of potential causal relationships. Natural Disturbances There were four years in which fires burned 500 acres or more in the watershed during the period of record. The greatest acreage burned in 1910, when 5562 acres burned, representing 6.7% of the total watershed area. Data collected in this study do not show evidence of large changes in lake sedimentation following any of these fires (Figure B-1). The sedimentation rate did increase during the time interval of the 1910 fire, and some of this in- crease may have been attributed to fire. How- ever, human land activities (described later) also occurred in the basin at this time; thus specific fire effects are difficult to discern. Moreover, no obvious ash layer was visible in the lake core around 1910. The mean sedimen- tation rates shown in Figure B-1 were in the midst of decade-long periods of decline during the time period of two other fires (1919, 1937). Sedimentation rates increased slightly during the period of the 1926 fires; however, extensive human disturbance of the watershed commenc- ing in the late 1920s complicates determination of actual cause and effect. Nevertheless, a thin layer of black ash is clearly visible in the sedi- ment core at a depth corresponding to the time period around 1926. (See photo in complete report.) This distinct ash layer undoubtedly resulted from transport of ash from the 1926 fires into the lake (either from the air or via streams). One of the 1926 fires, called the Hellroaring Fire, burned down to the shore of Whitefish Lake, and extended up into the Whitefish Range into the area that subsequently became the Big Mountain Ski Resort. Nonethe- less, changes in the sediment record resulting from the 1926 fire appear short-lived, and of small magnitude, compared to other distur- bance events described later. Flooding is another natural disturbance activity which may influence lake sedimenta- tion rates. It is well known that over the course of a given year, the of majority sediments car- ried by Rocky Mountain streams are trans- ported during spring run-off. Thus during Flathead Basin Cooperative Program Final Report Page 21 Evaluation of Historical Sediment Deposition Timber Harvest 1990 «Insert Figure B-l.» (5 Yr cumulative amortized acreage) 0 1000 2000 3000 4000 5000 1980 Flood— 1970- Flood— I960- 1950 Fire- 1940 Lazy Crk, Swift Crk logging begins- Fire- -1930 Fire- Logging declines- Fire" Logging along lakeshore Railroad construction' J 920 -1910 ~1900 Flood- First sawmill on Whitefish lake" ~1890 1880 1870 1860 ^^3=— Whitefish Lake 0 100 200 300 Mean Sedimentation Rate ,2 (mg/cm^/yr) Figure B-1. Mean sediment accumulation rates in Whitefish Lake over the last 125 years, and timber harvest activity over the time period (since 1929) when harvest records could be assembled. This latter activity is expressed as a 5 year cumulative acreage, with previous 4 years acreages amortized using the Flathead National Forest new road sediment delivery coefficients. Page 22 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition unusually large spring flood events, one might expect to observe increased sedimentation rates in the study lakes. Three major spring floods have been observed in the Flathead Basin dur- ing the period of record. These occurred in 1894, 1964 and 1974. The largest of the three occurred in 1894, when severe spring floods produced the highest flood waters ever recorded in the Rathead Basin. Whitefish Lake was raised to its highest recorded level, 9.5 feet above the low water mark (Schafer and Engel- ter, 1973). The sedimentation rate estimated for the core section spanning 1894, was 26 mg/cmV yr compared to 20 mg/cmVyr during the previ- ous period. For a flood of such magnitude, this increase in sedimentation rate was relatively small in comparison to subsequent changes. Furthermore, the 1894 flood was not the only land disturbance event in the watershed. Early land clearing and timber harvest activities commenced in the catchment in the late 1880s and early 1890s. The first sawmill on on the Whitefish River, just below the lake, was built in 1891. Nevertheless, in 1894, the vast major- ity of land in the Whitefish Lake watershed remained undisturbed by human activities. Thus it appears that major floods in undisturbed watersheds have a relatively small impact on deposition of fine sediments in lakes in com- parison to extensive human disturbance activi- ties (discussed later). More detailed analysis of the lake sediment record over shorter time increments (<1 year) would likely reveal increased sediment deposi- tion for a short time following major flood events such as 1894. However, such increases were not sufficient to elevate the mean sedi- mentation rates over the longer intervals meas- ured in this study. Shorter water retention time in lakes during flood events may lessen their impact on lake sedimentation rates. Neverthe- less, the majority of sediments entering White- fish Lake are deposited in the lake, even during spring run-off events (Golnar 1985). The 1964 flood was the next most intense flood event recorded in the valley. The sedi- mentation rate in Whitefish Lake reached levels up to 88 mg/cmVyr during this time period; however, there were significant timber harvest activities (described later) occurring in the basin over the same time interval (Figure B-1). Thus it is difficult to isolate the effects of the flood. The smallest of the three floods occurred in 1974. As with the 1964 flood, it is difficult to determine the effect of this flood on lake sedi- mentation rates. The sedimentation rate was well above background levels during this pe- riod; however, signiflcant timber harvest ac- tivities occurred during this time period as well. The 1964 and 1974 floods washed out numerous culverts and several bridges in the Swift Creek drainage. Observations by local foresters (T. Vars, D. Klehm) indicate that widespread surface erosion did not appear to have taken place in the Whitefish Lake water- shed during these fioods. Nevertheless, the floodwaters were reported very muddy and carried high levels of suspended sediment. These foresters speculated that increased sediment loads associated with these floods likely came from transport of sediment previously depos- ited in the stream channels, together with ero- sion of unconsolidated banks which are promi- nent along Swift Creek. More detailed discus- sion of these floods will follow in subsequent coverage of human disturbance activities. Another natural disturbance event which may have influenced the sedimentation rate in Whiteflsh Lake was the 1980 eruption of Mt. St. Helens. This eruption produced a fallout of volcanic ash across western Montana. There is no visible ash layer in the study cores, and lake sedimentation rates did not appear to increase during this time period. Nevertheless, it is pos- sible that the most recent sedimentation rates would have been lower in the absence of Flathead Basin Cooperative Program Final Report Page 23 Evaluation of Historical Sediment Deposition volcanic ash deposition, ^yman Di^t^rbanggs 1865-1929. Distinct correlations between natural disturbance events and changes in fine sediment deposition in Whitefish Lake are dif- ficult to make. By contrast, correlations be- tween human activities in the basin and lake sedimentation are more readily apparent. Prior to the arrival of the earliest Europeans settlers in the late 1880s, the Whitefish Lake watershed was heavily forested and virtually undisturbed by human activity. Human activities in the watershed apparentiy were limited to an occa- sional small Indian fishing encampment near the outlet of Whitefish Lake (Trippett 1956, Schafer and Engelter 1973). Mean sedimenta- tion rates in Whitefish Lake during this pre- settlement period were the lowest recorded over the 125 year period of record. Between 1865 and 1886, the estimated mean sedimentation rate was approximately 20 mg/cm7yr (See Figure B-1). Due to the low sedimentation rate during the pre- settlement period, the mean sedimenta- tion rate shown in Figure B-1 spans a relatively long time interval (21 years). Annual sedimen- tation rates likely varied within this long inter- val. However, I have no evidence to suspect that large changes occurred during the period from 1865-1886. In the first place, the mean sedi- mentation rate from 1865-1886 was compa- rable to the subsequent rate from estimated between 1886 and 19(X). Second, if some un- known, natural disturbance event occurred and resulted in large short-term changes in the seem- ingly stable, pre-settlement sedimentation rate then one would expect to see changes reflected in the visual character of the sediment core. No distinct visual changes were evident in the core over the 30 year pre-settlement period. The sedimentation rate increased slightly during the period from 1886-1900, concurrent with the 1894 flood and early logging activity around Whitefish Lake. The earliest European settlers arrived in the watershed in the late 1 880s (Schafer and Engelter 1973). These early residents were primarily trappers and lumber- men. A sawmill was constructed at the outlet of Whitefish Lake in 1891. Small logging opera- tions were concentrated at the southeast end of the lake. Trees were cut, and pulled to the lake with horses whereupon they were floated down the lake, or pulled on sleds across the ice to the sawmill. This early logging activity around the lake may have contributed to the slight increase in sedimentation rate recorded between 1886 and 1900. The 1894 flood also may have con- tributed to this increase. The sedimentation rate increased dramati- cally in the early 1900s, reaching a rate of 155 mg/cmVyr, representing a 7-8 fold increase over background levels. This increased sedi- mentation rate coincides with increased human activities in the basin including construction of a railroad line for the Great Northern Railroad along the entire 7 mile southern shoreline of the lake. The terrain along the south lakeshore is steep in places, thus considerable earth moving, and road bed leveling occurred along the lake- shore. Rail preparations along the lake also included blasting through bedrock, excavation of a tunnel near the head of the lake, and construction of a trestle and filling of Beaver Bay on the lake's southwest shore. Most of the construction occurred over a several year pe- riod prior to the official opening of the railroad in 1904. There is little doubt that substantial amounts of sediment were dumped and/or washed into Whitefish Lake during construction of the rail line. This activity likely contributed to the sedi- mentation increase measured in the lake core during this period. In addition, early logging activity around the lake during this time period Page 24 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition also may have contributed to increased sedi- ment deposition in the lake. Unfortunately, I have found no comprehensive early logging records which allow estimation of land areas or timber volumes involved in the Whitefish Lake basin. The lake sedimentation rate declined fol- lowing construction of the railroad. Neverthe- less, the deposition rate from 1902 to 1908 was 62 mg/cmVyr, which is 3 times higher than background levels. Following completion of the railroad along the lakeshore, timber harvest was the primary human land disturbance activ- ity in the Whitefish Lake watershed. Logging demand was fueled by the need for railroad ties as well as building materials for the new town of Whitefish (then called Stumptown), established in 1904. The Whitefish townsite, established in 1904, was located along the Whitefish River below the lake outlet. Thus clearance of land for the townsite did not likely contribute additional sediments to Whitefish Lake. A small sluice dam was operated at the outlet of Whitefish Lake for 10-15 years during this period, and lake level fluctuations associated with this dam may have increased shoreline erosion around the lake in the early 1900s.. During the early 1900s, logging activity in the watershed was limited to the immediately vicinity of Whitefish Lake and the adjacent railroad line. Little logging apparently occurred in upstream drainages since Swift Creek and other tributaries to Whitefish Lake were not suitable for carrying logs. Logging above Whitefish Lake in those days consisted of selec- tive cutting of the larger trees in the immediate vicinity of Whitefish Lake, which were pulled to the lake using horses, and then floated (or skidded across the ice) to the sawmills (Schaf- fer and Engelter 1973). By 1904-1905, there were several sawmills in the area including the one on Whitefish Lake near the outlet, one downstream on the Whitefish River, and sev- eral others farther downstream on the Stillwater and Flathead Rivers. Logging camps sprung up around Whitefish Lake, and along the White- fish River below the Lake. The lake sedimentation rate increased to 88 mg/cmVyrfrom 1908 and 1912. In addition to human land disturbance described above, the 1910 fires may have contributed to this in- crease. The sedimentation rate declined to 29 mg/ cmVyr from 1912 to 1922. This decrease corre- lates with a decline in land disturbance activity in the watershed. Desirable timber around the lake apparently was depleted by earlier logging activities, and logging efforts shifted to other areas of the Flathead Basin. 1929-Present. There was a large increase in sedimentation in Whitefish Lake for a short time period in the early 1930s (Figure B-1). Sedimentation levels increased 10-fold over background, reaching levels up to 212 mg/cmV yr. This large peak corresponds with extensive logging activities which commenced in 1929 in the Lazy Creek and Lower Swift Creek drain- ages above the head of Whitefish Lake. Most of this acreage was located on private lands owned by the Glacier Park Timber Company (which was subsequently incorporated into the Burling- ton Northern Company and then split off into Plum Creek Timber Company). A railroad spur was constructed from the head of Whitefish Lake up the Lazy Creek drainage during this time period. Trees were cut, pulled to the rail line using horses, and transported to the mills via railcar. The Lazy Creek railspur was operated for three years, and then removed in 1932 (D. Klehm, pers. comm.). Thereafter, logs were hauled out of the Lazy and Swift Creek drainages by truck. Logging roads and skid trails were constructed through- out the areas being logged. During this period, there apparently was little concern about water quality impacts, and few, if any, regulations Flathead Basin Cooperative Program Final Report Page 25 Evaluation of Historical Sediment Deposition existed. Stream crossings were unrestricted, culverts were used infrequently, and "cordu- roy" roads were constructed across wet areas simply by clearing the trees and then placing timbers, side by side, across the wetlands, cre- ating a road bed. These and other timber-related activities, including construction of the rail- spur, likely were major causes of enhanced erosion and sediment transport to surface wa- ters. The situation likely was compounded by large stream flows in the catchment during the spring of 1932 and 1933 (Schaffer and Engelter 1973). Comparatively low sedimentation rates following the large flood of 1 894 suggest mini- mal increases in sediment deposition in the lake from an undisturbed watershed. However, high stream flows in the early 1930s likely acceler- ated the erosion and transport of sediment from areas disturbed by various logging-related ac- tivities in the watershed. The correlation between logging activities and increased sedimentation in the early 1930s is striking and leaves little doubt concerning a cause and effect relationship between these two events. Thereafter, lake sedimentation rates declined in the mid to late 1930s in concert with declines in timber harvest activity . During the mid- 1930s the Civilian Conser- vation Corps (CCC) completed a road in the Whitefish Lake watershed which extended around the east side of the lake and up through the Swift Creek drainage. This road was used extensively for log hauling and appears to be the only other significant human-related land dis- turbance activity which may have affected sedimentation during the 1930s. However, the amount of land disturbance caused by the CCC road was small in comparison to area impacted by the network of roads, skid trails, and railspur used in the timber harvest efforts. The 1937 fire also may have contributed to sedimentation during this period. However the impact of this fire is believed small, due to the limited aerial extent of the fire (750 acres), and the fact that it left no visible ash layer in the lake sediments (unlike the ash layer corresponding to the 1926 fire). During the 1940s there was little timber harvest or road building activity in the catch- ment. Concurrent with the decline in timber harvest and associated activity, sedimentation rates in the lake declined to 42.5 mg/cmVyr. A road was built in the Whitefish Lake watershed up to the Big Mountain ski resort in 1947. Although there has been considerable develop- ment at the ski area, the vast majority of this activity is located in the Haskill Creek drainage, which is outside the Whitefish Lake Basin. Logging and associated road building ac- tivity in the Whitefish Lake watershed com- menced again in earnest in 1948, primarily on the Stillwater State Forest in the Swift Creek drainage. Harvest activities peaked around 1950, declined again, and then increased through the mid- 1 960s. This pattern of logging activity was correlated with similar changes in lake sedi- mentation (see Figure B-1). Lake sedimenta- tion rates peaked at 72 mg/cmVyr around 1950, declined to 52 mg/cm7yr in the mid 1950s, and then increased over a ten year period, peaking at 87.8 mg/cm7yr in the mid-1960s. During the two decade period from the late 1940s through the 1960s, timber harvest and accompanying road and skid trail construction occurted across large parts of the watershed on Stillwater State Forest Lands, and beginning in the early 1960s on Burlington Northern lands. During the first part of this time period, activi- ties were concentrated on the gently sloping valley bottoms. However, by the 1960s harvest and associated road building activities had moved up into steeper, more erosive areas. This twenty year period of timber harvest and asso- ciated activity appears cortelated with sediment deposition in Whitefish Lake. Page 26 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition The 1964 flood occurred during the latter part of this trend. The sedimentation rate did not increase dramatically following this flood, the second largest on record. Rather, the sedimen- tation rate continued an increasing trend which was initiated in the 1950s. Although the 1964 flood likely contributed to increased erosion and sediment transport in the basin, the impact of this flood on lake sedimentation appears small in comparison to previous land distur- bance activities. From 1967-1971, the sedimentation rate declined to 72.6 mg/cm7yr, concurrent with a decline in logging activity in the late 1960s. From 1971 to 1983, the lake sedimentation rate was reduced to 54.7 mg/cm7yr. Although tim- ber harvest was elevated at the beginning of this period, harvest declined to low levels during the latter part of this period. The 1974 flood also occurred during this period. Foresters report that the 1964 and 1974 floods did not cause extensive erosion and gullying throughout the Whitefish Lake watershed. Rather, these floods resulted in considerable erosion and channel destabilization in the main stream courses and floodplains. Although the flood of 1894 sug- gests that flooding of undisturbed watersheds may not result in greatly increased sedimenta- tion rates compared to human disturbance ac- tivities, it is unclear whether the same may be said for floods occurring in watersheds dis- turbed by timber harvest. For example, it is possible that the floods of 1 964 and 1 974 flushed out sediments from the larger streams which had accumulated there as a result of past activi- ties. Between 1983 and 1990, the mean sedi- mentation rate was 52.2 mg/cm7yr. Timber harvest activities increased during this period; however, as in the 1970s, this increased harvest was not accompanied by comparable increases in sedimentation as in previous years. In addi- tion to timber harvest activities, two other fac- tors may have contributed to changes in lake sedimentation in recent years. First, the 1980 eruption of Mt. St. Helens produced a fallout of volcanic ash across western Montana. There is no visible ash layer in the study cores, and lake sedimentation rates did not appear to increase during this time period. Nevertheless, sedi- ments were deposited in the basin as a result of this eruption. Second, recent increases in lake- shore housing and other developments along Whitefish Lake may have contributed sedi- ments to the lake. Considerable areas of the shoreline still remain undeveloped. Although lake sedimentation rates did not increase during recent years, it is possible that observed sedi- mentation rates would have been lower in the absence of volcanic ash deposition and lake- shore development. There are several factors which may ex- plain the reduced sedimentation rate in White- fish Lake during the 1970s and 1980s. First, this time period coincides with significant efforts on the part of government resource manage- ment agencies and the timber industry to at- tempt to reduce the impact of timber harvest activities on erosion and sediment transport to surface waters. A combination of mandatory and voluntary standards were adopted in an attempt to reduce the sedimentation risk. These efforts focused on minimizing erosion associ- ated with road construction, stream crossings, and restricting harvest activities on the most sensitive lands. The sedimentation data provide evidence that these more recent logging prac- tices may have reduced the rate of sediment transport to Whitefish Lake, in comparison to previous timber harvest activities in the basin. Although the recent reduction in lake sedi- mentation may support the effectiveness of newer logging practices in reducing sediment transport, this observation must be tempered by several important