Roosevelt wild life annals of the Roosevelt Wild Life Forest Experiment Station of the New York State College of Forestry at Syracuse University

Digitized by the Internet Archive
in 2015

https://archive.org/details/rooseveltwildlif02unse_4

Vol. II

OCTOBER. 1929

No. 2.

BULLETIN

OF

The New York State College of Forestry
At Syracuse University

FRANKLIN MOON. Dean

Roosevelt Wild Life Annals

VOLUME 2 NUMBER 2

OF THE

Roosevelt Wild Life Forest Experiment Station

Entered as second-class matter October 18, 1927, at the Post
Office at Syracuse, N. Y., under the
Act of August 24, 1912

[ 147 ]

Hill

ANNOUNCEMENT

The serial publications of the Roosevelt Wild Life Forest Experiment
Station consist of the following:

1. Roosevelt Wild Life Bulletin.

2. Roosevelt Wild Life Annals.

The Bulletin is intended to include papers of general and popular interest
on the various phases of forest wild life, and the Annals those of a more technical
nature or having a less widespread interest.

These publications are edited in cooperation with the College Committee on
Publications.

The editions of these publications are limited and do not permit of general
free distribution. Exchanges are invited. The subscription price of the Bulletin
is $4.00 per volume of four numbers, or $1.00 per single number. The price
of the Annals is $5.00 per volume of four numbers, or $1.25 per single number.
All communications concerning publications should be addressed to

The Director and Editor,

Roosevelt Wild Life Forest Experiment Station,
Syracuse, New York.

Copyright, 1929, by

Roosevelt Wild Life Forest Experiment Station

[148]

TRUSTEES OF THE NEW YORK STATE COLLEGE OF FORESTRY

Ex Officio

Dr. Charles W. Flint, Chancellor Syracuse University

Dr. Frank P. Graves, Commissioner of Education Albany, N. Y.

Hon. Alexander Macdonald, Conservation Commissioner Albany, N. Y.

Hon. Herbert H. Lehman, Lieutenant-Governor Albany, N.Y.

Appointed by the Governor

Hon. John R. Clancy Syracuse, N. Y.

Hon. Harold D. Cornwall Glenfield, N. Y.

Hon. George W. Driscoll Syracuse, N. Y.

Hon. William H. Kelley Syracuse, N. Y.

Hon. Louis Marshall New York City

Hon. Edward H. O’Hara Syracuse, N. Y.

Hon. Charles A. Upson Lockport, N. Y.

Hon. J. Henry Walters New York City

Hon. Edmund H. Lewis Syracuse, N. Y.

Officers of the Board

Hon. Louis Marshall President

Hon. John R. Clancy Vice-President

HONORARY ADVISORY COUNCIL OF THE ROOSEVELT WILD LIFE

STATION

American Members

Mrs. Corinne Roosevelt Robinson New York City

Hon. Theodore Roosevelt New York City

Mr. Kermit Roosevelt New York City

Dr. George Bird Grinnell New York City

Hon. Gifford Pinchot Milford, Pa.

Mr. Chauncey J. Hamlin Buffalo, N. Y.

Dr. George Shiras, 3rd Washington, D. C.

Dr. Frank M. Chapman New York City

Dean Henry S. Graves New Haven, Conn.

European Member

Viscount Grey Fallodon, England

ROOSEVELT WILD LIFE STATION STAFF

Franklin Moon, M.F

Dean of the College

Charles E. Johnson, A.M., Ph.D Director of the Station

Wilford A. Dence, B.S Ichthyologist and Ass’t Director

Miriam S. Mockford Secretary

Temporary Appointments *

Perley M. Silloway, M.S Field Ornithologist

Aretas A. Saunders, Ph.B Field Ornithologist

Robert T. Hatt, A.M Field Naturalist

M. W. Smith, A.B Field Naturalist

Myron T. Townsend, A.B., Ph.D Field Naturalist

Charles J. Spiker, A.B Field Naturalist

Dayton Stoner, Ph.D Field Ornithologist

Justus F. Muller, Ph.D Field Naturalist

Harley J. Van Cleave, Ph.D Field Naturalist

Le Roy C. Stegf.man, M.S Field Naturalist

Collaborators *

Richard A. Muttkowski, A.M., Ph.I) Field Naturalist

Gilbert M. Smith, Ph.D Field Naturalist

* Including only those who have made field investigations and whose reports are now in
preparation.

GENERAL CONTENTS

PAGE

1. The Ecology of Trout Streams in Yellowstone National Park.

Richard A. Muttkowski. 155

2. The Food of Trout Stream Insects in Yellowstone National Park.

Richard A. Muttkowski and Gilbert M. Smith. 241

ILLUSTRATIONS

FIGURES

Fig. 53. Yellowstone River above Cooke City bridge. End of “ Lower Canyon ”. August

8, 1921 154

Fig. 54. Clearwater River, Idaho, a short distance above the Oro Grande River, showing
type of shore, stream bed, etc. This region was swept by forest fires in 1915; the

trees on the slopes are fire-killed. August 30, 1924 154

Fig. 55- Junction of Yellowstone and Lamar rivers at low water stage. View downstream ;

Yellowstone River at left, Lamar entering at right. In spring the boulders in

the foreground are under water. September 5, 1921 159

Fig. 56. Oro Grande River, Idaho. About 300 yards above its entry into the Clearwater.

Showing rapids and boulders. In spring all except the largest boulders are

covered by water. August 31, 1924 159

Fig. 57. Tower Creek, above the canyon. September 6, 1921 160

Fig. 58. Slough Creek, showing connected pools. The placid surfaces indicate the pools,
the riffles indicate the gentle rapids connecting them. Photograph taken just
above gorge, through which the creek enters the Lamar River. August 11,1921.. 160

Fig. 59. Slough Creek meadows. August 29, 1921 163

Fig. 60. Smaller Peak northwest of Lake MacDonald, Glacier National Park, showing the
beds and confluences of temporary streams. While upper parts are dry, the

streams still contain water at the confluence. August 10, 1925 163

Fig. 61. Mt. Brown and others, Glacier National Park. Taken from Lake MacDonald, after

a summer blizzard. August 15, 1925 164

Fig. 62. Temporary stream resulting from summer blizzard; photograph taken on mountain

slope, about 2500 feet above Lake MacDonald, Glacier Park, August 17, 1925 ... . 164
Fig. 63. Bed of hillside stream (temporary stream), with trickle of water from spring. Sand-
stone formation, Clearwater River, Idaho. August 30, 1924 167

Fig. 64. Bed worn in sandstone by temporary stream, Clearwater River, Idaho. Area burnt

over by forest fire. August 30, 1924 167

Fig. 65. Lamar River eddy, looking upstream. To the left the high water mark of the spring
floods shows clearly. In June the stone in foreground was completely covered by

water. September 4, 1921 168

Fig. 66. Grand Canyon of the Yellowstone. Lower Falls at upper left. The characteristc

“ su'phur slides ” to the right. August 21, 1921 168

Fig. 67. “ Sulphur slides ” of Yellowstone Grand Canyon. Intermixed with these are
“ needles ”, which also are characteristic of the formations of Yellowstone Park.

August 21, 1921 171

Fig. 68. Lower Falls of the Yellowstone and rapids. Note intermixing of “ needles ” and

“slides”. August 21, 1921 171

Fig. 69. “ The Needles ”, looking directly across the Yellowstone in the Lower Canyon.

Note the layer of basalt rocks near the top. August 22, 1921 172

Fig. 70. Yellowstone River at Cooke City bridge. Lamar River Valley in upper middle.

August 8, 1921 172

Fig. 71. Rocks at eddy of Lamar River. Low water stage. In June these were completely

covered by the floods. September 4, 1921 177

Fig. 72. Stream bed of Lamar with recession pools. August 8, 1921 177

Fig. 73. Beaver pond on Lost Creek, showing hut in middle. June 26, 1921 178

Fig. 74. Lost Creek Falls. September 2, 1921 178

Fig. 75. Lost Creek, low water stage. In June all the stones were covered by water. Sep-
tember 2, 1921 1 81

Fig. 76. Lost Creek, low water stage. Showing the last pool in the bed of the stream. The

stream is dry for nearly a hundred yards above this point. September 2, 1921 . . 181

Fig. 77. Tower Creek, confluence with Carnelian Creek (right). September 6, 1921 182

Fig. 78. Tower Creek in Tower Creek Canyon, showing the thickets along banks and stones

in bed of stream. August 14, 1921 182

Roosevelt Wild Life Annals

PAGE

Fig. 79. Hot spring on shore of Yellowstone River. Covered in June. August 15, 1921 ... . 187
Fig. 80. Cleft boulder on meadow between Yellowstone and Lamar rivers. July 17, 1921. . . 187
Fig. 81. Brink of Lost Creek Canyon to right, and “ crater ” in foreground. The meadows
are said to be the bottom of the crater, the hills around forming the rim of an

extinct volcano. August 28, 1921 188

Fig. 82. Castle Rock (also called Cathedral Rock) on North Fork of Clearwater River,
Idaho. Showing types of shores and placid stretches of stream. September 2,

1924 188

Fig. 83. Clearwater River, Idaho. Showing sand deposits left by spring flood, and log jam

in river. August 31, 1924 193

Fig. 84. Clearwater River, Idaho. A closer view of log jam caused by spring flood. Also

showing “ The Gates ” and government bridge across river. August 31 , 1924. . . . 193
Fig. 85. Hibernating Coccinellidae. Photograph taken on a warm spring day when sun
brought out the beetles. Many had been washed into a creek about twenty feet

away. Photograph Moscow Mt., Idaho, about April 15, 1923 194

Fig. 86. Skeleton of winter-killed buffalo, near Lamar River eddy. Note the puparia of
flies on the scapula. The skull had been dragged into the eddy by some animal

(bear?). July 25, 1921 . 194

Fig. 87. Huge boulder in stream bed of Oro Grande River, Idaho. This boulder is fully
forty feet high. Smaller boulder can be seen on the slope along the river.

September 2, 1924 199

Fig. 88. Trick Falls, Glacier Park, in Two Medicine River, a short distance below Two
Medicine Lake. Photograph July 17, 1925. The same place was visited August
26; the lower part of the falls, which emerges from an underground conduit, was

then completely dry 199

Fig. 89. Keppler Cascade, Yellowstone Park. August 18, 1921 200

Fig. 90. Oro Grande River, Idaho. Freak roots of a tree on top of a boulder. One root
circles to the left and then turns under the tree. The boulder is surrounded by
water, and none of the roots passes into the stream bed; the boulder showed no

sign of being cleft 200

Fig. 91. MacDonald Creek, Glacier Park. Just above the falls the stream bed consists of

stratified rock. August 17, 1925 203

Fig. 92. Yellowstone River at Cooke City bridge. Showing rapids and marginal pools.

August 15, 1921 203

Fig. 93. Yellowstone River. Showing rock with “ splash flies ”. July 7, 1921 204

Fig. 94. Lamar River eddy, showing marginal pools. Later these pools became dry. Pool
upper left is completely cut off from stream. In the middle the moist bottom of

a recession pool shows clearly. July 16, 1921 204

Fig. 95. Yellowstone River, rapids and whirlpool, a short distance above Cooke City bridge,

July 7, 1921 ' 207

Fig. 96. Lost Creek, showing tangle and underbrush along stream. September 2, 1921 ... . 207
Fig. 97. Spider webs, with prey of aquatic insects. From rocks beside Lost Creek Falls.

September 2, 1921 208

Fig. 98. Details of a spider web, showing various insect adults. Lost Creek Falls. Sep-
tember 2, 1921 208

Fig. 99. Pteronarcys calif ornica nymphs on rocks, “ emerging ” from their nymphal skins.

July 25, 1921 211

Fig. 100. Rock with exuviae of Acroneuria. Oro Grande River, Idaho. July 3, 1924 21 1

Fig. 101. Cluster of Acroneuria adults in crevice of rock. Yellowstone River. July 6, 1921.. 212
Fig. 102. Stoneflies ( Acroneuria ) mating on shore grass along Yellowstone River. July

6, 1921 212

Fig. 103. Adult mayfly with parasitic worm ( Mermis sp. ?) emerging from the caudal end.

Enlarged times eight. Note the loop formed by the parasite within the abdomen
of the insect. From confluence of Lava Creek and Gardiner River, August 4, 1921. 217
Fig. 104. Stranded caddis worms {Brachycentrus and Rhyacophila). Lamar River. August

8, 1921 217

Fig. 105. Attachment of Brachycentrus cases. Some attached at caudal or small end, others

at broad or “ mouth ” end. Lamar River. July 25, 1921 218

Fig. 106. Larva of Bibiocephala, enlarged times 6. Note ventral suckers. From Snake
River, Idaho. At right, pupae of Bibiocephala, showing variation in size and

ventral attachments 218

Fig. 107. Bibiocephala pupae. From Lamar River. Note the tubes of Chironomids ( Tany -

tarsus ?) on the same rock. July 16, 1921 219

Fig. 108. Bibiocephala adults as “ splash flies ”. Oro Grande River, Idaho. September

3 , 1924 2I 9

Fig. 109. Deuterophlebia larva. Yellowstone River. July 30, 1921 220

Fig. no. Simulium sp. “ Combs ” of larva x 60 220

Ecology of Trout Streams

PAGE

Fig. hi. Atherix cluster. Lamar River, July 22, 1921. The insert shows a section of the

cluster enlarged about x 4 225

Fig. 1 12. Rock where Atherix cluster had formed, on underside in cleft of rock. Lamar River,
August 7, 1921. On July 22, when the cluster was photographed, the sharp edge

running from left to right was about twelve inches above the rapids 225

Fig. 1 13. Tipulid larva, with inflated “ baloon” at posterior end. Tower Creek, August 29,

1921 226

Fig. 1 14. “Spray flies” ( Chamaedipsis and Philolutra) on rock in rapids. Lamar River,

August 31, 1921 226

Fig. 1 15. Pamidae larva. Psepkenus leconlei enlarged x 10. From Lake Mendota, Wis. .. . 232
Fig. 1 16. Pamidae larva. Elmis viltatus enlarged x 10. From Lake Mendota, Wis 232

[i53]

Fig- 53 - Yellowstone River above Cooke City bridge. End of “Lower Canyon.”

Aug. 8, 1921.

Fig. 54. Clearwater River. Idaho, a short distance above mouth of Oro Grande
River, showing type of shore, stream bed, etc. This region was swept by forest
fires in 1915; the trees on the slopes are fire-killed. Aug. 30, 1924.

[ 154 ]

THE ECOLOGY OF TROUT STREAMS IN YELLOWSTONE

NATIONAL PARK

By Richard A. Muttkowski*

Collaborator, Roosevelt Field Naturalist, Roosevelt Wild Life
Forest Experiment Station, Syracuse, Nezv York

CONTENTS

PAGE

Introduction 156

Mountain Trout Streams 158

Classification of Mountain Trout Streams 158

Description of the Type Streams 161

Criteria of Mountain Stream Life 166

Stream Physiography 166

Stream Physiology 174

Qualitative List of Plants and Animals 179

Plants 179

Animals 180

Ecology of Mountain Trout Streams 206

General 206

Classification of Habitats 209

The Habitats 210

Adaptations to Mountain Stream Life 221

Food Relations 222

Trout Food 222

Insect Food 230

Seasonal Food Cycle 233

Comparisons and Summaries 234

Bibliography 238

* Professor of Biology and Head of the Department of Biology, University of Detroit,
Detroit, Michigan.

Roosevelt Wild Life Annals

is6

INTRODUCTION

Under the auspices of the Roosevelt Wild Life Forest Experiment Station
at the New York State College of Forestry, Syracuse, N. Y., the writer was
afforded an opportunity during the summer of 1921 to examine and study trout
streams in the northwestern section of Yellowstone National Park. Camp
Roosevelt, operated by the Yellowstone Park Camps Company, was the base
from which these studies were made, covering the period from June 18 to
September 9, inclusive.

Due to unforeseen complications, chiefly in the way of health and change of
position, the completion of the manuscript was delayed till 1927. In the mean-
time the writer has had ample opportunity to study other mountain streams in
a comparative way in Idaho, Washington and Montana. Chief among these are
the Palouse, Clearwater, Oro Grande, Snake, and St. Joe rivers in Idaho and
Washington; and the streams comprising the MacDonald drainage, Two Medicine
drainage, St. Mary’s drainage and Swift Current drainage in Glacier National
Park in Montana. The Glacier Park studies were made during the summer of
1925 and included a number of lakes and streams. The results as far as the
lakes of Glacier Park are concerned are to be published in a separate paper.
Data pertaining to streams are, however, incorporated in the present paper rather
than in a separate account.

Acknowledgments. In Yellowstone Park the writer was variously aided
by different persons, notably Dr. Gilbert M. Smith of the University of Wiscon-
sin, and more recently of Stanford University (see collaboration, Muttkowski-
Smith, the Food of Trout Stream Insects in Yellowstone National Park, in the
Roosevelt Wild Life Annals, Vol. 2, No. 2), Mr. Alvin G. Whitney of the
Yellowstone Park Forest and Trail Camp, Mr. E. R. Warren of Colorado
Springs, Colo., and his assistant Mr. Ellis L. Spackman. Mr. Horace M. Albright,
the Superintendent, facilitated the work by granting certain special permissions.
Mr. M. P. Skinner, then Park Naturalist, and the rangers of the Tower Fall
Junction Ranger Station assisted in various practical ways and by suggestions
as to favorable localities. Above all, I am indebted to Dr. Charles C. Adams,
until recently Director of the Roosevelt Wild Life Forest Experiment Station
for many suggestions and direct aid; and to Mr. Howard H. Hays, then President
of the Yellowstone Park Camps Company, for many courtesies, conveniences, and
facilities, without which this study would hardly have been possible. To the present
staff of the Roosevelt Station I am indebted for the laborious task of preparing this
paper for the press and seeing it through.

To the Zoological staff of Cornell University I am indebted for identifica-
tions of various specimens: Dr. Needham and Dr. A. Claassen identified the
Perloidea, Dr. Needham the Ephemeroidea, Mr. C. K. Sibley the Trichoptera,

Ecology of Trout Streams

x 57

Dr. A. O. Johannsen the Diptera, and Mr. H. Dietrich the Coleoptera. Most of
the insects were in the nymphal or larval stages and the task of identification was
therefore none too easy. To all these specialists I wish to express my grateful
acknowledgment.

In general, the plan of the present paper follows that of an earlier one,
the Fauna of Lake Mendota (Muttkowski, T8). That plan has received the
sanction and approval of a number of ecologists, so that I do not hestitate to
repeat it in the present study although slightly modified.

158

Roosevelt Wild Life Annals

MOUNTAIN TROUT STREAMS

Classification of Mountain Trout Streams. Trout streams are highly
aerated streams, found most abundantly in mountain regions. In the plains there
are many “meadow” streams which also contain trout. These, however, are
only of secondary interest here, since they are rare in mountainous regions.

Trout streams in Yellowstone Park and elsewhere in the Northwest can be
broadly classified into three groups, according to certain physical characters :

a. Constant streams, with large amounts of water, which are, relatively speak-
ing, only slightly diminished in summer. This group comprises the larger streams,
such as Yellowstone River (Fig. 53), Gibbon and Firehole rivers in Yellowstone
Park; the Snake, Boise, Clearwater (Fig. 54), Salmon, Coeur d'Alene and other
larger rivers of Idaho; the Flathead and St. Mary’s rivers in Glacier Park.

b. Variable or flood streams, which carry enormous masses of water for
brief periods in the spring, due to melting snow, and then rapidly recede to one
tenth or less of their flood level. These streams contain many rapids and in
flood period may attain a formidable size. As examples may be mentioned the
Lamar (Fig. 55), Gardiner, and Firehole rivers in Yellowstone Park, the forks
of the Elk and Clearwater rivers and the Oro Grande (Fig. 56) and Palouse
in Idaho and Washington, the Two Medicine River, MacDonald Creek and
Swiftwater Creek in Glacier Park.

c. Precipitous or cascading streams, with many falls and rapids. Like the
foregoing they are raging torrents during a brief spring, after which they shrink
to unimpressive proportions, indeed, often to little more than mere trickles. This
type is abundant throughout the Rockies. For the Yellowstone Park. I may
cite Soda Butte Creek, Deep Creek, Broad Creek, Hellroaring Creek, Lava Creek,
Lost Creek, etc. Tower Creek (Fig. 57) is an intermediate between b and c. but
faunistically like c.

Yellowstone Park contains another stream which is unique and the sole
representative of its type. This is Slough Creek, which consists of several
tiers of meanders, and is located in the northeast corner of the Park (Fig- 58 ).
There one finds splendid meadows, several miles in length, through which the
stream meanders placidly (Fig. 59). The stream leaps from one meadow to
the next through short hut magnificent gorges. But the chief characteristic
of the stream is probably glacial in origin ; for the meanders consist of a series
of deep holes, lined with huge boulders, which are connected by the quietly flow-
ing stream. The rapids are confined to the gorges. In the “holes" or quiet pools
one finds some of the largest trout in the Park.

Other mountain streams may course through long meadows and form
meanders ; but the “holes" are only occasional and do not consist of a series of
depressions such as one finds in Slough Creek.

d. Temporary streams, or snow flood streams. These are not trout streams
at all, but are included on account of their physical rather than biological rela-
tionship. This type includes the precipitous cascades and trickles that arise from
melting snows and which for a time may carry considerable quantities of water,
but which diminish to weak rills by the middle of summer and freeze up com-

159

Fig. 55. Junction of Yellowstone and Lamar rivers at low water stage. View
downstream; Yellowstone River at left, Lamar entering at right. In spring
the boulders in the foreground are under water. Sept. 5, 1921.

Fig. 56. Oro Grande River, Idaho. About 300 yards above it's entry into the Clear-
water. Showing rapids and boulders. In spring all except the largest boulders
are covered by water. Aug. 31, 1924.

i6o

Fig. 58. Slough Creek, showing connected pools. The placid surfaces indicate
the pools, the riffles indicate the gentle rapids connecting them. Photograph
taken just above gorge, through which the creek enters the Lamar River.
Aug. 11, 1921.

Ecology of Trout Streams

161

pletely in the fall (Fig. 60). Snowstorms and summer blizzards may give these
streams a temporary prominence (Fig. 61). They become conspicuous as the
snow melts, and the noise of their torrential fall can be heard at great distances
(Figs. 62, 63, 64). Quite generally they end in sumps or swamps on the moun-
tain sides or at the base of the slopes. Despite their temporary existence they
have a fauna and a flora ; and their terminal sumps supply abundant habitats
for the early summer plagues of mosquitoes so common in mountain regions.

In selecting streams for comparative study it was essential to consider
local convenience, accessibility, and type of stream. Four were selected for
their location near Camp Roosevelt, which was used as a base. It would be
difficult to find another point in the Park so favorably located, with repre-
sentative types of so many kinds of streams all within convenient walking
distance. After a preliminary survey, the following were chosen for intensive
study: Yellowstone River, Lamar River, Tower Creek, and Lost Creek.

Yellowstone River was chosen to represent the first type, the large “con-
stant” stream. The river is less than a mile distant from Camp Roosevelt and
offers many ideal habitats on both sides of the Cooke City Road bridge, within
convenient distance.

Lamar River was selected to represent the second type, the variable or
flood stream. The Lamar joins the Yellowstone about three miles from Camp
Roosevelt, and about a mile below the Yellowstone bridge. Excellent habitats
were found in the gorge and eddies (Fig. 65) within the first few miles above
the junction.

Lost Creek, although containing no fish at all, was selected for its accessi-
bility and to represent the cascading or precipitous type of stream. It passes
right through Camp Roosevelt, and its biota typifies such streams as Lava Creek,
Tower Creek, Buffalo Creek, Blacktail Deer Creek, Hellroaring Creek, etc.
Care was taken to check up Lost Creek frequently with Tower Creek, the fourth
stream.

Tower Creek at Tower Fall is about two miles distant from Camp Roosevelt.
But since it is much frequented by campers for about three miles above the
mouth, the distance to the undisturbed sections made it rather inconvenient
for frequent study. Besides, Lost Creek, as already noted, showed a practical
identity of biota; hence one trip a week was considered sufficient for the study
of Tower Creek.

The fauna of each type showed marked individuality. The work of com-
parison was begun at a favorable period, just preceding the summer emergence
or transformation of insect forms. This seems to follow shortly after the spring
flood period. Later on, after the first six weeks, as opportunity offered, hurried
studies were made of other streams for comparison, such as the Gardiner and
Firehole rivers, Lava Creek, Blacktail Deer Creek, Carnelian Creek, Elk Creek,
Pelican Creek, Plateau Creek, and Buffalo Creek on the Lamar and Yellowstone
rivers. Three days were also spent in a study of the biota of Yellowstone Lake,
particularly in its relation to Yellowstone River.

Description of the Type Streams, a. Yellowstone River. The Yellowstone
River arises far south of Yellowstone Lake. For the purpose of this paper only

Roosevelt Wild Life Annals

162

the stretches below the lake are considered. After leaving the lake, the river
is rather broad and in numerous places tends to become shallow and swampy,
particularly on the right shore. Notwithstanding the swift current, considerable
vegetation is present, such as Myriophyllnm, Ceratophyllnm, some Potamogeton,
and the plumes of algae (probably Cladophora) . This is confined to the lateral,
quite shallow portions of the stream. In the adjacent swamps Castalia, N ymphaea
and other swamp plants attempt to invade the stream, with little success.

About a mile above the Upper Falls the current increases its speed and
plants become very rare, except Cladophora, which persists nearly to the
brink of the falls. Just below the mouth of Alum Creek the river bed narrows
and the shores become higher and steeper. The water, till then smooth and
silent, now becomes turbulent, noisy and streaked with foam. Entering a short,
rocky canyon, the river leaps over the brink with a vertical drop of 109 feet,
forming the Upper Fall.

Here the aspect of the shores changes. For the Grand Canyon has begun,
with its multitude of colors, its precipitous “needles” and rocks, its gav slides
of sulphur sands — all thrown together in fantastic design. For twenty miles
the canyon extends, with alternating areas of strangely colored formations and
bluish gray cliffs (Figs. 66 and 67).

A third of a mile below the Upper Fall, the river takes a second leap of
308 feet, the Lower Fall (Fig. 68). The river then plunges into a series of
cataracts, rapids and lesser falls, for a distance of twenty miles, or the length
of the Grand Canyon. The total descent from Yellowstone Lake to this point
is from 7740 to 6400 feet elevation.

A short distance above Tower Fall the canyon flattens out and the river
becomes more placid. Tower Creek marks the entry into the Lower Canyon
(“The Needles.” Fig. 69). Here the shores are formed of steep, blue-grav
cliffs, interspersed with minor slides of gayly colored sulphur sands. Below
this is the Cooke City Road bridge (Fig. 70). Beyond the bridge the river is a
series of rapids and eddies, winding between low and high hills, with occasional
gorges and short canyons.

The sound of the river is a dull roar. If one approaches it — be it on a
low shore, or between high cliffs — one may feel quakes and tremors of varying
violence, giving one a curious feeling of uncertainty. Combined with this is
the stench of the river, of sulfuretted hydrogen, like foul eggs, due to the masses
of sulphur carried, and still more to the innumerable hot springs, small geysers,
and mud swamps along the shores. The sulfuretted odor and taste of the water
is never quite absent from the river; indeed, fish taken from the Yellowstone
River have a strongly sulphurous taste.

b. Lamar River. The Lamar River drains practically the entire northeastern
section of Yellowstone Park. More specifically, to the west and south it drains
the watershed between the Lamar and Yellowstone rivers and Y'ellowstone
Lake. To the east, it collects the waters from the Absaroka Range. For the
greater part it flows northwest, its bed lying in a basin between the foothills of
the Absarokas to the east, Specimen Ridge and Mirror Plateau to the west, and
a series of peaks to the south, including Pelican Cone, Mt. Chittenden, Cathedral
Peak, and Pyramid Peak.

163

Fig. 60. Smaller Peak northwest of Lake MacDonald, Glacier National Park, show-
ing the beds and confluence of temporary streams. While upper parts are dry,
the streams still contain water at the confluence. Aug. 16, 1925.

164

Fig. 62. Temporary stream resulting from summer blizzard; photograph taken
on mountain slope, about 2500 feet above Lake MacDonald. Glacier Park,
Aug. 1 7, 1925.

Ecology of Trout Streams

165

In its northward course it is joined by dozens of rills and creeks. Thus,
on the left it receives Cold Creek, Willow, Timothy, Clover, Flint, Opal,
Amethyst, Jasper and Crystal creeks ; on its right it receives Miller, Calfee,
Cache, Soda Butte and Slough creeks. Of these, Slough Creek and Soda Butte
Creek are the most important tributaries.

Lamar River has no conspicuous falls or cascades, but drops steadily and
gently from 7500 feet elevation at its point of entry into Yellowstone Park to
5800 feet at its junction with the Yellowstone. It is a beautiful stream, never
very wide, with alternating placid stretches and gentle rapids. At points where
tributaries enter, it widens and becomes turbulent, since most of the tributaries
are precipitous and the force of their waters impound those of the Lamar,
causing eddies, lateral gorges, and roaring rapids.

During the spring flood the Lamar rises to formidable heights. On June
20, 1921, the flood waters covered the rocks shown in figure 71 (see also Fig.
55). Examination of stranded debris indicated that the waters had been
several feet higher during the preceding week. Even then the river was

spectacular, both in appearance and sound — so much so, that as one stood on the
point above its junction with the Yellowstone, the question suggested itself,
“Does the Lamar feed the Yellowstone, or the Yellowstone the Lamar?” The
Lamar seemed to carry a greater volume of water at the time.

A few days later, June 25, the river had fallen considerably, the boulders
noted in figure 71 being wholly exposed. By July 16 the stream had fallen to
the level noted in the photograph. The total drop was approximately fifteen
feet.

Following the quick recession of waters, the Lamar forms an ideal habitat
for Cladophora; the whole river bed is covered with the green streamers, so that
from an elevated point some hundred feet away the waters appear a brilliant
green. With the recession a very marked silting occurs in the shallower places,
which has a striking effect on the stream life (Fig. 72).

c. Lost Creek. This is a typical mountain creek, carrying large quantities
of water for short periods, and gradually shrinking to an unimpressive trickle,
with perhaps less than one-fifteenth of its flood volume. Its main drainage
is from the northern slope of Folsom Peak. Its banks are, on the whole, steep
but not high. Smaller beaver works (Fig. 73) occur for six miles above Camp
Roosevelt. There a series of small pools is connected by trickles, with beaver
dams impounding the scant waters.

Lost Creek is not a trout stream, but its fauna and flora are typical of the
smaller trout streams. For a short distance above its falls it becomes quite
precipitous. It enters a short gorge, then tumbles over a cliff about sixty feet into
Lost Creek Canyon (Fig. 74). Its bed then comprises shattered rocks, and some
fair-sized boulders (Fig. 75). Over these it rushes in a series of steep cascades.
At the base of the hill it spreads out in a delta and then disappears (Fig. 76)
into the ground for several hundred yards — hence the name “Lost” Creek. It
reappears along the Cooke City Road, which it parallels for about 500 yards
down to Yellowstone River. Here it flows over a series of shelves, which in
1921 formed the site of a beaver colony, which was abandoned in 1922.

