INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated w ith a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 74-3266 MOERCHEN, S.M., Richard Frank, 1934- SPECIFIC ISOLATING MECHANISMS OF THE ORANGE- THROAT HARTER, SPECTABILE (AGASSIZ) , AND THE RAINBOW HARTER, ElHEOSraMA CAERULEUM STORER, IN CENTRAL OHIO, WITH CONSIDERATIONS OF THEIR HYBRIDIZATION.

The Ohio State University, Ph.D., 1973 Zoology

University Microfilms, A XEROX Company , Ann Arbor, Michigan

© 1973

Richard Frank Moerchen, S.M.

ALL RIGHTS RESERVED SPECIFIC ISOLATING MECHANISMS OF THE ORANGETHROAT DARTER,

ETHEOSTOMA SPECTABILS (AGASSIZ), AND THE RAINBOW DARTER,

ETHEOSTOMA CAERVLEUli STORER, IN CENTRAL OHIO, WITH

CONSIDERATIONS OF THEIR HYBRIDIZATION

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Richard Moerchen, S. M. , B. S., M. S.

* * * * *

The Ohio State University

1973

Reading Committee Approved By

Dr. Ted M. Cavender

Dr. Milton B. Trautman

Dr. Richard A. Tubb Adviser Department of Zoology ACKNOWLEDGMENTS

The present work was made possible with much assistance and

considerable encouragement on the part of many individuals. I wish to

first thank Dr. Ted M. Cavender for his advice, suggestion and con­

structive criticisms throughout the investigation. X am indebted to

Dr. Milton B. Trautman for originally suggesting the ecological aspects

of the problem, for his continued interest and for being so generous

and gracious with his time and advice. Thanks are also extended to

Dr. Richard A. Tubb who, with Drs. Cavender and Trautman, served on

the reading committee and was especially helpful in criticizing the writing style, and to Drs. Walter C. Rothenbuhler and Clarence E. Taft, who served on the examining committee. To all these Individuals I am

grateful for their assistance in helping me to more properly express

the findings and implications of the research.

I express profound gratitude to Dr. Barbara K. Dommert for an

Immense amount of clerical work, for assistance in the field and for her continual interest and encouragement. Without her generosity the

finalizing of the study would have been much longer in coming.

Mr. Thomas Whitney deserves my appreciation for his advice and patient assistance on computer work and statistical analysis.

X wish to especially acknowledge the confident and continual support of my superiors and colleagues of the Society of Mary

ii (Marianists, St. Louis Province) in terms of complete financial support and moral encouragement. For financial aid I also acknowledge a partial Mary Osburn Summer (1972) Fellowship from The Ohio State

University.

Thanks are due to Dr. David Stansbery, director of the Museum of

Zoology of The Ohio State University for furnishing space to work and for patient considerations in my behalf, to Mr. John Condit for his generosity in sharing aquarium and photography equipment, and to

Dr. Tony Peterle whose interest and patience were a source of encouragement.

Ify thanks are extended to Dr. Tony Echelle of the University of

Oklahoma for his correspondence and the use of the manuscript on j£. spectabile — JE. radiosum hybridization.

To my roommates Bernard Ploeger, S.M. and Dr. Frank Damm, S. M.

X am especially thankful. Their tremendous patience, forbearance, and charity made living and working with them a sincere privilege and real pleasure.

I am grateful to my parents for their confident anticipation of the completion of this study. To many close friends who, during the course of this study asked "How are your fish?", with sincere, empa- thetlc, and encouraging Interest, X express equally sincere gratitude and appreciation.

Final, though no less sincere thanks, are extended to Ms. Terri

Goga for her patience throughout the typing of the copy.

iii VITA

May 21,1934 ...... Born, Belleville, Illinois

1952 ...... Entered Religious Order, Society of Mary (Marianists), St. Louis Province

1955 ...... B. S., Chemistry St. Mary's University, San Antonio, Texas

1955-1956 ...... Chaminade College Preparatory Clayton, Missouri

1956-1957 ...... Don Bosco High School, Milwaukee, Wisconsin

1957-1961 ...... Chaminade College Preparatory Clayton, Missouri

1961-1964 .... Vianney High School, Kirkwood Missouri

1965 ...... M. S., Biology University of Notre Dame Notre Dame, Indiana

1965-1968 ...... McBride High School St. Louis, Missouri

1967 Summer ...... Instructor St. Mary's University San Antonio, Texas

1973 ...... Ph. D., The Ohio State University, Columbus, Ohio

FIELDS OF STUDY

Major Field: Zoology

Studies in Ichthyology. Associate Professor Ted M. Cavender

Studies in Fish Behavior. iv TABLE OF CONTENTS

Page ACKNOWLEDGMENTS ...... ±i

VITA ...... iv

LIST OF MAPS ...... vii

LIST OF TABLES ...... viii

LIST OF FIGURES ...... x

INTRODUCTION ...... 1

METHODS AND MATERIALS ...... 5

RESULTS ...... 31

E c o l o g y ...... 31

R a n g e ...... 31

Habitat: ...... 37

Reproductive behavior ...... 54

Location of spawning s i t e s ...... 54

Sexual dimorphism ...... 56

Spawning season ...... 58

Spawning habitat ...... 60

Pre-spawning activity ...... 69

Spawning behavior ...... 75

Hybridization...... 89

Meristic and morphometric analysis ...... 92

Chromosome morphology ...... 116

v DISCUSSION...... 119

SUMMARY ...... 126

APPENDIX ...... 128

t A Summary of meristic and morphometric d a t a ...... 128

B OSUM cataloging of specimens...... 141

LITERATURE CITED ...... 144

Vi LIST OF MAPS

Map Page

I. Detail of study stream a r e a ...... 8

II. Distribution of Etheostoma caeruleum and

Etheostoma spectabile in the United States ...... 32

III. Collection records of 15. caeruleum in O h i o ...... 35

IV. Collection records of E. spectabile in O h i o ...... 36

vli LIST OF TABLES

Table Page

1. Species list of Dry Run ...... 10

2. Species list of Peter's Run ...... 15

3. Habitat preference of 12. spectabile and E. caeruleum . . 43

4. Ratio of E. caeruleum and E. spectabile

taken at specific stream a r e a s ...... 51

5. Comparison of E^. caeruleum and jE. spectabile

pre-spawning behavior ...... 80

6. Similarities and differences related to

E. caeruleum and E.. spectabile spawning behavior . . . 87

7. Variances of characters of laboratory-raised hybrids

versus parental stock ...... 91

8. Characters of laboratory-raised hybrids

intermediate between parental stdck ...... 102

9. Characters of natural hybrids resembling

IS. caeruleum parents ...... 103

10. Characters of natural hybrids resembling

E. spectabile parents...... 104

11. Significant differences between E. caeruleum and

E. spectabile males ...... 107

12. Significant differences between JS. caeruleum and

E. spectabile females...... 108 viii Table Page

13. Significant differences between JS. caeruleum

males and females...... 109

14. Significant differences between IS. spectabile

males and females ...... 110

15. Significant differences between JS. caeruleum of

Scioto River drainage and Miami River drainage.... Ill

16. Significant differences between JS. caeruleum and

JS. spectabile from in study stream, Dry Run ...... 113

17. Significant differences between laboratory-raised

hybrids and their IS. caeruleum p a r e n t s ...... 114

18. Significant differences between laboratory-raised

hybrids and their JS. spectabile parents ...... 115

19. Meristic and morphometric data JS. caeruleum.

E. spectabile and their natural hybrids ...... 129

20. Meristic and morphometric data for IS. caeruleum,

JS. spectabile and their laboratory-raised hybrids . . . 135

21. OSUM cataloging of s p e c i m e n s ...... 142

ix LIST OF FIGURES

Figure Fage

1. Photograph of primary study site, Dry Run, Station IV * 12

2. Photograph of Peter's Run, Ohio Rte. 1 0 4 ...... 13

3. Photograph of glass bottom observation b o x ...... 17

4. Diagram of tanks and flow system used in simulated

natural environment ...... 19

5. Photograph of tank used for E. caeruleum observation . . 21

6. Photograph of infraorbital canal, E. caeruleum ...... 27

7. Photograph of infraorbital canal, IS. spectabile .... 27

8. Stream order classification of northwest Scioto

River drainage...... 40

9. Detail of study stream a r e a ...... 42

10. Diagram of relative abundance in terms of

rate of flow and stream o r d e r ...... 44

11. Relative abundance of _E. caeruleum and JE. spectabile

at certain stream sites of the Scioto River drainage . 45

12. Spawning areas of IS. caeruleum and E. spectabile

at Dry Run, Station IV ...... 55

13. A direct overhead view of riffle at Station IV, Dry Run 55

14. Photograph of IS. spectabile m a l e ...... 57

15. Photograph of E. caeruleum male ...... 57

x Figure Page

16. Photograph of tubercles on anal fin of

E. caeruleum m a l e ...... 59

17. Photograph of substrate at JE. caeruleum spawning area . 63

18. Diagram of stream and location of spawning sites of

JE. caeruleum and JE. spectabile at Station XV, Dry Run 64

19. Diagram of stream and location of spawning sites of

13. spectabile at Peter's Run, Ohio Rte. 1 0 4 ...... 65

20. Photograph of substrate at E. spectabile spawning area . 67

21. Photograph of IE. spectabile male showing

usual eye pigmentation ...... 74

22. Photograph of E_. spectabile male showing pigmentation

of eye during aggressive behavior ..... 74

23. Photograph of E. spectabile female showing usual eye

appearance...... 77

24. Photograph of E. spectabile female showing "eyespot"

pigmentation during spawning ...... 77

25. Photograph of 12. caeruleum female showing darkened

saddle bands during spawning ...... 78

26. Photograph of E. caeruleum in spawning aggregate .... 83

27. Ideal sequence of pre-spawning and spawning events . . . 88

28. Discriminant analysis of laboratory-raised hybrids

and their parental stock ...... 94

29. Discriminant analysis of E. caeruleum, E. spectabile.

and putative hybrids from study stream ...... 97

xi Figure Page

30. Simultaneous discriminant analysis of laboratory-raised

hybrids, putative natural hybrids, and parental stock 98

31. Photograph of 12. caeruleum female showing "stair-step”

pattern of pigmentation ...... 101

32. Photograph of natural hybrid male showing

disruptive pattern of pigmentation ...... 101

33. Laboratory-raised hybrid male showing

disruptive pattern of pigmentation ...... 101

34. Karyotypes of males 12. spectabile, 12. caeruleum,

and E. caeruleum x 12. spectabile h y b r i d ...... 117

xii INTRODUCTION

The rainbow darter, Etheostoma caeruleum Storer, and the orange-

throat darter, Etheostoma spectabile (Agassiz), are small, riffle-

dwelling fish of the family Fercidae, subfamily . Both

are members of the sub-genus Oligocephalus» a group of rather colorful

species often typified by their common names. In Ohio, 12. caeruleum

is especially abundant along the glacial boundary, the Allegheny Front

Escarpment and in the two Miami River systems (Trautman, 1957). It is

one of the most common species found in small to average size streams

of moderate to moderately high gradient. Etheostoma spectabile also

occurs in the glaciated westcentral Ohio streams, often sympatrically with 12. caeruleum, especially in the smaller streams of low to moderate

gradient.

The purpose of this study was to partially satisfy the need for

research into the determination of isolating mechanisms of closely

related, sympatric species from both a systematic and behavioral stand­

point; to attempt to clarify the means and mechanisms whereby E.

spectabile and E. caeruleum maintain their integrity as distinct

species, though they may seem, by appearance and behavior, to be quite

similar. To this end careful scrutiny was made of the similarities and

the subtlest differences in habitat occupancy, degree, and type(s) of

interspecific interactions, , chromosome morphology, and

especially reproductive behavior. 1 Systematic studies of IS. spectabile and E; caeruleum have been made by Dlstler (1968) and Knapp (1964) respectively, each giving the

synonymy and describing the specific range and general habitat. More

detailed local distribution patterns are described by Blair (1959),

Cross (1967), Gerking (1945, 1955), Pflieger (1971), Smith-Vanlz

(1967), and Trautman (1957). More qualitative descriptions of E. spectabile habitat are given by Larimore, Pickering, and Durham (1952) and Braasch and Smith (1967); of both species by Trautman (1957) and especially Winn (1958a,b). Reeves (1907) confused the two species in her study of E. caeruleum breeding behavior. Trautman C1930) was the first to publish a discernment of their specific distinctness.

Numerous workers have described the life history and/or repro­ ductive behavior of darter species, though not near to the extent of game and/or commercial species. The value of the study of smaller, less

"economically important" species has been well illustrated, for it was through the work on gradient in respect to darter habitats, that led to more enlightened stocking programs of smallmouth bass in Ohio and other states (Trautman, 1942). Stream gradient investigations have resulted in explaining sharp distributional patterns of 170 species of fish and enabled a more accurate evaluation of localities to be stocked with fishes or areas to be set aside for natural propagation purposes.

Most notable of the darter habitat and reproductive investigations is that of Winn's (1958a) comparison of fourteen species. Linder's

(1958) laboratory studies described the reproductive behavior and hy­ bridization of E. spectabile and JS. radiosum cyanorum, the orangebelly darter, and the embryonic development of their hybrids; Loos and

Woolcott (1969) reported on the hybridization and behavior of Percina

notogramma and P^. peltata.

Clark (1973) made a rather detailed statistical analysis of meristic characters of JE. spectabile and caeruleum taken in allopatry

and sympatry in Indiana streams and found evidence of introgression, as

did Echelle (in manuscript) with E. spectabile pulchellum and E. radio-

sum in Oklahoma*s Blue River system following work by Linder (1966) and

Branson and Campbell (1969).

Behavioral interactions between sympatric species have been inves­

tigated to a lesser degree. Meristic character analyses of darters are considerably scarcer than life history studies. Even more rare are hybrid investigations and comparisons of natural and laboratory induced hybrids. The works of Hubbs and Strawn (1957a,b) are outstanding exceptions to this paucity. Hubbs (1959) has shown that the gametes of a considerable number of intra- and intergeneric darters are compatible, producing viable, if not fertile, hybrids. Little research until recently has involved parental and hybrid meristic character compari­ sons. Cole*s (1957, 1965) systematic and distributional study of the two genera complexes of the subgenus Boleosoma (E. nigrum and E_. olm- stedi) has generated and/or stimulated some controversy regarding sub­ species designation (McAllister, 1972; Zorach, 1971). Part of Cole*s technique was the partial employment of Stone's (1947) use of discrimi­ nant function analysis. Nelson (1968) utilized discriminant analysis in the determination of hybrids between the catastomlds C. commersoni and C . macrocheilus in Canada. In the present study discriminant (multi­ variate) function analysis was used yielding individual specimen

Indices, and graphically Illustrating specific separation and intergradation.

As a systematic tool, Ross (1973) karyotyped five species of

Etheostomatine darters from five different subfamilies including E. caeruleum but not spectabile. The writer karyotyped both these parent species and their hybrid offspring.

Various approaches have been utilized to theorize on the evolution and phylogeny of darter species; Winn (1957; 1958a,b) and Winn and

Ficciolo (1960) by habitat and behavior; Bailey and Gosline (1955) by vertebral counts; Collette (1965) by the presence and distribution of breeding tubercles; and Hubbs (1959; 1967a) by hybridization. In the present work, evidence from a multiple disciplinary approach was used to determine either, the similarity of, or diversity between, the morphology and behavior of 12. caeruleum and 12. spectabile. METHODS AND MATERIALS

Distribution of _E. caeruleum and _E. spectabile was determined principally by reliance upon authorities in their respective geographi­ cal regions of expertise and experience and on individuals who made extensive studies on the species, notably Knapp (1964) on_E. caeruleum and Distler (1968) on JE. spectabile. These investigators made "spot" distribution maps of the respective species. Ohio distribution was determined by personal collections, records of The Ohio State Museum of

Zoology (OSUM) and Trautman's (1957) Fishes of Ohio.

After the riffle environment was determined to be the darters' most favored habitat, collections were made from approximately 25 different, principally average to less than average size, streams In central Ohio containing this environment. _E. caeruleum and E . spectabile and their associated species were collected with a 4 x 6*, 3/16" ace mesh seine.

When species surveys were done, a larger 6 x 12', 1/4" square mesh seine was used in main channels and deeper pools.

Study Area

Dry Run, in southwest Pickaway County, Muhlenberg and Monroe

Townships, Ohio, was chosen as the primary study stream on the following bases: abundant occurrence of both study species, absence of I£. caeruleum in the headwaters (not a rare phenomenon), predominance of E. caeruleum in the mainstream, relative constant rate of flow, rela­

tive lack of pollution and turbidity, well-defined riffle areas,

accessibility to several stations over most of the stream's course, and

the possibility of comparison with the secondary study stream and other

streams in the area.

The secondary study stream, Peter's Run, an intermittent stream in

northcentral Pickaway County, Scioto Township, was chosen because of its

large _E. spectabile population, the absence of E_. caeruleum and its

excellent sites of observation of breeding activity. Stream studies

and specimen collections extended from May of 1970 through May of 1973.

The primary and secondary study streams were investigated intensively

during the latter two of the four year period, with an average of week­

ly collections. Stream order classification was according to Kuehne

(1962). In collections used to indicate habitat preference, the para­

meters of relative abundance, water depth, temperature, rate of flow,

and substrate type were measured and noted. Relative abundance, whether

in the general stream area or in a riffle section (micro-habitat sam­ pling), was recorded in terms of the ratio of total numbers of Indivi­

duals of each species (caeruleum; spectabile) taken. In the study of a general stream area an average of all seine hauls was taken. In micro- habitat sampling an individual selne-haul ratio was taken or an average of selne-hauls of a particular micro-habitat was made; "micro-habitat" being defined as an area of stream substrate not more than 0.6 square meters. Water temperature was measured in degrees Centigrade. Air temperature was measured and recorded for comparison. Rate of flow was determined according to methods outlined in Lagler (1952) using a float and stop-watch. Measurements of depth and width were taken in metric

units. Substrate sizes and types were classified according to modifi­

cations of Wentworth (Welch, 1948). Gradient was determined for specif­

ic stream areas (see relative abundance data) from 7.5 min. U. S.

Geological Survey topographic quadrangle maps and compared with overall

stream gradient data (ft./mile) from the Gazetteer of Ohio Streams

(Krolozyk, 1960). Quantitative comparison (relative abundance) collec­

tions were made especially from the study stream area in Muhlenberg

Township, except Stations IV and V, in Monroe Township (see Map I).

Stations:

I - Near bridge at Rte. 316

II - Downstream from Cochran Rd.

Ill - Downstream from Rte. 56

IV - Downstream from farm bridge; farm property, John H. Dunlap; Carl Hooks, manager

V - Downstream from Dawson—Yankeetown Rd.

Dry Run, a moderate gradient stream, 2.2 m. /km. (after Krolczyk,

1960), eroded into thick limestone gravels of Illinoian and Wisconsin

glacial drift (Stout, Ver Steeg, and Lamb, 1943), eventually developing

into some of the most fertile soil of Ohio, in Pickaway County (Andrews,

1874) may well typify JE. caeruleum and _E. spectabile habitat. The thick

till plains, with very few bedrock (limestone and shale) outcroppings

(Schuster, 1952), (none in the study area) are to the greatest extent cultivated. Corn and soybeans are the major crops. The banks of the stream are riparian only to a limited extent. In many cases cultivation of land approaches the edge of the stream bank and, in the case of r ? 1 2.3 | KILOMETERS 1

(Soloto R1 t « | )

Map I. Westcentral Pickaway County, Ohio. Stations Indicated on principal study stream, Dry Run. pasture, the stream was used as a source of watering cattle and other

livestock.

Collection records clearly indicated that Dry Run possesses favor­

able habitat for both species of darters. Overall collecting was admit­

tedly biased by habitat collections, but _E. caeruleum was always one of

the most abundant species in all collections in Dry Run (Table 1, Species

List). The headwaters of Dry Run, in extreme northwest Muhlenberg Town­

ship, are formed by pralrie-ditch (field drainage) run-offs, becoming

quite dry during late summer. A few kilometers from its origin, at

Station I, Rte. 316, in Muhlenberg Township, it exists as a persistent i stream, achieving a summer normal rate of flow of approximately 150—300

liters per second. The first 4 to 5 kilometers of stream course has a

soft, silted substrate with few well-defined riffles with only scattered

rubble. It was in this stretch that E. spectabile was taken in greatest

preponderance among the darter populations, the johnny darter (E. nigrum)

and the fan-tail darter (E. flabellare), being the only other darters

taken in the stream's uppermost ramifications. Not until a seasonal normal rate of flow of three to four times this, found 5 to 6 kilometers downstream, at Station II, Cochran Rd., still in Muhlenberg Township, was JE. caeruleum taken. (In several collections a single male was taken.

E. caeruleum equalled spectabile in relative abundance where riffles were an average of 1.5 to 3.0 meters in width, the water over the riffles was

10 to 20 centimeters in depth, and where the substrate was more gravelly than silty, half-way between Station II and Station III— near the mouth of a tributary from the east. Station III is located just south of Ohio

Rte. 56, where E. caeruleum clearly outnumbers spectabile in overall 10

TABLE 1.

SPECIES LIST OF DRY RUN

1. Dorosoraa cepedianum F * 17. Ictalurus nebulbsus VF

2. Esox americanus F 18. Fundulus notatus C

3. Carplodes cyprlnus hinei VF * 19. Ambloplltes rupestrls F

4. Moxostoma anisurum VF * 20. Micropterus dolomieui F

5. Moxostoma erythrurum VF * 21. Mlcropterus salmoldes VF

6. Hypentellum nigricans C 22. Lepomls cyanellus C

7. Catostomus commersonl C 23. Lepomls macrochlrus F

8. Erimyzon oblongus C * 24. Lepomls megalotls C

9. Semotllus atromaculatus VA 25. Percina caprodes C

10. Notropls ardens A 26. Etheostoma nigrum C

11. Notropls chrysocephalus VA 27. Etheostoma blennloides C

12. Notropls spllopterus F 28. Etheostoma zonale C

13. Notropls stramlneus C 29. Etheostoma camurum VF *

14. Erlcymba buccata C 30. Etheostoma caeruleum VA

15. Pimephales notatus C 31. Etheostoma spectabile VA

16. Campostoma anomalum A 32. Etheostoma flabellare A

VA - Very abundant; A - Abundant; C - Common; F - Few; VF - Very Few.