observations. First, the recent sedimentation rates are still well above Flathead Basin Cooperative Program Final Report Page 27 Evaluation of Historical Sediment Deposition background sedimentation rates estimated for the period prior to human settlement. Second, the 1980s have been characterized by a series of relatively mild run-off years. It is possible that subsequent flood events could dislodge sedi- ments deposited in the flood plain and/or pools in the stream channel during the 1980s, and carry them into the lake. If this occurs, then lake sedimentation rates would increase, as these sediments, eroded during recent years, were finally transported into the lake. Another factor may have contributed to the reduced lake sedimentation rate observed since the early 1970s. Recent timber harvest in the Whitefish Lake watershed has been concen- trated on bottom lands where the erosion poten- tial is reduced compared to the steeper, more erosive lands in the upper regions of the water- shed which were logged during the 1960s. In fact, a significant portion of the land logged in the 1 980s lies on Plum Creek Timber Company lands in the Lazy Creek drainage, and repre- sents harvest of timber regrown since the area was first logged in the 1930s. Reduced sedi- mentation also may be due to the fact that new road construction, a major source of sediments, was reduced somewhat in the 1980s due to availability of pre-existing roads in the drain- age. Nevertheless, a portion of the recent log- ging activity lies on previously unlogged, steeper terrain (located on Stryker Ridge). It is clear that recent logging efforts concen- trated in the lower portion of the Basin have had less impact on sedimentation in Whitefish Lake, compared to the original roading, logging, and rail spur construction on many of the same areas in the early 1930s which produced a 10- fold increase in lake sedimentation. However, it is not clear from these data whether the recent logging practices including Best Management Practices (BMPs) have reduced sediment load- ings in comparison to activities in the 1 950s and 1960s. There is some evidence to suggest this; however before drawing definitive conclusions one must sort out the relative importance of improved practices together with other poten- tial causal factors such as use of pre-existing roads, a series of comparatively mild run-off years, and other factors which may affect sedi- mentation such as lakeshore housing develop- ment and the eruption of Mt. St. Helens. Fur- thermore, it is not clear if BMPs have been utilized for a long enough period to see any effect on lake sedimentation. Swan Lake The history of the Swan Lake Basin over the last 120 years includes a number of natural and human-related disturbance activities. There were large floods in 1894 and 1964. As in the other lakes, neither of these floods appeared to be accompanied by large changes in lake sedimen- tation. However definitive conclusions about the 1894 flood are speculative in Swan Lake since this flood occurred within the oldest dated core section. There were no large fires in the Swan Basin during the period of record. Rela- tively small fires did occur in 1910, 1919, 1936 and 1963. The largest of these occurred in 1919 when several thousand acres burned, represent- ing less than 1% of the Swan Lake drainage area. There were no visible ash layers in the sediment cores. Further discussion of natural disturbances is incorporated in the following chronological discussion of changes in the Swan Lake Basin. The lowest sedimentation rate measured in Swan Lake over the 120 year period of record occurred from 1874 and 1899, prior to human disturbance activities in the basin (Figure B-2). The mean sedimentation rate during this period was 16 mg/cmVyr. During the first two decades of the 1900s, the sedimentation rate increased slightly to 19 mg/cmVyr. There are a number of factors which may have contributexl to this Page 28 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition increase. First, the dating precision for these early dates is less certain than for more recent dates. Thus, this small increase in lake sedimen- tation may not be significant. Nevertheless, there were a number of land disturbance activi- ties in the basin in the early 1900s which could have contributed to an increase in lake sedimen- tation. A few small homesteads began to appear in the basin in the early 1900s. Cattle were brought into in the upper part of the basin near Condon around 1910; however, early cattle ranching efforts were abandoned in 1912 (Art Whitney personnel communication). Wildfires in 1910 fire burned a small portion of the basin above Condon. A road was built along the north shore of Swan Lake in 1916-1917. In addition, fires in 1919 bumed several thousand acres near the head of Swan Lake. Finally, early logging ac- tivity commenced near the head of Swan Lake in 1914. During the late teens and early 1920s, logging activities intensified. A 3 mile-long railroad spur was built from the head of Swan Lake up to S. Lost Creek. Logs were cut and then pulled by horses down to South and North Lost Creeks. In addition, a sluice dam was built on South Lost Creek during this period (Art Whitney personnel communication). Water was released by blasting out the dam during spring run-off and large numbers of logs were swept downstream into the Swan River and subse- quently into Swan Lake. The mean lake sedimentation rate increased to 33.4 mg/cmVyr for the period between 1920 and 1933. This increase represents a doubling over background levels during pre-setdement years. There are a number of factors which may have contributed to this increase. Timber har- vest activities including log drives and rail operation continued near the head of Swan Lake through the early 1920s. Log drives on S. Lost Creek and the Swan River likely caused considerable stream-bank erosion. The fires of 1919 and construction of a road along Swan Lake also likely increased sediment transport to the lake. Given the present information, it is not possible to estimate the relative importance of these various factors in contributing to increased lake sedimentation. The actual causes likely involve a combination of factors. Some of the early land disturbance activi- ties such as rail-line construction and early log drives appear to have preceded the lake sedi- mentation increase by several years or more. Delays in lake sedimentation could have been caused by the nature of the Swan River above Swan Lake. The Swan River passes through a broad meandering delta area which extends for several miles above the lake. This appears to be an area of sediment deposition. As such, this area could buffer, dampen, or delay transport of sediments from the river into the lake. The lake sedimentation rate declined to 27 mg/cmVyr from the mid- 1930s to the mid- 1940s. This decline corresponds with a decline in land disturbance activity in the basin. From the mid- 1920s to 1940, there was virtually no timber harvest or road construction in the Swan Lake basin. Fires burned a relatively small portion of the upper basin in 1936. The mean sedimentation rate increased to 37 mg/cmVyr for the time period between 1 946 and 1957. This increase corresponds with con- struction of the Swan Highway as well as a re- sumption of timber harvest activities. Work on the new highway began in the late 1940s and in- cluded relocation of portions of the old road closer to Swan Lake. In particular, this work included bank excavation and filling along the lakeshore just north of the Swan Lake camp- ground. By 1956, the road extended past the head of the lake, up to Goat Creek, some 15 miles above Swan Lake. The road reached the head of the drainage in the late 1950s. In addition, substantial logging and associ- ated road building activity commenced again in Flathead Basin Cooperative Program Final Report Page 29 Evaluation of Historical Sediment Deposition Timber Harvest • — • (5 Yr cumulative amortized acreage) 0 2000 4000 6000 8000 « Insert Figure B-2» Flood. Fire 1990- 1980 1970- 1960- Swan Highway Constructed 1950^ Logging resumed — 1940 1 Fire — 1930 Logging declines 1920 • Road built along Swan Lake, Fire — Log Drives & Railroad Logging begins Fire — 1910- 1900-1 flood — 1890 1880-1 1870 10 20 Mean Sedimentation Rate ,2 (mg/cm'^/yr) Figure B-2. Mean sediment accumulation rates in Swan Lake over the last 125 years, and timber harvest activity over the time period (since 1948) when harvest records could be assembled. This latter activity is expressed as a 5 year cumulative acreage, with previous 4 years acreages amortized using the Flathead National Forest new road sediment delivery coefficients. Page 30 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition the Swan Lake drainage in the 1940s. Although I was unable to obtain accurate timber harvest estimates for the 1940s, the level of harvest was comparable to the 1950s (Art Whitney, per- sonal communication). Logging roads and skid trails were built in the lower portion of the basin during this period. Timber harvest and associated road build- ing activities accelerated during the 1960s and lake sedimentation rates remained elevated during this period, reaching a mean rate of 38 mg/cmVyr from 1957 to 1972 (Figure B-2). The 1964 flood also occurred during this time interval, and these flood waters may have accel- erated sediment transport to the lake. However, as in the other two study lakes, this large flood did not appear to be correlated with a large increase in lake sedimentation. A small fire in 1963 burned less than 1(XX) acres in the upper portion of the Lost Creek basin and also may have contributed to lake sedimentation. The lake sedimentation rate increased again to 49.5 mg/cmVyr from 1972 to 1990. This sedimentation rate was the highest recorded over the 120 year period of record, some three- fold higher than background rates estimated for the late 18(X)s. This recent sedimentation in- crease is correlated with a large increase in logging activity and associated road building in the Swan Basin (Figure B-2). During the 1980s, timber harvest activities doubled over the levels attained during the 1960s and 1970s. During this period, timber harvest and associated road building expanded on bottom lands as well as on the flanks of the Mission and Swan mountain ranges that enclose the Swan Valley. Results from Module H (Potts 1991) indicated that a few areas in the Swan River drainage had an unusually high concentration of timber harvest activities. Much of this increased harvest oc- curred on private lands owned by Plum Creek Timber Company. Timber harvest on State and Federal lands declined slightly during this same time period. Although timber harvest and related road building represent the largest human land dis- turbance activity in the basin in recent years, there has been an increase in construction of lakeshore cabins and homes around Swan Lake. This activity also may have contributed to in- creased lake sedimentation. Nevertheless, large areas of lake shoreline remain undeveloped. Comparisons between Swan Lake and Whitefish Lake yield additional insight into the impact of lakeshore development on lake sedi- mentation rates. The lakeshore area around Whitefish Lake also has experienced home and recreational development. Recent development around Whitefish Lake may even exceed that around Swan Lake. However, in contrast to Swan Lake, recent sedimentation rates in White- fish Lake do not appear to have increased. Thus it is unlikely that lakeshore development is responsible for the recent large increase in sedimentation in Swan Lake. Rather, the con- trasting lake sedimentation responses appear more closely related to differences in recent timber harvest and road building activities in the two basins. As described earlier, recent timber harvest activities in the Swan Basin far exceed past acreages subject to harvest in the Basin. By contrast, recent timber harvest acre- ages in the Whitefish Basin fall within the level of activity reached several times over the last 60 years. As in Whitefish Lake, the lack of large flushing flows in recent years may have re- sulted in substantial accumulation of sediments in the river system. This could be more of a factor in Swan Lake than Whitefish Lake, given the low gradient depositional area in the river immediately above Swan Lake. If so, then subsequent flushing flows could carry large quantities of previously eroded material into the lake. Flathead Basin Cooperative Program Final Report Page 31 Evaluation of Historical Sediment Deposition Lake McDonald As in the other study lakes, the lowest sedimentation rate in Lake McDonald was re- corded at the beginning of the period of record. Between 1880 and 1910, the sedimentation rate was 7 mg/cm7yr (Figure B-3). The sedimenta- tion rate increased to 10 mg/cmVyr between 1910 and 1935, and then increased substantially to 29 mg/cm7yr over the period between 1935 and 1945. These increases follow construction of a two lane highway, called the Going to the Sun Road, which passes through a large part of the Lake McDonald watershed. This major road runs along the south shoreline of Lake McDonald for much of its 9 mile length, and continues up the drainage along McDonald Creek for 11 miles, closely bordering the creek in numerous locations. The road then switches back up the along steep exposed terrain within the Lake McDonald catchment leading up to Logan Pass on the Continental Divide. Roadbed prepara- tion included construction of numerous em- bankments along steep areas, blasting tunnels through bedrock, and considerable earth mov- ing activities, all of which undoubtedly contrib- uted sediments to surface waters. Although the initial road was completed in the early 1930s, the sedimentation rate in Lake McDonald did not peak until the late 1930s. This apparent lag in lake sedimentation may have resulted from a delay in transport of sedi- ments from the upper portion of the watershed down into the lake. Such delays could be due to the long distance between road building activi- ties on the erosive slopes near the continental divide and Lake McDonald. In addition, the heavily forested streams in the McDonald Creek Basin, may have higher sediment retention rates compared to logged watersheds. Natural down- fall in the stream bed serve as stream sediment traps. Furthermore, unlogged watersheds may have smaller maximum stream flows compared to logged watersheds, which could result in reduced sediment flushing capacity in undevel- oped watersheds (see Hauer, module A). Thus there are a number of factors which may have contributed to the apparent delay in sediment deposition in Lake McDonald. The sedimentation rate declined rapidly in the 1960s. Revegetation and stabilization of the original road cuts likely reduced sediment de- livery to surface waters along the road. This reduction also may have been partially due to stabilization of the road surface by paving in the 1950s. It is possible that periodic regrading of the road together with road dust stirred up by cars along this heavily travelled road may have contributed to elevated sedimentation levels in Lake McDonald. After the road was paved, the potential contribution of road dust to lake sedi- mentation would have been greatly reduced. The mean sedimentation rate for the 29 year period from 1961 to the present declined to 14 mg/cm7yr. This rate is roughly twice the rate estimated for the late 18(X)s. Reasons for the continued existence of sedimentation levels above background are speculative. The 1980 eruption of Mt. St. Helens left no visible band of sediments in the core. However, it is possible that this eruption caused increased sediment deposition in Lake McDonald. Thus, it is pos- sible that in absence of this eruption, the most recent sedimentation rate may have been closer to levels measured in the late 1 8(X)s. In addition, there have been limited human activities in the basin which could have contributed to recent sedimentation rates. These activities include ongoing maintenance on the Going to the Sun road, as well as limited construction projects around the lake. However, future plans call for significant "improvement" along much of the Going to the Sun road. Although the Going to the Sun Road repre- sents the largest human land disturbance activ- ity in the Basin, there has been some other Page 32 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition «Insert Figure B-3» 1990' 1980- 1970H Fire — Flood — Road paved- 1960 1950- Fire — Initial road completed — ■ 1940- 1930- Fire- Road construction begins — 1920- 1910- 1900 Flood — 1890- 1880 10 20 30 _i 40 Lake McDonald — r- 30 0 10 20 Mean Sedimentation Rate ,2 ' — I 40 (mg/cm'^/yr) Figure B-3. Mean sediment accumulation rates in Lake McDonald over the last 135 years. Flathead Basin Cooperative Program Final Report Page 33 Evaluation of Historical Sediment Deposition limited development around the lake. This in- cludes construction of several lodges along the lakeshore as well as small lakeshore cabins limited to the extreme east and west ends of the lake. Much of this activity occurred in the early part of this century, and does not appear to have noticeably influenced mean lake sedimentation rates shown in Figure B-3. Natural disturbance events did not appear to be closely correlated with major changes in sedimentation rate in Lake McDonald. As in the other study lakes, there were no large fires in the basin during the period of record. One of the larger fires in the watershed occurred in 1967, when approximately 5% of the watershed burned. In addition, the 1964 flood occurred during this interval. Sedimentation rates during the time interval following the flood and fire remained well below those achieved during the 1 940s and 1 950s. Sedimentation rates may have increased for a short time after these events, however any such increases were dampened out by subsequent sedimentation rates. Smaller fires in 1926, 1937 may have contributed to increased sedimentation in Lake McDonald; however, the impact of these natural distur- bances are likely masked by the road building activities. None of the McDonald Basin fires or floods left visible bands in the sediment core during the 110 year period of this study. However, a thick black band is visible in the Lake McDonald core at a depth of 1 8- 19 cm, which is below the oldest strata dated in this study (see complete report for details and photograph). This band may correspond to extensive fires which burned during the year 1 735. Several large fires burned portions of the Lake McDonald Basin that year, including much of the steep landscape leading down to the shoreline of Lake McDonald (Barrett 1988). Sedimentation rates must have increased dramatically in Lake McDonald following this large fire, given the appearance of this thick (1 cm) ash layer in the lake sediments. Unfortu- nately, I was unable to estimate the actual sedimentation rate during this period since current sediment dating techniques are not available for close estimation of dates over 1 50 years old. Management Implications This study shows that past land disturbance activities are correlated with increased sedi- ment deposition rates in all three study lakes. Sedimentation rates increased 3 to 10 fold over background levels, corresponding with logging, road building, and/or railline construction in the watersheds. Results of the BMP audits revealed that the greatest number of BMP deviations were related to road drainage and road maintenance(Potts, Module F). Lake sediment analyses suggest that over time, the sediment contribution from roads (such as the Going to the Sun Highway) is greatly reduced. Recent sedimentation rates in Whitefish Lake provide equivocal evidence for reduced sedimentation rates using newer forest practices. However, sedimentation rates in Swan Lake reached their highest estimated level within the last 15-20 years concurrent with a doubling of timber harvest activities in the basin in the last 10 years. Results from Potts (Module H) indicate that a few areas in the Swan drainage had an unusually high concentration of harvest activi- ties. Water quality violations resulting from recent timber harvest activities also have oc- curred in the Swan Basin. Therefore, if recent forest practices employed in the 1970s and 1980s do indeed reduce sediment delivery compared to older practices, then any such improvement appears to have been offset by the recent large expansion of harvest activities, at least in the Swan Lake Basin. In addition, sedi- ments resulting from past activities may still be in transit in the river systems above Swan and Page 34 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition Whitefish Lakes. Since increased sedimentation may have negative impacts on both stream and lake envi- ronments the large increases in sediment load- ings documented in this study represent an im- portant environmental concern regarding both past and projected activities. Hauer and Blum (Module C) reported that increased timber har- vest activities were correlated with increased suspended sediments, nutrients, and algal growth in streams. These results are consistent with studies of timber harvest in other areas (Likens et al. 1970, Lowe et al. 1986). Weaver and Fraley (Module D) showed that increased lev- els of fine sediments in fish spawning areas in streams may substantially reduce the spawning success and viability of bull trout and westslope cutthroat trout. These species are native to the Flathead Basin and have been designated sensi- tive species by the U.S. Fish and Wildlife Serv- ice and species of special concern by the state of Montana. Unfortunately there are nocomparable tech- niques available for quantifying historical changes in the sediment composition in impor- tant salmonid spawning and rearing areas of streams. However, since the majority of lake sediments transported to lakes enter via streams, past large increases in erosion, transport, and subsequent deposition of fine sediments in the lake environment would likely have been ac- companied by increased deposition of sedi- ments in portions of the stream channels. Thus I expect that past increases in sedimentation documented in this study had negative impacts on stream ecosystems in these watersheds. The fact that estimated