1 66

Roosevelt Wild Life Annals

d. Tower Creek. This creek drains the elevated plateaus between Mt. Wash-
burn and Folsom and Observation peaks, with headwaters arising at an altitude of
about 10,000 feet. In its course of about twenty miles it drops to 6400 feet at
Tower Fall, just before it empties into Yellowstone River.

In general Tower Creek is about twenty feet wide, and sixteen to twenty-
four inches deep, with clear, cold water. Its bed is composed of water-worn
cobbles, from a few inches to a foot in diameter, with occasional larger stones
or boulders five to ten feet in diameter. In its upper reaches the banks are flat
and frequently spread out to form broad pastures and willow swamps (Fig. 57),
with adjacent beaver works built around pools or springs. But the stream itself
is unobstructed, although frequent attempts at damming are evident. In its
course Tower Creek receives the drainage waters of innumerable springs, pools
and swamps, as well as of hillside sumps.

About seven miles above its mouth it receives the waters of Carnelian Creek
(Fig. 77), a narrow and shallow stream with considerable fall, which is used
extensively by beavers for a series of remarkable works (See Roosevelt Wild
Life Bulletin, Vol. 1, No. 2, pp. 187-221, and Roosevelt Wild Life Annals, Yol. 1,
Nos. 1 and 2, pp. 13-191, for E. R. Warren’s papers on the Yellowstone Beaver).
Two miles farther downstream the hills approach each other, and the creek
enters Tower Canyon, throtigh which it threads for about five miles (Fig. 78),
emerging in a short gulch which parallels the highway. Here the bed is strewn
with large rocks, up to ten feet in diameter, over which the water cascades in a
series of short, turbulent leaps. The gulch continues for approximately 150
yards, then twists sharply, and ends between the characteristic needles which
mark the entrance to the Lower Canyon of the Yellowstone. There the creek
plunges over the cliff for a fall of 132 feet, and after a turbulent passage of
100 yards, joins the Yellowstone.

CRITERIA OF MOUNTAIN STREAM LIFE

The biota of any environment is determined by physiographical and physio-
logical conditions. Physiography deals with the structural aspects of the envi-
ronment, physiology with the functional aspects. For the first of these there
are two viewpoints : — that of the geologist who deals with topographical classi-
fications, and that of the ecologist who classifies the environment in terms of
habitats. Hence under Physiography both the topography and ecology of streams
must be discussed.

Stream Physiography. The Physical Environment. Under this head the
sources of the streams and the structure of the shores and stream beds can be
considered.

Sources. The sources of water are primarily the melting ice and snows
of the mountains, and secondly the innumerable hot (Fig. 79) and cold springs
found in the Northwest. Besides these, many swamps, sumps, and ponds drain
into the streams.

The Shores. Geologically Yellowstone Park is of mixed origin — igneous,
volcanic, glacial and alluvial. This is shown in the character of the various shores

1 67

Fig'. 63. Bed of hillside stream (temporary stream),
with trickle of water from spring. Sandstone forma-
tion, Clearwater River, Idaho. Aug. 30, 1924.

Fig. 64. Bed worn in sandstone by temporary stream, Clearwater
River, Idaho. Area burnt over by forest fire. Aug. 30, 1924.

i68

Fig. 65. Lamar River eddy, looking upstream. To the left the high water mark of
the spring floods shows clearly. In June the stone in foreground was completely
covered by water. Sept. 4, 1921.

Fig. 66. Grand Canyon of the Yellowstone. Lower
Falls at upper left. The characteristic “sulphur
slides” to the right. Aug. 21, 1921.

Ecology of Trout Streams 169

and streambeds. Nearly every stream in the Park at some point or other passes
through rugged canyons composed of basalt cliffs. Decomposed lava slides are
frequent. Along the streams and on the uplands one finds numerous boulders
left by glaciers (Figs. 80 and 87), while in places there are great accumulations
of soil left by past floods. Much of the area in the region of Camp Roosevelt
is thought to be the crater of a huge extinct volcano (Fig. 81).

A feature of the shores of mountain streams is the changes taking place,
produced by two factors, (a) snow, and (b) the streams. The changes are
slow but they are nevertheless easily recognizable. The streams for the greater
part have eroded their beds considerably, so that the present shores are quite
steep, even perpendicular in many places (Fig. 82). During the spring floods
the present shores are undermined ; this tends to produce land slides, which in
turn, precipitated into the waters, affect the stream beds.

On the steep slopes the long winters deposit gigantic snow drifts, whose
weight, pressing on trees, bushes and the loose earth, dislodges these and grad-
ually forces them into the stream. During spring the melting snows may cause
great sections of the shore slopes to slip into the stream. This is particularly
true of Yellowstone River, where the slides are numerous, and the shores for
many miles are quite declivitous (Figs. 66-68). It is also true for streams
like the Lamar, Tower Creek, Lost Creek and many others. At many points
along' these streams there were signs of recent “slips” during the summer of 1921,
and of others of older date.

In places the trees, shrubbery, and debris thus precipitated into the stream
may become caught between or against protruding rocks and cause a damming
of the stream. Jammed trees are frequent in the great rapids of western streams
(Figs. 83 and 84).

Wintering animals, such as the Coccinellidae (Fig. 85), may be buried under
the snow and later swept into the water by the snow slides. Others, including
mammals such as buffalo, deer and elk, may become mired in the snow-drifts
and are suffocated (Fig. 86) ; these, too, may be carried into streams by snow
slides, since quite frequently one may find parts or whole skeletons of winter
killed mammals in the streams.

The Stream Beds. Change, too, takes place in the stream beds. The
force of the spring waters in particular brings about many marked annual
changes. The flood currents wash out all the lighter materials such as sand,
mud, silt, and deposits on rocks, so that the bottoms appear clean-washed or
scoured. Later with the recession of the floods, muck, silt, smaller pebbles and
sand are deposited along the shores among the rocks.

In midstream the floods tend to fill up the pools. In turn, they may wash
out new pools. Even during the summer, with water at lowest stage, and
depending somewhat on the size of the stream, there is a constant movement of
the bottom. Stones of varying size are displaced, ground together or battered
against each other, and whatever living creatures may come between will be
crushed. A water net, or still better, a plankton net, left in a stream just below
some rapids, will collect abundant evidence of this fact in the way of crushed
caddisworms, mayfly and stonefly nymphs, and occasional trout eggs and embryos.

i yo

Roosevelt Wild Life Annals

The Ecological Habitats. Like other ecological units, the mountain stream
may he subdivided into minor habitats, which can be readily recognized accord-
ing to physical characters and the biota. These habitats intergrade, here as
elsewhere. Following is a classification, based on comparative studies of the
streams of Yellowstone Park and other regions of the Northwest:

1. Permanent habitats, with native (endemic) biota.

a. White water habitats — falls, cascades, white rapids.

b. Clear rapids and stone bottoms — on and under rocks.

c. Placid water habitats- — pools and holes.

d. Marginal areas — on or under rocks, in soil.

2. Interrupted habitats, with native biota.

e. Deposits— on rocks, bottoms, or shores.

f. Splash areas — on rocks.

3. Temporary habitats — transient and transitional, with varied biota.

g. Marginal pools.

h. Recession areas.

The permanent habitats ( 1 ) are stable and are typical of the center of the
streams; their plant and animal life varies but slightly the year round.

Interrupted habitats (2) are found both in the center and at the sides. The
transient and transitional habitats (3) comprise relatively quiet spots, temporary
in existence, that dry up in low water stage and are destroyed by high water.
The biota in such areas is extremely varied.

Permanent habitats. As already noted, these comprise the central parts
of streams ; they offer the greatest degree of constancy in mountain streams.
In such places one may find the same fauna and flora practically the year round.

a. White water habitats. As indicated by the name, these are character-
ized by foaming or “white” water. That is, here the current and the mass of
w r ater are sufficiently swdft and great and the obstacles opposed to its flow
sufficiently firm to batter it into a white foam. Such white water is that of
falls, cascades, and major rapids. Falls (Figs. 66, 68, 74, and 88) indicate a
vertical flow of water; cascades an oblique flow (Fig. 89); white rapids (Fig.
95) are a mixture of horizontal, oblique and vertical currents. In mountain
streams all of these are intermingled. They constitute a highly specialized
habitat, combining strong and twisting currents and highest aeration.

b. Clear rapids and stone bottoms. Here are included rapids of less
violence, with fewer opposed obstacles, so that the water is transparent and
one may readily see the bottom. Very characteristically the bottom in these
rapids is nearly uniformly composed of rounded stones of various sizes (Figs.
87 and 90). But there may also be angular granitic rocks, and angular strati-
fied rocks (Fig. 91). Thus, in the Yellowstone, all types of bottom may be
found. At one point, just below the largest group of “needles" in the lower
canyon, there is evidence of a formation similar to those of Mammoth Hot
Springs ; this formation tips into the stream and extrudes at low water stage.
But the most common condition, which applies probably to more than 90 per
cent of the stream beds, is that of rounded, water-worn cobbles (Fig. 57).

Fig. 67. “Sulphur slides” of Yellowstone Grand Canyon. Intermixed with these
are “needles,” which also are characteristic of the formations of Yellowstone
Park. Aug. 21, '.921.

Fig. 68. Lower Falls of the Yellowstone and rapids.
Note intermixing of “needles” and “slides.”
Aug. 2i, 1921.

172

Fig. 69. “The Needles,” looking directly across the Yellowstone in the Lower
Canyon. Note the layer of basalt rocks near the top. Aug. 22, 1921.

Fig. 70. Yellowstone River at Cooke City bridge. Lamar River Valley in upper

middle. Aug. 8, 1921.

Ecology of Trout Streams

i/3

c. Placid water habitats. These are the “holes” and “pools” in the beds
of the streams and are of varying size. They may have been washed out by falls,
gouged out by stones born along by the waters, or they may be of glacial,
volcanic, or perhaps seismic origin. At the base of falls the pools are turbu-
lent and belong to the “white water” habitats; otherwise they are relatively
quiet at the bottom, since the current is strong only in the surface strata.
Often such pools have a scanty floor of gravel, or a sandy bottom — deposits
that have been made after the spring floods.

d. Marginal areas. The shores may comprise cliffs, boulders, or smaller
stones, and are rarely sand, pebbles, and mud. Here the impact of the water is
primarily lateral, not horizontal or vertical (Fig. 92). Friction by debris is
marked. This is an area offering many shelters, but too often of only tem-
porary nature. Many smaller organisms live under and between the rocks, while
others live in the moist or wet soil.

Interrupted habitats. These comprise primarily the deposits (e) on
the bottoms, shores, and sub-surface rocks. At first these were regarded as
transient in nature, since the deposits are absent during the flood period and
accumulate progressively during the summer. Still, one finds these deposits
in mid-stream, often exposed to a fairly strong current, and covering the huge
rocks that so often form the margin of the “holes” and pools. The deposits
persist during the winter and only the scouring spring floods sweep them away.
But during their existence they afford a habitat for numerous Chironomidae
and rock diatoms. The deposits extend from mid-stream to the shore rocks.
In Yellowstone Park they contain much sulphur sand.

f. Splash areas. These are the exposed rocks in rapids or along shores,
which are continually moistened by spray. Almost any time during the year —
spring, summer, fall, winter — one may find adult insects running up the rocks,
flying into the spray, and once more running upward. They may be found
chiefly on the downstream side of the rocks. Such are the Diptera, Trichoptera,
and more rarely the mayflies (Figs. 93, 108 and 114).

Temporary piabitats. These comprise (g) the marginal pools and shallows
which are found during a large part of the year, and (h) the recession areas,
following the spring flood. One might also refer to them as transitional and
transient areas ; — transitional as regards the biota, transient as regards their
period of existence. They are the lesser shore rapids, which silt in quickly,
and the marginal pools that appear with the recession of the flood waters.
Later on the interstices between the stones become filled in with sand and
debris (Fig. 94). Tiny rapids or trickles continue to fill in the connections,
until the pools are completely separated. Often a series of marginal pools is
thus formed. The trapped fauna either migrates or dies, and the temporary
habitat becomes populated with invading biota. It is remarkable how quickly
such a temporary habitat is invaded by aquatic beetles, water bugs, Chiro-
nomidae, Rotifera, Protozoa, filamentous algae, and scattering plankton ’Crustacea.
All of these types come primarily from the swamps, sumps, pools, and springs
found so abundantly in mountain areas. After a few weeks these places dry

W4

Roosevelt Wild Life Annals

up and with them the biota disappears. Or, if a pool is fairly deep and does not
dry up completely, then the fall rise of the waters will wash away the invading
hiota and restock the pool with native forms.

Habitat Distribution. A striking difference between mountain streams and
other water bodies such as lakes, ponds, and slow plains streams, is the absence
of regularity in the arrangement of habitats. In mountain streams the
habitats are all commingled; pools may follow falls, rapids alternate with eddies,
marginal pools may adjoin cataracts. In other words, there is a complete
absence of that regularity which permits a zonal classification for lakes, ponds,
and slow streams. On the contrary, the contrasts are quick and startling, one
extreme follows another. Hence in a very small area one may find all the
extremes from a vertical to a horizontal flow of water, from shallow, placid
stretches to deep, stone-lined pools. This, in a sense, is advantageous, since it
permits the study of diverse conditions in a rather short time. A second advant-
age arises from this, namely, that the deeper portions of the stream approxi-
mate the flood conditions, while shallower places permit more intensive study
of the biota and its relation to other communities.

In summary, there is in mountain streams only a horizontal distribution
of habitats, no vertical. The habitats must be studied in horizontal order, not
in the vertical order one finds in lakes and slow streams.

Stream Physiology. From a physical viewpoint, mountain streams are
distinctive. Parallel with this is the functional or physiological aspect, with
equally distinctive features, which depend upon extrinsic and intrinsic factors.
The extrinsic factors are such as affect the stream as an entity, namely, the
molar agents, including winds and currents, chemical makeup, temperature, and
circulation. Intrinsic factors are those which affect the makeup of the flora
and the fauna. These are currents, respiration, food supply, footholds,
locomotion, and special habits.

Extrinsic Factors. Molar Agents (Winds and Currents). — Winds do
not directly affect the mountain streams. Since the streams have bitten and
scoured their beds deeply, the shores are usually high and the waters are there-
fore sheltered from winds. The rush of the waters, however, generates definite
air currents (Fig. 95) which entangle with the shore-winds high above them
and create vortices and treacherous back drafts which readily draw unwary
insects and other winged life from their line of flight and spill them into the
hurrying waters beneath. This results in the surface toll or “surface bait" of
the summer months.

In the winter the surface winds carry snow into the gorges and canyons,
piling it on the shores of their streams in huge drifts, to be dislodged by wind
vortices or undermined from beneath. Avalanche after avalanche then crashes
down the slopes, dislodging trees, boulders, or causing huge land slides into the
streams. In the spring the melting snows are readily displaced by the winds and
slide into the streams. For weeks after the snow has disappeared from the lower
levels the streams are turbid from the sands and sediments washed along by the
currents.

Ecology of Trout Streams

175

To a lesser extent the currents, especially during the flood periods, under-
mine the shores and cause slides. For it is noteworthy that most slides start
from above and not from below.

Specifically, the currents are concerned with molding the bed of the stream,
the winds primarily with molding the shores (Fig. 96).

Chemical Composition. In general, mountain streams are noted for the
purity of their water. This is true also for the streams of Yellowstone Park,
except Yellowstone River and Firehole River. These two streams are both
distinguished by a sulphury taste and odor the year round, rendering the water
unpalatable and incapable of quenching thirst. Two factors seem to
cause this: (a) The nature of the river shores and bottoms, from which masses
of sulphur sand are constantly dissolved and swept along by the currents; (b)
the innumerable hot and cold springs and geysers along the shores, which con-
tain sulphur, and sometimes arsenic and other compounds. Some of the springs
may rise from the river bed ; hence they are submerged for a good part of the
year and become exposed only during the low water stage of summer. The
spring seen in figure 79 is of this type, and was exposed only when the water
had dropped about twenty inches.

Practically all other streams in the Park carry considerable amounts of
sulphur in suspension during the flood period and their waters are not very
palatable and refreshing during that time. But as the waters grow clear,
that is, as the sulphur sands have silted out, they improve in potability. More-
over, the fish then lose the sulphury taste they all seem to have during the
flood period.

Turbidity. Due to the materials in suspension, most mountain streams are
very turbid during the flood periods. Usually by the middle of July the waters
become clear. Temporary turbidity, however, is quite frequent, and is caused
by rains and by shore slides. Unless the rains have been very severe, the
temporary turbidity lasts only a few hours. Where a mountain stream passes
through large upland meadows, as is the case with the Gardiner River and
Slough Creek in Yellowstone Park, the Clearwater, Oro Grande, Wietas, and
Forks of the Elk River in Idaho, and the Flathead and St. Mary’s rivers in
Glacier Park, — the rains wash down considerable quantities of mud, and this
requires a much longer time to sediment out than do the sulphur sands.

As a matter of fact, local turbidity occurs quite often, due to the dislodg-
ing of some rock by the current or some passing animal. This temporary and local
turbidity seems to have the effect of making the moving fauna scurry into shelters.

Aeration. The continuous and marked descent of waters in mountain
streams results in a thorough churning and mixing of water with air, and hence
in a high oxygen content. Oxygenation reaches its highest level in mountain
waters since, besides the churning, the low temperature permits the absorption
of greater quantities of oxygen.

Temperature. A characteristic of a mountain stream is the uniformity of
its temperature. The constant churning and mixing does not permit a thermal
stratification such as one finds in quiet streams, lakes, ponds and swamps.

Roosevelt Wild Life Annals

176

Temperatures taken in Yellowstone River at the bottom of deep “holes,” in-
eddies and in rapids, were all identical. There is a difference among streams,
however. Small creeks such as Lost Creek rose rapidly in temperature ; here
the heat reflected from the slopes raised the temperature of the water as much
as 8° C.

An interesting fact is that the waters of Yellowstone Lake, before passing
into the outlet, absorb considerable heat from the sun. As a result, the river
in later summer showed a higher temperature than the Lamar River, or Tower
Creek. Streams that are fed by many springs do not freeze over in winter.

Circulation In lakes, ponds and slow streams circulation is affected'
chiefly by winds and convection currents. As a result a thermal stratification
is established and with it a stratification of dissolved gases. In mountain streams
circulation is direct and complete. The constant descent of water results in a
thorough mixing, so that both temperature and oxygen content are equalized
throughout all parts of the stream. This is one of the most marked physiological
differences between mountain streams and other aquatic bodies.

In summary, current caused by the descent of water is the primary extrinsic
factor, which in turn affects the aeration, temperature, circulation and
transparency of the stream.

Intrinsic Factors. Except for certain Tipulidae larvae burrowing along the
shores among rocks, the primary fauna of trout streams is confined to water
breathers, i. e., such as obtain their oxygen from the water. The speed of the
current precludes any sort of emergent vegetation, which latter is essential for
surface breathers, i. e., breathers of atmospheric air (see Muttkowski, ’21).
Among animals, therefore, only such as breathe oxygen in solution in the water
are found — in general, only such types, especially among insects, as require a
high degree of oxygenation.

Because of the currents, the biota is further limited to species that are
either strong swimmers or strong dingers — the latter directly by means of struc-
tural adaptations (claws, suckers) or indirectly by means of clinging devices
constructed by the various species (webs, jellies, attached cases).

The current causes a third limitation, namely, in the matter of food supply.
Species living here cannot limit themselves to plant or animal food, or as in
quiet habitats, to some one particular type of plant or animal. On the contrary,
favorable locations here are more isolated — “biotic islands,” one may say, where
a few plants and animals assemble ; and a hungry specimen must eat whatever is
available. Indeed, one finds that habitually carnivorous and herbivorous species
both become mixed feeders in mountain streams. On the whole, the highly special-
ized conditions call for specialization in structure, but for generalization in diet.

In summary, the current and resultant high oxygenation are the dominant
intrinsic factors which affect the physiology of a stream. These at once limit
the biota (a) to water breathers, (b) to active swimmers and dingers among
water breathers, and (c) to indiscriminate feeders, i. e., species of unrestricted
diet.

A further restriction arises from certain special habits. As noted in a
previous paper (Muttkowski, T8, p. 386), a special habit may eliminate a species

177

Fig. 7 1. Rocks at eddy of Lamar River. Low water stage. In June these were
completely covered by the floods. Sept. 4, 1921.

Fig. 72. Stream bed of Lamar with recession pools. Aug. 8, 1921.

178

Fig. 74. Lost Creek Falls. Sept. 2, 1921.

Ecology of Trout Streams

179

from a habitat which it might otherwise inhabit. This was instanced by
Odonata. many of which deposit their eggs on or in emergent vegetation.
Where such plants are absent, these Odonata will not be found, although they
are water breathers.

QUALITATIVE LIST OF PLANTS AND ANIMALS

The following lists cover only the more dominant or “type forms” of moun-
tain waters. Xo attempt was made to secure a complete collection of the species :
only enough to ascertain the dominant types, to learn something of their rela-
tions to the environment, and of their interrelations with one another. Further,
only the native biota is considered. Attention has already been directed to the
ecological distinction between primary and secondary inhabitants, or between
native and invading species. Such plants, for instance, as Elodea, Nymphaea,
Castalia, Potamogeton, may occur at some rare point in a mountain stream, where
for some reason or other the current is greatly diminished and permits a slight
amount of plant growth. But such single occurrences hardly warrant their
being listed as native species.

A further restriction must be noted because of the taxonomic difficulty.
This arises from the fact that in so many species one deals with immature stages,
many of which are as yet unknown to taxonomists. This fact was realized long
before these studies were begun. In the Park a distinct and prolonged effort
was therefore made to secure complete “life histories” of many insects. In
fact, I tried to breed many stages. Cans and jars, partly filled with stones and
covered with cheese-cloth, mosquito bar or wire screening, were placed at advan-
tageous points in Lost Creek, where human intruders were rare. Alas ! While
my security from humans proved complete, I had not anticipated the insatiable
curiosity and impish destructiveness of the bears. Day after day, despite my
best efforts in hiding my experimental jars, the bears searched them out
and despoiled or destroyed the jars and their contents. This lasted for more
than four weeks, after which I surrendered to the superior patience of the bears.

This is not offered as an excuse, but to signify a real condition that con-
fronts the investigator at times in Yellowstone Park. Visitors in the public
camps frequently use the streams to refrigerate some of their food supplies
during the night. This the bears seem to know and very often seek out the
“caches” for a toothsome morsel. Some of the bears are known to plunder the
food boxes which campers carry in their cars or on the running-boards. Being
possessed of curiosity, the bears would quite naturally seek to know what was
hidden in the jars I had planted in the stream bed. The results were dis-
astrous so far as my breeding experiments were concerned.

In view of this, I tried so far as possible to secure representatives of various
stages of insects, to be associated later on if possible. Such a method is unsatis-
factory, of course. But under the circumstances it offered the nearest approach
toward securing complete life histories.

The Plants. The list of plant species given below is copied from
Muttkowski and Smith’s, The Food of Trout Stream Insects in Yellowstone

i8o

Roosevelt Wild Life Annals

Park (Roosevelt Wild Life Annals, Vol. 2, No. 2). This list was based on
certain surveys made for the purpose of checking the food available in given
localities with that actually eaten by insects found in the same places.

A marked characteristic of mountain streams is the total absence of higher
plants. Lower plants, too, when present at all, are none too conspicuous and
generally are confined to one or two forms of algae. When these occur, they
are present for only short periods. Thus, Cladophora was found in the Lamar
River in considerable quantities for about four weeks; the entire bed of the
stream was decorated with bright plumes, so that from a distance the water
appeared a bright green. But in about four weeks most of this luxuriant growth
was gone and the remainder was a sparser and shorter growth of Cladophora.

The same held true for the Gardiner River. In Yellowstone River
Cladophora was very scarce, an occasional plume occurring in some sheltered or
less exposed spot. In Tower and Lost creeks, also, for some unexplained
reason (perhaps due to lack of carbonates), Cladophora was scarce. On the
other hand, in Lost Creek, Tower Creek, Lava, Blacktail Deer, Hellroaring,
and other creeks, the flat Prasiola was fairly abundant and associated with a
type of Nostoc. This same scarcity of Cladophora was noted in 1925 in the
streams and lakes of Glacier Park.

Besides Cladophora. smaller filamentous algae and various diatoms thrive
among the small accumulations between rocks ; this includes such forms as
Melosira, Ulothrix, Closterium, Cocconeis, etc. In the marginal pools Spirogyra
may thrive temporarily.

Following is the alphabetical list :

Cladophora sp.
Closterium sp.
Cocconeis sp.
Epithemia sp.
Gomphonema sp.
Melosira sp.

Moss — undetermined.

Nostoc sp.
Oscillatoria sp.
Prasiola sp.
Rhoicosphenia sp.
Spirogyra sp.
Synedra sp.

An occasional battered fragment of Potamogeton, Chara, or other lake plant
might also he found in Yellowstone River, washed down from the lake.

The Animals. The specialized conditions of mountain streams restrict the
number of animal species very markedly. In Yellowstone Park only three of
the Phyla can be called endemic, namely the Platyhelminthes, Arthropoda. and
Chordata. For Idaho, Washington, and Montana, certain additional phyla can
be noted — at least for some of the streams. Besides these representatives of
other phyla are of sporadic occurrence.

a. Invertebrate Phyla (exclusive of Arthropoda) . 1. Protozoa. — Only the
Yorticellidae and the parasitic Gregarina can be considered as native in the quick-
flowing mountain streams. One finds them more frequently, however, in the
marginal and sheltered pools left by receding flood waters. Where the current
is strong: thev do not occur, as a rule. Occasionallv, where small masses of
filamentous algae, e. g., Cladophora. Melosira, Oscillatoria, etc., have become

i8i

Fig. 75. Lost Creek, low water stage. In June all the stones
were covered by water. Sept. 2, 1921.

Fig. 76. Lost Creek, low water stage. Showing the last pool in
the bed of the stream. The stream is dry for nearly a
hundred yards above this point. Sept. 2, 1921.

Fig. 77. Tower Creek, confluence with Carnelian Creek (right). Sept. 6, 1921.

Fig. 78. Tower Creek in Tower Creek Canyon, showing the thickets along banks and
stones in bed of stream. Aug. 14, 1921.

Ecology of Trout Streams

183

wedged between rocks, or have formed accumulations at stagnant points, various
Protozoa do occur, such as the ubiquitous V orticella, Paramecium, several species
of Ameba, Oxytricha, Colpoda, etc.

A second condition under which one may find the Protozoa in rapid waters
is within the shelters formed by other animals, such as the cases of Trichoptera
and the tubes of Chironomidae. Here they may lodge, or attach themselves to
the occupant, a habit that applies particularly to the Yorticellidae. Empty
cases and tubes often contain a surprisingly varied microscopic life.

Among the deposits on stones, particularly those of the “holes,” the \ orti-
cellidae frequently form colonies of considerable size, numbering many
thousands of individuals.

Perhaps most surprising is the wide occurrence of Grcgarina in the various
aquatic insects, especially in Trichoptera. As noted in a second paper (see
Muttkowski-Smith), the specialized conditions of the habitat force a general-
ization in diet. It seems that insects and other animals are readily crushed,
and such fragments are avidly eaten by nearly all the different insects inhabit-
ing the mountain streams. Frequent catches made with plankton nets demon-
strate this fragmentation clearly. It is by eating these fragments that insects
spread the Gregarina among themselves ; for certainly the Grcgarina are entirely
passive in this matter.

2. Porifera, Coelenterata. Neither sponges nor hydrozoans are found in
the rapid w'aters, although they are of frequent occurrence in adjacent ponds
and small lakes. In European mountain streams several species of Hydra have
been noted in very placid stretches of their waters. These species can hardly
be considered as native.

3. Rotifera. Rotifers are of sporadic occurrence, being usually found
together with Protozoa in mats or wads of filamentous algae. Their occurrence
seems chiefly accidental. Insects feeding on the algal mats will readily devour
the rotifers and other small animal life. In the larger streams of the North-
west, such as the Snake. Clearwater, or St. Joe, one may find Rotifera in quiet
stretches, especially where the streams pass through great meadows.

4. Platyhelminthes. Two species of flatworms are very typical of the
western mountain streams, namely, Polycclis nigra Ehr., and Dendrocoelium
lacteum Mueller. The black Polycclis is very common, the white Dendrocoelium
much less frequent. The larger the stream, the less the abundance of these
flatworms. The optimal habitat seems to be that of moderate sized streams.
In Tower Creek most of the stones had numerous specimens on their under
sides, where the worms moved rapidly on their slime trails, feeding on the
various rock diatoms growing there and on the shore diatoms swept along by
the current and caught in the slime films. They reach a length of about 18 mm.

The scarcity of flatworms in Yellowstone River may possibly be due to
the great amount of sulphur carried by the stream. Still, Tower Creek and
Lost Creek both carry large quantities of sulphur, at least during the flood
period. On the other hand, the small number might be explained on the basis
of too strong a current. But here again it must be noted that comparatively
speaking the current in Tower Creek is faster than that of the Gardiner and

184

Roosevelt Wild Life Annals

Lamar rivers; yet Tower Creek is an ideal stream for flatworms and they occur
there in multitudes. In fact, in Lava Creek, which is still more rapid, the
flatworms are even more abundant.

5. Nemathelminthes. Except for a parasitic Mermis sp. found in
Simulium larvae and mayfly nymphs (Fig. 103), and an occasional Gordius
larva, the round worms and hair worms are absent. Thienemann finds that
Mermis and Gordius in their last or adult stages like to burrow in the wet soil
at the margins of spring-fed pools and also among the deposits on rocks.

6. Mollusca. To my surprise I did not find any snails or clams repre-
sented in any of the mountain waters. This is contrary to the findings of
Steinmann for Switzerland and of Thienemann for Germany. Xot even the
univalve Ancylus was noted. Only in quiet stretches where the streams flow
through the higher meadows, do occasional Planorbis and other snails occur.

7. Annulata. Annulate worms and leeches seem to be entirely absent from
the streams examined in Yellowstone Park. Only one specimen of a Pristina
was found on August 31 in Yellowstone River. Very probably this came from
an egg that had floated down from Yellowstone Lake. In this respect the larger
streams of Idaho show some difference. Thus in the Clearwater and Snake
rivers an annulate, probably a Limnodrilus , is quite abundant in the clear rapids.
Leeches, too, occur in the Clearwater, under the shore rocks. These shores
recall the lake shores markedly, since similar species occur in both, e. g., leeches,
flat mayfly nymphs ( Hcptagenia sp., etc.), Pscphcnus (Fig. 1 1 5 ) , flat
Trichoptera cases, hydroptilids, etc.

b. Arthropoda. The Arthropoda contain the dominant forms of the moun-
tain streams which could be studied more carefully in their interrelations.

