FWS/OBS-80/0J Biological Services Program FWS/OBS-80/08 June 1980 Gravel Removal Studies in Arctic And Subarctic Floodplains in Alaska NORTHERN INTERIOR Interagency Energy-Environment Research and Development Program OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY and Fish and Wildlife Service SOUTHERN INTERIOR U.S. Department of the Interior The Biological Services Program was established within the U.S. Fish and Wildlife Service to supply scientific information and methodologies on key environmental issues that impact fish and wildlife resources and their supporting ecosystems. Projects have been initiated in the following areas: coal extraction and conversion; power plants; mineral development; water resource analysis, including stream alterations and western water allocation; coastal ecosystems and Outer Continental Shelf development; National Wetland Inventory; habitat classification and evaluation; inventory and data manage- ment systems; and information management. The Biological Services Program consists of the Office of Biological Services in Washington, D.C., which is responsible for overall planning and management; National Teams, which provide the Program's central scien- tific and technical expertise and arrange for development of information and technology by contracting with States, universities, consulting firms, and others; Regional Teams, which provide local expertise and are an important link between the National Teams and the problems at the operating level; and staff at certain Fish and Wildlife Service research facilities, who conduct in-house research studies. W H 0 I DOCUMENT COLLECTION cO i ru i ru : cO /WHO' ill m □ a a : m = □ FWS/OBS-80/08 June 1980 GRAVEL REMOVAL STUDIES IN ARCTIC AND SUBARCTIC FLOODPLAINS IN ALASKA Technical Report by Woodward-Clyde Consultants 4791 Business Park Blvd., Suite 1, Anchorage, Alaska 99503 Contract Number FWS- 14- 16-0008-970 Nerval Netsch, FWS Project Officer Water Resources Analysis Project Biological Services Program U.S. Fisfi and Wildlife Service 1011 E. Tudor Drive Anchorage, AK 99503 Tfiis study was funded in part by tfie Interagency Energy-Environment Researcfi and Development Program Office of Researcfi and Development U.S. Environmental Protection Agency Performed for tfie Water Resources Analysis Project Office of Biological Services U.S. Department of tfie Interior Wasfiington, DC 20240 DISCLAIMER The opinions, findings, conclusions, or recoinmendations expressed in this report are those of the authors and do not reflect the views of the Office of Biological Services, Fish and Wild- life Service or the Office of Research and Development, U.S. Environmental Protection Agency. EXECUTIVE SUMMARY A 5-year gravel removal study was initiated in mid-1975 to evaluate the effects of gravel removal from arctic and subarctic floodplains in Alaska. The primary purpose of the project was to provide information that will assist resource managers in minimizing detrimental environmental effects resulting from floodplain gravel mining. To achieve this objective 25 ma- terial sites were studied by a team of scientists and engineers. Two major products of the project are a Technical Report which synthesizes and eval- uates the data collected at the sites, and a Guidelines Manual that aids the user in developing plans and operating material sites to minimize envi- ronmental effects. Data from the 25 study sites were collected and analyzed by the follow- ing six d i sc i p I i nes : • River Hydrology and Hydraulics • Aquat i c Biology • Terrestrial Ecology • Water Qual i ty • Aesthet i cs • Geotechnical Engineering Data Analysis compared the Physical Site Characteristics (drainage basin size, channel width, channel configuration, channel slope, and stream ori- gin) and the Gravel Removal Area Characteristics (type of gravel removal method, location of gravel removal, and age of the gravel removal site) with the measured effects of mining activities. The general conclusion reached was that proper site selection and project design facilitate gravel mining with minimal effects on the habi- tats and fauna of floodplains. The key to the successful mitigation of potential detrimental effects is to carefully match the material site design and operation (site location, configuration, profile, schedule, and rehabil- itation) with the Physical Site Characteristics of the selected floodplain. VARIABLES INFLUENCING MINING EFFECTS Physical Site Characteristics Among the Physical Site Characteristics, channel configuration was the most important. Potential floodplain change is least for a braided river and greatest for a straight river. Size of channel is an important factor, with the least change to be expected in a large system and the greatest in a small system (assuming equally-sized material sites). Combining these two variables (channel configuration and size), gravel removal operations can be expected to have the least effect on large braided rivers and the greatest effect on small straight rivers. Other influencing Physical Site Characteristics, which are related to configuration and size, are the availability and size of unvegetated gravel bars, floodplain width, and the distance that can be maintained between the mining site and the active channel. For example, in a small straight river system the floodplain is narrow and gravel bars are neither plentiful nor large. Thus, to extract gravel, either a significant length of active flood- plain or the adjacent inactive floodplain and terrace must be disturbed. In the latter case the narrowness of the floodplain forces the operation to closely encroach upon the active channel. In large river systems these problems can be less significant because gravel bars are larger and, if the inactive floodplain or terrace are used, the wider floodplain allows mainten- ance of a broader undisturbed buffer zone between the material site and act i ve f I oodp lain. IV Gravel Removal Area Characteristics All of the Gravel Removal Area Characteristics were found to signifi- cantly influence the effects of gravel mining. The location of the material site relative to the active channel is considered to be the most important factor. Whether a material site is scraped or pit-excavated is important, but often pits are located away from an active channel, avoiding the types of changes that can be associated with scraping in active floodplains. The major effects of pit sites located in inactive floodplains and terraces are the loss of vegetated habitat, the possibility for the occur- rence of fish entrapment, a change in the appearance of the floodplain, and long-term delay in the re-establishment of predi sturbance conditions. Where pit sites are situated close to active channels, particularly on the inside bends in meandering systems, the possibility exists for diversion of the channel through the pit, eventually forming a channel cutoff in the meander. This highlights the importance of providing a buffer between the material site and the active channel. Where pit sites are of suitable size, of suffi- cient depth, and have contoured perimeters, they can increase local habitat diversity and provide conditions suitable for fish and various species of terrestrial fauna. Scraped material sites in active floodplains have minimal effects on the floodplain environment when only exposed gravel bars are excavated above the water level, and when slope and contours are maintained (resem- bling those of natural bars). Removal of vegetated areas or banks, which resu I ts in decreased I atera I stab i I i ty of act i ve channel s or a I lows water to spread over a large area, is not desirable. Decreased water depth and veloc- ity increases sedimentation rates, alters water temperature, and alters dissolved oxygen levels. These changes in aquatic habitat usually affect the local distribution and community structure of benthos and fish. The effects of scraping in vegetated areas of inactive floodplains and terraces can be similar to those described for pits. However, long-term changes typically are minimal because the lack of standing water in the closed site will facilitate re-estab I i stiment of pre-mining vegetation con- di t ions. If material sites are located and operated to prevent or greatly mini- mize effects on channel hydraulics, and to utilize only exposed gravel bars, the probability of major localized changes to a floodplain generally is greatly reduced. Where exposed gravel bars are not available or are inadequate, a tradeoff decision between sites must be made that weighs the potential effects of aquatic disturbances against terrestrial disturbances. In these cases, minimization of hydraulic change to active channels should be important in the decision — major hydraulic changes can have a greater long-term effect on terrestrial systems than the controlled disturbances associated with a site located in a vegetated inactive floodplain or ter- race. RECOMMENDED FUTURE STUDIES During the present study a number of subject areas were identified that should be investigated. 1. Evaluation of gravel mining from coastal and upland sources; and, preparation of guidelines for users of these sources. These alternatives to sources have not been studied. 2. Evaluation of the effects of multiple sites on one river system. Such an investigation should be aimed at determining the critical, spatial, and temporal relationships of multiple sites. Gravel replenishment rate predictions should be an integral part of this investigation. 3. Several floodplain gravel removal sites should be investigated before, during, and after mining to assess the adequacy of the Guidelines Manua I . 4. Several topics of the Guidelines Manual should be studied in detail to assess their adequacy, (i.e., buffers, pit design, and active channel dredging) . VI This report was submitted in fulfillment of Contract Number 14-16- 0008-970 by Woodward-Clyde Consultants, Anchorage, Alaska, under sponsor- ship of the Office of Biological Services, U.S. Fish and Wildlife Service. Work was completed as of June 1980. VI I TABLE OF CONTENTS Page EXECUTIVE SUMMARY LIST OF FIGURES LIST OF TABLES ACKNOWLEDGMENTS INTRODUCTION, E. H, Follmann BACKGROUND PHILOSOPHY PROJECT DESCRIPTION REFERENCES APPROACH AND METHODOLOGY, E. H. Follmann .... SITE SELECTION DATA REVIEW FIELD STUDY OF SELECTED MATERIAL SITES DATA BASE TECHNICAL REPORT GUIDELINES MANUAL REFERENCES DESCRIPTION OF STUDY RIVERS, L. L. Moulton, Ed. SEWARD PENINSULA NORTH SLOPE NORTHERN INTERIOR SOUTHERN INTERIOR REFERENCES EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAULICS, L. A. Rundquist INTRODUCTION METHODS OF DATA COLLECTION METHODS OF DATA ANALYSIS HYDROLOGY HYDRAULICS QUANTIFICATION OF CHANGES RESULTS AND DISCUSSION CHANNEL CONFIGURATION AND PROCESS HYDRAULICS SEDIMENTATION ICE CHARACTERISTICS HYDROLOGY SUMMARY AND CONCLUSIONS CHANNEL CONFIGURATION AND PROCESS HYDRAULICS SEDIMENTATION ICE CHARACTERISTICS HYDROLOGY RECOMMENDATIONS REFERENCES vi i [ i 1 i X i i X 1 X :xi i 1 1 5 7 9 1 1 1 1 17 17 29 30 31 53 35 35 42 51 58 66 67 67 70 71 71 72 74 81 81 99 1 15 122 127 134 134 135 135 136 136 136 138 Page EFFECTS OF GRAVEL REMOVAL ON AQUATIC BIOTA, L. L. Moulton 141 INTRODUCTION 141 METHODS OF DATA COLLECTION 144 METHODS OF DATA ANALYSIS 145 RESULTS AND DISCUSSION 148 MAJOR GRAVEL REMOVAL HABITAT ALTERATION 148 EFFECTS OF HABITAT ALTERATION ON FISH POPULATIONS .... 167 EFFECTS OF HABITAT ALTERATION ON AQUATIC MACRO INVERTEBRATES 198 SUMMARY AND CONCLUSIONS 209 EFFECTS OF GRAVEL SCRAPING ON RIVERINE HABITATS 209 EFFECTS OF INUNDATED PIT FORMATION ON THE ASSOCIATED RIVER BIOTA 211 RECOMMENDATIONS 213 REFERENCES 214 EFFECTS OF GRAVEL REMOVAL ON TERRESTRIAL BIOTA, M. R. Joyce 215 INTRODUCTION 215 METHODS OF DATA COLLECTION 218 METHODS OF DATA ANALYSIS 220 RESULTS AND DISCUSSION 222 VEGETATIVE COMMUNITIES OF STUDY AREA FLOODPLAINS .... 222 VEGETATIVE COMMUNITY CHANGES AT GRAVEL REMOVAL SITES 225 FACTORS AFFECTING VEGETATIVE RECOVERY RATE 241 FAUNAL COMMUNITY CHANGES AT GRAVEL REMOVAL SITES .... 248 FACTORS AFFECTING RECOVERY RATE OF FAUNAL COMMUNITIES . . 253 PERMANENTLY PONDED SITES 254 SIMILARITIES OF RESPONSE BETWEEN BIOTA AND STUDY SITE PARAMETERS 259 SUMMARY AND CONCLUSIONS 268 VEGETATIVE REMOVAL 268 MINING DEPTH AND LOCATION 268 OVERBURDEN 269 PERMANENTLY PONDED HABITATS 269 RECOMMENDATIONS 270 REFERENCES 271 EFFECTS OF GRAVEL REMOVAL ON WATER QUALITY, L. L. Moulton 273 INTRODUCTION 273 RESULTS AND DISCUSSION 276 POST-MINING EFFECTS OF GRAVEL REMOVAL OPERATIONS .... 276 SUMMARY AND CONCLUSIONS 284 REFERENCES 285 IX Page EFFECTS OF GRAVEL REMOVAL ON AESTHETICS, D. K. Hardinger 287 INTRODUCTION 287 SCENIC QUALITY 289 VISUAL SENSITIVITY 289 DEGREE OF VISIBILITY 290 APPROACH 291 THE VISUAL RESOURCES OF THE STUDY REGIONS 292 SEWARD PENINSULA 292 NORTH SLOPE 295 NORTHERN INTERIOR 297 SOUTHERN INTERIOR 300 EFFECTS OF GRAVEL REMOVAL ON VISUAL RESOURCES . 304 SEWARD PENINSULA 504 NORTH SLOPE 305 NORTHERN INTERIOR 306 SOUTHERN INTERIOR 306 SUMMARY 308 GEOTECHNICAL ENGINEERING CONSIDERATIONS OF GRAVEL REMOVAL, H. P. Thomas and R. G. Tart, Jr 311 INTRODUCTION 311 APPROACH 314 SITE SELECTION AND INVESTIGATION 315 PRELIMINARY SITE SELECTION 315 SITE INVESTIGATION 317 FINAL SITE SELECTION 318 MINING PLAN PREPARATION 319 SITE PREPARATION 321 ACCESS 321 OVERBURDEN REMOVAL 324 CHANNEL DIVERSION 324 SETTLING PONDS 325 SITE OPERATION 326 EXCAVATION 326 TRANSPORTATION AND STOCKPILING 327 PROCESSING 328 SITE REHABILITATION 329 REFERENCES 330 INTERDISCIPLINARY OVERVIEW OF GRAVEL REMOVAL, E. H. Follmann 331 INTRODUCTION 331 PHYSICAL SITE CHARACTERISTICS 333 CHANNEL CONFIGURATION 333 DRAINAGE BASIN SIZE (CHANNEL WIDTH) 348 CHANNEL SLOPE AND STREAM ORIGIN 350 GRAVEL REMOVAL AREA CHARACTERISTICS 354 TYPE OF GRAVEL REMOVAL 354 LOCATION OF GRAVEL REMOVAL 565 DIKES AND STOCKPILES 576 Page SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS 379 SUMMARY 379 RECOMMENDATIONS 382 RECOMMENDED FUTURE STUDIES 384 APPENDICES A. SCIENTIFIC NAMES 385 B. GLOSSARY 395 XI L 1ST OF FIGURES Number Page 1 Location of the 25 gravel removal study sites in Alaska 13 2 Typical Seward Peninsula terrain 37 3 Arctic Coastal Plain wetlands 43 4 Northern portion of the Arctic Foothills ...... 44 5 Typical view of the White Hills section of the Arctic Foothi I Is 44 6 M. F. Koyukuk River valley looking upstream .... 52 7 Typical terrain of the Kokr i ne-Hodzana Highlands . . 53 8 Typical terrain in the Yukon-Tanana Upland Section . 59 9 G I ac i of I uv i a I deposits in Dry Creek floodplain ... 60 10 Typical view of Alaska Range section 61 11 Aerial photograph showing the two gravel removal locations at Sinuk River considered separately in the hydrology/hydraulics analysis . . 75 12 Aerial photograph of Washington Creek showing the upper and lower gravel removal areas 77 13 Aerial photograph of Oregon Creek showing the upper and lower gravel removal areas 78 14 Aerial photograph of Aufeis Creek showing upper and lower gravel removal areas 79 15 Aerial photograph of Middle Fork Koyukuk River- Upstream showing upper and lower gravel removal areas 80 16 Schematic diagram of the plan view and cross section of a typical braided river 83 17 Maximum depths and corresponding top widths of undisturbed major, side, and high-water chan- nels at four braided study sites 84 Number Page 18 Schematic diagram of the plan view and cross section of a typical split channel river 85 19 Maximum depths and corresponding top widths of undisturbed major, side, and high-water channels at four split channel study sites 86 20 Schematic diagram of the plan view and two cross sections of a typical meandering river 87 21 Maximum depths and corresponding top widths of undisturbed major, side, and high-water channels at 15 study sites with meandering, sinuous, and straight configurations 88 22 Schematic diagram of the plan view and cross section of a typical sinuous river 90 23 Schematic diagram of the plan view and cross section of a typical straight river 90 24 Schematic diagram of an al I uv i a I fan 91 25 Comparative aerial photography of the Nome River showing change in channel configuration resulting from gravel removal activities 96 26 Aerial photograph of the Ugnuravik River pit site showing the insufficient buffer zone 98 27 Aerial photograph of the Tanana R i ver-Upstream site with substantial buffer zone separating the pit from the active side channel 100 28 Aerial photograph of the Prospect Creek pit showing wide buffer zone separating the pit from the active channel 101 29 Schematic diagram illustrating definitions of channel geometric and hyraulic variables 102 50 Average hydraulic geometry of river channels expressed by relations of width, depth, and veloc- ity to discharge at two locations along a river (modified from Leopold, Wolman, and Miller 1964) . . . 104 31 Schematic diagram showing change in water surface slope in response to a change in water discharge 105 XIII Number Page 32 Schematic diagram illustrating the effects of a flow obstruction on the local hydraulics .... 107 33 Comparative aerial photography of the Penny River showing change in hydraulic character- istics resulting from gravel removal activities ... 113 34 Schematic diagram illustrating an example of a change in local water surface slope result- ing from an in-channel gravel removal operation ... 114 35 Schematic diagram showing degradation process .... 116 36 Upstream view of thermal and fluvial erosion in the access road at Ugnuravik River, acting as a long-term sediment source to the river 123 37 View of erosion of a diversion dam which acts as a long-term sediment source to Skeetercake Creek. Dunes in foreground are atypical of the undisturbed river 123 38 Large area of aufeis at the upper gravel removal area at Washington Creek as it appeared in early June 125 39 Aerial photographs of Washington Creek (top) and Aufeis Creek (bottom) showing material site loca- tions and approximate channel locations before the disturbance 133 40 Siltation resulting from extensive aufeis field at Oregon Creek mined study area, 20 June 1977 14-9 41 Removal of bank cover at Oregon Creek as observed on 24 June 1977 152 42 Removal of bank cover at Skeetercake Creek as observed on 18 June 1977 153 43 Washington Creek upstream and mined area on 9 September 1977 showing reduction of instream cover due to gravel removal operation (flow level [O. I I mVsec] = 20 percent of mean annual flow). Other habitat alterations include increased braiding, siltation, and intergravel f I ow I 54 X IV Number Page 44 Reduction of instream cover as provided by boulders at Sagavanirktok River, 3 August 1978 (flow level, 60 mVsec, = 155% of estimated mean annual flow) 155 45 Increased braiding at Sagavanirktok River study site caused by mining mid-channel gravel bars and a vegetated island in the active channel (mining operation conducted during the winter of 1974-1975) 157 46 Response of cross-sectional wetted perimeters to percentage of mean annual flow and percent- age of cross sections comprised of selected depth intervals at mean annual flow at three gravel removal study sites 158 47 Low velocity backwaters formed by gravel removal at Dietrich River-Downstream (13 July 1978) and Middle Fork Koyukuk R i ver-Upstream (18 July 1978), note extensive silt deposition in both cases .... 159 48 Creation of low velocity side channels and inundated pit following gravel extraction 160 49 Sequence of aerial photographs showing effects of overmining the inside of a meander bend at Middle Fork Koyukuk Ri ver-Upstream. Immedi- ately following mining (b) there was an increase in backwater areas. The next year (c) the meander was partially cut off, creating a vari- ety of low velocity habitats 163 50 Temperature and dissolved oxygen profiles at four deep gravel pit study sites 165 51 Ponded area at Kuparuk River study site where three seine hauls captured 61 Arctic grayling and 2 slimy sculpin, 9 August 1978 (pool I in Table 21 ) 179 52 Ponded area at Middle Fork Koyukuk-Upstream study site where one seine haul captured 28 Arctic grayling, 3 round whitefish and 3 slimy sculpin, 18 July 1978 (pool 2 in Table 21) 179 53 Potential migration blockages, aufeis fields at Washington Creek and Oregon Creek, June 1977 . . 182 54 Region where Aufeis Creek went subsurface creating migration blockage due to lack of surface f low 183 XV Number Page 55 Prospect Creek study site - shallow pond habitat supporting Arctic grayling, Chinook salmon juven- iles, round whitefish, northern pike, burbot, and slimy sculpin, 12 August 1978 188 56 West Fork Tolovana River study site - deep pond with extensive shallows providing northern pike and Arctic grayling habitat, 29 July 1978 188 57 Tanana Ri ver-Upstream upper pit showing exten- sive vegetation beds, 18 August 1978. Note difference in the extent of vegetative develop- ment in this 13-year old pit as compared to the 2 and 3-year old pits in Figures 55 and 56 190 58 Potential overwintering area at Willow Creek. This spring-fed tributary, open throughout the winter, had previously entered Penny River at a deep pool 192 59 Creation of a potential overwintering area at West Fork Tolovana River downstream from pit 196 60 Densities of selected aquatic macroi nver tebrates at Aufeis Creek study areas during 1977 sampling trips 206 61 Penny River undisturbed floodplain showing typical North Slope and Seward Peninsula floodplain charac- teristics of sinuous channel bordered with dense shrub thickets with incised outside meander bank, and narrow gravel point bar on inside meander .... 224 62 West Fork Tolovana River showing typical South- ern and Northern Interior medium river flood- plain characteristics with shrub thickets and white spruce-paper birch stands along the riparian zone 224 63 West Fork Tolovana River showing permanently flooded pit excavated adjacent to the active floodplain with a downstream connection 226 64 A view of Oregon Creek looking downstream through the mined area showing site conditions that remain 13 years after gravel removal 229 65 Penny River mined area looking upstream. Note the flooded conditions within the disturbed area, and the overburden piles in the center of the site 230 XVI Number Page 66 Close-up view of an overburden pile in the Penny River mined area. Note the development of herbace- ous and woody vegetation during the II years fol lowing gravel removal 234 67 Washington Creek mined area showing vegetative recovery only present on the overburden pile 13 years after gravel removal 234 68 Woody revegetation occurring through develop- ment of adventitious stems 235 69 Distribution of woody slash debris and other organics over the ground on the edge of the gravel removal area at Aufeis Creek 236 70 View of the upper pit at Tanana Ri ver-Upstream showing diversity of shoreline configuration and development of woody and herbaceous vegeta- tion 13 years after gravel removal 237 71 View of the Ivishak River floodplain looking downstream showing typical braided channel characteristics with extensive gravel bars and isolated, vegetated islands 239 72 View of both undisturbed (background) and mined (foreground) reaches of the Shaviovik River. Note that gravel removal maintained natural point bar contours and shapes and did not disturb riparian vegetative zones 240 73 Compacted surface gravels in an access road leading to the Dietrich River-Downstream site .... 244 74 Inorganic overburden piled on the edge of the Oregon Creek site which supported no vegetation 13 years after gravel removal 244 75 Close-up of dense and diverse vegetative devel- opment in an area of surface broadcast of woody slash and organics. Note the willow adventi- tious stem development 246 76 Distant view of a large silt depositional area at the Sagavan irktok River study site 247 77 A silt depositional area of the Kavik River supporting a well-developed pioneer vegetative community 247 XVI Number Page 78 Close-up of a concentration of willow seedlings at the shoreline of the Jim River ponded area .... 248 79 Vegetated organic mats that were washed down- stream and grounded during high water on Toolik River floodplain gravel bars 249 80 Tanana Ri ver-Upstream showing shoreline diver- sity and vegetative development in the upper pit . . . 256 81 Undisturbed buffer along the original stream channel at Aufeis Creek (downstream disturbed area only) 264 82 Gravel fill ramp used to protect the incised bank at the Sagavan irktok River study site 266 83 Thermal and hydraulic erosion of permafrost induced by multiple passes of a tracked vehicle across an unprotected incised floodplain bank and adjacent tundra 267 84 Armored bank protecting the West Fork Tolovana River pit from a channel diversion into the mined site 267 85 Typical Seward Peninsula landform at Penny River . . . 293 86 Typical view of an Arctic Coastal Plain floodplain . . 295 87 Dietrich River valley 298 88 Lower Middle Fork Koyukuk River valley 298 89 McManus Creek val ley 501 90 Phelan Creek valley 302 91 Gravel ramp at Shaviovik River site providing access over a permafrost river bank 322 92 Thermal erosion near Ugnuravik River resulting from compaction and destruction of the vegeta- tive mat overlying ice-rich permafrost soils . . . 323 93 Configurations of study rivers 335 XV I I LIST OF TABLES Number Page 1 Major Variable Matrix 15 2 Methods Used for Measuring Water Quality Parameters with the Number of Replicates Taken per Study Area 21 3 Aquatic Biology Sampling Methods Used at Each Study Site. 24 4 Size and Quantity Values of the 25 Study Sites 36 5 Quantification Ratings of Change in Channel Configuration Characteristics Resulting from the Gravel Removal Operation at Each of the 25 Sites 93 6 Values of Exponents for Hydraulic Geometry Power Relations 106 7 Quantification of Change in Hydraulic Variables Resulting from the Gravel Removal Operation at Each of the 25 Sites 109 8 Quantification Ratings of Change in Sedimentation Characteristics Resulting from the Gravel Removal Operation at Each of the 25 Sites 119 9 Quantification Ratings of Change in Aufeis Potential that Resulted from the Gravel Removal Operation at Each of the 25 Sites 126 10 Mean Annua! Flow Estimates at Each of the 25 Study Sites. 129 11 Calculated Discharges in m /s Corresponding to Selected Recurrence Intervals for Each of the 25 Study Sites . . 130 12 Quantification Ratings of Change in Quantity of Intergravel Flow Resulting from the Gravel Removal Operation at Each of the 25 Sites 131 13 Major Habitat Alterations Observed at Sites Mined by Scraping 150 14 Percent of Pit Area Composed of Selected Depth Intervals. 164 15 Effects of Cumulative Habitat Alterations on Fish Populations in the Mined Area of Study Sites Mined by Scraping 168 X I X Number Page 16 Estimated Densities and Blomass of Arctic Char and Slimy Sculpin at Washington Creek Study Site Based on Repeated E lectroshock i ng of Blocked Sections of Stream 21-23 June 1977 172 17 Estimated Densities and Biomass of Arctic Ctiar and Arctic Grayling at Kavik River Study Site Based on Repeated E I ectroshock i ng of Blocked Sections of Stream, 1976 . . 174 18 Comparison of Fish Densities in Wined and Undisturbed Areas as Determined by E I ectroshock ing Blocked Sections of Stream at Kavik River Study Site, 1976 . . 175 19 Catch of Arctic Grayling per Angler Hour at Kavik River Study Areas During Summer 1976 Sampling Trips . . 176 20 Change in Catch per Effort and Percent Composition of Indicator Species at Selected Study Sites 177 21 Summary of Catch from Ponded Water Areas Isolated from Active Channels at Two Study Sites 181 22 Mean Fork Lengths of Coho Salmon Caught by Minnow Trap at the Penny River Study Site During 1977 186 23 Differences of Coho Salmon Mean Fork Length Between Sample Areas and Associated Significance Levels, Penny River Study Site During 1977 187 24 Physical Conditions at Pits Visited During Winter .... 193 25 Response of Aquatic Riffle Macro i nvertebrate Taxa to Habitat Alterations Observed at Selected Study Sites . 199 26 Changes in Aquatic Macro i nvertebrate Densities at Sites Exhibiting Type I and 2 Substrate Alterations 203 27 Densities of Aquatic Macro i nvertebrates Collected at Inundated Pit Sites, 1976-1978 208 28 Quantitative Changes in Selected Terrestrial Biological Parameters at Gravel Removal Study Sites 223 29 Location, Response Time, and Community Characteristics of Vegetative Recovery at Selected Study Sites 232 30 Quantification of Change in Selected Hydrology Parameters Which Were Impeding Vegetative Recovery at Study Sites 243 31 Qualitative Evaluation of Habitat Quality and Fauna Use at Permanently Ponded Gravel Removal Sites 255 XX Number Page 32 Bird Observations by Habitat Type Within the Control and Disturbed Areas at Tanana R i ver-Upstream 3-7 June, 1978. Numbers Indicate Minimum Individuals Known to Occur in Each Habitat Type 257 33 Bird Observations by Habitat Type Within the Control and Disturbed Stations at West Fork Tolovana River 9-11 June, 1978, Numbers Indicate Total Individuals Known to Occur in Each Habitat Type 258 34 Two Way Coincidence Table Displaying a Hierarchial Clus- tering of Similar Sites and Similar Biotic Parameters . 260 35 Selected Alaska Water Quality Standards 274 36 Water Quality Parameters Measured at Gravel Removal Sites Which Exceeded Alaska Water Quality Standards 275 37 Changes in Turbidity and Suspended Solids Between Sample Areas at Selected Study Sites 278 38 Relative Change of Water Quality Parameters Between Up- stream and Downstream Sample Areas at Selected Study Sites 281 39 Average Measured Values of Selected Water Quality Param- eters at Study Sites with Inundated Pits 283 40 Interdisciplinary Rating of Cumulative Effect of Scraping, Using Various Indices of Change on Study Sites Visited from 1976 to 1978 338 4! Interdisciplinary Rating of Effects of Pits on Associated Floodplains at Selected Study Sites Visited from 1976 to 1978 Using Various Indices of Change 359 A-l Vegetation Identified in the Text 386 A-2 Mammals Identified in the Text 387 A-3 Birds Identified in the Text 388 A-4 Fish Species Reported and Caught or Observed in Major Geographical Areas Represented by the Twenty-Five Sites 390 A-5 Aquatic Macro i nvertebrates Caught at Study Sites During 1976-1978 Field Sampling 392 XXI ACKNOWLEDGWENTS Woodward-Clyde Consultants appreciates the contributions of a number of scientists and engineers. Dr. A. 0. Ott, now with the Alaska State Pipeline Coordinator's Office, conceived the aquatic biology field program and was responsible for much of its implementation. He also served as a principal investigator on the study for one and one half years of the field phase. Brent Drage, now with R & M Consultants, was responsible for early imple- mentation of the hydrology field program. Other scientists whose contributions have benefited the study are, James A. Glaspell of the Alaska Department of Fish and Game; Michael A. Scott of the U. S. Bureau of Land Management; and Dr. Keshavan Nair, Dr. Ulrich Luscher and Robert Pitt of Woodward-Clyde Consultants. A number of Woodward-Clyde personnel assisted in the field at various times, including Donald 0. McKay (now with the U. S. Fish and Wildlife Service), Kenneth E. Tarbox, Jonathan Isaacs, and Jerry P. Borstad. Thanks are also due to Alyeska Pipeline Service Company and the Alaska Department of Transportation for use of their photographs of some mining sites and the use of their mining plans. We are also grateful to Susan Ogle for graphics production and Marn ie Isaacs for editorial review. And finally, we are deeply indebted to Jean Borstad and Jayne Voorhis for report production. The U. S. Fish and Wildlife Service Project Officer appreciates the tech- nical and administrative assistance provided by Summer Dole throughout the project and to Dr. Norman Benson for technical advice during the data collec- tion and analysis phases. We thank the following individuals who reviewed and commented on some portions or all of the report in draft form: Bob Bowker , Hank Hosking, Ronald Kinnuner, Jim Lewis, Lou Pamp I i n, John Stout and Jera I d Stroeble, all with the U. S. Fish and Wildlife Service, various stations; Bill Gabriel and Earl Boone with the Bureau of Land Management, Anchorage, Alaska; Joe Childers and Bob Madison with U. S. Geological Survey, Anchorage, Alaska; Dr. Alvin Ott with the State Pipeline Coordinators Office, Fairbanks, Alaska; Bruce Barrett and Carl Yanagawa with the Alaska Department of Fish and Game, Anchorage, Alaska; Brien Winkley, U. S. Army Corps of Engineers, Vicksburg, Mississippi; and W. P. Metz and A. W. Schwarz with ARCO Oil and Gas Company, Anchorage, Alaska. XXIII INTRODUCTION E. H. Follmann This Technical Report and the accompanying Gravel Removal Guidelines Manual for Arctic and Subarctic Floodplains (Guidelines Manual) present data analyses and conclusions resulting from a 5-year study of 25 floodplain material sites in arctic and subarctic Alaska, and provide guidelines to insure minimal environmental degradation when siting, operating, and closing floodplain material sites. This study, its results and conclusions, and these reports directly relate only to floodplains, although several aspects may also be applicable in nonf I oodp I a i n locations. BACKGROUND A common denominator in all resource and industrial development is the need for granular material; gravel is used worldwide for construction projects and transportation routes. In the arctic and subarctic, however, the presence of permafrost creates special construction problems that place additional demands on the supply of gravel. Even slight alterations in the permafrost thermal regime caused by surface disturbances can cause thawing, thermokarst formation, subsidence, and erosional problems. Maintenance of the thermal regime is essential when building or operating in permafrost areas, but especially in regions characterized by fine grained soils with high water content. These latter areas are highly susceptible to subsidence when surface disturbance alters E. H. Follmann is presently associated with the Institute of Arctic Biology of the University of Alaska. the thermal regime. In these cases, the thawed ground becomes a morass in which vehicle passage can be impossible and maintenance of structural sta- bility of facilities becomes difficult. The current major solution for eliminating or greatly reducing perma- frost thaw is to use gravel as either pads for structures or as roadways. Although these demands exist elsewhere, the thickness of gravel required in permafrost areas is far greater than in nonpermafrost areas. The gravel pad in permafrost areas replaces the insulative function of the vegetative mat that was removed or compressed by the gravel fill. Since the insulative quality of the vegetative mat is greater than that of an equivalent thick- ness of gravel, a gravel pad must be considerably thicker to maintain an equivalent thermal regime. Under these circumstances the most important considerations for determining pad or road thickness are: climatic factors, soil surface temperatures, permafrost temperatures, and subgrade soil proper- ties (McPhail et al. 1975). The objective is to establish the freeze front in or slightly below the fill (McPhail et al. 1975). Where this is accom- plished, potential thaw problems can be greatly diminished. Arctic and subarctic regions have been the focus of attention during the past several decades because of the wealth of natural resources known or thought to occur in these regions. The discovery of oil and gas on Naval Petroleum Reserve No. 4 (now the National Petroleum Reserve-Alaska) in the I940's, at Prudhoe Bay in 1968, and in northern Canada has stimulated this interest and expanded it to include metallic minerals and coal. Expan- sion of exploration activities can be expected to continue. As resource development in remote arctic and subarctic areas becomes more economically feasible the region's resources will be utilized to meet society's energy and material needs. These future projects will require increased quantities of gravel to facilitate construction and to provide stable substrates for various permanent and temporary facilities. For ex- ample, the gravel requirement for the Trans-Alaska Pipeline System was about 49 mill ion cubic meters (m ) (Michael Baker, Inc. 1977) . Sma I ler projects requiring gravel, such as exploratory well drill pads and associated camps. typically use up to 75,000 m . If, tiowever, airstrips and roads are associ- ated with ttiese sites, quantities can increase to several hundred thousand cubic meters. Based on experience constructing the Yukon River to Prudhoe Bay Haul Road (Haul Road), approximately 31,000 m of gravel are required per kilometer of road construction, and maintenance requirements average about 700 m per kilometer (km) per year for about the first 5 years (Alson personal communication). Alyeska Pipeline Service Company requested about 1.5 million m of gravel for maintenance of their project over a 5 year period. The figures presented above for the large pipeline projects repre- sent gravel needs from both upland and floodplain sites. About half of the gravel used on the oil pipeline was from floodplains. Alluvial deposits found in broad floodplains offer one of the prime sources of gravel in northern areas. Individual material sites vary consider- ably in size, as indicated by the range of those considered for study in this project: 7,738 to 631,000 m of material removed. Several different sites may be necessary to supply material meeting the required project specifications because one site may not contain all types of material needed. For example, not al I potential sites wi I I have material suitable for topping. Also, since road and pipeline construction projects need materials throughout their lengths, one site or a series of sites in one area will not satisfy the demands of these projects. A haul distance of 6.5 km or less has been estimated to be economically efficient for construction in Alaska, and haul distances of 13 to 16 km or less are planned for mainten- ance of the Trans-Alaska Pipeline System (Alson personal communication). Therefore, material sites for these types of projects necessarily must be located at regular intervals due to economic considerations. To protect an environment from unacceptable disturbance, the elements comprising the environment must be known, the various elements of the pro- posed activity must be known, and the effects of the activity on the environ- mental elements separately and as a whole must be known. Where this infor- mation is available, guidelines to conduct the proposed activity with a minimum of environmental perturbation can be developed. Where information on one or more of these elements is lacking or is only partly understood, any guidelines that are developed are based on estimates and assumptions whose validity is dependent on the experience and predictive powers of those developing the guidelines. The latter condition is the rule in most cases where environmental impacts are concerned. Impacts from resource exploration and development have not been studied as much as is necessary to make intel- ligent decisions regarding environmental impacts. This lack of research is particularly true in arctic and subarctic regions. The remoteness of the area and the high cost of conducting research have not facilitated an ade- quate description of the environmental elements. Studies of the environ- mental effects of development have been similarly hindered. Extensive literature review revealed that the specific impacts of gravel removal had seldom been studied and, therefore, were poorly under- stood. Description of impact had been attempted in only a few cases (Bull and Scott 1974, Federal Water Pollution Control Administration 1968, Forshage and Carter 1975, Sheridan 1967); and these studies dealt specif- ically with only one aspect, e.g., fisheries. LaBelle (1973) reviewed gravel and sand availability in the Barrow area of the National Petroleum Reserve- Alaska and made recommendations on gravel extraction and evaluations of potential environmental impact. Northern Engineering Services Company Limited and Aquatic Environments Limited (1975) evaluated the material sites associated with the Trans-Alaska Pipeline System with reference to aquatic habitat. In addition, several reports identified problems associated with gravel extraction as one of many sources of environmental perturbations that could be expected from new and continued exploration and development in the north (Bliss and Peterson 1973, Klein 1973, Weeden and Klein 1971, West 1976). None of these latter reports presented results of any material site studi es. There have been few studies on the environmental effects resulting from construction of the Trans-Alaska Pipeline System. The Joint State/ Federal Fish and Wildlife Advisory Team (JFWAT) prepared a report on surveil- lance experience with gravel mining recommendations (Burger and Swenson 1977). The JFWAT also produced a series of reports dealing with experiences on the pipeline, including environmental effects studies. However, the major responsib i I i f-y of the majority of JFWAT staff was environmental surveillance of construction, not research on environmental effects. Weeden and Klein (1971:481) stated: "As with so many other problems of tundra management, the design of criteria for mining operations in gravel lags far behind present need because detailed knowledge of fish populations — where they are, when they migrate, where they spawn, their vulnerability to added silt loadings of river waters, etc. — is lacking". By early 1975, the state of knowledge had not progressed or expanded greatly. This fact, coupled with the dependence on gravel for arctic and subarctic construc- tion, stimulated the U.S. Fish and Wildlife Service to initiate a project to investigate the effects of gravel removal on floodplain systems. The project objective was to provide a comprehensive information review and data synthesis to form the basis for future mining of river and floodplain gravels. The purpose of the project is to provide an information base that will assist resource managers to formulate recommendations concerning oper- ations that will minimize detrimental environmental effects of gravel re- moval from arctic and subarctic streams. PHILOSOPHY Little is known about the natural changes which occur in riverine systems in arctic and subarctic regions. Therefore, determining the effects of resource exploitation in these regions is often difficult because of the interplay of natural changes and man-induced disturbances. The basis for this study was the assumption that gravel removal operations in a floodplain cause change, the magnitude of change depending primarily on the floodplain characteristics, the location of the site, and the method of gravel extrac- tion. Since almost all riverine systems in arctic and subarctic regions have evolved to the present through natural change and without man-induced dis- turbances, all changes due to gravel removal identified in this study were considered undesirable. To maintain a river system in its natural or near- natural state was considered the essence of guidelines development and provided the best conceptual base from which to minimize environmental degradation. However, it is recognized that there may be situations where resource managers may wish to exercise other options. Any site character- istics or methods that facilitated rapid recovery to pred i sturbance con- ditions were considered for implementation as guidelines. The presupposition that all changes due to gravel removal are undesir- able does not, by necessity, cause the data analyses and recommendations to be impractical. It is a foregone conclusion that changes will occur when gravel is removed from a floodplain. To note that changes from the natural state were less at one site than another suggests that the former site was operated more consistently with characteristics of the system than the latter, thereby reducing the magnitude of change. The floodplain and gravel removal characteristics at sites that produced these minor changes formed the primary basis for development of constructive guidelines to minimize change. Conversely, the floodplain and gravel removal character- istics at sites with major changes supported development of guidelines primarily of a precautionary nature. The analyses in succeeding chapters treat the changes that were meas- ured at individual study sites. There are sites, for example, where species diversity increased as a result of site disturbance. In some contexts, this increased diversity would be considered a beneficial effect of gravel removal. However, in the context of this project, this effect initially was evaluated equal to one which caused an equivalent decrease in species diver- sity because it reflected a change from the naturally evolved condition. This project treats al I changes consistently and objectively as a change from the natural, and special interest perspectives are neither recommended nor encouraged. However, it is recognized that a resource man- ager in certain circumstances may be greatly influenced by the need to consider a site from a multiple or optimal use standpoint. For example, subsequent to gravel removal a deeply dug site might be considered as a water source in areas where winter supplies of water are minimal. Several study sites were deep pits that contained water throughout the year. For- mation of a pit represents a major change from the natural situation and the site will not revert back to a natural situation for many years, if at all. In the context of this project, pits represent a major divergence from the natural. However, when considered from the standpoint of multiple use or habitat diversification, a resource manager may elect to recommend or ap- prove a permit for this form of gravel removal. In these situations the resource manager will be able to predict the results of such an operation by review of the following sections in this report. PROJECT DESCRIPTION A 5-year gravel removal study was initiated in mid-1975 to evaluate the effects of gravel removal from arctic and subarctic streams in Alaska. The primary purpose of the project was to provide an information base that will assist resource managers in formulating recommendations for minimizing detrimental environmental effects of removing gravel from arctic and sub- arctic streams. To achieve this the following objectives were met: • A comprehensive literature review and synthesis was conducted to evaluate known and conjectured effects of gravel removal and other similar disturbances on floodplain environments. • Physical, chemical, and biological characteristics of seven sites inhabited by fish after gravel removal were evaluated in moderate detail on a short-term basis. • Physical, chemical, and biological characteristics of 18 sites that reflected various removal methods, stream types, and times since completion of operations were determined in gross detail and on a short-term basis, • Relationships between parameters related to gravel removal operations, geomorphic characteristics of streams, water quality, and biota were eva I uated. The study of three sites prior to, during, and immediately after gravel removal was an original project objective that was eliminated due to a lack of suitable sites meeting project schedules. A thorough and broad-spectrum evaluation of the impacts gravel removal can have in floodplains requires assessment from a number of disciplines. To look at only one element could lead to conclusions and recommendations that might cause major changes to a riverine system on a long-term basis. Therefore, the approach taken in this study included analyses in the follow- ing six d i sc i p I i nes : River Hydrology and Hydraulics Aquat i c Biol ogy Terrestrial Ecology Water Qua I i ty Aesthet i cs Geotechnical Engineering This approach not only allowed analysis by individual discipline, but per- mitted consideration of the i nterdi sc i p I i ne trade-offs inherent in evalu- ations of disturbances to natural environments. For example, gravel mining techniques that would avoid effects on aquatic biota could require removal of important floodplain habitat used by terrestrial fauna or be impractical from geotechnical considerations. These disciplines were selected for the study because they were be- lieved to cover the various impacts that were known or surmized to be associ- ated with gravel removal. Due to a paucity of background information, it was not possible to be assured that al I significant impacts were addressed by these disciplines. Although the main purpose of this gravel removal study was to provide an information base for recommendations to be made by resource managers, another important contribution is to provide a base for subsequent long- term studies. For example, a problem needing extensive study is the effect of removing gravel from many sites in one river system, as occurs along highways and pipelines when they parallel floodplains for routing or geo- technical reasons. This problem is not treated in the present study and, in fact, was consciously avoided when sites were selected. REFERENCES Bliss, L. C, and E. B. Peterson. 1973. The ecological impact of northern petroleum development. Fifth International Congress. Arctic Oil and Gas: Problems and Possibilities. Le Havre. 26 pp. Bui I, B., and K. M. Scott. 1974. Impact of mining gravel from urban stream beds in the southwestern United States. Geology 2(4) : I 7 I- I 74. Burger, C. , and L. Swenson. 1977. Environmental Surveillance of Gravel Re- moval on the Trans-Alaska Pipeline System with Recommendations for Future Gravel Mining. Joint State/Federal Fish and Wildlife Adivsory Team Special Report No. 13. Anchorage, Alaska. 35 pp. Federal Water Pollution Control Administration. 1968. Sand and Gravel Waste Evaluation Study, South Platte River Basin, Colorado. Forschage, A., and N. E. Carter. 1973. Effects of Gravel Dredging on the Brazos River. Texas Parks and Wildlife Department, Inland Fisheries Research . Klein, D. R. 1973. The impact of oil development in the northern environ- ment. Proceedings I n terpetro I eum Congress 3:109-121. Rome, Italy. LaBelle, J. C. 1973. Fill Materials and Aggregate Near Barrow Naval Petro- leum Reserve No. 4, Alaska. The Arctic Institute of North America for the Office of Naval Petroleum and Oil Shale Reserves. Washington, D. C. 146 pp. McPhail, J. F., W. B. McMullen, and A. W. Murfitt. 1975. Design and con- struction of roads on muskeg in arctic and sub-arctic regions. Six- teenth Annual Muskeg Research Conference. Montreal, Quebec, Canada. 51 pp. Michael Baker, Inc. 1977. Accounting of sale/free use permit materials. Prepared for Alyeska Pipeline Service Company. Anchorage, Alaska. Northern Engineering Services Company, Ltd. and Aquatic Environments, Ltd. 1975. Reconnaissance of the Alyeska Pipeline-Material Source Borrow Methods and an Evaluation of These Methods with Respect to Aquatic Habitats. Canadian Arctic Gas Study, Ltd. Calgary, Alberta, Canada. Sheridan, W. L. 1967. Effects of Gravel Removal on a Salmon Spawning Stream. U. S. Department of Agriculture, Forest Service. 26 pp. Weeden, R. B. , and D. R. Klein. 1971. Wildlife and oil: a survey of cri- tical issues in Alaska. The Polar Record 1 5(9) :479-494. West, G. C. 1976. Environmental problems associated with arctic develop- ment especially in Alaska. Environ. Conserv, 5 ( 3 ) :2 I 8-224. APPROACH AND METHODOLOGY E. H. Follmann SITE SELECTION The site selection process began in July 1975 and initial work in- volved contacting various agencies and groups to locate potential study sites. Among those contacted, the main sources of information were the Bureau of Land Management, the Alaska Pipeline Office, the Alaska Division of Lands, and the State Pipeline Coordinator's Office. In addition, the Alaska Department of Highways (now Alaska Department of Transportation and Public Facilities) provided a considerable amount of information. A total of 575 potential sites were identified and subdivided into three areas north of Latitude 66 — the North Slope, the Yukon River Basin, and the Seward Peninsula — to obtain representative sites throughout arctic and subarctic Alaska. Later in the project the Yukon River Basin sites were separated into Northern Interior and Southern Interior sites. Following identification of these sites, field reconnaissance was initiated to assess the suitability of the sites for the study and to characterize those sites considered potential candidates for the study. Sixty-four sites remained as candidates following field reconnaissance. To augment the drainage and material site descriptions developed in the field for the 64 sites, additional information on gravel removal activi- ties and watershed characteristics was obtained from various agencies. E. H. Follmann is presently associated with the Institute of Arctic Biology of the University of Alaska. I I topographic maps, and other data sources. Based on more complete site des- criptions, preliminary variables were established with which to compare and se I ect s i tes. Site comparisons were restricted to sites within the same region to insure adequate representation of the North Slope, the Northern Interior, Southern Interior, and the Seward Peninsula. Six sites were selected to represent the Seward Peninsula, eight for the North Slope, six for the Northern Interior, and five for the Southern Interior (Figure I). The sites were categorized by the presence or absence of fish on the basis of field observation and reliable background information. The sites that were known to contain fish after gravel removal were compared to determine which should receive additional study. All sites were previously mined. As stated earlier, sites could not be identified which would allow studies (within project schedules) before, during, and after gravel removal operations. All sites were named in ac- cordance with the U.S. Board of Geographic Names. However, two sites oc- curred on unnamed streams and were assigned project names of Skeetercake Creek (unnamed tributary to the Too I i k River) and Aufeis Creek (unnamed tributary to the Kuparuk River). When two study sites occurred on the same river, they were designated upstream and downstream respective to their locat i ons. Major Variable Matrix Following site selection the preliminary variables used to compare sites were reviewed to determine which should be considered major variables. Initial ly, nine major variables identified as either site characteristics or mining characteristics were selected to describe each of the 25 sites (Woodward-Clyde Consultants 1976). These parameters were chosen because they were thought to be important from the standpoint of assessing gravel removal effects, they best described the sites, and they allowed selection of sites which exhibited the greatest variety of variables. The variety was especially important because it insured that sites were different, thus 12 NORTHERN INTERIOR SOUTHERN INTERIOR Figure 1. Location of the 25 gravel removal study sites in Alaska 13 permitting assessment of the effects of various gravel removal procedures on sites with different physical and biological characteristics. The major variables were again reviewed following the field inves- tigation, when detailed site characteristics were available to determine which were still suitable for comparing the 25 material sites. The seven variables selected for the final Major Variables Matrix included: • Drainage basin size, • Channe I > width, • Channel configuration, • Channel slope, • Stream origin, • Type of gravel removal, and • Location of gravel removal. These parameters were categorized as either Physical Site Characteristics or Gravel Removal Area Characteristics. Each of the sites was characterized according to these variables (Table I). Definitions of these variables are included in the Glossary. Physical Site Characteristics. Drainage basin size and channel width are significant because the impact of gravel removal could differ depending on the amount of disturbance in proportion to the size of stream and flood- plain. Also, systems having greater discharge and bed load movement could be expected to regenerate a material site more rapidly than a system with sma I ler discharge and less bed load movement assuming the amount of mining disturbance is proportionate in the two streams. Categories used were small, medium, and large based on the drainage area above the site and sma I I, medium, and large based on the channel top width within the study reach at mean annual flow. Although from a hydrological standpoint categorization only according to drainage basin area would have been sufficient, we con- sidered it important to include channel width because width is a tangible measurement that can be observed at a site location. 14 O c •qD UIOJJ lUBlSIQ |duueg3 jea(sj pue|S| japueauj spjsino japueaui 3p|su| jeq |8uueq3-p|^ jeq lejaiei jeq luioj lauueqo pauopueqv 'qo j3ieAK-q5iH lauueqo dAiioy paiDduuoo lou-t!(j pal33UU03-)ld u|e|d lejseoQ lliqioo-) uie)uno|/\| daais aiejapo;^ PI!K\I iqBjejis snonuis Bujjapuea^ J! Ids papiBjg aBje-] uinjpai^ IIPUiS aBje-) uunipaii^ lieuis O fO (O •- o t^mo^fonuip^ CM n ^ CM CM CM X X X X X X XX X X X X X X X XX X X X X X X X X X X X X XX X X XX X X X X X XXX XXX X X X X X XX XXX X CI S: "■ U: " XX X X XX X X X X X X X X X X X xxxxxx xxxxxxxx X X X X X X X X X X X X X X X X X X XXX 3 z> v> z z LU a. Q CE < in X XXX xxxx X X XXX XXX X X XX X xxxx X X X X X X XX X X X XX XX XX X X XX XX XX X at X oj (J J^ J S 2 xxxxxx X X xxxxxx X XX X X X X X X X X X X - O > OJ S S £ g 5 O al Z > to i s > I (J -^ " S -»: o "- s s " -2 "i I Z> < ^ vt i/i — i/) V 01 > s " ^ £ o o ; tt M M ^ ^ ~ S ^ ^ . . oc a S ^ IA-. 11., £ o b b s 5 ^ i "8 c 6 = ^ s ^ C O) k- w n, I. o) « > o > > 15 Channel configurations vary from straight mountain streams to braided rivers. Factors associated with various types of streams such as bed load movement, bank erosion, and water fluctuation were considered important. Configurations included in this study were braided, split, meandering, sinuous, and straight. Channel slope, along with other variables, is a major factor governing water velocity, discharge, and sediment transport. Therefore, streams with s I opes categor i zed as mild, moderate, and steep were included. Stream origin was considered because it governs aspects of stream hydrology. Stream origin also influences the amount of bed load material available for transport, thereby indicating the regenerative capacity of a stream, and the availability of suspended sediment that could deposit in a gravel removal area. Categories used were mountain, foothill, coastal plain, and g I ac i a I . Other factors such as stream bed material, bank vegetation, and water- shed characteristics are important, but were not considered as major vari- ables. To a large extent these factors are accounted for by the major vari- ables and the physiographic provinces occurring within the regions. Gravel Removal Area Characteristics. Two major types of gravel removal operations used in floodplain areas are pits and scrapes, distinguished primarily by depth of excavation and permanent inundation by water after site closure. During site visits it was apparent that pits were either connected or not connected to an active stream channel. Because magnitude of change to a system could be greatly influenced by this factor, pits were characterized as either connected or not connected. Location of gravel removal sites within a floodplain influences the degree of impact and the regenerative potential of a site. Therefore, dis- tinctions were made between sites located in a channel, adjoining a channel, and separated from a channel. 16 To determine the impact of gravel removal over time and the regenera- tive capacity of various types of streams, it was necessary to observe sites that were active during different years. Information was not available for sites associated with construction activity early in this century, but was for sites ranging back to the late I950's. Specific descriptions of the regional characteristics, physical site characteristics, and characteristics of the gravel removal operation at the 25 study sites occur in a subsequent section. DATA REVIEW Available information regarding the effects of gravel removal and other similar disturbances in floodplains was reviewed. Information was solicited from many Federal and most State agencies, from various Canadian groups, and from literature sources. Due to a minimum of information on the effects of gravel removal, particularly in arctic and subarctic re- gions, some of the processes involved had to be discussed from a theoretical standpoi nt . The results of this work were included in a Preliminary Report prepared in 1976 (Woodward-Clyde Consultants 1976). This report should be referred to if a review of available literature is desired. FIELD STUDY OF SELECTED MATERIAL SITES Preparation for the field program began in Spring 1976 and the last site was visited in March 1979. Site visits were split over three summers with 7 sites studied in 1976, 10 sites in 1977, and 8 sites in 1978. In addition, seven sites were visited during the winters of 1977-1978 and 1978-1979 to determine the presence or absence of fish, to record water quality parameters, and to describe the occurrence of icing conditions. During the 1976 field program field teams representing River Hydrology and Hydraulics, Aquatic Biology, and Terrestrial Ecology worked each site 17 simultaneously. The Aquatic Biology team also collected water quality data. Simultaneous effort of field teams was considered advantageous during the first field season to insure coordination of work where necessary. In ad- dition, simultaneous work permitted on-site discussion of methodology changes by all disciplines, thus further insuring coordination and co- operation. During subsequent field seasons, some of the sites were visited by individual field teams, but all teams visited the sites during the same summer. These individual visits allowed each team to visit sites during peak events for parameters associated with their discipline. Data and sample collection areas were flagged to facilitate collection of data at the same sites during subsequent visits by either the same or different teams. In addition, the hydrology and hydraulics and terrestrial teams placed semi- permanent posts at each site from which to initiate surveys for future studies. The following section includes a review of the field and laboratory programs conducted during the field effort. Programs are described only for River Hydrology and Hydraulics, Water Quality, Aquatic Biology, and Terrestrial Ecology because these were the only disciplines for which data were specifically collected. Geotechnical Engineering and Aesthetics relied completely on field information collected by other groups. River Hydrology and Hydraulics Introduction. Emphasis of the field program was on describing local fluvial geomorphic processes, obtaining evidence of past flood histories, measuring river hydraulic parameters, investigating sediment transport properties of the channels, describing river processes, and investigating specific effects of gravel removal on these factors. Photographs were taken for documentation of significant features. Hydraulic and hydrological data collection were coordinated with the water quality, aquatic biology, and terrestrial ecology studies. Hydrological and geomorpho I og i ca I literature pertaining to each site and its drainage basin, e.g., hydrological records, surficial geology. 