sedimentation rates in Swan Lake reached their highest levels within the last 15-20 years raises concerns about the potential negative impact of increased sedimentation in important bull trout and Westslope cutthroat trout streams above Swan lake, and the effect of future land disturbance activities on these streams. Another concern regarding increased sedi- mentation in surface waters is undesirable stimu- lation of algal productivity and lake eutrophica- tion. Sediments represent a major source of nutrients to surface waters (Mortimer 1941; Perry and Stanford 1982). Data from the Flathead Basin show a close correlation between sus- pended sediment concentrations in streams and stream nutrient (phosphorus and nitrogen) concentrations (Spencer and Hauer 1991, Ellis and Stanford 1988, Stanford and Ellis 1988, Stewart 1983, Golnar 1985). Detailed nutrient budget analyses from Whitefish and Flathead Lakes indicate that 60-70% of the annual phos- phorus and nitrogen loadings come from stream inputs, with the bulk of this input associated with turbid spring run-off and unrelated to point source inputs (Stanford and Ellis 1988, Stewart 1983, Golnar 1985). Lakes serve not only as sediment traps, but also nutrient traps. Golnar (1985) estimated that 74% ofthe phosphorus entering Whitefish Lake was retained in the lake. While some nutrients may become permanently buried in the lake sediments, a portion of the nitrogen and phos- phorus pool entering the lake environment remains in the water column. In studies on Flathead Lake, Dodds, Priscu, and Ellis (1991) showed that phosphorus and nitrogen could be recycled in the water column in a matter of hours or less. Thus, past increases in nutrient loadings are still likely affecting the lake eco- systems. Phosphorus and nitrogen availability have been shown to be the primary factors limiting algal production in lakes in the Flathead Basin (Dodds and others 1989, Spencer and Ellis 1990). Stanford and Potter (1976) and Perry and Stanford (1982) hypothesized that stream sedi- ments in the Flathead Basin, upon entering the lake environment, may settle out and strip phosphorus and algae from the lake water Flathead Basin Cooperative Program Final Report Page 35 Evaluation of Historical Sediment Deposition column. However Stewart (1983) disproved this hypothesis and concluded that algal pro- duction in Flathead Lake appeared to be stimu- lated by sediment additions. Furthermore, con- trolled bioassay experiments demonstrate that addition of sediments from a variety of loca- tions in the Flathead Basin stimulate algal growth (Perry and Stanford 1982, Ellis and Stanford 1988). Ahhough the bioavailability of phos- phorus contained in turbid spring run-off in the basin may be only 6% (Ellis and Stanford 1988), this still represents the largest single nutrient source to Flathead Lake (Stanford and Ellis 1988). Thus, enhanced erosion and sediment transport, as documented in this study, have undoubtedly contributed to lake eutrophica- tion, an undesirable process resulting from stimulation of algal growth, reduced water clar- ity, oxygen depletion and other related prob- lems (Wetzel 1988). At present, Whitefish Lake is in a transi- tional state between oligotrophy and mesotro- phy (Golnar 1985). Late-summer hypolimnetic oxygen depletion already is occurring in the lake. Golnar (1985) concluded that Whitefish Lake lies near a critical threshold of "exces- sive" phosphorus loading, as determined from the nutrient loading model of VoUenweider and Kerekes (1980). Other lakes in the Flathead Basin are threatened by increased nutrient load- ings. Flathead Lake also is undergoing eutro- phication as evidenced by increased algal pro- duction (Stanford and Ellis 1988). As with Whitefish Lake, scientists have described Flathead Lake as being on a threshold, such that increased nutrient loadings seriously threaten water quality in the lake (Bahls 1986, Stanford and Ellis 1988). Other lakes in the Flathead Basin including Ashley Lake and Lake Mary Ronan develop summer hypolimnetic oxygen depletion and are also at risk from increased nutrient loadings. During 1990, dissolved oxygen levels in Swan Lake declined to 0.5 mg/L near the bot- tom of the south basin (see data in the complete report). This appears to be the lowest dissolved oxygen measurement recorded in any of the large lakes in the Flathead Basin which are noted for their high water quality. There are smaller seepage-type lakes in the basin (for example. Echo, Loon, and Foy's Lakes) which have more severe dissolved oxygen depletion, together with algal blooms and reduced water quality which characterize these more produc- tive lakes. Two limnological studies of Swan lake conducted in the mid-1970s also reported dissolved oxygen depletion in Swan Lake but not nearly to the extent of the 1990. Unfortu- nately, the measurements made in the 1970s were taken from shallower depths than in 1990. Although the 0.5 mg/L measurment made in 1990 is significantly lower than any previously recorded level, there is insufficient data to say with any certaintly that DO concentrations have declined significantly since the 1970s. Nevertheless, reduction of hypolimnetic oxy- gen levels to near anaerobic conditions in Swan Lake (regardless of the past history) is surpris- ing and alarming, especially given the short water residence time in the lake. The existing data documenting substantial oxygen depletion in Swan Lake have led to the lake being described as an impaired lake (Loren Bahls, Montana Water Quality Bureau, per- sonal communication). Reduced oxygen con- centrations undoubtedly exclude trout and other aquatic organisms from portions of the lake in late summer and fall. Evidence from numerous scientific studies indicate that if oxygen con- centrations decline just a little bit more in Swan Lake, than one can expect a rapid increase in available phosphorus in Swan Lake (Mortimer 1941, Wetzel 1988). This would be caused by the release of sediment-bound phosphorus into the lake water when oxygen concentrations decline to 0 mgA- at the sediment-water Page 36 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition interface. If this happens, the eutrophication process would accelerate rapidly, leading to a deleterious cycle of further oxygen depletion, which in turn stimulates more widespread re- lease of sediment-bound phosphorus into the lake water. Such changes would likely lead to serious declines in water quality, including nuisance algal blooms, poor water clarity, and degradation of fisheries habitat. This negative scenario of events has been documented in numerous lakes in other parts of the country, frequently fueled by increased human activities and development in the lake basin. The USEPA (1976) estimated that 99.7% of the total phosphorus load to Swan Lake was from non-point sources. Unfortunately, there is insufficient data available at present to allow compilation of a comprehensive nutrient budget for Swan Lake in 1990, or to fully explain the cause of the dissolved oxygen depletion in the lake. There may be other important sources of nutrients and/or organic loadings to the lake that also have contributed to current water quality conditions (e.g. shoreline homes, upstream development, and natural sources). However, regardless of the actual cause, the reduced oxygen levels in Swan Lake raises serious concerns about any future increases in nutrient loadings to the lake. . Available data provide evidence that in- creased sediment loadings have negatively impacted stream and lake resources in the Flathead Basin. Symptoms of aquatic resource degradation include elevated fine sediment levels in key bull trout and westslope cutthroat trout spawning streams (see Weaver and Fraley , Module D) and increased sediment loadings to lakes. The cause-and-effect relationship be- tween increased fine sediment loading (from whatever source) and spawning habitat degra- dation is well understood. However, given the limited data available on nutrient budgets from the study lakes, it is more difficult to assign a direct causal relationship between increased sediment loadings and lake water quality. Nevertheless, there is no question that sedi- ments represent a significant source of nutrients to lakes in the Flathead Basin. Increased nutri- ent loadings to lakes typically lead to increased productivity, and given sufficient productivity, water quality problems such as periodic oxygen depletion and algal blooms. The present study provides evidence for a link between past human land disturbance ac- tivities (primarily related to timber harvest and road building) and increased fine sediment (and nutrient) loadings to lakes. These data provide evidence from the Flathead Basin that human disturbance activities increase fine sediment deposition to a greater extent than natural dis- turbance events. Similar conclusions have been drawn from studies in other regions (Hutchin- son and others 1970, Davis 1975, Batterbee and others 1985, see reviews in Berglund, 1986). Considerable efforts are presently being made to reduce the input of nutrients to lakes in the Flathead B asin, in an attempt to maintain the high water quality which characterizes many of its lakes. Examples of nutrient control measures being employed include a ban on the sale of phosphate detergents in 1985 and nearly $20 million dollars towards construction and/or expansion of wastewater treatments plants and collection facilities for phosphorus removal. New, or upgraded treatment plants, are either in place or under construction for all major com- munities upstream from Flathead Lake. Similar efforts should be directed at other controllable nutrient sources in the basin such as timber harvest and related road building and road maintenance. Flathead Basin Cooperative Program Final Report Page 37 Evaluation of Historical Sediment Deposition Conclusions ( 1 ) Lake coring analyses indicated that past human land disturbance activities were correlated with increased fine sediment deposition up to 10- fold in Whitefish Lake, 4 to 5-fold in Lake McDonald, and 3-fold in Swan Lake. (2) Lake McDonald (a) Initial road construction and up- grading of the Going to the Sun Road from Lake McDonald to the continental divide at Logan Pass during the 1930s and 1940s were followed by substantial increases in sediment deposition in Lake McDonald. (b) After the road was paved in the early 1950s the sediment deposi- tion rate in Lake McDonald de- clined substantially; however, sedimentation rates still remain above background levels. (3) Whitefish Lake (a) Large increases in sediment depo- sition occurred during the early part of this century (1900-1910) and were attributed to railroad construction along the lakeshore, logging activity around the lake, and the 1910 fires. (b) The largest sedimentation in- creases occurred in the early 1 930s when substantial logging and as- sociated road and railline construc- tion were concentrated in the Lazy Creek drainage and Lower Swift Creek, near the head of Whitefish L«J^9- (c) Sedimentation rates also were elevated near 1950 and again in the 1960s. These increases were largely attributed to substantial logging and associated road build- ing activity, which extended to upper portions of the Whitefish Lake drainage. (d) Recent logging activities in the Whitefish watershed appear to have had less impact on lake sedi- mentation that past activities. Possible explanations for reduced sediment impacts include use of pre-existing roads, logging on less- erodible lands, improved logging and road building practices, and a series of comparatively mild run- off years. (4) Swan Lake (a) Sedimentation rates increased during the 1920s following a number of land disturbance ac- tivities including road construc- tion, fires, and timber harvest ac- tivities that included sluice dams, log drives, and rail line construc- tion. (b) Sedimentation rates increased again in the 1950s in concert with a resumption of timber harvest and road building activities. (c) From the early 1970s up to the present, the lake sedimentation rate reached its highest level. This increase occurred as timber har- vest intensified, more than dou- bling the previous maximum har- vest level in the basin. Page 38 Flathead Basin Cooperative Program Final Report Evaluation of Historical Sediment Deposition (5) Results from the three study lakes suggest that roads represent the great- est disturbance activity resulting in increased fine sediment transport and deposition in the downstream lakes. Once constructed, the sediment contri- bution appears to decline. (6) Changes in deposition of fine sedi- ments directly attributed to natural stream banks, floods, fires, and other natural erosion processes during the past 150 years were much smaller that changes attributed to human distur- bance activities in these two water- sheds. Previous speculation that ero- sion of naturally unstable stream banks and other natural sources may mask sediment inputs attributed to human activities appear unfounded with re- spect to fine sediment deposition in Whitefish Lake, Swan Lake, and Lake McDonald. Flathead Basin Cooperative Program Final Report Page 39 Page 40 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quauty and Fisheries Cooperative Program Study Module C: The Effect of Timber Management ON Stream Water Quality By F. Richard Hauer and Christopher O. Blum This section of the Final Report summarizes a study of the same name published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^8® Page 42 Flathead Basin Cooperative Program Final Report The Effect of Timber Management ON Stream Water Quality By F. Richard Hauer^ and Christopher O. Blum^ Introduction The primary purpose of this study was to determine if past forest management practices (logging and related roads) have affected stream- water quality. Based on this primary purpose, we established the following goals: (1) to evaluate techniques that would be useful in examining the effects of for- est management practices on stream water quality, (2) to implement studies using those tech- niques to evaluate water quality and the conditions of streams affected by forest practices, and (3) to establish a baseline from which a longer monitoring plan might be de- veloped. We defined stream- water quality as includ- ing physical, chemical, and biological variables of stream ecosystems, but did not include issues specifically associated with fisheries manage- ment. We founded the hypotheses of this study on the concern that timber harvest and associ- ated activities may result in increased sediment and nutrient loading to streams. These physical and chemical factors may, in turn, affect the basic structure and function of the stream food web through increased stream-algae produc- tion and changes in stream invertebrates. Specific study objectives were to examine streams in watersheds with different levels of timber harvest and to determine if there have been measurable changes in: (a) sediment transport or nutrient concen- trations, (b) the accumulation of attached algae ex- pressed as surface density of Chloro- phyll a and ash-free-dry-mass, and (c) the structure and function of the stream invertebrate community. We chose 12 stream sites for study based on differing levels of timber management, stream size, and basin characteristics. Three study stream sites had no timber harvest or road building in the basin, five study sites were identified as having low to moderate activity in the drainage above the site, and four sites were distinguished as having a high level of timber harvest and roads in the watershed. Ideally, a study designed to answer the question, "Do current forest management practices affect stream water quality?" would entail a water sampling frequency of 15 to 20 times per year and seasonal sampling for algal growth and invertebrates over a 3 to 5 year period. How- ever, we were limited in this study by both time and financial resources. Nonetheless, we found 'Dr. F. Richard Hauer is a Research Associ- aieProfessor with the Flathead Lake Biological Station in Poison, Montana. K!hristopher O. Blum is a Research Assistant with the Flathead Lake Biological Station in Poison, Montana. Flathead Basin Cooperative Program Final Report Page 43 The Effect of Timber Management on Stream Water Quality distinct, quantifiable, and statistically signifi- cant differences between streams associated with differing levels of timber management activity for several of the water quality parame- ters. Sediment Transport and Nutrient Concentrations We compared sediment transport, nutrient concentrations, and various physical variables between watershed sites and upstream/down- stream study sites. We selected four sampling dates to represent time periods of distinctly different stream discharge dynamics — three during the spring runoff period (that is, rising limb of runoff, peak runoff, falling limb of runoff) and stable stream discharge (that is, autumn low flow). Total suspended sediment, both organic and inorganic combined, was stud- ied because of the broad effect of sediment on all stream biota (for example, microbial growth, attached algae growth, invertebrate communi- ties, fish reproduction). Dissolved and particu- late forms of nitrogen (NOj^; NH^; total persul- fate nitrogen — TPN) and phosphorus (soluble reactive phosphorus — SRP; total phospho- rus — TP) were analyzed because of their importance to algal production and their well- documented effects on eutrophication of down- stream lakes. We found that all study sites had low con- centrations of suspended particulates during base flow conditions and, not surprisingly, that maximum suspended materials occurred dur- ing June, the peak in spring runoff in 1990. However, results also indicated that total sus- pended sediment (TSS) was closely associated with stream gradient; and, that among paired streams, higher annual maximum concentra- tions of total suspended sediment was found among streams associated with moderate to high levels of timber management compared to streams from no to low activity watersheds. (See Table C-1.) We also compared annual mean nutrient values paired as no to low activity watershed sites with moderate to high activity watershed sites. Results of these comparisons revealed that statistically there was a very high signifi- cant difference (p < 0.001; Low Activity < High Activity) among all forms of nutrient variables combined and a significant difference (p < 0.1; Low Activity < High Activity) among TPN and TP concentrations. (See Table C-1.) Differences were particularly distinct for maximum concentrations of total phosphorus and total persulfate nitrogen. (See Figure C-1.) In this cross- watershed comparison, total phos- phorus (TP) and total persulfate nitrogen (TPN) concentrations were progressively higher be- tween respective no to low activity, moderate activity and high activity paired watersheds. Stream Algae We monitored attached algal growth and biomass in each of the study streams by placing artificial substrates in riffle habitats. Artificial substrates consisted of pre-leached, non-glazed clay tiles held above the stream bed by a solid frame to minimize invertebrate colonization and grazing. We removed three substrate samples from each stream after 2, 4, and 8 weeks of incubation and algal colonization. We analyzed the collected algae fi-om each sub- strate for Chlorophyll aconcentrations and ash- free-dry-mass. Analyses of these data revealed a general pattern of increased algal growth associated with higher levels of watershed man- agement activity. (See Figure C-2.) Stream sites within no to low activity watersheds had very low algal production. Sites located within streams of moderate watershed activity had Page 44 Flathead Basin Cooperative Program Final Report The Effect of Timber Management on Stream Water Quality Table C-1. Wilcoxon's signed-rank comparison of mean total suspended solids (TSS), three forms of nitrogen (NH3, NO2/3, and total persulfate nitrogen — TPN), and two forms of phosphorus (soluble reactive phosphorus — SRP and total phosphorus — TP) between no to low activity watershed sites paired to moderate to high activity watershed sites. Nutrient Variable Summary Statistic Low Activity < High Activity TSS p = 0.043 nutrient variables combined p = 0.0053 NH3 p = 0.715 NO2/3 p = 0.465 TPN p = 0.068 SRP p = 0.273 TP p = 0.068 Level of Significance *** ns ns ns ns = not significant * = significant (p < 0.10) ** = highly significant (p < 0.05 *** = very highly significant (p < 0.01) Flathead Basin Cooperative Program Final Report Page 45 The Effect of Timber Management on Stream Water Quality 40n 30 20 10 TP NO to LOW ACTIVITY MODERATE ACTIVFTY HIGHACTIVriY r 40 30 -20 10 '-o 600 500 400 300 O) =1 200 100 TPN (T cc a: O o o 7 ^ 8 ^ 8 1 U. Q => Z3 3 NO to LOW ACTIVITY MODERATE ACTIVITY HIGH ACTIVITY 600 500 -400 300 200 -100 1^ Figure C-1. Annual maximum total phosphorus (TP) and lota! persulfate nitrogen (TPN) at each often paired watershed sites. Site pairs appear as directly comparable sites with different levels of timber manage- ment activity in the watershed. Page 46 Flathead Basin Cooperative Program Final Report The Effect of Timber Management on Stream Water Quality comparatively moderate levels of algal density; while the highest Chlorophyll a concentrations were measured within streams in high timber management activity watersheds. An analysis of variance (ANOVA) of maximum Chloro- phyll a data, in which sites falUng into the three different activity levels were collectively com- pared, revealed that algal production was sig- nificantly less (p < 0.05) among no to low activity sites when compared to sites represent- ing high timber management activity water- sheds. Moderate activity sites were not statisti- cally significantly different from either high or low activity sites, although the general trend of increased algal density above that observed for low activity watersheds is readily apparent. (See Figure C-2.) This trend was not significant concerning algae mass accumulations because of the high variance between similar activity sites; however, mean values of ash-free-dry- mass were generally much higher at high activ- ity stream sites. Based on this study, which involved a range of stream types and sizes with a mix of open and closed canopy