Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Pattern and mechanism of resource partitioning between stream populations of juvenile coho salmon (Oncorhynchus… Glova, Gordon John 1978

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1978_A1 G44.pdf [ 10.53MB ]
Metadata
JSON: 831-1.0094633.json
JSON-LD: 831-1.0094633-ld.json
RDF/XML (Pretty): 831-1.0094633-rdf.xml
RDF/JSON: 831-1.0094633-rdf.json
Turtle: 831-1.0094633-turtle.txt
N-Triples: 831-1.0094633-rdf-ntriples.txt
Original Record: 831-1.0094633-source.json
Full Text
831-1.0094633-fulltext.txt
Citation
831-1.0094633.ris

Full Text

PATTERN AND MECHANISM OF RESOURCE PARTITIONING BETWEEN STREAM POPULATIONS OF JUVENILE COHO SALMON (Oncorhynchus kisutch) AND COASTAL CXTTTHROAT TROUT (Salmo clarki clarki) by GORDON JOHN GLOVA M. Sc., University of Victoria, 1972 A THESIS SUBMITTED IN PARTIAL RJLFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard The University of Briti s h Columbia October, 1978 @ Gordon John Glova, 1978 In presenting th i s thes is in p a r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make it f ree ly ava i lab le for reference and study. I fur ther agree that permission for extensive copying of th i s thesis for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th is thes is fo r f inanc ia l gain sha l l not be allowed without my writ ten permission. Department of Zoology The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date 19 October, 1978 i ABSTRACT Anadromous populations of juvenile coho and cutthroat frequently occur sympatrically in coastal streams of western North America. Their apparently similar external morphology and macrodistribution i n streams suggest they might broadly overlap i n resource use. This study examines resource partitioning between these two salmonids in streams and i n laboratory experiments. In small coastal streams, during late summer low flow, sympatric populations of juvenile coho and cutthroat were partially segregated in microhabitat use and diet. Abundance and biomass of coho were highest in low velocity microhabitats (pools, glides) whereas that of cutthroat were highest in r i f f l e s and lowest i n pools. In a l l microhabitats examined, adult insects were more common in the diet of coho than cutthroat, whereas chironomid larvae and pupae showed the reverse pattern. In laboratory stream experiments, partitioning of space between underyearling coho and cutthroat from sympatric populations was similar to that i n streams. When tested together i n summer at 12-14 °C, coho numerically dominated pools and trout dominated r i f f l e s . When tested separately, their microhabitat use was similar (60-75% of either species occupied pools). In winter, at 3 °C, both species showed strong preference for pools and overhead cover, whether tested separately or together. In laboratory stream experiments, coho and cutthroat fry dis-played a similar array of aggressive activity although non-contact aggression was more frequent in coho and nipping more frequent i n cutthroat. i i Both salmonids were most aggressive when food was presented, regardless of season. When tested together i n summer at 12-14 °C, intensity of aggression was high and cutthroat more actively defended territories i n r i f f l e s and coho in pools; in winter at 3 °C, aggression was low and both species weakly defended pools. Microhabitat use and aggressive behavior of allopatric and sympatric cutthroat tested separately i n the laboratory stream were similar, although sympatric trout defended r i f f l e territories more vigorously, responded to the feeding cycle with greater synchrony, and used components of aggressive display apparently more suited to high water velocity habitats. When fed ad libitum in the laboratory, underyearling coho grew faster than cutthroat i n winter, irrespective of temperature range (5-15 °C) and photoperiod (8-16 h), whether tested separately or together; in summer, growth was similar for both species at the same test conditions as above. When cohabiting during summer in two coastal streams, underyearling cutthroat grew faster than coho, possibly because of greater behavioral diversity in feeding and microhabitat u t i l i z a t i o n , and lesser social dissipation of energy. The data provide evidence of interactive segregation (Nilsson 1956) and ill u s t