Wildlife Monographs

WILDLIFE MONOGRAPHS

A Publication of the Wildlife Society

THE ECOLOGY OF A SPRING STREAM DOE RUN, MEADE COUNTY, KENTUCKY Ly W. L. MINCKLEY

SEPTEMBER1963 No. 11 I ___ . . C::". . ...;...... #...... ".e'IP__._" , eZ'"-.,..... :.l,.~..~.. ~;-~i-.

FRONTISPIECE. Spring source of Doe Run, Meade County, Ky., April 24, 1960. Photograph by Charles B. Stone. THE ECOLOGY OF A SPRING STREAM DOE RUN, MEADE COUNTY, KENTUCKY'

W. L. Minckley2 Department of Biology, University of Louisville, Louisville 8, Kentucky

CONTENTS INTRODUCTION ------6 ACKNOWLEDGMENTS ------6 DESCRIPTIONOF THE STUDY AREA ------6 Geographic, Geologic, and Historic Information ...... 6 Description of the Stream ------9 METHODS ------12 PHYSICALENVIRONMENT ------14 Characteristics of Flow in the Mainstream ------14 Suspended Solids and Sedimentation ------17 Temperature Relations ------19 Incident Light ------23 CHEMICALENVIRONMENT ------.------23 Dissolved Oxygen ...... ------23 Carbon Dioxide, pH, and Alkalinity ------26 Total Iron ------28 Phosphate Phosphorus ------29 Nitrate and Nitrite Nitrogen ------. ------29 THE BIOTA ------30 Algae and Higher Plants ------30 Attached Rheophilic Algae ------31 Encrusting Rheophilic Algae ------32 Algae of Quiet Waters and Slow Currents ------37 Subaerial Algae ------39 Emergent Higher Plants ------40 Submergent Higher Plants ------41 Spatial Relations of the Flora of the Upstream Area ----- 43 Invertebrates ------46 Observations on the Ecology of Abundant Invertebrates ---- 48 Quantitative Aspects of the Invertebrate Fauna ------58 Fishes and Other Vertebrates ------74 Annotated List of Fishes ------75 Summary of the Distribution of Fishes ------90 Vertebrates Other than Fishes ------94 Food Relations of the Fauna ------95 Invertebrates ------.-. - 96 Foods of Selected Fishes ------99 Trophic Structure ------.------101 SYNTHESIS ------104 Climaxes in the Vegetation of Streams ------104 Assemblages of Aquatic Invertebrates ------107 The Role of Fishes ---.------110 CONCLUDINGREMARKS ------110 SUMMARY ------.------112 LrERATURE CITED ------115

1 Contribution on No. 64 (New Series) from the Department of Biology, University of Louisville, Louisville 8, Kentucky. 2 Present Address: Department of Zoology, Arizona State University, Tempe, Arizona. 6 WILDLIFEMONOGRAPHS

INTRODUCTION from the area. Louis A. Krumholz, Uni- versity of was available at all In recent an amount of Louisville, years increasing times for discussion of problems that arose research has been directed toward the in the course of study, and his ideas and of because ecology flowing waters, mainly suggestions were factors in comple- of their in the of in- major importance disposal tion of this report. Jane S. Davis, Virginia dustrial and domestic wastes. Much of the Institute of Marine Science, prepared most research on streams in North America has of the figures. been oriented toward various of aspects assisted in identification of the search for indices of Many persons pollution control, invertebrates from Doe Run: Gerald A. pollution biological and physico- (both Cole, Arizona State University, and E. L. and toward the management of chemical), Bousfield, National Museum of Canada, streams for sport or commercial fisheries. identified malacostracans; William E. many problems that now confuse However, Ricker, Fisheries Research Board of Can- findings of applied research may be solved ada, Plecoptera; Herbert H. Ross, Illinois only after accumulation of basic data. Natural History Survey, Trichoptera; Lewis This paper concerns an unpolluted spring Berner, University of Florida, Ephemerop- stream, somewhat modified by man, and tera; and Stuart E. Neff, Virginia Polytech- is an attempt to describe and analyze its nic Institute, Diptera. Assistance in identi- physical, chemical, and biological char- fication of algae and higher plants was acteristics. Investigations of Doe Run, provided by Arland T. Hotchkiss and Don- Meade County, Ky., were begun in Feb- ald R. Tindall, University of Louisville. ruary and were intensified November 1959, This is dedicated to Bar- 1, 1959. A of was paper my wife, year study completed who assisted in in its before the downstream half of the creek bara, many ways completion, and who deserves special was prepared for impoundment, and in- recognition. tensive sampling ended on July 9, 1961, when impoundment was effected. DESCRIPTION OF THE STUDY AREA ACKNOWLEDGMENTS Geographic, Geologic, and Historic This report is based on research per- Information formed under Contract No. AT-(40-1)- The main spring of Doe Run (Front.) 2595 between the U. S. Atomic Energy lies about 3 miles east and 0.4 mile north Commission and the University of Louis- of Ekron, Ky., in eastern Meade County, ville, with Louis A. Krumholz as principal 37? 56' N and 86? 07' W (Fig. 1). The investigator. It is a revision of parts of a creek flows north-northeast and is 9.7 miles dissertation submitted to the graduate long, although the distance from source to faculty of the University of Louisville in mouth by airline is only 5 miles. Doe Run partial fulfillment of requirements for the enters the Ohio River about 3.5 miles east degree of Doctor of Philosophy. of Brandenburg, Ky. Many persons assisted in the various The stream is located on a belt of Missis- phases of study and I am grateful to all of sippian limestones that exhibit extensive them. Special mention should be made of karst topography and extend from central my fellow students, both graduate and un- Indiana through Kentucky and into central dergraduate, who gave their time and Tennessee (Swinnerton, 1942). In Ken- energy freely in the field and in the labora- tucky, this well-defined topographic area is tory. The landowners along Doe Run are called the Pennyroyal and is considered a to be given special consideration and major physiographic region of the state thanks. The Kentucky Department of Fish (Sauer, 1927). and Wildlife Resources granted the neces- The uplands into which Doe Run has sary permits for collection of vertebrates deeply incised lie approximately 680 feet ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 7 Olt,o

ile<9-~''">?>^ss^"^s ONE M I L E

N Station V _%Vb^^-> .-< Stream mlles Bridge Marshes or ponds C ) . oa Springs - Permanent streams

!/ ,.------Intermittent streams /IVf ,^4 | Towns A Weather stations

,(V d 40-

[38-

n IV

90 88 86 84 82 lma , \ -,~IIIc Insert A

' Station IlI I'. lt b

Mile 3

. Dam 3 Il \ia Dam 2 , \ * Station i Mile 2 lb 7 L ouisville

JEFFERSON COUNTY mileClt I Wolf 4

fotalo f') (r= 5 Brandenburg J Blue Spring Spring MEA D E BUDEULLITT COUNTY

/ Moin Spring \I \ CCOINTY Godman lb (_ (Source)el^ f Ivntn " \ u Field

\^ / HARDIN A / StatElizobthtown /

k \ / C5)COUNTY Insert B

\ Rushing River

\ .e \Cr FIc. 1. Map of Doe Run, Meade County, Ky., showing tributaries, creek miles, mill dams, collecting stations, and other features. Insert maps show location of the study area in Kentucky (A), and certain geographic features of Meade and adjacent counties mentioned in the text (B). 8 WILDLIFEMONOGRAPHS above the mean sea level (msl) near the structural joints of the limestone, and source; higher knobs rise to about 950 feet probably developed along a solution chan- 1.5 miles southeast of the source. The land nel. The downstreamsegment, where the surface slopes gradually toward the north streamflowed on the surface of Pleistocene to about 600 feet above msl at the tops of and Recent sediments, assumed a typical precipitous bluffs that border the Ohio meanderingchannel. River floodplain. Compared with some Prior to settlement of Meade County by parts of North America, this area shows white men, the area was used by Indians as little evidence of post-Mississippiansub- a major hunting ground (Ridenour, 1929). sidence (McFarlan, 1943). A drainagepat- The first settlers to explore the northern tern of northwesterly flowing streams Pennyroyal found dense forests along the probablywas formed in late Tertiarytimes, streams and on the knobs, and a prairie after reelevation in the early Tertiary upland that they called the "barrens"be- Period had stripped off nonmarine sedi- cause of its general lack of woody vegeta- ments deposited on the Mississippiansur- tion. These prairies appear to have been faces. The Ohio River formed in the early maintained as a fire disclimax by periodic Pleistocene, essentially in its present con- burning by Indians, and became covered figuration, by glacial damming of north- with woody plants soon after burning was flowing streams and subsequenttopping of discontinued (Allen, 1876; Sauer, 1927). cols toward the south and west by the The first settlements in Meade County ponded waters (Leverett, 1929; Netting, were near Doe Run and Otter Creek which 1956; Walker, 1957). The glacial waters have high gradients and thus provided a claimed preexisting stream channels, the source of power. Between 1780 and 1825, bedrock channel of the Ohio River was cut 2 water grist mills and a sawmill were to essentially its present level, and widen- established on Doe Run, and a large tan- ing of the valley was accomplishedin the nery was built (Ridenour, 1929). These Yarmouthian (Walker, loc. cit.). Borings industries necessitated modification of the made in the valley of Doe Run by the Doe creek by constructionof dams, 3 of which Valley Corporation,Brandenburg, Ky., in- are extant. After 1860, wells were drilled dicate that its rock floor lies about 320 feet to obtain brine from underlying strata for above msl, thus correspondingto the gen- the production of salt. Natural gas asso- eral elevation of the rock floor of the Ohio ciated with the brines allowed limited ex- River near Doe Run, 310 feet above msl pansion of industry in the Doe Run area (Leverett, 1929). Doe Run was a tributary (Jillson, 1922), but gas productiondeclined to the Ohio River either at the time of rapidly after 1900 and the fields were rapid downcutting prior to the Illinoian, closed. In the early 1930's, parts of Doe or no later than the Tazewell Glacial Sub- Run were used as recreationalareas. One stage of the Wisconsin Glacial Advance, of the mills, built in 1820 and operative when much of the Illinoian alluvial fill was until 1930, was converted to an inn and removed from the channel of the Ohio restaurant,and the area aroundthe inn was River (Walker, 1957). used for recreationalpurposes during 1959- Doe Run appears to have originated 1962, as were all places where roads pro- from a subsurface drainage channel. The vided access to the stream. main spring issues from a foundered cave The major alterationof Doe Run, during in the St. Louis Limestone. Probably the and following this study, was the prepara- upstream 1.5 miles of stream now flows tion for and impoundment of the lower through a "valley sink," the formation of creek by the Doe Valley Corporation.The which was discussed by Sauer (1927) for effects of clearing the riparian forest and the Pennyroyal area. Between Miles 1.5 of earth moving along the banks will be and 5.0 (Fig. 1) the creek seems to follow discussed more fully below. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 9

Descriptionof the Stream TABLE 1.-SOME PARAMETERS OF WIDTH, DEPTH, AND POOL-RIFFLE DEVELOPMENT IN DIFFERENT A map of Doe Run was made from an SECTIONSOF DOE RUN, MEADE COUNTY, KEN- aerial photograph and by reference to TUCKY. DATA MARKED WITH AN ASTERISK ( * ) ARE topographic maps of the Guston, Rock FOR IMPOUNDMENTS, BACKWATERS, OR SECTIONS OF SMOOTH, ALMOST-LAMINAR FLOW Haven, and Laconia (Indiana) quadrangles of the U. S. Geological Survey. Creek miles Maximum Maximum Mean Estimated were measured with Mile 0.0 Size of Depth of Depth of Riffle- beginning MileMI e SectionSec tio. Pools Pools Riffles Pool at the source and terminatingwith Mile 9.7 ( feet) (feet) (inches) Ratio at the mouth. Tributaries were assigned 0.0 -1.1* 35 X - 3.0 12.0 1-75 letter designations consecutively down- 1.1 -1.65* 40 X - 7.0 none 0- 1 stream from each of the 5 permanentsam- 1.65-2.1 45 X 60 5.0 6.0 1- 1.5 pling stations 1), so that each 2.1 -2.36* 50 X - 7.5 none 0- 1 (Fig. tributary 2.36-2.82* 60 X- 8.0 4.0 1-100 was known by the number of the sampling 2.82-5.3 35 X 85 7.0 6.0 1- 1.5 station closest upstream to it, and by suc- 5.3 -9.71 30 X 350 12.0 9.0 1-100 cessive letters of the If alphabet. a tribu- This section includes the long, slow-flowing portion of tary had a local name, that name was used Doe Run adjacent to the Ohio River (Miles 7.7 to 9.7) which in preference to the numberingsystem. is best considered as a backwater or pool habitat. The majorsource of Doe Run is 575 feet were developed, with extensive eddies and above msl, and the stream enters the Ohio quiet areas along the banks. River at 374 feet; the average gradient is In the upstream1.5 miles, the floodplain 20.7 feet per mile (fpm). Streamsflowing on was relatively broad (100 to 200 yards in limestones often have rapidly changing width), between Miles 1.5 and 5.0 it was gradientsin differentsections (Neel, 1951), narrow or absent and the stream flowed and this is exemplified by Doe Run (Fig. between precipitous bluffs of varying 2). height, and downstreamfrom Mile 5.0 the Doe Run is about 30 feet wide through- floodplain widened as a result of the allu- out its length, narrowing to as little at 6 vial fill, stretching 100 to 300 yards on feet at some riffles in the downstreamarea, either side of the creek. and widening at some large pools to more There was a fringe of trees along most of than 50 feet (Table 1). Generalrestriction the banks of Doe Run especially in areas of the creek to a 30-foot channel, and the of wide floodplain because of clearing and preponderanceof bedrock or marl bottoms cultivation of the "bottomlands." There in the upstream areas permitted little pool were few large trees in the area, and formation, and current velocities upstream stumps and slash implied heavy logging in from Mile 5.3 rarely were less than 0.5 previous years. ft/sec. Exceptions to this were found in Five permanentstations were established the impoundmentsof the 3 mill dams. In on Doe Run at the inception of the study. the lower part of the creek, typical pools Station I was located at the source and ex-

.00 tended about 0.1 mile downstreamfrom the main spring (Front.; Fig. 1). The stream sw below the spring was turbulent for about 520 DAMIII 60 feet where it graded into a long section $20 (Miles 0.03 to 1.1) characterizedby gener- \O-- tl ally smooth, almost laminar flow, and a '440 z bottom of sand and small gravel overlying : 400' bedrock (Fig. 3; Table 2). Between Miles

a ...... I ' g 1.1 and 1.65, the streamflowed throughthe T REAM M I L E S impoundment of the most upstream mill FIG. 2. Longitudinal profile of Doe Run, dam (Dam 1). Between Miles 1.65 and 2.1 Meade County, Ky. the creek fell rapidly over many cataracts 10 WILDLIFE MONOGRAPHS

FIG. 3. Doe Run, Meade County, Ky., at Mile FIG. 4. Doe Run, Meade County, Ky., at Sta- 0.17 looking downstream, November 11, 1960. tion II near Mile 1.75, January28, 1962. Note bed of watercress along left bank and rela- tivphrVV.Z thir-kiM.z..,z rinnrinn &uzL.fnrp.vtc &jp,az&az thirds was almost entirely a dark, malodorous muck and silt, apparently resulting and swift riffles Station II was (Fig. 4). from decomposition of leaves and other established near Mile but collections 1.9, organic debris. Downstream from Dam 3 made the swift section of creek throughout (Mile 2.82) the creek flowed 2.7 miles over were considered from Station II except marl, bedrock, and gravel-rubble bottoms where there were differences pronounced (Table 2), with mud deposits in some large in fauna. At Mile the creek entered the 2.1, pools. Station III was in the marl section held Dam the second of the pool by 2, at Mile 3.1 (Fig. 1), immediately above mill dams. This impoundment apparently and below a road crossing (Fig. 5). drowned a section of waterfalls and major rapids. The crossing was completed in 1958 with at the end of the Probing upstream pool construction of 3 culvert tubes through revealed a waterfall (Mile 2.1) submerged which the creek passed with current veloci- at least 6 feet Water over high. passing ties greater than 3 ft/sec even during peri- Dam 2 entered the formed (Mile 2.36) pool ods of modal discharge. High current by Dam 3 after flowing over a short sec- velocities in times of flood and shallowness tion of bedrock and marl waterfalls. The of water in periods of modal discharge, bottom third of this of the upper pool was apparently restricted movements of fishes covered with with some sand lateral marl, and other animals upstream. With the to the and in the downstream two- current, abrupt change in gradient near Mile 5.0 (Fig. 2), the bottoms gradually changed TABLE2.-ESTIMATED PERCENTAGES OF BOTTOM MIATERIALS IN DIFFERENT SECTIONS OF DOE RUN, MIEADE COUNTY, KENTUCKY. DATA MARKED WITH * AN ASTERISK ( ) ARE FOR IMPOUNDMENTS, BACK- WATERS, OR SECTIONS OF SMIOOTH, ALMOST-LAMI- NAR FLOW

Types of Bottom Materials Mile Section Marl Bedrock Rubble Gravel Sand Silt Clay 0.0 -1.1* - 10 5 25 55 5 - 1.1 -1.65* - 2 3 10 65 20 1.65-2.1 - 65 10 15 8 2 2.1 -2.36* - - 10 35 55 2.36-2.82* 30 - 5 10 5 50 2.82-5.3 65 15 10 5 3 2- 5.3 -9.71 trace - 2 5 18 65 10 FIG. 5. Doe Run, Meade County, Ky., at Sta- ' See footnote, Table 1. tion III looking downstream, January 28, 1962. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 11 from broken marl, rubble, and gravels, to silt and sand near the confluence of Doe Run with the Ohio River (Table 2). Sta- tion IV was at the downstream end of the marl section, near Mile 4.8, and Station V was in the downstream area near Mile 7.7 (Figs. 6 and 7, respectively). Three changes have occurred in the bottom materials of Doe Run in the recent past, probably as a result of human activity. (1) Prior to modification of the stream by damming, or at some earlier date, the marl area of Doe Run extended upstream at least FIG. 7. Doe Run, Meade County, Ky., at Sta- to Mile 1.65. Much of the material listed tion V looking downstream, June 13, 1961, when backwaters of the Ohio River were as rubble at Station II (Table 2), between present. Miles 1.65 and 2.1, was marl concretions, but there was no evidence of marl deposi- The permanent tributaries to Doe Run tion in that area during this study. Cut had gradients that corresponded to their banks near Station II revealed deposits of locations on the creek; where the gradient marl concretions as buried bars lateral to was relatively low in Doe Run, the tribu- the present channel. (2) The upper 1.5 taries were low in gradient, and between miles of Doe Run, where bedrock was Miles 1.5 and 5.0 (Fig. 2) where Doe Run thinly overlain with sand, gravel, and silt had high gradient, the tributaries had high during this study, probably had smooth gradient. Most of the permanent tributaries bedrock bottoms prior to cultivation of the originated in springs. There were, how- floodplain and erosion of the stream banks. ever, 2 kinds of intermittent tributaries: (1) (3) Maintenance of navigation pools in the subsurface tributaries whose waters rose Ohio River impounds the lower parts of after periods of prolonged or rapid rainfall, Doe Run and speeds the deposition of silt and (2) surface streams that developed in those areas. Damming effects of the after excessively heavy rains. Of the first navigation pool of the Ohio River were type, Tributaries I-c, I-c', and I-c", thought evident in Doe Run for more than a mile to be inactive springs at the beginning of upstream from the mouth. the study, carried large amounts of water at certain times. An intermittent tributary called Rushing River (Fig. 1) was the only surface stream that consistently contributed any significant surface runoff to Doe Run. The stream flowed directly into Doe Run on 5 or 6 occasions during the period of study, but may have contributed much water on many occasions by passage underground into 2 large, open sinkholes about 0.5 mile from the source of Doe Run. In periods of extremely heavy rain, Rushing River flowed into sinkholes and did not rise over a low divide until its volume exceeded the capacity of the subterranean channels. When that capacity was exceeded, surface water FiG. 6. Doe Run, Meade County, Ky., at Sta- tion IV looking upstream, April 24, 1960. Photo- connected with the headwaters of Tribu- graph by Charles B. Stone. tary I-e to enter Doe Run at Mile 0.5 (Fig. 12 WILDLIFEMONOGRAPHS

1). The fate of water that entered the open second series of 20 samples collected individually sinkholes is unknown, but presumably it from the current at Station I at intervals of about entered Doe Run either through the main 1 minute had a range from 71 to 86 units, with a mean of 77.6, standard deviation (SD) 4.46, and source or through the intermittent springs, V, 5.74. I-c, I-c', and I-c". Temperatures in Doe Run were measured with Permanent tributaries generally flowed a Whitney resistance therometer, maximum-mini- through steep, V-shaped valleys, with the mum thermometers placed at permanent locations along the creek, and with standard, vest-pocket exception of those upstream from Mile 1.5 thermometers. All instruments were calibrated and those in the extremelower portion and against a standard centigrade thermometer, and on the floodplain of the Ohio River. Per- appropriate corrections were made. Temperatures manent tributariesentering Doe Run in the were recorded in, or converted to degrees centi- areas grade. upper (Fig. 1) had bottoms of sand The amounts of light reaching Doe Run were and fine gravels and rose in springs gen- measured at Stations I, III, and V, on each erally at the same elevation as the main sampling date for 14 months following September of the creek. Farther 1959, with a PR-1, General Electric light meter, spring downstream, containing a photoelectric cell sensitive to the the bottoms of tributaries in the high- visible spectrum, and fitted with an incident light gradient section were bedrock, broken attachment. The light recorded is an index of rubble, and marl. Most of the latter streams intensities of solar radiation about 5 ft above the had stream surface at each station. One to 10 mea- gradients greater than 400 fpm and surements were made on each date, more being rarely were longer than a half mile. taken on days with little cloud cover than on overcast days. The data were converted to foot METHODS candles using tables furnished with the instrument, and all readings were corrected to 1:00 PM (EST) Physical.-Discharges were estimated at regularly by factors derived from data taken in June 1960. Stations I, III, and V, by a modification of the Chemical.-Determinations of dissolved cork-float method oxygen (Embody, 1927) for 14 months were made by the unmodified Winkler method following October 1959, and from water levels (Welch, 1948) after series of comparable samples recorded on stationary flow gauges and other ob- were the Rideal-Stewart and Alster- for analyzed by jects the remainder of the study. Measure- berg modifications of that method with no dis- ments of high-water marks known to have been cernible differences in results. formed Samples usually between sampling dates also were in- were obtained with a Kemmerer from the with the of sampler cluded, dates their occurrence docu- main current, unless studies were mented residents of the specific being by area. made on other parts of the stream. No corrections Time of flow was measured on 4 occasions were applied to compensate for displacement of using fluorescein or potassium permanganate as volume but all and sample by reagents, determinations dyes, one estimate of time of flow was made were corrected to mean barometric on the basis of debris pressure by drifting and silt, and one factors given by Mortimer (in by the of rotenone-treated water in Hutchinson, 1957). passage the No attempt was made, however, to correct for area between Stations IV and V. Rates are based variations caused by barometric pressures extant on time of average flow, by allowing about a at the time of sampling, and this undoubtedly fourth of the tracer material to pass a given point caused some error (Ricker, 1934b). Most samples downstream from the point of application before were titrated in the laboratory unless long-term recording the time it had been moving. studies dictated titration in the Turbidities were field. Oxygen measured against distilled saturations were derived from Mortimer's water blanks in a Bausch and nomo- Lomb Spectronic-20 graph (Hutchinson, loc. constructed from Colorimeter. from the cit.) Samples strongest currents data of et al. or were at each station were retained in and determi- Truesdale, (1955), calcu- jars lated from data the nations made in the laboratory after were given by latter authors. samples Carbon violently agitated. Data are presented as "turbid- dioxide in Doe Run, the free carbon ity units," roughly equivalent to mg/l, from tables dioxide plus carbonic acid (Hutchinson, 1957), obtained from the Hach Chemical Company, was determined from data on pH and alkalinity Ames, Iowa (Anonymous, n.d.). To determine (Ruttner, 1953). The concentration of hydrogen the accuracy of colorimetric turbidity measure- ions, expressed as pH, was determined electro- ments and variation in turbidity of stream water metrically, using pocket pH meters manufactured on a short-term basis, 2 series were run. First, 10 by Analytical Measurements Corp., and by Beck- aliquots from a single sample container showed man, Inc. On occasions when electrical apparatus a range from 79 to 80 units, with a mean of 79.3 was nonfunctional, pH values were ascertained and a coefficient of variability (V) of 0.71. The using colorimetric techniques. Total alkalinity, ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 13 expressed as mg/l CaCO3, was measured using thoroughly disturbing the enclosed quadrat by methyl purple as an indicator, and titrating with hand or with an appropriate object. Care was N/50 sulphuric acid. taken to prevent backwash at the mouth of the Total iron was determined by a modification of net, which was shown by Badcock (1949) to in- the 2,2' bypyridine technique of the Hach Chemical troduce 4 to 20 per cent sampling error. Marl Company (Anonymous, n.d.). Reagents for this riffles were sampled in the swiftest portion in and the following tests were obtained from the which the sampler could be effectively used, and Hach Company, and were checked against known always directly on marl; the poorly consolidated standards prior to full-scale use. Standards were surface material was broken and the surface prepared as recommended by the American Public scraped thoroughly. In riffles, each object in the Health Association (Anonymous, 1960) for each quadrat was scrubbed by hand into the net. Ek- analysis, and error was 5 per cent or less in all man dredge samples were obtained in soft-bot- tests used here. Duplicate series were run to tomed pools, and in beds of vegetation. The jaws indicate variations that might have resulted from of the dredge were sharpened and precise samples the kinds of sample containers used, or from the of vegetation were obtained by plunging the de- lapse of time between collection and analysis vice into the bed with enough force to cut through (Table 3). stems, roots, and substrate. On bedrock bottoms Total phosphates (ortho- plus metaphosphates) the sampler was used to cut Fissidens and was were measured spectrometrically by the ammo- held firmly to the bottom as withdrawn to scrape nium molybdate-stannous chloride method, after moss, including its holdfasts, from the substrate. boiling in acid (Anonymous, 1960). Samples were The relative shallowness of water in most of Doe run within 48 hours after collection, and the Run allowed use of the dredge in sand or detritus possibility of error introduced by absorption of bottoms, which usually clog the device or block phosphate by the polyethylene containers (Odum, complete closure of the jaws (Rawson, 1930). 1957a) or through bacterial action (Heron, 1962) Only those samples where complete closure of the was checked by running duplicate series at various jaws was effected were used. times after collection. No significant differences Surber samples were preserved in toto in 10 per were apparent 52 hours after collection (Table 3). cent formalin. Dredge samples were washed in the Nitrate and nitrite nitrogen were determined field in a 40-mesh sieve prior to preservation, and spectrometricallyby the brucine method and by a thus are comparable to those taken with the 40- modification of the naphthylamine hydrochlorine- mesh Surber netting. The preserved bottom ma- sulfanilic acid techniques, respectively (Anony- terials, and the included organism, were sorted in mous, 1960). the laboratory in white enamel pans or in glass Biological.-Quantitative samples of benthic in- dishes over a light-table, by flotation with satu- vertebrates were obtained with an unmodified rated salt (Lyman, 1943) or sugar solutions Surber Sampler (Surber, 1937), and a 6- by 6- (Anderson, 1959), and by hand-picking. A light- inch Ekman dredge. The Surber Sampler was table was useful in sorting small animals from used in riffles with marl, rubble, or gravel bottoms, vegetation in which they were entangled, allowed but its use in beds of Fissidens and Myriophyllum easy differentiation of inhabited from uninhabited, was precluded by the relatively great amounts of semitransparentcases of certain caddisfly larvae, vegetation lost. Surber samples were taken by and revealed some counting error based on diffi- pressing the frame tightly to the substrate and culty in distinguishing between some whole,

TABLE 3.-DUPLICATE ANALYSES OF WATER SAMPLES FROM STATION I, DOE RUN, MEADE COUNTY, KEN- TUCKY, SHOWING VARIATIONS DEVELOPED WITH TIME AFTER COLLECTION, WITH DIFFERING SAMPLE CONTAINERS, AND REPRODUCIBILITYOF RESULTS (DATA IN MG/L)

Sample Held in Glass, Analyzed Sample Held in Glass, Analyzed Sample Held in Polyethylene, 2 Hours after Collection 52 Hours after Collection Analyzed 52 Hours after Collection Dupli- Range of Mean SD V Dupli- Range of Mean SD V Dupli- Range of Mean SD V cations Analyses cations Analyses cations Analyses TOTAL IRON 5 0.60-0.62 0.61 0.01 1.6 5 0.61-0.62 0.61 0.01 0.8 - -

PHOSPHATE PHOSPHORUS 10 0.15-0.17 0.15 0.01 4.6 10 0.14-0.16 0.15 0.01 4.0 10 0.14-0.15 0.15 0.01 4.1

NITRATE NITROGEN 10 1.60-1.72 1.67 0.04 2.3 10 1.57-1.67 1.62 0.03 1.9 10 1.59-1.69 1.64 0.04 2.2 NrRITE NITROGEN 5 0.05-0.07 0.06 0.01 11.6 5 0.05-0.06 0.06 0.01 16.4 5 0.05-0.07 0.06 0.01 16.4 14 WILDLIFEMONOGRAPHS preserved crustaceans and their freshly cast exo- of gut contents and identifications were made by skeletons. That error was discovered late in the direct comparison of food items with known ma- study and amounted to 1.0 to 3.3 per cent terials collected from Doe Run. Data were re- (average 1.43) in 10 samples of isopods from corded by frequency of occurrence in the tracts, Stations I and II. No factor was applied because on a rough quantitative basis by spreading the of variability, but the error did not influence re- contents of a given tract over a 0.1-mm grid and sults greatly. Organisms were "rough sorted" counting the numbers of squares covered by each from bottom materials and placed in 70 per cent item, and by counts of individual items. General ethanol. Then they were sorted to taxonomic discussion of foods of various species uses the categories, the preservative was removed by a relative terms "rare," "scarce," and "abundant," modified vacuum filter using discs of conventional with each term implying an increasing change of filter paper, and the animals were immediately one order of magnitude from a base of one item weighed to the nearest milligram on a Mettler, in those animals feeding on diatoms, particulate electric, single-pan balance. Molluskswere weighed detritus, or other small-sized materials. In preda- in the shells. Trichopterans except those at Sta- tory animals, where digestive tracts rarely con- tion I were removed from their cases before tained more than 10 items, the terms are subjec- weighing; they were small, with fibrous, self-pro- tive. duced cases, and the time needed to remove them Variable numbers of the more abundant species was not warranted. Weights of crayfish and snails of fishes in Doe Run were analyzed for food re- are arbitrarily excluded from calculation of rela- lations; stomachs, or the anterior parts of the tive gravimetric abundance because of difficulty digestive tracts in catostomids and some cyprinids, in sampling the former in particular, and because were excised and the contents examined at various of inclusion of shell in weights of snails. Altera- magnifications with a binocular microscope. The tion of weights due to shrinkage in preservative volume of the contents of each stomach was esti- (Leonard, 1939) was not considered. mated visually, and the relative volumes of the Qualitative samples were taken by various various food items were recorded as estimated means and usually preserved in 70 per cent alcohol percentages of the total material. in the field. All major tributariesof Doe Run were Fishes were collected by seines from 4 to 30 sampled at least once, and samples were obtained feet in length and of mesh size ranging from 0.02- in the mainstream from Stations I through V on to 0.5-inch bar measure, by an electric shocker various occasions. Identifications were made by delivering 110 volts A.C., by scape nets of various standard keys such as Ward and Whipple's Fresh- kinds, by 3- and 4-foot hoop nets with pot meshes Water Biology (Edmonson, ed., 1959), Usinger of 1 inch, by experimental gill nets 125 feet long (ed., 1956), and Pennak (1953), with further with meshes ranging from 0.75 to 1.5 inches bar reference to literature cited therein. measure, by pole and line, and by both powdered Food relations of invertebrates were estimated and emulsifiable rotenone. from examination of gut contents of 10 to 20 individuals of a species from each of 2 collecting PHYSICALENVIRONMENT periods, if that many were secured. Major collections of animals were obtained for this purpose Characteristics of Flow in the Mainstream in September 1960, in a period of modal, presumably "optimum" conditions in the stream, and According to Meinzer's (1923) classifica- others in February 1960, at the middle of "winter tion, Doe Run is a spring of the second discharge." Some animals were not present in magnitude, that is, a spring with an average those periods and were analyzed from collections discharge between 10 and 100 cubic feet at other times. Field and laboratory observations were used to substantiate findings from analyses per second (cfs). At Station I, the average of digestive tracts when feasible, and food rela- discharge between October 1959 and Octo- tions of some invertebrates are included only on ber 1961 was about 65 cfs, and was roughly the basis of such observational data. doubled at Station III, at about 140 cfs. Specimens for food analyses were preserved in Data from Station V were be- weak with a small amount of incomplete formalin, copper cause of sulfate added. That compound tends to replace the sporadic presence of back- certain parts of the chlorophyll in algae and other waters from the Ohio River (Fig. 8), but plant materials, causing them to remain green and the computed average was near 250 cfs. thus aiding in identification and in determining In the light of field experience, the mean whether the material was living or dead when volumes of discharge seemed too great. ingested. The animals gulped the preservative as The overall data indicated a modal volume died. the anterior of they Only parts digestive of between 20 and 40 cfs at all tracts were examined in most species, although discharge further dissection was done in some instances. A stations, and this range represents a realis- binocular microscope was used for examination tic "normal discharge" for the stream. The ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 15

GODMAN FIELD, FORT KNOX

o JiL .LL A.. .!LL,it.1 ,,a ,.I -it.,.,i,,L . ,..,l .,.,.,, L 4 - ELIZABETHTOWN

4 - 0 I. s IAl J L L, L -iLiA i JIA JL .ak.I II IAI. Il I ll.1 *.. i JL lisitoljI.if A J 4 ?*~~~~~~~~~~~~~~-~IRVINGTON

lU i IIl .a L o,,,].i, DE , ~,1.'Ll;l.L- ~. A. ~4:..i ilU-.. JUNE, Al.~,.:,JUY.^..,i 1,au.,..!-, WI~I.lSE NV. DE. i ~,*U. ! ur!J~ ,.1'1I., .'-uY

FIG.8. Daily records of precipitation at 3 stations near Doe Run, Meade County, Ky., and records of discharge in Doe Run from November 1959 to July 1961. Dots denote records obtained at Station V when that area was not inundated by backwaters of the Ohio River.

large average discharge at each station re- The rapid movements of surface water sulted from no more than an accumulated into Doe Run is indicated by the close 2 months of extremely high discharge in agreement between some of the dates of the 2-year period (Fig. 8). Characteriza- high discharge in the creek and those of tion of streams on the basis of their average heavy rainfall in the vicinity (Fig. 8). discharge is acceptable only if long-term Rainfall in northern Kentucky varies greatly records are available. In relatively short- from place to place (Kendall, 1941), and term studies, such as this one, the modal such variation probably explains some high values give a more accurate estimate of the waters in Doe Run in the absence of "normal condition." Barring major changes recorded heavy rains, for example, in the in the stream itself, the longer that records period from mid-March through April 1961, are kept, the nearer the average of such and for lack of variations in discharge dur- data should approach the mode obtained ing other periods of relatively high, re- in the first few years. corded rainfall (August through September During 1959, 1960, and 1961, discharge 1960). Other factors contributing to the of Doe Run was variable and generally influence of rainfall on the discharge of high from December to June, then rela- the creek were levels of soil moisture prior tively constant from mid-June through No- to the precipitation, presence of water- vember (Fig. 8). In 1961, the winter holding ground cover in the form of litter period was culminated between February or standing vegetation, presence of frozen and May by extremely high volumes of dis- soils in winter that permitted almost total charge. runoff, and presumably the storage capac- 16 WILDLIFE MONOGRAPHS

TABLE 4.-VOLUMES OF DISCHARGE IN CUBIC FEET All the larger tributariesto Doe Run, the PER SECOND AT 2 STATIONS ON DOE RUN, MEADE main-sourcespring, and 3 presumably in- COUNTY, KENTUCKY, AND IN TRIBUTARIES LOCATED active were BETWEEN THOSE 2 STATIONS, UNDER CONTRASTING springs (Tributaries I-c-I-c"), CONDITIONS flowing violently, with water being forced up in boils as much as 3 feet high. Esti- October 15, mates of volumes of discharge for each Tributary March 6, 1961 1960 or Station Discharge Discharge tributary,and of Doe Run at StationsI and Station I 600 15 III, are in Table 4. The only evidence of Trib. I-a 4 trace great surface inflow during the rain was Trib. I-b 5 0.5 Rushing River. According to high-water Trib. I-c submerged, none marks,the flow observed was maximalfor flow ca. 50 Trib. I-c' 150 none that date at both Stations I and III, and Trib. I-c" 100 none about 85 per cent of the water passing Trib. I-d trace none Station III was accounted for in discharge Trib. I-e of the springs visited. The independence (= RushingRiver) trace' none Blue Spring 250 4 of Doe Run from surface water in periods Trib. I-g 1 none of normal or subnormaldischarge is illus- Trib. I-h and strated by data for October 15, 1960 Buffalo Spring unknown none (Table 4). Trib. II-a 10 0.5 In at times of low relative Tribs. II-b and II-b' unknown none summer, Tribs. II-c and II-d trace none humidity, high temperature,and maximal Station III 1,350 20 insolation, slight deficits in discharge were Total known discharge detected between Stations I and V, and of Tribs. and Sta. I 1,170 20 diurnal fluctuations in discharge were 1 Trace = less than 0.25 cfs; Rushing River had carried noted on 3 occasionsduring 24-hourstudies ca. 50 cfs earlier in the day as judged from high-water of water These variations marks. chemistry. apparently resulted from evaporation and transpiratoryactivities of the riparianvege- ity of the undergroundaquifer feeding the tation, and were most pronounced in the creek, the extent and geography of which upstream area where bedrock permitted is unknown. streamsidevegetation no more than 2 to 6 On occasions of high discharges, in- feet of soil in which to root. The greatest creases in water levels occurred rapidly. such fluctuation was 2 inches recorded on On a and June 23, 1960, during heavy ap- August 17-18, 1961, near Station II, and parently local thunderstorm,the discharge the diminutionof volume between increased from about 60 cfs at greatest 8:30 AM Stations I and V was about 1 cfs when (EST) to near 200 cfs at 10:00 AM. The maximum at Station I on that dischargeat Station I was 18 cfs. discharge It must date was 300 cfs, and occurred at about be reiterated here that water 12:30 PM. sampled at Station I cannot be considered At modal discharge there was less differ- the same as that sampled at Station III a ence between volumes of water at Stations short time later in the same day, or at I and V than when flood conditions pre- Station V later in the day. The concept of vailed, but the greater differences between a column of water, long used in the study stations in periods of flood reflect variation of lentic situations,is not applicable to the in discharge of downstreamsprings, rather study of flowing systems. However, in than variationsin surfacerunoff. On March recent years some work has been done on 6, 1961, after heavy rainfall on that and the aspects of stream ecology by following a preceding day (Fig. 8), Doe Run was given mass of water from the origin of the traversed on foot from Station I to Blue stream to a given point downstream,and Spring, and visited at Stations II and III. measuring certain factors as modified in ECOLOGY OF DOE RUN, MEAPE COUNTY, KENTUCKY-Minckley 17 the length of the section (Schmitz, 1955; TABLE 5.-SPEED OF PASSAGE OF WATER THROUGH OF Odum, 1956, 1957a, b; and others). Studies DIFFERENT SECTIONS OF THE CHANNEL DOE RUN, MEADE COUNTY, KENTUCKY, AT DIFFERENT on flood crests, sediment loads, and other LEVELS OF DISCHARGE features of flowing water have shown that a given mass of water tends to become dis- Discharge in cfs persed and diluted, both laterally and lon- Station I (Mile 0) 15-25 40-50 80-100 gitudinally, as it passes down a channel Station III (Mile 3.1) 20-30 60-70 130-150 Station V (Mile 7.7) 25-35 100-120 230-260 (Thomas, 1958). The use of the technique, however, gives a time parameter necessary Stream Section Average Speed of Flow, in ft/sec in a in- 0.0 -1.2 0.6 1.1 1.3 understanding constantly moving, 1.2 -1.65 0.2 0.5 0.6 tegrated system such as a stream, and 1.65-2.1 1.0 1.8 2.1 which is lacking in studies based on in- 2.1 -2.82 0.2 0.5 0.6 stantaneous sampling techniques. 2.82-5.3 1.0 1.6 2.0 Time of flow in Doe Run varied from 5.3 -9.7 0.3 0.5 0.6 section to section and in relation to dis- Time of passage, in charge (Table 5). The average time of ft/sec, from Mile 0.38 0.69 0.83 flow in the upstream 1.2 miles, where 0.0 to 9.7 smooth bottoms allowed almost laminar Time of passage, in streaming of currents, was about 0.6 ft/sec hours, from Mile 0.0 37.3 20.6 17.1 to 9.7 at a discharge of 20 cfs. This increased to more than 1 ft/sec at a discharge of 40 cfs. In the mill impoundments, the average the recognition of extreme variation in a of flow was about 0.2 speed ft/sec at 20 single riffle before attempting to interpret cfs, doubling with a doubling of discharge. upstream-downstream chemical data. In the swiftest section of the creek, between Miles 2.8 and 5.3, the average flow Suspended Solids and Sedimentation was about 1.0 at modal ft/sec discharge, Of 60 determinations of turbidity at Sta- and the decreased at Mile velocity sharply tion I, 8 contained more than 50 turbidity 5.3 to about 0.3 in ft/sec the low-gradient, units, and 3 more than 500 units (Fig. 9). downstream section of the creek. Turbidities greater than 100 units were as- These data illustrate the variability of sociated with periods of greatest discharge, time of flow in a given segment of stream and in periods of relatively constant dis- in relation of discharge, and point out the charge the slight increases in turbidity after marked variability of such data from sec- heavy rainfall probably indicated influx tion to section. Even more important, per- of surface water into the subterranean haps, in using time-of-flow techniques, is aquifer, even though no appreciable

I 5...045VMS--^-^ 5...Moo- - r- e~~~~~~~~~~~~~~~~~0

- I - - I - - I - I - . - - - - . - . - - - - I . - - I .- - - . - - I . .. . .I - - - -.. ------

FIG. 9. Turbidities of water at 2 stations on Doe Run, Meade County, Ky. Data from Station III were omitted because of similarity to those from Station I. 18 WILDLIFEMONOGRAPHS changes in discharge were detected (com- samples, collected at each of 3 stations: pare Figs. 8 and 9 in the latter part of Station I, 142; Station III, 155; and Station 1960). Minor fluctuations in turbidities V, 42 units. After clearing the downstream between stations on the same date may re- valley of Doe Run in preparation for flect errors in sampling, or may be a func- impoundment, there was a pronounced in- tion of time of flow and more long-term crease in turbidities at Station V. Compari- changes in turbidity. There appears, how- son of mean turbidities of 10 samples each ever, to have been some increase in the from Stations III and V, 12.6 units and 80 average turbidity from Station III to Sta- units, respectively, reflects the influence of tion V in periods of modal flow; the means runoff from the denuded slopes and flood- of 10 samples from each of 3 stations, taken plain, the activities of earth-moving equip- when discharges were less than 50 cfs, are: ment, and transport of soil pushed into the Station I, 12.7; Station III, 12.9; and Station stream during clearing operations. V, 15.3. Although the bottom types have been Generally, turbidities of less than 50 discussed, it seems appropriate to describe units were made up of suspended colloidal some specific aspects of sedimentation in clays. Less than 10 per cent of the turbid- Doe Run. Most sediments, excluding the ity settled from a sample retained in a marls that will be discussed later, were 1,000-ml graduated cylinder in a week derived from erosion of the immediate (original determination, 33 units). At tur- banks, and, to a greater extent, from bed bidities of less than 15 units, the water of load from the subterranean channels sup- the creek was somewhat "milky" in appear- plying the stream. The frequency of abrupt ance, particularly in open areas that re- bends in Doe Run (Fig. 1), and the degree ceived direct sunlight, and apparently was of pool-riffle development in some sections a phenomenon of light refraction by finely (Table 1), allowed for many features of divided, suspended solids. Filtration with deposition of bed load because of great "HA" membrane discs, pore size, 0.45 mi- variation in competence of water passing cron, resulted in total removal of turbidity through such an irregular channel (Braden, that could be detected with the spectrom- 1951). Also, differences in gradient allowed eter. At high turbidities, specifically those for deposition of deep, submerged beds of in June 1960, and during high dis- fallen leaves and other organic detritus in charge in late winter and spring 1961, a swift vortices downstream from obstruc- much greater proportion was noncolloidal. tions in the channel, and the formation of After 24 hours, the surface water in a 1,000- transverse sills of gravels downstream from ml graduated cylinder retained 30 per cent places where extremely swift water con- of the original determination of 5,000 units, tacted relatively quiet waters of large pools but after a week of undisturbed settling and resulted in various kinds of hydraulic the reading was 230 units. The largest jumps (Braden, 1958). Those sills, and particles found by microscopic examination other deposits in the channel of Doe Run, of the sediment were fine sand, with a were subject to movement in periods of maximum diameter no greater than 0.25 high discharge. mm (Twenhofel, 1939). The high turbidi- Some major deposits in the upper 3 miles ties in Doe Run probably originated in the of Doe Run were associated with beds of drainage of Rushing River where runoff macrophytes. Lateral bars of relatively fine from cultivated fields moved directly into sediments were formed in stands of water- subterranean watercourses. cress, Nasturtium officinale R. Br., and At times, backwaters of the Ohio River elongate bars were present beneath beds inundated Station V (Fig. 7) and allowed of Potamogeton, Myriophyllum, and Nitella. for considerable settling of sediments as One of the most puzzling deposits was illustrated by the average turbidities of 9 the thin film of amorphous sludge formed ECOLOGYOF DOE RUN, MEADE COUNTY,KENTUCKY-Minckley 19

in areas of moderate current after periods of high turbidity. Although the deposition of fine sediments was expected in quiet backwaters during flood, films frequently formed on stones, gravel, and vegetation in areas where velocities of flow were greater than 2 ft/sec. Violent agitation of samples of the material with distilled water resulted in suspensions frequently as stable, if not more so, than colloidal turbidities. At times, however, flocculation occurred, and the suspensoids were precipitated. Al- though flocculation of colloidal turbidities is known to be brought about by activity of various electrolytes, or by contact of one particle with another of opposite charge (Twenhofel, 1939; Irwin and Stevenson, 1951), it seems unlikely that adhesion to the substratum could be accomplished in the currents present, or that the negatively charged clays would be flocculated in the alkaline conditions generally persistent in Doe Run. The films appear to have been formed through biological activity and will be discussed later. Films exposed to the air dried rapidly, flaked, and fell or were blown into the water, where they sank to become incorporated into bottom sediments. Temperature Relations Seasonal air temperatures in the Doe Run area are moderate, the average Janu- ary temperature is between 1.1 and 2.2?C, and that for July about 26?C (Kendall, 1941). Air temperature recorded at Station I by a maximum-minimum thermometer in a shaded place about 8 feet above the water compared favorably with data provided by the U. S. Army Weather Station at Godman Field, Fort Knox, Ky., 8.5 miles east and 1.8 miles south of the source of Doe Run (Fig. 10). Records of air temperature at Station I were discontinued in April 1961. Throughout 1959 and 1960, water temperature at Station I was consistently between 13.0 and 13.5?C (usually 13.3?) in periods of modal discharge, but in spring 1961 it increased to between 13.5 and 14.5?C. The reasons for this change are

X Mw31SOl-WMuLVWU3N MlJ. 1-OVlWN383$3Mf l.VMUUL- WV not apparent, but may be related to 20 WILDLIFE MONOGRAPHS changes concomitant with the high volumes streams are interrelated. For example, the of discharge in 1961 (see Fig. 8). Tem- time of flow in Doe Run is decreased with peratures at Station I varied somewhat in increased discharge,thus the time water is times of high discharge; the lowest tem- subject to warming is reduced. Higher perature recorded was in January 1960, discharges in Doe Run had pronounced when a reading of 12.3?C was obtained, moderating effects on temperatures, as and the highest was on June 17, 1960, at shown by data for summer 1960 (Fig. 10), 15.6?C. Variations of the same magnitude when modal dischargebegan about August (from about 12.0 to 16.0?C) were recorded 1, water temperaturesthat had been rela- in the period of high discharge between tively low in May, June, and July immedi- late February and late May 1961. ately rose 2 or 3 degrees at Station III, and Maximum-minimum thermometers at as much as 10 degrees at StationV. At high downstream stations were plagued by levels of discharge, the speed and turbu- washout and by periodic inundation by the lence of the stream allowed little warming Ohio River at Station V. However, the in backwaters, and except in the extreme records illustrate the gross features of discharge of 1961, when the water spread temperature in the stream (Fig. 10). At from "bluff to bluff" in places, the narrow- Station III, the maximum temperature re- ness of the channel caused great reduction corded was 20.0?C in August 1960, and the of the surface-volume ratio. Although minimum was 6.1?C in March 1961. At turbulenceusually increasedwith increased Station V, the maximum and minimum discharge, the amount of surface rough- recorded temperatures were 25.6 and 1.7?C, ness, or "riffle,"decreased as obstructions respectively. Usually, the greatest ranges in the stream bed became more deeply of temperature occurred in late autumn and covered with water. Thus, increased dis- early winter at each of the downstream charge reduced the amount of spray stations. formed in passage over cataracts, dams, The major factor in warming natural and steep inclines, and reduced the cooling waters during the day is the absorption of effects of evaporation. In summer, careful solar radiation, whereas cooling is con- measurements above and below Dam 2 trolled largely by air temperature (Ricker, indicated a slight decrease in temperature 1934a). In Doe Run, the minimum water (0.05 to 0.1?C) downstream. The only ice temperatures did not fall below 13.3?C at that formed along Doe Run in winter, with Stations III and V from late May to October, the exception of thin layers in cut-off pools an indication that relatively high air tem- along the channel, resulted from the con- peratures at night were influencing water tact of spray with overhangingobjects, and temperatures, with a resultant positive de- subsequent aggregation into ice concre- viation from the temperature at the source. tions. In periods of modal discharge (see Table In periods of high turbidities, greater 4), water passing from tlie source of Doe absorption of solar energy may occur in Run in the afternoon was heated by insola- moving water (Ellis, 1936). Nevertheless, tion in 1 or 2 hours, and maintained a posi- the reductionin light penetrationmay have tive deviation throughout the night so that a cooling effect by shading and preventing heating began from a slightly higher point heating the bottom materials by absorbed the following morning. Water that left insolation. In shallow streamssuch as Doe Station I at night was delayed in the mill Run, in periods of low turbidity, warming impoundments and received considerable of bottom materials in daytime may in- insolation before reaching Station III the creasethe temperatureof the streamcon- following day, thus maintaining the posi- siderably. My few measurementsof the tive deviation. temperatureof sedimentsin upstreamsec- Most factors influencing the rates and tions of Doe Run indicatethat at maximal amounts of warming or cooling of water in solar radiationthe sedimentsare 0.1 to ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 21

0.9?C warmer than the water passing over Where cool density currents rose along them (16 measurements0.1 to 1 inch below the face of dams, a distinct line of demarca- the sediment-water interface in water 0.5 tion was visible 1 to 3 feet upstream,pre- to 1.5 ft deep; turbidities less than 10 sumably representing a thin boundary of units). Some heating may result from mixing between the lens of warm surface biological activity in the sediment. water and the density current. Little The impoundmentsin the upper section warming could be detected in water going of Doe Run allowed more warming of the over the dam. water in that region than elsewhere in the Wind action sometimes disrupted strati- stream. Preliminary investigations in the fied flow in the mill pools. Such an occur- mill impoundments in June 1960 yielded rence in Pool 2 (Fig. 11D) was studied on surface temperaturesabove 25.0?C, which August 3, 1960, from 8:10 until 11:30 AM. were much higher than any obtained at Air temperature ranged from 28.0 to Station III. Accordingly, detailed studies 31.0?C, discharge into the pool was about were made of water movements through 40 cfs, and the sky was cloudless. In early the pools and the concomitant modifica- morning the pool was relatively homo- tions of temperature. Data from each of geneous at its upstreamend, but stratifica- the 3 impoundmentsare similar, and find- tion and subsurface currents were begin- ings in one pool are applicableto the others ning to form by late morning (Fig. 11D). if minor variations in differential shading The study was resumed at 1:30 PM, and it by riparian vegetation and differences in was found that during the 2 hours in which configuration of the channels are disre- observationswere not made there had been garded. For convenience,the pools will be a rapid warming of surface water. About referred to by the number of the dam con- 12:45 PM a strong wind began blowing cerned (Pools 1, 2, and 3, referringto the downstream, and by late afternoon the impoundments of Dams 1, 2, and 3, re- warmed surface water was being pushed spectively). over the dam and cool waters were being On June 9, 1960, at an air temperature retained in the downstream part of the between 23.0 and 26.0?Cand a dischargeof pool. During the afternoon, temperatures about 40 cfs, a longitudinal series of tem- of 30.0?C were recorded at the surface in peratureprofiles was taken in Pool 3. The backwatersnear Mile 2.3. temperatureof the water entering the pool In winter, stratifiedflow in the mill pools was 15.3?C at the beginning of the study was also present, but the configurationof at 2:00 PM. The cooler water moved the isotherms was quite different. On through the pool by a general underflow February 7, 1961, under a totally overcast in areas of greatest depth, rose to the sur- sky and at air temperaturesbetween 4.0 face along a bar of sediments behind Dam and 6.0?C, the discharge at the head of 3, and passed over the dam into the stream Pool 2 was 32 cfs, and water temperature below (Fig. 11A). A lens of warm water was 11.2?. A longitudinalprofile at points was retained in the pool and dissipated or comparable with earlier studies indicated passed over the dam at night when cooling a density current in Pool 2 that must be occurred. Cross-sectionaltemperature pro- classed as an interflow (Coker, 1954) be- files indicated that lenses of warmed water cause of the presence of differing waters lay generallyin the center of the pool (Fig. both above and below the density current. 11B-C), and were influenced very little by This phenomenon was brought about by configuration of water masses beneath rapid cooling of surface water, its move- them. Discrepancies in isotherms resulted ment laterally and down the sides of the from movement of water during sampling, channel to contact colder water near the a serious drawback in the use of a single bottom and thus become concentrated in thermal element for temperaturemeasure- an area of maximum downstreamvelocity ments in moving water. in the center of the cross section (Fig. 11F). 22 WILDLIFEMONOGRAPHS

1-- LU LU L&.

