Bulletin 3r L Vi^-^. Wrr. 3<,-(L. atura.1 History Survey ..-iL •««»»•"*' "*"•••' 1981 blttlinKl Waterfowl Populations and the Changing Environment of the Illinois River Valley nlc C. Bellrose d L. Paveglio, Jr. laid W. Steffeck JUL I ^ m] TE OF ILUNOIS INOIS INSTITUTE OF NATURAL RESOURCES .TURAL HISTORY SURVEY DIVISION BANA, ILUNOIS VOLUME 32. ARTICLE 1 AUGUST 1979 us ISSN 0073^918 [SktvLi^SLl History Swrvey Waterfowl Populations and the Changing Environment of the inois River Valley nk C. Bellrose d L. Paveglio, Jr. nald W. Steffeck TE OF ILLINOIS INOIS INSTITUTE OF NATURAL RESOURCES iTURAL HISTORY SURVEY DIVISION BANA, ILLINOIS VOLUME 32, ARTICLE 1 AUGUST 1979 STATE OF ILLINOIS ILLINOIS INSTITUTE OF NATURAL RESOURCES BOARD OF NATURAL RESOURCES AND CONSERVATION Frank Beal, M.S.. Chairman; Thomas Park. Ph.D.. Biology; L. I. Sloss. Ph.D., Geology; H. S. Gltowsky, Ph.D., Chemistry; Stanley K. Shapiro. Ph.D., Forestry; W. L. Everitt, E.E.. Ph.D.. Representing the Presi- dent of the University of lUinois; John C. Guvon, Ph.D.. Representing the President of Southern Illinois Uni- versity. NATURAL HISTORY SURVEY DIVISION. Urbana, Illinois SCIENTIFIC AND TECHNICAL STAFF George Sprucel, Jr., Ph.D.. Chief Alice K. Adams. Secretary to the Chief Taxonomist and Section of Economic Entomology William H. Luckmann, Ph.D., Entomologist and Head James E. Appleby, Ph.D.. Entomologist Edward J. Armbrust. Ph.D., Entomologist Marcos Kocan, Ph.D., Entomologist Ronald H. Meyer. Ph.D.. Entomologist Stevenson Moore, IH, Ph.D.. Entomologist, Extension Michael E. Irwin. Ph.D., Associate Entomologist Donald E. Kuhlman, Ph.D., Associate Entomologist. Extension Joseph V. Maddox, Ph.D., Associate Entomologist Robert D. Pausch, Ph.D., Associate Entomologist RoscoE Randell, Ph.D., Associate Entomologist, Extension William G. Ruesink. Ph.D.. Associate Entomologist John K. Boiiseman, M.S.. Assistant Entomologist Catherine Eastman. Ph.D., Assistant Entomologist Allan Felsot, Ph.D., Assistant Entomologist Eli Levine. Ph.D., Assistant Entomologist Clarence E. White. B.S., Assistant Entomologist Luis R. Zavaleta, Ph.D., Assistant Entomologist Kevin D. Black. M.S.. Assistant Specialist, Extension David A. Gentry, M.S.. Assistant Specialist, Extension John Lublinkhof, Ph.D.. Assistant Specialist, Extension Kevin Steffey, Ph.D., Assistant Specialist, Extension Steven Troester, M.E., Assistant Systems Engineer Jean G. Wilson, B.A.. Siipemisory Assistant Lester Wei, Ph.D., Assistant Professional Scientist Charles G. Helm. M.S., Assistant Supffortive Scientist Stephen Roberts, B.S., Assistant Supportive Scientist John T. Shaw. B.S., Assistant Supportive Scientist Daniel Sherrod. M.S., Assistant Supportive Scientist Sue M. Hale, Junior Professional Scientist Robert J. Barney, B.S., Research Assistant Victoria Brunjes. B.S.. Research Assistant Michael Burke, M.S., Research Assistant Tzu-SUAN Chu, M.S., Research Assistant Marion Farris, M.S., Research Assistant Janet Harry, B.S., Research Assistant Jenny Kocan. M.S., Research Assistant Susan Post. B.S., Research Assistant Frank Ress, B.S., Research Assistant Barbara Stancer, B.S., Research Assistant Jo Ann Auble. Technical Assistant Ellen Brewer. M.S.. Computer Programmer Howard Ojalvo, M.S., Computer Programmer Section of Botany and Plant Pathology Glaus Grunwald. Ph.D., Plant Physiologist and Head Eugene Himelick, Ph.D.. Plant Pathologist Dan Neely, Ph.D.. Plant Pathologist D. F. ScHOENEWEiSS. Ph.D.. Plant Pathologist J. Leland Crane, Ph.D., Associate Mycologist Kenneth R. Robertson, Ph.D., Assistant Taxonomist Betty Nelson. Assistant Supportive Scientist Gene E. Reid, Junior Professional Scientist James E. Sfrgent, Greenhouse Superintendent Patricia O'Bryan, B.S., Research Assistant Robert A. Harrison, Technical Assistant Section of Aquatic Biology Robert W. Gorden. Ph.D.. Aquatic Biologist and Head D. Homer Buck, Ph.D.. Aquatic Biologist William F. Childers, Ph.D.. Aquatic Biologist R. Weldon Larimore, Ph.D.. Aquatic Biologist Robert C. Hiltibran, Ph.D., Biochemist Allison Brigham, Ph.D., Associate Aquatic Biologist Warren U. Brigham. Ph.D., Associate Aquatic Biologist Richard F. Sparks, Ph.D., Associate Aquatic Biologist John Tranquiili, Ph.D., Associate Aquatic Biologist Ken Lubinski, Ph.D., Assistant Aquatic Biologist Tfd W. Storck, Ph.D., Assistant Aquatic Biologist Richard J. Baur, M.S.. Assistant Supportive Scientist Dee a. McCormick, M.S., Junior Professional Scientist Eu(;ene Sons. Junior Professional Scientist Jana Lee Waite, M.S., Junior Professional Scientist David P. Philipp. Ph.D.. Research Associate Stephen W. Waite. ^LS.. Research Associate Harry W. Bergmann. B.S.. Research Assistant CONSULTANTS AND RESEARCH AFFILIATES: Systematic Fntomoio(.y, Roderick R. Irwin. Chicago. Illinois: Wildlife Research, Willard D. Ki.imstra. Ph.D., Professor of Zoology and Director of Cooperative Wildlife Research, Southern Itlinf>i\ l'niver\ily; Parasitoi o(.Y, Norm\n D. Levine, Ph.D., Professor of Veteri- nary Parasitology. Veterinary Research and Znology and Director of the Center for Human Ecology, l^nii-ersily of Illinois: Entomology, Rohfrt L. Metcalf, Ph.D.. Professor of Biology and Research Professor of Entomol- ogy. University of Illinois; and Gilbert P. Walobauer, Ph.D.. Professor of Entomology. University of Illinois; Statistics, Horace W. Norton, Ph.D., Professor of Statistical Design and Analysis, University of Illinois. Dale Burkett. B.S.. Research Assistant Kurt T. Clement, B.S.. Research Assistant Larry W. Coutant, M.S., Research Assistant Robert Graham, B.S., Research Assistant William Kraus, M.S., Research Assistant Richard Krause, M.S.. Research Assistant Sarah Liehr. M.S.. Research Assistant Lance Perry. M.S., Research Assistant Thomas Skelly, M.S.. Research Assistant Michael J. Sule, M.S.. Research Assistant Stephen O. Swadener, M.S.. Research Assistant Bruce Taubert. Ph.D., Research Assistant Carl Alde, B.S., Technical Assistant Elmer Atwood. B.S., Technical Assistant Paul Beaty, Ph.D.. Technical Assistant Bill Dimond. B.S.. Technical Assistant Kathryn Ewing, B,S., Technical Assistant Iefe Hutton, B.S.. Technical Assistant Sheila Magfe, B.S.. Technical Assistant Philip Mankin, Technical Assistant Dan Myrick, B.S., Technical Assistant F. J. Partenheimer^ B.A.. Technical Assistant R. Dan Sallee, B.S.. Technical Assistant Ifns Sandberger. M.S.. Technical Assistant Michael Sandusky, B.S.. Technical Assistant John J. Suloway, B.S., Technical Assistant Gary L. Warren. B.S.. Technical Assistant Mark J. Wetzel, B.S., Technical Assistant Ruth Wagner, Junior Technical Assistant Section of Faunistic Surveys and Insect Identification \Vallace E. LaBer{.f, IMi.D,. Ins, Head George L. Godfrey, Ph.D., Associate Taxonomist Larry M. Page, Ph.D., Associate Taxonomist John D. Unzicker. Ph.D.. Associate Taxonomist Donald W. Webb, M.S., Associate Taxonomist Bernice p. Sweeney, Junior Professional Scientist Section of Wildlife Research Glen C. Sanderson, Ph.D., WildUfe Specialist and Head Frank C. Bellrose. Sc.D., Wildlife Specialist William R. Edwards. Ph.D.. Wildlife Specialist Iean W. Graber. Ph.D., Wildhfe Specialist Richard R. Graber, Ph.D.. Wildlife Specialist Harold C. Hanson, Ph.D.. Wildlife SpeciaUst W. W. Cochran, Jr., B.S., Associate Wildlife Specialist Charles M. Nixon. M.S.. Associate Wildlife Specialist Kenneth E. Smith. Ph.D., Associate Chemist Ronald L. Westemeier. M.S.. Associate Wildlife Specialist I.ONNiE P. Hansen, Ph.D.. Assistant Wildlife Specialist Stephen P. Havera. Ph.D.. Assistant Wildlife Specialist Richard E. Warner, M.S.. Assistant Wildlife Specialist RoBFRT D. Crompton. Junior Professional Scientist Ronald E. Duzan. Junior Professional Scientist Iames W. Seets, Junior Professional Scientist Eva Steger, B.S., Junior Professional Scientist Eleanore Wilson. Junior Professional Scientist Supporting Services WiiM \ G. Dii I man. Properly Control and Trust Accounts Patty L. Duzan. Payroll and Personnel Robert O. Ellis. Assistant (or Operations Larr^' D. Gross. Operations Assistant L Wii.ii\M I.USK, Mailing and Distrihution Services Chris Rohl, Operations Assistant Melvin E. Schwartz. Fiscal Officer Publications and Public Relations RoBFRT M. Zfw\dski. \LS,. Technical Editor Shirley McClei.lan. B.S., Assistant Technical Editor Li.o^n LeMerf. Technical lllustralur Leslie Woodrum. Technical Photographer Technical Library Doris L. Sublette. \LS.L.S.. Technical Lihrarian Monica Li'sk, I.ihraiy Clerk CONTENTS Geological History 1 Historical Changes 2 C:hanges during the Present Century 3 Acknowledgments 5 Methods 6 Sedimentation 6 Water Depth 6 Wetland Vegetation 8 Waterfowl Populations 8 Moist-Soil Water-Level and Water-Stability Indices 8 Waterfowl Food Resources 9 River Levels 10 Wetland Vegetation 13 Evaluation of t-he Severance of Lakes from the River 20 Evaluation of Seasonal Changes in Bottomland Lake Water Levels 20 Effects on Moist-Soil Plants 20 Effects on Aquatic and Marsh Plants 21 Lake Chautauqua 21 Rice Lake 22 Cuba Island 22 Flat, Swan, and Gilbert Lakes 22 Changes in Water Depths , 23 Rice Lake 23 Douglas Lake 23 