sites among streams with differ- ing levels of upstream timber harvest, increased timber harvest results in increased stream algae production even in streams that are well cano- pied. These results are consistent with the pat- tern of increased nutrient concentrations among streams flowing from high activity basins. Thus, these data are mutually supportive from an ecological perspective. acterized among Flathead Basin zoobenthic species as having high biodiversity and broad trophic function. Thus, some of the differences between sites should be observable during this period. We identified a total of 67 taxa from the 12 study sites. Most taxa were identified to the species level of organization, particularly the mayflies {Ephemeroptera), stoneflies (JPlecop- tera) and caddisflies (Trichoptera). Our analy- sis of taxa frequency revealed high variability between sites that reflect the multitude of fac- tors affecting zoobenthic species distributions and abundance (for example, stream type, stream size, habitat, substrate, source of food, stream velocity, competition, predation). Measures of taxa richness and Simpson's index of equitabil- ity (Figure C-3) at each study site were highly variable. No specific pattern of either reduced richness or taxon dominance could be attrib- uted to any of the measured physical or nutrient variables, or specifically increased algal pro- duction. This was documented by no significant relationships being apparent employing non- parametric tests of taxon abundance among paired comparative streams. This does not mean that timber manage- ment has no effect on stream zoobenthos or that zoobenthic organisms are a poor indicator of change in stream ecosystems, but rather that within the constraints of this study, no statisti- cally significant patterns associated with the level of timber management activity could be determined. Zoobenthos Summary We collected benthic samples during au- tumn to determine quantitatively and qualita- tively the possible effects of timber manage- ment on benthic invertebrates. We were limited to a single sampling period because of time and financial constraints. However, autumn is char- ( 1 ) An experimental design of stream sites was chosen for study of the effects of timber management activity on the waterquality of FlatheadBasin streams. Stream sites were chosen to represent a broad cross-section of stream and wa- Flathead Basin Cooperative Program Final Report Page 47 The Effect of Timber Management on Stream Water Quality 0.6 0.5 cvj 0.4 H E o> =L 0.3 0.2 0.1 Chlorophyll oc oJ^ NO to LOW ACTIVITY MODERATE ACTIVmC HIGH ACTIVfTY r 0.6 - 0.5 0.4 0.3 0.2 0.1 ^0 I.Oi 0.001 -♦ NO to LOW ACTIVITY MODERATE ACTIVFTY HIGH ACTIVITY 1.0 rO.1 rO.OI ♦- 0.001 Figure C- 2. Summer maximum Chlorophyll a and ash-free-dry-mass (AFDM) at each of ten paired watershed sites. Site pairs appear as directly comparable sites with different levels of timber management activity in the watershed. Page 48 Flathead Basin Cooperative Program Final Report The Effect of Timber Management on Stream Water Quality UJ 30- ^ 20- 0 0.20 TAXA RICHNESS NO to LOW ACTIVITY MODERATE ACTIVITY HIGH ACTIVfTY Q Z 0.08- SIMPSON'S INDEX . bi •-0 0.20 Figure C-3. Autumn taxa richness and Simpson's index of riffle habitat macroinvertebrates at each of ten paired watershed sites. Site pairs appear as directly comparable sites with different levels of timber management activity in the watershed. Flathead Basin Cooperative Program Final Report Page 49 The Effect of Timber Management on Stream Water Quality tershed types; from very small streams to large creeks, from no to high timber harvest activities in the watershed, and from widely different geologies (Swan drainage. North Fork drainage, Upper Stillwater drainage. Upper Tally Lake drainage). (2) Concentration of sediment in transport was closely associated with the slope of the watershed and the gradient of the stream at the sample site. However, comparison of annual maximum sus- pended sediment concentration be- tween paired-stream study sites re- vealed that moderate and high activity watersheds had a general pattern of significantiy higher suspended sedi- ment concentration (p < 0.05) than their respective no to low activity wa- tershed site. (3) Nutrient concentrations, particularly N and P fractions of Total Phosphorus (TP) and Total Persulfate Nitrogen (TPN), were significantly higher (p < 0.1) among stream sites within high activity watersheds compared to streams from no or low activity water- sheds. (4) Attached algae growth was measured at each sample site using artificial sub- strates for colonization of algae. The density of Chlorophyll a on substrates incubated in streams with high timber management activities in the water- shed was significantly higher (p < 0.05) than in streams from no or low activity in the watershed. (5) No significant differences in inverte- brate trophic relationships, taxa rich- ness, or taxa equitability were observed between streams of differing water- shed timbermanagementactivity; how- ever, there was a general trend toward increased species richness associated with harvest activity. Conclusions Because the conclusions drawn from this research are best expressed holistically as they relate to the other empirical studies, most con- clusions and recommendations of this study appear in the "Summary of Conclusions" and "Summary of Recommendations" sections at the end of this document. However, the follow- ing points should be made here in the context of the cooperative studies: (1) Timber management activity has a quantifiable effect on stream water quality in several important areas: (a) increased maximum suspended sediment concentration during spring runoff, (b) increased mean annual concen- tration of algal growth nutrients (nitrogen and phosphorus), and (c) increased maximum algal den- sity on the stream bottom. These factors not only profoundly af- fect the ecology of the stream, but also that of downstream lakes. (2) Future monitoring of timber manage- ment effects on streams should con- centrate on measurement of suspended sediment, nutrient concentrations (TP, SRP, TPN, N0„, NH3), and attached algal growth (Chlorophyll a density). (3) Monitoring of macroinvertebrates will be most useful when used to define the most severe of impacts that result in catastrophic degradation of the stream biota. General baseline data of inverte- brates is therefore needed among streams at risk. Page 50 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module D: Fisheries Habitat AND Fish Populations By Thomas Weaver and John Fraley This section of the Final Report summarizes a study of the same name published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^^^ Page 52 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations By Thomas Weaver^ and John Fraley^ Introduction The fisheries study module focused on fish- eries habitat and fish populations. Westslope cutthroat trout and bull trout are native to the Flathead system. Montana recognizes these fish as "Species of Special Concern," and affords them special protection. Sediment originating from road building and other land management activities can reduce embryo survival and emer- gence of both species and decrease the available space in the stream bed used for rearing by bull trout. Spawning/incubation by both species and rearing by bull trout are the life stages most sensitive to sediment effects. Sediment deposi- tion can also affect rainbow and brook trout, as well as other fish species, by covering spawning gravel, filling in pools, and altering food habits. However, these other fish species are not good indicators of sediment effects on rearing due to specific behavioral differences. The objectives of this study were to ( 1 ) evaluate the relationship between sediments and westslope cutthroat and bull trout emer- gence success, (2) determine capabilities of several methods of measuring fish habitat qual- ity, (3) examine cause and effects relationships between forest practices and fish habitat and populations, and (4) recommend methods for a monitoring program for fish habitat and popu- lations in the basin, relative to forest manage- ment. In this study module, we concentrated on 29 tributaries. (See Table D-1 and Figure D-1.) These tributaries were located in the following drainages: Swan, Stillwater/Whitefish, North Fork, Middle Fork, and South Fork of the Flat- head River. These tributaries were chosen by the study team to be representative of the vari- ety of geography, habitat, land ownership and fisheries conditions in the basin. We selected specific variables to indicate quality of spawn- ing habitat (McNeil coring and Whitlock- Vibert box sampling), rearing habitat (substrate scor- ing), juvenile population levels (electrofishing estimates) and spawner use (redd counts). A complete description of the study design, meth- ods, and findings is available in the Final Mod- ule Report. Results and Discussion Embryo Incubation Studies Researchers examined emergence success and quality (length and weight) of westslope cutthroat trout fry in relation to varying levels of fine substrate materials in a natural stream en- vironment. We simulated natural incubation conditions in a stream (Chapman 1988) by constructing cells with particle sizes, egg pock- ets, and egg planting depths characteristic of natural westslope cutthroat trout redds. A sig- nificant negative relationship (p < 0.005) ex- 'Thomas Weaver is a Research Specialist with the Montana Department of Fish, Wildife and Parks in Kalispell, Montana. *John Fraley was a Fish and Wildlife Program Officer with the Montana Department of Fish, Widlife and Parks in Kalispell, Montana. V y Flathead Basin Cooperative Program Final Report Page 53 Fisheries Habitat and Fish Populations Table D-1: List of study streams showing activities scheduled for each stream included as part of the Flathead Basin Commission Cooperative Forest Practice, Water Quality, and Fisheries Study. Area/Drainage: Stream Fish Monitoring Activity Stillwater Fitzsimmons Creek MN WV SS _ P Chepat Creek MN WV SS - P Swan River . Elk Creek (MN) (WV) (SS) (RC) (P) Goat Creek (MN) WV SS (RC) (El Squeezer Creek (MN) WV SS (RC) p Jim Creek (MN) WV SS RC P Lion Creek (MN) WV SS (RC) (P) Piper Creek MN WV SS RC P Island Unit Freeland Creek MN WV SS RC P Tally Lake Fish Creek (MN) WV SS RC P HandCredk (MN) WV SS - P Swift Creek MN WV SS RC P Sheppard Creek MN WV SS - P Squaw Tributary MN - SS - P North Fork Big Creek MN WV SS RC (P) Coal Creek DH (MN) WV (SS) (RC) (P) NF Coal Creek (MN) (WV) (SS) (RC) in SF Coal Creek (MN) fWV) (SS) (RC> in Cyclone Creek MN WV SS RC p Red Meadow Creek (MN) WV SS - (p) Whale Creek (MN) WV SS (RC) (P) Trail Creek (MN) WV SS (RC) p Middle Foric Granite Creek (MN) WV SS fRQ - Challenge Creek (MN) WV SS (RC) (p) Ole Creek - - SS (RC) (P) Morrison Creek (MN) WV (SS) (RC) (p) South Fork Hungry Horse Creek (2) (MN) WV (SS) (RC) (p) Margaret Creek (MN) WV (SS) (RC) (p) Tiger Creek (MN) WV (SS) (RC) (p) Emery Creek (MN) WV (SS) (RC) (p) Rsh Monitoring Activity Codes: ( ) Denotes Montana Department of Fish, Wildlife and Parks (MDFWP) and/or Flathead National Forest (FNF) work supported by other funding. Denotes MDFWP work contracted by FNF. MN McNeil gravel core samples (12 cores at each site; 4 cores on each of 3 transects). SS Substrate scores (15 transects in a 150 m stream section). RC Redd counts (total number of spawning sites in a su-eam section). P Fish population estimates (electroflshing estimates for a 150 m stream section) Page 54 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Figure D-1: Watershed Study Sites ".".'." ,r,."i'."r".' ,Tr»T zz^ mites II S 10 n J FLATHEAD NATIONAL FOREST Fofesi Service Flatfiead Page 55 Fisheries Habitat and Fish Populations isted between fiy emergence success and the percentage of substrate materials less than 6.35 mm in diameter. (See Figure D-2.) Mean fry emergence success was 76, 55, 39, 34, 26, and 4 percent, respectively, in cells containing 0, 10, 20, 30, 40, and 50 percent materials less than 6.35 mm. We measured no distinct trend in emergence timing, and no significant differ- ences in length or weight of fry emerging from the six gravel mixtures. Biologists also examined emergence suc- cess of bull trout fry in relation to varying levels of fine substrate materials in a natural stream environment. We simulated natural incubation conditions in a stream (Chapman 1988) by constructing cells with particle sizes, egg pock- ets, and egg planting depths characteristic of natural bull trout redds. We found a significant negative relationship (p < 0.(X)5) between fry emergence success and the percentage of sub- strate materials less than 6.35 mm in diameter. (See Figure D-3.) Mean adjusted fry emer- gence success was 79, 64, 44, 39, 26, and 4 percent respectively, in cells containing 0, 10, 20, 30, 40, and 50 percent materials less than 6.35 mm. A major portion of the observed mortality resulted from fry entombment in the heavier sediment levels. Results from these studies showed an em- bryo mortality of about two-thirds when 35 percent of the gravel comprising the incubation environment is smaller than 6.35 mm. At 40 percent smaller than 6.35 mm, approximately three-quarters of the embryos deposited did not emerge successfully. Freeze Coring in Spawning Sites Project personnel collected frozen core samples from migratory westslope cutthroat trout redds in Hungry Horse Creek. This work was recommended to verify the locations and structure of egg pockets in natural spawning sites (Chapman 1988). This sampling confirmed the setup used for the incubation studies. We found an egg deposition depth of greater than 10.0 cm but less than 20.0 cm. We observed an undisturbed layer of large angular particles (mostly greater than 50.8 mm and some greater than 76.1 mm) in the frozen samples beginning around 17.8 cm below the surface and extend- ing downward. It is likely that these particles prevented deeper excavation by the fish and formed the "floor" of the redds. In all samples, less fine material (percent less than 6.35 mm) was present in the 10.0 cm depth band containing the eggs than in the one immediately above it. Likewise the geometric mean (Platts and others 1979) and Fredle index (Lotspeich and Everest 1981), two other meas- ures of particle size, were greater in the egg pocket strata. A comparison of the mean values for egg pocket depth bands and the 10.0 cm strata above it showed significant differences for both the geometric mean particle size and the Fredle index (p < 0.05). It is possible that the female fish cleaned the substrate in the egg bearing strata in the process of spawning (Chapman 1988). Although em- bryos may incubate in egg pockets having greater geometric mean particle size and less fine sedi- ment, emerging fry must still migrate up through the material covering the egg pocket to success- fully reach the stream. Entombment by this material resulted in the majority of the mortality observed during the incubation studies. Field crews collected frozen core samples from bull trout redds in Lion Creek. We found no eggs in these samples, but other interesting observations were made. Researchers have re- ported that some fish spawning in higher levels of fine sediment may build larger redds, but deposit eggs at a shallower depth (Everest and others 1987). Our work supported these find- ings although our sample size was small. Page 56 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Number 25 ■ of Fry Captured 20 15 10 5 + 0 Figure D-2: Relationship between numbers of westslope cutthroat trout fry successfully emerging from replicates of six gravel mixtures and the percentage of material smaller than 6.35 mm in each mixture Survival = -0.654812 (% < 6.35 mm) + 35.6749 p < 0.0005. r2 = 0.72. n = 17 10 15 20 25 30 35 Percentage of Material less than 6.35 mm 40 45 Number of Fry 40 Captured 30 + 20 10 0 Figure D-3: Relationship between number of bull trout fry successfully emerging from replicates of six gravel mixtures and the pereentage of material smaller than 6.35 mm in each mixture Survival = -1.29462 (% < 6.35 mm) + 72.4615 p < 0.0005. r2 = 0.91. n = 17 10 15 20 25 30 35 Percentage of Material less than 6.35 mm 40 45 50 Flathead Basin Cooperative Program Final Report Page 57 Fisheries Habitat and Fish Populations Bull trout redds are much larger than the cutthroat trout redds we sampled. The greater area of disturbed gravel makes it more difficult to obtain egg pocket samples. We feel this effort should continue to further refine embryo sur- vival modeling. Data from even a small number of bull trout egg pockets would help our predic- tive abiUty. Streambed Coring The size range of streambed material is indicative of spawning and incubation habitat quality. Most research shows negative relation- ships between fine sediment and embryo sur- vival to emergence (Chapman 1988). We used a McNeil Corer (McNeil and Ahnell 1964) to collect streambed samples which were dried and sieve analyzed to determine the particle size composition. As the percentage of fine material increased, habitat quality decreased. Median percentages (12 per site) of stream- bed material less than 6.35 mm ranged from 24.8 percent in Chepat Creek to 50.3 percent in Jim Creek. The values at our 29 sampling sites averaged 36.3 percent. The maximum variabil- ity observed within individual streams was about 20 percentage units. To obtain an indication of natural sediment levels, we averaged the McNeil core sampling results for the nine watersheds where the Sequoia index (Potts and Mclnemey 1990) was 0.00, indicating no disturbance. This calculation resulted in an average value of 3 1 .7 percent. Of these nine watersheds, we know of significant natural sediment sources with high levels of channel storage above the study areas at two sites — Elk Creek and Lion Creek. The Sequoia Model does not consider natural sedi- ment sources or other natural phenomena which may alter streambed conditions. Eliminating these two sites from calculations of the average condition in "undisturbed" watershed results in a value of 29.8 percent. Researchers have collected McNeil core samples in selected spawning areas annually during the past ten years. This period of record gives an idea of how spawning area gravel composition changes from year to year. The average annual change in the median percent- age (n = 12) of material smaller than 6.35 mm in samples from Big, Coal, and Whale Creeks has been 4.8 percentage units; changes ranged from 0. 1 to 10.7 percentage units. These drain- ages are mainly roaded timber lands. The aver- age annual change in the Trail Creek samples during this same ten year period was 2.1 per- centage units (range = 0.2 to 5.2). The Trail Creek spawning area is not subject to the ground- disturbing activities in the drainage above, which are occurring in Big, Coal, and Whale Creeks. The average annual change observed in "undis- turbed watersheds" (Sequoia index =0.0) where we have completed five annual samplings is 3.0 percentage units (range = 0.4 to 5.5). We believe McNeil core sampling is reflec- tive of streambed conditions in the specific sampling sites. Once one year of information exists, annual sampling provides an adequate monitoring tool for detecting changes in stream- bed composition. However, more cost efficient methods would allow us to increase the level of this activity basin- wide. Whitlock-Vibert Boxes As an attempt to validate a more cost effi- cient method of monitoring streambed compo- sition, field crews planted a total of 380 Whit- lock- Vibert (W- V) boxes during the study. Other researchers have reported good results in sub- stituting these slotted, plastic boxes for McNeil core sampling (Reiser and others 1987, Wesche and others 1989). We recovered 182, for a recovery rate of 48 percent. Other researchers reported a 58 percent recovery rate for boxes in a similar field test (Reiser and others 1987). Page 58 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Most of our box loss occurred during a Novem- ber, 1989, precipitation event. We noted sub- stantial movement of streambed material at this time. We lost all W-V boxes planted in several South Fork and Middle Fork Flathead tributar- ies. Some box displacement and loss occurred in the North Fork Flathead and Swan River drainages as well. In all, we lost well over 100 W-V boxes during this event. Of the 182 boxes recovered, we analyzed 109 in the laboratory for fine sediment accumu- lation and density. The overall linear regression of submerged weight of a W-V box against the percentage of material less than 4.75 in the associated McNeil core sample showed a sig- nificant positive relationship (p < 0.05), al- though considerable scatter existed (r = 0.48). When we compared the mean percentage less than 4.75 mm in the W-V boxes (12 per site) with the mean from McNeil coring (12 per site) at each study area, we observed significant differences (p < 0.05) in eight of the 12 com- parisons. We expected the mean percentages from the boxes to be in closer agreement with the coring data. Based on these findings, we question whether W-V boxes are suitable for monitoring streambed conditions at this time. However, we believe the potential advantages of this tech- nique over presendy used methods warrant greater effort at developing a W-V box program for the Flathead Drainage. Questions relating to box planting, box loss, timing of box planting, and best marble size will require more evalu- ation. A change in box design to use perfora- tions 6.35 mm in diameter would yield more usable results and might eliminate questions about the maximum particle size which could enter the boxes. Streambed Substrate Scoring Rearing fish, particularly juvenile bull trout, often occupy open spaces between or under streambed materials. Substrate score (Grouse and others 1981) is an index indicating the habitat's potential for rearing and overwinter- ing fish. Silt-free streams with large rocks and lots of hiding space would get a high substrate score. As the streambed becomes more imbed- ded with silt and sand, the substrate score would become less. Scores above 1 1 .0 generally indi- cate good rearing habitat quality. Other re- searchers have specified 9.0 as the minimum critical standard (Shepard and others 1984). Substrate