E Mile 2.36 Mile 2.3 Mile 2.2 Mile 2.1 v v v

3 I- .U 4 U. 5

6 -92 \ F 7- 19I ,, \/ 0

?8. POOL 2 *034

96 9

8 0 S 10 15 20 255 30

FIG. 11. Temperatures in mill impoundments of Doe Run, Meade County, Ky.: A-longitudinal profile of Pool 3, June 9, 1960; B, C-cross sections of Pool 3, June 9, 1960; D-longitudinal profile of Pool 2, August 3, 1960; E-longitudinal profile of Pool 2, February 7, 1961; and F-cross section of Pool 2, February 7, 1961. ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 23

60o ~ STATONI III and V. Light intensities at Station I ~~ ' r K - STATION III .. ^ V ;l.. k - '' o _. . -.STATION were similar to those at the 2 downstream stations only in January and December (Fig. 12). Light intensities varied from date to date, and on the same day at different stations, because of changes in angle

1 of the sun and cloud cover. There was an I I I I J.. F e Y AWE X <. oG M. T DEC. increase in incident light at all stations in February and March, but as the riparian measure- FIG. 12. Averages of incident-light in April, levels of ments at 3 sampling stations on Doe Run, Meade foliage developed County, Ky., 1960. light ceased to rise, and then declined throughout the summer. This seasonal se- In Pool 3, a deposit of gravel, sand, and quence was most pronounced at Station I, silt at the upstream face of the dam gave where large, overhanging trees completely the cool underflow an inclined surface over shaded the water most of the day, and an which to rise over the dam in summer. In understory of smaller trees and shrubs con- Pools 1 and 2, however, there were no such tributed to shading and scattering of in- deposits, and water rose vertically about 4 coming light. As the leaves fell in October and 7 feet, respectively, to clear the tops and November, the incident light at Station of the dams. Movements of water through I increased until it corresponded with Pools 1 and 2 in the manner shown in seasonal decreases at other stations. less Figure 11A was occasionally seen, but CHEMICAL ENVIRONMENT because frequently than in Pool 3, probably Dissolved Oxygen of the differences in vertical lift. Calcula- tion of velocities required to lift the density Concentrations of dissolved oxygen at current over Dam 2, using formulae of Station I were rarely less than 70 per cent Churchill (1958), indicate a range from of saturation, and exceeded 85 per cent on 0.4 to 1.0 ft/sec, varying with temperatures a few occasions (Fig. 13). Prior to the of the overlying water and the volume of period of high discharge in 1961 (see Fig. discharge. These values are somewhat 8), the dissolved oxygen concentrations at higher than those indicated in the data for Station I were relatively constant, but in- time of flow (Table 4), and the discrep- creased slightly in periods of highest dis- ancies probably reflect the sporadic occur- charge, and were lowest in periods of rence of the summer density currents in the modal or submodal discharge. After the pools, factors such as mixing of tracer dyes, flood in March 1961, the dissolved oxygen and accumulated error in both sets of data. content of the water at Station I became Stratification, or stratified flow, occurred less predictable, and the highest and lowest in some of the larger natural pools of the values, 9.12 mg/l and 4.3 mg/l, were ob- lower part of Doe Run, but temperature tained after that flood. The reason for differences were less pronounced than in these changes is unknown. the impoundments. Similar stratified flows Diurnal fluctuations in oxygen concen- were found in pools of a small creek near trations were erratic in samples obtained Lexington, Ky., by Neel (1951), and by directly in the spring source. However, on Slack (1955) in a creek in Indiana. In both July 22-23, 1961, the oxygen content 60 instances the streams were smaller than ft downstream increased during the day- Doe Run, with a greater pool-riffle ratio, light hours and decreased at night to and a lesser discharge. the levels of samples taken at the source (Fig. 14A). These diurnal variations must Incident Light be attributed to photosynthetic activities of There was much less light at Station I a relatively pure stand of the moss Fissi- for 10 months of the year than at Stations dens julianus (Mont.) Schimp., which cov- 24 WILDLIFEMONOGRAPHS

,s* ,ag

! 1. i r zc,'".

i,oo ., i; a1- n 1

iX'kO 1-bE I

FIG. 13. Levels of saturation and amounts of dissolved oxygen at 3 sampling stations on Doe Run, Meade County, Ky.

ered about half the stream bottom in that moderating extremes in dissolved oxygen area. below the dam. On June 7, 1960, when Upstream from Dam 1, prior to the high saturation values were extremely high discharges of 1961, there were large beds above Dam 1, dissolved oxygen content of of vegetation that quickly overcame the water in riffles below the dam was 10.43 oxygen deficit of the spring waters in peri- mg/I, about 103 per cent saturation. Con- ods of high insolation, and often produced versely, on August 7, 1961, when saturation extreme supersaturations. On June 7, 1960, values of 84 to 96 per cent prevailed above oxygen values ranging from 12.58 to 19.13 Dam 1 (Fig. 14B), saturation values below mg/l (saturations of 124 and 191%, respec- the dam varied from 97 to 103 per cent tively; water temperatures from 14.0 to (mean of 3 samples, 100%). The free fall 15.0?C) were obtained in Pool 1 between of water over small dams is a significant Miles 1.3 and 1.6; the average of 10 samples factor in maintenance of oxygen levels was 14.79 mg/l at an average temperature downstream in some polluted streams; flow of 14.4?C. After most vegetation in the of water clinging to the face of a dam is upstream areas was washed out, and at less effective in aeration (Barrett, et al., night, samples above Dam 1 retained an 1960). The latter phenomenon presumably oxygen saturation deficit, and after a few occurred in Doe Run in periods of ex- hours of darkness the oxygen values at Dam tremely low discharge, such as October 1 dropped nearly as low as those at Station 1961, when saturations between 85 and 90 I. This probably resulted from decomposi- per cent were present in riffles downstream tion of leaves and other detritus in the pool, from Dam 1, and water was flowing down respiration of plants and animals, and re- the steep face of the dam rather than fall- placement of water in Pool I (Fig. 14B). ing freely over it. The nearly vertical fall of water of about Samples from Mile 2.0, near the down- 7 feet over Dam 1 acted significantly in stream end of a riffle area (Fig. 2), gen- ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 25

NCx&N ,?N, r vN tain 100 per cent saturation downstream. APri 16, 1960 StationI Supersaturations were found most often in relatively water at the surface of A quiet CLOUDO 09 D5ARKNEDS%MO'90 |?HKENCLO9UOS I pools in summer. Density currents, when present, exhibited some increase in dissolved in their X oxygen passage through the A'- pools, but were not as great as those in Octob., 20-21, 1961 Station I warm, quiet lenses of surface water influenced by photosynthesis of aquatic vege- A, I..- , I tation along the banks. Dissolved oxygen content was more variable at Station III, where there were 2 seasonal highs in 1960 (Fig. 13). The first occurred in June and July and may be

2NO?D10 6ir.s ARKNESS.%" ,v fSTrM9 CLOUDS 0 explained by increased photosynthesis to - through rapid proliferation of algae and macrophytes, and the second occurred co- i.o incidently with cooling of water in autumn, / .9 9 following a long period of modal and sub- 7.0 -SePtember 15-16I Station IV _ I / / C 17 * 1 -. modal discharge. This latter high concentration is attributable to the cooling of the FIG.14. Diurnalvariation of dissolvedoxygen water and increasing photosynthesis result- at Station I (A), Mile 1.65 (B), and Station IV ing from reduced turbidities and relatively (C), Doe Run, Meade County, Ky., on various constant stream conditions. Depression of dates. Solid lines connect the absolute values; dotted lines, levels of saturation; and X's denote dissolved oxygen in September 1960, just values obtained directly in the source. prior to the surge in October and Novem- ber, is not easily explained. It may have erally yielded concentrationsof dissolved resulted from a combination of high water oxygen slightly above saturation. The high- temperatures and the first heavy accumu- est level recorded there was 13.5 mg/l in lation of fallen leaves in pools. In the October 1960 (134% saturation), and the high-gradient sections at Station III and lowest level was 9.21 mg/l in July 1961 between Stations III and IV, turbulence of (93%). The average of 15 determinations the many riffles and cataracts would seem was 10.31 mg/I, an average of 103 per cent to preclude oxygen depletion. However, in saturation. The highest amounts of dis- a 24-hour study on September 15-16 at solved oxygen were present in periods of Station IV, only 1 determination with more low discharge and high insolation, when than 100 per cent saturation was obtained, thick stands of Myriophyllum heterophyl- and most values were lower than 85 per lum Michx. and Fissidens were growing cent (Fig. 14C). The high oxygen demand most rapidly, and low discharge allowed of leaf accumulations was demonstrated more time for water in pools and eddies to when the oxygen content rapidly decreased become supersaturated. The lowest values with shading by storm clouds. A similar occurred in periods of high discharge and "cloud effect" has been noted in other turbidity, when physical and biological streams (Hornuff, 1957; Denham, 1938). aeration are least effective (Denham, 1938; Oxygen depletion as a result of decomposi- Churchill, 1958). tion of leaves in pools of small streams has Relations of dissolved oxygen in Pools been found in Indiana (Schneller, 1955; 2 and 3 were similar to those in Pool 1, with Slack, 1955), Illinois (Larimore, et al., supersaturations in daytime and slight def- 1959), and elsewhere. It is probably of icits developing at night. Passage of water common occurrence in smaller streams over the dams, however, tended to main- when accompanied by reduced discharge 26 WILDLIFEMONOGRAPHS and high water temperatures in autumn. result from discrepancies in stone-surface- Covering and shading of many beds of to-water-volume ratios. vegetation by fallen leaves, especially in At Station I alkalinities were highest in the shallow, upstream parts of Doe Run, late summer, autumn, and early winter presumably decreased oxygenation by in- periods of modal and submodal discharge, hibiting photosynthesis. and the related carbon dioxide and pH Downstream from Mile 5.0, the increased values were highest and lowest, respec- number and depth of pools (Table 1) tively, at those times. The amount of free allowed more accumulation of silt and carbon dioxide was less and 40 mg/l on few organic debris than in more upstream sec- occasions during periods of extreme dis- tions. In addition, a lack of extensive vege- charge (compare Figs. 8 and 15). The tation, higher summer temperatures, and diluting effect of the passage of large lentic conditions resulting from backwaters volumes of surface water through the of the Ohio River, caused decreased aquifer feeding Doe Run was obvious in amounts of dissolved oxygen at Station V depression of alkalinities, and often in ele- (Fig. 13). The seasonal curves represent- vation of pH values above 8.0. ing oxygen concentrations at Station V are Alkalinity and pH at Station I fluctuated more typical for streams (Butcher, et al., randomly with time (Fig. 16A). Down- 1930; Ricker, 1934a; Reid, 1961), in that stream from the source, pH, alkalinity, and the greatest amounts of oxygen were pres- carbon dioxide levels varied with dis- ent in winter as a result of water temper- charge, photosynthetic activity, temperature ature-oxygen saturation relationships. After change, and deposited organic debris. Of impoundment of Doe Run in July 1961, course, turbulence caused physical loss of seepage from the banks influenced deoxy- carbon dioxide from the water, but decom- genation of water near the mouth of the position and accrual from tributary springs creek. The seepage water was black and supplemented the amounts present. Physi- malodorous, apparently as a result of de- cal loss due to stream turbulence was not composition of layers of leaves that had as much as expected. On June 7, 1960, pH been buried previously by sedimentation values exceeding 7.3 were found only in from backwaters of the Ohio River. quiet backwaters in Pool 1, on the shoreward side of beds of Nitella (L.) Carbon Dioxide, pH, and Alkalinity flexilis Ag. In those areas, pH measurements Subterranean waters are notoriously rich greater than 8.0 were obtained (maximum in carbon dioxide collected during passage 9.1) indicating total use of free carbon of the water through a soil atmosphere dioxide by the plants (Ruttner, 1960). more generously supplied with that gas Alkalinities in the backwater areas ranged than the epigean atmosphere (Foster, 1942). from 175 to 200 mg/l, but no precipitation Combination of carbon dioxide with water of carbonates, expected with the marked results in formation of carbonic acid, which reduction in carbon dioxide, was found. then causes relatively insoluble carbonates However, Tindall (1962) found limited to go into solution as bicarbonates. A deposition of carbonate on Nitella flexilis dilute solution of bicarbonate, at equilib- from Doe Run. Alkalinities in the main rium, is generally basic in pH, but in some currents of the pools were reduced from springs, as in the Doe Run system, carbon 224 mg/l at Mile 1.2, to 214 mg/l near Dam dioxide is in excess, causing slightly more 1, but no consistent increase in pH was acid conditions (Shoup, 1948; Hem, 1959). detected (range 7.1 to 7.3). The slight re- This occurs when the amount of available duction of alkalinity in the main current gas is greater than the available carbonate, must have resulted primarily from mixing and in open solution channels like those of water from the sides, but the small presumably feeding Doe Run, this could amount of change emphasizes the relative ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 27

i -h-----~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~------| _ X _ - TO

FIG. 15. Dissolved carbon dioxide, hydrogen ion concentration (as pH), and methyl purple alkalinity at 3 sampling stations on Doe Run, Meade County, Ky. independence of lateral water masses from amounts of carbon dioxide enough to main- the main current in a stream. tain equilibrium and prevent deposition of Fluctuations in and were pH alkalinity 12 12 t 2 not so pronounced at Mile 1.65 on August July 22-23,1961 A 16-17, 1961, as they presumably were be- t.2 fore most of the vegetation was washed 220 210 - /0 -\X 20 away. they were erratic in time, Moreover, - with no obvious relation to daylight (Fig. a Statio0nI 16B), and may represent the net physical O2 obr 20-1 , 1961 - loss of carbon dioxide in the pool, and 1o.. . " \ - / 7 fluctuations at the main spring. The values were modified accrual of undoubtedly by M3'O*- 1 o0 from in Pool co carbon dioxide decomposition . .._ _x _F :: 2f.. >/. :..F*AIVI?C ...... Z:.::::10 _lcTru.l 1, but this was minimal because much of the detritus had been removed by high waters in spring 1961. At Mile 2.0, pH values ranged from 7.0 to 8.3 in 22 samples, with high alkalinities in of dis- persisting except periods high FIG. 16. Diurnal variations of pH and methyl charge. High pH values occurred in pe- purple alkalinity at Station I (A), Mile 1.65 (B), riods of low discharge and pronounced and Station IV (C) on Doe Run, Meade County, photosynthetic activity, and in periods of Ky., on various dates. Dotted lines connect pH lines connect and X's and values, solid alkalinities, (in high discharge. Decomposition respi- A) denote pH values obtained in the immediate ration in Pool 2 apparently bolstered the source. 28 WILDLIFEMONOGRAPHS carbonates. As noted before, accumulation unrecognizedform of soluble carbonate in of marl was obvious on the face of Dam 2 naturally hard waters. Hastings, et al. (Mile 2.36), but not upstream from there. (1927) and Eyster (1958) found that phos- Because samples from Doe Run usually phates in solution allowed conditions of were obtained in daylight, the influence of apparentsupersaturation of bicarbonatesto photosynthesis and warming of the water exist, and sewage effluents containing con- caused most series to show a reduction in densed phosphates and other materials alkalinitybetween StationsI and II. Never- inhibited deposition of carbonates in an theless, some samples on cloudy days, and English stream (Edwards and Heywood, at night, exhibited a slight increase in 1960; Heywood and Edwards, 1962). Elu- alkalinity in that section (on the order of cidation of processes involved in mainten- 5 to 15 mg/1), suggesting solution of car- ance of supersaturationsof bicarbonate bonate from the stream bed. is a study that may be of majorimportance Downstream from Dam 2 (Figs. 1 and in understandingthe carbon dioxide rela- 2), the balance between carbon dioxide tionships of limestone streams. and bicarbonate was disrupted causing Reduction in alkalinitybetween Stations precipitation and accumulation of car- I and III was most pronounced during bonate. There were marl accumulationsin periods of modal discharge and high water the upstreamtwo-thirds of Pool 3, but the temperature. Diurnal variationsin alkalin- bottom in the downstream third was silt ity and pH were marked in the single 24- and organic muck, with little carbonate hour study of the marl area (Fig. 16C), present. Again, as in Pools 1 and 2, an in- indicatinga significantinfluence of biologi- crease in carbon dioxide indicated by a cal activity in daylight hours. lowering of pH, occurredfrom the head of Downstreamfrom the marl area, carbon Pool 3 to near the dam. On June 9-10, dioxide content varied inversely with dis- 1960, the pH range in the pool was from solved oxygen. Data for Station V indicate 7.8 in the upstreamend .to slightly less than a slight increase in alkalinityat modal dis- 7.6 just above the dam. charge when compared with data from The relationshipsof alkalinity, pH, and Station III (Fig. 17), and on other occa- carbon dioxide in the marl area are exem- sions as in periods of high discharge and plified by data from Station III (Fig. 15). concurrentpresence of backwaters of the There, pH values were seldom below 7.5, Ohio River at Station V. This increase in and often exceeding pH 8.0. Similarsitua- alkalinity was accompaniedby increase in tions in a Virginia stream (Steidtmann, carbon dioxide content, and decrease in 1935), and in lakes in New York (Deevey, pH. Some fluctuations in pH at that sta- et al., 1954) and Germany (Ohle, 1934, tion and elsewhere may be attributableto 1952), were interpreted as indicating organic acids entering from seepage along presence of a stable, colloidal form of car- the banks (Wherry, 1920; Shoup, 1950). bonate, which allowed to supersaturations Total Iron exist. To test for such colloids, 6 aliquots of water from Station III were filtered The iron in Doe Run presumably origi- through "HA"filters about an hour after nates through leaching of ferrous com- collection. The original determination of pounds (primarilyferrous carbonate) from 239.5 mg/l alkalinity persisted after 3 fil- soil and limestones by anaerobic and car- trations, with pH at the beginning of bon dioxide-rich ground water, and from filtration ranging from 8.1 to 8.3. It must iron-richsoil particles carried as turbidity. be concluded, therefore, that if colloidal Determinations of total iron of 0.31 and carbonate existed in Doe Run, it was 0.36 mg/l at StationsI and III, respectively, smallerthan 0.45 micronin diameter. Rutt- were reduced to 0.02 and 0.03 mg/l after ner (1948) suggested the presence of an 3 filtrations through "HA" filter discs. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 29

I t V phates with resumptionof modal discharge 240 in June (Fig. 18). Slight, but consistent,

230 reduction in phosphate occurred from upstream to downstreamin the 1960 period of modal discharge, indicating some utilization of the material. Slack also 210 (1955) Avoer?* AlkalintityA noted a gradual downstream depletion of otult AI ago . CS. phosphate in 2 Indiana streams. In times of relatively high discharge, 22C phosphates were higher at the downstream stations than at Station I, although the in- Ii creased variability at downstreamstations

20C obscuresthe relationship. These data agree with what might be expected because re- S S '9C duction of phosphate occurred when con- 3 centrationsof dissolved solids were s! greatest and during periods of greatest growth of attached algae and macrophytes. Increases in phosphates downstream in winter may be attributedto decompositionand accrual from runoff and seepage. Phosphate phosphorus and total iron usually were present in inverse proportions in Doe Run (Fig. 18). Small amounts of were bound FIG. 17. Longitudinal variations in pH and phosphate probably with iron methyl purple alkalinity at various levels of dis- through oxidation of ferrous iron to some charge, Doe Run, Meade County, Ky. A-4-sam- form of ferric phosphate as in certain lakes ple series, flow 60 cfs at Station I, samples taken at turnover (Einsele, 1938). This is borne between 10:00 AM and 2:00 PM; B-5-sample out filtration series, flow 60-350 cfs at Station I, samples taken by experimentsthat resulted as in A; and, C-5-sample series, flow 40 cfs at in lowering of phosphate values by only Station I, samples taken between 11:00 AM and 25 to 29 per cent (from 0.32 mg/l to 0.23 2:00 PM. mg/l) accompanied by an almost total removal of iron. These results indicate a These values correspondto the 0.02 mg/l considerable range in the presumed iron solubility of ferric hydroxide given by phosphate component of total phosphate Stumm and Lee (1960) for highly oxy- phosphorus. genated waters. Some of the phosphate may come from Total iron was most abundant during commercial fertilizers used in the water- periods of high discharge and turbidities, shed, but a lack of correlation between reflecting the iron in suspended sediments. high phosphate and high discharge seems Trends in iron concentrationswere similar to preclude this. Probably, the phosphate at all stations; no significant differences is being slowly leached from detrital ma- existed between stations in the overall terial deposited in the limestones amounts (Graf, present (Fig. 18). 1960b). PhosphatePhosphorus Nitrate and Nitrite Nitrogen Phosphate phosphorus was highest at Concentration of nitrate nitrogen gen- Station I in periods of modal discharge in erally diminished throughout the study 1960. This resulted from concentrationor (Fig. 18). When data for the 3 periodi- lack of dilution of the subterraneanwaters. cally sampled stations are compared,there In 1961, there was little increase in phos- is a decrease in amounts of nitrate from 30 WILDLIFEMONOGRAPHS

r. - STATIONI .--- STATIONIII .-.... STATIONV

I II I I I t t 1 FIG.o f18 IAtsN. taJ M noN p. . WE. OCT. r OCtInit I U. SEPT. ..

FIc. 18. Amounts of total iron, phosphate phosphorus, nitrate nitrogen at 3 sampling stations, and amounts of nitrite nitrogen at Station I on Doe Run, Meade County, Ky. All values are given as mg/I. upstream to downstream, indicating some sewage from ruralfacilities has been known biological utilization over the length of the to pollute many such water tables (Foster, stream. Some decrease of nitrate was 1942). The decrease in nitrate throughout found in the course of an Indiana creek the study, however, implies that the rela- (Slack, 1955), and the changes were of the tively high levels found earlier were a re- same order of magnitudeas in Doe Run. sult of fertilizer application somewhere in Nitrite nitrogen was consistently less the drainage area, and subsequent move- than 0.1 mg/I, and because of its low con- ment into Doe Run through surface runoff. centrationwas to considerableerror subject THEBIOTA in determination. Nitrites were highest at the beginning of the study (Fig. 18), but Algae and Higher Plants those data have resulted from early may In small, swift streamssuch as Doe Run, lack of precision in technique at that time. benthic algae are of primary importance Although the amounts of nitrate and since true phytoplankton rarely occurs nitrite nitrogen in Doe Run generally are (Eddy, 1934). Blum (1956a, 1957, 1960) of the magnitude found in natural, unpol- provided excellent reviews on the ecology luted waters, there is presumptiveevidence of algae in flowing waters, and his papers for organic pollution of the aquifer from were used extensivelyin preparingthis sec- domestic sources. I know of one instance tion. Identifications of algae were made in Meade County where sewage was using keys of Drouet (1959) for the blue- shunted directly into a flowing subterra- greens, and of Tiffany (1930), Tiffany and nean stream, for convenient disposal, and Britton(1952), Prescott (1951), and Thomp- ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 31 son (1959) for other groups; abbreviations precarious substrates afforded by water- of the names of authorities are from those falls, swift chutes, dams, and areas subject works. Identifications of many diatoms to continuous impact by falling water. The were provided by Charles Brohm (pers. distribution of Lemanea began where Bat- comm.), who made a special study of them rachospermum left off, and ranged from in Doe Run. the face of Dam 1 to the swift, marl-en- The most usable method of classifying crusted riffles at Station IV. In the down- benthic algae from my point of view lies stream portion of this range, Lemanea was in the study of aspect (Symoens, 1951), or limited to the swiftest waters and was habit of growth, through consideration of seldom found in the marl area except on structural modifications of the species con- lips of cataracts where it was abundant, cerned. Algal groupings of Doe Run fall and where turbulent water caused physical into 4 broad categories, 2 of which were removal of deposited carbonate. In riffles modified from Cedergren's (1938) classifi- of the marl area, growth of Lemanea ap- cation: (1) those with a firm mode of peared limited in summer by rapid accu- attachment and a highly flexible thallus, mulation of marl after velocities were occurring in currents; (2) cushionlike or reduced and the creek assumed modal dis- encrusting species that grow appressed to charge. Greatest proliferation occurred in substrates in current; (3) matted or loosely winter and spring. arranged forms occurring attached or un- Cladophora glomerata (L.) Kuetz. ap- attached in quiet pools, eddies, or in areas parently represents, with its associates, a of almost laminar flow; and (4) subaerial stream community found throughout the species attached as encrusting mats in world in suitable habitats (Blum, 1956a). zones of spray, on waterfalls, on emersed In Doe Run, C. glomerata was rare at Sta- stones, or on muddy banks and bars along tion III, but was abundant downstream from the channel. These categories are not en- Mile 5.6 where it often covered all avail- tirely satisfactory since some ubiquitous able substrates in swift, gravel-bottomed species occur in more than one, or there riffles at modal discharge, with exception were gradational habitats where species of narrow channels directly in the swiftest were "out of place." currents. After the banks at Station IV and above were cleared of Attached Rheophilic Algae riparian vegetation, C. glomerata became the dominant alga in Large, macroscopic sexual thalli of most marl riffles where it shaded the en- and Lemanea Batrachospermum sp. spp. crusting communities completely, and ap- were of the flora major components algal parently killed them. This extension of the of the upstream 5 miles of Doe Run at cer- range presumably was in response to tain times. However, Chantransia stages greater amounts of light. Blum (1957) of both algae contributed most significantly noted that Cladophora in the Saline River, to communities of encrusting forms with of throughout the year. Batrachospermum Michigan, disappeared leafing trees in an was most abundant as macroscopic thalli riparian spring, indicating effect from October through March each winter, of shade and a high-light requirement by and occurred in sparse stands throughout the alga. The greatest abundance of C. the summer of 1961; it was absent as glomerata in Doe Run was correlated most macroscopic thalli in the summers of 1959 closely with modal discharge rather than and 1960 (Minckley and Tindall, 1963). with season (except in the indirect way The alga was restricted to the areas ad- that discharge was seasonal); however, jacent to springs in the Doe Run system, when large beds were found in winter, a distribution attributed to the generally they were invariably on unshaded riffles. constant conditions there. C. glomerata is markedly affected by dep- Lemanea, on the other hand, inhabited osition of silt in the thallus because of the 32 WILDLIFE MONOGRAPHS filtering action of the closely packed tenue occurred throughout late winter and filaments (Blum, 1957). Siltation was fol- spring in Tributary II-a, and also in lowed frequently by mechanical detach- Tributary I-j in April 1961. Generally, C. ment of the thalli during periods of modal elegans was attached to leaves, twigs, and discharge, causing initiation of a renewed bedrock where flow was shallow and rela- period of local growth. This was materially tively slow. C. incrassata, having a more augmented by accumulation of marl on compact, gelatinous thallus, occurred in some partially emersed thalli. swifter waters, but was also found adjacent Cladophora fracta (Dillw.) Kuetz. grew to colonies of C. elegans. Stigeoclonium on mill dams and in rapids adjacent to was an uncommon component in most col- dams in the upstream 3 miles of Doe Run. lections of Chaetophora. These algae ap- It was perennial in those habitats, and often peared very vulnerable to washout, but occurred, along with various encrusting rapidly recolonized open, sunlit areas, at- forms, on shells of the snail Goniobasis. taining maximum development in 2 to 4 The alga also colonized tags used to mark weeks. various fishes at Station II. It should be No collection of algae from Doe Run emphasized here that the pattern of dis- lacked diatoms, and filamentous genera tribution of C. fracta and C. glomerata in such as Melosira and Cymbella occurred Doe Run is in direct contrast with findings abundantly in swift waters throughout the of Anderson and Lundh (1948). They re- creek at certain times. Diatoms also made ported that C. fracta substituted for C. up a major part of the epiphytes on larger glomerata when the velocity of flow de- algae and higher plants, with the genera creased, and that C. fracta occurred down- already listed, plus Diatoma and stalked stream with C. glomerata most common species of Gomphonema. upstream. Encrusting Rheophilic Algae Characteristic associates of Cladophora in Doe Run were members of the diatom Nostoc verrucosum Vauch. was one of the genera Cocconeis, Diatoma, Synedra, and most widespread algae in both the main- Cymbella. Also, there were many species of stream and tributaries of Doe Run. It was blue-green algae and green algae, and the most abundant in spring on smooth stone rhodophyte Audouinella violacea (Kuetz.) or concrete tops of mill dams, and also as Ham. present. Associates of Cladophora in an epiphyte on Fissidens near Mile 0.5, and other parts of the world are discussed by on marl near Mile 5.6, in May 1960. A Margalef (1960) as "Series 'Cladophoreta- hard, compact alga, possibly Palmella mini- lia." ata Liebl., also was epiphytic on mosses at No other attached green alga was more Station I and in Tributary IV-b. It had the abundant than Cladophora in the main- superficial appearance of a Nostoc, but was stream. Others of frequent occurrence were composed of small, ovoid cells separated Microspora stagnorum (Kuetz.) Lagerh., by definite sheaths. Ulothrix zonata (Weber and Mohr) Kuetz., Many noncalcareous, encrusting algae and Basicladia crassa Hoff. and Tild., the are subaerial in some instances, but through latter from the carapaces of turtles. rapid development underwater in periods Green algae with thick, gelatinous sheaths of greater-than-normal discharge they re- that also may be included under this general tained fine silts and colloids as films on grouping, but were rare in the mainstream, other substrates. During June 1960, I ex- were: Stigeoclonium tenue (Ag.) Kuetz., amined film accumulations for the 2-week Tetraspora gelatinosa (Vauch.) Desv., and period following high discharges. One day Apiocystis brauniana Naeg. after the peak discharge the thin film, Chaetophora elegans (Roth) Ag., C. in- slightly more than 1 mm thick, was incrassata (Huds.) Hazen, and Stigeoclonium habited by large numbers of ciliates and ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 33 bacteria. Four days after the flood, diatoms Films as described above, thick, gelati- were the dominant organisms; Synedra, nous layers of diatoms, and colonies of blue- Achnanthes, Nitzschia, and many species green algae, were systematically grazed by of Navicula were present. Bacteria and Goniobasis at all stations when present. protozoans were not as obvious as before, Grazing by other aquatic organisms was and a few trichomes of an undetermined rarely seen in Doe Run, but has been found blue-green alga were on the surface of the to limit algal abundance in a number of 1-mm-thick film. After 9 days, Phormidium freshwater situations (Brooks, 1954, 1955a, autumnale (Ag.) Gom. was abundant as a b; Douglas, 1958; and others). thin, irregular layer over the surface of the Where substrates were submerged more film; the average thickness of the overall than a few inches in swifter waters, and deposit was about 1.5 mm, and many mostly in upstream areas (Stations I-III), diatoms were present beneath the blue- thin mats of Oscillatoria splendida and 0. green alga and among its loosely twisted amoena (Kuetz.) Gom. sometimes devel- trichomes. In 12 days, the common orga- oped. 0. splendida was more abundant in nism was P. autumnale, and a few other flowing water, however, than 0. amoena, species, including some diatoms, were pres- which usually grew in eddies and pools. ent on the surface. Oscillatoria splendida In the marl area of Doe Run, in places Grev. was present, and living diatoms were other than swift cataracts, chutes, or the abundant inside the covering layer of faces of dams, a calcareous, encrusting Phormidium. Fourteen days after the community of algae occurred. Scraping the flood, the mat was above water, but was poorly consolidated surface layers from a being saturated by water splashing over "fresh" marl exposed an underlying layer the stone. Phormidium autumnale was of greenish-yellow algae that included growing rapidly in the subaerial habitat, Phormidiumincrustatum (Naeg.) Gom., P. and there were no other algae at the sur- tenue (Meneg.) Gom., Oscillatoria sp., face of the mat. The Phormidium layer Gongrosira lacustris Brand, and rarely, was about 1 mm thick, and the overall film Chlorotyliumcataractum Kuetz. On stones deposit was 3 mm, including the alga. The recently recolonized after being exposed by next time the area was visited the film had low water, Gongrosira formed small, dried, flaked, and most of it had fallen into heavily calcified colonies that proliferated the stream. laterally to coalesce and develop a firm, Films of algae and amorphous "sludge," crustose layer. Phormidium subsequently presumably developing in the same man- invaded these colonies and became the ner, were common in Doe Run after dominant species in "mature" marls of the periods of high turbidities, and sometimes stream. In swifter areas, trichomes of blue- at modal discharge. Fritsch (1929) re- green algae appeared unable to persist corded encrusting communities of Phormi- projecting from the surface of the mass, dium autumnale and P. retzii (Ag.) Gom. and the pseudoparenchymatous Gongrosira in English streams, and suggested a suc- remained as the dominant organism. Fritsch cession from Chamaesiphon to Phormidium ( 1949, 1950) also noted invasion of Phormi- at his stations. Likewise, Nelson and Scott dium in developing marls, but into the tufts (1962) noted development of "microphyto- of Schizothrix that made up the bulk of the benthoic ooze" in summer in the Middle calcareous communities studied by him. Oconee River, Ga., and other such ooze Phormidium supposedly causes calcareous deposits have been recorded in streams by crusts to become "especially stony and re- many workers. Such deposits may afford sistant [Blum, 1956a]." a profitable line of research in understand- Other algae in calcified, encrusting coming stream ecology, at least from the stand- munities of Doe Run included many dia- point of productivity. toms, the Chantransia of Lemanea and 34 WILDLIFEMONOGRAPHS

Batrachospermum, various Oscillatoriaceae, the banks of the creek, often were marly to Cladophora, Vaucheria, and sexual thalli of the touch, probably as a result of evapora- Lemanea, but all are considered incidental tion of water broughtinto the mat by capil- to marl formation. Audouinella was rarely lary action (Gessner, 1959), and with taken in marls of the main channel, but was invasion of blue-green algae sometimes common in marl nodules on twigs and other were transformedinto soft, porous"biscuits" debris near dams, and once in spray from of marl. Epiphytes on Cladophora, Le- a cataract at Station IV. Gomotia sp. (prob- manea, and other large algae, on higher ably G. holdenii Coll.) was found on plants (rarely), and algae living on shells Goniobasis at Stations I and II, but snails and skeletons of invertebrates,apparently from Stations III and IV usually were en- resulted in the encrustations found there crusted with Gongrosira and Phormidium. (see Young, 1945). Little information is available on the Organisms contributing to marl deposi- ecology of deposition of stream marls, even tion in streams include members of most though they have been recorded in both divisions of algae and many higher aquatic recent and fossil deposits in North America plants (Gessner, 1959; Welch, 1935). It and elsewhere (Graf, 1960a, b). The larg- seems probable that many species involved est accumulations known to me occur in in the formation of stream marls require small, spring-fed streams of the Arbuckle currents because of an "inherent current Mountains, Okla., and are more than demand"(Ruttner, 1953; Whitford, 1960b). 100 feet thick and half a mile long (Emig, Marls from Doe Run, brought into the 1917). Other, larger deposits of this type laboratoryor displaced into backwatersof are known in central Europe and in other the creek, were quickly overgrown by parts of the world (Gessner, 1959). species characteristicof silted bottoms or Marl formation by activities of plants sediment films (e.g., Phormidium autum- may occur in 3 different ways (Gessner, nale, Oscillatoria spp., and diatoms). It 1959): (1) plants may physically increase seems logical that the need for rapid re- the surface area, allowing greater contact newal of films of water near thalli or marl- of the water with air and enhancing rapid encrusted algae, at least, and consequent loss of carbon dioxide and evaporation; (2) maintenanceof a high diffusion gradientof plants may extract carbon dioxide from the nutrientsnear the plant (Whitford, 1960b; water and thereby cause precipitation of Whitfordand Schumacher,1961) would be carbonates; or (3) they may extract carbon even more importantfor those half-buried dioxide from bicarbonates in the water, species than for the filamentous species elevate the hydroxyl-ion concentration, and usually cited as demonstratingsuch a "de- promote carbonate deposition. Deposition mand" (Batrachospermum,Lemanea, Cla- of carbonate on a given plant necessitates, dophora, etc.; see Whitford, 1960a, b). of course, that the plant grow faster than Also, the current assists in removal of silt marl accumulates, except where evaporation deposits and the preventionof such deposi- and loss of carbon dioxide, or evaporation, tion. are involved. Mosses were important in the Formation of marl on a given substrate deposition of marl by the first method in through algal activities occurs by secretion the Doe Run system, primarily in some or precipitationof crystals of monocarbon- tributaries and on the faces of Dams 2 and ate between algal filaments or as tubes 3. Marl formed in this way was soft and around the filaments (Blum, 1956a; Gess- tufalike, crumbling rapidly when subjected ner, 1959). In algae that grow in tufts, to freezing and thawing, or exfoliating in such as Schizothrix,crystals accumulate at slabs and shattering on contact with the the bases of the tufts, making it more com- ground. Mats of larger algae, such as pact than the terminal areas where marl is Cladophora and Vaucheria stranded along soft and relatively amorphous. In the ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 35

pseudoparenchymatousGongrosira of Doe Run, a similar effect was obtained by accumulationof crystallinecarbonates in and on the thallus of the plant, with trichomes of Phormidiumextending out of the main mass that allowed marl accumulationsimi- lar to that formed between projecting filaments of Schizothrix (Fig. 19). In this way, the algal mass grew in a lamellate way, with the greatest proliferation of Gongrosira in spring and early summer, and a following development of more shade-tolerant blue-green algae in mid- summer and autumn. As a result, dense highly organic marl alternated with the soft, light-colored bands formed by blue- green algae (Fig. 20). This general pattern was borne out by field observationson marl deposits in Doe Run, but there were many variables, such as flaking of surface layers on exposure to weathering by low water levels. There is significant physical erosion of marl from poorly consolidated surface accumulations,indicated by denser marls being found in swifter waters, and solution of carbonates must occur when decomposing leaves cause high carbon dioxide and lowered pH in the areas of marl deposition. The downstreamend of the marl area in Doe Run has been discussed from the standpoint of water chemistry, and seems to result from reestablishment of certain carbon dioxide and pH relations, perhaps through accumulationof organic materials in large pools. However, there is the possibility that the sharp line of demarcation near Mile 5.6 was more a function of the drop in gradient and a resultantunsuitabil- ity of the area for the algal species instru- mental in marl formationbecause of some factor related to decreased turbulence and C velocity of flow. Gongrosirawas not found FIG. 19. Marl algae from Doe Run, Meade downstreamfrom Mile 5.6 and Phormidium County, Ky.: A-photomicrograph of Gongrosira incrustatum was very rare in collections lacustris digested from marl with N/50 HCI; B- photograph of a "living marl" from Station IV from Station V. Greater amounts of light, illustrating the accumulation of carbonates in higher water temperatures,and abundance "tufts"; and C-composite sketch from photo- of the larger Cladophoradownstream could micrographs of marl algae illustrating the projec- tion of Gongrosiraand Phormidium from the marl also contribute to, or indicate, the unsuit- and some aspects of marl accumulation. ability of that area for encrustingalgae. 36 WIDLIEuEMONOGRAPHS

A C

B D

E FiG. 20. Marl concretions from Doe Run, Meade County, Ky.: A-gross appearance of a dried, "living marl" nodule from Station IV; B-gross appearance of a fossil marl from Station II illustrating the number of insect "burrows"revealed after some weathering in the stream; C, D-gross appearance of a fossil marl from Station II that had been subjected to weathering on the banks, and a cross section of the same nodule illustrating lamellate deposition of carbonate around another, highly weathered, marl nodule; and E-Goniobasis from Station III illustrating various stages of marl encrustation. ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 37 It is noteworthy that larger macrophytes diverse organicmaterials. In autumn,when in Doe Run, including MyriophyUum, leaf accumulations blocked riffles of the Potamogeton, and Nitella in the upper marl area, the leaves often were cemented reaches, and various emergentforms down- together forming low dams across the stream, showed little marl accumulationat stream. These were removed quickly with any time, although all have been recorded increased discharges, but fragments of with marl accumulations in other areas, leaves and leaf impressionswere common. especially lakes (Welch, 1935; Wetzel, Other detrital materials included twigs, 1960). Most of them lived in relatively seeds of terrestrial plants, and shells of swift waters in Doe Run, and Wetzel noted terrestrial snails in marls from near Sta- that water movement, such as wave action, tion IV. Twenty-fivenodules, rangingfrom was a majorfactor in preventionof surface 2 to more than 18 inches in diameter,were encrustations on large, flexible aquatic sectioned, and Goniobasis shells were the plants. center of aggregation of 7 of these. Most Marl accumulationcan be rapid in some nodules were formed aroundsmall pebbles, situations (Gessner, 1959), and was con- previously weathered marl, or twigs, but siderably more rapid in Doe Run than 4 showed no apparent center of aggrega- recorded by Blum (1957) for the Saline tion. Examinationof the various layers of River, Michigan (4-5 mm per year). No the nodules revealed few recognizablealgal specific studies were made in Doe Run to components. Apparentlythese are decom- ascertainthe rate of deposition,but 63 con- posed, reconstituted, or replaced by min- secutive rings of deposition were counted eralization. in a massive block of marl exposed around Marl nodules from Doe Run ranged from roots of a riparian sycamore tree that the size of small pebbles to more than 24 toppled in a windstorm of June 1960, and inches long and 18 inches wide. The smaller the rings ranged from 20 to 55 mm in ones were spherical, or followed the out- width, with an average of about 25 mm. lines of the object on which they were If these were annualrings of depositionthe formed, but larger ones tended to be uni- rate of accumulation was quite high. formly oblong in shape and flattened verti- Amountsof marl found on aquatic inverte- cally when lying in place. In many riffles brates also indicate rapid deposition. Many the bottoms consisted.of a totally cemented of these animals cast their exoskeletons conglomerate of marl, rubble, and gravel regularly,yet marl 1 mm thick was present (Minckley, 1961), and thicknessesof 2 feet on some stonefly naiads. In June 1960, a were not uncommon. partially submerged can was found with of Areas and marl deposits ranging from 1 to 4 inches Algae Quiet Slow Currents thick on the submergedportion. The total Blue-green algae were the most abun- weight of the can and wet marl was 3.5 dant benthic forms in pools, eddies, and pounds, and there was no evidence of rust backwatersdownstream from Dam 1, and or corrosionfrom exposureto water. Metal possibly in the impoundmentabove Dam 1 field equipment corroded in less than 2 at certain times. Of these, the most fre- months if left in the water continuously. quently encountered were Oscillatoria There was evidence in the period of my amoena, 0. splendida, and sometimes study, and from the presence of abandoned, Phormidiumautumnale. Lyngbya birgei G. marled stream beds lateral to the stream M. Smith occurred as isolated trichomes; channel, that accumulationsin swift riffles however, L. patrickiana Drouet and all caused shifting of the stream from side to others listed above were found at one time side, at least within the limits of the stream or another in relatively pure stands on silt bed. and detritus substrates. Mats of Oscilla- Marl formations in Doe Run included toria often were seen rising to the surface 38 WILDLIFEMONOGRAPHS and being carried downstreambecause of common in backwaters, where they as- accumulation of gases during periods of sumed a loose, erect branching growth high photosyntheticactivity (Blum, 1956a, form quite in contrast to the thick, felted 1957, 1960). Some buoyant mats contained mats in flowing waters. Vaucheriausually substantial numbers of small tendipedid washed into quiet areas from upstream. larvae, snails, and once a number of oli- One thriving colony of Vaucheria was, gochaetes. Thus flotation not only dis- however, found in the old millrace at Sta- perses algae downstream (Blum, loc. cit.) tion III growingon a massof iron hydroxide but may function in the downstreamdrift from 4 to 8 inches thick. Hormidiumkleb- of aquatic invertebrates. At times of modal sii G. M. Smith and 2 desmids, Netrium sp. dischargeand clear water in Doe Run, mats and Closteriumsp., were also found there, of Oscillatoriasometimes developed on the but in a place where no precipitated iron bottoms of pools in more than 7 feet of was present. water, causing them to appear deep, black- Riffles, and associated algae, were rare ish green in color. Samplesof bottom muck upstreamfrom Dam 1, although the rate of from downstream parts of Pools 2 and 3 flow other than in the impounded portion had surfaces covered by a thin, macro- usually exceeded 1 ft/sec in the main cur- scopic layer of Oscillatoria in June 1960, rents. This rapid flow tended to be smooth but samples at the same time from Pool 1 because of the uniform bottom and the did not have such algal growths. A thick relatively deep channel, and permitted the stand of 0: limosa (Roth) Ag. occasionally development of a distinctive flora of algae was found adjacent to a small sewage out- and higher aquatic plants. The swiftest fall near Station II. Sedimentation,or inflow was dominated by Fissidens and creased turbulence and velocity of flow in Batrachospermum(the latter seasonally), eddies and pools during spates, was quick but in long periods of modal discharge in to smother or displace algal growths of the summer and autumn of 1960 and 1961 these and the following kinds. (Fig. 8), and the autumn of 1959, dense Downstream from Dam 1, Spirogyra mats of Vaucheria geminata developed, varians (Hass.) Kuetz., 2 or 3 undetermined sometimes choking the channel. Whitford species of Spirogyra, Oedogonium sociale (1956) included stands of Vaucheria in Wittr., and a species of Mougeotia were Florida springs with "rapids"communities found in backwaters. Spirogyra demon- of higher plants rather than with any spe- strated little apparent habitat preference, cific algal community,and Butcher (1933) with the exception of usually being com- also mentioned Vaucheria in association monest in areas of open sunlight. Oedo- with his "torrential"flora of mosses and gonium was rare. Downstream from Mile Lemanea. 5.6, Mougeotia grew abundantly on the Maximal development of Vaucheria in shoreward side of dense Cladophorabeds Doe Run was from Mile 0.1 to 0.5. Growth when the latter were present, and was began in backwaters, downstream from usually heavily clogged with silt and epi- beds of watercressor behind logs and other phytic diatoms. debris, but mats sometimes grew inde- As already noted, filamentous and en- pendently in the channel, apparentlyfrom crusting diatoms were abundant in more the germination of zoospores, zygotes, or lentic areas of Doe Run. These usually from vegetative parts buried in the sub- were most noticeable after periods of win- strate. Germinatingzygotes of Vaucheria ter discharge, and were rapidly overgrown that had developed weak, rhizoidal fila- by blue-green algae and grazed by aquatic ments in addition to caenocytic thalli often invertebrates with onset of lower dis- were found in collections. These rhizoidal charges. Vaucheriageminata (Vauch.) D. filaments seemed to penetrate the softer C. and V. sessilis (Vauch.) D.C. also were substrates,and may have served to anchor ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 39 the developing thalli. Silt was rapidly de- Growth of Dichotomosiphon occurred from posited in beds of Vaucheria, even in permanent mounds of silt densely inter- periods of low turbidity, apparently as a woven with colorless, rhizoidal filaments. result of the filtering action of the densely The photosynthetic portions of the plant interwoven thalli. This may have assisted emerged from the substrate as short, inter- in fixing the beds to the stream bottom. woven, but erect, filaments on the surface The bottom inch of mats was usually com- of mounds that grew radially out from the pacted with silt, some of which was held by center. Dichotomosiphon did not wash out the rhizoidal filaments, but much was de- as severely as Vaucheria during floods, nor posited among living filaments of the was it damaged by overgrowth of Vaucheria plants. With development of massive when that occurred. growths, however, these attachmentswere Even when Vaucheria was essentially insufficient to hold the plants in current, absent from the channel of the upper area, and detachment often occurred, with the algae other than Dichotomosiphon were mats folding over one another. Slight in- uncommon. Encrusting communities on creases in discharge also detached massive larger objects in the stream bed were mats, rolling them as one would roll a dominated by Chantransia of Batracho- thick rug, sometimes blocking the channel spermum, and rarely by diatoms, in the until the rolls were moved against the area where Vaucheria became seasonally banks. abundant. The few riffles were dominated At times of maximum development, by Fissidens. This general lack of diversity Vaucheriaobviously grew over and buried of algal species may be related to low light other plants in the upstreamsection. Often intensities in the area (Fig: 1), to the the stream assumed a braided channel be- shifting nature of the bottoms (see Ceder- tween masses of Vaucheria,and the speed gren, 1938; Douglas, 1958; Margalef, 1960), of flow was markedly increased in these perhaps to the lack of diversity in environ- smaller channels. The accumulation was mental conditions and an accompanying rapidly removed by increased discharge lack of diversity of available niches, or to in winter, and large masses of algae were other unknown factors. washed downstreamor deposited on riparian objects. These masses remainedalive as Subaerial Algae long as parts were in water and kept moist These algae developed most, and con- by capillary action. Fruiting of Vaucheria tributed significantly to the flora of Doe did not occur in submergedmats, but was Run in periods of high discharge. Phor- commonly found in emergent mats which midium autumnale was the major species provided vegetative stock for subsequent that formed extensive mats in zones of periods of modal discharge. Vaucheria spray, but there were some mats of Oscil- beds similar to those in Doe Run, but ap- latoriaamoena, 0. splendida, and 0. tenuis parently not so extensive, have been de- Ag. Some large, filamentous green algae scribed in alkaline streams elsewhere and red algae, such as Cladophora fracta and Lemanea occurred 1957; Diatoms spp., respectively, (Symoens, Hornung, 1959). in zones of as a result of found in association with spray, usually those beds were seasonal decreases in water levels at onset similar to those found with Cladophora, of modal discharge in spring, but were un- there and in Doe Run. able to maintain themselves in other than Dichotomosiphon tuberosus (A. Br.) saturated conditions. Audouinella was Ernst was also abundant in the upstream characteristic of spray zones in late winter areas of Doe Run, but was more limited in and spring at mill dams, occurring as marl- its distributionthan Vaucheria. It formed encrusted colonies on twigs. Chlorotylium, thick, felted mats in silty areas downstream Gongrosira, and Phormidium incrustatum as far as Mile 0.75, and was perennial. also were found there, and Cladophora 40 WILDLIEEMONOGRAPHS fracta, Spirogyra, Hormidium klebsii, and L. between Stations II and III, and water Lyngbya taylori Drouet and Strick. were willow Justicia americana (L.) Vahl be- entangled on the surface of the nodules, or tween Miles 4.5 and 7.0. with filaments of other species. Diatoms, Beds of watercress were most highly de- sometimes making up thick gelatinous veloped in the spring tributaries of Doe layers on stones and twigs, became exposed Run, but extensive stands developed on after periods of high discharge, persisted sand and silt deposits lateral to the main in the spray as water levels fell, and dis- channel in the upstream portion of the appeared within 2 weeks after the water mainstream, and in the midst of stream- receded. drifted logs and branches at Station II. Gleocapsa sp. (= Anacystis sp.; Drouet, Downstream from Station II, beds of water- 1959), Cylindrospermummusicicola Kuetz., cress were present in winter and spring, Oscillatoria agardhii Gom., 0. tenuis, but became rare in summer and disap- Lyngbya patrickiana,and Nostoc commune peared when stranded by lowered water Vauch., occupied mud and sand bars along levels. Reproduction by seeds was most the main channel and many were found on evident in spring, when many young plants clay banks several inches above the water. grew along the banks submerged on de- Some aerial forms may have been included posits of silt and sand. Few seedlings, in samples from the creek because of high however, persisted to maturity because of water levels, but if such were the case it washout. The largest beds in the main- was not obvious at the time of collection. stream resulted from accumulation of Nostoc commune was abundant in tempo- stream-drifted plants in eddies, and prolif- rary pools along the floodplain of Doe Run, eration from broken stems and from plants as well as in the stream. Sphaerella lacus- lying on the bottoms. The beds rapidly tris (Girod.) Whttr. was found in an en- accumulated allochthonous and autoch- cysted stage on rock outcrops of Tributary thonous debris from the current, sometimes II-b and in another unnamed tributary near so extensively that it formed black, mal- Station IV. Likewise, Calothrix parietina odorous muck beneath larger beds. Nas- (Naeg.) Thuret. was in subaerial habitats turtium washed out readily when subjected on bedrock on Tributary II-a. to strong currents, and the large hollow stems floated in the water and were Emergent Higher Plants high transported for some distance downstream. The higher emergent plants along Doe Washout was somewhat prevented, how- Run included several species that occurred ever, by the occurrence of Nasturtium beds rarely, or were restricted in distribution. behind larger debris and on the inside of These were: Typha L., angustifolia repre- bends and by the accumulation of bars of sented by a sparse stand near Station III; organic and inorganic material that tended Sagittaria latifolia Willd., at Tributary II-d; to direct currents elsewhere. Watercress and the sedges Eleocharis compressa Sull., in the mill or in areas downstream E. obtusa (Lilld.), E. acicularis (L.) R and pools, from Station C, Carex spp., and Scirpus validus Vahl II, usually became lax, pros- that occurred sparsely in downstream areas. trate except for the terminal parts, and flowered earlier Smartweeds, Polygonum spp., also were than did plants near the scattered over bars downstream from Mile springs. Freezing affected the downstream 5.6. Equisetum variegatum Schleich. and plants more severely than those near the E. arvense L. both were abundant and springs. Water of the outflows caused functioned to stabilize banks and some microclimatic conditions that allowed con- sand and gravel bars along the stream. The tinued growth of plants in winter. The major emergent plants, however, were tops of larger plants were frost-killed, caus- watercress Nasturtium officinale near ing proliferation of subterminal branches springs, forget-me-not Myosotis scorpioides and generally thicker beds in winter and ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 41