Spring Lake 27 Anderson Lake 27 Effects of Turbidity and Sedimentation 28 Effects of Food Resources on Fall Waterfowl Populations 34 Summer Water Levels and Fall Waterfowl Abundance 37 Fall Water Levels 39 Water Levels in Relation to the Duration of Stay of Ducks 39 Management Practices and Considerations 42 Discussion 46 Summary 48 Literature Cited 51 List of Common and Scientific Names 52 Index 53 This report is published by aulhority of the State of Illinois. It is a contribiitinn from the Section of Wildlife Research of the Illinois Nnhiral History Survey. Dr. Frank C. Bellrose is a Wildlife Specialist and Fred L. Paveglio, Jr., and Donald W. Stefjeck are Research Assistants in the Section of Wildlife Research. (I3447-SM-8-79) c ^ \c * \ ^ < 4 ^ If > > K Waterfowl Populations and the Changing Environment of the Illinois River Valley Frank C. Bellrose, Fred L. Pavegllo, Jr., and Donald W. Steffeck THE VALUE OF THE ILLINOIS RIVER valley for waterfowl lies in its bottomland lakes that flank the rel- atively narrow river channel between Spring Valley and Meredosia and be- tween Pearl and Grafton. These lakes originated in the most recent glaciation (Wisconsinan), which profoundly al- tered the ancient Mississippi and Illi- nois rivers and created the unique Il- linois Valley that we find today. Up to the 1930's, the Illinois River valley was one of the nation's outstanding waterfowl hunting areas. Since then it has been in a state of decline for a variety of reasons, some of which will be discussed here. GEOLOGICAL HISTORY Prior to the Wisconsinan glaciation, the Mississippi River flowed down the Illinois Valley below the Big Bend at Hennepin, Illinois, by means of a now buried channel between Bureau and northeastern Rock Island counties (AVillman Sc Frye 1970). About 21,000 years ago, the most recent ice sheet, the Wisconsinan, moved westward past the Big Bend and diverted the Mississippi River westward to its present channel south of Muscatine, Iowa (Willman 1973). As the Wis- consinan glaciation retreated, flood waters formed the Des Plaines and Kankakee rivers, which joined near Chnnnahon to create the present Il- linois River. From Channahon west- ward to the Big Bend at Hennepin, tlic Illinois carved a new valley. At the Big Bend, the flood waters of the glacial melt entered an ancient valley of the Mississippi River and followed this pathway southward to the present Mississippi River at Grafton. Because the ancient Mississippi Val- ley liirough central Illinois had been broailened and deepened by repeated pre-Wisconsinan glacial melts, the Il- linois River entered a much deeper valley below Hennepin than was war- ranted by its volume of water. The valley has also been considerably filled with sediment carried by the glacial meltwater. This origin resulted in a river with an luiusually low rate of fall: 0.03 m per km (0.17 ft per mile) between Hennepin and Pekin, and 0.02 m per km (0.13 ft per mile) from Pekin to Meredosia. Forbes & Rich- ardson (1920) reported that at ordinary levels the flow varied from 2.01 to 4.02 km (1.25-2.50 miles) per hour. Since 1938, navigation dams have fmther re- tarded the river velocity to about 1.0 km (0.6 mile) per hour (Starrett 1971). The gentle slope of the river below the Big Bend has resulted, during postglacial times, in the valley's ag- grading rather than eroding, as do most river valleys. The Illinois River's low volume of flow for its channel capacity and its low rate of fall combined to form the imique bottomland lakes associated witii the Illinois Valley. Under over- flow conditions, the faster-moving wa- ters of the channel meet the slower- moving backwaters with the result that sediment is deposited more rapidly along this shear. Through eons of time natural levees rose, like barrier islands, to separate most of the chan- nel waters from the adjacent bottom- land waters. Thus evolved bottomland lakes, or backwater lakes, as the en- gineers term them (Fig. 1). The very creation of bottomland lakes also set the stage for their ex- tinction. The sedimentation that sep- Illinois Natural History Survey Bulletin Vol. 32, Art. I Fig. 1.—Illinois River north of Chillicothe, illustrating the natural banks that formed bot- tomland lakes in the Illinois Valley. The channel is shown meandering between Babbs Slough, Sawyer Slough, Big Meadow Lake, Wightman Lake, and Sparland Lake. arated the lakes from the river chan- nel is now rapidly engulfing them. Under pristine conditions, this extinc- tion of individual bottomland lakes would probably have taken himdreds of years, and while one generation of lakes was being filled with sediment, another generation would have been formed by the growth of new natural levees. Man, through intensive use of the land, has greatly accelerated the process. The Illinois Valley is especially sub- ject to sedimentation because its trib- utary .streams fall many limes faster than does the main stream. For ex- ample, the slope of the .S|ioon River, an important triljutary, varies from 0.19 to OM m per km (1-3 ft per mile) (Evans 8: Schnepper 1977). Conse- quently, the rivers and creeks feeding the Illinois River transport enormous quantities of finely suspended soils to its waters. Since it flows more slowly, tlie Illinois in turn deposits a sizeable proportion of this load in its network of bottomland lakes during floods. At lower water stages the river waters mingle very little with the backwater lakes. Only Peoria Lake, through which the river (lows, is contiiniously exposed to transported material. HISTORICAL CHANGES For 150 years following Pere Mar- quette's and Loin's Joliet's ascent of ihc Illinois River in 167.^, its valley was |)opulatetl largclv bv Indians and a few white traders and irajipers. In 1823 several while families lived at the present site of Peoria, and Chicago was known as a military and trailing post (Barrt)us 1910). However, the es- tablishment of steamboating in 1S2,S bioiighi an influx of immigrants, so thai i)v 1810 much of the Illinois Val- lev (ouiaincd fi-lS peo])le per square mile. ,\fter 1810 the population of the Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley valley grew apace. By 1900 there were 3.3 million people in the Illinois basin, 95 percent of which lies within the state of Illinois. The basin embraces slightly over half the area of the state. To facilitate transportation, a canal was constructed in 1848 from Lake Michigan at Chicago to the Illinois River at La Salle. In addition, three low navigation dams and locks were constructed on the Illinois River, at Henry in 1871, at Copperas Creek near Banner in 1877, and at La Grange in 1889. These changes appear to have had little effect upon the natural biota of the Illinois Valley; its lakes were still nearly pristine. Detailed maps of the Illinois Valley made by J. W. Woermann, Corps of Engineers, U.S. Army, from 1902 to 1904 indicate that slightly over one- third of the meadows and wooded bot- tomlands had been cleared and placed in some form of cultivation. During tlie late 1890's, the waters of the Illinois River still ran compar- atively clear. Kofoid (1903:151) de- scribed a boat trip across a series of bottomland lakes above Havana dur- ing high water in late May. Despite the high stage, waters were surprisingly clear; aquatic and marsh plants were besjinning to appear in abimdance, with coontail still "at some depth be- low the surface." In the waters of Thompson Lake (one of the largest lakes in the valley), Kofoid could see schools of young fry feeding upon plankton. Its waters were described as "somewhat turbid but more from plankton th.m silt." The river channel was more turbid from silt and plank- ton than were the adjacent lakes. Later in the summer, Kofoid (1903: l.')5) revisited the same lakes during low water. He found "the backwaters have been reduced to the lakes, sloughs, and marshes which abound everywhere in the bottomlands." Most of the lakes were choked with aquatic plants or rushes. His descriptions of bottom soils suggest that they were composed more of decayed plant material than of silt. CHANGES DURING THE PRESENT CENTURY From 1900 to the present, the nat- ural habitats of the Illinois Valley have experienced a sequence of over- lapping catastrophic events. Had the Illinois River and its backwater lakes been a fragile ecosystem, they would ha\'e passed into oblivion long ago. With varying degrees of success, the aquatic communities of the Illinois Valley have withstood the ravages of man. The first pronounced change oc- curred in 1900 with the completion of the Chicago Sanitary and Ship Canal, which diverted water from Lake Mich- igan and sewage effluetit from Chicago to the Illinois River. The diversion ranged from 82.1 cubic meters per second (cms) (2,900 cubic feet per sec- ond-cfs) in 1900 to 283.4 cms (10,010 cfs) in 1928; from 1900 to 1938, it av- eraged 204.5 cms (7,222 cfs). A U.S. Supreme Court decree limited diver- sion to 42.5 cms (1,500 cfs) after 1938, biu a second decree in 1961 increased the rate to 90.6 cms (3,200 cfs). Yearly water profiles