1. Crustacea. Crustacea are conspicuous by their absence from the moun-
tain streams of the Parks, Yellowstone and Glacier. Yet in the large trout
streams of Idaho and Washington, such as the Clearwater, St. Joe, Spokane,
the St. Mary’s in Montana, etc., one finds Cambarus quite frequently, and
occasionally Hyalclla and some forms of Gammarus. Smaller plankton
Crustacea, too, may occur. This distribution in Idaho may be due to the
greater amount of mud present in the streams, and the many long placid
stretches where the streams move through the “upland meadows.” Perhaps
the lower altitude (3000 ft. and less in the places examined), may have some-
thing to do with the Crustacean representation. Thienemann calls attention to
the fact that the chemical composition of the waters, particularly the carbonates,
affects the faunal composition in very marked manner. It is possible that the
paucity of Crustacea and Mollusca in the Park streams is to be associated with
the presence or absence of the carbonates. Unfortunately. I had neither the
time nor the requisite equipment to determine this.

From the parasitic viewpoint also the Crustacea do not seem to be estab-
lished in the Park. Of the several hundred trout examined only one was
found to have a copepod gill parasite.

2. Arachnida. Shore and rock spiders were observed quite commonly in
their hunts along the stream. Their main prey seemed to be the adult caddis-
flies. On overhanging rocks, on the bridges, and on trees along the shores.

Ecology of Trout Streams

185

webs of other species were found which contained caddisflies, mayflies, and
smaller stoneflies (Figs. 97 and 98). Such webs are often nothing more than
irregular tangles of silk, anchored here and there, hut not stretched out in
any regular fashion. The spiders seem to he able to run on the surface film
of the water for short distances. Thus, on September 1 a number of spiders
were observed on Tower Creek, running close to the water, or even skating out
on the surface film after caddisflies.

These spiders no doubt account for a considerable number of adult aquatic
insects. The caddisflies formed by far the major portion of the web contents.
In turn the spiders occasionally fall prey to the current and are eaten by the
fish.

Hydrachnida, like the Protozoa, Rotifera, Mollusca, and Crustacea, are not
native in the mountain streams of Yellowstone Park. In the smaller streams
they are sporadic, usually where a slight tendency to swampy conditions mani-
fests itself. In Lost Creek I found Hydrachnida in beaver ponds. In the
Lamar they occurred in empty Trichoptera cases and among the mosses in
very shallow water. Both Steinmann and Thienemann record a number of
species for the streams of Europe.

3. Insecta. Perloidea. — The stoneflies form the dominant invertebrate
animals of the mountain streams. While they also occur in other aquatic bodies,
such as lakes, ponds, and slow streams, their optimal habitat is rapid waters,
particularly the cold mountain streams, with rapids, cascades and falls and
therefore with a high degree of oxygenation — precisely the type of waters pre-
ferred by trout. One may say that trout and stoneflies are intimately
associated ; where one occurs, the other also is found.

In Yellowstone Park about a dozen species were captured ; and of these
about five were numerous and conspicuous, the remainder occurred more spar-
ingly. In habitus the larvae are very dissimilar, a fact which correlates with
their habits and habitats. Thus the largest species. Pterqnarcys calif ornica (Fig.
99), is conical and long, the sulphurous Acroneufia pacifica short and squat,
with Acroneuria tlieodora intermediate in shape, somewhat short, but less
markedly flattened. This divergence in shape is associated with divergence in
habits. A. pacifica is primarily a clinging species ; it hunts the undersides of
rocks in strong rapids. Ptcronarcys calif ornica is a roving type, and hunts
between rocks; it migrates freely from pools to rapids, and from midstream
to the shores in search of food. The various species of Allopcrla are elongate,
only slightly flattened, and are found at the bottom of pools or “holes” where
the surface current sweeps above them and leaves the bottoms relatively
undisturbed. They, too, are rovers.

Following are the species determined; Ptcronarcys calif ornica Newport.
Pteronarcclla badia Hagen, Pcrlodcs Hagen, Acroneuria pacifica Banks,
Acroneuria ( Donoroncuria ) tlieodora Needham and Claassen, Pcrla verticalis
Banks, Isoperla j-punctata Banks?, Allopcrla coloradcnsis Banks, A. lineosa
Banks, A. fidelis Banks, A. sp.

Of these Pteronarcys californica is most conspicuous, by virtue of its size
and bright salmon color, although it is really less numerous than its sulphur-

Roosevelt Wild Life Annals

1 86

colored relative, Acroneuria pacifica. This, the largest of stonefly species, has
received a number of popular or local names, according to color, region, shore
vegetation, and the kind of fish found; thus one meets with names such as
salmon flies, trout flies, mountain flies, red flies, river flies, willow flies, grass
flies, etc.

The larvae attain a size of three to three and one-half inches fFig. 9 9).
In color they are a dull blackish gray, with yellow or orange dorsal and lateral
stripes. Their form is conical, widest at the thorax and narrowest at the caudal
end. They are good swimmers, but move mainly by creeping among the rocks
in the weaker rapids, eddies, and bottoms of pools. Strange to say they eat
plant food almost entirely, dififering in this respect from other perloid nymphs.
Thus, the stomach contents of both young and full grown nymphs were diatoms,
bits of algae, wood, and bark, while those of other stonefly nymphs were chiefly
animal matter. The roving habits of this species make it a favorable food item
for trout, more so than any other perlid species.

Transformation appears to coincide with the recession of the spring floods,
when the waters are still turbid and a Secchi’s disk is visible for scarcely ten
inches. In 1921 transformation began on July 5 at the Cooke City Road bridge
(Fig. 70) over the Yellowstone, and lasted for about a week, although some
belated stragglers continued to emerge till July 26. But the peak of transforma-
tion was July 5 to 8. Curious to say, the transformation at this point did not
coincide with the upstream and downstream emergence. Upstream, in the stretch
of river between the lake and Upper Falls and in Yellowstone Lake final eedysis
was pronounced from June 25 to 30, according to reliable information received
from Mr. Ainesworth, of the Lake fish hatchery. Downstream, at the eddy at
the mouth of Elk Creek, this species transformed in large numbers on July 15 and
16. Opposite Hellroaring Creek transformation was very pronounced on July
19 for both this species and Acroneuria pacifica. Visitors at Camp Roosevelt
from points north of the Park reported that large numbers were emerging on
Inly 21, along the Gardiner-Livingston highway which parallels Yellowstone
River.

Interpreting these dates in a linear arrangement, it would seem that trans-
formation proceeds in a wave downstream, being earliest at Yellowstone Lake
and progressively later at points farther down. In how far these dates accord
with those of other years is problematical. According to Mr. A. G. Whitney,
in 1922 recession of water occurred earlier and the stoneflies appeared about
two weeks earlier than in the preceding year.

Transformation takes place early in the day, within the first four hours
after sunrise, the greatest number emerging at about eight o'clock. Stragglers
continue to emerge practically all morning — indeed, occasional specimens were
observed transforming in the afternoon. For eedysis the nymphs climb up
any convenient point along the shore — into grass, onto stones, brush, trees, etc.
(Figs. 100 and 102). On the whole, they prefer open places along the stream.
Also, the slack waters of eddies are sought. In rapids, the various emergent
rocks, and in smooth stretches some mid-stream boulder, form convenient places
for emergence. Eedysis then takes place on the shaded side of any object to
which the nymph has anchored itself by means of its legs.

18-

Fig. 80. Cleft boulder on meadow between Yellowstone and Lamar rivers.

June 17, 1921.

1 88

Fig. 81. Brink of Lost Creek Canyon to right, and “crater” in foreground. The
meadows are said to be the bottom of the crater, the hills around forming the
rim of an extinct volcano. Aug. 28. 1921.

Fig. 82. Castle Rock (also called Cathedral Rock) on North Fork of Clearwater
River, Idaho. Showing types of shores and placid stretches of stream. Sept.
2, 1924.

Ecology of Trout Streams

189

The slow pumping action and inflation to split the nymphal skin can be
readily observed. Eventually the skin splits on the dorsal side on a line between
the wing pads and the occiput, the thorax pushes up and through the slit, the
feet are partly withdrawn, then the head, the entire thorax and legs, and finally
the abdomen. The whole process may take from twenty minutes to an hour.

Spreading of the wings then proceeds, during which process the adults
generally cling to their exuviae. The adults take flight in a very few minutes,
while the wings are still very soft and pliable. They fly to some higher object
near the stream, generally to some tall bush or small tree ; and if a person be
in the vicinity, he too may serve as an alighting place and shortly may be
covered with these fluttering insects. Observation indicates that the points
sought are such as give suitable exposure to the light and to the breezes for
the purpose of drying their wings ; direct sunlight is never sought. The adults
spread and move their wings freely during this process. In addition, they exude
yellow or salmon-colored droplets from the anus, which may be blood or merely
waste products, it is unknown which.

Copulation in this species takes place late in the day, oviposition appar-
ently at any time in the day. For oviposition the female descends on rocks to
the water surface and dips her abdomen into the water and permits the eggs
to be washed off in the lesser currents along the shore.

A curious phenomenon noted for this species and for Acroneuria pacifica is
the crepuscular flight upstream. This begins immediately after sundown and
lasts till after dark. Swarms of thousands of specimens take flight, hovering
from 50 to 200 feet above the water, depending somewhat on the height of the
shores. Since both sexes participate in this flight, it might be regarded as a
“nuptial flight.” But, as already noted, copulation does not take place during
flight, but when the specimens alight in tall grass, low trees, and bushes, or
lacking these, in crevices of rocks, or grooves in the sand. It is therefore not
at all clear how this flight is to be interpreted. Practically, however, the. instinct
serves effectively to repopulate the headwaters of the various streams. During
the upstream flights numberless thousands are caught in the vortices, back

drafts, and wayward air currents, and are hurled to the stream beneath, where

they eventually form an important item of the surface food of trout.

Adult life normally lasts but three to four days. Relatively few specimens
die a natural death. Being awkward fliers, during their short flights they are
readily swept into the water by wayward air currents. Others form juicy tid-
bits for hundreds of birds. Still others may be eaten by small mammals. In

the morning one finds hundreds of severed wings along the shores of the

stream, indicating perhaps the active feeding of the birds during the early
hours. One morning, at a point near the Cooke City bridge, in a gravelly
depression about three feet in diameter, I counted the wings of over 200 P. cali-
fornica and A. pacifica. Most of these I removed. Next morning an approxi-
mately equal number of detached wings lay in the same depression. I was
unable to tell whether mammal or bird or other creature had feasted there.

Ptcronarcys calif ornica is primarily a river species, abundant only in the
larger mountain streams, and rather scarce in the precipitous creeks. It is
found throughout the Northwest.

Roosevelt Wild Life Annals

190

Acroneuria pacifica, which is very closely associated with P. calif ornica, was
even more abundant. In the main, what has been said about the larger species
applies also to A. pacifica, with the exception that the nymphs are quite squat
and smaller, and that they are much more active. On shore the adults also
seem to be more active than the larger species (Figs. 101 and 102).

For July 21 I have the following note: “Yellowstone River shores strewn
with wings of P. calif ornica and A. pacifica. Many caught in spider webs
with Tipulids and caddisflies. Many others water-trapped. Eaten by birds (night-
hawk, robin, etc.). Also snakes (garter snake) here that feed on the adults.
Ground squirrels seen twice feeding on Pteronarcys adults.”

Garter snakes were seen on various occasions feeding on stoneflies. Julv
15 six garter snakes were observed feeding on these insects in the stone rubble
that marks the Elk Creek eddy of the Yellowstone. On July 16 some twenty
garter snakes were noted along the Lamar, most of them feeding on the stone-
flies. A frog taken along the I^amar and examined for the stomach contents
showed several adult Acroneuria, a beetle, and some diptera.

Of other species, Acroneuria thcodora deserves mention. Its emergence
occurred early in August along the Gardiner and Lamar rivers. Very few
adults were seen and none of the nymphs, except for some exuviae. This is not
a common species.

The smaller species of stoneflies, with adults of pale greenish or yellowish
hue, are somewhat inconspicuous because of their small size. One finds the
nymphs most frequently among the rocks on the bottom of midstream pools,
and very sparingly among the rocks of rapids. Their food is chiefly animal
matter. Their emergence seems more prolonged, and not confined to a rela-
tively short period like that of the larger species. On the whole, the smaller
species are much more abundant in smaller creeks and along the shores of
mountain lakes than in the larger streams.

Insecta. Epherneroidca. — Of the mayflies the following have been identi-
fied: Am el etas sp., Callibaetis sp., Drunella grandis, Ephemerella eoloradensis,
Ephemerella sp., Hcptagenia sp., Iron longimanus. This list is rather frag-
mentary, since most of the species were available only in nymphal form. An
attempt to breed the nymphs has already been mentioned ; to the persistence of
the hears in wrecking my arrangements may be ascribed the lack of more specific
identifications.

Two types of habitus are present among the nymphs, the swimming and the
clinging. Thus, one encounters the cylindrical and slender nymphs as typified
by Bactis and Callibaetis, and the flat appressed type of Hcptagenia, Iron and
other forms.

The food, as indicated by examination of the stomach contents, is largely
offal. The food average for 25 Ameletus nymphs was: plant food 2 2.~ c c,
detritus 77.3% ; for twenty Drimclla, 18.6% animal food, 36.8% plant food, and
44.7% detritus ; for two Bactis, 50% plant, and 50% animal food ; for 28
Ephemerella, 3.6% animal, 36.3% plant food, and 60% detritus; and for 34
Hcptagenia, 24% plant food and 76% detritus.

Ecology of Trout Streams

HJl

Some of the Drunella nymphs were heavily infested with Gregarina. On the
other hand, two Heptagcnia nymphs were found to be parasitized by a nema-
tode parasite ( Mcrmis sp.). Figure 103 shows an adult mayfly with such a
parasite emerging. The specimen was taken at the entry of Lava Creek into
Gardiner River ; it was fluttering weakly and the weight of the parasite held
it from rising above the shallow rapids at that point.

Transformation for all species was very irregular and seemingly strung
out during the whole summer. One could expect to find a few adults trans-
forming every day. No attempt was made to secure a series of adults. This
was mainly for a personal or selfish reason. Usually laden with bottles, water
nets, cameras, luncheon, etc., an extra net and cyanide bottles might be more
of a load than would be comfortable for a tramp of many miles.

Insecta. Odonata.- — Relatively few dragonfly species are known from

rapid streams, and even fewer from mountain streams. Only a single species, an
Ophiogomphus (probably montanus Selys), was found, of which three adult
females and some eight or ten exuviae were collected. Strange to say, all of
these were found in one locality, some fifty yards below the Cooke City bridge
across the Yellowstone. On July 30 the emergence of a specimen was observed
in detail. The following is taken from my field notes of that date:

“10:37 a - m. — Gomphid climbed out of water. Immediate shore here is
somewhat sandy. Climbed onto rock.

10:45 — Thorax split open.

10:46.5 — Head and thorax thrust forth.

10:50 — Head, thorax, and legs out.

10:58 — Abdomen out to 4th segment, wings elongating.

1 1 : 02 — Abdomen out completely.

11:05 — Wings unfolded to one-half their adult length. Abdomen shows no
change.

11:12 — Wings open to stigma.

11 :I4 — Wings open to tip, now y?, longer than abdomen.

11 :i5 — Abdomen beginning to be inflated, insect clawing for hold on rocks
and exuviae.

11:17 — Insect leaves skin for hold on rock.

11:18 — Pumps abdomen with visible effort, swaying from one side to other.

11:20 — Abdomen extends to beginning of stigma on wings. The wings
are folded roof fashion to side of abdomen, which is pressed firmly against them,
to aid in process of inflation.

11:21 — Abdomen extends to end of stigma.

11:21.5 — Droplet forms at end of abdomen and drops into water.

11 :22 — A second droplet.

1 1 :22_3 — Third droplet.

11 :22.4 — Fourth droplet.

11:22.5 — -Fifth droplet.

11:22.8 — Sixth and seventh droplets.

11:23 — Eighth droplet.

192

Roosevelt Wild Life Annals

11:23.5 — Ninth droplet. Abdomen reaches to tip of hind wings.

11:24 — Three droplets in rapid succession.

11 :25.2 — Two more droplets. No more fell after this.

11 :28 — Abdomen extends slightly beyond tips of hind wings.

11:30 — Abdomen V) in. beyond tip of wing.

11:45 — Abdomen deflated.

11:50 — Leaves for top of rock and extends wings for flight. Captured
on ‘take off.’ ”

Two other females were captured the same day. Specimens of this species
were also seen on the following days along the Gardiner and Lamar rivers,
but were not captured.

In the neighborhood of mountain streams one frequently sees specimens
of Sympetrum, Aeschna, Lestes, Enallagma and other pond forms. These
come from nearby quiet waters, or. as in the case of Lost Creek, may even
occur in beaver ponds; that is, in such ponds as result from damming of the
flow of the creek by the beavers. The fauna of these ponds is of a mosaic
type, as a rule, consisting of a very few representatives of pond life.

Insecta. Hemiptera. — The aquatic bugs are hardly endemic to moun-
tain streams. Like many other aquatic insects, they may occasionally be found
in marginal pools, ponds, and in swamps and springs. Thus. Notonccta may
be found on rare occasions and with it Corixa, and more frequently some water
strider or surface skimmer. In Yellowstone Park I encountered water striders
on the Lamar, Tower Creek, Lost Creek, and Gardiner River; invariably they
marked the neighborhood of some spring-fed pool or clear pond. In Idaho I
have seen the skimmers along the Snake. St. Joe. Clearwater. Elk. and Oro
Grande rivers, in Glacier Park along MacDonald Creek. Two Medicine River,
St. Mary’s River, etc. There, too. they seemed to he migrating from some
neighboring pools.

Ixsecta. Trichoptera. — Only generic identification was possible for most
of the species taken, due to lack of adult stages. For here. too. the bears
interfered. For breeding cages I had used some of the familiar cone-like fly
traps of wire, which were fitted over stones harboring various caddisworms.
These the bears were not satisfied to tip over, but removed them completely ;
I did not discover what they did with the traps. Following are the species
determined : Brachyccntnis sp., Glossosoma sp., Gocra sp., Hydropsychc scalaris
Hagen, H yd ro psyche sp.. Limnophilus sp. (near rhombicus L.), Molanna sp..
Ncopliylax concinnas Hagen. Philo potamus sp.. Platyphylax sp.. Rhyacopliila
eoloradensis Banks, R. fnscula Hagen, R. tori'a Hagen, R. sp., Thrcnuna sp.,
Triacnodcs sp.

Of these species the four most conspicuous are Hydropsyche, R. torva,
Brachycentrus sp. and Thrcmnia sp. Of these the first is a web-spinner, the
following two have fixed cases, and the last a movable case (except during the
pupal stage, when the case is fastened). In the creeks Thrcnuna and Hydro-
psyche seemed abundant, in Yellowstone River the other two species.
Hydropsyche, strictly speaking, is of wider distribution than other species of
the mountain waters.

193

Fig. 83. Clearwater River, Idaho. Showing sand deposits left by spring flood,
and log jam in river. Aug. 31, 1924.

Fig. 84. Clearwater River, Idaho. A closer view of log jam caused by spring
flood. Also showing “The Gates” and government bridge across river. Aug.
3 b 1924-

194

Fig. 85. Hibernating Coccinellidae. Photograph taken on a warm spring day when
sun brought out the beetles. Many had been w'ashed into a creek about twenty
feet away. Photograph Moscow' Mt., Idaho, about April 15, 1923.

Fig. 86. Skeleton of winter-killed buffalo, near Lamar River eddy. Note the
puparia of flies on the scapula. The skull had been dragged into the eddy by
some animal (bear?). July 25, 1921.

Ecology of Trout Streams

195

The fixity of the cases results in an enormous death rate for Brachycentrus
and Rhyacophila. During the high water period their cases are attached to
rocks that are then submerged; with the quick recession of waters these cases
are “stranded” and their inhabitants die within them (Fig. 104). On several
occasions a number of these stranded cases were opened to see if the larvae
had left them in order to build new habitations. The stranded cases contained
dead larvae and pupae. Very quickly these form the food for scavenging shore
forms, such as Staphylinidae, a Silpha, Carabidae, and ants of various kinds.
This “death by stranding” was most notable along the Lamar River, where
the recession of water was very rapid.

For Brachycentrus two methods of attaching the square cases were noted.
One was to attach them by the base, i. e., the small end, so that the cases stuck out
from the rocks at right angles (Figs. 104 and 105). The second was an
attachment by means of a “lip” at the wide or head end of the case ; usually
this occurred close to the surface so that the surface layer of water , would
sweep directly into the case. When a stone carrying such cases was lifted
above the water surface or to a point where the water did not completely cover
the opening, the larvae would pull the case to an oblique or right angle to the
rock, bite off the secretion attaching the case and then crawl down the rock
to a point where the current was suitable. The cases were submerged from
one-fourth of an inch to six inches in fairly strong rapids.

Empty cases of Brachycentrus seem to afford opportunity for the develop-
ment of a micro-fauna of their own, particularly of Protozoa, Rotifera, and
an occasional Flydrachnid.

The food of caddisworms is quite generalized. Thus, 32 Rhyacophila con-
tained 49% animal food, 23% plant food, and 28% detritus; 27 Hydropsyche
had eaten 42% animal food, 54.3% plant food, and 3 .7% detritus; 33 Brachy-
centrus consumed 18.3% animal food, 72.7% plant food, and 9% detritus;
and 23 Thremma included 28% animal food, 64% plant food, and 36% detritus
in their diet. On the whole, the first three take their food as it comes to them,
while Thremma selects what food it consumes. All of the forms are cannibals
on occasion (see Muttkowski-Smith) .

Connected with the feeding one may note the extent to which the various
species are parasitized — Rhyacophila with Gregarina in about 40% of the
specimens examined, Hydropsyche about 60%, and Brachycentrus fully 100%.
None of the specimens of Thremma contained any parasites. This is possibly
correlated with the fact that their food was primarily selected, while the other
forms ate what food was swept along with the flotsam down the stream. Much of
this flotsam consists of crushed and torn specimens of insects.

The adults have the habit of flying above rapids, running up the wet rocks,
and also dipping into the water. This habit exposes them to capture by fish ;
indeed, they were numerous among the stomach contents of trout. One trout
was found to have gorged itself with caddisflies, the total number exceeding
500 specimens.

Insecta. Diptcra. — Only three families can be said to be typical in moun-
tain streams. These are the Blepharoceridae and Deuterophlebiidae, or net-

196

Roosevelt Wild Life Annals

winged midges, and the Simuliidae or Buffalo gnats. Besides these, isolated
genera of Chironomidae, Leptidae, Tipulidae, and Psychodidae breed in rapid
streams. In addition, the adults of some species must be listed as hovering near
swift water, chiefly on rocks that are sprayed by the stream.

The most typical residents of swift streams are the Blepharoceridae. This
family was represented by two species, Bibiocephala comstocki and B. cjrandis.
Where the current is swiftest, where their abiding places are perhaps the most
inaccessible, there one is likely to find the larvae of these species. Their
presence was first noted on the Lamar River, on June 23, when several dead
larvae were found stranded in a flood pool. Somewhat later, June 28. stranded
pupae were noted. The first adults were seen July 5, on the rocks in Yellow-
stone River, “on the shady side of the large boulders in the strongest current,
gathered in groups of forty or more. Very difficult to secure with a net (dip-
net). Some drop into the water, and appear to have little difficulty in reaching
a rock again. Apparently somewhat hydrofuge (notes of July 5).” The
adults were found from July 5 through the following months. The last were
observed September 6 (into October in Idaho streams). Pupae were found
quite frequently during the low water stages, but the larvae only very rarely
(Figs. 106, 107 and 108).

Because of their very firm means and methods of attachment one would
hardly expect to find the larvae and pupae in fish stomachs. Yet they do occur
in trout food. The adults occur much more frequently among food items of
fish than do the earlier stages.

One of the most noteworthy discoveries of the summer was that of a
Deuterophlebia larva (Fig. 109) in Yellowstone River on July 30. This enig-
matic larva for several years defied any attempt at identification. Very recently
its identity was established, through the courteous aid of Dr. Johannsen of
Cornell University. A detailed account of this new American record has been
published elsewhere (See Bull. Brooklyn Ent. Soc., Yol. 22, pp. 245-249, 1927).

Deuterophlebia was described from a pair of male specimens from India,
which formed the type of a new species, genus and family of Diptera. Several
years later the larval stages were described by Miss Pulikovsky from materials
received from the Altai Mountains in Siberia. The Yellowstone River larva
establishes this genus and family for North America. For the Park there are
three records :

[uly 10. Lost Creek at Camp Roosevelt. Pupae taken from sluice-dam.
Elevation about 6300 feet.

July 30. Yellowstone River, at Cooke City bridge. One larva caught in
plankton net. Elevation 5900 feet.

Aug. 2. Tower Creek Canyon, about four miles above Tower Fall. Pupae
taken from rocks holding Bibiocephala pupae. Elevation about 6700 feet.

While the pupae of both families are very similar, the larvae are very
unlike, as will be seen upon comparison of the figures 106 and 109. Bibio-
cephala is long and flat, with a series of ventral suckers; its locomotion is
extremely slow and clumsy. Deuterophlebia has a number of pro-legs with
terminal pads for attachment; although the locomotion was not observed, to my

Ecology of Trout Streams

19 r

mind the pro-legs operate much like the tube- feet of Echinodermata. The simi-
larity of the pupae is regarded by Miss Pulikovsky not as indicating relation-
ship, but as a clear case of convergence, similar to that shown by certain
Psychodidae. No adults were taken in Yellowstone Park. But I suspect from
the similarity of habitats and from what Miss Pulikovsky has written that the
adults probably show similar habits.

The larvae of Blepharoceridae (and Deuterophlebiidae) are found where the
conditions of mountain streams are most strenuous, namely, in the strongest
rapids and behind water-falls. The pupae are found under similar conditions.
The adults like to sit on rocks, near the surface of the water, where the spray
keeps the rock continually moistened.

The third family of Diptera that is confined to rapid waters is the family
Simuliidae. It is very probable that a number of species are represented in
the streams of Yellowstone Park. Of the fifty-five blackfly species of North
America listed by Dyar and Shannon (’27), some eighteen are regional; that
is, they occur in regions adjacent to Yellowstone Park. In fact, seven of these
regional species were actually taken in the Park, chieflv at Mammoth Hot
Springs and Old Faithful. It is very likely therefore that the specimens col-
lected and observed in the various streams represent more than one species.

In the Yellowstone the larvae and pupae were quite scarce after the spring-
flood, and none too frequent in the Lamar. At the same time the small tribu-
tary creeks and rills contained an abundance of Simulium From observations
made during the summer it seems that the larger streams are restocked every
year by specimens from the smaller tributaries. Thus, at the mouth of Elk
Creek, on July 15, the stones were covered with large patches of Simulium, while
the Yellowstone showed none of these larvae. About three weeks later the
river, too, showed an abundance of Simulium just below the mouth of Elk
Creek. Similarly in Lost Creek at the time of high water (about June 20!
Simulium was very scarce. On June 19 a small rill leading from the creek to
the meathouse of Camp Roosevelt was found heavily infested with Simulium
larvae and pupae. Lost Creek above and below the rill showed few traces of
the species. Turning the water from the rill for two days dried up the Simulium
and stopped their appearance at the camp during the summer. As the waters
of Lost Creek fell, Simulium became more and more abundant and conspicuous.

In Yellowstone River the larvae were found to contain a considerable portion
of lake plankton in their food. This was evidently swept downstream from
Yellowstone Lake, and the larvae combed this food from the waters (Fig.
no). In turn the larvae and pupae are eaten by fish. Small Pteronarcys
nymphs were observed feeding on them ; it is also probable that at least Hydro-
psyche among Trichoptera and the larger mayfly nymphs feed on Simulium.

A nematode parasite is quite frequent in the larvae. A field note of August
ir is pertinent: “Yellowstone River. Simulium was found extensively parasit-

ized. Certain individuals were noted with swollen caudal end and of greater
length than others. These specimens were invariably pale green on their venter.
A green Mermis ( ?) was noticed in the net. Comparison elicited the fact that
all swollen individuals were parasitized. It was not determined whether any

198

Roosevelt Wild Life Annals

pupae were parasitized or if parasitized individuals could pupate. It was evident
that the Mermis were leaving the larvae for their free-living adult stage. Just
at what point they left the Simulimn was not established. It was noted, how-
ever, that the parasite was coiled chiefly at the posterior end of the larva, with a
small coil near the head.”

Of families with occasional representatives in mountain streams, the Psy-
chodidae, Chironomidae, and Leptidae should be noted. A species of Psychoda
was caught quite freely along the various streams, but none of the larvae were
found. It is probable, according to Thienemann, that the larvae would be
found among the deposits on rocks.

Of Chironomidae the following species were found : Ccratopogon sp.,
Chironomus sp., Cricotopus varipes, Metriocncmius sp., Orthocladius sp.. Pro-
cladius sp., Tanytarsus exiguus. It is probable that many more species are
represented ; but since the main attention had to be given to the more dominant
and conspicuous forms, the chironomid collections were perforce somewhat
neglected. Cera'topogon larvae were found repeatedlv in the sandv spots in
Lost Creek, among the accumulations between rocks, and in the marginal areas.
No pupae were obtained. Cricotopus and Tanytarsus were conspicuous in the
sulphur and slime deposits on rocks along the shores and in midstream on sub-
merged rocks. There they gathered and built their tubes by the thousands
into a flocculent mat, among which rock diatoms and algae thrived. Here, too,
might be found the Yorticellidae (Hydra also, in European streams). Young
fish fed on these mats with eagerness ; but larger fish seemed to despise them
and came to them only when extremely hungry.

Adult midges could be seen flying along the shores among the grasses or
into the spray of the lesser and shallower rapids.

The Leptidae are represented by Athcrix variegata Walker. The larvae
have a rather wide distribution in mountain streams and are frequentlv found
in trout stomachs. For the adults. Malloch's account can be quoted 1 p. 363) :
“The females of the genus Atherix have a peculiar egg-laying habit. The
eggs are deposited upon branches or twigs of willows or other trees over-
hanging streams. After oviposition the female does not fly away, but dies and
remains attached to her egg-mass. The second female adds to the already
deposited mass both her eggs and her body, and gradually others do likewise, until
the combined mass of eggs and flies assumes considerable proportions, often con-
taining several thousand dead flies. The larvae which hatch drop from the mass
into the stream below, where they pass the immature stages. The Indians in
Oregon at one time collected the masses of eggs and flies and used them as
food. An interesting account of this aboriginal utilization of nature's resources
is given by Prof. J. M. Aldrich (Ent. News, Vol. 23, pp. 159-164).”

The cluster shown in figure in measured about seven by twelve inches and
was fastened to the underside of an overhanging rock. The position of the
rocks is indicated in figures 71 and 112; figure 65 shows the same rocks from
the distance. The photograph was taken under somewhat difficult conditions,
since at the time the base of the rock was surrounded by rapids. My notes
for that dav are as follows: “Flies in cluster arranged in shingle fashion, each

Fig. 87. Huge boulder in stream bed of Oro Grande River, Idaho. This boulder is
fully forty feet high. Smaller boulders can be seen on the slope along the river.
Sept. 2, 1924.