scores ranged from 8.8 in Freeland Creek to 13.2 in Piper Creek. Based on the above values, Jim and Freeland creeks are at or below the minimum recommended level. Squeezer, Lion, and Coal Creeks scored be- tween 9.0 and 10.0. Sixty-two percent of the scores basin- wide were 1 1 .0 or higher. Linear regression of juvenile bull trout den- sities (number of trout per 100 m^) against substrate score showed a significant positive relationship (r = 0.54, p = 0.05; n =15). (See Figure D-4.) Densities of other trout species in our study streams did not correlate with this index. This is probably due to behavioral differ- ences. These species are not strongly associated with the streambed at the time our estimates were made. During winter, long portions of stream chan- nel are completely ice- and snow-covered. Field crews have also observed extensive areas of anchor ice. In other areas, upwelling ground water keeps certain sections open, even during extreme conditions. These open areas may sup- port the majority of the winter rearing in the streams where they occur. It is likely that if we could obtain estimates of juvenile fish densities during this winter period instead of late sum- mer, a stronger relationship would result. Flathead Basin Cooperative Program Final Report Page 59 Fisheries Habitat and Fish Populations Figure D-4: Relationship between transformed substrate scores and juvenile bull trout densities (number of trout less than 75 mm per 100 m2) for 15 tributary reaches in the Flathead River Basin during 1989 12 T 10 - Number of Bull Trout greater than 75 mm per 100 meters squared 4 - 10 y = 0.0845433 (Transformed Score) +1.31704 p < 0.05, r = 0.54, n= 15 15 20 25 30 35 Transformed Substrate Score 40 45 50 Table D-2: Summary of 1989 rainbow trout spawning site inventories. Creek Redd Numbers 1989 Fish Freeland Hillburn (Ashley Lake) 27 (Lake Mary Ronan) 2(X) (Lake Mary Ronan) 19 Page 60 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations 3PAWNING Site Counts We completed rainbow trout spawning site inventories on Freeland and Fish Geek drain- ages. (See Table D-2.) The number of spawn- ers using these two streams during spring of 1989 appeared similar to observations made during past years. We completed westslope cutthroat trout spawning site inventories on all proposed streams. (See Table EX-3.) In the South Fork drainage, redd numbers in Hungry Horse Res- ervoir tributaries were up during 1989. Our index stream in the Middle Fork drainage con- tained fewer redds in 1989 than observed dur- ing 1982 surveys when we documented 24 redds. Since we had made no previous counts in Cyclone Creek in the North Fork drainage, comparisons with existing data are not possible. We conducted bull trout spawning site in- ventories in established monitoring areas basin- wide. (See Table D-4.) We identified 244 and 158 bull trout redds in North and Middle Fork Rathead tributaries, respectively. They were approximately nine percent above the ten-year average figure. (See Figure D-5.) We counted 37 1 redds in the Swan River tributary monitor- ing areas during 1989. This was approximately 57 percent above the seven-year average. (See Figure E)-6.) Our numbers do not represent the total annual spawning run. We estimate our annual counts represent about 35 percent of the annual Flathead Lake spawning escapement and about 75 percent of the Swan Lake run. Table D-3: Summary of westslope cutthroat trout spawning site inventories from 1986-1987 Redd Numbers Drainage Creek 1986 1987 1988 1989 South Fork: Hungry Horse 93 28 123 118 — ^ Margaret 18 10 37 43 Tiger 10 7 46 61 Emery 88 74 108 129 Middle Fork: Challenge - - - - - - 19 North Fork: Cyclone ■ ~ . " ■ " 31 Flathead Basin Cooperative Program Final Report Page 61 Fisheries Habitat and Fish Populations Table D-4: Summary of annual bull trout spawning site inventories between 1979 and 1989. Drainage/ Stream 1979 1980 1981 Redd Numbers 1982 1983 1984 1985 1986 1987 1988 1989 North Fork: Big Coal Whale Trail Total Middle Fork: Morrison Granite Lodgepole Ole Total Flathead Total Swan: Elk Goat Squeezer Lion Total 10 38 35 34» 20 34 45 31» 18 23 98 78 41 60 211 94 22 71 141 56 9 53 133 32 9 40 94 25 12 13 90 69 22 48 143 64 193 236 295 109 19 52 136 62 210 290 321 24 50 199 51 117 130 217 406 290 227 168b 184 277 269 324 25" 75 32" 86 67 38 99 52 49 50 63 14 34 14a 24 31 47 24 37 34 32 31 32 14 18 23 23 23 20 42 21 19 43 - 19 19 51 35 26 30 36 45 59 21 71 142 83 184 156 134 173b 167 149 160 158 188 272 300 590 446 361 341 351 426 429 482 _ _ _ 56 91 93 19 53 162 201 186 - - - 33 39 31 40 56 31 46 34 - - - 41 57 83 24 55 64 9b 67 - - - 63 49 88 26 46 33 65 84 371 founts may be underestimated due to incomplete survey. "High flows may have obliterated some of the redds. Page 62 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Figure D-5: Summary of annual bull trout redd counts in the North and Middle Forks of the Flathead River Drainage from 1979 through 1989. Number of Redds 600 T 500 - 400 - 300 - 272 200 100 - 300 -60Q 436 361 141 i51 426 429. 402 riH ■■■■■■■■ 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 Years D North Fork of Flathead River M Middle Fork of Flathead River Number of Redds Figure D-6: Summary of annual bull trout redd counts in the Swan River Drainage from 1982 through 1989. 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 Years Flathead Basin Cooperative Program Final Report Page 63 Fisheries Habitat and Fish Populations Population Esttmatton Mean total fish density for the 28 elec- trofishing areas was 13.1 trout greater than or equal to 75 mm per 100 m^ of stream surface area. Densities ranged from 2.1 in Jim Creek to 75.8 in Freeland Creek. (See Table D-5.) Tribu- tary streams to large lakes which support rear- ing populations of rainbow trout (Freeland and Fish Creeks) and westslope cutthroat trout (Hun- gry Horse, Tiger, Margaret, and Emery creeks) had the highest densities. Eastern brook trout populations in Sheppard and Hand Creek were also present at high densities. (See Table D-5 .) Juvenile bull trout are much more substrate oriented than these other trout species. Because of their close association with the streambed, bull trout are better indicators of the influence of fine sediment. We compared juvenile bull trout densities with substrate scores and found a significant positive relationship (r = 0.54; p = 0.05; n = 15) existed. (See Figure D-3.) Previ- ous work showed a similar but somewhat stron- ger relationship for Swan River tributaries sup- poning juvenile bull trout (Shepard and others 1984). We observed a mean juvenile bull trout density of 3.8 fish greater than or equal to 75 mm per 100 m^. Densities ranged from 0.4 in Jim Creek to 11.8 in Morrison Creek. (See Table D-5.) Swan River tributaries supported juvenile bull trout at an average density of 2.7 fish greater than or equal to 75 mm/100 m^ while the North and Middle Forks averaged 4.6 fish/100 m^. Eastern brook trout arc present in the Swan River tributaries but not in the North and Middle Fork sections. Information on juvenile bull trout densities and streambed conditions in winter rearing ar- eas may show stronger relationships than we obtained using late summer electrofishing. It is possible that winter rearing habitat may control juvenile bull trout densities in our study streams. Any ground-disturbing activities proposed above these critical rearing areas should be carefully planned and monitored. In general, these findings support the use of bull trout as an indicator species for future monitoring efforts. Linkage between Risk Assessment and Fisheries To demonstrate the linkage between land management activities and fisheries we used simple linear regression analysis. We used arc- sine transformations on the output from the Sequoia index and H^OY model. This is a stan- dard procedure used when data are percentages and the values are limited in range and close to zero. We evaluated the McNeil coring data and the substrate scores as percent differences. X = observed value - minimum value (100) minimum value This was necessary to expand the range of the data. Results showed significant relationships (p < 0.05) between McNeil coring results and output from both Sequoia and H^OY models. The correlation with the Sequoia index was slightly stronger than with the H^OY model. (See Figures D-7 and D-8.) Comparisons of substrate scores with output from both the risk assessment index and H^OY model also show significant relationships (p < 0.05). In this case the correlation with Sequoia was sUghtly weaker than with H^OY results. (See Figures D-9 and D-10.) Although these relationships are signifi- cant, considerable scatter exists. There are sev- eral reasons which may explain a portion of this scatter. First, to keep this analysis as simple as possible we elected to use linear regression techniques. By including more variables in a multiple regression analysis it is likely we could Page 64 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Table D-5: Comparison of total fish density, juvenile bull trout density, and juvenile cutthroat trout density (for those ^ 75 mm/100 m^) calculated from electrofishing estimates at 28 sites around the Flathead Basin during 1989. Stream Area of Section Total Fish Density # ^ 75 mm/100 m2 Juvenile Bull Trout # > 75 mm/100 m2 Juvenile Cutthroat Trout # S 75 mm/100 m2 Big 1695 5 4.9 Challenge 705 19.6 19.4 Chepat 540 13.9 13 CoaIDH 1410 7.4 4.2 Elk 1605 3 2.8 Emery 810 11.2 11.2 Fish 375 16.5 Fitzsimmons 645 9.1 4.5 Freeland 360 75.8 Goat 930 6.6 3.6 Hand 735 12.4 Hungry Horse 810 14.3 14.3 Jim 1155 2.1 0.4 Lion 1710 3 2.2 Margaret 465 25.6 25.6 Morrison 1095 11.8 11.8 North Coal 900 10.6 5.1 5.7 Ole 810 5.5 4.9 Piper 795 14.3 4.6 3.4 Red Meadow 1170 7.2 1.7 5.5 Sheppard 795 15.6 1.4 South Coal 810 9 1.8 7.3 Squaw Trib. 345 5.2 Squeezer 960 9.9 2.5 Swift 1245 4.4 4.2 Tiger 555 41.4 41.4 Trail 915 5.5 5.2 Whale 1545 2.4 2.1 Flathead Basin Cooperative Program Final Report Page 65 Fisheries Habitat and Fish Populations Figure D-7: Relationship between the arcsine transformations of Sequoia model output and transformed McNeil core results for 28 watersheds in the Flathead River Basin during 1989 120 J 100 - 80 Transformed McNeil 60 ± Results y = 1.25551 (Arcsine Transformation of Sequoia Model Output) + 33.7924 p < 0.05. r = 0.47. n = 28 5 10 15 20 25 Arcsine Transformations of Sequoia Model Output 30 35 120 - Figure D-8: Relationship between the arcsine transformations of H20Y model output and transformed McNeil core results for 28 watersheds in the Flathead River Basin during 1989 r 100 - ■ y = 1.69183 (Arcsine Transformation ■ of H20Y Model Output) + 33.0522 ■ p < 0.05. r = 0.45, n = 28 80 - - ■ ■ ■ ^^ Transformed McNeil Results 60 j i 1 40 - 20 - 1 1 ■ ■ ■ U " 0 5 10 15 20 25 30 Arcsine Transformations of H20Y Model Output Page 66 Flathead Basin Cooperative Program Final Report Fisheries Habitat and Fish Populations Figure D-9: Relationship between the arcsine transformations of Sequoia model output and transformed substrate scores for 28 watersheds in the Flathead River Basin during 1989 Transformed Substrate Scores y = -0.753617 (Arcsine Transformation of Sequoia Model Output) + 35.8244 p < 0.05, r = -0.48, n = 28 5 10 15 20 25 Arcsine Transformations of Sequoia Model Output Figure D-10: Relationship between the arcsine transformations of H20Y model ouq)ut and transformed substrate scores for 28 watersheds in the Flathead River Basin during 1989 Transformed Substrate 25 - Scores 20 15 10 - 5 - 0 0 y = -1.17647 (Arcsine Transformation of H20Y Output) + 37.3737 p < 0.05, r = -0.53. n = 28 5 10 15 20 25 Arcsine Transformations of H20Y Model Ouq}ut 30 Flathead Basin Cooperative Program Final Report Page 67 Fisheries Habitat and Fish Populations tion between independent variables could cloud the assessment of cause and effect. The Sequoia index assumes a ten year recovery period. It does not take into account older problems still having a major influence on the percentage of fines in spawning gravel in several watersheds. Sequoia does not consider catastrophic events such as fires or floods. The channel morphology and percentage of fine material in several of our spawning areas are still showing effects of the 1 964 flood. By eliminating the watersheds where assumptions dealing with recovery rates and natural events are not reflective, the fit im- proved (r = 0.65; n = 21). Conclusions (1) Results indicated a direct linkage be- tween ground-disturbing activity (Se- quoia and HjOY) and a measurable fisheries habitat parameter (percent- age of material less than 6.35 mm) which is linked to embryo survival by westslope cutthroat and bull trout. (2) Findings also illustrated a direct link- age between ground-disturbing activ- ity (Sequoia and H^OY) and an index of fisheries habitat (substrate score) which is linked to juvenile bull trout rearing potential. (3) Spawning area gravel composition in the nine watersheds with no develop- ment averaged 31.7 percent material smaller than 6.35 mm. This size class comprised an average of 39.0 percent in the 17 watersheds where disturbed area exceeded one percent of the drain- age. Forest management activities have had a quantifiable effect on streambed composition and fish populations in the Flathead Basin. (4) Monitoring streambed composition in known westslope cutthroat and bull trout spawning areas can provide fish- eries information useful in making land management decisions. Once an initial sampling is complete, McNeil coring is an adequate tool for quantifying streambed particle size composition in spawning areas. We can detect changes in gravel composition. (5) Monitoring streambed substrate score in known bull trout rearing areas can provide fisheries information useful in making land management decisions. Substrate scores are not adequate indi- cators of rearing potential for fish spe- cies other than bull trout. Behavioral differences between the trout species present in our study area makes use of a single index impossible. (6) A significant relationship exists be- tween substrate samples collected us- ing modified Whitlock-Vibert boxes and McNeil core samples taken at the box planting sites. However, more work is required before the W-V box tech- nique can replace McNeil coring in our streambed substrate monitoring pro- gram. Page 68 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module E: Application of the Montana Nonpoint Source Stream Reach Assessment in the Flathead Basin By Steve Tralles This section of the Final Report summarizes a study of the same name published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^^^ Page 70 Flathead Basin Cooperative Program Final Report Application of the Montana Nonpoint Source Stream Reach Assessment in the Flathead Basin By Steve Tralles^ Introduction Nonpoint source pollution is the major cause of aquatic degradation in Montana (Montana Water Quality Bureau 1990). Nonpoint source pollution originates from diffuse runoff primar- ily during extreme rains or snowmelt and is generally caused by land management activity. In order of significance, the following land management activities are the major causes of nonpoint source pollution in Montana: agricul- ture, stream modification, mining, and forest practices (Montana WaterQuality Bureau 1990). Forest practices only impair approximately 13 percent of Montana's streams. However, those streams are usually located in areas with the highest aquatic resource value. Streams in the Flathead Basin are no exception. The Flathead Basin exhibits some of Mon- tana's highest water quality and supports a variety of aquatic life, including westslope cut- throat and bull trout (both species of special concern). In the Flathead Basin, siltation is the leading cause of stream pollution and forest practices are the leading causes of nonpoint source pollution (Flathead Basin Commission 1990). Montana has developed a nonpoint source pollution management plan to identify and re- solve nonpoint source pollution problems. The Montana Department of Health and Environ- mental SciencesAVater Quality Bureau has de- signed a nonpoint source pollution stream reach assessment procedure. We use it as an initial screening method to identify and prioritize moderate and severe nonpoint source pollution problems across the state. In conducting the stream reach assessment, surveyors walk the stream and visually evaluate stream conditions. They note perceived impair- ment to beneficial uses caused by nonpoint source pollution and by natural or hydrologic influences. Impairment is the degree to which a stream or stream reach will support designated beneficial uses. For the purpose of this study, the primary beneficial use is fisheries. The surveyors also evaluate land management ac- tivities and application of best management practices. Since agency resources (people and money) are limited, the assessment procedure is rela- tively quick and easy — designed primarily to identify the most significant problems. Conse- quentiy, the assessment procedure is subjective and nonquantitative. Goals of the nonpoint source pollution stream reach assessment study were ( 1 ) to evalu- ate impairment to beneficial uses in the Flat- head Basin and (2) to evaluate the accuracy of the assessment procedure by comparing its re- sults with the results of quantitative studies performed by other Cooperative Program mem- bers. 'Steve Tralles is an Environmental Specialist with theWater Quality Bureau of the Montana De- partment of Health and Environmental Sciences in Helena, Montana. Flathead Basin Cooperative Program Final Report Page 71 Application of the Montana Nonpoint Source Stream Reach Assessment Methods One sxirveyor conducted all assessments on 30 streams containing 95 reaches. (See Figure E-1 .) We selected the streams to corre- spond to sample sites or drainages used by the other Cooperative Program studies. The sur- veyor assigned stream reaches and defined them by their relative homogeneity of factors (such as valley bottom shape, gradient, channel sub- strate, vegetation, and land use). Using information collected at one or more relatively accessible observation points within each reach, the surveyor completed one assess- ment form per stream reach. The form assesses reaches using 16 separate categories and sub- categories. (See Figure E-2 and Appendix F.) The first eight assessment categories best de- scribe conditions in the Flathead Basin. There- fore we present impairment ratings based on those eight categories only. We essentially rate the categories on a scale of good-to-bad relative to stream conditions. Surveyors assign scaled numerical values to each category. The overall impairment rating for a given reach is based on a straight percent- age derived from the sum of individual category ratings divided by the total rating possible for the categories that were rated. The impairment value for a stream is derived from an average of impairment ratings for all reaches on that stream. We defined and then calibrated impairment values by assessing several streams of known impairment (severe and none). The assessment form also provides space for narrative elaboration pertaining to individ- ual assessment categories and best manage- ment practices. The narrative information sum- marized specific management activities that the surveyor observed to be contributing to stream impairment or had a high potential for causing future problems. To evaluate the accuracy of this procedure. we compared the assessment results to the re- sults of other Cooperative Program studies. We compared the overall impairment value, aver- aged by drainage, to a relative activity level derived by Sequoia (cumulative runoff acreage — Module H) or H^OY (predicted water yield increase based in part on clearcut equivalent — Module J). We compared assessed stream reaches to corresponding reaches that had Region 1 , Chan- nel Stability Rating (CSR — Module J) and fish habitat data (Module D). We established an impairment value based on four categories from the assessment form to compare the stream reach assessment data against the Channel Sta- bility Ratings. The four categories were bank stability, substrate composition, channel stabil- ity, and channel modifiers. We selected these categories because of direct applicability to parameters on the Channel Stability Rating. The assessment category of Substrate Com- position evaluates stream substrate relative to sediment influence on fish spawning and rear- ing habitat. For corresponding reaches , we com- pared substrate composition to stream substrate score and percentage of fine materials (less than 6.35 mm) as reported by the fisheries study (Module D). We ranked all reported values and applied nonparametric statistical analysis. The Kendall Rank-Order Correlation (tau) test was used for analysis. A z test statistic appUed to the tau value, was significant at the 0.05 level when greater than ± 1.96. Results We conducted assessments on 30 streams containing 95 reaches. Table E-1 lists streams by order of impairment. Of the 30 streams assessed, 47 percent were rated as having mi- nor, moderate, or severe impairment. (See Figure E-3.) Of the 95 stream reaches as- Page72 Flathead Basin Cooperative Program Final Report Application of the Montana Nonpoint Source Stream Reach Assessment ^ ^ 1 y z,?- ^^^i 1 / lWalLil,^-4^^ \f^m. \ A \ M^O N T A N A 1 /T ^^^^^ ^% L^ __^J u \{ ^~^& \ — ^1 ^ jc ovT ^— ■•^-^''*~^^'^'*'^ "£^ .Jx ^ /v. \ ^ '^jfc'tV / \, \i/* J \ \^ ■ n Cr _a— — -" X \^ Vj J (^ -^^^ ^^ \ S / , Vtddfe J, ] •%, V If/itie/ish V jc:^^^^ /A^'^^^''''*' ^ ) \ Cr \ % , \ SincJoir-> r\ \\ ^ ^ { Tally \ \ ^.