at Peoria, 1868-1975, show that low-water levels climbed from 131.7 m (432 ft) mean sea level (msl) in 1899 to 132.1 m (433.5 ft) in 1900 and 133.2 m (436.9 ft) in 1902. Until the Peoria lock and dam became operational in 1938, low-water levels averaged aboiu 133.5 m (138.0 ft) msl. The diversion of water into the Illinois River appears to have increased low-water levels at Peoria between 1.5 and 1.8 m (5 and 6 ft). Forbes k Richardson (1919:140- 141) concluded that midsummer levels at Havana rose an average of 1.1 m (3.6 ft) above prediversion averages. The diversion caused the flooding of thousands of hectai'es of bottomland Illinois Natural History Survey Bulletin Vol. 32, Art. I forest during the growing season, kill- ing the trees and enlarging the back- water lakes. Pin oaks and pecans, which had grown extensively in the valley below Peoria, were especially vulnerable to higher water. Their loss greatly reduced the mast food supply for mallards and wood ducks. Overall, though, waterfowl benefited from tiie diversion of Lake Michigan water as the surface areas of lakes, sloughs, and marshes doubled. Prior to diversion, about 21,850 ha (54,000 acres) were in bottomland water areas, exclusive of the river channel. The diversion of Lake Michigan water ex- panded bottomland lakes and marshes to slightly over 48,560 ha (120,000 acres) in the 161,878-ha (400,000-acre) floodplain. This expanded water area did not persist long but shrank due to man's next alteration of the Illinois River valley. Levee and drainage districts, largely initiated between 1903 and 1920, placed 82,962 ha (205,000 acres) of bot- tomland behind levees (Mulvihill 8c Cornish 1929). Natural lakes and marshes within the leveed tracts were drained. The land in three ilrainage districts—Partridge, Cliautauqua, and Big Prairie, aggregating 3,238 ha (8,000 acres)—was not farmed successfully and reverted to a seminatural state (Fig. 2). Ironically, this drainage district loss in- creased the lake and marsh smface area to about 28,329 ha (70,000 acres), more than the predi version 21,854 ha (54,000 acres). Irretrievably lost to waterfowl were the pin oak and pecan groves that had provided important food resources during floods. At the same time that Illinois Valley lakes were being altered by the cre- ation of levees and drainage districts, the aquatic biota was being threatened by yet another serious menace—that of mban wastes discliargcd into the Illi- nois River. Although a small amount of pollution had occurred before 1900. the opening of the Chicago Sanitary and Ship Canal in 1900 created a prob- lem of catastrophic proportions for the river and its backwater lakes above Peoria (Mills et al. 1966). At first only the extreme upper reaches (above Mar- seilles) were affected. However, the zone of pollution steadily moved down- stream luitil by 1922 the upper Illinois was essentially a dead river, devoid of important aquatic life as far south as Chillicothe (Mills et al. 1966). Extensive beds of pondweeds, wild celery, and coontail, present in Peoria Lake between 1910 and 1914, had largely tlisappeared by 1920 (Richard- son 1921). In 1921, Peoria Lake con- tained slight growths of pondweeds and algae, so little as to be overlooked by the casual observer (Purdy 1930). A decline in urban and industrial pollution began with the operation of treatment plants by the Chicago San- itary District in 1922. Pollution con- trol was aided by the navigation dams tliat became operational in the upper river in 1933. These dams reduced the rate of flow, thereby resulting in bac- terial decomposition of waste products witliin a shorter distance downstream. A gradual reduction in the urban pol- lution of the Illinois River has contin- ued to the present time (Richard Sparks, Illinois Natural History Sur- vey, personal comnuuiication). The 9-foot (2.7-m) waterway for nav- igation became operational above the Starved Rock lock and dam in 1933 ami below that point in 1938. The water- way below Starved Rock was created by dredging the channel deeper ami raising tlie water levels by dams al Peoria, below Bcardstown. and on the Mississippi River at Alton. Even as the inbaii and industrial pol- lution abated, tlie Illinois River and its lakes were subjected to yet another degradation, that of pollution from sedimentation. Tlie ]3ermanent and in- siilious nature of silt pollution makes it more iiannful ih.in url).ui polluiion: ailiiougii not as ap])arenl, it is aciumu- lativc. I'his report examines the role ol I Aug. 1979 Bellrose et al.: Waterfowl and the Chancing Illinois Valley Fig. 2.—A portion of the abandoned Partridge Levee District located at the northern end of Upper Peoria Lake across from Chillicothe, showing a leveed area (center) dewatered for moist-soil plant development. sediiiieiuation in tlic Illinois Valley: its eflects on a(|uatii, niaisli, and moist- soil ]jlants and its eficds upon water- low! abmulance. ACKNOWLEDGMENTS We are indebted to these persons and organizations for their (ontribiitions to I his paper. Pari of the material ])rescnted, par- tiiularly that on sedimentation, was derived from stndies financed by the Department of the Army, Cliicago Distritt, (;orps of Enj^incers, Project DACW 23-70-6-0066 nmlcr project leader Edward Hanses, Dr. David L. Gross, Illinois State (.eological Siavey, reviewed the section on Geological History and made per- tinent siiggesiicjiis. Dr. Vernon Wright, formerly with the Illinois Department of (conservation, ])rovided guitlance in the statistical analyses. Eorrest Loomis, now with the Illi- nois Department of Conservation, con- (liKied field stndies and made vegeta- lioii maijs of botlonilaiul lakes, 1955- 1957. Drs. Leigh Eredcric kson of the Uni- versity of Missouri and David 'Eraugcr of the U..S. Eish and Wildlife .Service reviewed the pa|)er and offered sugges- tions for improvement. A number of Illinois Natural His- Illinois Natural History Sur\'ey Bulletin Vol. 32, Art. 1 tory Survey stafF members contributed to this paper: Lloyd LeMere, Tech- nical Illustrator, drew the graphs. Dr. Glen C. Sanderson, Head, Section of Wildlife Research, Mrs. Eva Steger, and Harold Henderson did preliminary editing. Final editing and preparation of the paper for publication was done by Robert M. Zewadski, Technical Editor. METHODS Three diverse aspects of Illinois River habitat were measured for this report: (1) the rate of sedimentation in various bottomland lakes, (2) the abundance and distribution of aquatic, marsh, and moist-soil plants, and (3) the abundance of waterfowl as related to the availability of aquatic and moist- soil vegetation and mollusks (Fig. 3). SEDIMENTATION Reports on the sedimentation of three bottomland lakes have been is- sued by personnel of the Illinois State Water Survey (M. T. Lee. 1976. Sedi- ment deposition of Lake Chautauqua, Havana, Illinois. Unpublished Xe- roxed report. M. T. Lee and J. B. Stall. 1976. Sediment deposition in Lake DePue, DePue, Illinois and its implications for future lake manage- ment. Unpublished Xeroxed report. M. T. Lee, J. B. Stall, and T. A. Butts. 1976. The 1975 sediment survey of Lake Meredosia, Meredosia, Illinois. Unpublished Xeroxed report). In ad- dition, we made sedimentation studies on these lakes: Senachwine, Sawmill, Billsbach, Sparland, Wightman, Babbs Slough, Upper Peoria, Rice, Chautau- qua, Anderson, and Meredosia. Comparisons between present-day water depths and earlier readings were possible because of soimdings con- ducted in the Illinois Valley in 1902- 1904 by J. W. Woermann, Assistant Engineer, Corps of Engineers, U.S. Army. Differences in water depths of lakes, sloughs, and marshes were used to determine the amount of sedimen- tation that has occurred during an in- terval of 73-75 years. Fortimately, at three lakes sedimen- tation could be separated into two periods during this time span. At Lake Chautauqua, Stall k Melsted (1951) studied sedimentation between 1926 and 1950, and Lee (1976 impublished) studied it during 1950-1976. The Peoria Project Office, U.S. Army Corps of Engineers, took soundings of Peoria Lake in 1965. These studies enabled us to calcidate the sedimentation at Peoria Lake for two periods: 1903-1965 and 1966-1976. Soundings of Meredosia Bay were taken in 1956 by the State Division of Water Resources and by us in February 1978. Sedimentation rates were calculated for the periods 1903-1956 and 1957-1978. WATER DEPTH In our 1976 and 1977 soundings, 3- 10 transects were established across each lake. Depths along the transects were taken at approximate intervals of 77.7 m (85 yards) except in Upper Pe- oria Lake, where the interval was 228.6 m (250 yards). The depth of water was related to the river stage (i.e., to the nearest gauge reading adjusted for the slope of the river). From these data the mean sea level (msl) elevation of each lake bottom was derived. The eleva- tions of lake bottoms on the 1902-1904 y. W. Woermann maps were based upon the Mempliis Datum (MD). The Memphis Datiun was an arbitrary el- evation at Memphis, Tenn., used as a reference point in measuring other el- evations. The Memphis Datimi was converted to msl (1929 general adjust- ment) by subtracting a correction fac- tor varying from 2.26 to 2.27 m (7.43- 7.45 ft) depending upon the location of the bench marks. The areas of bottomland lakes in the Illinois River valley were detcrmineil by using a planimeter to measure the basin areas as delineated by the Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 7 wooded-vegetation line on the U.S. series of the Woemiann 1902-1904 Army Corps of Engineers 1933 revised maps. LAKE MICHIGAN Swan and Flat MISSOURI ST, LOUIS 25 MILES 50 50 KILOMETERS Fig. 3.