Fig. 88. Trick Falls, Glacier Park, in Two Medicine River, a short distance
below Middle Two Medicine Lake. Photograph July 17. 1925. The Same
place was visited Aug. 26 ; the lower part of the falls, which emerges from an
underground conduit, was then completely dry.

200

Fig. 89. Keppler Cascade, Yellowstone Park. Aug.
18, 1921.

Fig. 90. Oro Grande River, Idaho. Freak root's of a tree on top of a boulder. One
root circles to the left and then turns under the tree. The boulder is surrounded
by water, and none of the roots passes into the stream bed : the boulder showed
no sign of being cleft.

Ecology of Trout Streams

201

straddling neighbor in part. All so arranged that at periphery of cluster the
individuals face outward. Unevenness of rock has caused several whorls.
Composed of thousands of specimens. All dead. These are this year’s flies.
Cause of death unknown. Came here for oviposition, for under them a fibrous
deposit is full of eggs, some of which are hatching. The mass is invaded by
a number of Tro.v adults, and Dermcstcs, also one or two Staphylinids.” My
last notation is of August 23 : “Winds and scavengers have stripped the patch
so that only very few dead Atherix remain. But Dcrmestes larvae are still
feeding on them.”

The Tipulidae, too, seem to be represented, for along the shores in the moist
sand or earth, between rocks, or submerged, one finds tipulid larvae of vari-
ous types. No attempt was made to collect them systematically, since they
appeared to be rather infrequent. One peculiar larva collected is shown in
figure 1 13, taken from Tower Creek, August 29. This larva inflated its pos-
terior end in balloon fashion ; for what purpose, is not known.

Diptera are further represented by what I have called “splash flies” or
“spray” flies. These are the adults of Philolutra simplex and Chamaedipsis
coniata Melander. They occur in fair numbers on the exposed rocks of streams
(Figs. 93 and 114). They make short flights to the water surface, alight at
the immediate edge of the current on the rock, and then run upward through
the spray. They seem to prefer the shady side of the rocks. Quite generally they
were mixed with Bibiocephala. Like these, I found them during the entire
summer, and even later in Idaho streams. The earlier stages are unknown
to me. Whether they are native forms or merely visitors I was unable to
ascertain. They are much more numerous than the Blepharoceridae, but despite
this they are much less frequent in fish stomachs.

Among casual visitors near the water, Boletma melcncohca, Mycomyia
mendax, Linanculus qucrulus, Leptoccra sp., Paralimna sp., Empid sp., have
been identified. It is very probable that many of these breed under moist
conditions, hence not in the water, but at the margins of streams.

Insecta. Colcoptcra. — The bettles like so many other aquatic types, are
practically absent from the mountain streams of the Park. Certain species
( Hydrobius scabrosus Horn, Agabus tristis Aube, Crenitis monticola Horn,
Dcronectes griseostriatus DeGeer, and Hydroporus sp.) were found in the
marginal shallows and pools, especially in the recession pools. The test for
the true fauna of the mountain streams lies in the food contents of fish stomachs.
Endemic species will be found there in the larval and adult form. Surface
drift can be readily distinguished. Thus, the beetles noted above are practically
entirely absent from the stomach contents either as larvae or adults. In fact,
one finds them only as adults, if present at all. If one wishes to find the larvae
of these species one must look to the ponds and small lakes and marshes.

What surprised me was that the streams did not show anv Parnidae and
Dryopidae (Figs. 105 and 106), the typical beetles of lake shores. Idaho streams
such as the Clearwater, the Palouse River and others, in places show an abund-
ance of Psephenus larvae and occasionally other parnid species. But in Yellow-
stone Park thev seem to be quite rare. I have one record of a Psephenus
skin, August ir, Yellowstone River, at Cooke City bridge, “taken from Simu -

202

Roosevelt Wild Life Annals

hum society.” This may have been carried down from Yellowstone Lake.
Another record was “June 28 — Parnid larva taken from Lost Creek.” This
was placed in a screened jar with the hope of breeding the specimen. With
this jar were a number of others, containing various stonefly nymphs, mayfly
nymphs, and Trichoptera larvae. Unfortunately, as already noted, the bears
of that region had too keen a scientific curiosity and wrecked my breeding
arrangements by tipping over the jars, scattering them downstream or removing
them altogether.

Shore beetles of various sorts play the parts of scavengers on various
stranded biota ( Cicindela , Dermcstes, Staphylinidae ; see notation under Trich-
optera). Another situation was of considerable interest. This was in connec-
tion with the Atherix cluster described (see Diptera). Although the cluster
could hardly have been more than a week old at the time of its discovery, still
the larvae and adults of Dermcstes talpinus Mann, and the adults of Trox sp.
and several forms of Staphylinidae were feeding on the dead Atherix adults
and on the developing eggs.

c. Chordata. Practically all the classes of Chordata are represented in the
mountain waters and affect its biota either directly or indirectly.

1. Fish. — The characteristic fish of the mountain streams are the native
red-throat trout ( Salmo clarkii), which are found in the highest elevations.
Besides these one may also find minnows, occasional bass, suckers, pea-mouth,
bullheads, Cottus, and graylings. Cottus has been found at surprisingly high
elevations; Cottus punctulatus (Gill) is recorded from the Yellowstone, and
other species from various rivers of the Northwest. The grayling ( Thy -
mallus) is held by Thienemann as typical of the lower elevations (below 1000
meters) of rapid streams. In Idaho streams I have found Cottus, bullheads,
pea-mouths, and other fish at elevations of 3000 feet and less, in the larger
streams.

In Yellowstone National Park (and to a lesser extent in Glacier National
Park) the planting of exotics (non-native fish) in the lakes and streams, includ-
ing various foreign trout, whitefish, pike, pickerel, rock bass, silver bass, perch,
etc., has perhaps affected some of the faunal complex, at least of the lakes.
Records of plantings are not available except for recent years. In places it
is therefore difficult to tell whether the fish present are native or foreign. On
the Pacific slopes the upstream migration of the salmon annually affects the
headwaters of streams. No food is eaten by the salmon, but the spawning and
the ensuing death, followed by decay of the adults, probably affects the life and
physiology of the stream in many ways. This is a problem that invites study.

2. Amphibia. — No amphibia are native to the mountain streams. Frogs
and in one case a salamander or newt (not captured) were found at points
where ponds, and marshes occur. Generally the frogs are only visitors to the
streams, or may have wandered away from their proper habitat.

On rare occasions one may find frogs among the stomach contents of fish.
I have only one record of a frog taken as food by trout, in the meadows of
Slough Creek. As a matter of fact, several attempts were made to use small
frogs as bait for trout; the fish seemed uninterested and refused it. In turn the
frogs eat much of the insect life along the streams — stoneflies. caddisflies. beetles,

203

Fig. 91. MacDonald Creek, Glacier Park. Just above the falls the stream bed
consists of stratified rock. Aug. 17, 1925.

k ig. 92. \ ellowstone River, at Cooke City bridge. Showing rapids and marginal

pools. Aug. 15, 1921.

204

Fig. 93. Yellowstone River showing rocks with “splash flies.” July 7, 1921.

Fig. 94. Lamar River eddy, showing marginal pools. Later these pools
became dry. Pool at upper left is completely cut off from stream. In
the middle the moist bottom of a recession pool shows clearly. July 16,
1921.

Ecology of Trout Streams

205

etc. The stomach contents of a frog taken along the Lamar showed several
Acroncitria pacifica, some beetles and flies.

3. Reptilia. — Snakes are frequent visitors along the shores of mountain
streams. Garter snakes were noted repeatedly — six on one day, over twenty
on another — feeding on whatever food good fortune might cast their way.
Preferably they would feed in places where rocks were heaped, or where such
rocks were sufficiently close and with spaces between and under them to pro-
vide proper shelter and good hunting ground. Repeatedly I observed the
garter snakes catching and eating adult stoneflies (. Pteronarcys , Acroneuria).

In no way connected with the work was the capture of a “rubber boa” on
the evening of August 10, along the road between Tower Fall Junction and
Tower Fall. Dr. C. C. Adams and Dr. G. M. Smith were with me at the time.
The specimen was sluggish in behavior, very inactive and rather docile when
picked up. It was later identified as Charina bottac (Blainv.). This species
is widely distributed along the Pacific Coast. Since then I have seen it in
Idaho, near Moscow Mountain ; and from descriptions of students, farmers,
hunters, etc., I find that it occurs throughout Idaho and western Montana.

4. Aves.— Of birds only one can be called a species of the mountain streams,
the water ousel or dipper ( Cinclus mexicanus unicolor). This is the character-
istic bird of the rapids. There one may observe it on exposed rocks, “dipping”
in characteristic fashion. From time to time it wanders into the water, through
the rapids, disappearing completely beneath the surface in search of food. Fre-
quently it emerges some twenty-five or more feet away from the point of entry.

Along the shores nighthawks ( Chord dies virginianus lies peris) robins
( Plancsticus migratorius propinquus), and other birds were seen feeding
on stoneflies. Thus, morning after morning during early July one might find hun-
dreds of stonefly wings (of Pteronarcys and Acroneuria) gathered in some
favorable spot, apparently stripped from the insects before they were eaten.
Although nocturnal observation was difficult, I feel certain that quite a number
of birds followed the crepuscular flight of the stoneflies and fed on the swarms
(see stoneflies, p. 189).

5. Mammalia. — \ isiting mammals are not at all infrequent along the moun-
tain streams. Most noteworthy are the moose (Alecs amcricamus shirasi Nelson),
the elk ( Cervus canadensis canadensis Erxleben), mule deer ( Odocoilcus
hemionus hemionus Rafinesque), White-tail Deer ( Odocoilcus virginianus
macrourus Rafinesque), various mice, beavers ( Castor canadensis canadensis
Kuhl), otters ( Lutra canadensis canademsis Schreber), black or cinnamon bears
( Ursus americanus Pallas), and water shrews (Neosore. r palustris navigator
Baird).

Of these the beavers, water shrews, and otters can be said to affect the
stream life directly, the others more or less indirectly. The bears were con-
stantly noted along Lost Creek, and Lamar River, plunging in for a bath, ford-
ing the stream, or perhaps looking for some toothsome fish. Water shrews were
observed only twice, swimming about actively in search of aquatic insects. The
beavers affect the smaller streams primarily by damming. The domiciles and
engineering feats of beavers may be observed at thousands of points along the
streams of the Northwest. Lost Creek, Carnelian Creek, Lava Creek, etc. are

206

Roosevelt Wild Life Annals

excellent illustrations of dammed streams. Along Tower Creek there are a con-
siderable number of large springs which have furnished sites for a number
of beaver works.

The results of such damming of streams are both direct and indirect. As
direct results may be indicated the destruction of current for long stretches
and hence a modification of the streambed and of the constitutive fauna and
flora. An indirect result is the drowning of adjacent forests, which in time
leads to the alteration of the stream banks.

ECOLOGY OF MOUNTAIN TROUT STREAMS

A. General. In the first part of this paper four types of mountain streams
were distinguished : the constant streams, the flood or variable streams, the
precipitous streams, and the temporary hillside streams. This classification does
not agree with that of Steinmann (’07) for the alpine streams of Switzerland.
Steinmann recognized two types of mountain streams, (a) the elevated moun-
tain streams (Hochgebirgsbache), which have their origin from the melted snows
of elevated fields and mountain sides and from glaciers; (b) middle mountain
streams (Mittelgebirgsbache), which are fed by springs. The latter might also be
called valley streams since they traverse the mountain valleys and there gather
the waters of tributaries.

Thienemann (’12) notes certain weaknesses in this classification, notably
the fact that it is based on altitude, and that insufficient distinction is made
between the major habitats and their subdivisions. Based on the study of
streams of the lower mountains of Germany. Thienemann adds certain elements
and offers the following classification :

A. The springs and their outlets, which are fishless.

B. The elevated and middle streams, which contain trout.

C. The lower streams, which contain grayling.

Springs and rills solely would he found in the higher altitudes. These
are followed by streams which contain Salmonidae ; these in turn by streams
in which the Cyprinidae predominate.

Both Steinmann and Thienemann consider the spring biota as ecologically
related to those of the mountain streams. While granting a physiographical con-
tinuity between spring outlets and streams, an ecological continuity appears to
me to exist only under exceptional conditions. In no place does this become so
apparent as in Yellowstone Park. Although in the present study the springs
received little attention, a word about them may not be amiss. After a number
of reconnaissances of the aquatic conditions in Yellowstone Park I recognized
the futility of obtaining even to a slight extent a fair picture of the true relations
of the springs to streams. For the springs of the Park are not only multitudinous
in number, but also multitudinous in variety One could employ a dozen different
criteria for their classification and still not exhaust them. One might classify
springs according to topography, whether their waters immediately fall away after
leaving the ground or whether they are caught in a pool of varying size ( Born-
hauser). One could also classify them according to size and temperature; as for
the latter, the springs range from icy chilliness to boiling point. One might arrange

207

Fig. 95. Yellowstone River, rapids and whirlpool, a short distance above
Cooke City bridge. July 7, 1921.

Fig. 96. Lost Creek, showing tangles and underbrush along
stream. Sept. 2, 1921.

208

Fig. 97. Spider webs, with prey of aquatic insects. From rocks beside Lost Creek

Falls. Sept. 2, 1921.

Fig. 98. Detail of a spider web, showing various insect adults. Lost
Creek Falls. Sept. 2, 1921.

Ecology of Trout Streams

209

them according to color, which not only differs among different springs, but often
in the same spring at different times. Again, one could classify them accord-
ing to substances in solution, such as carbonates, arsenates, chlorides, silicates,
sulphides, etc.; or according to substances in suspension, such as sulphur, mud,
sand, silt, etc. In fact, every conceivable type of spring occurs in Yellowstone
Park and feeds the various streams (see Brues, ’24).

In the main, the influence of these springs is only a local one, that is,
circumscribed and limited to a very small area of influence, although at least
in the Firehole River in the Upper Geyser Basin the temperature and chemical
composition of geysers and springs modify the river water sufficiently to affect
the faunal makeup. But all these various factors modify the composition of the
spring biota and of the plants and animals found in their outlets.

It was the realization of the great variegation of the springs that quickly
dissuaded me from attempting to study in detail the relation of the fonticolar
to the torrenticolar associations. The excellent studies of the fauna of thermal
springs in Yellowstone Park made by Brues in 1923 indicate many of these
difficulties of ascertaining this relationship.

B. Classification of Habitats. Even at first glance it is evident that
mountain streams, with their major and minor falls, their many winding and
tortuous rapids and channels, their indescribable intermingling of tumbled and
placid waters, represent a highly specialized habitat, indeed, the pinnacle of
aquatic specialization, comparable only with shores of oceans and larger lakes.
Yet it is also obvious that this biotic unit is composed of simpler associations,
each with its individual peculiarities. A classification of these must necessarily
be based partly along physiographical, partly along physiological lines. Thiene-
mann (’12) offers a classification, with which, as indicated in part II of this
paper, I am unable to agree. For the sake of clarity the two groupings are here
placed side by side ;

After Thienemann.

A. Types of streams.

1. Rills and Springs.

2. Streams.

a. Elevated streams and middle
streams, trout streams.

b. Lower streams, grayling
streams.

B. Ecological Habitats.

1. Springs and rills.

2. Stream.

a. Open water.

b. Bottoms and surface.

aa. Stone fauna,
bb. Plant fauna,
cc. Quiet inlets and bays.

— on the water.

— among leaves and dead
branches.

— in sand and mud deposits
(slimes) .

aiiojuouai, nun v a. 1

g. Marginal pools.

h. Recession pools.

Author’s Classification.

A. Types of Streams.

1. Rills and Springs.

2. Streams.

a. Constant.

b. Variable or flood.

c. Precipitous.

d. Temporary.

(Springs and rills)

B. Ecological Habitats

1. Permanent habitats, with native

biota.

a. White water habitats — falls,
cascades, white rapids.

b. Clear rapids and stone bottoms.

c. Flacid water habitats — pools
and holes.

d. Marginal areas.

2. Interrupted habitats, with native

biota.

e. Deposits — on rocks, etc.

f. Splash areas — on rocks

3. Temporary habitats — transient and

210

Roosevelt H'ild Life Annals

In the foregoing classification I have tried to recognize the important fact
that the trout streams have a marked seasonal variation which emphatically affects
the living organisms. In this respect the mountain streams do not differ from
other aquatic bodies, which are also affected seasonally, and by the same factor,
namely the spring floods.

A further criterion needs stressing. Both Steinmann and Thienemann make
little distinction between what I have called native or endemic (or primary)
organisms and invading (or secondary) organisms. Yet I believe this dis-
tinction requires emphasis. Certainly animals and plants found only in some
small and isolated locality in one stream and not elsewhere can hardly be called
endemic ; similarly, insects the larval life of which is spent in quiet water and
which in their adult stages visit the marginal pools, cannot be called native to
mountain streams.

In Yellowstone River, for several miles below the outlet from Yellowstone
Lake, there occur several species of plants, such as Potamogeton, Myriophyllum,
and even some tiny patches of floating Lemna. This despite the fair current of
the stream. But it is only in this locality that these plants occur at all. Farther
down one never sees such plants. Yet below Yellowstone Falls there are
considerable placid stretches where one might expect higher plants to effect a
luxuriant growth ; such, however, is not the case. No plants above the algae
and mosses have been able to establish themselves.

Only those species can be said to be native or endemic which can exist
under the average conditions offered by the particular ecological region. For
ecological clarity, this distinction between what is primary and secondary,
between what is native and what is foreign, between what is of constant and
year round occurrence and what is temporary or transient, must be recognized.
If not, then there is no purpose in ecological study of any sort, since then a
sporadic or single occurrence would be of equivalent value to constant and
wide-spread occurrence.

Quite logically it may be asked : Is there a criterion by means of which

we can judge what life is endemic and what is foreign to mountain streams?
Fortunately there is such a test which to a marked extent permits one to sep-
arate the native from the foreign biota. This criterion is the food, hence
the stomach contents, of trout and other fish. Perhaps the food of stream
insects should be considered also, since insects, aside from fisb, constitute about
98% of the bulk of the stream fauna. At any rate, endemic species will be
found quite regularly among the stomach contents and represented by both
their larval and adult stages, or at least the larval stages; transients will be
found only intermittently and as adults.

C. The Habitats. (Springs and Rills). As already noted, springs and
tbeir outlets received scant attention in this study. While physiographically con-
tinuous with mountain streams, they are physiologically separate. The follow-
ing is summarized from Thienemann and Bornhauser. Bornhauser (’13) in a
special study of springs groups them into two types — limnokren and rheokren.
Rheokren are those springs from which the water flows away at the mouth.

211

Fig. 99.

Pteronarcys ccilifontica nymphs on rocks, “emerging 1 ’
from their nymphal skins. July 25, 1921.

212

Fiji. ioi. Cluster of Acroneuria adults in crevice of rock. Yellowstone River.

July 6, 1021.

Fig. 102. Stoneflies {Acroneuria) mating on shore grass along Yellowstone

River. July 6, 1921.

Ecology of Trout Streams

213

down a hillside; there is no trace of any pool, just a direct outflow. Limno-
kren are springs which have a pool of varying size (Fig. 79). As characters
for springs Bornhauser notes (a) constant renewal of “filtered," and hence
“fresh” water, and (b) constant and low temperature. These two characters
may be affected by altitude, age, the plants, temperature, environment (e. g.,
woods, open field), and food supply.

Thienemann ('12) classifies the fauna into three groupings for springs:
(a) animals that are subterranean, (b) land forms and “moisture friends,”
(c) true spring forms. Bornhauser recognizes further distinctions among these
types and groups them as follows: (a) Strange or foreign elements, (b) cosmo-
polites and ubiquists, (c) stream forms, (d) alpine forms (profound and
boreal), (e) true spring forms, and (f) subterranean forms.

In the rills (Fig. 63) formed by spring outlets the fauna, according to
Thienemann. agrees with that of the more quiet portions of mountain streams,
i. e., that of the deposits and pools. From a physiographical standpoint such
spring outlets are distinguishable by the stones which for the greater part are
not submerged, but only moistened at their bases. Where such rills and outlets
enter the stream proper, there one finds the repositories of trout eggs.

1. Permanent Habitats. These are the habitats which exist throughout the
year, which neither the spring floods nor the summer droughts can extinguish.
The same biota may be found here throughout the year. Mingled with it at
times, chiefly at low water stage, one may find a foreign element, but such
intrusion is rare. In these habitats the typical life of the mountain stream is
found.

a. White Water. — Here are included the falls (Figs., 68, 74 and 88) r
cascades (Figs. 62, 66, and 89) and white rapids (Figs. 66, 68, 84 and 95)
recognizable by the flow and color of the water. In falls the flow of the current
is vertical, in cascades oblique, while in the white rapids it is a mixture of
vertical, oblique, and horizontal flows. In falls and cascades the current flows
chieflv in one direction ; in white rapids the current is twisting, whirling, tumb-
ling, hence simultaneously of many directions. Physiologically, one must note
pressure and impact. Pressure arises from the volume of the water ; impact
refers to the force with which this volume strikes an opposed obstacle. As a
result of the impact the water becomes white in color; this is merely the white-
ness of the foam and of the air in solution. The water here is thoroughly
saturated with oxygen, indeed, is really supersaturated. Impact and whitened
waters, and hence a lack of transparency, are the more obvious characteristics
of white water habitats.

Under such extreme and violent conditions one would hardly expect to
find anything living at all. Yet curious to say, directly behind the falling
water, and sometimes in the midst of it, if the volumes of water are not very
large, one may find moss and bits of Cladopliora caught in tiny crevices. In
similar situations I have found Blepharoceridae larvae and pupae, and also the
pupae of Deuterophlebia (Figs. 106 and 107). The distinct dorso-ventral
flattening of the larvae and pupae, the powerful suckers which serve for attach-
ment, and the very slow locomotion can be recognized as excellent adaptations.

214

Roosevelt Wild Life Annuls

Where large rocks extrude from the white rapids and are dashed with
spray one also finds the adults of several species of Diptera running up and
down the rock or sitting quietly in the spray (Fig. 114). These are the adults
of Blepharoceridae, Chamaedipsis and Pliilolntra (Figs. 93, 108 and 114).

b. Clear rapids. — The clear rapids (Figs. 57 and 78) are distinguishable
by the steadier flow of the water, the lack of opposed obstacles, and the trans-
parency of the water. Typically they have a bottom of rounded, water-worn
stones. An acoustic distinction might also be noted. Thus falls have a note of
thunder, a deep bass ; cascades have an overtone as of something rushing by :
white rapids have a tearing, ripping sound ; while the clear rapids have a
laughing, bubbling tone.

This is the typical habitat of the mountain streams. Here the plants may
find a hold on the immersed rocks. Here Cladophora, Prasiol a>, and Nostoc
thrive. Here also Oscillatoria and Melosira may form films on the rocks, in which
various rock diatoms grow abundant. Here, too, Thremtna becomes prominent
among Trichoptera, crawling about on the rock surfaces; while Rhyacophila
and Brachycentrus fasten their cases in numbers on the upper surfaces of the
rocks (Figs. 104 and 105). Behind and beneath the stones a luxuriant fauna
and micro-flora appears. Ulothrix, Melosira, Oscillatoria, etc. form clumps in
which smaller micro-organisms may establish a circumscribed residence. Blepha-
roceridae larvae (probably also Deuterophlebia) wander freely all over the
stones. Groups of Trichoptera such as Hydropsyche, Philopotamus, Goera, etc.
foregather between and under the rocks ; flatworms such as Polycclis and Dcn-
drococlium lay their slime trails and feed on the micro-life; mayfly nymphs
such as Heptagenia, Ecdyurns, Iron, Ephemerella, Bactis hunt diligently; and
Acroneuria among stonefly nymphs finds this the optimal hunting-ground.

Here, too, trout of medium size like to establish their feeding ground, eat-
ing whatever the currents bring them, or picking suitable food off the rocks.
Occasionally they will make quick forays into the white rapids. In precisely
this sort of habitat I found bullheads in the Clearwater River in Idaho (opposite
the mouth of the Wietas River) ; similarly in Glacier Park, in MacDonald
Creek, just above its junction with the Flathead River, I observed Cottas.
Suckers, too, may be found on rare occasion.

Where stones come sufficiently close to the surface, Siniulinm may settle
and thrive. A thin film of water flowing over them seems their ideal require-
ment (see Thienemann, fauna hygropetrica). Where stones extend above tbe
surface especially the larger stones, the adults of Blepharoceridae. etc., form
the splash or “spray fauna” (Figs. 108 and 114).

All of these forms show some sort of adaptations. Either they are dingers
in structure or by means of their housings, or they are strong swimmers. The
dingers are quite generally flattened dorso-ventrally ( Polycclis , Dcndrococliuni.
Iron, Heptagenia, Acroneuria. etc.), or their housings are attached (webs of
Philopotamus, Hydropsyche, cases of Brachycentrus, Rhyacophila, etc.). The
swimmers are more cylindrically built. Perhaps to call them “darters" would
be more appropriate. For no animals truly swim in these rapids. Mostly, they
hover behind some rock out of the way of the direct current, and dash or dart

Ecology of "Trout Streams

- T 5

forth quickly for the next promising shelter. This is the method of the young
trout, also of older trout, of bullheads, of insects such as Callibaetis among
mayflies, Pteronarcys among stoneflies.

c. Placid Water. — Where a depression occurs in the stream, the water

tends to become placid (Figs. 58 and 82). The depression may be in the center
or the side of the stream ; it may be shallow, or very deep, but always dis-
tinctly beneath the average level of the stream. Usually the depressions are

lined with rocks, quite generally larger in size than those in the clear rapids.
These rocks become covered with deposits of various types of sediment after
the spring floods, although the spring currents do not always wash them away
entirely. Physiologically, there are the following conditions : quiet water, except
for the surface current which is strong but smooth ; because of this quiet there
is less need for attachment ; there is consequent liberty of locomotion. The
temperature and the oxygen content are the same as in the clear rapids.

The plants of the pools, except for the tiny algae and diatoms noted in
the clear rapids, are negligible. Despite the lack of deeper currents, there are no
submerged or emergent plants. Among animals we find the larger fish seeking
these pools which afford them freer movement. Of Crustacea there is Cambarus
(found only in lower elevations, not in Yellowstone Park streams), Pteronarcys
calif ornica (Fig. 99) and the smaller stoneflv nymphs. Occasionally one finds
a rare ostracod and perhaps even a cladoceran or copepod. Among Protozoa
the Vorticellidae find this an ideal environment, and may be found plentifully
on the rocks. Thienemann lists various species of Hydra for such pools. I
did not find any, probably because I failed to look for them. Caddisworms
and mayfly nymphs are uncommon in the pools ; but the smaller stoneflies
and the large Pteronarcys seem to frequent the pools in preference to the
more turbulent localities.

d. Marginal Areas. — Here the shores are referred to. The current is
lateral in such places, with considerable friction by floating debris. Little bayous
or inlets are sometimes formed, whose inhabitants are in part typically that
of the clear rapids, in part that of other aquatic communities (springs, ponds).
Surface skimmers prefer such areas. In streams of lower altitudes in Idaho
I have found leeches, annulates (probably a Limnodrilus, sometimes a Gam-
rnarus or a Hyalella, and Hydroptilidae among Trichoptera in these shore 01-
marginal areas. In addition, the moist soil affords abiding places for various
surface breathing larvae, as of Tipulidae, Leptidae, Coleoptera, etc.

2. Interrupted Habitats. This division is admittedly arbitrary. It is
characterized by being seasonal. While the habitats noted thus far may be
found at all times of the year together with their biota, those falling under the
present head are wiped out completely by the spring floods and must be repopu-
lated or recolonized every year after the recession of the floods, when the
currents are less violent. In fact, even during low water stages, during a
summer freshet, a temporary flood of a few inches may wipe out the deposits.
An interrupted existence is therefore characteristic of these habitats.

e. Deposits. The deposits on rocks (Figs. 71 and 83) and bottoms, and
along the shores are formed by sediments after the spring floods. These sedi-

Roosevelt Wild Life Annals

216

ments may be of varied composition, such as sand, mud, sulphur, etc. The
spring floods remove these deposits completely, scouring the rocks until they
are completely cleaned. As the floods recede, the deposits accumulate and the
fauna is restocked in such places. Because of this annual interruption one
can hardly speak of a resident fauna. Yet an abundance of smaller chironomid
species establishes itself in these deposits together with most of the representatives
of the micro-life noted under the clear rapids.

For European streams Thienemann notes the regular moss caps on the
rocks, which during the spring floods are deeply covered with water, but emerge
partly or completely during the summer droughts. In these moss caps may be
found many Chironomidae, less motile types of mayfly nymphs, Trichoptera,
Tipulidae larvae, Athcrix larvae, the free living stages of parasites, Hydra-
clmida, etc., in fact a fauna that requires considerable protection and which
can not exist in the mountain streams unless such protection (namely the moss
caps) is present. In Yellowstone Park such moss caps occurred only in a few
places along the Lamar (Fig. 72) ; in Idaho the Palouse River contains many
such moss caps. In both places the fauna was much as described by Thiene-
mann. Yet since these moss caps are rather the exception than the rule in
the mountain streams of the Northwest, I am inclined to consider the biota as
not truly native.

f. Splash Areas. — Like the deposits, these, too, are interrupted by the
high water. They are the rocks that protrude from the water and are con-
tinually moistened by the spray. In this spray certain adults of Diptera ( Bibio -
cephala, Philolutra, Chainaedipsis) run up and down, fly to the water's edge and
then run up into the spray (Figs. 93. 108 and 1 1 4) .

3. Temporary Habitats. These are habitats characterized by an invading
fauna and flora. They can be regarded as both transitional and transient. —
transitional as regards the type of organisms found, and transient as regards
their existence. They are discontinued during high water or low water periods.
The spring floods invariably wipe them out, and great recession of water will
do the same (Figs. 54, 65. 71, 72, and 83). Perhaps no stream showed the
relations of the regressive migration of the stream fauna together with the
progressive invasion of foreign biota so typically as did the Lamar River. As
the spring floods receded, sand was deposited copiously in the shore areas
among the stones. In many places there were small pools and minor rapids,
all connected by trickles (Figs. 75 and 76). The inhabitants migrated toward
the deeper parts of the stream if possible; others, especially attached forms
( Rhyacophila, Rrachycentrus ) became stranded or at least isolated: simultane-
ously, these pools became populated w T ith various beetles, occasional Flemiptera
( Corixa , Notonecta) , copepods, Cladocera. and also various filamentous algae
( Spirogyra , etc.), all of which are pond types. These pools were mosaic in
their makeup; that is. only a few representatives of the pond communities were
present in each pool ; in fact, the pools differed strikingly among themselves.
Later still, these recession areas dried up completely (Figs. 71 and 72) and
various shore forms and scavengers appeared.

21

Fig. 103. Adult mayfly with a parasitic worm ( Mermis sp.?) emerging from
the caudal end. Enlarged eight times. Note the loop formed by the parasite
within the abdomen of the insect. From confluence of Lava Creek and
Gardiner River. Aug. 4, 1921.