18 and aerial photographic interpretations were also used in the analysis of each site. Geology and Geomorpho I ogy . Using topographic maps, stereo aerial photo- graphy, and surficial geology maps, a brief analysis of each drainage basin was made to evaluate the geomorpho I ogy of the river val ley, the river ter- races, and the present and past regime of the river. The morphological fea- tures pertaining to the general area around the material site were verified in the field. Hydrology. The U.S. Geological Survey Water Resources Records were reviewed for flow measurements within a study site's drainage basin. Where flow measurements were representative, various key discharges with the respective stages were estimated and documented. In the field, evidences of floods were investigated. Where sufficient data could be obtained at the study site or near vicinity, a stage-discharge relationship and flood frequency analysis were included in the data package. For the rivers that had no past flow records, the hydrology was synthesized using a regional flow analysis (Lamke 1979). Hydrau I i cs. Hydraulic parameters for each river channel and f I oodp I a i n were measured in the field. At each study site cross sections were surveyed upstream from, within, and downstream from the area of gravel removal (in conjunction with the aquatic ecology program) to measure the following hydraulic parameters: width, depth, and area. All cross section locations were documented and elevations referenced to temporary benchmarks. The longitudinal slope of the water surface and, where possible, the bed were surveyed. All surveys used standard surveying techniques. The discharge at the time of the survey was measured using standard techniques (Buchanan and Somers 1969). Materials and Sediment. Representative samples of the river's flood- plain surface material were obtained upstream and downstream from the gravel removal area using the photographic-grid method (Kellerhals 1971). These were considered to be representative of the channel bed material. The size 19 distribution was determined by the f requency-by-number method. In addition, the underlying material was measured using hydraulic sieves and the size distribution determined by percentage-by-weight. The river bank materials were described at cross section locations based on a subjective evaluation and photographed for documentation. Ma- terial gradation samples of river bank materials were not obtained. Channel Processes. The fluvial morphology at each site was assessed using comparative aerial photography. In the field, fluvial morphological features were verified and documented in more detail, e.g., gravel bar types, bed formations, scour holes, and sediment deposition. Degradation and/or aggradation upstream from, and downstream from the gravel removal site were investigated. River Ice. In the field, evidences of ice processes (breakup jams, ice scour, gouging, and aufeis) were documented to help evaluate the role of ice on the river morphology. Water Qua I i ty Water quality parameters measured were temperature ( C), dissolved 2 oxygen (ppm), conductivity (m i cromhos/cm ), turbidity (JTU), suspended solids (mg/S,), oxidation-reduction potential (MV), and pH (Table 2). Water quality measurements were taken at the aquatic macro i nvertebrate sample sites. Usually the measurements were taken along a transect across the river or pit with the number of replicates within a site adjusted to the size of the water body. The measurements were normal ly within 30 cm of the water surface, although depth profiles were taken in pits. Aquat i c Biol ogy Introduction. Field emphasis was placed on aquatic Invertebrates, changes in fish distribution in relation to the gravel mined area, and potential fish spawning and rearing habitat during the ice-free period. 20 Table 2. Methods Used for Measuring Water Quality Parameters with the Number of Replicates Taken per Study Area Parameter Method of determi nat i on Rep I icates per study area Dissolved oxygen Temperature Conduct i V i t y Tur b i d i ty Suspended so I ids Oxidation-reduction potent i a I pH YSI Model 57 DO meter YSI Model 57 DO meter Hach Model 2510 conductivity meter Hach Model 2I00A turbidimeter Millipore filter procedure (5 pm f i I ter ) Delta Scientific I 2 I 2-P2 ORP meter Delta Scientific 1212 pH meter Hach pH kit 3-15 3-15 3-15 2-11 I - 3 2-5 I - 5 Additional visits were conducted to specific sites if potential over- wintering habitat or suspected spawning areas were present within the mined area. Study sites were categorized into two groups. Eighteen sites were visited once during the open water season. Seven sites with known fish utilization in the mined area were subject to additional field study. These seven sites were visited on three separate occasions during open water conditions of I calendar year. In addition, seven pit sites where winter utilization by fish was suspected were visited to document overwintering. The 18 sites subject to a less intensive field program were visited only once. Selection of Sample Areas. Three sample areas were selected at all sites: upstream, within the mined area, and downstream. Selection of up- stream and downstream sample areas was based on similarity to the aquatic and terrestrial characteristics exhibited in the mined area prior to gravel removal. Selection of sample areas was made so that substrate, depth, width, velocity, and poohriffle ratio were similar at the upstream and downstream I ocat i ons. The upstream area was typical ly located at least 400 m above the mined area and the downstream area was between 400 and 800 m below the mined area. Selection of the 400 m criteria was based on the assumption that the hydrological effect of gravel removal would be minimal that far up- stream. Selection of a downstream area between 400 and 800 m below the mined area was based on the probability that changes occurred in this area either during or immediately after gravel removal. At sites with more than one mined area, additional sample areas were selected to assess effects. Similar selection criteria were used. Selection of Sample Gear. Fish and aquatic macrolnver tebrate sampling gear were selected relative to the types of habitat present. Features such 22 as width, depth, stream velocity, shoreline configuration, stream bank vegetation, obstructions, channel substrate, and presence of pits affected the gear selection process. Sample gear used at each study site is listed i n Tab I e 3. Sample Program. Information recorded in the field included stream name, sample location and description, description of the disturbed area, and the date, time, and existing weather conditions. Visual surveys were conducted within sampling areas to describe habitat and to record the pres- ence of fish. Sample Collection, Disposition, and Analysis. A variety of seines with square mesh (3.2 mm), 6 to 10 m long and 1.8 m deep, were used. Seines were extended across the stream from bank to bank and pulled downstream in narrow streams. In larger streams and pits the quarter-haul technique was used. Experimental, multifilament gill nets 15 x 1.8 m, with panels of 12.7, 25.4, 38.1, 50.8, and 76.2 mm square mesh, were anchor-set in pits, and, in one case, in the deep, slow-moving section of a large river. A backpack shocker, one of the least selective of all active fishing methods, was used in appropriate watercourses. Stream width permitting, a preselected length of stream was blocked with seines and the enclosed area shocked repeatedly until fish were no longer captured or observed. The area of the shocked section was usual ly measured to al low for density est ima t i on . Winnow traps selective for juvenile and small adult fishes were used to sample aquatic habitats. Traps were located in pools, riffles, and pits and were baited with salmon eggs. Traps were usually fished from 12 to 24 hours. A dip net was used at one site to capture juvenile fishes for identi- fication. Visual surveys were made at each site to record distribution and unusual events or critical habitats, such as spawning areas. 23 Table 5. Aquatic Biology Sampling Methods Used at Each Study Site Macro invertebr ate sampling gear ^ Surber Ponar Minnow Gi I I Electro- Hook & Set sampler g''ab trap Seine net shocker line line Study site F i sh samp I i ng gear Seward Peninsula Gold Run Creek S i nuk R i ver Washington Creek Oregon Creek Penny River Nome River + + + + + + + + + + + + + + + + + + North Slope Ugnuravik River Aufeis Creek Kuparuk River Skeetercake Creek Sagavan irk tok River Ivishak River Shaviovik River Kav i k R i ver + + + + + + + + + + + + + + + + + + + + + + + + + + + Nor ther n I nter i or Dietrich River-US Dietrich River-DS M.F. Koyukuk River-US M.F. Koyukuk River-DS Jim R i ver Prospect Creek Southern I nter i or + + + + + + + + + + + + + + + + + + + + + W.F. Tolovana River McManus Creek Tanana River-DS Tanana River-US Phelan Creek + + + + + + + + + + + + + + + 24 Captured fishes were identified, measured (fork length), weighed, and released except when preserved for reference. Data collected were used to determine species composition, size distribution, and relative abundance; estimates of density were made. These evaluations were compared within and between gravel removal sites. Macro! nvertebrates. A 30-cm square Surber sampler was used to collect macroi nvertebrates in riffle areas. Sampling areas were stratified by depth, bottom type, current velocity, and other variables that may have been corre- lated with benthic distribution. At most study areas three sampling sites were selected and five replicate samples were collected at each sampling site. Two sampling sites were selected in a few cases where there were multiple mined areas or where the river was not directly affected by gravel removal, e.g., a pit site away from the stream channel, with five replicates taken per site. A Ponar grab was used to collect macro i nvertebrates in pits. Single grabs were taken at several stations spaced to cover the main depth regions within the pits. Ponar grab samples were cleaned, separated (the slurry passed through a U.S. Standard No. 30 sieve), and placed in labeled con- t ai ners. Samples co I I ected wi th the Surber sampler were placed directly into labeled containers. Al I sample containers were f i I led with 70 percent al- cohol to preserve specimens for later examination. Samples were picked and sorted in the laboratory. Organisms were sorted into major categories and placed into labeled vials containing 70 percent alcohol. Identification was to the lowest practical taxonomic level. Data from quantitative samples were used to obtain total and individual taxon density. Data on standing crop and number of taxa were evaluated; comparisons were made within and between sample sites. Pit Sampling Program. Four pits were visited during March 1978 to assess the potential for fish entrapment and overwintering. During the 25 following summer these pits, plus three additional ones, were visited to assess if fish were present. The pits were then revisited during the 1978-79 winter to assess if fish remained in the pit after freezeup or moved into the river. If fish remained in the pit, subsequent visits were made to determine if fish could survive the winter. Sampling was conducted with a variety of gear types including minnow traps, set lines, gill nets, hook and line, and observation. In addition, an underwater television system was used for surveillance under the ice at two pits. Dissolved oxygen and temperature were measured when water was present. Ice thickness, presence or absence of flowing or open water, or both, and formation of aufeis by over- flow were recorded. Terrestrial Ecology Introduction. The terrestrial field program identified habitats af- fected by gravel removal operations and assessed the impact of habitat modification on associated wildlife. Qualitative and quantitative surveys were conducted during a 3-day field effort to characterize the plant com- munities and serai stages present on disturbed and undisturbed areas. Wild- life utilization of these habitats also was evaluated. The undisturbed sites encompassed serai stages likely to develop with time on the disturbed site, and were believed to be most representative of the disturbed areas prior to gravel removal. The program was expanded to 5 days at one representative study site (regional representative site) in each of five geographical areas: Arctic Coastal Plain (North Slope), Arctic Foothills (North Slope), Seward Penin- sula, Northern Interior, and Southern Interior. The increased time at these study sites allowed for additional sampling efforts using the same sampling procedures. Soils. Soil sampling was conducted within each habitat on disturbed and undisturbed sites to evaluate the growing conditions and the potential for revegetat i on. Within each habitat or definable soil unit, the character of the upper horizon, depth of organic layer, surface drainage, and domi- 26 nant vegetation were recorded. Approximately 15 subsamples were collected with a soil auger-tube sampler from the ground-cover rooting zone (approxi- mately the upper 20 cm). These subsamples were combined to form one compos- ite sample for each soil unit. Composite samples were air dried and ana- lyzed for pH, percent organic matter, and percent nitrogen, phosphorus, and potassium. A particle size distribution analysis was conducted to deter- mine the percent sand, silt, and clay in the composite sample. Vegetation. Vegetation surveys delineated the major cover types with- in the study area. Within each habitat, the serai stage of development was noted and the plant species were recorded. Qualitative site descriptions were augmented by limited use of quan- titative sampling methods that employed a systematic, nested plot design (James 1978). Strand or patch habitats required "spot" location of nested plots or qualitative description only. Description of the overstory vegetation included the following param- eters: dominant and subordinate tree species, average height and DBH (diam- eter at breast height) of the stand and stand components, and representative ages by species and height class. A limited number of circular plots (0.04 ha) were used to quantitatively sample each habitat. Forester's calipers or a diameter tape, or both, were used to determine tree DBH; tree height was estimated and an increment borer or cross-sectioning method was employed to determine the age of woody plants. Increment cores and cross sections were returned to the laboratory for staining and age determination when necessary. Shrub growth within each habitat was described by identifying species composition and relative density, average height by species, and representa- tive ages by species and height class. Stem and clump density counts were conducted on a limited number of systematically located, 0.004-ha circular plots. Selected shrubs were aged by cross-sectioning above the root collar. Evidence of herbivore browsing was noted. 27 Ground cover sampling identified species composition within each habi- tat and provided an estimate of percent surface coverage for each taxon. Percent surface coverage was visually estimated in systematically located, 0.0004-ha plots. Percent surface coverage was estimated as follows: if only one plant of a given taxon was present and its coverage was very sparse, it was rated at I percent coverage; if more than one plant of a given taxon was present, but its coverage was less than 10 percent of the plot's surface area it was rated at 5 percent coverage; the percent coverage of al I other taxa was estimated in increments of 10. Wildlife. Evidence of wildlife use of disturbed and undisturbed areas was recorded at each site. Direct observations and evidence of use (tracks, trails, nests, dens, runways, food caches, and scats) were keyed to their presence in specific serai stages. Historical use of a cover type was noted (i.e., hedged growth form of preferred browse species) and serai stages critical to certain life history stages of wildlife were inspected. The disturbed area was examined for the presence of special attractants or deter- rents to wildlife use of the site. An avian census was conducted in disturbed and undisturbed habitats at al I study sites; attempts were made to visit the five intensive study sites during the peak avian activity period. The census in homogeneous habitats employed a Modified Strip Plot technique for three consecutive morn- ings (five mornings at the intensive sites) to obtain data on the species present and habitat utilized. Sma II, isolated habitatswere qualitatively surveyed to ascertain avian species occurrence. Waterfowl, shorebirds, and game birds were inventoried by total counts when areas of concentration were clearly visible. Sma I I mammals (shrews, voles, and lemmings) were inventoried at al I sites in disturbed and undisturbed habitats using a trap and removal tech- nique. A "line" or "spot" trapping configuration was used in all cover types. Trapping was conducted for two nights at nonintensive sites and four nights at regional representative sites with the traps checked, baited, and reset each day. The species, sex, age, and weight of captured specimens were recorded to assess occurrence and characteristics by habitat. 28 Collection of terrestrial invertebrates was conducted at all fish intensive sites and at the regional representative sites. Collections were made adjacent to the watercourse at the disturbed site and near the upstream aquatic sampling station to assess the availability of potential food sources for the aquatic environment. Sweep nets were used to collect inverte- brates. Specimens were preserved in 40 percent alcohol and returned to the laboratory for identification. DATA BASE The data base, the third end product of the gravel removal study, (the Technical Report and Guidelines Manual are the first two end products) consists essentially of all information collected during site selection and field data collection. Information for each of the 25 study sites includes: • Case history information including mining plans and permits, if avail- able; • Biological, hydro I ogica I , and water quality field data; • Geotechnical evaluations; • Tabulation of data summations; • Computer printouts for aquatic ecology and hydrology and hydraulics; • Draft site description reports; • Site photographs, including both ground and aerial; • Topographic maps showing site location; and • Depiction of actual data collection areas within each site. 29 The information is in a form to allow any professional to evaluate where the data was collected, what data was collected, and the general conclusions of the original investigator. This data base is on file with the U. S. Fish and Wildlife Service. It will not be distributed routinely with the Technical Report and Guide- lines Manual. Due to the mass of information available, a specific need will have to be identified before the data relevant to that need can be provi ded. TECHNICAL REPORT Analyses of field data, beyond the immediate data reduction after site visits, began in winter 1977-78. This initial effort prepared descrip- tions of each of the study sites visited in previous summers and analyzed data specific to each site. Brief summaries of essential information rele- vant to each of the 25 material sites studied during this project are in- cluded in the subsequent chapter. These are included to orient the reader for the discussions that follow in the individual discipline chapters. Data syntheses for all sites did not begin until after the 1978 field season. Analyses of combined site data are contained totally in this report. Each of the six disciplines included in the project, (River Hydrology and Hydraulics, Aquatic Biology, Terrestrial Ecology, Water Quality, Aesthetics, and Geotechnical Engineering), is discussed in separate chapters. These chapters include some integration with other disciplines. For example, Aquatic Biology is dependent, for some of its data interpretation, on the Water Quality parameters measured, and on the physical changes that are described in the River Hydrology and Hydraulics section. An interdisciplinary overview of the effects of gravel removal follows the discipline chapters. This chapter reviews the analyses of the six disci- plines in terms of the similarities and differences that are evident. An important aspect of this chapter is discussion of the tradeoffs and com- parisons between disciplines that must occur with respect to the siting. 30 operation, and closing of material sites. Where possible, the similarities in approach of the various disciplines to minimize disturbance from gravel removal are emphasized because these conditions maximize protection of floodplain environments. GUIDELINES MANUAL The Guidelines Manual (printed separately) is based on the evaluations and recommendations contained in the Technical Report, on the preliminary guidelines developed in an earlier phase of this project (Woodward-Clyde Consultants 1976), and on stipulations and recommendations used by certain resource agencies when reviewing material site applications and projects. The guidelines are intended to provide guidance to the persons respon- sible for writing material site permits and for planning resource or indus- trial development in localized areas. The guidelines also are helpful to potential applicants for material site permits because they will help in planning a project characterized by minimal environmental perturbations. The guidelines are not designed as stipulations to be attached to each permit granted. If used in this manner contradictions in siting, opera- tional, and rehabilitation procedures could occur, thus negating the value of the guidelines. It is intended that the guidelines user evaluate the proposed project within the context of the guidelines, and the proposed area for the material site, to insure that it will develop in an environmen- tally acceptable manner. The guidelines were developed for use by personnel with some background in an environmental science. Ease of use was considered necessary because, at least on large projects such as pipelines and roads, permit agencies can be inundated with applications requiring quick consideration. A set of guidelines that are cumbersome and inefficient to use, under these circum- stances, could foster disregard of the guidelines or their misuse e.g., attaching the guidelines as stipulations to a permit. 31 The guidelines, as mentioned, were developed with the assumption that the potential user has some experience with environmental problems and issues and, thus, appreciates the potential complexities associated with a material removal project. It is strongly recommended that the user read the Technical Report and understand why and how the guidelines were devel- oped. A comprehension of the total project is considered necessary for intelligent, efficient, and expeditious use of the guidelines. Without this understanding, the guidelines could be viewed out of context and used inap- propr i ate I y . 32 REFERENCES Buchanan, T, J., and Somers, W. P. 1969. Discharge Measurements at Gaging Stations. Book 3, Chapter A8. Techniques of Water-Resources Investi- gations of the U.S. Geological Survey. 65 pp. James, F. C. 1978. On understanding quantitative surveys of vegetation. Am. Birds 52 ( I ) : 18-21 . Kellerhals, R. 1971. Sampling procedures for coarse fluvial sediments. J. Hydraulics Div. ASCE 97 ( HY8 ) : I I 65- I 180. Lamke, R. D. 1979. Flood Characteristics of Alaskan Streams. U.S. Geolog- ical Survey Water Resources Investigations 78-129. Anchorage, Alaska 61 pp. Woodward-Clyde Consultants. 1976. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska. U. S. Fish and Wildlife Service. FWS/OBS 76/21. Wash. D. C. 127 pp. 33 DESCRIPTION OF STUDY RIVERS L. L. Moulton, Ed. As previously mentioned, 25 sites were selected for study. These sites occurred in four geographical regions of Alaska and include a wide variety of Physical Site Characteristics and Gravel Removal Area Characteristics (Table I). Site locations are shown on Figure I. Table 4 summarizes dis- turbed area size, volume of gravel removal, and period of activity at each site. SEWARD PENINSULA General Description of Region The region of Seward Peninsula containing the six study sites is in the foothills of the Kigluaik Mountains, characterized by broad rounded hills with elevations of 250 to 700 m (Figure 2). The surficial geology at Sinuk River, Washington Creek, and Nome River is dominated by remnants of highly modified moraines and associated drift resulting from Pleistocene glaciation. Gold Run Creek however, is just outside the northern edge of glacial influence and the surficial geology is fine-grained alluvial and colluvial deposits with rare bedrock exposures. At Oregon Creek and Penny River the surficial geology is characterized by coarse and f i ned-grai ned deposits of alluvium and co I I uv i um associated with moderate to steep-sloped mountains and hills. Bedrock exposures are common on the upper slopes and crests. The region is general ly underlain with permafrost of variable thick- ness. Normal temperatures range from 3 to 13 C in the summer and -23 to -13 C in the winter. The annual precipitation of the region is about 30-40 cm, including approximately 130 cm as snow. 35 ir\ vo f'^ ir\ iT" in — Tt Tt 0\ Q\ O Cm >o ^O O O vO ^ — (0 r> .t) (0 n- _ _ f^ r^ r-' O^ 0) 0) — CN IN vO ^ N^ d d — d 6 K'. Csl 1^ ^ f\ W> lo CN — d d — d id CD r-- Oi CD CN ^ — - d d - - r^ O* CTi m vO — — ha — O CN CN O CD (N — fO r^ CD CD OtNO'^OOCNf^J O 0\ r* iTi O ^ CM — O O CN f^ CD O O O O - -^ — - d o o o — o> 1^ in lO ^ rO rsj 1^ — rO CO O CN — CM m o iTi in pr, r' — K> r-' — V f*^ — ^ ■ CO ro CO o CO ^ 03 C7i m oiinocNooo — p-inocooo — ov oj csj in r- ^o ** 00 03 ^ Kl O [^ O O r- CD Csj O O CD OD ■<* in « ^ — o CN O O fA O O 00 ■*- c « a o. o z> a I I u 0) c a; (u i_ ^ — o >• 3 £: en c a> c in Q) c e _ — (D i_ Qc (O gl 3 m (a (D o 3 ^ 10 > JZ (0 .fc. c Y If Ul 3 3 \ a. JC J£ () ■) n T 3 1 1 5^ >- L k O 0 Q) > 0) > ^ ^ r y V (T n- I O o ^ c r: ^ L. u. 0) OJ 0) r I i (T H o 2 o o £ £ ~> O 1 - Z Z < X u LEGEND • Active Channel • • AHigh-Waler Channe • a. • • • ■Side Channel • A • A .• A A A 8 A A AA ^ \ -.*i-^ • • AA ^ _% A (f A ^^ AA^ .s A £■ C, A D AA A A A imd T — TTTTTTT ;^ Y CHANNEL FULL TOP WIDTH (m) 120 1 50 180210 Figure 21. Maximum depths and corresponding top widths of undisturbed major, side, and high-water channels at 15 study sites with meandering, sinuous, and straight conf i gurat i ons. 88 a continuous process of erosion and deposition; erosion takes place on the outside bank, downstream from the midpoint of the meander bend and deposition occurs on the downstream end of the next point bar downstream. The rate of downvalley shifting varies from one river to another. The rate and direction of shifting is much more predictable than the lateral shifting of braided channels. A result of nonuniform shifting is channel cutoffs. The floodplain width of a meandering river is often roughly equal to the meander belt width, which is the average width from the outside of one meander bend to the outside of the next opposite meander bend (Figure 20). High-water channels on the inside of point bars are typical on meander- ing rivers. Sediment transport in meandering rivers is typically moderate. Sinuous Configuration. A sinuous river is similar in plan view to a meandering river except that its sinuosity is between I.I and 1.5 (Figure 22). In sinuous rivers, point bars are smaller and downvalley shifting is generally less than that of a comparable-size meandering river. Other than the greater stability, sinuous rivers are quite similar in form and hydraulic characteristics to meandering rivers. Straight Configuration. A straight river flows in a single channel with a sinuosity less than I.I (Figure 25). The thalweg, or deepest part of the channel, typically wanders back and forth within the channel with alternate ground bars formed by sediment deposition opposite those locations where the thalweg approaches the side of the channel. The alternate bars may or may not be exposed at low flows. Rivers with a long reach of straight channel pattern are much less common than rivers with other configurations. Banks of straight channels are expected to be relatively stable. Sediment transport is likely to be light to moderate in these systems. Other Processes. Rivers with any configuration may be found in narrow mountain valleys and on alluvial fans. Rivers in these locations have dif- ferent processes of erosion and deposition than those described for the' typical river with the same configuration. Channel configurations of moun- tainous rivers are typically not controlled by alluvial processes, but 89 PLAN VIEW SECTION A-A Figure 22. Schematic diagram of the plan view and cross section of a typica sinuous river. PLAN VIEW SECTION A-A Figure 23. Schematic diagram of the plan view and cross section of a typical stra i gh t r i ver . 90 rather are controlled by geological and morphological features of the val- ley. Mountainous rivers commonly have very little or no floodplain and consequently, have small quantities of gravel. Alluvial fans develop when a steep gradient stream flows onto a substantially less steep terrain; its sediment transport capacity is significantly reduced causing sediments to be deposited. This deposition fills the channel, thus forcing the flow to develop a new channel. This may occur by a gradual migration process or by a rapid abandonment of one channel to develop a new channel. Such processes develop a partial cone-shaped deposit of gravels with the apex being near the end of the steep gradient river val ley (Figure 24). The fan may or may not be vegetated; denser vegetation implies greater stability. Changes Due to Gravel Removal The most common change to the channel configuration resulting from gravel removal was a shift towards a more braided configuration as indi- cated, in part, by an increase in the number of channels. A decrease in lateral stability of the channels was often associated with changes to more numerous channels. These changes were mos1- prevalent in scraped sites and most prominent in single channel sites. Gravel removal at many scraped and pit excavated sites caused a diversion or a high potential for diversion of flow through the gravel removal site. These observed channel config- uration changes were given quantitative ratings for comparative purposes (Table 5). These changes in channel configuration are discussed in more detail in the following sections. Braiding Characteristics. The two braiding characteristics considered were the number and stability of the channels. The most significant changes in these characteristics resulted from scraping operations in straight, sin- uous, split, and meandering rivers with lesser changes observed in scraped braided rivers. This difference was expected, because braided rivers had such characteristics prior to gravel removal, thus, any change was compara- tively less significant. The locations of the gravel removal operations that caused the most significant change in the braiding characteristics were those which disturbed the bars adjacent to active channels or those which caused diversion of flow into the material site. Disturbance of the bars adjacent to active channels can hypotheti- cal ly reduce the flow within the channel during fl oods because flow spreads out through the mined area. The reduced flow within the channel would reduce the ability to transport sediments; sediment deposition within the channel may result. This deposition would potential ly aggravate the problem by further reducing the cross-sectional area available to the flow. This pro- cess can result in widening the channel and the development of mid-channel bars. Although the potential for this hypothetical process exists, it was not observed at the study sites. Braiding characteristics increased at many sites due to the diversion of flow through the site and the lack of a well-defined channel to confine the flow. The flow thus spread through the material site and likely did not have sufficient scour potential to develop a new channel. Thus, numerous poorly-defined channels flowed through the site. 92 13 0) a. o u 0) .»- Ul u 0) ro -(— 1- ■— (D 00 SI u in CM c o a> •— ^ -4— 4_ ro i. V4- D o CD ■ — SI M- u c (D o LU u -4— — ro 0) c c c o (0 x: 4— u ID 1- c 0) — Q. o 0) CD — c (D ro > ^ o u E c ro — 1- A- o (0 ce Q) JZ c .4- O — E -*— O (0 1_ o ^*- c u o ir\ 0) (O c o ^o — cri Q) in 3 CL L O (0 0) i- 1_ > SL O — -1- in c O SI — '3\J in D -1- i_ o - > ^ 0) .- c — c — ro ^ -c ro (_> ■>- ifl o 0) c • c o ro z SI o — (D ID > o Q) 10 E 1_ u 0) 10 o i_ 1_ > u3C^r^Oi£)Oii£iOOO\0'Or-OCTic^iOCTir-r~-UDOi^r^ I « I I O I I I I I r- I o I I I I I r- I I I I CO 03 I r^ r^ inro I OvtoinoOhoo^t^ — •^^incNiTiioirio-^ininininintnLn in-* I OiryCTir--criO>oaiirii^triOooioa)iriioirioaoiriiniriinir\ir\^ < CD < CO < OQ < CD < CD (_) c o — 0) (U 1_ u c o -*— en c SI U) ID (U — > cc — tr C Q) c E 0) o CL Z 0) > — jt: q: Q) Q) — o > 10 in I. — D ce 1_ a> — 0) ce 1_ (_> -^ o 0) -t- E 10 E 0) ID 1_ OJ ■•- i_ in -1- c in s Q. O 1- u 0) 0) > > E 10 Q> \. IS) CL I L. 0) > o j£ j^ ._ Q) 1- -^ ID lU CL 0) ID 1_ O — 1_ C Ct CC L. ID 10 > SI 10 in en — ID > SI sz o o +- 4- E ID Q) 1_ in c S o Q I 1_ > o is: -^ O L Ll 0) > 0) — — ce ■D ■o E > 10 c ID > -ii O Q) — ^ Q) O 01 i_ I— O) l_> I. ^ U o o in 0) u. =J CL c in -I- 10 O in S 1_ 0) o CL g S E 10 0) E 1- 10 -I- 0) in i_ C 4- g in o Q. Q Z) 1- i_ -bi: 0) 0) 0) > > 0) — — l_ ce cc o 10 10 c ID 10 ID — C C 0) 10 ID H cn 10 CL cn c i o c o in 01 o c O O 93 Table 5. Footnotes Number of channels ratings: Number of active channels in the mined area R Number of active channels upstream from the mined area 10 3 < B^ 9 2.5 < B„ £ 3 8 2 < B_, £ 2.5 7 I .5 < B^ < 2 K 6 1 < B < 1.5 or other B values if they are within normal variation ranges of the river 5 B = I of i f other data indicate no change K 4 0.67 < B < I or other B values if they are within normal variation ranges of the river 0-3 Not used Channel stability ratings: 6-10 Not used 5 No change in channel stability 4 Slight decrease in stability, but within natural stability vari- ation of the r i ver 3 Moderate decrease in channel stability due to gravel removal 2 Large decrease in channel stability due to gravel removal I Substantial decrease in channel stability due to gravel removal 0 Very substantial decrease in channel stability due to gravel remova I Flow diversion ratings: 10 High potential for river to divert all its flow permanently through the site 9 Diversion of a significant quantity of flow through the site oc- curred within several years 8 Moderate potential for river to di vert all of its f I ow permanent I y through the site 7 Moderate to high potential for some of the river flow to divert permanently through the site or for flow diversion through the site during flood events 6 Low potential for river diversion through the site 0-5 Not used 94 Ten sites had more than twice as many channels in the material site as were upstream. At four of these sites, Washington Creek, Nome River, Sagavanirktok River, and Middle Fork Koyukuk R i ver-Upstream, the numbers of channels increased more than three times due to gravel removal operations. Most sites (7 of 10) with large increases in numbers of channels also had a very substantial apparent decrease in the lateral stability of those chan- nels. Lateral stability evaluations were based on subjective judgements of stability indicators. Lateral stability indicators included the height and erodibility of the gravel bars at the edge of the active channels, the bed load transport characteristics evident at the time of the site visit, and the channel configuration. The Nome River is an example of a material site with increased braided characteristics (Figure 25). In this sinuous river, single channel flow was prevalent prior to the gravel removal operation; exceptions to this are the split in the channel immediately downstream from the material site location and two high-water or sma I I active side channels adjacent to the material site location. Approximately 20 years after the gravel was removed, the river was flowing in numerous, poorly-defined channels through the material site. The river apparently diverted into the scraped area soon after the operation was completed and has attempted to develop a well- defined channel since it diverted. The state of equilibrium between erosion and deposition in the Nome River was disturbed by the gravel removal opera- tion. To restore equilibrium it will probably take several decades from the time of the initial disturbance. Flow Diversion Through Site. Gravel removal operations caused flow diversion or a high potential for flow diversion at 12 of the 25 study sites. Sites with a high potential for the diversion of all of the flow permanently through the site included upper Washington Creek, Penny River, Nome River, upper Aufeis Creek, Skeetercake Creek, lower Middle Fork Koyukuk Ri ver-Upstream, and Phelan Creek. At most of these sites, all of the flow had already diverted when the site was visited. All of these sites were scraped and the lower Middle Fork Koyukuk R i ver-Upstream site was the only site where a buffer was known to have been used to separate the site from 95 August 1950 July 1977 Scale in Meters ' I 130 Figure 25. Comparative aerial photography of the Nome River showing change in channel configuration resulting from gravel removal activities. 96 the active channel. The vegetated buffer was approximately 30 m wide and roughly I m in height; vegetation was missing in and adjacent to a high- water channel which crossed the buffer. Low (0.3 m) dikes were used to block off this high-water channel. Flow began to divert through the material site during the first breakup following the removal of gravel. The buffer breached, apparently caused by overtopping and subsequent erosion of the top and downstream face during the flood. At the time of the site visit in 1978, 32 years after the mining took place, 85 percent of the flow was going through the material site. Scraped sites with a large amount, but not all, of the flow diverted through the material site by the time the site was visited included Sinuk River (in-channel site), upper Oregon Creek, Ugnuravik River, Sagavan i rk tok River, and Kavik River. None of these sites had a vegetated buffer. A major consequence of flow diversion through scraped sites was the development of braiding characteristics, as was discussed in the previous section. Another consequence was that flow in the former main channel (s) was eliminated or significantly reduced, thus affecting their hydraulics and their regime. Flow through scraped sites that had the potential to aid the replenishment of gravel within the site occurred at Sinuk River (in-channel site), Washington Creek, Oregon Creek, Ugnuravik River, Aufeis Creek, Kavik River, and Phelan Creek. At other sites, such as Penny River and Middle Fork Koyukuk R i ver-Upstream, flow through the site was probably eroding more sediments than it was depositing. Most (6 of 7) pit excavated sites had vegetated buffers separating the material site from the active channel (s). The exception is Ugnuravik River (Figure 26), which had only a 5- to I 0-m wide gravel bar separating the material site from the active channel. Therefore, the potential for flow diversion through this pit is high; flow has diverted through the site during floods, but the diversion has not yet been permanent. The two pit excavated sites on the Tanana River were judged to have moderate to high potential for some of the flow diverted permanently through the site within several decades following site closure. Both sites had 97 Scale in Meters I I 1 0 76 7 July 1977 Figure 26. Aerial photograph of the Ugnuravik River pit site showing the in- sufficient buffer zone. 98 approximately 30 m to 40 m wide vegetated buffers. The main ctiannel of the Tanana River has the capability to erode through such a buffer in less than a year. The side channel at the Tanana R i ver-Upstream site (Figure 27) eroded 3 m of the widest part of the buffer between early June and mid-September of 1978. At either of the Tanana River sites, it could take several years or several decades for the river to breach the buffer and flow through the pit, the length of time depending on the lateral direction of travel of the main channels. The Prospect Creek and West Fork Tolovana River sites were judged to have a moderate potential for all of the flow to divert through the pits. Both sites had vegetated buffers that included portions of abandoned channels. The upstream end of the abandoned channel, in both cases, causes a zone of weakness in the buffer. Even though, at both sites, the width and height of the buffers were likely sufficient to prevent breaching for several decades, zones of weakness in the buffers at the abandoned channels and channel aufeis development in the active channel may cause earlier flow diversion and buffer breaching. At the West Fork Tolovana River site, the upstream end was diked off and heavily riprapped; however, in spring of 1979, flow apparently overtopped the dike and scoured the channel lead- ing into the pit, leaving a large delta gravel deposit in the pit. Flood stage was probably high because of aufeis development in the channel. Channel aufeis development also influenced the Prospect Creek site (Figure 28). Aufeis developed in the channel reach upstream from the material site, reducing the channel capacity during the snowme I t runoff period. The runoff thus flowed directly down the valley, rather than following the ice-filled channel. The water flowed through the pit causing headcutting of the up- stream edge. The edge was subsequently riprapped to prevent further head- cutting. Doyle and Childers (1976) documented this April 1976 occurrence. HYDRAULICS Hydraulics, as used in this investigation, is the study of those param- eters which influence the mechanics of water flow through the study reach. The hydraulic parameters which were considered include hydraulic geometry. 99 Scale in Meters 1 1 1 0 168 /; t .. ' y> J' 11 July 1977 Figure 27. Aerial photograph of the Tanana R i ver-Upstream site with substan- tial buffer zone separating the pit from the active side channel. 100 Scale in Meiers I I I 0 145 11July1977 Figure 28. Aerial photograph of the Prospect Creek pit showing wide buffer zone separating the pit from the active channel. 101 channel slope, and local flow characteristics at flow obstructions. Hy- draulic geometry is defined as the geometric and hydraulic variables at a cross section that vary with changes in discharge. The hydraulic geometry variables discussed are top width, hydraulic depth, and mean velocity. Chan- nel slope (gradient) is the reduction of the water surface elevation in the downstream direction. A general discussion of these hydraulic parameters is presented in the following subsection, followed by a description of the effects on these parameters due to gravel removal. Undisturbed Condition The hydraulic geometry parameters considered herein are top width, hydraulic depth, and mean velocity. The top width is the width of the water surface at a given cross section and a given discharge (Figure 29). The Water Discharge ( Q ) Mean Velocity (V= x) Hydraulic Depth (d=:^) Cross- Sectional Area (A) Figure 29. Schematic diagram illustrating definitions of channel geometric and hydraulic variables. hydraulic depth is defined as the cross-sectional area of flow divided by the top width. The mean velocity is defined as the ratio of discharge to cross-sectional area of flow. An estimate of the carrying capacity of the channel is the conveyance, which is defined by: 102 K = CAR^ (I) where K = conveyance C = a coefficient related to ttie roughness of the channel A = cross sectional area of flow R = hydraulic radius X = a fractional exponent The discharge is directly proportional to the conveyance with the proportion- ality constant being the energy slope to a fractional power, usually i. The variation in the hydraulic geometry as a function of discharge at a river cross section is an indicator of the shape of the channel cross section. The shape primarily reflects the magnitude of the bank-full dis- charge which typically has sufficient sediment carrying capacity to shape a channel and occurs frequently enough to maintain the resulting shape. The top width, hydraulic depth, and mean velocity at a cross section are often expressed as a function of discharge in the form of power relations: W = a Q^ (2) D = c Q^ (3) V = k q"^ (4) where W = top width D = hydr au I i c depth V = mean ve I oci ty Q = di scharge a , c, k = coef f i c ients b, f , m = exponents Typical relations for a hypothetical river are shown in Figure 30. Sub- stituting the power relations for the hydraulic geometry variables into the flow continuity equation illustrates the interdependence of the variables: Q=AV=WDV (5) = (a Q^) (c Q^ (k q"^) , > n 'b + f + m) ,^, = ( a c k ) Q (6) 103 Thus, for continuity. and a X c X k = b + f + m = I (7) Discharge ^^^ Note All Scales Are Logarithmic Figure 30. Average hydraulic geometry of river channels expressed by relations of width, depth, and velocity to discharge at two locations along a river (modified from Leopold, Wolman, and Miller 1 964 ) . if a coefficient or exponent for one hydraulic geometry variable changes due to the gravel removal operation, at least one of the other variables must 104 also change to maintain continuity of flow. Generally speaking, if a channel is widened, it often satisfies continuity by becoming shallower. Similarly, if a channel slope, or gradient, is increased, thus increasing velocity, continuity is commonly satisfied by a reduction in depth. Exponent values for selected study sites and other rivers are given in Table 6. The ex- ponents exhibit a wide range of variability for different rivers; Rundquist (1975) found that the exponents and the coefficients can be expressed as functions of the bank-full discharge. The coefficient c and exponent f in the power relation for hydraulic depth were found in addition to be a func- tion of the median bed material size. The exponents in the power relations may change at a given site for discharges above bank-full because of the typically abrupt change in bank slope at bank-full conditions. The slope of the water surface profile for a typical river generally will parallel the bed slope at low f low, often producing a -sequence of r i f- fles and pools. At flood flows, the pool-riffle sequence is not apparent in the water surface profile (Figure 31). Flood Flow Water Surface Profile Low Flow Water Surface Profile m^^^}}^= Bed Profile Figure 31. Schematic diagram showing change in water surface slope in response to a change in water discharge. Naturally occurring flow obstructions in rivers can include vegetation, rock or snow avalanches, aufeis, and boulders. The effect of an obstruction on the hydraulics is to cause a local increase in velocity which often 105 Table 6. Values of Exponents for Hydraulic Geometry Power Relations Undisturbed Disturbed areas areas R i ver b f m b f m Kuparuk River 0.45 0.28 0.29 0.48 0.28 0.24 Sagavanirktok River 0.25 0.40 0.55 0.32 0.42 0.26 Shaviovik River 0.40 0.55 0.27 0.52 0.29 0.19 Middle Fork Koyukuk R i ver-Upstream 0.29 0.44 0.27 0.44 0.55 0.25 Middle Fork Koyukuk River-Downstream 0.54 0.28 0.18 0.57 0.29 0.54 Average values, midwestern United States'^ 0.26 0.40 0.54 Brandywine Creek, Pennsylvania 0.04 0.41 0.55 Ephemeral streams in semiarid United States^ 0.29 0.56 0.54 Average of 158 gaging stations in United States'^ 0.12 0.45 0.45 10 gaging stations on Rhine River 0.13 0.41 0.45 Average of 17 stations in Southcentral Alaska^ 0.19 0.59 0.42 Average of 50 stations in Upper Salmon River area, Idaho 0.14 0.40 0.46 ' W = a qJ D = c Q V= k Q^ Compiled by Leopold, et al. > 1 964 ) '^Emmett (1972) '^Emmett (1975) 106 results in erosion of the obstruction or bed scour adjacent to the obstruc- tion (Figure 32). Complete channel relocation is also a potential response to flow obstructions blocking a high percentage of the channel's cross- sectional area. Close Flow Line Spacing Indicates High Velocities Probable Scour Figure 32. Schematic diagram illustrating the effects of a flow obstruction on the local hydrau I ics. Changes Due to Gravel Removal Substantial changes in hydraulic geometry, slope, and flow obstructions resulted from gravel removal operations at roughly 60 percent of the sites. Typical hydraulic geometry changes in the mined area included increased channel top width, reduced hydraulic depth, reduced mean velocity, and increased conveyance. Changes in slope due to gravel removal operations took 107 the form of increases through the mined reach resulting from channel cutoffs and local slope redistributions affecting the pool-riffle sequence. Flow obstructions in the forms of material stockpiles, diversion dikes, and overburden piles have the potential for causing local scour, ice jam forma- tion, and si I tat i on. Hydraulic Geometry. Gravel removal operations caused changes in the natural cross-sectional shape of the active channels of approximately half of the rivers included in the study. The backwater analysis was not complete enough at some sites to confirm the hydraulic geometry change evaluation. A comparison of power equation exponents for cross sections in disturbed and undisturbed areas (Table 6) indicated a varied response to gravel re- moval. The coefficients in the power equations must also be considered to understand the effects of gravel removal. For example, at the Middle Fork Koyukuk River-Downstream site, the top width increased at a slower rate within the gravel removal area than outside of it. However, the coef- ficients in the power relations were greater for the disturbed than the undisturbed cross sections indicating that the top widths were larger at low flows in the disturbed areas than the undisturbed areas and were similar in both locations at higher flows. A qualitative evaluation of this effect can be made by comparing the relative channel widths in the material site at low flow and flood flow (Channel width and Flooded area. Table 7). The coefficient in the power equation for the top width was greater for the disturbed cross section than the undisturbed cross sections at eight of the sites; this difference resulted from a consistently greater top width at all discharges considered in the hydraulic analysis. The sites at which this occurred were Gold Run Creek, Washington Creek, Nome River, Aufeis Creek, Skeetercake Creek, Sagavan irktok River, and both sites on the Middle Fork Koyukuk River. At Sinuk River the exponent of the power relation for the top width was observed to be greater at the disturbed cross section than at the undisturbed cross section. This difference indicates that the gravel removal area had smaller top widths at low flows, but larger top width at high flows, than the undisturbed cross section. 108 c o X g o o => — i- u. ■>- m ^ O o c o '^I 3 — 0) n ro 0- — o O i. 0 — -t— _i ui m ■o 0) \- — J3 — Q) ID Q- i. 0 in O -DID 0) la "O Q) C 1_ O ro Q_ X) Q)(D X> (0 O 0) O i- — CD Ll- — (0 Q) n C -1- C X3 10 — ^ s u (D 01 > ro > O Q) (0 E 1. 1_ 0) (D o 1. > (rioocoCTiooooiriO\r-[^r-irir-~oiriini^Oi^irimir\intniriirioomLricri iTir- I r^irir^Lnir\ir\mt^triiriininir\ir\r^irioir\iriinr^inLr\ir\ir\tnir\ ^O ir\ QOir\kOir\or^r^i^iriirioyDm^oioioiriir\oi^ir\inirir-ir\iriir\ LnvooooocDmor^coo'Oioo«3«3ir\i£)irir-~o\cDr^ooininu3ir\tri*o '£ir-^criOCDaiuDOOOO^r~-aDO«3>£)r--irir~-vOCTiOOir\LnoirimcN coiooiriOOi^cDvou^incDioin^oirivomr^CTiOcniryiriiriiri^ < CD < m < DQ < m < DQ E (0 CO 1/5 1_ E Q) 3 Q lU E (0 I. 1 1 > ID Q. o js: ^ c c 4- -i: 0) •— 3 Q D 3 (0 g in 0) i_ Q) cc i_ 1 1 >- :^ > o Q. ^ 0) 1- U -!£ L > Q) 0) -^ •^ 0) -Jt: 1 1 0) (_) -^ — :^ CP 0 0) •— > > 0) o 0) i_ i_ ^ I. i. Q) L. cc (U > 0) 4— > ce 1_ — .— :^ ^ u 1— Q) a> Q) CD o 0) c Q) 0) U (U .— ^ ^ .— 0) ce cc \- u u i- > > Q) > o 1_ > CD J^ 1- CL 10 V- CL -i: > O o 1_ ^ O — .— U c — 4— u -— > •— o (J — ^ sz LL u_ 0) -(_ 1_ Ql CC l_) D ce CD ce •— > JC l- c j<: > cc o o > o o in cr c c cc (D in D 0) (0 (0 o — — 0) (D — Q) Ll D ID ID C js: •— o ^ 1_ — l- -»- > j= ^ 1- 1- ce a. C C C ID XJ =1 -C en c (U ZJ •— "O ■o m 4— CD (0 ID — — c Ul 0) c E c >-*- Q- 0) a> — (0 > 0) Q) TD ■D E o in s c £Z 0) o •— (0 1- Q) o O) Z) D -^ (0 > ^ 10 .— .— i_ 0) o ID (D JZ CO g o D- z z> < ^ CO iSl CO is: Q Q s s —3 Q- 5 s 1— h- D_ 0) CD ID Q- cn c g o c o in 109 Table 7. Footnotes Width and area ratings: _ Parameter in the mined area R Parameter upstream from the mined area where the parameter is: • top width of the channel Is) during the survey period for Channel Width • top width of the channel (si during floods of approximately Dank-full flood magnitude for Flooded Area • area of ponded water, excluding pits, for Ponded Area. * Wr, 2.5 < Wpj < 3 < Wp < 2.5 I . 5 < Wp < 2 6 I ^ *□ 1 '-5 or ether W values if they are within the natural range of variation of the river 5 W • I or i f other data indicates no change 4 0.67 1 W < I or other W values if they are within the natural range of variation of the river 5 0.50 ^ Wr '^ 0.67 0-2 Not used Overal I slope ratings: _ Length of disturped reach after gravel removal R ~ Length of disturbed reach before gravel removal 10 9 8 7 6 5 /L R .2 < S„ Lr ^°-^' 0.71 < L^ 0.77 I .3 or 0.77 < L„ < 0.83 K — — R I < S„ < I .2 or 0.83 < L„ < 0.91 .0 < S^ < I.I or 0.91 < Lp .0 Sr^Lr I or i f other data indicate no change 0-4 Not used "Local slope redistribution ratings: 10 Very steep slope followed by a very long pool 9 Steep slope followed by a long pool 8 Moderate slope followed by slightly longer than average pool 7 Slope and pool length slightly more than that in the undisturbed areas 6 Some local slope redistribution detected or likely to have occurred but not likely that of the natural river 5 No local slope redistribution 0-4 Not used Flow obstruction ratings: 10 Obstructions in an active low-water channel such that flow is di verted 9 Obstructions adjacent to an active low-water channel 8 Obstructions in or adjacent to high-water channels 7 Obstructions in the floodplain but away from any developed channels 6 Small obstructions not much different in size from those occurring naturally in the floodplain 5 No obstruct ions 0-4 Not used 110 Associated with the trend towards larger top widths in the gravel removal areas, the hydraulic depth in seven of these areas decreased. Sites with smaller hydraulic depths, in the mined area, for all discharges in- cluded Washington Creek, Nome River, Aufeis Creek, Skeetercake Creek, Sagavan irktok River, and both sites on the Middle Fork Koyukuk River. The mean velocity was consistently less at the disturbed cross section than at the undisturbed cross section at nine of the sites for the range of discharges included in the backwater analysis. These sites included Gold Run Creek, Washington Creek, Ugnuravik River, Aufeis Creek, Skeetercake Creek, Sagavan irktok River, Dietrich River-Downstream, and both of the Middle Fork Koyukuk River sites. At two sites, the rate of increase of velocity with discharge was different in the disturbed area than in the undisturbed area. At Sinuk River, the velocity increased at a lesser rate at the disturbed cross section than at the undisturbed cross section. At Middle Fork Koyukuk River-Downstream, the reverse was found. The conveyance, or carrying capacity of the channel, was consistently greater in the gravel removal area of eight sites compared with conveyances at undisturbed cross sections. These sites were Gold Run Creek, Sinuk River, Washington Creek, Aufeis Creek, Sagavan i rktok River, Kavik River, Dietrich River-Downstream, and Middle Fork Koyukuk R i ver-Upstream. The Sinuk River had a larger exponent or, equ i va I ent I y , a more rapid increase in conveyance with discharge than cross sections which were not disturbed by the gravel removal operation. Conversely, the conveyance at the downstream site on the Middle Fork Koyukuk River increased with discharge at a slower rate than did the conveyance of the undisturbed cross sections. Significant changes in hydraulic geometry were observed primarily at sites which were scraped, although not all scraped sites showed a signif- icant increase. Most of the significant changes were observed at meandering, sinuous, and straight rivers. Although no single gravel removal location caused a significantly greater change in hydraulic geometry than others, most of the sites that had significant change were those sites that were excavated by scraping in-channel and immediately adjacent-to-channe I loca- t ions. I N The area of ponded water, which includes those low-lying areas which accumulate water but are not effective in the conveyance of flow, was in- creased at roughly half of the study sites. This ponding indicated that the site was not smoothed during restoration, was excavated too deeply, or was not properly drained. Table 7 lists the relative effect of this parameter at the 25 study sites. The impact of the ponding to the hydraulics of the systems was not great. However, it was a concern to aesthetics and fish entrapment evaluations. Channel Slope. Channel slope changes took the form of an overall in- crease in slope or a local redistribution of slope. An overal I increase in slope was commonly due to the formation of a meander cutoff. A redistri- bution of slope without changing the overall slope occurred when the slope was increased leading into the gravel removal area and decreased through the gravel removal area. Table 7 indicates those sites which had slope changes. Study sites exhibiting an overal I increase in slope due to gravel removal were generally in small, nonbraided river systems that were exca- vated by scraping techniques. The location of gravel removal was an impor- tant factor affecting the overall slope of the system. Sites such as upper Washington Creek, Penny River, Skeetercake Creek, and lower Middle Fork Koyukuk R i ver-Upstream, that were excavated on the inside of bends, mean- ders, and islands most significantly affected the overall slope of the river system. This influence was expected because significant increases in slope are most likely to result from the development of a meander cutoff (reducing channel length and increasing slope). The Penny River gravel removal operation caused a significant increase in overall slope (Figure 33). The photograph of the site after the gravel was removed shows that the main channel flows in a relatively straight course along the inside of two broad meanders that were cut off in the excavation process. The channel length was reduced by a factor of two in the process, equivalent to doubling the overall slope through that reach. Doubling the slope has the effect of increasing the mean velocity by roughly 40 percent. 12 Scale in Meters I I I August 1950 July 1977 191 Figure 33. Comparative aerial photography of the Penny River showing change in hydraulic characteristics resulting from gravel removal activities. I 13 Gravel removal from active and high-water channels generally caused local slope redistribution. Removing gravel from bars and banks immediately adjacent to channels also appeared to cause a local redistribution of the water surface slope. An example of a local slope redistribution, which is similar to the situation at the Dietrich River-Downstream site, is schemat- ically illustrated in Figure 34. loodplain •Water Surface , Bed — Floodplain Water Surface Bed After Gravel Removal Figure 34. Schematic diagram illustrating an example of a change In local water surface slope that resulted from an in-channel gravel removal operat i on. Flow Obstructions. Flow obstructions in the form of material stock- piles, diversion dikes, and overburden piles had a larger potential for hydraulic disturbance on small rivers than those on medium and large rivers. This larger potential exists because the flow obstructions would have to be placed closer to the active channel due to the typically smaller floodplain width. There were no significant hydraulic impacts observed due to flow 14 obstructions, but the potential exists for bed scour at the base of the obstruction, erosion of the obstruction, and ice jamming at the obstruction. Erosion of a dike at Skeetercake Creek increased siltation as discussed in the following section. SEDIMENTATION Sedimentation includes the processes of erosion, transportation, and deposition of sediment. These are complex processes related to sediment and water flow properties. Attempts to quantify these processes provide, at best, estimates of the quantity. A very brief discussion of sediment size distribution, channel erosion, and sediment transport are given in the following section. Changes to these sedimentation characteristics due to gravel removal are then briefly discussed. Undisturbed Condition Sediment Size Distribution. An important factor influencing most sedi- mentation problems is the size distribution of the sediments. The typical descriptors of the size distribution of sediment are the median diameter and graduation coefficient of the material. Natural sediment distribution tends to be log-normal, which is a two parameter distribution. The median diameter of a distribution has 50 percent of the material sma I ler by weight and 50 percent of the material larger by weight. The second parameter, the grada- tion coefficient, gives the slope of the straight line resulting from plot- ting the distribution on log-probability paper. It is defined as a = 1 r^5o . vi 2 .^16 ^^50] (9) where a is the gradation coefficient and D is the particle diameter for which X percent of the material is finer. The gradation coefficient is related to the standard deviation of the material. The material can be described as uniform if its gradation is less than 1.3 or graded if its gradation Is greater than 1.5. 115 The median sedimenl" size in the floodplain generally decreases in the downstream direction along a river. Thus, the median size may be cobbles in the headwaters and fine gravel near the mouth. However, the median size can significantly vary around this general average within a sma I I area at a specified point along the river. This variation is a consequence of the variation in hydraulic forces from one point in the floodplain to an- other. Channel Erosion. Channel erosion in rivers is generally considered to be either local erosion (scour) or degradation. Both result from an increase in the sediment transport capacity, or a decrease in the sedi- ment load entering the area, or both. Local scour is most commonly a result of local increases in velocity due to flow obstructions or contractions. The increased velocity increases sediment transport capacity. Degradation can result if the channel bed is steepened in a short reach by, for example, a meander cutoff. The sedi- ment transport capacity would be increased through this reach causing ero- sion and a general upstream progression of the steepened slope (Figure 35). Upstream Progression of Steepened Slope Steepened Slope Figure 35. Schematic diagram showing degradation process. The progressive erosion continues upstream until equilibrium is reached. In theory, equilibrium is reached when the slope is equal to the slope prior to the occurrence of the cutoff, which would require the steepened slope to migrate to the headwaters. In practice, the steepened slope is reduced 116 during its upstream migration and gradually reaches an equilibrium con- dition. However, the degradation may extend over a long reach before equili- brium is ach i eved. Sediment Transport. Sediment transport is the movement of sediments past a specific cross section of a river. The sediment may be transported as suspended load or bed load. Suspended load is sediment that is trans- ported long distances suspended in the water column. Bed load is sediment that is transported by saltation (bouncing), or by rolling or sliding along the river bed. The sediment size distinction between bed load and suspended load varies with variations in discharge. At low flows, assuming the sedi- ments were available, silts and clays may be transported in suspension and sands and gravels transported as bed load. During floods, suspended load may include clays, silts, sands, and gravels, with cobbles and boulders transported as bed load. Often, the suspended load is assumed to include clays, silts, and sands and the bed load includes gravels, cobbles, and bou I ders. Changes Due to Gravel Removal Very little sediment data were collected at the study sites. Direct measurements or observations of bed or suspended transport were not made because site visits were scheduled during periods of low flow when the sites would be most workable. Because the sedimentation characteristics prior to gravel removal were also unknown, the upstream cross section was usually used as the undisturbed cross section. The effects of gravel removal were evaluated by comparing sedimentation features in the gravel removal area to those in the undisturbed upstream area. At six sites, a decrease in the median size of the surface layer, or armor layer, was observed in the mined area as compared with the undis- turbed area. Similarly, an increase or decrease was observed in the median diameter of the material underlying the armor layer at eight sites. In many cases it was difficult to evaluate whether the variation in median diameter was a result of the gravel removal operation or simply a result of the I 17 natural variation of the median diameter at a site. Degradation was also ob- served at a few sites although at other sites only causative evidence was available to indicate that this process can occur. Sediment transport changes were suggested at several sites where there were observations of bedforms in or downstream from the gravel removal area, observations of changes in the bed material size, computations of changes in shear stress, or observations of sediment sources which remained from the gravel removal operation. The effects of gravel removal activity on these sedimentation characteristics were evaluated and given quantitative ratings for com- parative purposes (Table 8). Sediment Size Distribution. The most common significant change in sediment size distribution resulting from gravel removal was a decrease in the size caused by fine material deposition in the material site. This change was reflected in the surface material at six sites and the subsurface material at six sites. Oregon Creek, Penny River, and Ugnuravik River had significant changes in both surface and subsurface material sizes. At Sinuk River, fine and medium sized gravels were nearly missing from the subsurface samples in the material site, causing an increase in the median size. The explanation for this is unknown. At Washington Creek, the subsurface ma- terial size was larger in the material site even though fine material depo- sition in the site reduced the median size of the armor layer. A pattern of correlation was not evident between increases or decreases in armor layer median diameter resulting from gravel removal and physical site or gravel removal area characteristics. One reason for this lack of correlation is that armor layer development is a complex function of several interrelated factors including degree of development of undisturbed armor layer, flooding history since gravel was removed, and flow characteristics in the gravel removal area. If the undisturbed size distribution of the armor layer was not significantly different from that of the material under- lying it, the relative change due to gravel removal would have been less and the time required for recovery to the undisturbed condition would also be less. The time for recovery is also a function of the floods during the recovery period; one large recurrence interval flood may be sufficient to 118 T3 0) -cro c -o 0) It! CL O (0 — :3 to o X3 X) ro Q) o CD — -D c o 0) -»- C (0 C -D (0 (D -1= 1- U CD 0) ■o 0) o — (D loro M- ._ 0) 1- L N D 0) — (/) -1- in JD (D D E l/^ ro 1- -1- (0 > O Q) (0 E V. 1_ 01 (0 o i_ 0) > oouD I roiTicoir^co-^ooiriinuDooioiOLnfAinK^ininhoyDiriiniriiriiniri cofo I (<~\irihO'<*rO'5tiriiriir\coO\iriiricoom'5tiriCTi'=tir\iriirimiriiriiri i£iiri I oirimintriiriir\ioiomirimiriir\inwcr!irivoiriir\iriiriooiriiri^o CTi ICD I— icNr^o— ii^CNi'^'d-iri'd-iriirimt^m iro i I I iTl ^■=t ICMhOOM ICNICNCNlO l-d-inl^VOI^I^LniTllTl— ITl IrO |«^ | |l^ < m < OD < m E ro Q) s_ 10 c s o Q I 1_ 0) > 13 q: "O D — c o — O CO Q) 0) i_ <_) c o c x: to (_> c o 0) 0) > q: C O) c E 0) o Q- Z 3 < > ID l_ 3 C CD 0) 0) I- o 1_ 0) ^ > Q) .