—Map of the Illinois River valley, showing many of the bottonnland lakes included in this study of wetland plants, sedimentation, and waterfowl abundance. Illinois Natural History Survey Billetin Vol. 32, Art. I WETLAND VEGETATION The areas of waterfowl food plants (aquatic, marsh, and moist-soil species) were taken from maps drawn by Frank Bellrose, 1938-1953, and by Forrest Loomis, 1955-1957. Waterfowl food plant beds were plotted by rough tri- angulation on the base maps provided by the U.S. Army Corps of Engineers 1933 map series scaled 1:12000 (Bell- rose 1941). Later vegetation maps pre- pared for 1959 and 1976 were based upon aerial photographs interpreteil by Bellrose. Mapped plant beds were measured by planimeter to determine their sizes in acres. To weight areas devoid of vegetation, acreages of plants were di- vided by the acreage of the bottomland lake basin mapped each year to obtain the percentage of the area covered by waterfowl food plants. To compare an- nual changes in vegetation composi- tion, the abundance of each plant spe- cies was determined and related to the entire lake basin. The measuretl acre- age of each lake basin varied slightly from year to year, depending upon the specific area mapped. Recent informa- tion on the status of wetland plants on Gilbert and Swan lakes was derived from the files of the Mark Twain Na- tional Wildlife Refuge, Quincy, Illi- nois. WATERFOWL POPULATIONS Censuses of waterfowl were taken in the Illinois Valley 1938-1970 by Frank Bellrose and 1971-1976 by Robert Crompton. However, the data are bet- ter since 1949, when a light aircraft made it possible to cover the valley completely 1 day each week from 1 Oc- toljer to 1 December. Waterfowl num- bers oijtained from the weekly censuses were multiplied by seven to obtain the duck-days of use for each fall. MOIST-SOIL WATER-LEVEL AND WATER-STABILITY INDICES To evaluate the relationship of river levels to the abiuuiance of wetland plants, a growth-period index was es- tablished for wetland plants and a water-level index was established bv computing weekly averages from daily water-level gauge readings. We converted National Oceanic and .\tmospheric Administration Environ- mental Data and Information Service daily gauge readings to weekly aver- ages. Because of tiie sluggish nature of the river, weekly averages adequateK reflect its rise and fall. The growth-period index was based upon the assumption that the earliest low-water stage that could benefit wet- land plants is 12 18 June, the latest, 14-20 August. After that date too few days remain before the average date of the first heavy frost for plants to pro- duce a meaningfid amount of seeds. The weeks of this 10-week period were ranked in descending order with the week of 12-18 June ranked 10 and that of 14-20 August ranked 1. Weekly average gauge readings were assigned water-level index values that increased as the water level receded from the wooded shoreline, exposing an area of lake basin for the tlevelop- ment of moist-soil plants. The lower the river level, the larger the assigned water-level index. Water levels at or above the wooded shoreline were desig- nated as 0. The upper limits of the wooded shoreline elevation on the Henry gauge is 135.0 m (443 ft) msl and on tiie Havana gauge it is 133.2 ni (437 ft) msl. Each water-level index was multi- plied by the growth-period index. The resulting products were squared and then totaled for the 10-week period. The result is the Moist-Soil AVater- Lcvcl Index. The products were squared because the exposure of muil rials is exponentially rather than lin- early related to declining water levels. With the exception of Peoria Lake, lake basins of the Illinois River are shallow and platier shaped. Therefore. as water levels decline, an ever larger proportion of the lake basin is exposed as unid flats. Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley The Moist-Soil Water-Level Index was first compared with the moist-soil vegetation acreage over a 22-year pe- riod, 1938-1959, by linear regression. Secondly, the Moist-Soil Water-Level Index was compaied by linear regres- sion with the fall duck-day use of Illi- nois Valley lakes, 1949-1976. On the other hand, to evaluate the effect of water levels on aquatic aiitl marsh plants, a Water-Stability Index was calcidated. A level 0.6 in (2 ft) above the prevailing low-water stage was judged to be optimum for aquatic plants. The absolute values of devi- ations from this level, based upon weekly gauge averages, were totaled for the growing season. The larger the Water-Stability Index, the more adverse were conditions for aquatic and marsh plant development. WATERFOWL FOOD RESOURCES A preliminary report on the water- fowl food resources of the Illinois Val- ley and the factors affecting them dur- ing the 1938-1940 period was made by Bellrose (1941). We present here those findings and others made since then. From the earlier study (Bellrose 1941) it is apparent that the four most im- portant factors affecting the abundance of duck food plants in bottomland lakes of the Illinois River are (1) fluc- tuating water levels, (2) turbidity, (3) water depth, and (4) competition by other plants that provide little or no duck food. In the present study we sought to examine further the effects of each of tliese factors upon the vegetation of valley lakes, but we also recognize that all of the factors are interrelated. Tur- bidity stems from sedimentation, which in turn alters water depth and may en- courage the growth of some weed .spe- cies at the expense of more beneficial plants. When the river rises, more sed- iment is deposited and turbidity in- creases. Fluctuating water levels can affect waterfowl habitats in opposite ways. Uncontrolled fluctuations result in min- imal development of aquatic, marsh, and moist-soil plants. However, con- trolled fluctuations can result in highly productive food resources if they occur within the optimimi 120-day summer growing period. Because the degree of control of wa- ter levels is so important to the water- fowl food plant resources of the Illinois Valley, we have separated bottomland lakes into four classes according to their association with the river (Ta- ble 1): Class I, areas whose water levels are completely controlled by chang- ing river levels Class II, areas that are separated from the river at low water stages Class III, areas that remain sep- arated from the river up to flood stage, when the river begins to over- top its natural banks Class IV, areas that are above the river's designated flood stage (bank full) Class III areas are further categorized according to management practices: Class IIIA, areas where water levels are stabilized at an optimum level below flood stage for aquatic and marsh plants, and Class IIIB, areas where the lake is dewatered for optimum produc- tion of moist-soil plants. Under severe flooding all areas but Spring Lake, near Pekin, come under the influence of the river. The entire .'>20-ha (1,285-acre) basin of Spring Lake is separated from the river by high levees; water control structures added in 1977 will make it possible in the fu- ture to manipulate water levels on 261 ha (645 acres) of Spring Lake. Other wetlands are separated to vary- ing degrees from the influence of the river, either becau.se of the elevation and formation of their basins, or be- cause low levees isolate the areas. Low levees have been built by private duck clubs, the Illinois Department of Con- .servation, and the U..S. Fish and Wild- life Service to provide .some low-level water control on wetlands under their 10 Illinois Natural History Survey Bulletin Vol. 32, Art. 1 respective ownerships. Most of the structures isolating tlie river waters are no higlier than the natural banks of the river. Pumps are frequently em- ployed to assist in the dcwatering and rellooding of impounded wetlands. Table 1.