r a t e the flexible behavior of these two salmonids, cutthroat slightly more so than coho, possibly due to subdominance. Behavioral f l e x i b i l i t y may counteract heterogeneity and ins t a b i l i t y of stream environ-ments, and may permit opportunistic exploitation of broadly overlapping niches when resources are pl e n t i f u l . i i i TABLE OF CONTENTS Title Page ABSTRACT i TABLE OF CONTENTS i i i LIST OF TABLES y i ' LIST OF FIGURES v i i i ACKNOWLEDGEMENTS x i i i CHAPTER I. GENERAL INTRODUCTION 1 CHAPTER I I . NATURAL POPULATIONS OF SALMONID AND NONSALMONID FISHES IN THE STUDY STREAMS INTRODUCTION 7 DESCRIPTION OF STUDY STREAMS 9 METHODS A. POPULATION ESTIMATES 14 B. DIETS OF SALMONID FRY 18 C. GROWTH OF SALMONID FRY 18 D. BREADTH AND OVERLAP IN RESOURCE USE 19 RESULTS SYMPATRIC POPULATIONS 20 COMPARISON OF ALLOPATRIC AND SYMPATRIC POPULATIONS 33 DIETS OF SYMPATRIC COHO AND TROUT FRY 39 GROWTH OF SYMPATRIC COHO AND TROUT 44 DISCUSSION RESOURCE PARTITIONING 47 LIMITING FACTORS: FOOD OR SPACE? 51 CHAPTER I I I . LABORATORY STUDIES OF STREAM POPULATIONS OF UNDERYEARLING COHO SALMON AND COASTAL CUTTHROAT TROUT INTRODUCTION 55 METHODS A. THE TEST FACILITY 56 B. THE FISH 60 C. EXPERIMENTAL PROCEDURE 61 D. PROCESSING OF DATA 64 i v TABLE OF CONTENTS (cont'd) Page CHAPTER I I I . (cont'd) RESULTS I. COHO AND TROUT OF SYMPATRIC ORIGIN General 65 MICRODISTRIBUTION Summer 69 Winter 79 AGGRESSIVE BEHAVIOR General 94 Summer 97 Winter 101 RELATIVE DENSITY OF SPECIES 108 I I . COMPARISON OF ALLOPATRIC AND SYMPATRIC TROUT TYPES Microdistribution 110 Aggressive behavior 113 DISCUSSION BEHAVIORAL ECOLOGY OF COHO AND CUTTHROAT TROUT IMPLICATIONS FOR MANAGEMENT OF STREAM SALMONIDS Habitat diversification Winter cover Superimposition of coho on trout Escapement control CHAPTER IV. LABORATORY GROWTH OF UNDERYEARLING COHO SALMON AND COASTAL CUTTHROAT TROUT INTRODUCTION 126 METHODS A. THE FACILITIES 128 B. THE FISH 130 C. EXPERIMENTAL PROCEDURE 130 RESULTS MORTALITY AND GROWTH 133 CONDITION FACTOR 138 HIERARCHICAL EFFECTS ON GROWTH 141 DENSITY EFFECTS ON WINTER GROWTH 143 116 121 122 123 125 TABLE OF CONTENTS (cont'd) Page CHAPTER IV. (cont'd) DISCUSSION TEMPERATURE AND PHOT0PERI0D EFFECTS ON GROWTH 147 ECOLOGICAL IMPLICATIONS 150 CHAPTER V. - GENERAL DISCUSSION AND CONCLUSIONS 152 REFERENCES 160 APPENDIX 171 •vi LIST OF TABLES Number Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Summary of some physical characteristics of the study streams. Standardized breadth of microhabitat use for populations of a) sympatric cutthroat trout, coho and sculpins, and b) allopatric cutthroat trout Overlap in microhabitat use between cutthroat trout (T), coho (C) and sculpins (S) for a) by year in Bush and Holland and b) by stream, using mean values in Bush and Holland creeks. Standardized breadth of diet for sympatric populations of coho and cutthroat trout for 14 food taxa in two streams, August-September, 1973. Overlap in diet between sympatric populations of coho and cutthroat trout by habitat type in two streams, August-September, 1973. Mean number of coho and cutthroat trout fry in the ri f f l e and pool habitats in summer at the two test velocities. Comparison of F-values (P<0.01 underlined) from factorial analyses of variance of the coho and cutthroat trout test series in summer. Vertical microdistribution of coho and cutthroat trout by size class in pools, combining the data of pre-, during- and post-feeding periods in summer at both test velocities. Horizontal microdistribution (mean ± S.E.) of coho and cutthroat trout in allopatry by size class (see Table 8) in the upstream and downstream halves in riffles and pools (upper third of water column) during-feeding period^ in summer. Horizontal microdistribution (mean ± S.E.) of coho (C) and cutthroat trout (T) in sympatry for a l l size classes combined, in the upstream and downstream halves in riffles and in pools (upper third of water column) during-feeding period in summer. Page 13 32 34 43 43 71 74 77 80 81 V l l LIST OF TABLES (cont'd) Table 11 Percent number of coho and cutthroat trout with respect to their position to rocks in riffles (UPS, upstream of; DNS, downstream of, and ALS, alongside of rocks) during-feeding period in summer. Table 12 Mean number of coho and cutthroat trout fry in the r i f f l e and pool habitats in winter at the