* Present as iuunatures only 11

abundance. Station IV, located on the farm property of John H. Dunlap

(Carl Hooks, manager) was chosen as the principal study site because of

the presence (and abundance) of both species and the particularly favor­

able habitat and observation opportunity. Just downstream of the farm

road bridge is a large, deep pool, gently sloping to a wide, shallow

riffle which is followed in turn by an extremely long (20 m.), narrow

(1 m.), deep (in excess of 25 cm.) riffle-raceway (Fig. 1). The pool area just preceding the head of the riffle forms a wide, flat "sand" bar, the substrate ranging in size from coarse sand (0.5 — 1.0 mm.) to fine gravel (2-4 mm.) to coarse gravel (4 - 20 mm.) to pebbles (20 -

64 mm.) and scattered cobbles (64 — 256 mm.) nearest the head of the riffle and the riffle proper, the types of substrate appearing in gradual gradation. Scattered boulders (25.0 + cm.) were found in the riffle proper and raceway.

The principal observation site at Peter's Run, the secondary stream, looking downstream from bridge on Ohio Rte. 104, shows the relatively wide, slightly currented pool, the narrow, shallow (4-6 cm.) riffle, and the elongated run (Fig. 2). Following and to the right of the run, though barely visible on the photograph (Fig. 2), just opposite the only tree in the vicinity is a wide, flat sand bar where breeding of the E. spectabile was observed. (See Results - Spawning Habitat.)

During the fall, winter, and early spring E. spectabile were taken in quantity from the riffles, which recalls Trautman*s (1957) emphasis on competition, or the lack thereof, in regard to riffle occupancy by one species in the presence or absence of its "competitor”. 12

Fig. 1.— Dry Run, Station XV. Pool, riffle and raceway; looking south. Fig. 2.— Peter's Run at Ohio Rte. 104. Riffle, run, and pool; looking downstream (east). 14 The substrate of the riffle area of Peter's Run Is essentially the

same as that described for Dry Run, limestone glacial gravels; the runs

and pools were more silted as typical of Ohio's Intermittent prairie

tributaries.

This portion of the stream was highly enriched by cattle waste run­

off from a farm yard just upstream of Rte. 104. In spring the algae-

covered rocks of the riffle actually teem with various species of aquatic insect larvae (especially Chlronomldae, Baetldee, and Ephemarl- dae). In mid- and late summer, portions of the stream were all but choked with Hydrodlctyon sp.» an alga common In enriched waters (Taft,

1968). The original speculation that the stream was Intermittent was supported by comments from a farmer living just west of Peter's Run on

Rte. 762 and by personal observation on August 17, 1972, when the riffles and shallow runs were quite dry. This condition lasted only a few days however; more than 5 centimeters of rain fell the evening of the day this was observed and the stream was not dry again during 1972, a year that averaged almost 25 cm. above the average annual rainfall.

A fairly wide variety of fish species was found In Peter's Run con­ sidering Its normal rate of flow. (Table 2, Species List) Enrichment

Is cited os a reason for their occurrence in such a small stream.

The specific areas of the stream— pool, raceway, riffle, rlffle- perlphery (head, above head, tall of riffle, etc.) were noted and recorded when micro-habitat collections were made. (The term "mlcro- habltat," as mentioned, connotes an area of approximately 0.6 square meters.) 15

TABLE 2.

SPECIES LIST OF PETER* S RUN

1. Catostomus commersonl C

2. Erimyzon oblongus VF

3. Rhlnlchthys atratulus VA

4. Semotilus atromaculatus A

5. Phoxlnus erythrogaster C

6. Notropis chrysocephalus VA

7. Ericymba buccata A

8. Pimephales promelas F

9. Pimephales notatus A

10. Campos toma anomalum C

11. Mlcropterus dolomleui F

12. Lepomls cyanellus C

13. Lepomls megalotls C

14. Lepomls macrochirus F

15. Etheostoma nigrum A

16. Etheostoma spectablle VA

17. Etheostoma flabellare F

18. Cottus balrdi F

19. Phoxlnus erythrogaster - Cllnos tomus sp. hybrid

(a single specimen) 16

In the majority of cases, most specimens collected were returned; some were kept live for laboratory observation; others were preserved for meristic and morphometrlc study. Those taken were chosen randomly, except in the case of particularly favorable breeding specimens, or in the case of an apparent or suspective hybrid. Specimens to be pre­ served were fixed in 10 per cent formalin and transferred to 70 per cent alcohol (AGW solution) and catalogued in The Ohio State Museum of

Zoology ichthyology collection by: Species identification, Accession numberj Collection number, Date, State, Locality, Range of standard length, and field data. See Appendix B for specifics on catalog numbers and collection data. Live laboratory—raised hybrids were kept for further study and experimentation.

Behavior

Fish behavior in the field was observed principally with the unaided eye while seated on a stool near or directly in a riffle area.

Natural foraging and reproductive behavior observations were noted at the time they were witnessed. In addition, a glass-bottom wood frame observation box, the viewing portion measuring 70 x 30 x 14 cm., was constructed and used to observe darter behavior in turbulent riffle environments. An "inverted" V on the front diminished turbulence around the box Itself. The box was tied to a stake driven into the stream bed (Fig. 3). This arrangement enabled easy observation of activity within even rather deep (more than 16 cm.) riffles. 17

Fig. 3.— Glass bottom observation box. 18

Simulation of natural habitat in the laboratory was preliminarily

(1971) carried out in two 10-gallon (38-liter) and one 15-gallon (56-

liter) glass aquaria divided into two and three equal sections respec­

tively. Usually no more than eight homospecifics (four pairs), with an occasional introduction of a heterospecific, were introduced into each section. In these preliminary investigations temperature, substrate, and waterflow were not scientifically controlled. Hore sophisticated equipment and controls were employed after It was observed thatvE. spectablle preferred a sandy to fine gravel substrate and IS. caeruleum preferred a bottom with pebbles, cobbles, and coarse gravel and would rarely breed In the absence of a current.

Three large narrow tanks (Fig. 4) were constructed:

A) 72-liter stainless steel frame (132 x 20 x 30 cm.) with an all glass front.

B) 132—liter plywood (152 x 30 x 20 cm.) with two (15 x 45 cm.) windows on one side.

C) 94-liter galvanized metal (130 x 30 x 23 cm.) with two (60 x 15 cm.) windows on one side, and one (92 x 15 cm.) window on the other.

A 255-liter tank, a commercially constructed "Living Stream" with a one-sixth horse power cooling compressor unit was purchased (Frigid

Units Inc., Toledo, Ohio) and proved adequate to serve as a holding tank and cooling system to maintain the temperature between 16° - 20° C. on all but the warmest summer days. ' Temperature regulation was not a problem, however, in breeding experiments as both species breed in the spring. It was not possible to induce breeding at any other time of the year. 19

-Compressor

B A.

P - Pump S — Siphon 0 - Overflow

Fig. 4.— Diagram of tanks and flow system used in simulated natural environment. (Arrows indicate direction of flow.) See text for tank capacities and description. 20

The 72-l±ter tank received water by means of a pump directly from

the holding tank at a rate of 16 liters per minute. The flow of water was piped into the tank at the level of the cobble—pebble substrate;

this tank was used primarily for caeruleum reproductive behavior observation (Fig. 5). Approximately one-third of a gravity overflow at

the opposite end carried water directly back to the holding tank; the remaining two-thirds was carried by continuous siphon flow to the ply­ wood tank. The latter was partitioned longitudinally with frosted glass and hardware cloth into two halves to keep the fish toward the observation (window side) half of the tank. With a substrate princi— * pally of sand, fine gravel, and scattered boulders, the wood tank was used for spectabile behavior study. Water from this tank was carried by gravity flow to a 75 liters sump tank.

Water was also moved from the holding tank by continuous siphon flow at a rate of 2 liters per minute to the galvanized metal tank, also divided longitudinally. One half was used for acclimation and temporary fish holding, the other half for observation of possible interspecific Interaction. Water was circulated and a current was maintained in this tank by means of a submerged continuously operating

Little Giant sump pump (Model it 1-AA) delivering 16 liters perr minute.

In this tank an attempt was made to set up optimum habitat for each species at opposite ends: cobbles, coarse gravel, pebbles, scattered boulders at the current source and a gradation to finer gravel and finally sand at the opposite end, with a few scattered pebbles and cobbles. caeruleum behavior. 22

Overflow from the metal tank emptied into the sump tank and water was returned to the holding tank, 3.5 gallons (13 liters) per minute, by means of a float—switch regulated pump. Figure 4 diagrammatically illustrates the tank set-up described.

The water depth in all tanks was approximately 18 - 20 cm. No aeration supplemented that obtained by running and partially turbulent water. During the months of April and May, approximately 25 fish in breeding condition were kept in each tank. After an acclimation period of less than 48 hours, most fish fed readily on finely chopped earth­ worms; breeding resumed within hours after introduction in laboratory tanks. Principally E. caeruleum were kept in the stainless steel tank,

J2. spectabile in the wood tank, and equal numbers of each species in the metal tank. Since breeding activity waned three to four days after

Introduction of fish into tanks, freshly caught specimens were used thereafter.

Tanks were illuminated by fluorescent lighting, though no attempt was made to accurately synchronize with the natural photoperiod and night-time illumination was employed for observation and photography.

Behavior in the field and the laboratory was recorded on color film using a Super 8 mm Bolex camera and either Ektachrome 160 or Dynacolor

(3M) film. Black and white film (Tri-X Reversal 7278) was also used in the laboratory. All color film was processed commercially; black and white movie film was processed by The Ohio State University Photography

Department. Black and white 35-mm. panchromatic film, whenever used, was developed and prints were made therefrom by the author in the laboratory of all phases of work in the investigation. Most movie 23

filming was done using 300-watt incandescent lighting for black and

white and EBV (3400° Kelvin) photofloods for color film (160 Ektachrorae-

ELA 464). Analysis and editing of the film was accomplished with the

use of a DeJur Eldorado Dual-8 mm. multi-motion projector, Model #86

and a Vernon Super-8 mm. film viewer-editor, Model #808. Notes on

foraging movements, reproductive behavior, striking morphological

changes and general behavioral activities were taken and, where feasi­

ble, comparisons between species were made immediately.

Laboratory intra- and interspecific crosses were performed accord­

ing to methods of stripping of small fishes described by Strawn and » Hubbs (1956). Fertilized eggs were kept in 1.5 to 3.0 cm. of water in

11.5 cm. culture dishes, which were floated in the water of the holding

tank until eggs hatched. Larvae were transferred to 2-liter square jars and were fed brine shrimp, microworms and later, white oligochaetes

(Enchytraeidae). Percentage of hatching was recorded and survival rate of offspring was monitored throughout the life of the controls and hy­ brid crosses. Airstone aeration was utilized through the life of the laboratory reared offspring. Temperature was not regulated except to avoid high extremes. These artificial crossings, especially the hybrid­ ization experiments, were performed in preparation for meristlc and morphometric characterization and chromosome morphology (karyology) of known hybrids.

Meristic and Morphometric Characters

Thirty-four E. caeruleum (19 males and 15 females) and thirty- three E. spectabile (15 males and 18 females) were used for counts and 24 measurements. Fifteen putative hybrids were tentatively listed subject to assignment after data analysis. Specimens used for this investiga­ tion were chosen randomly from collections taken from stations on the study stream, Dry Run. All but a few specimens examined were adults; no gravid females were used so as not to bias attempts to obtain information on sexual dimorphism outside of the breeding season.

Seven hybrids obtained by stripping methods and raised in the laboratory for a period of eleven months, were compared with ten (six males and four females, Including the parents) Individuals of each parental species; 12. caeruleum from Mac-o-chee Creek, Logan County,

Monroe Township, Miami River drainage, and _E. spectabile from Peter’s

Run, Pickaway County, Scioto Township, Scioto River drainage.

Characters of an eighth hybrid were compared with those of its parents.

Counts and measurements follow the procedures described in Hubbs and Lagler (1958) unless otherwise indicated. All measurements were taken with a dial caliper (accurate to 0.05 mm.) to the nearest 0.1 mm.

Sex determinations were made primarily on the basis of color pattern and the shape of the urogenital papilla. The characters examined in this study were chosen after perusal of methods by Hubbs and Lagler

(1958), Trautman (1948, 1957), Linder (1955), and Zorach (1972) and are as follows:

1. Standard length.

2. Head length.

3. Head width.

4. Head depth. 25

5. Snout length. From tip of snout to anterior margin of orbit.

6. Upper jaw length.

7. Interorbital width. Distance— least fleshy width.

8. Distance— union of gill membrane to tip of snout. The abbre­ viation U6M Is hereafter used to designate this character.

9. Body depth. Maximum.

10. Point of maximum body depth. Distance from tip of snout to point of greatest body depth.

11. Distance from occiput to dorsal origin. 12. Pectoral fin length.

13. Spinous dorsal fin base length.

14. Longest dorsal spine length.

15. Depressed soft dorsal fin length.

16. Longest soft dorsal ray length.

17. Predorsal length.

18. Preanal length.

19. Depressed anal fin length.

2 0 . Vent to caudal peduncle length. 21. Caudal peduncle length. 22. Caudal peduncle depth.

23. Dorsal spines.

24. Dorsal rays.

25. Lateral bars. All dark bands posterior to the opercular flap were counted whether partially or completely encircling body.

26. Pectoral rays. Combined count of rays on left and right sides.

27. Number of opercular scales.

28. Number of lateral line scales. 26

29. Pored lateral line scales. A single unpored scale within the series of pored scales was counted as pored.

30. Scales above lateral line.

31. Scales below lateral line.

32. Nape scalation. All scales on, above, and/or anterior to a line from the angle of the preopercle to dorsal origin were counted. The total number was divided by five giving an index ranging from zero to six.

33. Urogenital papilla pigmentation. The amount of pigmentation was given as a range (index) from zero to ten.

34. Infra—orbital canal. Completeness of the canal was assigned the value of one, incomplete zero. Determined for each side and totaled. Completeness (or incompleteness) of the canal was determined by introducing ca. 5 per cent methylene blue solution into the canal with a micro-pipette after clearing the canal of fluid with an air jet. See Figs. 6 and 7. This method was also used effectively on live specimens without permanent ill effects.

Infraorbital canal pores were divided into the categories:

lacrimal pores— usually four, in _E. spectabile separated from

subsequent postlacrimal pores by a discontinuity of the infra­

orbital canal. The normally "missing" pore of E. spectabile is

counted as a postlacrimal pore when found (in either species).

Completeness (or incompleteness) of the infraorbital canal seems

Independent of the number of lacrimal and/or postlacrimal pores.

35. Lacrimal pores. Combined count of the number on both sides of the head.

36. Postlacrimal pores. Combined count of both sides of the head.

37. Infraorbital bar. Intensity and degree recorded as an index ranging from zero to five.

All length measurements were computed by machine into thousandths of standard length. The .same calculation was made regarding the

following ratios: Fig. 7.— Infraorbital canal of 15. spectabile - Incomplete. Head length/Caudal peduncle length Head length/Distance to point of greatest body depth Head depth/Head length Head depth/Body depth Head width/Head depth Head width/Head length Snout length/Head length Snout length/Upper jaw length Interorbital distance/Head width UGM/Head length Body depth/Predorsal length Body depth/Head length Pectoral fin length/Depressed soft dorsal length Spinous dorsal fin base length/Depressed soft dorsal length Longest dorsal spine/Spinous dorsal base Longest soft dorsal ray/Depressed soft dorsal Longest dorsal spine/Longest soft dorsal ray Predorsal length/Preanal length Depressed anal fin length/Depressed soft dorsal length Caudal peduncle length/Caudal peduncle depth Caudal peduncle length/Distance— vent to caudal peduncle

The above parameters and ratios were chosen tentatively as possible species separating characters.

Analysis: Data was punched on cards and analyzed by computer according to the following tests on each of the fifty-seven parameters (measure­ ments and ratios) for:

a) Individuals of each specific group, by stream, and by stream

station: range, mean, analysis of variance, and standard

deviation.

b) Sexual dimorphism within each species. Fisher—Student t—test.

c) Interspecific character differences.

(Analysis of variance and Fisher-Student t-test were used to

test for significant differences. In the use of the t-test

the assumption of homogeneity of variance was verified using

an F-statistic (test of equality of variances). Only those 29

Instances where the probability of the F—statistic, (i.e. of ft the variance being equal) was greater than 10 per cent was

used.

d) Hybrid determination by specific character distribution and

individual intergradation. Discriminant function (multi­

variate) analysis based on selected parameters (see Results—

Analysis of Data).

One asterisk was used to indicate significant difference at the

0.01 level; otherwise, significant difference indicated was at

the 0.05 level.

The purpose of the analysis was to obtain a detailed, yet mean description of meristic and morphometric data for each species, so that future confusion regarding the identity of a fish of one of these two species can be avoided. The purpose of the discriminant analysis, in particular, was to determine: a) what variables were significantly different, one species from the other, b) using this information, determine what variables were of most use in deciding whether a given individual was 12. caeruleum. or 12. spectabile, or a hybrid between the two species, and c) to actually give a tri-modal distribution of fish tested indicating which were identified, statistically, as IE. caeruleum, or E. spectabile, or hybrids between the two species.

The hybrid index method of more popular use was shown to be essen­ tially the same, though more arduous method of obtaining the same results as discriminant analysis. 30

Chromosome Morphology

Karyotyping of somatic chromosomes of each species was performed according to the methods described in McPhail and Jones (1966) with the purpose of ascertaining on a somewhat gross genetic scale, the resemblance of the species' chromosome morphology to one another and to that of the hybrid. The thumb squash technique was used to reveal the metaphase plates. RESULTS

Ecology

Range

IS. caeruleum*s wide range extends from Minnesota, Iowa, and

Ontario, Canada, on the north through 14 states from Missouri and

Arkansas on the west to Alabama on the south and Kentucky, West

Virginia, Pennsylvania and New York on the east (Map II). It is of

frequent occurrence especially in the northern half of the Mississippi

River system: the drainages of Cedar River in Iowa and Minnesota, the

Rock and Wisconsin Rivers of Wisconsin, the Gasconade, White River

(which contains the subspecies E. c. oreium), and the Black and St.

Francis Rivers of Missouri and Arkansas (Pflieger, 1971; Knapp, 1964).

E. caeruleum is common in both the upper and lower drainages of the

Ohio River system: of the upper drainage, the Wabash of Illinois and

Indiana, the White and Whitewater of Indiana (Gerking, 1945), the

Scioto and Muskingum of Ohio and Pennsylvania respectively, and the

Allegheny River of Pennsylvania and New York (Knapp, 1964); of the

lower Ohio River drainage: the Cumberland, Kentucky, Green, Big Sandy,

and Licking Rivers of Kentucky, the Cumberland, Barren, and Tennessee

Rivers of Tennessee and Alabama (Smith-Vaniz, 1967), and the Kanawha

and Big Sandy Rivers of West Virginia (Lagler, 1966).

Besides this continuous range, the rainbow darter also occurs discontinuously in the Big Black River drainage of Mississippi and a

31 m

Etheostoma caeruleum

Etheostoma spectabile;

KILOMETERS

Map II. Distribution of Etheostoma caeruleum and Etheostoma spectabile in the United States. m 33

third subspecies, E. c. notiale is found in the Homochitto River

(Knapp, 1964; Cook, 1959),

Collection records, Knapp (1964), and Trautman (1957 and personal

communication) indicate extensive occurrence in various drainages of

all the Great Lakes notably the Grand River flowing into Lake Erie.

The fact that there is a thrust of distribution into the foothills of

the Catskill and Adirondack Mountains already suggests the preference

of this species for moderately high to high gradient streams, but Knapp

(1964) claims that "formidable gradients in the Tennessee River have

prevented E. caeruleum from invading headwater tributaries in eastern

Tennessee, North Carolina, and Virginia."

E. spectabile1s range is similar to that of IS. caeruleum and it

overlaps the latter in the east, and differs from IS. caeruleum with its

occurrence shifted southwesterly with the exclusion of the Appalachian

Plateau in the northeast and the inclusion of: the drainages of the

Central Lowlands of Kansas (Cross, 1967) and Oklahoma (Blair, 1957),

the Ozark Plateau of Missouri and Arkansas (Pflieger, 1971), and the

Ouachita Plateau of Oklahoma (Linder, 1955) and Texas on the west

(Map II), an overall physiography which suggests that the orangethroat

darter prefers a moderate to low gradient stream flow (Lagler, 1958;

Trautman, 1957). The eastern boundary of its range is the Allegheny

Front Escarpment and the Lower Appalachian Plateau. The range of the

subspecies IS. _s. spectabile is confined to the northeastern two-thirds

of the species' total range (Map II), limited on the west in the

Missouri River and White River drainages of Missouri and Arkansas

respectively. Like all other Ohio etheostomatine darters, neither 34

J2. caeruleum nor E. spectabile occurs commonly in the Gulf Coastal

Plains.

To summarize, major characteristics of the distribution patterns are as follows:

1) Both 12. caeruleum and 12. spectabile are all but absent from

the Coastal Plains.

2) Neither is found east of the Appalachian Divide.

3) JE. caeruleum occurs only sparingly in the upper St. Lawrence

drainage.

4) Both species are common throughout the Ozark uplift.

5) E. caeruleum does not enter the prairie states' drainage

while 12. spectabile is common there.

In Ohio _E. caeruleum is particularly abundant in the Miami, Scioto, and Muskingum drainages— in general in the glaciated regions of Ohio.

It is not found generally in unglaciated, low gradient streams of north­ west Ohio, nor is it particularly abundant in the very high gradient streams of eastern Ohio. (See Map III, Distribution map of Ohio after

Trautman's, 1957 Fishes of Ohio.)

JE. spectabile's western boundary in Ohio is, as previously men­ tioned, the western-most edge of the Appalachian Plateau. This boundary forms rather sharp limits to spectabile in Ohio, if one refers to its distribution in Ohio (Map IV) and compares this with Lobeck's physiographic map (Fennemann, 1938). 12. spectabile*s eastern limit corresponds precisely to the topographical delineation of Lowlands from the Appalachian Plateau in Ohio. This species is abundantly found in especially the western portion of Ohio in headwater streams and their 35

LAKE m m

Locality records

Since 1953 Glacial boundary

Map III. Collection records of Etheostoroa caeruleum In Ohio. Reprinted from The Fishes of Ohio, by Milton B. Trautman. Copyright ©1957 by the Ohio State University Press. • Locality records

A Since 1953

Glacial boundary

Map IV. Collection records of Etheostoma spectabile in Ohio Reprinted from The Fishes of Ohio, by Milton B. Trautman. Copyright ©1957 by the Ohio State University Press. 37

small tributaries of the Scioto and Miami drainages. It is of scattered

and isolated occurrences in the Muskingum and Hocking drainages

(Trautman, 1937). With respect to those in the Hocking drainage, in

Fairfield, Perry, and Hocking Counties, Trautman (loc. cit.) believes

"the orangethroat . . . invaded this drainage [in post—Pleistocene

times] when marshes connected the headwaters of the Hocking and Scioto

River drainages."