Fig. 104. Stranded caddisworms ( Brachyccntrus and Rhyacophila) .
Lamar River. Aug. 8, 1921.

Fig. 105. Attachment of Brachycentrus cases. Some attached
at caudal or small end others at broad or "mouth” end.
Lamar River. July 25, 1921.

106. Larva of Bibioccphala. enlarged 6 times. Note ventral suckers. 1 -rom
Snake River, Idaho. At right, pupae of Bibioccphala, showing variation in
size and ventral attachments.

219

Fig. 107. Bibioccphala pupae. From Lamar River. Note the tubes of
chironomids (T any tarsus?) on the same rock. July 16, 1921.

Fig. 108. Bibiocepliala adults as “splash flies." Oro Grande River, Idaho.

Sept. 3, 1924.

220

Fig. 109. Deuterophlebia larva. Yellowstone River. July 30, 1921.

Fig. no. Simulium sp. "Combs’’ of larva. xbO.

Ecology of Trout Streams

221

g. Marginal Pools. — The foregoing described (h) the recession areas.
The marginal pools can he primarily distinguished by their lateral connections
with the stream proper. Generally there is some small trickle that leads into
the pools (this includes the inlets or havous of Thienemann), or there may he
a broad lateral opening. At any rate, the current does not enter directly ; hence
there is a certain amount of quiet which permits less well adapted forms to
establish themselves. These are primarily species of the lake shores, springs,
and slow rivers. The fauna and flora can be called transitional, and is made
up of elements from rapid streams as well as the other aquatic bodies noted.
This is most notable in the types of Chironomidae, Trichoptera, and mayflies
that establish themselves, but most typically by the surface skimmers that hunt
in these spots.

The transition from stream to pond becomes marked in the various beaver
ponds, especially where these engineers succeeded in damming some mountain
creek. Such beaver ponds (Fig. 73) contain a rather sparse fauna, together
with scattered pond plants. Among the animals, Lestcs, Enallagma, and
Aeschna among Odonata; various beetles; swimming Trichoptera (Triaenodcs) ;
Notonccta and Corixa among Hemiptera ; and various semiaquatic Diptera are
present in varying numbers. As a rule, the heavers keep the ponds too well
cleaned to permit a rich pond life to establish itself. But at least some of
the types are represented.

D. Adaptations to Mountain Stream Life. The difficulties of mountain
stream life necessitate manifold adaptations on the part of the inhabitants.
Steinmann in his earlier paper (’07) has outlined these so thoroughly that I
can do no better than repeat them, with added instances. He lists seven types of
adaptations :

1. Dorso-ventral flattening. — By appressed structure, as found in Turbcl-
laria, leeches, Ancylus, stoneflv nymphs, mayfly nymphs (Iron, Hcptagenia,
Ecdyurus). Many cases of Trichoptera ( Goera , Molanna), Chironomus
( Tany tarsus) . Egg masses may he laid flat. The cocoons of leeches are flat.
Pupae of Blepharoceridae and Deuterophlebiidae also are very flat.

2. Enlargement of adhesive surfaces, e. g., Iron, other Ephemeridae nymphs.

3. Small body compass, tendency to dwarfing. Noticeable only in plants.

4. Attachments: Temporary and permanent. Permanent: with cases,

Brachycentrus, Rhyacophila, Thremrna, Glossosoma, Goera, etc.; without cases,
Simulium, leeches, snails, larvae' of Blepharoceridae, etc. Temporary, by means
of suckers, hooks, and anchors: various larvae of Diptera, Trichoptera, etc.

5. By weighting: web-spinning Trichoptera ( Hydropsyche , etc.).

6. Reduction of swimming hairs, which are reduced in number or
degenerate entirely.

7. Respiration. No surface breathers present.

Thienemann also notes an adaptation to temperature, namely the fact that
the Acarina of mountain streams lay fewer, but larger eggs.

To this I would add a final adaptation in habit, namely that of the food eaten:
the diet of inhabitants of mountain streams becomes generalized. Elsewhere (see

2 22

Roosevelt Wild Life Annals

Muttkowski-Smith) this has been formulated as an axiom: the more specialized
the conditions of the habitat, the more generalized the diet; the more generalized
the conditions of the habitat, the more specialized the diet.

Dodds and Hisaw (’24 and ’25) in their fine studies of the adaptations of
insects to swift currents, reach conclusions similar to those of Steinmann and
Thienemann.

FOOD RELATIONS

1. Trout Food. To arrive at a proper valuation of the stream population it
was necessary to examine the stomach contents of the fish. In the following
pages the stomach contents of a considerable number of trout are listed in tabular
form. From the contents, an estimate of which is given in cubic centimeters, the
bulk of the various food items may be inferred. To aid in this, the number of
specimens of a species eaten is also listed.

Table No. i. — Showing Food of Trout (Salnio clarkii ) from Yellowstone

River. June 28, 1921.

Number

I

234

5

6

7

8

9

10 1 1

12 13 14 15 16

Contents in cc. .

20

2 1 4 12

5

25

7

4

2.5

1 7

5 2 4 5 10

W ater Bait

Limnephilus . . .

20

■ ■ 3 i • ■

5

20

3

3

1

2

.... 44.

Rhyacophila.

......

4

......

Hydropsyche. . .

1

2 I

Pteronarcys ....

I

2

1

Aeroneuria ....

3

2 3 4

16

2

6

4 2 53

Baetis

I

2

I

Ephemerella . .

1 . . . .

4

4

4

I

Drunella

2

4

I

Blepharocera 1

5 20

3

Nematodes. . .

20

Tipulid lv

I

Surface Bait

Formica ad

I

1 1

Lg. red ant .

.... I

Sm. black ant..

.... 6

1

Vespa

.... I

Bee

I

Simulium ad. .

........ I

Culicid ad . .

7

Plecia ad

• •

• • 1 1 3

I

3

3

2

20

15 9 30

Mayfly

I

2

Cicada ad

I

. . 1 . . . . 1 1

Lucanid ad . . .

■ ■ 1 • •

Coccinellid ad. .

. .

. . . . j 2

Cerambycid ad .

. . . . I

I

Hydrophilid ad.

I

1

. . . .

Scarabaeid ad . .

......

I

.. .. .. .. ..

Ecology of Trout Streams

223

Table No. 2. — Showing the Food of Trout ( Salmo clarkii ) from Yellowstone

River

Number

June 29, 1921

June 30,

1921

I

2

3

4

5

6

I

2

3

4

I

5 6

7

Contents in cc

30

30

7

4

10

7

40

15

3

6

4 1

6

Pteronarcys lv

I

3

I

Acroneuria lv

28

23

8

4

16

6

4

8

I

6

4 2

I

Ephemerella nvmph

4

5

I

3

Rhvacophila lv

I

I

3

I

Limnephilus

2

Formica ad

I

I

Plecia

I

8

2

20

Lvmpvrid ad

I

Hvdropsyche lv

4

Black ant

I

I

Simulium ad

I

I

Culex ad

I

I

Plecia ad

44

4

Leptid ad

I

Muscid ad

I

Table No. 3. — Showing the Food of One Trout Taken from Yellowstone
River, July 6, 1921 by Mr. A. G. Whitney. Contents About 25 cc.

Pteronarcys larva

Acroneuria larva

Acroneuria adult

Mayfly adult (Ameletus)

Platyphylax ad

Simulium ad

Winged ant

Sawfly

Ichneumon

I

4

10

20

1

1

6

I

1

Plecia

Muscid ad

Lucilia. ad

Enellagma female ad

Raphidia ad

Capsid ad

Cerambycid ad

Scarabaeid ad

Lampvrid ad

30

1

1

1

1

1

1

1

1

Table No. 4. — Showing Food of Small Trout and Minnows from the

Lamar River, July 7. 1921.

Trout

T 3 .—

1 4

2| in. long

I

Minnows
J-2| in. long

Number

I

2

3

4

5

6

7

8

9

10

1 1

12

I

2

3

4

5

Ameletus sp. small

1

I

4

I

1

I

I

I

I

I

Chironomus lv

I

I

1

I

I

I

I

I

Alloperla sp

I

Ameletus sp

I

I

I

I

I

I

I

I

Chironomus sp

22 4 Roosevelt Wild Life Annals

Table No. 5. — Showing the Food of Cutthroat and Rainbow Trout from
the Yellowstone River. July 13, 1921.

I

2

3

4

5 6

Pteronarcys lv

5

1

Pteronarcvs ad

IO

2

2 ....

Acroneuria lv

— 4

Acroneuria ad

2

2

.... 2

Alloperla ad

I

Bibiocephala ad

4

2 2

Ameletus lv

.... ^

Limnephilid ad

I

2

Black ant

I

I I

Green bee

I

Cicada

3

Cerambycid ad

. I

Wood chunks

4

Young trout or sucker

The wood chunks were respectively 1x2x5 an< J 2 x 3 x 4 cm. The young
fish eaten by number 7 was six inches in length.

Table No. 6. — Showing the Food of Native Trout from Yellowstone River
at Mouth of Elk Creek. July 15, 1921.

Number 3 had a copepod gill parasite.

Fig. hi. Athcri . r cluster. Lamar River. July 22, 1921. The insert shows a section

of the cluster enlarged X4.

Fig. 1 12. Rock where Athcri . r cluster had formed, on underside in cleft of
rock. Lamar River, Aug. 7, 1921. On July 22, when the cluster was
photographed, the sharp edge running from left to right was about twelve
inches above the rapids.

226

/ ; -

Fig. 1 13. Tipulid larva, with inflated “baloon” at posterior end. Tower Creek.

Aug. 29, 1921.

Fig. 114. “Spray flies” ( Chamaedipsis and Philolutra ) on rock
in rapids. Lamar River. Aug. 31, 1921.

Ecology of Trout Streams

227

Table No. 7. — Showing Food of Trout from Yellowstone River at Mouth

of Elk Creek. July 16, 1921.

I

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Pteronarcys ad . . .
Acroneuria ad ... .
Acroneuria lv . .

2

2

5

5

1

2

I

I

5

4

4

1

3

5

5

3

4

1

1

3

I

5

1

3

2

2

1

2

2

2

1

Hydropsvche lv. .

Hydropsyche pupa
Hvdropsvche ad. .

6

Ameletus ad

I

I

1

1

Chironomus pupa.

Plecia ad

I

I

4

Tipula ad

Tipula pupa

I

I

I

Pentatomid ad . . .

Cicada ad

Cerambycid ad . . .
Scarabaeid ad. . . .

I

I

I

1

Black ant

4

8

Red ant

5

Argynnis ad

I

Attid spider

I

One of the specimens showed a Copepod parasite adhering to a gill.

Another contained a Nematode in its stomach. This may have been taken
in with food or as a free-living stage of the parasite.

Besides the sixteen specimens noted above, twelve others were noted for
their food content. In all cases the food comprised the two species of stone-
flies. One specimen showed an adult Lestcs sp.

Table No. 8. — Showing the Food of Trout from the Yellowstone River,
July 17, 1921. Numbers i-ii from Foot of Hot Springs Near the
“Needles,” Numbers 12-16 Near Mouth of Elk Creek.

I

2

3

4

5

6

7

8

9

10

II

12

13

14

15

16

Pteronarcys ad. .

I

4

2

4

I

3

1

4

2

4

3

7

1

3

1

5

2

6

Acroneuria ad ... .

I

2

500

3

5

I

2

3

Tipulid ad .

I

I

Chironomus pupa.
Cicada ad . .

1

I

I

Ant

I

Sawfly . .

Enallagma ad .

1

Carabid ad

1

Number 2 contained a huge number of adult caddisflies, approximately
500 specimens.

228

Roosevelt Wild Life Annals

Table No. g. — Showing Food of Native Trout from Yellowstone River at
Hellroaring Gap. Numbers i-io Collected by Mr. A. G. Whitney,
the Others by Various Fishermen.

Ji

LY 21 , I92I

July 23, 1921

I

2

3

4

5

6

7

8

9

10

1

2

3 4 5

Pteronarcys ad

4

6

3

2

I

4

3

2

8

5

12

8

4 2 1

Acroneuria ad

3

4

2

I

I

I

1

2

4

3

4

10

2 ... . 3

Noctuid moth

I

Hydropsyche ad ... .

I

. ... J .... |

July 2 4, 1921. Lamar River. Trout collected by Mr. Henry Lambert.
Three specimens, each with one minnow in stomach. No other contents.

July 25, 1921. Yellowstone River. Trout collected by fisherman. Two
specimens noted, each containing a minnow.

August 1, 1921. Yellowstone River, at Cooke City bridge. Three trout
noted feeding on caddises of Rliyaeophila.

August 1, 1921. Lamar River. One native trout with two Doroncuria and
two grasshoppers.

August 8, 1921. Lamar River. Two trout, each with two grasshoppers.

August 11, 1921. Yellowstone River. Trout taken by Mr. Arnold

V hitehouse at hot springs near the “Needles.”

I

2

3

4

5

6

7

8

9

10

1 1

12

Ameletus lv

3

3

2

3

2

2

I

Caddises

Acroneuria lv

I

I

Grasshopper

I

2

2

I

I

1

August 11. 1921. Slough Creek. Two specimens gathered by Mr. Robert
Hallenberg, one containing a field mouse, the other a green frog.

August 13, 1921. Yellowstone River. Two specimens collected by

fishermen, one containing 3 Ameletus and 10 Siviulimn larvae, the second with
about 200 Simulium larvae.

These 148 stomachs represent only a portion of those examined, probably
about one third. Since the others were practically identical as to their contents
it was not considered worth while to list them all individually. Toward the
middle of the summer only such specimens were noted as had fed on unusual food,
such as mice, frogs, pieces of wood, or on one type of food primarily.

From the listed data it will be seen that it is possible to follow the summer
history of the fish food, beginning with the spring floods, the summer lapse,

Ecology of Trout Streams

229

and the autumnal re-stocking of the streams. The general history has been
outlined in an earlier paper: “The Food of Trout in Yellowstone National

Park’’ (Muttkowski, ’25). Certain conclusions offered in that paper are rela-
vant and are quoted here: “From the examination of fish stomachs it is possible
to deduce much about the food habits of fish. First, fish will take food that
is easily captured ; secondly, that which is accessible with difficulty ; and lastly,
strange and unusual food. Of the food available in mountain streams, stone-
flies and mayflies are most easily obtained and constitute the major portion of
the food eaten. Here, too. the' eggs, fry and fingerlings should be listed, as
fish are cannibals when opportunity offers; and trout are no exception, hut
will eat other fish just as greedily as will bass. Caddisflies, well protected by
virtue of their tough attached cases, and moreover, even more inaccessible on
account of the appressed structure of many cases, rank in the second category,
and for that reason constitute a much smaller item of fish food than the other
two groups of insects. However, they are used extensively as food by the
stoneflies and mayflies, and in this respect become as important, though
indirectly so, as their enemies. Among strange and unusual food? can be listed
the surface bait.

“In general, fish are opportunists as far as their food is concerned. They eat
what animal food is available, regardless of the origin. As a result, if one knows
the animal life of a particular region, one can tell from the stomach contents
where a fish has fed. In a lake, for instance, the plant and animal life is dis-
tributed in regular ‘zones,’ most animals limiting themselves to particular depths.
Some are found only on the shores, others on the vegetation in the shallows, still
others, only in the muck at considerable depths. With a knowledge of the animals
found in these various zones it is possible to learn a good deal about the food
habits of a fish, his migrations, and his preferences.

“On the whole, fish are indiscriminate in their choice of food as far as
quality is concerned. They like to feed in a particular region, and stay there
until satiated. Thus, when feeding from plants, they eat whatever they can
find there ; and once they begin to feed from plants they continue feeding there
until their hunger is satisfied. At periods of plentiful food, fish do not migrate
while feeding.

“Curious to say, it often happens that a fish may find a certain type of
food so much to his liking that he will seek only that type. This may be worms,
leeches, snails, back swimmers, caddis-worms, or other kinds. Thus, we mav
find stomachs filled with dozens of individuals of one type of animal, such as
crayfish or snails. Even more striking, one may find stomachs filled with highlv
distinctive cases of some particular species of caddis-worm.

“This predilection for some particular food is more often observed in the
case of surface food than in water bait. I have found fish stomachs, including
trout stomachs, gorged with hundreds of specimens of a single type, such as
ants, grasshoppers, dragon flies, caddisflies, orl flies, mayflies, midges, etc.,
indicating that the particular fish had taken a fancy to this special type of
food, and had hunted assiduously for the delicacy. There is nothing abnormal
in such a predilection, not more so than in the case of a boy who makes a meal

- 3 °

Roosevelt Wild Life Annals

off desserts, be it ice cream, or fruit, or cake. But right there, in the longing for
the unusual, lies the weakness of the feeding habits of fish, the trait which lays
them open to capture by the angler. Since the unusual attracts, anglers have
made use of this phenomenon in the types of flies selected by them.

“A fish is easily deceived, for he is not very observant. His eyesight is
poor and he recognizes things chiefly through their movements. For instance
when an angler uses a fly, the fish is supposedly deceived by three factors, —
form, pattern, and movement. In the matter of form and pattern the fish’s
vision is too weak and nearsighted to recognize the bait for what it is. He
is used to certain distorted images which impress him more through motion
than by any other factor, and he captures or tries to capture such a moving
object. But it is also the unfamiliar, the unusual, which tempts fish, perhaps
more than the customary objects. How else can one explain the presence of
blocks of wood, of straws, twigs, leaves and the like, in fishes’ stomachs?
On more than a dozen occasions I have found blocks of wood in trout stomachs.
In at least half the cases there was not the remotest resemblance in the shape
of the block to any type of surface bait. An irregular cube has no resemblance
to any insect, while an oblong bit might well have the approximate outline
of a stonefly or a grasshopper. But it was probably the strangeness, the unusual-
ness of the block of wood which attracted and tempted the trout. The most
interesting feature of these instances was that in only one case were the blocks
of wood taken by a hungry fish ; that is. only once were the blocks of wood
the sole stomach content. In all other instances, blocks and sticks were gulped
by fairly well-fed fish. One might say that they were taken as a sort of
salad or dessert, indicating that their novelty tempted the fish."

2. Insect Food. A special study of the food of the insects was made
possible by the collaboration of Dr. Gilbert M. Smith, then of the University
of Wisconsin, who arrived early in August. Specimens were obtained from
each of the four type streams and their stomach contents studied, Dr. Smith
identifying the plant organisms and the writer the animal life. The results of
this particular collaboration are published in a separate paper (Muttkowski-
Smith : The Food of Trout Stream Insects in Yellowstone National Park,

Roosevelt Wild Life Annals, Yol. 2, No. 2).

For the purpose of the present paper the following may be noted from the
summaries given in the collaboration :

Ecology of Trout Streams

231

Table No. 10. Showing the Summaries of Food of Trout Stream Insects.

Yellowstone Park, 1921.

Name

Number of
specimens

Animal food

%

Plant food

%

Detritus

%

Perloidea

— Pteronarcvs

26

3.85

53 • 85

42.3

— Acroneuria

49

77-4

6-3

16.3

— Perla . .

8s.

IS-

Average for

80

= 54 -

22.3

23 -7

Ephemeroidea

— Ameletus ....

2 s

22 . 7

77-3

— Baetis sp

2

SO.

50.

— Drunella

20

18.5

36.8

44 7

- — Ephemerella

28

3-7

363

60

• — Hept agenia

34

24.

76

Average for

109

= 4-3

29.7

66 .

Trichoptera

■ — Rhvacophila

32

49

23 -

28.

— H vdropsyche

27

42.

54 3

3-7

— Brachycentrus

33

18.3

72.7

9 -

■ — Thremma

23

64.

36 .

Average for

105

= 28

54 -

18.

Diptera

— Chironomidae

6

71 .

29.

— Simulium

14

79

21

Average for

20

=

76 . 6

23 4

“Even from so brief a study as the foregoing certain facts can be gleaned.
The most notable point is that aquatic insects in rapid streams are opportunists
as regards food, and eat whatever is available. Secondary to this is the fact
that the aquatic insects forage extensively, i. e., migrate freely in search of
food.

“Both of these points become evident from the collections made in Lost
Creek, where special efforts were made to select various spots in the creek
for sample collections of food and of specimens in the vicinity or upon that
food. Reference to the stomach contents of the individual specimens and com-
parison with the food items listed for the particular spot shows at once that
the large majority of specimens contained food that did not occur within main-
feet of the particular locality. This indicates that these species must be rovers
and foragers to a marked extent, and that they are opportunists on the whole
and eat whatever is available.

“Environmental conditions in mountain streams are strenuous ; the strong
current in particular makes life somewhat precarious and selective feeding
difficult. Hence as a result the diet becomes diversified ; the insects take what-
ever comes along, be it plant, detritus, or animal matter. Their diet thus becomes

2 3 3

Fig. 1 1 5. Parnidae larva. Psephenus Iccoutci enlarged xio.
From Lake Mendota, Wis.

Fig. 1 1 6. Parnidae larva. Elmis z'ittatiis enlarged xio. From Lake

Mendota, Wis.

Ecology of Trout Streams

233

much more generalized than the diet of related species in more permissive
habitats, such as ponds and slow streams. In the latter the less strenuous con-
dition's permit the insects to select their food, that is. to restrict their diet to
favored food, and to search for that food.

“In other words, specialization of habitat leads to diversification of diet
(i. e., generalization of diet), while generalization of habitat permits a restriction
of diet, (i. e., specialization of diet).”

3. The Seasonal Food Cycle. Here. too. I prefer to quote from a prior
paper (The Food of Trout in Yellowstone Park) : “Besides these main items
of the normal trout diet in the mountain streams, the so-called ‘water bait,’
there is the surface drift or surface bait of water-trapped animals, chiefly insects.
This comprises especially the weak fliers such as moths, ants and grasshoppers,
while spiders, centipeds, mice and other animals may occur. But the life of
a trout stream is dependent on its normal inhabitants, not on the odds and
ends which a kind wind or accident may provide. It is only during the brief
summer period that surface bait becomes important ; and for a period of four
to six weeks the fish are largely dependent on this type of food for their exist-
ence. That the emergence of their natural water bait, with the resulting depletion
of this primary food supply, should be synchronous with the summer flights of
ants, moths, grasshoppers and other poor fliers that are easily water-trapped,
is one of many instances of the admirable provisions of nature.

“Indeed, this is carried still further at this period. At this time the minute
life of the shore waters, especially the shore diatoms, flat- worms, chironomids,
and the young stages of mayflies, stoneflies and caddisflies, receives a tremend-
ous impulse and becomes quite prominent. At this period also the young of
trout, suckers and other fish in the mountain streams can be found in the shore
pools and shallow rapids feeding on the minute organisms in these places. Here
lies the remarkable coincidence : the simultaneous appearance and growth of
fish fry and of a protected food supply for its use. For the older trout are
unable to get into these shallows, which therefore offer both protection and
food to the young fish.

“From the foregoing it is evident that there exist only two well-marked
periods in the annual cycle of mountain trout streams, namely, a ‘water food’
period covering nearly eleven months of the year, and a ‘surface food’ period
occurring during the summer, and lasting from four to six weeks. This is the
period when trout, as anglers put it, ‘rise to bait.’ These same periods might
also be called flood and ebb periods, or flood and drought periods, from the
fact that high water lasts from October to July, while the ebb or low water
stage of the summer is really very brief.

“With the fall rains the brief low water stage ceases and the conditions
revert to those existing during winter and spring, and continue to the time of
emergence described, that is, about the first week of July.”

234

Roosevelt Wild Life Annals

COMPARISONS AND SUMMARIES

Definition of Mountain Stream. Steinmann (’12, p. 125) in his fine paper
on alpine streams ofifered the following summary :

1. Conditions and character of mountain streams are:

a. Constant and low temperature.

h. Strong current and marked changes in volume of water.

c. Rich oxygen content.

d. Paucity of plants.

e. Stony bottoms.

2. These characters impress the animal world as follows :

a. Composed of resistant cosmopolites and stenothermic freshwater forms.

b. Body shape; morphological adaptations, arrangements for attachment

holdfasts, dorsoventral flattening, dwarfing, etc.

c. Life: low food requirements, embryonic development prolonged.

3. The fauna consists of

a. Cosmopolites, ubiquists.

b. Torrenticolar-profound elements.

c. True stream animals.

4. The fauna is composed of forms which originally belonged to still water.

The influence of stream life is recognizable to a different extent in each
form.

5. All true mountain stream forms can be regarded as glacial relicts, since they

agree more or less closely with Zschokke’s postulates in their :

a. Simultaneous occurrence in mountains in the north.

b. Simultaneous occurrence in mountains and in lake depths.

c. Absence from the warm, still and slow moving waters of plains.

d. Reproduction during low temperature.

This characterization, while very excellent so far as the fauna is con-
cerned, is primarily intrinsic, in that it stresses the responses of the animals
rather than the extrinsic factors which elicit this response. In the present study
I have attempted to outline what these external factors are. Based on the data
presented, the following definition of the mountain stream as an ecological unit,
including both positive and negative characters, is ofifered :

I. Physiocraphical Characters.

Positive. N egative.

1. Constant and considerable change in

elevation.

2. Slight depth.

3. Great and successive changes in types of

bottoms.

4. Volume of water relatively small, but im-

pact very great.

II. Physiological Characters.

1. Strong and rapid current, resulting in 1. Uninfluenced by winds.

2. Equal distribution (= circulation) of; 2. Do not freeze over in winter.

a. Low and constant temperature.

b. High oxygen content. 3. No stratification of temperature and oxygen

c. Substances in solution, affecting the content. Hence no thermocline.

chemical composition.

d. Substances in suspension, affecting

the turbidity or transparency.

Ecology of Trout Streams 235

III. Ecological Characters.

1. Types of habitats.

a. Permanent, with native biota : white

water, clear rapids, placid water,
marginal areas.

b. Interrupted, with native biota : de-

posits, splash areas.

c. Temporary, with transient and tran-

sitional biota : marginal pools,

recession areas.

2 . Distribution of habitats linear or hori-

zontal, all types intermixed.

3. Plant life confined to algae and mosses;

clinging types.

4. Animal life.

a. Respiration : all water breathers.

b. Locomotion : chiefly dingers and

crawlers. Swimmers move in
“leaps.”

c. Structural : usually well protected by

armor, cases, and webs.

d. Composition : stoneflies, Blepharo-

ceridae, Tricladida, are most
characteristic Invertebrates, trout
most characteristic Vertebrates.

e. Food habits generalized, other habits

specialized.

Even a superficial examination of these characteristics shows clearly that
whatever variability exists in different parts of a mountain stream is due to
physiographical factors; the physiological factors are quite constant. In this
respect the mountain stream as an ecological unit is outstandingly different
from other units, since in these the physiological factors are variable, and the
physiographical ones quite constant.

Of physiological factors only the current varies, a condition which depends
entirely on the physiography. All other factors are relatively constant ; thus
temperature, oxygen content, speed of circulation, chemical makeup, all tend
to be alike in all parts of a specific mountain stream.

Another characteristic is that these conditions are distributed horizontally,
namely, there is no vertical distribution, and hence a lack of the zonation one
finds in other ecological units.

Other Aquatic Bodies. As already noted, in this study the factors of
the environment have been stressed rather than the responses of the organisms
to these factors, since it is only through the study of these factors that the
responses can be understood. The following table is an attempt to present in
comparative form the factors that affect the various aquatic habitats, and the
chief responses to these factors. The arrangement is based on personal studies
covering some twenty years.

— No vertical distribution of habitats into
zones, as in lakes and slow streams.

— No emergent vegetation, no higher plants.

— Surface-breathers absent.

■ — Free swimmers few. Burrowers are scarce.

— Soft-bodied forms rare.

— Plankton organisms absent, Crustacea and
Annulata rare. No Amphibia.

Showing in comparative form the factors which affect the various aquatic habitats and the chief responses

to these factors.

Roosevelt Wild Life Annals

Table No. ii. ( Continued )

Ecology of Trout Streams

237

be

o

CQ

c c a C

| 1 C

1 1 1 1 M

O U uZ O Cj

rsoo°d>a>oo2 t ojurjd>

aj a; fli fli

»SS s cc"cc

rta3a3o2°oj22

p « o
C a) ft v S

£c?c

o O § o

u 1 ) u u
?CCCCCC;
OOOOOO;
c C C C C C--

a

2

a

£

c/3

%

■ S s s
>,000

•2 s s 3

•2 Ji O O

c c
o o
E E

tigdeSp

missels *' ee sieessHecc

^oooco^ooioooooooooo

-1 - r , . r 1 • . . ■ -1 - , r 1 1 ■ i~ r-

<0 0 0 0 0'

If

C C +»

oof.

rr r-> O

•. d C 3 m ii d
eJ}cS?dCC>H

SS88i8gg^8

KJ VI

-*TJ
J& C
s» 0

J,“

a c
.c c«
C/3

>> £ i* g

SgSssJ

C O C o D*
c o c u Od a

S — .ccc 2 ? ccccc~cccc~ t .c . .
oo .ooo$ooooo§oooog§o . •

■glfcESEsEEEEElEEEEllEdjo

2R>££EsEHEG£sHcGHtsScc
•O o j5 O O O O O O O O O O O O O O o O O o o

OOOUOOOOO'

OO O O

: u o c c

§ g § § 1 1 s : ■; ;||

gS55SgS<uoo£f

EsESGGtdccEE

0000000 OOO 00

O'o'oVoOO'oCCCoO

£ S

O ci

rr O

• •>*>>•>»

• ■ c c • c

£ £ oJ rt > aJ

G Ei2 G

00 .000 r; uuouu

Kx^GEEdEEEEE

SEES

£ D D

silEEE^SEEEEfeEEEEfefeEcc

oo§ooo£ooooo£oooooo§oo

I O O O O O

c . • c c .

o • • o ° ° N

c^iEEscddEE

00 ^ 0000222 °°

OCu<oOO<oCCCOO

& & g & & i
.a .2 ■£,.£.£ 8

& & 6 & &

C

o

E o

p.p. = t>f>^^S2§SgESSSEgE6E^

jifl)Oflj(j»i(iiO(l°[5Ort(SrtcilOUcjOO0J

Ooooo OUOUOl-i.l.LOCl.Olflo

e „

£ S
° 2
o C

-*-* -*->
c c
o o

d 3 a.