spring than in summer and autumn. source, and a predominantly rubble bot- The forget-me-not, although more ter- tom; a long area of shallow, particulate restrialthan watercress (Thompson, 1944), substrateover bedrock and relatively swift more or less replaced that species between currents;a transitionarea where deposition Dam 1 and Station III. Forget-me-nots of larger gravels and sands occurred; and usually occurredin associationwith a mint a semilentic situation in Pool 1, with sand Mentha sp. and various grasses, but were in the upper half grading to silt and muck most abundant on the streamwardside of near the dam. This area was inhabited, the community,whereas mint was thickest respectively, by: (1) Fissidens julianus on the shoreward side. Myosotis grew that also occurred on solid substrates in largest in areas of open sunlight (to 15 more downstreamareas; (2) seasonally by inches high), and flowered throughout Vaucheria, an alga capable of rapid pro- spring, summer, and autumn. In heavy liferation and not needing deep substrates shade the plant rarely grew more than 5 in which to root; (3) by Potamogeton inches high and was not observedto flower. foliosus Raf., var. macellus Fern., capable It would be interesting to know what of maintaining a foothold in coarse, un- native riparianplants, if any, were replaced stable substrates by massive development by Nasturtium and Myosotis scorpioides. of subsurfacerhizomes; and (4) by Nitella Both were introduced to North America flexilis in the sand-silt areas caused by ag- from Europe (Fernald, 1950), where they gradation in the mill pool. Dichotomo- are part of the common streamsidevegeta- siphon tuberosus,discussed above, is a form tion (Gessner, 1955). capable of maintainingits own substrateby Waterwillow was scarce in Doe Run, compact matting of subsurface filaments, distributed in colonies downstream be- and was found throughoutthe upper 0.75 tween Miles 4.5 and 6.0. The plant ap- mile. Batrachospermumwas also a major peared in May, flowered, and usually was componentof the communityin its seasonal gone by Septemberor October;it occurred occurrence on solid substrates throughout on the gravelly parts of riffles, seldom in the area (Minckley and Tindall, 1963), marl, and may have been limited in its and Nasturtium officinale was the major distributionby the presence of marl. emergent form. Downstream from Dam Fissidens SubmergentHigher Plants 1, again was a major part of the community, The charophyte,Nitella flexilis, usually is occurring with Myriophyllum heterophyl- included with algae in studies of this kind, lum in a mosaic patternapparently dictated but is included here for convenience be- by the presence of bedrock riffles and ex- cause of its habit of growth. Potamogeton tremely swift waters for the moss, and pool Raf. and Callitriche sp. oc- diversifolius situationsfor the milfoil, usually with cur- curred rarely in Doe Run. The former was rents 1 ft/sec, however. Nastur- represented by a near the exceeding single clump tium was a minor of this com- inflow of TributaryI-c', but was absent at component the next visit and none was found subse- munity, quickly colonizing exposed bars quently. The species is commonin sinkhole and developing stands lateral to the major ponds of the surroundingarea. Callitriche currents. was found at 2 places in the mainstream, Nitella flexilis was the subject of an once in a pool near Station III and again intensive ecological study in Doe Run by in Pool 2, but was abundant in February Tindall (1962), and only an outline of its 1961, in a small spring-fed swamp con- relations is presented here. The greatest nected to Tributary IV-a. proliferationof Nitella occurredin periods In Doe Run, the upstreamarea presented of modal discharge;after spates other than a grading environment downstream, with those of 1961, beds of Nitella in Doe Run swift, turbulent water adjacent to the resembled illustrations given by Gessner 42 WILDLIFEMONOGRAPHS

swiftest currents throughout Doe Run. It could, however, persist and grow in relatively slow waters and on gravel in periods of modal discharge. It proliferatedrapidly early in those periods only to be covered later by Vaucheria or other plants. The moss often was damagedby high discharge, and removed down to its holdfasts,but the beds rapidly developed from the basal portions. In a sense, Fissidens was one of ------Ic-lc- -_.------the most constant inhabitants of the upstream section, and probably had a much wider distribution upstream from Dam 1 prior to impoundmentand consequent covering of the bedrock bottoms of that area by various sediments. F. julianus fruited rarely in late winter and early spring. Re- production was primarily by vegetative proliferation from basal portions of the plant. The largest beds of Potamogetonfoliosus grew on elongate hummocksor mounds in FIG. 21. Illustration of the formation of a the channel (Fig. 21). Most reproduction mound of bottom materials by a higher plant in by P. foliosus in Doe Run resulted from Doe Run, Meade County, Ky., and its partial wash- spreadingof rhizomes or by rooting of de- out (bottom). tached plants in downstreamareas. Flower- ing was noted in almost all seasons, but (1955) for stands of Paspalum repens Berg appeared most pronouncedin late summer in the Orinoco River of South America. and autumn. No seedlings were found. Beds in the main channel had been re- Myriophyllum heterophyllum formed moved and currents had trimmed project- thick, channel-blockingbeds in the swift ing portionsof plants growing on the inside area between Dam 1 and Mile 2.0 where of bends into an almost vertical, convex it replaced Fissidens in areas of rubble and form that followed contours of the banks. in "pools." The thick, adventitious roots of 1961 ex- The extreme flooding nearly allowed for rapid enlargementof beds, and tirpated N. flexilis from Doe Run, but by mounds were formed similar to but not so 1962 it was from February developing clear cut as those in beds of Potamogeton. remnants left in the sediments, the few was amazingly resistant to remaining attached plants, and possibly Myriophyllum from of Nitella con- washout, and rarely was damaged by high germination spores. other than in habitats. sistently grew on soft, silt-sand deposits discharge peripheral and did not downstreamfrom Dam 1 The beds of Myriophyllumwere almost as grow in at any time despite substantial amounts extensive in the rapids areas early 1962 carried there by floodwaters. This charo- as they were at the beginning of the study, phyte is most often found on silty bottoms notwithstanding the severe flooding of of clear, sluggish streams, where streams 1961. Reproduction occurred by adventi- enter lakes, and sometimes deep in lakes tious rooting, rhizomes, and by rooting of (Wood and Muenscher, 1956). It appears broken plants. No seedlings were found, adapted to low-light environments (Daily, but flowering occurred regularly in spring 1958). and summer after beds were exposed by Fissidens julianus usually grew in the receding water levels. When elevated from ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 43 the water, the leaf segments grew broad unpubl.), who called them "Schwaden," and serrate, quite unlike the submerged, and by Butcher (1933). The coarsest sedi- filiform leaf segments (Muenscher, 1944). ments in such mounds usually were on the upstreamends, serving to protect the struc- Spatial Relations of the Flora of the ture the current. In of Area against periods Upstream modal discharge, long "spits"of sand and Stream communitiesof plants, or mono- silt formed at the downstream ends of specific stands of higher plants or algae in mounds, and were rapidly colonized and streams, are governed primarily in their stabilized by adventitiousrooting of plants spatial relations by velocity of the current. trailing in the current,or by Nitella in the This determinesthe nature of the substrate, downstreampart of the Potamogetonarea. which, in turn, serves to determine the Also, mounds were formed by original ag- kinds of plants present (Butcher, 1933; gradation of material near a stone or log Blum, 1956a, et seq.). The persistence of with subsequent colonizationand stabiliza- a plant communityin a stream depends on tion of the area by plants. Those formed structuraladaptations of the species to the downstreamfrom Dam 1 by Myriophyllum situation, or if the community can modify were larger than those formed by Pota- its environmentin the amount of time it mogeton, possibly because of protection has to become firmly established. It is afforded by large stones in the stream bed, especially those plants with linear or finely or by greater width of the stream with divided leaves that occur in swiftly flowing more space for lateral encroachment by waters (Nielson, 1950a). Mosses, many of beds and for movementof water in periods which are typical of torrential streams of extreme discharge. In some instances, (Gessner, 1955), tend to flatten against the successionfrom Fissidens to Myriophyllum substrate and present a smooth surface to occurred with dense beds of moss on bed- current. Algae also demonstrate many rock being used for attachmentof adventi- similar structuraladaptations. tious roots of trailing milfoil. Subsequent Most larger plants in Doe Run modified accumulationsof sand and gravel around their substrates and encroached on unfav- the milfoil smotheredthe moss and allowed orable habitats by doing so. Plants, such expansion of mounds. Nasturtium also as Nasturtium and Nitella, growing along rooted on dense moss during low flow, but the banks, persisted after high discharges was removed by higher discharges. in an aggradingenvironment where stream- Spatial relations of plants in the upper carried silt and debris were deposited by part of Doe Run are to be presented else- sudden decreases in current. As the beds where, but the dispersionat the beginning enlarged they augmented physical slowing of this study is illustratedin Figures 22-24. of the water, and accumulationsof silt and The extensive vegetation in the upper part other materials soon formed bars lateral to of Doe Run in 1959 apparently had been and parallel with the currents. Prior to developing for at least 10 years without flooding of 1961, some bars apparently influence of discharges as great as those formed by Nitella in the upper end of Pool recordedin June 1961 (Curtis Brown, pers. 1 extended two-thirds of the way across comm.), when the stands were severely the channel on the inside of bends, and decimated. High discharges in June 1960 were more than 8 feet wide where the damaged the stands, to be sure, but recov- channel was straight. ery was rapid in the succeeding period of Myriophyllumand Potamogeton in cur- modal discharge. Furthermore,the data on rent soon formed elongate mounds through recolonization of the mainstream after deposition of silt, sand, and gravel among spates in 1959 and 1960 indicate that pat- the plants (Fig. 21). Mounds of this type terns of substrateand/or currentin the area were discussedby Gessner(1955, from Ruess, usually were changed so little after a major 44 WILDLIFE MONOGRAPHS

A A c c' 0 0

2I 4 . . 10 20 30 to0 20 o 0ES9 8'

2 . . 1o 20 30

' 0e D 0e Ffr. ,

. 2 I E E' 10 20 30 to 20 30

I 0 20 30

6 G' 0^

21 >t 2 2 . . t ' 10 20 30 10 20 30 0 o 30

LEGEND

xx BATRACHOSPERMUM DICHOTOMOSIPHON FISSIDENS POTAMOGETON VAUCHERIA I I VEGETATIONLACKING NASTURTIUM --- DIRECTIONof CURRENT

FIG. 22. Dispersion of aquatic vegetation in various areas upstream from Dam 1, Doe Run, Meade County, Ky., October 1959.

A A' c c

I 20 30 10 20 9 S tO0 ~ 2 3 1 0 0

1020 0203

10 20 30 ;o 22 10 20 30

40 4. - 4. I 10 20 30 MILE 50 1.1 LEGEND 140 Ft xx BATRACHOSPERMUM DICHOTOMOSIPHON i FISSIDENS NITELLA VAUCHERIA POTAMOGETON NASTIRTIUM II VEGETATIONLACKING ) DIRECTIONOF CURRENT

FiG. 23. Dispersion of aquatic vegetation in various areas upstream from Dam 1, Doe Run, Meade County, Ky., October 1959. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 45

n England (1960-1962), however, the higher plant communities usually were mosaics of small, monospecific stands, apparently developing in response to microvariations in current or substrate. The presence of a given stand of plants and the attribute of I0 most aquatic plants for developing exten-

0 MILE 1.7 sive beds from a single individual through rhizomes, adventitious rooting, and other vegetative means, would result in mosaic

20 4G 20 0 community structure in a mosaic environment. The mere presence of an established I ~30 ~ ~50 20 I 0 40 stand of aquatic plants is a deterrent to invasion by another species. However, the JO 20 30 40 50 influence of that plant may encourage

LEGEND colonization of the area by another species, XX BATRACHOSPERMUM NASTURTIUM resulting in a succession of sorts. 3 FISSIDENS E VAUCHERIA MYRlOPHYLLUM E | VEGETATIONLACKING The upper portion of Doe Run may be - DIRECTIONOF CURRENT compared with other, larger springs where FIG. 24. Dispersion of aquatic vegetation at essentially monospecific stands of vegeta- Station II, Doe Run, Meade County, Ky., October tion persisted for a long time (Odum, 1959. 1957a, b), and to lakes where the changes in substrate from stony, wave-washed shore- washout that recovery occurred in a pat- lines grade to silt-covered bottoms of the tern approximating the original. Indica- deeper littoral zone. In springs, steady- tions are that reestablishmentof the plants state conditions may prevail with constant after the scouringfloods of 1961 will repeat flow and ample supplies of nutrients main- the pattern once again, provided another taining remarkably stable subsurface mead- catastrophicflood does not occur too soon. ows. In lakes, edaphic conditions are The plant distribution upstream from primary factors influencing the dispersion Dam 1 differed from others described in of aquatic plants if depth relations permit the literaturein that the few species present (Pearsall, 1920; Wilson, 1935, et and occurred in stands in a seq.; relatively pure Roelofs, 1944). Temperature and light re- linear sequence. The distribution of the lations in lakes seem to have little effect thus the dis- sediments and currents, and unless related to tributions of the plants, were perhaps depth (Pearsall, 1920), and in communities result artificially induced to some extent by the changes plant constructionof Dam 1. Thus the area re- largely from specific sedimentation coupled flects the disturbanceof the streamby man. with certain chemical and physical condi- Be that as it may, the gradational picture tions that develop with the maturation of produced by this situation approximates the basin (Wilson, 1939). Doe Run lies the distribution of sediments and vegeta- somewhere between these extremes. It tion that might be found in the length of exhibits the more dynamic aspects of a much larger river if other factors were mound formation, washout and recoloniza- equal. It includes examples of 4 of the 5 tion, and development of a mosaic com- principal higher plant communities recog- munity or "mixing" of habitats for the nized by Butcher (1933) in streams of various species involved. The area between England. In streams studied by Butcher Dam 1 and Mile 2.1 may illustrate a more (1927, 1933), Roll (1938, et seq.), Ruess typical stream condition in the dispersion (in Gessner, 1955), Gessner (loc. cit.), and of the plants (Fig. 22), but retains a con- by R. W. Edwards and his colleagues in stancy usually not found in other studies 46 WILDLIFEMONOGRAPHS of stream communities of higher aquatic nematodes, common in silty bottoms and plants. It seems unfortunate that the study as parasites on larger aquatic plants; the of hydrophyte zonation in both lakes and nematomorph Gordius sp., rare; a few streams has been bypassed for more spe- Gastrotricha, Chaetonotus sp. and unde- cialized forms of ecological research. The termined species; rotifers, Keratella coch- latter studies are based largely on estimates learis Ehrenberg, Epiphanes sp., and unde- that must, through necessity, be derived termined loricate forms; conchostrachans, from more investigation of the former Cyzicus sp., found in drying pools near (Swindale and Curtis, 1957). Station V; undetermined ostracods; copepods, edax (Forbes) and Invertebrates Mesocyclops Bryocamptus sp.; and diverse aquatic mites Major obstacles to the study of benthic (Acri). Few planktonic invertebrates were animals in North American streams are the taken in limited sampling (see Pennak, lack of knowledge of taxonomy of most 1943). groups, the great diversity of habitats, and The following macrobenthic animals were the corresponding diversity of animals collected: about which little specific information is PLATYHELMINTHES available. Futhermore, the tendency for each species to reside in or to select spe- TURBELLARIA cific areas or habitats (Dodds and Hisaw, Dugesia sp. (near D. dorotocephala 1925b; Percival and Whitehead, 1929; [Woodworth]) Wene Mottley, et al., 1939; and Wickliff, D. tigrina (Girard) 1940), along with the concentrations and Phagocata velata (Stringer) movements of certain organisms in response Phagocatasp. (near P. subterraneaHy- to food, shelter, and other factors (Bad- man) cock, 1949, et seq.; Mueller, 1954; Roos, 1957; Macan, 1957; Armitage, 1961; Minck- ANNET IDA ley, 1963), present serious difficulties in OLIGOCHAETA are sampling. Stream-sampling techniques Unidentified branchiobdellids often were critical undergoing scrutiny by many found attached to the underside of the workers, but additional research is badly abdomen and inside the gill chambers of needed (see Leonard, 1939; Badcock, et Needham and crayfish (Cambarus). 1954b; Gaufin, al., 1955; Aelosomaspp. 1956; Macan, 1958; and others). Usinger, Nais spp. The most reliable to be technique appears Lumbriculusvariegatus (Mueller) the consistent use of a device by the given L. inconstans (F. Smith) same person (Surber, 1937, et seq.). Using this the data will at least be com- Chaetogastersp. approach, serpentina within themselves if local condi- Ophidonais (Mueller) parable "Lumbricus spp." This designation was tions at the time of sampling are taken into used for large, amphibious and terrestrial account. The benthos of Doe Run was earthworms that sometimes were common sampled in this way, and was supplemented beneath stones in shallow water throughout by selected sampling by myself and others. the stream system. Early in the study, a number of qualitative were examined for microben- Tubifex sp. samples Limnodrilus Claparede thic and a list of those is given hoffmeisteri organisms undetermined oligochaetes by Minckley (1962). Other microbenthic animals were Hydra littoralis Hyman, at HIRUDINEA Station II and in vegetated areas down- Placobdella sp. Leeches were attached stream from the Fissidens section between to the carapace, plastron, and fleshy parts of Station I and Dam 1; various undetermined large snapping turtles (Chelydra). ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 47

ARTHROPODA Isonychia sp.* I. sadleri Traver* ISOPODA Baetis spp.* Asellus bivittatus Walker*' B. vagans Traver* A. stygius (Packard)* Pseudocloeon sp.* Lirceus fontinalis Rafinesque* P. carolina Banks* Lygidium longicaudatum Stoller* Cloeon sp.* AMPHIPODA Stenonema sp.* Heptagenia sp.* Crangonyx forbesi (Hubricht and Mackin) * ODONATA Cragonyx n. sp. (near C. minor Bous- Agrion sp. (near A. maculatum IBeau- field)* vais) Cragonyx n. sp. (near C. rivularis Argia sp. * Bousfield) Gomphus sp. Gammarus minus (Say) * G. bousfieldi Cole and Minckley PLECOPTERA Synurella dentata Hubricht* Nemoura delosa (Banks) * Nemoura DECAPODA sp.* Allocapnia sp.* Cambarus bartoni (Fabricius) Perlesta placida (Hagen ) * Orconectes rusticus rusticus (Girard) Paragnetina media (Walker) * 0. pellucidus pellucidus (Tellkampf) Isogenus subvarians ( Banks) * COLLEMBOLA HEMIPTERA Podura L. aquatica Locally common Microvelia sp. (near M. pulc'hella upstream from Station IV, most often asso- Westwood) ciated with stream-drifted debris along the Gerris sp. banks but also in beds of watercress and G. remigis Say forget-me-not, and rarely on the surface of Metrobates sp. open pools and eddies. Predation by Gelastocoris oculatus (Fabricius) young-of-the-year creek chubs, Semotilus, Benacus sp. and water striders, Gerris and Microvelia, Belostoma sp. was heavy on populations in open water. B. fluminea Say Ranatra sp. EPHEMEROPTERA Notonecta undulata Say Ephemera sp.* undetermined Notonectidae E. simulans (Say)* Sigara spp. Hexagenia atrocaudata McDunnough undetermined Corixidae Ephemerella sp.* E. subvaria McDunnough* MEGALOPTERA E. needhami McDunnough Sialis sp. Leptophlebia sp.* Chauliodes sp. (near C. penticorn is L.) P. Paraleptophlebia sp. (near praepe- COLEOPTERA dita [Eaton]) Paraleptophlebia sp.* Peltodytes sp. Laccophilus sp. 1 Crustaceansand insects identifiedfor me by Agabus disintegratus Crotch others (see ACKNOWLEDGMENTS) are marked with Gyrinus sp. an asterisk.However, determined specimens were Psephenus herricki DeKay used in further identificationand interpretation of the fauna, so that I am responsiblefor any Stenelmis sp. errors. Optioservus sp. 48 WILDLIFEMONOGRAPHS

TRICHOPTERA Pomatiopsis lapidaria Say Glossosoma nigrior Banks* Goniobasis sp. Psychomyia sp.* PELECYPODA Diplectrona modesta Banks* Splaerium spp. Hydropsyche sp.* Pisidium H. betteni Ross* spp. Musculium sp. Represented only as Cheumatopsyche spp.* "fossil" shells washed from deposits along Agraylea multiplicata Curtis the stream, but living in the adjacent Ohio Pychnopsyche sp. River. Helicopsyche borealis Hagen Observations on the of Abundant DIPTERA Ecology Invertebrates Tipula sp. (near T. nobilis Alexander)* T. abdominalis (Say)* Data on the ecology of specific animals Pilaria tenuipes (Say) * in Doe Run varied greatly from group to Antocha sp. group because of the vagaries of field in- Paradixa sp.* vestigations. Descriptive adjectives indi- Anopheles punctipennis (Say) cating abundance of various kinds of Aedes sp. animals are based mostly on data in follow- Culex pipiens L. ing sections, but are somewhat subjective Chaoborus punctipennis (Say) and based on field experience, especially Simulium sp.* in species that were inadequately sampled Pentaneura sp.* by the quantitative methods attempted. Procladius sp.* Turbellaria.-Flatworms were abundant Psectrocladius sp.?* above the marl area of Doe Run and in Trichocladius sp.* spring tributaries throughout the system. Calopsectra sp. (near C. dissimilis Marl deterred development of large popu- Johannsen) * lations, but flatworms were numerous on Microtendipes sp. (near M. pedellus gravel bottoms between Mile 5.6 and Sta- [Johannsen]) tion V on occasions; no turbellarians were Tanytarsus sp.* found at Station V or below. I. (Tanytarsus) sp.* Phagocata velata was most abundant in T. (Stictochironomus) sp.* the mainstream near Stations I and II be- T. (Endochironomous) sp.* neath the upper surfaces of beds of Fissi- Cryptochironomus sp.* dens and on undersides of flat stones or Tendipes (Tendipes) sp.* logs, twigs, and other debris, but always T. (Limnochironomous) sp.* in relatively swift currents. At Station II, it was often associated with a turbel- T. (L.) modestus (Say)* larger larian, Dugesia dorotocephala?, which also Glyptotendipes senilis Johannsen* lived downstream from Station II, through- undetermined Empididae* out the marl area in small and Culicoides numbers, sp.* below. Similar distributions of these 2 Tabanus sp.* species were found in the larger tributaries. Atherix Walker* variegata Another species, D. tigrina, was present in Scatella sp. downstream parts of smaller tributaries, Limnophora sp.* throughout intermittent streams, and some- MOLLUSCA times in the downstream portion of Doe Run. GASTROPODA Oligochaeta.-Few notes were obtained on Physa spp. oligochaetes because of the difficulties in Ferrisia (Ferrisia) sp. identifying them in the field. Aelosoma ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 49 spp. were fairly common near Station I on by Walker (1961, 1962) as Asellus brevi- sand and silt bottoms, but were not taken caudus Forbes. elsewhere. Nais, on the other hand, was Minckley (1961) recorded Asellus styg- most abundant in silt deposits at Stations ius, a cavemicolous species, at 15 localities III and IV. Chaetogaster was found in silt- in the Doe Run system, including only 3 detritus deposits at Station IV. occurrences in caves. Since that publica- Tubifex sp. and Limnodrilushoffmeisteri tion, the species has been taken at 4 differ- frequently occurred downstream from small ent places along the mainstream of Doe sewage outfalls near Station II and Dam 1, Run. It was never abundant in the open and abundantly in thick deposits of decom- stream, but occurred frequently enough to posing leaves in pools. Larimore, et al. be considered a consistent, but minor ele- (1959) found tubificids the most common ment of the epigean fauna. A. stygius was animals in intermittent pools of Smith's most common in interstices of gravel beds Branch, Illinois, when heavy leaf accumula- and beneath large marl slabs in current. tions occurred in autumn. In Doe Run, the Lirceus fontinalis was the largest aquatic occurrence of Ophidonais and "Lumbricus" isopod in Doe Run, but rarely occurred in was also associated with fallen leaves, but the mainstream. In tributaries, however, those fed that dried those animals colonized wet, packed leaves especially by seepages in L. was the along the banks. summer, fontinalis only malacostracan It was not taken - Asellus an present. Isopoda. bivittatus, unusually from the marl area in the main- small was the most upstream aquatic isopod, abun- stream, but occurred in some dant upstream invertebrate collected in Doe Run. tributaries,and reached its abun- The greatest largest individuals reported by Walker dance beneath dry gravel and rubble of (1962) were less than 6 mm in total length. small islands and bars, where a slight flow The species is characterized by its small of water persistedin the interstices.Asellus size, by a general reduction of setation on stygius was also found in that habitat. In the body and appendages, and by its gen- autumn, small clumps of leaves collected eral dark coloration marked off by unpig- upstream from stones and other obstruc- mented bands crossing its body near the tions in the channel (termed "leaf packets" head and on the abdomen (Walker, 1961). by Badcock [1954a]), and leaf dams It appears to be adapted for life in closely formed on some riffles were rapidly colo- packed vegetation, and the greatest num- nized by Lirceus, amphipods, and aquatic bers were found in beds of Fissidens. Re- insects. The isopod was often seen moving production by A. bivittatus occurred about in open pools of Doe Run and its throughout the year but was most intense tributaries, especially when fallen leaves in April, May, and June, coincident with littered the bottoms. Reproduction oc- drops in winter levels of discharge and curred throughout the year, but seemed in resurgence of growth by flood-thinned greatest spring (Cole, 1957). a Fissidens. The number of young fe- Lygidium longicaudatum, semiaquatic per inhabited the immediate banks male was small, more than but isopod, of rarely 8, the creek and its most abun- females more as be- tributaries, produced young they in moist leaves, but also in beds of came Walker dantly larger. (1962) recorded a mosses and liverworts. When alarmed,the maximum of 18 young. A large, somewhat animal scuttled beneath debris on the aberrant form of A. bivittatus at Station V banks or into the water with little prefer- was most common in thin beds of gravel ence for either medium (Minckley, in Cole, overlying clay bottoms, and was extirpated 1959). Lygidium foraged on the tops of by impoundment of the lower part of the emersed beds of Fissidens and in clumps creek. That population was incorrectly of watercress 2 to 3 feet from the water's identified by Minckley (1961, 1962) and edge. 50 WILDLIFEMONOGRAPHS

Amphipoda.-Gammarus bousfieldi and G. (Hutchinson, 1959; Cole, 1961; and others); minus were the most abundant amphipods this problem is being studied. in Doe Run. Synurella dentata was the Although G. bousfieldi was not found in only other amphipod that occurred in sig- tributaries of Doe Run above the first nificant numbers in the stream system, and major riffle, it occurred in those with gen- Crangonyx spp. were represented by fewer erally smooth, low-gradient channels in the than 40 specimens in all collections made. upstream area (Tributaries I-a, I-b, Blue Gammarus minus was represented by a Spring, Buffalo Spring, and II-a) that had somewhat aberrant form in the mainstream thick beds of watercress and/or Myosotis. of Doe Run, but material from the steep- The species was often taken in bottom gradient tributaries to the creek and from samples or by seine, from smooth, appar- adjacent streams (Cole, 1957) appears ently undisturbed bottoms, and subsequent typical. G. minus lived in Fissidens beds observation indicated that it frequently at Stations I and II, but was rarely seen on burrows into soft, silty bottoms. It is a the surface of the beds and apparently strong swimmer, and cruises much of the moved about within the mass of vegetation. time in open pools or along the banks. When disturbed from that habitat, it darted After periods of high discharge, G. bous- into the current and back into the vegeta- fieldi and G. minus both appeared more tion (Cole and Minckley, 1961). In high- abundant at downstream stations than be- gradient tributaries, however, G. minus fore flooding, indicating a significant occupied almost all conceivable habitats: downstream "drift" (see Waters, 1961). in mosses on steep inclines and riffles, in After such spates, G. bousfieldi moves back interstices of gravel on riffles, in moist upstream en masse (Minckley, 1963), leaves far from the flowing water, and whereas G. minus either maintains its up- often in pools. This species joined Lirceus stream populations by rapid reproduction, in colonizing leaf packets in the main- or recolonizes upstream areas with more stream,a habitat utilized by G. pulex L. in subtle, short-term movements in the stream streams of Sweden (Badcock, 1954a, b), channel. and also was present in interstices of gravel Synurella dentata was abundant in Trib- bars and small islands downstream from utary III-b, and nowhere else in the system. Station IV, its typical habitat in areas of That stream had a bedrock bottom and was the mainstream that lacked vegetation. intermittent in summer. Cole (1957) found The typical habitat of Gammarus bous- the largest populations of the species in fieldi was pools, thus contrasting sharply and around Louisville, Ky., in spring- with G. minus in this respect. G. bousfieldi brooks with dense stands of watercress. He is a large, setiferous, banded species, most also noted the presence of Synurella and abundant in beds of emergent vegetation Lirceus fontinalis in shallow, rocky head- along banks near Stations I and II, and in waters of creeks, a habitat similar to Trib- mats of leaves, roots, and detritus along the utary III-b. banks and in pools downstream. The gen- Crangonyx spp. were scattered through eral restriction of habitats of both species the Doe Run System and no distributional in the mainstream of Doe Run, their gen- patterns are obvious. C. forbesi occurred erally close relationships (both belong to at Stations I, IV, and V; Crangonyx (near C. the subgenus Rivulogammarus), and the minor) at Tributary III-a"; and Crangonyx presence of a somewhat aberrant form of (near C. rivularis), beneath a slab of marl G. minus in the mainstream with "typical" at Station III. populations of that species in tributaries Decapoda.-The 3 species of crayfishes in not inhabited by G. bousfieldi, brings up Doe Run occupied distinct ranges in the the possibility of character displacement stream system, with some seasonal fluctua- and partitioning of niche by the 2 species tions. Cambarus bartoni was the largest ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 51 form, ranging up to 200 mm in length and nymphs and imagos from the mainstream 90 grams in weight (Minckley and Crad- of Doe Run are thought adequate for de- dock, 1961), and was found only in the limiting distributions, but only 7 tributaries immediate vicinity of springs in summer, were collected intensively enough to merit with juveniles and some adults ranging discussion. The greatest number of species downstream to Station IV in winter, and and individuals were found downstream also in caves of the area. Large C. bartoni from Dam 3 in and below the marl area foraged actively in daytime, but juveniles (Fig. 25). seldom were seen in the open. Large ani- Ephemera simulans, a large, burrowing mals resided in beds of watercress, in species, was found from Station IV down- Myriophyllum near Station II, or in de- stream to the mouth of Doe Run. Nymphs tritus. In summer, however, they presum- listed as Ephemera sp. probably are E. ably moved into extensive burrows in the simulans; they were small, or had indistinct banks. Downstream, the species lived in color patterns on their wing pads. Nymphs burrows in the stream bed in summer, lived in large pools and eddies, and always building crude chimneys near flat stones in gravel bottoms that contained consider- similar to those described by McManus able amounts of silt and sand. None was (1960) in New York. Young animals lived found on marl bottom. Ricker (1934a) and in Fissidens or Myriophyllum and in beds Ide (1935) both noted E. simulans as very of emergent vegetation along the banks; abundant in bare marl gravels of Ontario they were rare in summer. No female streams, with populations as high as 250 Cambarus found carried eggs, although per square foot in spring and early summer. young less than 10 mm long often were Adults emerged from Doe Run in late June taken in the mainstream near Station I. and throughout July, but always in small Orconectes rusticus was abundant down- numbers. stream from Station II, at least in the Nymphs of Hexagenia atrocaudata were mouths of all major tributaries below that collected downstream from Station IV on area, and throughout the permanent tribu- 2 occasions, and were relatively abundant taries. At Station II, 0. rusticus was pres- in the terminal pool of 7-Springs Branch at ent from Tributary II-a upstream to a low its confluence with Doe Run in June 1961. waterfall at Mile 2.0, but none was taken They had burrows in about 2 inches of fine farther upstream despite intensive collect- silt overlying bedrock in a pool just lateral ing in that area. Likewise, in spring and to a relatively swift current. summer, specimens of 0. rusticus were rare Ephemerella was one of the most abun- at Station II, apparently having moved dant genera of mayflies at Station II, repre- downstream into Pool 2. They returned to sented by E. subvaria, E. needhami, and a populate the area thickly in autumn and small, undetermined species. E. subvaria winter. Reproduction was obvious in April, was by far the most abundant, and the May, and June, but occurred earlier at the largest individuals inhabited the more open downstream stations than at Station II. 0. riffles with gravel bottoms. Smaller nymphs rusticus lived in diverse habitats, but was were in beds of moss or milfoil, generally concentrated in vegetation when it was on the upper surfaces in association with present. E. needhami and the undetermined species, The eyeless, unpigmented Orconectes however, they did not enter the bases of pellucidus was common in caves of the the plants as did crustaceans and some area, where it occurred with Cambarus other mayflies. Ide (1935) considered E. bartoni. A single specimen was taken in subvaria to have the lowest lethal tempera- the open creek about 25 feet from the ture threshold of any of 3 closely related source. species of Ephemerella. This was postu- Ephemeroptera.-Collections of mayfly lated on the basis of early emergence dates 52 WILDLIFEMONOGRAPHS

FIG. 25. Longitudinal distribution of ephemeropterannymphs in Doe Run, Meade County, Ky., and occurrences in some of its tributaries. The width of the line indicates abundance of a species, with the widest place where the greatest numbers were present. No comparisons between species abundance are intended.

when compared with the other species. warmertrout streamsin Michigan. In Doe The restriction of the species to the area Run, the nymphs swam actively when dis- of StationII (Fig. 25) may be related to its turbed, and usually were obtained from need for low developmental temperatures. mosses on the tops of stones. I found no imagos of E. subvaria. Large, active nymphs of Isonychia were The genus Leptophlebiawas represented prevalent in riffles from Station III to Sta- by a few nymphs from Station II down- tion V in late winter and early spring, and streamto below Station III. They occurred a few records were obtained at Station II in eddies and soft-bottomed pools, princi- (Fig. 25). Imagos of I. sadleri were found pally in the upper parts of mill ponds, but at Stations III and IV in June and July. also in dredge samples from the down- On the basis of general color differences, I stream,mud-bottomed part of Pool 3. Large believe that there were at least 2 species of numbers of nymphs also were present in Isonychia in Doe Run, one in the main- the downstreampool of 7-Springs Branch. stream (I. sadleri) and the other in the Paraleptophlebia spp. were most com- tributaries. Isonychia nymphs were almost mon, though rare at best, at the source, in never in quiet waters, but usually were TributaryI-b, and in Blue Spring. Nymphs abundantin small riffles lateral to the main usually are found in gravel-bottomed,swift currentsrather than in the swiftest currents. streamsof small size (Burks, 1953), and P. Tiny nymphs of the genus Baetis were praepedita was thought by Leonard and the most abundant mayflies in Doe Run, Leonard (1962) to be characteristic of occurring from Station II to riffles at Sta- ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 53 tion V. There were at least 3 species of current. In the upper marl area, and be- Baetis, of which B. vagans was most abun- tween Mile 7.0 and Station V, a few were dant and most easily recognized in the found lateral to the main currents under field. It was abundant in the marl area, logs, loose stones, and other debris. Fewer where it lived in the swiftest currents, in than 10 nymphs of Stenonema were ob- chutes of marl riffles, and directly on the tained at Station II during the study. upper sides of stones or gravel in riffles. Nymphsrepresenting species differentfrom At StationII, B. vagans often occurredwith those in the mainstreamwere abundant in nymphs of Ephemerella more abundantly 4 or 5 downstream tributaries. An avoid- on beds of vegetation in swift water than ance of springsis indicated by the absence on open, gravelly bottoms. The reverse was of Stenonema from the upper parts of true for the Ephemerellaspp. Percival and tributaries, and in Tributary I-b, Blue Whitehead (1929) found a direct contrast Spring, and Tributary II-a (Fig. 25). In in English streams with other species of small, downstream tributaries the animals those genera; recently hatched nymphs of lived side by side with Lirceus and various Baetis and Ephemerella occurred in beds insect larvaebeneath stones in mere trickles of aquatic vegetation, and along the banks of water. more often than larger nymphs. Harker Nymphs of Heptagenia sp. were ob- (1953) and Macan (1957) found smaller tained rarely at Station III, but were more nymphs of B. rhodani (Pictet) closer to the abundant in gravel bottoms overlain by banks and farther upstream than larger silt between Miles 6.4 and 8.0. individuals of the same species. Odonata.-Three genera of Odonata, 2 of According to Ide (1935), B. vagans has which were damselflies, were identified 2 periods of emergence in Ontariostreams, from Doe Run; however, after impound- one in early summer and the second from ment of the downstreamarea a rich drag- mid-Augustto autumn.B. vagans emerged onfly fauna developed in the lake as it from Doe Run in January and February, filled. Gomphussp. was the only dragonfly with perhaps a smaller emergence in June. collected from Doe Run and was taken only More likely, however, is an extended period at Station V where it burrowed in silty of emergence from January through June, bottoms of large pools. with early and late peaks (Murphy, 1922; Adults of Agrion maculatumwere found Leonard and Leonard, 1962). On occa- at all stations but I and along most tribu- sions, terminal-instarnymphs, seined from tary streams. They were most common at the stream in midwinter and warmed by Station II where naiads were associated sunlight for a few minutes, emerged in the with Myriophyllumin the currentand also net. in emergent vegetation along the banks, or Pseudocloeonsp. and P. carolinawere in in beds of detritus. Males actively de- swift parts of Doe Run at StationII, usually fended territoriesagainst other males near in association with aquatic plants or roots Station II and along the banks of Tributary and debris along the banks. P. carolina II-a each spring of the study. Females inhabited open riffles in association with oviposited on the undersides of Myosotis Baetis (Ide, 1935), or, when larger, along and Nasturtium leaves that hung in the the banks in shallow, rapidly flowing water water. (Burks, 1953). Argia sp. was rare and was found only as Flattened, sprawling nymphs of Steno- naiads in beds of emergent vegetation nema occurredfrom StationIII throughthe along the banks at Station III and in 7- marl area to about Mile 7.0, and in rubble Springs Branch. riffles at StationV. This distributionreflects Plecoptera.-Allocapnia sp. and Nemoura the importanceof rubble or other loose ma- spp. were rare or difficult to secure from terial to which the animals can cling in the mainstreamof Doe Run. Nymphs oc- 54 WILDLIFEMONOGRAPHS curred mostly in the tributaries and also secluded habitat that allowed less chance in leaf packets in the mainstream in for marl formation on the exoskeleton. autumn. Frison (1935) noted the occur- However,more frequent ecdesis could have rence of various species of Allocapnia in the same effect. leaf deposits of Illinois streams,and Hynes Other species of stoneflies in high-gra- (1941) noted certain Nemoura nymphs in dient tributaries to Doe Run were not English streamsmost often in leaf packets. identified. Large numbers of Nemoura delosa were Hemiptera.-Four species of surface-dwell- found at Stations III and IV from April to ing bugs were identified. Microvelia pul- June of 1960 and 1961. Adults of Allocap- chella?was commonand often mistakenfor nia and the other Nemoura were not ob- springtails when concentrated in quiet tained. areas. It usually was associated with Perlesta placida, considered the most stream-drifteddebris. Metrobates sp. was abundant and widely distributed stonefly low in numbers,but also occurredin quiet of Illinois by Frison (1935, 1942), was water or in beds of vegetation. The large commononly in tributariesof Doe Run, but Gerris spp. were bolder, venturing across occurred at Stations III and IV in the swift riffles and inhabiting other, relatively mainstream. Adults were captured from swift eddies and pools throughout the May through early July. stream. Of the 2 kinds present, G. remigis Paragnetina media was the largest and was usually downstreamwhereas the unde- most widespread stonefly of Doe Run, termined species was common upstream. ranging from Station I as a few tiny young Gerris spp. apparentlydid not seek shelter on 3 occasions to Station V. The species for overwinteringin Doe Run as reported was not found in tributaries,however, ex- by Hungerford(1920) in Kansas,probably cept in winter in the downstreamparts of because of favorable microclimatic condi- Tributary II-a and 7-Springs Branch. It tions maintainedthroughout the winter by was most abundantin the marl area, where the relatively warm spring water, at least it was often heavily encrusted with marl at the upstream stations. and covered by small colonies of Gon- Water scorpions, Ranatra sp., were taken grosira and Phormidium. Large nymphs on 4 occasions in seine hauls from Myri- lived in extremely swift water, often on ophyllum beds at Station II. All were in tops of smooth slabs of marl in rapids, or quiet water within 3 feet of the banks. in rubble or broken marl on riffles at Sta- Other bugs recorded from Doe Run were tions III and V. Its abundance decreased taken only after periods of high discharge. sharplywith the beginning of greater pool- Belostoma sp. was seen at Station II in to-riffle ratios and a lack of rubble or marl April 1960, and B. fluminea, Benacus sp., bottoms near Mile 5.3 (Tables 1-2), but it and several specimens of Sigara sp. were was again common in rubble riffles at Sta- taken at Station III in mid-June 1960. tion V. Adults were found onibridge abut- Notonecta undulata, undetermined noto- ments and streamsidevegetation from early nectids and corixids, and additional speci- April to June. mens of Sigara were found only after the The second largest and second most severe flooding of 1961 at Station I. All are abundant stonefly, Isogenus subvarians, abundant in sinkhole ponds of Meade was found only in the marl area, where it County, and presumably were displaced lived on the undersidesof stones somewhat into the creek during heavy rains. The lateral to the marl chutes in riffles with insects were concentrated in quiet water relatively uneven bottoms of rubble and near Station I. gravel. It is worthy of note that this species Megaloptera.-Two adult alderflies, Sialis was less often coated with marl than sp., were found at Station III in spring Paragnetina,perhaps because of its more 1960. The fishfly, Chauliodes, however, ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 55 was abundantin suitable habitats through- were not separated in the field and oc- out Doe Run downstreamfrom Station II. curred both as larvae and adults. Collec- Larvae were beneath gravel and rubble tively, they were found at all stations, but except in autumn when they lived among reached peak abundancein Fissidens beds fallen leaves along the creek. They were and on small, gravel-rubble riffles with often in interstices of dry gravels of small sparse moss cover at Station II. They islands and bars downstreamfrom Station were very uncommonin the marl area and IV, and in small tributarieslived beneath were almost always coated with marl stones and leaves long after flow had (Leonard, 1939). ceased. In the last situation,larvae usually Trichoptera.-Glossosoma nigrior was were found beneath large stones in small scarce at Station I, common at II, and in chambers. Pupae were not found in the most of the spring tributariesof Doe Run, stream bed, but occurred beneath stones but was rare downstreamin the marl area. and decomposinglogs lateral to the water's Larvae of Psychomyiasp. were even more edge. Pupal chambersof Chauliodes were rare, however, occurringin small numbers seldom more than 5 feet from water and at Stations I, II, and III. were immediately beneath a stone or log, Net-building hydropsychids were the in contrastto hellgramitelarvae (Corydalis most abundant trichopteranlarvae in the spp.) whose burrows for pupation are stream,occurring most often on riffles, marl sometimes 30 feet from water and at the chutes, tops of mill dams, and on beds of end of a definite tube (Chandler, 1956). submergent plants in swifter water. Di- Fishflies emerged the last week of May and plectronamodesta was the least commonof first 2 weeks of June in both 1960 and 1961. these, and was taken only downstreamfrom Coleoptera.-A number of aquatic beetles Station IV. This species has been taken was found only after the floods of 1961, only from small, heavily shaded brooks in including a species of Dytiscus, and 2 un- Illinois (Ross, 1953). Hydropsyche sp. determined dytiscid adults. Adult Pel- occurred rarely at Station I (a total of 7 todytes sp. were rare in beds of aquatic larvae), and with H. betteni was common plants at Station II, and no larvae were from Station II to riffles at Station V. obtained. Two resident dytiscids, Lacco- Cheumatopsyche spp., distributed in the philus and Agabus, were distributed from sameparts of the creekas was Hydropsyche, Station II to Station V as adults, with the was somewhat less common than Hydro- latter more abundant downstreamand the psyche downstreamfrom Station IV. former upstream, but larval Laccophilus A micro-caddisfly,Agraylea multiplicata, were taken only at StationsI and II in beds was the only insect other than certain of vegetation. Whirlygig beetles, Gyrinus, dipterans that was abundant at Station I. were rare upstream from Station IV, but The small, purselikecases of Agrayleawere were locally abundant downstream. As attached firmly to moss at Station I, and with water striders, Gyrinus was active to a lesser extent at Station II, in such a throughoutthe winter in Doe Run, appar- position that they lay appressedto the beds ently in response to the relatively warm and thus offered little resistanceto current. water temperatures. Sometimesthe cases were abundantenough Psephenus herricki was not present in to make the beds appear yellowish. Cases the mainstreamof Doe Run, but a series also were attached to small stones on the was collected in the downstream part of stream bed, to twigs or other debris, and Tributary III-i in spring 1961, a habitat once on open sand-silt bottom near Station later destroyed by impoundment. I; the few occurrencesat Station III were Elmids were the most widespread group on marl riffles, and the cases were heavily of beetles in Doe Run, represented by encrustedwith marl. Adults were obtained Stenelmis sp. and Optioservus sp. These from March through September. After 56 WILDLIFEMONOGRAPHS emergence, cases remained on the moss for characteristicof rich organic muds at the some time and caused difficulty in sorting edges of water, but usually in swamps, of bottom samples. marshes, or lakes (Alexander, 1931). An- Pychnopsyche sp. was obtained at all tocha sp. was common only upstream,and stations downstreamfrom Dam 1, but was inhabited open, relatively soft bottoms never abundant. It was most often in beds along the banks and in pools. of stream-drifteddetritus in eddies, where Paradixa sp. was taken only at Station its case was made of small twigs, pieces of IV, in the shallow, lateral riffles. bark, and other vegetative debris. At Sta- Mosquitoeswere quite abundantat times tions IV and V, however, the cases con- along Doe Run, and larval or pupal ma- sisted of a large proportionof small gravel, terial of Anopheles punctipennis, Culex with woody fragments attached to the pipiens, and an unidentified Aedes were posteriorhalf. obtained. The former was quite abundant Helicopsyche borealis was rare, being upstream from Dam 1, especially in thick collected once at Station III, and twice at beds of watercressor rafts of fallen leaves Station II. Empty cases were taken in along the banks. Anopheles punctipennis bottom samples from Myriophyllumbeds characteristically breeds in clear, cool, at Station II on 3 occasions,but subsequent shaded pools of streamsor brooks (Mathe- to my study, Helicopsyche was common at son, 1944). Culex pipiens, on the other Station II in August 1962. hand, usually breeds in rain-filled cans, Diptera.-Larvae of Tipula nobilis? and T. hoofprints,and the like, and is apparently abdominalis were among the largest in- unusual in stream situations. It was con- vertebrates other than decapods in Doe siderablymore rare than Anophelesin Doe Run. T. nobilis?was found only at Stations Run. Aedes sp. was recorded on the basis I and II where it was always intimately of 2 larvae from Station II. associated with Fissidens. The greatest Chaoboruspunctipennis usually is found emergences of T. nobilis? occurred in late in lakes of Kentucky and other parts of August, September, and October, with North America,but to my knowledge it is pupation beginning in June and lasting very rarely found in smaller streams. A through September. The pupae occurred number of larvae were collected from a most frequentlyon the tops of larger stones deep, quiet pool downstreamfrom Station covered by thick mats of mosses (usually IV in June 1961. The species rapidly colo- species other than Fissidens). On Septem- nized Doe Valley Lake after its impound- ber 7, 1960, I carefully removed the moss ment in July 1961,with a massiveemergence layer from a stone located near the bank approximately 3 months after the first in Doe Run and found 33 cast pupal skins, inundationof the valley. 15 pupae, and 51 larvae ranging from small Simulium sp. was found from Station II to large in 180 square inches of moss. All to Station V on swift, open riffles, and at- stones emersed or along the banks, and tached to most solid substrates. Its abun- with significant amounts of vegetation on dance was somewhat reduced in summer them, had many pupal skins projecting in the marl area, possibly because of from the vegetation. Tipula abdominalis encrustationby carbonateson pupal cases, was found from Station III downstream, and it was most abundant in the high- and inhabited interstices of gravel bars, gradient tributaries of the middle portion islands, or packets of fallen leaves that of Doe Run (Fig. 1). collected in autumn. Fifteen species of tendipedid dipterans Pilaria tenuipes occurred from Station I were identified from Doe Run, but many to Station V, and usually in beds of emer- additional forms undoubtedly occur. Be- gent aquatic vegetation or in detritus accu- cause of the difficultyin identifyingmidges, mulations in pools. The genus is thought and the problems of their taxonomy in ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 57