Topography (physiography), the resultant stream gradient, and geology are limiting factors in the distribution of 12. spectabile even within its general range. Striking examples are the absence, except for possible strays, in Alum Creek*s Devonian shale ravine in east- central Ohio and the sandstone substrates of eastern Ohio.

Habitat

E. caeruleum is more common in terms of abundance, while specta— bile penetrates a greater area (more tributaries) in any one stream system due to its greater tolerance to oxygen and turbidity levels.

IS. caeruleum is sometimes found in intermittent streams, but it does not thrive there, whereas spectabile survives and even prospers. A case in point is the abundance and size of spectabile, and absolute absence (in terms of numbers collected) of caeruleum, in Peter's Run, an Intermittent stream in Pickaway County (Scioto Twp.). Specimens collected from this stream rival (or excel) in size, abundance, and color those taken elsewhere in Ohio. Other small streams in the area confirm the hypothesis that intermittency Is a limiting factor in the distribution of _E. caeruleum, since Peter's Run and other prairie 38

streams similar in gradient, substrate, and water quality are, during

the fall and spring favorable habitat in every respect and flow directly

into the Scioto River where caeruleum is known to occur.

In Knapp's (1964) use of Trautman*s (1957) description of Ohio

caeruleum habitat as ’’moderate-sized streams of moderate or high gra­

dients, wherever the riffles were 15-70 feet in width, the average depth was one foot, and the bottom was sand, gravel and boulders", he stated

that this habitat observation "aptly applies to the species over its entire range."

E. spectabile, in most given streams, was found in greater abun­ dance in the headwater, prairie tributaries, and caeruleum in faster, deeper waters. It was usually possible, even in the mainstream, to collect a majority of either species, depending upon the ability of the collector of making a truly micro-habitat sampling. In a short, narrow riffle, though the overall numbers of each species were proven by collection to be approximately equivalent, careful observation revealed that each species manifests a rather distinct habitat preference, confirming Winn's (1958a) observations. The two species are considered sympatric, however, in the broadly conceived sense of the word, in that they are populations "the individuals of which are within cruising range of each other during the breeding season." (Cain, 1953; Mayr,

1963) To further illustrate habitat preference and stream type in reference to habitat, it seems advisable to employ Kuehne's (1962) modification of Horton's stream order classification system based on branching. Extreme headwater streams, intermittent or permanent, were ranked as first order. The union of two such streams forms a second 39

order stream. Whenever two streams of equal order join, they form one

of the next higher order (the rank of any stream is not increased by

the entrance of another of lower order than Itself). "Both pools and

riffles occur as components of individual streams of any order. Physico­

chemical and biological characteristics do not construct the classifi­ cation, they fit into it" (Kuehne, loc. cit.). Kuehne adopted the branching system of stream classification when he found a high correla­ tion between longitudinal succession (length in miles), gradient (ft./ mile), and stream order. Utilizing this method he compared occurrence of species with stream order. In B u c k h o m Creek, Kentucky, for example, he found "the creek chub, (Semotilus atromaculatus) was the only Order 1 fish, the most abundant Order 2 and 3 fish, but a rather uncommon

Order 4 species. The arrow darter (E. sagitta) was typical of Order 2 and to a lesser extent Order 3. The stoneroller (Campostoma anomalum) occurred in every Order 2, 3, and 4 collection" and so on.

This scheme can be applied with great efficacy to Ohio streams, especially in the current study involving _E. caeruleum and 12. spectabile distribution. Figure 8 illustrates its use in classifying streams of the Scioto drainage. In Ohio spectabile is found in Order 1 and 2 streams and caeruleum is found in Order 2, 3, and 4 streams, most especially in 3 and 4. The study streams themselves and the collections made therefrom were easily exemplified. Peter's Run, Plum Run, and in­ termittent tributaries of Dry Run and Blackllck Creek are clearly

Order 1 streams; and it Is from these that spectabile have been taken in greatest preponderance and abundance; while the mainstreams of Dry

Run, Big Darby, and the Olentangy River (upstream of impoundments) are *

A« Mill Creek B. Scioto River C. Olentangy River D. Alum Creek E. Big Walnut Creek F« Rocky Fork Creek 6a Blaoklick Creek Peter*s Run I* Big Darby Creek Ja Buskirk Creek K. Dry Run La Deer Creek

Fig. 8— Kuehne's (1962) stream order classification applied to northwest Scioto River drainage. Inset: area of study stream* Dry Run* and Buskirk Creek (detailed in Fig. 9). 41

Order 2, 3, and 4 streams respectively, all of which yield caeruleum in

abundance. (Figure 9 illustrates more detail of Dry Run.)

E. spectabile was found most prevalently in prairie stream head­

waters and tributaries (Order 1 streams), having a rate of flow of

0.15 - 0.30 kilollters per second. E. caeruleum were found in greater

abundance in riffles of streams having a rate of flow of 0.3 - 1.0

kilollters per second (principally streams of the second order).

Table 3 outlines the preferred habitat regarding stream order

(Kuehne, 1962), substrate (Welch, 1948), and rate of flow (Lagler,

1952). This information is Illustrated more graphically in Figure 10,

where relative abundance and rate of flow measurements are superimposed

on the left and right ordinate axes respectively. Since rate of flow

obviously fluctuates greatly with the weather, daily and seasonally,

determination of annual norms are the result of averaging rate of flow

readings taken throughout the year(s).

Examination of the parameters of rate of flow, substrate type and

size, and the comparative number of individuals of each species taken

or observed revealed an easily recognizable gradation— an increase and

decrease in relative abundance of j3. caeruleum and E. spectabile

respectively. E. caeruleum numbers increased in direct proportion to

the increase in rate of flow, and substrate size; the converse was true

regarding 13. spectabile. This phenomenon is illustrated in Figure 10 and more precisely by actual counts and percentages of their occurrence

(relative abundance) at stream sites studied in Fig. 11. The simple ple-graph is used to illustrate relative abundance in terms of percent­ age of each species; the light portion that of spectabile, the dark 42

KILOMETERS

(Scioto R»i- )|

Fig. 9.— Detail of inset (Fig. 8)— study- stream areal Stream order classification. 43

TABLE 3.

HABITAT PREFERENCE OF E. SPECTABILE AND E. CAERULEUM.

Habitat Substrate Rate of flow Stream Stream parameters order area

E. spectabile Silt, sand 0.15-0.30 I (& II)* Riffle fine gravel kiloliters/ periphery sec. (5-10 c.f.s.)

E. caeruleum Glacial gra­ 0.3-1.0 (II) & III Riffle vels , cob.bles kilollters/ proper boulders, sec. limestone (10-30+c.f.s.) bedrock

* It is especially in streams of the second order that E. caeruleum and E. spectabile appear together; neither in optimum numbers. Relative abundance r e d r o m a e r t S ntrso eaieaudne sra re n aeo flow. of rate and order stream abundance, relative of terms in i. 0—Darmai niaino eea curneo E ceuemadE spectabile E. and caeruleum E. of occurrence general of indication Diagrammatic 10.— Fig. aeo fo (liters/second) flow of Hate . caeruleum E. . spectabile E.

- I I I ■ - I I I

.O S ■o Ox O •P o § © U 03 © n •> Mill Creek W. Br* Alum Creek Turkey Run Un-named trib. 01.R. Alum Creek Scioto River Flint Ravine Creek Big Walnut Creek Rocky Fork Creek Sugar Run Blacklick Ck. Sycamore Ck. Poplar Ck. CP Grant's Run Plum Run Peter's Run Big Darby Creek Oppossum Run Clark's Run Buskirk Creek Dry Run V. Sgiggg

Percentage of E. caeruleum

Percentage of E. spectabile E* caeruleum not {filjfji abundant; £. spec- Vllll) tabile absent. Arabic numerals « Number of Fig. 11.— Sites of collection and relative abundance of JE. caeruleum and 12. spectabile. 46 portion that of caeruleum respectively. Streams are identified by higher case of the alphabet; the number of collections and/or observa­ tions at that site from which the percentages obtained were based, is indicated within or nearest the graph for that location. One can see at a glance that 12. spectabile is found in much greater abundance (than

.E. caeruleum) in the smaller, slower streams; IS. caeruleum in larger, deeper streams. In all cases tested, with the single exception of Alum

Creek tributaries, the upper reaches of streams within the geographical range of E. spectabile, this species was found to exceed IS. caeruleum in abundance. (However, Alum Creek’s northern-most tributaries are actually outside the range of E. spectabile.) Where both species occurred in the same stream, caeruleum supplanted spectabile in abun­ dance, with the exception of Grant's Run in Pickaway County, though the downstream-most site studied was approximately two miles from its mouth at the Scioto River where, of the two species, IS. caeruleum was clearly more abundant (OSUM collection records). Information outlined on the map (Fig. 11) is the result of collection and observation data taken on approximately 130 different occasions at more than 60 different stream sites.

Associated with headwaters and intermittent tributaries was a silt substrate and low gradient. IS. spectabile was taken in greatest numbers from small, slow, even sluggish streams having a low to moderate gradient with a soft, silted substrate, while IS. caeruleum was collected

In faster, moderate to high gradient streams in water broken by riffles composed of cobbles, boulders, glacial gravels, and bedrock cleaned of silt by fast flowing water. These findings are well documented by The 47 Ohio State University Museum of Zoology collection records. Where notes were made on substrate and gradient and three or more individuals of the species were taken, 68 of 76 (89.5 per cent) JE. spectabile collection

records indicated the substrate as "mud" or "muck" (silt?], silt, and/or

fine (glacial) gravel with less than 10 per cent mention of boulders or bedrock. In 63 of 68 cases (92.6 per cent) the gradient was noted as slight (low) to moderate. In the case of JE. caeruleum, the substrate was noted to be predominately, or a combination of, limestone bedrock, boulders, cobbles, large gravel, and/or glacial gravels in 161 of 181

(89 per cent) collections. Silt, when mentioned, was most often as:

"some silt," "little silt," or "sand silt." In 158 of 173 cases (91.3 per cent) the gradient was given as moderate, or moderate to steep, or steep; the gradient was indicated as low in only the remaining 8.7 per cent.

The same preference as illustrated by E. spectabile*s frequency of occurrence in headwaters and E. caeruleum * s abundance within faster, more voluminous waters and riffle areas, determined by selective seining, was indicated on a smaller scale within a given riffle. Care­ ful and repeated observation especially, revealed rather distinct pref­ erential habitats: IS. caeruleum in water broken by cobbles, rubble and boulders of the riffle proper, and _E. spectabile on the finer, sandy, gravel area just ahead of the riffle and similar substrate on sand- gravel bars and shallow pools following the tail of the riffle and the more silted areas of the raceway margins. As mentioned, where caeruleum was absent or scarce, spectabile was found in greater numbers in the riffle proper. These observations confirm those made by Winn (1958a) 48 who suspected that spectabile buried itself in the mud when frightened.

In this study it was observed that this darter will not infrequently bury itself under loose pebbles in an aquarium when attempts are made to capture it. E. caeruleum occupies the riffle proper, while spectabile selects the more flattened area ahead of and/or below the riffle where

the substrate is finer, where there is less turbulence. These comments are based on personal collections and observations from a total of approximately 40 different stream sites, and/or stations and are sub­ stantiated by The Ohio State University Museum of Zoology (OSUM) ich­ thyology records and behavior studies of darters in laboratory natural habitat simulation.

In two temperature controlled, circulating current tanks, 12. caeruleum. almost without exception, occupied the areas with most cur­ rent, whereas E. spectabile chose areas of quieter water, either in front of, or some distance beyond (downstream of) the current source.

Information based on collection records alone could be misleading, however, in terms of determination of a species—specific niche. For example, quite frequently it was recorded, and has been pointed out

(Trautman, 1957; Distler, 1968) that both E. caeruleum and E. spectabile were taken from the riffle area, which was indeed the case. But very cautious, quiet observation of a favorable habitat revealed the separate micro-habitat selections. When a sudden movement or vibration frightens the fish, they dash rather indiscriminately into the cover of, either the turbulence of the riffle, or the depth of the adjoining pool.

Obviously, the splashing and commotion involved in conventional col­ lecting methods on a riffle, makes determination of a micro-habitat 49

preference all but impossible. Seated on a stool in mid-stream at the

head of the riffle proved rewarding. In a few moments, if the observer

remained very quiet, the fish resumed occupancy of their preferred habi­

tats. Here E. caeruleum foraged, and between April and June, bred

freely and apparently undisturbed even between the feet of the observer.

This phenomenon was observed also by Reeves (1907) who made a special

point of mentioning a reduction in shyness during the breeding season.

Normally, observation of the turbulent portion of a riffle is

obscured by splashing and/or light refraction. This problem was elimi­

nated by the construction and effective use of a glass-bottom observa­

tion box. This device enabled excellent observation of activity within

even rather deep (more than 15 cm.) riffle. By this method it was

observed that El. spectabile was sometimes found feeding in peripheral

areas of the riffle. Obviously, the interphases of preferred habitats

are "encroached" upon by adjoining populations. In streams where E.

caeruleum predominates, there was evidence of competition for the most

optimum habitat in terms of food supply, especially if the food source

was benthos embedded in, or drift emerging from a riffle. In streams where j2. spectabile predominated or was present exclusively (in relation

to its sympatric partner), occupancy by the species of a riffle habitat

in question was more noticeable.

Thus, while holding that there was a clear and distinct habitat

preference per species, it is also true that some competition occurs.

Trautman (1957) especially places much emphasis upon this competition.

Describing the habitat of E. caeruleum, he states: . . marked inter­ specific competition was apparent in many riffles, especially in the feeding and nursery areas. . . . Whenever Interspecific competition from one or more of these species [J5. spectabile and other darters] was lessened or absent, the Rainbow Darter numbers showed a definite in­ crease, and in the few localities where such competition was absent, the

Rainbow Darters were phenomenally great." Young of both species occupy generally the same habitat, especially small riffles and the periphery of the faster more turbulent riffles. As maturity is attained so is the choice of the species specific habitat. The stream area occupied by given populations is outlined in Table 3 as the riffle proper for

E. caeruleum and the somewhat slower waters of the riffle periphery for

E. spectabile (above the head of the riffle, especially in the breeding season, and the runs following the tail of the riffle, when the riffle is flat rather than deep and narrow, and the water covered sand bars to either side of the turbulent water of the riffle). Data relating to

» this preference from collections carefully made from sections of rif­ fles and their surroundings are summarized in Table 4. This information, expressed as the ratio of caeruleum to spectabile in the two different habitats (riffle and riffle periphery) is compared to the ratio of these darters in the general stream area, the same data used to indicate rela­ tive abundance in Figure 10. Though, these overall total ratios of individuals taken from the general area show, in some cases, a high predominance of one species, the ratio of caeruleum to spectabile darters in a riffle is high in comparison, and the reverse situation of caeruleum to spectabile in peripheral areas, is low in comparison to the ratio of their numbers in that general stream area. TABLE 4.

STREAM AREA PREFERENCE. RELATIVE NUMBER OF E. CAEBULEUM TO E. SPECTABILE.

* Technique Collection Observation

Stream and General Riffle Riffle General Riffle Riffle station area proper periphery area proper periphery

Dry Run 2.5:1 23:1 1:5 2:1 10:1 1:5 Station III (Spring)

Dry Run 1.7:1 3:1 1:4* 2:1 5:1 1:5 Station III (2:1) (Fall)

Dry Run 2:1 35:1 1:5 6:1 30:1 1:10 Station IV (Spring)

Dry Run 8:1 16:1 1.3:1 6:1 10:1 1:10 Station IV (Fall)

Blacklick 1.4:1 8:1 1:8 1:1 10:1 1:10 Creek, C. College Rd. (Spring) Cn V * This ratio includes a number of immature E. caeruleum. 52

Data, based on collections, Indicated at least a tendency in

preference of habitat, but this information is not very reliable because

micro—habitat collecting is extremely difficult. When darters of either

species are disturbed, they can rather easily elude netting by diving

into the sand or among the cobbles and boulders. The tearing up of the

stream bed enabling their capture makes determination of their original

precise "micro-habitat" virtually impossible.

It is more rewarding esthetically and. more reliable statistically

to observe the darters undisturbed. This is possible with patience, and especially in the case of E. caeruleum, the use of the glass bottom observation box. It was principally by this approach that the ratios were obtained in the observation column of Table 4. Quiet observation requires a more precise knowledge of the markings and movements charac­

teristic of each species, but the resulting ratios are probably more reliable than those obtained by collecting techniques. Since all indi­ viduals can not be seen at once, the method is still an estimation, as is the collecting procedure.

Comparing the general stream area and riffle data of spring and fall darter ratios in Table 4, one finds a noticeable difference sug­ gesting a decrease in spectabile abundance, a decrease in caeruleum abundance or both. The obvious explanation would be an emigration or immigration respectively, or again, possibly both. Total numbers of each species taken during all seasons of the year reveal that of the two populations, E. caeruleum was more static in its local habitat occupancy. 53

Relative numbers of _E. caeruleum remained fairly constant (except

during the breeding season, when they were found in large numbers,

exceeding six individuals per 30 sq. cm. on the riffle proper). 13.

caeruleum was not found in greater or lesser abundance upstream ruling

out the possibility of a continual upstream (or downstream) migration.

Based on percentage of recaptures of marked (fin clipped) caeruleum in

three Pennsylvania streams, Reed (unpublished Master* s thesis; 1968) in

a 15—week study, found their populations "to be relatively stable within

a riffle habitat".

JE. spectabile, on the other hand, is found in fewer numbers in

riffles and riffle peripheries during summer, fall, and winter. At

these times it is taken more frequently from pools. Braasch and Smith

(1967), in discussing associated species in the life history of the slough darter (E. gracile) in Illinois, found that the slough darter, the bluntnose darter (E. chlorosomum), the orangethroat darter (E. spectabile), and the johnny darter (E. nigrum), all retreat to the deeper pools during the winter, where they, along with the blackside darter, (Percina maculata) may possibly compete for the common food supply. 54

Reproductive Behavior

Location of Spawning Sites

At no time during the year was E. caeruleum found in greater numbers on shallow riffles than during its breeding season, April through May. In 1 to 2 weeks following the breeding season and through­ out the fall, and especially the winter months, the adults were more abundant in the deeper parts of the riffle, the raceways, and shallow pools with current. These observations confirm those of Winn (1958a,b) in his studies of the reproductive habits of darters including _E. caeruleum and _E. spectabile.

During the spring _E. spectabile congregated in large numbers on shallow waters covering sand and fine gravel bars and riffles. At other times of the year E. spectabile was found in greater numbers along riffle margins, the riffle proper, especially in the absence of

15. caeruleum, and in adjoining pools.

Figures 12 and 13 illustrate the position of the breeding grounds at Station IV of Dry Run where the two species occur sympatrically. In this case, where a pool 0.8 m. deep preceded the riffle area, the breeding area of spectabile corresponds to the arms of a "Y", the shallow margins of the pool preceding the riffle, while the breeding area of caeruleum corresponds to the upright portion, the riffle proper.

Even where spectabile occurred alone (Peter's Run), breeding took place where the current was slow and the substrate was pebbles, fine Fig. 12.

Fig. 13.

Figs. 12 and 13.— Dry Run, Station IV. Different views of the same riffle area. Spawning areas of 12. caeruleum ( ) and E. spectabile 56

gravel, and sand, not In fast riffles having a cobble-rubble substrate.

Sexual Dimorphism

From late November through the following May, the caeriileum male

resumes a more brilliant and Intensified coloration, manifested most

strikingly by: a darkening of the rather uniformly parallel blue-green,

lateral bars; a similar darkening of the blue-green of the breast,

pelvic fins, and the marginal and sub-marginal bands of the spinous and

soft dorsals; a brightening of the Interradlal red of the caudal and

anal fin; and an overall deeper bluish-green cast. (Figure 15 contrasts

the markings of E. caeruleum with those of E. spectabile in Figure 14.)

Larger specimens, especially those taken from deeper, faster waters, had the darkest pigmentation. There is some individual variation in

degree and distribution of pigmentation locally In each species as

Zorach (1972) found with the bluebreast darter, IS. camurum, and the

greenfln darter, IS. chlorobranchium, and Scalet (1973), with the orange- belly darter, IS. radiosum. Females are an olive drab, color throughout

the year with little or none of the above described fin and body

coloration.

Typical of its name, the male E. spectabile is characterized during the breeding season by a brilliant orange throat, large blotches of orange on the posterior trunk and caudal peduncle separated by four or five blue-black often obtuse sub-triangulate blotches, the apices of which point anteriorly (Fig. 14), and irregularly shaped mid-lateral blotches below the spinous dorsal. As in IS. caeruleum, the spectabile female has an almost complete absence of color other than a beige cast. 57

Fig. 15. — IS. caeruleum male. 58

Somewhat linearly arranged dorso-lateral melanophores are more obvious

on the female than the male. Subtle, localized color changes, to be

described, in the females occur during the spawning process itself.

All specimens of both species which were used in morphoroetric

analysis, plus additional large series, were examined for the presence

of tubercles. No tubercles were found on E. spectabile. Only male E.

caeruleum, particularly large individuals taken from Mac-o-chee Creek,

Miami River drainage, possessed any visibly distinct tubercles. Tuber­

cles were found on the pectoral, pelvic and anal fins (Fig. 16) as well

as on some ventral scales. No tubercles were found on 13. Caeruleum

from the Scioto River drainage. Knapp (Collette, 1965) discovered

tubercles on belly scales, and on mid-ventral rox*s of the caudal

peduncle of E. caeruleum principally from Illinois collections. He

found no tubercles on fins of caeruleum from any location. Collette

reported tubercles on the anal fin, ventral scales, pelvic, and pectoral

fins, and the lower part of the caudal fin of spectabile males taken in

Missouri, Tennessee, and Oklahoma. Winn (1958a) found no tubercles on

Michigan spectabile. Collette (1965) suggested that the tubercles assist

the male in maintaining his position while in contact with the sides of

the female during spawning.

Spawning Season

Females of both species show the onset of ovarian development by

abdominal distention in early January, but otherwise manifest no color

change. Eggs were determined to be ripe according to methods described by Strawn and Hubbs (1956); "Ripe eggs are large, clear, adhesive and Fig. 16.— Tubercles on anal fin of IS. caeruleum male from Mac-o-chee Creek, Miami River drainage. 60

flow with gentle pressure on the abdomen." Both caeruleum and specta­

bile eggs are ripe In most Individuals tested by April 5, as judged by

these criteria, as well as their viability in laboratory hybridization

and control crosses. Actual matings in the field were not observed

until mid—April when the water reached and maintained a temperature of

15-20° C. At the latitude of central Ohio the breeding season reaches

a peak for both species during the end of April and beginning of May,

tapers off and spawning all but ceases by the end of May. Exception

was found in fish obtained from more northern streams (the Kokosing

River and its tributaries) having cooler water. Some of these specimens

brought back to the laboratory mated until the end of the first week in

June. However, even here most of the females were spent by this time.