- Q) CC la 0) CL Q> 3 J^ ^ — c j<: > ^ O) — (o > E (0 0) 1_ (/) Q. 3 I 0) > i_ — 0) cr > — -C E (O Q) V. U) Q. 3 E 1- ro 0) 0) > i_ — O -^ Q 3 O 1_ o — J^ I. i_ ^ W) ui — CO > O) (O — ■D 0) "O o O 1_ Ll. Q) > -D E 0) > (O c 10 > ^ o 0) — O) o u 1— u -1- 1_ U O 0) Ll Q. I/) ■<- O to i. 0) Q- S E (0 0> If) c s o Q -^ I 0) 1_ 0) 0) i. > o — (0 (O S c E (0 0) i_ U) D- 3 1 1_ ^ 0) 0) > 0) — u ct o (0 c C CO CO — C Q) CO ^ I- CL 0) cn (O CL O) c 5 o c o c/) OJ o c o o I 19 Table 8. Footnotes D _ Wedian size in the gravel removal area R ~ median size upstream from ttie gravel removal area 10 10 <_ 0- (due to gravel removal activity) 9 2 < D < 10 (due to gravel removal activity) 8 1.2 < 0 < 2 (due to gravel removal activity) 7 1.2 £ D (cause uncertain! 6 I < Op < 1.2 5 D^: I 4 0.8 < Dp £ I 3 D„ £ 0.8 (cause uncertain) 2 0.5 < 0 < 0.8 (due to gravel removal activity) I 0.2 < D < 0.5 (due to gravel removal activity) 0 D < 0.2 (due to gravel removal activity Channel degradation ratings: 10 Very substantial degradation upstream of ttie disturbed area 9 Substantial degradation upstream of the disturbed area 8 Large amount of degradation upstream of the disturbed area 7 A noticeable amount of degradation upstream of the disturbed area, but not unl ilie degradation which could occur naturally 6 Slight degradation upstream of disturbed area observed or implied; may not be a result of gravel removal 5 No degradation, observed or implied by the data 0-4 Not used Bed load ratings: 10 Substantial increase in bed load by erosion In the gravel removal area 9 Large increase in bed load oy erosion in the gravel removal area e Increase in bed load by erosion in the gravel removal area 7 Bed load increase due to gravel removal activity expected but not verified by direct evidence 6 Slight bed load increase potentially due to gravel rerroval activity 5 No bed load change evident 4 Slight bed load decrease by deposition in the gravel removal area 3 Moderate bed load decrease by deposition in the gravel retroval area 0-2 Not used Suspended load ratings: 9-10 Not Lsed 8 Large temporary and/or moderate long term increase in suspended load 7 Temporary increase in suspended load as a result of disturbance of armor coat 6 Potential slight increase In suspended lead resulting from gravel r emova I act I v i ty 5 No apparent change in suspenreO iced 4 Potential slight decrease In suspended lead resulting fronr deposition 3 tfoderate amount of deposition of suspended ir.atcrlal C-2 Not used 120 develop an armor layer comparable to that in ttie undisturbed area. Ttie development of an armor layer in the gravel removal area is also greatly dependent on the location of the area relative to the active channel and the resulting flow characteristics through the site. The location and extent of gravel removal may be such that an armor layer may not develop until the area fills in sufficiently to have appropriate hydraulic character- istics for armor layer development. Channel Erosion. Channel erosion in the form of local scour was not observed at any of the study sites. The potential exists for local scour to develop as a result of flow obstructions in the form of material stockpiles, overburden piles, and diversion dikes. This potential was discussed in the previous section discussing hydraulics. Channel degradation was observed at four sites and may have been devel- oping at three other sites. At Washington and McManus Creeks, obvious degra- dation had occurred upstream from the site in the main channel. At the two Dietrich River sites, degradation was occurring in high-water channels; at the downstream site, one of the high-water channels developed into an active side channel after work completion. Channel degradation resulting from gravel removal activity has been documented elsewhere (Woodward-Clyde Consultants 1976b, Li and Simons 1979). Li and Simons (1979) suggest that the installation of check dams can restrict upstream degradation. Sheridan (1976) discusses in-channel gravel removal, noting that the pits filled in with sediment; a similar situation occurred on Sinuk River with no apparent degr adat i on . Sediment Transport. Changes in sediment transport due to gravel removal were difficult to evaluate. The ratings given in Table 8 are thus highly subjective. A few possible changes which were suggested by the sedimentary features in and around the material sites are discussed below. It is likely that most scraped sites exhibited an increase in suspended load during the first flood event and possibly during one or two subsequent events as the material in the gravel removal area was washed clean of the fine grain sizes. This increase was thus likely a temporary increase common at most 121 scraped sites. Long-term increases in suspended load were implied at sites with disturbed areas which contributed fine materials to the flow. Examples of such long-term increases were the access road degradation at Ugnuravik River (Figure 36), the diversion dam at Skeetercake Creek (Figure 37), and several sites with overburden piles or berms containing fine-grained ma- terials. Similar increases in suspended load could occur from accelerated bank erosion at the site. Deposition of fine-grained sediments in several of the gravel removal areas was also observed. Sites with changes in suspended load showed no pattern with the physical site or gravel removal area charac- teristics. Apparent changes in bed load were observed at some sites in the form of gravel dunes or loose gravel deposits in and downstream from the gravel removal area. When these deposits occur in the gravel removal area, they could indicate the inability of the flow through the area to carry the sediment load delivered to it or generated within it. Deposition occurring downstream from the gravel removal area would imply that the flow through the area is sufficient to erode the loose gravel from the gravel removal area. It is possible that when these gravels reach the main channel they are transported in the form of another bed form or possibly in suspension. Bed load changes occurred most often at scraped sites in active and high-water channels, and in locations immediately adjacent to such channels. ICE CHARACTERISTICS Undisturbed Condition Ice jamming can occur during breakup when ice floes moving down the river are blocked, thereby blocking subsequent ice floes and eventually creating a surface dam to the flow of ice. Ice jams can cause scour due to increased velocity beneath the ice dam; they can also cause the water level to rise, resulting in increased flooding. Ice jams are normally caused by a constriction in the channel width or depth, a reduction in flow velocity, or manmade structures in the floodplain. 122 Figure 36. Upstream view of thermal and fluvial erosion in the access road at Ugnuravik River, acting as a long-term sediment source to the river. ■^i *^«>-..- iL Figure 37. View of erosion of a diversion dam which acts as a long-term sediment source to Skeetercake Creek. Dunes in foreground are atypical of the undisturbed river. 123 Aufeis is defined as areas of ice which have developed by a sequence of events of overflowing water on top of the previous ice surface. The general mechanism for the growth of aufeis involves an increase in the hydrostatic pressure due to a reduced flow area; when the pressure exceeds the elevation of the ice surface, overflow onto this surface results and subsequently freezes. The overflow causes the pressure to decrease and ice surface ele- vation to increase. This sequence continues to repeat until the source water cannot produce sufficient pressure to exceed the elevation of the ice sur- face. Three requirements for the formation of aufeis are given by Carey (1973); (I) significant ground water or under-ice flow, (2) growth of ice to the channel bed or near the bed, and (3) subsurface constriction such as bedrock, less pervious soil, or permafrost. Changes Due to Gravel Removal An organized program of winter and spring observations of aufeis and breakup were not included in this study. Therefore, much of the following discussion is based on observations of auf e is and ice jamming potential, rather than of actual aufeis and ice jams. However, at two sites, Washington Creek (Figure 38) and Oregon Creek, large areas of aufe is were observed in early June. Incidental winter observations at a few other sites documented the existence of aufeis. Ice jams could be caused by several aspects of floodplain gravel re- moval. In rivers which are increased in width and depth by the gravel re- moval, such as by in-channel mining, the velocity would decrease causing the ice floes to gather. At the downstream end of the gravel removal area these floes could jam where the channels constrict back to the natural width. This ice jam could cause flooding in and upstream from the gravel removal area and possible bed scour beneath the ice jam. River channels which are widened causing shallower depths, such as by removing bars adjacent to the channel, could cause ice jamming by grounding the ice floes. Another potential mechan- ism for ice jam formation resulting from a gravel removal operation is the blocking of ice floes by flow obstructions in the form of overburden piles, stockpiles, or dikes. 124 Figure 58. Large area of aufei s at the upper gravel removal area at Washington Creek as it appeared in early June. In evaluating the potential for aufeis development at each of the study sites, it was assumed that wide, shallow channels were more likely to develop aufeis than narrow, deep channels. This assumption is probably valid because shallow channels are more likely to freeze to their bed and to have a shallow talik (unfrozen zone) than deep channels carrying equivalent flow. The results of this evaluation of aufeis potential are listed in Table 9, along with the identification of those rivers with aufe is activity or po- tential aufeis activity prior to the gravel removal operation. Most of the observations of increases or potential increases in aufe i s activity were associated with mining activities in straight and sinuous rivers, although some activities in braided, split, and meandering rivers also caused potential increases. Increases in aufeis activity were associ- ated with scraping operations. Increased aufeis activity or potential aufeis activity often occurred at those sites where the gravel removal operation was located in active or high-water channels and in locations immediately 125 Table 9. Quantification Ratings of Change in Aufe i s Potential that Resulted from the Gravel Removal Operation at Each of the 25 Sites Gravel removal Au f e i s River area potential Gold Run Creek 6 Sinuk River A 5 B 6 Washington Creek A 10 B 6 Oregon Creek A 10 B 5 Penny River 6 Nome R i ver 6 Ugnuravik River 6jj Aufeis Creek A 6. Kuparuk River ° Skeetercake Creek 5 Sagavan irktok River 6^ Ivishak River 6 Shaviovik River 5^^ Kavik River 6jj Dietrich Ri ver-Upstream 7 Dietrich River-Downstream 5 Middle Fork Koyukuk Ri ver-Upstream A 5 Middle Fork Koyukuk River-Downstream 8 J im Ri ver ^^ Prospect Creek 5 West Fork Tolovana River 5 McManus Creek 6 Tanana River-Downstream 5 Tanana Ri ver-Upstream 5 Phelan Creek 5 Aufe is potential ratings: 10 Large aufeis development observed in the disturbed area where no aufeis was previously recorded 9 Moderate sized aufeis development observed in the disturbed area where no aufeis was previously recorded 8 SmaTT aufe is development observed or a strong potential for aufe is occur- rence is inferred 7 Relocation of an existing aufe i s area by gravel removal activity 6 Potential increase in aufeis activity resulting from gravel removal act i v i t y 5 No change in aufe is characteristics 0-4 Not used ''Rivers with a high potential tor icing activity prior to the gravel removal operat ion. 126 adjacent to the channels. Such locations, when excavated for gravel, tend to increase channel width, decrease depth, and allow for freezing down to the channel bed. As noted earlier, large areas of aufeis were observed in the Washington Creek and Oregon Creek study sites. Both of these sites had been extensively scraped and that caused numerous channels to form and loss of surface flow to intergravel flow because of loosely compacted gravels. The aufeis may be retarding the recovery of the surface flow by protecting the loose gravels from the flood flows during the snowmelt runoff period. At both sites, the channels flowing during the survey were not flowing where the channel had previously been; it is thus likely that the talik was not as deep beneath the newly formed channels, thereby providing the auf e i s requirement of a subsurface constriction. The shallow channels would likely freeze to the bed, thereby satisfying another requirement for aufeis formation. The third requirement, a water source, was already available. Thus, at these two sites the gravel removal operation changed the channel location and cross section sufficiently to provide two of the three requirements for aufeis format i on . HYDROLOGY Hydrology is the study of the origin, distribution, and properties of water during the time it is at or near the earth's surface. Of concern in this section is the distribution of the water. More specif ical ly, this section discusses briefly the quantity of water that can be expected at the 25 material sites during low flow and flood flow conditions and poten- tial effects on the quantity due to the removal of gravel. Undisturbed Condition The mean annual flow of a river at a specific point is, as the term implies, the mean flow during any 12 month period. It is an indication of total annual runoff and may also be used as an approximation of the typical low summer flow. Estimates of mean annual flow for the 25 study 127 sites are listed in Table 10. They range from 0.09 m /s at McManus Creek to 540 m /s at Tanana River-Downstream. Flood frequency curves show the expected frequency of occurrence of different magnitude floods at a specific point on a river. The frequency of occurrence is commonly referred to by the recurrence interval of the flood, which is the average number of years between floods of that magnitude. The reciprocal of the recurrence interval is the probability of occurrence of a given magnitude flood in any year. Flood frequency curves were developed for each of the study sites. Discharge values corresponding to selected fre- quencies of occurrence are shown in Table II. Changes Due to Gravel Mining Hydrologic characteristics are, to a large extent, governed by basin- wide parameters such as climate and geology. Gravel removal operations did not have a significant effect on these characteristics. However, local changes in the ratio between surface flow and subsurface flow occurred at several sites. The local changes were not measured; quantitative ratings shown in Table 12 were assigned based on a subjective evaluation. A local reduction in mean annual flow occurred at the upper Washington Creek and upper Aufeis Creek sites as a result of a loss of surface flow to inter- gravel flow. At Washington Creek, the flow entered the gravel removal area and spread out through loose, uncompacted gravel; a large percentage reduc- tion in surface flow resulted at low flows. This intergravel flow component was sti I I evident in the site 13 years after the site was worked. The rela- tive effect of the loss of surface flow during flood events was likely minimal. At Aufeis Creek, surface flow appeared to cease entirely for a period of 2 years, although continuous surveillance was not available to verify this. Thus, the mean annual flow of Aufeis Creek in this local region was reduced to near zero for 2 years. The effect on flood flows was unknown. Two other sites, the upper Oregon Creek and Penny River sites, had a potential for a similar, but not as extensive, decrease of surface flow lost to intergravel flow. No observations or measurements were available 128 Table 10. Mean Annual Flow Estimates at Each of the 25 Study Sites River Gold Run Creek S i nuk R i ver Washington Creek Oregon Creek Penny River Nome River Ugnuravik River Aufeis Creek Kuparuk River Skeetercake Creek Sagavanirktok River I V i shak R i ver Shavi ov i k R i ver Kavik River Dietrich Ri ver-Upstream Dietrich River-Downstream Middle Fork Koyukuk Ri ver-Upstream Middle Fork Koyukuk River-Downstream J im R i ver Prospect Creek West Fork Tolovana River McManus Creek Tanana River-Downstream Tanana Ri ver-Upstream Phelan Creek Unit mean annua 1 f 1 ow 3 2 (m /s/km ) Mean annual (m^/s) f low 0.013 0.9 0,033 18.0 0.018 0.5 0.023 0.7 0.023 1 .4 0.033 4.3 0.0023 0.6 0.0044 1 . 1 0.0045 38 0.0035 0.3 0.0083 39 0.0066 24 0.0040 1 .6 0.0062 5.5 0.006 3, 1 0.006 4.0 0.0054 13 0.0054 22 0.010 7. 1 0.010 2.6 0.0062 4.7 0.0062 0.09 0.012 559 0.012 468 0.063 5.2 129 Table II. Calculated Discharges in m /s Corresponding to Selected Recurrence Intervals for Each of the 25 Study Sites Recurrence interval ( years ) River 1 .25 2 5 10 25 50 100 Gold Run Creek 1 1 .2 19.2 32. 1 42.8 53.6 70.2 91 .0 S i nuk R i ver 1 13 171 256 323 391 481 589 Washington Creek 2.58 5.63 10.7 16.6 28. 1 39.5 54.9 Oregon Creek 6.21 1 1 . 1 19.4 26.3 33.5 44.8 59.3 Penny River 18.2 23.7 31.7 37.0 43.7 50.2 57.0 Nome River 32.4 53.3 86.3 1 14 142 182 232 Ugnuravik River 31.4 46. 1 71 .5 92. 1 121 149 180 Auf ei s Creek 39.2 56.8 89.3 1 16 160 196 235 Kuparuk River 905 1355 2165 2848 3906 4840 5912 Skeetercake Creek 10.6 16.7 28.4 38.4 54.6 69.8 87.0 Sagavan irktok River 376 462 592 665 785 863 970 1 V i shak R i ver 267 333 432 489 579 641 726 Shaviovik River 35.8 59.6 98. 1 130 164 212 272 Kavik River 108 171 27 1 353 444 559 701 Dietrich River-Upsti -eam 35.6 58.6 102 140 195 253 322 Dietrich River-Downstream 46.9 75.9 131 178 247 318 402 Middle Fork Koyukuk R- -US 126 189 302 396 534 661 808 Middle Fork Koyukuk R- -DS 190 276 428 552 736 896 1079 J im Ri ver 101 125 156 178 204 228 25 1 Prospect Creek 33.3 43.6 57.6 67.3 78.5 90.4 102 West Fork Tolovana R i ver 63.9 89.2 130 159 203 242 282 McManus Creek 1 .65 3.32 7.48 12.0 20.6 29.8 42. 1 Tanana River-Downst ream 1562 1752 1992 2120 2356 2460 26 19 Tanana R i ver-Upstream 1341 1518 1738 1857 2069 2169 23 18 Phelan Creek 49.3 65.3 92.8 1 14 146 171 197 130 Table 12. Quantification Ratings of Change in Quantity of Intergravel Flow Resulting from the Gravel Removal Operation at Each of the 25 Sites River Gravel removal area Intergravel flow Go I d Run Creek 5 S i nuk R i ver A 5 B 5 Washington Creek A 9 B 5 Oregon Creek A 7 B 5 Penny River 7 Nome River 5 Ugnuravik River 5 Aufeis Creek A 10 B 5 Kuparuk River 5 Skeetercake Creek 5 Sagavan ir k tok River 5 I V i shak R i ver 5 Shaviovik River 5 Kavik River 5 Dietrich R i ver-Upstream 3 Dietrich River-Downstream 5 Middle Fork Koyukuk R i ver-Upstream A 5 B 5 Middle Fork Koyukuk River-Downstream 5 Jim R i ver 5 Prospect Creek 5 West Fork Tolovana River 5 McManus Creek 5 Tanana River-Downstream 4 Tanana R i ver-Upstream 4 Phe I an Creek 5 a ntergravel flow ratings: 10 All surface flow converted to intergravel flow for one summer or more 9 Substantial long-term loss of surface flow to intergravel flow 8 Moderate long-term loss of surface flow to intergravel flow 7 Implied long-term loss of surface flow to intergravel flow 6 Small quantities of surface flow lost to intergravel flow 5 No apparent change 4 Implied increase of surface flow and decrease of intergravel flow 3 Known increase of surface flow and decrease of intergravel flow 0-2 Not used 131 to estimate the magnitude of the decrease. The location of the gravel re- moval area may provide an explanation for the significant intergravel flow at Washington Creek and Aufeis Creek. At these two sites the scraping occur- red near the downstream end of a sharp meander bend (Figure 39). It appeared that the scraping in this location caused most of the flow to leave the confinement of the channel. The lack of a well defined channel caused the flow to spread over the gravels in the material site and deposit the sedi- ment load that it was carrying. These deposits were quite loose and un- stable, and thus were very conducive to intergravel flow. Other sites having a similar specific location of scraping were slightly different in configur- ation from that shown in Figure 39; either the bend upstream from the scraped area at these sites was not as sharp or the scraping occurred fur- ther downstream on the bend, thus allowing some of the flow and likely much of the bed load to be retained in the original channel. Three possible explanations for the continued loss of surface flow at Washington Creek are (I) that the suspended load is not sufficient to fill the openings in the gravel, (2) the presence of aufeis in the site protects the gravels from the significant snowmelt floods, and 13) water freezes in the gravel, expanding and separating the gravels in the process. Pit sites, such as Dietrich Ri ver-Upstream and the two Tanana River sites, had a potential to locally increase the mean annual flow as a result of intercepting intergravel flow and allowing it to surface at the pit. However, the percentage increase in the mean annual flow at these sites is probably quite small. 132 Sr«!*'Active Channel Prior ; to Gravel Removal High-Water Channel Prior to Gravei Removal Scale in Molers 17Junlf73 7 July 1977 Figure 39. Aerial photographs of Washington Creek (top) and Aufeis Creek (bottom) showing material site locations and approximate channel locations before the disturbance. t35 SUMMARY AND CONCLUSIONS Various physical characteristics of arctic and subarctic rivers were affected by gravel removal operations. These characteristics were divided into five categories: 1. Channel configuration and process, 2. Hydrau I ics , 3. Sedi mentat i on , 4. Ice characteristics, and 5. Hydrology. One or more characteristics from these categories were observed to have changed as a result of removing gravel from the 25 floodplain study sites. CHANNEL CONFIGURATION AND PROCESS Channel configuration and process characteristics that changed as a result of gravel removal operations included braiding characteristics, such as increase in the number of channels and decrease in lateral stabil- ity of the channels, and the potential for diversion of flow through the gravel removal area. The greatest changes in braiding characteristics occurred at 10 study sites and resulted from gravel removal operations that disturbed the bars adjacent to active channels or that diverted flow through the material site. Flow diversion through the mined site resulted from having insufficient buffers or no buffers at all. Gravel removal operations caused flow diversion or a high potential for flow diversion at 12 of the 25 study sites. 134 HYDRAUL ICS Hydraulic characteristics exhibiting changes as a result of gravel removal operations included the hydraulic geometry (including width, depth, velocity, and conveyance), overall channel slope, local slope redistri- bution, flow obstructions, and area of ponded water. Increases in channel width, conveyance, overall slope, flow obstructions, and ponded water were typical responses to gravel removal, as were decreases in channel depth and velocity. One or more of these effects from gravel removal were observed at al I of the sites except those pit excavated sites that were separated from the active channels by a buffer. Sma I I river systems typ- ical ly had sma I ler f loodplains which forced the gravel removal operation closer to active or high-water channels, causing hydraulic changes. SEDIMENTATION Sedimentation characteristics which appeared to have changed as a result of gravel removal operations included armor layer and subsurface material site distributions, channel degradation, and suspended and bed loads. The most common significant change in sediment size distribution resulting from gravel removal was a decrease in the size caused by fine material deposition in the material site. This change was reflected in the surface material at six sites and the subsurface material at six sites, three of which were different from those with surface material changes. Channel degradation was observed at four sites and may have been develop- ing at three other sites. Changes in sediment transport due to gravel removal apparently took the form of increases as we I I as decreases, with apparent changes occurring at II sites. Most changes in the sediment char- acteristics resulting from gravel removal operations occurred at scraped sites in or immediately adjacent to active and high-water channels and at those sites where fine sediment sources were left in the floodplain near the channe I . 135 ICE CHARACTERISTICS Two ice characteristics were identified as potentially being increased as a result of gravel removal activity. They are ice jamming and auf e i s formation. These can be affected by a widening of the channel followed by a rapid reduction in width, a reduction in depth, obstructions in the floodplain, and relocating the channel through an area which was previously dry. Aufeis formation was observed at four study sites. HYDROLOGY The only characteristic related to the hydrology of the river which was identified as potentially changing as a result of gravel removal opera- tions was a change from surface flow to groundwater flow or vice versa. This change, although relatively minor at most sites, can have a local effect on the mean annual flow, flow duration curve, and potentially, on the flood frequency curve. Significant reduction of surface flow occur- red at two study sites. RECOMMENDATIONS Listed below are several recommendations concerning gravel removal operations, the purpose of which is to reduce the number or magnitude of changes to the physical characteristics of rivers: 1. Small rivers should not be considered as gravel sources. 2. Braided rivers should be considered as primary gravel sources; other river configurations, listed in order or likelihood of caus- ing the least physical change, are split, meandering, sinuous, and s tr ai gh t . 3. Pit excavations should be located on terraces or possibly inactive floodplains and should be separated from the active floodplain by a buffer designed to maintain this separation for two or more decades . 136 4. Material sites within the active f I oodp I a i n should: • Not disturb the edge of the active channel (s); • Maintain a high-water channel shape, within the material site, similar to that which enters and leaves the site; • Not increase the bed slope of active or high-water channels local ly to more than that of natural ly occurring slopes; • Form new high-water channels through the site if flow is expected through the site; • Be shaped and contoured to provide proper drainage; • Have material stockpiles, overburden piles, and dikes removed from near active channels unless they have a specific purpose for being there and are designed to withstand the hydraulic forces; and • Be protected from low flow channels until the occurrence of the first flood after the site is completed. 157 REFERENCES Carey, K. L. 1973. Icings Developed from Surface Water and Groundwater, U.S. Army Cold Regions Research and Engineering Laboratory. Mono- graph I I I-D3. 71 pp. Chow, V. T. 1959. Open-Channel Hydraulics. McGraw-Hill Book Company, New York, 680 pp. Doyle, P. P., and J. M. Childers. 1976. Channel Erosion Surveys Along TAPS Route, Alaska, 1976. U.S. Geological Survey Open-Pile Report. 89 pp. ;ams Emmett, W. W. 1972. The Hydraulic Geometry of Some Alaskan Strec South of the Yukon River. U.S. Geological Survey Open-File Report. Anchorage. July. 102 pp. Emmett, W. W. 1975. The Channels and Waters of the Upper Salmon River Area, Idaho. U.S. Geological Survey Professional Paper B70-A. 116 pp. Karaki, S., K. Mahmood, E. V. Richardson, D. B. Simons, and M. A. Stevens. 1974. Highways in the River Environment - Hydraulic and Environmental Design Considerations. Prepared for Federal Highway Administration by Civil Engineering Department. Colorado State University. Fort Collins, Colorado. 453 pp. Lamke, R. D. 1979. Flood Characteristics of Alaskan Streams. U.S. Geo- logical Survey Water Resources Investigations 78-129. 61 pp. Leopold, L. B., M. G. Wolman, and J. P. Miller. 1964. Fluvial Processes in Geomorpho logy . W. H. Freeman and Company, San Francisco, Cali- fornia. 522 pp. Li, R. M. , and D. B. Simons. 1979. Mathematical modeling of erosion and sedimentation associated with instream gravel mining, pp. 420-429. In Conservation and Utilization of Water and Energy Resources. ASCE Hydraulics/ Energy Division Conference. San Francisco. 8-11 August. Rundquist, L. A. 1975. A Classification and Analysis of Natural Rivers. Dissertation. Colorado State University. Fort Collins, Colorado. 377 pp. Sheridan, W. L. 1976. Effects of Gravel Removal on a Salmon Spawning Stream. U.S. Department of Agriculture. Forest Service. 26 pp. 138 I Woodward-Clyde Consultants. 1976a. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-arctic Streams in Alaska. U.S. Fisti and Wildlife Service. FWS/OBS 76/21. Wasti. D. C. 127 pp. Woodward-Clyde Consultants. 1976b. Aggregate Extraction Management Study, County of Yolo California. Prepared for the County of Yolo. Planning Department. Aggregate Resources Management Committee. 128 pp. 139 EFFECTS OF GRAVEL REMOVAL ON AQUATIC BIOTA L. L. Moulton INTRODUCTION Populations of organisms are conlTolled by physical and chemical fac- tors, often termed their environment, and by biological factors, including predation and competition. Environmental constraints on a particular species determine the usable habitat available to that population and the size of the population is often restricted by the amount of usable habitat. After the maximum number of individuals a particular habitat can support (termed the carrying capacity) has been reached, the population cannot increase with- out an increase in usable habitat. Predation and competition can act on a population to limit numbers below the carrying capacity, thus undisturbed populations are not necessarily fully utilizing the available habitat. Alterations to the habitat can alter the quality of the habitat, lead- ing to direct changes in the carrying capacity, and consequently, to reduc- tions in the affected populations. Decreases in habitat diversity may reduce the carrying capacity for one species while leaving that for another un- changed. If the two species were in competition, the reduction or removal of one may al low the other species to increase. General ly, decreases in habitat diversity will result in an increased carrying capacity of one species which is able to efficiently utilize the more uniform habitat. Conversely, in- creases in habitat diversity general ly cause increases in the number of species or life history stages present as new habitat types are added. These species increases are often accompanied by decreases in the populations which had formerly been utilizing the more uniform, less diverse, habitat. 141 The decreases may be due either to less available habitat or to competition from species which more efficiently utilize the newly created habitats. The types of habitats present in a river are determined by the loca- tion, size, configuration, and water quality characteristics of the river. Features which define specific habitats include depth, velocity, substrate, and cover. Alterations to a river which affect any of these features will also affect the habitat available in the river and may impact habitats downstream from the alterations. Habitat alterations may affect the quality or diversity of the habitat, or both. Reduced habitat quality makes the area less desirable to the species present prior to alteration, while altered habitat diversity may favor one species or life history stage over another. Reduced habitat quality implies alteration of a single habitat type whereas reduced habitat diversity implies reduction in the number of available habitats but the two responses are not independent. Several types of habitats may be used in the life cycle or even sea- sonal cycle of an organism, and there is often a critical habitat which controls the size of the population. In the arctic and subarctic environ- ment, the critical habitat for fish populations is often the amount of overwintering habitat. Other critical habitats often controlling fish popu- lations are spawning and rearing areas. Critical habitats vary from stream to stream and species to species depending on the characteristics of the streams and the life cycle requirements of the species. Recent studies have been aimed at quantifying the effects of habitat alteration on stream populations (Stalnaker and Arnette 1976, Bovee and Cochnauer 1977, Binns and Eiserman 1979). Two of the basic requirements of these efforts are detailed measurements of appropriate habitat parameters and an intimate knowledge of the habitat requirements of the species in ques- tion. The emphasis of the present study was on a multiple-disciplinary survey of the effects of floodplain gravel removal on a broad geographical scale. Because of the limited data on many species and complete lack of data on many of the river systems studied, a detailed habitat analysis was not possible. The 3 to 4 day surveys at each site allowed for gathering of basic 142 physical and biological data but not the type of detail required for sophis- ticated correlation analysis. For these reasons the present analysis was confined to analysis of trends and subjective evaluations of habitat alter- ations and their effects on aquatic organisms. The material sites were visited 2 to 20 years after mining was com- plete, thus the immediate effects of grave> removal operations were not studied. The changes evaluated during the present study were those which persist over a number of years rather than those affecting the biota during the year of disturbance. A literature review of impacts at the time of actual gravel removal was presented by Woodward-Clyde Consultants (1976). 143 METHODS OF DATA COLLECTION As detailed in APPROACH AND METHODOLOGY a variety of standard sampling methods were utilized at each study site with the specific methods used dependent on the type of river system and habitat being studied. 144 METHODS OF DATA ANALYSIS The data from each of the 25 sites were first analyzed on a site- by-site basis to determine the effect of gravel removal operations on the aquatic environment at each study site. These individual site evaluations provided the basis for further analysis to identify trends and correlations relating to major site variables (Table I, Major Variable Matrix). These individual site evaluations are not included because of space limitations but are part of the permanent data base maintained by the U. S. Fish and Wi I d I i f e Service. The various physical and biological parameters measured at the dif- ferent sites varied greatly in magnitude and the variation made the direct comparison of data among sites impractical. The various parameters recorded at the study sites were standardized on a scale of 0 to 10 to obtain a relative measure of the degree of change. A rating of 5 indicates that a parameter measured in the mined area had not changed from the same parameter in the upstream area; ratings of 0-4 and 6-10 indicate decreased and in- creased parameter values in the mined area relative to the upstream area. The rating was determined by calculating the percentage change in the mined area relative to the upstream area for each site and subjectively assigning rating values to various percentage intervals such that all or most of the 0-10 scale was utilized for those sites at which the parameter was evalu- ated. Data from study sites with similar ratings were examined for similar alterations that might lead to a similar parameter response. The analysis of habitat alteration was based on field notes from the site surveys, ground and aerial photographs, direct measurement of habitat parameters, results of hydraulic analysis, and visual observations. Habitat parameters considered in the analysis included changes in substrate type. 145 substrate porosity, configuration of adjoining banks, bank and instream cover, number of channels, pool-riffle frequency, depth, velocity, and wetted perimeters at different flow levels. Additional habitat alterations were noted where appropriate, such as excessive siltation, auf e i s formation and creation of new aquatic habitats. Much of the analysis was subjective because many habitat parameters were difficult to quantify, consequently, the analysis was kept conservative. The results of hydraulic analysis, as described in the EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAULICS, allowed for a certain amount of habitat parameter quantification and these results supported the subjective evaluations whenever comparisons were available, indicating that subjectivity was not a major source of error. Analysis of changes in fish populations was accomplished by evalu- ating the types of habitat alterations occurring in the mined area relative to the upstream area. Then the measured parameters that appeared to be most important at the particular site were examined to determine if there had been a change in fish distribution, as indicated by a difference in catch rate between the upstream and mined areas. In this manner the combi- nations of habitat alteration could be evaluated for their cumulative effect on the population of fish present during the site visit. Additional effects were postulated based on known life history requirements of the various spec i es. The large number of benthic sample replicates obtained at each study site during the field surveys allowed for an analysis of variance to de- termine if significant differences existed in the densities among sample areas within a study reach. All Surber sample data were computer coded and the densities were subjected to an analysis of variance and multiple classi- fication analysis (Nie et al. 1975). A nonpar ameter i c procedure, the Mann- Whitney U-test (Zar 1974), was also used to evaluate differences in density. The results of the two tests were compared and, where the results of the two tests differed, the more conservative nonparametr i c test was used. Addi- tional computer analysis included the calculation of various indices of diversity and similarity, such as the Bray-Curtis and Raabe similarity 146 indices, and Shannon-Weaver and Simpson density indices. The indices respond differently to changes in density and diversity and were used primarily to search for changes in the aquatic macro i nvertebrate assemblages vulnerable to Surber samplers. Because the level of identification was to the generic level at best and often only to family or order, the indices were applicable only to the present study. Comparison with results of other studies and extensive anal- ysis of the data are not justified. Often multiple species within a genus were recognizable but the absence of suitable taxonomic aids for arctic aquatic macro i nvertebrates inhibited identification. A list of collected taxonomic groups by phyletic classification, with associated common names, is included in Appendix A. 147 RESULTS AND DISCUSSION MAJOR GRAVEL REMOVAL HABITAT ALTERATIONS Habi tat Qual i ty Alterations of habitat quality observed at many of the sites consisted primarily of substrate alteration and removal of both instream and bank cover. Siltation, commonly associated with instream disturbances, was ob- served at a few sites, but was not a major factor because most of the sites were visited several years after mining had been completed. At three sites where siltation was observed it was caused by eroding berms (Kavik River) or melting aufeis fields (Washington Creek, Oregon Creek) (Figure 40). Two types of substrate alteration were observed: (I) a shift from a moderately compacted gravel substrate to a very loose, unconsolidated sand- gravel substrate, usually with considerable intergravel flow and (2) a shift from a smooth, paved substrate which produced near laminar flow to a more porous, irregular substrate producing turbulent flow. Most of the substrate alterations recorded were Type I alterations with only two Type 2 alterations observed. Type I alterations occurred at four of the eight sites where scraping was conducted in an active channel (Washington Creek, Oregon Creek, Penny River, McManus Creek) and at four where flow subsequently increased or diverted to inundate a scraped area (Sinuk River, Kuparuk River, Sagavan irktok River, Ivishak River) (Table 13). The effects of this type of alteration appear to be long-term, because this alteration was noticeable at McManus Creek 16 years after mining. The effect on the sub- strate was caused by removal of the armor layer, loosening of the gravels, and subsequent washing out of fine materials. Formation of ice in the mined areas appeared to prolong the recovery time of this type of alteration. 148 a) Sediment being released by melting aufeis. b) Silt deposited in substrate downstream from aufeis field. Figure 40. Siltation resulting from extensive aufeis field at Oregon Creek mined study area, 20 June 1977. 149 Table 13. Major Habitat Alterations Observed at Sites Mined by Scraping (5 = No Change, 6-10 = Trend Towards Parameter, 0-4 = Trend Away From Parameter ) Q) CL c o (0 (U Site ID o u c QQ XJ 0) o D ■o 0) > o o E 0) c •D T3 0) 0) (0 It) ID ro 0) Q) 1_ I- o o c c — i- cr> Q) c 4- .— la ■D ? .— ^ ID o i- 10 m QQ T3 0) I/) la 0) o c S ■o 0) •D c o Seward Peninsula Gold Run Creek S i nuk Ri ver Washington Creek Oregon Creek Penny River Nome River a 5 5 8 5 6 5 7 5 8 - 6 9 8 10 5 10 10 10 10 10 10 5 10 10 9 9 8 9 5 10 8 9 10 10 — 5 5 - 10 10 7 North Slope Ugnuravik River Aufeis Creek Kuparuk River Skeetercake Creek Sagavan ir ktok River I V i shak R i ver Shav i ov i k R i ver Kavik River 9 6 5 8 5 5 5 5 5 3 5 5 5 5 9 5 5 5 5 5 8 5 8 9 5 5 9 6 10 8 9 10 10 8 7 6 5 8 10 0 10 6 8 6 5 6 6 5 8 7 5 Northern Interior Dietrich Ri ver-Upstream Dietrich River-Downstream M.F. Koyukuk River-US M.F. Koyukuk River-DS 5 5 5 5 5 5 5 5 5 5 5 6 7 7 5 5 10 5 10 9 9 5 5 5 5 8 10 7 Southern I nter i or McManus Creek Phelan Creek 7 5 5 4- 10 2 Dash means parameter not evaluated at this site. 150 Type 2 substrate alterations were documented at two locations, both on medium size North Slope rivers (Table 15). In one case, Ugnuravik River, the upstream area showed near laminar flow that was changed to turbulent flow while in the other case, Shaviovik River, the reverse occurred - the upstream flow was turbulent whereas the flow through the mined area was laminar. Such changes would be expected naturally where localized substrate or slope differences alter flow characteristics. Bank cover is provided by structures on or features of the stream bank that provide shelter from surface predation and reduce visibility. Ex- amples of bank cover include overhanging vegetation and incised or undercut banks, thus bank cover was eliminated when mining removed these features (Figures 41 and 42). These types of bank cover were typically present in straight, sinuous, meandering or split channel rivers, but were less common in braided rivers. Significant bank cover loss was observed at 6 of the 21 scraped sites, Sinuk River, Washington Creek, Oregon Creek, and Penny River sites on the Seward Peninsula, at the Skeetercake Creek site on the North Slope, and at the Middle Fork Koyukuk R i ver-Upstream in the Northern In- terior (Table 13). Instream cover is created by obstructions, such as boulders or logs, that provide slack water where fish can hold position with minimal energy expenditure and reduce predation from above by being less visible. Water depth can also function as cover, because deep pools and runs offer more overhead protection and often lower velocities than shal low riffles. Certain species, such as Arctic char and Arctic grayling, are often associated with instream cover. Instream cover was reduced at five sites, Washington Creek, Oregon Creek, Penny River, Kavik River, and Sagavan ir k tok River, as a result of directly removing boulders and large cobbles or altering flow such that new channels did not possess this habitat (Figures 43 and 44). At six sites. Gold Run Creek, Washington Creek, Oregon Creek, Aufeis Creek, Skeetercake Creek, and Sagavan i rk tok River, the channel configuration was altered so that the channel was wider and shallower in the mined areas, thus the in- stream cover provided by depth was reduced by lowering the ratio of pools to r i f f I es. 151 a) Undercut vegetated bank typical of Oregon Creek upstream study area. b) Oregon Creek mined study area mu 1 1 i p ie channel s. - notice lack of bank cover, Figure 41. Removal of bank cover at Oregon Creek as observed on 24 June 1977. 152 a) Skeetercake Creek upstream study area - note undercut vegetated bank. b) Skeetercake Creek mined study area - bank cover absent, flow spread over wide, shallow area. Figure 42. Removal of bank cover at Skeetercake Creek as observed on 18 June 1977. 153 a) Washington Creek upstream study area showing predominance of boulders. b) Washington Creek upper mined study area, note spread of flow, multiple channels, lack of surface water. Figure 43. Washington Creek upstream and mined area on 9 September 1977 showing reduction of i nstream^cover due to gravel removal operation (flow level [O. II m /sec] = 20 per- cent of mean annual flow). Other habitat alterations include increased braiding, siltation, and intergravel flow. 154 Sagavan i rk tok River upstream study area, note predominance of bou I ders. b) Sagavan i rk tok River mined study area showing extensive sedimentation and backwaters. Figure 44. Reduction of instream cover as provided by boulders at Sagavan i rk tok River, 3 August 1978 (flow level, 60 m /sec, = 155% of estimated mean annual flow). 155 Hab i tat D i vers! ty The result of decreasing habitat diversity, that is, creating uniform habitats by gravel removal operations, was to favor certain species or life history stages over others. One of the main indicators of reduced habitat diversity was increased braiding in the mined area caused where gravel deposits were scraped to below the water line or where flow subsequently increased to inundate the mined area. This type of habitat alteration oc- curred at 10 study sites (Washington Creek, Oregon Creek, Penny River, Nome River, Aufeis Creek, Kuparuk River, Sagavan irktok River, Ivishak River, Kavik River, and Middle Fork Koyukuk River-Downstream) (Figures 43 and 45, Table 13). The channels in a braided area usually have a uniform depth, velocity, and substrate with minimal bank cover. The areas were general ly characterized by increased wetted perimeter, reduction in channel depth, and reduced mean velocities (Figure 46). At Washington Creek (Figure 46a), for example, the cross section in the upper mined area (Cross Section 3) had the greatest wetted perimeter at al I flow levels, but most of this was in shal- low open channels with little cover. Similarly, at Oregon Creek (Figure 46b) the wetted perimeter at cross sections in the mined area (Cross Section 2 and 3) was considerably greater than that in the upstream area and ap- proached or exceeded that of the Cripple River cross sections, a river with greater than three times the estimated mean annual flow of Oregon Creek. Again, the Oregon Creek mined area channels were wide and shallow, providing low quality and low diversity habitat. The final example, Sagavan i rktok River (Figure 46c), showed a similar pattern with the mined area cross sections having a greater wetted perimeter, but a shallower depth profile than cross sections in undisturbed areas. Habitat diversity was increased in some other mined areas by the crea- tion of new habitats. Three types of new habitats were usually found: (I) low velocity backwater areas, (2) a side channel off the main river, and (3) a flooded pit forming a pond habitat (Figures 47 and 48). Low velocity back- water areas were found at five sites (Sinuk River, Skeetercake Creek, Sagavan irktok River, Dietrich River-Downstream, and Middle Fork Koyukuk Ri ver-Upstream) ; side channel formation occurred at three sites (Skeetercake 156 f- «p ^ a) 27 July 1973 - pre-mining b) 2 August 1976 - post mining Figure 45. Increased braiding at Sagavan i rk tok River study site caused by mining mid-channel gravel bars and a vegetated island in the active channel (mining operation conducted during the winter of 1974-1975). 157 70-1 E MAF= 0 53 m^/sec ^ 3 0:60- LU Cross Section ^ 50- CO Depth 1 5 Inten/al US 2 3 US UM 4 BM 5 6 LM DS 5 40- --^ t-0-10 42% 2 10-20 35 26% 88% 22 12 42% 26 53% 39% 26 28 O n/^ P 20-30 24 19 18 10 14 UJ 30- 1— ^ £30-40 15 11 7 10 UJ — « — ■ 'J 40-50 11 3 4 6 J 20- 10- 6 50-60 60-70 6 0.7 0.4 2 /.^^'^^^ 2 10 30 60 100 % OF MEAN ANNUAL FLOW a. Washington Creek 30-, y3 S MAF= 0 69 nr^/sec Oregon y^ Cross Section IT UJ 2 32 m^/sec Cripple y 1 Depth 2 3 4 C- 5 6 5 201 X -5 CO Internal 0-US 0-UM O-LM Bridge C-LM C-DS 2 £ 0-10 39% 66% 64% 24% 29% 31% UJ £ 10-20 29 32 29 21 24 28 a. 2 20<30 24 4 9 16 19 24 S 101 5 30-40 9 02 15 15 15 1 UJ 40-50 0.3 " 5060 60-70 15 8 0.4 11 2 2 UJ 5 /£(' 10 30 60 100 % OF MEAN ANNUAL FLOW b. Oregon Creek — Cripple River Cross Section Depth 1 2 3 4 Interval US UM LM DS 150- MAF= 39 m3/sec ^^^2 0-10 14% 10-20 13 23% 21 % 21 19 21% 19 20<30 13 19 17 17 ? 30-W 12 16 14 14 —3 40-50 11 10 11 11 C -- — ^* 50€0 7 6 10 7 IIMEJE ^ 60-70 5 7080 4 3 6 1 2 5 3 2 80-90 4 0.2 0.4 2 UJ Q. ^ 90-100 3 1 UJ 100-110 3 1 a UJ ^ 3 110-120. 3 3 t-50- 1 z 120-130 3 0.2 UJ S 130-140 2 0.07 5 140-150 2 150-160 1 160-170 1 170-180 0.4 180-190 04 190-200 us UM BM LM DS 0 KEY Upstream Uppermined Between Mined Lower Mined Downstream Oregon Creek 10 30 60 °/o OF MEAN ANNUAL FLOW 100 c. Sagavanirktok River C •MAF Cripple River Mean Annual Flow Figure 46. Response of cross- sect i ona 1 wet ted Derimeters to percentage of mean annual flow and percenta ge of cross sections comprised of se lected depth intervals at mean annua 1 f low at three g ravel removal study s i tes. 158 a) Dietrich River-Downstream - inundated mined study area. b) Middle Fork Koyukuk R i ver-Upstream - backwater in lower m i ned area . Figure 47. Low velocity backwaters formed by gravel removal at Dietricti River-Downstream (13 July 1978) and Middle Fork Koyukuk R i ver-Upstream (18 July 1978), note extensive silt deposition in both cases. 159 Cut-off channel created by mining Original channel a) Skeetercake Creek showing cut-off channel, 4 September 1975. b) Jim River showing side channel created by mining in a high- water channel, 12 August 1978. c) West Fork Tolovana River pit created by deep excavating in an abandoned channel, 29 July 1978. Figure 48. Creation of low velocity side channels and inundated pit following gravel extraction. 160 Creek, Middle Fork Koyukuk R i ver-Upstream and Jim River); and flooded pits were created at seven sites (Penny River, Ugnuravik River, Dietrich R i ver- Upstream, Prospect Creek, West Fork Tolovana River, Tanana River-Downstream, and Tanana R i ver-Upstream) . The changes in habitat diversity were determined by the location of mining and, to some extent, the type of mining. Braiding (decreasing habitat diversity) occurred where the majority of flow went through a mined area, such as where a meander was eliminated (two sites: Penny River, Middle Fork Koyukuk River-Downstream), an inchannel island or gravel bar was removed (five sites: Washington Creek, Kuparuk River, Sagavan irk tok River, Ivishak River, Kavik River) or where excavation occurred in an active channel (five sites: Washington Creek, Oregon Creek, Penny River, Nome River, Aufeis Creek). Removal of gravel in active channels created braided areas in what had previously been pool-riffle habitats, thus, in these cases there was often a loss of instream and bank cover, substrate alteration, depth alter- ation, spreading of flow combined with decreased velocity, and loss of pools and riffles. Habitat diversity increased at two sites with incomplete meander cutoffs forming backwater and ponded areas or side channels (Skeetercake Creek, Middle Fork Koyukuk R i ver-Upstream) and with gravel removal in a high-water channel to below the water table such that it con- tained ponded water (Sinuk River) or annual flowing water (Jim River). Habitat diversity also increased at three sites where recent gravel extraction or channel changes created low velocity backwater areas and braided characteristics were not well established ( Sagavan irk tok River, Dietrich River-Downstream, Middle Fork Koyukuk R i ver-Upstream) . Ponded areas or low velocity backwaters were characterized by a sand to silt substrate. The low velocity with associated clear water often allowed increased growth of filamentous algae. Water temperatures were usually increased over those in the active channel because of the dark substrate and poor circulation. Similar effects, although not as great in magnitude, were observed where side channels were formed at Jim River and Middle Fork Koyukuk River- Upstream. Water velocities were reduced and increased silt deposition was observed in the main channel. 161 The three sites with increased habitat diversity due to recent flow were 3 to 4 years old and, in two cases (Dietrich River-Downstream and Middle Fork Koyukuk R i ver-Upstream) , flow had only entered the site within a year or two of the site study (Figure 49). The habitat diversity in these areas will probably decrease within a few years as meander cutoffs are completed and braiding characteristics are established. Inundated pits were formed when gravel removal was conducted away from the active channel and the depression, usually deeper than I m, filled with water either by direct connection to the river or through intergravel flow. These areas developed characteristics typical of pond habitats, i.e., mud bottom, rooted aquatic vegetation around shorelines, high density plank- ton communities, and macroi nvertebrates typically associated with a lentic environment. Two types of pits were included in the study: shallow (< 2 m) and deep (> 2 m) pits (Table 14). Shallow pits (Penny River, Ugnuravik River, Prospect Creek) normal ly froze to the bottom in the winter wh I le deep pits (Dietrich R I ver-Upstream, West Fork Tolovana River, Tanana River- Downstream, Tanana R I ver-Upstream) contained water year-round. Two of the deep pits (West Fork Tolovana River, Tanana R i ver-Upstream) showed dissolved oxygen and temperature stratification in the summer of study while the other two (Dietrich R i ver-Upstream, Tanana River-Downstream) did not (Figure 50). The time at which stratification would be most pro- nounced was missed at Dietrich Ri ver-Upstream and Tanana River-Downstream and it is possible that there was some stratification mid-SLimmer; however, the Tanana R i ver-Upstream and West Fork Tolovana River were thermally strati- fied from early June to mid-September. Al I pits except the Tanana River- Downstream pit were connected to the associated rivers. The Tanana River- Downstream pit was on a vegetated island and connection to the river was inundated only during annual high water events. This pit had clear water (bottom visible to deeper than 5 m) , very little mud or silt even In the deepest area, and virtually no thermal stratification. Aquatic vegetation was absent except along the shoreline, despite the extreme water clarity. Four of the five deep pits had extensive shal low areas, with over 25 percent of the area less than I m deep. Only at the Tanana River-Downstream was a majority of the area deeper than 2 m (Table 14). 162 vf a ) 16 Sep t ember I 972 b) 2 August 1976 Figure 49. Sequence of aerial photographs showing effects of overmining the inside of a meander bend at Middle Fork Koyukuk R i ver-Upstream. Immediately following mining (b) there was an increase in backwater areas. The next year (c) the meander was partially cut off, creating a variety of low velocity habitats. 163 (0 > - 6? (D Ul — C c (0 S 1— o o i ID n c ro — > S5 4- o — in — a> o ^ (_> j_ o 0) Q. — 01 S5 O 1- Q_ CC E n (a o a> .— L. V- -1— ^- -*- in 65 0) O- — .— 3 Q C 6S 0) — Q. (D i: > — -^ \- E O. Q) — Q) -1- Q C ro — in O O O CNJ ^0 •^ «3 CD 00 ^ rsl tA ro in O O O in c^ r- ■* O in in O O CN (N O I^ •^ CN lO "* ^0 O in K> in (0 10 o o o o o o a\ — d o o ^ f^ ^ OJ O _■ ^ _ O CNI 1^ Osl in 1^ o o o o o o U3 o o Q- 0) CM ro "* in in 1 1 1 1 c — CM rO ^ A £ ID 1_ E ^ ID 3 4- ^ E Q- — (D OJ (D ^ X ■o -*- ^ (0 o 13 01 (D E 164 1 + 2 ~ 3 E 44 I t 5 UJ ° 6 T and DO X X i T and DO ? 1 ^ ? .le ^^ a. Dietrich - Upstream. 10 July 1978 b.West ForkTolovana,13 Sept 1978 1 2 3 E - 4 I m Q 6 T arxj DO 2 4 6 8 10 12 14 16 i 1 X I cTanana— Downstream, 9 Sept 1976 1 2 • Q. Q 3 T and DO 2 4 6 ^ 10 1g 14 16 18 X // d.Tanana — Upstream, 19 Aug 1978 X' — X Temperature (°C) • • Dissolved Oxygen (mg/t) Figure 50. Temperature and dissolved oxygen profilesat four deep gravel pit study sites. 165 Water Qua I i ty Water quality measurements reflected habitat alterations in several ways. First, dissolved oxygen and temperature responded in a predictable fashion to increased braiding. The spreading and shallowing of flow and loss of cover led to an increased rate of heat exhange, with the temperature, and therefore dissolved oxygen, responding more quickly to ambient air tem- peratures in the mined area than in the upstream area. Similarly, areas with ponded water showed increased temperatures and reduced dissolved oxygen (Skeetercake Creek, Dietrich River-Downstream). An increase in dissolved oxygen and decrease in temperature which was not caused by flow alteration was recorded at Dietrich Ri ver-Upstream where a spring was uncovered during gravel removal operations. As mentioned, inundated pits functioned as pond habitats with corresponding water quality characteristics. These included higher temperature and lower dissolved oxygen than the associated rivers and in some cases, thermal and oxygen stratifications. A second type of water quality change was a change in conductivity between the upstream and mined areas. A change in conductivity may indicate the existence of a spring water source near or exposed by the gravel removal operation. Such changes were recorded at Aufeis Creek, Skeetercake Creek, Dietrich Ri ver-Upstream and Penny River. As already mentioned, the Dietrich R i ver-Upstream was an identified spring exposure. The Penny River had a spring-fed tributary entering the floodplain in the mined area. Springs were not recorded at Aufeis Creek or Skeetercake Creek, but the conductivity changes may indicate their existence. A third type of water quality change was alteration in turbidity or suspended solids, or both, in the mined area compared to the upstream area. These changes probably indicate erosional or depositional characteristics of the mined area, but the sampling was insufficient to reach definite conclu- sions on an individual site basis. 166 EFFECTS OF HABITAT ALTERATION ON FISH POPULATIONS Observed Alteration of Summer Distributions or Densities Several types of changes in summer fish distribution were observed in the mined areas; specific types of distributional changes were related to certain types of habitat alterations caused by gravel removal. These changes included: (I) reduction in the numbers of all fishes in a disturbed area, (2) replacement of one species by another species, (3) replacement of one age group by another age group, and (4) increase in the number of fish or species, or both (Table 15). A list of all species caught during the study and their scientific names is included in Appendix A. Density Reductions. Reductions in numbers of all fish populations occurred at Washington Creek, Aufeis Creek, and Kavik River sites. The habitat in the upper mined area of Washington Creek was altered in several ways, reducing habitat quality and diversity to an extent that few organisms could utilize the newly created habitat. The density and biomass of Arctic char was significantly reduced downstream of the upstream sample area (Table 16). The slimy sculpin density and biomass was also reduced in the upper mined area, but increased in the lower sample areas to densities exceeding those in the upstream area. The sculpin biomass remained low, indicating the slimy sculpin captured below the mined area were smaller than those captured above. Thus, there was a replacement of Arctic char habitat by a habitat more suitable for slimy sculpin in the lower three sample areas. The spe- cific habitat alterations that led to a loss of Arctic char habitat were removal of bank and instream cover and possibly reduced water quality (i.e., increased turbidity) caused by siltation from the melting aufeis field. At the Aufeis Creek site, there was only one life history stage of Arctic grayling present during each sampling trip, thus any changes would have to be density reductions rather than species or age-group shifts. Density reductions were recorded in the upper mined area during the first trip and al I disturbed areas in the second trip. Specific habitat altera- tions that led to reductions in Arctic grayling habitat were: (I) the reduc- 167 (0 I. < ■o 0) > 3 E U o 0) in to o (D -4- - u c in in - — ro O (0 in XI en ■o 1- .— sz — c 4- c c < — >■ i- in ■o j3 — (0 "D c — (0 0) :^ E o 0) E i- c o i_ cn 10 H- •. c >^ cr> 10 •— (0 CT) (0 0) — \- — T3 •— 1- — — 4- cu c ^ in cn in — 0) E (D (D E en 10 1- •— o c c ro 4- U c 1- ■o o o — u - •— (0 5 o 4— ^ cn •— a> c c — Q. — — •D 4— c O Q. ro L. c i_ 4— o (0 XI in :^ 4- 0) C .— E 4— (0 .