two test velocities. Table 13 Comparison of F-values (P<0.01 underlined) from factorial analyses of variance of the coho and cutthroat trout test series in winter. Table 14 Vertical microdistribution of coho and cutthroat trout by size class (see Table 8) in pools, combining the data of pre-, during- and post-feeding periods in winter. Table 15 Horizontal microdistribution (mean ± S.E.) of coho and cutthroat trout by size class (see Table 8) in the upstream half (Rl + PI) and the downstream half (R2 + P2) during-feeding period in winter. Table 16 Mean level of aggression (number of encounters per fish per 100 min) of coho and cutthroat trout in allopatry by size class (see Table 8) of fish in pools in winter at 3 °C, combining the data for pre-, during- and post-feeding periods. Table 17 Mean number of cutthroat trout fry of allopatric and sympatric origin, tested separately in the ri f f l e and pool habitats in summer at the two test velocities. Table 18 Coho and cutthroat trout fry mean rate of growth ± 95% confidence limits for a 3 X 3 temperature and photoperiod test space. Table 19 Mean winter growth of coho and cutthroat trout fry at three different densities, fed a submaintenance ration for seven weeks.: C, coho; T, trout. - v i i i LIST OF FIGURES Number Page Figure 1 Plan view and stream gradient of each of the 10 streams containing sympatric populations of coho, cutthroat trout and sculpins. Heavy solid line denotes section of stream sampled; dotted line, perimeter of watershed; hatched area, ocean; • , barrier f a l l s . Figure 2 Plan view and stream gradient of each of the 12 three study streams containing allopatric populations of cutthroat trout. See Fig. 1 for caption details. Figure 3 Mean monthly temperature ( A ) and discharge 15 (0) i n Bings Creek; vertical lines represent range (data collected by Water Survey Canada: temperature 1961-1976; discharge 1976). Figure 4 Density of coho (solid), cutthroat trout 21 (hatched) and sculpins (open) i n the lower-, mid- and upper regions of Bush and Holland creeks, 1973. The data represent the average mean of pools, glides and r i f f l e s combined. Numberscare total f i s h sampled. Figure 5 Absolute (upper) and relative (lower) biomass 23 of coho (open) and cutthroat trout (hatched) in Bush Creek. The number of samples are in-dicated and those i n parentheses indicate total number of fish sampled. Figure 6 Absolute (upper) and relative (lower) biomass 24 of coho (open) and cutthroat trout (hatched) i n Holland Creek. See Fig. 5 for details. Figure 7 Relative abundance of coho (solid), cutthroat 25 trout (hatched) and sculpin (stippled) populations in Bush and Holland creeks, 1973-75. Figure 8 Relative biomass of coho (solid), cutthroat 26 trout (hatched) and sculpin (stippled) populations i n Bush and Holland creeks, 1973-75. Figure 9 Mean relative biomass ±95 confidence limits of 28 coho (solid), cutthroat trout (open) and sculpin (hatched) populations i n Ayum, Bush and Holland creeks. ix LIST OF FIGURES (cont'd) Page Figure 10 Relation between a) fish biomass (coho, trout 30 and sculpins combined) and stream water velocity, and b) salmonid biomass (coho and trout combined) and pool surface area i n Ayum ( A ) , Bush (A) and-Holland ( A)...creeks. Figure 11 Mean biomass ±95% confidence limits i n pools 36 (1), glides (2) and r i f f l e s (3) of a) allopatric populations of cutthroat trout, b) sympatric populations of coho, cutthroat trout and sculpins combined, and c) sympatric populations of coho (solid) and cutthroat trout (open) . Figure 12 Histograms of fork length frequency of the 38 sympatric populations of coho, cutthroat trout and sculpins by habitat type i n Bush, Holland and Ayum creeks, 1975. Figure 13 Histograms of fork length frequency of the 40 allopatric populations of cutthroat trout by habitat type i n Bings (1976), French (1976), and Shawnigan (1975) creeks. Figure 14 Diet analysis of underyearling coho and cut- 41 throat trout i n Bush and Holland creeks, August, 1973; i n pools (open), glides (hatched) and r i f f l e s (solid). Figure 15 Mean fork length ± 2 S.E. (horizontal lines) of 45 underyearling coho (0) and cutthroat trout (A) i n Bush Creek, 1974. Vertical lines are size ranges: • , denotes stream discharge. Figure 16 Mean fork length of underyearling coho and 46 cutthroat trout and stream