These dates apply to central Ohio. Reeves (1907) and Winn (1958a,b) in

Michigan, Linder (1958) and Dlstler (1968) in Oklahoma, and Clark

(1973) in Indiana found approximately the same time span, April through

May, as the reproductive period of both species with that of IS.

caeruleum slightly longer.

Spawning Habitat

For several weeks prior to the breeding season of E^ caeruleum

and JS. spectabile, especially where the latter species occurred alone, males outnumbered the females in the shallow riffle areas. Repeated

collections and observations have shown that when water was high and

rather turbid, fewer darters of either sex were found on, than below,

the shallow riffle areas, often being quite abundant in eddies and

pockets to the downstream-side of a bank abuttment or sand bar. Males 61

of the two species were the first to occupy their respective breeding grounds. Xt was observed that, in the 3 to 4 weeks before the actual spawning, males predominate on what will be the breeding areas. In deeper parts of riffles and pools, the ratio of females to males was found to equal or exceed twenty to one, whereas the proportion in the shallow riffle was approximately reversed. From the last of March through May, males continued to predominate on the riffle areas, except when females moved up from the seclusion of the deeper riffle or pool to spawn on the shallow riffles. Data gathered from field and labora- tory observations indicated that males of both species established territories, which were oriented to a large stone or depression in the gravel. Either these sites were, in reality, optimum spawning sites based upon water flow, depth, and substrate type, or they became optimum in that the presence of (a) male(s) defending such a territory attracts females to the site. In laboratory tanks having a rate of flow of

16 liters per minute, dominant males defended territory against other males— areas "downstream" of a boulder or areas with scattered boulders.

A dominant male caeruleum defended such an area in a laboratory tank against other males for 3 days leaving it only a short distance (10 -

15 cm.) when a female "invited" his services in the near vicinity. But more often than not, the breeding females chose this same place to spawn. Obviously, in the natural environment there is an increased availability of such favorable sites.

Females, therefore, lagged behind the males in movement to the breeding sites. Both species moved to riffle areas, but a more critical evaluation of habitat selection was necessary in order to deduce means 62

or mechanisms of specific isolation. The breeding habitat preferences were based upon numerous observations at stream sites and on stream

habitat simulation studies In the laboratory.

E. caeruleum occupied as their breeding grounds primarily the flat, wide downstream-most half of the shallow riffle portion where the water was from 10 to 15 centimeters deep. The breeding ground substrate

corresponded roughly to that described as general habitat— coarse gra­ vel, pebbles, cobbles, but often oriented, to one or more cobbles or boulders scattered in this area (Fig. 17).

Reeves (1907) and especially Winn (1958a,b) adequately described the movements of the females to the breeding site and the sham (display) battles of the males. None of these descriptions, however, admittedly distinguished between the two species. A particular attempt was made to ascertain subtle differences between all of these-—behavior, habitat selection, and morphology.

Figure 18 illustrates the type and distribution of substrate materials and water depth at the observed (preferred) breeding sites of both species at the principal observation Station IV on Dry Run.

Figure 19 outlines the same parameters regarding 15. spectablie alone at Peter's Run.

E• spectablie chose shallower water, 5 to 15 centimeters deep, a finer substrate (coarse sand, gravel, and pebbles), and areas immedi­ ately preceding, adjoining or following the turbulent water of the rif­ fle proper. They were found at the head of the riffle (Winn, 1958a,b) in a smooth gradually sloping area with a fine substrate, and the water not more than 22 cm. deep. The breeding sites of spectablie in Dry Run 63

Fig. 17.— Substrate of coarse gravel, pebbles and cobbles typical of JE. caeruleum spawning area. Surface Stream area Water depth Substrate velocity Bridge

Sand­ bar A

Coarse sand Silt

POOL

0*8 m,

10-20 cm. sloping to bank

Coarse gravel steep Pebbles bank Scattered cobbles and RIFFLE * Gently boulders sloping bank

Gravel Cobbles RACEWAY Boulders Sand-silt

20 m. Silt Scattered boulders

E. caefuleum; x E# spectabile: <

POOL o-5

Fig. 18. — Sketch of stream and location of spawning sites of E . spectabile and E_. caeruleum at Station IV of Dry Run. Substrate, surface velocity and depth data are superimposed. ace Stream area Water depth Substrate ve.

Riffle Cobbles Boulders

Slow raceway

Sand, fine and coarse Sand-gravel 10-12 cm. gravel; bar scattered pebbles

Silt

Scattered pebbles Fine & coarse gravel Riffle Breeding area: o

1 meter

Fig. 1 9 .— Sketch of stream and location of spawning sites of E. spectabile at Peter’s Run, Ohio Rte. 104. Substrate, surface velocity and depth data are superimposed. 66 were the gradual slopes of the large pool preceding the riffle and the area of slower, shallower water (Fig. 18). In Peter's Run, where there was no competition with caeruleum, spectabile occupied this same type of habitat for breeding. Early in the spring (mid- to late March) some male and female spectabile were found in the riffle proper, in a cur­ rent where one would expect to find caeruleum, but more were found in slower water of the shallow pools. This particular stream is atypical in that it is highly enriched by run-off from a livestock (cattle) yard less than one-half kilometer upstream from the study site. By mid-May the stream contained thick growths of algae, Cladophora and Hydrodictyon sp., veritably teeming with aquatic insect larvae. The riffle was, as in most streams, one of the principal food source areas (Hynes, 1970;

Waters, 1965). Yet spectabile were not observed breeding in the riffle proper where the substrate was principally cobbles and boulders, even in the absence of caeruleum. Figure 20 illustrates the breeding habitat of spectabile in Peter's Run: flat, fine-gravel bars preceding or to one side of the turbulent riffle or raceway. Reeves (1907) was undoubtedly referring to E. spectabile (not to JE. caeruleum) when she relates: "In spring the darters leave their lurking places in the rapids and congregate on the gravel sheets which are spread out at the head of the rapids where the stream leaves a pool. Here the water is from 1.5 to 4 or 6 inches deep with a current moving at the estimated velocity of about 75 feet per minute. The pebbles of the bottom are small, averaging 0.5 inch in diameter while the largest is not over

2 Inches." This description coincides precisely with this researcher's observations for E. spectabile confirming those of Winn (1960). In 67

Fig. 20.— Gravel bar, spawning site of IS. spectabile. studies of substrate plots In Illinois, Larlmore et al (1952) found "the

rainbow and the greenside [Eh blennioldes] showed preference for the

larger, faster rapids, whereas the orangethroat and the fantail [Eh

flabellare] showed a preference for the smaller and slower rapids."

These findings concur with those of Trautman (1930, 1957, and personal

communication). Neither these investigators, nor the literature search

revealed the fact that, in many cases, the female caeruleum chooses a site to spawn immediately adjacent to, or very near, a large cobble cor­ responding, especially in laboratory tank observations, to a "marker" bordering an area defended by a male. In the field, coarse gravel has accumulated here, and more than likely, the current of this micro­ habitat (choice) was considerably less than that measured at the surface.

This makes it understandable why a rather small, though definite current in a laboratory tank with scattered boulders placed among finer bottom materials sufficiently simulated the natural environment to elicit the limited territorial defense and breeding behavior of IS. caeruleum in captivity. Unlike E. spectabile which bred readily in tanks without current, caeruleum rarely bred in tanks without at least some current.

Any orientation by a female spectabile to a visual cue seems accidental, if present at all, since its breeding habitat is considerably more uni­ form (fine gravel with only sparsely scattered cobbles). In the great majority of matings observed in the laboratory, females selected areas of uniform sand and/or fine gravel quite distant from any such marker. 69

Pre-spawning Activity

_E. caeruleum

About 2 to 3 weeks before the height of the spawning season, the

deep blue-green and red pigmentation of the males reached its greatest

intensity. Portrayal of these colors was exaggerated during sham

battles. Ritualized fighting between the males occurred most often

without any fixation on a particular area although it may have origi­

nated when an interloper challenged the territorial holdings of a

particular Individual. In the field and in the laboratory, battles were manifested as "moving" territorial displays and continued over a dis­

tance of as much as a meter or more, normally in an upstream direction,

except when in a protected area (of less current) or in an observation

tank with little current. But these displays were not territorially

defensive as no specific area was protected during these parallel moving

displays. During these encounters the pelvic, caudal, dorsal, and anal

fins are stiffly erect. In less current the display was more elaborate.

Often the body, with the head lowered, assumed an angle of 25-30° with

the substrate. The pelvic fins and head ventrum formed the legs of the

trlpodal attitude. Unlike the elevated caudal of the male spectabile

to be described below, there was a decided effort to keep the caudal fin

spread and touching, or near touching, the substrate. In the former position the posterior half of each male including the soft dorsal,

caudal, and anal fins served as a veritable flag as each of these fins were widely expanded. In the caeruleum male, especially during the breeding season, when these three fins are diffuse with red, this 70 display was reciprocated between males. The dark vertical bars of the posterior trunk seem to serve as the target, as this region was lunged at alternately by the pugnacious participants. The region about midway between the anal fin origin and the caudal peduncle was butted by the opponent. Contact was often made, but biting was not observed, and injury was rare. In the field these battles were observed as the fish swimming parallel in almost continual display, fins erect, except for maintenance of equilibrium with the current. The combatants maintained a distance from each other of only about 2 to 3 centimeters or almost touching, as described by Reeves (1907). In an aquarium this behavior was modified to the extent that there was more lateral display, head to tail circling, more actual butting contacts, and the distance between individuals varied considerably from actual contact to a tolerance distance of 8 to 10 centimeters.

One of the two males eventually retreated and the dominant male seemed to be the victor, but in a sense both had won, if the theory also expressed by Winn (1958b) is correct, that this mutual agonistic activity results in a state of frenzied sexual excitation for all fish, including females in, or approaching, the vicinity.

E. spectabile

Differences between caeruleum and spectabile agonistic behavior possibly resulted from their evolution in slightly different habitats.

The parallel display between caeruleum males observed was modified in spectabile in that there was considerably more circling, "carouselling", a term used by Barlow (1962) in describing the circling of one 71 Individual about the other during the reproductive behavior of male to

female to male Asian teleost Badis badis: "The movement follows natu­

rally from the initial antiparallel orientation combined with reciprocal

ramming of one another." In Barlow's Badis the ramming gradually

changes to persistent pushing which is a continual head to tail circling

of the pair. Possibly spectabile, in slower water, has evolved a con­

comitant more elaborate agonistic behavior. spectabile was found

breeding on a sand-gravel bar, in water so shallow that the dorsal and

caudal fins often break the surface.

In spectabile there was considerably more movement and pugnacious

activity. In addition, spectabile elevated itself on its pelvics and

anal fins in a slightly more horizontal position than caeruleum but,

at the same time, raised Its caudal fin definitely above the substrate.

Ramming attacks from the opponent seemed actually to be solicited by a

waving, undulating motion of the caudal region while the were

propped on pelvic and anal fins. During these agonistic encounters,

the melanistic blotches in the frontal trunk region (below the spinous

dorsal) all but disappeared, accentuating those of the posterior por­

tion, again surrounded by the colorfully diffused soft dorsal, caudal,

and anal fins. In this case, however, the anal is suffused with blue-

green rather than red as In the caeruleum male.

The patterns of the two species, which draw such intra-specific

agonistic behavior are sufficiently different to serve as an explanation why no inter-specific interaction (display) between males was observed in the field or the laboratory. This observation concurs with Winn's

(1958a,b) observations. 72

The response to the waving display of the caudal region of E.

spectabile was a vigorous lunge, a butting or ramming action which often

tumbled the recipient dashing against the substrate causing a flurry of

sand and gravel. No specific area was defended as this activity was

reciprocated as many as 10 to 12 times before it suddenly abated, as

did the display attitude.

In both 12. spectabile and caeruleum, the latter especially in

slower water, there was a definite head to tail arrangement, each appar­

ently fixated on the caudal target area. These aggressive displays

varied in extent and degree from one encounter or one pair of indivi­

duals to the next. Barlow (1968), in studying fixed action patterns

and their occurrence in B. badis stated that "the hostile displays are

under continuing modulation during their expression. The degree of the

spreading of the fins, the angle at which the body is held, the pres­

ence or absence and timing of head Jerking, are all responsive to the

continuing changes in behavior of the opponent." The point applicable

here is, that if a displaying male fish is not challenged, a sham bat­

tle will not occur; if it is strongly challenged, a strong response

(encounter) will ensue. Exactly what amount of valuable feedback is

achieved was not determined. Barlow observed (loc. cit.), "When two

JJ. badis engage in hostile displays, they characteristically pitch for­

ward at an angle of about 45° while maintaining a head—to-tail orienta­

tion; this behavior is recognized through its orientation to the sub­

strate and to the rival." From the vertical headstand display of the

stickleback (Morris, 1958) to the slighter angles to substrate of IS.

caeruleum and J2. spectabile, this orientation seems to serve as a 73 signal; possibly it is a remnant of a previously more complex ritualis­ tic behavior or the evolution of one yet to be embellished and refined.

Perhaps it is already refined but not yet understood. Certainly, realization of fragments or fine units of behavior is at least a clue to a more highly integrated complex of which it is but a part.

A phenomenon not mentioned by darter behaviorists is the change

(increase) in pigmentation of the eye, especially of spectabile during aggre.sslve behavior. Figure 21 compares the normal eye color of spectabile with that attained during the combat period in Fig. 22. In the '’normal" male spectabile eye the sub-circular pupil is surrounded by a golden-colored iris. Three light brown lines radiate from the orbit, one toward the snout and two posteriorly, the first at an angle of approximately 20° and the second roughly parallel to the body axis.

These three lines are all but obliterated (diffused) during the male's agonistic behavior. Instead, the anteriodorsal and anterioventral por­ tion of the iris darken and this darkening continues into the hyper­ orbital tissue making the eye appear larger and considerably more noticeable, creating a more visible disruptive pattern. This phenomenon no doubt ties in with the orientation and stereotyped display including the■head-to-tail position and elevated body. Barlow (1972) in dis­ cussing fish eye-line patterns, suggested that such an emphatic pattern is one solution to solving the problem of low visibility through water.

The overall darker color of the aggressive caeruleum male suggests less

"need" of any added aposematic coloration. (This phenomenon was defi­ nitely less striking in E. caeruleum.) In 15 to 20 seconds following the conflict abatement, the eye returned to its usual color. Fig. 21.— Usual eye appearance of male 13. spectabile.

Fig. 22.— "Slash-eye" pigmentation of male 13. spectabile during aggressive behavior. 75 At times, brief but violent exchanges between males resulted in

the pursuit of, and mating with, a female which was in the vicinity; a fact supporting the "excitation" theory (Winn, 1958a). While the sham battles do not seem to be a necessary preliminary to mating, or spawning a necessary consequence of these battles, they could well serve as social releasers, or stimulators to reproductive activity for both sexes. Social releasers here were comparable to that discussed by

Tinbergen (1959) in connection with gull displays. Though female J2. spectabile did not form a spectator circle around the displaying males, they were attracted to the area and, though no test of their response was devised, they are most probably affected socially.

Spawning Behavior

Even though the males of both species become involved in what might be considered pre-spawning behavior, the females actually initiated spawning. One of the first signals of receptivity by the females of either species is their movement from the deeper portion of the riffle, raceway, or pool and their presence on the "breeding grounds". Aside from recognizable markings described below, each species manifests an

"agitated" type of movement which elicits the "following" reaction from the respective males. The precise difference in behavior per species involves subtle associated changes in coloration and type of movement.

The distinction between the movement of caeruleum and spectabile females can best be termed "pseudo-feeding" and "twitching" respectively.

E. caeruleum females appeared to be feeding or in search of food as they move from deeper parts of a riffle to the more gravelly portions of the riffle. Upon more critical observation, however, it was revealed

that they pick up granules of sand and gravel and momentarily eject them

as if testing the substrate. In this position, head down, with back

arched, the male fish rushes to the side and follows her closely in her

sporadic, jerky movements. For her part the spectabile female exhibits

a "twitching" motion when at rest. This movement ensued especially

when a male was in the near proximity. It lasted barely more than a

second, five to six vibrations per second, and was repeated at two to

five second intervals. The twitching may suddenly cease, and after a

short forward dart on the part of the fish, begin again. The movement

begins as a side-ways vibration of the head and may continue backward

vibrating the entire body.

The JE. spectabile female, disposed to spawning, acquires two kinds

of localized color changes. The iris of the eye and lines radiating

from the eye darken considerably resulting in the appearance of a much

larger "eyespot" (Figs. 23 and 24). Associated with, or at least coin­

ciding with, the "eye darkening" is an intensification of the pigmenta­

tion of the three saddle bands of the caeruleum female. The difference

in coloration of females of the two species is the greater intensity of

eye darkening of JS. spectabile and the more pronounced saddle bands of

_E. caeruleum respectively (Fig. 25). Again, though these markings may seem of little significance to the human observer, they may afford appreciable discernment to conspeclfic males. There is marginal overlap in the breeding habitats of the two species so, what may seem to be undue emphasis on specific distinction, is necessitated for species

Isolation. Interspecific matings have often been observed in the 77

Fig. 23.— Usual eye appearance of J3. spectabile female

Fig. 24.— "Eyespot" pigmentation of 12. spectabile female in spawning disposition. Fig. 25.— Saddle bands of 15. caeruleum female in spawning disposition. 79 laboratory where spawning areas were admittedly more cramped.

Present, therefore, in the pre-spawning phase, was a series and combination of visual cues which could well serve as stimuli and re­ leasers to synchronize the male with the readiness of the female. The movements, markings, and the distention of the abdomen of the female elicit the "following" reaction of the male. Specific differences in pre-spawning behavior of both sexes are summarized in Table 5.

Spawning

The actual spawning behavior of the two species is so similar that

< the description to follow describes both species except where otherwise specified.

The receptivity of the female is acknowledged by the male's

"following" response. If the female does not respond in turn by imme­ diately "diving" into the gravel, the male makes continual passes over and in front of the female, sometimes coming to rest on her dorsum, sometimes nudging her side. The female may or may not immediately respond to this attention and contact from the male. Males now violent­ ly defend an area surrounding a receptive female, identified, as described, by the agitated movement, "pseudo-feeding," and dark saddle bands of JE. caeruleum, and the "twitching" movement and dark eyespot of

E. spectabile. Whether the passing over and/or touching of the female by the male causes her any further stimulation is not known, for the amount of time she "waits" before "diving", varied from a few seconds to several minutes, although in the field it was more a matter of a few seconds. Specific releasers were different as mentioned and summarized 80

TABLE 5.

COMPARISON OF PRE-SPAWNING BEHAVIOR AND ASSOCIATED LOCALIZED CHANGES IN PIGMENTATION OF E. CAERULEUM AND E. SPECTABILE.

Males E. caeruleum E. spectabile

Agonistic behavior

Body position Angle of body to Angle of body to substrate 20—30° substrate 10-20°

Sham battles Limited "carrou- More selling" in quieter "carrouselling" water

Coloration Overall Lateral blotches intensification of anterior trunk (bluish cast) reduced in intensity

Females

Type of activity

Movement and Agitated, sporadic Agitated, sporadic marking "Pseudo-f eeding" "Twitching"

Localized Coloration Dark saddle bands Dark eyespot

Position in Dorsum exposed Head exposed. gravel Dorsum not always exposed. 81

In Table 5 along with more subtle differences In mating.

In spawning the female thrusts herself headlong Into the sand or

gravel by a rapid undulation of the caudal fin. This action usually

results in the burial of the anterior half of the fish. A second or

third effort, often necessary on the part of jE. spectabile, thrusts the

fish about 2 cm. forward causing the head to partially emerge and the

remainder of the body to be almost completely buried in the gravel with

the dorsum barely exposed, if exposed at all. Though no test was made

of this, E. caeruleum seemed to enter the gravel more rapidly (with

fewer unsuccessful attempts) than spectabile. The action and vibration

involved in the diving was detected by males in the area which rushed

to the scene. The dominant male vigorously defended a circular area of

approximately 10 - 12 centimeters radius of tolerance against rival males surrounding the buried female. Defense against conspeclfics was always more vigorous than interspecific aggression although both occur.

Female conspeclfics seemed to be chased from the area only by accident as the excited agonistic male darted back and forth protecting the area surrounding the buried female against male rivals. This defense was not totally effective, since a second or even third male joins in after the original pair had already begun the act of spawning. These intru­ sions were usually made by smaller, sometimes juvenile (yearling) males.

In the case of E. caeruleum, the dorsal region except for the caudal area is more exposed above the gravel, revealing rather clearly at least two of the three darkened saddle bands. E* spectabile is not as robust and deep-bodied as caeruleum, and though the dorsum may be exposed, it is not as obvious as caeruleum*s. What jte obvious in the more than half—buried female spectabile, is the greatly exaggerated pigmentation of the eyespot. Experiments with recently sacrificed female caeruleum decoys and models, show that the saddle bands are recognizable signals to the males and females as well. Once an indi­ vidual female has positioned herself with all but her dorsum buried in the gravel, often near a cobble or boulder, another female dives in immediately alongside, followed by a third, fourth, or even fifth so that four or five females will be aligned side by side. A similar grouping, each associated with numerous males, can be found just inches away. This phenomenon was confirmed by field observations (Fig. 26).

By using models partially buried and colored with pronounced saddle bands, it was learned that 1) females align and bury themselves immedi­ ately adjacent to the burled model and 2) males have mounted the models, but mating did not follow, possibly because some vital movement was lacking— however, on one occasion a male attempted to spawn with a dead female thus buried. The aggregation of females reminds one of what Winn and Plcciolo (1960) call "communal" spawning (a huge concen­ tration of eggs on a single object or smaller area) as it occurs in the glassy darter (E. vltreum). In this case a large aggregation, 50 or more darters spawn at a single site on a large rock or under a log or ledge. Though neither caeruleum nor spectabile spawns at a single site or object, not in quite such large numbers, their spawning behavior is reminiscent of communal spawning. Large numbers of spectabile are found spawning at one time on the same gravel bar. As many as 60 darters have Fig. 26.— Aggregate spawning of IS. caeruleum. Note position of females (arrows) partially buried in gravel. 84

been observed In an area no greater than a square meter, starting with

males displaying during sham battles, followed by Increasing numbers of

both sexes. After 2 to 5 minutes of quiet observation, this same area

became concentrated with much displaying and spawning. Xt Is suggested

that such spawning aggregations result In the mutual sexual stimulation

of Individuals involved.