—Surface areas of Illinois River valley lakes in relation to the degree of separation from the influence of the river, 1977.' Figures are in hectares (2.471 acres/hectare). Aug. 1979 Bellrose et al.: Waterfowl and the Chancing Illinois Valley 11 The arrangement of levees often sep- arates wetlands into more than one category. Through tlie years 1938- 1977, the percentage of the lake basins Table 1. Continued 12 Illinois Natural History Survey Bulletin Vol. 32, Art. associated with the river has changed as some levees have been destroyed by floods and as others have been newly constructed. Table 1 shows the waters in the four categories of control as of the summer of 1977. However, for pur- poses of evaluating wetland vegetation in association with water levels, the de- gree of water level control at the time of the study is discussed. Local aher- ations in water level were made on cer- tain lakes during this period, compli- cating their evaluation. RIVER LEVELS Three navigation dams, at Peoria, La Grange, and Alton, influence water levels, and thus bottomland lakes, in the Illinois River valley. These dams began operation late in 1938 and have influenced river levels and vegetation from the 1939 growing season to the present. Although the Alton dam is on the Mississippi River 24.1 km (15 miles) below the mouth of the Illinois River, its navigational influence extends 128.7 km (80 miles) up the Illinois to the La Grange lock and dam. Similarly, the La Grange navigation pool extends 125.5 km (78 miles) to the Peoria lock and dam, and the Peoria navigation pool extends 117.5 km (73 miles) to the Starved Rock lock and dam near Utica. Other navigation dams—Mar- .seilles, Dresden Island, Kraiulon Road, and Lockport—form adilitional naviga- tion pools upstream, but they are only of incidental concern in this report. Because of its great height, which re- sulted in raising the river level 8.4 m (27.7 ft), the ,*\llon dam has had the most influence on the water levels of lakes upstream from the mouth of the Illinois River. Both La Cirange and Peoria are low-level ilams, 3.0 m (10 ft) ami 3.4 m (11 ft), respectively, and have had relatively minor cllects upon ri\cr levels. Of course, as with all ilams, their greatest control over river levels is at the clam site, and their influence upstream dcdines according to their height and the rate of fall of their pools. River control progressively de- clines to the tailwaters of the next dam upstream, where water levels behave aboiu as they did in preimpoiuidment days. As far as we can ascertain, the Alton dam has little influence over river lev- els above Pearl, the La Grange dam above Havana, and the Peoria clam above Henry. One reason that they ap- pear to have had so little influence on present-day water levels is that when they became operational in 1938 the diverted flow of water from Lake Mich- igan into the Illinois River was reduced from 14I.f3 cms (5.000 cfs) to 42.5 cms (1,500 cfs). In 1961, the discharge was raised to 90.6 cms (3,200 cfs). Both the Havana and Henry gauges are about midway in their respective navigation pools, and they thus provide the best record of changing water le\els that in- fluence wetland plants. We consider the period 15 Jiuie-12 October to be the most favorable season for the development of marsh, aquatic, and moist-soil plants. ,\n earlier period ^voulcl he still better lor ac^uatics, but the earlier the date, the less likely is low water, because of spring floods. The minimum period of dewatering (between 15 June and 12 Octolier) within which moist-soil plants can pro- duce mature seed (although not at top yields) is 70 clays. A review of river levels at Henry, 1938-1976, shows that during the period 15 Jiuie-12 October seasonal means ranged from 1.2 to 2.3 m (3.8- 7.5 ft) (Table 2). Downstream at Ha- vana during the same period, seasonal means ranged more widely, from 1.8 to 3.6 m' (5.9-11.9 ft) (Table 2). This greater fluctuation results from large iribiuary streams adding their sizeable discharges to the Illinois River between Peoria and Ha\ana. The weekly mean gauge records for 15 Jiuie-12 October were averaged for each year, and siand.nci deviations were computed. The lower the mean and its standard deviation, the more mud Hats were exposed, and the longer Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 13 Table 2.—Water level fluctuations (in meters) in the Illinois River 15 June and 12 October, 1938-1976. at two stations between 14 Ilunois Natural History Survey Bulletin Vol. 32, Art. 1 to facilitate our discussion, we use the following groups. The lakes are grouped according to their water-level stability (Classes I-IV) and by periods: 1938-1942, 1943, and 1944-1959. The highest flood on rec- ord occurred during May and June of 1943. We sought to determine the ef- fect of this tremendous flood on wet- land vegetation both in the year it oc- curred and in following years. Table 3 lists the areas covered by wetland plants in six lakes connected with the river at all stages (Class I) from 1938 to 1942. River bulrush was the most abundant species, covering an average of 7.3 percent of the lake basins; American lotus averaged 6.1 percent; and marsh smartweed and thick potato each had about 2 percent. Other wetland species were of only minor occurrence, and all species to- gether covered only 24.2 percent of the lake basins. Average areas covered by wetland plants in six lakes that were separated from the river by barriers at moclerately low stages (Class II) are shown in Ta- i)le 4. Many of the plant species in these lakes belong to an ecological com- Table 3.—The average abundance by area of wetland plants in bottomland lakes connected with the Illinois River at all water stages (Class I), 1938—1942. Figures are in hectares (2.471 acres/hectare). Toul Aver- age Hec- tares Species Saw- Aug. 1979 Bellrose et al.: Waterfowl and the Chancing Illinois Valley 15 o •- i3 4> 03 c 2 — 00 ^ en en in in en 2 ^N CT) W CO CO o '^ re -^ (?-i Cf Oi o -^ '^ O ir- O — M* flJ Cv| O ^H OO ra d <>i d d ^ CJ O QJ I-- -- ^ cn O -• ^ cj d d -^ o d fljTfHo^cooicip ;aj CC CI to ^ ; (>{ ^ p cj> Of O) o^ ! to d i> CO »ri ^ r^ -H r- CI -^ c^i to ;oC5'c*j-^i—"^-©o^-r^i—.cDtot^iot^ |p cicn coci o-^d'-Hc4.-^dddddc<^o6aicoci-^d--cic<^-<*^"inod .^ rr^citoco ID cj -" cj to CI r- — cj -^ ^ c-i -^f CO oi 1-^ -^ lO— ^ ^ TJH ft^ CT CJ Tfl (O • C30 I d to i> CI o to in p (N d lO d cj ifi CO "^ o^ c^ — "^ pd CO CO d d 00 ^ Oi --; ^p 00 -^ OO d ' -- Oi OJ O} p CO ^ d to d cri d ^ I crj -^ 04 CO ;tO';;;;;;;p;;co;-;;o^^c^;;*;^ Cfoo oo in in -^ 00 'rr "^ •"^ o d to qj(M ' r*- I *co •eoojoo ; •t^'^woo e'^»f>^I*'^.^.'*.^*»o*-oo*. ddd^^ON orito —to cj oq^^o^oo*^"'^ ^^ o ^ ! ! * ! ^ * I ! ! d ! ." ! i ^ I CO * ^ iri ^ ! ! '^ Ci »— i to r>- in o\ i> cj ;O-'00^H'*U%0O^COI!IIt!!I'I!I!iII*!oo!tdc6I^cs iri^ r* in *n oo to • to "^ 00 to ^ -^ CO o — . cTt in CO CI o 16 Illinois Natural History Survey Bulletin Vol. 32, Art. 1 Fig. 4.—Extensive mud flats exposed on the west shore of Lake Chautauqua (Class II) a few miles north of Havana. Mud flats like these are colonized by plant species that belong to an ecological community termed moist-soil plants. Fig. 5.—Lush stands of moist-soil plants at Spring Lake. These stands occurred when mud flats were exposed early during the summer growing season, 1978. Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 17 u^ 18 Illinois Natural History Survey Bulletin Vol. 32, Art. 1 munity associated with mud flats, and are termed moist-soil plants (Fig. 4 and 5). These plants include: teal grass; rice cutgrass; wild, Japanese, and Walter's millet; nmgrasses; nodding and largeseed smartweed; water hemp; cocklebiirs; and Spanish needles. They covered II. 3 percent of the lake basins in 1938-1942, and 13.3 percent in 1944-1959. In contrast, the moist-soil species present in Class I lakes (Table 3) were limited to rice cutgrass, wild millet, nntgrasses, nodding smartweed, and water hemp; they aggregated a mere I.l percent of those lake basins from 1938 to 1942. Arrowleaf covered averages of 0.8 and 1.8 percent of the Class II lake basins in 1938-1942 and 1 944-1 959, re- spectively. In both Class I and Class II lakes river bulrush covered almost the same average proportion of lake basins (7.3 and 5.3 percent, respectively), but American lotus expanded its coverage from 6.1 percent in Class I lakes to 17.1 percent in Class II lakes. Increases also occurred in the areas covered by aquatic pondweeds, southern naiad, waterweed, and coontail: they totaled 5.2 percent of Class II lake basins, 1938-1942, and 12.2 percent, 1944-1959, up from 2.5 percent present during the former period in Class I lakes. From 1938 to 1942 marsh smartweed more than doubled in abundance in Class II over Class I lakes, ijut duck potato was less abundant (probably because of competition with marsh smartweed). All wetland plants coveretl 46.8 percent of Class II lakes in 1938-1942, and 58.4 percent in 1944-1959. Class III lakes are only alTectcd by river levels when flood waters top the natural banks that lie ijetwcen lake and river. Four lakes in Class IIIA were studied during 1938-1959 (Table 5). Species of wetland plants in Class IIIA lakes were similar to those found in C;iass II lakes; moreover, the propor- tion of wetland vegetation in the lake basins was similar to the percentage (46.8) in Class II lakes during the first period (48.9 percent) but was less in the second period (36.5 percent). However, there were pronounced dif- ferences in the relative abinidance of certain species of plants between Class IIIA and Class II lakes. Pondweeds, southern naiad, and waterweed were