E.h.h § « H E 2 ,«? o

a: a

• 3 <u

•C

;||S1
• E 1 § 8

S c c c c

o o o o o

c c c c c ! c c

Souuuu .OOOOC - OO

§■§ E E E E lEEEEE^EE

~'cccc.cic.ccl: -

rscEEE^SS&E

OOOOOO -
<o q o o o o

G^EGEEcdEEd^

o53°°°°°2°°2oo

o d 00000 C 00 gt£<£

c

- O -

0) 2 o o
c £ e c
o o o o
c o c 2

eg...
2 35 o O O
c 5 c c c

oo222

o <o c c c

: : ‘ o 2 :
O 0 O 2. E O
c c c % £ e
o o o ° o o
C C Cv y C

o o

15 2

d«

o 0

> c c

r* n n

* S3

• o

^ ? d k ^

o 2 ° o o

C o u u ■

. s 33 d)

o v. 2 * a* 2

b ^ S « o 2

c S

a <v

£S

O „ . • O O 2 <y a,

2°>ccgcc

°Eo22°22

o u. <a c c o c c

EdJJc H
o 2 oj 2 03

3

CT >

C id S3 u 'o :

C ™

o

I 1 1 1

1 1 & S' - 1 s's;seEj;j;s;|s|Ecs;si;||gE££ e£

J§^.i2§ 2i2822222g2S2c282gcg8SS 8c

« o* « c c
^.^«22

35 35 4) a;

uSiCC

t3 o o 2 2

u <o vi C S

C

c« a> -
H be." ^

cJ

• o

o

_ u
T3 Qj

• CU

E 2

> v n t cj !/i ~

2 S^’E^p °.GP

a W&hc/:<SS

gc^ w . TO _

53 § =•“

2 fe « <u S oDSg^

' U V- •— t •- rr-l 1

5 0$-^,

J, EL 'O +->

: o >,«

Z c §1 §1 2 |ll| |'g-c!ii <*«S S & &.& 3 .T & 1 | gs

<e,Ka.Z3s<jScaaa.o<KWOffiZLiH^m oua, JhflOQMo<

238

Roosevelt Wild Life Annals

From the table it would be possible to list separately the characteristics of
each type of aquatic habitat. But since these are already emphasized in the
table by italics, it would be mere repetition to go through such a procedure.

In so far as the main work of this study is concerned, it is frankly admitted
that the results are by no means complete, or even remotely so. The streams
of the Northwest, in fact, throughout the United States, offer a multitude of
problems which need prolonged investigation. If the present paper succeeds
even to a slight extent in stimulating interest and study of some of these prob-
lems, it will be a recompense for the considerable effort, time, and money spent
by the writer and other agencies noted in the introduction, in carrying on the
present study.

BIBLIOGRAPHY

An inclusion of all the directly or indirectly relevant literature would
require the listing of more than a thousand titles. Further, such a list would
for the larger part be a repetition of what is listed in recent and readily acces-
sible bibliographies. Hence only directly pertinent literature is cited, together
with some titles not listed in prior bibliographies. Special attention is called
to the various papers in Ecology, University of Toronto Studies (publications
of the Ontario Fisheries Research Laboratory), Illinois Natural History Survey,
Notes of the Biological Laboratory of the Wisconsin Geological and Natural
History Survey, Annales de Biologie Lacustre, Archiv fiir Hydrobiologie
und Planktonkunde, International Revue der gesamten Hydrobiologie und Hydro-
graphie, Abderhalden’s Handbuch der biologischen Arbeitsmethoden.

Alexander, Charles

1925. An Entomological Survey of the Salt Fork of the Vermillion River
in 1921. with a bibliography of aquatic insects. Bull. 111 . Nat. Hist.
Surv., 15, pp. 439-535. (Literature prior to 1925 listed.)
Bornhauser, Konrad

1912. Die Tierwelt der Quellen in der Umgebung Basels. Int. Rev. d.

ges. Hydrobiol. u. Hydrogr., Biol. Suppl. 5. pp. 90. (310 titles listed.)
Brues, Charles T.

1924. Observations on Animal Life in the Thermal Waters of Yellowstone
Park, with a consideration of the thermal environment. Proc.

Am. Acad, of Arts and Sci., 59. pp. 371-487 (good bibliography).

1924a. Observations on the fauna of Thermal Waters. Proc. Nat. Acad, of

Sci., 10, pp. 484-486.

Carpenter, George H.

1923. Insect Transformation. Pp. 1—280. Dodd, Mead & Co., New \ork.
Dvxlds, G. S. and Hisaw, F. L.

1924. Ecological Studies of Aquatic Insects. I. Adaptations of mayfly

nymphs to swift streams. Ecology, Vol. 5. pp. 137-148. II. Size
of Respiratory Organs in Relation to Environmental Conditions.
Ecology, Vol. 5, pp. 262-271. III. Adaptations of caddisfly larvae
to swift streams. Ecology, Vol. 6, pp. 123-137. 1925.

Ecology of Trout Streams

239

Dyar, Harrison G. and Shannon, Raymond C.

1927. The North American Two-winged Flies of the Family Simuliidae.

Proc. U. S. Nat. Mus., Vol. 68, art. 10, pp. 1-54. A monograph.

Greene, Charles T.

1926. Description of larvae and pupae of two-winged flies belonging to the

family Leptidae, Proc. U. S. Nat. Mus., Vol. 70, art. 2, pp. 1-20.

Hatch, Melville

1925. An outline of the Ecology of Gyrinidae. Bull. Brooklyn Ent. Soc.,
Vol. 20, pp. 101-114.

Kennedy, Clarence H.

1927. Some non-nervous factors that condition the sensitivity of insects to

moisture, temperature, light and odors. Ann. Ent. Soc. Am., Vol.
20, pp. 87-106.

Malloch, John R.

1917. A preliminary classification of Diptera, exclusive of Pupipara, based

upon larval and pupal characters with keys to imagines of certain
families. Bull. 111 . State Lab. Nat. Hist., Vol. 12, pp. 161-409.
Muttkowski, R. A.

1918. The Fauna of Lake Mendota: A Qualitative and Quantitative Survey,

with special reference to the Insects. Trans. Wis. Acad. Sci.,
Vol. 19, pt. 2, pp. 374-482. (Bibl. of aquatics).

1920. The Respiration of Aquatic Insects: A collective review. Bull. Brook-

lyn Ent. Soc., Vol. 14, pp. 89-96, 131-141. (Bibliography).

1925. The Food of Trout in Yellowstone National Park. Roosevelt Wild

Life Bull., Vol. 2, pp. 470-497.

Muttkowski, R. A. and Smith, Gilbert M.

1929. The Food of Trout Stream Insects in Yellowstone National Park.
Roosevelt Wild Life Annals, Vol. 2, No. 2, pp. 241-263.

Needham, James G. and others

1922. A Biological Survey of Lake George, N. Y. State of N. Y. Conserv.
Comm., pp. 1-78.

Needham, James G.

1927. The Rocky Mountain species of the mayfly genus Ephemerella.

Ann. Ent. Soc. Am., Vol. 20, pp. 107-116.

Pearse, A. S.

1926. Animal Ecology. Pp. 1-417. McGraw-Hill Co., New York.

(Literature listed, but without titles).

Platt, Ernilie L.

1916. The Population of the “blanket-algae” of freshwater pools. Am.
Nat., Vol. 59, pp. 752-762.

Rousseau, E.

1921. Les Larves et nymphes aquatiques des Insectes d’Europe. Vol. 1.

pp. 1-959.

240

Roosevelt ll'i/d Life Annals

Steinmann, Paul

1907. Die Tierwelt der Gebirgsbache. Eine faunistisch-biologisch Studie.
Ann. Biol. Lac., II, pp. 1-137.

Thienemann, A.

1912. Der Bergbach des Sauerlandes. Int. Rev. d. ges. Hydrobiol. u.

Hydrogr., Biol. Suppl., 4, pp. 1-125. (Extensive bibliography).

1926. Der Nahrungskreislauf im Wasser. Verb. d. D. Zool. Ges., Vol. 31.

Jahresversammlung zu Kiel., pp. 29-79. (Summary of literature
since about 1918. Bibliography).

1926a. Hvdrobiologische Untersuchungen an den kalten Quellen und Bachen
der Halbinsel Jasmund auf Riigen. Arch. f. Hydrobiol. Vol. 17,
pp. 221-336.

1926b. Der Bach und der Fluss. in Abderhalden, Handb. d. Biol. Arbeits-
methoden. Abt. IX, teil 2, lste Halfte, heft 1, pp. 83-96.

Tillyard, R. J.

1920. Report on the Neuropteroid Insects of the Hot Springs Regions, X. Z.,
in relation to the problem of trout food. Proc. Linn. Soc. X. S. \\ .,
Vol. 45. pp. 205-213.

Ward, Francis

1920. Animal Life Under Water. Pp. 1-178. Funk and Wagnalls, X T e\v
York. (Of great interest because of studies by photographic
methods).

Yellowstone National Park

Bull. U. S. Dept. Interior. Publ. annually. With descriptions of forma-
tions, photographs, maps, list of vertebrate animals, plants and
bibliography of popular and scientific papers.

THE FOOD OF TROUT STREAM INSECTS IN YELLOWSTONE

NATIONAL PARK

by Richard A. Muttkowski * and Gilbert M. Smith **

Collaborators, Field Naturalists, Roosevelt Wild Life Forest Experiment Station,

Syracuse, New York

CONTENTS

PAGE

Introduction 242

Description of Localities 243

The Plants and Animals 244

General Factors 244

List of Biota 245

The Food of the Insects in Trout Streams 246

Food of the Perloidea or Stoneflies 246

Food of the Ephemeroida or Mayfly Nymphs 249

Food of the Trichoptera or Caddisflies 253

Food of the Diptera 260

Comparative Summary of the Food of Insects 261

List of References 262

* Professor of Biology and Head of the Department of Biology, University of Detroit, Detroit,
Michigan.

** Professor of Botany, Stanford University, Palo Alto, California.

[ 241 ]

242

Roosevelt Wild Life Annals

INTRODUCTION

The present collaboration is based on certain joint investigations of the authors
in Yellowstone National Park during the summer of 1921, under the direction
of the Roosevelt Wild Life Forest Experiment Station, of the New York State
College of Forestry at Syracuse. During that summer the senior author was
engaged in the study of the broader problem of the ecology of trout stream
organisms. In considering the trout stream as an ecological unit, it was necessary
to study the biota not merely as related to trout, hence as a potential food supply,
hut also as related to each other. In other words, what is the food supply of the
various animals inhabiting the trout streams?

For the study of this phase it was necessary to make a survey of the plant
life of the streams and the part it plays in the food of the animals. For obvi-
ously, in the last analysis, all life in trout streams, quite as much as elsewhere, is
dependent on the plant life. This phase was taken up more intensively after the
arrival of the junior author in Yellowstone Park in early August, when collec-
tions were made of the plants of different streams together with associated ani-
mals. Collections were made from different parts of the streams in order to
obtain as varied a series of selections as possible in these highly specialized habi-
tats. For instance, from Lost Creek samples of plant life were obtained from
the swiftest rapids, from the pools formed by the rapids, from relatively quiet
waters, from exposed and shaded spots, and also from the lateral pools.

Specimens were collected jointly from four streams typifying the various
kinds of trout streams found in Yellowstone Park, — indeed, found throughout
the Northwest. These are: 1. Yellowstone River, typical of the larger and per-
manent rapid streams ; 2. Lamar River, a good type of “variable” or fluctuating
stream, carrying enormous masses of water for short periods, followed by equally
short periods of rapid recession — on the whole, a stream with a relatively low.
equable flow; 3. Lost Creek and Tower Creek, both precipitous mountain streams
of rather short length, with many falls and rapids. Their type is very numerous
and form the tributaries of the larger streams. Tower Creek is the best repre-
sentative of this type ; but Lost Creek, a smaller stream, comprises similar con-
ditions and biota, and was selected because of its convenient location at Camp
Roosevelt, our field base.

In each case two comprehensive survey collections were made covering as many
different habitats as the stream seemed to offer. Besides these, a number of
isolated collections made previously by the senior author were available for com-
parison. The collected material was studied while still fresh. This applied
also to the stomach contents of the various insects.

In the following pages, the notes on animals are by the senior author, those
on plants mainly by the junior author.

Food of Trout Stream Insects

2 43

DESCRIPTION OF LOCALITIES

Yellowstone River, within Yellowstone National Park, has a continuous fall,
thereby increasing the force of its current enormously. Generally the current
attains a speed of eight to ten miles per hour, in places up to twenty miles. The
many rapids, short and long falls and the tortuous bed, make the stream un-
navigable within the Park. In width it varies from fifty to three hundred feet,
especially in the “eddies” where the waters are dammed somewhat by the inflow
of some tributary. The depth may be considerable, some of the “holes” attain-
ing thirty and even forty feet, and rarely less than six feet, except for a few
fords. The bed of the stream contains many huge boulders, granite rocks, and
smaller stones. Hardly anywhere along the Yellowstone does one find sand or
gravel except in the interstices between the huge boulders and lesser stones along
parts of shores that are less directly exposed to the current.

The shores of the Yellowstone are greatly varied, from volcanic rock to allu-
vial and glacial soil, the latter with the “sulphur slides” so characteristic of the
canyons of the Yellowstone. In the spring the melting snow carries great quan-
tities of this soil into the stream which then silts out as the period of high water
passes. Such slides may, however, occur at any time in the summer and the
river will then be turbid for a few hours or days, until the loose soil has settled
or is washed away.

The Lamar River varies from fifty to one hundred and fifty feet in width,
and at the ordinary stage is about two feet in depth. Its rapids are of a much
milder type than those of the Yellowstone, although during the flood period they
rage with tremendous force and appear formidable. In spring the melting snows
bring enormous masses of mud and sand with them, which silt out among the
shore rocks and boulders as the waters recede. — in fact, nearly covering the
rocks and thus giving little idea of the ferocity of the spring torrents. Yet each
year, with the spring floods, all these deposits are washed out, to be replaced
by others coming down from the mountains.

The bed of the stream is generally clear, composed of granite boulders, and
rocks usually not smaller than a cubic foot, but of smaller size at the fords.
The Lamar is essentially a shallow stream and fordable in numerous places.
Yet it is an excellent stream for trout and supports a rather luxuriant plant
and animal life.

The precipitous streams, such as Lost Creek and Tower Creek, because
of their great vertical fall, have few cobble stones or gravel or sand in their
beds, except at the margin of the pools. Tower Creek is of practically constant
width, about twenty to twenty-five feet, for at least seven miles above the falls,
the junction point with the Yellowstone. The depth varies from eighteen
inches to two feet. The bed is nearly uniformly composed of rounded cobbles
varying from pebble size to a cubic foot, with some huge boulders scattered
along the bed and shores of the stream.

Lost Creek is ideal for the study of the recession of waters. In flood periods
(late June) it is an energetic and precipitous torrent which impresses one by the

244

Roosevelt Wild Life Annals

amount and force of the descending waters. A few weeks later, in middle
August, the creek is a thin trickle carrying a flow of hardly more than a gallon
per second. Even this disappears entirely into the ground about five hundred
yards below the pretty Lost Creek Falls in a bed of mixed gravel and cobbles,
a short distance above Tower Fall Junction Ranger Station. Half a mile down
the stream reappears as a sump along the Cooke City road, leading down to
the Cooke City bridge across the Yellowstone River.

THE PLANTS AND ANIMALS

General Factors. In such a highly specialized habitat as mountain trout
streams both physiographical and physiological conditions act restrictively on the
biota. The constant fall of the water and the resulting tremendous force of the
current constitute the major factors in limiting the aquatic population. Secure
places for temporary attachment are few : shelters against the current are none
too many; there is little opportunity for swimming. Hence only plants and
animals with holdfasts, — -either natural or artificial, such as webs, claws, suckers
— or powerful swimmers can establish and maintain themselves. In addition,
the force of the water makes their resting places rather uncertain, since offal,
sticks, and stones are constantly being whirled along and strike against the
rocks. Even larger rocks may he moved by the current ; indeed, well into the
summer, weeks after the spring floods are over, one can hear the slow' rubbing
and grinding of the rocks in the stream beds.

On the other hand, the speed of the water, and the frequent rapids and
falls make for high oxygen content. Most of the fauna seem to require this.
If for any reason the oxygen is diminished, or the constant flow ceases, the
animals die.

Due to these influences, the fauna is surprisingly sparse in all of the
streams — that is, in number of species, but not in number of individuals of each
species. The specialized conditions limit the number of species ; but they also
tend to increase the numbers of individuals of a species vastly. Hence one finds
the endemic forms present in astonishing numbers.

The major portion of the fauna is composed of insects, to exceed 99%, —
that is, if fish he excluded. One is surprised at the almost complete absence
of Annelids, leeches, Crustacea, and Mollusca, and at the rarity of Protozoa and
Rotifera, all of which are so abundant under less violent aquatic conditions.

The plants are restricted to algae ; all higher plants are absent. Smaller
algae, diatoms and desmids find shelter in tiny crevices, in the films of emergent
shore rocks, or behind rocks where the current is not too strong.

Certain faunistic peculiarities of the streams may be noted. Thus, both
Tower Creek and Lost Creek are characterized by the abundance of planarians
and of the caddisworm Hydropsyche, and by the nearly total absence of Simulimn.
On the other hand, Simulimn is present in quantities in the Yellowstone and the
Lamar rivers, while planarians are exceedingly rare. The larger Pcrloidca are
also very characteristic for these two streams. In smaller and precipitous
streams they seem to be somewhat infrequent.

food of Trout Stream Insects 245

List of Biota.

Alyac and Diatoms.

Cladophora sp.

Oscillatoria sp

Closterinm sp.

Prasiola sp.

Cocconeis sp.

Rlwicospheniutn sp.

Li pit hernia sp.

Rhisoclonium sp.

Gomphonema sp.

Spirogyra sp.

M cl 0 sir a sp.

Synedra sp.

Nostoc sp.

( Moss — undetermined )

Animals.

Protozoa

Ameba sp.

Col pod a sp.

Gregarina sp. — as parasites in

Trichoptera, Perloidea, and Ephemeroidea

Rotifera

Several spp.

Nematoda

Merrnis (?) sp. — as parasites

in mayfly nymphs and in Simnlium.

Arthropoda

Perloidea.

Pteronarcys calif ornica

Doroneuria theodora

Pteronarcella badia

Isopcrla 5- punctata

Acroncuria pacifica

Perl a verticalis

Alloperla eoloradensis

Pcrlodes signata

Alloperla fidclis

Pteronarcella badia

Alloperla lincosa

Pteronarcys calif ornica

Alloperla sp.

Ephemeroidea

Ameletus sp.

Ephemerella eoloradensis

Bactis sp.

Ephemerella sp.

Callibaetis sp.

Hcptagenia sp.

Drunella grand is

Iron longimanus

Trichoptera

Brachycentrus sp.

Philopotamus sp.

Glossosoma sp.

Platyphylax sp.

Goera sp.

Rhyacophila torva

Hydropsychc sp.

Thremma sp.

Limnophilus sp.

Triaenodes sp.

Neophylax concinnus

Diptera

Ath erix varie ga ta

Orthocladius sp.

Bibiocephala comstocki

Philolutra simplex

Bibiocephala grandis

Procladius sp.

Chamaedipsis sp.

Psychoda sp.

Chironomns sp.

Rhamphoniyia sp.

Cricotopus varipcs

Simulium sp.

Metriocnemus sp.

Tanytarsus exignus

246

Roosevelt Wild Life Annals

THE FOOD OF THE INSECTS IN TROUT STREAMS

In the following tables the first numbers are the collection numbers. The
succeeding number refers to the individual insect whose stomach contents are
listed. The estimates are given in percentages. “Detritus” signifies unidenti-
fiable refuse. The figures in parenthesis are the numbers of individual items
of food.

The Food of Perloidea or Stoneflies. The table below gives in detail the
food percentage of the stoneflies as determined from the specimens examined.

Table No. i. — Showing the Food of the Perloidea or Stoneflies.

Collec-

tion

Num-

ber

Number and Name of
Individual Specimens

Date

1921

Locality

Habitat

5517

1. Pteronarcys californica . .

July s

Yellowstone R.

Rapids near shore.

5517

2. Pteronarcys californica. .

July s

Yellowstone R.

Rapids near shore.

5517

3. Pteronarcys californica . .

July 5

Yellowstone R.

Rapids near shore.

5 5 bb J

3. Acroneuria pacifica

Aug. 10

Lost Creek

Rapids of fair
violence.

5S 6 7a

1. Acroneuria pacifica

Aug. 1 1

Yellowstone R.

Strong rapids

5567a

2. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

5567a

3. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids

5567a

4. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

5567a

5. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

S567a

6. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

5567a

7. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

SS67a

8. Acroneuria pacifica

Aug. 11

Yellowstone R.

Strong rapids ....

5567a

9. Acroneuria pacifica

Aug. ti

Yellowstone R.

Strong rapids ....

5567a

10. Pteronarcys californica . .

Aug. 11

Y ellowstone R

Strong rapids ....

5567a

11. Pteronarcys californica . .

Aug. 11

Yellowstone R.

Strong rapids ....

5567a

12. Pteronarcys californica . .

Aug. 11

Yellowstone R.

Strong rapids ....

5572a

1 . Perla verticalis ?

Aug. 12

Lost Creek . . .

Strong rapids ....

5574*1

1. Acroneuria pacifica

Aug. 13

Lamar River. .

Feeding among

rocks in quiet
current.

5574*1

2. Acroneuria pacifica

Aug. 13

Lamar River. .

Feeding among

rocks in quiet
current.

5574*1

3. Acroneuria pacifica

Aug. 13

Lamar River .

Feeding among

rocks in quiet
current.

5574*1

4. Acroneuria pacifica

Aug. 13

Lamar River. .

Feeding among

rocks in quiet
current.

S 5 74*1

5. Acroneuria pacifica

Aug. 13

Lamar River. .

Feeding among

rocks in quiet
current

5574d

5575a

6. Acroneuria pacifica

1. Acroneuria pacifica

Aug. 13
Aug. 14

Lamar River. .
Tower Creek. .

Feeding among

rocks in quiet
current.

Mild rapids

5575a

2. Acroneuria pacifica

Aug. 14

Tower Creek. .

Mild rapids

5575a

3. Acroneuria pacifica

Aug. 14

Tower Creek . .

Mild rapids

5575a

4. Acroneuria pacifica

Aug. 14

Tower Creek. .

Mild rapids

5575a

5. Acroneuria pacifica

Aug. 14

Tower Creek .

Mild rapids

5585a

20. Acroneuria pacifica

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

5585a

21. Acroneuria pacifica

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

Food Items in Percentages

Intestines with wood fibers, 100.

Intestines with wood fibers. 1 00.

Plant matter, 100.

Heptagenia (2), 75; diatoms, 25.

Drunella, 25; Heptagenia, 25;
Rhyacophila, 45; wood frag-
ments, s.

Tanytarsus pupa, 100.

Gregarina.

Tanytarsus larvae. 99; wood
fragment, .5; pollen grain of
pine, .s.

Tanytarsus (5), 50; Hepta-
genia, 20; Rhyacophila, 28;
detritus, 2.

Tanytarsus (3), so; Ephemerel-
la, 49; Cladophora, 1.

Chironomus in tube, 100.

Empty.

Detritus, 100.

Tanytarsus, 100.

Tanytarsus (3), 5; Cladophora,
75; shore diatoms, 15; wood
and bark fragments. 5.

Small moth, 100.

Detritus, 100.

Ephemerella, 40; Heptagenia,
SO; Chironomus, 9; diatoms,
I.

Chironomus larvae, 1 00.
Gregarina.

Mayfly fragments. 100. Gre-
garina.

Tanytarsus (5), 20; mayfly
nymph. 50; insect fragments,
20; detritus, 10. Gregarina.

Tanytarsus pupae (2), 100.

Tanytarsus pupae (2), 90;

wood fragments, 10. Grt-
garina.

Bachycentrus (2), 100.

Heptagenia, 60; Ephemerella,
20; Perla, 15; detritus, 5.
Ameletus (4), 100.

Heptagenia, 60; Rhyacophila
(a), 40.

Chironomus pupa, 60; caddis-
worm, 40.

Chironomus (2), 40; Pro-

tenthes, 50; wood fragment,
10.

Perla nymphs (5), 35; Chir-
onomus pupa, 15; Melosira,
48; wood fragments, 2.
Insect fragments, is; Melosira,

8s.

Food of Trout Stream Insects

247

Table No. i. — Showing the Food of the Perloidea or Stoneflies. —

(Continued) .

Collec-

tion

Num-

ber

Number and Name of
Individual Specimens

Date

1921

Locality

Habitat

Food Items in Percentages

5585a

22. Acroneuria pacifica |

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

Chironomus, 5; Chironomus
pupa, 20; perla nymphs (12),
65; Melosira, 5; detritus, 5.

5585a

2 3. Acroneuria pacifica

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

Insect fragments, 100.

5585a

24. Acroneuria pacifica

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

Cocconeis, 10; Melosira, 5; dia-
toms, 5; detritus, 80.

5585a

25. Acroneuria pacifica

Aug. 26

Lost Creek . . .

In shade. No

vegetation.

Mayfly fragments, 30; Melo-
sira, 5; detritus, 65.

5585b

20. Perla (verticalis)

Aug. 25

Lost Creek . . .

Sunlit parts of
stream.

Empty.

5585b

21. Perla (verticalis)

Aug. 25

Lost Creek . . .

Sunlit parts of
stream.

Heptagenia, 60; Chironomus
pupa. 25; perla nymphs, 14;
melosira, 1.

5585b

22. Perla (verticalis)

Aug. 25

Lost Creek. . . .

Sunlit parts of
stream.

Perla nymphs (7), 99; Melo-
sira, 1.

5585 b

23. Perla (verticalis)

Aug. 25

Lost Creek . . .

Sunlit parts of
stream.

Perla nymphs (13), 100.

5585b

24. Perla (verticalis)

Aug. 25

Lost Creek . . .

Sunlit parts of
stream.

Chironomus pupa, 30; Melo-
sira, 50; said, 10; wood
fragments, 10; all in clearly
marked zones in stomach.

5589a

12. Pteronarcys californi a..

Aug. 26

Lamar River. .

Strong rapids

among moss and
Cladophora.

Moss, 60, bark, 35; diatoms, 5.

5589a

13. Pteronarcys californi. a. .

Aug. 26

Lamar River . .

Strong rapids

among moss and
Cladophora.

Moss, 80; Epithemia, 20.

5589a

14. Pteronarcys californi a..

Aug. 26

Lamar River. .

Strong rapids

among moss and
Cladophora.

Moss, 50; diatoms, 10; Epi-
themia, 40.

5589a

15. Pteronarcys californica . .

Aug. 26

Lamar River. .

Strong rapids

among moss and
Cladophora.

Moss, 85; diatoms, i; Epi-
themia, 14.

5589a

16. Pteronarcys californica. .

Aug. 26

Lamar River..

Strong rapids

among moss and
Cladophora.

Moss, 50; diatoms, 10; Epi-
themia, 40.

5589a

17. Pteronarcys californica. .

Aug. 26

Lamar River. .

Strong rapids

among moss and
Cladophora.

Epithemia, 100. Small amount.

5589a

1 Acroneuria pacifica

Aug. 26

Lamar River. .

Strong rapids

among moss and
Cladophora.

Ephemerella nymph, 99; moss,

1.

5589a

19. Acroneuria pacifica

Aug. 26

Lamar River . .

Strong rapids

among moss and
Cladophora.

Insect fragments, 100.

5589b

8. Acroneuria pacifica

Aug. 26

Lamar River . .

Minor rapids,

among decaying
Cladophora.

Chironomus larvae (16), 40;
Perla, 40; sand, 20.

5589b

9. Acroneuria pacifica

Aug. 26

Lamar River. .

Minor rapids,

among decaying
Cladophora.

Chironomus larvae (19), 95;
sand, 5.

5589b

10. Acroneuria pacifica

Aug. 26

Lamar River. .

Minor rapids,

among decaying
Cladophora.

Chironomus larvae (16), 90;
Epithemia, 5; diatoms, 5.

5589b

1 1. Acroneuria pacifica

Aug. 26

Lamar River. .

Minor rapids,

among decaying
Cladophora.

Chironomus larvae (6), 40;
Perla nymphs, 50; insect
fragments, 10.

5589b

12. Acroneuria pacifica

Aug. 26

Lamar River. .

Minor rapids,

among decaying
Cladophora.

Perla numphs, 45; mayfly
nymph fragments, 50; sand,

S-

5589b

13. Acroneuria pacifica

Aug. 26

Lamar River. .

Minor rapids,

among decaying
Cladophora.

Chironomus larvae (2), 100.

559 oa

1. Acroneuria pacifica

Aug. 26

Yellowstone R.

Moderate rapids

Chironomus larvae (4), 40;
Rhyacophila, 50; sand, 10.
Gregarina.

5 5 90 a

2. Acroneuria pacifica

Aug. 26

Yellowstone R.

Moderate rapids. .

Insect fragments, ioo. Ore-

5500a

3. Acroneuria pacifica

Aug. 26

Yellowstone R.

Moderate rapids. .

Rhyacophila, ioo.

5590 a

4. Pteronarcys californica . .

Aug. 26

Yellowstone R.

Moderate rapids..

Detritus, 100.

5590 a

5. Pteronarcys californica. .

Aug. 26

Yellowstone R.

Moderate rapids..

Detritus, ioo.

5590 a

6. Pteronarcys californica . .

Aug. 26

Yellowstone R.

Moderate rapids. .

Detritus, 100.

5590 a

7. Pteronarcys californica. .

Aug. 26

Yellowstone R.

Moderate rapids. .

Detritus, ioo.

5590 a

8. Pteronarcys californica. .

Aug. 26

Yellowstone R.

Moderate rapids..

Trebonema, 5; moss fragments,
10; detritus, 85.

55903

9. Pteronarcys californica . .

Aug. 26

Yellowstone R.

Moderate rapids..

Trebonema, 15; moss frag-
ments, 5; diatoms, 10;
detritus, 70.

2 4 8

Roosevelt Wild Life Annals

Table No. i. — Showing the Food of the Perloidea or Stoneflies. —

( Concluded ).

Collec-

tion

Num-

ber

Number and Name of
Individual Specimens

Date

1921

Locality

Habitat

Food Items in Percentages

559oa

10. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Moderate rapids. .

Bark fragments, 85; detritus.

5590a

11. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Moderate rapids..

*5-

Trebonema, 3; detritus, 97.

5590b

1 . Pteronarcys californica . .

Aug.

26

Yellowstone R.

Moderate rapids. .

Wood fragments, 85; detritus.

5590b

2. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Moderate rapids.

!5-

Detritus, ioo.

5590b

3. Pteronarcys californica . .

Aug.

26

Yellowstone R.

Moderate rapids. .

Sand, s; diatoms, 5; detritus.

5 5 90b

4. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Moderate rapids..

90.

Wood fragments, 40; diatoms,
10; Trebonema, 5; detritus,

5 59ob

5. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Moderate rapids. .

45*

Wood fragments, 3; diatoms, 5;
Trebonema, 2; detritus, 90.

5590b

6. Acroneuria pacifica

Aug.

26

Yellowstone R.

Moderate rapids. .

Insect fragments, 95; diatoms.

559ob

7. Acroneuria pacifica

Aug.

26

Yellowstone R.

Moderate rapids..

5-

Rhyacophila pupa (2), 100.

5590b

8. Acroneuria pacifica

Aug.

26

Yellowstone R.

Moderate rapids..

Chironomus (2), 90; sand. 10.

5590b

9. Acroneuria pacifica

Aug.

26

Yellowstone R.

Moderate rapids..

Empty.

5590C

1. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Strong, violent

rapids.

Bark, 95; detritus. 5.