discharge i n Holland Creek, 1974. See Fig. 15 for details. Figure 17 a) Schematic of stream simulator Rafter Hartman 57 1965a) with heavy arrows indicating direction of stream flow; b) plan view of the experimental section showing outline of rocks, logs, and undercut bank (stippled); r i f f l e s - Rl, R2; pools - PI, P2; c) side view showing physiographic profile of the experimental section. Figure 18 Summer aggression (a),,and microdistribution (b), 67 of coho (circles) and cutthroat trout (triangles) in pools (open) and r i f f l e s (closed) during their i n i t i a l 4 days i n the stream simulator. Symbols are means ± S.E. X LIST OF FIGURES (cont'd) Page Figure 19 Cumulative plots of the summer microdistri- 70 bution (one replicate only) of coho and cut-throat trout during the pre-feed periods i n a) allopatry and b) sympatry. The percent number of fi s h per habitat is shown for each species. The open circles indicate position of f i s h above the undercut bank ( i .e. flood-bank area) . See Fig. 17 for details of the experimental section. Figure 20 Relative microdistribution of coho (solid) 72 and cutthroat trout (open) i n allopatry and in sympatry i n a) summer and b) winter. Figure 21 Relative microdistribution of coho (solid) 75 and cutthroat trout (hatched) i n summer ind a) sympatry and b) allopatry i n relation to size of fish (1, large; 2, medium; 3, small), feed-period (4, pre-; 5, during-; 6, post-feed) and test velocity (7, low; 8, high). Open portion of bars refer to fish i n under-cut areas i n pools. Figure 22 a) Common pattern of body alignment and 78 vertical positioning of coho (upper) and cutthroat trout (lower) i n the stream simulator; b) cutthroat trout i n high intensity lateral threat posture. Figure 23 Cumulative plots of the winter microdistri- 84 bution (of one replicate only) of coho and cutthroat trout at 3 ©C during the pre-feed periods in a) allopatry and b) sympatry. See Fig. 19 for caption details. Figure 24 Relative microdistribution of coho (solid) and 89 cutthroat trout (hatched) i n winter i n a) sympatry and b) allopatry i n relation to size of fish (1, large; 2, medium; 3, small), feed-period (4, pre-; 5, during-; 6, post-feed), water velocity (7, low; 8, high) and water temperature (9, at 3 oC; 10, at 5 °C). See Fig. 21 for additional details. Figure 25 Mean aggression ±S .E . of allopatric coho (solid) 98 and cutthroat trout (open) i n a) summer and b) winter (3 °C) i n relation to the feeding cycle (1, pre-; 2, during-; 3, post-feed period) and water velocity. LIST OF FIGURES (cont'd) Page Figure 26 Upper: summer aggression .(mean ±S.E.) of 99 sympatric coho (solid) and cutthroat, trout (open) i n r i f f l e s and pools. Numbers relate to the feed cycle as i n Fig. 25. Lower: relative frequency of the components of aggression i n intra- and interspecific cases for coho (solid) and trout open: IM, intention movement; DT, drive toward; CH, chase; TN, threat nip; CN, contact nip; L, lateral dis-play, WW, wig-wag display; F, frontal display; PS, parallel swimming; C, c i r c l i n g ; B, biting. Figure 27 Aggression of sympatric coho and cutthroat 103 trout i n winter at 3 °C (upper) and relative frequency of the components of aggression %'lower) . Symbols as i n Fig. 26. Figure 28 Winter aggression (mean number of encounters 105 per f i s h per 100 min) of coho (0) and cut-throat trout (•) i n pools i n relation to the feed cycle for the high test velocity at 5 °C. Duration of each observation period was 5 min at each successive 10 min intervals. Vertical lines indicate range. Figure 29 Aggression of sympatric coho and cutthroat 107 trout i n winter at 5 °C (upper) and relative frequency of the components of aggression (lower). Symbols as i n Fig. 26. Figure 30 Mean ±S.E. of aggression and.of microdistri- 109 bution of sympatric coho (circles) and trout (triangles) i n summer i n pools (PI, P2) and r i f f l e s (Rl, R2) for the low (open) and high (closed) water velocity at two different species relative densities: a) 10 coho, 30 trout; b) 30 coho, 10 trout. Figure 31 Upper: .summer aggression (mean ±S.E.) of cut- 114 throat trout of allopatric (heavy line) and of sympatric (light line) origin i n r i f f l e s and pools, tested separately i n the stream simulator. Lower: relative frequency of the components of aggression for trout