more abundant in Class III,-\ lakes, cov- ering 7.8 percent of the lake basins in 1938-1942, and 9.6 percent in 1949- 1959. compared with 2.3 and 5.0 per- cent, respectively, in Class II lakes (Fig. 6). Coontail was more abundant in Class IIIA lakes than in Class II lakes during the first period but was less abundant in the Class IIIA lakes than in Class II lakes during the second period (Table 5). Wild rice and pickerel weed, unusual marsh plants in Illinois, were noticeably present prior to the 1943 flood. River bulrush was also proportionately more abiuidant in Class III,\ lakes than in lakes of the other classes. Moist-soil plants were almost absent (0.6 percent) from Class III.A lakes dur- ing the first period and of slight abini- dance during the second period (3.7 percent), much less than in Class II lakes. Class IIIB lakes must be de^^•atered by 1 .August in central Illinois to produce moist-soil plants consistently. Otherwise, the growing season is too short for the plants to mature and pro- duce an abuniiant seed crop. Dewater- ing b\ 15 July provides a still more fa\orablc growing perioil. Prolongetl spring floods mav make dewatering in- feasii)le in certain vcars, and mitisiun- nicr Hoods uiav wipe out moist-soil plants well along in growth. Some (loods occur so late that the time re- maining in the growing season is too short for the de\elopmcnt of another plant generation. Table 6 shows wetland vegetation areas, averaged for a number of years, at three Class 1 1 Hi lakes. Moist-soil species covered 41.7 percent of the lake Aug. 1979 Bellrose et al.: Waterfowl and the Chancing Illinois Valley 19 Fig, 6.—Abundant growth of aquatic plants, primarily coontail, present in Rice Lake (Class IIIA) during the summer of 1953 Table 6.—The average abundance by area of wetland plants in bottomland lakes separated from the Illinois River at levels below flood stage and intensively managed by dewatering (Class IIIB), 1946-1959. Figures are in hectares (2.471 acres/hectare). 20 Illinois Natural History Sur\'ey Bulletin Vol. 32, An. 1 basins, and 6.2 percent was covered by other species. Artificially sown Jap- anese millet composed nearly half of all moist-soil plants; the other species occurred naturally. Nutgrasses, rice cutgrass, teal grass, water hemp, and Walter's millet and wild millet were the principal species, listed in descend- ing order of abundance. Marsh plants—river bulrush (3.1 per- cent), arrowleaf (1.3 percent), marsh smartweed (1.0 percent), and duck po- tato (0.2 percent)—aggregated 5.6 per- cent (Table 6) in Class IIIB, a much smaller proportion of the area than they covered in other classes of lakes. Aquatic plants were almost nonexis- tent in Class IIIB lakes. Longleaf and sago pondweed (0.1 percent and trace, respectively) and American lotus (0.6 percent) totaled only 0.7 percent of the lake basins. These aquatics occurred because the lakes were never completely drained; usually from one-third to one- half of the lake basin retained a shal- low residue of water surrounded by bare mud flats and shoreward, by zones of moist-soil plants. EVALUATION OF THE SEVERANCE OF LAKES FROM THE RIVER Data presented in Tables 3-6 show that the greater the separation of the lake basin from the Illinois River, the more productive the lake was of wet- land plants. Although after the 1943 flood, the percentage of the lake basins covered by aquatic and marsh plants in Class IIIA lakes declined below levels in Class II lakes, this difference oc- curred for other reasons that will be discussed later imder water deptii and turbidity. Not only did the quantity of wetland plants increase with the degree of isola- tion from the river water, but the ()ual- ity of vegetation for waterfowl food also improved. In Class I lakes, river bulrush, American lotus, and marsh smartweed made up 64 percent of the wetland plants, but their production of waterfowl food was minute (Low & Bellrose 1944). River bulrush produces a negligible amount of seeds; American lotus pro- duces a moderate seed crop, but its seeds are too hard for ducks to eat dur- ing the fall (Bellrose &; Anderson 1943); and marsh smartweed fails to produce seed when growing out of water (Low k Bellrose 1944), as it does in most Class I lakes. Class II lakes produced more moist- soil plants than did Class III.\ lakes, but Class IIIA had a greater abundance of aquatic and marsh plants. The maximum waterfowl food plant pro- duction occurred in Class IIIB lakes because of the extensive development of moist-soil vegetation resulting from controlled drawdowns of water. Most moist-soil plants yield large crops of seed (Low &: Bellrose 1944) that are preferred foods of most dabbling ducks (Bellrose &: Anderson 1943). Plants most typical of all classes of Illinois Valley lakes are river bidrush, American lotus, and marsh smartweed. These species in their respective niches are obviously the most adaptable to a regimen of fluctuating water levels. Within each class of lakes, the abun- dance of wetland plants changes from year to year as the water level fluctu- ations vary. EVALUATION OF SEASONAL CHANGES IN BOTTOMLAND LAKE WATER LEVELS Effects on Moist-Soil Plants A comparison of the yearly Moist- Soil Water-I.evcl Index (described un- der Methods) and the development of uioist-.soil plants in the lake basins of tlie lower Illinois River valley is shown in Fig. 7. (At the time of the study, nuul Hats were not extensively exposed in the \alley above Peoria because the Peoria lock and dam had raised water levels.) As the yearly Moist-Soil Water- Level Index rose, the proportion of all Aug. 1979 Bellrose et al.: Waterfowl and the Chancing Illinois Valley 21 R = 0.62 r =0.79 0.07 0.06- 0.05- 0,01- 0.03 0.02- 0,01 200 I 600 1 1000 I WOO I 1800 I 2200 I 2500 too 800 1200 1600 2000 2400 MOIST-SOIL WATER-LEVEL INDEX Fig. 7.—The linear relationship of the Moist-Soil Water-Level Index to the abun- dance of moist-soil plants per hectare of lake basin in the Illinois River valley. lake basins covered with moist-soil plants likewise increased. The coef- ficient of determination (R^ = 0.62) in- dicates tliat about 60 percent of the annual coverage of moist-soil plants was related to the Moist-Soil Water- Level Intlex. Small rises in water during the growing season of moist-soil plants destroyed extensive areas of these plants, the extent depending upon the lieight of the rise and the lateness of the season. The havoc that these mid- sununer fluctuations wrought on the development of moist-soil plant beds is only partially measured by the Moist- Soil Water-Level Index and is the ])rincipal reason that it accounted for only 60 percent of the plants' coverage of lake ijasins. Once eradicated by a brief inundation, moist-soil plants might not have sufficient time to regen- erate, or mud flat areas might be re- duced. Effects on Aquafic and Marsh Plants Through the years, 1938-1959, only a small coefficient of determination oc- curred between the area covered yearly by aquatic and marsh plants and the yearly variation in the seasonal fluc- tuation of water levels. The Water- Stability Index for seasonal stability on all lakes studied compared with the development of marsh vegetation gave a coefficient of determination of 0.11 (11 percent of the variable accounted for by fluctuating levels); for aquatic plants it was 0.108 (10.8 percent). Two items were proposed for further analysis: (1) Water levels had various effects on aquatic and marsh plants in different lakes that we attributed to difl^erent elevations and bottom topog- raphies of the lake basins. (2) Factors other than the degree of water-level stability appeared to be more impor- tant in regulating the growth of aquatic and marsh plants. To evaluate further the effects that yearly variation in the fluctuation of water levels had on aquatic and marsh plants, we examined particular Class IIIA lakes more closely. Lake Chautauqua.—One of our longest series of aquatic plant studies was made at Lake Chautauqua, a national wildlife refuge. The 1,416-ha (3,500-acre) lake is enclosed by levees that minimize low-water fluctuations. A spillway at 133.4 m (437.5 ft) msl permitted river waters above that Iieight to enter the lake during the years of the study. Unfortunately, ref- uge gauge records are available for only 8 of the 14 years studied. The Water-Stability Index for those 8 years was compared to the occurrence of sago pondweed by linear regression. The coefficient of determination was 0.31, suggesting that water stability ac- counted for aboiu 30 percent of the yearly change in abundance of sago pondweed. American lotus, coontail, river bulrush, and duck potato showed 22 Illinois Natural History Survey Bulletin Vol. 32, Art. 1 only a slight correlation between their abundance and the Water-Stability Index. The highest flood on record in the Illinois River valley occurred in late May 1943. At Lake Chautauqua the flood crested on 25 May, 5.3 m (17.5 ft) above the normal lake level. It inundated other Class IIIA lakes to similar depths. Vegetation studies were made in late summer after the water level had returned to normal. The flood eradicated all pondweeds from Lake Chautauqua (Table 5); coontail declined greatly in 1943 and never recovered in subsequent years; marsh smartweed and river bidrush also suffered losses in abundance that were never regained. American lotus increased in abundance but subse- quently died out in the early 1960's. Rice Lake.