In a stronger current the aggregate formation of caeruleum takes

on the different form described above. Of probable signalling value to

both sexes of caeruleum are the darkened saddle bands of the female.

While this aggregate spawning does not result in the huge concentration

of eggs on a single object as in 13. vitreum, it cannot be denied that

E. caeruleum and IS. spectabile congregate on small (one to two square meters) optimal areas. The group spawning behavior of the latter two

species will be referred to as "aggregate" spawning to distinguish it

from the communal spawning of the former.

Approaching usually from the tail end, the male assumes a position

parallel to and over the back of the female. With the head of the male

above and just behind the head of the female, with the pelvic fins of

the male straddling the female's dorsum, he begins a very rapid vibra­

tion of the pectoral fins; the force driving the trunk portion against

the female's soft dorsal. Both fish are now in a head to head orienta­ tion as the male has moved slightly forward. Almost immediately after the vibrating action of the male upon the female, the latter begins vibrating her pectoral fins. The combined vibrating action of the male's and especially female's pectorals sends sand and gravel flying 85 a distance of 10 or more centimeters above and around the breeding pair.

As mentioned, usually a second or third male joins in the mating process.

Simultaneous with the vibrating motion the fins of the male are held stiffly erect. The anal fin wedges between the gravel and the abdomen of the female as the anal (abdominal) regions of both fish dip deeper into the gravel.

During the actual spawning, the female's mouth is agape possibly adding pressure to that already exerted by the force of the male above and that of the substrate from below and the sides upon the abdominal region, thus releasing 7 to 12 eggs per mating (Winn, 1958a) into the gravel. Most often, the downward and forward thrust by the male forces the female forward. This action often continued to a point where the female was forced completely free of the substrate, or, with a single jerking motion, she moved 3 to 4 centimeters forward and rested, venti­ lating gills rapidly, as if exhausted. Especially in the case of j2. caeruleum, matings of the same pair may be repeated after 1 or 2 minutes until abruptly, a) the female is no longer receptive and no longer buries herself for a period of time in the gravel (no doubt receptivity in the.female is signalled by the presence of ripe eggs released from the ovary), b) an interloping male engages the dominant (breeding) male in a sham battle, or c) either the male or female is frightened off or temporarily loses orientation with the preferred spawning site.

After the caeruleum female was no longer in the spawning disposi­ tion, the agitated, "nervous" movement, abated; the eye and the saddle bands returned to their normal coloring; the spectabile female ceased the twitching movement and the eye resumed its usual color. Tables 5 and 6 summarize the specific behavioral differences of 15. caeruleum and

IS. spectabile and Figure 27 illustrates the ideal sequence of pre-

spawning and spawning events. The spawning behavior in either species

cannot be considered elaborate when compared to complex ritualistic

courtship as observed, for example in the Gasterosteidae (sticklebacks),

(Sevenster, 1968; Morris, 1958) or the Anabantidae (bettas, paradise

fish), (Miller and Hall, 1968; Miller and Miller, 1970; Barlow, 1968).

Thirty-five IS. caeruleum and 23 IS. spectabile matings were recorded

on film (multiple male matings were counted as a single spawning). Many more were observed, though not filmed in the laboratory. Approximately

50 matings of each species were observed in the field. Four cross­ matings were filmed in the laboratory involving a male spectabile and a

female caeruleum; six others were observed though not filmed. Four

Interspecific matings involving a male spectabile and female caeruleum were observed in the field in marginal habitat which was optimal breeding habitat for neither species. A single attempted mating between a male caeruleum and a female spectabile in the laboratory was

thwarted by replacement by a conspecific male. No male caeruleum- female spectabile matings were observed in the field. It is probable

that the scarcity of space and the combination of approximately equal numbers of each species accounted for a certain amount of "unnatural" behavior in the laboratory. Whether any of the interspecific matings resulted in viable fertilization was not determined. 87

TABLE 6.

COMPARISON OF SPAWNING OF E. CAERVLEUM AND E, SPECTABILE.

E . caeruleurn E. spectabile

Habitat Riffle Head of riffle

Micro-habltat Flat portion, pebble, Slope from pool to cobble substrate riffle, sand, fine- gravel bar

4 Male aggressive (sham battles) . display

a) coloration 8—9 vertical bars 3—4 blotches of intense posterior trunk and caudal region intense

b) attitude caudal touching caudal waved above substrate substrate

c) territoriality "moving", around "moving", around female, one side of female cobble or boulder

Female orientation Near cobble, boulder Open area ? or depression

Visual cues Another female none already buried

Female receptivity indicators

Color Dark saddle bands Dark eye spot

Behavior "Feeding" "Twitching" Agitated movement Agitated movement

Actual spawning No detectable difference Hale Female

Sham battles Presence

Presence Agitated movement

(Dark saddle bands)

(Eyespot)

v Twitching1 Following

Swimming over

touching, nudging<

Diving

Burying

Territorial defense

(around female)

Mounting Vibrating

Vibrating

Spawning

Emergence aE. caeruleum behavior

^E. spectabile behavior

Stimulus------>■ Response

Fig. 27.— Ideal sequence of events in pre-spawning , and spawning of IS. caeruleum and JE. spectabile. 89

Hybridization

* -r

The difficulty of correctly identifying hybrids between j5.

caeruleum and _E. spectabile lends support to the evidence already

accumulated that these two species are closely related. It is not

illogical to hypothesize a frequent occurrence of hybrids between the

closely related species, but immediate, conclusive identification of

such natural hybrids has been extremely difficult. This difficulty is

not circumvented or alleviated by observations on hybrid intermediacy

made by Hubbs (1955):

It has proved to be an almost universally valid rule that natural interspecific hybrids are intermediate between their parental species in all characters in which those species differ, whether they be external or internal, of shape, color, form, structure, or numbers of parts (vertebrae, gill- rakers, fin rays, teeth)— except for some features that reflect hybrid vigor.

While a hybrid may be an overall, or average, blend of parental

characters, over-reliance on purely Mendelian inheritance, that is,

expecting general blending of parental traits resulting in recognizable

Intermediacy in the hybrid or backcross, is often disappointing in that

the intermediacy may seem considerably less obvious than desired by even

the subjective analyst.

Forced hybridization techniques (stripping) provide a test to eli­ minate speculation or conjecture regarding a putative hybrid (Hubbs and

Hubbs, 1932). Experiments show, however, that laboratory-raised hybrids 90

sometimes deviate from the mathematically and even genetically predicted

blend of parental traits to the extent that they often exceed in degree

the expression of particular inherited traits. The adverse effects of

sub-optimal developmental environmental conditions, whether as diet,

thermo—regulation, photoperiod, or oxygen-carbon dioxide levels are

difficult to evaluate even by the use of controls. Linder (1958) and

Strawn (1961) have shown that both hybrids and controls of species crossed exhibit variances, especially in lateral line scales and un­ paired fin rays, that exceed those of wild specimens. Similar results were found in seven artificially induced hybrids from _E. caeruleum and

E. spectabile parents. Of 57 ineristic and morphometric characters, 40 of which were length measurement ratios, the variance of hybrid characters exceeded that of the J2. caeruleum parent in nine cases and of the E. spectabile parent in eight cases (Table 7). In another cross, a single hybrid between sympatric E. caeruleum and 12. spectabile, the hybrid, of 57 characters, resembled its paternal and maternal parent in twelve and eleven cases respectively, was intermediate in seven, exceeded its caeruleum parent in five characters, and its spectabile parent in eleven characters. The parents were similar to each other and to the hybrid in five characters. The variance of characters of the population was not determined, but in regard to the parents alone, the hybrid was aberrant in the (reduced) number of pored lateral line scales and in the (abnormally high) ratio of the length of the soft dorsal ray to the length of the depressed soft dorsal.

More reliable conclusions on character Inheritance can be drawn from sample populations of the parent stock and the seven hybrid fish, 91

TABLE 7* CHARACTERS IN WHICH VARIANCE OF LABORATORY-RAISED HYBRIDS EXCEEDED VARIANCE OF PARENTAL STOCK

Hybrid character variances exceeding that of both, E« caeruleum and E. spectabile parents,____ Head depth/SL Upper jaw length/SL Body depth/SL Point of maximum body depth/SL Caudal peduncle length/SL Head length/bistance to point of maximum body depth Head depth/Head length Head width/Head depth UGM distance/Head length Body depth/Predorsal length Body depth/Head length Infra-orbital bar index Number of dorsal spines Number of lateral bars Number of pored lateral line scales Number of scales above lateral line Number of scales below lateral line UGH to maxillary tip/SL Hybrid character variances exceeding that of the caeruleum parent.

Upper jaw length/SL Occiput to dorsal origin/SL Vent to caudal peduncle/SL Head width/Head length Length of pectoral fin/Length of depressed soft dorsal Length of soft dorsal ray/Length of depressed soft dorsal Number of lateral line scales Nape scalation index Number of post—lacrimal pores hybrid character variances exceeding that of the £. spectabile parent. Head width/SL Interorbital distance/SL First dorsal base/SL Head depth/body depth Predorsal length/Preanal length Number of soft dorsal rays Urogenital pigmentation index Number of lacrimal pores 92 even though they are of allopatric rather than sympatric origin.

The male parent is JS. caeruleum taken from Mac-o—chee Creek, a

tributary of the Mad River in Logan County, Miami drainage. The female

JS. spectabile parent was taken from Peter*s Run, a prairie stream in

Pickaway County, Scioto drainage. _E. caeruleum does not occur in this

small stream. Techniques of stripping were according to Strawn and

Hubbs (1956). Temperature was regulated (16-20° C.) only until hatch­

ing. Thereafter, care in raising the larvae and young was as described

in Methods and Materials.

Meristic and Morphometric Analysis

One principal tool of systematic taxonomy is a statistical evalua­

tion of meristic and morphometric characters. Through its use ichthyol­

ogists can approach the problem of the phylogeny of fishes. In the past

the traditional technique of comparison of meristic characters was used

to draw inferences regarding the relationship of particular species

(Ouellette and Qadri, 1968). The present study employs these tradi­

tional methods and supplements them with an added analytical approach.

By the use of discriminant analysis, utilized effectively by Hill

(1959) in classifying races of American shad and by Stone (1947) and % Cole (1965) on species of the subgenus Boleosoma, it was possible to

establish a multiple-character specimen index. As described by Stone

(1947), "each measurement is multiplied by a factor and the products of measurements and factors are added, yielding a single specimen index."

The technique closely resembles multiple regression and is essentially

a series of analyses of variances. Its most common use is in the 93

handling of taxonomic data (Mather, 1951). With the exception of

Nelson's (1968) use of this method on the catostomlds C. commersoni

and C. macrochellus In Canada and the above mentioned works, Its use

In lchthyological taxonomy seems rather limited.

Since the artificially induced series of "lab" hybrids were con­

sidered to constitute a separate group, they were classed as such in

the first series of discriminant function analysis (Fig. 28). Initially, with all 57 characters, and later with the more discriminating charac­

ters, the final print-out, based on six characters, was as illustrated

in Figure 28. The characters selected, from most to least discrimi­

nating are as follows:

Infra-orbital bar index Number of lateral bars Number of pectoral rays UGM to snout distance/SL Head length/Distance—point of maximum body depth Length of longest soft dorsal ray/Depressed soft dorsal fin

More discriminating between species than any of those characters mentioned, is the completeness versus the incompleteness of the infra­ orbital canal. However, since this character was consistently the same

(without variance), that is, always complete in JE. caeruleum and always incomplete in E. spectabile, it could not be used in discriminant analysis, which is based on the use of analysis of variance. In all laboratory hybrids the canal was incomplete. In one putative natural hybrid the canal was complete on one side of the head and incomplete on the other.

In Figure 28 the ordinate and abscissa values are those determined by computer analysis. A secondary abscissa (X-axis) is introduced to ■ P * 0

np 1.3.8 1

- +2 1r M ■F M T M P * P T ; - 0.0 p t ; * fl 11 P P p ♦ i i » 4 --2 P T 1 t T T

P - IE. spectabile; Peter's Run, Pickaway Co,, Scioto River drainage, T - 'True" Laboratory-raised hybrids; caeruleum (Mac-o-chee Creek), X E, spectabile (Peter's Run), M - E, caeruleum; Maoo-chee Creek, Logan Co,, Miami River drainage, * - Group mean,

-7.5 -4.5 -1.5 +1.5 +4.5 +7.5 Fig. 28.— "Three group" discriminant analysis on six characters of laboratory-raised hybrids ^ and parental stock. Horizontal line within graph indicates hybrid index scale. Mean hybrid index * value is 43.8. 95 convert the graphical description to the more popular hybrid index

determinations according to Hubbs and Kuronuma (1942) and employed by

Hubbs, Hubbs, and Johnson (1943), Trautman (1948) and Linder (1955) on

suckers, catfish, and darters respectively. Using the mean value of

characters of L caeruleum as 100 and the mean value of IS. spectabile

as 0 (zero), a hybrid index can be obtained for each specimen as well

as a mean hybrid index for the group. The mean hybrid index for this

group is 43.8, indicating a slight overall leaning toward IS. spectabile.

The parental stock of the "lab" hybrids is represented by a sample * of ten of each species (six males and four females) Including the

parents themselves.

One of the purposes in analyzing confirmed hybrids was to establish

at least the semblance of a norm or standard by which to judge sus­

pect ive natural hybrids. After determining which characters had abnor-

j t mally high variance in laboratory specimens, these characters were elim­

inated from the analysis of both "lab" and "stream" (putative) hybrids.

The characters so involved were lateral line scales, pored lateral line

scales, and unpaired fin rays. Since none of these characters proved

to be significantly discriminating, neither the analysis nor the hybrid determination was in any way impaired.

Fifteen putative hybrids were taken from Dry Run, the principal study stream. Speculation that these fish might be hybrids was based primarily on color and color patterns and body shape. As measurement of various characters proceeded, it became more and more evident that half of these were backcrosses, possibly F^ individuals from previous hybridization or bona fide "pure" E. spectabile or 12. caeruleum with 96 slight variation in one or more characters. They were possibly even aberrants— individuals deviating significantly from the norm in one or more characters. Still, none of the original 15 suspects was deleted in comparison with the confirmed hybrid standard and with a sample of the parental stock, 33 JS. spectabile and 34 IS. caeruleum, from the same stream.

The determination of hybrids was based upon the theory that the space discriminating the gene pools, expressed phenotypically by the meristic and morphometric characters of 13. spectabile and 13. caeruleum, is sufficiently wide to allow the occupation by hybrids and/or back- crosses between hybrids and the two species. Figure 29 shows the dis­ criminant analysis print-out when considering all fish in the group as one of two species.

Figure 30 Illustrates the results of simultaneous discriminant * function analysis of six groups of fish: confirmed hybrids and their parental stock and putative hybrids and their presumed parental stock.

Even though the first three groups contain known hybrids, the analysis was run as four groups, two groups of _E. caeruleum and two groups of

13. spectabile and all hybrids (confirmed and putative) considered as unknowns.

Using the confirmed hybrids as limits, seven of the putative hybrids seem to be valid natural hybrids. (See "Y" subscripts—

Fig. 30.) This does not mean that some of the other suspects are not hybrids, nor that one of the other “pure" stock specimens was not a hybrid or backcross (see E. caeruleum specimen (R) at lower right- center of Fig. 30.) OSUM catalog numbers of hybrid specimens are ♦3.9-

♦2.7-

0 0 0 +1.5- °o £ 2 00.0 o R ~ofl - 1 1 - 0 ■» R 0 0 R R R R R R R r R 0 0° R r -0.9- H

R

-2.1-

-3.3-

0 - E. spectabile; H - Putative hybrids R - E, caeruleum -4.5-

—4.4 -2;2 "oTcf +2.2 "+4.?4 " ® T Fig. 29.— Discriminant analysis of E. caeruleum, E, spectabile, and putative hybrids from Dry Run on six characters. Horizontal line indicates hybrid index scale. Only the seven "central" putative hybrids were used to determine the mean hybrid index— 53.1 (larger asterisk). The overall mean (including all specimens) "accidentally" (?) falls on the mid-point (50.0) of the hybrid index scale. +3.5

D +2,0 p M M

P 0 M Yl R M * Y - R D D 0 Y2 0 R ' R D 00 tf T R H * T R R D o P D n R R H 0 # T R Y R R - 1.0 D Do 0 dd D p R * R R M Y Y3 R D Y4 R j. p ° “ p w R r R Do T R R R “2.5 R R R d D D y5 F 0 Yo Y R

-4,0 • Y7 R Y Kean

D - E. spectabile; Dry Run T - Laboratopy-raised hybrids R - E. caeruleum; Dry Run “5*5" P - E. spectabile; Peter’s Run Y - Putative hybrids; Dry Run M - E. caeruleum; Mao-o-chee Creek ” ... — .. ■ i ■■ . — i — - i —..... i . **i “7*2 -5*1 -3*0 -0.9 + 1.2 +3.3 + 5*4 Fig. 3 0 . --Simultaneous discriminant analysis of six groups of fish. " Y " subscripts ( 1 - 7 ) used to indicate specimens used to find mean hybrid index value. See Figure 3 0 . \o CO 99

listed in Appendix B.

Figure 29 also shows the conversion of hybrid index values using

only those specimens that fall within the range of the "true" hybrid

index values established (see Fig. 30). The mean hybrid index of the

seven putative hybrids is 53.1.

Color characters were not involved in determining hybrids because

some of the specimens faded prior to the time of their use in this

analysis.

The vertical bars of _E. caeruleum are quite regular, while those

of IS. spectabile are more irregular blotch-like patches. IS. spectabile

has a series of broken horizontal lines below the nape and especially

below the first of the two dorsal fins. Specimen Yc (Fig. 30) has both

the vertical bars and the broken horizontal pigmentation. In addition,

besides the spectabile green, it has the slightest amount of red pig­

mentation on the anal fin, a character never found in JS. spectabile.

Therefore, with the added use of color characters, specimen Y 0 might

be considered a hybrid as validly as the seven falling within the

"confirmed hybrid standard." Only one other hybrid suspect has red on

the anal fin, that being specimen Y 3 . Completeness and incompleteness

of the Infraorbital canal was found in only one putative hybrid,

specimen Y 2 *

Neither pattern nor degree of pigmentation of unpaired fins was used as a morphometric character since the writer believed that the

tremendous variation and distribution of color based on sex, size, and maturation of specimens studied, did not warrant its use either as a discriminating or intermediating character. 100

In addition to evidence gleaned from discriminant analysis, it was determined that the mean of the seven natural hybrids, as a group, was intermediate between that of the parental species In 16 other

characters (Table 8 ). Overall, of the 57 characters examined, the natural hybrids were intermediate in 16 characters, and resembled their caeruleum and spectabile parentage in 14 and 27 cases respectively

(Tables 9 and 10). Considering all factors, the hybrids tend to resemble J£. spectabile more than _E. caeruleum. Branson and Campbell

(1969), in studying hybrids between IS. spectabile and IS. radiosum, had similar findings, that the hybrids resembled E. spectabile more than

IS. radiosum.

The data indicates that the following characters are the ones best suited for the field determination of hybrids: (1 ) the mosaic of pig­ mentation of both species expressed in a single individual as a com­ bination of both vertical bars (of E. caeruleum) and broken horizontal lines (of IS. spectabile)j (2) an irregular "stair-step" pattern; and

(3) intermediate body shape. (See Figs. 14, 15, 22, 23, and 24 for

"typical" color patterns.)

In most female E. caeruleum a fairly regular and repetitious

"stair-step" pigmentation pattern is manifested in the anterior lateral trunk region (Fig. 31). In the hybrid, whether male or female, it appears that this design is highly irregular, sometimes broken, some­ times forming a continuous curving, series of meandering lines.

Figure 32 illustrates this pattern in one of the stream (natural) hy­ brids and it was observed in all artificially induced, laboratory- raised hybrids (Fig. 33). It was either this erratic, mosaic pattern Fig. 31.— "Stair-step" pigmentation pattern of E. caeruleum female.

Fig. 32.— Pigmentation pattern of stream hybrid male.

Fig. 33.——Pigmentation pattern of laboratory- raised hybrid male (imm.). Actual size is one-half that of specimens in Figs. 31 and 32. 102

TABLE 8.

CHARACTERS IN WHICH LABORATORY-RAISED HYBRIDS ARE INTERMEDIATE BETWEEN THEIR PARENTS

UGM to maxillary tip/SL Point of maximum body depth/SL First dorsal base length/SL Depressed anal fin length/SL Distance— vent to caudal peduncle/SL Head length/Distance— point of maximum body depth Head depth/Body depth Snout length/Upper jaw length UGM dis'tance/Head length Longest soft dorsal ray/Length of depressed soft dorsal Snout length/SL Number of dorsal spines Number of lateral bars Number of pectoral rays Number of opercular scales Nape scalation index 103

TABLE 9.

CHARACTERS OF NATURAL HYBRIDS RESEMBLING THEIR E, CAERULEUM PARENTS.

Upper jaw length/SL UGM to maxillary tip/SL Occiput to dorsal origin/SL Predorsal length/SL Longest soft dorsal ray/SL Length of longest dorsal spine/Longest dorsal ray Predorsal length/Pre-anal length Caudal* peduncle length/Caudal peduncle depth Caudal peduncle length/Distance— vent to caudal peduncle Number of soft dorsal rays Number of lateral line scales Number of scales above lateral line Urogenital pigmentation index Number of lacrimal pores 104

TABLE 10.

CHARACTERS OF NATURAL HYBRIDS RESEMBLING THEIR E. SPECTABILE PARENTS.