— 0) E I. 0) (D o U Q) E 0) 0) ■ — L. ■o 4— to 4— c I- 1- 1_ 3 4- ^ 4- XI 0 c c (0 in o en < ■D o — (0 Q- ro •— in in • — 0) 0) Q- (1) c \- en -C E L g 0) i_ 4- — o •^ 1_ •— 0 c a> 4- 1_ -— o o (0 J3 — u • *. u Q. M- •— l_ 4- in 0) o X -4- I- ro 4- (0 >- ■o (0 — •o (D > 0) o O) ^ o JZ U - 01 - — O M- — >- c in — c ^ U) C 1- E l- 4— E ^4- 4- E 0) 4- 0) O in (0 c Q) (0 (0 u (0 CD o 4— T3 o 4- o in 0) o -«- 4— I. O SI i- (D C C l_ — 0 c l_ o — o l- — (0 o o z o < U) D — < in Q- (D < Q. XI _] 10 4- 4- Q. Q- 3 :^ in Q) c Q) — l_ c U Q) Q- c D ■o ce I- (D TD s — 0) o in to ». in ^ E 1 1 ■o 1- 1- i- ID ^ X -^ c 1 Q) XI u 1 01 O •^ E (0 XI 4- > m 4- 1- 0) 10 01 in ID cn ID — D 10 o in (0 4— in X X 1_ 10 X x> c Q) — 1_ 1_ •— ^ 0) 0) c 0) •« • ^ > \- ■^ in XI 4— CD 0) o j<: 1- 4- "D — u cn in o o cn ID — in 4- -. — ID c o ID Q) o c •— o c c 0) ID ■o c (0 — X X 10 c 1- •o XI c •— 0) — - — i_ 1- d) — S Q) ID X — 4— C c — X M- E X u X 3 C 4— -. in o 10 .— 3 ID • ■^ .— 01 1- C •o •o c in en X) • -^ X Q- ■ ■^ 10 ID 0) 14- ID X 1- X Q) .— 0) 0) (0 c c 0) — D ^ — 1_ 1- ^4- 1_ 01 01 4— 4- in ■o X D .— in Q) in •. c 0) X •V 4- o X • -s > in ID C (0 c o "O ID C in ID c i_ in 4— i- o ID g o 0) o TD .— Q) c 0) 1- X c ■^ 0) c 3 -. 0) o a> tj u Q- i_ ID ID s_ 10 — 0) ID Q) 4- •— u 0) 4- \- X) u a> in l_ u X X 4- XI X 4- 10 4- 10 E o 0) Q> ■o 4— (0 X m o ID 10 0) u 10 g X 1_ 10 g ID c X -o 0) (0 a> ■a 4- g in l- c Q) l- 01 .— c in g i- 4— m in ^ 10 0) X ID T3 X) L. o . .^ (D 1 u X >>s > c o in 0) C in 0) 4- • - CL in U ■o 1_ c 1- 10 ^4- > 1- in 4- 0) in 4— c 1_ t) O o o c 4- o o - — 1- 3 o o c ID i_ 3 c ID ID o ID 1/5 o CO 10 in CL lo in X ID o to 3 g o S 3 g X u-> X 1_ OJ > 0) Q) I. tj c o cn c X in 10 0) 0) I. I. o > >- c c (U Q. I- 0) E O 168 Q) C O O ro ->- a> i_ C (0 O — T3 -1- CD (D C l_ — (u e — (/) (D c -4— O o ■— -(_ 4- u CD l_ 1_ CD CD D — o CD (D D 1_ \- •* < o CD o — 4- a. » X) o CD CD M- O c •4- >■ en "". 4- CD c "O c ^4- (D CO 0) cn c l_ •— c •— o — 4— JZ s 1_ D M- CD 10 (0 — ID o o Q) en "D o 05 c CD •- CD :^ in .— 1- ^4- (0 — (J> 4— E (0 in 4- < 4- 4— o Q) 1_ 4- E 10 in i_ o c c (0 -C .— a> (D cn o in o — CD — CD c ID CD 4- (0 XI T} 3 — (0 4— o >- 3 o -^ E ^ CD .— O *. o — (0 1- u s 3 i- CD c cn in 4— .— ^ CL CD ■4- c o < o o C o cz •— c ID 1_ 4— in in .— o — — 4— •— ■D ■— 3 (0 o -— >. (0 >~ i- < CD (D 1- a. "O CD JZ (0 1- (0 CD V4- 4_ 4_ o in 4- — c CD U c i- 4— 1- > O ■D .— .— CD c o =1 0 4— (D O CD C7) in CJ> O CD JZ XI c o 0) u — c M- •— ^4- o (0 in 5 ID .— c ^4- if) 4- CD i- 4- • — o -D o CD (0 JZ •o 4— ^4- o E 3 o — CD •— CD 1- 0) ■o c -C ■^ 0) c D D cn D 4— 4— 4- 1- ID s_ c 1- 10 CD CD L ■D u "D ■D u c o CD u 3 10 1_ .— c o o Q) o M- CD C 1- CD s_ JZ C c o ^ 4— — ._ z ^ cr ■D o a: (0 < E < S — 1- o in CO — 10 — ._ • •* i_ ^ XI 1 CD 1- cu (D CD ^ M- -> T> a> 4- 4— cn O 3 CD CD CD 4— ■V CO ^ in •v CD ID ID c CD in CO CD in 5 c c in .- CD XI .— c 5 ID g — i_ ^ ID 3 L in l_ ^ -o ID O 0) XI CD in o X3 CD — o ". — •— ^ — L. "O CO 4— — ID T3 4- CD c CD X3 10 o ^4- o CD 4- ID CD XI T3 CD CD -4- .— C — s_ c ■a in 5 M- ID in g 3 — in Xl ••s 4— •— c c ^ 3 TD in CO ^ CD ■ ■^ XI in i- C o 1- 3 o CO CD 10 CD o 1- in — O -o 0) CD • - CL CD ID in CD V- CD >- CD i_ CO CL. CD 4— — — > ■D X] >. ID 1- o Xl — > ID \- in 10 3 CD - O CD — ID o c XI O XI n ID s XI c in o in •. X) 1_ CD — ■V • — u i- CD i- c 1- CO CD •— o -o CD in — - CD 1- "D 3 (0 CD E CD C in c 1- • •. C in E ID CD -1- o CD 4— ^ 4- ID l_ — in — •■^ CD XI • — o ID c 4- CO c "O o 10 CD o ■o o l- > c X! CL ID c CD g •— c o S l_ c •— Q- • - CD O CO — 1_ ID L. o 4- CD ^ 4- .— CO M- 4— u — ID - 4- x: 4- X) • *. CL > o in 1_ ■s M- ID cn \- in cn o cn CD cn i_ •— CO cz ■ - XI 1_ o g E •— XI 1- c 1 XI XI i_ "^ 10 4— X} •— L CD 4— ID ID — XI 3 c (0 in c o CO •^ 4— 3 "D CD V4- -. 4- cn o XI v_ •— 10 ■V "D XI CD 10 o CD 1- o CD ID XI E Q- CD E CD CD 4— S ■D 4— g CD CD ■o 4- 10 X) c in "D ID 1_ C in — 10 in X) CD (0 — CD •— (0 CD l_ ■D CD o c CO 1- X) CO CD XI in Q. 5 CL "D CD CL. 4- CD TO CL •— > ID CD CL ID 1_ (D _^ E ID • — L. CO in "O c O in -o i_ 10 4- CD 1_ o O 1_ ID o i_ XI c 10 "O ■a E X) c o 1_ in 4- o ID i- o i- ID o 3 o CD c c CD 3 O CD u C ID lyo X) '*- uo XI T) 00 in CL S ID ID cc in CL XI CO => g CD > 1- — j^ CD cr CD > CD .— Jx: 1- ce — O > ^ ID in 3 1_ .— i_ 3 CD ID C •4- CL CD 3 3 n < ^ i. CD ^ > CD — CU cc I. o -!£ o Q) -f- ^ js; CO l- o L. c IV CD 4- > CU CD cu D> -^ CD to to (U > CD (0 XJ cu 3 C o u 169 0) c c o o (0 -f- Q) ^ c ro o ■o ^— OJ (D c 1- .— 0) E 4— — c <□ — U) (D c ■*- o o -1- 4- (0 V4- — o 3 Q. (/) O -»— Q. o cu ^ (fl CD C c E >- JD ■D in D o J3 (O in 01 en c (O r: u J3 I. ZJ en (O Q) E c (0 a> o 1_ < -o c ■o (D 0) in 1- (0 (0 (U ^ I. u o (U o -o ■^ ^ u 4- 1- 0 < n 10 1. 1 C CP ID .— -*- 4— -o E X) (0 ■o CL > Q) J2 0) o in in -■ o 10 -a ID in Q) c 0) 1- E 1_ (0 i- 0) ID u u -*— a> c c ID i_ .- CD •— S 4- c ji: in • •. •— ■ >^ o c in -o c ID — L. — .— n (0 ID ID "O JD 1_ — »■ 0) n CL CD in ■o c ID C -4- O .— Q) — ^ O T3 1- O en — .— u Q- — ^4- ID Q) (U I- XI ■D in in ■D n 0) (0 Q) • •^ Q. - 0) CL -. 1_ ID S 1_ <0 C Q) i. o u \- o 4— o — c U .— ID to ^ — in -"- s ■o c 1 1 D C CD o o — C (0 Q. i. •* c 1 — 0) 1- CD 1 '~ Q) X> v_ 1^ - (U C Q. X) C (0 a> c j«: •— c — ID c O o -o O 3 CD U CD — D c >- — O C 4- c 4- in ID 4- 4- in — in >- — I- — — — L. (0 in Q) 4- in in >- >- i_ O in in \- O E (0 o en c CL o Q) Q- — i_ "^ c E c T3 > E - O) o -o O O) 0) •— O in — i — c u c 4— X) o U ID l_ 4- ID o C -• •— •— (U O I. in — Q) 4- in sz 4— 4— > I. -t- 0) E ID 0) in o c o < (fl — -a 3 4— •— •— 1- - I. 0) to o J3 0) 0) o o ^ o Q- 13 (D CL 4- ^ CL *4- in -c ^ ■D in — XI c a> in ». c — O C 1_ o CO -t- < 5 o o — < 1- — u 1 ■ •. X V_ a> c XI 0) 4- ^ .^ JD o in ID — Q) 10 s Q> 0) X) m Q) C c c in 1- ■o C c o CD o 0) ID ID Q- t- c XI SZ 11 CL — c o u XI o D 0) > Q) 1 4— o o 4— o ^ 4- -C ID — ID u CD (D CD s "4- V- 5 ? .— ^ Q) ^ J^ H IZ o >^ X) u c CD ID — c ID ID C C c n S 10 n J3 ._ 0) 0) 1_ XI c E •s "O CD Q- CD Q) CD 0) C in c in 1 — C in ._ ID 10 — (0 CL X) CL 0) L .— XI 0) ID a> 10 u 0) 4- •— 1- L in 1_ o 4- 1- 10 o O o o c ID ID l_ — s Q- JD TD ID lO 3 1 1 I. CO CO 0) O ID Q > l_ — 1 1 '— Q) 0) (U > > -^ cc 1_ C •— •— D Q) Q) — cc ce J^ 4— J^ > D c SI ^ >- in > ct 1- o u o o Q) '— •— :*: >~ J«: x: u 1- X) > . — 4— 4— 4- • =j ID > 1_ Q) a> Ll SI ID o •— •— • CO CO •^ z Q Q s ■o 1 0) C T3 — C o •- Q. 0) - c m C I. (0 0) ^ -t- o ro S I. ^ 0) o ■•- (0 (0 ^ ? 1 •> x: CD CD C ^ X) c w — 1_ ^ en C X) ._ 0) Q_ in l- ID 10 01 \- 01 4— U 1. ID CO o 5 10 o I > 3 o 2£ X) u ^ 0) E l- 0) .- D — — -»- J3 in Q- 10 10 I. U 13 a> in c SI 10 — m in in 1- 10 10 *- o c S ro c -^ 1. — ■D U -1- 10 in ... ID -Q £1 in 1_ D i_ X) T3 in 10 0) XI ■o in 0) Q) 10 — ■D in Q) n C 10 I. 10 10 0) O -l- I. c in in u — c — 0) D Q) T3 • •^ C in - c ... 1- 1- 10 1_ (0 CD X Q) n -i- u -1- in 10 ID I. cn g CD 5 0) c c — "a in .— "O ID CL 0) — Cl- • 0) — (0 CM D E E ^ lO c O O - O n — 1^ O lO o O in vo 0) 13 O) E > cn < — 1 CM in — 1 GO O 1 03 ID — CM — 1 CN — rO ■* — 1 >* CM ^ — O CT> o in c ■ — O 0) >■ >- CM •t- c E ■4- E a> D — — O c —> — in o (/) CM to c — -O -£= 2 ,^ ITl in CN r- (0 1 s - CN -*- in CD — > -*- < — CM 1 CM lO 1 CN O — lO 1 in o — 1 O ■* — CM _ (0 -s E C (D ._ Q) in , Q- 1. in — ^ (0 CM CD 3 C/3 E E V O O O — ■ (J 1/5 M- O E CD < — ^O O ro 00 CN — O iTl 4— o 4— u (D •o o — XI C Q) -t— >~ — Q> to If) -1- CM — E X \- T} < in o J- (D (U c o ^ J£ Q) — ■ S X) CJ O O X CN CN CM 1 O — — OQ cr> in > — 1 ■* lO 1 — o 1 CM — OM O CN CM (D -4- < '^ — — - '-' "^ Q. . o *- ■ E in u o (D 0) < in D cr> ^ c S (D in ID O — u- -^ l_ (D > Ji 0 in a> g «) o (0 x: • in o o in o CM CO o VO r- E (D ■o (D E in Z <0 SI CD ID O 0 Q. in l- E — i_ CD ■<- in J— U in •o CU •o ^4- (D c — Q) o (D LU U) "D UJ 0) TJ 1_ ■*— 0) -♦- ro . CO (0 Q) c E ID O CL c 0) I- (0 E (0 CD E c 0) 'e I. 1_ 0) 0) (0 in CL 1- 0) CL CL CD 5 (D u (D s o in c X E D Z 3 > CO Z) 3 CD _J Q (D X 172 tion of the pool-riffle frequency, and (2) increased braiding characteris- tics with the associated loss of bank cover and altered flow regime. At the Kavik River site, habitat quality was altered by the erosion of berms left in and along active channels, channelizing one section of the river, and creation of a more braided configuration. The densities of Arctic char and Arctic grayling for each study area were estimated by repeated shocking of blocked channels (Table 17). Total fish densities in the mined area were reduced by a factor of three or greater when compared to the undisturbed areas (Table 18). The catch of adult Arctic grayling, as de- termined by angling, was also lower in the mined area (Table 19). The den- sity reductions occurred in both Arctic grayling and Arctic char with neither species apparently favored by the habitat alteration. Removal of instream cover appeared to be a major habitat alteration affecting reduction of fish densities because a channel that contained boulders adjacent to the mined area supported densities of both species comparable to those in un- disturbed areas. Species and Age Group Alteration. Species shifts were observed at nine sites (Washington Creek, Oregon Creek, Penny River, Kuparuk River, Sagavan i r k tok River, Ivishak River, Dietrich River-Downstream, Middle Fork Koyukuk R i ver-Upstream, and Middle Fork Koyukuk River-Downstream) because alterations in the type of habitat allowed other species to populate an area (Table 20). A similar response is a change in the age structure of fish inhabiting a reach of river, as was observed at Kuparuk River, Skeetercake Creek, and Middle Fork Koyukuk Ri ver-Upstream. In these areas newly created habitats favored or excluded certain age groups in the areas affected by gravel removal operations. On Kuparuk River, the mined area had a more uniform habitat than the upstream area and numerous small channels of simi- lar velocity. Age-0 and age-l Arctic grayling and several age groups of slimy sculpin were present in the upstream area while only age-l Arctic grayling were captured in the mined area. At the Sagavan i rk tok River, Arctic grayling juveniles were confined almost exclusively to the mined area, while the upstream area catch was dominated by round whitefish and an unmined channel adjacent to the mined area contained adult (-300 mm) Arctic gray- ling. Again, the mined area was changed from a large single channel to an 173 to it) ^- U) CN ro E E O O O cn .— — c OD ^ E — cn a> >- > — ro < 1- C7> . — . J3 CN U ::^ E 4— O 4_ .— O o Ul u c ^ < CD -C -D 10 CT> IZ > ^ < m — U) CM ro E E O O O I. J3 ^ (0 E JI 05 cr> o > — < u 4- :^ CM o -4- E 1_ •— O < en o c — 0) \ ■D SZ m en — > -4- < — g (O ^, i_ in 0) 0) -!■: in o in o ID ^ CL in ■o 0) 10 — ^^ d) Q. CN i- E E < ID — in in ■D c 0) 0 _ CL ■*- E u ID Q) in in lO I ITl CN O CN I ro O O r~ • I ^ 0\ CD -^ • I o o r- CN ■* 00 — 00 r- — o — CO r^ — — in r- o ITi iTl O ro 1^ t^ O t^ — d 6 rr\ lO d 6 O d o CN O CN CM VO 00 ^ CN N~\ CM in o ^ 03 (J. ■^ CN — K^ CN in 00 rri — o in o en — in CN • I • I O 1^ o o CO CN VO ID Q) 1- ID >- 3 I- 0) E (U CL 0) Vi >- 3 —5 E in 10 CM 0) D 1_ < ■*— in 00 CL CM ZD CN a ■D 0) C in in 0\ o\ o d in — • I rA O 00 lO t^ lO 0\ 1 00 CN 1 t ^ o 1 O ^ rA CN a\ o m o OJ CN CTi CN r-~ CN CTi K-> 1 -I -I — o o o o o CM Cvl CO CM in 't OO ^ in VO E ID 0) 1- in c 5 O Q 1_ 0) 0) E — D ID Z > ID X) 174 a> c JC o o £ U) o \. 4- u 0) Lll >- J3 TJ 0) C 'e lO l- 1^ Q) ON ■^- — 0) Q •. 0) U) 4> (0 .— CO U) (0 >- Q) ■o 1- D < -*- 01 "O 0) 1_ JD 0) 1_ > D — 4- ce Ul — -St: ■D C > Z) ID :^ TD C 4- ro (D -o E 0) (0 c . =J ID o 4- 4- Q> — .— in \- \ in — C 0) — < • — 1_ \ S ■a (U Q. - E - Spacias composition % Char % Gray l i nq Minnow Eiactro- All Major spacies trap shock gaar typas lost/gainac Washington Ck U UM BM L* D I .3 0.7 0.0 0.2 2.2 26 5 12 22 26 100 100 55 il u 9 I I •'SS ♦ss *ss ♦ss Or agon Ck - J una 22 I I 56 93 -SS August 4.6 2.7 137 164 93 78 68 65 *SS Saptampar u 1 .7 2.9 14 30 100 85 36 45 »SSI«IIT1 -SSIESI Penny R • Jung 0.06 0.20 0.50 0 33 3 ♦SS ♦CS. ♦SS August U W P D 15.2 40.5 7.4 24.8 71 64 4 85 ♦CS ♦CS -CS Saptatnber U W P 0 ie.2 9.8 12.6 1.3 42 34 0 67 ♦CS ♦CS -CS Kuparuk R U M 6. J 12.8 1.0 43 94 75 -SS -SS Segavanirktok R U Hi D 0.56 1.12 0. 14 10 61 0 ivisnak R u W 0 3.1 3.3 1.9 80 87 87 17 8 5 ♦AC ♦AC Dietrich R-Oownstraam u M 0 1.65 0.65 1.50 54 79 48 -SS ♦RWP Miodla Fork Ksyukuk River-Opstraam U UM LM OC 0 4.2 1.9 4.4 1.8 3.1 64 44 ♦RWF 59 ♦RWF.^LNS 55 ♦RWF 25 ♦SS, ♦RWF Middle Fork koyukuk Ri var-Downstraam u M D 1.3 2.7 25 38 ♦SS, ♦RWF ♦SS, ♦awF U ■ upstream UW « upper mi ned BM LM * lowe*- mined P « pit OC * « increased feiative to upstream CS » coho salmon RwF » round whitefish AC between mined D ■ downstream or iginai channel decreased relative to upstream SS Arctic char LNS S [ imy SCu I p in longnose sucker 177 area criss-crossed with numerous shallow small channels. At Skeetercake Creek, gravel removal in the upper mined area created an extensive backwater which was utilized by adult Arctic grayling; at the middle mined area, bank cover and pools were removed and this led to a reduction in the population density of Arctic grayling. At the lower mined area of the Middle Fork Koyukuk Ri ver-Upstream site, the single-channel sinuous configuration of the river was changed to a split channel with extensive backwater areas. The catch and species present were similar between mined and undisturbed areas, but the age structure was more complex in the areas affected by gravel removal. Age-0, age- I , and age-2+ Arctic grayling, age-0 round whitefish, and age-l and adult longnose sucker were captured in the mined areas while the species caught in undisturbed areas were primarily represented by a single age group. Only round whitefish exhibited a more diverse age struc- ture in the undisturbed areas. Similarly, at the Middle Fork Koyukuk River- Downstream site the river was changed from a single channel to a multiple channel braided system with numerous backwater areas. Arctic grayling domi- nated the catch at the upstream area, but were replaced in the mined area by round whitefish and slimy sculpin. Potential for Entrapment. Gravel removal in active floodplains created areas of ponded water which were isolated from the active channel. Typically these ponded areas were inundated during high water and became isolated as the water level receded (Figures 51 and 52). Fish often entered these ponded areas during high water and became stranded as the water level dropped. The mortality rate of these fish was assumed to be high because they were sub- jected to increased temperature, decreased dissolved oxygen, greater vulner- ability to surface predation, desiccation if the area dried completely, and freezing. There were 15 scraped areas at which ponded areas were observed: Sinuk River, Washington Creek, Oregon Creek, Penny River, Nome River, Ugnuravik River, Aufeis Creek, Kuparuk River, Skeetercake Creek, Sagavan irktok River, Dietrich River-Downstream, Middle Fork Koyukuk River- Upstream, and Middle Fork Koyukuk River-Downstream (Table 13). Sampling in these ponded areas revealed significant entrapment at some sites. At Sinuk River the mined area was not heavily utilized by fish. Pink and chum salmon spawn in the river and considerable numbers of chum salmon fry were captured 78 Figure 51. Ponded area at Kuparuk River study site where three seine hauls captured 61 Arctic grayling and 2 slimy sculpin, 9 August 1978 (pool I in Table 21). Figure 52. Ponded area at Middle Fork Koyukuk-Upstream study site where one seine haul captured 28 Arctic grayl ing, 3 round whitefish and 3 slimy sculpin, 18 July 1978 (pool 2 in Table 2 1). 179 above and below the mined area. Pink and chum salmon are often associated with low velocity water and there was high potential for entrapment of downstream migrants of these two species. The same two species, plus coho salmon, were vulnerable to entrapment at the Penny River site. At Washington Creek, Oregon Creek, and Penny River, the dominant species, Arctic char, are probably not greatly affected by entrapment because they are generally associated with high velocity water and instream cover and would tend to avoid the type of areas which are prone to ponding. At the Kuparuk River site, a natural ponded area, apparently enlarged by gravel excavation, contained a high density of age-l Arctic grayling (Table 21, Figure 51). At the latter site both natural and ponded areas created by gravel removal were present in the study reach. At the Middle Fork Koyukuk R i ver-Upstream, considerable stranding was documented when several isolated pools were sampled (Table 21, Figure 52). The primary species subjected to entrapment in the Middle Fork Koyukuk River system was Arctic grayling. Migration Blockage. Two types of potential mi n i ng- i nduced migration blockages were observed during the study: (I) blockage due to aufeis for- mation, and (2) blockage due to lack of surface flow. Possible temporary migration blockage due to aufeis formation may have occurred at the Wash- ington Creek and Oregon Creek sites (Figure 53). The principal migrations that could be affected in these particular systems would be upstream and downstream movements of juvenile Arctic char and juvenile coho salmon moving from overwintering areas to feeding areas and downstream migrations of adult Arctic char returning to the sea from upstream overwintering areas, if present. A short-term delay in these migrations may not have a critical effect on these particular species, but a similar blockage for another species, such as an upstream spawning migration of Arctic grayling, may have a great effect on the population in the river. A blockage due to lack of surface flow can occur where flow is spread over a wide area and there is considerable intergravel flow. Under such conditions, all surface flow may cease. Such a condition occurred at the Aufeis Creek site (Woodward-Clyde Consultants 1976) (Figure 54) and possibly could occur at the Nome River site (K. Tarbox, personal communication). The potential for such a blockage 180 Table 21. Summary of Catch from Ponded Water Areas Isolated From Active Channels at Two Study Sites Locat i on Catch per hau I Pool No. of seine Arctic Slimy Round Longnose hauls grayling sculpin whitefish sucker Kuparuk River 20.3 0.7 Middle Fork 2 28 3 3 0 Koyukuk River- 3 20 1 0 •j Upstream 4 0 0 0 0 5 0 0 0 0 6 2 0 0 0 7 2 9 0. 5 0 5 181 a) Washington Creek aufeis field, 21 June 1977. b) Washington Creek aufeis field, 21 June 1977. Note sediment layer on ice inside cavern. c) Oregon Creek aufeis field, 7 June 1977. Note sediment layer on melting ice in foreground. Figure 53. Potential migration blockages, aufe is fields at Washington Creek and Oregon Creek, June 1977. 182 "^*Tr*" ■ a) Aerial view of Aufels Creek middle mined study area, 21 July 1977. b) Aufeis Creek upper study area where surface flow disap- peared for three years, 22 July 1977. Figure 54. Region where Aufeis Creek went subsurface creating migration blockage due to lack of surface flow. 183 existed at several additional sites, such as Washington Creek, Oregon Creek, Penny River, and Skeetercake Creek, but a specific blockage was not ob- served. Creation of New Habitats New aquatic habitat was created at eight sites where mined areas sep- arated from the active channel were flooded subsequent to site closure. These include the Dietrich River-Downstream and Jim River sites as well as the pit sites at Penny River, Dietrich R i ver-Upstream, Prospect Creek, West Fork Tolovana River, Tanana River-Downstream, and Tanana River- Upstream. At the Dietrich River-Downstream site, a wide shallow backwater was created in the spring immediately prior to the site survey, 3 years after mining, and was quickly utilized by round whitefish and Arctic gray- ling. Less mobile species, such as slimy sculpin, had not moved into the area by the time of the survey (12-13 July) but would probably immigrate into the mined area over the summer period. In the river, the most abundant species was juvenile Arctic grayling; the second and third most abundant were slimy sculpin and round whitefish. Removing gravel in an abandoned channel at the Jim River site created a large pool habitat that contained a high density of adult Arctic grayling during the summer. Other species captured included juvenile Chinook salmon, burbot, and slimy sculpin. In the main river, the catch was dominated by Arctic grayling. The present configuration of the Penny River apparently resulted from two separate periods of mining. Originally, the floodplain was scraped adjacent to the channel. The channel subsequently diverted through the scraped site and gravel was removed from the original channel, leaving a shallow pit. During the site visit the present Penny River channel, formed by flow diversion through the original scraped area, was heavily utilized by Arctic char juveniles. The pit, created by excavating in the original chan- nel, provided rearing area for coho salmon juveniles and spawning and rear- ing areas for Alaska blackfish and ninespine stickleback. The catch in undis- turbed areas was dominated by Arctic char and coho salmon with Arctic char dominant in the spring and coho salmon dominant in the fall. The occurrence 184 of both species in undisturbed areas, compared to the single species dom- inance in the mined areas, again reflects the reduced habitat diversity in areas disturbed by gravel removal. The Penny River pit provided coho salmon rearing habitat, which was lim- ited in the river. Arctic char appeared to be more suited to the river environment than coho salmon, and avoided the pit. The pit thus provided ideal rearing conditions for coho with little competition from Arctic char. There was a significant difference in size of coho using the pit as compared to those using the river possibly indicating increased growth rate by those in the pit (Tables 22 and 23). During the winter the coho left the pit and moved to other areas where they possibly would be in direct competition with char for space. If overwintering space is limiting in this river system, the increased number of larger coho could lead to displacement and subsequent reduction in the numbers of char. The Prospect Creek pit, a shallow pond habitat previously not present in the immediate area, was used as a rearing area by Arctic grayling, round whitefish, Chinook salmon, burbot, and slimy sculpin, and also provided a feeding area for adult northern pike (Figure 55). In the upstream area of Prospect Creek the catch in 1977 was dominated by round whitefish, Arctic grayling, and slimy sculpin listed in diminishing order of abundance. In 1978 juvenile Chinook salmon appeared to dominate the fish populations in the creek. The Dietrich R i ver-Upstream pit and associated channels provided a deep-water, spring-fed system utilized principally by adult Arctic grayling and Arctic char while the main river contained juvenile Arctic grayling, slimy sculpin, and round whitefish. The West Fork Tolovana River pit contained extensive vegetated shallow water areas which sloped off rapidly to deep water areas up to 6 m deep, thus creating excellent spawning, rearing, and feeding areas for northern pike and feeding areas for adult Arctic grayling (Figure 56). Arctic gray- ling were the only species captured in the river during three sampling trips, while northern pike were abundant in the pit. The only Arctic gray- ling captured in the pit were adults longer than 225 mm; smaller Arctic 185 I. > a >- c c 0) Q. 0) Q. (0 I. s o c c r- Q 0) "D (0 ■I- a> c 0) l- o c (0 (0 I cn < (1) — 0) Q- N E .— ro U) cyo c "O o i_ .— ro 4- "O 10 c .— <0 > 4- 0) to X) C -I- — ro O) E 0) c E 0) — O I Q) < 4- Q) co ■o c -t- — ro cn E 0) c E 0) — ro < CN in 0\ CN a 00 fo CD O CN O ■* — — vo CM (Ji o in cyi (3\ in in vo in hO «3 in O (Js in C^ rO r-~ CO CD r- CD 00 CO o C7^ 1^ O CO — in lO ^o hO in CM r-- 00 in CO O ^ o < r- in CN in I. .— . 0) ■o £1 *- < E m 0) CO E ro • . 1- ro o 0) K^ a> L ID Ul — \_ 0) c ■*- 1 if) 4_ C S 1 U) Q. .— o Q- ^ ZD Q- S Q 0\ Z) E ro a> S o Q 186 1 — (/) 0) u Q) c (1) (D 0) U O 0) n Ul < ^^ -o o c _^ en Ul 1- Q) 1— 1. < to 0) -»- O- tu CO ^ c ^ C CN CM i.^ 0) U) 0) CO _ _ ■'- r-- ro r, — c o U. c 0) — ^ I. ■<- Q C 0) _1 CO XI CO cn c o :^ E "O < u CO i_ 0) > >- c c a> Q. a> < > c - CD < 0) u c ro u — ■^ > — Q) C — cr> CO in O O O O O O CO CO CO z z z V V V Q- CL Ol r- ro r^ ■* ■* O 00 in in m o r<^ O K^ f<^ ■* ^O CM — — — in o o o o O O O O CO CO V V V V Q. Q. a. Q. o o O O 00 ■z V V CL Q. lO — in in ■* — r^ CN PA — — O vo in CO vo -^ — E E ro ro 5 D ro i_ 1_ in o 4— ro 1- O CD 0) -*- -♦— c "O CL 0) T) D i- ■o in in g 1 Q) i_ in 1 < 4- 0) c CL o E CO -♦— c E (/) c 5 3 ■D ro in S ro o CL •— O 1 1 0) N^ CL o 0) — 1 E 1 ■o 1 ■o 0) i_ — 1 •D 1 1- 1 -*- -t— C c in C3. 1 in CL ■* CL Q- Q- S S 3 Ol CL CL =) 187 Figure 55. Prospect Creek study site - shallow pond habitat supporting Arctic grayling, Chinook salmon juveniles, round whitefish, northern pike, burbot, slimy sculpin, 12 August 1978. Figure 56. West Fork Tolovana River study site - deep pond with extensive shallows providing northern pike and Arctic grayling habitat, 29 July 1978. 188 grayling either were not entering the pit or were consumed by pike soon after entering. Northern pike were apparently spawning in the pit because many age-0 pike were caught or observed in the shallows throughout the summer. During September, age-0 pike were observed in the river in a large pool opposite the pit outlet, apparently moving from the pit to the river. Thus, the pit may be increasing the number of pike in the river system in general and, given the high density of age-0 and age-l Arctic grayling observed in the river near the pit, may lead to a localized increase in the density of river-dwelling northern pike near the pit. Studies by Alt (1970) and Cheney (1972) indicate that movements of northern pike in the rivers of the nearby Minto Flats region may not be extensive. On a small river, such as the West Fork Tolovana River, a local increase in the northern pike population may lead to local reductions in the Arctic grayling population. The upper pit at the Tanana Ri ver-Upstream site had a similar habitat and also provided a spawning, rearing, and feeding area for northern pike as well as a feeding area for least c i sco and humpback whitefish (Figure 57). On a large river, as at the Tanana Ri ver-Upstream pit, the effects of the increased numbers of northern pike must be minimal when compared to the river population. The main effect of a deep pit on this type of river system is providing a clear water feeding area that increases the availability of desirable species to sport fishing. The lower pit was a more uniform depth with minimal littoral area and was used as a spawning and feeding area by I ongnose sucker. The connection between the two pits, a shal low (8 cm deep) stream, was used by longnose sucker fry, lake chub, and juvenile chum salmon as a rearing area. The lower pit was also utilized as a feeding area by humpback whitefish, least cisco, northern pike, and burbot. The Tanana River-Downstream pit was a deep (maximum depth = 9.4 m) Clearwater pit with apparently very low productivity. Fish species captured in the pit were longnose sucker, Bering Cisco, and Chinook salmon. There was no connection to the river, thus, the fish apparently immigrated during high water and became trapped after the water level dropped. 189 a) Upper Tanana Ri ver-Upstream Pit, note extensive shallow areas. w b) Upper Tanana R i ver-Upstream Pit p i ke dens i t y . - area of high northern Figure 57. Tanana R i ver-Upstream upper pit showing extensive vegetation beds, 18 August 1978. Note difference in the extent of vegetative development in this 15-year old pit as compared to the 2 and 3-year old pits in Figures 55 and 56. 190 Effects on Overwintering Areas Possible effects of gravel removal on fish overwintering areas were observed at several of the study areas. Potential overwintering areas were created at the deep pit sites — Dietrich R i ver-Upstream, West Fork Tolovana River, Tanana River-Downstream, and Tanana R i ver-Upstream — by the pits themselves. The Dietrich R i ver-Upstream pit has been reported as an over- wintering area (W. Anderson, personal communication to A. Ott). In addition, outflow from the West Fork Tolovana River pit created a potential overwinter- ing area approximately 50 m downstream from the outlet where a deep natural pool with a 1-2 cm ice cover existed into March 1979. A possible overwinter- ing area on the Penny River was altered as a spring-fed tributary; Willow Creek, that had previously entered the main channel at a deep pool, now entered the river through the scraped area in a series of shal low braided channels (Figure 58). The pattern of freezing observed during winter studies on six of the pit sites indicated that fish entrapment was not a problem during the 1978- 1979 winter (Table 24). In those pits studied, the outlet remained open well intowinter with outlet flow velocities increasing as the still water at the edges of the pit froze, reducing the volume of the pit. Fish appeared to move to the open water found at the outlet areas and the increased veloc- ities may have induced the fish to move downstream to areas of reduced velocity. If fish were holding at an outlet pool and the outlet closed downstream from the holding fish, entrapment could occur. The outlet area in the pits examined was general ly quite sma I I. However, the number of fish affected compared to the numbers using the pit in the summer would be mini- ma I . The outlets of the Prospect Creek and Jim River sites remained open at least until late January and possibly into early February, thus fish had ample opportunity to emigrate as flow decreased during freeze-up. Fish were present (caught and observed) at both sites in early November but were not evident in late January. Both sites were frozen to the bottom in March. At the Penny River pit site, fish were caught in the pit in late December and ORIGINAL CHANNEL WILLOW CREEK - a) Willow Creek, a tributary of Penny River, stiowing flow diversion following gravel removal operations, September 1975. b) Willow Creek as it entered Penny River on 20 March J979. Figure 58. Potential overwintering area at Willow Creek. Tti i s spring-fed tributary, open throughout the winter, had pre- viously entered Penny River at a deep pool. 192 — S -a — — S 13 - O d z E 10 ID T3 d Oi r^ — E O ^ T3 E c 3 O C U 193 0) o — U J£ *- Q. E — U O 0) u — » -D - C 0) -- — 0) — cn E >- Q. Qj (0 — £ o o Z i- (0 c S c "D — - the outlet was flowing at that time. By March all flow in the pit had ceased and the pit and outlet were frozen to the bottom. The spring-fed tributary, Willow Creek, however, remained open and flowing into March, but fish were not detected either in the tributary or in the Penny River downstream from where the tributary entered the mined area. At West Fork Tolovana River, the outlet was blocked at the time of the first winter visit, 29 November 1979, because the deep, low velocity arm connecting the pit to the river was frozen and the other arm flowed through a beaver dam. Flow out of the pit through the beaver dam persisted through March (Figure 59). Fish were not detected during any of the winter visits. There was sufficient water and dissolved oxygen to support overwintering fish in mid-March 1979 and the persisting outflow through the beaver dam indicates the pit may be receiving some intergravel flow from the river. The Tanana River-Downstream pit was visited only on 6-7 March 1979; fish were not captured but as emigration after the previous September visit was not possible, fish were probably present. The dissolved oxygen should not have been depleted because of the depth, limited phy top I ank ton pro- duction, and absence of littoral vegetation, and, in fact, was 6.0 mg/d- in March (Table 24). At the two Tanana R i ver-Upstream pits, a more dynamic pattern of freezing was observed. On 27-28 November 1978, the connection between the two pits was frozen solid, thus isolating the upper pit. The surface of the ice in the upper pit was approximately 1.5 m higher than the surface of the lower pit. A burbot and possible lamprey were observed with an underwater television system. The outlet of the lower pit was open to the Tanana River with a school of juvenile salmon and two species of whitefish holding in the outlet current. Burbot were captured by setline in the lower pit. On 6-7 March 1979, the ice surface of the lower pit had risen to the level of the upper pit and the connection between the two pits was open, approximately 30 cm deep and flowing at about 0.1 m/sec into the lower pit. The outlet to the lower pit was frozen solid. Dissolved oxygen at the upper pit had increased from 3.4 to 6.0 ppm between November and March. Fish were not detected in either pit in March. 195 •^ V a) Flow out of beaver dam at pit outlet, 29 November 1978. HjB™ V ■ -1 1 ■ vf. rPt^ '/V ^ / v^ Jr^ TtK^ 1 tt flA /^ ' P J ■■^-H "T 'j(i" > -f \ "' A ■UK" ^ iP ■ -A., ■■* % f i b) Deep pool (>l m) witti thin ice cover approximately 50 m downstream from beaver dam, 15 March 1979. Figure 59. Creation of a potential overwintering area at West Fork Tolovana River downstream from pit. 196 The above observat i ons indicate that after November the outlet froze, then the side channel of the Tanana River adjacent to the pit started flow- ing through gravel into the upper pit, opened the connection between the two pits and flowed back into the side channel through an intergravel pathway. The raising of the surface of the lower pit appeared to have been caused by overflow on top of the existing ice and snow. Oxygen depletion was a poten- tial problem at the upper pit because of the dense stands of aquatic vege- tation (the March 1978 dissolved oxygen was 3.2 ppm) but these were absent in the lower pit and the dissolved oxygen was consistently higher than that of the upper pit. The net effect was the creation of one and possibly two overwintering areas, depending on the minimum winter oxygen levels at the upper pit. Assuming an adequate water depth, the main factor determining the suitability of a pit as an overwintering area is an adequate level of dis- solved oxygen through the winter. A pit with sufficient depth for over- wintering but with an extensive, heavily-vegetated littoral area may ex- perience an anoxic period following the initial snow cover. Barcia and Mathias (1979) found that winterkill in eutrophic prairie lakes was closely correlated to the mean depth of a lake and developed a method to estimate the potential for winterkill based on the initial oxygen storage, rate of oxygen depletion and the mean depth. The critical mean depth for the lakes studied was approximately 2.0-2.5 m. Lakes with an average depth less than 2.0 m experienced regular winterkill, lakes 2.0-2.5 m experienced occasional winterkill, and lakes with an average depth greater than 2.5 m generally did not experience winterkill. The indications were that a productive pit with an average depth of less than 2.5 m may have marginal uti I ity as an over- wintering area, especially during years of early heavy snowfall. The upper Tanana R i ver-Upstream and West Fork Tolovana River pits had the characteristics to fit this type of pit (Table 14). The 6 m deep area in the latter pit may have provided sufficient volume to maintain a suitable dissolved oxygen level, but both of these pits should be considered marginal overwintering areas. Intergravel flow from the adjoining river, however, adding a continual supply of oxygenated water, could maintain sufficient 197 oxygen levels throughout the winter. The lower Tanana R i ver-Upstream pit did not contain a great average depth, 1.7 m, but the lack of littoral vege- tation reduced the probability of oxygen depletion. The water in the pit was turbid during the summer, limiting production of aquatic vegetation. The lower pit maintained higher dissolved oxygen than the upper pit during the winter (Table 24). The Dietrich R i ver-Upstream and Tanana River-Downstream pits both contained deep, clear water regions and did not have well-devel- oped littoral vegetation. Oxygen levels probably remained high through- out the year. The depth and lack of productivity combined to make these two pits excellent overwintering areas; the same features limited their value as rearing areas. There are other possible effects of grave! removal on overwintering areas, but they are difficult to assess because of the absence of data on the study sites before gravel removal. A primary effect is the loss of overwintering areas due to diversion of flow from an original channel, as occurred at four sites (Penny River, Dietrich River-Downstream, Middle Fork Koyukuk R i ver-Upstream, and Middle Fork Koyukuk River-Downstream). In these cases, complete or partial diversion of flow could lead to loss or reduction of overwintering habitat. Another effect is the loss of overwintering hab- itat due to increased braiding and the associated changes — loss of pool- riffle sequence and reductions in depth and velocity which promote rapid freezing. In some areas, gravel removal created or aggravated the formation of aufeis fields, thus leading to a reduction in water available for over- wintering downstream (Washington Creek, Oregon Creek, McManus Creek, pos- sibly some of the North Slope sites). EFFECTS OF HABITAT ALTERATION ON AQUATIC MACRO I NVERTEBRATES Observed Effects on Density and Species Assemblage Habitat alterations expected to affect assemblages of riffle macro- invertebrates would be changes in velocity, substrate, depth, and water quality. During the present study, habitat alterations resulting in a change of each of these parameters were accompanied by changes in the riffle com- mun i ty (Tab I e 25 ) . 198 +- o •— c - M- 4- o • — c m ro o c 0) .— 0) s_ 4— ■D ro ro cn c a> c — ->- g (/) ro O Q) J= W i_ (/) ro 0) O ro 1- ro X o E ro 1- 0) cr> c c c o — ■>- — E ro -I- •«- ro >- — 1_ JD ro -1- X3 X — 0) ro (n D ro o ro ■a Ul D tu l- ro -♦- ■D C •— 01 X) o -— :^ Q. JZ E u LU ro i- •^ m (U ro ■V "D ro .— — E D O E c Ol o >~ 1_ c -— -— -C u o 0) -(— 01 4— cm M- U) M- m c =1 ^ Z) n .— ro D ro 3 -o ■o If) "D in ro 01 0) CD 0) 0) i_ en — C en — JD ro ^ .— ro n 0) ro ■D 01 ro ^ L -t- • — i- 4— -(— O m ro o in — c c L. c c • — — D -Q — D in 0) c 3 0) 0) 1_ o c o 01 o 0) 0) ro i_ c 0) — E 4- 0) 0) ^ ro Q. m LU 0) ro •^ * ■o ro ro c c =3 Q. Q- a. ro ro ._ u O 1- 0) ro ro ■- E o c o ^ ro in ro ^ D i- Zl 1_ Zl O o O E 0) E Q) Q- 0) Z LU z ro m s_ ^ 01 ro CL — O i_ in 0) — Q- ro •s 1- ro ro 4— Q_ 0) ro jz ro o 1_ O 3 O) O — E — 0) O 2 (_) ro •. ■V ro 01 E ro O c in .— o 4- in en ^ en Q) c 0) 01 c J3 n ^ XI X3 ro .— ro ro 4— ro 4- 4— ro in s_ in 05 I. c n c C XI 3 ^ 3 3 ^ ■o 0) "D 0) ■o 0) 0) 4- 0) 4— 01 4- in ro in ro in ro ro i- ro i_ ro 1- in 0) 4- Q) 4— 01 .— 1. in L. in L. in 0) o jn U n U XI c 3 c 3 c 3 3 — in — in — in ro 1- Q) -Q 4- E m 01 3 4- 0) CO Q. c 3 01 3 < 1/1 — ) c CL ro T3 I. 0) - Q) 0) ro (0 T3 c — -•- 3 (U Q. (0 — CD H 3 E - O) ro >- 4- c 0) — ro u 0) ro "O - E " o c o 1. X 1/) u •^ 3 o ro 1- en — o •— I. 0) — 0) a. O Cl LU in cn 3 < 0) L. Q- 0) ro ro — Q. ro i_> ro E •^ O ro in u o 3 in o in E o 0) — z CD 0) X3 E 01 4- Ql Q) m (U (□ "O o •V . — in 1_ 3 01 i. X o 4- 01 < CL ro ^ — ro — — 01 1- 1- 01 0) CL E O 0) — SI — CL < LU ■^ •, X ro ro l_) 4- • — ro 0) >~ ■V E ro E ro O x: O ■— in o m c o O — ro in CD — 4— in — u ro o — o CL O LU < O O) ai OJ c 01 c cn c J3 T3 JD "D ro — ro .— ■o J- ro 4- ro U) 1- in I. ro C XI c XI 1- 3 3 XI ^ •^ "O fl) TD 01 T3 Q) ■•- 0) 4— 0) in ro in ro in ro i_ ro i_ ro Q> 4- 0) 4— 0) 1- i/i i_ in 1- O XI u XI o C 3 c 3 c — in — in — > 0) E o "D 01 3 C c o u 199 x> 03 D C C O u ITl in c 0) c — -^ g I/) ro o 0) JI in i- in ID Q) o (0 1- (0 X o E to 1- 0> CD c c c o — ■•- — E (0 -1- -t- ID >- — 1_ X3 ID -I- "O X — Q) (D in D ID O ^ (D (D — (D 3 C S — CD 4— ::^ Q) c ID •— CD (_) Q) C o CD C (D JD "O 0) in (D Q) I. o c ■^ J^ — o in Irt >- .— "O 01 D ^*- 4- D 1/1 < Q) (D "D 13 01 Q- C 0 ID XI X) Q) in CD 01 1_ o in cn D < L. Q) > I. 10 Q. x: Q- 01 C E OJ (D ■o . D ID E c CD 01 ^ CD c O •— 1_ u x: -1— >, . — 01 cr 0) 01 (D (D •O X) E - O — C 3 O E 1_ — (D in C - 3 .— ID i_ 4— — -»— Q) — c ro 01 0) CD u o - (D 01 (D 01 — ~ 0) -a CO E E (D — ^ O O c — o c c ._ — o o o -t- 3 CD i_ 1- Q) 0) E C (0 — — ^ s: o CD to o u <_) z ■o CD 01 0) c 0) in ._ — CD n "O n cu ID .— CD \- 4— CD -t- o in i- U) c c J3 c — 3 — 3 ^ Q) •V •D 01 C "O 01 0) A- c 0) -»— C31 m (0 CD in CD c CD 1- ^ CD 1. •— 0) O 0) ■♦- ■o u in 1- in — o JD g o JD ID c 3 lU c 3 i_ in z — in n I. 0) ^ > Q) .— CD cc 1- O :^ O 0) A~- ji: ^ ID 1- o •— 1_ c 0> ID 4- > Q) CD 0) CD Jt CD in CO 01 (0 -D I ■- O — C 3 0 Q- 1_ — — I— sz 0) •- CO 01 n CO — c — *- 3 cu E CO — CD CO 0) (O O O cn CD C CD n X3 01 Cfl CD 0) 1_ o c O) CD - E CD c O- (□ u 01 c o 0) c o 0) c o 3 E >. c U 01 CO - -a ID 1_ 3 o E CD 01 CD X3 3 E 01 ID "D 3 E ■o 01 c o u ID i_ 1 en 1 1 -»— CO c ID ID in -t— CD ■*- CD XI — ■o — c — c 3 in in ID in ■D in •D c l_ • — — X) o -o XI ■D ID ■o CD 01 01 01 V. 01 1_ in 4_ m ■^ in x> cn XI CO CO CO C ID ID 01 1_ CD o 01 -> 0) ■• i_ 01 L. .— 1. c v. c u O -»— u o u o c — c ID c •— c — CO — ■♦- — +- — -1- in -«- 3 in ai 3 D cn < 3 < i_ >. 0) >~ — 0) i_ > — 1- 4- 01 3 CD CD > cr —3 01 — cn > q: ' 1 CO o x: — -^ in > •— ,_ ID > > H CD CO ■^ 200 "D T3 O c o >- •4- -1- o — c U) (0 0 c 0) ■— (U L. -1— ■o (0 It) CD c - - E (a -1- ■t- (O >^ — 1_ J3 n 0) (D -1- X) X — Q) (D tn D (0 u (0 (0 1- cu JZ Q- Q- O O — O — - -C ^ q: 0) ro (D — -C Z! u E o en en ^ — c O u (/) ro •— i_ 0) 4- M- (/) u -O (0 3 U) X) Ji: -*- <_> c/) w 3 >~ C ■o ro D S (D (D S_ — 3 C o ro E 4— Q) ID Z Q. < 0 (0 ■o Q) (D _C CD a -C O o >- ID O O cn ^ ID +- m c 3 XJ CD 0) 4- in ID ID 1_ a> -1— u 1/) o JD c 3 — in 3 Q. O i. -4— C Q O 0) ID Q) "D ID — X> E — O — C 3 O Q- U — U - - ID ID ID 1_ 3 Q> E Q- cn O >- — c < u ■o o -C u — c >- in CL 0) CL ID 1_ ID CL 0) — ID JZ -C CL. u o o o o cc JD ID -4— in c 3 ■D Q) 0) 4- in ID ID l- (U 4— 1- in o ^ c 3 (U n E 0) 4— Q. a> to cn c Q) > C ID ID c x> > O > c I. o in 0) in c ID 1_ XI 0) i_ 0) > o ID E S o >- 1_ Q) > ID "D 0) in to (D 201 Response to Substrate Alteration. The two types of substrate alter- ations observed during the study (a shift to unstable substrate and change from laminar to turbulent flow) significantly affected the total numerical densities of aquatic macro i nver tebrates in the mined area as compared to undisturbed areas (Table 26). At Washington Creek, Oregon Creek (June and August), al I Penny River, Kuparuk River, and McManus Creek (May) site visits, macro i nver tebr ate densities in mined areas were significantly less than those in the upstream area. At al I five sites there was a shift from a moderately compacted gravel substrate to a very loose, unconsolidated sand- gravel substrate (Table 25). A similar habitat change at the Sagavan i rk tok River and Ivishak River sites resulted in a significant increase in the density of aquatic macro i nver tebr ates. In five of the eight cases in which there were total density decreases, there were density reductions in the ephemeropt eran genus Cinygmula while in seven of the eight cases, there were reductions in the dipteran family Ch i ronomi dae. The density increases at the Sagavan irk tok River and Ivishak River sites both contained density increases in the ephemeropt eran subfamily Baetinae and dipteran family Ch i ronomi dae, as we I I as some other taxa. At two sites there was a change from laminar flow to turbulent flow caused by substrate alteration. At both Ugnuravik River and Shaviovik River sites, there was a significant decrease in total macro i nver tebr ate density, primarily because of a decrease in Simuliidae densities. At Ugnuravik River, the laminar flow was in the upstream (control) area, while at Shaviovik River, laminar flow occurred in the mined area. At three of the five sites where there were decreased densities in the mined area (Washington Creek, Oregon Creek, McManus Creek) there were also aufeis fields associated with the mined area (Table 25). All three sites were visited early in the summer so that any aufeis effects would have been measured at their greatest magnitude. Later visits at two of the sites (Oregon Creek, McManus Creek) indicated that densities in the mined area increased to levels similar to those in the upstream areas. At Oregon Creek, the summer recovery from aufeis effects was not complete for population densities of Nemoura and Cinygmula, which remained below the densities 202 c^ ir> r- \D r^ t^ .. ^ .. * ^ * «o t VO ^ VO — — O K^ o o -a- o o o o o o — in '^ in in in ^ — in 03 CD (D in in — r- vo CTi — CD CN — ^ Oi — CN CN O o CN o o in in CN o r"' lo o d t oi m in •^ L-1 CO 1 o fo r~- r- O o in r^ — CNJ r^ o o o — o o Qj — — (a II in 3 in l_ C Q) 01 ID u C >- ^ o 0- (D ID ;_ CO m (D (D i- 3 0) E 1. QJ 0) 2 E (P II £ Q. H-» LU O- - ■— c o u 2C3 reached by the same genera in the upstream area. The August and September population densities of Capnia and Baetinae, however, exceeded those re- corded in the upstream area. At McManus Creek, the mined area densities of Oligochaeta and Rhyacophila did not reach those recorded in the upstream area; the mined area densities of Alloperia, Ch ironom i dae, and Tipulidae exceeded the up- stream area densities on each of the two succeeding trips. The failure of the mined area densities of some taxa to reach upstream densities, while those of other species exceeded the upstream densities, indicated that there was a long-term habitat alteration which has led to an alteration in species composition of the mined area. Another site which showed a similar response, but where an aufeis field was not identified, was the Penny River site, where mined area densities of Oligochaeta, Nemoura, C i nygmu I a, Ch i ronomi dae, and others were general ly lower than upstream densities. In the Penny River mined area, population densities of Tipulidae and, at times Capnia, Baetinae, Ephemerella, and Athericidae were higher than those in the up- stream area. The shift in taxa at the above sites appeared to be related to the occurrence of unstable substrate possibly aggravated by an aufeis field. Other sites with a similar substrate alteration (Washington Creek, Kuparuk River) also showed density reductions of most organisms but the site was only visited once and this precluded any analysis of recovery or sea- sonal patterns. At Kuparuk River, densities of al I species were lower in the mined area than in the upstream area while at the Washington Creek upper mined area, only Tipulidae densities exceeded those in the upstream area. In summary, certain taxa, primarily Oligochaeta, Nemoura, Cinygmula, and Chironomidae were reduced in areas of unstable substrate while others, primarily Tipulidae, but also Capnia and Baetinae, showed increased den- sities. Response to Increased Braiding. Aquatic macro i nvertebr ate responses to these alterations were colonization by taxa which are more suited to lower velocity waters with higher organics. Clinging ephemeropt erans, as found in the family Hept agen i i dae (Cinygmula, Epeorus), were replaced by sprawlers 204 and climbers, e.g., Baetidae. Tr i chopferans often increased in these areas and the dipteran family Tipuliidae was often associated with the finer sediments found in mined areas. At two sites on large rivers showing in- creased braiding as well as altered substrate ( Sagavan i rk tok River and Ivishak River) there was an increase in the density of virtual ly al I taxa in the mined area as compared to the upstream area (Table 24). The riffles in the mined area in these two cases were in sma I I shal low channels with exten- sive riffle area while the riffles in the upstream area were in large chan- nels, were less extensive, and composed of a more coarse material. The riffles in the mined area had greater detrital accumulation, and the de- creased depth and velocity associated with the braided areas may have allowed greater periphyton production. Such a situation would increase the quality of the habitat for most of the species unless a critical parameter, such as velocity, had been lost or altered. The increased braiding at other sites, such as Oregon Creek and Penny River, may have contributed in a similar manner to the altered species composition. The increased braiding at many of the sites led to changes in the water temperature and dissolved oxygen in the mined area. An examination of the seasonal variation in the riffle macro i nvertebrates at Aufeis Creek revealed a pattern of density changes which indicated a possible effect of the al- tered temperature and dissolved oxygen regime on the apparent densities of certain macro i nvertebrates (Figure 60). In the ephemeropt eran taxa, Baetinae and C i nygmu I a, the densities in the upstream area increased from the July to August trip while those in and below the mined area decreased. Simuliidae densities decreased between the two trips in the upstream area with simu- liids absent in and below the mined area in August. The temperature at the area between the two mined areas was 2.8 C (July) and 1.2 C (August) higher than that in the upstream area. The immature stages of the three taxa ap- parently emerged earlier in the areas affected by gravel removal than in the unaffected upstream area. The altered water quality parameters may have altered the emergence times of these three taxa because temperature and dissolved oxygen can affect developmental rates (Hynes 1972). 205 5000 1000 500 100 E 50 in E < — i- t — t I I I I I I I I I I I I I «£> I I I 9i I I I t I I I III il^ I I I I I I <7i O O « — E in 3 « Ifl 3 ■ L. o « HI 3 E <3 o a C ai U) m i!) o » o _ ■ O O I. — — (D l_ _ 0) o — •— (D X ^ (A a. a. o ■ — — Q. O *- Q. V O ■• . — O 41 >^ O -I EU — OI— -(( tfil = ^ ol a u « trt c c o Q> CT U a> o » £: O o a. CL c ^ a 3 >■ >• i. o 0.0 U Ui tn ■ o — ._ o SUMMARY AND CONCLUSIONS EFFECTS OF GRAVEL SCRAPING ON RIVERINE HABITATS Gravel removal by scraping in floodplains resulted in a number of alterations to aquatic habitats with the biota showing a variety of re- sponses to these habitat alterations. Important habitat alterations in- cluded: (I) the creation of braided channel areas with associated changes in various habitat parameters, (2) removal of bank and instream cover, (3) increased habitat diversity, (4) creation of potential migration blockages, and (5) creation of potential entrapment areas. Increased Braiding Characteristics This habitat alteration occurred at 15 study sites where active channel deposits were scraped to below the water line or where flow subsequently increased to inundate the mined area. The main effect of braiding on spe- cific habitat parameters was to reduce velocity and depth by spreading flow over a wider area. The populations of both aquatic macro i nvertebrates and fish utilizing these areas were altered with shifts in species and life history stages. The reduction in velocity led to increased detrital accum- ulation, deposition of fine materials, and often altered the temperature and dissolved oxygen regime. The altered temperature regime led to altered emergence periods of aquatic insects; the effect of this alteration on reproductive success and overall population stability is unknown. Fish populations responded to increased braiding in a number of ways, but the general pattern was a reduction in the diversity of the fish com- munity. The number of species and age groups usually decreased in the braid- ed areas. 209 The increased braiding also increased the probability of aufeis forma- tion in the mined areas. This effect was documented at Washington Creek and Oregon Creek and was indicated at McManus Creek and Penny River. There may have been additional ice formation at some of the North Slope sites, such as Kuparuk River, Sagavan i rk tok River, and Ivishak River. The formation of aufeis fields seemed to prolong the recovery of the site as the channels and substrate remained unstable and siltation persisted during the melting process. In addition, the water needed to create the aufeis field became unavailable downstream, thus reducing water available for overwintering, often the factor limiting fish populations in arctic rivers. Removal of Bank and Instream Cover Reduction of bank cover occurred whenever a portion of incised or undercut bank was removed. At sites with this habitat alteration, the bank was scraped to remove overburden in order to access underlying gravel de- posits. The former bank with cover was changed to a gravel bar following removal operations. Certain species, such as Arctic char and Arctic grayling were strongly associated with bank cover and the loss of this cover led to reduced population densities in the mined areas. Similarly, loss of instream cover led to reduced densities in mined areas. Increased Habitat Diversity Habitat diversity increases were documented at three scraped sites, but these were viewed as temporary increases at newly inundated sites. The habitat diversity will decrease as braiding characteristics are established, the channel cutoffs are completed, and the habitats become more uniform. Migration Blockages The combination of increased wetted perimeter and decreased depth in mined areas created a situation that could lead to migration blockages during periods of low flow. Such a situation occurred at the Aufeis Creek site and possibly could occur at the Nome River site. The potential for 210 migration blockage was present at sites, including Oregon Creek and Washington Creek, where the entire active channel was scraped. Because of the known complexity of fish movements throughout arctic watersheds, migra- tion blockages can have a significant, but as yet unstudied, effect on popu I at i ons. Potential Entrapment Areas The potential for fish entrapment was high at areas with extensive backwater, as was found at newly inundated areas (Dietrich River-Downstream, Middle Fork Koyukuk R i ver-Ups tream) and areas with increased braiding (many sites, including Sinuk River, Kuparuk River, Sagavan i rktok River, Ivishak River, and Middle Fork Koyukuk River-Downstream). At these sites, areas of ponded water became isolated from the active channel as the water level dropped, trapping fish and invertebrates that had moved or been carried into these depressions during the high water. Mortality of stranded fish and invertebrates is assumed to be high because they are subjected to high summer water temperatures, low dissolved oxygen, increased predation from terrestrial predators, winter freezing, and total loss of aquatic habi- tat as the isolated pools often dry up if the river continues to drop. EFFECTS OF INUNDATED PIT FORMATION ON THE ASSOCIATED RIVER BIOTA The direct effects of pit excavation on the river biota were difficult to assess because the river habitat was not directly affected; inundated pits were created from previously terrestrial habitat. Because of this, the pits represented a new habitat and the fauna inhabiting the pits was con- siderably different from that inhabiting the associated river. Summer Utilization by F i sh Two of the pits, Dietrich Ri ver-Upstream and Tanana River-Downstream, were deep clear water pits with low productivity and fish utilization. At Tanana River-Downstream this low utilization was easily explained because there was no connection to the river and immigration into the pit occurred 21 only at infrequent h i gti water levels. The Dietrich Ri ver-Upstream pit, however, was connected to the active channels but fish were apparently not utilizing the pit for feeding. Benthic macro! nvertebrate densities in both these pits were low when compared, to those of other pits. The spring-fed channels upstream from the Dietrich River pit were utilized by adult Arctic grayling and the pit itself was reported to be an overwintering area. All other pits were highly productive and heavily utilized by fish as summer rearing areas. The shallow pits. Penny River, Prospect Creek, and Jim River side channel (this site had some characteristics of a pit) supported high densities of juvenile salmon (coho in the Penny River, chinook in the latter two) as well as some species associated with both a lacustrine environment (Alaska blackfish, burbot, northern pike) and stream environment (round whitefish, Arctic grayling, slimy sculpin). The productive Southern Interior deep pits. West Fork Tolovana River and two Tanana R i ver-Upstream, contained a more lacustrine fish fauna with northern pike dominating the fauna and humpback whitefish, least cisco, and burbot also present in the Tanana R i ver-Upstream complex. Potential for Winter Mortality and Winter Survival Areas The creation of shallow pits and subsequent heavy summer usage by fish created the possibility for entrapment during freezeup and subsequent winter mortality when the pit freezes solid or decay of vegetation consumes the dissolved oxygen. The pattern of freezing observed during winter studies indicated that during the year of observation, entrapment was minimal and probably not a significant problem. The creation of deep pits connected to the river could create over- wintering areas; this was documented or suggested at several study sites. All pits studied, with the exception of Tanana River-Downstream, however, had a mean depth insufficient to preclude winter mortality. Intergravel flow appeared to maintain the ability of some pits to support winter fish sur- vival, but this is an unpredictable factor in the design of pits. 212 RECOMMENDATIONS I. It is recommended that mining practices leading to an increased braided configuration be avoided. This is best achieved by avoiding active channels and by mining above the water table. 2. Undercut and incised vegetated banks should not be altered. 3. Critical habitats, such as spawning and overwintering areas should be avo i ded. 4. Formation of isolated ponded areas that cause entrapment should be avoided by contouring and sloping to provide drainage. 5. Pits should be excavated to a sufficient depth to preclude winter mor- tality. Generally, a mean depth of at least 2.5 m should ensure winter sur- vival . 213 REFERENCES Alt, K. T. 1970. Sheefish and pike investigations of the upper Yukon and Kuskokwim drainages with emphasis on Minto Flats drainages. Alaska Dept. of Fish and Game. Fed. Aid in Fish Restoration, Annu. Prog. Rept. 1969-1970, Proj. F-9-2, 11:32 1-330. Barcia, J., and J. A. Mathias. 1979. Oxygen depletion and winterkill risk in small prairie lakes under extended ice cover. J. Fish, Res. Bd. Canada 36(8) :980-986. Binns, N. A., and F. M. Eiserman. 1979. Quantification of fluvial trout habitat in Wyoming. Trans. Am. Fish. Soc. 108 ( 3 ): 2 I 5-228. Bovee, K. D., and T. Cochnauer. 1977. Development and Evaluation of Weighted Criteria, Probab i I i ty-of-Use Curves for Instream Flow Assess- ments: Fisheries. Instream Flow Information Paper I^o. 3. Coop. Instream Flow Serv. Group, Fort Collins, Colorado. 39 pp. Cheney, W. L. 1972. Life history investigations of northern pike in Tanana River drainages. Alaska Dept. of Fish and Game. Fed. Aid in Fish Res- toration. Annu. Prog. Rept. 1971-1972, Proj. F-9-4, 13:1-30. Hynes, H. B. N. 1972. The Ecology of Running Waters. University of Toronto Press, Toronto, Canada. 555 pp. Nie, N. H., C. H, Hull, J. G. Jenkins, K. Ste i nbrenner , and D. H. Bent. 1975. Statistical Package for the Social Sciences. Second Edition. McGraw-Hill, Inc. 675 pp. Stalnaker, C. B., and J. L. Arnette (eds.) 1976. Methodologies for the Determination of Stream Resource Flow Requirements: An Assessment. Utah St. Univ., Logan, Utah. 199 pp. Woodward-Clyde Consultants. 1976. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska. U. S. Fish and Wildlife Service. FWS/OBS 76/21. Wash. D. C. 127 pp. Zar, J. H. 1974. Biostatical Analysis. Prent i s-Ha I I , Inc. Englewood Cliffs, N. J. 620 pp. 214 EFFECTS OF GRAVEL REMOVAL ON TERRESTRIAL BIOTA M. R. Joyce INTRODUCTION The ecological importance of floodplain and riparian terrestrial habi- tats in temperate regions has been well documented in the ecological litera- ture. These habitats, particularly the riparian zones, have high primary and secondary biological productivity and typically support a diverse and abun- dant flora and fauna. These biotic zones frequently provide temporary and permanent refuge for many of our rare and endangered species. The signifi- cance of these floodplain and riparian habitats has recently been recognized and incorporated into the management plans of several Federal agencies (Johnson and Jones 1977; U.S. Army Corps of Engineers 1979). Arctic and subarctic floodplain and riparian habitats are no less significant in their importance and ecological value. The riparian zones develop dense shrub thickets dominated by willows and alder in all four study regions. Overstory forest dominated by white spruce and paper birch also frequently inhabit the riparian zones of the Northern and Southern Interior regions. (Scientific nommenc I a t ure for terrestrial flora and fauna is presented in Appendix A.) High primary productivity in these zones pro- vides optimum feeding, nesting, and cover habitat for a diverse fauna usu- ally dominated by small mammals and passerines. These riparian habitats in interior Alaska frequently support over 100 birds per 40 ha during the nesting season (Spindler and Kessel 1979). Some birds, such as the yellow warbler and northern waterthrush, very seldom nest in habitats other than riparian shrub thickets. These zones also are preferred habitats for tundra voles and singing voles. The more dense riparian shrub thickets provide critical feeding and cover habitats for moose and ptarmigan during winter. 215 The unvegetated and sparsely vegetated areas within arctic and sub- arctic floodplains provide equally valuable habitat for a different segment of fauna. Many of the major floodplains provide key migratory corridors for large numbers of waterfowl, shorebirds, and caribou moving to and from wintering zones and summer nesting and calving territories. Unvegetated areas of larger floodplains are used as prime nesting and feeding habitat by numerous shorebirds, gulls, terns, and waterfowl. The delta areas of larger rivers also are prime Juvenal rearing habitats for shorebirds and waterfowl. Along coastal regions, these river deltas also are key nesting sanctuaries for geese, brant, swans, gulls, terns, and shorebirds, and during late summer and early fal I they provide protected habitat for large concentra- tions of molting waterfowl. Due to the high secondary productivity of these areas, predators including bears, wolves, eagles and jaegers also frequently concentrate their feeding activities along floodplains. Unfortunately, from a biological viewpoint, floodplains also provide easily accessible gravels that are available in large quantities and fre- quently close to development sites. As previously noted, arctic and sub- arctic conditions, primarily associated with the presence of permafrost, place large demands upon gravel resources by all development projects. During the construction of the Trans-Alaska Pipeline System, over 3,300 ha of unvegetated floodplain habitat and approximately 1,000 ha of riparian habitat were affected by gravel removal operations (Pamplin 1979). The proposed construction of a gas pipeline through Alaska, depending upon final route selection and the degree of use of existing construction facilities, could require similar gravel supplies. Other development projects are expec- ted to increase the future demand upon gravel resources. Previous to this study, natural resource managers had little indepth knowledge, relative to arctic and subarctic terrestrial floodplain eco- systems, of how to best mitigate the use of floodplains as gravel removal sites. The short-term effects of gravel removal operations were believed to be associated with reduction of habitat, probable decrease in local fauna population sizes, and potential indirect effects through reduced habitat quality in adjacent and downstream habitats. However, the variations in the 216 levels of influence and the durations of influence between differing grave! removal sites and methods of operation were not completely known. Also, there were no data on long-term effects in the arctic or subarctic. Factors such as the size and location of the site, and the characteristics of the stream and floodplain were believed to be influencing parameters, but their relationships to short-term and long-term detrimental effects were not understood. To help answer these questions, a terrestrial study was incorporated into this project. The study was designed to be compatible with the hydrol- ogy and aquatic biology programs and organized to provide answers on: (I) the degree of flora and fauna change resulting from gravel removal opera- tions; (2) the rate of habitat recovery at disturbed sites respective to the characteristics of the gravel removal operation and the characteristics of the river and floodplain system; and (3) how the detrimental affects of gravel removal operations could best be mitigated. 