-O"^ \ Lake \ } S "^ 0*/ r v^ ) \ ( f ^ ^y^ X I ( l^ 1 h^ C/laIienge-^S. Cr i('7-i I \ Ashley \{ Ic? //unjrylHk ^ Worse ^^L 1 \ \% f \ ^NRIS iDformatioD System "AS \ 1 ^^ Fjfontana Stale Itirari/ t*^ Y ^t [ • = ^ Assessed Streams '^ 1 > \ 0 20 40 Miles Figure E-1: Map: Completed NSPS Stream Reach Assessments Upper Rathead River Basin Flathead Basin Cooperative Program Final Report Page 73 Application of the Montana Nonpoint Source Stream Reach Assessment Figure E-2: Nonpoint Source Stream Reach Assessment Overview INTRODUCTION: NONPOINT SOURCE (NPS) POLLUTION IS HIGHLY DIFFUSE IN NATURE, UNLIKE POINT SOURCE POLLUTION UHICH CAN BE ATTRIBUTED TO A SINGLE POINT. NPS POLLUTION DISCHARGES OCCUR OVER DISPERSED PATHWAYS AND ARE GENERALLY ASSOCIATED WITH PRECIPITATION AND RUNOFF. NPS POLLUTION IS RELATED TO NUHAN LAND USES UHICH ARE CONTROLLABLE. BEST MANAGEMENT PRACTICES (BNP'S) ARE METHODS INCORPORATED WITH LAND USES THAT AVOID OR MINIMIZE IMPACTS TO WATER RESOURCES CAUSED BY J15% i Variable Slope i STOP AT FORMULA" DISTANCE-USE AVERAGE SLOPE, Wetland >50' ■> Slope <159{ ■>- Uniform Sk^ie ^Bench>50' from bank Bench<60' £rombank STOPatedgfe of wetland STOP at 50' MINIMUM STOP AT FORMULA* DISTANCE STOP AT EDGE OF BENCH STOP AT SO- MINIMUM C * FORMULA: SMZ = % Slope X 100 X Erodibility Factor (2, 1, 3 or 4) J Figure G-1: Field guide for locating the edges of the SMZ Page 98 Flathead Basin Cooperative Program Fined Report Management Guidelines for Riparian Forests of the current Best Management Practices (Mon- tana Environmental Quality Council 1989). Variability of field conditions will still re- quire professional judgement in delineating the SMZ. Wetlands often occur as pockets or string- ers reflecting microsites in bottomlands. Wet- lands may also extend up adjacent slopes re- flecting unstable conditions. Microtopography also changes within management units. The decision tree process allows for varying the SMZ within the unit to provide stream protec- tion. This needs to be balanced with other management considerations as long as the mini- mums at any one point along the stream are not violated. The user should remember that the above criteria are to meet minimum standards and that the current BMPs allow considerable flexibility in management practices within the Streamside Management Zones. In order to ensure protect- ing water quality, the Audit Teams and the Management Guidelines Working Group made several suggestions that should be considered in any specific operation: (1) If the existing stand has little woody debris and little undergrowth vegeta- tion to trap and filter sediments, the SMZ should be expanded appropri- ately to provide an effective filter or use other mitigating measures. (2) Large trees in or near wetlands are especially susceptible to windthrow. Potential windthrow should be antici- pated and planned for to minimize soil disturbance and bank damage caused by uprooted trees. This may mean re- moval of some high risk streambank trees and/or minimizing edge effects of certain harvesting practices. (3) Trees should be marked and reserved along those stream segments where woody debris recruitment is needed for stream protection. The selection of re- serve trees must be integrated with stream type, existing woody debris, desired woody debris, windthrow risk, silvicultural prescription, harvesting practice, and projected stand develop- ment. (4) A sediment filter buffer may be needed adjacent to certain wetlands where routed sediments from roads or skid trails have a potential of being depos- ited in the adjacent wetlands. (5) Headwater basin sites may have a pat- tern of many small streams where it is difficult to define SMZs for. each tiny intermittent stream. However, these sites are critical in terms of watershed management. In many situations, it may be most practical to simply treat entire cutting units as a Streamside Manage- ment Zone. One of the first steps to ensure protection of water quality is to recognize and delineate the SMZ and make sure every person involved in a forestry operation knows where the boundaries are. The next step is to make sure the SMZ is not violated inadvertently during logging, slash dis- posal, or site preparation. It does little good to develop a sound prescription and conduct care- ful logging if the operator piling slash or the people conducting prescribed burning do not use the same care. Estimating soil erodibility potential has been done in different ways and the Management Guidelines Working Group considered alterna- tives such as soil texture. Current studies are underway to obtain better estimators, but are not yet available for routine application. Until better estimates are available, we decided to stay with the original estimators. When better estimates become available (or local expertise is well documented) they can simply be substi- Flathead Basin Cooperative Program Final Report Page 99 Management Guidelines for Riparian Forests MONTANA "RISK" MATRIX (3/9i) TYPE CLEARCUT PARTIAL CUT SITE PREP ROADS SLOPE SOIL EROD. «!3 H 2** 0-5% M >*4 H 4* 5-20% M H 4* 5* 5* 3** 3*4 20-40% M 4* 4* H > 40% M 3* ( * Indicates Mean Value Used, ** Final Group Consensus Adjustment for Internal Consistency) Figure G-2: Montana "Risk" Matrix (3/91) Page 100 Flathead Basin Cooperative Program Final Report Management Guidelines for Riparian Forests tuted to place soils in one of the three classes. The practical guidelines for delineating Streamside Management 2^nes are based on current knowledge. As additional knowledge is gained from current studies and experience, SMZ criteria and BMPs should both be modi- fied to provide a proper balance between pro- tection of water quality and efficientiy meeting objectives of the landowners. A 25-foot minimum SMZ was originally recommended by the Environmental Quality Council ' s Best Management Practices commit- tee. The Montana Riparian Association Man- agement Guidelines Working Group discussed the minimum at their final meeting in March, 1 99 1 . After weighing the minor advantages and disadvantages of 25 feet vs. 50 feet, they agreed to recommend a 50 foot minimum as a reason- able compromise in the context of current BMPs within the Streamside Management Zone. Results - Soil Erodibility Matrix Risk Assessment The results of a group expert opinion con- sensus process (Delphi approach) to develop the soil/slope risk assessment matrix are sum- marized as Chapter One of Lull's Masters thesis (Lull 1990). TTie round two matrix was also used as an aid in conducting the 1 990 state- wide BMP field audits (Schultz 1990). The Working Group revised the matrix to obtain internal consistency at their meeting on March 6, 1991. This revised matrix is shown in Figure G-2. The Management Guidelines Working Group considered several alternatives to the parentmaterialclassesofClineandothers(1981) for estimating soil erodibility. Fairly close agree- ment was reached on soil texture in the round two Delphi questionnaire. However, current research by Ed Burroughs (USDA-Forest Serv- ice-Intermountain Station, Moscow, ID) sug- gests that the amount and shape of coarse fi-ag- ments may be equally important. Also, USDA- Northem Region Hydrologists are currently working on revised information on soil erodi- bility for input into improved sediment models. The Working Group decided (August 6, 1990) to defer further work on soil erodibility and wait for the results of the current efforts. When better estimates of soil erodibility become available they can be substituted to place soils in one of our three classes. We think it is important to keep the methods for estimating soil erodibility consistent for risk assessment, streamside man- agement zone formulas, and sediment models. Conclusions (1) Habitat type-specific management in- formation has now been provided in a 1990 draft publication (Boggs and oth- ers 1990). This information is far more suitable than previous generalities and better suited to aid field people in mak- ing site-specific evaluations. (2) Criteria have now been developed and evaluated for consistent field deline- ation of Streamside Management Zones to meet the intent of the Best Manage- ment Practices. Additional suggestions are provided for special situations where the general formula should be adjusted to provide additional protection. (3) A forest practices risk matrix in rela- tion to slope and soil erodibility is now available to evaluate relative risks. It can be applied for watershed level risk assessment of proposed alternative practices and for several other applica- tions and interpretations. Acknowledg- ment of relative risk is the first step in prescribing activities and possible miti- gation. Flathead Basin Cooperative Program Final Report Page 101 Management Guidelines for Riparian Forests Acknowledgments The following professionals have actively provided input and review for the development of these field guides as well as several related topics. They have all been working together as a Management Guidelines Working Group within the Montana Riparian Association: Larry Brown Vito Ciliberti Mike Enk Paul Hansen John Joy John Mandzak Greg Munther Bill Pumam Donald Potts Gordon Sanders Bill Schultz Dean Sirucek Dan Svoboda Page 1 02 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module H: Application of the Sequoia Method for Determining Cumulative Watershed Effects IN THE Flathead Basin By Don Potts This section of the Final Report summarizes a study of the same neime published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Paee 103 Flathead Basin Cooperative Program Final Report ^ Page 1 04 Flathead Basin Cooperative Program Final Report Application of the Sequoia Method FOR Determining Cumulative Watershed Effects IN THE Flathead Basin By Don Potts^ A cumulative watershed effects risk assess- ment procedure has been applied to the entire Swan River watershed and to 30 smaller Flat- head Basin watersheds. The Swan and the 30 smaller watersheds were selected for study by the Flathead Basin Water Quality and Fisheries Cooperative because of their high fisheries val- ues and growing concerns over possible im- pacts from forest management practices. The risk assessment model, Sequoia, was developed by the hydrology, soil science and forest management staff on the Sequoia Na- tional Forest in 1980. It is basically an account- ing system for areal disturbance. The various forest management activities are assigned a Runoff Coefficient that varies with the degree of site compaction and soil exposure. The coef- ficients range from a high of 1.0 for permanent harvest system roads to a low of 0. 1 for cable system partial cuts and low intensity fires (ten percent soil exposure). (See Table H-1.) The Runoff Coefficient times the area disturbed is called the Cumulative Runoff Acreage (CRA, or in other Region 5 methods, the Equivalent Road Acreage). Various assumptions are made about the actual areas disturbed by roads, trails, skid systems, and landings. (See Table H-2.) The procedure assumes that disturbance from all timber harvest-related activities recovers within ten years except for roads, trails, recrea- tion, and administrative sites which never re- cover. The basic premise of Sequoia is that soil compaction and soil exposure effectively in- crease the drainage efficiency of a watershed, thus increasing the magnitude of peak flows, which in turn may cause destabilization of chan- nels and deterioration of fisheries and water quality. Based on research conducted in Oregon and California, the procedure recommends a Threshold of Concern (TOC) for watersheds with "average sensitivity" when the Cumula- tive Runoff Acreage reaches 12 percent of the watershed. The Swan River watershed was partitioned into 54 analysis units ranging in size from roughly 1,400 to 23,000 acres. (See Figure H- 1.) Many of the analysis units, particularly those at higher elevations, have boundaries corresponding to actual watershed boundaries. All forest management activities during the past decade and all existing road information in each of the units was obtained from the land owners. Sequoia estimates of areal disturbance ranged from a high of nearly 40 percent to a low of zero. Thirteen of the analysis units had dis- turbance greater than Sequoia's 12 percent threshold of concern. Nearly 1 1 percent of the Swan River watershed received some sort of harvest treatment during the 1980s. This in- volved over 750 miles of temporary or perma- nent roads totalling a Cumulative Runoff Acre- 'Dr. Don Potts is a Professor with the Univer- sity of Montana's School of Forestry in Missoula, Montana. Flathead Basin Cooperative Program Final Report Page 105 Application of the Sequoia Method ^^^ Table H-1. Runoff Coefficients and Recovery Rates YEARS - ACnVITY 0 1 2 3 4 5 6 7 8 9 10 rractor Clearcut .4 .4 .35 .2 .1 .1 .1 .1 ,1 1 1 Cable Clearcut .2 .2 .2 .15 .1 .1 .1 .1 .1 1 1 Tractor Partial .2 .2 .15 .1 .1 .1 .1 .1 ,1 1 1 Cable Partial .1 .1 .1 .1 0 _ . . _ Site Prep Mech, .7 .7 .6 .5 .3 .2 .1 .1 .1 .1 1 Median. Release .5 .4 .4 .3 .25 .15 .1 .1 0 . . Aband. Roads .9 .9 .9 .9 .9 .9 .9 .9 .9 ,9 9 Perm. skid sys. .9 .9 .9 .9 .9 .9 .9 .9 .9 ,9 9 Bums 10% soil .1 .1 .1 0 _ _ . . Bums 80% soil .4 .4 .35 .3 .2 .1 0 _ ORV Trails .9 .9 .9 .9 .9 .9 .9 .9 .9 ,9 9 System Roads 1 1 1 1 1 1 1 1 1 1 1 Page 106 Flathead Basin Cooperative Program Final Report Application of the Sequoia Method Table H-2. Extent of Activities — Equivalent Acres Activity Tractor Clearcut Cable Clearcut Tractor Partial-cut Cable Partial-cut Mechanical Site Prep. Mechanical Release Abandoned Roads Rec. & Admin. Sites Perm. Skid System/landings Bums ORV Trails System Roads Equivalent Acres Harvested Acres Harvested Acres Harvested Acres Harvested Acres Treated Acres Treated Acres Miles X 2 Acres Acres of Sites Harvested Acres x 27% Acres Burned Miles x 1.5 Acres Miles X 3.5 Acres age of about eight percent of the watershed in 1989. Sequoia was similarly applied to the 30 smaller watersheds located within the Flathead Basin. Cumulative Runoff Acreages in these critical fisheries ranged from 0 percent in Elk Creek and Lion Creek to over 30 percent in Freeland Creek. In addition to Freeland Creek, Cumulative Runoff Acreages in Jim Creek, Fish Creek and Sheppard Creek were above Sequoia's 12 percent threshold of concern. The Squaw Creek tributary and Hand Creek are approaching the threshold. (See Table H-3.) The Flathead National Forest and the Mon- tana Department of State Lands currently use similar water yield models (H^OY), based on the Equivalent Clearcut Area (ECA) concept. Rather than measuring compaction and soil exposure in anticipation of changes in peak discharges, the ECA models measure canopy removal in anticipation of changes in average annual water yield. In the Flathead Basin, it is assumed that channels with "normal" stability can withstand an increase of ten percent in annual water yield. The ECA and the CRA model predictions should be correlated — you can't remove canopy without soil compaction and exposure. Nevertheless, the models meas- ure different impacts and have different under- lying assumptions. The ECA model was applied to the same 30 critical fisheries watersheds as Sequoia, and in the same rank order Freeland, Fish and Shep- pard Creeks were judged to be above the thresh- old of concern. The ECA model did not, how- ever, find Jim Creek, the Squaw Creek tributary or Hand Creek to be at or near the threshold. Hopefully, these results, when compared and Flathead Basin Cooperative Program Final Report Page 107 Application of the Sequoia Method Figure H-IA: Swan Watershed Analysis Units — North Pag»108 Flathead Basin Cooperative Program Final Report Application of the Sequoia Method Figure H-IB: Swan Watershed Analysis Units — South Flathead Basin Cooperative Program Final Report Page 109 Application of the Sequoia Method Table H-3. Comparison of model results - WATERSHED SEQUOIA (% DIST) H2OY {% INCR) 1A ELKCREEK 0.00 0.00 1B ELKCREEK 0.00 0.00 2 QCATCREB< 0.01 0.90 3 SQIFF7FR 0.04 0.46 4 UON 0.00 0.00 5A JIM 12.20 4.49 5B JIM 13.00 4.26 6 PIPER 0.00 0.00 7 FREELAND 30.90 22.00 8 FISH 22.60 16.78 9 HAND 9.50 1.55 10 UPPER EF SWIFT >0.00 NA 1 1 SHEPPARD 15.70 10.24 12A UPPERBIG 1.20 2.93 12B LOWERBIG 3.00 3.32 13 LDWERCOAL 2.30 1.58 14 COALCREEKNF 6.70 3.03 15 COAUCREEKSF 3.90 1.70 17 REDMEADOW 3.40 1.83 18 WHALE 1.20 1.27 19 TRAIL 2.30 0.65 20 GRANIIH 1.90 1.72 21 CHAIIFNGE 2.50 1.03 23 MORRISON 2.50 1.30 24A HUNGRYHORSE >0.00 0.60 24B HUNGRYHORSE >0.00 0.50 25 MARGARET >0.00 1.64 26 TIGER >0.00 0.11 27 EMERY 2.00 3.14 29 SQUAWTRIB 11.10 3.62 Page 110 Flathead Basin Cooperative Program Final Report Application of the Sequoia Method correlated with the other Cooperative studies, will allow us to gain understanding of cumula- tive watershed effects in the Flathead Basin. Summary The Sequoia Method for cumulative water- shed effects risk assessment was chosen for use in the Flathead Basin for a number of reasons. The procedure is representative of the official U.S. Forest Service Region 5 (California) pro- cedures. This model or others similar in appli- cation and data requirements are currently used by the National Forests in Region 5. Sequoia is computationally straight-for- ward. The most difficult obstacle is obtaining all of the information on harvesting, site-prepa- ration, and roads from the various land-owners in mixed ownership watersheds. All of the data required by Sequoia arc routinely maintained by landowners. Finally, since Sequoia is an accounting sys- tem for areal disturbances, it is actually compat- ible with our growing Geographic Information Systems capabilities. All of the information required by Sequoia is found on the Flathead National Forest Geographic Information Sys- tem (GIS). Just as Region 5 has made the CRA methods official analysis tools, the Montana Cumulative Watershed Effects Cooperative has made a model much like H^OY the new "official" model for Region 1 . The understanding in the coopera- tive is that members will work with the new model to resolve its shortcomings. Neverthe- less, many of the ECA logic flaws will never be worked out. The most critical of these is that increases in average annual flow are related to channel forming processes. Again, Sequoia is an accounting system that does measure things we know to be related to peak flows and channel forming processes. We have begun working on a computer version of Sequoia for use with microcomputers. When it becomes available, the model will be in a form easily usable by all land managers in the Flathead Basin. Flathead Basin Cooperative Program Final Report Page 111 Page 112 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module I: A Forest Management NoNPOiNT Source Risk Assessment Geographic Information Systems Application ' By Don Potts This section of the Final Report summarizes a study called Development of a Geographic Information Systems Application for Assessment ofNonpoint Source Pollution Risk on Managed Forest Lands published by the Flathead Basin Commission (723 Fifth Avenue East, Kalispell, Montana 59901). Flathead Basin Cooperative Program Final Report ^^® Page 1 14 Flathead Basin Cooperative Program Fined Report A Forest Management Nonpoint Source Risk Assessment Geographic Information Systems Application By Don Potts' Introduction Risk Matrix Construction The assessment of nonpoint source pollu- tion and cumulative watershed effects resulting from forest management is proving to be a difficult task. Application of Best Management Practices (BMPs) does not provide an assess- ment of condition, monitoring can be prohibi- tively expensive, and our quantitative modeling skills are in their infancy. An alternative ap- proach, watershed risk analysis, may, however, provide resource managers with a powerful tool to assist in land-use decision making. In addi- tion, Geographic Information Systems (GIS) allow the correlation of land cover and topo- graphic information such as terrain configura- tion and drainage networks, thus making GIS useful in assessing the potential effects of land- use activities on water resources (Walsh 1985). We have developed a methodology to assist forest managers in assessing nonpoint source pollution and cumulative watershed effects by combining watershed risk assessment with GIS . The information required by the procedure is available in the Rathead National Forest GIS. Making a map with GIS is easy, but including appropriate landscape attributes is difficult. The key is the construction of a nonpoint source pollution risk matrix. The results of such an application are presented on a heavily impacted, mixed-ownership watershed. Leopold and others (1971) were among the first to propose the use of an environmental matrix to evaluate potential environmental im- pact. The matrix provides a comparatively simple system intended to be a guide in the impact assessment process. This approach was adopted by Rickert and others (1978) to evalu- ate erosion potential for the Oregon 208 non- point source assessment project and by Brown in and others (1979) as a guide for land-use planning. The "risk" matrices developed in these stud- ies assessed the relative impacts of various land-use activities on different combinations of slope and substrate erodibility. Forest manage- ment practices were among the land-use activi- ties considered, but Brown in and others sug- gested that matrices should be developed re- gionally to reflect local conditions, practices, and socioeconomic considerations. Therefore, we contacted a large group of soil and water specialists and silviculturists, representing land management agencies and the forest products industry in Montana, to build the Montana erosion-impact matrix. We quickly learned that it was not going to be as easy as we thought. There was considerable 'Dr. Don Potts is a Professor with the Univer- sity of Montana's School of Forestry in Missoula, Montana. Flathead Basin Cooperative Program Final Report Page 115 A Forest Management Non point Source Risk Assessment disagreement over the relative risk values, on a simple scale of 1 to 5, to be used in the body of the matrix as well as the basic combinations of slope and erodibility that would produce differ- ent potential impacts. We convened the group of experts four times and, through the Nominal Group Tech- nique, arrived at the consensus matrix appear- ing in Figure I-l. The horizontal axis is com- posed of 4 principal forest management prac- tices: (1) clearcutting, (2) partial cutting, (3) site preparation, and (4) road construction. Each principal practice is further split into spe- cific methods, treatments, or designs. The group decided that four slope classes on the vertical axis were sufficient and distinctly different in inherent erosion risk, and that three basic soil erodibility classes were also ade- quate. These classes were based primarily on the geologic erosion factors described in the R1-R4 Sediment Yield Prediction Procedure (USDA Forest Service 1981). Everyone agreed that alluvial and granitic