Head length/SL Head depth/SL Head wldth/SL Interorbital dlstance/SL Body depth/SL Pectoral fin length/SL Longest dorsal spine/SL Depressed soft dorsal/SL Pre-anal length/SL Depressed anal fin length/SL Caudal peduncle depth/SL Head length/Caudal peduncle length Head depth/Head length Head width/Head depth Head width/Head length Snout length/Head length Interorbital distance/Head width Body depth/Predorsal length Body depth/Head length Length of pectoral fin/Length of depressed soft dorsal Length of first dorsal base/Length depr. soft dorsal Length longest dorsal spine/Length spinous dorsal base Length of depressed anal fin/Length depr. soft dorsal Number of pored lateral line scales Number of scales below lateral line Infra-orbital canal (complete vs. incomplete) index Post—lacrimal pores 105 or the combination of lateral bars and horizontal broken (seemingly dotted) lines that served as initial clues to field identification of a hybrid between IS. caeruleum and E. spectabile. Intermediacy of body shape between the more terete spectabile and the more robust caeruleum may also lend support to preliminary identification. 106

Intraspeci£ic and Interspecific Character Differences

The Student t-test was used to compare the means of meristic

characters of E. caeruleum and E. spectabile of the same sex to deter­ mine specific differences. Tables 11 and 12 list the significant dif­

ferences between JS. caeruleum and jS. spectabile males and females

respectively. The differences in characters between males and females of the same species was tested and the results are listed in Tables 13 and 14 for 13. caeruleum and E. spectabile respectively. To determine what differences, if any, existed between 12. caeruleum taken from two different drainages, those specimens from Dry Run, Scioto River drain­ age, were compared with E. caeruleum from Mac-o-chee Creek, Miami River drainage. These allopatric conspecific populations were found to differ significantly in 15 characters (Table 15). (None of these character differences were of the magnitude to make them morphologically obvious.

Those from Mac-o-chee Creek did have a more intense pigmentation typical of E. caeruleum taken from higher gradient streams.) This phenomenon is also illustrated in Figure 30 (for six characters) where the Mac-o-chee Creek E. caeruleum (M) are slightly shifted from those taken at Dry Run (R). Perhaps these two populations are descendants of a previously single ancestral population prior to the Illinoian and

Wisconsin glaciation which purportedly caused disruption of the once westward flowing Teays River. Subsequent to glacial recession, two separate southerly flowing drainages were formed: the Scioto and the TABLE 11. SIGNIFICANT DIFFERENCES BETWEEN E. CAERULEUM AND E, SPECTABILE MALES.

Character E. caeruleum E. spectabile t-value Mean (N=15) Mean (N=19)

Head depth/SL * 0.189 0.176 -5.6548 Head width/SL * 0.155 0.135 -8.2898 Snout length/SL * 0.072 0.065 -3.8636 Upper jaw length/SL 0.085 0.082 -2.1071 Distance— UGM to maxillary tip/SL * 0.157 0.123 -7.8134 Body depth/SL * 0.216 0.199 -5.4659 Predorsal length/SL * 0.364 0.348 -4.5278 Caudal peduncle depth/SL 0.104 0.100 -2.0816 Head length/Caudal peduncle length 2.934 3.087 2.5403 Head depth/Head length * 0.622 . 0.573 -5.4299 Head width/Head depth * 0.820 0.769 -3.9995 Head width/Head length * 0.510 0.440 -8.1702 Snout length/Head length * 0.235 0.211 -4.7379 Snout length/Upper jaw length * 0.845 0.796 -2.9382 UGM distance/Head length * 0.518 0.401 -8.8053 Body depth/Predorsal length 0.595 0.573 -2.4877 Body depth/Head length * 0.714 0.648 -5.8370 Longest soft dorsal ray/Length depr. s. dors. 0.511 0.547 2.5869 Predorsal length/Pre-anal length 0.571 0.554 -2.0395 Length of depressed anal/Length depr. s. dors. * 0.818 0.873 3.0070 Infraorbital bar index * 0.400 4.526 21.0195 Number of lateral bars * 9.800 5.684 -8.9674 Number of pectoral rays * 25.467 23.368 -6.4763 Number of pored lateral line scales * 30.667 25.789 -3.7386 Number of scales above lateral line 4.933 4.579 -2.4604 Urogenital pigmentation index * 6.200 0.263 -10.4798 107 Post-lacrimal pores * 7.733 6.789 -3.1495 i TABLE 12.

SIGNIFICANT DIFFERENCES BETWEEN E. CAERULEUM AND E. SPECTABILE FEMALES.

Character E. caeruleum E. spectabile t-value Mean (N=19) Mean (N=14)

Head width/SL * 0.144 0.133 -3.8875 Snout length/SL * 0.068 0.063 -3.5478 Distance— UGM to maxillary tip/SL * 0.156 0.130 -6.7817 Body depth/SL * 0.207 0.195 -2.7747 Caudal peduncle depth/SL * 0.104 0.099 -3.2858 Head length/Caudal peduncle length * 2.861 3.026 3.1754 Head width/Head depth * 0.800 0.759 -3.6128 Head width/Head length * 0.484 0.444 -3.7895 Snout length/Head length * 0.231 0.211 -3.7017 Snout length/Upper jaw length * 0.861 0.793 -3.3578 UGM distance/Head length * 0.524 0.436 -6.8345 Body depth/Predorsal length * 0.583 0.544 -3.0566 Infraorbital bar index * 1.500 4.143 7.7414 Number of soft dorsal rays 12.900 12.429 -2.2838 Number of lateral bars * 10.300 5.214 -17.3829 Number of pectoral rays * 25.300 23.500 -5.3615 Number of lateral line scales * 44.450 42.571 -2.7344 Number of pored lateral line scales * 30.150 24.286 -5.8656 Urogenital pigmentation index 1.700 0.000 -2.4566 TABLE 13* SIGNIFICANT DIFFERENCES BETWEEN E. CAERULEUM MALES AND FEMALES.

Character Males Females t-value Mean (N=15) Mean (N=19)

Head length/SL 0.303 0.297 -2.0872 Head depth/SL * 0.189 0.180 -3.2003 Head width/SL 0.155 0.144 -3.8512 Upper jaw length/SL * 0.085 0.080 -3.2850 Pectoral fin length/SL * 0.270 0.252 -3.8055 Depressed soft dorsal fin length/SL * 0.304 0.280 -3.8780 Predorsal length/SL 0.364 0.355 -2.6277 Depressed anal fin length/SL 0.248 0.236 -2.3292 Head width/Head length * 0.510 0.484 -2.7949 Length spinous dorsal fin base/Length depr. s. dors. 0.961 1.028 2.6469 Nape scalation index 2.067 2.800 2.0518 Urogenital pigmentation index 6.200 1.700 -5.3171

o to TABLE 14. SIGNIFICANT DIFFERENCES BETWEEN E. SPECTABILE MALES AND FEMALES*

Character Males Females t-value Mean (N=19) Mean (N=14)

Occiput to dorsal origin/SL * 0.142 0.153 3.4515

Longest dorsal spine/SL * 0.129 0.117 -2.7717

Length— depressed soft dorsal fin/SL * 0.292 0.274 -3.0440

Length— depressed anal/SL * 0.255 0.239 -2.8945

Longest soft dorsal ray/SL 0.160 0.151 -2.0819

UGM distance/Head length * 0.401 0.436 2.8843

Body depth/Predorsal length 0.573 0.544 -2.6691

Number of lateral line scales 44.105 42.571 -2.2022 TABLE 15.

SIGNIFICANT DIFFERENCES BETWEEN E, CAERULEUM OF DRY RUN, SCIOTO RIVER DRAINAGE AND MAC-O-CHEE CREEK, MIAMI RIVER DRAINAGE.

Character Dry Run Mac-o-chee Ck. t-value Mean (N=34) Mean (N=10)

Head length/SL * 0.300 0.284 4.5078 Head depth/SL * 0.184 0.199 -4.4053 Snout length/SL 0.070. 0.075 -2.6689 Body depth/SL * 0.211 0.241 -7.3105 Pectoral fin length/SL * 0.260 0.283 -4.2900 Depressed soft dorsal/SL 0.290 0.309 -2.2774 Predorsal length/SL 0.359 0.351 2.2888 Pre-anal length/SL 0.636 0.605 2.4560 Head length/Caudal peduncle length * 2.892 2.652 4.2514 Head depth/Head length * 0.613 0.701 -7.1960 Head depth/Body depth * 0.870 0.826 3.0824 Head width/Head depth 0.809 0.777 2.5543 Number of dorsal spines 9.886 9.500 2.0326 Number of opercular scales 13.314 10.800 2.3131 Nape scalation index * 2.486 3.600 -2.9069

H H H 112

Miami River systems (Stout and Lamb, 1938; Ver Steeg, 1946).

Tests for significant differences between E. caeruleum and 13. spectabile revealed a surprising number of differences, considering their similarity in appearance (Table 16).

The Snedecor-Fisher F-test and analysis of variance indicated that the variance of characters of JE. spectabile taken from Dry Run exceeded those taken from Peter*s Run, where JE. caeruleum does not occur, in 43 of 57- cases. This fact suggests the occurrence of introgression between

E. caeruleum and 12. spectabile where they occur sympatrically, as Clark

(1973) found in Indiana. In a study of sympatrlc versus allopatric

_E. caeruleum and _E. spectabile, he found that in allopatry 11 characters would separate at least 80 per cent of the specimens. Of these, two characters were found which would separate all individuals of the two species. Where the species were sympatrlc, he found only seven charac­ ters that separated 80 per cent of the individuals and only one, com­ pleteness or Incompleteness of the infra-orbital canal, which would separate all individuals.

Tests for significant differences between laboratory-raised hybrids and their parental stock were performed and indicated that, overall, the hybrids resemble their spectabile more than their caeruleum parents

(Tables 17 and 18). These findings agree with those indicated in

Tables 9 and 10 listing the characters in which the hybrids resemble their parental stock. TABLE 16. SIGNIFICANT DIFFERENCES BETWEEN E. CAERULEUM AND E. SPECTABILE TAKEN FROM DRY RUN.

E. caeruleum E. spectabile - Character t Mean (N=34) Mean (N=33)

Head depth/SL * 0.184 0.176 -4.2411 Head width/SL * 0.148 0.134 -6.9593 Snout length/SL * 0.070 0.064 -4.8895 UGM to maxillary tip/SL * 0.156 0.126 -10.5208 Point of max'm body depth/SL * 0.389 0.353 -4.1146 Body depth/SL * 0.211 0.197 -5.0012 Pectoral fin length/SL 0.260 0.268 2.1899 Caudal peduncle depth/SL * 0.104 0.099 -3.8113 Head length/Caudal peduncle length * 2.892 3.061 4.2639 Head depth/Head length * 0.613 0.578 -4.8030 Head depth/Body depth 0.870 0.890 2.3568 Head width/Head depth * 0.809 0.765 -5.1210 Head width/Head length * 0.495 0.442 -7.6266 Snout length/Head length * 0.233 0.211 -5.9733 Snout length/Upper jaw length * 0.854 0.795 -4.6197 UGM distance/Head length * 0.521 0.416 -11.1085 Body depth/Predorsal length * 0,588 0.561 -3.4480 Body depth/Head length * 0.705 0.651 -5.3827 Length longest dorsal spine/Length— Bp. dors, base 0.418 0.443 2.1150 Length soft dorsal ray/Length— depr* soft dors. 0.522 0.549 2.5035 Length of depressed anal/Length— -depr. soft dors. * 0.833 0.873 3.3769 Infra-orbital bar index * 1.029 4.364 15.3950 Number of soft dorsal rays * 12.886 12.455 -3.0971 Number of lateral bars * 10.086 5.485 -16.9101 Number of pectoral rays * 25.371 23.424 -8.5432 Number of pored lateral line scales * 30.371 25.152 -6.4229 Number of scales below lateral line 7.714 7.394 -2.2354 Urogenital pigmentation index *______3.629 0.152 -5.9223 TABLE 17. SIGNIFICANT DIFFERENCES BETWEEN LABORATORY-RAISED HYBRIDS AND THEIR E. CAERULEUM PARENTS.

Character Hybrids E. caeruleum t-value Mean (N=7) parents Mean (N=10)

Head depth/SL * 0.176 0.199 3.5481 Snout length/SL * 0.068 0.075 3.0692 Body depth/SL * 0.211 0.241 5.6429 Pectoral fin length/SL * 0.254 0.283 4.7380 Predorsal length/SL * 0.367 0.351 -3.7587 Caudal peduncle length/SL * 0.213 0.247 4.7889 Caudal peduncle depth/SL * 0.083 0.108 7.4160 Head length/Caudal peduncle length * 3.687 2.652 -10.6232 Head depth/Head length * 0.573 0.701 5.1749 Head width/Head length * 0.467 0.544 5.1685 Snout length/Head length * 0.223 0.264 4.8402 Snout length/Upper jaw length * 0.839 0.916 5.5436 Body depth/Predorsal length * 0.622 0.688 6.4007 Body depth/Head length * 0.689 0.850 6.5417 Length longest dorsal spine/Longest d. ray . * 0.864 0.773 -4.7315 Caudal ped. length/Caudal ped, depth * 2.556 2.304 -3.3136 Infra-orbital bar index * 2.714 0.300 -6.1467 Number of soft dorsal rays * 14.429 13.300 -3.6783 Number of pectoral rays 24.714 25,800 2.2008 Number of lateral line scales * 39.571 44.200 5.9548 Number of pored lateral line scales * 15.857 28.800 6.6716 Number of scales above lateral line 4.571 5.100 2.5691 Number of scales below lateral line * 7.000 8.000 3.9295 Number of lacrimal pores 7.429 8.000 2.3302 Post-lacrimal pores 6.143 7.900 10.4172 UiH TABLE 18. SIGNIFICANT DIFFERENCES BETWEEN LABORATORY-RAISED HYBRIDS AND THEIR E. SPECTABILE PARENTS.

Character Hybrids E. spectabile t-value Mean (N=7) parents Mean (N*10)

Head length/SL * 0.307 0.291 3.4045 UGM to maxillary tip/SL 0.150 0.130 2.4911 Pectoral fin length/SL 0.254 0.269 -2.2031 Predorsal length/SL * , 0.367 0.349 3.5936 Caudal peduncle length/SL * 0.213 0.242 -3.9761 Caudal peduncle depth/SL * 0.083 0.107 -7.0676 Longest soft dorsal ray/SL * 0.141 0.165 -4.0879 Head length/Caudal peduncle length * 3.687 2.745 8.8145 Head width/Head length 0.467 0.499 -2.8933 Body depth/Predorsal length 0.575 0.622 -2.9052 Body depth/Head length 0.689 0.744 -2.7027 Length of soft dors, ray/Length— depr. s. dors. * 0.489 0.563 -4.2084 Length longest dors. spine/Longest dors, ray * 0.864 0.724 4.4572 Predorsal length/Pre-anal length * 0.582 0.553 3.5094 Caudal peduncle length/Caudal peduncle depth 0.128 2.284 2.3719 Caudal peduncle length/Vent to caudal ped. * 0.538 0.612 -3.5325 Infra-orbital bar index * 2.714 4.400 -4.2258 Number of soft dorsal rays * 14.429 12.500 7.3831 Number of lateral bars 11.286 15.200 -2.6279 Number of lateral line scales * 39.571 44.500 -5.5783 Number of scales above lateral line * 4.571 5.400 -3.2103 Number of scales below lateral line 7.000 7.600 -2.2480

H H 116

Chromosome Morphology

In addition to using chromosome morphology to Indicate species

relationship (Roberts, 1967), karyotyping (determination of the number

and shape of chromosomes) has been used as a clue to compatibility of

potentially parental species and evidence of predictability regarding

sterility in the hybrid product (White, 1954). With the development of

evolutionary cytogenetics, chromosome techniques have become more

refined and methods of obtaining and preparing suitable tissues have

been elaborated (McPhail and Jones, 1966; Denton and Howell, 1968;

Sharma and Sharma, 1972).

In the present study, karyological techniques were used to cor­

roborate already considerable evidence of the proximal taxonomic rela­

tionship of E. spectabile and E. caeruleum. Techniques of preparation were according to methods described in McPhail and Jones (1966) on both

parental species and laboratory raised hybrids. In Figure 34 the

karyotypes were formed by arranging the chromosomes somewhat arbi­

trarily by length. Each species had a diploid complement of 48 chromo­

somes. No significant heteromorphism was observed regarding sex

chromosomes. The karyotypes illustrated are typical of _E. caeruleum

and E. spectabile taken at various locations in Ohio.

15. caeruleum exhibits four (two pairs) rather clearly defined metacentric chromosomes. In IE. spectabile and the hybrid, metacentric chromosomes are not obvious. Ross (1973) discerned one pair of sub— metacentric chromosomes in E_. caeruleum. In the current study, with the exception of E. caeruleum, separation of metacentrics, Stheostoma apectabile & 2n ■ US 117 XXJtXAA A fl AX X* X AJl A 1UAAAXAHAAXHAAAA A/I/I AAll/l/cii

Stheostoma caeruleum x Etheostoma apectabile & 2n * US Xfixxft/innjuiAAxiUkii h/iJLAtt/ij<.iinn/*>KaAXA X A /iK/ion n /v —.k xn*%

Etheostoma caeruleum & 2n ■ U8

??7f)CK AA AAAJUW\. AAAA/tn

w * v A ^ O t * A

> i 5 . w Fig. 34.--Karyotypes of males of E. spectabile, caeruleum x jS. spectabile hybrid and IS. caeruleum. 118 submetacentrics, subtelocentries and acrocentrlcs was not performed.

In E. caeruleum there are two pairs of rather obvious metacentries— listed first in the karyotype (Fig. 34).

The Individual chromosomes making up the largest pair are unevenly matched in both E. caeruleum and E. spectabile, but seemed more exag­ gerated in the latter. In the JS. caeruleum-E . spectabile hybrid, this pair is more evenly matched. There is little disparity suggested in the hybrid karyotype regarding the morphometry and comparative size of homologues from the parental species. This information suggests that the F^ hybrid would be fertile and that introgresslon could occur where the two species are sympatrlc. This hypothesis is borne out in the greater variance of characters in fish taken from Dry Run, than Peter’s

Run, which are sympatrlc and allopatric respectively, regarding the occurrence of E. caeruleum and IS. spectabile. DISCUSSION

Etheostoma spectabile and E. caeruleum are members of the same

subgenus, Oligocephalus, (Bailey and Gosline, 1955; Collette, 1963),

they have similar breeding behavior, are often sympatrlc in occurrence,

and their breeding sites are adjacent to, or even overlap, one another.

Under these conditions hybridization is normally more likely to occur.

However, extensive hybridization of _E. spectabile and E. caeruleum is

not the norm in the streams of central Ohio.

A situation somewhat in contrast to the findings regarding E.

spectabile and JS. caeruleum has occurred between IS. radiosum and JS.

spectabile in the Blue River system of Oklahoma (Linder, 1958; Distler,

1968; Branson and Campbell, 1969). The most recent follow-up of this

study is that being done by Echelle et al (in manuscript) who suggested

considerable hybridization: "occurring In virtually ‘every Blue River

locality where spectabile and radiosum [members of the same subgenus]

are together." The resulting introgression, whether originating from

E. radiosum or IS. spectabile crossbreeding, is manifested in "popu­ lations" having a variance of characters exceeding that of the parental stock, a phenomenon found in the present study of E. spectabile and

E. caeruleum. Similar results were found in a number of darter hybrids of interspecific crosses (Hubbs and Strawn, 1957). In hybrids of the minnows Notropis lepidus x N. proserpinus (Hubbs, 1956) also found 120 considerably more variability than In parental controls. Hubbs and

Strawn (1957b) found F^ hybrids of IS. spectabile x Percina caprodes

"much more variable than their parental species In most taxonomic characters and grade into the parental types In those characters which vary most. Not only are these hybrids more variable, but also they are somewhat extreme to the parental species in some taxonomic characters."

(Hubbs, 1956). He suggested three possible causes for such increased variability of hybrids: previous introgression, resulting in exchange of chromosome segments, modifying genes Influencing the meristic characters, and, variation in developmental rates. It is not likely that differences in developmental rates played an important part in the large variation of F^ hybrids between E. caeruleum and _E. spectabile.

Some hybridization between E. caeruleum and 12. spectabile does occur. Of greater interest is, what enabling or facilitating factors allow its occurrence and what factors prevent or inhibit the gene flow and help maintain the relative integrity of the respective species.

Factors Favoring Hybridization

A pre-requisite for the occurrence of hybridization is the suf­ ficiently proximal phylogenetic relationship of the species involved.

Second, mechanisms must exist to enable the gametes of the species to unite and produce a viable zygote. Third, an environment sufficiently favorable for the development of the hybrid must exist, even if sub- optimal, regarding the parental stock. 121 The data and observations Indicate that E_. caeruleum and 12.

spectabile possess a ready potential to hybridize in nature. It is

not a foregone conclusion, however, that even sympatrlc species, ful­

filling all these "requirements", necessarily hybridize (Hubbs, 1955;

Miller, 1968), but some E. caeruleum-E. spectabile putative hybrids

examined, deserve assignment to the "valid hybrid" category. Granted

a close phylogenetic relationship, the three principal factors enabling

or facilitating the process of hybridization are: restricted and simi­ lar breeding sites, disturbance of the environment, and scarcity of one parental species and abundance of the other (Hubbs, 1955; Hubbs and

Laritz, 1961).

Particularly intimate sympatry, expressed in the form of similar breeding behavior on overlapping or identical breeding sites, enhances the possibility of interspecific gametic union. In the case of 12. caeruleum and 12. spectabile, the area separating the preferred breeding sites is sometimes transitory. There is considerable ecological compe­ tition between the two species for occupancy of even sub-optimal breed­ ing sites when the number of riffle areas and gravel sheets are sparse, especially where the zone of 12. spectabile predominance ends and that of

E. caeruleum begins. This condition existed in the principal study stream where this portion was partially dredged.

In contrast to more obvious factors allowing hybridization in 12. caeruleum and _E. spectabile, there exists a more specific breakdown in interspecific Isolation. 12. spectabile males sometimes seem to par­ tially lack "integrity" in the choice of a conspecific mate, and will mate with a female E. caeruleum. Especially in overlapping, crowded, 122 sub-optimal breeding sites, this results in a breakdown of specific reproductive isolation. Overcrowding, combined with the almost iden­ tical mating behavior of the two species and the three "typical" hybrid-conducive factors, it is believed, explain the existence of the hybrids collected at the principal study sites of Dry Run. Based on observations, only one other stream, Blackllck Creek, in NE Franklin

Co., Ohio, produced hybrids to any appreciable extent. The upper portion of Blacklick Creek, the zone of 12. spectabile predominance, was relatively undisturbed, followed by a dredged section, followed by a zone of j2. caeruleum predominance.

Fortunately, for the survival of the species, the number and potency of the isolating mechanisms far outweigh the effects of those factors tending to break down the genetic, ecological and behavioral barriers. Observations on specific habitat and breeding site prefer­ ence gave considerable insight into hybridization barriers on an ecological scale. When the separation effect of habitat preference was diminished by the disturbance or obliteration of either of the preferred habitats, species competition resulted in at least partial occupancy of the same habitat by E. caeruleum and E. spectabile. This * was the case where dredging of parts of the study stream facilitated a certain degree of hybridization between these two species. The fact that substantial numbers of hybrids were not taken even under these circumstances, is, no doubt, accounted for by the reproductive behavioral isolating mechanisms. Even when E. caeruleum and _E. spectabile were in disproportionate numbers, in terms of their natural occurrence and habitat preference, cross-mating was rare. Female 123 pre-spawning behavior triggered species-specific response on the part of the male in the greatest majority of cases. In the case of E. caeruleum matings, the observed species-specific breeding integrity was complete.