217 METHODS OF DATA COLLECTION As previously described in APPROACH AND METHODOLOGY, terrestrial data were collected at all 25 study sites, with individual site visits occur- ring either during the summer of 1976, 1977, or 1978. Standard procedures were used to collect field data on flora, soils, birds, and mammals. Site locations are identified on Figure I. Sites occurred on the Seward Peninsula, North Slope (in both the coastal plain and Arctic foothills). Northern Interior (between the Brooks Mountain Range and Yukon River), and Southern Interior (between the Alaska Mountain Range and the Yukon River). One study site, selected as being most representative with respect to river type and biological conditions in each regional study area, was sampled during a 5-day visit. We attempted to coincide this visit with the peak of the avian nesting season. All other sites were surveyed during a 3-day visit. Within each region, the 3-day visits were spaced throughout the spring, summer, and fall to measure seasonal fluctuations in species compo- sition and abundance. The selected approach to meet the objectives of this project was to document the presence and establish the habitat relationships of the flora and fauna of the disturbed area and compare these to pred isturbance flora and fauna populations and habitat affinities. A control area which was most representative with respect to physical site characteristics (i.e., inside or outside meander) and habitat characteristics (i.e., dense riparian shrub thickets, or unvegetated f I oodp I a i n ) was selected to establish pre-gravel removal biological conditions and flora-fauna relationships. In addition, surveys were conducted in floristic serai stages representative of the disturbed area during the time of the field visit, and in serai stages representative of anticipated future disturbed-area vegetative development. 218 These areas were surveyed to identify flora-fauna re I at i onsti i ps during various site recovery stages. The Major Variable Matrix Table (Table I) identifies the variety of sites studied. Study sites varied from large braided rivers to small, single-channel streams located in four major geographical regions of Alaska. Selected sites were studied from 2 to 20 years after disturbance, allowing data gathering on short-term and long-term response and recovery by the terrestrial biota. Characteristics of gravel removal areas included: scrap- ing operations of surface gravels within and adjacent to active channels; scraping in areas separated from the active channels; and pit excavations separated from active channels. This range of sites allowed comparison of the effects of different techniques and site locations on terrestrial biota. 219 METHODS OF DATA ANALYSIS Data analysis initially resulted in the identification of the degree of change in measured parameters at each study site. A numerical rating ranging from 0 to 10 was assigned to indicate an increase (ratings 10 through 6), no change (rating of 5) or a decrease (ratings 4 through 0). These ratings indicate the degree of change at the time of the site visit between the pre-gravel removal conditions (i.e., extent of shrub thicket cover, or number of passerines present) and the post-gravel removal conditions. Each numerical unit increase (6 through 10) or decrease (4 throijgh 0) approxi- mates an alteration similar to a 20 percent level of change in that param- eter. Each site was analyzed to determine how measured parameters (vege- tation, soils, birds, and mammals) interacted, and how they responded as a whole to the Physical Site Characteristics (such as river size and config- uration) and Gravel Removal Area Characteristics (such as type and location of gravel removal). After individual site analysis, all sites were compared to evaluate similarities and differences in the degrees of change in biolog- ical parameters. Fauna directly respond to the presence (and type) or absence of vegeta- tive development, consequently, the degree of change and the rate of re- covery at the gravel removal sites received major emphasis in the vegetative data analysis. Factors that influence vegetative recovery (e.g., soil con- ditions and aufeis development), also were thoroughly reviewed. Selected biological data were subjected to a computerized hierarchical clustering routine to identify similar responses in a measured biological 220 parameter between rivers. This analysis grouped similar sites and similar responses (increase or decrease) by biological parameters. All data were thoroughly reviewed to identify any correlations between Physical Site Characteristics, Gravel Removal Area Characteristics, degree of change by the terrestrial biota, and short-term and long-term recovery rates. The following sections include the results of data collection and analysis. 221 RESULTS AND DISCUSSION Changes in selected terrestrial parameters ttiat were induced by gravel removal are identified in Table 28. These changes were based upon measured levels of variation in each parameter at each site. In general, the degree of both short-term and long-term changes in local faunal communities strong- ly reflected the extent of disturbance to floodplain and riparian vegetative commun i t i es. VEGETATIVE COMMUNITIES OF STUDY AREA FLOODPLAINS Vegetative communities of floodplain and riparian zones at the study sites were typical of those occurring throughout arctic and subarctic regions. In general, the Seward Peninsula rivers and the smaller North Slope rivers usually were meandering or sinuous in configuration with well-defined (incised) outside meander banks (Figure 61). This configuration and profile created a relatively narrow floodplain (30 to 60 m) and al lowed extensive development of mature shrub thickets adjacent to single channel rivers. These shrub thickets usual ly were dominated by Sa I ix a I axensi s. On Inside meanders (point bars) and in more active portions of floodplains (lateral and mid-channel bars) herbaceous, woody pioneer and early willow communities occurred adjacent to unvegetated gravels bordering the river. Meandering and sinuous rivers of the Northern and Southern Interior were similar in pattern and were characterized by extensive shrub thickets with dense stands of advanced and mature successional stage boreal forest communities at the edges of active floodplains (Figure 62). White spruce usually dominated these stands, but paper birch and balsam poplar also were common. Similar pioneer and early shrub successional stage communities occupied point bars and edges of lateral and mid-channel gravel bars. 222 >iaaj3 uBiayj Sn--J9'^ ! a euBuej. SQ-jaAiy Bueuci jdAiy cuBAojoi -d'W >iaajD ioedsojd J9A j y uif SG-*a iinjinAoH -j-w Sn-*d HnnnAox -j-w SQ-jaAia uDiJiaiQ Sn--J«A!y yoij^aiQ jaAiy MIAOIA849 jaAiy iiOMS!A| jdAjy >to;)4J!ueAe6es Maaj^ a^B3ja4ad)(s jaAiy >injBdn>j iiaajQ s la^nv jSAiy HiABjnuBn a a. 0) a. u ffl jaAiy auioN jaAj^ Auudd >(a«J3 uo6ajo )(aaj3 uo46uigsBM jBAia iinujs >(aaj3 una piog r- r-' — I/) Ui c c o « o « o> I. c « — -o L. c o « ■*- u X a. « U 3 — *- • I. l_ 3 « «) O o o i £ u O E V) O £ u E 223 Figure 61. Penny River undisturbed floodplain showing typi- cal North Slope and Seward Peninsula floodplain character- istics of sinuous channel bordered with dense shrub thick- ets with incised outside meander bank, and narrow gravel point bar on inside meander. Figure 62. West Fork Tolovana River showing typical South- ern and Northern Interior medium river floodplain character- istics with shrub thickets and white spruce-paper birch stands along the riparian zone. 224 The larger rivers within all four regions typically flowed in braided or split channel configurations. These floodplains were more hydr au I i ca I I y dynamic, with much wider active areas, and contained advanced serai stage vegetative communities only along floodplain borders and on isolated mid- channel islands. Much of the floodplain in these large, braided rivers contained expansive areas of unvegetated gravels or were sparsely vegetated with herbaceous and woody pioneer or early shrub thicket communities. This very briefly describes In general terms the normal vegetative patterns of floodplains in the area of study. For a more detailed descrip- tion of normal patterns, refer to the "Preliminary Report Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska" (Woodward-Clyde Consultants 1976) and for a detailed description of the vegetative structure which occurred at each study site refer to the Project Data Base. VEGETATIVE COMMUNITY CHANGES AT GRAVEL REMOVAL SITES The observed changes in vegetative communities of the study sites varied from no significant change to long-term loss of habitat. Habitat loss and alteration (both short-term and long-term) repeatedly resulted in signif- icant secondary changes within the bird and mammal populations that inhab- ited study area floodplains. These faunal responses are discussed in a fo I lowi ng sect i on. Significant areas of existing floodplain vegetative cover were removed at 18 of the 25 sites (Table 28). Lost vegetative habitats usually consisted of mature shrub thickets on the Seward Peninsula and North Slope sites, and a mixture of shrub thickets and advanced successional stages of boreal forest floodplain communities in Northern and Southern Interior regions. At all sites these habitats supported a diverse and abundant fauna dominated by passerines and sma I I mammals prior to clearing and gravel removal activ- ities. Refer to the Project Data Base for a complete listing of recorded flora and fauna at each study site. Vegetative habitat removed at these 18 sites averaged 10 ha and ranged from approximately I ha at Gold Run Creek to 35 ha at Dietrich River- Upstream (Tab le 28) . 225 In general, sites separated from the active floodplain frequently disturbed the most vegetative habitat as a percentage of the total disturbed area. For example, Table 28 identifies seven sites that were entirely (100%) vegetated prior to gravel removal and all were separated from the active floodplain. At all seven sites vegetative cover and associated organic overburden were completely cleared prior to gravel removal. Long-Term Loss of Vegetative Habitats Long-term loss of terrestrial habitat occurred at those sites where: (I) the gravel extraction method (either pit excavation or deep scraping) removed gravel to depths that resulted in permanent flooding; or (2) the specific site location and material site characteristics resulted in river hydraulic changes which annually affected the site. Permanently Flooded Material Sites. Eight of the study sites were excavated pits, either totally or in part (Figure 63). Pits varied from an Figure 63. West Fork Tolovana River showing permanently flooded pit excavated adjacent to the active floodplain with a downstream connection. 226 average of 1.5-m in depth at the Penny River to over 7 m deep at the Dietrich Ri ver-Upstream, West Forl< Tolovana River, and Tanana River- Downstream sites. The pits were either connected or unconnected to adjacent active river channels, however, in all cases they were permanently filled with ponded water (Figure 63). Surface areas ranged from 7.5 ha at Tanana Ri ver-Upstream to 0.1 ha at Ugnuravik River. Six of the eight sites were separated from the active floodplain and were completely vegetated with mature white spruce-paper birch and/or willow and alder shrub thickets prior to excavation. At these sites the depth and subsequent flooding created aquatic habitats that led to long-term loss of terrestrial habitats. At the two other pit sites, the excavations occurred in unvegetated point bars (Ugnuravik River) and unvegetated lateral bars (Kavik River). Thus, no vegetated habitat disturbance occurred. Excavation of deep pits, however, was not the only gravel removal method that led to development of permanently ponded water and consequently the long-term loss of terrestrial habitats. The combined gravel removal and site location characteristics at the Jim River and Dietrich River-Downstream sites also led to permanent ponding. At the Jim River, gravel was scraped from within and immediately adja- cent to a high-water channel. The resulting profile at the completion of the scraping operation resulted in an almost circular depression in the middle of the worked area. The high-water channel traversed this depression. Since this channel carries summer flow, it consequently had formed an annually ponded area of approximately 4.5 ha over this centrally depressed portion of the II ha site. Before clearing and gravel removal, with the exception of the approximately lO-m wide high-water channel, this site contained a di- verse complex of mature and intermediate-aged white spruce-paper birch stands with scattered willow and alder thickets. The Dietrich River-Downstream site was scraped to an average depth of I to 1.5 m in a rectangular shaped 7.5 ha. The area was separated from the active floodplain by approximately 150 m prior to the activity. However, the depth of excavation was the probable cause of a permanent channel change by a major side channel of the Dietrich River. This channel entered the pre- 227 viously dry site during the second spring breakup following the activity. This channel change caused flooding of approximately 90 percent of the material site. This condition will remain as long as this side channel flows through the site. Thus, at both the Jim River and Dietrich River-Downstream sites, mining depth and site location characteristics also created permanently ponded aquatic habitats which will lead to long-term loss of terrestrial habitats. Annual Hydraulic Stress. In addition to the creation of permanently ponded sites, long-term loss and alteration of habitat occurred at sites where the gravel removal operation resulted in significant changes in river hydraulics. Examples of such changes include shifted channels, annually flooded sites, and aufeis development within the material site. On the Seward Peninsula, the Penny River and Oregon and Washington Creeks are small rivers with relatively narrow, densely vegetated flood- plains. Penny River and Washington Creek flowed in a sinuous configuration, while Oregon Creek flowed in a straight configuration. The portion of the total disturbed area which was vegetated by dense, mature shrub thickets prior to disturbance at each site was extensive (Oregon Creek 65 percent; Penny River 80 percent; and Washington Creek 85 percent) (Table 28). At all three sites, the working area (which was scraped to a level equal to or slightly below normal water levels) extended across the entire floodplain and at Washington and Oregon Creeks the disturbed area extended approxi- mately 9 to 15 m beyond the floodplain banks and into the adjacent shrub- tussock tundra. The resulting effect of these scraping operations created: an unvegetated, flat floodplain which was 2 to 3 times wider than upstream or downstream reaches; a floodplain that was equal to, or only slightly higher in elevation (10 to 20 cm on the average) than normal summer flows; and a wider channel with increased braiding, straighter configuration and shallower flow (Figure 64). The effects of these induced hydraulic changes created direct impedi- ments to vegetative recovery and thus they also resulted in long-term altera- tion of the habitat structure of the disturbed reach in these floodplains. 228 Figure 64. A view of Oregon Creek looking downstream through the mined area showing site conditions that remain 13 years after gravel removal. The specific changes that retarded vegetative recovery and development at these sites were related to induced aufeis development and increased annual high-water stresses. At Washington and Oregon Creeks, extensive aufeis fields annually developed within the material sites. This ice, which is known to last until late June throughout the disturbed areas, severely impeded vegetative recov- ery at these sites. No significant vegetative communities had developed within the disturbed areas of either site during the 13 years following the gravel removal operations. There is no evidence of aufeis development at the Penny River site. However, the area was scraped in an irregular surface pattern over 15 ha to a depth equal to or slightly below normal summer flow levels (Figure 65). The site was visited II years after gravel was removed. As a result of the depth of scraping, much of the site contained either small pools of ponded 229 Figure 65. Penny River mined area looking upstream. Note the flooded conditions within the disturbed area, and the overburden piles in the center of the site (circled on photograph ) . water or water saturated soils. A small 0.6 ha, 1.5 m deep pit was dug in the southeast corner of the site. The hydraulic analysis shows that the Penny River site is flooded for short durations during higher flows on an annual and possibly semiannual basis. Flows of only approximately 150 per- cent of mean annual flow begin to flood the material site. During the II growing seasons following the disturbance, only sparse, scattered pioneer and early willow floodplain communities had developed within the scraped portions of the Penny River site. These early succes- sional habitats were not present in the undisturbed floodplain reach which, as previously stated, consisted almost entirely of mature shrub thickets. Thus, the structure of the vegetative community within the mined site changed for the long-term from one dominated by dense mature shrub thicket habitats to one dominated by scattered and low-density immature herbaceous and woody species that are adapted to wet soil conditions. Repeated stress from annual or semiannual high water, combined with the continuously 230 water-saturated soils over much of the Penny River site, were probably the key factors impeding vegetative recovery (especially by woody species). Another example of gravel removal and site location characteristics which resulted in known short-term (the site was visited 3 years after disturbance), and probably long-term annual hydraulic stress occurred at the Sagavan irktok River study site. At this site 20 ha of a complex mixture of mature and advanced, seral-stage shrub thickets was removed and the under- lying gravels excavated to an average depth of 1.5 m. This area was located between a high-water channel and the main river channel. The Sagavan irktok River was a large river with moderate channel slope that flowed in a sinuous conf i gur at i on . This gravel removal operation resulted in a permanent shift of much of the main channel through the material site. Hydraulic analysis at this site shows that extensive flooding is expected to occur on an annual basis with water potentially influencing the site for up to 70 days each year. The site was visited during the third growing season after disturbance, and no vegetative recovery had occurred. As long as the river continues to flow through and annually flood the material site, it is not expected that significant vegetative recovery will occur in the long-term. Short-Term Alteration of Vegetative Habitat Structure Short-term alterations, in the types of vegetative habitats present within disturbed areas, occurred at those sites where vegetation was re- moved, but where some natural vegetative recovery began within I or 2 years post-mining and continued thereafter unimpeded. At no instance did an entire disturbed area naturally revegetate over the short-term. However, in por- tions of 13 sites pioneering communities became we I I estabi ished within I or 2 years (Table 29). This development most frequently occurred in those portions of the disturbed areas which: were not influenced by normal or high water flows; had a plentiful seed source or contained root stocks and other woody slash; and/or consisted of well drained but moist soils with high silt 23! >- i. > o V q: > CD 0) > o 1- Q) 4- U (0 U) 1_ ^ >- ■o D c 3 ■4— to E E 0 U O T3 0) C 0) CO ^ 01 E 1- 0) in c o a. in 0) cr ^ c o 4- ro o o CM 4- o >- \- 0) > o o 0) — i_ in lU 10 > 0) — >~ — ID C *- .- 0) a> 0) > in c u E ->- E O O <0 O 1_ CJ >^ -4— 1- in Q) 1. > .— O ^4- O 0) M- u o 0) c > o .— .— 4- 4_ 10 ro -»- o 0) 0 CD _i 0) > CO c c s s o o c c ^ J£ c c in OJ CN — in in n >- :::^ >~ >- XI Z! 13 ■D "D ■o D 1_ O O O o 1_ SZ O O 0 o ^ in 5 s 5 s in s ^ ^ JZ ^ S 01 -♦- -♦— -*- ■*- 0) ^4— . — . — . — . — M— — S 5 5 5 — in in in in in in u D D D D D o o 0 o o o Q) CD 0) 0) 0) 0) O U in u in o in o in u ID ID JD (0 J3 (0 n ID n 10 n n D XI D JD D n D n 1- \_ 1- l- 1- V. 1- u V- i_ 0) Q> ^ 0) JZ 0) X Q) ^ 0) X X in X in X in X in X en c c (U 0) XI ■o 1- u D D n JD 1- 1_ 0) Q) > > o o in 0) c 0) ■o D i_ 0) > o c Q) "O 13 r> i_ 0) > o 01 0) l- l- u 0) > c -— :3 q: cc -!<: ■D D — C o — o CO in 0) "O C "D ID C — ID Q. ■o -C O in o ID — in V4- ■«- o in ID a> O CD ■o •o 10 0) o m ID 0) 0) u l_) j l_ -4- •— U co a c in .— >~ .— x: c Q) in c M- 10 0) D s Q_ < 0) c o "O c 10 n ID 0) -D C 10 0) E _ 0) 0) T3 C — C in ID c ^ — u 0) 0) I. u o en ■D "O ID 01 o \- -t- DD ID E ID 01 i_ -*— in c 5 O Q I 1_ > D o 0) x> c o CL > o XI ID in ^ c ID -Q CD C — 1_ O. IV O 4- — ID CO 5 01 "D c O Q. 0) > o n ID in C 10 XI en c — 1_ a. 0) o -1- — ID 1/5 5 01 > E 0) 0) o o 01 Q. in o 1. Q- ■D 01 X) C O CL 0) > o XI 10 c ID X} CD c — i. Q- 01 O -^ — 10 1/5 s > on ID c (0 > o in 01 c 0) "O 1_ D JD I. 0) > o CN >^ ^ :^ >- >- >- ■D ■D ■o •o XI XJ O O o 0 O o O O o o o o g s in s in g in s 5 JZ ^ 01 SI 01 n 01 ^ x: -t— -f- 0) A- 01 -t- 01 -4- -♦- •— — i_ — i_ •— i_ — — 5 5 -^ g -t— g ■^ g in g in in "D in -o in X) in 3 in D D C D c 3 c 3 I. D o o ID o ID o ID o X o 0) Q) OJ 0) 01 in 01 o in O in o U) o in o in o in ID J3 10 XI ID XI ID XI 10 XI >- 10 XI -O D XI D XI D XI 3 XI D XI XI D 1_ 1. i_ 1_ I. l_ 1_ u 1- U o 1- V. 01 x; 01 x: 0) ^ OJ Sl 01 X o 01 X X in X in X in X in X in s X in CD CD c c XI XI c c D D o o 1_ \_ u \. 3 D in in in in 01 0) CL i- Q- S- 0) 01 c -f_ c •*- Q) 10 Q) 10 XI g •o g \_ 1. D XI D XI ^ 0) J3 01 1_ X! 1_ X) 0) c 0) c > 0 > o o Q. o CL o 01 01 1_ o in D C ID E (0 01 i- in c g o Q l_ 01 > E 10 01 I. in CL n 0) > ID C 10 c ID 232 and sand content. The results of soil sample analysis indicated soil nutri- ents were not limiting factors influencing vegetative recovery at any of ttie 25 study sites. The initial reco I on i zat i on of these disturbed areas most frequently oc- curred by seed development; at several locations, however, willows had reinvaded through development of adventitious stems and roots from old woody slash and root stocks. Adventitious stem development occurred most often in overburden piles where woody slash was placed. All overburden piles occurred in sites developed before 1971. More recent regulation of gravel removal activities require overburden and woody cover to be removed completely from f I oodp lain s i tes . In general, herbaceous species dominated in those pioneer communities which were developing from seed. However, Salix alaxensis was a frequent member of these communities in all four geographic regions, and seedling Betu I a papyr i f er a and Popu I us balsamifera commonly occurred in pioneer communities at several Northern Interior sites. Taxa that most often were dominant in these invading communities included Epilobium latifolium, Salix a I axens i s , Sa I i x spp., Equisetum variegatum, Stellaria spp., Hedysarum Mackenz i i , Astr aga I us spp., Oxytropis spp., Juncus spp., Carex spp., Eriopho- rum spp., Ca I amagrost i s spp., and Poa spp. In soils that were less moist and more coarse, Ar temis i a spp., Crepis nana. Aster sibiricus, and Erigeron spp. frequently occurred as initial invaders. Overburden was piled either within the disturbed area or at its edge at many of the older sites. At the Penny River and Washington and McManus Creeks these overburden piles contained many organics and woody slash, root stocks, and debris. At Penny River, three piles of material were located within the 15-ha site (Figure 65). At Washington Creek, one pile was placed in the middle of the 3-ha site and one on its edge, and at McManus Creek the organic overburden was all piled on the edge of the 4-ha disturbed area. These piles averaged I to 2 m in height, however, a few were 5 to 7 m ( F i gure 66 ) . At all three sites, herbaceous and woody vegetation were well estab- lished on the overburden piles within I year after disturbance. Development 233 -'^fJm^; Figure 66. Close-up view of an overburden pile in the Penny River mined area. Note the development of herbaceous and woody vegetation during the II years following gravel remova I . on these piles preceded other disturbed area revegetation at Penny River and McManus Creek by approximately 6 to 7 years. At Washington Creek, which was visited 13 years after disturbance, the only significant revegetation of the site occurred on overburden piles (Figure 67). At all sites, the initial ^ Figure 67. Washington Creek mined area showing vegetative recovery only present on the overburden pile 13 years after gravel removal. 234 shrub development was through adventitious stems (Figure 68). Willows, primarily S. a I axensis, most frequently developed from old slash and root stocks . a. View of broadcast slash and 2-year-old stems. b. View of old root stock with new stem. Figure 68. Woody revegetation occurring through develop- ment of adventitious stems. 235 Similar rapid development of woody shrubs through adventitious stem development occurred in I- to 2-ha areas at both Middle Fork Koyukuk River- Downstream and Aufeis Creek study sites. However, at these sites the slash and woody debris were not piled, but were spread over the ground at the edge of the disturbed areas (Figure 69). Figure 69. Distribution of woody slash debris and other organics over the ground on the edge of the gravel removal area at Aufeis Creek. At the Tanana River-Downstream site overburden from the 5-ha pit was placed in contoured banks surrounding the flooded pit. These overburden piles were approximately 2 to 3 m deep inversely piled (top material covered by bottom material), and consequently contained no organics or woody remains near the surface. However, an early shrub community dominated by Popu I us balsamifera, S. alaxensis, and AInus crispa, with a density of 230 stems per 0.004 ha, was present during the fourth growing season following gravel removal. This shrub community developed from seed and invaded in mass during the first growing season. The shrubs occurred in uniform density over approx- imately 60 percent of the gently-sloped, 20 to 25 m wide overburden banks surrounding the pit. Rapid natural reco I on i zat i on of disturbed areas was not always limited to overburden piles. At the Jim River, West Fork Tolovana River, and Prospect Creek, pioneer communities were well developed at the end of the 236 first full growing season following di sfurbance. At these sites the com- munities were developing on the contoured side slopes of the permanently ponded areas. An average of 13 species, with a range of 7 to 21 species, occurred in 0.0004-ha sample plots located in these habitats during the second (Jim River and Prospect Creek) and third (West Fork Tolovana River) growing seasons following disturbance. Willows, alders, birch, and spruce occurred with the herbaceous taxa in these habitats at all three sites. Although these sites have not been inspected since 1978, the pioneer com- munities will probably develop unimpeded and quickly lead to early and advanced serai stage shrub communities. The Tanana R i ver-Upstream site was very similar to the West Fork Tolovana River site with respect to Physical Site Characteristics and Gravel Removal Area Characteristics. The mined site was 10 years old during site inspection, and 15 years old at the time of data collection (summer 1978). Shrub thickets dominated by Sal ix arbuscu loi des and A I nus tenui folia had developed surrounding much of the pit and on spits and islands which remain- ed above the water level of the upper pit (Figure 70). These communities had Figure 70. View of the upper pit at Tanana Ri ver-Upstream showing diversity of shoreline configuration and develop- ment of woody and herbaceous vegetation 13 years after gravel removal . 237 reached an advanced shrub stage with densities as high as 990 stems per 0.004 ha by the 13th year. Thickets averaged 2 to 3 m in height. During site inspection these thickets most likely were equally as dense and practically as tall. At most above mentioned sites, following rapid invasion and development of pioneer communities (both by seed and adventitious stems), early shrub communities usual ly were wel I estabi ished in 3 to 5 years. The majority of these areas were small (0.5 to 2 ha) and were usually scattered throughout the scraped sites or surrounding the flooded sites. Usually only one to three isolated patches of early shrub communities occurred in the scraped sites. Those sites that were of sufficient age (including Penny River, Oregon Creek, Washington Creek, Sinuk River, McManus Creek, and Tanana Ri ver-Upstream) began to provide sufficient cover for nesting and feeding passerines and summer and winter cover for sma I I mammals about 10 years after initial disturbance. Thus, at sites that provide areas (of various sizes) for revegetative growth without severe stresses from flooding or aufeis scour, habitats that provided food and cover for passerines and sma I I mammals (primary shrub thicket occupants) were naturally replaced about 10 years after completion of gravel removal activities. No Significant Change in Vegetative Habitats Contrasted to long-term loss of habitat and short-term alteration of habitat structure are gravel removal operations that resulted in no measur- able change in the vegetative structure of the study areas. Gravel mining did not affect vegetation at 5 of the 25 study sites, either because of the disturbance location, or the floodplain character- istics, or both (Table 28). At two additional sites, the Nome River and Kavik River, only slight reductions in vegetative cover were observed. Three of the five sites with no vegetative disturbance were large flood- plains with large- and medium-width channels flowing in braided patterns. At 238 all three sites large quantities of gravel were removed by shallow scraping surface layers over a broad area. Specifics on these sites are: Study site Scraped surface area Quant i t y of grave I removed Ivishak River Kuparuk River Phelan Creek 40 ha 14 ha 70 ha 120,000 m' 42,000 m^ 575,000 m' Although Phelan Creek was a wide (approximately 1,000 m) unvegetated floodplain, and the Ivishak and Kuparuk Rivers also had extensive unvege- tated gravel bars, the latter two sites also contained numerous islands with densely vegetated shrub thicket stands (Figure 71). At the Ivishak River and Figure 71. View of the Ivishak River floodplain looking downstream showing typical braided channel characteristics with extensive gravel bars and isolated, vegetated islands. Kuparuk River sites, operators conformed the configuration of their gravel removal areas to avoid the vegetated islands. At the Phelan Creek site, gravel was scraped from a uniformly shaped and contiguous area, because the floodplain was entirely unvegetated within the work area. 239 The best example of avoiding disturbance to vegetated areas on a mean- dering or sinuous river occurred at the Shaviovik River study site (Figure 72). This river flowed in a medium width, single channel and in a sinuous *™i^^^^3cr Figure 72, View of both undisturbed (background) and mined (foreground) reaches of the Shaviovik River. Note that gravel removal maintained natural point bar contours and shapes and did not disturb riparian vegetative zones. configuration. With these characteristics the floodplain consisted of broad (averaging approximately 40 to 50 m in width) unvegetated point bars at every inside bend and numerous unvegetated lateral bars located between point bars. Gravel removal consisted of shallow scraping on every point bar and lateral bar over a distance of several river kilometers. Small quan- tities were taken from each location, however, a total of 116,000 m was removed. The actual scraping of unvegetated gravel deposits throughout most of the Shaviovik River site was conducted in a manner that caused minimal, or no biological disturbance. Gravel bars were scraped only in their unvege- tated portions and riparian shrub thickets were not disturbed. Also, the mining operation maintained natural contours and shapes on gravel bars and 240 did not mine adjacent to the river. Thus, the Shaviovik River has maintained its natural channel and configuration. FACTORS AFFECTING VEGETATIVE RECOVERY RATE Several factors found to be influencing vegetative recovery already have been discussed. The composition of fauna! communities using disturbed areas was directly related to the habitat types available, thus, an under- standing of how factors at the study sites influenced the rate of natural vegetative recovery warrants further discussion. Overburden piles, woody slash, and debris, an abundant seed source, and displaced organic mats enhanced recovery rate. Hydraulic stress such as aufeis development, perman- ent ponding, actual channel shifts, and increased flooding impeded develop- ment. Soil conditions and growing season, depending upon site specific characteristics, either enhanced or impeded vegetative recovery. Imped iment s Among the factors believed to be impeding vegetative recovery, hydrau- lic stress influenced most sites and had the strongest and most long-term effect. These stresses resulted from changes induced by gravel removal in floodplain elevations, dimensions, and configurations. They included: • Permanent or annual flooding, • Increased frequency and duration of temporary flooding, • Long-term channel changes (increased braiding and channel width and decreased channel stability), and • New or increased aufeis development. The specific known causes for these induced hydraulic changes are presented in detail in EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAULICS. In general, they most frequently resulted because sites were excavated too deeply (excluding pit sites) without maintaining buffers or stable channel banks, or because the gravel removal method and character- istics were not correct for the chosen location. 241 At 13 sites the gravel removal method led to significant hydraulic changes that secondarily impeded the vegetative recovery rate (Table 30). Permanently ponded water and aufeis development caused the most significant impediment. Permanently ponded water occurred at those sites where the mining plan called for excavated pits, but also at sites where depressions were scraped below summer water levels. The latter occurred at sites that were directly connected to an active channel (Jim River); at sites that were not directly connected to an active channel (Penny River); and at sites that were originally not connected, but where gravel extraction caused an active channel to reroute through the deep depression (Dietrich River-Downstream). Aufeis impeded vegetative recovery at four sites (Washington Creek, Oregon Creek, Middle Fork Koyukuk River-Downstream, and Jim River), all of which were directly connected to active channels. Aufeis development is believed to occur annually at all sites, and affects the entire disturbed area at Washington Creek and Oregon Creek and most likely affects much of the disturbed areas at Jim River and Middle Fork Koyukuk River-Downstream. Two additional factors were impediments to vegetative recovery under certain conditions: soil condition and length of growing season. Vegetative recovery was occurring to some degree under a wide variety of soil type, texture, nutrient, and moisture levels. Differences in the degree of develop- ment and the species composition reflected the wide range of xeric and mesic soil conditions. Soil nutrients were not found to be limiting factors at any site regardless of its age, original condition, or final condition. However, vegetative invasion was restricted by very compacted surface layers at several of the more recent sites. These areas most frequently were associ- ated with access routes over gravel surfaces leading to and from the mined sites. At Dietrich River-Downstream, heavy equipment compacted the flood- plain gravels approximately 25 cm adjacent to the gravel removal area (Figure 73). This site was visited 3 years after completion and vegetation had not invaded this access road although the unflooded banks of the materi- al site were supporting pioneer communities. Another soil condition which restricted vegetative development 13 years after site work, occurred at Oregon Creek. Inorganic materials were scraped 242 05 c ■o 0) Q. E 0) 0) (0 L. <0 4- 0) e in CT) O o Z3 1_ "O CO ::^ X j_ (0 "O 0) >~ o 0) > (1> O u CO 0) ce c — 0) > 0) O) J c ro (D en 0) > c o ID O (0 o o I- 0) Q> E ro i_ ro CL >~ en O O 1_ o Q. ■o 0) ro ■u 0) c \ o ro Q. CD ^ C ->- C "O ro — -c 5 (_) ro CD CD < 0) ro -1- 0) •D CO CO 0\ CD O Oi 0\ Ch 1 E ro E 0) ro u ro 0) i_ 0) 0) c S E Q. o ro Z) Q rA 01 c 5 o -it: o jk: — — to 0) -— Q D D — 0) — O j«: (D cr 1 :^ >^ 0) 1 CN 0) 1 1- i_ o o 1 1- 1 0) U -i: 0) >i: ^ CN u -i£ 1 i_ -i: O > i- o i- > 0) -^ 1_ ro \- O o 1. — ■t- o •— > •— (_) o ^ Ll u. 0) cc cn q: — > i_ c o > c c cc ro Ul 0) ro — ~ I. .— -»— > 1- — cr D ^ O) c a> D 0) Q> ID -*- "D "O C 01 0) c E C ^4- Q> cn - ■o (0 ■D > 0) •D 3 3 ^ — "O — c — (0 JD — -a to o > o 0) M- 255 appeared to provide adequate habitat with sufficient food supplies but both received low fauna use. They were both 1.0 ha or less in size. The Tanana R i ver-Upstream pit, which was 13 years old, provided the most desirable lacustrine habitat. This 7.5-ha pit had a very irregular shoreline with heavy vegetative cover; contained numerous shrub-thicket vegetated islands in its southern half (upper pit) and graveled islands in its northern half (lower pit); had an abundant food supply dominated by fish and macroinver tebrates; and had a variety of deep and shallow water zones (Figure 80). During the site visit 147 individual birds of 39 species were Figure 80. Tanana Ri ver-Upstream showing shoreline diver- sity and vegatative development in the upper pit. recorded in the entire study area and four individual beaver, at least two muskrats, and three moose were observed using the pits. The avifauna observ- ed are identified in Table 32. The West Fork Tolovana River pit was smaller (4.5 ha) and not as old (5 years) but otherwise was similar to the Tanana R i ver-Upstream pit. Avifauna observed at this site are identified in Table 33. Due to the young age and sparse vegetative cover, the avifauna in the disturbed area included few 256 0) _ _ 10 » ■o in - — s c - . O — O 0) E C71 £ u — 1- ■c S (0 10 a> 3 I- it) u 10 a Q) 1 — O Q. 0) (D CN U) ^ di _ _ E OJ >~ a> C X) o a (0 »_ V © o I. Q. 0) -•- k- a> a> > ■^ a> in ■o L. > o E O 0 Q. tfi -. « o l- a> — Q. dj — (0 E < tn Q. LO Q. 33 < a. (fl— o Q.(rt~— o >■ D CM a— O — C Q. (0 ■ cc— to o._j — * «to- I — o c — E l_ E — 1_ — O to O — (D — -ta S O — Q — Q. U T3 (0 CD 0) o — jr u — T3 y c W -^ 3 c K\ Q.— O — 1 ® £ t^ (J > r — >^ -o o >- I. —wi- (D ll-—*. SI) o c o - u 1 g — £1 « ^ E >. O — U T3 t_ E « a — c 0 O >- S < o u 257 in — o u ■a •- o 1* -^ U) .- — Q. - 01 _ n n 0 a £ u I 3 Z c >* 0 ^ o c ■•- — — (D q: — 5 ta *- — s CM _ Qi tn 1 c — l_ c c o — J3 (D - (1» 01 O ■*- u 0) q: £ c a> (u — (/) — in — CO CVi c Cl> O tfl i_ ^ E „ a> — cn 10

(D — (0 CT O Q. E 0 0) (0 — c 5 E a> -0 l- a. — 0) — U 0) to 10 Cl 0 >- in 0) . o (U S — 03 O < — 3 trt < >- X 0) — . (0 2 — o -"t I- c \- ^ (0 £ Q. ^ (rt yi 3 l- I. X 0 £ 0 z u. .o — — — — (N •>- -o o I/) — (0 O 258 passerines. However, vegetative recovery had become well established on the gravel islands and shoreline and it is believed this site will soon provide the same quality of habitat as the Tanana R i ver-Upstream. One colony of beaver also were using the West Fork Tolovana River pit. Permanently ponded material sites of sufficient size (at least larger than I to 2 ha) will provide a high quality habitat if they have: • A diversity of shoreline configuration and water depth, • Dense border cover, • Islands or peninsulas or both, and • An abundant fish and macro i nvertebrate food supply. SIMILARITIES OF RESPONSE BETWEEN B I OT I C AND STUDY SITE PARAMETERS A computer analysis for similarities in response between terrestrial biotic parameters and study site characteristics was conducted (Table 34). Ten biotic parameters were selected for analysis. The analysis demonstrated that responses of biotic parameters could be categorized into three groups. Each parameter within each group displayed a similar reaction to specific gravel removal operations. When comparing the responses of the biotic param- eter groups for al I 25 sites, 5 site response combinations were found (Table 34). After these analyses, the material site characteristics were compared for each site response group. Biotic Parameters The biotic parameters reacted in three groups of similar response to gravel removal induced changes. Group I included passerines, shrub thickets, moose habitat, and ptarmigan habitat; Group II included soil nutrients, ground squirrels, early shrub communities, and small mammals; and Group III included soil texture and water birds. 259 Table 34. Two Way Coincidence Table Displaying a Hierarchical Clustering of Similar Sites and Similar Biotic Parameters Biotic parameters Group I Group I I Group I I I £) ^ *- Site response group Shaviovik R Ptielan Ck Ugnuravik R Kuparuk R Ivishak R M.F. Koyukuk R-DS Nome R Dietrich R-DS Kavik R WcManus Ck M.F. Koyukuk R-DS Tanana R-OS Oregon Ck Dietrich R-US W.F. Tolovana R Sagavan irk tok R J im R Prospect Ck Aufei s Ck Tanana R-US Penny R Gold Run Ck Washington Ck Sinuk R Skeetercake Ck Symbols used for computer analysis were adapted from quantification of change ratings ITable T- I ) as follows: (0,11 equals =; (2,31 equals -; (4,5,61 equals b; (7,81 equals +; and 19,10) equals *. Note: all b's (no response or weak response) were eliminated from this table to remove c I utter. Responses by group were: A - essential ly no response. B - minor decreases in biotic parameter Group I; minor increases in biotic parameter Groups II and III. C - significant decrease in biotic Group I; minor decrease in biotic Group II; increase in biotic Group IM. D - significant decrease in biotic Group I; increase in biotic Group 11; significant Increase in biotic Group III. E - decrease in biotic Group I; Increase in biotic Groups II and III. 260 In general. Group I parameters either stiowed no response, or displayed a significant decrease resulting from gravel removal induced changes. This was directly related to clearing of significant quantities of vegetation which passerines, moose, and ptarmigan used as primary habitat. Group II parameters displayed no response at sites where vegetative habitats were not disturbed. However, all parameters except soil nutrients decreased at sites that were subjected to permanent or frequent hydraulic stresses (aufeis, ponding, and flooding) and did not contain overburden piles. At sites that were subjected to hydraulic stress but which contained overburden piles, small mammals, ground squirrels, and early shrubs increas- ed. Soil nutrients basically displayed no response. Group III parameters either displayed no response at sites where the floodplain character was not significantly disturbed, or they increased. Both parameter responses were once again directly related to removal of extensive vegetative cover. Water birds increased in response to the in- crease in aquatic, gravel bar, and mud flat habitats, while soil texture increased due to the removal of organic, silt, and sand overburdens and the exposure and deposition of coarse gravels and cobbles. Physical Site Characteristics The Physical Site Characteristics that were analyzed are those identi- fied in the Major Variable Matrix Table (Table I). They included: drainage basin size, channel width, channel configuration, channel slope, and stream or i g i n . Responses of biotic parameter groups at the 25 study sites displayed five basic combinations. These are labeled Site Response Group A through E on Table 34. Eight sites occurred in Group A, where no significant responses were measured in any of the biotic parameter groups. These sites were mostly of medium to large channel widths, of braided or sinuous configuration, and of mountain or foothill origin. However, these site characteristics were not considered to have significantly contributed to the minimal disturbance at these sites. Of greatest significance was the minimal vegetative disturbance which occurred during the gravel removal operations. 26 1 Site Response Groups B through E did not display any apparent similar Physical Site Characteristics. Thus, it was judged that drainage basin size, channel width, channel configuration, channel slope, or stream origin were not significant factors in governing the responses of terrestrial biota. Gravel Removal Area Characteristics The most significant similarities in Gravel Removal Area Character- istics were those that led to permanent or frequent hydraulic influence within the disturbed area. This annual stress led to a significant and often long-term impediment of site vegetative recovery. Two similar Gravel Removal Area Characteristics were observed that produced this result. They were: scraping within the active channel at any location along the river coarse; and scraping adjacent to an active channel primarily on an inside bend, and without an adequate buffer along the channel. Scraping Within the Active Channel. Wherever gravel was scraped from within the active channel, the scraping also extended beyond the original channel to adjacent gravel bars. In these areas gravel was scraped to depths equal to or slightly below normal water levels. This characteristic produced a long-term decrease in Biotic Group I (primarily shrub thickets and passer- ines). The hydraulic changes that occurred in these areas were the prime factor found to be influencing site vegetative recovery. These changes are discussed in further detail in EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAUL ICS. Scraping Adjoining the Active Channel on an Inside Bend. At seven sites gravel removal occurred on a point bar or inside meander but did not extend into the adjacent active channels. All sites were of sinuous or meandering configuration and were scraped on sharp inside bends. At five of these sites (Penny River, Ugnuravik River, Skeetercake Creek, Middle Fork Koyukuk R i ver- Upstream, and Middle Fork Koyukuk River-Downstream) the scraping occurred to within or below the water level. Except at Middle Fork Koyukuk River- Upstream, no buffer was maintained between the scraped area and the main river channel. At the Middle Fork Koyukuk R i ver-Upstream site a 30-m wide vegetated buffer was maintained. However, within a few years the rivers had formed cut-off channels through the scraped areas at all five sites. 262 Thus, scraped sites located on sharp inside bends led to the formation of cut-off channels unless extensive vegetated buffers (Jim River) or natur- ally contoured channel slopes (Shaviovik River) were maintained during the gravel removal operation. These cut-off channels subjected the mined areas to frequent or permanent ponding and flooding which impeded vegetative recovery . Additional Similarities. Overburden piles, as previously discussed, were a positive addition at sites annually subjected to ponding, flooding, and aufe i s deve I opment . At sites where piles occurred, Biotic Group II (primarily small mammals, ground squirrels, and early shrub communities) increased (Site Response Group B, D, and E, Table 34). However, at sites where overburden piles did not occur, but the site received annual hydraulic stress of flooding, permanent ponding, or aufeis development, Biotic Group II decreased (Site Response Group C). Overburden piles occurred in a variety of shapes and sizes and were placed in various locations within the material site. From a revegetative viewpoint the most effective pile compositions were those with a mixture of silts, organics, woody slash, root stocks, and debris. These piles only occurred at the older sites and all were at least I to 1.5 m above normal water levels. It is not known if additional piles of lower height originally occurred and had been eroded and removed by flood waters. Also, all piles that were within the central portions of the mined areas were either not directly in the path of main currents or were placed in windrows oriented parallel to the current. Overburden piles that remained in the middle of large scraped sites were judged to be of more overal I benefit than those placed on the edge of the disturbed areas. These piles provided immediate denning habitat for ground squirrels and, within several years, began to provide cover and nesting habitat for small mammals and passerines within the central portions of large mined areas. The effectiveness of natural buffers was related to their location and dimensions in relation to river size and configuration. Twelve of the 25 study sites included some use of buffers. Two types were employed: 263 • Undisturbed gravel bars separating scraped sites in active flood- plains from active channels, and • Incised banks and associated riparian zones separating scraped and pit sites located in inactive floodplains and terraces from active f I oodp I a i ns. The level of understanding that was obtained regarding the effective- ness of these buffers does not allow conclusions to be drawn. Accurate data describing original buffer characteristics (such as width, height, veget- ative structure, and soil composition) were not available for many sites, however, several trends were observed. At sma I ler rivers of sinuous and meandering configuration, buffers (primarily incised banks and associated riparian zones) of widths in the range of 10 to 15 m were effective in containing active channels at sites that were 5 to 16 years old (Figure 81). Figure 81. Undisturbed buffer along the original stream channel at Aufeis Creek (downstream disturbed area only) 264 In larger rivers, most natural buffers that were maintained to protect scraped sites in active floodplains failed within a couple years. At Middle Fork Koyukuk R i ver-Upstr earn a 30-m wide, I- to I . 5-m high heavily vegetated buffer protecting an inside meander site was breached in I year; at Sagavan- irktok River, a 30-m wide, 0.5-m high gravel buffer protecting a mid-channel site was breached in I year; and at Dietrich River-Downstream a 50-m wide and 0.5- to I -m high gravel and sparsely vegetated buffer protecting a site on the edge of the active floodplain of a braided river was breached in 2 years. These buffer failures have all created permanent channel changes through the mined areas of these sites. At pit sites located in inactive floodplains and terraces, buffers composed of incised banks and heavily vegetated riparian zones ranging from 50 to 90 m in width were sufficient in protecting the pits from active channel diversion at sites up to 13 years old. However, most of these sites (three of five) are located on sma I ler rivers with relatively stable chan- nels and are on the inactive side of the floodplain. On the other hand, at the oldest pit site (Tanana River-Downstream) a 50-m wide buffer separated the pit from an erosional zone of a side-channel of this braided river. During 1977 and 1978 this buffer was being actively eroded. It is not known how wide the buffer was at the completion of the mining activity. One mining method (pits) and one site location (separated from the active floodplain) frequently led to the creation of high quality habitat that resulted in an increase of water birds (Biotic Group III). As previous- ly discussed, this method created a habitat type that frequently was not readily available in adjacent floodplain reaches. The quality of this habi- tat was related to its size, shoreline diversity (configuration), water depth diversity, shoreline cover, presence of islands, and food ava i I ab i I i t y . Other characteristics occurred that were not directly related to the location or operation of the material site but that reduced detrimental impacts to the terrestrial biota. At those sites where access to the flood- plain had to pass an incised bank, gravel fill ramps (Figure 82) reduced the overal I impact. At sites where incised banks were cut for access severe 265 ^ :;ai;uPifi^*. Figure 82. Gravel fill ramp used to protect ttie incised bank at ttie Sagavanirktok River study site. erosion frequently resulted. In permafrost areas both thermal and hydraulic erosion induced by surface travel on unprotected banks can, and at the Ugnuravik River site did, create uncontrollable problems (Figure 83). At sites separated from active channels by buffers, a heavy layer of rip rap on the buffers significantly increased their effectiveness (Figure 84). 266 Figure 83. Thermal and hydraulic erosion of permafrost induced by multiple passes of a tracked vehicle across an unprotected incised floodplain bank and adjacent tundra. IK' Figure 84. Armored bank protecting the West Fork Tolovana River pit from a channel diversion into the mined site. 267 SUMMARY AND CONCLUSIONS Overall, gravel removal from floodplalns frequently had a detrimental long-term effect upon local terrestrial biota. Specific site locations coupled with the depth of scraping proved to be the most influencing fac- tors. VEGETATIVE REMOVAL At 18 of the 25 study sites gravel removal operations cleared signif- icant quantities of riparian vegetated habitat. This loss most significantly affected passerines and small mammals which rely upon these riparian zones for primary feeding, nesting and cover habitats. At most of these sites this habitat reduction led to long-term changes in fauna utilization and com- mun i t y structure. At 4 of the 25 sites, gravel removal operations did not alter existing vegetative communities, and consequently did not lead to changes in local faunal communities. Three of these sites were located in floodplains with large and medium width channels that flowed in a braided pattern. At all three sites large quantities of gravel were removed by shal low scraping of surface layers over a broad area. The fourth occurred on a sinuous to mean- dering river. At this site a large quantity of gravel also was removed by shallow scraping unvegetated portions of lateral bars and point bars. This scraping maintained natural point bar profiles and subsequently did not induce any channel changes. MINING DEPTH AND LOCATION Gravel removal operations that scraped to within or slightly below the water table and that occurred at inside bends or immediately adjacent to, or 268 within the active channel also produced a long-term negative response (de- crease in numbers) from terrestrial biota. At 13 of the 25 study sites gravel removal operations with these characteristics caused hydraulic changes (such as permanent channel shifts, aufeis development, or increased flooding) that impeded subsequent vegetative recovery of the disturbed areas. However, at those sites where gravel removal created permanently ponded areas, or extensive gravel and mud flat habitats with pockets of ponded water or backwater areas, water birds (including waterfowl, shore- birds, gulls, and terns) frequently increased utilization of the area. OVERBURDEN Overburden piles containing silts, organics, and woody slash and debris facilitated rapid and continued vegetative recovery within the mined site. These areas provided islands of useable passerine and small mammal habitat within a relatively short-term period. At many sites overburden piles were providing vegetated habitats that were being used by these species within 10 years after gravel removal. Ground squirrel populations frequently showed immediate response to available denning habitat provided by overburden piles. At most sites where piles occurred these animals were significantly more abundant within the mined site than in adjacent floodplain reaches. When this overburden material was broadcast over the ground in areas where it would not be washed downstream it was equally effective in facil- itating rapid vegetative recovery and development. PERMANENTLY PONDED HABITATS At eight sites the gravel removal operation (primarily through pit exca- vation) created permanently ponded habitats. Although this operation led to a long-term change from natural terrestrial conditions, at several sites this mining result led to the development of a diverse habitat that provided high quality feeding, nesting, and cover areas for passerines, small mam- mals, water birds, and furbearers. Factors that were found to influence the fauna response to these areas were: shoreline configuration, shoreline vegetative cover, water depth profiles, presence of islands, pit size. 269 availability of food, and connection to an active ctiannel. Fauna utilization of the area significantly increased at several sites with a high diversity of these factors. RECOMMENDATIONS Gravel removal operations in floodplains should attempt to incorporate the following recommendations into site selection and site operation de- cisions in order to minimize long-term disturbance to terrestrial flora and fauna : 1. Whenever possible, avoid vegetated habitats. 2. When scraping in active or inactive floodplains, maintain buffers that will contain active channels to their original locations and conf i gurat i ons . 3. When small quantities are required (approximately 50,000 m ), select sites that will scrape only unvegetated gravel deposits. 4. When large quantities are required (approximately in excess of 50,000 m ), select large rivers containing sufficient gravel in unvegetated areas, or select terrace locations on the inactive side of the floodplain and mine by pit excavation. 5. If pit mining, design a configuration with high shoreline and water depth diversity and provide islands. 6. If mining in vegetated areas, save al I overburden and vegetative slash and debris to use during site rehabilitation to facilitate vegetative recovery. This material should be piled or broadcast in a manner so it wi I I not be washed downstream. Detailed elaboration and expansions of these recommendations are pre- sented in the Guidelines Manual. 270 REFERENCES Johnson, R. R., and D. A. Jones, (tech. coord.) 1977. Importance, preservation and management of riparian habitat: A symposium. Tucson, Arizona. July 9, 1977. USDA For. Serv. Gen. Tech. Rep. RiV\-43, 217 pp. Pamplin, W. L., Jr. 1979. Construction-related Impacts of the Trans- Alaska Pipeline System on Terrestrial Wildlife Habitats. Joint State/Federal Fish and Wildlife Advisory Team. Special Report No. 24. 132 pp. Spindler, M. A., and B. Kessel. 1979. Forty-second breeding bird census: Census 167. American Birds 33( I ) :99- 100. United States Army, Corps of Engineers. 1979. Wetland Values: Con- cepts and Methods for Wetlands Evaluation. Research Report 79-R I . 109 pp. Woodward-Clyde Consultants. 1976. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska. U. S. Fish and Wildlife Service. FWS/OBS 76/21. Wash., D.C. 127 pp. 271 EFFECTS OF GRAVEL REMOVAL ON WATER QUALITY L. L. Moulton INTRODUCTION Water quality parameters were measured in conjunction with the aquatic biological studies at the 25 gravel removal sites. Since the sites were visit- ed from 2 to 20 years after gravel removal had been completed, the results of the monitoring program reflect only long-term effects on water quality con- ditions. The sites selected for study represented a broad range of Physical Site Characteristics and Gravel Removal Area Characteristics, which are des- cribed in the Major Variable Matrix (Table I). Instruments and procedures used are described in APPROACH AND METHODOLOGY. Changes in water quality during gravel extraction were not measured because active gravel removal sites were not available for study. A review of available information on this aspect was included in an earlier report (Woodward-Clyde Consultants 1976). 273 Table 35. Selected Alaska Water Quality Standards Beneficial use Parameter Water supply Aquatic life Recreation Dissolved oxygen >75% saturation >7 mg/ft >5 mg/S- (mg/il) or >5 mg/H Temperature n - B" o 0) > i_ - c 0) E c o. ^ « O "^ ii - E (J I/) — 3 O » O - o o o O o s O g o o O o 8 o CM o o CM g g g O o I I r~ I I — I CM 03 03 \n o» — ''> t*^ Q. o — K> CN "O o o cN in K\ CD I I o u S E _ o <0 4- X) 0) K^ CN 03 CN N^ CD CN 278 1) 0) E > D 0 B^ o 01 £ > u — CN CN (D -*- E en _ 10 c L cu § CD -4- Q (D cn E CL c (0 3 *~ 1- C 01 0) in CL Q) (0 E s CD (0 i_ CD Q) u 0) 1_ CD Q II c/) in c i_ ^ 5 0) 1 O o Q (U E 01 c ^ CD ro c CD Q_ (D cn jr (0 >. o CD .t: o o — c II c o cj CD '^ >^ (0 CD CU CD CD c O 'jl o D -D — CD ro 1_ •— 1 -f_ C c o CD — -1— 4— O (D Cl ■D X O "O (D X5 CO c T3 CD • — Q. — cn o Zl If! CO X) "D C o CJ CD 1_ D (D l_ CD Q_ E Q) (D > O cn cn >- "D CD =1 -1- co cn LniriinLntniTiirim tri iTi [^ C3^ O CM UD r- in — in CO ir> O CTi ic^ ro iTi in in CO ro CO CO c c o o hO CO in r<^ CTi cTi in CN ON r^ CO cji CM CO CTi CN in 3 .;_ 01 D .*_ CO cx c cn Q. D CD Z) D CD < (Jl ^) < in JH u c O ^ c -f- u 3 IX CJ) cn cr c c ^ •— o >- "O D ^ zn c — C in a> c o • — (D i_ (U CD CO S o CL CD E O — CJ5 =) D ~3 < cn ^ -^ — o > ro cn s_ .— u CD c ^4- a> U Z3 < q: CD 0 ^ -»- q: (D ^ O — ^ 1- c 3 CD CO i- -t— > CD CD (0 Q- CD CD D j«: CO ^ CO CO 28 1 0) "O 3 U c o o 00 13 a c o '^ o 3 •o — 01 (0 I. — 1 4~ c c o 0) — 4- 4- o <0 a. x> X o •D (U T3 in c ■o « — Q. —• (A o 3 (/) to I. O D ■o c o (J (0 I. d) Q. E 0) (- 0) > c — 0) o O) U) >- V) X — o in »o in in — CO lO y3 in in in in I CM in in in in in in in in 00 in lo in — vo in in c\ r- — tA OD •O 9) 3 -t- 00 rO VO 00 0\ lO in in £ (0 E E O) (0 (0 1_ £ (U 0) 4- (0 I. t- (fl • 0) +- •*- c 0) >~ 4- 4- I. l- w to S l- c — Q. in ■ o -»- c Q. o o 3 3 O) >- 3 4_ in s 3 o — —i —3 to ^ C7> Q. t- Ql o 1 1 u 3 3 0) cc c o o 3 3 c U) (D o i. — •— >' >. k. 3 ^ ^ (U l- I. o o 0) C U) > £1 ••- -♦- ^ ^ ^ (D (D > 4- 0) (1) ■♦- s > ^ (0 I. U. Li- 3 U to iC o z Q Q £ s o to £ c ID 0) 282 Table 39. Average Measured Vatues of Selected Water Quality Parameters at Study Sites with inundated Pits Act i ve Parameter Inundated pit channel Sfudy si te (unl tsl surface bottom 1 upstream) Fenny R T ("C." 2.0 „ 4.0 6 June 1977 00 (mg/ll"^ Cond Ipmhos/cm) 12.8 — 12.0 UO — 65 8 August 1977 T (°C) 12.1 ._ 10.6 DO Img/i) 9.8 — 11.6 Cond (pmhos/cml 510 — 250 Dietr i ch-Upstream T ;°ci 4.5 4.2 (5. .8ml 14.1 8-11 July 1978 00 (mg/l) 10.6 9.7 15. .8ml 8.5 Cond Ipmhos/cml 400 — 275 J im River T (°C, 13.2 _« 9.1 5-5 July 1977 00 (mg/ll 9.8 10.2 Cond (umhos/cm) 60 — 55 Prospect CK T (°Ct 16.7 «_ 11.5 7 July 1977 00 Img/tl 8.4 — M.9 Cond lumhos/cml 70 — 55 WF Tolovana R T <°CI 17.8 7.1 (4. ,3m) 7.5 8-12 June 1978 DO (mg/l) — _ 11.4 Cond (iimhos/cinl 320 — 225 11-13 Sept. 1978 T (°C) 10.4 7.5 (4, .3ml 8.0 DO Img/ll 9.3 0.2 (4. .3ml 10.2 Cond l|iinhos/cml 185 — 235 Tanana R-Oownstream T (°C) 13.0 12.9 (7, .2ml 7.0 9-10 Sept. 1976 DO (mg/tl 10.2 9.9 (7, .2ml 12.7 Cond (umhos/cml 280 — 85 Tanana R-Upstream T (°C) 17.2 14.0 (2, .7ml ._ 4 June 1978 DO Img/l) 10.7 9.8 (2, .7ml — Cond (iimhos/cm) 286 ~ — 18 August 1978 T l°C) 15.2 8.2 (2, .7ml _. DO Img/ll 9.4 5.6 12. .7ml — 18 Sept. 1978 T ("o 9.0 6.2 ( 1. .4ml .»_ DO (mg/H 10.0 4.6 (1. .4ml — Cond (umhos/cml 280 -— •"" Sample sizes and variance estimates omitted to simplify Table. T = temperature. 00 dissolved oxygen. Cond = conductivity. 283 SUMMARY AND CONCLUSIONS Few changes in water quality parameters were measured that could be attributed to gravel removal; most of the observed changes were within the range of that expected by natural variation. The major reason for a lack of measurable effects was the age of the sites, as most were visited several years after mining had ceased. The few changes that were observed were related to physical changes in the rivers, generally due to a reduction in velocity and spreading of flow. 284 REFERENCES National Academy of Sciences. 1973. Water Quality Criteria 1972. Environ- mental Studies Board, National Academy of Engineering, Wash. D. C. 594 pp. Woodward-Clyde Consultants. 1976. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska. U. S. Fish and Wildlife Services. FWS/OBS 76/21. Wash. D. C. 127 pp. 285 EFFECTS OF GRAVEL REMOVAL ON AESTHETICS D. K. Hardinger INTRODUCTION Aesthetics pertains to manmade modifications of natural landscape fea- tures to a degree that public concern may be expressed. Aesthetic concerns of State and Federal government include maintenance of visual resource values by minimizing undesirable modifications to natural landscapes. Visual resource values of natural landscapes are the particular physical components of an area that have been identified as having high value based on any number of measurable criteria. These could include unique cultural, historical, recreational, geological, or biological significance. Typically the management objectives of an agency having statutory powers for maintaining visual resource values are to protect land areas identified as having high aesthetic values. The agencies may do this by diverting proposed construction to less valued locations, modifying the construction plan, or requiring the application of mitigating measures where construction-related visual impact proves unavoidable. Maintenance of visual resource values has become increasingly important to the American people. Federal legislation has recognized this concern by establishing the visual resource as an integral and coequal resource under the multiple-use concept of land management. At the same time, there is an increasing demand for other resource developments that may not be compatible This section was reviewed and input was provided by B. Sharky of Land Design North. 287 with the management of visual resources. In order to resolve potential con- flicts, it has become necessary to develop a system than can identify visual resources and provide measurable management standards that are practical to imp I ement . Numerous systems for identifying visual resource values and evaluating visual impact have been developed. The systems vary considerably both in procedures followed and criteria applied. On Federal lands there are two principal visual resource management (VRM) systems in use today. One was developed by the U.S. Forest Service and the other by the U.S. Bureau of Land Management (BLM). Both systems have the capability to: • Identify areas of significant visual resource value; • Establish land units with each unit having measurable, homogeneous qua I i t i es; and • Prioritize the land units through establishment of units of low visual quality, hence requiring minimal management protection, and units having high visual quality requiring maximum management protection. The major components of each system involve a systematic field inventory including (I) scenic quality or visual variety, (2) visual sensitivity, and (3) degree of visibility. Generally, the field inventories are conducted from an on-t he-ground perspective. Visibility from the air is generally not considered except under specialized circumstances. Definitions of the three key VRM inventory components of scenic quality, visual sensitivity, and degree of visibility follow. Inventoried systemat- ically using the BLM system, these components yield a land unit rating system divided into five classes. Each class provides various degrees of resource management control over prospective resource development proposals, including gravel removal operations from arctic and subarctic floodplains. 288 SCENIC QUALITY Establishing a scenic quality rating begins by using physiographic prov- inces to distinguish landscape character units having common visual qualities and to provide a regional context for the specific area being evaluated. Within each major landscape unit there may be areas having significant visual differences. These differences might include variations of typical landforms that would be classified as character rating units. Each rating unit is fur- ther classified according to the degree of scenic quality or variety as being distinctive, common, or minimal. Generally any landscape has recognizable parts that can be described in terms of form, line, color, and texture. These basic visual elements exert various degrees of influence and their composition will determine the scenic quality of a given landscape unit. The premise is that landscapes with the most variety or diversity have the greatest po- tential for high scenic value. Several key factors are inventoried in determining the scenic quality of the landscape and are used to delineate VRM land classes. • Land form. • Vegetat i on . • Water. • Co I or . • The influence of adjacent scenery. • Scarcity (distinctive features) or uniqueness. VISUAL SENSITIVITY Visual sensitivity levels measure the public concern for the scenic quality of the landscape and for the changes that may alter the existing landscape character. The degree of sensitivity is determined by user attitude and use demand (volume). User attitude can be measured by a survey of private citizens and public officials, or indirectly by public documents such as recreation plans, trail systems, scenic highways, and other items. These documents indicate areas of general concern. Use volume identifies areas of pedestrian and motorized vehicular use and rates them high, medium, or low 289 based upon frequency and duration of use. User attitude and use demand are frequently combined in a matrix to determine final sensitivity levels. DEGREE OF VISIBILITY A distance zone is ttie area that can be seen from a sensitivity area, and is described as foreground, middleground, background, or seldom seen. Distance zones are delineated on the premise that the ability to perceive change or detail in the landscape is a function of distance. Specific site information (Scenic Quality, Visual Sensitivity, and Degree of Visibility) is initially displayed on separate topographic maps. A hier- archy of importance is established and the maps are combined. The resulting classifications are the basis for defining minimum management objectives and the degree of acceptable alteration for each landscape classification. The determination of the degree of acceptable alteration for each landscape unit is defined utilizing a numerical rating system that enables a decision maker to see exactly what feature (landform, water, vegetation, structures) is being af fee ted and to what extent. This method a I lows some flexibility in determining appropriate mitigation measures. 290 APPROACH The aesthetic analysis of gravel removal from the 25 project study sites utilized the premises and criteria of the VRM system developed by the Bureau of Land Management. However, an actual VRM inventory and classification was conducted on a site by site basis rather than on a regional basis as would normally occur. Each project study site was analyzed for scenic quality, visual sensitivity, and degree of visibility. Project aerial and on-site ground photography, USGS togography maps, and project site descriptions were the primary data source for the scenic quality and degree of visibility anal- ysis. Visual sensitivity data sources are limited in Alaska; therefore, user attitude and use volume were interpreted from the public documents cited in the bibliography and by communications with persons familiar with the loca- tions under study. After the sites in each general region were inventoried for existing visual resources, a contrast evaluation was conducted. The contrast evaluation outlines specific visual effects of gravel removal according to BLM def i n i t i ons. 291 THE VISUAL RESOURCES OF THE STUDY REGIONS Characteristic landscape descriptions are needed in order to assess the degree of change or contrast that is created by floodplain gravel removal. The following section describes the physical characteristics of each region or site location in terms of the basic visual elements of form, line, color, and texture. Although site specific physical descriptions are found else- where in this text, the purpose here is to create an overal I impression of the landscape quality in the vicinity of the study sites. When available, information documenting public concern and use (or visual sensitivity) in each region is also included in this section. SEWARD PENINSULA Seen i c Qua I i ty Seward Peninsula sites include Gold Run Creek, Sinuk River, Washington Creek, Oregon Creek, Penny River, and Nome River. The typical landform in the vicinity of all sites is characterized by broad, smooth textured, rolling hills with moderate to gentle slopes (Figure 85). The hills are separated by sharp V-shaped valleys near stream headwaters; these valleys become wider near the coast. Al I study sites on the Seward Peninsula are located in narrow val leys or at the point where a narrow val ley opens into a broad val ley. The panorama at these sites includes both gentle and moderately steep slopes. Angular, rugged mountains are visible in the distance from al I Seward Penin- sula sites, but do not significantly influence or enhance the local scenic qua 11 ty. The study site rivers on the Seward Peninsula usually flow in sinuous configuration with moderate to swift currents. The Sinuk River is the largest 292 ""^'aariifiir " ^4^ Figure 85. Typical Seward Peninsula landform at Penny River. river and it flows in braided pattern through the study reach. The other rivers have a single well-defined active channel with occasional side channels or islands. The presence of occasional reaches of steeply eroded river bank do not create strong, visibly apparent vertical lines. Some river edges are of coarse texture with cobbles and boulders. All river systems enhance the scenic quality of the immediate surroundings, but they are not the most dominant element in the large scale landscape. In the Seward Peninsula, riparian vegetation grows in various densities and heights. In most cases low-growing shrubs (1-2 m) are interspersed with other ground cover species (herbaceous and woody). Islands frequently are vegetated with similar vegetative communities. The Penny River in particular has extensive, wide bands of tall (2 to 3 m) riparian willow. The greener shrub thicket vegetation also extends up adjacent val leys providing a sharp color and texture contrast with the matted brown tundra on the surrounding hillsides. Dense shrub thickets also are a common feature along old diversion ditches, seeps, and other water sources; these create contrasting bands and clumps of dense green color across the brown hillsides. 293 The predominant summer colors of the region are provided by the vege- tative patterns. Common patterns include: bright green near water sources and dull green or brown on the hillsides. During fall the floodplains turn bright yellow, while red and golden yellow colors dominate the hillsides. Ridges of nearby hills are barren and appear gray in color with occasional dark brown rock outcrops. Cultural modifications are visible from every site in the Seward Penin- sula. The Nome-Teller Highway intersects and/or parallels five of the region's study sites, and the Nome-Taylor Highway parallels the Nome River near the sixth study site in this region. The roadways are the most visible cultural modifications, but the lines they create generally blend into the lines of surrounding landscape. Several streams are crossed by bridges of varied de- sign. These bridges create vertical and horizontal lines that are not fre- quently found in these landscapes. Access roads frequently lead from main highways to river floodplains. Drainage ditches constructed during early gold mining periods frequently can be seen as they follow the contours of adjacent hillsides. These ditches were constructed to collect and provide water at upland gold mining sites. Several trails traverse the local terrain and are visually disruptive. Some cabins are situated within sight of roadways, but none are noticeable from within the study sites. There also is evidence of other gravel removal and gold mining sites throughout the region. Visual Sensitivity and Degree of Visibility The Seward Peninsula study sites are located within immediate or fore- ground view of the Nome-Teller and Nome-Taylor Highways. There are only three established highways for vehicle travel on the Seward Peninsula and all radi- ate from Nome, the largest population center on the peninsula. All of the study sites are within a 40 km radius of Nome. There is an established BLM campground about 24 km north of the Nome River study site. This campground and the historical gold mining districts near Nome attract additional summer tourist travel along these routes. Commercial tours of the peninsula usually begin in Nome and branch out along these roadways. Any changes or alterations of the landscape that occur in the foreground along these roadways would be 294 highly visible. However, lower use volume than in other parts of the State, and less resource agency concern for the quality of this landscape (no wild- life refuges, wild and scenic rivers, etc.), give the study areas only a moderate visual sensitivity. NORTH SLOPE Scenic Qua I i ty North Slope study sites include the Ugnuravik River, Aufeis Creek, Kuparuk River, Skeetercake Creek, Sagavanirktok River, Ivishak River, Shaviovik River, and the Kavik River. The Kuparuk River and the Ugnuravik River sites are located on the Arctic Coastal Plain which is characteris- tically flat to slightly rolling. The steeply incised river banks accentuate the strong horizontal line of the coastal plain and also provide vertical relief (Figure 86). The remaining sites are located in the Arctic Foothills Figure 86. Typical view of an Arctic Coastal Plain f loodp lain. which is a transition area between the coastal plain and the Brooks Mountain Range. Gentle, undulating slopes with occasional isolated, round and rolling 295 hills characterize the landform of the foothills (refer to Figures 4 and 5 in DESCRIPTION OF STUDY RIVERS). Incised river banks or terrace banks establish horizontal lines that contrast with the characteristic undulating terrain. The landform features appear to be smooth with few surface rock outcrops. Rivers, tributaries, lakes, and ponds are common features of the North Slope landscape. On the coastal plain the abundance of these water features comprise approximately 75 percent of the land's surface. However, no single landform or water feature stands out or is visual ly significant. The braided river systems with their islands create variations in line, texture, and color that contrast with the surrounding homogeneous landscape. The rivers of foothill region study sites are more visually significant elements in the landscape due to the diminishing frequency of other water features and their prominent, focal location traversing foothi I I val ley floors. The vegetation of the North Slope study sites is relatively rich in color and texture. Riparian vegetation usual ly consists of low-growing com- munities of dense willow thicket interspersed with herbaceous and woody ground cover species. These riparian communities develop irregular outlines created by irregular channel patterns and uneven texture. Occasionally there are concentrated stands of tal ler, more mature wi I low that become a visual focus due to the contrast in height with surrounding low-growing vegetation. The color variation of the North Slope landscape is varied particularly in the fall. The most significant color contrast exists between the greens of the riparian shrub thickets and the tans and browns of unvegetated flood- plains. Some form of cultural modification is evident near all North Slope sites. Most modifications are the result of oil and gas exploration. Several gravel access roads parallel and intersect the floodplains near many of the study sites. Gravel drill pads, camp pads, and airstrips are adjacent to several sites. These surface materials with various buildings sharply contrast the form, line, color, and texture of the surrounding undisturbed landscape. 296 In addition, the Trans-Alaska Pipeline and Haul Road are within II km and 1.5 km, respectively, of the Ivishak and Sagavan i rk tok River sites. These features are visible from the floodplain banks at both sites. The dominant visual feature of elevated sections of the Trans-Alaska Pipeline consists of the vertical pipe supports and the horizontal pipe. The rigid lines of both elements contrast sharply with surrounding undulating landscape. The North Slope scenery is unusual and intriguing. This vast landscape with its subtle variety provides a sustaining viewer interest and, therefore, yields a fairly high scenic quality rating. Visual Sensitivity and Degree of Visibility At the present time, there is little visitor or public use near the North Slope study areas. However, several sites are located within or adjacent to areas identified by various groups as lands of national interest. The Ivishak River, for instance, has been recommended as a wild and scenic river. These designations do not guarantee increased public use, but they are an expression of public concern for preservation of scenic quality. Increased use could result if and when the Haul Road is opened for public access. Mater- ial sites within view of the Haul Road would have increased degree of visi- bility and therefore higher visual sensitivity. NORTHERN INTERIOR Seen i c Qua I i t y The landscape of the Northern Interior is among the most spectacular scenery in Alaska. It includes the Dietrich River (two study sites). Middle Fork Koyukuk River (two study sites), Jim River, and Prospect Creek. The sites on the Dietrich River and Middle Fork Koyukuk R i ver-Ups tream are located in flat glaciated val leys surrounded by steep, rugged mountainous terrain (Figure 87). The steep angular mountain walls are often crested with massive light colored rock outcrop and cut by jagged ravines. Near the Middle Fork Koyukuk River-Downstream site and the Jim River and Prospect Creek sites 297 Figure 87. Dietrich River valley. the val ley widths fluctuate and mountainous features diminish in visual domi- nance (Figure 88). The slopes are more gentle and the surrounding mountains are more rounded in form. '■:*i:. Figure 88. Lower Middle Fork Koyukuk River valley. 298 River systems of the Northern Interior exert varying degrees of influence on overall scenic quality. The large, active f I oodp I a i n of the Dietrich River covers nearly one half of the valley floor. This river flows in braided pat- tern over much of its length. Numerous light colored unvegetated gravel bars in the active floodplain sharply contrast with the remaining vegetated valley f loor and val ley wa I Is. The Middle Fork Koyukuk River varies from a large, sinuous single channel to a braided system with a large main channel. Through- out, there are many abandoned channels, vegetated islands, and terraces. Both Jim River and Prospect Creek are smaller, sinuous to meandering and less dominant in local scenic quality than the Dietrich and Middle Fork Koyukuk Rivers. AM Northern Interior study sites are in an enclosed landscape where the rivers become a focal point given their prominent and central location. The vegetation along the floodplains and hillsides is a diverse mixture of coniferous and deciduous trees of varying ages and densities. Dark-green white spruce trees contrast with the rounded, lighter green deciduous trees and willow thickets. High-water and abandoned river channels have created broken patterns in the vegetation throughout the floodplain. A rich, complex visual texture has developed because of the variable heights and colors of the vegetative communities. Color variety is further enhanced by the gravel deposits in the flood- plains, local patterns of vegetation, and in some areas extensive rock out- crops. During fall, vegetative changes introduce another dimension of color variety with the seasonal colors of red, orange, and ye I low added to the land- scape. The most noticeable cultural modifications in the Northern Interior are those associated with the Trans-Alaska Pipeline System. Facilities adjacent to the study sites include construction and maintenance camps, airstrips, ma- terial and disposal sites, and elevated and buried pipeline. Spur dikes have been built into the floodplain in several locations along the Dietrich and Middle Fork Koyukuk Rivers. The light colored gravel materials used to con- struct the pipeline work pad, Haul Road, and camp facilities sharply contrast 299 with the rich natural color variety of this region. The pipeline and Haul Road often create contrasting lines in the natural landscape. The scenic quality of the Dietrich and Middle Fork Koyukuk Rivers can be characterized as a region of high diversity. This diversity is a result of a rich and complex texture of color, landform, and contrasts. The degree of diversity provides the region with a somewhat unique capability of accom- modating limited manmade encroachments in comparison with the North Slope landscape where manmade structures would produce highly visible results. Although the scenic quality is not as distinctive, Jim River and Prospect Creek have greater recreation potential than the Dietrich and Middle Fork Koyukuk Rivers. This recreation potential may have an overriding influence on the final outcome of the visual analysis. Visual Sensitivity and Degree of Visibility The Northern Interior (at the time of this evaluation) is accessible to the recreation and tourist oriented public only by air or by foot; hence, public use is limited at the present time. The Bureau of Land Management has several proposals that would affect the use patterns in this region if the Haul Road is opened to the public. Most development would be restricted to presently disturbed areas with an emphasis on maintaining scenic quality. Not all study sites are easily visible from the Haul Road because of screening qualities of the natural vegetation. However, current and proposed river recreation use would increase the amount of visible area. In addition, lands of national and State interest are adjacent to the Trans-Alaska Pipeline System Utility Corridor (proposed "d-2" lands). Hence, there is strong public interest in maintaining the scenic quality of this region. SOUTHERN INTERIOR Seen i c Qua I i ty Most study sites of this region (West Fork Tolovana River, McManus Creek, and Tanana River) have some similar landform characteristics. Rounded foot- 300 hills with moderately steep slopes surround the flat-bottomed West Fork Tolovana River valley and the narrow McManus Creek valley (Figure 89). Lower, gently rolling hills border one side of the Tanana River, while the opposite Figure 89. McManus Creek valley. side consists of a broad, flat plain. Rock outcrops and barren soil are usual- ly confined to the tops of the higher foothills surrounding these sites. Pheian Creek, however, is located in a mountainous river valley (Figure 90). The valley walls are steep and angular with rugged ridges of rock out- crop. Mountain glaciers provide added visual interest to the surrounding landscape. The Tanana River and Pheian Creek flow in braided configuration. The Tanana River has numerous gravel bars and vegetated islands in the active floodplain that contrast with each other in visual appearance. On the other hand, Pheian Creek has a gravel floodplain with little contrasting vegetation. The contiguous gray-white color sharply defines the Pheian Creek valley floor. 301 Figure 90. Phelan Creek valley. The West Fork Tolovana River and McManus Creek flow in sinuous configuration through heavily vegetated, more narrow floodplains and do not strongly domi- nate the surrounding landscape. The vegetation at most Southern Interior locations is a diverse mixture of deciduous-coniferous forest and riparian shrub thickets. The rounded decid- uous shrubs and trees contrast with the dark, slender white spruce. The West Fork Tolovana River and Tanana River floodplains have a particularly lush understory that increases the variety of texture patterns. The valley walls near most Southern Interior sites are less obviously patterned with a more sparse understory except near drainages. However, contrasting patches of dark and light green can still be seen in most locations. The color variety near the Southern Interior sites includes a complex mixture of greens, browns, grays, and tans with fall vegetative foliage adding reds, oranges, and yellows. 302 The Southern Interior sites are in the vicinity of many manmade modifi- cations. The Trans-Alaska Pipeline System is near the West Fork Tolovana River and Phelan Creek sites, with State highways, rural communities, and recrea- tional facilities present in the vicinity of all Southern Interior sites. These facilities, with their modifications of landform and vegetation pat- terns, detract from the overall scenic quality of the surrounding natural I andscape. Southern Interior sites, with the exception of Phelan Creek, have minimum or common scenic qualities because landforms are not unique and there are a relatively high number of cultural intrusions. Phelan Creek has more landform variety and in some sections is highly distinctive. Visual Sensitivity and Degree of Visibility The Southern Interior sites are located in the vicinity of some of the most heavily used recreation, tourist, and scenic areas in Alaska. In addi- tion, most sites are close to major Alaskan highways connecting the largest population centers in the state. Increasing recreational use of rivers (lead- ing to increased view area) is facilitated by convenient road access. Nearby campgrounds and waysides increase the viewing time in the landscape. All of these factors contribute to high visual sensitivity in the Southern I n ter i or . 303 EFFECTS OF GRAVEL REMOVAL ON VISUAL RESOURCES Gravel removal activities caused alterations in the landscape that in many cases were not visual ly harmonious with the surrounding landscapes. These alterations are discussed in this section in terms of contrast. Contrast is determined by the change in the form, line, color, and texture of character- istic landscape features such as landform, water, vegetation, and structures. The degree of contrast can vary widely; however, the significance of each contrast will depend upon the scenic quality and visual sensitivity of the characteristic landscape. The contrasts presented in the following sections generally denote a negative effect unless otherwise stated. Similar contrasts frequently exist at separate study sites in each region, hence discussions have been grouped by region with exceptions noted. SEWARD PENINSULA Gravel removal activities in the Seward Peninsula created the most signif- icant contrasts in local landform and water features of all study areas. The uneven texture or angular lines, or both, of gravel stockpiles and over- burden piles present at most Seward Peninsula sites, visually disrupt the existing smooth lines of the surrounding homogeneous landscape. Scraping and pit excavation have left contrasting rigid, rectangular lines at several site locations. The presence of water located throughout much of the gravel removal areas in unnatural shapes and configurations ac- centuates this contrast. The construction of access roads has introduced an additional contrasting form and line in this landscape. These features are particularly disruptive if there are several at one site (Nome River, Oregon Creek). Landform contrasts are more evident in this region because the vegeta- tion is relatively low growing and cannot effectively screen gravel removal 304 activities. The overall color contrast has been increased at all sites by removing riparian vegetation. However, gravel removal has not created signif- icant overall contrasts with the form, line, and texture of the existing vegetation patterns except at Penny River where the vegetation is much taller. Rigid blocks of vegetation now define some borders of the gravel removal area at Penny River, thus producing a contrast with the existing random pattern and height variations of the natural vegetation. NORTH SLOPE Very few significant contrasts are visible on the braided rivers of the North Slope. The rivers are large enough to visually absorb the changes in cnannel and island configuration. The banks, however, are a strong visual focus in many places and are more visual ly sensitive to change. The height of incisea banks necessitated the use of gravel fill ramps in many locations. Some ramps were partial ly removed after mining was completed and the remnants are still visible. In either case, the ramps produce a moderate contrast with the form and line of the river bank. The Kavik River is an example of strong contrast in the form and line of the I andf orm-water feature. Large portions of the bank were altered at this site. In addition, a large rectangu- lar scraped area adjoins the river. These lines are not unlike those of the nearby airstrip, but in this case they disrupt the visual linear flow of the river's edge. The removal of vegetation and overburden in this area has produced a color contrast that accentuates the unnatural rectangular lines of the disturbed area. Gravel removal created stronger contrasts along the smaller and/or single channel rivers in the North Slope region. The creation of additional warer channels and/or ponds at the Aufeis Creek and Skeetercake Creek has signifi- cantly disrupted the natural lines of each system. Removal of vegetated over- burden and stockpiling of gravel created additional contrasts in color and texture. The resulting broken textures and configurations at these sites contrast sharply with the existing natural landform and vegetation patterns. 305 NORTHERN INTERIOR The Northern Interior sites are general ly located in areas where patterns of manmade activity already exist and are visibly apparent. Gravel removal sites in vegetated floodplains developed the most significant visual con- trasts. Rectangular excavation boundaries contrast with the curvilinear shape of naturally vegetated floodplains. The removal of vegetation and overburden created color contrast at the Dietrich R i ver-Upstream, Middle Fork Koyukuk R i ver-Upstream, Jim River, and Prospect Creek. This contrast distinguishes the rectangular lines of the disturbed areas from the surroundings. Color contrast would not be as signif- icant at these sites if the disturbed area boundaries were developed in config- uration to reflect surrounding landform and vegetative patterns. Sites that have filled with water (Prospect Creek, Jim River, Dietrich Ri ver-Upstream, and Dietrich River-Downstream) have produced line and form contrasts because ponding is not a common element in the floodplains of this region. Angular diversion channels at Dietrich River-Upstream were equally contrasting with natural channel patterns. The abrupt and block-like shape of existing gravel stockpiles at Dietrich River-Upstream sharply contrasted with the flat terrain of Northern Interior river valleys. SOUTHERN INTERIOR The presence of tall white spruce-paper birch stands associated with specific site locations make the study sites of this region less visible from public roadways than sites studied in other regions. However, the Southern Interior is a high recreational use area and natural screens between roads and gravel removal areas are not total ly sufficient to keep the dis- turbed areas from public view. Landform contrast is the most obvious change in visual quality resulting from gravel mining at the Southern Interior sites. The West Fork Tolovana River, Tanana River-Upstream, and Tanana River-Downstream sites have rectan- 306 gular, flooded pits with steeply sloped banks. The angle of bank slope and pit shape contrast with the natural flat floodplain form and the curvilinear lines of the river systems. Where gravel stockpiles remain within the visible por- tions of the study site (such as at Phelan Creek) they create a contrasting unnatural form. 307 SUMMARY After studying the effects of gravel removal on visual resources at specific sites, some overall generalizations can be made. Certain landscape features or conditions will be similarly effected by gravel remova I i n a I I regions. The deciding factor in determining total impact will be the relative public sensitivity to the specific landscape. The same impact in two different areas may be judged differently depending upon public priority. Theoretically, the landscapes that are highly visible and highly regarded by the public will be more seriously affected than landscapes of lesser priority. The following summarizes the effect of gravel removal on generalized landscape features and briefly discusses public priority. Small, single channel rivers bordered with low-growing vegetation experi- enced the most dramatic visual impact. The location of gravel deposits on these rivers usually requires the removal of riparian vegetation and over- burden along incised banks. Along meandering and sinuous systems this pro- cedure frequently results in significantly altered river configuration. The vegetation removal causes a color change that clearly brings attention to the disturbed area. The remaining low-growing shrub vegetation is not of sufficient height to screen the disturbed area. Braided rivers with or without vegetated islands usual ly can visual ly absorb mining induced changes 'if the gravel removal occurs between the flood- plain banks. Any changes to the banks create noticeable visual contrasts. The most frequently observed contrast to river banks result from access roads and fill-ramps, cut banks, and mined banks. Tall, dense vegetation buffers surrounding the work area often screen many mining sites from public view at ground level. However, the removal of 308 vegetation from buffer areas at most study sites has caused unnatural line and color contrasts that draw attention to the disturbed areas. Color con- trasts are more visible from an elevated position where a viewer is looking down onto the site. Rectangular, water-filled excavation pits, due to their unnatural shape, generally create significant contrasts in all floodplain landscapes. The contrast is accentuated when the vegetation bordering the pit is tall and conforms to the rectangular shape. Sites that can be viewed from above, where the viewer is able to look down onto a site, generally results in high visibility potential particularly in areas of sparse or low-growing vegetation. Access roads also have resulted in significant contrasts in many study sites. Access roads frequently create a high degree of visual prominence and contrast where they traverse perpendicularly across existing slope con- tours. This contrast is more disruptive in regions of rolling or steep ter- rain, having sparse or low-growing vegetation, as exists on the Seward Penin- sula and North Slope. The presence of more than one access road can produce a multiplying effect with respect to increasing visual prominence. The presence of stockpiled gravel and overburden piles often increase visual prominence to a site. Often due to their height or linear shape, or both, the piled material tends to attract the viewer's attention to a site even though the site itself may not be clearly visible. Large stockpiles are detractive in most landscapes although less noticeable in broad floodplains surrounded by tall, highly patterned, mixed stands of vegetation. Tall vege- tation and terrain features can provide a visual screening effect particularly where the viewing location is at ground level. Areas having more or less homogeneous vegetation and terrain generally are more highly visible than those areas that are more diverse. The diverse landscape character types general ly can accommodate gravel removal partic- ularly at locations where the potential viewer is at a substantial distance 309 from the site or is at a similar elevation (ground level with respect to the site) . Visual prominence of a site tends to increase where vegetative clearing occurs along straight, long lines. This pattern is generally true in regions of both high and low landscape character diversity. Less visual contrast results where irregular clearing patterns have been accomplished. Site visi- bility is further reduced where natural vegetative recovery has occurred on sites cleared on irregular patterns. Four different regions of Alaska were included in this study and each region evokes a different public response to visual resources. The regions that appear to be the most publicly sensitive to change are the Northern and Southern Interior regions because of exceptional scenic quality or inten- sive public use. The visual effect of gravel mining activities is expected to be more scrutinized by the public in those areas. Visual standards for gravel removal areas should recognize this public sensitivity. 310 GEOTECHNICAL ENGINEERING CONSIDERATIONS OF GRAVEL REMOVAL H. P. Thomas and R. G. Tart, Jr. INTRODUCTION The initial geotechnical effort on ttie project consisted of a litera- ture review and evaluation of questionnaires sent to highway departments around the United States. Results of this effort were presented in a prelim- inary report (Woodward-Clyde Consultants 1976). This section presents the findings of a geotechnical review that consisted of an office evaluation of the limited data from the 25 study sites made available to the project geotechnical engineers. This section identifies general geotechnical consid- erations that should be considered in gravel removal projects. The major data sources were: the mining plans that varied greatly in detail from site to site (for some sites no mining plans are available); aerial photography that varied from site to site in scale, coverage (both historical and areal), and quality; and site photographs collected during biological and hydrological field inspections. This section is, in many cases, generic and general in its treatment because of the limitations of the available data. The objectives of this evaluation were to identify: 1) Engineering techniques that led to efficient development and opera- tion of gravel removal areas; 2) Engineering techniques that mitigated environmental disturbance; and 3) Engineering techniques that could have been used in various condi- tions that would have led to more efficient operation with less environmental disturbance. 3! I Volumes of gravel removed from each site ranged from approximately 8,000 m to 630,000 m , with the largest volumes removed from Dietrich Ri ver-Upstream, Phelan Creek, Aufeis Creek, and Sagavan irk tok River. Refer to Table 4. Scraping was the most common removal method used, but four sites were operated as pits and another four sites were operated as combinations of scrapes and pits. Nine of the sites were developed in connection with construction of the Trans-Alaska Pipeline System. Most North Slope sites were opened in connection with oil exploration and drilling activities, wh i le al I Seward Peninsula and most Southern Interior sites were developed in connection with local highway projects. More detailed information on site use is presented in DESCRIPTION OF STUDY RIVERS. Permafrost conditions at most of the study sites are unknown. There normal ly is a thaw bulb associated with rivers in permafrost areas. In continuous permafrost, the thaw bulb may be a transitory feature present only during summer flows. However, in discontinuous permafrost and for large rivers in continuous permafrost, the thaw bulb persists year-round although it may shrink considerably in winter. A 1969 study on the Sagavan i rk tok River I I km south of Prudhoe Bay (Sherman 1973) showed that in summer the thaw bulb associated with the main channel was 12 m deep and had a cross- 2 2 sectional area of 762 m . In winter, this thaw bulb shrank to 167 m with a maximum 7 m depth. Depending especially on whether underflow occurs, thaw bulbs may or may not be present outside the main channel. A major gravel use in arctic and subarctic Alaska is directly related to the need to provide a gravel overlay sufficient to carry traffic and to prevent permafrost degradation (progressive thawing). The minimum overlay thickness to prevent thawing can be calculated as a function of the local thawing index. The thickness is 1.5 m at Prudhoe Bay and increases as one moves southward (e.g., it is 2.1 m at Galbraith Lake and in Fairbanks it would approach 6m). A 1.5 m gravel overlay has generally been used for roads, drillpads, airstrips, and other permanent facilities at Prudhoe Bay. However, it has been shown that a 60-cm thick gravel overlay with 5 to 10 cm of polystyrene insulation is thermally equivalent to 1.5 to 2.1 m of gravel. This represents a 60 percent reduction in gravel thickness and a 64 percent 312 reduction in gravel quantity, considering a typical gravel pad with 1^:1 side slopes and a crest width of 10 m. Gravel needs during construction of the Trans-Alaska Pipeline System were reduced by using this solution for 110 km of the pipeline workpad on the North Slope. Depending upon relative costs of gravel and insulation, synthet i ca I I y- i nsu I ated embankments may or may not be less costly than their all-gravel counterparts (Wellman et al. 1976). 313 APPROACH The main factors considered in the geotechnical evaluation were site selection, access, operation, and rehabilitation. Primary information re- viewed for each site included mining plan information from permitting agen- cies, aerial photographs, ground photographs, and field notes taken by the project hydro I og i sts. Early in the review effort, a geotechnical fact sheet and evaluation form were developed and filled out for each site. The purpose of these forms was to assemble relevant information, to draw out observations of project personnel who had visited the sites, and to general ly focus the review effort. Although the geotechnical data base was very limited at a number of the study sites, it was bel ieved to be sufficient overal I to allow certain meaningful judgments to be drawn. The following sections contain geotechnical discussions related to gravel removal during principal stages in the life of a material site. 314 SITE SELECTION AND INVESTIGATION Selection of a gravel removal site often begins with a comparison of candidate floodplain and/or upland sites in the immediate use area. Upland sites are beyond the scope of this report and will not be further considered. The site selection process includes preliminary selection, site investigation, final selection, and mining plan preparation. PRELIMINARY SITE SELECTION Preliminary selection of one or more candidate sites results from assembling and reviewing available information followed by implementation of an appropriate selection procedure. Sources of Information Primary sources of information used in preliminary site selection are topographic maps, sur f i c i a I geologic maps, and aerial photographs. Topographic maps of 1:250,000 and 1:63,360 scale are available from the U.S. Geological Survey (USGS). Similar topographic maps are also available for Canadian arctic and subarctic regions. From these maps, one can obtain a general impression of the size and type of river, potential gravel availa- bility, desirable access routes, and proximity to the use area. The only currently available surficial geologic map of Alaska is the 1964 USGS map entitled "Surficial Geology of Alaska". With a scale of 1:1,584,000, this map does not show much detail. However, USGS recently published a potentially useful set of maps which cover the Trans-Alaska Pipeline route from'Prudhoe Bay to Valdez. 315 Aerial photographs frequently are the most useful sources of informa- tion. Stereo pairs are needed to show relief (e.g., height of banks) and a scale of not more than 1:12,000 is preferred. Color photographs are avail- able for some areas of the State, and black and white photography is avail- able for most areas of the State. For some areas, pre-existing aerial photo coverage can be purchased from local aerial survey companies. However, it is frequently worthwhile to have the area in question flown and photographed in order to obtain the needed coverage. From adequate aerial photographs, one can normally distinguish such features as the physical characteristics of the floodplain (e.g., channel configurations, flow regime, gravel availa- bility, vegetation patterns) and can select potential access routes and f ac i I i ty locat i ons. Preliminary Selection Procedure The procedure for selecting a gravel removal site usually involves identifying two or three alternative sources that appear to have sufficient quantities of gravel. These alternates are then compared either in an in- formal basis (usually minimizing haul distance) or in a more formal pro- cedure involving establishing criteria, evaluating significant factors, and ranking sites. The criteria would be specific to the situation, however, factors that may be considered include physical properties of the material available, haul distance, material site size and configuration needed to produce desired quantities, equipment available and equipment needed, re- quired site preparation (e.g., ramps, berms, dikes, overburden), river hydraulics, and floodplain access from nearest point. At this stage the anticipated life-span of the material site also should be considered. If it is desired to use the site for several consecutive years, or for two or more periods separated by inactive periods, the potential bed-load replenishment rate should be incorporated into site selection. It is generally assumed (See EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAULICS) that rivers of glacial and mountain origin, particularly near their headwaters, have greater potential for gravel replenishment than streams of foothill or coastal plain origin. Non-engineering aspects of site selection are dis- cussed in other sections of this report. 316 SITE INVESTIGATION The import-ance of an adequate on-the-ground site investigation cannot be overemphasized. At the Ugnuravik River site, the investigation stopped with an interpretation of aerial photographs. Subsequent site operations discovered that the gravel was merely a veneer and not present in sufficient quantities to meet project needs. In contrast, before construction of the Trans-Alaska Pipeline System rather extensive site investigations were conducted which significantly increased the knowledge of site gravel quan- tity and qua I i t y . Types of Data Several different types of data need to be obtained in a material site i nvest i gat i on. Aerial Extent and Depth of Deposit. Estimating the volume of material available depends on establishment of the aerial extent and depth of the deposit in question. If this volume is less than the needed volume, the site will be inadequate to satisfy the material needs. Hence, this is one of the most important types of data to be obtained. Thickness and Aerial Extent of Overburden. Gravel sites frequently have a covering of silt or organic material, over all or part of the site, which must be removed in order to expose underlying gravel. Mining may not be economical if more than about I m of overburden is present over most of the site. Homogeneity of Deposit. A deposit which appears suitable on the surface may be unsuitable at depth. This change in deposit quality frequently is a result of fluvial processes involving channel shifting, alternating erosion and deposition, and overbank flows associated with periodic flooding. Test pits or borings from several locations within the site should be analyzed to determine deposit quality. 317 Groundwater Table. It is important to establish the depth to the ground- water table together with spatial and temporal variations in this parameter. Groundwater conditions may vary widely throughout the year in response to changing river levels, thus, several measurements are preferable. The date of measurements should be carefully recorded. Extent of Permafrost. Although permafrost occurrence in the vicinity of rivers and streams can be highly erratic, it should be anticipated in arctic and subarctic regions. The presence or absence of permafrost can be an important factor in developing a gravel removal site. Field Techn i ques Both borings and test pits can be used for geotechnical exploration. Test pits are generally preferred in granular soils because of the diffi- culties of drilling and sampling in sma I I -d i ameter borings. However, borings can provide a good indication of overburden thickness, water table, perma- frost conditions, and presence and extent of unacceptable (e.g., silty) materials. These borings or test pits should extend to the depth of the anticipated gravel removal. The number of pits or borings would depend upon the size and variability of the site. Laboratory Testing The required laboratory testing effort varies. Sieve analyses are needed, as a minimum, to classify the material and establish its suitability for its intended use. For these tests, rather large (50 to 100 kg) bulk samples are desirable. Other tests that may be needed include hydrometer tests (if frost-susceptibility is a concern) and compaction tests if the gravel will be used to support structures. FINAL SITE SELECTION The final site selection is based upon the criteria analysis of the alternative sites. This analysis compares the characteristics of the ma- 318 terials found at the available sites to the needs of the project. A major portion of this analysis is the cost-benefit trade off of the options devel- oped during the site investigation process. Sites further from where the material is needed may have gravel that requires less processing; the re- duced processing cost may lower total costs despite the added cost of trans- port and road construction. In another case a more distant site may have an existing access road which would, on a cost basis, justify use of the more distant site rather than a closer site. In some instances, such as pipeline bedding and padding, rounded well-graded gravel might be preferable. Spe- cific gradation requirements may be necessary for subsurface drains. Uni- formly graded angular gravel may be a requirement for asphalt pavement aggregate. In final site selection the engineer makes trade offs to choose the site that will provide the required material at the least cost. This engineering analysis is then reviewed and biological resources, hydraulic factors, and aesthetic concerns are considered before the final site se I ect i on. MINING PLAN PREPARATION The agency having jurisdiction will generally require preparation and submittal of a mining plan. Minimum elements of the mining plan are: • Planned use of gravel, • Basis for determination of material quality and quantity (e.g., bor- ings, test pits, laboratory tests.) • Site configuration and depth, • Quant i t y limits, • Project schedules, • Overburden presence, • Access to site, • Buf f er I ocat i ons, • Operation plan, and • Rehabilitation plan. 319 Specifically, the mining plans should include at least the following i nf orma t i on : • A site sketch drawn to scale showing: project I ocat i on cross-sections of borrow areas, gravel source locations, existing or planned haul road locations, test pit or boring locations (if any); • An estimate of the volume of material that is needed; • An estimate of the volume of material that is anticipated at the ava liable s i tes; • An estimate of the properties of the material required; • An estimate of the properties of the in-situ materials; • An estimate of the type and amount of processing that will be required; • Project schedules for al I major activities; • Preliminary design features of any required support structures, such as access roads, processing plants, culverts, and bridges; and • Description of operational and rehab i I i t at i ona I aspects of site use. Plans prepared as described above should provide sufficient information to evaluate the appropriateness of the planned development of the gravel sources. Mining plans were prepared and submitted to the appropriate government agency for most of the 25 study sites. However, no mining plan information was found for the Washington Creek, Nome River, or Skeetercake Creek sites. The mining was apparently a trespass action at the upstream Aufeis Creek site and for initial gravel removal at the Kavik River site. Only results of a very limited site investigation were found for the Penny River site; only some correspondence was found for the Ugnuravik River site; and only a right-of-way permit was found for the McManus Creek site. Mining plan infor- mation reviewed ranged from sketchy (for the Seward Peninsula sites) to quite detailed (in the case of the Trans-Alaska Pipeline System sites). 320 SITE PREPARATION Having selected and gained approval to develop a gravel removal site, site preparation activities can begin. These activities may include construc- tion of access roads, removal of overburden, and construction of channel diversions and settling ponds. ACCESS As a part of most floodplain gravel removal operations, haul roads must be built to connect the site to the use location or existing roads. This construction poses no special engineering problems in non-permafrost areas or in areas where the permafrost is thaw-stable. However, in areas of ice- rich permafrost, protection of the tundra is of vital importance. From an engineering standpoint, t undr a- i nsu I a ted permafrost, as long as it remains frozen, is an excellent base or foundation for structures whether they be drill pads, roadways, pipelines, or other structures. When the permafrost begins to thaw two critical things happen. First, there is a tremendous loss in strength, and second, the thawing process is very difficult to stop. Thus, after the tundra is disturbed enough to allow the permafrost to begin this progressive thawing, the same area that formerly was an excellent base for structures becomes a very difficult, if not impossible, foundation problem for any engineering purpose. Drainage and other related problems also begin to develop and these can have significant adverse impacts on engineered structures. Access roads traversed ice-rich permafrost at several of the study sites with varying degrees of success. In general, where at least 0.5 m of gravel depth was used, permafrost integrity was maintained. However, at several sites (Ugnuravik River, Aufeis Creek, Skeetercake Creek, and Kuparuk 321 River) the access roads were less than 0.5 m in depth and subsidence fre- quent I y occurred. Access roads to a given site should be limited in number and confined to prepared surfaces. Both season of operation and long-term effects need to be considered in planning. Access to most of the study sites seemed to be appropriate and usually consisted of short gravel ramps and haul roads, sometimes including gravel bars within the river floodplain. The practice of constructing temporary gravel ramps, as at the Kuparuk, Sagavanirktok, Ivishak, and Shaviovik Rivers sites to provide access over incised permafrost river banks, reduces bank disturbance (Figure 91). How- Figure 91. Gravel ramp at Shaviovik River site providing access over a permafrost river bank. ever, cutting into permafrost banks, as was done at the Kavik River, can lead to severe thermal erosion and is not recommended. 322 Winter-Only Access Winter access to a floodplain site is generally easier than summer access because the surrounding terrain is frozen and river levels are low. However, even frozen organic mats need to be protected from mechanical crushing and ripping created from multiple passes over an unprotected access road while building snow or ice roads. The Ugnuravik River site provides an example of adverse long-term effects: access to the site was via a temporary winter trail across the frozen North Slope tundra. As far as is known, the trail was used only during the last week of March 1969. However, as was commonly done, the tussocks may have been bladed off to provide a smoother riding surface. Compaction and destruction of the vegetative mat started an irreversible process of thermal erosion. When the site was visited in summer, 1977, the road had eroded to a depth of 1.5 to 2.5 m over a distance of 90 to 120 m. Erosion was continuing, and a permanent scar had been created on the land- scape (Figure 92; also refer to Figure 83). Based on the current state of Figure 92. Thermal erosion near Ugnuravik River resulting from compaction and destruction of the vegetative mat over- lying ice-rich permafrost soils. 523 knowledge, a better solution would have been to construct a snow or ice road (Adam 1978). Year-Round Access A substantial gravel 11-3 m thickness) overlay is required where year- round access to a site is needed over ice-rich permafrost. However, place- ment of insulation beneath the gravel would reduce the thickness of overlay required. Year-round access roads must also be above flood stage of the river, which may require placement of culverts at high-water channels crossed by the road. OVERBURDEN REMOVAL The stripping of overburden involves the removal of any material cover- ing the gravel deposit. The overburden material, usually topsoil and or- ganics, is normally removed from the site and either stockpiled for later use in site rehabilitation or hauled to approved disposal sites. Stripping is normal ly done with graders, scrapers, or dozers. Overburden depths were not recorded at all of the study sites. However, where information was available, the depths ranged from a thin veneer (at six of the sites) to 0.9 m (at one of the sites) and the average was 0.3 m. CHANNEL DIVERSION For efficient gravel removal at some floodplain sites, it may be desir- able to divert river flows, especially those associated with subchannels, away from the area from which gravel is to be removed. This diversion is normally done by constructing earthen dikes or levees upstream from the site. Armoring of the upstream face and outer end of these structures may be necessary to provide erosion resistance. Erosion prevention is discussed further in EFFECTS OF GRAVEL REMOVAL ON RIVER HYDROLOGY AND HYDRAULICS. 324 SETTL ING PONDS It is necessary to wash gravel if the mined material has an appreciable silt content. When gravel is washed, it is essential that settling ponds be provided to allow silt to settle out before the wash water re-enters the river. These ponds should be of sufficient capacity to handle the daily volume of wash water or stream flow, or both, considering the settling velocity of the entrained silt particles. Design considerations for settling ponds can be found in Appendix F of the Guidelines Manual. 325 SITE OPERATION The basic elements of a gravel removal operation are excavation, trans- portation, and material processing. The details of equipment selection, scheduling, and operation procedures are dependent on the composition of the gravel, the season of operation, the topography, the haul distance, and the environmental characteristics of the site. EXCAVATION The two basic gravel removal techniques used at the 25 study sites were scraping and pit excavation. Table I identifies the technique used at the respect i ve si tes. R i pp i ng and Blast i ng Frequently, site operators prefer removing gravel in winter because water levels are low and access is easier. However, winter mining means excavating gravel in a frozen, possibly ice-saturated condition. At the study sites, if the gravel deposits were well above water levels and were low in frozen moisture, excavation by scraper was normally not difficult. Ripping frozen gravel was required at at least three of the sites (Middle Fork Koyukuk Ri ver-Upstream, Prospect Creek, and Phelan Creek). It is not known if blasting was utilized to remove gravel at any of the sites. Scrap i ng Scraping at larger sites is usual ly done with be! ly-dump scrapers. At smaller sites or remote sites, or both, D-9 or smaller caterpillar tractors 326 are frequently used. Scraped sites are usually dry when worked, however, caterpillar tractors can work in shal low water (possibly up to 0.5m). P i t Excavat i on Pit excavation is generally done with draglines or backhoes. Dewatering may or may not be necessary. At the study sites some of the more shal low pits were dewatered, but deeper pits, e.g., Dietrich R i ver-Upstream, West Fork Tolovana River, and Tanana River-Downstream were excavated underwater. Comparison of Techniques Some engineering and economic advantages and disadvantages of removing gravel via pits versus scraping are listed below. Advantages of Pits Versus Scraping • Greater quantity from sma I ler area. • Can work within confined property limits (if necessary). • Less clearing required. • Less stripping required. • Can provide silt trap. Disadvantages of Pits Versus Scraping • Dewatering or underwater excavation required. • May provide less gravel per unit time than scraper operation. • Cannot be restored as closely to original condition. TRANSPORTATION AND STOCKPILING Transportation of gravel from the material site to the stockpile or processing plant may be done with scrapers or front-end loaders and dump trucks. Stockpiling gravel removal operations greatly reduces scheduling problems. It is possible to load trucks directly for long-haul transport to ultimate-use areas without stockpiling, but a great deal of coordination is 327 required between the excavating and transporting activities. It is advan- tageous to maintain a stockpl le of at least moderate size to serve as a buffer between excavating and transporting. Gravel stockpiles remained on or immediately adjacent to nine of the study sites, however, only Dietrich R i ver-Upstream, Jim River, and Phelan Creek stockpi les were sti I I being used. PROCESSING Gravel processing can involve screening, washing, crushing, mixing, or combinations of these. Materials of the study sites frequently were fairly uniform, subrounded to well-rounded, hard gravels with varying amounts of sand and cobbles. Such materials are suitable for road embankments with little or no processing. However, silt content should be limited to approxi- mately 10 percent to minimize frost susceptibility. Processing apparently was only conducted at those study sites used for construction of the Trans- Alaska Pipeline System where screening and some crushing were done to pro- duce bedding and padding material for the below-ground pipeline. 328 SITE REHAB I L I TAT I ON Engineering concerns contribute to rehabilitation mainly if future site development (e.g., erecting of structures) is planned. In this situation, long-term integrity of structures is the primary concern of site rehabilita- tion. Otherwise, the primary purpose of site rehabilitation is erosion control. The main function of erosion control is to prevent degradation of disturbed and adjacent areas. Some rehabilitation was done at all study sites worked s i nee 1972. There was no evidence of rehabilitation having been done at any of the older sites. Where final site grading was conducted, it typically included sloping or flattening of stockpiles and overburden piles to blend with the terrain, contouring the site to a maximum 2: I slope, and removal of gravel ramps (not done at the Ivishak and Shaviovik Rivers). 329 REFERENCES Adam, K. M. 1978. Winter Road Construction Techniques, pp. 429-440. In Proceedings of ASCE Conference on Applied Techniques for Cold Environ- ments. Vol. I. Anchorage, Alaska. Sherman, R. G. 1973. A Groundwater Supply for an Oil Camp near Prudhoe Bay, Arctic Alaska, pp. 469-472. In Proceedings of the Second Inter- national Conference on Permafrost. Yakutsk, USSR. Wellman, J. H., Clarke, E. S., and Condo, A. C. 1976. Design and Construc- tion of Synthetically Insulated Gravel Pads in the Alaskan Arctic, pp. 62-85. In Proceedings of Second International Symposium on Cold Regions Engineering. Fairbanks, Alaska. Woodward-Clyde Consultants. 1976. Preliminary Report - Gravel Removal Studies in Selected Arctic and Sub-Arctic Streams in Alaska. U. S. Fish and Wildlife Service. FWS/OBS 76/21. Wash. D. C. 127 pp. 550 INTERDISCIPLINARY OVERVIEW OF GRAVEL REMOVAL E. H. Follmann^ INTRODUCTION This chapter presents a general overview of the effects of gravel removal in contrast to the preceding disciplinary chapters that rely more heavily on analytical treatments of data collected at the 25 study sites. Each of the Major Variables identified in the Matrix (Table I) is discussed relative to its influence on the effects of a gravel removal operation. These characteristics directed the early phases of the study, including the site investigations, and form, for the most part, the framework of the gravel removal guidelines. The disciplinary chapters on gravel removal effects did not necessarily treat each of these characteristics because some were not relevant or they did not influence the evaluations or syntheses sufficiently to warrant individual attention. Thus, this overview chapter constitutes the functional bridge between the Guidelines Manual and the Technical Report. Few problems were encountered in the discussion of the Physical Site Characteristics and their interaction with gravel removal projects because the categories are mutual ly discrete, i.e., a river cannot be both meander- ing and straight within the study reach. The categories under each of the Gravel Removal Area Characteristics, however, are not mutually exclusive and, thus, cause difficulty in the development of that discussion. The sites selected encompassed at least several individual locations from which gravel E. H. Follmann is presently associated with the Institute of Arctic Biology of the University of Alaska. 