North America (James, 1959; and others), may be carnivorous,thus comprisinga gen- there are little data on their ecology. A eral exception to the other Tendipedinae carnivorous group (Leathers, 1922), the (Curry, 1958; Wirth and Stone, 1956). subfamily Pelopiinae (= Tanypodinae), in- Of the 2 remainingspecies of tendipedids cluded Pentaneura sp. and Procladius sp. in identified, Microtendipes pedellus? oc- Doe Run. The former occurred at all curred throughout the creek in samples stations downstream from Station II, from pools, and Glyptotendipessenilis was whereas the latter was taken only at Station taken at Station IV in a similar habitat. III. Both were found in eddies and in beds Biting midges, Heleidae (= Ceratopo- of emergent vegetation or detrital deposits. gonidae), were represented by Culicoides Usually they are associated with lotic sp. in shallow pools at Stations III, IV, conditions in streams (Curry, 1954). The and V. remaining tendipedids identified are gen- Tabanus sp., the horse-fly, occurred as erally considered phytophagous or detritus larvae at Stations III and IV in interstices feeders (Leathers, 1922). Those of the of gravel, usually slightly above the water subfamily Hydrobaeninae (= Orthocladii- levels of the creek. This larva was rare nae), Psectrocladius?sp. and Trichocladius when compared with Tipula abdominalis sp., were generally distributed in the stream, in the lower parts of Doe Run, but the 2 and usually were identified from samples large dipterans often were taken side by obtained in swift waters. Specimens from side in beds of stream-driftedleaves and Doe Run referred to Psectrocladius? could debris in autumn. be the closely related Eukiefferiella; how- Atherix variegata is a large, common in- ever, they lack a small hook on the basal habitant of the marl area of Doe Run, and portion of the mandible which should be was at Stations II and V in small numbers. present in the latter genus (Stuart E. Neff, This species often was seen moving openly pers. comm.). on marl riffles, and was capable of nego- The subfamily Tendipedinae (= Chiro- tiating swift currents. It also inhabited nominae) was represented by 11 species in small pits or "burrows"in marl surfaces. Doe Run. Of these, Calopsectra dissimilis?, Scatella sp. was found at Stations I and Tendipes (Tendipes) sp., and T. (Limno- II along the banks on riparian roots, and chironomus) sp. were identified only from often in emersed or slightly submerged samples at Station II in association with beds of Fissidens. A few individuals also submerged vegetation. A third species, T. were taken at StationIII in the small riffles (Limnochironomus) modestus, was taken lateral to the channel. The anthomyiid in pool habitats at Station V. Limnophora was common downstream Tanytarsussp. and T. (Tanytarsus) sp. from Station II in small riffles and in small occurred only at Station III, but T. (Sticto- pits in the surface of marls, and a small, chironomus) sp. and T. (Endochironomus) unidentified species of Empididae was col- sp. were found throughout the stream. lected at Station IV from a similarhabitat. They were in samples obtained from small Mollusca.-Pulmonate snails were notably riffles lateral to the channel at each station, absent from Doe Run, but Physa spp. oc- but also occurred in samples from Fissidens curred in many smaller, seepage-fed tribu- at Stations I and II, and in emergent vege- taries, and was the only snail of this type tation or detritus accumulations at other found in the stream system (excluding stations. semiterrestrialforms). These snails were Cryptochironomus sp. occurred only at found in the mainstream only at dams, Station III, in a sample from a riffle lateral where they inhabited spray areas and to marl. The species of this generally shallow pools, and downstream from Sta- northern genus are flexible in their habitat tion V, in backwatersof the navigationpool requirements (Curry, 1958), and many of the Ohio River. Physa was also absent 58 WILDLIFEMONOGRAPHS from the larger, spring-fedtributaries, thus and their shells were incorporatedinto the implying a general avoidance of cool bottom sediments. waters. The eggs of Goniobasisin Doe Run were Living Ferrisia were not collected from laid on undersides of stones in packets of Doe Run, but appeared in laboratorytanks 16 to 27 eggs in 20 packets counted. The of vegetation stocked from Station II. eggs were enclosed in a jellylike substance Twice there was no doubt as to the prov- covered by tiny sand grains and other enance of the material. Empty shells of debris, presumably through action of the this mollusk were found in sediments from current. I do not know the time required Stations I, II, and III. for hatching, but eggs were most common Pomatiopsis lapidaria, a predominantly in bottom samples from July through Sep- terrestrialspecies (semiaquatic), was taken tember 1960 at Station I. Eggs also were in bottom samples from deep water at found in October,January, and April, indi- Station V in 1959 (Kaplan and Minckley, cating sporadic or continuous breeding, at 1960), and was relatively abundant along least by part of the population. Egg masses the banks of the lower creek prior to im- at Stations III and IV were covered with poundment. marl on occasions, but this did not appear Goniobasiswas the largestand most abun- to deter hatching. dant gastropod in Doe Run, occurring Sphaeriidclams were present throughout throughoutthe creek to Station V, and in the Doe Run system, and usually were 5 of the spring tributaries: Tributary I-b, abundant wherever there were relatively Blue Spring,Buffalo Spring,Tributary II-a, firm deposits of silt-sand, detritus beds, or and 7-Springs Branch. It was also found relatively firm substrates beneath beds of in those portions of some smaller spring vegetation. Reproduction was greatest in tributariesthat flowed on the floodplain of summer, from sizes of animals se- Doe but did judging Run, not occur above the first cured, but this may be an expression of incline or riffle. Goniobasiswas ex- major better survival of young in periods of abundant in Doe tremely locally Run, modal flow or movement of larger indi- in the and especially upstream areas, in viduals in sediments with of modal flow the animals covered deeper receding periods water levels of summer. the bottoms near Mile 0.5 in an almost perfect randomdistribution. At those times QuantitativeAspects of the Invertebrate populations comprised 243 to 329 indi- Fauna viduals per square foot (10 quadrats with an of Descriptions of Sampling Areas.-Although a counted), average density 296, general description of Doe Run and the principal with considerably larger populations per collecting stations has been made, it seems appro- unit area in beds of vegetation. Equally priate to discuss in detail the areas where quanti- large populations of this snail were found tative sampling was attempted. Station I.-At Station I during modal discharge, in the Hocking River, Ohio, by Ludwig most bottoms were covered by vegetation (Fig. (1931). After floods, Goniobasis moved 22), but during high discharge the amounts var- from pools into which they had been dis- ied with the severity of washout. Gravel and placed, back onto shallow areas. The rubble were present to about 60 ft downstream from the source. Sand and some gravel were species in Doe Run was not a riffle animal, common in backwaters, and were the principal but generally inhabited vegetation or open bottom materials downstream from a small rubble areas where currents on obstruction at the lower end of the stony bottoms. eddying persisted There was some silt in the mouths of Tributaries firm bottoms. On soft, silty bottom the I-a and I-b, and along the banks, and there was snail sinks deep into the substrate and its a thick deposit of detritus just below the inflow movements are obviously impaired. After of I-b during modal flow. Samples were taken from the following habitats: floods, many Goniobasiswere stranded on Fissidens Beds in the Channel. Beds of aquatic bars and small islands, where they expired, mosses tend to reduce turbulence by presenting a ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 59 smooth surface over which the water passes formation of mounds or bars as in the latter, but the (Margalef, 1960). This surface results from im- finely divided leaves, flexibility, and tendency to brication of the tiny, flattened leaves, and pro- flatten in current to form cushionlike mats, are duces a thin, cushionlike mat beneath which there mosslike. The bottom in beds of milfoil usually is only nominal current and a rich supply of was gravel, sand, and snail shells, with silt and detrital foods filtered from the water. Fissidens detritus on the downstream sides. Samples were was damaged by washout or by abrasion by water- taken from near the center of the mounds. Cur- borne materials in 2 different ways: (1) the rents ranged as high as 2-3 ft/sec, but conditions moss with its contained invertebrates was removed within the beds were undoubtedly more quiet. en masse from recently colonized, unstable sub- Open Pools with Varied Bottom Types. Larger strates and washed away, and (2) the action of pools at Station II usually had bottoms of sand, the water plucked the terminal portions of the gravel, and shells of Goniobasis, mixed with silt plants, thinning the beds without removing them and detritus. In autumn, leaves were a major and without extirpating the fauna. Such thinning constituent of the bottom materials, but usually was sustained by plants firmly attached to larger were washed away by floods that sometimes substrates and the amount of damage depended cleared the materials to bedrock. As noted above, on the duration and severity of flooding. With currents in pools in the upper parts of Doe Run alleviation of flood condition some silt was de- usually exceeded 0.5 ft/sec. posited in Fissidens beds in the channel, but with Station III.-Station III was typical of the marl subsequent clear-water flow the moss was rela- area, except that it was frequently disturbed by tively free from silt within a few weeks. man. The station may be characterized by 2 Unvegetated Riffles, or the Least-Vegetated types of pools, 1 with sharp, precipitous marl rims Riffles Present. There were small open channels and rubble or marl bottoms, and the other with between major moss beds throughout the year, sand, gravel, and silt banks that sloped gently to presumably because the substrate was unsuitable. the bottoms, and with considerable eddying of In 7 of the 15 months of sampling, however, no current. Such pools were adjacent to waterfalls completely unvegetated riffles were present. Sub- and had swift, turbulent current. Riffles at Station strates consisted of small rubble and gravel over- III consisted of solid marl sheets or of gravelly lying larger rubble and bedrock. Currents ranged areas maintained by human activities such as from 1 to 3 ft/sec and usually were turbulent. swimming and wading. Samples were taken in Nasturtium Beds, or in Such Areas After Wash- areas not visibly disturbed: (1) in a relatively out. Watercress beds at Station I were subject inaccessible marl riffle upstream from the recrea- to severe washout and samples from those areas tional area; (2) in small, rubble-bottomed riffles might be considered to be from silt-sand bottoms. lateral to the marl riffle; (3) in beds of Myosotis Seven of 17 samples included Nasturtium, the growing in deep silt bottoms lateral to pools; and others were taken in areas of former beds after (4) in a large, silt-detritus-bottomed eddy. Of washout, but always immediately downstream these, number 3 is comparable to watercress beds from Tributary I-b. Accumulations of silt and at Station I and will not be described further. detritus in the beds were sometimes black and Marl Riffles in the Channel. The term "marl malodorous. After washout some roots and other riffle" refers to riffles almost completely cemented organic materials remained and were covered by into a solid slab by marl deposition. The depth sand and silt. of water at the point of sampling usually ranged Shifting Sand Bottoms. The shifting sand bot- from 2 to 8 inches in periods of modal flow, and toms at Station I were immediately downstream currents exceeded 3 ft/sec. Currents and depths from the rubble obstruction and were continuously were much higher at higher discharges. subjected to eddying currents exceeding 0.5 ft/sec. Riffles Lateral to Marl. Substrates were small During modal flow the bottoms were somewhat marl concretions, gravel, and snail shells, with silt stabilized, and supported sparse stands of Batra- and detritus deposition between stones. Currents chospermum and diatoms. were variable, ranging to a low of 0.5 ft/sec. In Station II.-This station was similar to Station low-water periods many large stones were exposed I in that plants covered a large part of the bottom to air, especially in daytime when evaporation and (Fig. 24). Most quantitative sampling was done transpiration of riparian vegetation caused slight immediately below Dam 1, but some samples reductions in water level. were obtained in comparable habitats near the Eddying Pools. Samples from the large eddy at mouth of Tributary II-a. Two habitats studied Station III were from diverse bottom types, some- at Station II, Fissidens beds and unvegetated what comparable to the open pools at Station II. riffles, were similar to those at Station I. Addi- The bottom ranged from small gravel and large tional areas were: sand after periods of flood, to deep, soft silt and Myriophyllum Beds. Beds of milfoil are inter- detritus after summer flow. Currents in the sam- mediate between Fissidens and Nasturtium in pling area flowed upstream at the rate of 0.25 to habitat type. The relatively large plants allow 0.5 ft/sec except in extreme flood. The water was for considerable accumulation of debris and the about 18 inches deep at modal flow. 60 WILDLIFEMONOGRAPHS

Station IV.-The main study area at Station IV fauna and few comparative data can be derived was a large, circular pool with a series of swift from them. Substrates ranged from clay to coarse marl riffles upstream and a small, marl-rubble sand, depths from a few inches to greater than 6 obstruction at an old bridge abutment downstream. feet in periods of partial inundation by the Ohio As with most pools in Doe Run, much of the bot- River, and velocity of current ranged from negli- tom was bedrock, with silt deposits in eddies. gible to about 1 ft/sec. Deep beds of detritus settled out on the bedrock bottom in periods of constant flow. Samples were StandingCrops and NumericalRelations of obtained from 4 habitats: (1) marl riffles that the Fauna.-The fluctuant nature of inver- differed from those at Station III only in the tebrate populations in a stream is well presence of a few large, limestone slabs; (2) from data from Doe Run. This is a small lateral riffle generally comparable to that exemplified by sampled at Station III, but with slightly more silt true even though the stream originates in and sand deposition; and from 2 additional habi- a large spring with a consequent ameliora- tats, descriptions of which follow. tion of extremes other than discharge. Al- Silt-Bottomed Eddy. This habitat was unique error is actual in that it remained essentially unchanged through- though sampling great, out the study. The silt deposit lay at the mouth changes in benthic populations are subject of Tributary IV-a, but apparently was in an to rapid fluctuations (Hynes, 1960), and abandoned channel of Doe Run rather than an are manifest in 3 ways: (1) by recruitment eddy. Tributary IV-a was a composite of a num- of new individuals through reproduction; ber of springs and a small seepage-fed marsh in a low part of the old channel (Fig. 1). The silt (2) by death of the organisms in situ, from deposit lay about 3 feet from the main inflow of catastrophe, old age, predation, or other the upstream riffle, but was protected by marl agents; and (3) by movement of organisms deposits on roots of a streamside tree. Eddying into or out of the habitat, under their own currents flowed at 0.5 to 1.0 ft/sec, but there was or little removal or deposition of silt. power by physical forces in the environ- Gravel-Rubble-Bottomed Eddy. This area has ment. Interpretation of variations in the sufficient current to warrant consistent use of the populations is complicated by seasonal as- Surber sampler, and might easily be classed with pects of reproduction of the different riffle habitat if it were not for its relatively great and related to depth (12 to 18 inches) and general lack of turbu- species phenomena repro- lence. Some deposition of sand and silt occurred duction such as emergences of certain after major flooding, but was removed by sub- insects, by differences in response of vari- sequent flow. ous kinds of animals or their vulnerability Station V.-Because Station V was inundated to diverse in the or by the Ohio River, quantitative data on the ben- changes environment, thos are sparse. The station consisted of a single by many other factors. series of riffles, bordered both up- and downstream Riffles of the Main Channel.-Fissidens by long, relatively quiet areas. Few stones were Beds. In beds of Fissidens at Station I, present, even on riffles. An amazingly resistant animal populations ranged from 18,708 clay comprised the bottom in most swift areas, and was overlain by silt, sand, and detritus in individuals per square foot (ft2) in Decem- pools. Samples were obtained from the following ber 1959, to 2,332/ft2 in July 1960. The areas: average number of animals in 15 samples Swift, Rubble-Bottomed Riffle. A short section was 8,628/ft2 of which isopods made up of rubble bottom was present at the upstream end of the riffle area; however, velocity of the water 68.9 and amphipods 24.1 per cent of the was rarely less than 3 ft/sec in modal discharge total. Dipterous larvae were third in nu- and sampling was difficult at best. Also, the merical abundance, followed by caddisflies, rubble bottom consisted of large, angular boulders, turbellarians, and oligochaetes. The great- not conducive to effective sampling. est of in this habitat was Rubble-Gravel-Bottomed Riffle, Lateral to the weight organisms Channel. This habitat was similar to those at Sta- 28.695 g/ft2 in December 1959, weights of tions III and IV lateral to the channel, except that crayfish and mollusks excluded, of which most of the gravel at Station V was smooth, 25 g/ft2 were amphipods. The least weight rounded erratics eroded from adjacent Pleistocene was in 1960, 0.926 g, when the smallest Velocities of current were July deposits. variable, but numbers of animals the riffles were generally quite turbulent. were present. Average "Pools," in the Channel. Samples from these weight was 8.075 g, decapods and mollusks areas were taken more or less as a survey of the excluded, and only 8.750 when they were ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 61 included. Amphipods were by far the most Major changes in animal numbers in abundant by weight, making up 58.8 per Fissidens, and in other habitats of Doe cent of the total, and were followed in Run, apparently resulted from changes in order by isopods, 24.8 per cent, and by discharge or related phenomena. For ex- dipteran larvae, trichopterans, turbellarians, ample, at Station I the smallest populations and oligochaetes. were present in July-August 1960 (Table Moss beds at Station II were somewhat 6), following extreme discharge in June, more stable than at Station I, apparently and in June 1961 following severe flooding because of more extensive bedrock bot- through spring of that year. At Station II, toms, and contained a larger standing crop however, the populations were higher in of animals, both in numbers and weight. Fissidens after the June floods than before, Isopods and amphipods retained numerical mostly because of great numbers of caddis- dominance in the fauna, comprising 47.0 flies (principally Hydropsyche and Cheu- and 22.4 per cent, respectively, of the matopsyche) that colonized upper surfaces average 9,295 individuals/ft2. It is note- of the beds. Numbers of turbellarians in worthy that isopods were both actually and Fissidens beds generally increased with in- relatively more numerous at Station I, but creases in discharge at both stations, as did thle number and relative abundance of oligochaetes at Station I, but at Station II, amphipods were the same at both stations the largest numbers of oligochaetes were (2,082/ft2, 24.1% at Station I, and 2,085/ft2, present during the 1960 period of modal 22.4% at Station II). Trichoptera larvae, discharge. Isopods usually were the most turbellarians, oligochaetes, and mayflies abundant animals, but their populations were more abundant in moss at Station II fluctuated considerably as a result of wash- than at Station I, but dipteran larvae were out and thinning of Fissidens, restriction reduced in relative numerical position, fall- of the major breeding season to 3 months ing to fourth most abundant at Station II. in early summer (Walker, 1962), and The greater average weight of standing movement and concentration of the animals crop at Station II, 10.810 g/ft2, was aided in response to unknown factors (see Allee, substantially by the abundance of large 1929). Some inconsistencies may relate to tipulid larvae and hydropsychid caddisflies. the chance collection of such aggregations Dipterans contributed the most weight, of isopods and other forms by the sampler. followed by amphipods (25.6%), trichop- Numbers of amphipods, mostly Gam- terans, isopods, turbellarians, and oligo- marus minus, were remarkably constant at chaetes. A major difference between the both stations in Fissidens, and apparently fauna of moss riffles at Stations I and II were much more independent of discharge was the greater abundance of snails at the in maintaining their populations than were latter, comprising an average of 12.642 g/ft2 isopods. This may be partly explained by (not included in biomass data given above). upstream movements after displacement by When mollusks and crayfish are included flood (Minckley, 1963). in the biomass at Station II it was almost The small trichopteran, Agraylea, at Sta- 3 times that at Station I (23.467 and 8.750 tion I was abundant only from March to g/ft2). Numbers and weights of inverte- June 1960, and its sudden decrease may be brates in this habitat at Station II ranged attributable to emergence in June. Dip- from maxima of 20,004 individuals in De- terans also were most abundant at that cember 1959 and 48.937 g/ft2 in January station from March to June, but maintained 1960 (38 g of which were Tipula), to considerable populations to the September- minima of 1,691 individuals and 1.658 October period. Dipterans were generally g/ft2 in November 1959. The last values, abundant in Fissidens at Station II through- however, were obtained by Surber sampler out the study. and are not strictly comparable. Increase in the numbers of Turbellaria, 62 WILDLIFEMONOGRAPHS

TABLE 6.--NUMBERS OF ANIMALS PER SQUARE FOOT IN RIFFLES OF THE MAIN CHANNEL OF DOE RUN, MEADE COUNTY, KENTUCKY. DATA ARE GIVEN AS AVERAGES OF 2 SAMPLES, 1 FOR EACH MONTH INDI- CATED EXCEPT WHERE SAMPLES ARE LACKING; THOSE CASES ARE MARKED WITH AN ASTERISK (*). DATA FOR JUNE 1961 ARE BASED ON SINGLE SAMPLES THROUGHOUT

November- January- November Animal Group December February March- May- July- September- (1960 )- June (1959) (1960) April June August October January (1961) (1961) STATION I: Fissidens Beds Turbellaria 7.5 68.0 316.0 170.0 108.0 152.0 50.0 168.0 Oligochaeta 8.0 - 28.0 148.0 24.0 134.0 66.0 - Isopoda 9,283.5 6,596.0 6,764.0 4,230.0 2,912.0 8,690.0 4,908.0 1,096.0 Amphipoda 2,340.0 2,124.0 2,902.0 852.0 1,294.0 2,216.0 3,624.0 524.0 Plecoptera - - - 2.0 - - - - Ephemeroptera - - 2.0 2.0 - - - - Trichoptera 0.5 - 308.0 998.0 30.0 2.0 86.0 128.0 Coleoptera - - - 20.0 4.0 - - - Diptera 4.5 6.0 6.0 404.0 573.0 266.0 36.0 472.0 Decapoda 3.0 - - - - - 2.0 - Gastropoda 3.0 4.0 - 8.0 - 2.0 2.0 - Pelecypoda 0.5 4.0 - - - - TOTALS 11,650.5 8,802.0 10,344.0 6,834.0 4,945.0 11,462.0 8,774.0 2,388.0

STATIONII: Fissidens Beds Turbellaria 391.0 1,284.0 306.0 156.0 732.0 56.0 644.0 92.0 Oligochaeta - 6.0 - 28.0 1,028.0 - 8.0 Isopoda 7,050.0 6,792.0 1,424.0 4,330.0 5,450.0 2,602.0 4,598.0 996.0 Amphipoda 2,550.0 2,004.0 956.0 2,152.0 2,828.0 650.0 3,984.0 1,024.0 Plecoptera - - 8.0 2.0 -- -- 8.0 Ephemeroptera 353.0 104.0 54.0 206.0 120.0 236.0 332.0 72.0 Megaloptera - -- -2.0 Trichoptera 348.0 3,506.0 1,321.0 536.0 2,148.0 494.0 5,140.0 208.0 Coleoptera 30.0 42.0 10.0 26.0 24.0 54.0 76.0 96.0 Diptera 119.0 450.0 268.0 910.0 914.0 1,450.0 98.0 708.0 Acri 1.0 - - - 2.0 Decapoda ------2.0 - Gastropoda 7.0 - 2.0 44.0 20.0 1,130.0 112.0 4.0 Pelecypoda - - - 4.0- 2.0 - - TOTALS 10,849.0 14,188.0 4,349.0 8,394.0 12,240.0 7,702.0 14,986.0 3,216.0 STATIONIII: Marl Riffle Turbellaria 0.5 0.5 0.5 0.5- Oligochaeta - 0.5 0.5 0.5 0.5 4.0 Isopoda 3.5 1.5 3.0 3.5 4.5 2.0 -- 1.0 Amphipoda 2.0 -3.5 3.0 2.5 -- 1.0 Plecoptera 3.5 2.0 3.5 2.0 2.0 1.0 - Ephemeroptera 30.5 16.5 116.5 84.5 187.0 57.0 10.5 6.0 Trichoptera 9.5 5.5 3.5 11.5 96.5 16.0 9.5 Coleoptera 2.5 - - 0.5 1.0 -- 2.0 Diptera 19.0 10.0 40.0 114.5 167.5 18.0 36.5 - Acri -- - 0.5- 1.5 Decapoda 0.5 - 1.0 1.5 Gastropoda 2.0 0.5 1.5 2.0 TOTALS 73.5 37.0 174.0 222.0 464.5 93.0 58.0 14.0 STATIONIV: Marl Riffle

Turbellaria '------1.0* Oligochaeta 0.5 4.0 2.5 2.5 - 1.0* Isopoda 2.0 2.5 1.0 -- ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 63

TABLE 6.-Continued

November November- January- March- May- July- September- (1960)- June Animal Group December February April June August October January (1961) (1959) (1960) (1961) Amphipoda 2.5 - 2.0 -0.5 - 0.5 Plecoptera 9.5 2.0 11.0 2.0 4.0 4.0* 1.0 Ephemeroptera 31.5 25.5 42.0 48.5 63.0 20.0* 32.0 88.0 Megaloptera - 0.5 - 1.0* 0.5 Trichoptera 78.0 28.0 4.0 15.5 91.0 4.0* 5.0 25.0 Coleoptera 1.0 -- - 1.0 - Diptera 50.0 25.5 87.0 40.0 96.0 21.0* 16.0 614.0 Acri 2.5 - --- - Decapoda - -0.5 0.5 1.0-- - Gastropoda 2.0 0.5 0.5 0.5- - TOTALS 179.5 86.0 152.0 110.5 256.5 51.0* 56.0 727.0

STATION V: Rubble Riffle Oligochaeta - 9.0 - 1.5 0.5 1.0* Isopoda 2.5 5.0 17.0* 32.0 4.0 1.5 2.0* Amphipoda - 1.0 - 4.0 - 2.0* Plecoptera 1.0 -2.0* - - Ephemeroptera 10.0 5.0 6.0* 35.0 26.5 18.5 22.0* Megaloptera - - - 1.0* Trichoptera 13.5 0.5 3.0* 8.0 7.5 12.5 7.5* Coleoptera 0.5 1.0 - 2.0 2.0-- Diptera 14.0 2.0 66.0* 38.0 2.5 75.0 8.0* Acri -.- 0.5 - Decapoda - -- 1.0 0.5 Gastropoda - 2.0 1.0 1.0 TOTALS 41.5 25.5 94.0* 122.5 43.5 109.0 43.5*

Oligochaeta, and dipteran larvae in moss entire stream from such sheltered areas in during periods of above-normaldischarge spring after recession of winter floods may be attributableto movement into rela- (Whitehead, 1935) possibly was important tively stable habitats during extremecondi- in Doe Run. An increase in the diversity tions. With oligochaetes and dipterans, of the fauna in winter in Fissidens,particu- however, these increases may have been larly at StationI, seems to indicate general related to deposition of sand and silt in the emigrationby certain forms in summer. In moss by the highly turbid waters. Removal long periods of modal flow the fauna of these sediments in periods of modal gradually changed to almost entirely A. discharge was gradual, and probably was bivittatus and G. minus, but substantial assisted by burrowing activities of the numbers of dipterans and oligochaetes re- animals. Vertical movements by aquatic mained in the sediments beneath and larvae through as much as 12 centimeters within the beds. of gravel were noted by Badcock (1954a, As a general rule, dipterans and oligo- b) as possible causes of rapid change in chaetes increased or maintained relatively numbersof animals in a given habitat, and large populations throughout the creek, as a mechanismfor avoiding severe winter including the Fissidens habitat, during discharges. Frost (1942) recordedpeaks in modal flow. This may have resulted from abundance of chironomid (= tendipedid) receding water levels and movement of dipterans in winter similar to those found animals from the silt and clay banks, small in moss habitatsof Doe Run, and attributed gravel bars, and winter-submergedislands changes in populationsto immigrationfrom in downstreamarea, into the channel of the less suitable areas. Recolonization of the creek to avoid desiccation. This presum- 64 WILDLIFEMONOGRAPHS ably would be counteracted by lateral Diptera were most numerous on marl movement on resumption of winter dis- riffles at Station IV. Mayfly nymphs were charge conditions, and would result in second in abundance, averaging 42 indi- decreased populations in the channel dur- viduals/ft2 (24.2% of the total, and caddis- ing those periods. Increases in numbers of flies were third with 19.2%). The average some burrowing forms in pools may also numerical standing crop was 175/ft2, result from drift and accumulation there slightly higher than the average in compa- during flood, but would be at the expense rable habitat at Station III (147/ft2); the of those in the current. average weight, 0.595 g/ft2, was higher than Unlike the fauna at Station I, that in the average of 0.404 g/ft2 at Station III. Fissidens at Station II was relatively di- Trichoptera made up a major part of the verse throughout the year (Table 6). Eight weight, 33.9 per cent, similar to that at groups of organisms comprised at least 1 Station III. Ephemeroptera and Diptera per cent of the sample at Station II, were present in almost equal weights at whereas at Station I there were only 5 Station IV (21.6 and 20.8%, respectively) such groups. At the specific level, the and were followed by stonefly nymphs disparity between the 2 areas was even (17.3%). The range of standing crops was more pronounced (see Fig. 25). This as follows: maxima-727 individuals in situation apparently resulted from an over- June 1961 and 1.647 g in late February lapping of faunal elements indigenous to (= "March"); minima-8 individuals in the source of the creek, and of forms more July 1960 and 0.032 g in January 1961. common downstream (such as caddisflies, Marl riffles at Stations III and IV are various mayflies, etc.). There were, how- comparable, not only in averages, but in ever, some species at Station II that were numbers of individuals (Table 6). Mayfly rare or absent at stations upstream and nymphs were least abundant in January- downstream. February 1960 and in November-January Marl Riffles. The general absence of 1960 and 1961, at both stations, which cor- vegetation at Station III and the resulting responds generally to the major emergence paucity of plant-inhabiting crustaceans was of the most abundant form, Baetis vagans. reflected in numbers of animals per unit Caddisflies, principally hydropsychids, be- area. Samples from marl riffles averaged came less abundant in September-October 145 organisms/ft2 of which mayflies made 1960 at each station, and this may relate to up 44.0 per cent. Dipterans inhabiting the their emergence, although no definite pat- porous marls (tendipedids, Limnophora, terns were recorded. Dipteran larvae in- and others) and some living on top of the creased in summer at both stations, but marl such as Simulium, made up 36.6 per were absent from marl riffles at Station III cent, and Trichoptera, 13.3 per cent. By in June 1961, presumably as a result of weight, however, trichopterans were first severe scouring by spring floods. At Sta- with 37.1 per cent, followed in order by tion IV, however, dipterans (mostly tendi- mayflies and dipterans. Plecoptera nymphs pedids) were more abundant after the 1961 comprised only 1.3 per cent of the animals floods than ever before, apparently because taken, but made up 13.4 per cent by of clearing in that area. Siltation of some weight because of their generally large riffle areas, and especially the luxuriant sizes. The average standing crop on marl growth of Cladophora on formerly marl riffles was 0.404 g/ft2 (1.607 when mollusks riffles, may have produced more favorable and decapods are included). Numbers and conditions for the larvae. weights of organisms ranged from no ani- Stonefly nymphs were a consistent, mals in October 1960, the reasons for which though minor, part of the fauna on marl are unknown, to 679/ft2 in August of that riffles at both Stations III and IV, but were year, and 1.522 g/ft2 in July (1.193 of totally absent from those habitats in June which were Hydropsychidae). 1961. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 65

Rubble Riffles. As noted earlier, sam- Diptera, Turbellaria, and Oligochaeta. pling at Station V was complicated by Snails,beetle larvae,mayflies, and sphaeriid sporadic presence of backwaters of the clams each comprised at least 0.1 per cent Ohio River, and data are not strictly com- of the total. As in Fissidens, amphipods parable to those from other stations. The were outnumbered by isopods, but con- swift, rubble-bottomedriffle at Station V, tributed a major part of the weight however, yielded average numbers of ani- (43.2%). Isopods were second by weight mals quite similar to those found in marl (21.6%), and the occurrence of a number riffles (146/ft2 compared with 147 at Sta- of large Tipula larvae caused dipterans to tion III and 175 at Station IV). The rank third. The average weight was 0.347 biomass was much smaller at Station V, g/ft2 when mollusks and decapods were averaging 0.136 g/ft2 in 12 samples. Cad- excluded, but 1.090 when included. disflies, dipterans, and mayflies were the Unvegetated riffles at Station II often most abundant kinds of animals, and col- contained a small amount of moss; 9 of 15 lectively made up 81.7 per cent of the total samples contained at least 10 per cent esti- weight. The numbers of animals per mated moss cover. The largest number of sample ranged from 9 to 185/ft2, and organismswas in November1960, 2,250/ft2, weights ranged from 0.003 to 0.252 g. when the least vegetated riffles found bore Invertebrateson the riffles at Station V 30 per cent moss cover. Crustaceansmade were subjected to inundation by the Ohio up 59.3 per cent of the standing crop, of River backwaters on occasions, which re- which amphipods comprised 34.0 per cent sulted not only in semilentic conditionsbut and isopods 25.3 per cent. Aquatic insects in relatively heavy siltation. The fauna made up 35.3 per cent of the fauna, thus consisted of organisms somewhat resistant contributinga much higher proportionthan to ponding, such as Stenonema, various in comparable samples at Station I. The dipteran larvae, and isopods. Organisms average standing crop was 505 individuals/ such as Isonychia, Simulium, and hydro- ft2. Amphipods were most abundant by psychids were rare, and increased in abun- weight (40.8%), followed by caddisflies, dance in periods of modal flow, when riffle isopods, dipterans, and mayflies. The conditions persisted. average, overall weights for 15 samples Riffles Lateral to the Channel.-Unvege- were 0.698 g/ft2, 1.257 if mollusks and tated riffles at Stations I and II. The decapods are included, with a range in standing crop of invertebrateson unvege- standing crop from 0.011 in July 1960, to tated riffles was smaller than in Fissidens, 2.980 g/ft2 in November 1960. averaging264 individuals/ft2in 15 samples The abundance of animals in Fissidens from Station I. The greatest number of beds at both StationsI and II was strongly animals was present in September 1960, reflected in the fauna of unvegetatedriffles 1,363/ft2, when the "unvegetated riffle" not only at times when moss was present contained an estimated 25 per cent moss but throughout the study period. At Sta- cover. Proportionately high populations tion I, animals that were relatively scarce were present in June 1960 at an estimated in Fissidens were much more importanton moss cover of 20 per cent (1,043/ft2). The the gravel-rubblebottoms of smallerriffles, smallest populations occurred when no principallybecause of an avoidance of the moss was present, and after majorwashout latter by crustaceans (or a preference by (minimum, 26/ft2 in January 1961). Iso- crustaceansfor moss). In some instances, pods were most numerous,comprising 53.3 however, absolute differences indicate a per cent of the overall total, and their preference for "unvegetated areas." For numerical abundance was a reliable index example, mayfly nymphs were found on of the amount of moss. Amphipods made open riffles twice as frequently as on moss, up 25.9 per cent, followed by Trichoptera, and beetles were present much of the time 66 WILDLIFEMONOGRAPHS

TABLE 7.-NUMBERS OF ANIMALS PER SQUARE FOOT IN RIFFLES LATERAL TO THE MAIN CHANNEL IN DOE RUN, MEADE COUNTY, KENTUCKY. SEE TABLE 6 FOR FURTHER EXPLANATION

November- January- November Animal Group December February March- May- July- September- (1960 )- June (1959) (1960) April June August October January (1961) (1961) STATIONI: Unvegetated Riffles Turbellaria 1.0 -7.5 52.0 2.5 6.0 10.5 3.0 Oligochaeta 1.0 -0.5 7.0 -5.5 - - Isopoda 62.0 50.5 82.0 254.5 46.5 501.5 53.0 11.0 Amphipoda 19.0 27.0 88.0 55.5 32.5 186.0 95.5 18.0 Plecoptera --- 0.5 ---1.0 Ephemeroptera 0.5 1.0 3.0 - - 1.0 Trichoptera 1.5 61.0 100.5 1.0 1.5 1.5 1.0 Coleoptera - - 1.0 3.0 2.0 3.0 2.0 Diptera 0.5 -0.5 101.5 5.5 4.0 1.5 13.0 Acri - -.- - 0.5 - Decapoda - - 0.5 - 0.5-- Gastropoda 1.5 --3.0 1.5 2.5 0.5 1.5 2.0 Pelecypoda 2.0 - ---- TOTALS 88.5 78.0 243.5 577.5 93.5 707.5 167.0 52.0 STATIONII: Unvegetated Riffles Turbellaria 10.0 5.0 8.5 1.5 60.5 5.5 72.0 Oligochaeta 5.0 0.5 5.0 1.5 1.0 1.5 1.0 - Isopoda 313.0 260.0 191.0 20.0 151.5 26.0 78.5 4.0 Amphipoda 114.0 96.0 100.0 17.0 324.0 27.0 608.0 9.0 Plecoptera - - 0.5 - Ephemeroptera 14.0 46.0 45.5 3.5 32.5 39.0 24.0 4.0 Megaloptera - 0.5 ----- Trichoptera 22.0 59.0 142.0 8.5 310.5 17.5 332.0 2.0 Coleoptera 12.0 11.0 3.5 3.5 10.0 31.5 8.5 Diptera 8.0 4.5 8.5 8.0 66.5 44.0 18.0 5.0 Acri 0.5 - - 0.5- Decapoda 0.5 1.0- 0.5 - - Gastropoda 2.0 0.5 2.5 0.5 0.5 - 11.5 2.0 Pelecypoda - -- 0.5 - - TOTALS 501.0 483.5 507.0 64.0 958.5 1,925.0 1,153.5 26.0 STATIONIII: Gravel-Rubble Riffles, Lateral to Marl Turbellaria - - - 0.5 Oligochaeta 7.5 2.5 - 39.5 - 0.5 Isopoda 3.0 - 3.0 29.5 1.5 2.0 6.0 Amphipoda 3.5 - 1.5 11.5 2.0 1.0 4.0 2.0 Plecoptera - - 1.0 5.0 - Ephemeroptera 5.5 13.5 39.5 544.0 69.0 57.5 30.0 39.0 Trichoptera 1.0 4.5 1.5 7.0 13.5 10.0 1.0 Coleoptera 1.0 - - 1.0 0.5 0.5 0.5 Diptera 9.5 4.0 24.0 28.5 10.5 58.5 8.0 9.0 Decapoda 0.5 - 1.0 6.0 0.5 0.5 Gastropoda 1.0 - 1.5 11.0 ---0.5 - Pelecypoda 0.5 -- ... - TOTALS 33.0 24.5 73.0 678.0 97.5 135.0 50.0 51.0 STATIONIV: Gravel-Rubble Riffles, Lateral to Marl ?..1. Oligochaeta - 62.0 3.0 3.0 2.0* 1.0 Isopoda 1.0 - -- -- 1.0* - Amphipoda 0.5 0.5 -- 3.0 1.0* - 12.5 Plecoptera LO.0 0.5 - - 0.5 Ephemeroptera 25.0 3.5 221.5 30.5 18.0* 12.0 21.0 4.0 ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 67

TABLE 7.-Continued

November- January- November Animal Group December February March- May- July- September- (1960)- June (1959) (1960) April June August October January (1961) (1961) Odonata - -- - 0.5- - - Megaloptera - 1.0* - 1.0 Trichoptera 6.0 -4.5 13.0 2.0* - 1.0 Coleoptera 0.5------Diptera 9.5 19.5 93.0 126.5 2.0* 2.0 10.0 3.0 Acri - - 1.5 -- - - Decapoda - -.5 1.0 0.5 1.00.5 Gastropoda 4.5 6.5 1.5 1.0 - - 1.5 - TOTALS 47.0 92.0 134.5 180.0 25.0* 15.5 49.5 7.0

STATION V: Gravel-Rubble Riffles Oligochaeta 0.5 1.5 1.0 0.5 2.0- Isopoda 96.5 1.5 3.0 5.0 1.0 23.0* Amphipoda 15.0 --4.5-- Plecoptera -- 2.0--- Ephemeroptera 32.0 4.5 15.0 13.0 20.0 66.0* Trichoptera 28.0 1.5 2.0 5.0 0.5 54.0* Coleoptera 3.0 1.0 - 0.5 0.5 2.0* Diptera 4.5 3.0 4.5 10.0 18.5 22.0* Decapoda 0.5 - -- Gastropoda -- 0.5 0.5-- TOTALS 180.0 13.0 28.0 39.0 42.5 167.0* on open riffles, though never so abundant nymphs made up an even greater part of as in 1 of 2 occurrences on moss (com- the total fauna than in marl riffles (69.7%), pare Tables 6 and 7). The greaterdiversity followed by dipterans, oligochaetes, and of the fauna at Station II and the presence isopods. There was a considerablygreater of some species of aquatic insects that diversityof fauna in this habitatthan in the apparently preferred moss habitats when marl riffles. Mayflies were most important small then moved to open riffles as mature in weight (46.2%) and were followed by nymphs (such as Ephemerellaand possibly plecopterans, dipterans, and caddisflies in Baetis) somewhat masks the numerical decreasing order of importance. The last comparisons of the animals. In relative 3 groups made up 42.3 per cent of the terms, however, the fauna of bare riffles average biomass. Standing crops ranged at StationII was characterizedby a greater from 13 to 1,324 individuals/ft2, and from proportion of aquatic insects, and a re- 0.005 to 0.752 g. duced number of crustaceans and turbel- Dipteran larvae made up 49.2 per cent larians. Animals utilizing soft substrates, of the fauna of small riffles at Station IV, such as those seasonallypresent in Fissidens with mayflies second (23.3%), and oligo- and in some habitats yet to be discussed, chaetes third (12.8%). The number of were also reduced in numbers in the un- animals at Station IV was lower than at vegetated riffles at both Stations I and II Station III, with 76 and 149 individuals/ft2, (see oligochaetes,Tables 6 and 7). respectively. By weight, however, they Riffles Lateral to Marl or Rubble. The were more similar (IV, 0.189 g; III, 0.199 average number of animals obtained in g). Oligochaetesmade up a majorportion samples from small riffles lateral to marl of the weight (32.8%), followed by may- at Station III was almost identical to that flies, dipterans, and amphipods. Variation from marl riffles, 149/ft2, but their weight in this habitat was high, ranging from 5 was only half as much (0.199 g). Mayfly individualsand 0.004 g/ft2 in October 1960, 68 WILDLIFEMONOGRAPHS

to 333 individuals in May 1960, and 0.688 whereas dipterans,mayflies, and caddisflies g in January 1961. Mollusks and decapods prevailed at Station IV. Standing crops at were represented by an overall average of Station V ranged from 7 to 260 individuals 1.165 g/ft2 (not included above). and from 0.009 to 0.743 g/ft2. Although habitats discussed here are The numbers of animals at Station V included under the same general category, were greatest at the end of modal flow in the conditions in each were somewhat dif- 1959, and again in June 1961. However, ferent. At Station III, the small riffles the small numbers present in the interim were continuously supplied with water, and the lack of samples preclude discus- although with considerable fluctuation in sion. It is significant that the gradual in- velocity and depth; however, at Station IV, crease in numbersof animalspresent in the the riffles were sometimes exposed by re- 1960 period of modal flow comprised ceding water levels and contained greater mostly riffle species (principally Simulium amounts of silt in interstices. Mayflies and and Isonychia). caddisflies both were more abundant, rela- Pools and Eddies with Generally Ag- tively and absolutely, at Station III than at grading Bottoms.-Nasturtium Beds. Beds Station IV, and the reverse was true of the of Nasturtiumpresent an almost-lenticen- burrowing forms such as oligochaetes and vironment. However, washout altered the dipterans. As at upstream stations, how- areas by removal of plants, so that average ever, the abundance of many of the organ- figures are actually a composite of more isms in the swift marl riffles was reflected than 1 type of habitat. Samplesfrom water- in the lateral riffles. For example, stone- cress were second only to Fissidens at Sta- flies were found on lateral riffles only in tion I in numbers and weights of animals, periods of relatively high flow, except for but greatly exceeded all other habitats if small individuals in late summer and au- weights of mollusksare included. Numbers tumn when leaves accumulated (mostly ranged from 24/ft2 in June 1961, to a maxi- Allocapnia and Nemoura, Table 7). mum of 2,652. The average for 15 samples Ephemeropteran nymphs showed similar was 893/ft2, of which oligochaetes were trends in abundance at both stations, in- most abundant (27.1%), followed by am- creasing during and just after winter dis- phipods and dipterans in almost equal charges and gradually declining during numbers, isopods, and caddisflies; other receding water levels and modal flow groups comprised less than 0.5 per cent (Table 7). The increase in abundance of each. The average biomass was 1.581 g/ft2 this group in winter is directly attributable (17.464 if mollusks were included). Am- to immigration from swifter riffles of the phipods were the most abundantby weight main channel, as indicated by Denham (47.8%), and were closely followed by (1938) in tributaries of the White River, oligochaetes (41.7%). Ind. The prolonged, severe floods of 1961 The large numbers of animals in this markedly reduced the numbers of indi- habitat were closely linked to the presence viduals and diversity of fauna in small of the watercress. Vegetation was totally riffles. The influence of high water was absent or covered less than 10 per cent of amplified at Station IV by siltation result- the bottom in March-April 1960, July- ing from clearing the banks. Few animals August 1960, and June 1961, and the small- of any kind were present on small riffles est numbers of animals were present in at Station IV in June 1961. those periods except for the extremely Standing crops of small riffles at Station small population in September-October V were almost the same as those at Station 1960. Crustaceans,primarily G. bousfieldi, IV in average numbers of animals, but were most abundant, but oligochaetes and slightly less in weight. Isopods and may- dipterans were least abundant in the pres- flies were most common at Station V, ence of plants and most abundant when ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 69 the plants were extirpated (Table 8). The Gastropods comprised 14.4 per cent of small number of animals present in Sep- the numerical total in Myriophyllum habi- tember-Octoberoccurred after Nasturtium tat of Doe Run, and were followed by had covered 20 to 50 per cent of the bot- dipterans, mayflies, and caddisflies. The tom, apparently effecting a decrease in combined weights of amphipods and iso- numbers of burrowing forms, and no am- pods were 60.3 per cent of the total. Dip- phipods or isopods were present (Table 8). teranswere third, followed by oligochaetes. Crustaceans were generally concentrated in The weights ranged from 5.138 g/ft2 in Fissidens beds at that time (Table 6). January 1961 to 1.063 in February 1960, The increase of oligochaetes and dip- and numbersranged from 1,364 in January terans in summerin the absence of Nastur- 1960 to 9,304 in August 1960. tium may be related to seasonal fluctuation The amazing resistance of milfoil to in the populations, as may the coincident, washout was reflected in the considerably marked increases in Fissidens beds and greater stability of the standing crops of other habitats. A relation between the organismsthan in other habitats (Table 8). appearance of Nasturtium and decreases As in Fissidens and Nasturtium,the num- in annelid and fly populations, however, bers of dipteransand oligochaetesincreased seems clear, and possibly may be explained in summerin the Myriophyllum,and both by oxygen deficiencies caused by accumu- decreased with the advent of autumn and lation of organic debris and the marked the developmentof almost lentic conditions slowing of water within the beds. Unpub- in some of the larger beds. Amphipods lished data of D. R. Tindall and myself were consistently abundant except in Sep- indicate that in daytime there is as much as tember-October and November 1960-Jan- 40 per cent less dissolved oxygen in beds uary 1961 samples;they were lowest in the of Nasturtiumthan in the water immediately former and highest in the latter period adjacent. At night, oxygen deep in the (Table 8). Isopods gradually decreased beds may reach levels that would select from a high of near 3,000/ft2 in November- against animals inhabiting those strata, but December 1950, to only 48/ft2 in Septem- would not influence those in the upper ber-October 1960, but again increased parts. during the 1960 modal flow to nearly 2,000. Myriophyllum Beds. Beds of milfoil It is interesting to note changes in abun- were about a third as productive as moss dance of caddisfly larvae on milfoil com- riffles at Station II on the basis of average pared with those on Fissidens at Station I. numbers of animals (3,423/ft2), and about Apparently, the lack of swift currents in a fourth as much on the basis of weight beds of milfoil, and possibly the relatively (2.820 g). However, with mollusks and soft nature of the beds, caused hydro- decapods included, there was 1.4 times as psychid larvae to avoid the habitat, and much weight in Myriophyllumas in Fissi- small Agraylea larvae predominated. The dens (32.217 g and 23.467 g, respectively). seasonal emergence of Agraylea is similar Isopods were numerically most abundant at Stations I and II, but decreases in the (30.4%)and amphipods made up 21.6 per larval population at Station II occurred cent. Unlike the Fissidens the habitat, ac- somewhat earlier in the season, presumably cumulationof and in silt, sand, sand-gravel as a to of the the mounds of was stable response slight warming Myriophyllum water and earlier enough to allow fairly large populationsof emergence. burrowing oligochaetes to persist. It is The effects of flooding in 1961 were less notable, however, that Krecker (1939), in evident in milfoil than in other habitats, studying animal populations on various but the fauna was nevertheless damaged, aquatic plants, found oligochaetes abun- with the lowest numbers recorded during dant in Myriophyllumsamples that did not June 1961. include substrates. Myosotis Beds. Forget-me-nots were ab- 70 WILDLIFEMONOGRAPHS