—At Rice Lake, coontail abiuidance increased markedly in 1943, apparently because of less competition from American lotus (Table 5). The high water lessened the area of river bidrush on lower sites, enabling marsh smartweed to invade stands of bulrush and increase in abundance. Flood wa- ters submerged the growing plants of lotus so deeply that they failed to re- cover after the water subsided, perhaps enabling white water lily to increase. Cuba Island.—Flood waters adversely aff^ected the abundance of rice cut- grass, coontail, and American lotus at Cuba Island (Table 5). Marsh smart- weed, however, increased as competi- tion from other plants decreased. Flat, Swan, and Gilbert Lakes.— Flat, Swan, and Gilbert lakes, part of the Calhoun Division of the Mark Twain National Wildlife Refuge, lie in the lower Illinois River valley a few miles above the river's confluence with the Mississippi River. Several relatively small lakes oc- curred there before 1939. At that time the Alton navigation dam raised water levels, increasing the lakes' surface area from 445 to 1,943 ha (1,100-4,800 acres). By 1941, wetland plants had developed in a sizeable area where lit- tle but lotus had grown before the expansion of Flat, Swan, and Gilbert lakes in 1939 (Table 7). However, the flood of 1943 caused a marked reduction in wetland vegeta- tion, and the plants appeared late in the growing season after the flood had subsided. The comeback of wetland plants was retarded in 1944 by another high flood in the late spring. During most of the 1960's, sago and longleaf pondweeds, southern naiad, and coontail flourished in Swan Lake. Beds of these acjuatics aggregated 405 ha (1,000 acres) in 1965, but a late September flood reduced their fall availability to waterfowl. A high flood in April and a minor one in June Table 7.—The abundance by area of wet- land plants at Flat, Swan, and Gilbert lakes. 1941-1944. Figures are in hectares (2.471 acres/hectare). Species Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 23 1967 resulted in turbid water through much of the growing season, limiting the growth of submerged aquatics but not that of American lotus or duck po- tato. Stable water levels prevailed dur- ing the growing season and fall of 1968, promoting an excellent growth of sago and longleaf pondweeds, south- ern naiad, and coontail. Three floods occurred in 1969: spring, July, and October. After the spring flood, beds of submerged aquatics and American lotus began to ap]5ear, only to be destroyed by high, turbiil water in July. In spite of favorable water levels in 1971 and 1976, acpiatic plants did not reestablish themselves in Flat and Swan lakes. Large depositions of new sediments were resuspended by wave and fish ac- tivity, greatly increasing turbidity. Yearly fluctuations in water levels liave an influence on aquatic and marsh plants, but such fluctuations are less important to the welfare of tiiese plants than they are to moist- soil species. It is also apparent that factors other than water-level stability have influenced the growth of aquatic and marsh plants. Therefore, we eval- uated several other factors. CHANGES IN WATER DEPTHS Semipermanent changes in water deptji profoundly affected the vegeta- tion at several Illinois Valley lakes: Rice, Douglas, Spring, and Anderson. Rice Lake In 194.5 an earthen dam was con- structed at the outlet of Rice Lake, increasing its normal minimimi depth l)y 0.4 m (1.3 ft). In 195.3 the dam was raised an additional 0.5 m (1.5 ft) and another 0.4 m (1..3 ft) in 1961. These increases in water level at Rice Lake severely altered its plant communities. With a depth increase of 0.4 m (1..3 ft) in 1945, river bulrush de- clined from covering over 33 percent of the basin to merely 1.7 and 1.0 per- cent in 1950 and 1953, respectively (Table 8). With low water levels in 1977 and 1978, aboiu 10.1 ha (25 acres) of river bulrush have become reestab- lished. The abiuidance of marsh smart- weed varied from year to year with spring water conditions; it covered a larger area in 1955-1956 after the wa- ter was raised an additional 0.5 m (1.5 ft) during 1953. Marsh smartweed in- vaded areas formerly occupied by river bidrush but could not tolerate the ris- ing water and disappeared in the 1960's. American lotus declined in abun- dance from 15.1 percent of the Rice Lake area in 1944 to a trace in 1956 antl none in 1957. Small beds of lotus reappeared in 1977 and 1978 as water levels were lowered. In the 1950's, coontail markedly in- creased in area as a result of the deeper water (Table 8). It invaded areas va- cated by declining beds of lotus and river bulrush. However, the disap- pearance of river bidrush and lotus increasingly exposed a greater expanse of open water to wave action. Waves resuspend highly organic bottom ma- terials, causing the waters to become much more turbid than before. As a residt of increased turbidity, coontail vanished from Rice Lake in the mid- 1960's and lias not become reestab- lished. White water lily increased in abun- dance as competition from lotus less- ened and as water levels remained more stable. However, it could not tolerate the increasing wave action and turl)idity of the early 1960's and even- tually disappeared from the lake. Douglas Lake Witii the inauguration of the Peoria lock and dam in December 1938, the minimum water level at Douglas Lake was raised aliout 0.6 m (2 ft). At that time the area supported the largest river bulrush marsh in the state (364.1 ha, 899.7 acres; Table 9). Over the 24 Illinois Natural History Survey Bulletin VoL 32, Art. 1 CN 3 o> in 1 oo CO a a c a u c n •o c D Percent of Lake Basin Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 25 ; Q 5 o. o X 2 u «- cq &H o 6 a K 2 ffi 2 c ^ c On O S 2 X 2 u ^ pg CL, o S 2 • OO -^ CM C^ lO ; o o o o to t-j iO ; 00 eo (M I ^ M o; ; p —1 ens because of the greater standing column of wa- ter from which sediments precipitate. The exceptionally high correlation co- efficients (r) confirm this relationship. Upper Peoria Lake has a lower sed- imentation rate than the other lakes because the river transports more sed- iment through it than through lateral lakes. At 2.1 m (7 ft), the yearly rate of sedimentation is 0.01 m (0.056 ft) as compared with 0.024-0.027 m (0.078- 0.090 ft) in the other lakes. The lateral lakes are inundated during floods. Re- duced current velocity in these lakes allows more sediment to precipitate than can drop out at the higher ve- n2 locities present in LTpper Peoria Lake. There is amazingly little difference among the sedimentation rates in lat- eral lakes above Peoria. Fig. 9 indi- cates that at a water depth of 0.3 m (1 ft) in Upper Peoria Lake wave and ice action have an erosive effect on the shallow margins of the lake bed and transport bed material to other areas. The same action has affected the shore margins of some of the lateral lakes. Below Peoria, sedimentation studies were made at Rice Lake, Lake Chau- tauqua, Anderson Lake, and Meredosia Bay (Fig. 10). These lakes have lower and more diverse sedimentation rates than the upper lakes. The differences in sedimentation rates among lakes can be accounted for by their relative isolation from the river. The lakes above Peoria have only low, narrow peninsulas of land separating them from the river at lev- els below flood stage. All of the lat- eral lakes studied above Peoria have = SENACHWINE, R n2 +0.050 -1 +0.045 ^ +0,040 UJ 5 +0,035 ~ +0.030 CD S +0.025 -0.005 -0.010 • = SAWMILL, R = BILLSBACH, =SPARLAND, R A = WIGHTMAN, R O = BABBS SLOUGH, R = UPPER PEORIA LAKE 0.83, r = 0.91 < 0.01 0.98, r = 0.99 < 0.01 0.98, r = 0.99 < 0.01 = 0.96, r = 0.98 < 0.01 0.84, r = 0.97 * 0.01 0.95 < 0.01 0.92, r = 0.96 < 0.01 0.90 2 Fig. 9.—Linear relation- ships between the depth of water and the yearly rate of sedimentation in seven bot- tomland lakes in the upper Illinois River valley between Hennepin and Peoria. T r 1 1 1 1 '0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 DEPTH OF WATER IN METERS 31 0.51, r = 0.71 -: 0.01 = CHAUTAUOUA. 1926-1976. R^ = 0.72, r = 0.85 ' 0.01 _ ANDERSON, 1903-1977, R^ = 0.56, r = 0.75 < 0.01 = MEREOOSIA, 1903-1978, R^ = 0.76, r ' 0.87 < 0.01 Fig. 10.