331 was removed. Sites such as Aufeis Creek on the North Slope and Penny River on the Seward Peninsula each included 8 of the 12 specific site locations that were possible (Table I). This complexity made it difficult to identify any specific floodplain changes with specific gravel removal locations. For these sites, the overall effect on the floodplain resulted from the total gravel removal operation and specific effects were masked. The problem of sites with multiple Gravel Removal Area Characteristics was unavoidable because almost all of the over 500 sites originally considered reflected the same situation. The major result is that, in some cases, generalities are discussed with little or no reference to specific material sites. If none of the sites clearly exhibited the relationship being discussed, none were cited as examples. However, the generalities discussed are considered ac- curate because of the analyses and conclusions reached in the preceding disciplinary chapters. 332 PHYSICAL SITE CHARACTERISTICS The Physical Site Characteristics considered in this project were: drainage basin size, channel width, channel configuration, channel slope, and stream origin (Table I). Following study of the 25 material sites and analyses of data, it was established that channel configuration was the most important floodplain characteristic affecting environmental change when combined with gravel removal activities. Drainage basin size (channel width) was found to be less significant, and channel slope and stream origin were found to have little influence on the effects of gravel removal. The follow- ing discussion is subdivided according to these categories. CHANNEL CONFIGURATION The channel configuration or pattern of a river is the shape of the river channel (s) as seen from the air. The channel configurations considered in this study were braided, split, meandering, sinuous, and straight. Bra i ded A river with a braided channel pattern typical ly contains two or more interconnecting channels separated by unvegetated gravel bars, sparsely vege- tated islands and, occasionally, heavily vegetated islands. Its floodplain is typically wide and sparsely vegetated and contains numerous high-water channels. The lateral stability of these systems is quite low within the boundaries of the active floodplain. Four braided systems used for material sites were studied. Ivishak River on the North Slope, Dietrich River in the Northern Interior, and Tanana River and Phelan Creek in the Southern Interior. These systems usu- 333 ally contain large quantities of gravel and, therefore, are often utilized as gravel sources (Figure 93). The bed load carrying capacity of these rivers is large, thus facilitating the replenishment of extracted gravels after site closure. Braided river systems are dynamic and lateral shifting of channels from year-to-year is common, therefore, any channel shifting resulting from lower- ing bars through gravel removal would be similar to the natural processes. For example, any diversion of a channel through an area that was lowered by the removal of gravel possibly would have occurred naturally sometime in the future. Material sites in these areas typically are scraped because required quantities of gravel usually can be obtained over large areal extents and it Is more efficient to work a site above the existing water level. Due to the bed load carrying capacity of these systems, the typical shal low scraped sites are subject to sedimentation rates similar to natural depressions occurring in these floodplains. Therefore, the minded sites can return relatively quickly to near natural conditions. This recovery is particularly true if the site is located near the active channel. An example of rapid recovery is the Ivishak River site, which was shallow scraped over a large area of unvegetated gravel bars. After several years the only evidence of gravel mining is the presence of access roads and fill ramps that connected the material site with an airstrip and drill pad. Long-term effects of gravel removal on water quality were not evident at the four sites located in braided systems. Due to the relative insta- bility of channels in a braided river system, any channels routed through an abandoned material site probably would be affected in a manner similar to a channel being rerouted due to natural hydraulic processes. An exception would be where an aliquot of a material site was used as a settling pond during a gravel removal operation. The accumulated fines could be suspended during subsequent high flows if this material was not armored and was left in the depression during site closure. None of these situations was en- countered at the study sites, however, the possibility would exist in simi- lar site conditions. 334 Stra i gh t Figure 93. Configurations of study rivers. 335 The aquatic organisms in braided systems are adapted to the seasonal dynamics of the channels and, therefore, any channel changes resulting from gravel removal operations provide situations for which the organisms are already adapted. An exception to this generalization occurs where a pit is separated from the active channel (Tanana River-Downstream) or is within the floodplain (Dietrich R i ver-Upstream) and connected to an active channel. In these cases, organisms that are more adapted to lentic environments become established. Also, certain fish species may use the calmer waters of these pits for spawning, rearing, and feeding areas. These pit sites are the excep- tion, because scraping is the usual procedure selected to excavate sites in braided systems. Excavating aspects are discussed further in the following section on Types of Gravel Removal. Terrestrial species that utilize braided river systems similarly are little affected by the usual scraping operation. Since non-vegetated bars are favored gravel removal sites, few sma I I mammals or passerines are af- fected. The water-associated birds that use the various channels and back- waters for feeding are also little affected by the material sites because the usual result of these operations is to provide habitats already present. Due to the dependence of sma I I mammals and passerines on vegetated islands, gravel bars, and banks present in braided systems, any removal of vegetation to expose a gravel deposit would total ly displace birds and eliminate small mammals from the disturbed site. Similarly, these areas, which often have associated dense shrub thickets, are used by moose and ptarmigan, especially during winter. Loss of this habitat would cause lo- calized displacement of these animals. Maintenance of the scenic quality of an area can be achieved by de- signing a material site to complement the natural setting. Material sites in braided systems did not detract from the visual quality of the floodplain where gravel removal was restricted to unvegetated gravel bars. The ex- pansive floodplains typical of these systems are somewhat uniform in ap- pearance, yet the numerous channels and gravel bars endow these areas with a complexity that permits material sites to be located with little effect. 336 The usual mining technique for these sites is to scrape unvegetated gravel bars rather than to excavate deeply, thus, any rearrangement of channels through an abandoned site would closely resemble the natural annual pro- cesses of lateral channel migration. In summary, braided river floodplains can be desirable locations for extracting gravels (Table 40). The abundance of well graded materials and the potentially small effect on the physical, biological, and aesthetic char- acteristics suggest the desirability of these areas for material sites. Th i s conclusion assumes that the procedures of shallow scraping of unvegetated gravel bars with minimal disturbance to active channels, banks, and vege- tated areas, and complete rehabilitation of sites during site closure, are adhered to. Sp I i t Channe I A river with a split channel pattern has numerous islands dividing the flow into two channels. The islands and banks are usually heavily vege- tated and stable (Figure 93). The channels tend to be narrower and deeper and the floodplain narrower than in a braided system. Four split channel rivers were included in this study: the Kavik, Kuparuk, and Sagavan irktok Rivers on the North Slope and the Sinuk River on the Seward Peninsula. Although the bed load carrying capacity of split channel rivers is less than for braided systems, they often have a greater carrying capacity than equivalently sized meandering or sinuous rivers. The narrower flood- plains and lack of numerous gravel bars restrict the extent of potential gravel removal areas. Channels, islands, and banks are often used for extrac- tion, as was the case at the four sites studied. Islands and banks typically are vegetated and relatively stable, consequently, there is a direct effect on small mammals, passerines, ptarmigan, and moose utilizing these areas. The long-term terrestrial disturbance is directly related to the extent of vegetation removal and the rehabilitation practices used during site clo- sure. 337 o c o O — CT) X l_ (0 c aj .- ^ (o T3 > c i: c c o o COrOODO aSinCNCD J — — d J — f^ - 3) 00 in r-. r^ CN ; r>j — o K-» N-i fN (*-"0(n CTl "O *- (U >. — in — ro in ^ ^ in m fo ^ o (U 0) o (1) o 0) 0) v o 0 o o z z 2 CO l/l z z z z z to to to in z Z z z to in *- ^ *- c 4- ^ ■*. «- «- <0 a Kfi (jy J£ 3 o 1 1 O cr cc JC a: j£ cc \- £ £ (J J£ o u a: JC c a: «) c 3 (D £ I. k. 10 j£ U > ^ in 3 o a — > ^ — 3 flj 10 q: cr 3 3 to i: irt ic o JC J£ 3 3 u >. >- OTtX > > o O in c cr o o X it: 3 - >' :- — C J= c 10 in c C ffl u. u. S (0 0) en £ u S Q- z 3 to S S £ II o E •" II (0 — 1. II 10 Q. — 336 Lowering islands and banks by removing gravel, even if maintained above the existing water level, can result in reduced stability of channels during high water. Material sites will then be inundated at least tempo- rarily. Spreading water over a broader area reduces its velocity, causing deposition of suspended and bed load materials. Some of this reduced vel- ocity may function to replenish materials in the abandoned material site but this process would probably require a longer period than would be expected in a braided system. Spreading of water and reduction of velocity is conducive to changing water temperatures during the open water season. Altered water temperatures may influence the abundance and diversity of aquatic biota by altering the amount of usable habitat for particular species. The reduced stability of the channels that could occur after site closure could be detrimental to the establishment of permanent biotic popu- lations, in particular, benthic organisms. In addition, entrapment of fish in pockets and pools in the disturbed site may occur as water recedes into the active channels following high-water conditions. The increased deposition of both suspended and bed load materials could be detrimental to the establishment of benthic communities. Fine materials would likely be deposited in these areas, thus changes in the structure of benthic communities could be expected. These changes would be from organisms adapted to coarse substrate to those able to exist on finer less stable substrate. Changing channel configuration by removing islands, removing gravel deposits from banks, and locally widening the active f I oodp I a i n will affect the scenic quality of an area. This aesthetic effect was quite noticeable at the Sinuk and Kavik River sites where care was not taken to preserve natural contours and channel configurations. In addition, stockpiles and remnants of diversion berms were left in place. The net effect of these conditions was to form a major contrast with the natural conditions occurring both upstream and downstream of the site. 339 In summary, the split channel system is one that contains a relatively large quantity of gravel material, but its narrow floodplain with stable islands and banks restricts the areal extent where gravel can be easily obtained. Use of vegetated areas will directly affect terrestrial organisms by either complete removal or displacement to undisturbed areas. Similarly, the tendency for localized widening of the floodplain will reduce lateral stability of channels, facilitate the possible formation of a braided chan- nel pattern, decrease water velocity, increase sedimentation rates and, perhaps, increase water temperature. These changes will affect aquatic organisms by increasing secondary productivity, by changing benthic com- munity structure, by providing rearing areas for some species of fish, and perhaps by affording situations conducive to fish entrapment (Table 40). Meander i ng A meandering river winds back and forth within the floodplain. The meandering channel shifts downval ley by a regular pattern of erosion and deposition. Few islands are found in this type of river and gravel deposits typical ly are found on the point bars at the insides of meanders (Figure 95). Sediment transport in meandering systems is usually less than for braided and split-channel river systems of equivalent size. The size of individual gravel deposits in a meandering river depends on the size of the river. On a large river, point bars can be quite ex- tensive wh i le on sma I ler rivers the point bars are char acter i st i ca I ly smal- ler. The areal extent of these gravel bars determines, to a large extent, the degree of change which gravel extraction has on a meandering system. For example, if a large point bar is used to supply gravel for a sma I I project, the operation of a material site may cause little change to the river sys- tem. However, when projects with large gravel requirements are situated close to a sma I I meandering river or where the gravel requirements exceed that available on a large point bar, potential effects to the river system increase greatly. The alternative mining procedures are to completely remove the point bar, use several point bars, or remove vegetated deposits back from the channel. In al I cases, varying degrees of impact can be expected, but all will depend on the manner in which the gravel is extracted. 340 Four material sites on meandering systems were studied on this project (Table I). Two were dug as pits and two were scraped. Pit Sites. The material sites at Prospect Creek and West Fork Tolovana River were dug in abandoned channels. In neither case was there a change in the lateral stability of the active channel. There was loss of terrestrial vegetation and associated fauna because the material sites were located back from the active channels. Aquatic fauna in the active channel apparently did not change. Change, if any, was due to the presence of an adjacent flood- ed pit. Similarly, water quality did not change in the active river channels but, as expected, water quality in the pit was different from that in the active channel. These differences and changes are discussed in the section on Type of Gravel Removal because they were not unique to meandering sys- t ems . Formation of a permanently flooded pit within a floodplain, that other- wise contains few ponds or lakes, changes the appearance of the area by in- creasing the diversity of physical features. These pits are quite visible when seen from the air or from a high terrestrial vantage point. Ta I I vege- tation in the areas of these two material sites contributed greatly to blocking view of the sites. Many meandering river floodplains contain a multitude of oxbow lakes that are formed by channel cutoffs. In these cases, a pit could blend easily into the natural landscape, thus greatly reducing the visual effect of gravel removal operations. However, most pits are dug with angular perim- eters which create a visual contrast in the floodplain. This contrast is a generic problem and will be discussed further under Type of Gravel Removal. Scraped Sites. The material sites on Aufeis Creek and Skeetercake Creek were scraped. The environmental changes were quite different at the two sites resulting principally from differences in their locations relative to the channel (Table 4-0). The gravel at Aufeis Creek was scraped from across the entire channel, which changed the channel from a single to a braided configuration. The short-term influence was so severe that surface flow was 341 nonexistent the year following site closure but, over 3 years surface flow v^jas re-established. Although the site was not studied when surface flow was absent, the effect on fish would have been to prohibit passage. Epibenthic communities would have been reduced due to the lack of surface water. Fol- lowing re-establishment of surface flow, benthic communities characteristic of riffle zones would be most common due to channel spread and reduced water depth. The change from a single channel to a braided channel can significantly affect the local distribution of aquatic organisms. The altered community would be similar to that typically found in a naturally braided system. Reduced water velocity enhances sediment deposition and can alter water temperatures. During the study, changes in water temperature were noted between the upstream and disturbed sample areas, but a difference in sus- pended solids was not found. The impact on the terrestrial environment frequently entails removal of vegetation and other habitats along the bank. Little change to the ter- restrial environment would be expected when gravel is mined only on unvege- tated gravel bars, unless the hydraulic characteristics of the channel are changed significantly following site closure. Also, little change would be expected in the scenic quality of an area as a result of gravel removal, unless vegetation is removed. At Aufeis Creek, changes in both the ter- restrial environment and scenic quality resulted from the gravel removal operation because of the area disturbed, the site location, and operating procedures that were used, none of which complemented the floodplain char- acter i st i cs. At Skeetercake Creek the hydraulic changes were somewhat different. The exposed gravel deposits were limited because this was a small river. Thus, gravel was mined from vegetated areas in the floodplain, with concom- itant effects on the terrestrial fauna. The gravel removal activity affected channel stability by facilitating a channel cutoff, however, the channel did not braid due, at least in part, to the restricted floodplain. The cutoff formed an oxbow lake in the abandoned site. The floodplain in this reach of 342 the river had few oxbow cutoffs, consequently, mining changed the appearance of the area. However, the presence of overburden and gravel stockpiles detracted far more than the altered channel. Aquatic habitat changes at Skeetercake Creek were not as great as would be expected if the channel had become braided. The narrowness of the natural channel imparted a greater significance to the value of bank vegetation. Loss of this cover can change the distribution of fishes. The change from an incised channel to a shallow riffle area through the abandoned site caused the water temperature, during the study, to be higher in the disturbed area than upstream. However, changes in suspended solids were not noted. Summary. Scraping point bars can have little environmental effect assuming that the operation is conducted in a manner that minimizes changes to the hydraulic characteristics of the channel and adjacent vegetated areas. If change is minimized, the effects on aquatic and terrestrial biota, and water and scenic quality are greatly minimized. Meandering rivers provide usable deposits of gravel from point bars, in inactive floodplains, and terraces. The potential effects on such a system vary depending on whether only point bars are used or whether the adjacent inactive f I oodp I a i n and terrace also are mined. Sites in inactive flood- plains and terraces often are dug as pits while point bars in active flood- plains are scraped. Pit sites remote from the active channel have caused some problems during spring breakup at sites visited during site selection, but not studied as primary sites in this project (unpublished data). When channels are blocked with ice, melt water must flow over the ice and may overflow the bank and spread across the entire floodplain. Pits located in these flood- plains are then subject to filling which can facilitate diversion of flow through the site. This diversion is particularly possible where pits are dug within the inside of a meander. Depending on the size and inherent stability of the undisturbed buffer between the pit and channel, the flow may cut 343 through the buffer zone and permanently divert flow. Ultimately, the meander will be cut off through sediment deposition and form an oxbow lake. Other effects can be anticipated when pits are dug in the f I oodp I a i n of meandering systems, however, they are characteristic of pit mining. There- fore, these aspects are discussed under Type of Gravel Removal. S i nuous Sinuous channels are similar to meandering channels except that the winding pattern is less pronounced. The channel may contain smaller point bars and have less tendency for downval ley shifting. Also, the channels are more stable with respect to lateral shifting. Ten of the sites studied on this project were on sinuous rivers (Figure 95). Their similarity to meandering channels suggests that the effects from gravel extraction are also similar, with the major influence determined primarily by the site location and the removal method. Due to this simi- larity only a few characteristics of mining gravel at sinuous channels are d i scussed . The sma I ler point bars in sinuous rivers, as compared to meandering rivers, limit the quantity of exposed gravel that is locally available for removal. This limitation can magnify the need for using multiple point bars or vegetated areas back from the channel to fulfill the gravel requirements of larger projects. Floodplain areas adjacent to the channel contain gravel deposits that are typically overgrown with vegetation. Floodplain width usually is roughly equivalent to the meander belt width, thus, the floodplain of a sinuous river tends to be narrower than in a meandering system. Therefore, the area in the floodplain that is available for gravel extraction is more limited. This places restrictions on the areal extent of potential gravel resources, and may require that a greater length of floodplain be used to extract grave I . 344 The potential effects of removing gravel from sinuous ctiannel rivers are increased because of these limitations. If point bars are scraped too deeply, or if incised banks and the adjacent floodplains are disturbed, the potential for decreasing channel stability is greatly enhanced. The initial disturbance from site clearing, and the changes resulting from a poorly located and operated site, will have multiple effects. The decreased channel stability and tendency for braiding will affect both benthos and fish by altering aquatic habitats. Benthic communities adapted to riffles, fine sediment bottoms, and a relatively unstable bottom, will become established. Loss of bank cover and potentially reduced current in the disturbed site will affect fish distribution and perhaps species composition. In addition, reducing water depth and velocity could change water temperatures and affect the level of dissolved oxygen. Fish could become trapped in the disturbed site when water recedes following high f I ows. Terrestrial vegetated habitat will be destroyed when the floodplain adjacent to the channel is used as a material site. This destruction of vegetation will cause either elimination or displacement of terrestrial fauna. If the stream banks are affected the decreased hydraulic stability in the area could reduce the potential for re-establishment of vegetative com- munities, thus creating a long-term rehabilitation problem. Gravel removal from a sinuous river will have effects on the scenic quality similar to those discussed for a meandering system. The degree of effect is fully dependent on the diversity of landforms in the area of the site and the amount of disturbance. Single channel river systems are scen- ically more sensitive than multiple channel systems particularly those single channel rivers located in areas with low growing vegetation, such as on the North Slope. In summary, the amount of environmental change that can be anticipated in a sinuous river system is largely dependent on the location of the ma- terial site and the methods of operation. Anticipated effects are similar to 345 those for a meandering system but, because floodplains generally are more narrow and contain smaller point bars, the potential for permanent altera- tion is generally greater (Table 40). Proper placement of the material site and operational procedures can minimize permanent change and these should be selected to prevent or minimize changes to the hydraulic characteristics of the channel . Stra i ght Straight channel patterns are less common than other types. The thalweg of a straight river typically winds back and forth within the channel. Gravel bars form opposite where the thalweg approaches the side of the channel (Figure 93). These gravel bars may not be exposed during high flow. Banks of straight systems typically are stable and floodplains are usually narrow. These river systems are considered to be an unusual configuration in transition to some other configuration. Only the material site studied at Oregon Creek was situated on a straight channel system. As with other types of single channel systems the major potential effect from scraping floodplain gravels is decreased stability of the chan- nel and a tendency to develop a braided configuration. These are probable occurrences because of the typically narrow floodplains and the limited number of exposed bars available. Often the adjacent floodplain will have to be disturbed, or even the channel itself, because of the limited area avail- able. The Oregon Creek site typified the extensive long-term changes that can occur when gravel is removed from within the channel and the adjacent floodplain (Table 40). The channel stability was greatly reduced and the channel had become braided within the confines of the abandoned site. These conditions exist 13 years after the site was closed and probably will remain in that condition for many more years. The change from a single to a braided channel alters water quality parameters and aquatic biota as discussed in previous sections on sinuous and meandering systems. These alterations include the potential for changing water temperature and increasing sedimentation in the disturbed site where 34-6 the water fans out and becomes shal lower and slower in velocity. Dissolved oxygen and conductivity levels can also be altered. Benthic communities may change from a community associated with the relatively stable channel of a straight river to one that is better adapted to the less stable substrate characteristic of braided areas. Removal or alteration of vegetated banks and changes in pool:riffle ratios can alter the distribution of fish within the immediate vicinity of the disturbance. Fish passage is obstructed if the spreading of water sufficiently reduce its depth. The disturbances at the Oregon Creek site provided a situation con- ducive to the formation of aufeis. Aufeis could have direct effects on fish by eliminating or greatly reducing the flow downstream from the ice field, thus threatening overwintering areas and spawning beds. Similarly, during breakup, delayed thawing of the ice field could obstruct fish passage. Benthic communities would be later in establishing at the disturbed site due to the delayed melt of the ice field. The terrestrial environment will almost always be subject to distur- bance for any site situated on a straight channel river. This vulnerability is due to the rarity of large exposed gravel bars in the channel which necessitates mining the adjacent vegetated f I oodp I a i n banks or terrace. At the Oregon Creek site the vegetated overburden was removed and placed in a row at the edge of the terrace. The gravel was removed from the exposed area and from within the channel causing extensive spreading of the flow through the exposed floodplain. Inundation of this area during high flow and the build-up of an auf e is field greatly minimized the potential for stabili- zation and revegetation of the disturbed area. This stabilization and revege- tation had not occurred after 13 years, thus the likelihood of the site revegetating in the near future is remote. The appearance of the floodplain was greatly affected at the Oregon Creek site. This altered appearance will exist for a long time and will only diminish when the channel begins to narrow and when adjacent areas revege- tate. The potential for major changes in the appearance of a straight chan- nel floodplain, that is mined, is great because of the limited availability 347 of exposed gravels, which necessitates the disturbance of adjacent vegetated areas. The magnitude of effect increases with a decrease in river size. In general, the rarity of straight channel rivers probably is fortunate from the standpoint of gravel requirements. The relatively few exposed gravel deposits and the narrow floodplains suggest the major problems that can result from gravel removal operations in these systems. Major distur- bances probably will occur in any river of this type unless precautions are taken to protect the area. When mining is restricted to exposed gravel deposits a major length of floodplain will be disturbed if gravel require- ments are large. The latter problem can be prevented by restricting mining to the adjacent vegetated floodplain. Straight channel systems should be avoided where it is possible to select alternate areas to mine. DRAINAGE BASIN SIZE (CHANNEL WIDTH) Drainage basin size and channel width are closely related from a hydro- logical standpoint and analysis of only the former would be sufficient for assessing change from gravel removal activities. However, channel width was included in the Major Variable Matrix (Table I) because it is a measure- ment easily obtainable in the field while drainage basin must often be estimated from topographical maps. Because of the close relationship between these two parameters, the following discussion applies to both. Drainage basin size (channel width) was considered to be the second most important Physical Site Characteristic influencing the amount of change in a floodplain from gravel removal activities. In general, the effects of mining were considerably greater on sma I I rivers than on large ones. The determining factor is the amount of exposed gravel material available within the floodplain. In larger systems, gravel deposits can be numerous and any given deposit usually contains a large quantity of material. The situation is the opposite in a sma I I river - the few exposed deposits general ly do not contain much material. 348 In large rivers, a given amount of gravel can be removed from exposed deposits with relatively less effect on the floodplain than at a small river. If gravel requirements are very large, the alternatives are to use multiple gravel deposits along the channel, or to expand the areal extent of one site to include adjacent vegetated areas. In a small river system, there are no real options. Gravel has to be removed from adjacent vegetated areas, or from the active channel, or both. This solution was the case for seven of the small rivers studied. The Gold Run Creek site exhibited less change than the other small river systems (except for the site at Phelan Creek where vegetation was not removed). At Gold Run Creek the gravel removal operation was restricted principal ly to gravel bars and an island in the channel. A bank was removed but the degree of floodplain disturbance was less than for the sites on Washington, Oregon, and McManus Creeks, and Penny River. At these latter sites, extensive adjacent floodplain disturbances tended to either greatly expand the channel width or divert the channel. Phelan Creek is a braided system and has a small drainage basin above the material site. Although the site is situated near the headwaters, the channel is of medium width because of flow carried in the summer during glacial melt. In this case the large exposed gravel deposits were scraped and the material site included neither vegetated areas nor channels carrying flow. Even though this is a small river system, the long-term effects are minimal because of other overriding factors. Minimal effects are usually not the case, however, on small rivers. Location of the material site is most critical on small river systems because of the limited availability of exposed gravel deposits and the rela- tively narrow floodplain. Extensive damage can occur to the entire flood- plain reach being mined in these systems, while on large rivers the effects are not as great because the material sites cover a sma I ler proportion of the floodplain. Location of sites and potential effects are discussed in a subsequent section. 349 CHANNEL SLOPE AND STREAM ORIGIN Neither of these Physical Site Characteristics was found to greatly influence the effects of gravel removal in f I oodp I a i n environments. Both channel slope and stream origin are closely related to such factors as drainage basin size and channel configuration, therefore, their influence on the effects of gravel removal are dependent on these factors. The Physical Site Characteristics are discussed separately because of specific impli- cat i ons i nvo I ved. Channel Slope. Removal of gravel from a channel wi I I affect the channel slope within the site and, perhaps, immediately upstream and downstream. Usually this effect entails increasing the slope, which can have localized effects on the floodplain. The main effect is to increase water velocity. Localized changes that can be expected due to the relationship of increased velocity and increased slope are scour and alterations of aquatic communities. Increased scour in a disturbed site can increase downstream deposition of bed load materials where the water slows to the velocity characteristic of the undisturbed channel. The greater scour potential in the disturbed site decreases the stability of bed materials thus affecting habitat for benthic organisms. Increased water velocity can directly affect benthic organisms by displacing those not adapted to higher velocities and favoring those adapted to these conditions. Similarly, fish may become redistributed locally be- cause of water velocity changes. Those fish species or age groups preferring lower velocities may displace to areas upstream or downstream. Altered velocity is not expected to change the terrestrial environment or the scenic quality of an area. Indirectly, an effect might occur to water- associated birds that are dependent on benthic organisms as a food source. Any alterations to benthic communities could alter feeding sites for these b irds. 350 Significant changes in slope most often reflect changes in channel length. If a channel is shortened by mining then the slope is increased; if the channel is lengthened, the slope is decreased. At all study sites the slope was either unchanged or it increased. The likelihood of decreasing channel slope by lengthening the channel is slight because water tends to flow downvalley over the shortest distance. However, if channel lengthening should occur by diversion through a site, then the effects would reflect reduced velocities. Stream Origin. The origin of the stream was found to have little or no relationship to the effects of gravel removal activities. Origin can in- fluence, at least in part, other characteristics of a river system, e.g., channel configuration and shape. Therefore, the preceding discussions are indirectly related to this characteristic. The origin of a stream determines greatly the quality and quantity of gravel materials available in downstream areas. The original purpose for including stream origin in the study was to maximize diversification of the types of sites to be studied. The origins of streams included were mountain, foothill, coastal plain, and glacial. Twelve of the sites studied were of mountain origin, 9 were of foothill origin, and only 4 were of glacial or coastal plain origin. The availability of gravels in streams of coastal plain origin is general ly low and the materials are finer in texture than those found in other systems. Within the geographical limits of our study, only the Seward Peninsula and North Slope have coastal plains. The coastal plain of the Seward Peninsula is so narrow it precludes the existence of such river systems. On the North Slope material sites were located on the Sakonowyak, Putuligayuk, and Ugnuravik Rivers, but only the latter was studied. Gener- al ly, these sites are not favored and are only used if alternative sites are not aval lable. The lack of rock in the headwaters and the low mean annual discharges are the reasons that gravel materials are only minimally avail- able in coastal plain streams. If these sites are utilized, the potential for replacement of gravel sources is very low even over extended time 351 periods. The minimal areal extent- of exposed gravel bars also generally leads to extensive damage to the river system either by use of extended lengths of river channel or by disturbing vegetated floodplains. Glacial origin streams are not common in the area of study; only three sites situated on this type of river were studied. These were on Phelan Creek and the Tanana River. Because these systems are of mountain origin, the availability of weathered parent materials is not limiting and usually large quantities are available. The Phelan Creek site was situated near the glacier and gravel was abundant across a wide area. The proximity of the site to the glacier strongly influenced the seasonal fluctuations in dis- charge. During winter, water flow from the glacier is greatly reduced and is supplemented by that from associated springs. This reduced flow exposes vast expanses of gravel for extraction. The Tanana River sites are well downstream from the river origin, there- fore, water flows throughout the year because of the numerous spring- and groundwater-f ed tributaries entering the river. Affects include those associ- ated with braided channels that flow in winter. In these systems, however, ice cover on channels is more of a factor than on a system like Phelan Creek, near its origin. The availability of gravels in glacial origin rivers makes them a viable source of materials even when needed in large quantities. This is basically true for systems of all sizes although on smaller rivers the localized deposits are more restricted. Most rivers in northern and interior Alaska are of mountain or foothill origin. The weathered parent material in the headwaters provides large quantities of gravels, particularly in the mountain systems. These rivers are fed by springs, melt water, and runoff and, therefore, discharge fluc- tuates seasonally. Spring-fed systems can be expected to have at least Intergravel flow in winter. Moderate to steep channel slopes are normal in the headwaters but these slopes are influenced by the length of the river and the topography through which it flows. Bed load movements are usually 552 higher than in rivers with mild slopes. These rivers generally have large quantities of gravel available even near the mouth. The size of the system and other hydrological and hydraulic factors also influence availability of gravel. The abundance of mountain and foothill origin rivers and the fre- quent availability of suitable gravel materials generally combine to favor the location of material sites in these systems. The geographical location of these rivers, and the topography through which they flow, directly affect the type of channel configuration, a factor discussed in a previous section. 353 GRAVEL REMOVAL AREA CHARACTERISTICS In the preceding section on Physical Site Characteristics it was ap- parent that not all characteristics were important in evaluating the po- tential floodplain change caused by gravel removal activities. In contrast, all of the factors discussed in this section were found to greatly influence the amount of change to a river system. The three main features discussed are type of gravel removal (pit or scrape), location of the material site relative to the active channel (s), and the occurrence of dikes and stock- piles. Singularly and in combination these factors caused varying degrees of change at the 25 study sites, in some cases, irrespective of the specific physical site characteristics. TYPE OF GRAVEL REMOVAL There are two basic types of material sites: pits and scrapes. Pits are dug deeply, usually with draglines or backhoes, and are flooded year- round after site closure. In many cases pits are flooded during gravel extraction unless water is pumped out to keep the site relatively dry. Eight pit sites were studied and they represented two types, those connected to an active channel and those completely separated from an active channel by a buffer zone. Pits usually are situated away from an active channel. In a scraping operation, gravel deposits are removed with bulldozers or scrapers in active and inactive floodplains and terraces. Gravel is extracted by successive removal of thin layers, and scraping depths usually are sufficiently shallow to minimize the occurrence of surface water. At certain study sites, gravel was extracted below the water table, thus water ponded in the site. This situation is not conducive to a scraping operation and, therefore, is usually avoided unless it is required for other reasons. 354 Pits Pits are usually excavated away from an active channel and cause little or no change to the natural hydraulic processes of the channel. Where pits are connected to a channel, either year-round or seasonally, some change to the hydraulics of a river can occur. The most obvious alteration occurs when spring breakup or other high water flows spread throughout the floodplain; much of the water can flow out of the channel because it is often filled with ice. A pit in the floodplain probably would fill during high flows and then, through erosional processes at the upper and lower ends, function as a channel. The inlets or outlets (or both) connecting the pit to the channel could enlarge significantly and reroute flow through the excavated pit. Depending on site conditions this could be only temporary, for example, where a pit is adjacent to a relatively straight reach of channel. In this case, following high breakup flows, the water would again flow down the original channel because the downvalley distance is shorter than if the water flowed through the channel formed by the pit. A permanent alteration to flow is more likely to occur where a pit is located on the inside bend of a meandering stream. Even with undisturbed buffer zones separating the pit from the channel, spring breakup flows can overflow the pit and exit into the downstream reach of the meander surround- ing the pit. If the stability of the buffer zone is low, erosion can breach the buffer zone, thus, connecting the pit to the active channel. The down- val ley distance is shorter through the pit, consequently, there would be a tendency for permanent redirection of flow through the pit and eventual cut off of the meandering channel. Excavation of a pit separate from the channel does not affect the water quality of the active channel. As would be expected, however, the water qual- ity is different in a flooded pit than in the channel. In comparison to channel waters, pit waters typically have higher temperatures during ice free conditions, the dissolved oxygen levels are lower, and sometimes there is stratification of both temperature and dissolved oxygen. Differences in water quality parameters could be less in situations where channel flow is 355 through a pit. This difference depends on the size of the pit and the amount of mixing. A pit could facilitate deposition of suspended and bed load ma- terials if flows are through a pit and velocity is decreased. The aquatic biota of pits differ depending on whether there is an opportunity for exchange between the pit and the active channel. Those pits that are separated (e.g., Tanana River-Downstream) or have little potential for exchange (Dietrich R i ver-Upstream) typically are unproductive. The Tanana River-Downstream pit is situated in the middle of an island and is completely surrounded by a broad undisturbed (except for an access road) timbered buffer zone. The likelihood for injection of nutrients and organ- isms into this pit is remote, except during high flows. The aquatic surveys reflected this. The occurrence of a few fish suggests that overflow may occur at irregular intervals. The Dietrich R i ver-Upstream pit, on the other hand, is connected by its outlet to the channel. A spring, exposed during excavation, floods the pit and exits through a channel. The pit system has been used by overwintering fish but the pit itself is relatively unpro- duct i ve. All other pits studied were highly productive and the diversity of the fish community was usually increased over that in the river channel. All of these pits were connected to the river channel through either inlets or outlets and thus exchange was possible between the two systems. The still waters in the pit, which are warmer than the river water, provided con- ditions more suitable for primary and secondary productivity. Fish such as Arctic grayling entered presumably to utilize the pit as a feeding area. This situation is particularly good for feeding by fish of younger age classes because of the greater supply of food available and the lack of a current. Fish we I I suited to a sti I I water environment, such as northern pike and burbot, also did we I I in some of these pits and, being piscivorous, had an abundance of young age classes of other fish to feed upon as they entered the pits to feed and rear. Northern pike also utilized two of the pits as spawning areas. The potential for the pits to provide a more diversified 556 fish community in the river also exists because of the connection between the two systems. This increased community diversity may be restricted to the area of the channel in the immediate vicinity of the pit. Pit depths are important to fish utilization. Obstructions to movement are not a factor during open water periods if either an inlet or outlet are available for fish movement between the river and the pit. A potential for fish entrapment exists, however, during winter when ice cover is present on the river, the pit, and the interconnecting channel. In the latter situation the pit must be sufficiently deep so it does not freeze to the bottom and decomposition of aquatic vegetation does not decrease the oxygen content of the water below that necessary for fish survival. The creation of a pit in a floodplain constitutes a major change to the local terrestrial environment. Pits are usually situated on vegetated flood- plains, consequently, terrestrial habitat is almost always destroyed. The depth of excavation and the permanent inundation that results also greatly retards or prevents on the long-term, the re-establishment of pred i sturbance conditions. What most frequently occurs, however, is the creation of a more diverse habitat with concomitant changes in faunal communities. The creation of a pit in meandering river floodplains, that contain oxbow lakes, merely adds to the habitat diversity in a localized area. Where pits are located in floodplains lacking natural lakes and ponds, the effect is again principally local, but has implications that affect a much larger system. In these cases, the newly formed body of water can attract migrant waterfowl and shorebirds and perhaps even provide habitat suitable for nesting and rearing that did not previously exist. The higher aquatic produc- tivity of many of these ponds could afford a significant food source for those species adapted to feeding in pond and lake environments. The effect of creation of a pit, on the scenic quality of an area, is totally dependent on the diversity of the floodplain environment. A pit will have less effect where lakes and ponds occur naturally than where these types of aquatic systems do not occur. Where lakes and ponds do not 357 occur location should be selected so that view of the site is blocked from vantage points. For example, the Tanana River-Downstream pit, which is large and contains very clear water, is in a floodplain where the river channels are highly turbid, thus, offering a dramatic visual contrast. However, the site is situated on an island completely surrounded by a heavily wooded buffer zone which blocks view of the site from the Richardson Highway. The pit is visible only from the air. Pits are often excavated with angular perimeters that ignore natural land contour. Since angularity is not characteristic of naturally formed aquatic systems the usual pit site offers some contrast even in areas where lakes and ponds occur natural ly. Excavating these sites with perimeters that blend with natural land contours, such as in abandoned river channels, de- creases the visual diversity that will result from development of pit sites. The West Fork Tolovana River and Tanana R i ver-Upstream sites are excellent examples of this management technique (refer to Figures 63 and 70). Pit sites require considerably less area to obtain a given amount of gravel than do areas that are surface scraped. Because of the depths nor- mal ly required, subsurface waters are exposed, usual ly fill ing the pit during site operation. This water poses problems for the efficient extrac- tion of materials but, since draglines or backhoes are usually used for excavation, the presence of water does not prevent the removal of gravels. Pumping is the only method used to eliminate the water but even this is impossible in some systems because of the volume of subsurface flow through floodplain gravels. During mining, the water in a pit is usually highly turbid and should not be pumped into adjacent channels. In summary, there is little doubt that the excavation of a pit materi- al site creates significant change in a floodplain environment (Table 41). If situated and operated properly, the hydraulics of the river system are little affected whereas significant changes occur to the terrestrial system and the scenic quality of the area. Differences in water quality and aquatic biota can be expected between a pit and the adjacent channel regardless of whether they are connected. The increase in both aquatic and terrestrial 358 O CD it) 0\ l_ ■— 10 in c — in — — m C *- — tn -- — D ._ j3 -D -■N r^ CO p- ►A -^ f^ iTi m CD u^ w "^^ \0 "O fA — "^ Tj- CD lO CO 0) 5 ■*- o O — Q_ «- t. — 10 0) 1^ (?» o o L. \- i- U O O O O i- L. I- L. 0} 3 U 3 S U U 0 O O 0 0 o 0) o o Z IT) i/i (/i Z «1 Ul z z 3 Q. 5 Q. 3 t/) 3 1 to (>^ tO CE o z> D £ cr cc q: u (0 CD (0 'L. c c c (0 (0 10 — .- i_ — 10 c C ^ CD *- — (0 359 habitat diversity is reflected in a more diverse faunal community. Pit sites are a viable alternative for material extraction in areas where changes to the river hydraulics can be avoided or greatly minimized. When major hy- draulic changes occur the effects on the environment can be damaging from many standpoints. Scraped Si tes Scraped sites can occur essentially anywhere in a floodplain from within the active channel to vegetated areas in the inactive floodplain and terrace. Location of the site greatly affects the potential impacts that can be expected from a scraped site. Although scraping implies that material sites are operated by shallow removal of gravel, certain sites studied on this project were excavated below the water table and thus resulted in permanent flooding. These sites, however, were worked with scrapers or bulldozers and not draglines or backhoes as might be implied by depth of excavat i on . Scraped sites have several operational advantages; usually the sites are dry, providing better working conditions and more efficient gravel extraction. Additionally, excavated materials require less handling when using scrapers to remove the gravel because only one machine is normally used to excavate, transport, and deposit at the construction site. This is not feasible using a bulldozer on a scrape or when digging pits with drag- I i nes or backhoes. Given the same gravel requirement, the scraped site will generally disturb a larger area than a pit site because the excavation is more shal- low. In the study sites, the large area affected was often the greatest problem of scrape-mining because there were few restrictions regarding avoidance of channels and areas adjacent to channels. Locations of extrac- tion sites are discussed in the subsequent section. Scrapes are generally situated in active floodplains adjacent to active or high-water channels. Lowering these areas spreads water flow, at least 560 during high flows, and in some cases forms a braided configuration through the disturbed site. When this occurs on unvegetated gravel bars in braided systems, the effect on the f I oodp I a i n is relatively minor because the ef- fects are similar to natural hydraulic processes. After site closure, unless stockpiles or dikes are present, the disturbed site can return to a rather natural configuration within a maximum of a few years. This, however, is not the case where lateral bars are excavated to include removal of adjacent banks. Bank removal is discussed in the subsequent section. The potential for causing braiding from scraping operations within the active floodplain, is usually insignificant in a river system that already has a braided channel configuration. However, in split channel and single channel systems braiding constitutes a significant change to the aquatic environment and alters the aquatic biota; species which benefit are those better adapted to riffle areas, to less stable substrates and, perhaps, to substrates less granular than those found in the natural system. These habitat changes primarily affect the distribution of organisms. This study general ly found a local decreased diversity of the fish community as a result of braiding. There is a potential of blockage to fish passage, at least during low flow conditions, as occurred at the Aufeis Creek site because the water flows over a wider area than in the undisturbed channel. Blockage is most severe if the entire active floodplain is disturbed, not just the lateral bars. Entrapment of fish, in depressions created by scrap- ing, is also possible during periods when water is receding from high flows. Effects on the terrestrial environment depend greatly on the river type involved and on the location of the work area within the floodplain. In braided systems mined in the active floodplain, there essentially is no effect. However, on split and single channel systems, braiding caused by gravel mining can provide feeding habitat for shorebirds that utilize ben- thic organisms. Destruction of banks with associated vegetation removes habitat used by terrestrial fauna; the effects are the same as removal of vegetation for pit sites. 361 The potential for re-establishment of natural configurations and flow patterns after site closure are totally dependent on the degree of change to the hydraulic processes characteristic of the river system. Long-term effects can be expected where major changes to the stability of channels occur. The major terrestrial effect of scraping resulted where deep scrapes occurred in areas immediately adjacent to the channel. Channel flow often diverted through these depressions and caused year-round ponding which retarded the re-establishment of vegetation. These deep scrapes usually were inadequate as quality habitat for waterfowl and shorebirds and unsuitable for fish. To minimize short- and long-term effects, scraped sites should not be excavated beyond certain depth limits. These restrictions are discussed in the Guidelines Manual. The effects of scraping operations on the scenic quality of a braided floodplain can be minimal if the material sites are restricted to the active floodplain. Where banks and vegetated areas are altered, significant effects can be anticipated. In split and single channel systems the establishment of a braided configuration in the disturbed area produces an unnatural condi- tion in the floodplain, thus affording a visual contrast. Properly located scraping operations that avoided or minimized disturbances to the hydraulic characteristics of a river, minimized long-term environmental change. How- ever, where sites were poorly located and caused significant changes to the channel hydraulics, major long-term effects were evident on the scenic qua I i ty of the area. In summary, scraping operations typically occurred in both active and inactive floodplains. Both vegetated and unvegetated areas were used but the fewest long-term disturbances occurred where only exposed gravel de- posits were scraped. The potential for broadening or diverting channel flow in split and single channel systems is great if depths of excavation are excessive and locations of sites are poor. The potential for braiding in these situations was increased with concomitant changes in aquatic biota. Terrestrial effects were greatest when the depth of excavation was excessive and led to permanent ponding which retarded recovery to predi sturbance 562 conditions. Visual effects of scraping operations depend greatly on the type of river system, the location of the site, and the areal extent of the site within the f I oodp lain. LOCATION OF GRAVEL REMOVAL Location of a gravel removal operation in relation to the channel of a river was found to be the most important aspect influencing long-term change to a floodplain environment. Whether a pit or scrape, in general, the loca- tion of the site was a more important consideration than the type of site. Site location in this section is discussed with minimal reference to the type of site although the latter is a factor influencing the extent of change. In-Channel Locations As used in this project, in-channel gravel removal includes areas in the active channel, high-water channels, and abandoned channels. Fourteen of the sites studied on this project were situated in high-water channels and 7 of the 8 sites located in the active channel also included areas in high- water channels. From hydraulic and hydrological standpoints, material sites in active and high-water channels caused the greatest long-term change to the floodplain environment. Active Channel. Gravel removal operations in the bed of an active channel cause a series of changes all basically related to changes in the depth and location of the thalweg. The degree of change depends on the type of channel configuration, principally whether it is a braided or a single channel. In a braided system the channels generally shift throughout the active floodplain on an annual basis. This is due to the lateral instability of the individual channels. In these systems removal of gravel has the effect of perhaps causing greater instability in the area of the distur- bance. Changes occurring in a single channel river caused by removing bed material are unknown because all seven sites with this mining location had substantial alteration to adjacent deposits or banks. 363 Removing gravel from within the channel is accomplished either by dredging or by scraping the bed after flow has been diverted. Either method can result in a deepening of the thalweg and, if the edges of adjacent gravel bars or banks are removed, a widening of the channel. Depending on the location of the material site, this operation could alter the pool: riffle ratio in the river. Where the channel is dredged, turbidity in and downstream of the site will increase greatly during mining. Turbidity should reduce quickly after the operation has ceased. If the channel is diverted during mining, the effects on water quality entail suspension of the fines exposed during mining when water is diverted back through the site. This suspension will result in a temporary increase in turbidity. Reduction in the velocity of water entering the excavated hole will cause sedimentation of both bed load and suspended materials. This will aid in rapid replenishment of the gravel materials removed from the site. Being in the active channel, the replenishment rate is considered high compared to other areas in the f I oodp I a i n . Excavation of the channel bed can remove spawning areas. During a dredging operation fish probably will redistribute to less turbid waters. Benthic organisms adapted to silt-laden areas will establish following excavation and remain until the natural gravel bed becomes established. Assuming that the disturbances resulting from gravel removal are re- stricted to the channel, and do not include the banks or edges of gravel bars, little long-term effect on the terrestrial environment is expected. Changes could occur if hydraulic changes in the channel affect adjacent banks. Aesthetically, the in-channel material site has little or no effect. Hydraulic changes resulting from in-channel disturbance that affects banks can cause some effect. 364 High-Water Channel. High-water channels flow only during high-water periods. The hydraulic effects of removing gravel from high-water channels are not as great as they are in the active channel where the disturbed area is subjected to flow throughout the year. The changes that can be expected are similar to those described for the active channel although they occur only during the period when the site is subjected to flow. Effects on water quality are only evident during the high-flow period. Localized widening or deepening of the high-water channel would slow the water velocity and thus facilitate deposition of both bed load and suspended materials. Depending on the degree of change to the channel this deposition would reduce the time required to re-establish near-natural conditions in the area. Also, any fines exposed during mining would be available for suspension during high flows. Removing gravel from a high-water channel could trap fish and benthic organisms in the depressions of the disturbed areas as flow recedes. Many benthic organisms that are adapted to a riffle community and most fish species would not be able to survive in such a habitat. Since high-water channels are subjected to less flow than active chan- nels, they tend to be more stable and are usually bordered by established terrestrial vegetation. Any disturbance to these channels causing lateral instability during high flows could facilitate erosion of adjacent banks and thus serve to reduce the areal extent of vegetated areas. Loss of habitat would cause localized elimination of small mammals and displacement of birds and larger mammals. Having water pooled in the high-water channel during low-flow periods could attract shorebirds, particularly where a benthic fauna has become established to serve as a potential food source. The most serious effect from a gravel removal operation in a high-water channel is bank destruction which often occurs with this type of operation. This aspect is discussed in a subsequent section on removing gravel from banks . 