5590C

2. Pteronarcys californica . .

Aug.

26

Yellowstone R.

Strong, violent

rapids.

Bark, 25; detritus, 75.

5590C

3. Pteronarcys californica. .

Aug.

26

Yellowstone R.

Strong, violent

rapids.

Bark, 50; detritus, 50.

5590C

4. Pteronarcys californica . .

Aug.

26

Yellowstone R.

Strong, violent

rapids.

Empty.

559ia

7. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Mavfly fragments, ioo.

559ia

8. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Detritus, 100.

559la

9. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Empty.

5591a

10. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Detritus, 100.

559ia

1 1 . Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Mayfly fragments, 95; detritus.

559ia

12. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

5*

Moth scales, 50; detritus. 50.

5591a

13. Acroneuria pacifica

Aug.

27

Tower Creek. .

Minor rapids

Chironomus larvae, 90; detri-
tus, 10.

559lb

6. Acroneuria pacifica

Aug.

27

Tower Creek. .

Strong rapids ....

Digested matter, ioo.

559lb

7. Acroneuria pacifica

Aug.

27

Tower Creek. .

Strong rapids ....

Detritus, 100.

S59lb

8. Alloperla sp

Aug.

27

Tower Creek. .

Strong rapids ....

Digested matter (plants?), 100.

Table No. 2. — Showing Summary of Food of Perloidea. (Empty Stomachs
are not Included in Computing the Averages).

Name

Locality

Number

of

Specimens

Animal

Food

Plant

Food

Detritus

Pteronarcys

Yellowstone River

20

s

43

90

53 85

I

52

10

Pteronarcys

Lamar River

6

Average for

Acroneuria

Yellowstone River

26 =

14

14

3.85

90

95 5

50

60

42.3

9

Acroneuria

Acroneuria

Lamar River

Lost Creek

1 . 5
30
5

63

15

3

20

Acroneuria

Tower Creek

11

3 S

Average for

Perl a verticalis (?). . .

= 77-4

85

16.3

Food of Trout Stream Insects

249

Undoubtedly, the stoneflies are the dominant insect forms of the mountain
trout streams, particularly in Yellowstone and Lamar rivers. Previous to their
final eedysis, about the time the spring floods abate, the nymphs are extremely
abundant and appear to constitute the hulk of the insect fauna. The stomachs
of trout taken from the stream at this period are largely filled with stonefly
nymphs.

But where are all these nymphs to obtain their food? They are carni-
vores, one generally reads. But if their numbers are so great that their hulk
exceeds the available animal food supply, how can they subsist? This was
one of the puzzles the senior author met early in the work, — one which was
not solved until the present data were obtained.

From the examination of the stomach contents of Perlid nymphs several
surprising results were obtained : ( I ) That Perloidea are not exclusively

carnivores, but that their diet contains an admixture of plant matter and
detritus, i. e., predigested and decomposed matter. (2) That the largest species,
Ptcronarcys calif ornica, is largely a vegetarian. (3) That about 12% of those
Perloidea whose main diet was animal matter, were parasitized by gregarines.

The diet of Ptcronarcys was perhaps the most surprising result. It’s first
notice was so unexpected that thereafter special efforts were made to secure a
considerable number of specimens from different localities for examination.
W hether this species is really a vegetarian, or whether it is so only on occasion,
.cannot be stated positively. Only this much seems clear: The specimens

examined, about thirty in number, had fed chiefly on a plant diet, averaging
less than 4% animal diet.

The other prominent perlid species, Acroneuria pacifica, is quite evidently
a carnivore. The admixture of plant food and detritus may be accounted for
on the basis that they were taken in with the insect food. It is also possible,
that the gregarines found in six of the Acroneuria were not “resident” parasites,
but were taken in with some insect hosts, such as mayfly nymphs or caddisworms.

In a very interesting study of Nemoura, C. F. Wu (’23, p. 39) remarks
as follows: “Besides some fine sediment and the half digested fragments of

decaying leaves, there were found great varieties of unicellular algae, chiefly
diatoms and desmids. No remains of animal tissue have ever been detected,
so that naiads are herbivorous in their food habits.” Relative to the food of
the adult Wu ( 1 . c.) states, “Of the various kinds of living plant leaves found
around the water and fed to the adults, the young leaves of Touch-me-not are
eaten.” Newcomer (T8) reports the adults of Taeniopteryx pacifica as feeding-
on the buds and leaves of plants and causing considerable injury.

Food of Ephemeroidea or Mayfly Nymphs. The following table shows
in detail the food percentages of the mayfly nymphs examined.

250

Roosevelt Wild Life Annals

Table No. 3. — Showing Food of Mayfly Nymphs.

Collec-

tion

Num-

ber

Number and Name of
Individual Specimens

Date,

1921

Locality

55Mb

3. Ameletus

Aug. 10

Lost Creek ....

5566b

I. Drunella sp

Aug. 10

Lost Creek ....

5566d

1. Drunella sp

Aug. 10

Lost Creek ....

5566d

2. Drunella sp

Aug. 10

Lost Creek ....

5Sb6d

5. Ephemerella sp. . .

Aug. 10

Lost Creek

5566(1

6. Ephemerella sp. . .

Aug. 10

Lost Creek ....

5S66j

I. Heptagenia sp . . . .

Aug. 10

Lost Creek ....

SS f >6h

4. Ephemerella sp. . .

Aug. 10

Lost Creek ....

55 b 6 i

i . Ephemerella sp . . .

Aug. 10

Lost Creek ....

5566i

2. Ephemerella sp. . .

Aug. 10

Lost Creek ....

5566 j

2. Heptagenia sp . . .

Aug. 10

Lost Creek ....

5566)

7. Ephemerella sp. .

Aug. 10

Lost Creek ....

ssftftj

8. Ephemerella sp. .

Aug. 10

Lost Creek ....

5566k

1. Heptagenia sp . .

Aug. 10

Lost Creek ....

5566 k

2. Ameletus sp

Aug. 10

Lost Creek ....

55661

1 . Ephemerella sp . . .

Aug. 10

Lost Creek ....

55661

2. Ephemerella sp. .

Aug. 10

Lost Creek ....

5567a

IS- Ameletus sp

Aug. 11

Yellowstone R

S567a

16-19. Ameletus sp. . .

Aug. 11

Yellow'stone R

5567a

20. Ameletus sp

Aug. 1 1

Yellowstone R

5567a

21, 22. Ameletus sp . .

Aug. 1 1

Yellowstone R

5567a

23. Heptagenia sp

Aug. 11

Yellowstone R .

5567a

24. Heptagenia sp . . .

Aug. 11

Yellowstone R .

5567a

25. Heptagenia sp.

Aug. 11

Yellow'stone R .

5567a

26. Heptagenia sp . . .

Aug. 11

Yellowstone R .

5567a

27. Heptagenia sp. . . .

Aug. 11

Yellowstone R

5567a

28. Heptagenia sp . . .

Aug. 11

Yellowstone R .

5567a

29. Ephemerella sp . .

Aug. 11

Yellowstone R.

5567a

30. 51. Ephemerella sp.

Aug. 11

Yellowstone R .

5567a

32. Ephemerella sp. . .

Aug. 11

Yellowstone R .

5567a

33, 34. Ephemerella sp.

Aug. 11

Yellow'stone R

5572a

4. Ameletus sp

Aug. 12

Lost Creek ....

5572a

5, 6. Ameletus sp . .

Aug. 12

Lost Creek ....

5572a

7. Heptagenia sp . .

Aug. 12

Lost Creek ....

5572b

5. Ameletus sp

Aug. 12

Lost Creek ....

5 572b

6. Ameletus sp

Aug. 12

Lost Creek ....

5572b

7. Ameletus sp

Aug. 12

Lost Creek ....

5572b

8. Ephemerella sp

Aug. 12

Lost Creek ....

5572b

Q. Ephemerella sp . .

Aug. 12

Lost Creek ....

Habitat

From rocks with clusters of
vegetation, chiefly Melo-
sira and some Gompho-
nema and Closterium.

Pure Melosira

From Prasiola and Oscil-
latoria.

From Prasiola and Oscil-
latoria.

From Prasiola and Oscil-
latoria.

From Prasiola and Oscil-
latoria.

From Prasiola

From barren region

Barren region w'ith few
diatom snells.

Barren region with few
diatom shells.

Barren, with few' diatoms.

Barren, w'ith few diatoms.
Barren, with few diatoms.

Barren with few' diatoms,
some Closterium.

Barren w'ith few' diatoms,
some Closterium.

Mostly Oscillatoria.

Traces of Melosira.
Mostly Oscillatoria.

Traces of Melosira.

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Strong rapids

Stroi g rapids

Strong rapids

Strong rapids

Rapids. Vegetation: Melo-
sira, Prasiola, Oscilla-
toria, Calpoda and Roti-
fera among plants.
Rapids. Vegetation: Melo-
sira, Prasiola, Oscilla-
toria, Calpoda and Roti-
fera among plants.
Rapids. Vegetation: Melo-
sira, Prasiola, Oscilla-
toria, Calpoda and Roti-
fera among plants.
Among rocks in rapids.
Much detritus and

Melosira.

Among rocks in rapids.
Much detritus and

Melosira.

Among rocks in rapids.
Much detritus and

Melosira.

Among rocks in rapids.
Much detritus and

Melosira.

Among rocks in rapids.
Much detritus and

Melosira.

Food Items in Percentages

Melosira, 90; Gompnonema,
10.

Melosira, 99; Gomphonema,
•5; protozoan. .5.

Melosira, 99; Gomphonema,

1.

Melosira, 80; Closterium, 1;
Oscillatoria, 19.

Prasiola, 99; Oscillatoria, 1.

Melosira, 50; empty diatom
shells, 50.

Prasiola, 100.

Melosira. 60; Oscillatoria, 40.

Empty.

Melosira, 75; mixed diatoms,
2 5*

Melosira, 95; mixed diatoms,
5.

Melosira, 50; Oscillatoria,
45; diatoms. 5.

Prasiola, 99; Oscillatoria. 1.

Gomphonema, 85; diatoms,
I 5-

Melosira, 85; diatoms ard
Closterium, 15.

Melosira, 100.

Oscillatoria, 99; diatoms, 1.

Oscillatoria, 99; diatoms, 1.

Sand, 95; diatoms, 5.

Detritus, 100.

Fresh diatoms, 15; sana, 85.

Botn witn detritus, 100.

Detritus, 100.

Sand, 50; diatoms, 5; detri-
tus, 45.

Detritus, 100.

Sand. 15; diatoms, 5; detri-
tus, 80.

Sand, 5; diatoms, 5; detri-
tus, 90.

Diatoms, 5; detritus, 95.

Cladophora. 5; sand, 10;
detritus, 85.

Botn with detritus, 100.

Detritus, 75; adult dipteran,

25.

Both with detritus, 100.

Melosira, 50; detritus. 50.

Both with detritus, 100.

Rock diatoms, 15; sand, 5;
detritus, 80.

Melosira, 50; detritus, 50.

Detritus, 50; sand, 25; rock
diatoms, 25.

Rock diatoms, 75*. detritus,
20; sand, 5.

Rock diatoms, 85; detritus,
5; diatoms, 10.

Rock diatoms, 75; detritus.
23; sand. 2.

Food of Trout Stream Insects

251

Table No. 3. — Showing Food of Mayfly Nymphs. — (Continued) .

Collec-

tion

Num-

ber

Number and Name of
Individual Specimens

Date,

1921

Locality

Habitat

Food Items in Percentages

5572b

10. Drunella sp

Aug. 12

Lost Creek ....

Among rocks in rapids.
Much detritus and

Melosira.

Wood fibers, 35; detritus, 65.

5572e

1. Heptagenia sp

Aug. 12

Lost Creek ...

From rocks in rapids.
Moss and Prasiola.
Melosira caught in moss.

Detritus, 95; wood frag-
ments, 2; mixed diatoms,
3-

557 2 e

2, 3. 4. Heptagen a
sp.

Aug. 12

Lost Cee'c

From rocks in rapids.
Moss and Prasiola.
Melosira caught in moss.

All with de r.tus, ioo.

557 2 e

5. Heptagenia sp. . . .

Aug. 12

Lost Creek ....

From rocks in rapids.
Moss and Prasiola.
Melosira caught in moss.

Detritus, 95; mixed diatoms,
5.

i. Baetis sp

Aug. 14

Tower Creek .

Mild rapids

Sand, 20; detritus, 80.
Detritus, ioo.

2. Baetis sp

Aug. 14

Tower Creek .

Mild rapids

5575c

1 . Ephemerella sp . . .

Aug. 14

Tower Creek . . .

From lateral rapids

Sand, 20; detritus, 80.

5575c

2. Ephemerella sp. . .

Aug. 14

Tower Creek . . .

From lateral rapids

Sand, 10; detritus, 50; wood
fragments, 15; insect frag-
ments, 25.

5575c

5575d

3, 4. Ephemerella sp.

Aug. 14

Tower Creek

From lateral rapids

Detritus, ioo.

i . Ameletus sp

Aug. 14

Tower Creek . . .

Mild rapids

Diatoms, 1 ; wood fragments,
4; detritus, 75; sand, 20.

5575e

1. Ephemerella sp. . .

Aug. 14

Tower Creek . . .

From lesser rapids

Heptagenia, 50; detritus, 49;
wood, 1.

5584a

15. Heptagenia sp. . . .

Aug. 25

Lost Creek ....

Shaded area. Xo vegeta-
tion.

Detritus. 30; Cocconeis. 65;
wood fragments, 2; mixed
fragments, 3.

5584a

16. Heptagenia sp

Aug. 25

Lost Creek ....

Shaded area. Xo vegeta-
tion.

Cocconeis, 25; Rhoicos-
phenia, 25; mixed diatoms,
5; detritus, 40; sand, 5.

5584a

17. Drunella sp

Aug. 25

Lost Creek ....

Shaded areas. No vegeta-
tion.

Cocconeis, 40; Rhoicos-
phenia, 5; mixed diatoms,
5; detritus. 50.

5584a

18. Drunella sp

Aug. 25

Lost Creek ....

Shaded areas. Xo vegeta-
tion.

Cocconeis, 35; Melosira, 15;
Synedra, 3; mixed dia-
toms, 7; sand. 5; detritus.

5584a

19. Drunella sp

Aug. 25

Lost Creek ....

Shaded areas. X'o vegeta-
tion.

35-

Sand, 25; Cocconeis, 25;
Melosira, 5; mixed dia-
toms, 5; detritus, 40.

5589a

20. Drunella sp

Aug. 26

Lamar River . . .

Strong rapids. Feeding on
moss and Cladophora.

Epithemia, 80; Cocconeis,.

10; mixed diatoms, 10.
Epithemia, 50; detritus, 50.

5589a

21. Drunella sp

Aug. 26

Lamar River . . .

Strong rapids. Feeding on
moss and Cladophora.

5589b

14. Ephemerella sp . .

Aug. 26

Lamar River . .

Minor rapids. Feeding

among decaying Clado-
phora.

Chironomus (2), 5; Epi-

themia, 50; mixed diatoms,
10; detritus, 35.

Sand, 5; Epithemia, 50;
mixed diatoms, 10; detri-
tus, 35-

5589b

15. Ephemerella sp . . .

Aug. 26

Lamar River . .

Minor rapids. Feeding

among decaying Clado-
phora.

5589b

16. Ephemerella sp . .

Aug. 26

Lamar River. . .

Minor rapids. Feeding

among decaying Clado-
phora.

Sand, 5; Epithemia, 15; dia-
toms, 10; detritus, 70.

5589b

17. Ephemerella sp. . .

Aug. 26

Lamar River. . .

Minor rap ds. Feeding

among decaying Clado-
phora.

M:x d d atoms, 5; detritus,.
95-

5589b

18. Ameletus sp

Aug. 26

Lamar River . . .

Minor rapids. Feeding

among decaying Clado-
phora.

Epithemia, 75; diatoms, 10;
detritus, 15.

5589b

19, 20. Ameletus sp. .

Aug. 26

Lamar River . . .

Minor rapids. Feeding

among decaying Clado-
phora.

Both with Epithemia, 10;
diatoms, 15; detritus, 75..

5590a

13. Heptagenia sp. .. .

Aug. 26

Yellowstone R .

Moderate rapids

Cladophora, 90; sand, 10.

5590a

14. Heptagenia sp . . .

Aug. 26

Yellowstone R .

Moderate rapids

Detritus, 100.

5590b

9. Ephemerella sp . . .

Aug. 26

Y ellowstone R .

Minor rapids

Detritus, ioo.

5590b

10. Ephemerella sp. . .

Aug. 26

Yellowstone R .

Minor rapids

Detritus, 85; diatoms, 5;
Trebonema, 10.

5590b

11. Ephemerella sp. . .

Aug. 26

Yellowstone R .

Minor rapids

Detritus, 90; diatoms, 5;
Trebonema, 5.

559ia

14. Heptagenia sp . . . .

Aug. 27

Tower Creek . .

Minor rapids

Detritus, 100.

559ia

15. Heptagenia sp. . . .

Aug. 27

Tower Creek . . .

Minor rapids

Detritus, 99; diatoms, 1.

559ia

16. 17, 18. Heptagenia

Aug. 27

Tower Creek . . .

Minor rapids

All with detritus, ioo.

559ia

sp.

19. Ameletus sp

Aug. 27

Tower Creek . .

Minor rapids

Sand, 15; detritus, 85.

559ia

20. Ameletus sp

Aug. 27

Tower Creek . . .

Minor rapids

Sand, 18; detritus, 80; dia-
toms, 2.

559ia

21. Ephemerella sp. . .

Aug. 27

Tower Creek . . .

Minor rapids

Sand, 1; detritus, 97; dia-
toms, 2.

559ia

22. Ephemerella sp . . .

Aug. 27

Tower Creek. . .

Minor rapids

Sand, 1; detritus, 97; dia-
toms, 2.

559ia

23. Drunella sp

Aug. 27

Tower Creek . . .

Minor rapids

Insect fragments, 50; wood
fragments, 45; detritus, 5..
Gregarina as parasites.

252 Roosevelt Wild Life Annals

Table No. 3. — Showing Food of Mayfly Nymphs. — (Concluded) .

Collec-

tion

Num-

ber

Number and Name 0*
Individual Specimens

Date,

1921

Locality

Habitat

Food Items in Percentages

5591a

5591a

24. Drunella sp

25. Drunella sp

Aug. 27
Aug. 27

Tower Creek . . .
Tower Creek . . .

Minor rapids

Minor rapids

Wood fragments, 10; detri-
tus. 90. Gregarina.
Insect fragments. 80; dia-
toms, 2; detritus, 18.
Gregarina.

5591a

5591a

26. Drunella sp

27. Drunella sp

Aug. 27
Aug. 27

Tower Creek . . .
Tower Creek . . .

Minor rapids

Minor rapids

Detritus, 100. Gregarina.
Insect fragments, 5; wood
fragments, 10; sand, 4;
diatoms, 1 ; detritus, 80.

5591a

28. Heptagenia sp . .

Aug. 27

Tower Creek . . .

Minor rapids

Diatoms, 20; detritus, 70;
sand, 10. With nematode
(Mermis sp.?) as parasite.

5 S 9 ia

29. Heptagenia sp . .

Aug. 27

Tower Creek. . .

Minor rapids

Diatoms. 20; detritus, 70;
sand, 10. With two nema-
tode parasites (Mermis
sp.?).

Detritus, 100.

559 ib

9 - Ameletus sp

Aug. 27

Tower Creek. .

Strong rapids

559 ib

10. Ameletus sp

Aug. 27

Tower Creek .

Strong rapids

Detritus, 95; diatoms, 5.

559 ib

1 1 . Ameletus sp

Aug. 27

Tower Creek .

Strong rapids

Detritus, 94; sand, 5; dia-
toms. 1.

5591 b

12. Drunella sp

Aug. 27

'Power Creek . .

Strong rapids

Insect fragments, 75; sand.
5; detritus, 20. With

gregarina.

559 lb

559 lb

13. Drunella sp

14. Drunella sp

Aug. 27
Aug. 27

Tower Creek . . .
Tower Creek . .

Strong rapids

Strong rapids

Insect fragments. 10; sand,
5; detritus. 85. Gregarina.
Bark fragments. 20; sand, 5;
detritus, 75- Gregarina.

559 lb

15. Drunella sp

Aug. 27

Tower Creek . . .

Strong rapids

Diatoms, 2; sand, 23; detri-
tus, 75 - Gregarina.

559 lb

16. Heptagenia sp. . .

Aug. 27

Tower Creek

Strong rapids

Detritus. 100.

559 lb

17. Heptagenia sp .

Aug. 27

T ower Creek . .

Strong rapids

Diatoms, 5; sand, 5; detritus.

559 lb

18. Heptagenia sp. .

Aug. 27

Tower Creek. .

Strong rapids

90.

Sand. 5; detritus, 95 -

559 ie

2. Heptagenia sp . . . .

Aug. 27

Tower Creek. . .

Lateral pools. With mod-
erate current.

Detritus, ioo.

55919

3, 4. Heptagenia sp.

Aug. 27

Tower Creek. . .

Lateral pools. With mod-
erate current.

Both with detritus, 100.

559 ie

5. Drunella sp

Aug. 27

Tower Creek . . .

Lateral pools. With mod-
erate current.

Insect fragments, 100. Gre-
garina.

559 ie

6 . Drunella sp

Aug. 27

Tower Creek . .

Lateral pools. With mod-
erate current.

Insect fragments, 50; wood
fragments, 20; detritus,
30. Gregarina.

Food of Trout Stream Insects

Table No. 4. — Showing Summary of Food of Mayfly Nymphs.

253

Name

Locality

Number

of

Specimens

Animal

Food

Plant

Food

Detritus

Yellowstone R.
Lamar River . .
Lost Creek. . .
Tower Creek. .

Ameletus sp

Ameletus sp

Ameletus sp

Ameletus sp

Average for . .

Drunella sp

Drunella sp

Drunella sp

Average for

Baetis sp Tower Creek

Lamar River.
Lost Creek. .
Tower Creek.

Ephemerella sp . . . .

Ephemerella sp

Ephemerella sp

Ephemerella sp . . . .
Average for

Heptagenia sp

Heptagenia sp

Heptagenia sp

Average for ....

Yellowstone R .
Lamar River . .
Lost Creek . . . .
Tower Creek. .

Yellowstone R .
Lost Creek. . .
Tower Creek. .

8

3

8

6

25

2

7

11

20

9

4

8

7
28

8
13

13

34

33-6

18.5

2.6

125

11

3-7

2

45

50

3

22 . 7

75

68

10

36.8
50

3-4

37

96.25

10

36.3

14

50

4

24

98

55

50

97

77 -3

25

32

56.4

44-7

50

94

6i.75

3-75

79

60

86

50

96

76

From the predominance of detritus in the food of mayfly nymphs, as
indicated by the averages, it would seem that these nymphs are primarily
scavengers. They appear to feed on the flotsam and jetsam that is caught
between rocks, on diatoms, and the bits of filamentous algae that grow on
rocks. The animal matter eaten may possibly be dead specimens caught with
the flotsam.

Of Drunella ten specimens, and these all from Tower Creek, were parasit-
ized by gregarines. Two Heptagenia are noted as parasitized by a nematode,
perhaps a species of Mermis; these, too, were taken from Tower Creek. In
another paper (Muttkowski, ’25, Fig. 124, facing p. 485) is shown an adult
mayfly with such a nematode emerging from the caudal end. This adult was
taken from Gardiner River, at its junction with Lava Creek, a short distance
from Mammoth Hot Springs.

Whether these records indicate that only Drunella is parasitized by
Grcgarina, and Heptagenia by a nematode parasite, is conjectural. The records
are too few to permit any definite conclusions.

The proportions of plant and animal matter and of unidentifiable detritus
accord well with the findings of Needham (’20) for midwestern lakes and
streams.

The Food of Trichoptera. The following table shows in detail the food
percentages of the specimens of caddisworms studied.

254

Roosevelt Wild Life Annals
Table No. 5. — Showing Food of Trichoptera.

Collec-
tion
Num- 1
ber

Specimen Number
and Name

Date,

1921

Locality

5566a

1. Rhyacophila

Aug. 10

Lost Creek ....

5566h

1. Thremma

Aug. 10

Lost Creek ....

5566d

3. Thremma

Aug. 10

Lost Creek ....

5566d

1. Hydropsyche

Aug. 10

Lost Creek ....

5566d

2 . Hydropsyche

Aug. 10

Lost Creek ....

5S66d

3. Hydropsyche

Aug. 10

Lost Creek ....

SS66d

5. Thremma pupa.

Aug. 10

Lost Creek ....

5566i

3. Hydropsyche

Aug. 10

Lost Creek ....

556ti

4* Hydropsyche

Aug. 10

Lost Creek ....

5566i

5. Rhyacophila

Aug. 10

Lost Creek ....

S566j

1. Rhyacophila

Aug. 10

Lost Creek ....

5566j

2. Hydropsyche

Aug. 10

Lost Creek ....

5566j

6. Rhyacophila

Aug. 10

Lost Creek ....

55661

3. Rhyacophila

Aug. 10

Lost Creek ....

5 5 661

4. Rhyacophila

Aug. 10

Lost Creek ....

55661

5. Hydropsyche

Aug. 10

Lost Creek ■ • • ■

5587a

35. Hydropsyche

Aug. 11

Yellowstone R .

5587a

36. Hydropsyche

Aug. 11

Y ellowstone R .

5587a

40. Brachycentrus . . .

Aug. 11

Yellowstone R.

5587a

48. Hvdropsyche

Aug. 11

Yellowstone R.

5587a

49. Philopotamus? . . .

Aug. 11

Yellowstone R .

5587a

50. Thremma

Aug. 11

Yellowstone R .

5572a

2. Hydropsyche

Aug. 12

Lost Creek. . . .

5572b

1. Hydropsyche

Aug. 12

Lost Creek ....

5572b

2. Hydropsyche

Aug. 12

Lost Creek ....

5572b

3. Hydropsyche

Aug. 12

Lost Creek. . . .

5572b

4. Hydropsyche

Aug. 12

Lost Creek ....

5574a

1. Brachycentrus....

Aug. 13

Lamar River . . .

Habitat

Food in Percentages

Rocks splashed with spray
Vegetation consisting of
Melosira, chiefly. Some
Gomphonema and Clos-
terium.

From Prasiola and Oscilla-
toria.

Rotifera, 60; Oscillatoria, 40.
Melosira, 99; Colpoda, 1.

Melosira, ioo.

From shady area

From shady area

From shady area
F rom shady area
Prasiola and diatoms

Empty.

Spirogyra, 50; empty diatom
shells, 50.

Melosira, 80; diatom shells,
17; Closterium, 3.

Diatom shells, 100.

Empty.

Melosira, 95; mixed diatoms.

Prasiola and diatoms

5-

Melosira, 65; mixed diatoms.

Prasiola and diatoms

Barrens. Few diatom

shells.

Barrens. Few diatom

shells.

Barrens. Few diatom

shells.

Oscillatoria with trace of
Melosira.

Oscillatoria with trace of
Melosira.

Oscillatoria with trace of
Melosira.

Rapids

Rapids

Rapids

Rapids

Rapids

Rapids

Rapids. Oscillatoria, Pra-
siola, and Melosira.
Colpoda and rotifers
feeding among plants.

Rocks in rapids. Accumu-
lations of detritus and
Melosira. Ameba and
Colpoda feeding.

Rocks in rapids. Accumu-
lations of detritus and
Melosira. Ameba and
Colpoda feeding.

Rocks in rapids. Accumu-
lations of detritus and
Melosira. Ameba and
Colpoda feeding.

Rocks in rapids. Accumu-
lations of detritus and
Melosira. Ameba and
Colpoda feeding.

Rapids

* 5 :>*

Diatom shells, 100. Gregar-
ine parasites.

Rotifers, 1; chiror.omids. 4;
Melosira, 40; diatoms, 10;
Oscillatoria, 45.

Mayfly, 1; ad. Chironomus.
90 ; Melosira, 5 ; diatoms, 4.

Mayfly nymph. 30; Chirono-
mus ad., 60; Melosira. 5;
diatoms, 5; Gregarina.

Oscillatoria, 95; diatoms, 5.

Caddisworm, 30; Chirono-
mus ad., 30; diatoms, 10;
Oscillatoria, 30.

Mayfly nymphs (5). 30;

Chironomus larvae (10),
50; Melosira, 15; Oscilla-
toria. 5.

Chironomus. 10; stem paren-
chyma. 15; diatoms (Melo-
sira chiefly), 15; detritus,
60.

Mayfly nymph. 95; sand. 5 -

Wood fragments, 50; dia-
toms, 15; detritus, 35.

Tanypus pupa, 40; shore
diatoms, 20; lake plank-
ton. 5; Rhizoclonium, 5;
detritus, 30.

Empty.

Shore diatoms. 50; Triton-
ema, 5; detritus. 45.

Ameletus, 25; Ephemerella,
35; Melosira, 40.

Glossosoma (?) pupae (3),
40; Glossosoma (?) larvae,
10; Ephemerella, is:Chir-
onomus adult, 25; Melo-
sira, 20.

Ephemerella (5). 50; Glosso-
soma (?), pupa. 35; Melo-
sira, 10; detritus, 5.
Gregarina.

Chironomus (2), 20; caddis-
flies (6), 65; Melosira, 14;
sand and detritus. 1.

Chironomus (2). 50; Chirono-
mus adult, 45; Melosira,
3; detritus, 2. Gregar.r.a.

Caddisflies (7), 75; Acro-
neuria sp. leg, 7; wood
fragment, 10; trichop-
terous pupa, 5; Melosira,
3. Gregarina

Trebonema, 20: Cymbella,
40; Cocconeis, 20; Syne-
dra, 10; diatoms. 10; Gre-
garina.

255

Food of Trout Stream Insects

Table No. 5. — Showing Food of Triciioptera. — ( Continued ).

Collec-

tion

Xu Ti-
ber

Specimen X'umber
and Name

Date,

1921

Locality

Habitat

Food in Percentages

5574 a

2. Brachycentrus. . . .

Aug. 13

Lamar River. . .

Rapids

Tanytarsus, i . ;Nauplii (?) .
30; wood fibe u 50. Gre-
garina.

5574a

3. Brachycentrus....

Aug. 13

Lamar River . . .

Rapids

Chironomus, 5; mayfly

nymph, 20; Cocoo. eis, 50;
Synedra, 20; minor dia-
toms, 5. Gregarina.

5574a

4. Brachycentrus. . . .

Aug. 13

Lamar River. . .

Rapids

Tanytarsus pupa, 30; Coc-
coneis, 30; Synedra, 30;
minor diatoms, 10. Gre-
garina.

5574 a

5. Brachycentrus....

Aug. 13

Lamar River . . .

Rapids

Mayfly fragments, 20; Tre-
bonema, 5; Cocconeis, 65;
Melosira, 5; minor dia-
toms, 5. Gregarina.

5574 a

6. Brachycentrus. . . .

Aug. 13

Lamar River . . .

Rapids

Synedra, 50; mixed diatoms,
10; detritus, 40. Gregar-

5574 a

7. Brachycentrus....

Aug. 13

Lamar River. . .