Visual cues and releasers appeared to be sufficiently inter- specifically exclusive by themselves. Henceforth, even with a break­ down of ecological barriers, behavior differences significantly isolate the two species and effectively act to maintain the respective species integrity.

In a rather lengthy study of Etheostomatine hybrid survival, * Hubbs (1967b) discussed data regarding interspecific gamete inter­ actions. He found that "the sperm of E^. caeruleum seems to be inhibited by 12. spectabile eggs, but the 12. spectabile sperm seems to be invigorated by 12. caeruleum eggs. The inhibition is slightly greater than the apparent invigoration . . . indicating an overall inhibition."

In a study of gamete compatibility of interspecific crosses involving JE. lepidum (the greenthroat darter) Hubbs (1961) found much greater success between JE. lepidum and allopatric populations of JE. caeruleum and _E. spectabile than with crosses of sympatrlc populations of these species. Though it is not stated specifically, it can be assumed that, based on the proximal relationship of these three species, crosses between sympatrlc 12. spectabile and 12. caeruleum would not be as successful as crosses of allopatric populations in terms of per­ centage of fertilization. 124

Finally, in addition to, or in conjunction with, gamete inhibition

and hybrid vitality Hubbs (1967b) proposed another safeguard against

speciation breakdown, that of proximal inhibition* According to this

hypothesis, hybrids of closely related species, in particular of

crosses between E^. lepidum and IS. spectabile and between E. spectabile

and JE. caeruleum, have low hybrid vitality and are essentially doomed.

In addition, no male E. caeruleum x E. spectabile hybrids were fertile.

Some females were fertile. If this hypothesis is true, it means that

_E. caeruleum x E. spectabile hybrids are not heterotic, as is the case

so often of F-^ hybrids of Etheostoma tine origin (parentage), that the

possibility of introgression would be severely limited, and the

probability of finding backcrosses would be as low or lower than that

of finding an F^ hybrid.

It is not surprising, therefore, to find a relative rarity of

natural hybrids between IS. caeruleum and E. spectabile. The choice of

a study stream was made after many observations. It was considered to

have factors particularly conducive to hybridization between the study

species and may be one of the very few locations in central Ohio where

caeruleum and IS. spectabile hybridize to an appreciable extent. The

fact that natural hybrids could occur was substantiated by the success­

ful laboratory crossing and raising of the hybrid offspring and the

observation of a number of cross-matings between E^. caeruleum and IS.

spectabile in both the field and laboratory.

The secondary study stream, Peter’s Run, presents a situation which should readily lend itself to the investigation of the question 125 of natural hybridization between 13. caeruleum and 13. spectabile. While

this stream has a thriving population of 13. spectabile, 13. caeruleum is

totally absent. The stream is only occasionally intermittent, that is,

in only the driest summers. The introduction of a few hundred E.

caeruleum females during the fall, when the danger of dessication has

passed, would seem to establish conditions ideal to hybridization. It might also help answer the question as to the existence of pheromones as releasers, more sophisticated isolating mechanisms. If pheromones were released by the female 13. caeruleum, it is possible that they would elicit an upstream migrationi of male 1 caeruleum" 1 from the mouth of the stream at the Scioto River. Primarily, hybrids could easily be identi­ fied, preferably during the following fall or even summer, to deter­ mine if they are, or are not, heterotic, which habitat they prefer, etc. A comparison could be made of their meristic and raorphometric characters with that of the parental stock. The female JE. caeruleum used could be taken from Dry Run in the same (Scioto) drainage, or from

Mac-o-chee Creek of the Miami drainage system. (In either of these cases, meristic and morphemetrie studies have already been started.)

It seems an ideal opportunity to use a natural system to study the hybridization of allopatric occurring 13. caeruleum and 13. spectabile in a sympatric setting. The problem of introgressive hybridization par­ ticularly needs more study. SUMMARY

The Isolating mechanisms that must be broken down before natural hybridization can occur rarely appear to be violated under normal condi­

tions. These isolating mechanisms maintaining E_. caeruleum and E. spectabile as distinct and integral species are as follows: 1 ) specific habitat preference, 2) specific breeding site preference, 3) orienta­

tion by IS. caeruleum (males and females) to visual environmental cues,

4) intraspeclfic recognition and response to reproductive behavioral releasers, especially those sensed visually, 5) minimal chances of accidental (cross—) fertilization due to simultaneous deposition of eggs and sperm in the substrate, 6 ) inhibition of E. caeruleum sperm by 12. spectabile eggs, and 7) low compatibility of gametes of sympatrlc populations as compared with allopatric populations.

The most enlightening revelations of this study have been the specific isolating mechanisms involved in the reproductive behavior of the respective species. Aspects of habitat preference and environ­ mental stimuli, especially during the breeding season, serve as princi­ pal ecological barriers to hybridization. Behavioral mechanisms rein­ force ecologically influenced responses in that different specific releasers elicit equally specific responses. These factors are espe­ cially striking in pre-spawning and spawning behavior.

E. spectabile prefers streams and stream parts having a relatively

126 127 low rate of flow and fine substrate, while _E. caeruleum prefers faster, deeper waters with a coarse, even cobbled substrate.

Previously undetected signals were recognized In observation of the specific reproductive behavior which enhance the more physical isolating mechanisms. The "twitching" movement and the dark eyespot pigmentation of the pre-spawning 12. spectabile female loomed in signi­ ficant contrast to the "pseudo—feeding" (arched back) and darkened saddle bands of the 12. caeruleum female. Beyond this, the actual reproductive process of spawning is essentially the same.

Meristic and morphometric analysis revealed both similarities and significant differences between E. caeruleum and _E. spectabile.

Discriminant function analysis of putative hybrids in comparison with laboratory-raised hybrids and samples of parental stocks confirming the occurrence of at least seven natural hybrids, all from the same study stream. Analysis of variance of sympatrlc and allopatric popu­ lations of E. spectabile indicated the probability of introgresslon in sympatrlc situations.

Karyological studies disclosed the same chromosome complement and similarities in chromosome morphology supporting taxonomic categori­ zation. Hybridization experiments in the laboratory proved that the two species are reproductlvely compatible. Finally, characters were described which would facilitate the identification of E. caeruleum-

E. spectabile hybrids in the field and laboratory. APPENDIX A TABLE I?. MERISTIC AND MORPHOfCTRIC DATA FO R E. CAERULEUM, E, SPECTABILE, AND THEIR NATURAL HYBRIDS TAKEN FROM DRY RUN, PICKAWAY CO., SCIOTO RIVER DRAINAGE.

Character E, caeruleum (N n 34 ) Hybrids {N ■ 7) E. spectabile (N ■ 3

Infraorbital bar index Range 0,000-4,000 2.000-5.000 . 3.000-5,000 Mean 1.029 3.000 4.364 SD 1.071 1.000 O.653 Number of dorsal spines Range 8,000-11.000 9,000-10,000 9.000-11.000 Mean 9.686 9.714 9.788 SD 0.530 0,488 0.545 Number of soft dorsal rays Range 12.000-14.000 12.000-13.000 11.000-14.000 Mean 12,886 12.571 12.455 SO O.530 0.535 0.617 Number of lateral bars Range 6.000-12.000 7.00&-9.000 4.000-7.000 Mean 10.086 8.429 5.485 SD 1.380 0.787 0.755 Number of pectoral rays Range 23.000-26.000 23,000-26.000 21.000-25.000 Mean 25.371 24,286 23.424 SD 0.973 0.951 0.902 Number of opercular scales Range 7.000-19.000 10.000-20,000 5,000-23,000 Mean 13.314 15.286 12.909 SD 3.104 3.498 3.703 Number of lateral line scales Range 40,000-49,000 42 .000-46.000 40.000-49.000 Mean 44.429 43.714 43.455 SD 2.279 1.704 2.093 Number of pored lateral line scales Range 25.000.36.000 22.000-30.000 17.000-33.000 Mean 30.371 27.143 25.152 SD 3.049 2.968 3.641 Number of scales above lateral line Range 4.000-5,000 4 ,000-5,000 4.000-5.000 Mean 4,866 4.714 4.667 SD 0.323 0.488 0.479 Number of scales below lateral line Range 7.000-9.000 7.000-8.000 7.000-8.000 Mean 7.714 7.714 7.394 SD 0.667 0.488 0.496 \o TABLE 19. Continued.

Character E. caeruleum (N «* 34) i-lybrids (N - 7) spectabile (N ■ 33)

Nape scalation index Range 1.000-3,000 1.000-8.000 1,000-7.000 Mean 2.486 4.714 3.354 SO 1.095 2.928 1.619 Urogenital pigmentation index Range 0.000-10.000 0.000-7.000 0.000-3.000 Mean 3.629 1.429 0.152 SO 3.326 2.573 O.566 Infraorbital canal index Range 2.000 0.000-2.000 0.000 (complete vs. incomplete) Mean 2.000 1.571 0.000 SO 0.000 0.787 0.000

Number of lacrimal pores Range 7,000-8,000 8.000-9.000 8.000-10.000 Mean 7.371 8.143 8.091 SO 0.169 0,378 0.384 Number of postlacrimal pores Range 6.000-9.000 6.000-9.000 5.000-9.000 Mean 7.771 8.000 6.606 SO 0.598 1.000 0.998 Head length/SL Range 0.282-0.327 0.298-0.321 0.271-0,325 Mean 0.300 0,308 0.304 SO 0.009 0,008 0.013 Head depth/SL Range 0.168-0.202 0.166-0.192 0.159-0,187 Mean 0.183 0.181 0.175 SO 0,009 0.009 0.006 Head width/SL Range 0.130-0.165 0.127-0.160 0.123-0.147 Mean 0,148 0.139 0.134 SO 0,010 0,012 0.007 Snout length/SL Range 0.061-0.082 0.061-0.072 0.058-0.074 Mean 0.070 0.068 0.064 SO 0.005 0.005 0.004 Upper jaw length/SL Range 0.072-0.091 0.077-0.090 O.O75-O.O89 Mean 0.082 0,083 0,081 SD 0.005 0.004 0.004 Interorbital distance/SL Range O.O36-O.O54 0.037-0,060 O.O35.O.O67 Mean 0,046 0.047 0.048 SO 0.005 0.009 0,008 TABLE 19. Continued.

Character E^ caeruleum {N ■ 34) (N ■ 7) spectabile (N ■ 33)

UGM to maxillary tip/SL Range 0.130-0.181 0. 117- 0.164 0. 107- 0,147 Mean 0.156 0.144 0.126 SO 0.013 0.016 0,011

Body depth/SL Range 0. 191- 0,242 0.176-0.209 0.177-0.219 Mean 0,211 0.199 0.197 SD 0,012 0.012 0.011

Point o f maximum body depth/SL Range O.3IO-O.432 0. 339- 0.453 0. 283- 0.427 Mean 0.389 0.379 0.353 SD 0,035 0.045 O.O36

Pectoral fin length/SL Range 0. 231- 0.304 0. 243 - 0.285 O.23O-O.3IO Mean 0.260 0.266 0.268 SD 0.016 0.017 0.016

Occiput to dorsal origin/SL Range 0.126-0.171 0.125-0.166 0.123-0.167 Mean 0.147 0.154 0.147 SD 0.011 0.016 0.010

Spinous dorsal base length/SL Range 0.256-0.315 0. 243 - 0.310 0.246-0.323 Mean 0,289 0.282 0.280 SD 0.013 0.026 0.019

Longest dorsal spine/SL Range 0. 105 - 0.149 0, 121- 0,145 0.106-0.155 Mean 0.120 0.130 0.124 SD 0.010 0,010 0.013

Depressed s o ft dorsal fin length/SL Range 0. 243 - 0.337 0. 271- 0.337 0. 240 - 0.333 Mean 0.290 O.361 0.284 SD 0,022 0.025 0.019

Predorsal length/SL Range 0. 341 - 0.382 0. 351 - 0.373 O.327-O.396 Mean 0.359 O.36I 0.353 SD 0.010 0.009 0.015

Preanal length/SL Range 0.579-0.680 O.605-O.67I 0.601-0.658 O Mean 0,636 0.644 ro & SD 0.021 0.025 0.014

Oepressed anal fin length/SL Range 0.207-0.272 0,217-0.282 0. 214 - 0, 2/9 Mean 0.241 0.249 0.248 SD 0.017 0.027 0.017 TABLE 1^» Continued.

Character E. caeruleum (N ■ 34) Hybrids (N ■ j) ju spectabile (N - 33)

Vent to caudal peduncle/SL Range 0.361-0.441 0.374-0,414 0.184- 0.443 Mean 0.408 0.394 .0,388 SO 0.018 0,014 0.041

Caudal peduncle length/SL Range 0,220-0,295 0.225 -0.254 0.219-0.297 Mean 0.257 0.237 0.255 SO 0,017 0,010 0.016

Caudal peduncle depth/SL Range 0.093-0.113 0.093-0.110 0,087-0.108 Mean 0.104 0.103 0.099 SO 0,005 0,006 0.087

Longest soft dorsal ray/SL Range 0.128-0,201 0.150-0.187 0.130-0.185 Mean 0.151 0.163 0.156 SO 0.014 0.012 0.013

Head length/Caudal peduncle length Range 2.651 -3.439 2.837-3.302 2.811-3.469 Mean 2,892 2.997 3.061 SD 0.145 0,205 0,181

Head length/Distance-point of Range 0.702-0,932 0 . 6 6 9 4 . 9 1 9 0.678-1.060 greatest body depth Mean 0.776 0.821 O.869 SD O.067 0.100 0.094

Head depth/Head length Range 0.559 -0.687 O.547-O.623 0.521-0.652 Mean 0.612 O.589 0.578 SO 0,030 0.026 0.028

Head depth/feody depth Range 0,812-0.940 O.869-O.945 O.8I5-O.975 Mean O.87O 0.909 0.890 SD 0.032 0.027 0.040

Head width/Head depth Range 0.740-0.878 O.7I8-O.853 0.707-0.825 Mean 0,809 O.767 O.765 SO O.O36 0.044 O.O35

Head widthAtead length Range 0.429-0.549 0.421 -0,512 0.398-0.506 Mean 0.495 0.452 0.442 SO 0.030 0.038 0.027

Snout length/Head length Range 0.198-0.271 0,198-0,240 0.189-0.234 Mean 0.233 0.221 0.211 * SD 0.017 0.016 0.013 TABLE 19* Continued,

Character E. caeruleum (N - 34 ) ftybrids (N a 7) E. spectabile (N -

Snout lengtf^Jpper jaw length Range 0.735-0.972 0.735-0-917 0.703-0.920 Mean 0.854 O.8I9 • 0.795 SD O.059 O.O63 0.045 Interorbital distance/Head width Range 0.253-0.382 0,294- 0.400 0.264 -0.459 Mean 0.309 0.340 O.36I SD 0.035 0,043 0.053 UGM/Head length Range 0.424 -0.596 O.376-O.526 0.359-0.485 Mean 0.521 0.470 0.416 SD 0.040 0.052 0.038 Body depth/Predorsal length Range 0.529-0.685 O.5OO-O.59O 0.500-0.628 Mean O.588 0.553 0.561 SD O.O33 0.035 0.033 Body depth/Head length Range 0,638-0.784 0.579-0.697 0.580-0.775 Mean O.705 0.649 0.651 SD 0.037 0.039 0.046 Pectoral fin length/Depressed Range 0.797- 1.024 0.830-1.026 0.768-1,111 soft dorsal fin length Mean 0.897 0.917 0.946 SD 0.055 0.074 0.071 Spinous dorsal fin base length/ Range 0.787-1.149 0.844-1.125 0.858-1.196 Depressed soft dorsal fin length Mean 1.000 0.974 0.989 SD 0,081 0.135 0.084

Longest dorsal spine/Spinous Range O.353-O.583 0.397-0.512 0.35 CW ).576 dorsal fin base length Mean 0,418 0.464 0.443 SD 0,045 0.047 0.051 Longest soft dorsal ray/Depressed Range 0.444 -0,708 0.533-0.537 0.440 -0.617 soft dorsal fin length Mean 0.522 0,558 0.549 SD 0,049 0,021 O.O38

Longest dorsal spine/Longest dorsal ray Range 0.537-0.970 0.762-0.848 . 0.689-0.943 Mean 0.800 0,800 0.794 SD 0.075 O.O33 0.060

Predorsal lengtli/Preanal length Range 0.519-0.610 0.531 -0.582 0.515-0,621 Mean 0.565 O.560 0.558 SO 0.024 0.022 0.024 TABLE 19. Continued.

Character E. caeruleum (N - 34 ) Hybrids (M ■ 7) E. spectabile (N - 33)

Depressed anal fin length/ Range O.6B6-O.932 ' O.756-O.947 O.760-O.943 Depressed soft dorsal fin length Mean O.833 0.853 0.873 SD 0.054 0.068 0.043 Caudal peduncle length/ Range 2,071-2.822 2.146 -2,517 2.098-3.133 Caudal peduncle depth Mean 2.479 2.313 2.568 SD 0.205 0.144 0.211

Caudal peduncle length/Distance- Range 0.536 -0.736 O.579-O.652 0.577-1.402 vent to caudal peduncle Mean O.63I 0.603 0.669 SD 0.043 0.026 0.136

I-* tj JS- TABLE 20. MERISTIC AND MORPHOfCTRIC DATA FOR E. CAERULEUM (MAC-O-CHEE C R „ LOGAN CO., MIAMI RIVER DRAINAGE), E. SPECTABILE (PETER'S RUN, PICKAWAY CO., SCIOTO RIVER DRAINAGE), AND THEIR LABORATORY-RAISED HYBRIDS*”

Character E. caeruleum (N - 10) Hybrids (N ■ 7) E» spectabile (N « 10)

Infraorbital bar index Range 0.000-2.000 1.000-4.000 3.000-5.000 Mean 0.300 2.714 4.400 SD 0.675 0.951 O.699 ' Number o f dorsal spines Range 9.000-10.000 9.000-11.000 9.000-11.000 Mean 3.500 9.857 9.900 SO 0.527 0.690 O.568 Number o f s o ft dorsal rays Range 12.000-14.000 14 .000- 15,000 12.000-13.000 Mean 13.300 14.429 12.500 SD O.675 0.535 0.527 Number of lateral bars Range 10.000-12.000 5.000-9.000 4 .000-6.000 Mean 10.600 6.714 5.000 SD O.699 1.496 0.471 Number of pectoral rays Range 24 ,000-28.000 24.000-26.000 21.000-26.000 Mean 25,800 24.714 23.800 SO 1.135 0.756 1.398 Number of opercular scales Range 6.000-15.000 10.000-15.000 12.000-24.000 Mean 10,800 11.286 15.200 SD 2.741 2.215 3.458

Number of lateral linB scales Range 42 ,000-47.000 37.000-42.000 43,000-48.000 Mean 44,200 39.571 44.500 SD 1.549 1.618 1.900 Number of pored lateral line scales Range 25.000-33.000 6,ooo-a,ooo .23,000-30.000 Mean 28.800 15.857 26.200 SD 2.6l6 5.336 2.150 Number o f scales above la te r a l lin e Range 5.000-6,000 4,000-5,000 5.000-6.000 Mean 5.100 4.571 5.400 SD 0.316 0.535 0.516 Number of scales below lateral line Range 7.000-9.000 6.000-8.000 7.000-8.000 Mean 8.000 7.000 7,600 ■ SD 0.471 0.577 0.516 uh* ui TABLE 20. Continued.

Character E, caeruleum (N » 10) tfybrids (N ■ 7) E. spectabile (N ■ 10)

Nape scalation index Range 2.000-5.000 2.000-6.000 2.000-8.000 Mean 3.600 4.000 4.500 SD O.366 1.414 2.121 Urogenital pigmentation index Range 0.000-10.000 0.000-8.000 0.000-1.000 Mean 4.300 4.423 0.300 SO 4.111 3.155 0.483 Infraorbital canal index Range 2.000 0.000 0.000 (complete vs. incomplete) Mean 2.000 0.000 0.000 SO 0.000 0.000 0.000 Number of lacrimal pores Range 8.000 6,000-8,000 8.000-3.000 Mean 0.000 7.423 8.200 SO 0.000 0.787 0.422 Number of postlacrimal pores Range 7.000-3,000 6.000-7.000 5.00CU7.000 Mean 7.500...... 6.143 6.200 SO 0.316 0.378 0.632 Head length/SL Range 0.271-0.300 0.302-0.315 0.272-0.305 Mean 0.284 0.307 0.232 SO 0.011 0.005 0.011 Head depth/SL Range 0.172-0.213 0.153-0.130 0.163-0.200 Mean 0.133 0.176 0.185 SO 0,012 0.015 0.003 Head width/SL Range 0.141 -0.163 0.134-0.156 0.135-0.155 Mean 0.155 0.143 0.145 SD 0.003 0.007 0,006 Snout length/SL Range o,o66-o.o81 0.064-0.073 0.062-0.072 Mean 0.075 0.068 0.067 SO 0.005 0.003 0.003 Upper jaw length/SL Range 0.075-0.030 . 0.075-0.086 0.076-0.083 Mean 0.082 0.082 0.081 SO 0.004 0.004 0.004 Interorbital distance/SL Range 0.033-0.056 0.033-0.051 0.043-0.053 Mean 0.048 0,046 0,046 SD 0.006 0.004 0.003 TABLE 20. Continued.