—Linear relation- ~ ships between the depth of 2 water and the yearly rate of S sedimentation in four bot- g tomland lakes in the lower E Illinois River valley between l^j Banner and Meredosia. u- -0.010 -I broad inlets-outlets at their lower ends, connecting tiieni to the river at all stages. Hence, they are subject to sed- imentation from river waters at all stages above the minimum. The de- gree of separation of lateral lakes from the river is greater below Peoria, where sedimentation rates among the lakes studied are lower and more variable. In this group, Rice Lake is the most isolated from the river, and it has the lowest sedimentation rate of all lakes studied. Anderson Lake is protected at its upper end by a large agricultural levee and on its river side by a natural levee 1.8 m (6 ft) above normal river level. Meredosia Bay is partially iso- lated from the river by an agricultural levee at its upper end and by a broad lateral peninsula of land that permits little ingress of river water until the river has risen 3 m (10 ft). Although surrounded by levees, Lake Chautauqua spillway elevations are 133.4 m (437.5 ft) msl, 2.2 m (7.3 ft) above normal river level. However, water was let into the lake through gates when pro- jected river rises threatened to overtop the spillways; in recent years erosion has reduced the spillway level to 132 m 1 I I I I I 0.5 1,0 1.5 2,0 2.5 3.0 3.5 4.0 DEPTH OF WATER IN METERS (433 ft) msl. Once river water enters, it is trapped in Lake Chautauqua to a greater degree than in any natural lake; consequently, the sedimentation rate is higher than at Rice and Ander- son lakes. Data previously obtained at Upper Peoria Lake, Lake Chautauqua, and Meredosia Bay by the U..S. Army Corps of Engineers, the Illinois State Water Survey, and the State Division of Wa- ter Resources, and data obtained by us for 1976 ami 1978 enabled us to cal- culate linear regressions for the two groups of years to determine periodic sedimentation rates (Fig. 11, 12, and 13). Sedimentation rates were appre- ciably liigher in the more recent period at all three lakes. The rise in sedimentation rates in the Illinois River valley in recent years indicates a pronounced increase in the sediment load of the river. The sed- iment load appears to have increased as a result of the intensified growing of row crops on the waterslied of the river and an increase in tributary stream bank erosion. In the early 1900's, lakes in the Illi- nois Valley had bottom profiles that Illinois Natural History Survey Bulletin Vol. 32, Art. 1 y-int. slope i" 1903-1956 -0.2599 0.01129 0.93' 1957-1978 -0.1466 0.01665 0.a5 : 0.5 1,0 1.5 2.0 2.5 3.0 3.5 1.0 DEPTH OF WATER IN METERS Pig 11.—^The linear relationship between the yearly rate of sedimentation and the depth of water in Upper Peoria Lake in each of two time periods, 1903-1965 and 1966-1976. = 1926 - 1950, R^ = 0.41, r = 0.64 < 0.01 = 1951 - 1976, R^ = 0.50, r = 0.71 < 0.01 +2.5 u- >- u_^ +2,0 - O CXI I— UJ , r §a +1.0 i +0.5 - -I—*—I— I 0,5 1.5 2.5 DEPTH IN METERS Pig 12.—The linear relationship between the yearly rate of sedimentation and the depth of water in Lake Chautauqua in each of two time periods, 1926-1950 and 1951-1976. presented a diversity of elevations (Fig. 14 and 15). By the mid-1970's the bot- tom profiles had flattened into platter- shaped basins because of the greater deposition of sediment in the deeper waters. Because sedimentation is dy- namic, the amount of silt deposited 0.5 1.0 1.5 2.0 2.5 3.0 DEPTH OF WATER IN METERS Fig 13.—^The linear relationship between the yearly rate of sedimentation and the depth of water in Meredosia Bay in each of two time periods, 1903-1956 and 1957-1978. in a backwater lake changes as lake depths change. Other factors (such as the sediment load carried by the nver) being equal, the total amount of sed- iment deposited per year in backwater lakes lessens as the lakes becotne shal- lower. Inasmuch as lake beds have different elevations, resulting in differ- ent degrees and time spans of inunda- tions, sedimentation rates differ from lake to lake at comparable depths. Table 13 shows the total amount of sediment deposited in 11 bottomland lakes during this century. The lakes that were originally deepest have re- ceived the greatest amounts of sed- iment. The total fill is staggering, leaving most bottomland lakes with shallow basins that are rapidly being filled as sediments "rain down'" from the waters above. Because of us depth. Upper Peoria Lake in all like- lihood will be the last lake remaining in the Illinois Valley. Rice Lake prob- ably will fill more slowly than the other lateral lakes as a result of its dis- tance from the river and the relatively high elevation of its basin. Sawmill. Billsbach, Sparland, Wight- man, Babbs Slough, and many other Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley = 1903 WIGHTMAN LAKE -, TRANSECT A ly/b RIVER MILE 188.2 . . . = IQJ-J WATER SURFACE = 440.05 msl 33 0.5 1,0 1.5H 2.0 2.5 EAST SHORE 100 200 300 1400 500 600 700 METERS 800 gdo WEST SHORE fig. 14.—A cross section of each of three lakes — Wightman, Babbs Slough, and Rice—delineating the bottom contours in 1903 and in 1976 or 1977. VBABBS SLOUGH TRANSECT C RIVER MILE 185.2 WATER SURFACE 440.0 msl 200 300 "iioo" METERS -T- "T" 500 600 700 WEST SHORE 0.5 1.0 1.5- 2.0- EAST SHORE RICE LAKE TRANSECT D RIVER MILE 137. WATER SURFACE = 200 too ~600~ 100 300 500 700 800 1000 1200 WEST I I SHORE 900 1100 METERS Fig. 1 5.—A cross section of Senachwine Lake, show- ing its bed in 1903 and 1976, and a cross section of Upper Peoria Lake, show- ing bottom elevations in 1903, 1965, and 1976. The deep trench in Upper Peoria Lake is the channel of the river. At Senachwine and other bottomland lakes, the river channel is lateral to their basins. SENACHWINE LAKE TRANSECT B RHER MILE 201.5 UATER SURFACE = 440.05 BSl "A BOTTOM ELEVATION -1903 = 1976 V /^^ EAST SHORE 0.5 1.0 1.5-1 2.0 2.5- 3.0 3,5 4.0 H.5 5,0 5.5 I 100 I 800 i 1200 I 1600 I 2000 I 2100 west 200 600 1000 WOO 1800 2200 ^"""^ METERS UPPER PEORIA LAKE TRANSECT D RIVER HILE 172.1 WATER SURFACE = 440.0 msl -- 1903 • = 1965 = 1976 BOTTOM ELEVATION I 3201EAST I 100 I 800 I 1200 I 1600 I 2000 I 2100 I 2800 I 3200 west SHORE 200 600 1000 1100 1800 2200 2600 3000 ^"""^ METERS 34 Illinois Natural History Survey Bulletin Vol. 32, An. 1 Tabte 13.—The total sedimentation, annual sedimentation, and average water depth in 1976 of 11 bottomland lakes in the Illinois River valley. Sediments and depths are recorded in centimeters at normal water levels. Lake Span of Lake Measurement Surface ofSed- Elevation" imentation msl in in Years meters Total Sedimen- tation'' in Centimeters Average Sedimen- tation Per Year in Centi- meters 1976 Average Water Depth in Centi- meters Senachwine Aug. 1979 Bellrose et al.: Waterfowl and the Changing Illinois Valley 35 variables affecting the local abundance of waterfowl. Moreover, as will be dis- cussed, a regression analysis of the an- nual availability of natural food re- sources is difficult. It is almost impossible to quantify seed yield, palatability, and availability of waterfowl food plants. A study of seed and vegetative yield by Low k Bellrose (1944) illustrated a wide diver- sity among species of wetland plants. The yield of any one sp>ecies may vary from year to year, depending upon the growing period and competition from weed species. Fall floods may make seeds and other vegetative parts un- available to dabbling ducks. Seeds pro- duced in one year by moist-soil plants may still be available in bottom soils in the following year for ducks to feed upon at times when current moist-soil plant development is at a minimum. Mallards obtain a large proportion of their food from the waste corn left after harvest (Anderson 1959) and are, therefore, only partially influenced by the availability of naturally occurring foods. Many diving ducks, especially the lesser scaup, are influenced more by the availability of animal than plant food. Local duck populations are also influenced by yearly fluctu- ations in the continental and flyway populations of the various species. These complex variables, combined with the failure of vegetation maps adequately to sample waterfowl food plant availability, make it difficult to compare the abundance of wetland plants with that of waterfowl. Never- tiieless, we have found a relationship ijetween the abundance of wetland plants and the abimdance of some spe- cies of waterfowl in the Illinois Valley. Fig. 16 shows a significant correlation between the abundance of moist-soil and marsh plants and the number of pintails. Anderson's (1959) study of the food preferences of the pintail in the Illinois Valley pointed up the primary importance of the seeds of moist-soil CO ca a: >- ca 125 -, 100- 75- 50- 25- R = 0.45 r = 0.67 P< 0.05 1 1 1 1 0.025 0.050 0.075 0.100 0,125 MOIST- SOIL AND MARSH PLANTS PER HECTARE OF LAKE BASIN I 0.10 I 0.20 I 0.05 0.15 0.25 ALL WETLAND PLANTS PER HECTARE OF LAKE BASIN Fig. 16.—The linear relationship between Fig. 17.—The linear relationship between the yearly abundance of pintails and the the yearly abundance of American green- yearly abundance of moist-soil and marsh veg- winged teals and that of all wetland plants etation in the Illinois River valley. in the Illinois River valley. 36 Illinois Natural History Survey Bulletin Vol. 32, Art. I plants and the secondary importance of marsh plants in the diet of this species. Green-winged teal numbers were in- fluenced by the relative abundance of all wetland plants (Fig. 17) rather than CO CQ 3 0.50 0.71 0.05 UJ a. oo >-