TABLE 8.-NUMBERS OF ANIMALS PER SQUARE FOOT IN VEGETATED, SILTED, OR OTHERWISE AGGRADING POOL OR EDDY AREAS IN DOE RUN, MEADE COUNTY, KENTUCKY. SEE TABLE 6 FOR FURTHER EXPLANATION

November- January- November Animal Group December February March- May- July- September- (1960)- June (1959) (1960) April June August October January (1961) (1961) STATIONI: Nasturtium Beds, or in the Area after Washout Turbellaria 4.0 10.0 4.0 Oligochaeta 24.0 356.0 382.0 268.0 466.0 130.0 192.0 Isopoda 220.0 72.0 12.0 146.0 22.0 12.0 34.0 4.0 Amphipoda 734.0 150.0 46.0 100.0 18.0 152.0 444.0 4.0 Ephemeroptera 2.0 ------Trichoptera - 2.0 198.0 Coleoptera - 2.0 4.0 -2.0 Diptera 62.0 40.0 328.0 764.0 236.0 116.0 64.0 8.0 Decapoda - 2.0 - - 4.0 - Gastropoda 58.0 46.0 54.0 58.0 112.0 60.0 252.0 8.0 Pelecypoda 16.0 26.0 8.0 5.0 122.0 8.0 28.0 TOTALS 1,120.0 694.0 832.0 1,553.0 976.0 482.0 1,020.0 24.0 STATIONII: Myriophyllum Beds

Turbellaria 5.0 - 10.0 2.0 6.0 74.0 Oligochaeta 21.5 18.0 126.0 634.0 3,712.0 916.0* 716.0 248.0 Isopoda 2,762.0 1,094.0 498.0 237.0 744.0 48.0* 1,858.0 124.0 Amphipoda 573.5 726.0 906.0 439.0 634.0 140.0* 1,636.0 388.0 Ephemeroptera 74.0 50.0 150.0 6.0 116.0 - 10.0 64.0 Odonata - 2.0 - 12.0* Trichoptera 9.0 120.0 136.0 4.0 20.0 32.0 Coleoptera 0.5 2.0 10.0 - 4.0 8.0 8.0 Diptera 7.0 10.0 72.0 144.0 836.0 44.0* 66.0 8.0 Decapoda 1.0 2.0 - 2.0 - 4.0 4.0 Gastropoda 37.0 304.0 270.0 514.0 1,058.0 1,256.0* 344.0 596.0 Pelecypoda 2.0 -4.0 48.0 26.0 8.0* 8.0 TOTALS 3,492.5 2,326.0 2,182.0 2,032.0 7,156.0 2,404.0* 4,836.0 1,440.0 sent at Station III only 3 of the 15 times and 5.776 g in March 1960. Fluctuationsin the area was sampled, and contained a abundance of the various animals were fauna similarto that in Nasturtiumbeds at similarto those in Nasturtiumbeds (Table Station I, and in pool habitats at other 8). stations. The standing crop was only 1.362 Silt-Bottomed Eddy. The thick silt de- g less than that in Fissidens habitat at posit was the most productive habitat Station II when the mollusksand decapods sampled at Station IV. There were 418 are included (22.105 g/ft2), and oligo- organisms on the average, that weighed chaetes made up slightly more than half 0.642 g/ft2 (or 5.647 g if mollusks are the total weight when mollusks and dec- included). Oligochaetes were most abun- apods are excluded. Diptera and amphi- dant by numbersand weight, similarto the pods comprise 41.5 per cent to include situation in beds of emergent vegetation, most of the remainder. Oligochaetes were and dipteran larvae were second. The likewise the most abundant numerically, standingcrop varied from 1,136 individuals averaging429/ft2 (44.9%of the total), and and 1.801 g/ft2 in June 1960 to 56 indi- were followed by dipterans, gastropods, viduals in January 1960 and 0.022 g in and amphipods.The smallest standing crop Februaryof that year. was in November 1960, with 208 indi- This eddy was similar to Fissidens and viduals and 0.385 g/ft2. Maximal popula- Myriophyllumin that it remainedessentially tions were 2,224 individuals in April 1960 unchanged throughout the study. Oligo- ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 71

TABLE 8.-Continued

November- January- November Animal Group December February March- May- July- September- (1960)- June (1959) (1960) April June August October January (1961) (1961)

STATIONIII: Myosotis Beds Turbellaria ------4.0 Oligochaeta 30.0 172.0 358.0 734.0 1,456.0 168.0 94.0* 92.0 Isopoda 2.0 12.0 44.0 4.0 4.0 - Amphipoda 74.0 146.0 246.0 36.0 46.0 12.0 14.0* 160.0 Plecoptera - 4.0 - - - - 4.0 Ephemeroptera 6.0 6.0 86.0 10.0 24.0 6.0 Odonata 6.0 4.0 - -10.0 -4.0 Trichoptera - -2.0 - - - Coleoptera 2.0 2.0 - - 2.0* Diptera 124.0 120.0 862.0 372.0 40.0 256.0 214.0* 12.0 Hemiptera - -- - 4.0 Decapoda 10.0 - 4.0 12.0 4.0 2.0* 12.0 Gastropoda 168.0 286.0 52.0 76.0 80.0 4.0 24.0* 80.0 Pelecypoda - 2.0 2.0 98.0 94.0 10.0 26.0* 4.0 TOTALS 422.0 750.0 1,660.0 1,330.0 1,752.0 474.0 648.0* 376.0 STATIONIV: Silt-bottomed Eddies Oligochaeta 588.0 2.0 124.0 496.0 192.0 46.0 434.0 196.0 Isopoda - 4.0 - --- Amphipoda - -4.0 20.0 2.0 6.0 4.0 8.0 Ephemeroptera - 60.0 2.0 4.0 2.0 2.0-- Odonata -- - - 2.0 Trichoptera 2.0 20.0 ------Coleoptera - 6.0 - - - Diptera 18.0 62.0 168.0 250.0 74.0 28.0 6.0 80.0 Gastropoda 6.0 16.0 16.0 131.0 44.0 48.0 28.0 32.0 Pelecypoda 20.0 2.0 4.0 6.0 - 12.0 10.0 4.0 TOTALS 634.0 172.0 318.0 907.0 314.0 142.0 484.0 320.0 chaetes increased in summer as in other (3.228 g if mollusks are included). Isopods habitats, but also were high in the latter and amphipods were almost equally repre- part of modal flow in November 1960- sented in the habitat (42.1 and 41.6%, January 1961 (Table 8). Diptera behaved respectively), followed by dipterans, snails, similarly, but did not show a significant caddisflies, oligochaetes, and turbellarians. increase in numbers late in 1960 as did By weight, dipteran larvae were first in oligochaetes. The fauna was generally importance because of a few Tipula, fol- consistent in its composition, perhaps more lowed in decreasing order by isopods, so than any other habitat sampled. caddisflies, amphipods, and oligochaetes. Open Pools and Eddies with Variable Because of the generally depauperate nature Bottoms.-Shifting Sand. Sand bottoms of the fauna of this area, little discussion are generally considered the least produc- seems warranted, however, it is noteworthy tive habitat in streams because of their that the fauna was numerically largest in constant shifting and consequent unsuit- periods of modal flow and smallest in ability for growth and development of periods of high discharge or following such vegetation and the lack of cover for animals periods (Table 9). Also, the size of the (Hynes, 1960). At Station I, the fauna of animals on such bottoms was generally sand bottoms ranged from 16 individuals small, since they were mostly young amphi- on 2 occasions to 724/ft2 in June 1960; the pods, isopods, and gastropods, with lesser average was 183 at a weight of 0.170 g numbers of other groups. These facts 72 WILDLIFEMONOGRAPHS

TABLE 9.--NUMBERS OF ANIMALS PER SQUARE FOOT IN OPEN-POOL AND EDDY HABITATS OF DOE RUN, MEADE COUNTY, KENTUCKY (INCLUDING THOSE WITH VARIABLE BOTTOM TYPES, BUT WITH NO AQUATIC VEGETATION). SEE TABLE 6 FOR FURTHER EXPLANATION

November- January- November Animal Group December February March- May- July- September- (1960)- June (1959) (1960) April June August October January (1961) (1961) STATIONI: Shifting-sand Bottom Turbellaria - -- - 2.5 10.0 4.0 - Oligochaeta 8.0 6.0 4.0 10.0 8.0 20.0 12.0 Isopoda 26.0 8.0 12.0 10.0 58.0 178.0 162.0 Amphipoda 20.0 - 12.0 10.0 8.0 62.0 56.0 4.0 Trichoptera - 2.0 - 102.0 2.5 24.0 6.0 Coleoptera - -2.0 - - - - - Diptera 14.0 6.0 - 118.0 64.0 26.0 2.0 20.0 Gastropoda 8.0 2.0 22.0 12.0 30.0 20.0 76.0 4.0 TOTALS 76.0 24.0 42.0 386.0 173.0 340.0 318.0 28.0

STATIONII: Pools; Bottoms Ranging from Silt to Gravel-Sand with Differing Discharges Turbellaria 6.0 - 8.0 Oligochaeta 26.0 10.0 40.0 248.0 68.0 438.0 88.0 20.0 Isopoda 20.0 4.0 132.0 2.0 12.0 2.0 16.0 Amphipoda - 6.0 170.0 14.0 10.0 4.0 48.0 Plecoptera ------2.0 Ephemeroptera 2.0 -2.0 - - 2.0 Trichoptera - 2.0 - 10.0 - - - - Coleoptera - - - 2.0 - - - Diptera 88.0 50.0 114.0 370.0 198.0 16.0 12.0 Decapoda - -- - 2.0 Gastropoda 2.0 2.0 6.0 52.0 52.0 62.0 418.0 496.0 Pelecypoda - -- - 58.0 2.0 10.0 TOTALS 140.0 74.0 470.0 696.0 410.0 884.0 264.0 516.0

STATIONIII: Pools; Bottoms Ranging from Silt to Sand-Gravel with Differing Discharges Turbellaria 2.0 - - - - Oligochaeta 186.0 27.0 86.0 524.0 454.0 244.0 504.0 116.0 Isopoda 24.0 2.0 4.0 6.0 - - 8.0 Amphipoda 68.0 2.0 166.0 70.0 2.0 - 10.0 Ephemeroptera 32.0 - 4.0 2.0 2.0 - 2.0 Trichoptera 2.0 - 4.0 2.0 - 2.0 Diptera 6.0 10.0 204.0 700.0 34.0 314.0 30.0 12.0 Hemiptera 2.0 - - - - Decapoda 2.0 ------Gastropoda 52.0 24.0 22.0 58.0 82.0 38.0 52.0 52.0 Pelecypoda 18.0 4.0 - 4.0 10.0 6.0 40.0 4.0 TOTALS 392.0 69.0 486.0 1,368.0 588.0 602.0 648.0 184.0

illustrate the dispersal of bottom inverte- ft2), but differed markedly in their faunal brates into peripheral habitats in periods composition. Oligochaetes, dipterous lar- of low discharge, and dispersal of young vae, and snails made up 66.4 per cent of rather than adults from areas of optimal the total fauna, but amphipods retained habitat. some dominance (26.2%). Isopods, how- Large Pools with Variable Bottoms.- ever, were relatively rare (4.7%). The Samples from pools at Station II contained standing crop by weight was small at 0.283 numbers of animals similar to those found g/ft2, of which oligochaetes made up 52.7 on unvegetated riffles at that station (531/ per cent. Amphipods were second in ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 73

TABLE 9.-Continued

November- January- November Animal Group December February March- May- July- September- (1960)- June (1959) (1960) April June August October January (1961) /1961) STATIONIV: Eddy; Gravel Bottom, Silted after Periods of High, Turbid Discharge Oligochaeta 36.0 4.5 19.0 37.5 8.0 12.5 21.5 220.0 Isopoda 1.5 - 0.5 0.5 --- - Amphipoda 2.0 2.5 2.5 2.5 2.5 3.0 - - Plecoptera 0.5 0.5 ------Ephemeroptera 6.5 7.0 3.5 10.5 3.0 2.5 3.0 4.0 Megaloptera - 1.0 ------Trichoptera 1.0 2.0 - -0.5 --- - Coleoptera - 0.5 - -1.0 - - Diptera 16.5 6.0 82.5 15.5 9.0 34.0 23.0 12.0 Hemiptera - - 0.5 -- - Decapoda - 0.5 - 4.0 -0.5 - - Gastropoda 12.5 3.0 10.0 6.0 2.5 8.5 22.0 76.0 Pelecypoda 0.5 -- 0.5- - TOTALS 77.0 27.5 118.0 77.5 25.5 62.0 69.5 312.0 STATIONV: "Pool"; Various Bottom Types in Stream Channel Oligochaeta 43.0 2.0 -4.0 4.0 1.0 Isopoda - 4.0---- Ephemeroptera 3.0 ----- Trichoptera - 2.0 - - 2.0 - Coleoptera 2.0 2.0---- Diptera 16.0 4.0 2.0 60.0 58.0 87.0 Gastropoda 2.0 - 2.0 - 14.0 - Pelecypoda 1.0 ----- TOTALS 67.0 14.0 4.0 64.0 78.0 88.0 weight followed by dipterans and isopods. remained relatively abundant in all subse- The standing crop ranged from 32 to 1,412 quent samples. The greatest abundance of individuals, and from 0.010 to 0.598 g/ft2; amphipods and isopods in pools at Station this range, however, was not as great as in II and amphipods at Station III occurred unvegetated riffles. during the winter discharge of 1960. These An average of 566 individuals/ft2, with data indicate drift of the crustaceans down- a mean weight of 0.493 g if mollusks and stream and accumulation in pools with crayfish were excluded, was obtained in sustained but moderate flooding. The samples from the large eddying pool at rapid numerical decrease in pools with Station III. The fauna of this habitat was resumption of modal flow may indicate similar to that of pools at Station II in movement of isopods and G. minus back composition, both by numbers and by into riffle areas. weight. Ranges were from 56 to 1,776 Bottoms in pool habitats at Station V organisms in February and late April ranged from solid clay to shifting sand. (= "May") 1960, respectively, and from The standing crop varied in response to 0.028 to 1.449 g/ft2 in those same months. these conditions, and ranged from no Diptera at Stations II and III and organisms from clay bottoms on 2 occa- Oligochaeta at Station II, as in pools as sions, to a maximum of 172 animals in elsewhere, increased in summer and de- November 1960 and 0.292 g in September creased in autumn and winter (Table 9). 1960 (the latter was the weight of a single At Station III, however, numbers of oligo- large Pychnopsyche). Also, the fauna was chaetes increased in June 1960 after alle- limited in taxons represented (Table 9). viation of winter discharge and they Faunal Affinities of Doe Run.-Springs are 74 WILDLIFEMONOGRAPHS sometimes unique in possessing relict or crustaceans. It must be noted, however, endemic faunas (Nielsen, 1950b). Tem- that this group is the only one studied in- perature moderation in a spring, with tensively from a taxonomic standpoint, yet warming of water downstream in summer, the presence of other undescribed species but maintenance of cool conditions up- is probable. Populations of Gammarus stream, allows development of a recapitu- minus and Asellus bivittatus in the stream lation of south-to-north distributions similar system are aberrant in some respects, and to those found on mountains. Such distri- different from other local populations (Cole bution of animals in a spring stream such and Minckley, 1961). as Doe Run is effected by the physiological Although not unique to Doe Run itself, limitations of the species present. Low the highly specialized, cavernicolous ani- temperatures themselves generally are not mals of the limestone region of Kentucky detrimental to the well-being of an animal and adjacent states should not go unmen- of temperate zones, but the maintenance of tioned. Three of the many forms, Asellus cool temperature may limit completion of stygius, Orconectes p. pellucidus, and a the life cycle so that its distribution in the species of Phagocata, were collected from stream will be limited by the minimum Doe Run, and Amblyopsisspelaea DeKay, temperature tolerance for growth, devel- a blind cavefish, has been collected from opment, and reproduction, and by the subterranean waters of Meade County minimum lethal high temperature (Ide, (Clay, 1962; Barr and Kuehne, 1962). 1935). Although some animals of Doe Run are Fishes and OtherVertebrates widespread, ubiquitous, or southern in As with many collecting techniques, their geographic relations, many are north- those used in studies of fishes have little ern in their distribution or are "spring quantitative basis except where a body of inhabitants" in the southern parts of their water can be completely drained and the ranges. The turbellarians Phagocata velata animals counted individually. "Minnow and Dugesia dorotocephala? (Pennak, seines" tend to catch smaller fishes, except 1953), the Crangonyx group of amphipods, those small enough to go through the mesh and amphipods of the subgenus Rivulogam- (Paloumpis, 1958), the use of electric marus (see Bousfield, 1958) were restricted shockers tends to select for larger, light- to the upstream parts of Doe Run. Of the colored fishes (Larimore, 1961), hoop and mayflies, Ephemera simulans usually is a gill nets provide various sizes of catch northern or northeastern form, whereas depending on mesh size and the kind of Baetis vagans, Ephemerella subvaria, and "set" (Starrett and Barnickol, 1955), and perhaps Pseudocloeon carolina (see Ide, rotenone is limited somewhat by the dif- 1935), may represent relict or outlying ferential susceptibility of various species populations considerably removed to the (Krumholz, 1948; and others). Larimore south or the west of their normal ranges (1961) concluded that behavior of fish to Ricker (Burks, 1953). According (1959; species, their diverse habitats, and their pers. comm.), both media and Paragnetina morphology all directly affected efficiency subvarians are "Appalachian and Isogenus of electrofishing in Jordan Creek, Ill., northern species" of stoneflies. Glossosoma and those seasonal in is a northern trichopteran inhabiting springs factors, plus changes in the southern parts of its range, as does behavior and distribution, changing condi- in Diplectrona modesta (Ross, 1953, 1959). tions of collecting (such as variations Also Hydropsyche betteni and Agraylea discharge and turbidity), and relative ef- multiplicata are northern forms, the latter ficiency of the collectors, combine with being holarctic in distribution. other factors to make quantitative interpre- A relatively great geologic age for Doe tation of data on diverse stream populations Run is implied by the presence of endemic difficult. ECOLOGYOF DOE RUN, MEADE COUNTY,KENTUCKY-Minckley 75

Annotated List of Fishes for the most upstream record, and the other by rotenone downstream from Station A total of 66 of fishes was col- just species IV in 1961 (Fig. 26). lected from Doe Run. Estimates of the July Dorosoma (LeSueur): giz- relative abundance of each are cepedianum species zard shad. This species was present at all based mostly on frequency of occurrence times downstream from Station V, and prob- in collections, and to a lesser degree on the ably occurred in small numbers throughout actual numbers collected. About 250 col- the low-gradient section (Fig. 2) below lections or observations of fishes were made, Mile 5.3 (Fig. 26). The gizzard shad was and those in half or more species occurring collected as far upstream as Station III in of the collections are considered abundant; the spring of 1960 and 1961, but did not in more than a third of the collections, but ascend swift water in culverts of the bridge less than are in more than a half, common; at that station. Shad were watched ascend- tenth, but fewer than a third, scarce; and ing very swift waters at Station V in May in fewer than a tenth, uncommon to rare. 1961, however. They milled about in Some in species restricted portions of the eddies at the foot of the riffle and as- creek may have been common or even cended in a rush, directly in the center of abundant in relation to other species in the swiftest flow and very near the surface. those limited areas, but cannot be so Gizzard shad appeared to be spawning considered in the entire population. Num- in a large pool near Mile 3.75 on 5, bers of individuals alone were not a June 1960. About 25 individuals were reliable index of overall abundance because moving rapidly around a small, stream-drifted tree of sporadic of great numbers of emigration in the center of the pool, and presumed certain species from the Ohio River. My males were most often positioned on either collections of fishes from Doe Run were side of a female (Langlois, 1954). Water culminated by eradication of the area temperatures at the upstream end of the downstream from Station IV with rotenone pool were near 16.0? C, which may be in 1961 prior to impoundment. July just somewhat low for spawning by the species However, data collected subsequently are (Miller, 1960), but water in the center of included where pertinent. Comparisons of the pool have been warmer. relative abundance of some fishes of the may slightly No eggs were found, however, and larval lower of Doe Run with that in the part shad were never obtained from Doe Run. Ohio River are made in reference adjacent Trautman (1957) noted movement of adult to data published by Krumholz, et al. gizzard shad into clear, low-gradient waters (1962) and to unpublished data collected for spawning in Ohio, but rarely into high- in 1957, 1958, and 1959 by Krumholz and gradient streams. myself. I watched a number of D. Lepisosteus osseus (Linnaeus): longnose cepedianum feeding along a bank in about a foot of gar. A single longnose gar was seen near water and over compact clay bottom at Mile 8.5 after application of rotenone, but Station V. They moved in a somewhat was not captured. The species is present in much of the Ohio River and often is spiral path with their bodies angled about found in creek mouths, especially as young 45 degrees to the substrate. A given indi- individuals. vidual stayed near the bottom for perhaps Alosa chrysochloris (Rafinesque): skip- 3 to 5 feet, then moved upward and for- jack herring. Small numbers were in the ward to repeat the performance. Appar- mouth of Doe Run in periods of flooding ently they were selecting certain items as of the Ohio River, and several were seen judged by clouds of silt expelled from the pursuing minnows near Station V in June opercles at short intervals. Turbidity pro- 1961. Two specimens were taken upstream duced by the feeding aggregation was ob- from Station V, 1 by shocker near Mile 4.0 vious for about 25 feet downstream. 76 WILDLIFEMONOGRAPHS

MILE I 4 5 6 7 8 9 1 - I I I I~ Catostomus commersoni ~ry7 ?t???r//j/?-?l?- -r------I- - 4- I I Cottus carolinae S*W/*PgER+//Wi/fifXRRW/S/EPmMiSS/iZEBS/WUXXhmSREWERS/anM Rhinichthys atratulus s.-o,_1 Semotilus atromaculafus x I 2A- Sc/mo gairdneri Lepomis cyanellus J - - .. Etheostoma f/abel/lare I Pimephales notatus .=- ? . 1 ...

Notropis atherinoides I _- t-7f......

N. spilopterus - mO - 7~

Hypentfeium nigricans I!/P-/ 7 "-III Ameiurus melas 0** 0 0 I l- Lepomis megalotis - 5f5757/ m,57

/e mv L. macrochirus 0 _.rW "-w*zy,Z . -

Micropterus punctulatus IX.M _talc crrrT 7 _ M Campostoma anomalum t- ' MD Dorosoma cepedianum Etheostoma blennioides E. caeruleum '---_

Minytrema melanops 0@ Notropis chrysocephalus _0 Moxostoma breviceps M. erythrurum -_ am am M E Cyprinus carpio Moxostoma duquesnei Carp/odes carpio Aplodinotus grunniens 0. -0- - Micropterus dolomieui I Amblop/lies rupestris NVotropis blennius 0. _ _ I N. volucellus Micropterus salmoides 0 Ictalurus punctatus Alosa chrysochloris ^ I I _ I _mi io Sta. I Sta. II Sta. Ii1 Sta.IV Sta.

FIG. 26. Longitudinal distributions of some fishes in Doe Run, Meade County, Ky. Broken lines denote rare or occasional occurrence; width of the line, the presence (wide) or absence (narrow) of reproductive success in the area; vertical lines, occurrences in tributaries; dots, occurrences of single individual or of fewer than 5 individuals in no more than 2 collections from the area; and open circles, occurrences when backwaters from the Ohio River were present.

Eradication of fishes from the lower part cofferdams, and removed much of the fill of Doe Run was aimed principally at re- to form a short, swift rapids down the rip- moval of this species and the carp (Cyprinus rap of the downstreamcoffer. On July 18, carpio L.), but must have failed because a number of large carp were observed many young-of-the-year individuals of both ascending these rapids, and gizzard shad species were seen in Doe Valley Lake in may also have used this route, though none August 1961. This may have resulted was seen. from failure to kill all adults upstream Salmo gairdneri Richardson: rainbow from the new dam site, but it must trout. Trout, presumably of this species, be pointed out that subsequent to the were stocked in Doe Run in 1939 (Brown, rotenone treatment, flooding on July 16-17, pers. comm.), with additional stockings in 1961, filled the incipient lake, topped the May 1955 and in mid-October 1961 after ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 77 impoundment of the downstream section. tures exceeding 26? C in daytime, and A small population of trout may have per- there was a marked thermocline below sisted from the 1939 stocking through which there were only small amounts of naturalreproduction. In spring 1958, while dissolved oxygen though temperatureswere collecting for the University of Kansas, I considerablylower (17-18? C). There was saw a large trout near Station III, and an- a small density current, however, where other individual, much larger than those water of Doe Run flowed in its original knpwn to be from the 1955 stock, was seen channel, that may have been suitable for a at Station IV in spring 1961. few trout. Conditionswere therefore such Generally,the range of trout in Doe Run that most trout were forced to move into was from Dam 1 downstreamto immedi- Doe Run proper, and considerablecrowd- ately below Station III (Fig. 26). A single ing occurred in pools and eddies. Move- specimen was seined at Station I in 1959. ments noted on October 17 were mostly The fish were found below waterfalls to a upstream over riffles, with the small fish great extent, where they lived beneath rapidly returning downstream when dis- overhangingledges, but a few were in open turbed. At least 2 screens were erected pools and were taken in hoop nets from the across the creek by the Doe Valley Corpo- mill impoundments. Many of the 1955 ration to prevent upstreammovement, but stock were present near Station II in 1959, they soon clogged with leaves and other but in succeeding years the population de- debris and became relatively ineffective. clined more rapidlythan could be explained Small trout were found at Station III on by fishing pressure. Moreover, in spring October 18. Subsequent increases in dis- 1959, trout were taken frequentlyat Station charge dislodged the screensand many fish III, but they apparentlymoved downstream moved into the area near Station III. The and into the Ohio River since none was presence of so many fish in such a small taken subsequentlyat Station III or below. area apparently attracted many predatory They could not have moved upstream animals,and losses the first few weeks after because of mill dams. Newell (1957) found stockingwere heavy. Numeroustrout were that stocked rainbow trout moved less than found along the banks partially eaten, pre- brook trout, Salvelinusfontinalis (Mitchill), sumablyas a result of mammalianactivities. but that most movement by each was Hiodon tergisus LeSueur: mooneye. A downstream. One individual in Doe Run, single mooneye was taken by rotenone fin-clipped in January 1960 near Mile 1.8, downstreamfrom Station V in 1959. was taken 7 times in June and July 1961 in Esox americanusvermiculatus LeSueur: Pool 2, and each recapturewas progressively grass pickerel. Two pickerel were caught, farther downstream. In June 1961, it was 1 by gill net in a large, shallow pool just caught in a hoop net from Pool 3, near Mile downstreamfrom Station IV, and the other 2.7, and was retained. below Station IV by rotenone in July 1961. I detected no successful reproductionby Carpiodes carpio carpio (Rafinesque): rainbow trout of the 1955 stock, but a river carpsucker. This large, deep-bodied single redd was found at Mile 1.68 in early catostomid is generally found in deeper May 1960; slight digging in the shallow waters of the Ohio River (Trautman,1957), gravels yielded 2 fertilized eggs. An in- but often occurs in creek mouths through- crease in dischargein mid-Mayscoured the out the Kentucky section of that stream. bottom and destroyed the redd. River carpsuckers occurred sporadically On October14, 1961, about 65,000finger- upstreamas far as Mile 5.0, that is, to the ling rainbow trout were introducedinto the area of the beginning of high gradient,but upstream part of Doe Valley Lake by the probably were present at all times in the Doe Valley Corporation(near Station IV). downstream area below Station V (Fig. At that time the lake had surface tempera- 26). This and other carpsuckersapparently 78 WILDLIFEMONOGRAPHS moved into Doe Run in considerable num- certain catches in gill nets. A group of at bers when backwaters of the Ohio River least 17 adults was seen beneath a log at were present. Mile 0.4 in June 1961. They rested near the C. cyprinus hinei Trautman: quillback. bottom with their fins extended, maintain- Only young quillback were taken in Doe ing their positions in the slight current by Run (young-of-the-year and presumed 1- regular, measured strokes of the pectoral year-old individuals), and most were caught fins; no movement of the caudal fin was below Station V. seen. When disturbed, aggregations in C. velifer (Rafinesque): highfin carp- pools move quickly into debris or other sucker. This species was collected by cover (Fowler, 1906). It may be that rotenone near the mouth of Doe Run in "schools" of this type in streams are more a 1959. function of common use of a given area of Catostomus commersoni (Lacepede): cover and tolerance of the species for white sucker. White suckers were abun- others of its kind rather than an actual dant in Doe Run, especially at Station II schooling phenomenon indicated in studies and upstream from Dam 1. This species of lake populations (see Adams and Hank- is to be the subject of a later, more detailed inson, 1928). Young suckers in the stage paper dealing with its life history, but a that involves surface feeding (Stewart, general review of its relations in Doe Run 1926) likewise appear to congregate in re- is given here. sponse to gentle currents in eddies and White suckers exhibit amazing tolerance backwaters. to diverse environmental conditions (Traut- In late 1959 and in summer 1960, 103 C. man, 1957; and others), but usually reach commersoni were marked by finclipping or their peak abundance in headwater creeks by sewing small, numbered beads to the in Indiana (Gerking, 1945), and in the area caudal peduncle with monofilament line. Of of Doe Run. Numbers in Doe Run were these, 19 were recovered at least once after greatest in the deeper pools, especially in being free from 1 day to 11 months and 4 the mill ponds as judged by hoop netting, days, a total of 28 recoveries of the 19 indi- with lesser numbers occurring in swifter, viduals. Only 3 individuals had moved shallower waters. In vegetated areas up- more than 0.2 mile from the point of mark- stream from Dam 2, C. commersoni was ing, and were the only ones that had moved often in shallow water with appreciable over substantial riffles. The fish in Pool 2, current, but usually in the shelter of vege- where most hoop netting was done, were tation (see Larimore, 1961). Downstream, more motile than those in the pool-riffle however, where vegetation was lacking, the area between Miles 1.65 and 2.1. Stream suckers lived in deepest pools or in smaller, fishes generally regard riffles as barriers to shallow pools that had accumulations of movement, or as limits of their home ranges debris or overhanging riparian vegetation. (Gerking, 1953), and the lack of obstruc- White suckers generally swam ahead of the tions in the mill pool probably allowed electrical field of the shocker, as noted by greater freedom of movement. Total linear Larimore, et al. (1952), but would not movements, that is, distances on either side cross riffles and were quickly affected by of the point of initial marking, were 0.26 the charge as they turned into the field mile (the total length of Pool 2) for a male (Minckley and Craddock, 1961). Indi- marked at Mile 2.15, taken at Mile 2.1 a viduals in beds of vegetation moved vio- week later, and then at Dam 2 (Mile 2.36) lently when "hit" with the electric current, on the succeeding day. Two fish marked in became immobilized, and usually floated Pool 2 (Mile 2.1) were recovered at Miles belly-up. 1.95 and 1.7 a week and almost a year, Schooling of large suckers was seldom respectively, after the initial catch. The seen, but was indicated by bunching of former had passed over a riffle and a ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 79 waterfall about 2 feet high, and the latter stage (ranging from 12 to near 20 mm in had moved over several riffles and 3 water- total length; Stewart, 1926). Males began falls that collectively amounted to about 6 to display breeding colors and tubercles in vertical feet. The latter individual had the Januaryeach year of the study, and some opportunityto move over obstructionsin the "high" males were present through June floods of 1960 and 1961. Fish marked in 1960, although high water in 1961 pre- the pool-riffle area at Station II usually vented observation. Both males and fe- were recovered in the same pool in which males stripped easily from February to they were initially taken, but 4 had moved early July 1960 (fewer after mid-June), over small riffles to an adjacent pool (3 and females with fully formed ova were upstream, 1 downstream), and 2 had as- presentthroughout the year, but in reduced cended waterfalls to the pool immediately numbers between July and December. upstream. These data substantiatethe con- Some females with ripe ovaries in late clusion of Gerking (1953) in his statement summer may have matured late in the that "the fish population of a small stream spawning season-they were usually small with riffle-pool development may be con- -or egg formationmay occur more rapidly sidered as a series of discrete, naturalunits in fish of Doe Run than in others with rather than a single homogeneous, freely- which I have had experience. Actual mixing group." However, his conclusion spawning was not observed, but shallow was based on species other than the white depressionsin a small area of coarse-sand sucker, and there is evidence that it and bottom about 40 feet downstreamfrom the some species studied by him in Indiana source of Doe Run yielded eggs and lar- may behave differently in certain parts of vae of white suckers on May 14, 1960, and Doe Run. It is significant,however, that at again 2 weeks later. The latter was a dif- Station II there was a strong tendency for ferent spawning as judged from the de- white suckers to remain in restricted areas velopmental stage of the eggs. for relatively long periods. It is noteworthythat suckersmore or less Between Station I and the head of Pool "trapped"above the mill dams of Doe Run, 1 there were few areas that could be especially upstream from Dam 1, do not classed as pool, and suckerswere not found experience marked changes in water tem- upstreamfrom Mile 0.4 in periods of modal perature. This factor is considered very flow. In periods of higher discharge and important in the spawning cycle of many turbidity suckers were commonly taken in species (Adams and Hankinson, 1928), seines at Station I, and they spawned there, including white suckers,but it seems hard presumablyat night. I have little informa- to believe that the magnitude of change at tion on movements of suckers downstream upper stations of Doe Run is sufficient to from Dam 2, but it is significant that no initiate gonadal maturationand spawning. marked individuals from Station II were Reproductionin response to changing light obtained by hoop nets or other means in conditions is indicated for species main- Pool 3, or in extensive collecting by seine, taining populations in the upper parts of shocker, or rotenone application down- Doe Run (see Harrington, 1950, 1957; stream from Dam 3. I found no evidence Hoar, 1957). of upstream movement of suckers for As noted before, the mill dams of Doe spawning or for other reasons downstream Run are more than 100 years old, and have from Dam 3. acted as a barrier to upstream movements Young-of-the-year white suckers ap- of fishes since their erection. Also, a water- peared at Station I in mid-April 1960 and fall at least 6 feet high was present near were found there and at downstream sta- Mile 2.1 prior to flooding by Dam 2, and tions (excepting Station V) from that time may have acted as a barrierto fishes for an through early July in the surface-feeding unknownperiod of time priorto disturbance 80 WILDLIFEMONOGRAPHS

by man. Suckers from the upstream section thereabouts (Fig. 26). Spawning upstream of Doe Run appear to be separable from may be restricted by the general lack of those in the downstream sections and in gravel bottoms. Hypentelium must be con- the Ohio River by a number of morphologi- sidered a scarce component of the fauna cal characters. Research on this problem is of Doe Run when compared to other now under way. species, but was common in the area in Hypentelium nigricans (LeSueur): north- which it occurred. ern hog sucker. Although Trautman (1957) Ictiobus bubalus (Rafinesque): small- considered the northern hog sucker highly mouth buffalo. Immature I. bubalus were migratory, spending winters in deep slow- present in the area downstream from Mile moving streams and migrating into creeks 6.7, and most common when backwaters in spring to spawn, my data indicate a from the Ohio River were present. continuous population between Miles 3.3 I. cyprinellus (Valenciennes): bigmouth and 7.0, with no extensive movements. buffalo. A large individual was affected by Young hog suckers were collected in late rotenone near Station V, but was not cap- August outside the area of greatest abun- tured. dance (Fig. 26), and some adults were Minytrema melanops (Rafinesque): taken at Station III. Gerking's (1953) data spotted sucker. Minytrema was rare, oc- from Richland and Stott's creeks, Ind., im- curring in the low-gradient section down- ply that hog suckers inhabit a definite stream from Miles 5.3 to 5.6, and as 2 or 3 home range usually consisting of a single specimens upstream (Fig. 26). Individuals large pool. In an earlier study (Gerking, from Doe Run were generally large, but a 1950) he stressed the stability of stream few were less than 6 inches total length. populations of fishes, including hog suckers, The population may have been maintained even after major flooding. In Doe Run, no by emigration from the Ohio River, where hog suckers smaller than about 25 mm total it is generally rare in the mainstream, or length were collected, but individuals from reproduction occurred in the part of Doe that size to about 100 mm were usually in Run downstream from Station V and young shallow riffles or in eddies over clean did not move upstream until considerably gravel, sand, or even bedrock bottoms. larger. Minytrema was invariably in the Large adults showed a marked preference largest, deepest pools as adults, but smaller for deeper pools, where they were often individuals sometimes were in the lower shocked from beneath semifloating deposits parts of slow riffles or in open, shallow of stream-drifted detritus. Large fish sel- pools over silt-gravel bottoms. Its greatest dom were found in appreciable current, abundance was near and below Mile 7.0 except in areas where 2 riffles came to- over soft bottoms. Preferences for soft, silt- gether around an obstruction to form a detritus bottoms in streams were noted by strong eddy. Several large H. nigricans Gerking (1945), but Trautman (1957) con- were surprised early one morning in May sidered the species more characteristic of 1960, on a riffle near Mile 3.75 where they lakes or base-gradient streams with hard may have been feeding. Since the species bottoms of sand, gravel, or clay. generally spawns in shallow, cuplike de- Moxostoma anisurum (Rafinesque): sil- pressions in gravel (Gerking, 1950) or on ver redhorse. Two specimens were taken open gravel riffles, and sometimes in pools by rotenone in the extreme mouth of Doe (Hamilton, in Raney and Lachner, 1946), Run, and 2 others were obtained by rote- it seems unlikely that they were spawning none near Mile 7.0. on solid marl. Judging from the distribu- M. breviceps (Cope): shorthead redhorse. tion of young hog suckers in summer, This species was uncommon in Doe Run, breeding occurred in a somewhat limited but occurred most frequently in collections area between Station IV and Mile 7.0 or between Miles 5.5 and 7.0 (Fig. 26) where ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUcKY-Minckley 81 it inhabited areas of moderate current in vidualswere collected at StationV. Golden shallow, open pools or at the foot of riffles, redhorsein Doe Run did not seem as easily and was somewhat more abundant in the shocked as those reported from Jordan upper part of that section. It was almost Creek, Illinois, by Larimore (1961), pos- always over clean gravel bottoms. The sibly because of deeper pools in Doe Run species was present in Doe Run throughout and more variable conditions of collection. the year, and there was no evidence of a It is of note, however, that intensive elec- "run"of shortheadredhorse into the stream, trofishing in a given pool apparently and no reproductiveactivity was indicated depleted the populationof golden redhorse by condition of gonads or by occurrenceof there; subsequentcollections rarely yielded young fish. as many, and sometimes none in those I watched a number of M. breviceps pools. This may indicate relatively seden- feeding in a shallow pool near Mile 6.0 in tary habits. Gerking (1953) found it re- June 1960. They moved with their bodies sembled the hog suckerin distances moved held at a slight angle to the bottom in a in Richland Creek, Ind., quite different compact group,flicking the fins rapidlyand from the contention of Trautman (1957) pushing between small stones and gravels that the species is "highlymigratory." The with their snouts. This behavior contrasts species inhabited the same range in Doe with my observations of the related M. Run as M. breviceps (Fig. 26), but was aureolumpisolabrum Trautman and Martin more often in pools and over soft, silty in the Marais des Cygnes River, Kansas,in bottoms. 1958. Those fish held their bodies almost Ameiurus melas (Rafinesque): black perpendicularto the bottom in a shallow bullhead. I chose to retain the generic riffle and actually rolled stones and gravel name Ameiurusfor the bullhead catfishes, while feeding. M. a. pisolabrumhas a pro- at least until data are published substantiat- tuberance,or "pea"on its upper lip (Traut- ing synonymization of that genus with man and Martin,1951), which may function Ictalurus by Taylor (1954). The black in this type of feeding. Hubbs and Lagler bullhead is commonin most sinkholeponds (1958) indicated that the name aureolumis of Meade County that I have seined, but unavailable in the genus Moxcastoma,but was rarely taken in Doe Run and did not failed to give reasons for their action; it is reproducethere, at least not upstreamfrom retained here pending detailed clarification Station IV. Specimens were taken at all (Minckley and Cross, 1960). stations, but the numbers upstream were M. duquesnei (LeSueur):black redhorse. especially small. Hoop nets set in Pool 2 Black redhorsefrom Doe Run all exceeded caught 6 black bullheads,all of which were 10 inches total length and usually were as- marked and released only to be caught sociated with M. erythrurumin the deepest repeatedly without addition of new indi- pools. They were most common between viduals to the catch. Bullheadsare difficult Miles 5.0 and 7.0 (Fig. 26), but were a rare to seine because of their bottom-dwelling component of the fish fauna. This species habits and the tendency to live in deeper never was found in swift waters of Doe pools amidst debris. Shocking is relatively Run. It was considered a semimobile ineffective because they stiffen and sink species in Missouristreams by Funk (1957). to the bottom ratherthan moving into sight M. erythrurum (Rafinesque): golden (Larimore, 1961). The lack of specimens redhorse. This species was the most abun- in collections downstreamfrom Station IV dant redhorsein Doe Run, but was no more must be attributedto these factors because than an uncommon component of the substantialnumbers were found in 2 large fauna. Only large individuals, from 6 to pools near Mile 5.0 during the rotenone 18 inches total length, were found upstream operation of July 1961, and occasional from Station V, but young-of-the-yearindi- specimenswere obtained down to Mile 7.5. 82 WILDLIFEMONOGRAPHS

The species presumably occurred through- Marl deposition between the 2 locations out the downstream area but is uncommon may have prevented spawning. The species in the Ohio River at this point. I believe constructs nests at the heads of riffles on that those specimens obtained from the mill clean gravel bottoms (Langlois, 1937; Len- ponds passed through underground waters non and Parker, 1960), sometimes markedly into Doe Run, or entered through the Rush- altering the bottoms in the process. ing River system. I have seen black bullheads Stonerollers in Doe Run inhabited shal- deep in caves in Breckinridge County, lower pools, swift eddies downstream from Ky., and elsewhere. Populations in the marl riffles, and marl- or gravel-bottomed downstream part of Doe Run may come riffles in spring. Young individuals were from the Ohio River, or some reproduction most often taken on riffles, usually in the could have occurred in the mouth with most turbulent parts. Metcalf (1959) im- subsequent movement upstream (Fig. 26). plied an association of young stonerollers A. natalis (LeSueur): yellow bullhead. with more turbulent areas in Spring Creek, Two yellow bullheads were taken in hoop Kansas, rather than with the smoother-flow- nets in Pool 2, 1 was preserved and the ing riffles. other fin-clipped. The latter was recovered About 85 per cent of the stonerollers col- again near Mile 7.0 by rotenone in July lected from Doe Run were more than 3 1961. It had been free for almost a year inches total length, but none exceeded and had moved downstream a distance of 7 inches. I consider large stonerollers more 4.9 miles between captures. difficult to catch than small, at least by Ictalurus punctatus (Rafinesque): chan- seine, but the reverse is probably true for nel catfish. Channel catfish were very rare collections made by shocker (Larimore, in Doe Run, occurring only downstream 1961). Nevertheless, the paucity of young from Mile 6.5 with the exception of 1 small fish in collections indicates a marked dif- individual taken near Mile 5.3 (Fig. 26). ference in reproductive success of the According to fishermen in the downstream species in Doe Run when compared with area, the species was never abundant in other localities (Minckley and Cross, 1959; Doe Run, but was commonly caught in Metcalf, 1959; Lennon and Parker, 1960; other stream mouths in the vicinity. It is and others). abundant in the adjacent Ohio River. Chrosomus erythrogaster (Rafinesque): Noturus flavus Rafinesque: stonecat. A southern redbelly dace. This species was single N. flavus was collected by electric represented by 2 specimens from Station shocker near Mile 6.0. II, a single individual from immediately Pilodictis olivaris (Rafinesque): flathead downstream from Station III, and many catfish. Few flathead catfish were col- specimens from Tributary IV-e. In the lected, and most were in the area down- latter, C. erythrogaster inhabited the downstream from Station V. The most upstream stream portion of the small rubble-bottomed record was a juvenile obtained from a large stream, which was destroyed by impound- pool at Mile 5.3. ment of Doe Run. The stream was perma- Campostomaanomalum anomalum (Rafi- nent in the sense that it was spring fed and nesque): stoneroller. Stonerollers were summer pool levels were maintained by caught from Station III downstream to ap- subsurface percolation of water through proximately Mile 5.5, and occasionally at coarse bottom materials. Station V. Highly colored, tuberculate Ericymba buccata Cope: silverjaw min- males were found in April and May, but now. Three silverjaw minnows were ob- young-of-the-year individuals were scarce tained, 1 each from Stations II and V, and in Doe Run, and occurred only downstream the other near Mile 7.0 by rotenone. The from Mile 5.0 and at Station III where species is locally common in creeks near gravelly bottoms were present (Fig. 26). Doe Run, but rare in the Ohio River. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 83