soils were highly erod- ible, those formed from Belt Series metamorph- ics had low erodibility, and that most other substrates fell into the moderate erodibility class. Howard Creek The Howard Creek watershed, totalling 501 5 hectares, is located approximately 32 kilo- meters southwestof Missoula, Montana. Eleva- tion in the watershed ranges from 1 190 to 1770 meters. Average annual precipitation at mean elevation is about 100 cm, and roughly 60 percent of that falls as snow. The forest cover is primarily Douglas-fir mixed with sub-alpine fir at the higher elevations and Ponderosa pine at the lower elevations and south aspects. The main stem of Howard Creek is oriented east- west while the three sub-basins. Tepee Creek, North Fork of Howard Creek, and Krystal Creek, are oriented north-south. (See Figure 1-2.) This map, and the following maps were prepared using PAMAP (1989) CIS software and infor- mation hand-digitized at a scale of 1:24000. The Lolo National Forest Land Systems Inventory maps the primary geologic groups found in Howard Creek as metasedimentary and undifferentiated materials. (See Figure I- 3.) Metasedimentary parent materials are de- rived from Belt S uper Group quartzite , argillite, and siltite. Rock fragment hardness is variable depending upon the degree of rock weathering. Weathering is dependent on associated faults, preponderance of argillites, and calcium car- bonate content. These materials were classified as either L-low or M-moderate erodibility for the erosion-impact matrix. Undifferentiated geology is composed of materials derived from Belt Super Group metasedimentary rocks or weakly weathered granitic rocks. Materials include alluvium on terraces and flood plains; shallow soils on flood scoured foot slopes and stream breaklands, strongly frost churned broadly convex ridges, and glacial outwash on plains. These are classi- fied as "H" — highly erodible in the Montana erosion-impact matrix. Like much of western Montana, the water- shed has a "checkerboard" pattern of land own- ership. Champion Timberlands owns 34 per- cent of the watershed. Plum Creek Timber owns 23 percent of the watershed, and the remaining 42 percent is managed by the Lolo National Forest. Ownership was mapped by the GIS and is shown in Figure 1-4. Timber harvest-related activities between 1981 and 1986 impacted 17 percent of the watershed. The location of these activities is shown in Figure 1-5. In a report to the Lolo Forest Supervisor (Munther and others 1987), cumulative watershed effects in Howard Creek were estimated to have produced a sedi- ment load increase of 50 percent and water yield Page 116 Flathead Basin Cooperative Program Final Report A Forest Management Nonpoint Source Risk Assessment MONTANA "RISK" MATRIX (3/9i) TYPE CLEARCUT PARTIAL CUT SITE PREP ROADS SLOPE SOIL EROD. PS OSCQ 09 H 2** 0-5% M H 4* 5-20% M H 4* 5* 3** 20-40% M 4* H > 40% M ( * Indicates Mean Value Used, ** Final Group Consensus A4justment for Internal Consistency) Figure I-l: Montana "Risk" Matrix (3/91) Flathead Basin Cooperative Program Final Report Page 117 A Forest Management Nonpoint Source Risk Assessment Page 118 Flathead Basin Cooperative Program Final Report A Forest Management Nonpoint Source Risk Assessment Flathead Basin Cooperative Program Final Report Page 119 A Forest Management Nonpoint Source Risk Assessment Page 120 Flathead Basin Cooperative Program Final Repor A Forest Management Nonpoint Source Risk Assessment Flathead Basin Cooperative Program Final Report Page 121 A Forest Management Nonpoint Source Risk Assessment increase of 8 percent. The Forest Supervisor immediately imposed a 10-year timber harvest moratorium on Forest Service lands. Industry has cooperated, and there has been little activity in the watershed during the past three years. A QjMULATivE Effects Risk Intdex Value An undisturbed watershed has a zero risk index value. The risk index value for a managed watershed depends on the areal extent of activi- ties, the age of the treatments, and the type of terrain on which they are located. The calcula- tion of a cumulative effects risk index value is straightforward and is outiined in the following eight steps: Step 1: Determine the year/type/acreage of past (or planned) forest management activities. Permanent road disturbance is calculated at 0.87 ha/km; temporary road disturbance is calculated at.37 ha/ km. Digitize this information for use in the GIS if it has not already been done. Step 2: Determine the erosion classification of the terrain on which the activity has taken (or will take) place. This is gen- erated by the GIS. Step 3: Determine slope category on which the activity has taken (or will take) place using GIS. Step 4: Obtain a risk value for each activity from the Erosion-Impact Matrix that has been developed. Step 5: Determine the areal percentage of the watershed on which the activity took place. Step 6: Use a recovery coefficient (RC — a value between 1 and 0) to reduce the risk associated with older activities. In this application, we assumed a linear 10-year risk recovery for all activities except roads, which always maintain a RC of 1. Thus, current year activities have a RC of 1 , four- year-old activities have a RC of 0.6, and so on. Step 7: Multiply the risk value times the percentage of the watershed times the RC to obtain the total risk for that activity. Step 8: Sum all past and planned land-use totals. Table I-l contains a summary of distur- bance information, by year, for Howard Creek. Harvest, site preparation and road data were obtained directly from the three land owners in the watershed. To avoid double accounting, only the last entry onto a site was counted. Usually this was for site preparation, which was almost always dozer piUng of slash. If no site preparation had been completed, then harvest- ing was the only disturbance counted, and this was almost always partial cutting with tractor skidding. Note that in 1990, the cumulative risk index value is 0.2284, but if the risk index value had been calculated in 1986, the year the timber harvest moratorium was called, it would have been 0.3998. The differences between the 1990 and 1986 weighted risk columns in Table I-l are due only to the effect of the recovery coef- ficients. Thresholds? How much disturbance a watershed can tolerate before reaching a level that produces significant environmental damage is both tech- nically and politically difficult to answer. In the Forest Service's Region 5, a watershed's "Threshold for Concern" is typically an areal disturbance (actually, equivalent road acres) of between 10 percent and 20 percent depending Page 122 Flathead Basin Cooperative Program Final Report A Forest Management Nonpoint Source Risk Assessment Table I-l. "Risk" Calculations for Howard Creek, Lolo National Forest, Montana. % of Total 1990 1986 Area Area Weighted Weighted Year Activity (ha) Risk Risk 1981 Site Prep. 14 0.27 .0004 .0020 1983 Site Prep. 177 3.53 .0342 .0798 1984 Harvest only 36 0.72 .0065 .0130 Site Prep. 390 7.78 .0650 .1300 1985 Harvest only 92 1.83 .0275 .0495 Site Prep. 84 1.67 .0160 .0288 1986 Harvest only 6 0.12 .0021 .0035 Site Prep. 53 1.06 .0247 .0412 Permanent Roads 61.4 1.22 .0360 .0360 Temporary Roads 21.3 0.42 .0160 .0160 Unrecovered Disturbance 18.62% 1990 Cumulative "Risk" 1986 Cumulative "Risk" .2284 .3998 Flathead Basin Cooperative Program Final Report Page 123 A Forest Management Nonpoint Source Risk Assessment on the watershed's sensitivity (Coboum 1989). These values are based on research conducted by Hair and others (1975). If Howard Creek were in Region 5, its combination of slopes and geology would proba- bly classify it as moderately sensitive, and there- fore would have a "Threshold for Concern" of 15 percent. Note in Table I-l that six years of management produced an unrecovered distur- bance of nearly 19 percent. The "yellow flag of caution" would not yet have been raised be- cause of recovery and only some percentage of the disturbance was equivalent road acres. The specialists on the Lolo National Forest raised their own "yellow flag of caution" based on monitoring information. Nevertheless, there were some indications of changes in the system, so perhaps a 10 percent to 15 percent distur- bance threshold is legitimate. Similar rationale for assigning a risk index value threshold may be used. A moderate risk for generating cumulative effects has a value of 3 in the Erosion-Impact matrix. If we were to chose a 15 percent areal disturbance limit, and be willing to accept moderate risk, then the risk index target would be 0.45 (0.15 X 3). A ten percent disttirbance limit would have a risk index of 0.3. Similarly, acceptance of higher or lower risk could also increase or decrease the target. In either event, the Howard Creek water- shed was being stressed, or nearly so, in 1986. Recovery is rapid, however, so deferring activ- ity or selecting low risk options (a combination of sites and methods) would allow the cumula- tive risk value to fall. Note that by 1990, the cumulative risk index value fell to about 0.23. Summary This methodology has not yet been applied to any watersheds in the Flathead Basin. Care was taken, however, to make sure that all infor- mation required for the procedure is ahieady available on the Flathead National Forest GIS. We see distinct advantages that this risk assess- ment procedure offers over other cumulative effects analyses. First, machine processing of data tends to minimize "human errors." Sec- ondly, the technique allows us to weight our assessment of risk by our understanding of both erosion and peak flow generation processes and the inherent impacts associated with various management activities. We strongly recommend testing the proce- dure in the Flathead Basin, using the Flathead National Forest's GIS capability and the data sets obtained for the other watershed analyses conducted by the Cooperative, and comparing the results with those of the other techniques. Page 124 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Study Module J: Linear Correlation/Regression Analysis of Forestry Models, Risk Assessment, and Water Quality and Fisheries Data By Dean Sirucek, Elizabeth Hill, F. Richard Hauer, John Fraley, Thomas Weaver, Don Potts, and Steve Tralles Flathead Basin Cooperative Program Final Report ^^® Page 1 26 Flathead Basin Cooperative Program Final Report Linear Correlation/Regression Analysis OF Forestry Models, Risk Assessment, AND Water Quality and Fisheries Data By Dean Simcek,^ Elizabeth Hill,^ F. Richard Hauer,^ John Fraley/ Thomas Weaver,^ Don Potts,^ and Steve Tralles^ Introduction Animportantaspectof thecooperative study was to determine whether there are meaningful, quantitative linkages between the type and in- tensity of forest management activities within a watershed and measured effects to the stream ecosystem. Some of the nine previous studies were designed to investigate quantitative or qualitative water quality or fisheries habitat monitoring parameters. Other studies applied modeling procedures to test their validity in the Flathead Basin. However none of the studies focused on directly linking computer generated predictions to the measured field parameters. We designed Module J to compare computer model predicted effects of timber harvest di- rectly to actual quantitative and qualitative water quality or fisheries habitat parameters. Several state-of-the-art watershed computer models are used in modem forestry to predict changes in stream flow volume, regime, and sediment yields from proposed forest manage- ment activities. Different watershed models examine various parameters using a number of methods to predict potential changes in the aquatic ecosystem. The Sequoia model (as dis- cussed in Module H), is an index of cumulative disturbed areas. The procedure assumes that disturbance from all timber harvest-related ac- tivities recovers within ten years except for roads, trails, recreation, and administrative sites which never recover. The Module H study team developed Sequoia CRA (Cumulative Runoff Acreage) on the 28 fishery watersheds investi- gated in Module D. Currently the Flathead National Forest and the Department of State Lands use similar com- puter models to predict water yield increase. These models are based on an Equivalent Clearcut Area (EC A) concept — that is, an increase in average annual water yield will occur following removal of trees from a site. (As water yield increases, the potential for streambank erosion increases. As the trees grow 'Dean Sirucek is aForest Soil Scientist with the Flathead National Forest in Kalispell, Montana. ^Eli2abeth Hill is a Forester with the Flathead National Forest in Kalispell, Montana. 'Dr. F. Richard Hauer is a Research Associate Professor with the Flathead Lake Biological Station in Poison, Montana. *John Fraley was a Fisheries Biologist with theMontana Department of Fish, Wildlife and Parks in Kalispell, Montana. Thomas Weaver is a Fisheries Officer with the Montana Department of Fish, Wildlife and Parks in Kalispell, Montana. 'Dr. Don Potts is a Professor with the Univer- sity of Montana School of Forestry in Missoula, Montana. 'Steve Tralles is an Environmental Specialist with the Water Quality Bureau of the Montana Department of Health and Environmental Sciences in Helena, Montana. Flathead Basin Cooperative Program Final Report Page 127 Linear Correlation/Regression Analysis of Forestry Models back on the site, the water yield is reduced over time.) The ECA model used on the Flathead National Forest is namedH^OY (Isaacson 1977). In 1990, the Montana Cumulative Effects Cooperative (which includes the Flathead Na- tional Forest, Department of State Lands, and Plum Creek, L.P.) adopted and then updated a watershed computer model for analysis of tim- ber harvest effects in multiple-ownership wa- tersheds. This updated model (named WATSED) combines a sediment yield model (Cline and others 1981) and an ECA water yield predictive model (U.S.D.A. Forest Service Re- gion 1 1972). WATSED predicts the natural levels of water yield and sediment production along with increases due to timber harvest The WATSED computer model develops predic- tions for a water yield model based upon an ECA concept. The sediment yield portion of WATSED predicts soil erosion from roads, timber harvest, and fire. The purpose of this module was to deter- mine the relationships between the quantitative studies and current forest management model- ing tools. We applied land use data to the three computer models. Numerical outputs were then applied as independent variables to linear cor- relation/regression analyses. Results from the water quality and fisheries modules were ap- plied as dependent variables. The goal of this effort was to test the statistical significance of those regressions and to determine whether there is a relationship between what is predicted by the management models and the empirical data collected in the water quality and fisheries studies. Linear Correlation/Regression Analysis The quantitative field measurements re- viewed included the following: suspended sedi- ment (Flathead National Forest database), total suspended solids (Module C), streambed sedi- ment (Module D — McNeil sediment cores and Whitlock- Vibert box sediment), nutrients (Mod- ule C), maximum algae growth (Chlorophyll { — Module C), and frequency of bull trout and westslope cutthroat. trout redds (Module D). Qualitative procedures were conducted in the stream reaches immediately upstream from the water quality or fishery monitoring sites. These procedures included the following: Region-1 stream channel stability rating (Flathead Na- tional Forest database, Pfankuch 1978), sub- strate score (Module D), and Montana Water Quality Bureau's state-wide stream reach as- sessment procedure (Module E). This repre- sents all available historic and current informa- tion collected for each watershed. The ECA model HjOY was run on the 28 fisheries study watersheds. Model output for HflY is water yield increase, which was used for comparison to the water quality and fisheries data. The cumulative watershed effects risk assessment procedure. Sequoia (CRA = Cumulative Run- off Acreage — Module H) was applied to the 28 fisheries study watersheds. Because WATSED was being updated, it was unavailable for use until late in this coop- erative effort. Therefore we did not have time to run data sets on all 28 fishery watersheds. For this reason, we chose from the 28 fishery water- sheds, the 10 that had the longest history of water quality and fish habitat monitoring data. There were five variables from WATSED that were used for the correlations: (1) equivalent clearcut acreage — cumulative, (2) water yield increase (WYI) — annual basis, (3) water yield increase — 75 percent peak flow duration change, (4) natural sediment yield — annual basis, and (5) sediment yield increase — annual basis. See Table J-1 for the list of study water- sheds and the variables measured. Refer to Figure D-1 for watershed locations. The Region 1 Channel Stability Ratings (CSR) were conducted on the same ten stream Page 128 Flathead Basin Cooperative Program Final Report Linear Correlation/Regression Analysis of Forestry Models Table J-1: List of study sites by watershed and the application of water yield increase and/or sediment yield models (H2OY and WATSED), the Sequoia risk analysis, and the study sites that were investigated in the water quality module (C), the fisheries module (D), and the Montana Water Quality Bureau's qualitative reach assessment module (E). WATERSHED H2OY WATSED SEQUOIA MODULE C MODULE D MODULE E Lion Creek X X X X X X Chepat Creek X X X X X Coal Creek X X X X X Challenge Creek X X X X X Upper South Fork Coal Creek X X X X X X LowCT South FoTk Coal Creek X X North Coal Credc X X X X X Hand Creek X X X X X X Squaw Creek X X X X X Jim Creek X X X X X X Fish Creek X X X X X Elk Creek X X X X Goat Creek X X X X Squeezer Creek X X X X Piper Creek X X X X Freeland Creek X X X Shq)pard Creek X X X X Big Creek X X X X UppCT Red Meadow Creek X X X X Whale Credc X X X X Trail Creek X X X X Granite Creek X X X Morrison Creek X X X X Ole Creek X X X X Hungry Horse Creek X X X X Tiger Creek X X X X Margaret Creek X X X X Emery Creek X X X X Lower Fitzsimmons Creek X East Foric Swift Creek X X X X Upper Fitzsimmons Creek X X X X Lower Red Meadow Creek X X Deano Creek X Flathead Basin Cooperative Program Final Report Page 129 Linear Correlation/Regression Analysis of Forestry Models Figure J-1: Watershed Study Sites FLATHEAD NATIONAL FOREST Page 130 Linear Correlation/Regression Analysis of Forestry Models Table J-2: Independent and dependent variables applied to the linear correlation/regression analysis. Land use measures, risk assessment value, water yield model output, qualitative reach assessment, and qualitative estimates of substrate score were applied as independent variables. The quantitative in- stream measures were applied as dependent variables. Independent Variables Dependent Variables % of Watershed Harvested % of Watershed in Roads Sequoia risk procedure H2OYWYI WATSED cumulative acres WATSED annual WYI WATSED WYID75 WATSED increased sediment * Region 1 Channel Stab. Rating * Change in CSR * WQB stream reach assessment * Substrate Score FNF suspended sediment Total Suspended Solids McNeil Core Sediment Whitlock- Vibert box Sediment Total Nitrogen (persulfate) Total Phosphorus Algae Growth (Chi a) Bull Trout Redds * Region 1 Channel Stab. Rating * Change in CSR * WQB stream reach assessment * Substrate Score The Water Quality Bureau's qualitative stream reach assessment (Module E), the fishery module's substrate score (Module D), and channel stability ratings were examined both as independent and dependent variables. reaches directiy upstream from the WATSED sites. (See Table J-1.) The Channel Stability Rating is a qualitative method of assessing the vulnerability of a stream reach to degradation if increased water flow were added to the stream. Whenever possible, historic CSR values are researched for the same stream reach to see if any major changes occurred over time. All pair- wise regression analyses were con- ducted on twelve independent and twelve de- pendent variables. (See Table J-2 for the vari- ables and Appendix E for the data set.) Model data output in a percentage format was trans- formed using an arcsine transformation. The regression analyses were performed using Stat- View II statistical computer program. All sig- nificant relationships and accompanying statis- tics are given in Table J-3. Non-significant relationships were not listed. Flathead Basin Cooperative Program Final Report Page 131 Linear Correlation/Regression Analysis of Forestry Models Table J-3: Statistically significant regressions of independent and dependent variables; where n = number of cases, p = significance level (only those < 0.1 were considered significant), and r^ = coefficient of determination. Figure Comparison n P r2 J-2 Sequoia CRA vs. McNeil Core J-3 Sequoia CRA vs. Substrate Score J-4 Sequoia CRA vs. Maximum algae growth J-5 Sequoia CRA vs. Maximum algae growth J-6 H20Y WYI vs. McNeil Core J-7 H20Y WYI vs. Substrate Score J-8 H20Y WYI vs. Total Persulfate Nitrogen J-9 H20Y WYI vs. Maximum algae growth J-10 WATSED WYI vs. Total Persulfate Nitrogen J-11 WATSED 75% Pk Flow Duration vs. Total Persulfate Nitrogen J-12 WATSED WYI vs. Maximum algae growth J-13 WATSED 75% Pk Flow Duration vs. Maximum algae growth J- 14 % of Watershed Harvested vs. Total Phosphorus J- 1 5 % of Watershed Harvested vs. Total Persulfate Nitrogen J-16 % of Watershed Harvested vs. Maximum algae growth J-17 H20Y WYI vs. Change in Channel Stability Rating J-18 Sequoia CRA vs. Change in Channel Stability Rating J-19 WATSED WYI vs. Change in Channel Stability Rating J-20 WATSED 75% Pk Flow Duration vs. Change in CSR J-2 1 WATSED Predicted Sediment vs. FNF Suspended Sediment 28 0.02 0.19 28 0.01 0.23 10 0.08 0.34 10 0.08 0.34 28 0.03 0.16 28 0.004 0.28 10 0.04 0.43 10 0.03 0.49 6 0.07 0.59 6 0.08 0.57 6 0.05 0.66 6 0.002 0.92 12 0.015 0.46 12 0.002 0.63 12 0.04 0.36 6 0.003 0.91 6 0.004 0.90 6 0.012 0.824 6 0.001 0.943 8 0.0002 0.92 Page 132 Flathead Basin Cooperative Program Final Report Linear Correlation/Regression Analysis of Forestry Models Discussion We found statistically significant relation- ships among several paired variables. Sequoia CRA was positively correlated with increased McNeil core measurements (Figure J-2) and substrate score (Figure J-3) at the 28 fisheries study watersheds and with total nitrogen (Fig- ure J-4) and algae growth (Figure J-5) at ten water quality study sites. There was also a statistically significant regression between HjOY water yield increase and these same four dependent variables. (See Figures J-6 to J-9.) Because of data availability, we were only able to match 6 study sites with the WATSED water yield increase model. WATSED WYI and WATSED 75% peak flow duration were sig- nificantly correlated with total nitrogen con- centration (Figures J-10 and J-11) and maxi- mum algae densities (Figures J-12 and J-13). No significant relationship was observed be- tween die WATSED WYI or WATSED 75% peak flow duration values and any sediment measurements. We also found that the percent of harvest within a watershed was positively correlated with total nutrient concentrations (total phosphorus and total nitrogen) and with algae growth. (See Figures J-14 to J-16). The Montana WaterQuality Bureau's stream reach assessment was done on 26 of the study watersheds. We found no statistically signifi- cant correlations between the stream reach in- dex and any other independent or. dependent variables. Because Channel Stability Rating and fish habitat measurements have common parame- ters, it was hypothesized that there would be a correlation between the variables. However, there were no statistically significant correla- tions between the CSR and any of the three fish habitat measurements. CSR values for those study sites that had historic CSR values were compared to the new CSR values to measure the change over time. This change in CSR was hypothesized to be a result of water yield increase. We did not find any statistically significant relationships be- tween the change in CSR and any of the quan- titatively measured dependent variables. How- ever, there were statistically significant rela- tionships between the change in CSR and the water yield models and risk assessment (H^OY water yield increase. Sequoia CRA, WATSED water yield increase, and WATSED water yield increase — 75 percent peak flow duration). (See Figures J-17 to J-20.) We observed a highly significant relation- ship between WATSED predicted sediment and the measured sediment yields from the Flathead National Forest water quality moni- toring. In order to do the correlation of sampled suspended sediment to WATSED predicted sediment, the predicted natural background values and the management-caused sediment increases were added together. The correlation between the measured suspended sediment val- ues and the predicted values was highly signifi- cant with an r' = 0.919 and p < 0.