365 The effect of mining gravel from a high-water channel on the scenic quality of an area is minimal if the disturbance is restricted to the chan- nel. If banks are destroyed the effect would be more significant. Since the high-water channel is active only part of the year re-establishment of pre-existing conditions will require a longer time. Formation of pits in high-water channels would have effects similar to those described in the section on Type of Gravel Removal. Abandoned Channel. Abandoned channels carry water only during major flood events. Normally, these channels are considered to be dry during most years. Since they represent old river channels they usually contain reason- ably large quantities of gravel, depending on the type of river with which they are associated. Only two of the sites studied were located on an aban- doned channel. Prospect Creek and West Fork Tolovana River, both in meander- ing systems. Abandoned channels are common in this type of floodplain be- cause of the formation of cutoffs that result from the fluvial processes of meandering channels. Location of material sites in abandoned channels causes little problem with regard to changes in river hydrology and hydraulics because the sites are separated from active flow. Where pits are dug in abandoned channels and are connected to the active channel, flow can be diverted through the site during high flows. The magnitude and duration of this change is dependent on the nature of the connection between the material site and the channel and the integrity of the undisturbed buffer zone separating the site from the active channel. Where the once-abandoned channel carries water annually during high-flow stages, the effects to the floodplain would be similar to those described for sites in high-water channels. Where an abandoned channel is scraped and the water table is not reached, water quality does not become a problem. Where pits become flooded, the water quality would be different than that occurring in the active channel, as is discussed in the section on pits. 366 Aquatic biota will not be affected in a scrape operation located in an abandoned channel, however, if a pit is dug, aquatic biota could become established. In these cases the effect depends on whether the gravel removal operation alters the site sufficiently to cause it to be subjected to annual high flow or whether it is connected to the active channel. In the former case, there is potential for entrapment of fish during high flow as was discussed for high-water channels. In the case of a site connected to a channel, the effects are those discussed in the section on pits. The effects of removing gravel on the terrestrial environment can be greater in an abandoned channel than in other in-channel locations. Aban- doned channels are rarely subjected to hydraulic forces, consequently, vegetation usually is established, and the stage of succession is dependent on the time since the channel ceased to carry flow. Thus, vegetation must be removed from these sites to expose gravel deposits. Removal of this habitat results in a loss of feeding, nesting, and cover habitat for those sma I I mammals and passerines that utilize riparian shrub thickets. Larger mammals, being more mobile, are displaced to adjoining areas. If the abandoned channel is scraped above the water table, the dis- turbed site will initiate primary plant succession following site closure. The time required to reach the pr ed i stur bance stage of vegetational succes- sion is dependent on the geographical region and the vegetative charac- teristics of the area. This process is the same as occurs in other recently abandoned high-water channels and entails the same vegetational and faunal communities. If the site is a pit that is permanently flooded, the site would not return to a terrestrial environment in a relatively short time. However, overall habitat diversity is increased. Further discussion of these aspects is included in the section on pits. The effects of siting a gravel removal operation in an abandoned chan- nel, on the scenic quality of an area, reflect the changes occurring to the terrestrial vegetation. The short-term effect is to expose an area that was previously vegetated. The long-term effect in a scraped site depends on the rate of revegetation of the disturbed area. Where a pit is dug the altera- 367 tion is long-term but, in fact, could blend more with the i nterspers i on of cutoffs and lakes occurring naturally in the floodplain. Adjoining Channel Locations The Major Variable Matrix (Table I) includes four subdivisions under adjoining channel locations. These are: point bar, lateral bar, mid-channel bar, and bank. To thoroughly characterize the 25 study sites it was neces- sary to utilize all of these subdivisions but the gravel removal effects are similar for some. Therefore, the following discussion combines the three bar locations and discusses banks separately. Remember, at a given material site these bars and banks are associated with one of the three channel types discussed in the previous section. Point, Lateral, and Mid-Channel Bars. This discussion only considers removing gravel from unvegetated bars with exposed gravel deposits. All three gravel bars are usually numerous in braided systems but, in single channel systems, usually only point and lateral bars are found. The effect of removing gravel from a bar is to lower the elevation of the bar thus allowing flow to inundate an area that was previously above the low-flow water line. These sites are usually scraped. Maintenance of the integrity and conformation of the bar will cause little permanent change to channel hydraulics and will facilitate replenishment of the gravel during subsequent high flows. Changes in the active channels can and probably will occur where bar integrity is not maintained. In a braided river system this change will be similar to the natural processes and the long-term effects will be minimal. In a single-channel system redistributing flow by removing bars can have long-term effects by changing the local hydraulics of the channel. This hydraulic change could either decrease the lateral stability of the channel or widen or deepen the flow because the cross-sectional area is larger. Where the banks are stable, the river eventually will equilibrate itself by reforming gravel bars as upstream bed load materials become avail- able during subsequent high flows. Where banks are less stable it is pos- 368 sible that subsequent high flows will cause erosion due to the hydraulic forces acting on the once protected banks. This could significantly alter the local reach of a river. This effect is less likely to occur in straight and perhaps sinuous river systems because the flow is relatively unidirectional down the flood- plain and direct hydraulic forces on the banks would be less than in a meandering system. The effect on a meandering river could be to facilitate the formation of cutoffs by increasing the hydraulic force on the inside bank at the upstream end of a meander. Removal or lowering of gravel bars will facilitate the spreading of river flow when water levels are higher than during the gravel removal opera- tion. This flow spread has the effect of reducing the depth and velocity of the water and will increase sedimentation rates of both bed load and suspen- ded materials. Additional ly, water temperature and dissolved oxygen contents could change. Benthic communities would develop that are adapted to riffles and less stable substrate. Fish would become redistributed with younger age classes perhaps being attracted to the disturbed site where currents would be less. The effects to the terrestrial environment, of removing gravel from a bar, are minimal if the integrity of the bar is basically maintained. The only changes that could be expected are if the hydraulic regime of the river channel is altered, thus, causing changes in adjacent vegetated areas. The spreading of flow between the banks when bars are removed might attract shorebirds for purposes of feeding. These effects would only be expected in single-channel systems. Removing gravel from isolated material sites using accepted mining techniques from bars in braided river systems would have little or no effect on the scenic quality of a floodplain. The lateral instability of the chan- nels that characterize these systems would cause any changes resulting from gravel removal to blend in with natural processes. Removal of bars in a single channel system wi I I local ly affect the appearance of the river sys- 369 tern, the magnitude of effect depending completely on the degree to which the bar was disturbed. Any significant changes to the hydraulic geometry of the reach causing subsequent disturbance to adjacent vegetated areas will lo- cally alter the appearance of the floodplain. Banks. Probably the most consistent long-term changes to a floodplain occurred when banks were destroyed or greatly modified during a gravel removal operation. In these cases significant changes to the hydraulic geometry of the river occurred. Banks typical ly are stable and function to restrict the flow of the river to the channel except during high flows. When these are removed or disturbed the river is no longer contained and it begins to wander and erode the adjacent floodplain. This wandering results from the hydraulic forces of the river impinging on newly exposed bank ma- terial. Where banks are made of stable materials the degree of erosion should not be greater on the newly exposed bank than what occurred natural ly before the site disturbance. Where the newly exposed bank materials are not stable erosion will occur at a rate faster than occurred previously. Also, if the newly exposed bank is situated at an angle to the flow different than what occurred natural ly in that reach of the river, erosion could be aggra- vated because of the increased hydraulic force on the bank. Generally, channel width increases with bank destruction. Previous discussion identified that increased channel width can result in reduced water velocity, reduced water depth, changes in water temperature, and dissolved oxygen, and increased sedimentation. Aquatic biota would reflect these altered habitat conditions by changes in benthic communities to those that are adapted to riffle areas with unstable substrate and changes in distribution of fish in the reach affected by the disturbance. Undercut and vegetated banks are heavily utilized by fish as cover and removal of this habitat can greatly reduce the local abundance of certain species. The effects on the terrestrial environment include destruction of riparian habitat during site clearing with resultant effects on faunal distribution. The decreased lateral stability of the channel can cause more destruction after site closure if hydraulic forces erode newly exposed 370 areas. In addition, even if the newly exposed banks are stable the hydraulic forces occurring over the disturbed site would retard the re-establishment of terrestrial floodplain habitat. The effect on the scenic quality of the area will reflect the changes occurring to the terrestrial environment and to the hydraulic geometry of the river channel. Major changes to these aspects will greatly alter the appearance of the floodplain in the affected reach. Locations Separated From Active Channel The five specific site locations identified in the Major Variables Matrix (Table I) that are separated from the active channel are not mutually discrete locations. That is, a site can exhibit a combination of these locations by for example, being located near the channel on the outside of a meander. Hence it is more difficult to assess the potential impact for these locations than for those previously discussed. The following discussion has been separated into two sections: inside and outside of meanders, and islands. These then are discussed from the standpoint of whether a material site is near or distant from the active channel. The essential factor with sites in all of these locations is whether diversion of the water out of the active channel and through the site is possible. The distance between the material site and the active channel is of major concern, but the height of the intervening bank certainly would be a necessary consideration in this evaluation. Inside and Outside of Meanders. The location of a site on the outside of a meander is possible on any sized river system regardless of the areal extent of the material site. This, however, is not the case on the inside of meanders. In small river systems the areal extent of the floodplain or terrace circumscribed by the meander can be quite small. In cases where these were used for material sites, the surrounding areas, including the channel, were often disturbed by the gravel removal operation. Therefore, to 371 limit activities to the inside of a meander and maintain undisturbed buffers the site must be located on at least a medium sized river. Any activity inside a meander, that would reduce the integrity of the banks or weaken the cross-sectional area, could lead to premature cut off of the meander. In many Alaskan rivers during breakup, water often flows over the ice in the channel and, if sufficiently high, over the banks and down the floodplain. A depression resulting from a material site located near the channel on the inside of the meander would aid in channeling the water through the site. Depending on the erodability of the soil separating the material site from the channel, a channel could erode at both the upstream and downstream portion of the meander and thus eventually establish a cut- off. The erodability of the soil would govern the length of time required for this natural event to occur. When a pit material site is connected to the active channel, the probability of a cutoff occurring could be enhanced greatly, even in a very short time. Such an event occurred at Skeetercake Creek on the North Slope. The inside of a meander of this sma I I river was mined for gravel and when the site was studied II years after site closure, a cutoff had occurred. The time required for this event to occur is unknown. A pit visited during site selection, but not studied in this project, that showed a potential for channel diversion, was located at Hess Creek in the Southern Interior region. The buffer strip was breached during the first spring breakup following site opening while the site was being op- erated. The initial breach was temporary and the water remained in the active channel when the flow receded. The key point of concern when mining in the inside of a meander is maintenance of a sufficiently wide undisturbed buffer zone between the active channel and the perimeter of the material site. The size will depend greatly on factors such as the discharge of the river, flood frequency, and soil erodability and must, therefore, be determined on a site-specific basis. In order to maintain the integrity of the channel over the long-term it may be necessary to dig deeper to obtain needed gravel volumes, rather than decrease the buffer width. 372 Buffer zones are similarly important to separate the active channel from material sites located on the outside of meanders. A breach occurring in this situation would lengthen the meander. This breach probably would be a temporary event during high flow periods and the river would maintain its main flow through the active channel during lower water levels because of the shorter downval ley distance. Periodic and aggravated damage to the area between the material site and the active channel and perhaps the creation of a backwater area in the material site, would occur from an outside meander breach. It is obvious that the closer a material site is to the active channel the greater the probability of a permanent breach occurring in a short time. Placement of a material site either on the inside or outside of a meander has no effect on water quality, regardless of the distance sep- arating the site from the channel. However, if water is ponded the water in the pit would differ from that in the channel, as described in the section on pits. Changes in water quality could result if a breach occurs. These also are discussed in the section on pits. Change will not occur to aquatic biota when material sites are located away from the active channel. However, if high flow conditions reach a material site, and cause either temporary or permanent ponding, fish could become trapped in the site when the water recedes. Effects similar to those described for connected pits could occur where the buffer is breached and a pit site becomes connected to the active channel. In general, locating material sites back from the active channel will necessarily entail destruction of vegetative habitat. This will result in local ized loss of sma I I mammals and displacement of birds and larger mam- mals. If the area is scraped and does not become flooded during high water the site eventually will return to the pr ed i sturbance condition through processes of primary and secondary plant succession. The length of time required will depend on the regional characteristics. If the site is flooded 373 because it was dug as a pit, or because depressions are at least temporarily flooded, vegetative re-establishment will be retarded. Because of the soil binding characteristics of vegetation, maintenance of the vegetation on the buffer zone between the material site and the active channel is important. The wider this zone the less the likelihood of a breach. If a buffer breaches, the progressive erosion of soils and loss of overlying vegetation will result in prolonged impact to the terrestrial environment. Concern for maintenance of the natural hydraulic geometry in the floodplain while selecting a material site location, and while operating the site, will limit terrestrial change to the area of the disturbance. The usual need to remove vegetation to operate a site away from the active channel will affect the scenic quality of the floodplain environment. The magnitude of effect will depend much on the shape of the site, whether it conforms to natural land forms, and what the vegetative structure is in the area. If the site is not visible from a road or other accessible vantage point, the overal I impact wi I I occur only from the air. The distance of the site from the active channel would not necessarily be related to the mag- nitude of impact on the scenic qual ity but this would be determined on a site-specific basis. Islands. Material sites located on islands require the removal of vegetation. The distance between the perimeter of the material site and the active channel is the major consideration in the development of these sites. Islands are situated in the active channel most of the time, thus, the maintenance of buffer zone intregity is of greatest concern. If buffer zones are removed or greatly disturbed the net long-term effect could be the loss of the island, perhaps changing the hydraulic geometry significantly enough to cause other changes within the floodplain. Sites that have been located on islands where the banks were disturbed or eliminated have had greater effect on the floodplain t-han those where the site was developed total ly separate from the channel (e.g., Tanana River-Downstream). In the latter case there was no change detectable to the 374 hydraulic regime of the channel. In the other cases, induced erosion of the disturbed banks has had more prolonged effects than where this erosion has not occurred. Again, of prime concern with material sites on islands, as with other sites separated from the channel, is maintenance of the natural hydraulic geometry of the river channel. If natural hydraulic forces erode islands in a given reach of a river, the presence of a material site, whether a pit or scrape, wi I I weaken the integrity of the island after natural bank erosion reaches the perimeter of the site. Development of material sites on islands where the perimeters of the sites are separated from the channel, will have little effect on water quality and aquatic biota. If the material site is flooded because it was deeply dug, the contained water will be different than the water in the active channel, as discussed under pit sites. If the site is flooded regu- larly during high-flow conditions there is a potential for fish entrapment as the water recedes. The long-term effect on aquatic biota depends on whether the site is permanently flooded and the depth of the water. If the site becomes connected to the active channel by breaching of the buffer zone, the effect may be development of a braided section with the accom- panying changes. Flooding of depressions in the disturbed area could cause fish entrapment before the establishment of a braided pattern. Terrestrially, the loss of vegetated habitat would result in loss of both small mammals and perhaps some larger ones. Loss would depend on the size relationship of the material site to the island, but would occur regu- larly where a large proportion of the island is disturbed for the material site. The mortality would occur as a result of animals not being able to cross the river channel (s) to adjacent floodplain habitat. The loss of vegetation on an island reduces the amount of bird nesting habitat. This could affect the total productivity of an area more than if an equivalent amount of vegetation were removed along the edges of the flood- plain. This assumes that the island provides some protection from mammalian predators unable to cross the intervening channels. Otherwise, the mobility 375 of birds allows them to redistribute in the floodplain just as large mammals do that are dependent on floodplain habitat. Material sites on islands will affect the scenic quality of the flood- plain, but the type of vegetation characteristic of the area would determine the long-term visibility of the site. Where stands of timber block view of the site except from the air, as with the Tanana River-Downstream site, little change would occur. Where such timber is not present the material site could be quite conspicuous and affect the appearance of the floodplain environment more than if the site was located along the edge of the flood- plain. In either case, maintenance of an undisturbed buffer zone between the material site and the active channel reduces the induced disturbances that could further detract from the natural appearance of the floodplain. Summary. The problems associated with material sites located separate from the active channel are essential ly dependent upon maintenance of the integrity of intervening buffer zones. Where this is maintained, and the hydraulic geometry of the river is not affected, very little or no change would be expected relative to hydrology-hydraulics, water quality, and aquatic biota. The terrestrial system and scenic quality of the floodplain will be affected because usual ly vegetation must be removed to expose under- lying gravel deposits. Generally, sites located back from the channel are favored from a practical standpoint because they can be operated in a dry condition making for a more efficient and easier operation. Excavating a pit would be an exception because the depths of excavation would normally be below the water table. DIKES AND STOCKPILES The location of certain material sites and the gravel removal opera- tions require the construction of a protective structure and/or the stock- piling of overburden and gravel in or near the material site. Protective structures prevent water from entering the material site and include channel plugs and diversion dikes. Overburden piles consisting of brush, slash, groundcover, and organic soil are located either permanently or temporarily. 376 usual ly at the edges of sites. Gravel stockpi les are considered to be tem- porary and are located within the material site. Dikes and stockpiles of unused gravel were sometimes left intact when the site was abandoned, thus, contributing to the long-term effect of the gravel removal operation. Any dikes or stockpiles deflecting or otherwise modifying flow patterns could aggravate the long-term hydraulic effects of the material site. Flow alterations could significantly modify the hydraulic forces in the local reach of the affected floodplain and cause other damage. Alterations to natural flow patterns in the winter could induce or aggravate auf e i s forma- tion. The water quality of an area could be affected by the location of these structures in the floodplain. Any erosion of overburden piles by active flow could introduce large quantities of organic materials for suspension and eventual downstream deposition. Also, any structures that would impound waters, after high flows have receded, would result in differences in the water quality between the active channel and impounded waters. Aquatic biota could be affected by the presence of obstructions. Fish could become entrapped behind any structures that impound water. The suspen- sion of fines in the water column as a result of erosion could cause redis- tribution of fish and reduction of riffle invertebrates. Overburden piles provided a nucleus for revegetation of abandoned material sites. The organics, and particularly the root stocks and slash, facilitated re-establishment of vegetation in localized areas of the site. Overburden piles were used for denning by ground squirrels and, because they were vegetated, provided habitat for small mammals and nesting passerine birds. Abandoned stockpiles of gravel were less prone to provide these cond i t i ons. In the long-term, any alterations of flow patterns that resulted from abandoned structures probably would be detrimental to vegetative recovery 377 on the site. Revegetation in these cases would only occur on the area above the h i gh f I ow I eve I s. Abandoned structures in most cases further detract from the already af- fected scenic quality of a floodplain. Where the site is hidden from view except from the air abandoned structures would not alter the overall impact. However, in places characterized by tundra and low riparian vegetation, these abandoned structures can attract attention to the floodplain site. 378 SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS SUMMARY Not al I of the major variables used to characterize the 25 material sites were significant determinants of gravel removal effects. Among the Physical Site Characteristics, channel configuration was the most important. Potential floodplain change is least for a braided river and greatest for a straight river. Size of channel is a significant factor, with the least change to be expected in a large system and the greatest in a small system. This assumes equally sized material sites. Combining these two variables, (channel configuration and size) gravel removal operations can be expected to have the least effect on large braided rivers and the greatest effect on small straight rivers. Influencing Physical Site Characteristics related to configuration and size are the availability and size of unvegetated gravel bars, floodplain width, and the distance that can be maintained between the mining site and active channel. For example, in a small straight river system the floodplain is narrow and gravel bars are neither plentiful nor large. Thus, to extract gravel, either a significant length of active floodplain or the adjacent inactive floodplain and terrace must be disturbed. In the latter case the narrowness of the floodplain forces the operation to closely encroach upon the active channel. In large river systems these problems can be less signif- icant because gravel bars are larger and, if the inactive floodplain or terrace are used, the wider floodplain al lows maintenance of a broader undisturbed buffer zone between the material site and active floodplain. 379 In the present study, channel slope and stream origin did not correlate with changes resulting from gravel mining. However, channel slope influences the bed load carrying capacity of a stream — steeper slopes indicate greater carrying capacity. This relationship is useful in evaluating po- tential replenishment rates in a disturbed site after mining. Also, stream origin has an influence because rivers of mountain and glacial origin charac- teristically have larger quantities of gravel available than do rivers of coastal plain or i g i n. All of the Gravel Removal Area Characteristics were found to signifi- cantly influence the effects of gravel mining. The location of the material site relative to the active channel is considered to be the most important factor. Whether a material site is scraped or pit-excavated is important, but often pits are located away from an active channel, avoiding the types of changes that can be associated with scraping in active floodplains. The major effects of pit sites located in inactive floodplains and terraces are the loss of vegetated habitat, the possibility for fish entrap- ment, a change in the appearance of the floodplain, and long-term delay in the re-establishment of pr ed i sturbance conditions. Where pit sites are situated well away from active channels they have little effect on the active channel and, there is little chance of contributing to channel diver- sion. When situated close to active channels, particularly on the inside bends in meandering systems, the possibility exists for diversion of the channel through the pit, eventual ly forming a channel cutoff in the meander. This problem highlights the importance of providing a buffer between the material site and the active channel. Where pit sites are of suitable size, of sufficient depth, and have contoured perimeters, they can increase local habitat diversity and provide conditions suitable for fish and various species of terrestrial fauna. Scraped material sites in active floodplains have minimal effects on the floodplain environment when exposed gravel bars are only excavated above the water level and slope and contours are maintained resembling those of natural bars. Removal of vegetated areas or banks, which results in 380 decreased lateral stability of active channels, or allows water to spread over a large area, is not desirable. Decreased water depth and velocity increases sedimentation rates, alters water temperature, and alters dis- solved oxygen levels. These chan9,es in aquatic habitat usually affect the local distribution and community structure of benthos and fish. The effects of scraping in vegetated areas of inactive floodplains and terraces can be similar to those described for pits. However, long-term changes typically are minimal because the lack of standing water in the closed site will facilitate re-establi shment of pre-mi ning vegetation con- d i t i ons. In-channel locations that are dredged have the potential for causing the least change to channel hydraulics, terrestrial biota, and aesthetics; however, they can have the greatest effect on water quality and aquatic biota. Gravel replenishment rates are highest in this location. Mining exposed gravel bars in active floodplains potentially has the least effect on terrestrial systems. Sites in inactive floodplains and terraces affect the terrestrial biota and scenic quality most, but potentially have no affect on the aquatic system. In general, the farther a material site is located from a channel the greater the potential effect on the terrestrial biota and scenic quality and the smaller the effect on the channel hydrology-hydraulics, aquatic biota, and water quality. This relationship constitutes the major tradeoff consideration in locating material sites in f I oodp I a i ns. If material sites are located and operated to prevent or greatly mini- mize effects on channel hydraulics, and to utilize only exposed gravel bars, the probability of major localized changes to a floodplain is generally greatly reduced. Where exposed gravel bars are not available or are inade- quate, a tradeoff decision between sites must be made that weighs the poten- tial effects of aquatic disturbances against terrestrial disturbances. In these cases, minimization of hydraulic change to active channels should be important in the decision — major hydraulic changes can have a greater long-term effect on terrestrial systems than the controlled disturbances 38 1 associated with a site located in a vegetated inactive floodplain or ter- race. Dikes and stockpiles of gravel and/or overburden left in a material site after closure, have potential effects on the floodplain. These struc- tures can alter channel hydraulics locally if they are subject to high flows. During high water the fines and organic debris may be introduced into the water and result in downstream sedimentation. Depending on their position and orientation relative to flow, dikes and stockpiles can also cause fish entrapment. Where overburden piles are above high-water levels, they can facilitate the establishment of vegetation after site closure. This vegetation provides habitat for small mammals and passerine birds. In some cases, revegetation at a site was found only on such overburden piles. This observation suggests that, as long as the piles are situated where they are not subject to inundation or hydraulic erosion, they can provide a source for revegetation of the site. Overburden piles may detract from the scenic quality of a floodplain. RECOMMENDATIONS The recommendations developed for each of the disciplines are generally in agreement, with several exceptions. All recommendations are generally designed to minimize change to the floodplain and to enhance re- establishment of pred i sturbance conditions. I. River types that should be used in order of decreasing preference are: braided, split, meandering, sinuous, and straight. The major consideration in this preference is the availability of gravel from exposed bars. The largest volumes are available from braided systems and the least from straight systems. An additional factor is the decreasing floodplain width of the configuration series identified above. If areas adjacent to the channel must be used for gravel mining, greater overall change will result in straight systems. 582 2. River sizes that should be used in order of decreasing preference are: large, medium, and small. The rationale is the availability of gravels and width of floodplain. Larger systems have more gravel. The proportionally smaller disturbance in large systems will reduce the overall effect of grave I r emova I . 3. Mining gravel from active channels should be avoided to reduce detrimen- tal effects on water quality, aquatic habitat, and biota. However, if hy- draulic changes can be minimized, in-channel sites will replenish more rapidly than other areas and effects on the terrestrial biota and scenic quality of the floodplain will be avoided or greatly minimized. 4. Changes to channel hydraulics should be avoided in all cases, es- pecially the establishment of a braided configuration in the disturbed site. 5. When possible, exposed gravel bars in large active floodplains should be considered for mining. A properly operated material site in these areas can minimize changes to channel hydraulics during low-flow periods, minimize changes to water quality and aquatic biota, minimize or eliminate affects on terrestrial biota, and maintain the scenic quality of the floodplain. In addition, the probability of gravel replenishment is increased. 6. Although pits reflect a major change from predi sturbance conditions, they can increase local habitat diversity if suitably located and developed. They should be located to minimize the probability of channel diversion through the site. Adequate undisturbed buffers should be maintained between the material site and the active channel. 7. Organic debris and overburden should be spread over or piled in the abandoned site to promote revegetation and establishment of predi sturbance conditions. This procedure must be conducted only in situations where there is a low likelihood of this material being eroded into active channels. 383 RECOMMENDED FUTURE STUDIES During the present study a number of subject areas were identified that should be investigated. 1. Evaluation of gravel mining from coastal and upland sources; and, preparation of guidelines for users of these sources. These alternatives to floodplain sources have not been studied. 2. Evaluation of the effects of multiple sites on one river system. Such an investigation should be aimed at determining the critical, spatial, and temporal relationships of multiple sites. Gravel replenishment rate pre- dictions should be an integral part of this investigation. 3. Several floodplain gravel removal sites should be investigated before, during, and after mining to assess the adequacy of the Guidelines Manua I . 4. Several specific topics of the Guidelines Manual should be studied in detail to assess their adequacy, i.e., buffers, pit design, and active channel dredging. 384 APPENDIX A Scientific names of flora and fauna identified in the text are presented in Tables A- I through A-5. References are: Herbaceous Vegetation - Hulten, E. 1968. Flora of Alaska and Neighboring Terr i tor ies. Stanford Univ. Press. 1,008 pp. Woody Vegetat ion - Viereck, L. A., and E. L. Little, Jr. 1972. Alaska Trees and Shrubs. U.S. Dept. Agric. Handbook 410. 265 pp. Mammals - Hall, R. H., and K. R. Kelson. 1959. The Mammals of North America, Ronald Press Co., New York. 2 vols. Birds - American Ornithologists' Union. 1957. Check-list of North American Birds. Port City Press, Inc., Baltimore. 69 1 pp. American Ornithologists' Union. 1976. Thirty-third supplement to the AOU check-list of North American Birds. Auk 93(41:875-879. Fish - Alaska Department of Fish and Game. 1978. Alaska's Fisheries Atlas. Vol. I and II. Alaska Dept. Fish and Game, Juneau, Alaska. 83 pp. + maps . Bailey, R. M. , J. E. Fitch, E. S. Herald, E. A. Lachner, C. C. Lindsey, C. R. Robins, and W. B. Scott. 1970. List of Common and Scientific Names of Fishes from the United States and Canada. Third edition. American Fisheries Soc. Spec. Publ. No. 6. 150 pp. McPhail, J. D., and C. C. Lindsey. 1970. Freshwater Fishes of North- western Canada and Alaska. Fish. Res. Bd. Canada. Bull. No. 173. 381 pp. Morrow, J. E. 1974. Freshwater Fishes of Alaska. Alaska Northwest Publishing Co., Anchorage, Alaska. 78 pp. 385 Table A- I . Vegetation Identified in the Text Common Name Her set a i I Reed Bent Grass Poa Cotton Grass Sedge Rush Ba I sam Pop I ar Pel t leaf Wi I low Littletree Willow Paper Birch American Green Alder Th i n I eaf Al der Ch i ckweed Mi Ik Vetch Oxy trope Sweet Pea Dwarf Fireweed S iber i an Aster F I eabane Wormwood Hawk's Beard Sc i ent i f i c Name Equisetum variegatum Ca I amagrost i s spp. Poa spp. Er iophor urn spp . Carex spp. Juncus spp. Populus balsamifera Sa 1 ix a 1 axens i s Sa 1 ix tule nus arbuscu 1 oi des Be a papyr i \ Fera Al cr i spa Al nus tenu i f 0 1 ia Ste I I ar i a spp. Astragalus spp. Oxytropis spp. Hedysarum Mackenzii Epilobium latifolium Aster s ib ir icus Erigeron spp. Artemisia spp. Crepis nana 386 Table A-2. Mammals IdenHfied in the Text Common Name Sc ient i f i c Name Arctic Ground Squirrel Red Squirrel Beaver Tundra Vole S ing i ng Vo I e Muskrat Gray Wo I f Black Bear Grizzly Bear Moose Car i bou Spermoph i I us undulatus Tamiasciurus hudsonicus Castor canadensis M icrotus icrotus ndatra . oeconomus M Ol mi urus z ibeth icus Can i s I upus Ursus amer icanus Ursus horr ib i 1 is A 1 ces Alces Rang i f er tarandus 387 jable A-3. Birds Identified in the Text Common Name Wh i st I i ng Swan Trumpeter Swan Canada Goose Black Brant Mai lard P intai I Green-winged Teal Common Goldeneye Barrow's Goldeneye Buff lehead Red-breasted Merganser Semipalmated Plover Ruddy Turnstone Semipalmated Sandpiper Western Sandpiper Spotted Sandpiper Northern Phalarope Gl aucous Gu I I Herr i ng Gu II Mew Gull Arctic Tern Alder Flycatcher Tree Swal low Violet-green Swal low Bank Swa I I ow Gray Jay Black-capped Chickadee American Robin Sc i en t i f ic Name 01 or col umb i anus 01 or bucci nator Brant a a canadensi s Brant Anas n i gr i cans 1 atyrhynchos Anas acuta Anas caro I i nens i s Bucephala c I angu I a Bucephala i s I and i ca Bucephala a I beol a Mergus serrator Charadrius semi pha Imatus Arenaria i nterpres Caldr Caldr Act it i s i s is = s h' pusi 1 1 us maur i macu 1 ar i a Lob ipi lobatus Larus yperboreus Larus argent at us Larus canus Sterna par adi saea Empidonax ainorum Ir i doprocne b i co I or Tachyci neta tha I assi na Ripar i a r ipar i a Perisoreus canadens i s Parus atr icap i I I us Turdus mi grator i us cont i nued 388 Table A-3. [ Cone I uded ) Common Name Sc i ent i f i c Name Gray-cheeked Thrush Ye I I ow Wagt a i I Orange-crowned Warbler Ye I low Warbler Ye I I ow-r umped Warbler Northern Waterthrush Wi I son ' s Warb ler Common Redpo I I Dark-eyed Junco Tree Sparrow White-crowned Sparrow Fox Sparrow Catharus minima Motaci 1 1 a f 1 ava Vermi vor a ce 1 ata Dendroica petech i a Dendroica coronata Seiurus novebor acens i s Wi I son i a pus i I I a Acanth is f I ammea Junco hyema I i s Sp i ze I la arborea Zonotrichia leucophrys Passere I I a i I i aca 389 U) 10 0) I. < (0 o SI D- (D 1_ C7) O 0) tD 1- O * > (D s in c 0) — -t- ■D 1/5 Q) > 0) i_ > 0) ■— in Ll JD 1 O ^ -*- u c o 0) S -1- 1- ^ O) Q) D x: (0 4— o >~ ■D JD C (D ■o Q) ■D 4- OJ C -1- Q) L. in O OJ CL i_ tu Q- cc Q) ce in 0) o 0) CL I/O ^ U) I < J3 (0 c >■ Q) TD in 3 Q) -*— 1- in D. (D O ■o •— I. 1- 0 o o 0) in 1- c :^ Q) n in D - 0) ■o in D Q) 4- i_ in Q- (0 O XI •— u i- o o u 4- ■ 0) T3 in ■3 0) 4— i_ in D. (0J3 O ■D •— i_ 1_ o o o 4- 0) in i- in 0) E ID C ■D o c — K} ^ C 4- o c E 0) E O o u in CN CN CN CN - in — (0 x: 3 (0 ID 1- 4— JD D 10 C ■D ^ U c o E c o O •— — D — •— CD 10 Q. c 4- .C (0 — o ^ — o ZJ 0) in c — ^ D in u (0 sz Q. (0 10 .— 10 — o Q) 3 in m (0 4- £ ^4- -— ^ — > in 4- in 0) ^ C in E u o c c ^ 3 O <\> D (U .— in 4- u • — .— 3 in c 4- O (0 o 5 C >~ L. 10 0 C o 1_ ^+- 3 • — in ■D ^*- -— 3 ^ (0 3 E ^ E (0 • — 3 c — CI. o S- o .c .^ • — ^ o 0) u — u — o E o • 5 • u • O • 5 i_ D 4- (0 c 10 ■ (0 . 0) . .^ • (0 1- • O 10 o o a> CJl Ul (0 Ol^ ol D. C co in o in Olin 0l>-0|0 OIU 1/5 l/ll - _l .— c ■a n •D C Q) O .— 4— 4— •— (D Q- in C o J^ E o J^ c 4- 0) o u 1_ o E (0 u o c 3 sz o o .^ u 1. 0) i_ 3 ^ ~ fl3 «J I. E O^ >. £ u h- £ e . O 1/1 0) Q. E >- 5 (/) X O O W) 3 E 0 m 3 (rt O o ^ E (D D U CL l_ tftf Q_ ». if) O 3 Ul JC E 3 u O 3 Ul 01 10 J3 0) O r o en 3 c c U 0 — 3 in O — ' Q. Q. <_) c to t/J (0^03 > U 1 I Ji: o q: cc u — i/i O 3 ID (0 c E Lu £ c C 01 (□ o (0 (0 £ O S S h— t- a. ■a 1! II II II II c (D rsi rO •^ in CN OJ CN CM CN '^ VI cc ct JZ 3 3 ■*- \^ i^ ic cr a. Q S S -> Q_ It) 0 ct I ji q: u cc j: ^ £ • r* C J£ > q: o - o ro (□ 0 O - > £ jc u r^ — ID (/) > Oi CT) — flj > £: J£ 1- O ■a ct «3 U) 3 a> 1- c (A l_ ~ i_ ■*- 0) 3 0) (0 OJ -J E C •«- Q. 0) (0 o cn 3 3 JC 0) ■o m z 3 < ^ tn E c n (0 c o V) 11 II II II II — u •O r^ CO O O >- — 39 1 Table A-5. Aquatic Macro i nvertebrates Caught at Study Sites During 1976-1978 Field Samp I ing Taxon Common name Nematoda 01 i gochaeta P I ecoptera A I I oper I a Arcynopter yx Capn i a Di ura Hastaper I a I sogenus I soper I a Nemour a Par aper I a Ephemeroptera Ame I etus Baet i nae Caeni s Ca I I ibaet i s Centropt i I um Ci nygmu I a Epeorus Ephemere I I a Heptagen i a Rh i throgena S i ph I onur us Odonata Ena I I agma I schnur a L i be I I u I i dae Tr i choptera Apatan i a Arctopsyche Brachycentrus Ecc I i somy i a G I ossosoma Homophy I ax Hydatophy I ax Lep i dostoma Leptoce I I a L i mneph i I us round worms ear thworms stonef I ies may f I i es dragonflies and damsel flies caddl sf I i es Cont i nued 592 Table A-5. I Cone I uded ) Taxon Common name Oecet is Onocosmoecus Phryganea P 1 atycentropus Pol ycentropus Pseudostenophylax Psychog 1 ypha Rhyacoph i I a Hemi ptera Cor i X i dae Col eoptera Dy t i sc i dae Hal i p I i dae D iptera Ather i c i dae Ceratopogon i dae Ch i ronomi dae Eph i d i dae Emp I d i dae Psychod i dae S imu I i i dae T i pu I i dae Hydracar i na Mol I usca Lymnaea Physi dae P i s i d i um P I anorb i dae Va I vata Amph i poda Gammar i dae water bugs waterboatman beet I es diving beet I e f I i es biting mi dge mi dge b I ackf I y crane f I y mi tes mo I I uses snai I snai I f i ngerna i I c I am snai I snai I amph i pods 393 APPENDIX B GLOSSARY abandoned channel — A channel that was once an active or high-water chan- nel, but currently flows only during infrequent floods. active channel — A channel that contains flowing water during the ice-free season. active floodplain — The portion of a floodplain that is flooded frequently; it contains flowing channels, high-water channels, and adjacent bars, usually containing little or no vegetation. aesthetics — An enjoyable sensation or a pleasurable state of mind, which has been instigated by the stimulus of an outside object, or it may be viewed as including action which will achieve the state of mind de- sired. This concept has a basic psychological element of individual learned response and a basic social element of conditioned social atti- tudes. Also, there can be ecological conditioning experience because the physical environment also affects the learning process of attitudes. algae — Primitive plants, one or many-celled, usually aquatic and capable of elaborating the foodstuffs by photosynthesis. aliquot — A portion of a gravel removal area that is worked independently, often sequentially, from the other portions of the area. alluvial river — A river which has formed its channel by the process of aggradation, and the sediment by which it carries (except for the wash load) is similar to that in the bed. arctic — The north polar region bounded on the south by the boreal forest. armor layer — A layer of sediment that is coarse relative to the material underlying it and is erosion resistant to frequently occurring floods; it may form naturally by the erosion of finer sediment, leaving coarser sediment in place or it may be placed by man to prevent erosion. auf ei s — An ice feature that is formed by water overflowing onto a surface, such as river ice or gravel deposits, and freezing, with subsequent layers formed by water overflowing onto the ice surface itself and f reez i ng . 395 backwater analysis — A hydraulic analysis, the purpose of which is to compute the water surface profile in a reach of channel with varying bed slope or cross-sectional shape, or both. bank — A comparatively steep side of a channel or floodplain formed by an erosional process; its top is often vegetated. bank-full discharge — Discharge corresponding to the stage at which the overflow plain begins to be flooded. bar — An al I uv i a I deposit or bank of sand, gravel, or other material, at the mouth of a stream or at any point in the stream flow. beaded stream — A sma I I stream containing a series of deep pools intercon- nected by very small channels, located in areas underlain by permafrost. bed — The bottom of a watercourse. bed load — Sand, silt, gravel or soil and rock detritus carried by a stream on, or immediately above its bed. bed load material — That part of the sediment load of a stream which is composed of particle sizes found in appreciable quantities in the shift- ing portions of the stream bed. bed, movable — A stream bed made up of materials readily transportable by the stream flow. bed, stream — The bottom of a stream below the low summer flow. braided river — A river containing two or more interconnecting channels separated by unvegetated gravel bars, sparsely vegetated islands, and, occasionally, heavily vegetated islands. Its floodplain is typically wide and sparsely vegetated, and contains numerous high-water channels. The lateral stability of these systems is quite low within the boun- daries of the active floodplain. carrying capacity, biological — The maximum average number of a given organ- ism that can be maintained indefinitely, by the habitat, under a given regime (in this case, flow). carrying capacity, discharge — The maximum rate of flow that a channel is capab I e of pass i ng . channel — A natural or artificial waterway of perceptible extent which periodically or continuously contains moving water. It has a definite bed and banks which serve to confine the water. configuration — The pattern of a river channel (s) as it would appear by looking vertical ly down at the water. contour — A line of equal elevation above a specified datum. 596 cover, bank — Areas associated with or adjacent to a stream or river that provide resting shelter and protection from predators - e.g., undercut banks, overhanging vegetation, accumulated debris, and others. cover, fish — A more specific type of instream cover, e.g., pools, boulders, water depths, surface turbulence, and others. cover, instream — Areas of shelter in a stream channel that provide aquatic organisms protection from predators or a place in which to rest, or both, and conserve energy due to a reduction in the force of the cur- rent . cross section area — The area of a stream, channel, or waterway opening, usually taken perpendicular to the stream centerline. current — The flowing of water, or other fluid. That portion of a stream of water which is moving with a velocity much greater than the average or in which the progress of the water is principally concentrated (not to be confused with a unit of measure, see velocity). datum — Any numerical or geometrical quantity or set of such quantities which may serve as a reference or base for other quantities. An agreed standard point or plane of stated elevation, noted by permanent bench marks on some solid immovable structure, from which elevations are meas- ured, or to which they are referred. dewater — The draining or removal of water from an enclosure or channel. discharge — The rate of flow, or volume of water flowing in a given stream at a given place and within a given period of time, expressed as cu ft per sec. drainage area — The entire area drained by a river or system of connecting streams such that all stream flow originating in the area is discharged through a single outlet. dredge — Any method of removing gravel from active channels. drift, invertebrate — The aquatic or terrestrial invertebrates which have been released from (behavioral drift), or have been swept from (catas- trophic drift) the substrate, or have fallen into the stream and move or float with the current. duration curve — A curve which expresses the relation of al I the units of some item such as head and flow, arranged in order of magnitude along the ordinate, and time, frequently expressed in percentage, along the abscissa; a graphical representation of the number of times given quantities are equaled or exceeded during a certain period of record. erosion, stream bed — The scouring of material from the water channel and the cutting of the banks by running water. The cutting of the banks is also known as stream bank erosion. 397 fines — The finer grained particles of a mass of soil, sand, or gravel. The material, in hydraulic sluicing, that settles last to the bottom of a mass of water. f lood — Any flowwhich exceeds the bank-full capacity of a stream or chan- nel and flows out on the floodplain; greater than bank-full discharge. floodplain — The relatively level land composed of primarily unconsolidated river deposits that is located adjacent to a river and is subject to flooding; it contains an active floodplain and sometimes contains an inactive floodplain or terrace(s), or both. flood probability — The probability of a flood of a given size being equaled or exceeded in a given period; a probability of I percent would be a 100-year flood, a probability of 10 percent would be a 10-year f lood. flow — The movement of a stream of water or other mobile substances, or both, from place to place; discharge; total quantity carried by a stream. flow, base — That portion of the stream discharge which is derived from natural storage - i.e., groundwater outflow and the draining of large lakes and swamps or other sources outside the net rainfall which creates the surface runoff; discharge sustained in a stream channel, not a result of direct runoff and without the effects of regulation, diversion, or other works of man. Also called sustaining flow. flow, laminar — That type of flow in a stream of water in which each par- ticle moves in a direction parallel to every other particle. flow, low — The lowest discharge recorded over a specified period of time. flow, low summer — The lowest flow during a typical open-water season. flow, uniform — A flow in which the velocities are the same in both magni- tude and direction from point to point. Uniform flow is possible only in a channel of constant cross section. flow, varied — Flow occurring in streams having a variable cross section or slope. When the discharge is constant, the velocity changes with each change of cross section and slope. fork length — The length of a fish measured from the tip of the nose to the fork in the tail. freeze front — A surface that may be stationary, which has a temperature of 0 C and is warmer on one side of the surface and colder on the other. frequency curve — A curve of the frequency of occurrence of specific events. The event that occurs most frequently is termed the mode. 398 gage — A device for indicating or registering magnitude or position in spe- cific units, e.g., the elevation of a water surface or the velocity of flowing water. A staff graduated to indicate the elevation of a water surface. geomorphology — The study of the- form and development of landscape fea- tures. habitat — The place where a population of animals lives and its sur- roundings, both living and nonliving; includes the provision of life requirements such as food and shelter. high-water channel — A channel that is dry most of the ice-free season, but contains flowing water during floods. hydraulics — The science dealing with the mechanical properties of fluids and their application to engineering; river hydraulics deals with mechanics of the conveyance of water in a natural watercourse. hydraulic depth — The average depth of water in a stream channel. It is equal to the cross-sectional area divided by the surface width. hydraulic geometry — Those measures of channel configuration, including depth, width, velocity, discharge, slope, and others. hydraulic radius — The cross-sectional area of a stream of water divided by the length of that part of its periphery in contact with its contain- ing channel; the ratio of area to wetted perimeter. hydrograph — A graph showing, for a given point on a stream, the discharge, stage, velocity, or another property of water with respect to time. hydrology — The study of the origin, distribution, and properties of water on or near the surface of the earth. ice-rich material — Permafrost material with a high water content in the form of ice, often taking the shape of a vertical wedge or a horizontal lens. impervious — A term applied to a material through which water cannot pass or through which water passes with great difficulty. inactive f I oodp I a i n — The portion of a floodplain that is flooded infre- quently; it may contain high-water and abandoned channels and is usually lightly to heavily vegetated. island — A heavily vegetated sediment deposit located between two channels. 2 large river — A river with a drainage area greater than 1,000 km and a mean annual flow channel top width greater than 100 m. lateral bar — An unvegetated or lightly vegetated sediment deposit located adjacent to a channel that is not associated with a meander. 399 Manning's equation — In current usage, an empirical formula for the calcula- tion of discharge in a channel. The formula is usually written Q = l^R 2/3 31/2 ^_ n mean flow — The average discharge at a given stream location computed for the period of record by dividing the total volume of flow by the number of days, months, or years in the specified period. mean water velocity — The average velocity of water in a stream channel, which is equal to the discharge in cubic feet per second divided by the cross-sectional area in square feet. For a specific point location, it is the velocity measured at 0.6 of the depth of the average of the velocities as measured at 0.2 and 0.8 of the depth. meander wave length — The average downvalley distance of two meanders. meandering river — A river winding back and forth within the floodplain. The meandering channel shifts downvalley by a regular pattern of ero- sion and deposition. Few islands are found in this type of river and gravel deosits typically are found on the point bars at the insides of meanders. 2 medium river — A river with a drainage area greater than 100 km but less than 1,000 km and a mean annual flow channel top width greater than 15 m but less than 100 m. microhabitat — Localized and more specialized areas within a community or habitat type, utilized by organisms for specific purposes or events, or both. Expresses the more specific and functional aspects of habitat and cover that allows the effective use of larger areas (aquatic and ter- restrial) in maximizing the productive capacity of the habitat. (See cover types, habitat). mid-channel bar — An unvegetated or lightly vegetated sediment deposit lo- cated between two channels. parameter — A variable in a mathematical function which, for each of its particular values, defines other variables in the function. permafrost — Perennially frozen ground. pit excavation — A method of removing gravel, frequently from below over- burden, in a manner that results in a permanently flooded area. Gravels are usually extracted using draglines or backhoes. point bar — An unvegetated sediment deposit located adjacent to the inside edge of a channel in a meander bend. pool — A body of water or portion of a stream that is deep and quiet rela- tive to the main current. 400 pool, plunge — A pool, basin, or hole scoured out by falling water at the base of a water fa I I . profile — In open channel hydraulics, it is the water or bed surface ele- vation graphed aganist channel distance. reach — A comparatively short length of a stream, channel, or shore. regional analysis — A hydrologic analysis, the purpose of which is to esti- mate hydrologic parameters of a river by use of measured values of the same parameters at other rivers within a selected region. riffle — A shal low rapids in an open stream, where the water surface is broken into waves by obstructions wholly or partly submerged. riparian — Pertaining to anything connected with or adjacent to the banks of a stream or other body of water. riparian vegetation — Vegetation bordering floodplains and occurring within f I oodp I a i ns. riprap - Large sediments or angular rock used as an artificial armor layer. river regime — A state of equilibrium attained by a river in response to the average water and sediment loads it receives. run — A stretch of relatively deep fast flowing water, with the surface essentially nont urbu I ent . scour — The removal of sediments by running water, usual ly associated with removal from the channel bed or floodplain surface. scrape - A method of removing floodplain gravels from surface deposits using tractors or scrapers. sediment discharge — The volumetric rate of sediment transfer past a spe- cific river cross section. sinuous river — Sinuous channels are similar to meandering channels with a less pronounced winding pattern. The channel may contain smaller point bars and have less tendency for downvalley shifting. The channels are more stable with respect to lateral shifting. sinuousity — A measure of the amount of winding of a river within its flood- plain; expressed as a ratio of the river channel length to the corres- ponding valley length. slope — The inclination or gradient from the horizontal of a line or sur- face. The degree of inclination is usually expressed as a ratio, such as 1:25, indicating one unit rise in 25 units of horizontal distance. 401 2 sma I I river - A river with a drainage area less than 100 km and a mean annual flow channel top width of less than 15 m. split river — A river having numerous islands dividing the flow into two channels. The islands and banks are usually heavily vegetated and stable. The channels tend to be narrower and deeper and the f I oodp I a i n narrower than for a braided system. stage — The elevation of a water surface above or below an established datum or reference. standing crop — The abundance or total weight of organisms existing In an area at a given time. straight river — The thalweg of a straight river typically winds back and forth within the channel. Gravel bars form opposite where the thalweg approaches the side of the channel. These gravel bars may not be ex- posed during low flow. Banks of straight systems typically are stable and f loodplains are usual ly narrow. These river systems are considered to be an unusual configuration in transition to some other configura- tion. subarctic — The boreal forest region. suspended load — The portion of stream load moving in suspension and made up of particles having such density of grain size as to permit movement far above and for a long distance out of contact with the stream bed. The particles are held in suspension by the upward components of turbu- lent currents or by colloidal suspension. talik — A zone of unfrozen material within an area of permafrost. terrace — An abandoned floodplain formed as a result of stream degradation and that is expected to be inundated only by infrequent flood events. thalweg — The line following the lowest part of a valley, whether under water or not; also usually the line following the deepest part or middle of the bed or channel of a river or stream. thermokarst — Landforms that appear as depressions in the ground surface or cavities beneath the ground surface which result from the thaw of ice-rich permafrost material. top width — The width of the effective area of flow across a stream chan- ne I . velocity — The time rate of motion; the distance traveled divided by the time required to travel that distance. wash load — In a stream system, the relatively fine material in near-perman- ent suspension, which is transported entirely through the system, without deposition. That part of the sediment load of a stream which is composed of particle sizes smaller than those found in appreciable quantities in the shifting portions of the stream bed. 402 water quality — A term used to describe the chemical, physical, and biolog- ical characteristics of water in reference to its suitability for a par t i cu I ar use. wetted perimeter — The length of the wetted contact between the stream of flowing water and its containing channel, measured in a plane at right angles to the direction of flow. wildlife — All living things that are neither human nor domesticated; most often restricted to wildlife species other than fish and invertebrates. 403 50272-101 REPORT DOCUWENTATrON. PAOC 1. REPORT NO. FWS/08S-80/08 3. Recipient's Accession No. 4. Title and Subtitle GRAVEL REMOVAL STUDIES IN ARCTIC AND SUBARCTIC FLOODPLAINS IN ALASKA - TECHNICAL REPORT 5. Report Date June 1980, Pub. date 7. Author(s) 8. Performing Organization Rept. No. WOODWARD-CLYDE CONSULTANTS 9. Performing Organization Name and Address Woodward-Clyde Consultants 4971 Business Park Blvd., Suite #1 Anchorage, Alaska 99503 10. Project/Tssk/Work Unit No. II. Contract(C) or Grant(G) No. (C)FWS 14-16-0008-970 (G) 12. Sponsoring Organization Name and Address U. S. Fish and Wildlife Service 1011 East Tudor Road Anchorage, Alaska 99503 13. Type of Report & Period Covered Final Report 1975 - 1980 14. IS. Supplementary Notes This report is part of Interagency Energy - Environment Research and Development Program of the Office of Research and Development, U.S. Environmental Protection Agency 16. Abstract (Limit: 200 words) A 5-year investigation and biological charact is described. Twenty-f sites were selected su a wide range of river covered the major disc quality, and terrestri tics site reviews were biological changes wer observed at some sites hydraulics, sedimentat macroi nvertebrates, fi mammal usage. of the effects of f er ist ics of r i ver sy ive sites were studi ch that within each and mining character iplines of hydrology al biology. In addit conducted. A wide r e observed in respon , whereas other site ion, ice regime, aqu sh ut i I izat ion , vege loodplain gravel mining on th stems in arctic and subarctic ed within four geographic reg of the regions the group of s istics. The field data col lee /hydraulics, aquatic biology, ion, geotechnical engineering ange of magnitude and type of se to mining activity. Little s exhibited changes in channe atic habitat, water quality, tation, soil characteristics. e physical Alaska ions. The ites exhibited tion program water , and aesthe- physical and change was I morphology, benth ic and bird and Two major products of the project are a Technical Report which synthesizes and evaluates the data collected at the sites, and a Guidelines Manual that aids the user in developing plans and operating material sites to minimize environmental ef fects. 17. Document Analysis a. Descriptors Gravel Removal, Alaska, Arctic, Subarctic, Floodplains, Streams, Scraping, Pit Excavation, Environmental Impacts, Hydrology-Hydraulics, Aquatic Biology, Terrestrial Ecology, Water Quality, Aesthetics, Geotechnical Engineering, Site Selection, Site Design. b. IdentlFiers/Open-Ended Terms c. COSATI Field/Group 18. Availability Statement Release un I imi ted 19. Security Class (This Report) Unci assi f ied 20. Security Class (This Page) Unci assi f ied 21. No. of Pages 403 22. Price (See ANSI-Z39.18) See /nsfructions on Reverse OPTIONAL FORM 272 (4-77) (Formerly NTIS-35) Department of Commerce # %