Hybognathus nuchalis nuchalis Agassiz: 5.3. Large schools of emerald shiners silvery minnow. This species was rare in usually swam ahead of an electric field, Doe Run, and occurred downstream from and the bottom of the stream became liter- Station IV. The greatest numbers were in ally silver as the fish turned into the field. deep pools near streamside debris such as Emerald shiners were especially abundant tangled roots or drift. No specimen smaller in Doe Run when backwaters of the Ohio than 4.0 inches was obtained, and many River were present, and large schools re- approached the maximum size of 6 inches mained in pools of the drying creek in 1961, given by Forbes and Richardson (1920) for downstream from the impoundment. Illinois specimens. Reproduction by N. atherinoides in Doe Hybosis amblops amblops (Rafinesque): Run apparently was limited to the area bigeye chub. Two specimens were ob- downstream from Station V; individuals tained in the rotenone operation of July taken upstream at all times of year ex- 1961. They were not recognized in the ceeded 2.5 inches total length. field but were presumably obtained near N. blennius (Girard): river shiner. This Mile 6.5. species was rare in Doe Run but occurred H. storeriana (Kirtland): silver chub. throughout the area below Mile 5.3 and a The most upstream record of this species single specimen was caught below the cul- was in Tributary IV-e during backwater verts at Station III (Fig. 26). conditions. Two additional specimens were N. buchanani Meek: ghost shiner. Three caught by seine near Station V, and the N. buchanani were taken by rotenone in species occurred in rotenone collections July 1961 near Mile 7.2. from the extreme mouth of the creek. N. chrysocephaluschrysocephalus (Rafi- Trautman (1957) stated that silver chubs nesque): central common shiner. This often avoid silty water and sedimentation species was considered a subspecies of N. in larger bodies of water by movement into cornutus (Mitchill) until Gilbert (1961) clear streams of higher gradients. presented adequate evidence for its specific Notemigonus crysoleucas (Mitchill): identity. In Doe Run, it occurred in a re- golden shiner. A single golden shiner was stricted area as a small population of mostly seined downstream from the swift bridge large adults. There were rare occurrences culverts at Station III in association with of young up- and downstream from Miles other minnows. 4.6 and 6.9, respectively. The limited range Notropis ardens lythrurus Jordan: rosefin in Doe Run may relate more to water tem- shiner. Five specimens of rosefin shiner perature than to gradient since its range were collected from Doe Run, 1 in 1958 overlaps both the high- and low-gradient while I was procuring fish for the Univer- sections (Fig. 2). Gilbert (1961) consid- sity of Kansas and another in 1961 at Station ered chrysocephalus a southern species in III, 2 at station V in 1960, and the last by the process of extending its range north- rotenone in July 1961. It is relatively abun- ward (see Trautman, 1957). dant in Otter Creek, Meade County, Knob N. photogenis (Cope): silver shiner. Creek, Jefferson and Bullitt counties, and One specimen was taken at Station IV by other streams near Doe Run, but very rare seining at night, but repeated attempts on in the Ohio River. that and subsequent nights failed to obtain N. atherinoidesatherinoides Rafinesque: others. emerald shiner. This species probably is N. spilopterus spilopterus (Cope): spot- the most abundant fish in the Ohio River fin shiner. This species was common at near Doe Run, and it was classed as "abun- Station III in winter, but rarely farther dant" in Doe Run. It occurred throughout upstream than Station IV in summer. the stream from Station III to the river Spawning apparently occurred in August (Fig. 26), but was most common in the and September (Starrett, 1951), and young- low-gradient portion downstream from Mile of-the-year individuals were present at Sta- 84 WILDLIFE MONOGRAPHS tions IV and V in small numbers (Fig. 26). Martin and Campbell, 1953; and others), Spotfin shiners showed a definite preference and the species was most common in those for eddy habitats in Doe Run, but large habitats between Miles 5.3 and 7.0 in Doe adults also were typically associated with Run, with lesser numbers occurring up- riparian debris, such as roots or overhang- and downstream (Fig. 26). Large P. notatus ing bushes. Young were more often found were taken often in current during winter, in relatively swift water and sometimes but occurred in the deeper pools in sum- joined schools of N. atherinoides in open mer; similar seasonal relations were noted pools. by Starrett (1950a) in Iowa and by Metcalf N. stramineusstramineus (Cope): sand (1959) in Kansas. shiner. A number of sand shiners, appar- Young bluntnose minnows were found at ently comprising a single school, were Stations III, IV, and V, but in considerably taken by rotenone at Station V in April greater numbers downstream. Spawning 1959. apparently occurred in late June, July, and N. volucellus subspp.: mimic shiner. early August. A single adult specimen col- Specimens identifiable as N. v. volucellus lected in a muddy, cutoff pool at Station II (Cope), N. v. wickliffi Trautman, and (Fig. 26.) may have been introduced from probable intergrades between the two were a fisherman's bait bucket. present in the lower part of Doe Run, most P. promelas Rafinesque: fathead minnow. of which apparently moved into the creek This species is represented by a single col- from the Ohio River. The form recognized lection of 2 specimens from Tributary IV-e. as N. v. wickliffi was considered an "eco- Rhinichthysatratulus meleagris Agassiz: logical subspecies" by Hubbs and Lagler blacknose dace. This species was abundant (1958), and is generally more restricted to in Doe Run and occurred mostly upstream the Ohio River, while the typical form is from Mile 6.0 (Fig. 26). The subspecific wide-ranging in small tributaries, lakes, and name is applied mostly on geographic some larger streams (Trautman, 1931, 1957). grounds; specimens from Doe Run appear Notropis volucellus ranged farthest into to have large mouths and other characters Doe Run in periods of backwater from the tending toward R. a. obtusus Agassiz of the Ohio River (Fig. 26), and no young-of-the- Tennessee River System. Breeding males year individuals were taken at any time. from Doe Run differ from those described On the basis of frequency of occurrence by Trautman (1957) in having rust-red this species must be considered rare but lateral bands rather than a suffusion of pink large schools were present in the mouth of or red, and in the lack of color on the cen- Doe Run, in drying pools after impound- tral pads of the pectoral fins, as opposed ment in 1961, and the species was very to a deep orange or orange-red color in abundant in some rotenone collections from Ohio specimens. the extreme downstream sections. Delimiting the habitat of R. atratulus is N. whipplei (Girard): steelcolor shiner. complicated by its general abundance in This species was rare in Doe Run, but was Doe Run. Larger individuals live close to collected as single specimens throughout the bottoms in large, deep, eddying pools, the creek from Station III to Station V, and but also on riffles in June, presumably to below. spawn. Some large fish were found in un- Phenacobius mirabilis (Girard): sucker- usual situations. For example, in 1959 a mouth minnow. Two specimens were seined probe of the electrofishing gear was pushed at Station V. deep into the spring source of Doe Run and Pimephales notatus (Rafinesque): blunt- 6 specimens were washed out by the cur- nose minnow. Large, open pools appear to rent. Young-of-the-year blacknose dace be preferred by stream-inhabiting blunt- usually were found in quiet, shallow waters nose minnows throughout their range over mud bottoms until 1.0 to 1.2 inches in (Thompson and Hunt, 1930; Starrett, 1950a; length, and larger individuals (to 2 inches) ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 85 inhabited progressively deeper pools and Cyprinus carpio: carp. The presence of stronger eddies. Large numbers of young carp in Doe Run and in Doe Valley Lake and yearling fish were left in overflow has already been discussed (see Dorosoma). pools along Doe Run after floods. The Cyprinus was consistently present in Doe species forms loosely knit aggregations, Run downstream from Mile 7.0, but occur- usually of roughly equal-sized fish, and rences upstream from there were unusual apparently as a response to movement and limited in the largest pools. (Kuehne, 1958). Smaller blacknose dace Anguilla rostrata (LeSueur): American were more pelagic than large adults, and eel. Twelve eels were obtained from Doe were observed feeding at the surface on Run between Station III and Mile 7.0. Al- many occasions. The species was caught though the occurrence of this species in easily on small hooks baited with almost Doe Run is not surprising, the habitats any animal small enough to be swallowed. from which they were taken were remark- Large numbers of young R. atratulus ably consistent, and in contrast with that were produced in 1960; however, in 1961 described by Forbes and Richardson (1920) few young-of-the-year dace were found. and others. All specimens collected by The severe flooding in 1961 may have electrofishing were in tangled roots of affected spawning to some extent. High riparian trees, and in areas where strong discharge and siltation were major factors currents entered deep, hard-bottomed pools. hindering reproductive success in the Des Those seen making initial moves after being Moines River, Iowa, by a numberof species affected by rotenone were also in this type of minnows (Starrett, 1951). of habitat, but rapidly swam away from Semotilusatromaculatus (Mitchill): creek rotenone-treated water, often deserting the chub. Semotilus, although most abundant stream and moving rapidly up the banks. in the upstreamarea of Doe Run near Sta- Ambloplites rupestrisrupestris (Rafines- tion III, did not occur above Dam 1, and que): rock bass. Rock bass were scarce in was relatively rare in the stream below Doe Run when compared with other fishes Station V (Fig. 26). Its propensity for but were common between Miles 5.5 and ascending small streams (Shelford, 1911a; 7.1. Downstream from Mile 7.1 several Starrett,1950a) was indicated in Doe Run young-of-the-year individuals and 2 or 3 by its occurrencesometimes in considerable small adults (Fig. 26) were collected. The numbers in almost all tributariessampled. rock bass usually is considered a species of Large chubs in Doe Run sometimes ap- relatively swift waters in the Ohio River proached 11 inches total length and usually Basin (Forbes and Richardson, 1920), and were taken in nets from the mill hoop is associated intimately with rock ledges, and other the ponds deep pools throughout undercut banks, and other streamside sections or roots, upper (above IV), by shocker or debris Traut- in swift eddies at the foot of riffles and vegetation (Gerking, 1945; It is below waterfalls. The man, 1957; and others). interesting species provided that in the Ozarks the is considered considerable sport for fishermen in those species areas. Smaller chubs often schooled with an upstream fish (Funk and Campbell, R. atratulusas reported by Kuehne (1958) 1953), as it is in Iowa (Harlan and Speaker, but were somewhat more secretive and 1956); a species of medium-sized rivers in tended to occur near overhangingriparian Illinois and Ohio (citations above); but as vegetation or "cut-banks." Young-of-the- characteristic of downstream areas in year chubs appeared in May, and often colder streams of Ontario (Hallam, 1959); were found in association with surface- and of stream mouths and thick beds of feeding stages of Catostomus. The resem- aquatic vegetation of lakes (Adams and blance of small chubs at pool surfacesnear Hankinson, 1928). vegetation to certain top-feeding cyprino- In Doe Run, the habitats of large rock dontiform fishes was striking. bass were similar to those utilized by 86 WILDLIFE MONOGRAPHS

Anguilla-tangled roots along undercut L. machrochirus Rafinesque: bluegill. banks where there was considerable cur- Lepomis macrochirus was uncommon in the rent. Smaller individuals were in quieter ichthyofauna of Doe Run, with the greatest waters, but always near drifted debris or part of its population occurring between other cover. The species is quite sedentary Mile 5.3 and Station V. A few individuals, in streams (Scott, 1949; Gerking, 1953, however, were taken at Station II, and 2 1959; Funk, 1957; and Trautman, 1957). were caught at Station I, all of which were Chaenobryttus gulosus (Cuvier): war- large and were presumably involuntarily mouth. This species was rare in Doe Run displaced individuals from adjacent sink- and ranged from 4 to 5 inches total length. holes where they were very abundant. The Specimens were obtained singly from larger general preference of bluegills for lentic pools near Station IV and downstream to waters was expressed in their occurrence in near Station V. The species reaches its long, shallow, slowly flowing pools in Doe greatest abundance in lakes; small numbers Run; however, large individuals were some- such as in Doe Run are more or less typical times taken in swift eddies if cover was of stream populations, at least in Illinois present. (Larimore, 1957). L. megalotis megalotis (Rafinesque): Lepomis cyanellus Rafinesque: green longear sunfish. "Longears" were only sunfish. Green sunfish were a rare, but slightly more abundant than bluegills, but consistent component of the fauna of Doe their numbers were more evenly distributed Run downstream from Dam 3, but occur- between the up- and downstream limits of rences upstream were even more infre- their principal range. The largest popula- quent, and included only large adults. In- tions, however, were in the area of distinct dividuals taken in the upstream area may pool-riffle development between Miles 5.3 have entered Doe Run from adjacent sink- and 7.0 (Fig. 26). holes through subterranean watercourses, Although green sunfish and longear sun- or by passage through surface streams. No fish are more or less associated throughout young-of-the-year sunfishes of any kind the Middle West, in the sense that both were obtained upstream from Station IV, inhabit smaller streams, I have been unable but small green sunfish were more fre- to find reference to a stream population of quently taken in the downstream area than sunfishes including substantial numbers of were of the other sunfishes. any those species along with bluegills. In Green sunfish ascended tributaries of sporadic sampling of Doe Run in 1959, Doe Run farther than other fish any except prior to the beginning of intensive study, Semotilus. Stocks remaining in tributaries I noted increased numbers of bluegills after the rotenone operation of 1961, July and green sunfish downstream from swift- and probably some fish upstream from the flowing culvert tubes of the at Sta- area, reproduced successfully in Doe Valley bridge tion III. These increases occurred in Lake in August. This species apparently April, and and were preferred shallow, quiet pools where rubble, May, June, surprising because of the stream-drifted debris, or overhanging vege- general dearth of sunfishes in tation provided thick cover. my previous collections, but corresponding L. humilis (Girard): orangespotted sun- decreases of the populations occurred fish. One orangespotted sunfish was col- throughout the summer and they were lected at Station V, another near Mile 7.0 again rare by the end of the year. In by rotenone, and a third in the unlikely February 1960, after a short period of win- area of Station I. The last specimen con- ter discharge, those species were again forms with L. humilis in most respects, but below the bridge, and became progres- may be a hybrid involving L. cyanellus. sively more abundant in succeeding months. Its origin and occurrence at Station I are That summer they again diminished in enigmatic. numbers in July, occurring only rarely until ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 87 after high discharges of January. The bedrock bottoms of pools in the marl area large population accumulated in spring certainly were not suitable, and areas of 1961 remained near Station III throughout particulate bottom, excluding silt or detri- the summer, fall, and winter. Lepomis tus, were generallly in swift currents. How- megalotis also became more abundant in ever, in spite of the presence of "ripe" the large pool each spring, but was present adults at Station III, no breeding sunfishes all year throughoutthe downstreamsection, or young-of-the-year individuals were ob- and at the bridge. served or collected. It must be concluded Of these sunfishes, L. megalotis occupies that very little, if any, reproduction by remarkably stable home ranges of about centrarchidsoccurred upstream from Sta- 100 to 200 feet of stream,usually delimited tion IV. by riffles (Gerking, 1959), and is capable The most logical explanation for this of returningto those areas when displaced, situation is the low water temperature at apparently by olfactory recognition of the Station III. With maturationof the gonads "home pools" (Gunning, 1959). Of 45 in spring, and the accompanyingsecondary specimens of L. cyanellus recapturedafter sexual characteristicssuch as male breed- tagging in Missouri streams (Funk, 1957), ing colors and presumed searching for 77.8 per cent made movementsof less than suitable nest sites, the fish became concen- 1 mile (= "no movements"), and of those trated below the barrier produced by the that had moved, 13.3 per cent had gone swift, shallow culvert, or considered the upstream and 8.9 per cent downstream. pool "suitable" (the former seems more Green sunfish in shallow ponds of Wiscon- likely). Subsequently, water temperature sin returned to the same area each spring, failed to rise sufficiently to permit repro- whereas displaced individuals,followed by duction, and the fish resorbed their ova an attachment of a floating bobber with (noted in L. megalotis and L. macrochirus), nylon thread, returned "home"with con- or moved downstream until higher tem- siderable precision (Hasler and Wisby, peratures were encountered. No major 1958). Gerking (1950) found green sun- downstream movement of sunfishes was fish living in the same pools of an Indiana observed, but it seems likely that most creek for a considerabletime, and persist- moved slowly downstream to reoccupy ing or returning there (Gunning, 1959) variouspools and eddies without spawning. after severe flooding. Bluegills in Sugar- Sunfishesresident between StationsIII and loaf Lake, Michigan, returned to the half IV presumably represent "mobile" com- of the lake in which they were originally ponents of the downstreampopulations that trapped after being displaced to the oppo- gradually ascended the creek and stopped site side (Cooper, 1953). I know of no in suitable habitats. The of fluctuation studies concerning movements of that sequence population in the in 1961, with the fishes species in streams. "bridgepool" remaining there throughout the year, in- Movements of many fishes are greatest validates some of the above discussion. in and spring (Funk, 1957;Trautman, 1957; However, no reproductiontook place de- others), which may be related to searching spite the presence of adult fish. In the for areas suitable for nest building and winter of 1960-1961 a number of large other reproductiveactivities (Winn, 1958a; stones fell into the pool from riprap of an Gunning, 1959). Likewise, the possibility extensive road fill, and substantially in- exists that certain individuals move to the creased the suitability of the area for a same places each year for spawning (Hasler larger number of fish. It is possible that and Wisby, 1958). The large pool down- clearing the downstream portion of Doe stream from the bridge at Station III pre- Run, siltation, and continual disturbance sented the most likely spawning site by machinery and felling of trees caused available upstream from Station IV. The movement of fish upstreamand concentra- 88 WILDLIFE MONOGRAPHS tion below the barrier. spotted bass. Scattered individuals were These interpretationsare based on the caught from Station III to the mouth of assumption that certain minimal tempera- Doe Run, but young were obtained only tures are required for reproductionby the at Station V. Its typical habitat appeared species concerned. Such records are rarely to be shallower, quieter pools than that of less than 21.1?C (Witt and Marzolf, 1954; the smallmouth bass, but it was taken also Swingle and Smith, 1950; and others), a in swift eddies upstream and in the almost- water temperature never attained at Sta- lentic waters near the Ohio River. Traut- tion III. The developmentof the gonads of man (1957) noted migration of spotted sunfishes in Doe Run and elsewhere, how- bass upstream in spring, but in Doe Run ever, occurs at cooler temperatures pre- specimens were collected in spring and sumablyin responseto day length. Detailed autumn at Station III. The presence of the studies of spawning cycles in warmwater species in shallow pools throughout the fishes are badly needed to evaluate the high- and low-gradient sections also con- influence of light, water temperature,and trasts with reports of Howland (1931) and other environmentalconditions. Trautman (1942, 1957) that the species Micropterus dolomieui dolomieui La- usually inhabits deep, well-shaded pools in cepede: smallmouthbass. This species was areas with gradients less than 4 fpm. With rare in Doe Run and restricted in its disregard to movement, Funk (1957) recap- tributionbetween Miles 5.3 and 7.2, except tured 24 individuals from Missouri streams, for an occasional small individual taken at 22 of which had moved no more than a Station V (Fig. 26). Distributionof small- mile from the point of original tagging, mouth bass in streams has been correlated and Gerking (1953) found that only 1 of with gradient (Trautman,1942; Burton and 16 fish recaptured after marking had moved Odum, 1945), with the largest populations from the pool in which it was first caught. existing in streams with gradients between Habitats and behavior of this species seem 4 and 25 fpm. Upstream from Mile 5.0 in to vary considerably throughout its range. Doe Run the gradient is about 35 fpm (to M. salmoides salmoides (Lacepede): Mile 1.5; Fig. 2), and in the area where largemouth bass. A single largemouth bass smallmouthbass occurred it is 4 to 6 fpm, was caught on pole and line at Station III becoming higher downstream. Smallmouth in June 1961 and no additional specimens bass may have avoided the downstream were found until the rotenone operation of area because of the generally soft bottoms July 1961. Twelve specimens were col- and the sluicelike nature of the stream. lected then between Miles 6.2 and 7.35 Seldom more than a single individual was from large pools with silt-covered gravel caught from a given pool in Doe Run, bottoms. usually near logs, roots, or other cover, Pomoxis annularis Rafinesque: white where swift currentsprevailed. The species crappie. This species was not obtained is rather sedentary in smaller streams until 1961. Two were taken by shocker (Gerking, 1950, 1953), and appearscapable near Mile 5.5 when backwaters of the Ohio of returningto a home pool when displaced River were present, and most of the re- (Larimore,1952, 1954). mainder were obtained in the rotenone Youngsmallmouth bass were caught only study of July. The species was present, at Station V, and in very small numbers. however, downstream from the Doe Valley Most specimens obtained upstream were Dam in small numbers, where it was seined more than 8 inches total length. The larg- from drying pools in association with large est individual I saw weighed 2 pounds, 2 schools of Notropis atherinoides and N. ounces after viscera and scales were re- volucellus. moved. That fish was caught by a fisher- P. nigromaculatus (LeSueur): black man near Mile 5.4. crappie. A single black crappie was taken Al. punctulatuspunctulatus (Rafinesque): by rotenone near Mile 6.9 in July 1961. I ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 89 found this species relatively abundant in and spawning apparentlyoccurred in June the Ohio River near Doe Run in 1959 while and July. fishing hoop nets. E. flabellare Rafinesque: fantail darter. Etheostoma blennioides Rafinesque: Two subspecies of fantail darter, E. f. greenside darter. Greenside darters were flabellare and E. f. lineolatum (Agassiz), rare inhabitantsbetween Miles 4.5 and 6.2 appear to intergrade in Doe Run and in (Fig. 26) with single occurrences at Sta- other streams of the vicinity. tions III and V and a few specimens from E. flabellare was abundant in Doe Run, Miles 6.2 to 6.9 in the rotenone operation occurringfrom StationV to StationII (Fig. of 1961. 26). Its upstreamdistribution was abruptly In Ohio, greenside darters usually terminated near Mile 2.05 immediately spawned in April when water temperatures upstream from Tributary II-a at a low were slightly below 18.3?C (Trautman, waterfall, the reasons for which are ob- 1957), but spawning in Doe Run occurred scure. It may be that temperaturesin Doe in June as indicated by presence of young Run at Station II were seldom, if ever, in collections, and condition of gonads and high enough for breeding of this species, adults. Later reproduction in Doe Run with most young-of-the-year and "high" presumably was in response to the rela- males being taken in the downstreampart tively cool water temperatures earlier in of Tributary II-a or immediately down- the year. Winn (1958a) found spent E. stream from its mouth. That tributarywas blennioides in the Green River System of spring-fed, but small and relatively un- Kentucky the first week of April. No repro- shaded, and water temperaturesaveraged duction in Doe Run was detected upstream 2?C higher in its mouth than in Doe Run from Mile 5.3. on bright, warm summer days. The habitat of E. blennioidesin Doe Run Young-of-the-yearfantail dartersfirst ap- was usually the swiftest riffles or fast peared in May and June, and moved im- eddies at the lower ends of the swiftest mediately to quiet, warm, shallow waters, riffles. During winter some were obtained living there until nearly an inch long. in pools when higher discharges produced Those areas have soft, semiflocculent bot- significant current velocity there. There toms that support a rich fauna and flora of was no marked propensity by the fish to microscopicorganisms, thus providingfoods inhabit vegetation on riffles in Doe Run, for small fish; secondly, small darters are though such a preference has been re- relatively alone in that habitat, which is corded elsewhere (Forbes and Richardson, inhabitedonly by very young Cottus on the 1920; Trautman, 1957). The use of vegeta- bottoms and postlarval Catostomus and tion for spawning (Fahy, 1954; Winn, Semotilusnear the surface. Water temper- 1958a, b; and Trautman,1957) may have atures in those areas sometimes reached caused the apparent lack of reproduction 30?C in summer,but the small darters re- in the marl area and its occurrence down- mained there anyway. stream where Cladophora was locally abun- Larger E. flabellare inhabited eddying dant on riffles. pools, swift rubble riffles, smaller riffles in E. caeruleum Storer: rainbow darter. Doe Run, and often were taken in mere This species was as rare as the greenside trickles of water between and beneath darter in Doe Run and occupied a similarly small stones at Station IV. Downstream restricted distribution though ranging some- from Station IV, where Cladophoradevel- what farther downstream (Fig. 26). It oped lateral to the main currents of the inhabited swift riffles with gravel bottoms riffles, there was a segregationof sizes. The and apparently avoided vegetation on those smallest individuals were in less than 0.25 riffles. Young-of-the-year rainbow darters inch of water along the banks, slightly were taken as far upstream as Station IV, larger ones were in Mougeotia of the quiet 90 WILDLIFEMONOGRAPHS

areas lateral to Cladophorabeds, and pro- in gradually diminishing numbers. Many gressively larger fish were found farther preliminaryaspects of its life historyin Doe from the shore. Collections from the main Run, including habitats, food habits, channel, over clean gravels, yielded only a growth, and certain aspects of the accumu- few large fantail darters,with E. caeruleum lation of radioactive materials through its and occasionally E. blennioides. food web, were presented by Minckley, et At Station II, E. flabellare was common- al. (1963), and furtherstudies on its biology est in beds of Myriophyllumin relatively are being made by James E. Craddock. quiet currents near the upstream end of Sculpins comprised more than 75 per Pool 2. cent of all fish collected at Station I, and Hadropterus maculatus Girard: black- occurred at downstream stations in some- side darter. On the basis of characters what smaller numbers. Young individuals given by Hubbs and Lagler (1958), and inhabited quiet, warm backwaters over others (see Winn, 1958b), I consider silt-detritus bottoms (see E. flabellare), Hadropterus as distinct from the genus then gradually invaded swifter waters as Percina,with which it was synonymizedby they became larger to inhabit the fastest Bailey (in Bailey and Gosline, 1955). Two currentsas adults. Extremely large adults, specimens were obtained by rotenone in however, preferred pool habitats where July 1961. Both were large adults and were strong currents persisted as eddies or tail- captured in large pools near Mile 7.0. waters of upstreamriffles. Percina caprodes caprodes (Rafinesque): Young-of-the-year Cottus first became logperch. This species was rare in Doe abundantin Doe Run in April, with repro- Run, occurring between Station IV and duction extending at least through May. Mile 7.4, usually as single individuals. The smallest one obtained was about 4.8 inches Summaryof the Distributionsof Fishes total length and the largest near 7.0 inches. As with the vegetation and invertebrates The species is rare in the mainstream of in Doe Run, the fishes were found in longi- the Ohio River, but relatively consistent in tudinally restricted ranges in relation to its occurrencenear Doe Run. physical features of the stream and specific Stizostedion canadense (Smith): sauger. differences in the behavior of the diverse A single sauger was captured by rotenone animals present. Shelford's (1911a) work within 50 yards of the mouth of Doe Run on the distribution of stream fishes near in 1959. Chicago, III., formed much of the basis for Aplodinotusgrunniens Rafinesque: fresh- physiographicinterpretation of longitudinal water drum. This large species occurred succession in streams. In a later paper consistently in shocking and rotenone col- (Shelford, 1911b) he demonstrateda suc- lections downstreamfrom the high-gradient cessional series of pond fishes, and found section of Doe Run, and a few individuals a primary dichotomy between lakes and were taken upstream as far as the large streams in the great influence of biotic pool below the bridge at Station III (Fig. factors in the successional series of the 26). The largest specimenseen was at Mile former as opposed to geologic factors with 3.5; its weight was estimated at 5 pounds the latter (Allee, et al., 1955). Studies in or more. It was shocked in a long, shallow Europe have prompted recognition of pool, but was not captured. On the basis "trout"(Forelle), "grayling"(Asche), "bar- of frequency of occurrence the drum was bel" (Barbe), and "bream"(Blei) regions an uncommoncomponent of the fish fauna occurring successively from up- to down- of Doe Run. stream (Leger, 1945; Ruttner, 1953). Ac- Cottus c. carolinae (Gill): banded scul- cording to Huet (1959), these zonations pin. Cottus carolinae was the most wide- result mostly from stream gradient, cur- spread and abundant fish in Doe Run, rents, temperaturerelations, physical nature ranging from Station I to below Station V of the substrate through which the stream ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 91 is cutting, composition and abundance of fish, except the largemouth bass, that in- the flora, and size and abundanceof stocks habited a fixed range in Doe Run and was of food organisms. Thompson and Hunt more abundant upstream than near the (1930) found that stream width where mouth. Still, there was no evidence of marked differences in gradient did not spawning and the populations may have occur between streams or stream sections occurred because of a preference for the was also an important factor in fish dis- open, clean bottomsof the creek as opposed tribution. Gerking (1949) obtained data to those of the Ohio River. It is noteworthy indicating that volume of the stream was that many of the most upstreamcatches of important in the weight of fish present, a river fishes were at the barrierimposed by stream with frequent deep pools being culvert tubes below Station III, and that more productive than one of the same size most were single specimens. but having little deep water. The marked A second group of fishes in Doe Run importance of gradient in the distribution appeared most concentrated in the pool- of streamfishes was furtheremphasized by riffle area immediately downstream from Trautman (1942) and Burton and Odum Miles 5.3 to 5.6, with scattered occurrences (1945). up- and downstream. These included In Doe Run, the 3 major factors con- Lepomis spp., Pimephales notatus, Hypen- trolling the distributionof fishes are gradi- telium, Ameiurus melas, Campostoma, ent, temperature,and the presence of marl Etheostomablennioides, E. caeruleum,No- in the midsectionof the stream. The effects tropis chrysocephalus,Moxostoma erythru- of gradient are most pronouncedwhen the rum, M. duquesnei,and Ambloplites. Some "river fish" components of the fauna are of them, principallythe suckers (excluding examined (Fig. 27). Thirteen of the 31 Hypentelium), N. chrysocephalus,and the species that I consideredmore or less typi- larger centrarchids,M. dolomieui and Am- cal of the Ohio River fauna near Doe Run bloplites, may have been restricted to the maintained definite ranges in the creek, area downstream from Mile 5.3 by the and all but 1 of which merged into the presence of shallow, open pools upstream, Ohio River populations (Fig. 26). The and their innate habitatpreference for deep single exception, Micropterus salmoides, pools or more shelter (Larimore, et al., occurred in very small numbers in Doe 1952;many others). Their downstreamdis- Run, and its inclusion in Figure 26 would tributions were presumably restricted by not have been justified had it not been for the lack of pool-riffle development in the the sharplycircumscribed area it inhabited. lower areas,and its sluicelikeconfiguration. The designation of Micropteruspunctula- The absence of gravel for spawning areas tus as a "riverfish" is questionable, but it also may have limited habitation of the was consistentlymore abundantin the Ohio marl area by the latter species, and prob- River near the mouth of Doe Run than in ably is a factor limiting populations of other nearby streams I sampled. However, Hypentelium, Campostoma,and E. caeru- the reverse is true in other regions, at least leum there. Campostoma,however, repro- in spring and early summer (Trautman, duced successfully at the limited gravel 1957), when both types of stream were area of StationIII. Moreover,the cemented sampled by me. Minytremaseems charac- marl riffles and bedrock-bottomed pools terstic of stream mouths along the Ohio between Station III and Mile 5.3 may have River (Krumholz, et al., 1962), and is been unsuitable habitat for these bottom- therefore included with river fishes since dwelling, secretive forms. most stream mouths along the Ohio River Two species also included in the above are flooded by navigation installationsand category, Etheostomaflabellare and Semo- are effectively parts of the backwaters. tilus, were more successful in maintaining Moxostoma breviceps was the only river populations throughout the high-gradient, 92 WILDLIF MONOGRAPHS marl section (Fig. 26), and to some distance Because of lack of reproductionof many below Mile 5.3. The penetration of Doe fishes in and above the marl area, even Run by these species to Station II, with the after maturation of ova by females and darter not occurring upstream from Mile attainmentof breeding colorationby males, 2.05, but the chub present to immediately it is evident that low water temperature downstream from Dam 1, invites examina- and the presence of marl restrictedspawn- tion in light of the possible differentiation ing and distributionof many species. In- of Catostomuscommersoni in the upstream cluded here are the "creek fishes" (Fig. area. The first mill dam built in Doe Run 27) not known to reproduce above Mile was at Mile 1.65 (Dam 1), upstreamfrom 5.3, but which reproduce downstreamand the now-drownedwaterfall at Mile 2.1. It presumablymaintain populations by move- effectively blocked upstream movement of ment upstream. A few creek inhabitants, fishes from below that waterfall even Percina caprodes, Moxostoma breviceps, though they could cross it after the con- M. duquesnei, and a few others, did not structionof Dam 2. Both E. flabellare and successfully reproduce in Doe Run insofar Semotilus occur in other springs of Ken- as I could determine; their populations tucky and other parts of their ranges must have depended on influx of individ- (Forbes and Richardson, 1920; Gunning uals from the Ohio River. and Lewis, 1956), but it is not known Only 3 of the 66 kinds of fishes collected, whether they reproduce there or merely Catostomus,Rhinichthys, and Cottus,main- maintain populations by upstream move- tained populationsthrough reproductionat ments of young hatched downstream in Station I. These fishes ranged throughout warmerwaters. It is conceivable that both most of Doe Run in significant numbers, species invaded the area of Doe Run above but Rhinichthyswas the most restrictedin Mile 2.1 after Dam 2 was built, and have its downstream distribution. Salmo gaird- maintainedsmall populations but are unable neri was introducedand did not reproduce to pass over Dam 1. Exclusion of other successfully. It was most abundant at species could have occurred easily because Station II, and it presumably was limited the constant, low temperatures pre- to the upstream area in response to the cluded or lowered reproductive success. cooler waters. According to Curtis Brown (pers. comm.), It is of note that some species appearing Micropterus dolomieui was stocked up- closely associated in Doe Run, such as streamfrom Dam 2, but none was taken by Micropterus dolomieui and Hypentelium, me or in recent years by fishermen. That have also been found closely associated in species usually spawns at temperatures other streams (Thompson and Hunt, 1930; greater than 17.8?C (Adams and Hankin- Larimore, et al., 1952; and many others). son, 1928), or with little success at lower Trautman (1942) indicated that M. dolo- temperatures(Henderson and Foster, 1957). mieui rarely was found in Ohio streams A certain segment of populations stocked havinga gradientless than 4 fpm, but in the in Jordan Creek, Ill., by Larimore (1954) area of Doe Run inhabited by that species remainedthere, althougha significantnum- the gradientwas only slightly greater (Fig. ber moved downstream and disappeared 2). The restrictionof upstreamrange may from the population. Also, all Lepomis spp. also have been a function of gradient, be- and Ameiurus spp. taken in the upstream cause few smallmouth bass occur in area of Doe Run were relatively large. streamswith gradient greater than 25 fpm, Possibly these are maintainedby very rare and the marl area of Doe Run averaged reproduction that was not detected there more than 30 fpm throughout. during this study, but I am inclined to be- The rare species of Doe Run, those oc- lieve they entered from adjacent sinkhole curring only as single specimens or in ponds, and persisted there temporarily fewer than 10 collections, comprise al- without reproducing. most half of the 66 species obtained. ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 93

RIVER FISHES RIVER CHANNEL ? BACKWATER

Aloso chrysoch/oris Hybopsis storeriono Moxostomo breviceps Stizosted/on conadense Hiodon fergisus Notropis otherinoides N. blennius N. straminous Lepisosteus osseus /cfolurus punctotus Hybognothus nucho/is Notropis buchanani N. vo/uce//us N. whipplei Pi/odictis o/ivoris Dorosomo copedionum Ap/odinotus grunniens Cyprinus corpio Corpiodeos corpio Ictiobus bubo/us Moxostomo oni/surum Notemigonus crysoteucos Microp/erus punctulatus Caorpiodes cyprinus C. vwlifer Ictiobus cypri/ellus Minytremo molenops Lepomis mocrochirus Micropterus so/moides Pomoxis onlnu/oris P nigromoculotus

CREEKFISHES RIFFLES * POOLS

Phenoacobius mirobi/is Noturus flovus Eithostomo blennioides So/mo goirdneri Cottus carolinoe Compostoma onomolum Notropis photogen/is Rhifnchthys atrotulus Micropterus dolomi/eui Etheostomo coeruleum Semotilus otromocu/otus Hypente/ium nigricons Hybopsis omblops Etheostomo flaobellre Notropis ordens N. chrysocepholus N. spilopterus Cotostomus commersoni Moxostomo duquesne/i Ame/urus notolis Anguillo rosrotlo Ambloplites rupostris Lepomis cyonellus L. megalotis FIG. 27. Ecological relations of fishes in Doe Hodropferus moculatus Esox omoriconus Run, the adjacent Ohio River, and in other small Porcina Chrosomus creeks in the vicinity of Doe Run, Meade County, coprodes orythrogoster buccoto Ky. Fishes are grouped according to the apparent Ericymbo relations to current, depth, and bottoms, beginning Pimepho/es noto/us with those of relatively swift, deep (in the Ohio P. prome/os River) or shallow (in creeks), hard-bottomed Moxostomo orythrurum areas to the left, and terminating with those in- Ameiurus molos habiting relatively deep, quiet, soft-bottomed areas Choenobryttus gulosus to the right in each column. Lepom/is hum/ils 94 WILDLIFEMONOGRAPHS

Eighteen of the 32 species classed as rare histories involved specific interactionswith are "riverfishes" (Fig. 27), and must have Doe Run or its biota are included here. been strays. The remaining 14 species are Rana catesbiana Shaw, R. clamitans more or less typical "creekspecies" in other melanota (Rafinesque), and R. pipiens streams near Doe Run, and their presence sphenocephala Cope were abundant along in such small numbers must indicate that the tributariesof Doe Run, but were rarely the spring stream is not suitable for them. seen along the mainstream. R. palustris The discussionsof Grinnell (1922 a, b) of Le Conte was representedin our collections the possible roles of rare and accidental by a single specimen. The most abundant birds in California were applied to the frog along the banks of Doe Run was the occurrence of stray or rare fishes in Iowa cricket frog, Acris crepitans blanchardi streams by Starrett (1950a). Rare species Harper. Larval anurans, including Bufo are available to fill niches or habitats that spp., Pseudacris triseriata (Wied), and might become vacant by failure of year Hyla spp., were abundant in nearby sink- classes of resident species, for example, or hole ponds and in overflow pools along the possibly more importantly, to fill voids creek, but none was collected in Doe Run. produced by changed environmentalcondi- Specimensof hellbender,Cryptobranchus tions in or near a given aquatic habitat. alleganiensisalleganiensis (Daudin), mud- Such changes are easily demonstrated by puppy, Necturusmaculosus maculosus (Ra- changes of fish populationsfollowing intro- finesque), Eurycea bislineata bislineata duction of pollutantsinto streams (Thomp- (Green), Pseudotriton ruber ruber (La- son and Hunt, 1930; many others), or the trielle), and Desmognathus fuscus fuscus changing from lotic to lentic situations by (Rafinesque), were obtained from Doe impoundingstreams. Such a change usually Run or its tributaries in small numbers. is followed by rapid increase in species of Adult Desmognathuswere the most abun- backwaters and coincident decrease of dant. A single newt Diemictylusviridescens riffle-inhabiting species (Trautman, 1939; viridescens (Rafinesque), typical of sink- Finnell, et al., 1956;and others). Moreover, hole ponds in the area, was caught near the sudden or progressive occurrence of a Mile 1.4. "rare"or "accidental"species in a given Snapping turtles were collected at Sta- area, such as the eastward movements of tion III, and in Pools 2 and 3. The stomach Phenacobiusmirabilis and Lepomis humilis of a large specimen from Station III con- (Trautman, 1957), or other changes in the tained at least 143 individual Orconectes dispersion and dispersal of populations rusticus. Only the chelae were counted, (Black, 1949; Minckley and Cross, 1959; thus considerably more than is indicated and others), may indicate climatic or land- may have been eaten. Other turtles re- use changes of importance not only to corded from the creek were Graptemys fishes, but as an index of the effects of pseudographica(Gray), and a "sightrecord" man's activities. Such data are valuable of Trionyx sp. in predicting and evaluating conditions of Natrix sipedon Linnaeus was the only the naturalenvironment, on which man de- abundant snake along Doe Run. It fed pends for his livelihood. principallyon Acris along the banks, but 1 individual contained a small VertebratesOther Than Fishes unidentified minnow, and another was seen attempting Information on vertebrates other than to ingest a trout near Station IV. Tham- fishes was obtained incidental to other nophis sirtalis sirtalis Linnaeus, Heterodon field work. However, observations and platyrhinosLatreille, and Coluber constric- collections of amphibiansand reptiles were tor Linnaeus, were also taken along the made in some detail, and have been pre- stream, and undoubtedly ate such stream- sented elsewhere (Craddockand Minckley, side animals as frogs. 1963). Only those animals whose life Birds of many kinds were numerous ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 95 along Doe Run. Kingfishers, Megaceryle inside the culverts at that station. Riggs alcyon (L.), were abundant throughout (1947) saw purple martins, Progne subis the length of the creek, and preyed on (L.), feeding on emerging mayflies on fishes and larger invertebrates such as Shafer Lake, Indiana, and Burks (1953) crayfish. Greatblue herons,Ardea herodias noted that birds sometimes fed heavily on L., were frequent in the downstream sec- mayflies at times of peak emergences. I tions, especially after the area was cleared also saw robins, Turdus migratorius (L.), prior to impoundment, and just after the feeding on adult caddisflies and stoneflies eradication of fishes by rotenone. There along the banks of Doe Run in 1960. were 6 or 8 great blue herons along Doe Muskrat,Ondatra zibethicus (L.), inhab- Run the day after the rotenone operation, ited bank dens throughoutDoe Run except presumably feeding on the fishes missed in the area sporadicallyflooded by the Ohio by pickup crews;none had been seen in the River. Near Dam 2 I found old stumps area for several weeks precedingthat study. and cuttings undoubtedly produced by Green herons, Butorides virescens (L.), beaver, Castor canadensis Kuhl, but no were relatively common in the downstream recent signs were observed. The species section of the stream throughoutthe study, must have been present on Doe Run within and apparently nested there as indicated the last 25 years or so, judging from the by the numerous juvenile birds in late condition of the cuttings. summer. The presence of raccoon, Procyon lotor Several species of ducks were observed, (L.), and mink, Mustella vison Schreber, both in the downstreamsection at times of has already been noted. Tracks of raccoon flooding by the Ohio River, and in the were especially abundant along Doe Run, heavily vegetated area upstreamfrom Dam and numerous crushed exoskeletons of 1. The species most commonly seen near Cambarus and Orconectes on the tops of the Ohio River were mallard, Anas pla- stones from StationII to Station IV attested tyrhynchosL., and black ducks, A. rubripes to its predationon aquatic animals. Brewster, but a few lesser scaup, Nyroca Bats were seen often in daytime flying affinis (Eyton), rested on open waters of slowly along Doe Run a few feet above the downstream section when it was in- the water. A single specimen, Lasiurus undated. Those ducks frequenting the up- borealis (Mueller), was shot near Station stream area included the black duck, wood III and its stomach contained remains of duck, Aix sponsa Linnaeus, and bluewing various dipterans, a few mayflies, and a teal, Querquedula discors L. The black single adult of Paragnetina. duck and teal usually were present only at Food times of migration, but a single brood of Relationsof the Fauna bluewing teal was produced near Blue It is apparent from the abundance of Spring in 1960, and the wood duck was a aquatic animals in Doe Run, the thick and consistent breeding bird along the creek persistentstands of macrophyticvegetation, in all years of the study. and the overall abundance of algae, that Bronze grackle, Quiscalus quiscula (L.), the stream presents a highly productive and other "blackbirds" frequented the system. Actual estimates of productivity banks of Doe Run, especially in spring. In were beyond the scope of this study, but April 1960 I saw a number of grackles some discussion of the food relations of foraging among mats of aquatic vegetation animals may give insight into the trophic stranded on riparian roots by receding structure. Odum (1959), Damell (1961), water levels. Barn swallows, Hirundo Ivlev (1961), and others have pointed out erythrogaster Boddaert, were also fre- that animals in nature rarely comply with quently seen along the stream, and on 2 the conceptual "trophic levels" prevalent occasions were watched feeding on emerg- in some current ecological thinking, but ing mayflies near Station III. They nested that behavior and relative abundance of 96 WILDLIFE MONOGRAPHS animals resembles more the "food web" tained in September indicated that foods concepts of workersprior to publication of were similar to those in Hydropsyche. the trophic concepts by Lindeman (1942). However, particulate detritus was rela- It is increasingly evident that local varia- tively more abundant than Melosira both tions in habitats determine the position of in summer and winter. Sand was classed a given species in a trophic system. Such as abundant in winter, but was absent in variationsare reflected in differentialavail- September. abilities of food, seasonal changes in foods Digestive tracts of 13 Simuliumsp. from in a given habitat, and changes in food Station III in June 1960 contained finely habits of an animal with changes in age, divided detrital materials,but 1 individual size, or physiological condition (Brown, also contained a small Navicula. 1961a; Darnell, 1961), and a myriad of Nymphs of the mayfly genus Isonychia other factors. In addition, there are pro- feed passively much of the time by using nounced specific differencesbetween kinds their setiferous front legs as a funnel for of animals in a given habitat. It must be capture of stream-drifteddiatoms and de- pointed out that I use the term trophic in tritus (Clemens, 1917). Nymphs of Isony- the nutritional sense which, although it is chia sadleri?from Doe Run had food habits an energy transfer, it appears to involve similar to Hydropsyche. Fourteen Isony- many levels of utilization and transfer, chia contained Melosira,detritus, and sand, some of which are little understood. in order of decreasing abundance; 2 contained fragments of insects, possibly ex- Invertebrates uviae of Baetis?and an unidentifiedmayfly Passive feeders.-Mueller (1955) separated (see Leonard and Leonard, 1962), and 3 stream animals into passive feeders and contained blue-green algae and some active feeders, the former depending on Gongrosira. In the latter group there was the amounts of food extant in the drifting a considerableinclusion of marly sediment state, and the latter predominatingin small (estimated 50%). The nymphs were col- streams usually devoid of such "net plank- lected in September 1960. ton." Filter- and detritus-feedingorganisms Active feeders.-Oligochaetes feed largely are generallyabundant in turbid streamson on organiccomponents of the substrateinto solid substrates. which they burrow,passing large quantities Hydropsyche betteni fed mostly on the of material through their digestive tracts. diatom Melosira which was abundant in A number of Nais sp. and Ophidonais guts of those larvae from Station III in serpentina from Station IV contained no summer and winter. Ten animals taken in identifiable materials other than sand and winter also contained Cocconeis, Gom- particulatedetritus in their guts. Intestines phonema, several species of Navicula, and of Nais taken from Myriophyllumplants a few trichomesof Phormidiumor Oscilla- rather than from the substrate were filled toria, all of which were rare or occurred with diatoms of many genera and a large singly. Particulate detritus was a scarce proportionof epidermal tissue of the mil- component in larvae from both periods, foil. The guts of Limnodrilushoffmeisteri and sand was scarce in winter animals but from beds of decomposing leaves near absent to rare in summer. Five larvae from StationIV in October 1960 were filled with StationII, collected in June 1960, contained leaf particles, sand, and a few frustules of essentially the same things as those from diatoms. Station III in September. The only excep- Walker (1962) reportedthe gut contents tion was the presence of a scarce com- of 30 Asellus bivittatus. Her data indicated ponent of Fissidens from Station II, and that the species feeds primarilyon epiphy- that was considered detrital. tic diatoms associated with Fissidens, but Analysesof 17 larvae of Cheumatopsyche she also found a few fragments of blue- sp. from Station III in winter and 4 ob- green algae and some partially digested ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 97

TABLE 10.-PERCENTAGE FREQUENCYOF ITEMS the isopod fed indiscriminately on most FOUND IN DIGESTIVETRACTS OF 57 GAMMARUS organic materialsplaced in the tank. MINUS AND 66 G. BOUSFIELDI FROM STATION I, Somewhat more detailed studies were DOE RUN, MEADECOUNTY, KENTUCKY, SEPTEMBER 1960 made of the foods of the amphipods (Gammarusspp.) in Doe Run in connection Item(s) G. minus G. bousfieldi with other researchon those animals(Table Blue-green algae 3.5 28.8 10). Usually, G. bousfieldi was somewhat Diatoms 45.6 10.6 more omnivorousthan G. minus, and the Nasturtium officinale foods of the former are those most often in Epidermis 1.8( ? ) 63.6 the where the lived Root hairs - 43.9 pools amphipod (Cole Fissidens julianus 22.8 4.5 and Minckley, 1961). The most frequent Particulate detritus 42.1 75.8 items, sand and particulate detritus, indi- Animal remains 1.8 10.6 cate considerablegrazing on the open bot- Amorphous material 89.5 28.8 toms of pools, which was substantiated Sand 47.4 80.3 by field and laboratory observations. Water- cress also was utilized heavily, and quanti- remains of Fissidens. Verification of the tatively was the most abundant item. last-mentioned material indicates that about Gammarusminus fed on an amazingly 10 per cent of the food consists of epidermis small amountof Fissidens and diatomscon- of Fissidens. Although some unidentifiable sidering its habitat in dense mosses of the detrital material was present, the amount channel. It apparently fed mostly on the was not significant. undifferentiated, semiflocculent materials Twenty individuals of Asellus bivittatus collected beneath the moss beds through from Station V in September 1960 fed al- its own decomposition,and throughdeposi- most entirely on small leaf fragments and tion of feces by the extremely abundant particulate detritus. One individual con- Asellus bivittatus. The great numbers of tained fragments of isopod exuviae, pre- separated diatom frustules in the digestive sumably of its own species. Specimens in tracts of G. minus, with no evidence of laboratory aquaria fed readily on exuviae chlorophyll, may have resulted from pre- of their fellows, on commercial fish foods, vious passage of the diatoms through the on debris at the bottom, and on Vaucheria, digestive tract of Asellus and secondary but made no attempts to feed on scraps of ingestion by the amphipod. Brown (1961a) meat or on dead invertebrates placed in the noted a similar situation in certain habitats tanks. of England, where defecation by great Digestive tracts of 31 Lirceus fontinalis numbers of Asellus aquaticus L. supplied anaylzed on various occasions indicated a rich source of detrital materialfor a may- that this species is omnivorous. Detrital fly (Chloeon dipterum). materials such as small leaf particles, pieces Of the crayfishesof Doe Run, Cambarus of grass, and undetermined particles were bartoni was seen feeding on various kinds abundant in all animals examined. In addi- of vegetation near Station I, on smaller tion, there were large numbers of diatoms crayfish,on isopods (in laboratoryaquaria), in 18 individuals from the mainstream. on detritus, and on fishes (Minckley and Filamentous green algae (Oedogonium?) Craddock, 1961; Minckley, et al., 1963). were found in 3 specimens from 7-Springs Orconectes rusticus was apparently more Branch, and blue-green algae were in 2 saprophagous,feeding on dead animals of of 5 individuals from Station IV in Septem- various kinds, but also on algae and other ber 1960. Fragments of undetermined ar- plant materials. Of special interest was the thropods, a small portion of an oligochaete, picking up and rapid rotating of snails and head capsules of two tendipedids were (Goniobasis) by 0. rusticus,and to a lesser in a large Lirceus from another spring-fed extent by C. bartoni,with apparentgrazing tributary to Doe Run. In the laboratory, of the algae on its shell. 98 WILDLIFEMONOGRAPHS

Tipula nobilis? from Stations I and II exception that the guts of 3 collected in fed almost entirely on Fissidens. That was February were empty, whereas all others the only food in 11 individuals collected at had their entire guts filled. Baetis was the Station I in June 1960, but 3 specimens most important single food item, followed from that station in February 1960 and 2 by tendipedids, smaller stoneflies, and from Station II in September 1960 con- Simulium. Of course, a large proportion of tained appreciable amounts of detritus and the animal material in a given gut could sand (estimated 10% by volume) in addi- not be positively identified, but appeared tion to the moss. to be arthropod material. Analyses of the Among the mayflies, 10 Ephemerella guts of 12 Isogenus subvarians taken in subvaria utilized basically the same foods September and 4 in February indicated as Baetis vagans. Both fed mostly on that the foods were similar to those of epilithic algae and diatoms, with some Paragnetina. Dipterans, however, were detritus. E. subvaria sometimes contained much more numerous than Baetis as the a few fragments of Fissidens or Myriophyl- identified component of the food of Iso- lum at Station II where those plants were genus, probably reflecting its occupancy of abundant. Summer samples of 17 Baetis a habitat more lateral than the midchannel from Station III contained a much larger habitat of Paragnetina. There was a large proportion of detrital material than those amount of particulate detritus in some I. collected in February 1960. Perhaps the subvarians from September, which may higher discharges in winter remove small have been ingested intentionally; leaf par- detrital films from stones or riffles and ticles were larger than those usually found. forces greater utilization af algae (see Surface-feeding gerrids were seen prey- Jones, 1951). Brown (1961a), studying ing on springtails (Podura), a Gerris was Baetis rhodani in streams of Great Britain, seen attacking and eating a Microvelia, but found an excellent correlation between for the most part they fed on terrestrial foods available and gut contents at a given forms that fell into the water and were season. That species fed largely on detritus, transported along by the current (Pennak, supplemented by algae when available. At 1953). Most Hemiptera in Doe Run are Stations III and IV in Doe Run, the gut predaceous. On one occasion Gerris sp. contents of B. vagans were marly, and in- was observed trying to capture free-swim- cluded Gongrosira, Phormidium, various ming amphipods near Station III, but none diatoms, and some Chantransia stages of a was successful. Amphipods trapped in the red alga (probably Lemanea). Stenonema surface film after fish-seining operations, sp., at Station IV in September 1960, fed however, were quickly captured by water almost entirely on diatoms and amorphous striders. detrital materials, with sand being the most Two of the "active feeding" caddisflies abundant item. A few trichomes of a blue- of Doe Run, Agraylea multiplicata and green alga were found in an individual Pychnopsyche sp., were examined for food from Station III. relations. Agraylea contained large num- Four damselfly naids, Agrion macula- bers of diatoms of various genera, domi- turn?, from Station II contained parts of nated by the common epiphyte Cocconeis. Asellus bivittatus, Gammarus sp., an un- The 16 Agraylea were obtained in Septem- identified mayfly, 2 elmid larvae, and ber 1960 at Station II, and were very small. numerous tendipedids. Pychnopsyche was relatively rare in the The large, active stonefly, Paragnetina stream, but 8 individuals were dissected media, was totally predaceous, any vegeta- and the guts contained marly, particulate tion present having been ingested pre- detritus, along with a few diatoms and viously by prey species. No apparent some trichomes of blue-green algae. differences exist in the foods of individuals Snails of the genus Goniobasis grazed on collected in different seasons, with the epilithic algae, diatom slimes, and on thin ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 99 mats of blue-green algae. Grazing was undifferentiatedmud, detritus,tendipedids, particularly heavy on Chantransiaof Ba- Cladophora,a few oligochaetes, and some trachospermum near Station I. In the Stenonema. Animals made up 50 per cent laboratory, Goniobasis grazed on higher by volume and plants 25 per cent. Five aquatic plants only when placed in a newly Moxostomaerythrurum from the same area filled aquarium where algal development as M. breviceps contained mud, detritus, was minimal;in tanks with significantalgal tendipedids,oligochaetes, about 60 sphaeri- growth the higher plants were rarely eaten. ids, some mayflies and amphipods, and a Snails collected at Stations III and IV def- single Tipula. ecated large numbers of small, oblong, The stonerolleris the classic example of predominatelymarl pellets when placed in a feeder on the bottom "ooze"of streams jars. (Kraatz, 1923; Starrett,1950b), and in Doe Run their foods consisted of blue- Foods of Selected Fishes diatoms, green algae, and calcareous bottom ma- Seventeen of the more abundant species terials in the 7 fish examined. A few tiny of fishes were selected for analysis of gen- tendipedids were in each of 2 stomachs, eral food relations because of their con- and leaf fragmentswere in 3. sistent occurrence,at least in the lower part Ten N. atherinoidesfrom Station V con- of Doe Run. tained 65 per cent terrestrial insects in Stomachs of 10 Dorosoma cepedianum, summersamples; the remainingfoods were from near Station IV, contained mostly tendipedid larvae and mayfly nymphs. calcareousbottom materials,diatoms, blue- Five specimens from Station V in Decem- green algae, some Spirogyra, and much ber 1960 contained amphipods, mayflies, detrital material. Six shad from Station V and adult caddisflies. Similar foods, were contained bottom muds and included ten- noted in specimens from northern Illinois dipedid larvae, oligochaetes, and sparse by Forbes and Richardson(1920). detritus. Foods of 17 N. spilopterusfrom Stations Foods of 23 adult white suckers,10 from III and IV were similarto those reportedby Station II and 13 from near Station III, Starrett (1950b) in the Des Moines River, were mostly crustaceansand insects. Fish Iowa. Five collected at Station IV in June from Station II contained about 25 per cent contained substantial amounts of seeds estimated volume of Myriophyllum,Fissi- (Platanus), terrestrialinsects, and a smaller dens, diatoms, and detritus,and those from number of aquatic nymphs and larvae Station III contained 25 per cent silt, marl, (mostly tendipedids). Four from Station diatoms, and fine detrital material. Crusta- III contained75 per cent trichopteransand ceans comprised 28 per cent of the esti- 25 per cent mayflies in January1960, dur- mated volume, and included mostly Gam- ing a periodof high discharge.Tendipedids, marus bousfieldi from Station III, but from mayflies, and terrestrial insects were the StationII aboutone-third was G. minusalong principal items in the remaining 8 speci- with a few Asellus bivittatus. The remain- mens, 4 of which were taken in autumn, der of the foods was large dipteransas the and the others in spring. most abundant insects, followed by mayfly The foods of 7 bluntnose minnows from nymphs and hydropsychid larvae. Other Doe Run were similar to those of Campos- animals were Goniobasis,Sphaeriidae, and toma, mostly calcareous bottom material, oligochaetes. Five hog suckersfrom Station diatoms, and a few filamentousalgal forms IV contained mostly silt, aquatic insect such as Spirogyra and Mougeotia. Foods larvae, and detritus, with the insects being of 37 Rhinichthys atratulus, collected at characteristicof the riffle habitat of Doe various times at the 3 upstream stations, Run (Baetis, Simulium, Paragnetina, and consisted of amphipodsand isopods (39%), tendipedids). Stomachs of 5 Moxostoma mayfly nymphs (21%), and tendipedids, breviceps from near Station IV contained trichopterans, and oligochaetes combined 100 WILDLIFE MONOGRAPHS