(X)1 — an extremely high level of significance. (See Fig- ure J-21.) We have the following observations of the relationships that were tested on the sample watersheds. First, we observed a statistically significant relationship between increased tim- ber harvest (as calculated by Sequoia and H^OY) and increased substrate sediment within known bull trout spawning areas. Although only a fifth of the variation in the substrate sediment values and substrate scores was explained by timber harvest disturbance (Sequoia and H^OY), the regressions were positively correlated with sig- nificance levels better than 95%. However, to use the watershed disturbance models to predict timber harvest impacts to fish habitat would be undesirable because of the high level of vari- ance. Therefore, we recommend that more re- Flathead Basin Cooperative Program Final Report Page 133 Linear Correlation/Regression Analysis of Forestry Models search effort be put into developing sediment routing relationships (that is, delivery of eroded soil to the stream channel) so that upland activi- ties and changes in streambed materials can be better understood. Second, stream algae growth as measured by Chlorophyll a was significantly correlated to most of the model, risk assess- ment, or direct measurements of disturbance. However, to date changes in stream algae pro- duction have not been used as an index of water quality change or effects on beneficial use. We recommend that more study is needed of algae growth relationships to the stream food web and the potential use of stream algae growth as an indicator of water quality. Third, the hypoth- esized relationship that channel stability ratings would change over time due to water yield increases was statistically established. Figure J-2 00- 0 0 80- 00 0 0 60, ^° ^^^^"^^ ^^^^J-*"© 0 40-1 r^ 0 ° ' 20 1 ° ( 10 20 30 40 bO-i Figure J-3 50< 40- 1 0 30t "^^ (DO 20- 10- 0 0 ^^\0 0 0 ^\.^^^ 0 0 10 20 30 4C 300- 0 Figure J-4 % ^^ S 200- ( 100- ^^ 0 ^^ 0 o ° ( ( 1 c 10 20 3C 0 Figure J-5 05- 04- 0 >^ 0 ^/^ 03- y^ 02- 01- /^ 0 OOI - 0 • 1 ■ Page 134 Flathead Basin Cooperative Program Final Report Linear Correlation/Regression Analysis of Forestry Models (1 1 ^ 0 Figure J-8 -^^"'^0 0 0 0 2 4 6 8 10 15 300- y Figure J- 10 y 1 / ^ I 2 200- < 100- ( 0- 1 / 0 ,.,.,. 0 0.5 . Figure J-9 04 . s 0^ 0.3 ^^ 02 ^^ 01 ^^ 0 00^ ^y^^ ^ , . ° , 0 2 4 6 8 10 12 H20YWYI% WATSED ANNUAL WYI% WATSED WYI % (75% PEAK DURATION) Flathead Basin Cooperative Program Final Report Page 135 Linear Correlation/Regression Analysis of Forestry Models Ob- o Figure J-12 0.4- / y 0.3- ^ 0,2- 0.1- i 0.0' / 0 1 / / 0 Figure J-13 0.4- 0 y / 0 0.3- / 0.2- / 0.1- / 0.0 4x:-^ — 1 — ^ — 1 — ^ — WATSED ANNUAL WYI% WATSED WYI% (79% PEAK DURATION) 10 20 % of BASIN HARVESTED E 200- <' 0 0 0 0 Figure J-15 0 0 % of BASIN HARVESTED ChanrMi Stability Rating Figure J- 17 30. o/ 20 . / 10 ■ 0 / in / 0. / -10 ■ Z " -20- %ol BASIN HARVESTED Page 136 Flathead Basin Cooperative Program Final Report Linear Correlation/Regression Analysis of Forestry Models ll O<0 10 20 WATSED ANNUAL WY1% 30- Figure J-20 y/O ? 70- y^ •s^ y / 5^ 10- o y^ "Nin^«-coo^a>ooOM*- oo oooooooooo o — tA^oo)r^ako>ttOkhk9 o o o o o o o o o o o o d ^ ri at tf) ^ r> CM o ^ (0 CO on ^ r« oooooooooooooooooooooo oooooooooooooooooooooo ..._ cvioin«-v(oc»(or)ninNN'NM>- h>«ioo o o o o 2 r»ooiniocM^eMCMKCj«oo «io«r>onincMrvcMOP}a>iD«CMr>io ^cxnv oooooooooooooooooooo OOOOOOOOOOOOOOOOOOOO — vvvvv w>oj^o>«-a>or^naooa>^oor^*-^rva> r*^mT-o>« 5 r^ N fii "T rg in c» o o' rg V V o co r-' >o o> V ri ^ o o o o o o o o o o o o CMOoo^ vvcgncMn »♦ oooooooooo oooooooooo OOI^<00c^h«(»(»oc V »-" •- »- »- pg CM * o a> o o -^ o d at o oooooooooo oo(fi^<^r^co(Oir>otom(DCM t^ o «- CMrO'«tf) oooooooo > ooa>r~n OOCW10'-*0>0 §s in n i>." r«.' — oooooooooo > 6000000000 ^ ooiAr^ooioNnio O dd^MoiMoioino S •- ^ N ••- CM n in ■» o o o b b K ni ♦ »• n 10 N ni ♦ »• on* b b b 00000 00000 o) r^ CO CO CO r* to o *- 10 *- 000000000000 000000000000 r^ o> r« m o GO a» n in in CO »- »- c» ooi«Dc*>mr>>«(Mr)0«-*r *- «r.f^ r^ioin«iDinioinr^r^r-ieoer>r^o■nn(^lr>lNOOOOOP) 00 ooi^^noo>v--p>oin<-ot»fOO><-ooooT- 00 oocooi'-mr-oi'-woO'-orji •- ^ •- T- CM CM o>oioor>'(»ooooco r> CM T- OOOOOOOQ oooooooo OOOOOOOO b r«' rj o' b CM CD b ior*ioo>CMrtO>*- 'Page 202 ■-- (M rs ^ »n ^»■>a>o«-CMcn^lnto^«•CDC^o^cM ^^^*-*-*-*-w-^*- fVI W OJ N CM N Ci ^i M CM r3«o « Data Set for Module J OOOOOOOOOO OOOOOOOOOO ooior^moooocM o o h* *- ^ o o in in ^ oo'coo>«-ioooa>oo> o oooooo o to mt^'-c«j lo *» o«>O'-h-*0 *- OOOOOOOOOO OOOOOOOOOO OOOOOOOOOO o ^ ^ en r) a> ■« N CM « « ^ N ^ ^ *" < s o o o o o o o o o o n o o o o o o o o o »{ o o o o o o o o r^ .- T- in OOOOOOOOOO oonNcg^Noiooo) »- tM r) » lo .o60>o»-CM«'«io«rveoo»o»-iMr>«io«r^«»o»-CM ^^^^..-^«>^T-^NCMCMeM01WT SOURCE STREAM REACH ASSESSMENT HOWTANA WATER QUALITY BUREAU FIELD REPORT FORM CENERAL INFORHATION STREAM NAME RIVER BASIN' COUNTY(S) ASSESSED BY (NAME) DEPT. /AGENCY/OTHER DATE ASSESSED_ REACH LEGAL DESCRIPTION T./R./S. 1/4SEC. BEGINNING END, REACH NUMBER OR LETTER (ASSIGNED BY SURVEYOR) PLEASE DESCRIBE ANY PROMINENT LANDMARKS THAT MAY DELINEATE REACH BREAKS. GRADIENT VALLEY SHAPE >3X 'V SHAPED 1-3X -U- SHAPED ERATURE AT AM/PM FLOg STAGE (RELATIVE TO BANK FULL 1CX3X) FLOOD, 100+X HIGH,75X MODERATE. SOX LOW.<25X NO WATER STREAM DESCRIPTION: CHECK THE ITEM IN EACH COLUMN THAT BEST FITS CHANNEL PATTERN CHANNEL EROSION STREAM TYPE SINGLE, STRAIGHT VERTICAL EROSION DOMINATES MOUNTAIN SINGLE, SINUOUS VERTICAL ( HORIZONTAL COMMON TRANSITIONAL BRAIDED, STRAIGHT HORIZONTAL EROSION DOMINATES PRAIRIE BRAIDED, SINUOUS FISHERIES: CHECK THE ITEM THAT BEST FITS COLD WATER (TROUT SPECIES ONLY) UARM/COOL WATER DOMINATED BY SPECIES SUCH AS BASS, CATFISH, SUNFISH, PIKE, WALLEYE. COMBINATION OF COLD/COOL/WARM WATER FISHERIES ALL WELL REPRESENTED STREAM ORDER: FIRST SECOND THIRD FORTH , YOU SHOULD USE A TOPO. MAP TO DETERMINE THIS. TUO OR MORE FIRST ORDER STREAMS BECOME A SECOND ORDER. TWO OR MORE SECOND ORDER STREAMS BECOME A THIRD ORDER. TUO OR MORE THIRD ORDER STREAMS BECOME A FOURTH ORDER. ETC... FLOU REGIME: EPHEMERAL, FLOWS ONLY DURING RAINSTORMS OR PEAK SPRING RUNOFF. INTERMITTENT, SEASONAL FLOWS DURING WET WEATHER, SPRING RUNOFF OR FLOWS YEAR ROUND EXCEPT FOR POINTS OF SUBSURFACE FLOW. ^PERENNIAL, FLOWS YEAR ROUND, LIHLE OR NO SUBSURFACE FLOU. MOITIONAL INFORMATION - COMPLETED BY FIELD PERSONNEL IF KNOWN, OTHERWISE COMPLETED BY UQ8. MT WATER BODY ID »: COUNTY FLAG: STREAM CLASSIFICATION: A, B OR C EPA REACH #_ BENEFICIAL USES: DRINKING WATER , STOCK WATER , FISHERIES , OTHER IMPAIRMENT SEVERITY: MINOR OR NO INTERFERENCE ^THREATENED, NO INTERFERENCE BUT VOMIUD T9BO MOCEWATC. SOME IKIERFERBCE BUT USE NOT PRECLUDED SEVERE, USE PRECLUDED WERE WATER SAMPLES OBTAINED? YES, NO. IF YES, WHAT ANALYSES WILL BE PERFORMED? WHAT PRESERVATIVES WERE USED? GIVE RESULTS, IF ANY RIVER THAT THE STREAM BEING ASSESSED IS A TRIBUTARY TO. Flathead Basin Cooperative Program Final Report Page 207 Nonpoint Source Stream Reach Assessment Form PLEASE CHECK THE HOST APPROPRIATE ANSWER FOR EACH PARAMETER. HAKE SURE TO ELABORATE WHEN ASK OR WHENEVER NECESSARY TO CLARIFY YOUR ANSWER. ! 6ENERAL ASSESSHENT PARAMETERS 1. NATURAL EROSION: (NATEROS) NATURAL EROSION OCCURRING OUTSIDE OF HIGH WATER MARK BUT WITHIN 200' OF THE STREAM. DO NOT BASE THIS RATING ON HUMAN OR STREAM CHANNEL/BANK EROSION. CONSIDER EROSIVE FEATURES SUCH AS LAND SLUMPS/SLIDES, GULLIES/RILLS/SHEETS ETC. EROSION EVIDENT: NONE OR MINOR SOME MODERATE SEVERE [PLEASE ELABORATE] LAND USE: (LAND) HUMAN ACTIVITIES WITHIN THE WATERSHED THAT HAVE THE POTENTIAL TO CREATE NPS POLLUTION. CHECK ONE OR MORE ITEMS TO DESCRIBE DISTANCE FROM STREAM CHANNEL THAT ACTIVITIES HAVE OCCURRED: NO ACTIVITIES HAVE OCCURRED ^GREATER THAN 200' 50-200' LESS THAN 50' LIST PREDOMINANT ACTIVITIES AND HOW RECENT THE ACTIVITIES ARE (<5, 5-10, >10 YEARS) FOR EACH DISTANCE CATEGORY. > 200' 50 - 200' < 50- NONPOINT SOURCE POLLUTION: (NPSP) CONSIDER ONLY THE POLLUTION THAT APPEARS ATTRIBUTABLE TO HUMAN ACTIVITIES. [PLEASE ELABORATE] | POLLUTION EVIDENT: NONE SOME, THOUGH NOT COMMON MODERATE SEVERE (EXAMPLES MIGHT Be SEDIMENT DEPOSITION IN ROAD DITCHES OR BELOW EPHEMERAL, MANURE, IRON PRECIPITATES, TRASH ETC.) BANK STABILITY: (BANKS) CONSIDER ALL OBSERVED STREAM BANK EROSION. [PLEASE ELABORATE] NONE OR MINOR EVIDENCE OF BANK EROSION OR FAILURE (<10X OF TOTAL). MODERATE INSTABILITY, MODERATE AREAS OF BANK EROSION OR BANK FAILURE (10-20X OF TOTAL). SUBSTANTIAL INSTABILITY, FREQUENT AREAS OF BANK EROSION OR FAILURE (21-405! OF TOTAL). UNSTABLE, SEVERE BANK EROSION AND FAILURE (>40X OF TOTAL). EROSION OCCURS ON STRAIGHT SECTIONS AND BENDS. • AVERAGE SIZE OF ERODED BANKS; HEIGHT <1 ' 1-3' >3' LENGTH <5' 5-20' >20' • ARE BANKS VERTICALLY EROSIONAL UNDERCUT SLUMPS OTHER (DESCRIBE) • DOES THE EROSION APPEAR TO PRIMARILY BE THE RESULT OF HUMAN ACTIVITY ? YES NO SUBSTRATE COMPOSITION: (SUBCOMP) CONSIDER ONLY THE TOP (VISIBLE) UYER OF SUBSTRATE MATERIAL. RIFFLE « RELATIVELY SHALLOW, FAST MOVING, WITH BROKEN OR TURBULENT SURFACE, BUN " MODERATE FLOW, WATER SURFACE NOT TURBULENT TO THE POINT OF BEING BROKEN; POOL - SLOW FLOW, SMOOTH SURFACE, GENERALLY DEEPER THAN A RUN OR RIFFLE. • LIST THE DOMINANT SUBSTRATE MATERIAL FOR EACH FEATURE: RIFFLE RUM POOL FINES <1/4''DIA., GRAVEL ^/l,•-Z'DlA. . COBBLE/RUBBLE 3"-12"DIA., BOULDERS/BEDROCK >12"DIA • ESTIMATE PERCENTAGE OF FINE MATERIAL (<1/4") fOR EACH FEATURE: ■ • INDICATE THE AREA OF THE CHANNEL WHERE THE MAJORITY OF FINE MATERIAL IS DEPOSITED: EDGE, MIDDLE, DOWNSTREAM OF ROCKS/LOGS, THROUGHOUT • IS SUBSTRATE MATERIAL CEMENTED (SURROUNDED BY TIGHTLY PACKED FINES)? YES, NO. MISC. NOTES: Page 208 Flathead Basin Cooperative Program Final Report Nonpoint Source Stream Reach Assessment Form STREAM NAME AND COUNTY_ GENERAL ASSESSMENT PARAMETERS CONTINUED... 6. CHANNEL STABILITY: (CHANSTAB) CONSIDER SHIFTS IN SUBSTRATE MATERIAL (I.E. SCOURING, DEPOSITION, POOL FILLING). SCOURING AND DEPOSITION NONE OR MINOR, BARS STABLE AND WELL VEGETATED (<5X AFFECTED BY EITHER PROCESS). SOME SCOUR AT CONSTRICTIONS OR STEEP GRADES, AND/OR POINT BAR ENLARGEMENT BY COURSE GRAVEL. (5-JOX AFFECTED). MODERATE SCOUR AND/OR DEPOSITION, POINT BARS ENLARGING BY GRAVEL AND SAND AND SOME NEU BARS FORMING. (J1-50X AFFECTED). SEVERE SCOUR AND/OR DEPOSITION OCCURS IN MOST AREAS OF THE STREAM. NEU BAR DEVELOPEMENT BY FINE MATERIAL IS COMMON. (>30X OF THE CHANNEL IS IN THE STATE OF FLUCTUATION) [NOTE: "NEW DEPOSITINAL MATERIAL SHOULD BE CLEAN (LIGHTER COLORED), LOOSE AND UNVEGETATED] POOL DEVELOPMENT POOLS COMMON AND/OR WELL DEVELOPED. MINOR POOL FILLING, MOSTLY BY SILT. POOL SIZE DECREASING DUE TO FILLING. POOLS SEVERELY REDUCED OR LACKING DUE TO FILLING. CHANNEL CAPACITY , CHANNEL CONTAINS PEAK FLOUS PLUS MORE IF NECESSARY (U-D >7) , CHANNEL CONTAINS HOST PEEK FLOUS {U-0-8 TO 15) . CHANNEL BARELY CONTAINS PEAK FLOUS (U-D-16 TO 25) , CHANNEL CAPACITY GENERALLY INADEQUATE FOR PEAK FLOW. OVERBANK FLOUS COMMON (U-D»25) STREAMSIDE VEGETATION: (STRVEG) CONSIDER VEGETATION ON BOTH BANKS, UITHIN 100" OF ACTIVE (SCOURED) CHANNEL. BRIEFLY DESCRIBE THE STREAMSIDE VEG. COMMUNITY (I.E. DOMINANT SPP., AGE, WIDTH, ETC.) VEGETATIVE BANK STABILITY: ABILITY OF SANK VEG. TO RESIST EROS. GOOD FAIR POOR PERCENT BARE GROUND: AMOUNT OF BARE GROUND (EXPOSED MINERAL SOIL) U/IN VEG. ZONE <15X 15-30X 31-45X ^>45X AVERAGE WIDTH OF VEGETATIVE ZONE; >100' 76-100' 51-75' 25-50' <25' ESTIMATE MINIMUM UIOTH ' AND MAXIMUM WIDTH ' USE AND/OH DAMAGE: CONSIDER NATURAL (I.E. UINDTHROU, DEAD TREES ETC.) OR HUMAN FACTORS (I.E. GRAZING, LOGGING ETC). PLEASE ELABORATE* BY DESCRIBING TYPE OF USE OR DAMAGE OBSERVED. NONE OR MINOR MODERATE SUBSTANTIAL SEVERE STREAK SHADING: CONSIDER HOU MUCH OF THE STREAM IS SHADED BY BANK VEGETATION. >75X 51-75X 25-50X <25X SEDIMENT FILTERING CAPACITY CONSIDER FACTORS SUCH AS SLOPE AND SEDIMENT SOURCES ^GOOO FAIR POOR CHANNEL HOOIFIERS: (CHANMDF) CONSIDER NATURAL AND ARTIFICIAL MATERIALS AND/OR STRUCTURES (IE. ANY MATERIAL OR STRUCTURE THAT HAS FALLEN IN OR BEEN PLACED IN THE STREAM CHANNEL). EXAMPLES OF MODIFIERS ARE BRIDGES, CULVERTS, DEBRIS JAMS, BEAVER DAMS, CAR BODIES, RIP-RAP ETC... [PLEASE ELABORATE] STRUCTURES/MATERIALS STABLE AND IMPROVING STREAM CONDITION (EX. SED. TRAP, FISH HABITAT, BANK STABILITY.) OR OTHERUISE NOT CAUSING FISH PASSAGE OR STABILITY PROBLEMS. STRUCTURES/MATERIALS GENERALLY STABLE EXCEPT DURING FLOODS; REUTEO STABILITY PROBLEMS MINOR. STRUCTURES/MATERIALS HAVE GOOD POTENTIAL TO BE MOVED UITH HIGH TO MODERATE FLOUS. THESE FEATURES MAY RELEASE SEDIMENT INTO THE SYSTEM AND/OR CAUSE BANK OR CHANNEL EROSION. STRUCTURES/MATERIALS UNSTABLE AND/OR CAUSING SEVERE BANK OR CHANNEL EROSION. DESCRIBE THE MATERIALS OR STRUCTURES THAT MAY BE MODIFYING THE CHANNEL. DESCRIBE ANY POSITIVE CONTRIBUTIONS THAT MODIFIERS ARE MAKING TO THE STREAM (EX. fISH COVER, STABLE SED. TRAPS, BANK STABILIZATION ETC.) DESCRIBE ANY NEGATIVE CONTRIBUTIONS THAT MODIFIERS ARE MAKING TO THE STREAM ( EX . BANK/CHANNEL EROSION, FISH BARRIERS, UNSTABLE SEDIMENT TRAPS, ETC) Flathead Basin Cooperative Program Final Report Page 209 Nonpoint Source Stream Reach Assessment Form IMPAIRHENT INDICATORS (IF IMPAIRMENTS ARE EVIDENT AND CONCENTRATED IN A SPECIFIC AREA, PLEASE NOTE LOCATIONS ON HAP) TURBIDITY CLEAR SLIGHTLY OFF COLOR CLOUDY OPAQUE COLOR: HATER SURFACE OILS NONE_ SLICK , SLIGHT , SHEEN , MODERATE FLECKS , SEVERE OTHER SOURCE, IF KNOWN PRECIPITATES OTHER THAN SEDIMENT ON CHANNEL BOTTOH NONE SLIGHT , MODERATE , PROFUSE COLOR: SALINIZATION NONE EVIDENT: SOME: EVIDENCE OF SALINITY IS PRESENT IN THE WATERSHED HOWEVER NO SALT CRUSTS WERE OBSERVED WITHIN 100' OF THE STREAM. MODERATE: SALT CRUSTS PRESENT <100' FROM THE STREAM. MAY BE MINOR EVIDENCE OF SALTS ON THE STREAM BANK. PLANT DIVERSITY AND ABUNDANCE MAY BE REDUCED OR DOMINATED BY SALT TOLERANT SPECIES (I.E. SALTGRASS, RABBIT BRUSH, FOXTAIL BARLEY, W. WHEATGRASS) . SEVERE: SALT CRUSTS COMMON WITHIN 1(X3' OF THE STREAM OR ON STREAM BANKS. VEGETATION HAY BE SEVERELY REDUCED DUE TO SALT. SALT DEPOSITS MAY BE EVIDENT ON STREAM ROCKS. FOREIGN DEBRIS: CONSIDER MATERIALS SUCH AS GARBAGE, MANURE, LOGGING SLASH, ETC. AND POSITION REUTIVE TO STREAM CHANNEL. LOCATION OF DEBRIS RELATIVE TO DISTANCE FROM CHANNEL AMOUNT POTENTIAL TO REACH CHANNEL (N = NONE, (P = PRESENT, C= COMMON, E = EXCESSIVE) (L » LOU, M = MOD., H » HIGH) >200' FROM STREAM CHANNEL. 50-200' OF STREAM CHANNEL. - < 50' FROM THE STREAM CHANNEL. MATERIAL PRESENT ON BANKS AND IN THE STREAM CHANNEL. • DESCRIBE THE MATERIAL OBSERVED. HATER ODOR NONE , SLIGHT , MODERATE , STRONG SEWAGE , PETROLEUM , CHEMICAL , NATURAL , OTHER DEUATERING: AS A RESULT OF IRRIGATION OR NATURAL FACTORS SUCH AS SUBSURFACE FLOWS. NO APPARENT WATER LOSS (IRRIGATION MAY BE SUPPLEMENTING BASE FLOW) WATER LOSS NOTICEABLE, HOWEVER FLOWS ARE ADEQUATE TO SUPPORT AQUATIC ORGANISMS. FLOW WILL SUPPORT AQUATIC ORGANISMS, THOUGH HABITAT, ESPECIALLY RIFFLES IS DRASTICALLY REDUCED. CHANNEL MAY BE DRY OR FLOW LOW ENOUGH TO PRECLUDE OR SEVERELY IMPAIR AQUATIC ORGANISMS. • SUSPECTED CAUSES: NATURAL HUMAN-CAUSED • WERE IRRIGATION RETURN FLOWS OBSERVED? YES NO IF YES, RATE TURBIDITY H M_ L_ • STREAM TEMPERATURE? ABOVE RETURN BELOW RETURN AQUATIC ORGANISMS FISHERIES • WERE FISH OBSERVED? YES NO IF YES, PRESENT, COMMON, ABUNDANT. SPECIES) IF KNOWN^ • DID FISH EXHIBIT ABNORMAL BEHAVIOR? YES NO • WERE ANY FISH BARRIERS (IMPENETRABLE BLOCKAGES >3' HEIGHT) OBSERVED? YES NO BE SURE AND LOCATE BARRIERS ON MAP. AQUATIC PLANT GROWTH ABSENT PRESENT/NOT COMMON COMMON ABUNDANT DOES PLANT GROWTH APPEAR NORMAL GIVEN THE SEASON AND NATURAL STREAM CONDITIONS? YES NO IS ALGAE FILAMENTOUS? YES NO WHAT COLOR IS THE ALGAE? PLEASE ANSWER THE FOLLOUING OUESTIONS - WAS WATER SURFACE FOAM >3" HIGH OBSERVED? YES NO - HAS CHANNEL BEEN ARTIFICIALLY MODIFIED? CONSIDER ONLY PRACTICES SUCH AS DREDGING. STRAIGHTENING. AND CHANNELIZATION. YES NO IF YES, INDICATE THE TYPE OF MODIFICATION AND APPROXIMATELY HOW RECENT? <5YRS. 5-1QYRS. >10YRS. - HAVE THE MODIFIED PORTIONS OF THE CHANNEL STABILIZED? YES NO . - WERE POINT SOURCES OBSERVED? YES NO IF YES, PLEASE DESCRIBE AND RECORD LOCATION ON MAP Page 210 Flathead Basin Cooperative Program Final Report Nonpoint Source Stream Reach Assessment Form STREAM NAME AND COUNTY, BEST MAHAGEMEKT PRACTICES (BMP'S) AfC LAND USE CHARACTERISTICS (200' EITHER SIDE OF STUEAM) PLEASE RATE THE FOLLOUING TO THE BEST OF YOUR KNOWLEDGE. GENERALLY CONSIDER BMP'S TO BE ANY MEASURES INCORPORATED WITH LAND USE ACTIVITIES THAT HAY HELP PROTECT UATER RESOURCES FROM NPS POLLUTION. EST Z OF TDTJ TO DISTAM &-50' LAND USES U. AREA RELATIVE :E FRCN STREAM S0-2C0- HAVE Bi APPLH O-50 VS BEEH ED Y/H* 50-200' RATE EFFECTIVENESS OF BWS« 0-50' 50-200- NON-IRRIGATED CROP IRRIGATED CROP GRAZING FEEDLOTS TIMBER HARVEST ROADS (Z IS OPTIONAL) SURFACE MINING SUBSURFACE MINING URBAN OTHER OTHER .. ll •APPLICATION OF BMP'S Y (YES) THE MAJORITY OF EACH PARTICULAR LAND USE IN THE REACH HAS BMP'S APPLIED. N (NO) THE MAJORITY OF EACH PARTICULAR LAND USE IN THE REACH DOES NOT HAVE BMP'S APPLIED. • BMP EFFECTIVENESS 1 MORE THAN ADEQUATE PROTECTION 2 ADEQUATE PROTECTION 3 MINOR AND/OR SHORT TERM IMPACTS 4 MAJOR AND/OR LONG TERM IMPACTS PLEASE ELABORATE: UHICH BMP'S UERE OBSERVED? UHAT BMP'S NEED TO BE APPLIED TO IMPROVE THE SITUATION? WHAT IMPACTS UERE OBSERVED? UHAT IS THE POTENTIAL FOR CORRECTION? G'GOQO. F-FAIR, P"P0OR Flathead Basin Cooperative Program Final Report Page 211 Page 2 1 2 Flathead Basin Cooperative Program Final Report Flathead Basin Forest Practices Water Quality and Fisheries Cooperative Program Appendix G: List of Flathead Basin Streams with Major Bull Trout Spawning and/or Rearing Areas or Concentrated Use by Spawning Westslope Cutthroat Trout Flathead Basin Cooperative Program Final Report ^^e 2 1 3 Page 214 Flathead Basin Cooperative Program Final Report List of Flathead Basin Streams with Major Bull Trout Spawning and/or Rearing Areas or Concentrated Use by Spawning Westslope Cutthroat Trout The following non-wilderness streams are considered critical spawning and/or rearing habitat for bull trout in the Flathead Basin: Swan River Drainage: Elk Creek' Goat Creek' Squeezer Creek* Lion Creek' Piper Creek Jim Creek Cold Creek North Fork Lost Creek South Fork Lost Creek North Fork Flathead River Drainage: Big Creek' Hallowatt Creek Coal Creek' South Fork Coal Creek' Mathias Creek Red Meadow Creek Whale Creek' Shorty Creek Trail Creek' considered critical spawning habitat for westslope cutthroat trout: North Fork Flathead River Drainage: Langford Creek Cyclone Creek Red Meadow Creek Moose Creek Middle Fork Flathead River Drainage: Challenge Creek' Dodge Creek South Fork Flathead River Drainage: Hungry Horse Creek' Tiger Creek' Margaret Creek' Emery Creek' The above lists will be subject to change as more information becomes available. Annual updates will be available upon request from the Montana Department of Fish, Wildlife and Parks. Middle Fork Flathead River Drainage: Bear Creek Granite Creek' Morrison Creek Puzzle Creek The following non-wildemess streams sup- port areas of concentrated use by spawning westslope cutthroat trout. These streams are •These streams are annually monitored by the Montana Department of Fish, Wildlife and Parks as part of the Flathead Basin Commis- sion's Master Monitoring Plan. We recom- mend continued long term annual sampling in these areas. Flathead Basin Cooperative Program Final Report Page 215 500 copies of this public document were published at an estimated cost of $5.50 per copy, for a total cost of $2750.00, which includes $2750.00 for printing and $0.00 for distribution. ® PRINTED ON RECYCLED PAPER