Rapids

ina.

Tanytarsus, 25; Cladophora.
20; diatoms, 10; detritus,
55. Gregarina.

5574 a

8. Brachycentrus....

Aug. 13

Lamar River . . .

Rapids

Mayfly fragments, 30; Melo-
sira, 20; Cladophora, 20;
diatoms, 10; detritus, 20.
Gregarina.

5574 a

9. Brachycentrus. . . .

Aug. 13

Lamar River . . .

Rapids

Mayfly fragments, 10; Coc-
coneis, 5; Melosira, 65;
diatoms, 10; detritus, 10.
Gregarina.

5574 ^

1. Brachycentrus sp.

Aug. 13

Lamar River . . .

Violent rapids

Tanytarsus pupae, 80; plant
tissues, 15; diatoms, 5.
Gregarina.

5574 b

2. Brachycentrus sp .

Aug. 13

Lamar River . . .

Violent rapids

Synedra, 5; detritus, 95.
Gregarina.

5574 b

3. Brachycentrus sp .

Aug. 13

Lamar River . . .

Violent rapids

Mayfly fragments, 10; dia-
toms, 1; detritus, 89.
Gregarina.

5574b

4. Brachycentrus sp .

Aug. 13

Lamar River. . .

Violent rapids

Mayfly fragments, 20; Syne-
dra, 60; Cocconeis, 10;
detritus, 10. Gregarina.

5574 b

5. Brachycentrus sp

Aug. 13

Lamar River . . .

Violent rapids

Tanytarsus larvae, 5; Tany-
tarsus purpae, 15; Syne-
dra, 10; Cocconeis, 10;
detritus, 60. Gregarina.

5574c

1. Brachycentrus sp

Aug. 13

Lamar River. . .

From Cladophora

Tanytarsus pupae, 30; Clad-
ophora, 40; Cocconeis, 25;
diatoms, 5. Gregarina.

5574 C

2. Brachycentrus sp .

Aug. 13

Lamar River. . .

From Cladophora

Detritus, 5; Cladophora. 70;
Cocconeis, 20; diatoms, 5.
Gregarina.

5574c

3. Brachycentrus sp .

Aug. 13

Lamar River . . .

From Cladophora

Cladophora, 85; Cocconeis,
14; diatoms. 1 . Gregarina.

5574c

4. Brachycentrus sp .

Aug. 13

Lamar River . . .

From Cladophora

Synedra, 5; Cladophora, 30;
Cocconeis, 60; diatoms, 5.
Gregarina.

5574c

5. Brachycentrus sp .

Aug. 13

Lamar River. . .

From Cladophora

Cladophora, 85; Cocconeis,
14; diatoms, 1. Gregar-

5574c

6. Brachycentrus sp .

Aug. 13

Lamar River . . .

From Cladophora

ina •

Synedra, 4; Cladophora, 20;
Cocconeis, 75; diatoms, 1.
Gregarina.

5575 ^

1. Thremma

Aug. 14

Tower Creek . . .

Rapids

Cocconeis, 5; detritus, 95.

SS 75 f

2. Thremma

Aug. 14

Tower Creek . . .

Rapids

Detritus, ioo.

5575 g

1. Rhyacophila

Aug. 14

Tower Creek . . .

Lateral pools

Rhyacophila (3), 80; detri-
tus, 20. Gregarina.

5575 g

2. Rhyacophila

Aug. 14

Tower Creek. . .

Lateral pools

Glossosoma (?) pupa, 90;
detritus, 10. Gregarina.

5575 g

3 . Rhyacophila

Aug. 14

Tower Creek . . .

Lateral pools

Chironomus, 20; mayfly
caudal setae, 80.

5585a

i. Hydropsyche

Aug. 25

Lost Creek ....

Shaded area. X'o vegeta-
tion.

Chironomus (3), 10; mayfly
nymph, 10; Melosira, 75;
diatoms, 5. Gregarina.

5585a

2. Hydropsyche

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Mayfly fragments, 10; Melo-
sira, 85; diatoms, 5.
Gregarina.

5585a

3. Hydropsyche

Aug. 25

Lost Creek ....

Shaded area. No vege-
tation.

Chironomus, 2; mayfly frag-
ments, 18; Melosira, 75;
diatoms, 5. Gregarina.

Roosevelt Wild Life Annals

-5

6

Table No. 5. — Showing Food of Trichoptera. — (Continued ).

Collec-

tion

Num-

ber

Specimen Number
and Name

Date,

1921

Locality

Habitat

Food in Percentages

5585a

4. Hydropsyche

Aug. 25

Lost Creek . . . .

Shaded area. No vege-
tation.

Chironomus, 5; mayfly frag-
ments. 10; Glossosoma (?)
pupae. 20; Hydrachnid, 3;
Melosira. 60; diatoms. 2.
Gregarina.

5585a

5. Hydropsyche

Aug. 25

Lost Creek ....

Shaded area. No vege-
tation.

Melosira, 18; diatoms, 2;
Glossosoma (?) pupa. 40;
mayfly fragments, 40.
Gregarina.

5585a

6. Rhyacophila

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Emptv. Gregarina.

5585a

7. Rhyacophila

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Mayfly fragments. 50; Melo-
sira, 25; wood fragments,
20; diatoms, 5. Gregar-

5585a

8. Rhyacophila

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Insect fragments, 100. Gre-
garina.

5585a

9. Rhyacophila

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Glossosoma (?) pupa. 50;
wood fragments, 49; Melo-
sira. 1. Gregarina.
Rhoicosphenia, 96; Coc-
coneis, 3; diatoms, 1.
Gregarina.

5585a

10. Glossosoma (?)...

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

55 » 0 a

11. Glossosoma (?)...

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Rhoicosphenia. 92; Cocco-
neis. 8. Gregarina.

5585a

12. Glossosoma (?)...

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Empty.

558 sa

13. Glossosoma (?)...

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Rhoicosphenia. 90; Cocco-
neis, 4; wood fragments,
4; mixed diatoms. 2.
Gregarina.

5585 a

14. Glossosoma (?)...

Aug. 25

Lost Creek ....

Shaded area. No vegeta-
tion.

Melosira. 95; Cocconeis. 2;
Synedra, 2; otner diatoms.
1. Gregarina.

5585 b

1. Hydropsyche

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Mayfly fragments. 25; Glos-
sosoma (?) pupa. 25;
Melosira, 50. Gregarina.

5585b

2. Hydropsyche

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Mayfly fragments, 5; Syn-
edra, 5; Melosira, 90.
Gregarina.

5585 b

3. Hydropsyche

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Mayfly fragments, 15; dia-
toms, 2; Melosira, 83.
Gregarina.

5585b

4. Hydropsyche

Aug. 25

Lost Creek ....

Mayfly fragments, 15; Glos-
sosoma ( ? ) pupa . 1 5 ; M elo-
sira. 68; diatoms. 2.
Gregarina.

5585 b

5. Hvdropsyche

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Mayfly fragments. 40; Chi-
ronomus pupa. 5; Perlid
young (2). 15; Melosira.
38; mixed diatoms. 2;
Gregarina.

5585b

6. Rhyacophila sp. . .

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Rhyacophila, 50; perlid
young. 10; mayfly frag-
ments. 25; Melosira. 15 ■

5585b

7. Rhyacophila sp. . .

Aug. 25

Lost Creek ....

Mayfly fragments, ioo.
Gregarina.

5585b

8. Rhyacophila sp. . .

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Perlid young (3), 60; insect
fragments, 40. Gregar-
ina.

5585b

9. Rhyacopnila sp.. .

Aug. 25

Lost Creek . . . .

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
1 ola.

Mavflv fragments. 99; Melo-
sira, 1. Gregarina.

5585b

10. Rhyacophila sp. . .

Aug. 25

Lost Creek. . . .

Melosira (small amount),
100. Gregarina.

5585b

11. Thremma

Aug. 25

Lost Creek

l

Melosira, 50; Cocconeis. 5;
diatoms. 5; detritus. 40.

Food of Trout Stream Insects

2 57

Table No. 5. — Showing Food of Trichoptera. — ( Continued ).

Collec-

tion

Num-

ber

Specimen Number
and Name

Date,

1921

Locality

Habitat

5585b

12. Thremma

Aug. 25

Lost Creek ....

Sunlit areas. Melosira
predominant. Some

Oscillatoria and Prasi-
ola.

5585b

13. Thremma

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

5585b

14. Thremma

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

5585b

15. Thremma

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

5585b

16. Thremma

Aug. 25

Lost Creek ....

Sunlit areas. Melosira

predominant. Some

Oscillatoria and Prasi-
ola.

5589a

1. Brachycentrus. . . .

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

2. Brachycentrus. . . .

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

3. Brachycentrus. . . .

Aug. 26

Lamar River . . .

Strong rapids. From

moss and Cladophora.

5589a

4. Brachycentrus. . . .

Aug. 26

Lamar River .

Strong rapids. From

moss and Cladophora.

5589a

5. Brachycentrus. . . .

Aug. 26

Lamar River. .

Stong rapids. From

moss and Cladophora.

5589a

6. Brachycentrus. . . .

Aug. 26

Lamar River . . .

Strong rapids. From

moss and Cladophora.

5589a

7. Brachycentrus.. . .

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

8. Thremma

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

9. Thremma

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

10. Thremma

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589a

11. Thremma

Aug. 26

Lamar River. . .

Strong rapids. From

moss and Cladophora.

5589b

i. Brachycentrus. . . .

Aug. 26

Lamar River . .

Minor rapids. Decaying
Cladophora.

5589b

2. Brachycentrus. . . .

Aug. 26

Lamar River . .

Minor rapids. Decaying
Cladophora.

5589b

3. Brachycentrus....

Aug. 26

Lamar River . .

Minor rapids. Decaying
Cladophora.

5589b

4. Brachycentrus. . . .

Aug. 26

Lamar River. . .

Minor rapids. Decaying
Cladophora.

5589b

5. Brachycentrus. . . .

Aug. 26

Lamar River . . .

Minor rapids. Decaying
Cladophora.

5589b

6. Brachycentrus. . . .

Aug. 26

J

Lamar River. . .

Minor rapids. Decaying
Cladophora.

Food in Percentages

Rhoicosphenia, 10; Cocco-
neis, 35; Melosira, 10;
diatoms, 5; detritus, 40.

Cocconeis, 50; Melosira, 30;
diatoms, 5; detritus, 15.

Rhoicosphenia, 5; Cocco-
neis, 10; Melosira, 10;
diatoms, 5; detritus, 70.

Rhoicosphenia, 65; Cocco-
neis, 15; diatoms, 5; detritus,

15.

Cocconeis, 5; diatoms, 5;
detritus, 90.

Moss, 90; Cymbella, io.
Gregarina.

Chironomus pupae (4), 40;
Chironomus larva, 5 ; may-
fly nymph, 30; Clado-
phora, 10; Cocconeis, 10;
Cymbella, 3; diatoms, 2.
Gregarina.

Trebonema, 50; Cladophora,
10; Cocconeis, 35; dia-
toms, 5. Gregarina.
Chironomus, 15; Clado-
phora, 5; Cocconeis, 5;
diatoms, 5; mayfly frag-
ments, 30; moss fragments,
30; Epithemia, 10. Gre-
garina.

Moss, 40; Trebonema, 30;
Epithemia, 15; Cocconeis,
5; Cladophora, 5; dia-
toms, 5. Gregarina.
Mayfly fragments, 10; Coc-
coneis, 30; Chironomus, 5;
diatoms, 5; Epithemia, 50.
Gregarina.

Cocconeis, 14; diatoms, 1;
Epithemia, 85. Gregar-
ina.

Moss, 70; Epichemia, 30.

Moss, 70; Epithemia, 30.

Moss, 70; Epithemia, 30.

Moss, 70; Epithemia, 30.

Mayfly fragments, 50; Melo-
sira, 35; Cocconeis, 5;
Epithemia, 5; diatoms, 5.
Gregarina.

Mayfly fragments, 30; Chiro-
nomus, 5; detritus, 15;
Cocconeis, 25; Epithemia,
20; diatoms, 5. Gregar-
ina.

Melosira, 10; mayfly frag-
ments, 20; Chironomus, 5;
Trebonema, 20; Cocconeis,
30; Epithemia, 10; dia-
toms, 5; Gregarina.
Melosira, 5; detritus, 20;
Cocconeis, 30; Epithemia,
40; diatoms, 5. Gregar-
ina.

Melosira, 5; Cocconeis, 20;
Epithemia, 65; diatoms,
10. Gregarina.

Ameletus fragments, 40;
Chironomus pupae, 5;
Cocconeis, 10; Epithemia,
40; diatoms, 5. Gregar-
ina.

Roosevelt Wild Life Annals

Table No. 5. — Showing Food of Trichoptera. — (Continued ).

Collec-

tion

Num-

ber

Specimen Number
and Name

Date,

1921

Locality

Habitat

Food in Percentages

5589b

7. Thremma

Aug. 26

Lamar River. . .

Minor rapids. Decaying
Cladophora.

Epithemia, 100.

5590a

12. Rhyacophila

Aug. 26

Yellowstone R .

Moderate rapids

Cladophora, 5; detritus. 95.

5590b

5. Rhyacophila

Aug. 26

Y ellowstone R .

Rapids along shore

Ephemerella fragments. 40;
Chironomus, 40; Trebon-
ema. 5; detritus, 15.

5590b

6. Rhyacophila

Aug. 26

Yellowstone R .

Rapids along shore

Wood fragments. 50; Tre-
bonema, 5; detritus. 45-

5590b

7. Rhyacophila

Aug. 26

Yellowstone R .

Rapids along shore

Insect fragments, 15; detri-
tus, 85.

5590b

8. Rhyacophila

Aug. 26

Yellowstone R .

Rapids along shore

Trebonema, io; detritus, 90.

5590b

9. Rhyacophila

Aug. 26

Y ellowstone R .

Rapids along shore

Detritus, 50; Trebonema, 20;
Cladophora, 5; moss
leaves, 15; wood frag-
ments, 10.

5 5 90 b

10. Rhyacophila

Aug. 26

Yellowstone R .

Rapids along shore

Chironomus. 20; detritus, 80.

5590 b

11. Rhyacophila

Aug. 26

Yellowstone K..

Rapids along shore

Insect fragments, 90 Chiro-
nomus, 10.

559 ia

I. Hy drops vche

Aug. 27

Tower Creek. .

Minor rapids

Insect fragments, 100.

559 ia

2. Rhyacophila

Aug. 27

Tower Creek . . .

Minor rapids

Unrecognizable matter (de-
tritus?), 100. Gregarina.

559 ia

3. Rhyacophila

Aug. 27

Tower Creek . .

Minor rapids

Insect fragments, 80; detri-
tus, 20. Gregarina.

5591 a

4,5. Rhyacophila

Aug. 27

Tower Creek . . .

Minor rapids

Detritus, ioo. Gregarina.

559 ia

6. Thremma

Aug. 27

Tower Creek . .

Minor rapids

Detritus, 100. Gregarina.

5591 a

7. Thremma

Aug. 27

Tower Creek . . .

Minor rapids

Detritus. 100. Gregarina.

5591 a

30. Rhyacophila

Aug. 27

Tower Creek . .

Minor rapids

Insect fragments, ioo.

5591 b

i. Hydropsyche

Aug. 27

Tower Creek . . .

Strong rapids

Mayfly fragments, 98; Chiro-
nomus, 2.

559 lb

2. Rhyacophila

Aug. 27

T ower Creek . . .

Strong rapids

Detritus, ioo. Few Gregar-

559 lb

3. Rhyacophila

Aug. 27

Tower Creek. . .

Strong rapids

Insect fragments, ioo. Few
Gregarina.

559 lb

4. Thremma

Aug. 27

Tower Creek . . .

Strong rapids

Diatoms, 1 ; detritus. 99.

559 ib

5 . Thremma

Aug. 27

Tower Creek . .

Strong rapids

Diatoms, 5; detritus, 95.

559IC

1. Rhyacophila

Aug. 27

Tower Creek. . .

Lateral pool, with moder-
ate current.

Detritus, 100.

Table No. 6. — Showing Summary of Food of Trichoptera.

Name

Locality

Number

of

Specimens

Animal

Food

Plant

Food

Detritus

Yellowstone River.

Lost Creek

Tower Creek

Rhyacophila

Rhyacophila

Rhyacophila

Average for

Hvdropsyche | Yellowstone River

Hydropsyche Lost Creek

Hydropsyche Tower Creek

Average for

Brachycentrus | Yellowstone River

Brachycentrus Lamar River

Average for

Thremma ! Yellowstone River

Thremma Lamar River

Thremma Lost Creek

Thremma Tower Creek

Average for

8

14

10
32

3

22

2

27

1

32

33

1

5

11

6
1 1

57

55

49

48

35

100

42

19

18.3

5 /

45

28

32

i>

35
8

9

45

20

98

36

3-7

Food of Trout Stream Insects

259

Of the results obtained from the study of the stomach contents of insects,
those from trichopterous stomachs are perhaps of most interest. Of these
results the following might be said: (1) Each species must be judged by itself.
Some species seem to have a large percentage of animal matter in their diet,
others little or none at all. (2) Those feeding on animal matter are inclined
to be cannibalistic. (3) Local conditions beget local results. A species may
have a large animal diet in one locality, and a large plant diet in a different
locality. (4) About sixty per cent of the caddisworms are parasitized by
gregarines. It was thought possible to establish a correlation between the
number parasitized and the amount of animal food taken. Thus, Rhyacophila
with an animal diet of 49% was parasitized to about 45%, and Hydropsychc
with an animal diet of 42%, to 60%. In contrast to this, Brachycentrus with
the much smaller percentage of 18.3 animal matter was parasitized practically
100%. Hence no correlation can be said to exist.

The findings of other writers seem to corroborate the foregoing conclusions,
especially 1, 2, and 3. Thus Muttkowski (T8, p. 442) calls attention to the
fact that caddisworms readily exchange a phytophagous for a sarcophagous
diet. Felber (’08) remarks on the avidity of Halesus larvae. These are car-
nivores, and not only do they not content themselves with smaller animals as
food, but may even attack larger animals. Thus on one occasion Felber noted
that some fifteen larvae clung to a Triton in an aquarium. Next day the
salamander was dead and the skeleton had been practically stripped.

Lloyd (’21) in a series of detailed studies of various caddisflies brings
together considerable data as to the food of the larvae. In successive order
the food is noted for the following species :

Neuronia postica, stygipcs, and pardalis — leaves in all stages of preservation.

Phryganca interrupta, and vestita — green plant tissues in natural
environment, in the laboratory any plant food.

Limnophilus combinatus — vegetable matter, some diatoms.

L. indivisus — vegetable matter, without discrimination, decaying tissues in
greater abundance.

D. submonilifer — raspings from sticks and plants, diatoms and other
microscopic organisms, as well as wood or plant fragments.

Arctoccia consicia — shallow raspings from sticks and vegetation.

Astcnophylax argus — dead bark and wood.

Pycnopsyche scabripennis — raspings of decomposed wood.

Platyphylax designata — young larvae with diatoms ; large larvae with
diatoms, sand, and fragments of higher plants. Vorhies lists “watercress and
watermilfoil.”

Halesus guttifer — fine raspings of decomposed wood.

Chilostigma difficilis — fragments of wood and leaves.

Brachycentrus nigrisoma — diatoms at first, then green algae and seed plants ;
carnivorous at the end of six weeks.

Mystacides sepidchralis — masticated pulp of vegetable origin.

Roosevelt Wild Life Annals

2 6o

Hydropsychidae — young larvae on green and blue-green algae. Older larvae
tend to be carnivores.

Polycentropus — plankton and small insects.

Rhyacophilidae — filamentous algae and small larvae.

The last three are the types noted for their animal food in Yellowstone
Park. It would seem, therefore, that the animal diet of these species is not
confined to inhabitants of trout streams of the West.

The Food of Diptera. The following table gives in detail the food
percentages of the specimens studied.

Table No. 7. — Showing the Food of Diptera.

Collec-

tion

Num-

ber

Specimen Number
and Name

Date,

1921

Locality

Habitat

Food Material Percentages

5566b

2.

Chironomus sp. . .

Aug.

IO

Lost Creek ....

From Melosira

Melosira, 50; Cymbella, 50.

5566 j

4 -

Chironomus sp . . .

Aug.

10

Lost Creek ....

From barrens with a few
diatom shells.

Sand, 80; diatoms, 20.

5566k

3 -

Chironomus sp . . .

Aug.

10

Lost Creek ....

From barrens with a few
diatom shells. Some

Closterium.

Diatoms, 5; sand, 95.

5 5 661

6 .

Chironomus sp . . .

Aug.

10

Lost Creek ....

Nearly pure Oscillatoria
with a trace of Melosira.

Oscillatoria, 95; diatoms. 5.

556 -a

I 3 -

Tipulid larva

Aug.

II

Y ellowstone R .

Rapids

Juices of other animals?, 100.

5567a

14 .

Tipulid larva

Aug.

1 1

Yellowstone R .

Rapids

Juices of other animals?, ioo.

5567a

37 -

Tany tarsus larva .

Aug.

II

Yellow'stone R .

Rapids

Diatoms, 95; sand, 5.

5567a

38 .

Tany tarsus larva .

Aug.

1 1

Yellowstone R..

Rapids

Diatoms, 100.

5567a

41.

Simulium sp

Aug.

II

Y ellow'stone R .

Rapids

Shore diatoms, 45; Rhizo-
clonium, 5; lake plankton,

5567a

42, 43, 44, 45. Simu-
lium sp.

Aug.

II

Yellowstone R.

Rapids

Shore diatoms. 45; Rhizo-
clonium, 5; lake plankton.

5567 a

46.

Simulium sp

Aug.

II

Yellowstone R.

Rapids

Shore diatoms, 80; lake
plankton, 20.

5567a

47 -

Simulium sp

Aug.

II

Y ellowstone R .

Rapids

Shore diatoms, 80; lake
plankton, 20.

5589 a

22.

Simulium sp

Aug.

26

Lamar River . . .

Strong rapids on moss and
Cladophora.

Trebonema. i; diatoms. 60;
Cladophora, i; detritus,
38.

Detritus, 50; diatoms, 45;
green algae, 5.

5590 C

12.

Simulium sp

Aug.

26

Yellowstone R .

Strong rapids. Violent

current.

5590 C

13. 14. 15- Simulium
sp.

Aug.

26

Yellowstone R .

Strong rapids. Violent

current.

Detritus, 50; diatoms, 45;
green algae, 5.

5 5 90 C

16.

Simulium sp

Aug.

26

Yellowstone R .

Strong rapids. Violent
current

Detritus, 50; diatoms. 45;
green algae. 5. Parasitized
by Mermis (?).

5590 C

17 -

Simulium sp

Aug.

26

Yellowstone R..

Strong rapids. Violent
current.

Detritus, 50; diatoms, 45;
green algae, 5. Parasi-
tized by Mermis (?).

Table No. 8 — Showing Summary of Food of Diptera.

Name

Locality

Number of
Specimens

Plant Food

Detritus

Chironomids

Lost Creek

4

58

42

Chironomids

Yellowstone River

2

97

3

Average for

6

= 71

29

Simulium

Lamar River

1

62

38

Simulium

Yellowstone River

14

80

20

Average for

15

= 79

21

Food of Trout Stream Insects

261

Probably more than three times the number of specimens listed of both
chironomids and Simulium were examined, but since the contents were very
much alike, it was not considered worthwhile to record them separately. What
is noted here for the food of chironomids agrees in the main with prior data

recorded by the senior author (Muttkowski, 18, p. 41 1) for species from Lake

Mendota. The diet comprises primarily the micro-food so abundant in the slimy
deposits on rocks.

For Simulium taken from Yellowstone River the large percentage of lake
plankton is of interest. The source of this is Yellowstone Lake, some forty

miles above the point where the Simulium were taken. A check made with

catches from a fine plankton net showed a somewhat similar proportion of lake
plankton and river algae and diatoms.

A considerable number of the Simulium were parasitized by a nematode
(Mermis?) . In the field notes of the senior author, under date of Aug. 15,
1921, Yellowstone River, the following remarks are relevant: “Simulium taken

with Mermis (?) emerging. Young Pteronarcys nymphs (black at this stage)
found feeding among the larvae and pupae. Enemies feeding on Simulium are
fish and Perlids, and perhaps mayflies. Parasites were found in about one
third of the larvae.” Another note, dated Aug. 1 1 , is as follows: “All swollen
individuals (of Simulium ) were parasitized. It was not determined whether
any pupae were parasitized or if parasitized individuals could pupate. It was
evident that the Mermis were leaving the larvae for their adult free-living stage.
Just at what point they left the Simulium larvae was not determined. It was
noted, however, that the parasite was coiled chiefly at the posterior end of the
larva, with a small coil near the head.”

Comparative Summary of the Food of Insects. Even from so brief a study
as the foregoing certain facts can be gleaned. The most notable point is that
aquatic insects in rapid streams are opportunists as regards food and eat what-
ever becomes available. Secondary to this is the fact that the aquatic insects
forage extensively, and migrate freely in search of food.

Both of these points become evident from the collections made in Lost
Creek, where special efforts were made to select various spots in the creek
for sample collections of food and of specimens in the vicinity or upon that
food. Reference to the stomach contents of the individual specimens and
comparison with the food items listed for the particular spot shows at once that
the large majority of specimens contained food that did not occur within many
feet of the particular locality. This indicates that these species must be rovers
and foragers to a marked extent, and that they are opportunists on the whole
and eat whatever is available.

Environmental conditions in mountain streams are strenuous ; the strong
current in particular makes life somewhat precarious and selective feeding
difficult. Hence as a result the diet becomes diversified : the insects take
whatever comes along, be it plant, detritus, or animal matter. Their diet thus
becomes much more generalized than the diet of related species in more per-
missive habitats, such as ponds and slow streams. In the latter the less strenu-

26 2

Roosevelt Wild Life Annals

ous conditions permit the insects to select their food, that is, to restrict their
diet to favored food, and to search for that food.

In other words, specialization of habitat leads to diversification of diet,
while generalization of habitat permits a restriction of diet.

The following table gives a comparative summary of the food of the insects
studied.

Table No. 9. — Showing Comparative Summary of Food of Insects. (Empty
Stomachs not Included in the Computations).

Name

Number

of

Specimens

Animal

Food

Plant

Food

Detritus

Pteronarcys .
Acroneuria .
Perla

Perloidea

Perloidea

Perloidea

Average for . .

Ephemeroidea. . .

Ephemeroidea . . .

Ephemeroidea. . . .

Ephemeroidea. . .

Ephemeroidea. . .

Average for. .

Trichoptera

Trichoptera

Trichoptera

Trichoptera | Thremma

Average for

Diptera | Chironomidae

Diptera 1 Simulium....

Average for

Ameletus sp . . .

Baetis sp

Drunella sp. . . .
Ephemerella sp .
Heptagenia sp .

Rhyacophila. .
Hydropsyche. .
Brachycentrus .

26
49

5
80

25

2

20

28

34

109

32

27

33

23

115

6
14
20

3.85

77-4

85

54

18

3

4-3

49

42

18.3

28

53 85
6-3
15

22.3

22.7
50

36.8

36.3

24

29.7
23
54-3

72.7

64

54

71

79

76 . 6

42

16

23-7

77-3

50

44-7

60

76

66

28
3-7
9

36

18

29

21

23 -4

LIST OF REFERENCES

References to food of aquatic insects are exceedingly scant. As a rule they
consist of brief notations in papers devoted to the biology and metamorphosis of
aquatic forms.

Felber, Jacques

1908. Die Trichopteren von Basel und Umgebung, mit Beriicksichtigung der
Trichopterenfauna der Schweiz. Arch. f. Naturgesch 74, pp. 80.

Lloyd, J. T.

1921. The Biology of North America Caddisflv larvae. Lloyd Librarv Bull.
21, p. 124.

Muttkowski, R. A.

1918. The Fauna of Lake Mendota: A Qualitative and Quantitative Survey
with Special Reference to the Insects. Trans. Wis. Acad. Sci..
Vol. 19, pt. i. pp. 374-482.

1925. The Food of Trout in Yellowstone National Park. Roosevelt Wild
Life Bulletin, Vol. 2, No. 4, pp. 470-497.

Food of Trout Stream Insects

263

Needham, J. G. and Claassen, W. C.

1925. Monograph of Plecoptera or Stoneflies of America north of Mexico.
Pp. 397. (Am. Ent. Soc. Monographs, Lafayette, Ind.)

Needham, J. G.

1920. Burrowing Mayflies of our Larger Lakes and Streams. U. S. Bureau
Fisheries Bull. 36, pp. 269-292.

Newcomer, E. J.

1918. Some Stoneflies Injurious to Vegetation. Jour. Agr. Research, No.
13- PP- 37-41-
Wodsedalek, J. E.

1912. Natural History and General Behavior of the Ephemeridae Nymphs
Heptagenia interpunctata (Say). Ann. Ent. Soc. Am., No.
pp. 31-40.

Wu. C. F.

1923. Anatomy and Ethology of Nemoura. Lloyd Library Bull. Ent. Ser.
No. 3, pp. 81.

IO

THE ROOSEVELT WILD LIFE MEMORIAL

As a State Memorial

The State of New York is the trustee of this wild life Memorial
to Theodore Roosevelt. The New York State College of Forestry at
Syracuse is a State institution supported solely by State funds, and
the Roosevelt Wild Life Forest Experiment Station is a part of this
institution. The Trustees are State officials. A legislative mandate
instructed them as follows:

“To establish and conduct an experimental station to be known as
‘ Roosevelt Wild Life Forest Experiment Station,’ in which there shall
be maintained records of the results of the experiments and investiga-
tions made and research work accomplished; also a library of works,
publications, papers and data having to do with wild life, together with
means for practical illustration and demonstration, which library shall,
at all reasonable hours, be open to the public.’’ [Laws of New York,
chapter 536. Became a law May 10, 1919.]

As a General Memorial

While this Memorial Station was founded by New York State, its
functions are not limited solely to the State. The Trustees are further
authorized to cooperate with other agencies, so that the work is by
no means limited to the boundaries of the State or by State funds.
Provision for this has been made by the law as follows :

“ To enter into any contract necessary or appropriate for carrying
out any of the purposes or objects of the College, including such as
shall involve cooperation with any person, corporation or association
or any department of the government of the State of New York or
of the United States in laboratory, experimental, investigative or
research work, and the acceptance from such person, corporation,
association, or department of the State or Federal government of
gifts or contributions of money, expert service, labor, materials,
apparatus, appliances or other property in connection therewith.”
[Laws of New York, chapter 4 2 . Became a law March 7, 1918.]

By these laws the Empire State has made provision to conduct forest
wild life research upon a comprehensive basis, and on a plan as broad
as that approved by Theodore Roosevelt himself.

Form of Bequest to the Roosevelt Wild Life Memorial

I hereby give and bequeath to the Roosevelt Wild Life Forest
Experiment Station of The New York State College of Forestry at
Syracuse, for wild life research, library, and for publication, the sum
of , or the following books, lands, etc.