TABLE 11.--PERCENTAGE FREQUENCYOF ITEMS a larval Chauliodes. A single small rock FOUND IN STOMACHSOF 22 LEPOMIS MEGALOTIS bass contained2 grasshoppers(Locustidae) (5 EMPTY), 30 L. CYANELLUS (4 EMPTY), AND 20 and a Stenonema. L. MACROCHIRUS(3 EMPTY) FROM STATIONSIII AND IV AND BETWEENTHOSE STATIONS, DOE RUN, Foods of green sunfish, Lepomis cyanel- MEADE COUNTY,KENTUCKY, JUNE 1960-1961 lus, in Doe Run were comparedwith those of longear sunfish and bluegill (Table 11). Item L. L. L. megalotis cyanellus macrochirus The fish were selected to conformgenerally in size, locality of collection, and date of Oligochaeta - 3.3 10.0 Isopoda 9.1 6.6 5.0 collection, and thus should be comparable. Amphipoda 13.6 26.4 5.0 The foods of L. cyanellus were more di- Decapoda 18.2 36.3 - versified than those of the other species, Ephemeroptera 59.1 19.8 20.0 and especially more so than those of the Odonata 4.5 3.3 smaller Plecoptera 4.5 3.3 bluegill. Although aquatic organ- Coleoptera 4.5 3.3 isms made up a consistentpart of the diet, Trichoptera - 3.3 crayfish were found in almost 40 per cent Diptera 31.8 6.6 55.0 of the stomachs, and probably comprised Gastropoda 4.5 3.3 more than the estimated 80 cent of the Terrestrial arthropods' 22.7 52.8 65.0 per Fishes - 16.5 volume. The occurrence of fishes in the Algae and higher food is also of note since that food item was vegetation - - 35.0 not found in the other 2 species. The most Unidentified materials 13.6 6.6 15.0 frequent items were terrestrialarthropods, 1 Includes Hymenoptera (ants), Hemiptera (reduviids), which made up 8 per cent of the estimated Arachnida (spiders), Homoptera (leaf-hoppers), Coleoptera (scarabaeids and carabids), Orthoptera (locustids), and volume. Bluegills fed heavily on terrestrial winged dipterans that have larval stages in water (tendi- which made more than 60 pedids, simuliids, and tipulids). invertebrates, up per cent of the volume, followed by algae and higher vegetation at 25 per cent. Dip- (30%o). A number of Goniobasis were tera and mayflies made up most of the found in larger fish, along with Gammarus remaining volume of food in that species, bousfieldi as the most common amphipod. with the former occurring in more than Smaller blacknose dace contained a larger half the stomachs examined. Longear sun- proportion of tendipedids and oligochaetes fish reflected their habit of frequenting than the larger fish, reflecting somewhat eddying currentsby the preponderanceof their habitat over bottoms. Five soft, silty mayflies in their stomachs (55%by volume, creek chubs, ranging from 1.0 to 1.3 inches but also fed on total had filled with mostly Baetis), heavily length, guts aphids, terrestrial insects, dipterans, and amphi- collembolans, and mosquito larvae. Larger fish fed on more diverse foods, such as pods (Table 11). Etheostoma from crayfish, trichopterans, tendipedids, Tipula, Thirty-six flabellare Station IV were and 2 crustaceans, terrestrial arthropods, and a analyzed, items, considerable proportion of vegetation. One Baetis and chironomid larvae, made up individual about 8 inches long contained 99.4 per cent of the total estimated volume 27 fish eggs and a smaller individual of its (the former was more common at 84.9%). own species. Diverse food habits of S. Two of the fish had eaten small stonefly atromaculatus have been found elsewhere nymphs. (Forbes and Richardson, 1920), and are The smallest Cottus carolinae in Doe indicated by the extreme plasticity of the Run subsist mostly on tendipedids. At species in its tolerance of different environ- Station I, food items change gradually to mental conditions. consist mostly of amphipodsand isopods in Stomachs of 9 rock bass contained 95 per intermediate-sizedfish and then to cray- cent crayfish by volume. Other animals fish and fishes in larger adults. At Station found were 2 Notropis atherinoides? and IV, food habits followed the availabilityof ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 101 food by utilizing mayfly nymphs in the in- on algae, higher plants, and detritus,and a termediate size ranges, and then crayfish well-defined group that preys on the "pri- and fishes as adults. These generalizations mary consumers" and uses very little are based on data obtained from about 300 vegetable material. Among the fishes, only sculpins and reported by Minckley, et al. 3 of 17 species whose foods were studied (1963). can be considered primary consumers, Variationsin stomach contents of Cottus Campostomaanomalum, Pimephales nota- in Doe Run, with size of the fish, time of tus, and Dorosoma cepedianum, and each day, locality of capture, and other factors utilized a large amount of detritus in addi- emphasize the survey nature of this study tion to the algae and other vegetation. of the other animals of the stream, and The assemblage of animals abundant at should be kept in mind when interpreting StationI is shown in Figure 29. The trophic data given here. A review of the determi- structure there was relatively simple and nation of food habits of fishes was reported self-contained: Fissidens and its associated by Hynes (1950), and points out factors of epiphytes were the major producers;Asel- error. lus bivittatus fed on the epiphytes and Tipula nobilis? directly on the Fissidens; Trophic Structure Gammarusminus fed on feces of A. bivitta- When estimated percentages of volumes tus and on detritus from decompositionof of food in digestive tracts of the animals Fissidens and associatedplants; and Cottus studied are arranged in "trophic spectra" carolinae was the "top carnivore,"or sec- (Darnell, 1961), there are distinct differ- ondary consumer. Gammarus bousfieldi, ences between nutritional relations of in- on the other hand, was more omnivorous, vertebrates and fishes (Fig. 28). The and somewhat bridged the hiatus between invertebrates comprise 2 fairly distinct the "detritusfeeder" (G. minus) and the groups, the primaryconsumers that depend primary consumers (Asellus and Tipula).

BENTHIC INVERTEBRATES FISHES

"PASSIVE FEEDERS" "ACTIVE FEEDERS FOOD CATEGORIES : { : - : I i l i: l l i 7i

(25) (4) (13) (4 (14)( 4 D1)(0) (57) (-) (- 1) (-( 1 l (11313)4) 4)( 01 1 0 ) ((I ) (13)1-) (10 (23) (5) (S)(9) (15 7) (M) 5 ( 7) (07) CM((0 (3 (21 (22)I (2-)

FISHES

TERRESTRIAL - -| INVERTEBRATES i I I

INVERTEBRATES |||-||i||ll - -- PLANTS |llIi ll-*l:i I

INORGANIC - DETRITUS I I

FIG. 28. Trophic spectra of some consumer species from Doe Run, Meade County, Ky. Horizontal bars represent the percentage of a given food category in the digestive tracts of a species. Those species for which the number of specimens used in food analyses is not given are included on the basis of observational data or on data of Walker (1962) and Minckley, et al. (1963). Asellus bivittatus was separated as "A" from Stations I and II, and "B" from Station V, because of obvious differences in food habits. 102 WILDLIFEMONOGRAPHS

COTTrrS CAROLINAE the 17 fishes ate substantial amounts of detrital material, and 2 of the 7 predatory species ate significant amounts of algae in CAMBARMUSBARTON7 addition to their animal diets. Examination of the primary foods of in Doe Run reveals ALLOCHTHONOUSDETRITUS. A.AMMARUSMINUS - predators (Table 12) relatively few prey species, and the most

x - AUTOCHTHONOUSDETRITUS CAMMARU/SBOISIELOI-- abundant ones bear the brunt of most feeding activity. When the foods of the princi- NISTURTAIUMOFfICAIALE /7 pal prey species are examined (Fig. 28),

- rIPLA NOBLlS 7, \)FSS/DENS ULIAHS' / it is apparent that detritus is at least as important as primary production in the - ASELLUSBIVITTATU' EPIPHYTICALGAE/ food chains or webs (see also Muttkowski and Smith, 1929; Jones, 1950, 1951). Ani- LIGHT,WATER *nd DISSOLVEDNUTRIENTS mals of "higher" trophic levels in Doe Run FIG. 29. Food relations of the most abundant generally show no restriction to feeding on species of animals at Station I, Doe Run, Meade primary producers and depend heavily on County, Ky. (modified from Minckley, et al., detrital materials, both autochthonous and 1963). allochthonous. When secondary consumers are considered as a whole both The large, vagile, aggressive Cambarus (including fishes bartoni encompassed all nutritional levels predatory and invertebrates), there in its range of food selection. is no distinct point of separation between the and levels of nutri- When the species just discussed are ex- primary secondary cluded from Figure 28, the trophic struc- tional organization. This is especially ap- ture of more downstream parts of Doe Run parent if terrestrial invertebrates, listed as is illustrated. Detritus was more important a separate category in Figure 28, are in- as a food downstream than at Station I and cluded as "detritus." Such action probably was utilized by almost all consumers not is justified because their provenance is 'the markedly predaceous. The "top carnivores" same as fallen leaves. are well separated from the primary levels This analysis suffers from lack of specific when fishes are excluded; however, 10 of information on caloric values of detritus,

TABLE 12.-PRIMARY ANI) SECONDARY COMPONENTS OF THE MAJOR FOOD GROUPS USED BY CONSUMERS IN DOE RUN, MEADE COUNTY, KENTUCKY, BASED ON ANALYSES OF DIGESTIVE TRACTS

Major Food Groups Primary Components Secondary Components Fishes Notropis spp., Etheostoma flabellare Rhinichthys atratulus, Semotilus atromaculatus, Cottus carolinae, fish ova Terrestrial Ants (Hymenoptera), leaf-hoppers Hemiptera, Coleoptera,Orthoptera, Diptera, "Lum- invertebrates (Homoptera), spiders (Arachnida) bricus,"others Benthic Tendipedidae, Gammarusminus, G. Oligochaeta, Turbellaria, A. stygius, Lirceus fon- invertebrates bousfieldi, Asellus bivittatus, Baetis tinalis, Crangonyx, Stenonema, Ephemerella, vagans, Cambarusbartoni (as Ephemera, Baetis, Agrion, Paragnetina media, young), Orconectes rusticus Isogenus subvaria, Hydropsyche, Cheumatopsy- che, Elmidae, Laccophilus, Tipula nobilis?, Ath- erix, Simulium, Antocha, Goniobasis, Sphaeriidae, others Algae and Diatoms, Vaucheria, Fissidens Cladophora, Gongrosira, Cyanophyta, Ulothrix, higher plants julianus,Nasturtium(?) other algae, Myriophyllum, Nasturtium(?) Detritus Autochthonous: Feces of inverte- Drifted Fissidens (downstream), drifted Myrio- brates phyllum, drifted Nasturtium, algal materials(?) Allochthonous: fallen leaves of Humus, "mud,"inorganic materials ripariantrees ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 103 the small numbers of animals examined, very capacious alimentarycanals, feed con- and informationon specific food relations tinually and yet may display slow growth." of tendipedids. On the basis of published Teal (1957) found that larval Anatopynia reportsand cursoryexamination of the rela- (Diptera) feeding on vegetable detritus tive abundance of certain groups of these lost considerable weight, indicating low animalsit appearsthat detritus-feedingspe- conversion of the material to a usable cies make up majorparts of their numbersin form. Brown (1960) showed that algal Doe Run. This is borne out by the pre- foods were more thoroughly digested than ponderanceof species of the tendipedidsub- vegetable detritus by the mayfly Chloeon families Tendipedinae and Hydrobaeninae dipterum, and indicated that more caloric that are predominantlydetritus feeders and value was derived from the former. How- phytophagous (Leathers, 1922). ever, in a later paper, Brown (1961b) in- The detrital componentof food of stream dicated that algae required chewing and animals is a heterogeneous mixture of the expending of energy by the insect, allochthonous and autochthonous debris, whereas small-particleor amorphousdetri- primarily of vegetable origin. Generally, tus was merely brushed up by the mouth- detritus is an amorphous complex formed parts and passed directly into the gut. by the decomposition of plants that accu- Darnell (1961) pointed out that much mulates in interstices of the stream bed detrital materialingested by fishes or other such as in gravels or in beds of aquatic animals includes a large proportion of vegetation (Jones, 1950). In streams such protoplasm in the form of bacteria (see as Doe Run that flow through wooded also ZoBell and Feltham, 1938, 1942; Gneri areas of temperate zones, fallen leaves and Angelescu, 1951). Probably much of contribute the greatest mass of detrital the food value of detritus is in the form of material (Jones, 1951; Hynes, 1961). As bacterial protoplasmand large numbers of floating leaves become saturatedthey settle other microorganismssuch as ciliate pro- in eddies and pools along the streamcourse, tozoans. This was noted by Minckley,et al. or form "leaf packets" (Badcock, 1954) (1963) in their designation of "detrital against obstructionsin riffles. The distinc- materials (allochthonous, autochthonous, tion between thoroughly digested plant and contained food chains)" in the food foods and true detrital materials is not an webs of Cottus carolinae in Doe Run. easy one, but the use of copper sulfate to Hynes (1961), furthermore, found the fix chlorophyll materially aids in the final abundance of animals in a small stream in decision. Also, collections of food materials Wales related to the abundanceof detritus from the stream may be used as points of (fallen leaves), with algae-feeding species reference. The presence of finely divided occurring in summer and detritus-feeding "grit"also seems to be an indicator of a species developing rapidly in winter (see true, amorphousdetritus in the diet of an also Jones, 1951). Unfortunately,my data animal (Jones, 1950). are not precise enough to permit such an Brown (1961a) recently reviewed the analysis for Doe Run. importance of detritus to aquatic inverte- It is apparent that accumulation of or- brates (principally mayflies), and noted ganic detrital materialsin streamsincreases that for much of the year detritus may be downstreamfrom the source (Odum, 1956, composed of crude fibers and inorganic 1957a,b), both throughthe developmentof particles that may be indigestible by most animals and plants in the stream itself, and species (see also Birch and Clark, 1953). through continual accumulationof leaves, Jones (1950) noted the high proportionof twigs, and other terrestrialmaterials. Often, cellulose and "valueless"mineral matter in accumulationsof leaves in Doe Run were detritus, and remarked". . . it would not sufficientto block riffles in the downstream seem to form a nourishing diet . . . " and areas, and to cause oxygen deficits in large "Insects feeding upon it [detritus] have pools at night or on cloudy days in spite of 104 WILDLIFEMONOGRAPHS

the relatively swift nature of the stream. systems offer a great variety of orders of Leaf accumulationsoften have been found magnitude in the amount and intensity to cause "blackwaters" in streams (Schnel- of "trophic unbalance"-differences seem ler, 1955; Slack, 1955; Larimore, et al., greatest when comparing riffle and pool, 1959), and although such situations are and least when comparingside and center extremein that they may destroy the fauna of pool. Therefore, they illustrate some of through deoxgenationor other means, they the complex interrelationsof stream ecol- serve to point out the magnitude of accu- ogy, and merit further research and inter- mulationsthat may occur. It is logical that pretation. the rich source of organic material so pro- SYNTHESIS vided would be utilized as it by animals, Concepts derived from stream studies apparentlyis, and that downstreampopula- contrast with those formed from of tions in streams would tend to be study much successional, or other level of more on that source of food community, dependent integration that occur else- than animals of headwater or situa- phenomena spring where because of the prevalenceof geologi- tions. It follows, therefore, that the lower cal influence in water as reaches of creeks and rivers are flowing opposed essentially to pronounced biotic effects in lakes in a state of unless the trophic unbalance, (Forbes, 1925) and in most terrestrial total allochthonous and import (Nelson situations (Allee, et al., 1955). Streamsare Scott, 1962) is considered in the stream's and as a of the unique, open-ended systems (Thieneman, economy, part ecosystem. 1953). from This condition with They pass youth, through develops physiographic maturity,to old or base level at a and biotic successsion in streams toward a age given point in their length, if sufficient time is base level, both of which cause the flowing allowed waters to tend more and more toward a (Adams, 1901). They grow by upstream erosion, and ultimately produce lentic situation. It is not appropriate to a drained a stream consider all detrital materials sedimented peneplain totally by near base level (Nelson and Scott, 1962). in the lower parts of streams as being A of a continuum of in a concept "youthful" placed condition "below the active stream conditions is tenable, however, not- metabolic level" (Odum, 1956) of the withstanding ultimate peneplanation of a system. In Doe Run at least, these deposits given area, because of the sporadic eleva- are actively moved into and through the tion of land masses that would cause a biota on the macroinvertebrateand verte- persistence of the habitat type if spring brate levels, in addition to the microorgan- waters or runoff were present. Thus the ismal utilizations. Heterotrophic bacteria idea that the downstream parts of rivers and associated protozoans probably form are the only places where stream climaxes an intermediatelevel of utilizationbetween can occur (Shelford, 1911a, 1913; Shelford, primaryproducers (both auto- and alloch- in Gersbacher, 1937; Odum, 1956, 1957a; thonous) and the nutrition of the macro- Nelson and Scott, 1962) may be seriously biota (Darnell, 1961; Nielsen, 1962). questioned, especially since there are large Gradients in trophic structure in Doe numbersof plants and animalsthat display Run, from a more or less autotrophictype adaptations for a riffle habitat (Gessner, to a more heterotrophictype, also appear 1955, 1959; Butcher, 1933; Hora, 1930; when analyses are made of animals living Nielsen, 1950a; many others). A riffle is on riffles and in the adjacent pool, on ani- in essence a short section of "youthful" mals living on the sides of pools and those streameven in a watercoursethat is mostly living in the center where sedimentationof at base level. debris is more pronounced, and from the Climaxesin the Vegetation of Streams swift "oligotrophic"center of a riffle to its The concept of climax in terrestrialsitu- quieter, more "eutrophic" sides. These ations apparently has been interpreted in ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 105 different ways by different investigators was not solved by the formal, latinized (Whittaker, 1956; Selleck, 1960), and in designations of associations or communities aquatic habitats the presence of climax vege- of aquatic plants proposed by Margalef tation has been debated (Eddy, 1925, 1934; (1958, 1960) and others, even though the Shelford and Eddy, 1939; Blum, 1956b, et communities appear extraregional in nature. seq.; and others). If one accepts the mono- I feel that much disagreement among in- climax approach, with reservations as out- vestigators lies in the fact that streams are lined by Oosting (1953), that is, with a biological and chemical individuals. There preclimax and a postclimax, one on each is an absence of specific knowledge of side of a given climactic area, the vegeta- physics and chemistry of streams in which tion of a lake could be considered as one plant communities occur, and a general or the other of these, as a hydrosere, or failure to recognize that the "weather" in perhaps most logically, as a subclimax. a creek or river, that is, the variations of Streams are influenced by the climax vege- chemical and physical conditions from day tation of the regions through which they to day, certainly averages out to a climate pass (Margalef, 1960), and have innate in a manner similar to meterological data chemical and biological differences not from terrestrial stations. Thus, streams only along their individual courses, but have a climate governed by the origin of from stream to stream, even though their their water, the area through which they basic physical characteristics may be simi- flow, biological activities occurring in them, lar (Leopold, 1962). and the meteorological climate of the re- Whether the relatively permanent as- gion through which they pass. These semblages of benthic algae and higher factors may be compared, respectively, plants in streams can be considered "climax" with the influence of mountain ranges on is debatable if one adheres to some impli- provenance of air masses to a terrestrial cations of the term as it applies to terres- region, the influences of local topography trial vegetation (Blum, 1956b, 1960). It on temperatures and rainfall, the influence may seem more realistic to consider these of biological activity in modifying the assemblages in the same perspective as a chemical composition of the air, and the perpetuated disclimax, that is, as represent- influences of the meteorological climates ing a set of communities adapted in such that surround the region. Current is a a manner that they can quickly recover major climatic factor in streams, acting in from chronic and severe depletion as soon a manner similar to rainfall on land in in- characteristics. as favorable conditions recur, whereas the fluencing edaphic An climate is not with- true climax vegetation is suppressed by aquatic proposed out and 1929; some decimating factor. The use of the precedent (Shelford Eddy, Clements and Shelford, 1939; Burton and term disclimax, however, necessitates recog- 1945; and and to nition of a climax and such a Odum, others) appears vegetation, differ from that of terrestrial situations not only single, specific ecological stage may in the medium involved and the methods exist in waters. Also, the definition flowing used in measuring its different components. of the term climax a successional implies This is not to say that terrestrial climates of or events that sequence vegetation many do not influence aquatic ones, and vice times do not occur in streams. The first versa, but the tendency for water to resist colonist of a substrate in given a stream rapid change in temperature, for example, often maintains its dominance until a major makes it considerably different from air, change occurs (Blum, 1956b). Hence, the and the presence of nutrients in the medium problem lies in whether or not a "climate" surrounding an aquatic organism, makes occurs in flowing waters, and in the defini- water an even more different kind of place tion of climax that implies arriving at a in which to live. A major influence of the terminus through succession. This problem terrestrial climate on that of streams is the 106 WILDLIFEMONOGRAPHS

occurrence of abrupt, torrential rainfall, change in the averages of various chemical which allows full expression of the climac- and physical factors. teric nature of stream climate in the oc- A physiographic approach to the study currence of floods. of stream organisms was first used by Shel- Succession in streams is, however, more ford (1911a), and in his early work (Shel- difficult to demonstrate than climate, ford, 1913) considered riffle communities mostly because of seasonal periodicities of as "quasi-climaxes," generally equivalent various species (Transeau, 1913) and be- to lichen community colonizing a granite cause of catastrophic events in the stream. boulder as a climax. However, later on, However, many successional series may be Shelford (in Gersbacher, 1937) dismissed detected if communities are examined in the riffle as a possible climax community, detail (Eddy, 1925). The realization of and considered only larger, base-level succession in streams often is a community streams or relatively stable pools of smaller that lasts not more than a season (Blum, streams in that category. If I adhere to 1956b), but which recurs time after time this concept, my arguments about the pres- indicating attainment of some kind of ence of smaller climaxes on riffles and the climax condition in response to the physical reassessment of Shelford's physiographic properties of the habitat. Communities approaches to stream ecology are to no that are relatively permanent in time often avail. However, since streams are an inte- are found in large springs. Examples of gral portion of the land masses of the earth, this condition are the established or season- and continual or periodic rejuvenation of ally prominent Cladophora, higher plants, those land masses cause a continuum of and calcareous algae in Doe Run. erosion cycles, it seems illogical to assume Because stream climates tend to be in- that riffles have ever become nonexistent. dividualistic, they should be approached The concept of climax ecological groupings, from stream type rather than from regional specifically those organisms morphologi- aspect (Whitford, 1960a). Equivalent habi- cally suited for existence in swift water, tats in streams support ecologically equiva- should immediately follow recognition of lent communities of algae and higher the continual presence of streams per se. plants, and in this respect are comparable The categorical dismissal of the importance of histories in to terrestrial situations. Regional relations physiographic interpretation of stream succession and are observed in that plant communities in climaxes Gersbacher must have resulted widely separated localities consist of the by (1937) from a lack of recognition of the same species, or at least genera or families, primary dichotomy between lakes and streams, and and there is a tendency for rapid appear- was an attempt to the two. I feel ance of "characteristic" forms where suit- equate also that the outline of stream succession, able climatic or microclimatic conditions as given by Margalef (1960) and suggested similar prevail (Whitford, 1960a). Springs by Clements and Shelford (1939), is un- to Doe Run should, and apparently do, acceptable in the sense that it includes a have similar communities of plants, and final transition to a terrestrial community. these may represent certain subclimaxes or Conversion of a given stream to dry land climaxes in the extant environments. There- occurs, to be sure, but is usually abrupt in fore, when one accepts the concept of a geological sense, and most often is a climate in flowing waters he may seek the function of major climatic change or modi- climax vegetational condition of a stream fication by man, rather than one of and investigate succession in that habitat. vegetational succession in the stream itself. Linear "succession" in streams illustrates However, this statement does not apply the above arguments very well, since it wholly to headwater streams fed by surface results mostly from the response of differ- runoff that may show fluctuant or sporadic ent plants or animals to the linearity of cycles of flow and intermittency. Allorge ECOLOGY OF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 107

(1921) gave the only example of vegeta- around a hill and observing changes in tional succession from streams to dry land vegetation or fauna resulting from differ- that I have found in the literature. ences in exposure and variation in amounts of insolation or rainfall. Invertebrates Assemblages of Aquatic My criteria for designating an assem- Armitage (1961) pointed out the fluctu- blage of stream organisms as given above ation in populations of riffle organisms were more or less physical because I at- and the extreme variation of riffle habitats tempted to discern a sharp break in a given in streams, and concluded that rather than system, then to examine it to find whether there being discrete communities, ". . . it biological groupings in one area did not seems more likely that stream habitat con- occur in the other, or vice versa. Although sists of a variety of ecological factors this approach may seem subjective in first interacting in numerous and complex ways analysis (Andrewartha and Birch, 1954), to produce many environmental gradients there are rather definite boundaries to along which species are distributed." His habitats in streams that are related to actual conclusion may be true regarding the fauna differences in fluid masses such as hy- of riffles of the swift, turbulent Firehole draulic jumps at the foot of riffles, obvious River in which he worked, but I do not see areas of shear between current and eddy, where its acceptance necessitates disregard and the obvious "spillage line" at the point of natural assemblages of organisms that where water of a pool drops suddenly to exist in streams where pool-riffle develop- the riffle downstream. ment predominates. The 3 strata of habita- I consider 3 basic assemblages of benthic tion in a riffle (Shelford, 1913; Pearse, invertebrates in Doe Run, 2 of which have 1926 )-the surface of the stones in current, long been recognized as a "pool fauna" between the stones in open interstices, and and a "riffle fauna," and the third as the the sediment-filled interstices below the "spring fauna" defined below: influence of current-may, however, be (1) The animals of swift, unvegetated considered the basic habitats from which riffles downstream from Station II, includ- communities of riffle invertebrates may be ing the marl riffles at Stations III and IV described. and the rubble riffles at Station V, consti- Fluctuations or sporadic changes in in- tute a relatively cohesive unit (Tables 6-7). vertebrate populations resulting from local The assemblage was characterized by a variations in environment (Moffett, 1936; preponderance of aquatic stages of may- Mottley, et al., 1939; Muttkowski, 1929; flies, caddisflies, dipterans, and stoneflies. in Badcock, 1954b; Armitage, 1958, 1961) Many animals living in this habitat are to diverse life response cycles (Hynes, passive feeders (Mueller, 1955) in the 1961; Tebo and Hassler, 1961; and others) sense that they depend on current to bring demonstrate the dynamic nature of the their provisions. This fauna and the habitat rather than a total grouping corresponds generally to the lack of community structure. I can see no "Cladophora-Hydropsyche- Etheostominae" community of basic differences between changes in the Shelford, which he considered as the numbers or kinds of animals on a given essentially ". . . of water bottom com- riffle, in response to receding water levels only type rapid at low attitude in or to life cycles of the constituent species, munity middle North and changes in a decomposing log with America" (Shelford and Eddy, 1929). onset of summer desiccation, and realiza- (2) The second assemblage included the tion of life cycles of the animals present. animals of vegetated or unvegetated, sand- Likewise, the changes in composition of or silt-sand-bottomed pools. Gersbacher fauna from the sides to the center of a (1937) concluded that there was a single riffle (Needham and Usinger, 1956) are no major pool community characterized by greater than those found while passing chironomid dipterans in small streams, and 108 WILDLIFEMONOGRAPHS by certain mollusks, the mayfly genus current, and usually are inhabited by a Hexagenia, and oligochaetes (Limnodrilus) fauna similar in most of its constituents to in large creeks and small rivers. In Doe that of the main riffle. Run, pool communities were similarly Such transitory habitats occurred in the dominated by dipterans, oligochaetes, and small riffles lateral to marl and rubble gastropods, except in Nasturtium beds at riffles at Stations III, IV, and V. Here, Station I (Table 8). there was a general increase in the relative (3) The third assemblage was that of the abundance of animals typical of pool situa- spring, comprising a fauna dominated by tions, and a shift in the proportion of active- crustaceans in Fissidens (Table 6). This feeding riffle animals such as mayfly major assemblage caused modification of nymphs, when compared with the passive- lateral and intermingled patches of incip- feeding trichopteran larvae. In spite of ient communities more similar to those these changes, the areas lateral to the main downstream to the extent that they were riffle channels are obviously "riffle" when sometimes obscured, even at Station II. their faunal composition is compared with Such situations may be typical of cool other habitats. limestone springs, with a fauna limited in Pools comprise 2 distinct habitats for numbers of species but characterized by a benthic invertebrates, the surface of the large number of individuals (Odum, 1957a, substrate and the area beneath that surface. b). Insect larvae are generally rare at the Pool animals show fewer adaptations in sources of springs (Berner, 1950; Sloan, their morphology, habits, and overall re- 1956), and the absence of that group alone sponses to their environment than animals is significant. characteristic of riffles. This is indicated I do not mean to imply a lack of transi- by the greater numbers and diversity of tional habitat between pool and riffle, nor species that reach their greatest abundance that all animals inhabiting riffles are pres- in pools but can also inhabit certain parts ent on the surface of the substrate and of riffles. The morphology (Dodds and under influence of currents. On a given Hisaw, 1924a et seq.; Hora, 1930; Nielsen, riffle there are flattened or otherwise modi- 1950a) and an actual physiological need fied animals that inhabit the tops of stones, of many riffle animals for current (Wu, the interstices of the bottom where a flow 1931; Ruttner, 1953; and others) restricts of water persists, and beneath stones or in them to that habitat. In many forms the sediment-filled interstices (Shelford, 1913). passive-feeding adaptations, such as those These last-mentioned organisms may in- found in net-building caddisflies, dictates clude forms common in softer bottoms of their distribution to areas where current pools. Also included on riffles are animals will supply food (Mueller, 1955). Appar- living in special habitats such as the leaf ently, few pool animals suffer from these packets colonized by Lirceus and other restrictions. Moreover, there is evidence invertebrates in Doe Run, or in temporary that basal metabolic rates of riffle-inhabit- deposits downstream from obstructions ing mayflies (Baetis) are several times which may be colonized by a multitude of greater than that of mayflies living in ponds animals from pools and riffles alike. These or pools (Fox and Simmonds, 1933), and special or transitory habitats technically are that the food intake of the former is pro- not part of the riffle assemblage, but repre- portionately greater. sent peripheral environments that occur Pool situations at Station II and down- because of the linear nature of flow in stream were remarkably consistent in the streams. Areas lateral to riffles or to the overall composition of their fauna, in spite main center of velocity may be considered of the specific bottom type or the presence as transition zones between riffle and pool, or absence of vegetation. Vegetation in but are under relative control of the riffle pools, however, provides more surface area because of fluctuations in water level and for habitation by animals (Krecker, 1939; ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 109

Rosine, 1955), protection from predation, of their faunal dominants. Because of the and food for many invertebrates, either very large numbers of animals in the Fissi- directly (Frohne, 1956) or indirectly by dens, these large populations must have the development of periphyton (Young, influenced adjacent, smaller habitats, es- 1945; Rosine, 1955). Vegetation in pools pecially when maximalamounts of vegeta- augments a lentic tendency already present tion were present. Animals characteristic by physically slowing the water, and con- of downstream riffles, however, formed a tributes to eutrophicationof the habitat by more significant part of the fauna of un- accumulationof autochthonousand alloch- vegetated riffles than of vegetated riffles thonous debris. Vegetation in streams at StationsI and II. If the crustaceanswere and lakes rarelytends to alter the composi- extirpated,the faunal compositionon riffles tion of the fauna (except, possibly, in upstream would approximate that of the Nasturtiumbeds of Doe Run), but almost downstream areas (with the exception of invariably results in increased biomass mayflies at Station I). Myriophyllumbeds (Hynes, 1960). Nasturtiumbeds at Station at Station II, whose intermediatecharacter I and open pools at Station II were very has been stressed,somewhat bridge the gap similar in faunal composition,but differed between assemblages with faunal constit- from pool habitats downstream mostly in uents of both riffle and pool and the abundanceof Gammarusbousfieldi. I con- crustaceansof upstreamhabitats. sider these variations from the basic pool The continuity of these communitiesin assemblages to be produced by influence time and space is maintained by rate of of the special fauna of the spring. flow, which depends on discharge. Silt All the swift-water areas at Station II deposits on riffles ultimately are removed and upstream were dominated by the by moving waters and deposited in pools. closely matted Fissidens, and this, in effect, Maintenance of longitudinal patterns of produced a pool environmentin the midst distribution of animals in Doe Run was of riffles. Evidence of this was obtained effected by the presenceof suitable habitat, by introducingsmall amountsof fluorescein Fissidens, for those species characteristic into the moss beds with a pipette and ob- of the spring. Also, because of the linearity serving the time required for the dye to in the system, those species were found in disappear. Dye was visible issuing from reduced numbers downstreamprimarily as smooth moss beds upstream from Dam 1 a function of drift during increased dis- for from 3 to 30 minutes when velocities charge. Depletion of invertebratepopula- over the beds ranged from 2 to 4 ft/sec. tions in upstream areas by flood or by After the dye had apparently dissipated, innate drifting is rectified by movement pressure applied to the area forced some upstream, as in Gammarus bousfieldi residual dye from the deeper parts of the (Minckley, 1963) and Goniobasis,from the beds. mill ponds and other pools (Beauchamp, The animals prevalent in the Fissidens 1932;Dendy, 1944), and by upstreamflight community, Asellus bivittatus and Gam- of egg-laden females of aquatic insects in marus minus, were not adapted to a riffle response to down-valley winds or other environment,but ratherto the close-spaced, factors (Roos, 1957). Larvae and nymphs quiet habitat inside the protective upper in upstream areas must become somewhat layer of moss. Conditions for large popu- crowded for space, and drift of these ani- lations were enhanced by the great surface mals is presumably a mode of preventing area of the moss, the protection offered, or reducing competition, and a means of and by the abundant epiphytes and de- allowing the animals to disperse down- tritus. stream into all areas suitable for their Unvegetated riffles at Stations I and II development (Mueller, 1954). This was obviously were part of the Fissidens or substantiated by Waters (1961) who re- spring assemblageof Doe Run on the basis ported that the drift of animals with 2 or 110 WILDLIFEMONOGRAPHS more generations per year in certain Minne- waters to inhabit the deepest pools as sota streams was correlated directly with adults, although riffles were utilized for their standing crops, and he proposed the spawning and possibly for some feeding. measurement of stream drift as a new ap- Because of their greater mobility, size, proach to estimating the productivity of and greater plasticity of ontogenetic and in- streams (Waters, 1962). stantaneous response to the stream environ- Development of longitudinal distribu- ments, fishes do not fit into the assemblage tional patterns of animals of the down- systems discussed above. Rather, they are stream parts of Doe Run did not correlate superimposed on those systems and utilize with substrates. The upstream limits of a greater proportion of the overall ecologi- some species may have been a function of cal resources of the creek than do other the relative lack of drifting foods presum- organisms. This is indicated by the broad ably required by passive feeders (simuliids, spectrum of foods utilized by fishes (Fig. hydropsychid larvae, and some Ephemerop- 28). The apparent lack of such plasticity tera). Temperature, however, was a major in other animals, however, may be artificial. factor in Doe Run and elsewhere (Dodds I will not be surprised if more detailed and Hisaw, 1925b; Ide, 1935; and Sprules, research on the ecology of aquatic inverte- 1947), either by its constancy at the spring brates indicates somewhat comparable that may have allowed a species to persist ontogenetic changes in habit or foods by but not to complete its life cycle (Ide, some, at least in the larger species. Such 1935), by being so low that it was unsuit- changes are indicated by limited data on able for life of many animals of warmer, crayfish, and on some mayflies (Harker, more fluctuant stream climate. Species in 1953; Macan, 1957; Brown, 1961a) and the spring assemblage were adapted to stoneflies (Hynes, 1941). those conditions, occurred as stragglers, or CONCLUDINGREMARKS represented more northern forms as relicts or outliers of their major centers of abun- I hope to clarify my concept of biotic dance. dispersion in streams by stressing ecologically equivalent communities rather than The Role of Fishes communities based on taxonomic entities. Fishes resident in Doe Run, specifically Therefore, I approached Doe Run from those characteristic of the stream (i.e., the standpoint of ecological equivalence of Cottus, Rhinichthys, Catostomus), and the animals and plants in physically similar "creek species" (Fig. 27) that maintained habitats. The composition of the flora and populations through reproduction (Fig. 26) fauna changed in passing downstream from occupied distinctly different habitats in the source, in response to temperature, different stages of their life cycles. For gradient, marl formation, or other factors example, Cottus carolinae, generally con- dependent on physiological responses and sidered an inhabitant of riffles, moved plasticity of the species concerned. How- directly into quiet, shallow, hot, shoreline ever, the general morphology of animals areas immediately after hatching, then as occurring in riffle habitats, for example, they grew they progressively invaded was similar at all stations in that they dis- swifter waters only to return to pool situa- played equivalent adaptations to the rapid tions in deeper currents as large adults movement of the dense medium in which (Minckley, et al., 1963). Larvae of Catos- they lived. This is not to say that precise tomus commersoni likewise moved to quiet adaptations, such as marked flattening of areas, but in gently eddying currents, the body, were displayed by each, but that where the early stages fed on neuston and the organisms occurring on a given riffle terrestrial materials floating on the surface collectively displayed adaptation of some (Stewart, 1926) for a time. Then the kind, be it the presence of strong hooks for suckers progressively moved into deeper attachment, flattening of the body or ap- ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 111 pendages, elaborate nets and tube construc- doubtedly is affected by the bottom ma- tion, or a proclivity for living in burrows terials, but the deposition of marl in the maintained in marl or constructed in silt- midsection of the stream may be a recipro- filled interstices. cal effect where the chemistry of the water, The interrelations of physical, chemical, modified by physical and biological phe- and biological phenomena in streams have nomena, is a causative factor in the type of been stressed repeatedly and are illustrated bottom material present (Fig. 30). diagramatically in Figure 30. Arrows point- There may be some ambiguity in con- ing from causative factors to their results clusions and concepts formulated from the are deleted primarily because of reciprocity study of Doe Run since they are based in most instances (see Major, 1961). This primarily on a single stream fed by a large is not true in some instances, such as the spring. However, after the spring water effect of rainfall on discharge of a stream leaves its subterranean channel, it is im- in which there could be little reciprocity, mediately influenced by the epigean en- but most show some degree of reversibility. vironment through which it flows and by For example, the gradient of a stream organisms and physicochemical phenomena affects its turbulence, depending on dis- within it. It is a stream that should be charge. Conversely, the turbulence of a comparable to other streams in its ecology stream influences its competence for carry- notwithstanding certain features of con- ing sediments, which in turn may greatly stancy or uniqueness superimposed by its affect the speed of downcutting and thus origin. It is hoped that the nature of this the speed of change in gradient, assuming study and any ambiguous conclusions no rejuvenation of the land mass. The drawn from the data will serve to illustrate chemistry of the water in Doe Run un- and stimulate further research in descrip-

FIG. 30. Causes and effects of various factors of stream environment with special reference to Doe Run, Meade County, Ky. 112 WILDLIFEMONOGRAPHS tive phases of the ecology of flowing May 1961, with discharges as great as 600 waters. cfs at Station I (ca. 2,000 cfs at Station V) in the latter SUMMARY period. Turbidities rarely exceeded more than 1. Chemical, physical, and biological 50 "turbidity units" (approximately 50 mg/1) features of Doe Run, Meade County, Ky., except in periods of unusual discharge. were studied from February 1959 to After the riparian vegetation was cleared May 1962, with the most intensive investi- at the downstream section in preparation gations from November 1959 through July for impoundment, turbidities showed a 1961. The stream originates as a large marked increase at Station V. spring, flows for 9.7 miles, and empties Aspects of transport of the bed load are directly into the Ohio River. The surround- discussed in relation to fluctuations of dis- ing land has a rolling, karsted topography, charge. Major deposits were associated underlain with limestones of Mississippian with beds of macrophytic vegetation in the age. upstream areas. After periods of relatively 2. Doe Run was mapped from topo- high turbidities, deposits of sludgelike ma- graphic quadrangles and an aerial photo- terials occurred on boulders in the channel, graph. "Creek miles" were designated be- and were apparently formed through suc- ginning with Mile 0.0 at the source. Five cessional development of an unique assem- stations were established as follows: Station blage of organisms. I, Miles 0.0-0.3; Station II, Miles 1.65-2.1; Water temperatures at Station I were Station III, Miles 3.0-3.2; Station IV, Miles consistently between 13.0? and 13.5?C in 4.7-4.9; and Station V, Miles 7.7-8.0. periods of modal flow, but increased to 3. Bottoms downstream from the source between 13.5? and 14.5? after the extreme graded from rubble adjacent to the spring, discharges of spring 1961. Ranges from through gravel and sand, to soft silt in an 12.0? to 16.0? were recorded in periods of impoundment by an old mill dam at Mile unusual discharge conditions. The maxi- 1.65. Bottoms at Station II were generally mum water temperatures recorded at Sta- bedrock and rubble. The areas between tions III and V were 20.0? and 25.6?C, Miles 2.1 and 2.36 and Miles 2.36 and 2.82 respectively, and the minima were 6.1? and also were influenced by impoundments 1.7?. The greatest temperature ranges oc- above dams at Miles 2.36 and 2.82. A short curred in late fall and early winter at each section of marl bottom existed between of the downstream stations. Miles 2.36 and 2.5. Marl deposition domi- Surface water in the mill impoundments nated the bottom deposits from Dam 3 often approached 30?C in summer, but downstream to near Mile 5.3, and the waters downstream never exceeded 20.0? gradient throughout that area was about at Station III. This was explained by the 35 fpm. From Mile 5.3 to the Ohio River presence of density currents in the pools, the average gradient was about 9 fpm and which flowed along the bottom in summer the bottoms ranged from broken marl and and near the surface in winter. gravel in the upper part to deep silts near 5. Dissolved oxygen at Station I ranged the river. Pool-riffle development was most between 70 and 85 per cent of saturation. pronounced between Miles 5.3 and 7.0, Supersaturations of oxygen were found in below which the stream resembled a deep, the larger pools in periods of strong insola- slowly moving sluice. tion and large stands of vegetation; passage 4. Discharge in Doe Run was "normal" of water over dams tended to reduce satu- at 20 to 40 cfs at all stations based ration levels to about 100 per cent. Samples on modal conditions. Discharge was gen- from riffle areas usually ranged near 100 erally greatest in December-March and per cent saturation except in autumn when lowest in August-October. Severe fluctua- large deposits of fallen leaves apparently tions occurred in June 1960 and in March- increased oxygen demand in pools and ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 113 caused deficits in downstream areas. The factors. Special attention is given those photosyntheticactivity of algae and higher species associated with marl deposition. plants masked other factors influencing Factors influencing marl deposition in the dissolved oxygen levels except after floods creek, and the influence of the marl on when vegetation was relatively sparse. In aquatic invertebrates,are discussed in de- those periods oxygen tended to follow tail. temperaturerelations and respond to oxy- Higher plants occurred mostly upstream gen demand in pools by remaining some- from the first mill dam, and formed certain what below saturation. depositional features and spatial associa- Carbon dioxide, pH, and alkalinity relations among themselves. Emergent vege- tions were complex and generally consisted tation was relatively sparse because of the of high carbon dioxide, circumneutralpH, stony banks, but Nasturtiumofficinale was and high alkalinity (200 to 280 mg/1) at a major component of the upstream com- the spring, a gradual loss of carbon dioxide munities. Four major submergent plants and increase in pH between Stations I and were Myriophyllum heterophyllum, Pota- II, then rapid elevation of pH, almost total mogeton foliosus, Nitella flexilis, and Fissi- loss of carbon dioxide, and precipitationof dens julianus. calcium carbonate from Mile 2.82 to Mile The gradingenvironment of the upstream 5.3. Downstreamfrom Mile 5.3, large pools section (above Mile 2.0) was occupied by and their accumulated organic materials monospecific stands of vegetation in a apparently caused restablishmentof pH- linear sequence that apparently resulted carbon dioxide-alkalinityrelations in such from the distributionof bottom sediments. a way that marl deposition was inhibited It is concluded that the upstream section in spite of a general, persistentsupersatura- represents an environment for the plants tion of bicarbonate in the water. Diurnal intermediate between large, constant variation in these factors at stations in the springson one hand and "typical,"fluctuant marl area indicated considerableactivity of streams on the other. Thus the stream the vegetation in the deposition of marl, portraysboth the dynamic aspects of cata- presumably through assimilation of half- strophic washout of the vegetation and bound carbon dioxide. recolonization, and a constancy of plant Determinations of total iron, phosphate dispersiontypical of larger, less fluctuating phosphorus,nitrate and nitrite nitrogen are spring systems. presented and discussed. 7. A list of invertebrates is presented. 6. Algal communitiesof Doe Run were Specific ecological relations of certain placed in 4 broad categories: (1) those abundant invertebratesare given in detail, algae with a firm mode of attachmentand including distributionin the stream and its a highly flexible thallus, occurring in cur- tributaries, reproduction, times of emer- rent; (2) cushionlike or encrusting species gence of aquatic insects, and behavior in that grow appressed to the substrate in relationto the varioushabitats in the creek. current; (3) matted or loosely arranged Faunal affinities include a ubiquitous algae occurring attached or unattached in group of animals and some species char- quiet pools, eddies, or in areas of almost acteristic of or related to other species laminar flow that approximate pools and occurringin areas to the north and east. eddies because of lack of turbulence; and Quantitative sampling of the benthos (4) subaerial species attached as encrust- was done monthly from February 1959 to ing mats and occurring in zones of spray, November 1960, with additional series in on waterfalls, on tops of emersed stones, Januaryand June 1961. Samples were ob- or on muddy banks and bars along the tained from preselected habitats in the channel. Algal species are discussed in re- creek, such as beds of Fissidens at Stations lation to their spatial distributions in the I and II, open pools at Stations II, III, IV, creek and to certain physical and chemical and V, and beds of emergent vegetation at 114 WILDLIFEMONOGRAPHS