Character E. cacruleum (N « 10) Hybrids (N - 7) E. spectabile (N ■ '

UGH to maxillary iip/SL Range 0.121-0.168 0.119-0.168 0.102-0.145 Mean 0.150 0.150 0,130 SO 0.016 0,020 0,014 Body depth/SL Range 0.221-0.260 0,195-0.222 0.201-0.231 Mean 0.241 0,211 0.217 SD 0,011 0,010 0.010 Point of maximum body depth/SL Range 0.387-0.421 0.319-0.470 0,320-0.387 Mean 0.408 0.408 0.365 SO 0.012 0.059 0.0a Pectoral fin length/SL Range O.256-O.3OI 0.238-0.271 0 . 2 4 5 4 . 2 8 9 Mean 0,284 0.254 0.269 SD 0.013 0.011 0.015 Occiput to dorsal origin/SL Range 0.136-0.168 0.126-0.158 0.136-0.174 Mean 0.147 0,142 0.149 SD 0.009 0.011 0,012 Spinous dorsal base length/SL Range 0.261-0.319 O.258-O.307 0.254 -0.304 Mean 0.2B5 0.285 0.277 SD 0.020 0.018 0.014 Longest dorsal spine/SL ] Range 0.102-0.131 0.115-0.129 0.095-0.140 i Mean 0,116 0,122 0.120 ] SD 0.010 0.005 0.014 Depressed soft dorsal fin length/SL Range 0.270-0.341 O . 2 7 6 A 3 O 5 0.2594.3a Mean 0.309 0,288 0,295 SD 0.028 0.012 0.023

Predorsal length/SL Range 0.335 -0.365 O . 3 5 6 4 . 3 7 8 0.3354.372

Mean ' 0.351 O.367 O.349 SO 0.009 0,008 0,012

Preanal length/SL Range 0.421 -0.650 0,611-0.644 0.593-0.655 Mean 0.605 O.63I O.63I SD 0,067 0.013 0.018 Depressed anal fin length/SL Range 0,220-0,287 0.233-0.271 0.2234.285 Mean 0.254 0.254 0,257 SD 0.023 0,012 0,022 TABLE 20, Continued,

Character E» caeruleun (N - 10) Hybrids (fJ ■ 7) E, spectabile (N ■ 10)

Vent to caudal peduncle/SL Range 0,368-0.445 0.384- 0.408 0.383-0.425 Mean 0.415 O.396 0.396 SD 0.023 0.009 0.013 Caudal peduncle length/SL Range 0.230-0.270 0.193-0.234 0.208-0.262 Mean 0,247 0.213 0.242 SO 0.014 ‘0.015 0.015 Caudal peduncle depth/SL Range 0.091-0,115 0.075-0.090 0.094-0.121 Mean 0.108 ■ O.O83 0.107 SO 0,003 0.005 0.008 Longest soft dorsal ray/SL Range 0.133-0.167 0.130-0.152 0,145 -0.194 Mean 0.151 0.141 O.I65 SO 0.011 0.006 0.015 Head length/Caudal peduncle length Range 2.388-3.022 3.417-4.036 2.421 -3.143 Mean 2.652 3.687 2.745 SO 0.199 0.195 0.230 Head length/Distonce-point of Range 0.642-0.755 0.642 -1.000 O.73O-O.939 greatest body depth Mean 0.696 O.768 0.801 SO 0.039 0.128 O.065

Head depth/Head length Range 0.621-0.795 0.056 -0.486 0.588-0.674 Mean 0.701 0.573 O.636 SD 0.046 O.O56 O.O32 Head depth/Body depth Range 0.719-0.892 0.779-0.863 0.798-0.911 Mean 0.827 0.831 0.855 SO 0.060 O.O34 0.035 Head width/Head depth Range ■ O.735-O.8I9 O.725-O.972 0.747-0.829 Mean 0.777 0,819 0.785 SO 0.029 0.061 0.026 Head width/Head length Range 0.496-0,615 0.433-0.505 0.463-0.543 Mean 0.544 0.467 0.499 SD O.O33 0,025 0.021 Snout length/Head length Range 0.227-0.290 0.212-0.236 0,203-0.254 Mean 0.264 0.223 0.231 SD 0.021 0.009 0.015 TABLE 20. Continued.

Character E, caeruleun (N ■ 10) hjybrids (N o J) E, spectabile (N • 10)

Snout length/Upper jaw length Range 0,861-0.957 0.793-0,862 0.781-0.075 Mean 0.916 0.840 0.834 SD 0.030 0.024 0.035 Interorbital distance/Head width Range 0.268-0.379 0.289-0.340 0.293-0.364 Mean 0.309 0.320 0.319 SD 0.034 0.016 0.022

UGM/head length Range 0.414 -0.603 0.335 -0.559 0.340 -0.504 Mean 0.530 0.491 0.4 66 SD 0.058 0,072 0.053 Body depth/Predoreal length Range O.62I-O.763 0.524 -0.624 0.579-0.676 Mean 0.688 0.575 0.622 SO O.038 O.033 0.032 Body depth/Head length Range O.774-O.960 0.619-0.735 O.69I-O.8O4 Mean 0.850 0.688 0.745 SD 0.055 0.041 0,042 Pectoral fin length/Oepressed Range 0.820-1.040 0.781-0.982 0.793-1.061 soft dorsal fin length Mean 0.922 0.883 0.917 SD 0.072 O.O69 0.084 Spinous dorsal fin base length Range 0.815-1.132 0.910-1.055 0.879-1.023 /Depressed soft dorsal fin length Mean 0.929 0.988 O.943 SD 0.104 0.050 0.052 Longest dorsal spine/Spinous Range 0.329-0.477 0.374-0.500 0.356 -0.493 dorsal fin base length Moan 0.409 0.429 0.432 SD 0.040 0.042 0.052 Longest soft dorsal ray/Oepressed Range 0.46 a-0.526 0.462 -0.518 - 0.494 -0.664 soft dorsal fin length Mean 0.488 0.489 O.563 SD 0.018 0,024 O.O54 Longest dorsal spine/Longest dorsal ray Range 0.714-0.847 O.8II-O.915 0.598-0.816 Mean 0.773 O.863 0.724 SD 0.040 O.O37 O.076 Predorsal length/Preanal length Range 0.516-0,834 0.559 -0.603 0,519-0,572 Mean 0.588 0.582 0.553 SD Q.O89 0.014 0,018 TABLE 20. Continued.

Character E, caeruleum {N - 10) fyb rid s {N - 7) E, spectabile (N * 10)

Depressed anal fin le n g th / . Range 0.730-O.936 O.763-O.982 0,786-0.936 Depressed s o ft dorsal fin length Mean 0,823 O.883 0.874 O O SD 0,070 •

•S? 0.042

Caudal peduncle le n g th / Range 2.097-2.600 2. 389- 2.786 1. 825- 2.714 Caudal peduncle depth Mean 2.304 2.556 2.284 SO 0.164 O.139 0,278

Caudal peduncle length/Distance Range O.556-O.694 0. 486- 0.609 0. 528- 0,684 vent to caudal peduncle Mean O.596 O.539 0.612 SO 0.042 0.043 0.041 APPENDIX B 142

TABLE 21.

SPECIMENS STUDIED

Location: Ohio. SW Pickaway Co., SW Muhlenberg Twp. and W. Monroe Twp. Dry Run, trlb* of Deer Cr., Scioto River drainage; 3 5 km. E of Five Points, Ohio. Collected in 5 km. stretch of creek S of Ohio Rte. 316.

Species Date OSUM catalog Number of collected number specimens

Etheostoma caeruleum: 22 May 1971 21798 4 (34) 27 May 1971 21799 8 2 Aug 1971 21800 5 1 1 Dec 1971 21801 1 29 Sept 1972 21802 16

Etheostoma spectabile: 22 May 1971 21803 1 (33) 27 May 1971 21804 2 1 1 Nov 1971 21805 8 29 Feb 1972 21806 3 12 May 1972 21807 1 21 May 1972 21808 1 7 June 1972 21809 1 2 0 Sept 1972 21810 16

E. caeruleum—E. spectabile Code Station hybrid suspects: (cataloged individually) 22 May 1971 21811 22 May 1971 21812 22 May 1971 21813 * YS V 27 May 1971 21814 27 May 1971 21815 * Y4 III 29 Febr 1972 21816 * 3 IV 29 Febr 1972 21817 27 Apr 1972 21818 * y7 III 21 May 1972 21819 7 June 1972 21820 * y6 IV 2 0 Sept 1972 21821 2 0 Sept 1972 21822 ^Specimens designated as 2 0 Sept 1972 21823 * valid natural hybrids 2 0 Sept 1972 21824 Y 1 IV subsequent to analysis 2 0 Sept 1972 21825 * 2 IV 143 Location: Ohio. Logan Co., SW Monroe Twp. Mac—o-chee Cr., trib. of Mad River, Miami drainage. 3.5 km. ENE of West Liberty, Ohio.

Etheostoma caeruleum (10) 20 Apr 1972 21826

Location: Ohio. N Pickaway Co. N Scioto Twp. Peter*s Run, trib. of Scioto River; 3.2 km. ENE of Commercial Point, Ohio. Collected in 100 m. stretch of creek E of Ohio Rte. 104.

Etheostoma spectabile (10) 15 May 1972 21827

IS. caeruleum—E . spectabile laboratory-raised hybrids (7): Mad River Peter's Run May 1972 — July 1973 Locations described above.

0SUM catalog numbers : 21931 144

LITERATURE CITED

Andrews, E. B. 1874. Geological survey of Ohio. Vol. II. Surface Geology of Southeastern Ohio, pp. 441-452.

Bailey, R. M. and W. A. Gosline. 1955. Variation and systematic significance of vertebral counts in the American fishes of the family . Mus. Zool. Univ. Michigan. Pub. 93:44 p.

Barlow, G. W. 1962. Ethology of the Asian teleost, Badis badis. IV. Sexual behavior. Copeia 1962(2):346-360.

. 1968. Ethological units of behavior. In D. Ingle (ed.). The central nervous# system and fish behavior. pp. 217-232. Univ. Chicago Press, Chicago. 27 2 p.

. 1972. The attitude of fish eye-lines in relation to body shape and to stripes and bars. Copeia 1972(1):4—12.

Blair, A. P. 1959. Distribution of the darters (Percidae, Etheos— tomatinae) of northeastern Oklahoma. Southwest. Nat. 4(1):1-13.

Braasch, M. E. and P. W. Smith. 1967. The life history of the slough darter, Etheostoma gracile, (Pisces, Percidae). Biol. Note., 111. Nat. Hist. Surv. 58. 12 p.

Branson, B. A., and J. B. Campbell. 1969. Hybridization in the darters Etheostoma spectabile and Etheostoma radiosum cyanorunm. Copeia 1969(1):70-75.

Cain, A. J. 1953. Geography, ecology, and coexistence in relation to the biological definition of the species. Evolution 7:76—83.

Clark, D. E. 1973. Morphological and ecological relationships of Etheostoma caeruleum and _E. spectabile in west central Indiana. Ph. D. dissertation. Indiana State University. 170 p.

Cole, C. F. 1957. The taxonomy of the percid fishes of the genus Etheostoma, subgenus Boleosoma, of eastern United States. Ph. D. dissertation. Cornell Univ. 366 p.

. 1965. Additional taxonomic evidence for the separation of Etheostoma olmstedi Storer from Etheostoma nigrum Rafinesque. Copeia 1965(1):8-13. 145

Collette, B. B. 1963. The subfamilies, tribes and genera of the Percidae (Teleostei). Copeia 1963(4):615-623.

______. 1965. Systematic significance of breeding tubercles in fishes of the family Percidae. Proc. U. S. Nat. Hist. Mus. 117(3518):567-614.

______. 1967. The taxonomic history of the darters (Percidae: Etheostomatini). Copeia 1967(4):814-819.

Cook, Fannye A. 1959. Freshwater fishes In Mississippi. Miss. Game and Fish. Comm. Jackson, Miss. 239 p.

Cross, F. B. 1967. Handbook of fishes of Kansas. Univ. Kansas Mus. •Nat. Hist. Misc. Pub. 45. 357 p.

Denton, T. E. and W. M. Howell. 1969. A technique for obtaining chromosomes from the scale epithelium of teleost fishes. Copeia 1969:392-393.

Dlstler, D. A. 1968. Distribution and variation of Etheostoma spectabile (Agassiz) Percidae-Teleostei). Univ. Kans. Sci. Bull. 48(5):143-208.

Echelle, A. A., J. R. Schenck, and L. G. Hill. 1974. Etheostoma- spectabile-E. radiosum hybridization in Blue River, Oklahoma. (Manuscript, to be published in Am. Midi. Nat., 1974.) 23 p.

Fenneman, N. M. 1938. Physiography of eastern United States. New York and London, McGraw-Hill Book Co. 714 p.

Gerking, S. D. 1945. Distribution of the fishes of Indiana. Invest. Indiana Lakes and Streams 3:1-137.

______. 1955. Key to the fishes of Indiana. Invest. Indiana Lakes and Streams 4:49-86.

Hill, D. R. 1959. Some uses of statistical analysis in classifying races of American shad (Alosa sapidissima). U. S. Fish Wildl. Serv. Fish. Bull. 147,5 9:269-286.

Hubbs, C. L. 1955. Hybridization between fish species in nature. Syst. Zool. 4:1-20.

______. 1961. Isolating mechanisms in the speclation of fishes. In W. F. Blair, (ed.) Vertebrate Speclation. Univ. Texas Press, Austin, pp. 5-23.

______» and Laura C. Hubbs. 1932. Experimental verification of natural hybridization between distinct genera of sunfishes. Papers Mich. Acad. Sci., Arts, and Letters, 15(1931):427-437. 146 Hubbs, C. L., Laura C. Hubbs, and R. E. Johnson. 1943. Hybridization in nature between species of catostomid fishes. Contr. Lab. Vert. Biol., Univ. Michigan, 22:1-76.

_ , and K. Kuronuma. 1942. Analysis of hybridization in nature between two species of Japanese flounders. Papers Mich. Acad. Sci., Arts, and Letters, 27(1941);267-306.

_ , and K. F. Lagler. 1958. Fishes of the Great Lakes region. Cranbrook Inst. Sci. Michigan 26. 213 p.

Hubbs, C. 1956. Relative variability of hybrids between the minnows, Notropis lepidus and N. proserpinus. Texas J. Sci. 8(4):463-469.

• 1958. Fertility of F^ hybrids between the percid fishes, Etheostoma spectabile and 12. lepidum. Copeia 1958(1) ; 57—59.

. 1959. Laboratory hybrid combinations among Etheostomatine fishes. Texas J. Sci. 11:49-56. * ______. 1961. Differences between the gamete compatibility of Etheostoma lepidum sperm and eggs of related percid fishes from allopatric and sympatric populations. Texas J. Sci. 18:427-433.

______. 1967a. Analysis of phylogenetic relationships using hybridization techniques. Bull. National. Inst. Sci. India. (Reprint). 1967(34):48-59.

______. 1967b. Geographic variation in survival of hybrids between Etheostomatine fishes. Bull. Texas Mem. Mus. 13:1-72.

______. 1970. Teleost hybridization studies. Proc. California Acad. Sci. Fourth Series 38(15):289-298.

_ , and C. M. Laritz. 1961. Occurrence of a natural inter­ generic Etheostomatine fish hybrid. Copeia 1961(2):231-232.

_ , and K. Strawn. 1956. Interfertility between two sympatric fishes Notropis lutrensis and Notropis venustus (Cyprinidae). Evolution 10:341-344.

______, and K. Strawn. 1957a. Relative variability of hybrids between the darters Etheostoma spectabile and Percina caprodes. Evolution 11(1):1-10.

______, and K. Strawn. 1957b. Survival of F^ hybrids between fishes of the subfamily Etheostomatinae. J. Exp. Zool. 134:33-62.

Hynes, H. B. N. 1970. The ecology of running waters. Univ. Toronto Press. 555 p. 147 Knapp, L. W. 1964. Systematic studies of the rainbow darter Etheo­ stoma caeruleum Storer and the subgenus Hadropterus (Pisces, Percidae). Ph. D. Dissertation. Cornell Univ. 207 p.

Krolczyk, J. C. 1960. Gazetteer of Ohio Streams. Second printing. Ohio Water Plan Inventory Report No. 12. Ohio Dept. Nat. Res. Div. of Water. Columbus, Ohio. 179 p.

Kuehne, R. A. 1962. A classification of streams, illustrated by fish distribution in an eastern Kentucky creek. Ecology 43(4):608-614.

Lagler, K. F. 1952. Freshwater fishery biology. Wm. C. Brown Co., Iowa. 360 p.

Larimore, R. W., Q. H. Pickering, and L, Durham. 1952. An Inventory of the fishes of Jordan Creek, Vermilion County, Illinois. Biol. Note. 111. Nat. Hist. Surv. 29. 20 p.

Linder, A. D. 1955. The fishes of the Blue River in Oklahoma with descriptions of two new percid hybrid combinations. Amer. Midi. Nat. 54:173-191.

______• 1958. Behavior and hybridization of two species of Etheostoma (Percidae). Trans. Kansas Acad. Sci. 61:195-212.

Loos, J. J. and W. S. Woolcott. 1969. Hybridization and behavior in two species of Percina (Percidae). Copeia 1969(2):374-385.

Lorenz, K. 1970. Studies in and human behavior. Harvard University Press.

Mather, K. 1951. .atistical analysis in biology. 4th ed. Methuen & Co. Ltd. London. 267 p.

Mayr, E. 1963. Animal species and evolution. Belknap Press of Harvard University Press, Cambridge. 797 p.

McAllister, D. E., P. Jolicoeur, and H. Tusyuki. 1972. Morphological and myogen comparison of johnny and tessellated darters and their hybrids, genus Etheostoma, near Ottawa, Canada. J. Fish. Res. Bd. Canada 29(8):1173-1180.

McPhail, J. D. 1961. A systematic study of the Salvelinus alpinus complex in North America. J. Fish. Res. Bd. Canada 18(5):793- 816.

______, and R. L. Jones. 1966. A simple technique for obtaining chromosomes from teleost fish. J. Fish. Res. Bd. Canada 23(5): 767-768. 148 Miller, R. J. 1968. Speclation in the common shiner: an alternate view. Copeia 1968(3):640-647.

______» and D. D. Hall. 1968. A quantitative description and analysis of courtship and reproductive behavior in the anabantoid fish Trichogaster leeri (Bleeker). Behaviour 32:85-149.

______, and Helen C. Miller. 1970. Studies on the agonistic behavior of anabantoid fishes. Proc. Oklahoma Acad. Sci. 49:60-68.

Morris, D. 1958. The reproductive behaviour of the ten-spine stickleback (Pygosteus pungitius L.). Behaviour supplement 6: 1-154.

Nelson, J. S. 1968. Hybridization and isolating mechanisms between Catostomus commersonii and C^. macrocheilus (Pisces: Catostomidae). J. Fish. Res. Bd. Canada, 25(1):101-150.

Ouellette, R. P. and S. U. Qadri. 1968. The discriminatory power of taxonomic characteristics in separating salmonoid fishes. Syst. Zool. 17:70-75.

Pflieger, W. L. 1971. A distributional study of Missouri fishes. Univ. Kans. Mus. Nat. Hist. Publ. 2(3):225-570.

Reed, R. J. 1953. A study of the behavior of fish in three streams of Crawford County, Pennsylvania. Master's thesis. Univ. Pittsburgh. 57 p.

______. 1968. Mark and recapture studies of eight species of darters (Pisces, Percidae) in three streams of northwestern Pennsylvania. Copeia 1968(1):172—175.

Reeves, C. D. 1907. The breeding habits of the rainbow darter (Etheostoma caeruleum Storer), a study in sexual selection. Biol. Bull. 14:35-59.

Roberts, F. L. 1967. Chromosome cytology of the Osteichthyes. Prog. Fish Cult. 29:75-83.

Ross, M. R. 1973. A chromosome study of five species of Etheostomine fishes (Percidae). Copeia 1973(1):163-165.

Scalet, C. G. 1973. Reproduction of the orangebelly darter, Etheostoma radiosum cyanorum (Osteichthyes: Percidae). Amer. Midi. Nat. 89(1):156-165.

Schuster, R. L. 1952. The glacial geology of Pickaway County, Ohio. Master's Thesis. Ohio State Univ. Columbus, Ohio. 149 Sevenster, P. 1968. Motivation and learning in sticklebacks. Xn D. Ingle (ed.) The central nervous system and fish behavior, pp. 233-246. Univ. Chicago Press, Chicago. 272 p.

Sharma, A. K. and Archana Sharma. 1972. Chromosome techniques: theory and practice. Second edition. University Park Press. London. 575 p.

Smith-Vaniz, W. F. 1968. Freshwater fishes of Alabama. Auburn University. Alabama. 211 p.

Stone, F. L. 1947. Notes on two darters of the genus Boleosoma. Copeia 1947(2):92-96.

Stout, W. and G. F. Lamb. 1938. Physiographic features of south­ eastern Ohio. Ohio J. Sci. 38(2):49-83.

______, K. Ver Steeg and G. F. Lamb. 1943. Geology of water in Ohio. Geol. Surv. of Ohio Bull. 44. Ser. 4: pp. 1-694.

Strawn, K. and C. Hubbs. 1956. Observations on stripping small fishes for experimental purposes. Copeia 1956(2):114-116.

______and ______. 1961. A comparison of meristic means and variances of wild and laboratory-raised samples of the fishes Etheostoma grahami and Etheostoma lepidum (Percidae). Texas J. Sci. 13:127-159.

Taft, C. E. 1968. Algae as indicators in domestic water supplies. Address to Phycological Soc. Am. Sept. 1968.

Tinbergen, N. 1959. Comparative studies of the behaviour of gulls (Laridae): A progress report. Behaviour 15:1-70.

Trautman, M. B. 1930. The specific distinctness of Poecilichthys coeruleus (Storer) and Poecilichthys spectabills Agassiz. Copeia 1930(1):12-13.

______. 1942. Fish distribution and abundance correlated with stream gradient as a consideration in stocking programs. Trans. Seventh North Amer. Wildl. Conf. Washington. pp. 211-223.

______. 1948. A natural hybrid catfish Schilbeodes miurus x Schilbeodes mollis. Copeia 1948(3):166-174.

______. 1957. The fishes of Ohio. Ohio State University Press, Ohio. 683 p.

Ver Steeg, K. 1946. The Teays River. Ohio J. Sci. 46(6):297-307. 150 Waters, T. F. 1965. Interpretation of invertebrate drift in streams. Ecology 46:327—334.

Welch, F. S. 1948. Limnological methods. Blakiston Company, Pennsylvania. 381 p.

White, M. J. D. 1954. Animal cytology and evolution. Cambridge University Press. 454 p.

Winn, H. E. 1957. Egg site selection by three species of darters (Pisces, Percidae). Brit. J. Anira. Behav. 5(l):25-28.

______. 1958a. Comparative reproductive behavior and ecology of fourteen species of darters (Pisces-Percidae). Ecol. Monograph. 28(2):155-191.

______. 1958b. Observations on the reproductive habits of darters (Pisces—Percidae). Am. Midi. Nat. 59(1):190-212.

______and A. R. Picciolo. 1960. Communal spawning of the glassy darter Etheostoma vitreum (Cope). Copeia 1960(3):186-192.

Zorach, T. 1971. Taxonomic status of the subspecies of the tessel­ lated darter, Etheostoma olmstedi Storer, in southeastern Virginia. Chesapeake Sci. 11:254-263.

______. 1972. Systematics of the percid fishes, Etheostoma camurum and Etheostoma chlorobranchium new species, with a discussion of the subgenus Nothonotus. Copeia 1972(3):427—447.