Stations I and III. Four such habitats were movements are annotated. The 3 major sampled on each date at each station; data factors controlling the distribution of fishes from Station V were fragmentary because were gradient, temperature, and the pres- of sporadic occurrences of backwaters of ence of marl in the midsection of the the Ohio River. stream. Thirteen of 31 species considered The greatest number of taxonomic groups more typical of the Ohio River than of of animals occurred at Stations II, III, and smaller creeks maintained definite ranges IV, with decreases in both the numbers of in the creek, most of which merged into groups and the numbers of individuals at populations of the Ohio River. Many of Station V, but the greatest numbers of the most upstream records of those "river animals per square foot at Stations I and fishes" were at a barrier imposed by a II. Beds of Fissidens were the most pro- bridge and culverts at Station III, and most ductive, as determined from standing crops, were single specimens. A second group of with maxima of 18,708 and 20,004 indi- fishes seemed somewhat restricted to the viduals/ft2 at Stations I and II, respectively. section between Miles 5.3 and 7.0 in the The maximum weights recorded there were area of pronounced pool-riffle develop- 28.695 and 48.937 g/ft2, respectively, with ment, apparently avoiding the open, shal- weights of gastropods and decapods in- low, bedrock pools and swift currents of cluded. Average values from Station I were the upper area, and also the muddy, sluice- about 8,600 individuals and 8 g/ft2, and for like lower portion of the stream. There Station II, about 9,200 individuals and 11 was little reproduction by most species of g (crayfish and gastropods excluded from fishes upstream from Mile 5.3. Marl depo- the weights). Standing crops of inverte- sition may have deterred reproduction of brates graded from largest in Fissidens, those species usually building nests in through Myriophyllum beds, to beds of gravels or other loose bottom materials, and emergent vegetation, thence to stable silty the relatively low temperatures of the up- bottoms, riffles in the channel, riffles lateral stream area appeared to inhibit reproduc- to the main channel, to unstable bottoms in tion of many, particularly sunfishes. Only pools, and the least was found on shifting- 3 species, Catostomus commersoni, Rhi- sand bottoms at Station I. In the upstream nichthys atratulus, and Cottus carolinae, areas, all habitats were greatly influenced maintained populations at Station I, and by the tremendous populations of animals those introduced into the upper area of in moss riffles, and showed increased stand- Doe Run by fishermen or by displacement ing crops in periods of modal discharge from adjacent ponds failed to reproduce. presumably because of movement of ani- Amphibians, reptiles, birds, and mam- mals from moss to more peripheral habitats. mals that appeared to influence, or be in- The most important groups were Crustacea fluenced by, the stream are listed with at Stations I and II, and insect larvae and annotations. nymphs in more downstream areas. Oligo- 9. Examination of the trophic structure chaeta and Mollusca occurred throughout of Doe Run revealed a major dependence the stream system in significant numbers. of primary consumers on allochthonous and Most changes in invertebrate populations autochthonous detrital material. The role seemed related to changes in discharge, of detritus in nutrition of aquatic animals either directly or indirectly, through wash- is reviewed. Analyses of digestive tracts of out or movements into or out of a given invertebrate and vertebrate consumers ten- area. Some changes in populations could tatively indicate that the most "trophically be attributed to seasonal emergence of balanced" section of Doe Run is nearest aquatic insects. the spring, where allochthonous debris is 8. Sixty-six species of fishes were col- minimal. Downstream from the source the lected from Doe Run, and data on the amount of "trophic unbalance," in the sense distributions, abundance, reproduction, and of the fauna depending more on detrital ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 115 materials produced in areas other than of streams, or segments of streams, on the those which they inhabit for their nutrition, basis of taxonomic groupings of organisms. increases with the accumulation of more Similar arguments and illustrative data and more detritus. are derived from analyses of invertebrate 10. Arguments are presented for the assemblages in Doe Run. Two well-defined existence of climates in flowing waters assemblages, based on the concept of eco- analogous to those found in terrestrial logically equivalent adaptation, occur in situations. Climate in flowing waters is the area downstream from Station II: that defined as an average or mode of the of the riffles, and that of pool situations physical and chemical characteristics of a without considering bottom types or pres- given area in a period of time (an average ence of vegetation. Station II is considered "weather"). The major differences between an area of transitions from conditions char- stream and terrestrial climates are pointed acteristic of the spring and those down- out as those of magnitude of size and stream, but is predominantly a spring or change, with many climates being present Fissidens-dominated assemblage on which throughout the length of a stream and certain elements of the downstream fauna locally due to variations in gradient and are superimposed. The spring community water chemistry even in periods of modal is characterized by animals adapted to live discharge. Also, swift and radical varia- in the close-spaced beds of Fissidens and tions may decimate the communities and in the cool, constant-temperature conditions. necessitate a total resumption of succession, Even at Station I, however, the downstream perhaps several times in a single season. pool assemblages are represented in Nas- The communities or assemblages of plants turtium beds. Myriophyllum beds appar- found in streams should be grouped eco- ently are a truly intermediate habitat, logically, according to their habitats and whose fauna includes elements of all other structural adaptations, and climax condi- communities in the stream. tions may be designated as based on Fishes, because of their plasticity of climates present. ontogenetic instantaneous response to their The physiographic approach of earlier environment in streams, apparently are workers is reapplied to the study of Doe superimposed on the other systems and are Run, and put forth as a more realistic ap- not an integral part of any given assem- proach than those involving classification blages of plants or invertebrates. LITERATURECITED

ADAMS,C. C. 1901. Base leveling and its faunal ALLORGE, P. 1921. Les associations vegetables significance. Am. Nat., 35:839-852. du Vexin francais. Rev. Gen. Bot., 33:481- , AND T. L. HANKINSON. 1928. The ecol- 583 (cited in Blum, 1956a). ogy and economics of Oneida Lake fishes. ANDERSON,A., AND A. LUNDH. 1948. Algstudier IV. Annotated list of Oneida Lake fishes. i Vegean. Bot. Notiser, (1948) :284-304. Bull. N. Y. State Coll. For., Roosevelt Wild ANDERSON, R. C. 1959. A modified flotation Life Ann., 1:283-521. technique for sorting bottom fauna samples. ALEXANDER, C. P. 1931. The crane-flies (Tipu- Limnol. Oceanogr., 4:223-225. lidae, Diptera). Deutsche Limnologische ANDREWARTHA,H. G., AND L. C. BIRCH. 1954. Sunda-Expedition. Arch. Hydrobiol., Suppl., The distribution and abundance of animals. 9:135-191. Univ. Chicago Press. xv + 782pp. ALLEE, W. C. 1929. Studies in animal aggrega- ANONYMOUS. n.d. Water and sewage analysis tions of the isopod, Asellus communis. Ecol- methods manual. Hach Chemical Co., Ames, ogy, 10:14-36. Iowa. 52pp. , A. E. EMERSON,0. PARK, T. PARK, AND . 1960. Standard methods for the ex- K. P. SCHMIDT. 1949. Principles of animal amination of water and wastewater. 11th ed. ecology. W. B. Saunders Co., Philadelphia. Am. Publ. Health Assn., New York. xxi + xii + 837pp. 626pp. ALLEN, J. A. 1876. The American bisons, living ARMITAGE, K. B. 1958. Ecology of the riffle and extinct. Mem. Geol. Surv. Ky., l:i-ix, insects of the Firehole River, Wyoming. Ecol- 1-246. ogy, 39:571-580. 116 WILDLIFEMONOGRAPHS

. 1961. Distribution of riffle insects of the 1955a. The attached algal flora of slow Firehole River, Wyoming. Hydrobiol., 17: sand filter beds of waterworks. Ibid., 7: 152-174. 103-107. BADCOCK,R. M. 1949. Studies on stream life . 1955b. The aquatic fauna as an eco- in tributaries of the Welsh Dee. J. An. Ecol., logical factor in studies of the occurrence of 18:193-208. fresh-water algae. Rev. Algol. (n.s.), 1:142- 1954a. Studies of the benthic fauna in 145. tributaries of the Kiivlinge River, southern BROWN,D. S. 1960. The ingestion and diges- Sweden. Inst. Freshwater Res., Drottning- tion of algae by Chloeon dipterum L. Hydro- holm, Rept. 35(1953):21-37. biol., 16:81-96. . 1954b. Comparative studies in the . 1961a. The food of the larvae of populations of streams. Ibid.:38-50. Chloeon dipterum L. and Baetis rhodani BAILEY, R. M., AND W. A. GOSLINE. 1955. Vari- (Pictet) (Insecta, Ephemeroptera). J. An. ation and systematic significance of vertebral Ecol., 30:55-75. counts in the American fishes of the family . 1961b. The morphology and function- Percidae. Misc. Publ. Mus. Zool., Univ. ing of the mouthparts of Chloeon dipterum Michigan, 93:1-44. and Baetis rhodani. Proc. Zool. Soc. Lond., BARm,T. C., JR., AND R. A. KUEHNE. 1962. The 136:147-176. cavefish, Amblyopsis spelaea, in northern BURKS,B. D. 1953. The mayflies, or Ephemer- Kentucky. Copeia, (1962) :662. optera, of Illinois. Bull. Ill. Nat. Hist. Surv. BARRETT, M. J., A. L. H. GAMESON,AND C. G. 26:1-216. OGDEN. 1960. Aeration studies at four wier BURTON, G. W., AND E. P. ODUM. 1945. The systems. Water and Water Eng., Sept., (1960): distribution of stream fish in the vicinity of 1-7. Moutain Lake, Virginia. Ecology, 26:182- 194. BEAUCHAMP, R. S. 1932. Some ecological factors and their influence on competition be- BUTCHER, R. W. 1927. A preliminary account of the vegetation of the River Itchen. tween stream and lake-living triclads. J. An. J. Ecol., 1:175-190. Ecol., 15:55-65. BERNER, L. 1950. The mayflies of Florida. 1933. Studies on the ecology of rivers. Univ. Fla. Biol. Sci., 4:1-267. I. On the distribution of macrophytic vegetation in the BIRCH, L. C., AND D. P. CLARK. 1953. Forest rivers of Britain. Ibid., 21: 58-89. soil as an ecological community, with special reference to the faunas. Quart. Rev. Biol., , F. T. K. PENTLOW, AND J. W. A. WOOD- 28:13-26. LEY. 1930. Variations in the composition of river waters. Int. Rev. BLACK, J. D. 1949. Changing fish populations ges. Hydrobiol., as an index of pollution and soil erosion. Ill. 24:47-80. Acad. Sci. Trans., 42:145-148. CEDERGREN, G. R. 1938. Reofila eller det rin- nande vattnets algsamhallen. Svensk, Bot. BLUM, J. L. 1956a. The ecology of river algae. Bot. Rev., 22:291-341. Tidskr., 22:362-373. . 1956b. The application of the climax CHANDLER, H. P. 1956. Megaloptera. In concept to algal communities of streams. Aquatic insects of California, with keys to Ecology, 37:603-604. North American genera and Californiaspecies. Univ. California Press, Berkeley and Los . 1957. An of the ecological study algae Angeles. Pp. 229-233. of the Saline River, Michigan. Hydrobiol., 9:361-408. CHURCHILL, M. A. 1958. Effects of impoundments on oxygen resources. In Oxygen re- . 1960. Algal populations in flowing in streams. Robt. A. Taft San. waters. In The lationships ecology of algae. Spec. Publ. Eng. Cent., Tech. Rept. W58-2:107-129. 2, Pymatuning Lab. Field Biol. 11-21. Pp. CLAY,W. M. 1962. A field manual of BOUSFIELD, E. L. 1958. Fresh-water amphipod Kentucky crustaceans of glaciated North America. Can. fishes. Ky. Dept. Fish Wildl. Res., Frankfort. Field-Nat., 72:55-113. vii + 147pp. BRADEN,G. E. 1951. Turbulence, diffusion, and CLEMENS, W. A. 1917. An ecological study of sedimentation in stream channel expansions the mayfly Chirotenetes. Univ. Toronto Stud., and contractions. Proc. Okla. Acad. Sci., 31: Biol. Ser., 17:5-43. 72-77. CLEMENTS, F. E., AND V. E. SHELFORD. 1939. . 1958. The hydraulic jump in natural Bio-ecology. John Wiley & Sons, Inc., New streams. Ibid., 38:78-79. York. vii + 425pp. BROOKS,A. J. 1954. The bottom-living algae of CoKER,R. E. 1954. Streams lakes ponds. Univ. slow sand filter beds of waterworks. Hydro- North Carolina Press, Chapel Hill. xviii + biol., 6:333-351. 327pp. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 117

COLE, G. A. 1957. Some epigean isopods and , AND . 1925b. Ecological studies amphipods from Kentucky. Trans. Ky. Acad. on aquatic insects. IV. Altitudinal range and Sci., 18:29-39. zonation of mayflies, stoneflies, and caddis- . 1959. A summary of our knowledge flies in the Colorado Rockies. Ibid.:380-390. of Kentucky crustaceans. Ibid., 20:66-81. DOUGLAS,B. 1958. The ecology of the at- . 1961. Some calanoid copepods from tached diatoms and other algae in a small Arizona, with notes on congeneric occur- stony stream. J. Ecol., 46:295-322. rences of Diaptomus species. Limnol. Ocean- DROUET, F. 1959. Myxophyceae. In Fresh- ogr., 6:432-442. water biology. John Wiley & Sons, Inc., New , AND W. L. MINCKLEY. 1961. A new York. Pp. 95-114. species of amphipod crustacean (genus Gam- EDDY, S. 1925. Fresh-water algal succession. marus) from Kentucky. Trans. Am. Micro- Trans. Am. Microscop. Soc., 44:138-147. scop. Soc., 80:391-398. . 1934. A study of fresh-water plankton COOPER,G. P. 1953. Population estimates of communities. Ill. Biol. Monogr., 12:1-93. fishes in Sugarloaf Lake, Washtenaw County, EDMONDSON, W. T. (Ed.) 1959. Ward and Michigan, and their exploitation by anglers. Whipple's fresh-water biology. John Wiley Pap. Mich. Acad. Sci., Arts, and Lett., 38: & Sons, Inc., New York. xx + 1,248pp. 163-186. EDWARDS,R. W., AND M. W. BROWN. 1960. An CRADDOCK, J. E., AND W. L. MINCKLEY. 1963. aerial photographic method for studying the Reptiles and amphibians from Meade County, distribution of aquatic macrophytes in shal- Kentucky. Am. MidI. Nat. In press. low waters. J. Ecol., 48:161-163. CURRY,L. L. 1954. Notes on the ecology of the , AND J. HEYWOOD. 1960. The effects midge fauna (Diptera: Tendipedidae) of of a sewage effluent discharge on the deposi- Hunt Creek, MontmorencyCounty, Michigan. tion of calcium carbonate on shells of the Ecology, 35:451-550. snail Potamopyrgusjenkensi (Smith). Nature (London), 186:492-493. 1958. Larvae and pupae of the species AND M. OWENS. 1960. The effects of of Cryptochironomus (Diptera) in Michigan. , Limnol. Oceanogr., 3:427-442. plants on river conditions. I. Summer crops and estimates of net productivity of macro- DAILY,F. K. 1958. Some observations on the phytes in a chalk stream. J. Ecol., 48:151-160. occurrence and distribution of the Characeae , AND . 1962. The effects of of Indiana. Proc. Ind. Acad. Sci., 68:95-107. plants on river conditions. IV. The oxygen DARNELL, R. M. 1961. Trophic spectrum of an balance of a chalk stream. J. Ecol., 50:297- estuarine community, based on studies of 320. Lake Pontchartrain, Louisiana. Ecology, 42: EINSELE, W. 1938. Uber chemische and kol- 553-568. loid-chemische Vorgange in Eisenphosphat- DEEVEY, E. S., JR., M. S. GRoss, G. E. HUTrcHIN- systemen unter limnochemischen und lim- SON, AND J. L. KRAYBILL. 1954. The nogeologischen Gesechtspunkten. Arch. Hy- natural C14 contents of materials from hard- drobiol., 33:361-387. water lakes. Proc. Nat. Acad. Sci., 40:285- ELLIS,M. M. 1936. Erosion silt as a factor in 288. aquatic environments. Ecology, 17:29-42. DENDY, J. S. 1944. The fate of animals in EMBODY,G. G. 1927. An outline of stream stream drift when carried into lakes. Ecol. study and the development of a stocking Monogr., 14:333-357. policy. Contr. Agri. Lab., Cornell Univ., 27pp. DENHAM, S. C. 1938. A limnological investi- EMIC, W. H. 1917. Travertine deposits of of gation the West Fork and Common Oklahoma. Okla. Geol. Surv., Bull., 29:1-76. Branch of White River. Invest. Ind. Lakes C. 1958. Streams, 1(5):17-71. EYSTER, Bioassay of water from a concretion-forming marl lake. Limnol. DODDS,G. S., AND F. L. HISAW. 1924a. Ecologi- Oceanogr., 2:455-458. cal studies of aquatic insects. I. Adaptations FAHY, W. E. 1954. The life history of the of mayfly nymphs to swift streams. Ecology, northern greenside darter, Etheostoma blen- 5:137-148. nioides blennioides Rafinesque. J. Elisha , AND . 1924b. Ecological studies Mitchell Sci. Soc., 70:139-205. of aquatic insects. II. Size of respiratory FERNALD,M. 0. 1950. Gray'smanual of botany. organs in relation to environmentalconditions. 8th ed. American Book Co., New York. lxiv Ibid.:262-271. + 1,632pp. , AND 1925a. Ecological studies FINNELL, J. C., R. M. JENKINS,AND G. E. HALL. on aquatic insects. III. Adaptations of 1956. The fishery resources of the Little caddisfly larvae to swift streams. Ibid., 6: River System, McCurtain County, Oklahoma. 123-137. Rept. Okla. Fish Res. Lab., 55:i-ii, 1-82. 118 WILDLIFEMONOGRAPHS

FORBES,S. A. 1925. The lake as a microcosm. GERSBACHER, W. M. 1937. Development of Ill. Nat. Hist. Surv. Bull., 15:537-550. stream bottom communities in Illinois. Ecol- (Emended reprint of Forbes, S. A. 1887. ogy, 18:359-390. The lake as a microcosm. Bull. Sci., Arts, GESSNER, F. 1955. Hydrobotanik. I. Energie- Peoria Coll., 1887:77-87.) haushalt. Veb Deutsch. Ver. der Wiss., Ber- , AND R. E. RICHARDSON. 1920. The lin. xii + 517pp. fishes of Illinois. (2nd ed.) Ill. Nat Hist. 1959. Hydrobotanik. II. Stoffhaushalt. Surv., Urbana. cxxxvi + 357pp., many pls. Ibid. xiv + 701pp. FOSTER,M. D. 1942. The chemistry of ground GILBERT,C. R. 1961. Hybridization versus inte- water. In Hydrology, physics of the earth. gradation (sic): an inquiry into the relation- IX. McGraw-Hill Book Co., New York. Pp. ships of two cyprinid fishes. Copeia, (1961): 646-655. 181-192. FOWLER, H. W. 1906. The fishes of New GNERI, F. S., AND V. ANGELESCU. 1951. La Jersey. Ann. Rept. N. J. Mus., 1905:35-447. nutricion de los peces iliofagos en relacion Fox, H. M., AND C. A. SIMMONDS. 1933. Meta- con el metabolismo general del ambient bolic rates of arthropods from different acuatico. Rev. Inst. Nac. Invest. Cien. Nat., habitats. J. Exp. Biol., 10:67-74. Mus. Argentina. Cien. Nat., 2:1-44. FRISON,T. H. 1935. The stoneflies, or Plecop- GRAF, D. L. 1960a. Geochemistry of carbonate tera, of Illinois. Bull. Ill. Nat. Hist. Surv., sediments and sedimentary carbonate rocks. 20:i-vi, 281-471. I. Carbonate minerology, carbonate sedi- 1942. Studies of North American ments. Ill. Geo. Surv., Circ., 297:1-39. Plecoptera, with special reference to the fauna . 1960b. Geochemistryof carbonate sedi- of Illinois. Ibid., 22:233-355. ments and sedimentary carbonate rocks. II. FRITSCH, F. E. 1929. The encrusting algal Sedimentarycarbonate rocks. Ibid., 298:1-43. community of certain fast-flowing streams. GRINNELL,J. 1922a. The role of the "acciden- New Phyto., 28:165-196. tal." Auk, 39:373-380. . 1949. The lime-encrusting Phormidium 1922b. The trend of avian populations community of British streams. Verh. int. ver. in California. Science, 56:671-676. theoret. angew. Limnol., 10:141-144. GUNNING, G. E. 1959. The sensory basis for . 1950. Phormidium incrustatum (Naeg.) homing in the longear sunfish, Lepomis mega- Gom., an important member of the lime- lotis megalotis (Rafinesque). Invest. Ind. encrusted communities of flowing water. Lakes Streams, 5:103-130. Biol. Jaarboek, 17:27-39. , AND W. M. LEWIS. 1956. Age and FROHNE,W. C. 1956. The provendering role growth of two important bait species in a of the larger aquatic plants. Ecology, 37: cold-water stream in southern Illinois. Am. 387-388. Midl. Nat., 55:118-120. FROST,W. E. 1942. River Liffey Survey. IV. HALLAM, J. C. 1959. Habitat and associated The fauna of the submerged "mosses" in an fauna of four species of fish in Ontario acid and an alkaline water. Proc. R. Irish streams. J. Fish. Res. Bd. Can., 16:147-173. Acad., B, 47:293-369. HARKER,J. E. 1953. An investigation of the FUNK,J. L. 1957. Movement of stream fishes mayfly fauna of the Lancashire stream. J. in Missouri. Trans. Am. Fish. Soc., 85:39-57. An. Ecol., 22:1-13. , AND R. S. CAMPBELL. 1953. The popu- HARRINGTON, R. W., JR. 1950. Preseasonal lation of larger fishes in Black River, Mis- breeding by the brindled shiner, Notropis souri. In The Black River studies. Univ. Mo. bifrenatus, induced under light-temperature Stud., 26:1-136. Pp. 69-82. control. Copeia, (1950) :304-311. GAUFIN, A. R., E. K. HARRIS, AND H. J. WALKER. . 1957. Sexual photoperiodicity of the 1955. A statistical evaluation of stream cyprinid fish, Notropis bifrenatus (Cope), in bottom sampling data obtained from three relation to the phases of its annual reproduc- standard samplers. Ecology, 37:643-648. tion cycle. J. Exp. Zool., 135:529-555. GERKING,S. D. 1945. Distribution of the fishes HARLAN, J. R., AND E. B. SPEAKER. 1956. Iowa of Indiana. Invest. Ind. Lakes Streams, 3: fish and fishing. Iowa Cons. Comm., Des 1-137. Moines. x + 377pp., many pls. . 1949. Characteristics of stream fish HASLER, A. D., AND W. J. WISBY. 1958. The populations. Ibid., 3:283-309. return of displaced largemouth bass and green 1950. Stability of a stream fish popula- sunfish to a "home" area. Ecology, 39:289- tion. J. Wildl. Mgmt., 14:193-202. 293. 1953. Evidence for the concepts of HASTINGS, A. B., C. D. MURRAY, AND J. SENDRAY. home range and territory in stream fishes. 1927. Studies of the solubility of calcium Ecology, 34:347-365. salts. I. The solubility of calcium carbonate 1959. The restricted movements of fish in salt solutions and biological fluids. J. Biol. populations. Biol. Rev., 34:221-242. Chem., 71:723-781. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 119

HEM, J. C. 1959. Study and interpretation of IDE, F. P. 1935. The effect of temperature on the chemical characteristics of natural water. the distribution of the mayfly fauna of a U. S. Geol. Surv., Water-supply Pap., 1473: stream. Univ. Toronto Stud., Biol. Ser., 39: i-ix, 1-269. 9-76, 10 pls. HENDERSON, C., AND R. F. FosTER. 1957. Studies IRWIN, W. H., AND J. H. STEVENSON. 1951. of smallmouth black bass (Micropterus dolo- Physicochemical nature of clay turbidity with mieui) in the Columbia River near Richland, special reference to clarification and pro- Washington. Trans. Am. Fish. Soc., 86:112- ductivity of impounded waters. Bull. Okla. 127. A. & M. Coll., 48:1-54. HERON, J. 1962. Determination of phosphate IVLEV,V. S. 1961. Experimental ecology of the in water after storage in polyethylene. Lim- feeding of fishes. Yale Univ. Press, New nol. Oceanogr., 7:316-321. Haven. viii + 302pp. (translated by Douglas HEYWOOD,J., AND R. W. EDWARDS. 1962. Some Scott). aspects of the ecology of Potamopyrgus jen- JAMES, M. T. 1959. Diptera. In Fresh-water kinsi Smith. J. An. Ecol., 31:239-250. biology. John Wiley & Sons, Inc., New York. HOAR, W. S. 1957. The gonads and reproduc- Pp. 1,057-1,079. tion. Chapt. VII. In The physiology of fishes. JILLSON,W. R. 1922. The conservation of nat- Academic Press Inc., New York. Pp. 287-321. ural gas in Kentucky. J. P. Morton Co., Louis- HORA, S. L. 1930. Ecology, bionomics, and ville. 152pp. evolution of the torrential fauna, with special JONES,J. R. E. 1950. A further ecological study reference to the organs of attachment. Phil. of the River Rheidol. The food of the com- Trans. R. Soc. Lond., 218:171-282. mon insects of the mainstream. J. An. Ecol., HORNUFF, L. 1957. A survey of four Oklahoma 19:159-174. streams with reference to production. Okla. 1951. An ecological study of the river Fish. Res. Lab., Rept., 63:1-22. Towy. Ibid., 20:68-86. HORNUNG, H. 1959. Floristisch-bkologischeUn- KAPLAN, M. F., AND W. L. MINCKLEY. 1960. tersuchungen an der Echaz unter besonderer Land snails from the Doe Run Creek area, Beriicksichtigung der Verunreinigung durch Meade County, Kentucky. Nautilus, 74:62- abwiisser. Arch. Hydrobiol., 55:52-126. 65. HOWLAND, J. H. 1931. Studies on the Kentucky KENDALL,J. L. 1941. Climate of Kentucky. In black bass (Micropterus pseudaplites Hubbs). Climate and man. U. S. Dept. Agri., House Trans. Am. Fish. Soc., 61:89-94. Doc. 27, 77th Congress, 1st Session. Pp. 884- HUBBS, C. L., AND K. F. LAGLER. 1958. Fishes 893. of the Great Lake Region (revised ed.). KRAATZ,W. C. 1923. A study of the food of Cranbrook Inst. Sci., Bull., 26:1-213, 44 pls. the minnow Campostoma anomalum. Ohio HUET, M. 1959. Profiles and biology of western J. Sci., 23:265-283. European streams as related to fish manage- KRECKER,F. H. 1939. A comparative study of ment. Trans. Am. Fish. Soc., 88:155-163. the animal population of certain submerged HUNGERFORD, H. B. 1920. The biology and aquatic plants. Ecology, 20:553-562. ecology of aquatic and semi-aquatic Hemip- KRUMHOLZ, L. A. 1948. The use of rotenone tera. Univ. Kans. Sci. Bull., 11:1-341. in fisheries research. J. Wildl. Mgmt., 12: HUTCHINSON,G. E. 1957. A treatise on limnol- 305-317. ogy. I. Geography, physics, and chemistry. ., J. R. CHARLES, AND W. L. MINCKLEY. John Wiley & Sons, Inc., New York. xiv + 1962. The fish population of the Ohio River. 1,015pp. In Aquatic-life resources of the Ohio River. . 1959. Homage to Santa Rosalia or Why Ohio River Valley Water Sanitation Commis- are there so many kinds of animals? Am. Nat., sion, Cincinnati. Pp. 49-89, 143-152, 166- 93:145-159. 180, 200-210. HYNES, H. B. N. 1941. The taxonomy and ecol- KUEHNE, R. A. 1958. Studies on the schooling of ogy the nymphs of the British Plecoptera, behavior of the minnows, Semotilus and with notes on the adults and Trans. R. eggs. Rhinichthys. Diss. Abstr., 19:606. Ent. Soc. Lond., 91:459-557. LANGLOIS, T. H. 1937. Bait culturists 1950. Food habits of freshwater stickle- guide. Ohio Div. 137:1- backs (Gasterosteus aculeatus and Dept. Agric., Cons., Bull., Pygosteus 23. pungitius), with a review of methods used in study of fishes. J. An. Ecol., 19:36-58. 1954. The western end of Lake Erie . 1960. The biology of polluted waters. and its ecology. J. W. Edwards, Inc., Ann Liverpool Univ. Press, Liverpool. xiv + Arbor. xx + 479pp. 202pp. LARIMORE,R. W. 1952. Home pools and hom- 1961. The invertebrate fauna of a ing behavior of smallmouth black bass in Welsh mountain stream. Arch. Hydrobiol., Jordan Creek. Ill. Nat. Hist. Surv., Biol. 57:344-388. Notes, 28:1-12. 120 WILDLIFEMONOGRAPHS

. 1954. Dispersal, growth, and influence . 1960. Ideas for a synthetic approach of smallmouth bass stocked in a warm-water to the ecology of running waters. Int. Rev. stream. J. Wildl. Mgmt., 18:207-216. ges. Hydrobiol., 45:133-153. . 1957. Ecological life history of the MARTIN, R. G., AND R. S. CAMPBELL. 1953. warmouth (Centrarchidae). Bull. Ill. Nat. The small fishes of Black River and Clear- Hist. Surv., 27:1-83. water Lake, Missouri. In The Black River . 1961. Fish population and electrofish- studies. Univ. Mo. Stud., 26:1-136. Pp. 45- ing success in a warm-water stream. J. Wildl. 66. Mgmt., 25:1-12. MATHESON,R. 1944. Handbook of the mos- , Q. H. PICKERING,AND L. DURHAM. 1952. quitos of North America. (2nd ed.) Comstock An inventory of the fishes of Jordan Creek, Publ. Co., Inc., Ithaca. viii + 314pp. Vermillion County, Illinois. Ill. Nat. Hist. MCMANUS, L. R. 1960. An occurrence of Surv., Biol. Notes, 29:1-26. "chimney" construction by the crayfish Cam- , W. F. CHILDERS, AND C. HECKROTTE. barus b. bartoni. Ecology, 41:383-384. 1959. Destruction and reestablishment of McFARLAN, A. C. 1943. Geology of Kentucky. stream fish and invertebrates affected by Univ. Ky. Press, Lexington. xviii + 531pp., drought. Trans. Am. Fish. Soc., 88:261-285. 117 pls. LEATIERS,A. L. 1922. Ecological study of MEINZER,0. E. 1923. Outline of ground-water aquatic midges and some related insects, with hydrology. U. S. Geol. Surv., Water-supply special referenec to feeding habits. Bull. U. Pap., 494:1-52. S. Bur. Fish., 38:1-61. METCALF, A. L. 1959. Fishes of Chautauqua, LEGER, L. 1945. Economic biologique et pro- Cowley, ald Elk counties, Kansas. Univ. ductive de nos riveres a Cyprinides. Bull. Kans. Publ., Mus. Nat. Hist., 11:345-400, 2 Franc. Piscicult., 139:49-69. pls. LENNON, R. E., AND P. S. PARKER. 1960. The MILLER, R. R. 1960. Systematics and biology stoneroller, Campostoma anomalum (Rafines- of the gizzard shad (Dorosoma cepedianum) que), in Great Smoky Mountains National and related species. U. S. Fish Wildl. Serv., Park. Trans. Am. Fish. Soc., 89:263-270. Fish. Bull., 173:i-iv, 371-392. LEONARD,J. W. 1939. Comments on the ade- MINCKLEY, W. L. 1961. Occurrence of subter- quacy of accepted stream bottom sampling ranean isopods in the epigean environment. technique. Trans. 4th N. Am. Wildl. Conf.: Am. Midl. Nat., 63:452-455. 288-295. . 1962. Studies of the ecology of a spring , AND F. A. LEONARD. 1962. Mayflies stream: Doe Run, Meade County, Kentucky. of Michigan trout streams. Cranbrook Inst. Unpubl. Ph.D. dissertation, University of Sci., Bull., 43:i-x, 1-139. Louisville. 374pp. LEOPOLD,L. B. 1962. Rivers. Am. Sci., 50:511- . 1963. Upstream movements of Gam- 537. marus (Amphipoda) in Doe Run, Meade LEVERErr, F. 1929. The Pleistocene of northern County, Kentucky. Ecology. In press. Kentucky. Ky. Geol. Surv., Ser. IV, 31:i-xiv, , AND J. E. CRADDOCK. 1961. Active pre- 1-80. dation of crayfish on fishes. Prog. Fish-Cult., LINDEMAN, R. L. 1942. The trophic-dynamic 23:120-123. aspect of ecology. Ecology, 23:339-418. , J. E. CRADDOCK, AND L. A. KRUMHOLZ. LUDWIG,W. B. 1931. A limnological survey of 1963. Natural radioactivity in the food web the Hocking River. B. The bottom inverte- of the banded sculpin, Cottus carolinae (Gill). brates of the Hocking. Ohio J. Sci., 31:267- In Radioecology. Reinhold Publ. Co., New 268 (Abstr.). York. Pp. 229-236. LYMAN,F. E. 1943. A pre-impoundment bot- , AND F. B. CROSS. 1959. Distribution, tom fauna study of Watts Bar Reservoir area habitat, and abundance of the Topeka shiner, (Tennessee). Trans. Am. Fish. Soc., 72:52-62. Notropis topeka (Gilbert), in Kansas. Am. MACAN,T. T. 1957. The life histories and mi- Midl. Nat., 61:210-217. grations of the Ephemeroptera in a stony , AND . 1960. Taxonomic status of stream. Trans. Soc. Brit. Ent., 12:120-156. the shorthead redhorse, Moxostoma aureolum . 1958. Methods of sampling the bottom (LeSueur), from the Kansas River basin, in stony streams. Mitt. int. ver. Limnol., 8: Kansas. Trans. Kans. Acad. Sci., 63:35-39. 1-21. , AND D. R. TINDALL. 1963. Observa- MAJOR,J. 1961. Use in plant ecology of causa- tions on the ecology of Batrachospermumsp. tion, physiology, and a definition of vegeta- (Rhodophyceae) in Doe Run, Meade County, tion. Ecology, 42:167-169. Kentucky. Bull. Torrey Bot. Club. In press. MARGALEF, R. 1958. "Trophic" typology versus MOFFETT, J. W. 1936. A quantitative study of biotic typology, as exemplified in the regional the bottom fauna in some Utah streams var- limnology of northern Spain. Verh. int. ver. iously affected by erosion. Bull. Univ. Utah Limnol., 13:339-349. Biol. Ser., 3:1-33. ECOLOGYOF DOE RUN, MEADECOUNTY, KENTUCKY-Minckley 121

MOTTLEY, C. McC., H. J. RAYNER, AND J. H. . 1957b. Primary production measure- RAINWATER. 1939. The determination of ments in eleven Florida springs and a marine the food grade of streams. Trans. Am. Fish. turtle grass community. Limnol. Oceanogr., Soc., 68:336-343. 2:85-97. MUELLER, K. 1954. Investigations on the or- OHLE, W. 1934. Chemische and physikalische ganic drift in north Swedish streams. Inst. Untersuchungen nordeutscher seen. Arch. Freshwater Res., Drottningholm, Rept., 35 Hydrobiol., 26:386-464. (1953):133-148. . 1952. Die hypolimnische Kohlendi- 1955. Produktionsbiologische Unter- oxyde-Akkumulation als producktionsbiolo- suchungen in Nordschwedischen Fliessege- gischer Indikator. Ibid., 46:153-285. wassem. Ibid., 36(1954):148-162. OOSTING,H. J. 1953. The study of plant communities. W. H. Freeman & San W. C. 1944. of Co., MUENSCHER, Aquatic plants the Francisco. United States. Comstock Publ. Co., Ithaca. 389pp. OWENS, M., AND R. W. EDWARDS. 1961. The x + 374pp. effects of plants on river conditions II. Further MURPHY,H. E. 1922. Notes on the biology of crop studies and estimates of net productivity our North American of species mayflies. in a chalk stream. J. Ecol., 49:119-126. Lloyd Library (Cincinnati, 0.) Ent. Ser., A. A. 1958. Measurement of some 2:1-46. PALOUMPIS, factors affecting the catch in a minnow seine. MuTTowsSKI,R. A. 1929. The ecology of trout Proc. Iowa Acad. Sci., 65:580-586. streams in Yellowstone National Park. Bull. PEARSALL, W. H. 1920. The aquatic vegetation New York State Coll. For., Roosevelt Wild of the English lakes. J. Ecol., 8:163-199. Life Ann., 2:155-240. PEARSE,A. S. 1926. Animal ecology. McGraw- , ANDG. M. SMITH. 1929. The food of Hill Book Co., Inc., New York. ix + 417pp. trout stream insects in Yellowstone National PENNAK,R. W. 1943. Limnological variables in Park. Ibid.:241-263. a Colorado mountain stream. Am. Midi. Nat., NEEDHAM, P. R., AND R. L. USINGER. 1956. 29:189-199. Variability in the macrofauna of a single riffle 1953. Fresh-water invertebrates of the in Prosser Creek, California, as indicated by United States. Ronald Press, New York. ix the Surber sampler. Hilgardia, 24:383-409. + 769pp. NEEL, J. K. 1951. Interrelations of certain PERCIVAL, E., AND H. WmTEHEAD. 1929. A physical and chemical features in a head- quantitative study of the fauna of some types water limestone stream. Ecology, 32:368- of stream bed. J. Ecol., 17:382-414. 391. PRESCOTT, G. W. 1951. Algae of the western Great Lakes Cranbrook Inst. NELSON,D. J., AND D. C. SCOTT. 1962. Role of region. Sci., detritus in the productivity of a rock-outcrop Bull., 31:1-924. community in a piedmont stream. Limnol. RANEY, E. C., AND E. A. LACHNER. 1946. Age, Oceanogr., 7:396-413. growth and habits of the hog sucker, Hypen- telium in New York. NETTING,M. G. 1956. Geography of the upper nigricans (LeSueur), Ohio Valley. In Man and the waters of the Am. Midl. Nat., 36:76-86. Upper Ohio basin. Spec. Publ. 1, Pymatuning RAWSON, D. S. 1930. The bottom fauna of Lab. Field Biol. Pp. 2-13. Lake Simcoe and its role in the ecology of NEWELL,A. E. 1957. Two-year study of move- the lake. Univ. Toronto Stud., Biol. Ser., 40: ments of stocked brook trout and rainbow 1-183. trout in a mountain stream. Prog. Fish-Cult., REID, G. K. 1961. Ecology of inland waters and 19:76-80. estuaries. Reinhold Publ. Co., New York. NIELSEN,A. 1950a. The torrential invertebrate xvi + 375pp. fauna. Oikos, 2:176-196. RICKER,W. E. 1934a. An ecological classification of certain Ontario streams. Univ. of 1950b. On the of zoogeography springs. Toronto Stud., Biol. Ser., 37:1-114. Hydrobiol., 2:313-321. . 1934b. A critical discussion of various NIELSEN,C. 0. 1962. Carbohydrasesin soil and measures of oxygen saturation in lakes. Ecol- litter invertebrates. Oikos, 13:200-215. ogy, 15:348-363. ODUM,E. P. 1959. Fundamentals of ecology. . 1959. Plecoptera. In Fresh-water biol- W. B. Saunders Co., Philadelphia. xvii + ogy. John Wiley & Sons, Inc., New York. Pp. 546pp. 941-957. ODUM,H. T. 1956. Primary production in flow- RIDENOUR, G. L. 1929. Early times in Meade ing waters. Limnol. Oceanogr., 1:102-117. County, Kentucky. Western Recorder Co., . 1957a. Trophic structure and produc- Louisville. 108pp. tivity of Silver Springs, Florida. Ecol. Monogr., RIGGS,C. D. 1947. Purple martins feeding on 27:55-112. emerging may-flies. Wilson Bull., 59:113-114. 122 WILDLIFEMONOGRAPHS

ROELOFS,E. W. 1944. Water soils in relation to , AND S. EDDY. 1929. Methods for the lake productivity. Mich. St. Coll., Agri. Exp. study of stream communities. Ecology, 10: Sta. Tech. Bull., 190:1-31. 382-391. ROLL,H. 1938. Die Pflanzengesellschaften 6s- SHOUP, C. S. 1948. Limnological observations tholsteinischer Fliessagewiisser. Arch. Hydro- on some streams of the New River watershed biol., vol. 34 (cited in Blum, 1956a). in the vicinity of Mountain Lake, Virginia. . 1939a. Die Entwicklung der Potamo- J. Elisha Mitchell Sci. Soc., 64:1-12. botanik. Int. Rev. ges. Hydrobiol., vol. 39 1950. Field chemical examination of (cited in Blum, 1956a). the waters in Tennessee streams. J. Tenn. . 1939b. Einige Waldequellen Holsteins Acad. Sci., 25:4-55. and ihre Pflanzengesellschaften. Bot. Jahr- SLACK, K. V. 1955. A study of the factors bucher, vol. 70 (cited in Blum, 1956a). affecting stream productivity by the compara- Roos, R. 1957. Studies on the upstream migra- tive method. Invest. Ind. Lakes Streams, 4: tion in adult stream-dwelling insects. I. Inst. 3-47. Freshwater Res., Drottningholm, Rept., 38 SLOAN,W. C. 1956. The distribution of aquatic (1956):166-193. insects in two Florida springs. Ecology, 37: ROSINE,W. N. 1955. The distribution of inver- 81-98. tebrates on submerged aquatic surfaces in SPRULES,W. M. 1947. An ecological investiga- Muskee Lake, Colorado. Ecology, 36:308- tion of stream insects in Algonquin Park, 314. Ontario. Univ. Toronto Stud., Biol. Ser., Ross, H. H. 1953. The caddisflies, or Trichop- 56:1-81. tera, of Illinois. Bull. Ill. Nat. Hist. Surv., STARRETT, W. C. 1950a. Distribution of the 23:1-326. fishes of Boone County, Iowa, with special . 1959. Trichoptera. In Fresh-water biol- reference to the minnows and darters. Am. ogy. John Wiley & Sons, Inc., New York. Midl. Nat., 43:112-127. Pp. 1,024-1,049. . 1950b. Food relationships of the min- RUTTNER,F. 1948. Die Veriinderungern des nows of the Des Moines River, Iowa. Ecol- Aquivalentleitvermogens als Mass der Kar- ogy, 31:216-233. bonatassimilation der Wasserpflantzen. 1951. Some factors affecting the abun- Schweiz. Z. Hydrol., 11:72-89 (cited from dance of minnows in the Des Moines River, Hutchinson, 1957). Iowa. Ibid., 32:13-27. 1953. Fundamentals of Limnology. , AND P. G. BARNICKOL. 1955. Efficiency Univ. Toronto Press, Toronto. xvi + 242pp. and selectivity of commercial fishing devices (translated by D. G. Frey and F. E. J. Fry). used on the Mississippi River. Bull. Il1. Nat. . 1960. Uber die Kohlenstoffaufnahme Hist. Surv., 26:325-366. bei Algen aus der Rhodophyceen-Gatung STEIDTMANN, E. 1935. Travertine-depositingwa- Batrachospermum. Schweizer. Z. Hydrol., ters near Lexington, Virginia. Science, 82: 22:280-291. 333-334. C. 0. 1927. SAUER, Geography of the Penny- STEWART, N. H. 1926. Development, growth, royal. Ky. Geol. Surv., Ser. VI, 5:i-xii, 1- and food habits of the white Catosto- 249. sucker, mus commersonii LeSueur. Bull. U. S. Bur. ScHMITZ,W. 1955. Physiographische Aspeckte Fish., 42:147-183. de limnologischen fliessgewdssertypen. Arch. Hydrobiol., 22:310-523. STUMM, W., AND G. F. LEE. 1960. The chem- Suppl., of SCHNELLER, M. V. 1955. Oxygen depletion in istry aqueous iron. Rev. Suisse d'Hydrol., Salt Creek, Indiana. Invest. Ind. Lakes 22:295-319. Streams, 4:163-176. SURBER, E. W. 1937. Rainbow trout and bot- ScoTT, D. C. 1949. A study of a stream popu- tom fauna production in one mile of stream. lation of rock bass. Ibid., 3:169-234. Trans. Am. Fish. Soc., 66:193-202. SELLECK, G. W. 1960. The climax concept. Bot. . 1939. A comparison of four eastern Rev., 26:534-545. smallmouth bass streams. Ibid., 68:322-335. SHELFORD, V. E. 1911a. Ecological succession. . 1951. Bottom fauna and temperature I. Stream fishes and the method of physio- conditions in relation to trout management graphic analysis. Biol. Bull., 21:9-34. in St. Mary River, Augusta County, Virginia. . 1911b. Ecological succession. II. Pond Va. J. Sci., 2(n.s.):190-202. fishes. Ibid., 22:1-38. SWINDALE, D. N., AND J. T. CURTIS. 1957. 1912. Ecological succession. IV. Vege- Phytosociology of the larger aquatic plants in tation and the control of land animal com- Wisconsin lakes. Ecology, 38:397-407. munities. Ibid., 23:59-99. SWINGLE,H. S., AND E. V. SMrTH. 1950. Fac- . 1913. Animal communities in tem- tors affecting the reproduction of bluegill perate America. Univ. Chicago Press, Chicago. bream and largemouth black bass in ponds. xiii + 362pp. Agri. Exp. Sta., Ala. Poly. Inst., Circ., 87:1-8. ECOLOGYOF DOE RUN, MEADE COUNTY, KENTUCKY-Minckley 123

SWINNERTON,A. C. 1942. Hydrology of lime- 1942. Fish distribution and abundance stone terranes. In Hydrology, physics of the correlated with stream gradients as a consid- earth. IX. McGraw-Hill Book Co., New eration in stocking programs. Trans. 7th N. York. Pp. 656-677. Am. Wildl. Conf., 211-233. SYMOENS, J. J. 1951. Equisse d'un systema des . 1957. The fishes of Ohio. Ohio State associations algales d'eau douce. Verh. int. Univ. Press, Columbus. xvii + 683pp., 7 pls. ver. theoret. angew. Limn., 11:395-408. , AND R. G. MARTIN. 1951. Moxostoma 1957. Les eaux doces de l'Ardenne et aureolum pisolabrum, a new subspecies of des rigions voisines: Les milieus et leur vege- sucker from the Ozarkian streams of the tation algale. Bull. R. Soc. Belg., 89:111-314. Mississippi River system. Occ. Pap. Mus. TAYLOR,W. R. 1954. Records of fishes in the Zool., Univ. Mich., 534:1-10, 1 pl. John N. Lowe collection from the Upper TRUESDALE,G. A., A. L. DOWNING, AND G. F. Peninsula of Michigan. Misc. Publ. Mus. LOWDEN.1955. The solubility of oxygen in Zool., Univ. Mich., 87:1-50. pure water and sea-water. J. Appl. Chem. TEAL, J. M. 1957. Community metabolism in a Lond., 5:53-63. temperate cold spring. Ecol. Monogr., 27: TWENHOFEL,W. H. 1939. Principles of sedi- 283-302. mentation. McGraw-Hill Book Co., New TEBO, L. B., AND W. W. HASSLER. 1961. Sea- York. x + 610pp. sonal abundance of aquatic insects in western USINGER, R. L. (ED.) 1956. Aquatic insects North Carolina trout streams. J. Elisha of California with keys to North American Mitchell Sci. Soc., 77:249-259. genera and California species. Univ. Calif. THIENEMANN, A. 1953. Fluss und See, ein Press, Berkeley and Los Angeles. ix + 508pp. limnologischer Vergleich. Gewiisser und Ab- WALKER, B. A. 1961. Studies on Doe Run, wasser, 1:13-30. Meade County, Kentucky. IV. A new species THOMAS,H. A., JR. 1958. Mixing and diffusion of isopod crustacean (genus Asellus) from of wastes in streams. In Oxygen relationships Kentucky. Trans. Am. Microscop. Soc., 80: in streams. Robt. A. Taft San. Eng. Cent., 385-390. Tech. Rept. W58-2:97-103. . 1962. The ecology of the epigean THOMPSON,D. H., AND F. D. HUNT. 1930. The isopod Asellus bivittatus in Doe Run, Meade fishes of a Champaign County, study of the County, Kentucky. Unpubl. Master's Thesis, distribution and abundance of fishes in small University of Louisville. streams. Bull. Ill. Nat. Hist. 19:1-101. 32pp. Surv., WALKER, E. H. 1957. The deep channel and 1-101. alluvial deposits of the Ohio Valley in Ken- THOMPSON,J. W. 1944. A survey of the larger tucky. U. S. Geol. Surv., Pap., and bank Water-supply aquatic plants flora of the Brule 1411:i-iii, 1-25, 6 pls. River. Wis. Acad. and Sci., Arts, Lett., WATERS, T. F. 1961. Standing crop and drift 36:57-76. of stream bottom organisms. Ecology, 42: THOMPSON,R. H. 1959. Algae. In Fresh-water 532-537. biology. John Wiley & Sons, Inc., New York. 115-170. . 1962. A method to estimate the pro- Pp. duction TIFFANY,L. H. 1930. The a rate of a stream bottom invertebrate. Oedogoniaceae: Trans. Am. Fish. 91:243-250. monograph. Publ. by the author, and Soc., Spahr WELCH,P. S. 1935. McGraw-Hill Glenn Co., Columbus. 326pp. Limnology. Book Co., New York. xiv + 471pp. , AND M. E. BRITTON. 1952. The algae . of Illinois. Univ. Chicago Press, 1948. Limnological methods. Blakis- Chicago. ton xviii xiv + 407pp. Co., Philadelphia. + 381pp. WENE, AND E. L. WICKLIFF. 1940. Modi- TINDALL, D. R. 1962. Life history and ecology G., of Nitella fication of a stream bottom and its effect on flexilis (Characeae). Unpubl. the insect Master's Thesis. University of Louisville. fauna. Can. Ent., 72:131-135. WETZEL, R. G. 1960. Marl encrustations on 77pp. in E. N. 1913. The hydrophytes several Michigan lakes. Oikos, TRANSEAU, periodicity of algae 11:223-236. in Illinois. Trans. Am. Microscop. Soc., 32: 31-40. WHERRY, E. T. 1920. Soil acidity and a field method for its measurement. Ecology, 1: TRAUTMAN, M. B. 1931. Notropis volucellus 160-173. wickliffi, a new subspecies of fish cyprinid WHITEHEAD, H. 1935. An ecological study of from the Ohio and Upper Mississippi rivers. the invertebrate fauna of the chalk stream Ohio 13:468-474. J. Sci., near Great Driffield, Yorkshire. J. An. Ecol., 1939. The effects of man-made modi- 4:58-78. fications on the fish fauna in Lord and Gor- WHITFORD, L. A. 1956. The communities of don Creeks, Ohio, between 1887-1938. Ibid., algae in the springs and spring streams of 39:275-288. Florida. Ecology, 37:433-442. 124 WILDLIFEMONOGRAPHS

. 1960a. Ecological distribution of fresh- WINN, H. E. 1958a. Comparative reproductive water algae. In The ecology of algae. Spec. behavior and ecology of fourteen species of Publ. 2, Pymatuning Lab. Field Biol. Pp. darters (Pisccs-Percidae). EcoL Monogr., 2-10. 28:155-191. . 1958b. Observations on the . 1960b. The reproduc- current effect and growth tive habits of darters Am. of fresh-water Am. (Pisces-Percidae). algae. Trans. Microscop. Midl. Nat., 59:190-223. Soc., 79:302-309. WIRTH, W. W., AND A. STONE. 1956. Aquatic , AND G. J. SCHUMACHER. 1961. Effect Diptera. In Aquatic insects of California, of current on mineral uptake and respiration with keys to North American genera and by a fresh-water algae. Limnol. Oceanogr., California species. Univ. Calif. Press, Berke- 6:423-425. ley and Los Angeles. Pp. 372-482. WITT, A., AND R. C. MARZOLF. 1954. WHITTAKER, R. H. 1953. A consideration of JR., Spawning and behavior of the longear sun- climax theory: the climax as a population and fish, Lepomis megalotis megalotis. Copeia, pattern. Ecol. Monogr., 23:41-78. 1954:188-190. WiLSON, L. R. 1935. Lake development and WOOD, R. D., AND W. C. MUENSCHER. 1956. plant succession in southern Vilas County, The Characeae of New York. Cornell Univ., Wisconsin. Ecol. Monogr., 5:207-247. Agri. Exp. Sta. Mem., 338:1-77. Wu, Y. F. 1931. A contribution to the . 1937. A biology quantitative and ecological of Simulium (Diptera). Pap. Mich. Acad. study of the larger aquatic plants of Sweeny Sci., Arts, and Lett., 13:543-599. Lake, Oneida County, Wisconsin. Bull. Tor- YOUNG,0. W. 1945. A limnological investiga- rey Bot. Club, 64:199-208. tion of periphyton in Douglas Lake, Michigan. . 1939. Rooted aquatic plants and their Trans. Am. Microscop. Soc., 64:1-20. relation to the limnology of freshwater lakes. ZOBELL, C. E., AND C. B. FELTHAM. 1938. In Am. Assn. Adv. Aci., Publ., 10:101-122. Bacteria as food for certain marine animals. J. Mar. Res., 4:312-327. . 1941. The larger aquatic vegetation ,AND . 1942. The bacteria of of Trout Lake, Vilas County, Wisconsin. Wis. flora of a marine mud flat as an ecological Acad. Sci., Arts, and Lett., 33:135-146. factor. Ecology, 23:69-78.