of allopatric (solid) and sympatric (open) origin. Symbols are as i n Fig. 25,26. x i i LIST OF FIGURES (cont'd) Page Figure 32 Details of the test f a c i l i t y used in laboratory 129 growth experiments: a) diagrammatic represent-ation of the 9 test combinations of temperature and photoperiod including the 3-diagonal test points (along broken line) and 2 points tested in sympatry i n winter ( ^ f - ) , b) cross-section of the entire test f a c i l i t y and c) plan detail of one test point unit. Figure 33 Mean ±95% confidence limits (three replicates) 135 of specific growth rates of coho and cutthroat trout fry tested separately for the 9 test combinations of temperature and photoperiod. The growth rates are chronologically shown: i n summer at each 3-wk intervals (a-d); i n winter at 4-wk after i n i t i a l time (a) and at each sub-sequent 2-wk intervals (b-d). Figure 34 Mean condition factor of coho and cutthroat 139 trout fry tested separately for the 9 test combinations of temperature and photoperiod. I n i t i a l f i s h condition i s shown at a). Sub-sequent time intervals (b-e) are as i n Fig. 33 (a-d). Figure 35 Growth comparisons by length of the largest 142 coho (•), the largest cutthroat trout ( A ) and the group mean of coho (o) and of cutthroat trout ( A ) fry i n allopatry, at each of the three diagonal points of the test space i n summer and winter. Figure 36 Winter growth comparisons of the largest coho 144 (•), the largest cutthroat trout (A), and the group mean of coho (o) and of cutthroat trout .(A) fry i n sympatry, at two temperature . and photoperiod combinations i n the test space. x i i i ACKNOWLEDGEMENTS I express my appreciation and thanks to my supervisor, Dr. T. G. Northcote for his "masterful guidance" during the course of this study and constructive criticism of earlier drafts of this manuscript. Dr. J. C. Mason at the Pacific Biological Station, Nanaimo, B. C., who as my off-campus supervisor provided a stimulating environment of intellectual enthusiasm, consultation and criticism during a l l phases of this study. I also thank others who served on my supervisory committee and made help-f u l comments toward the f i n a l version of this manuscript. I owe special thanks to Dr. J. E. Mclnerney and Mr. R. Scheurle of the University of Victoria, Department of Biology, for their cooperation and kindness extended to me during my use of the stream simulator f a c i l i t y located in their laboratory. Thanks are also due to a number of other people: to R. Ptolemy and A. Savory for their assistance i n some of the laboratory work; to F. Nash for computer work, to B. Bartlett, G. Bates, A. Campbell, S. Donaldson, J. Elford, S. Glova, B. Hughes, P. Naysmith, B. Pick, D. Straughn and D. Sutherland for their assistance i n the f i e l d ; to my wife for typing this manuscript. This work was joint l y supported by Federal and Provincial Fisheries Agencies. I thank the Director and Dr. J. C. Mason of the Pacific Biological Station, Nanaimo, B. C, for use of research f a c i l i t i e s and logistic support while conducting this study; R. C. Thomas, D. W. Narver, J. C. Lyons, G. E. Reid, a l l of the B. C. Fish and Wildlife Branch for their * interest and financial assistance, without which this study would not have been possible. 1 CHAPTER I. GENERAL INTRODUCTION Studies on resource partitioning between closely related or ecologically similar taxa have gained considerable momentum (see review by Schcener 1974) ever since Hutchinson (1959) posed his query: "- - -why are there so many kinds of animals?" While Hutchinson's concept of the "ecological niche" stimulated renewed interest i n the structure of communities and i n species interaction, the foundation for such works had developed from an interplay of theoretical and experimental studies since the early 1900's. Joseph Grinnell (1904), an American naturalist, was the f i r s t to appreciate the concept that no two ecologically similar species could coexist indefinitely on a single resource. This concept was later acknowledged i n the mathematical equations of competition by Lotka (1925) and Volterra (1926) . During this era, experimental evidence supporting Grinnell's concept had been generated by several workers, although most notably by Gause (1934),,. a Russian' biologist. Gause found that when two closely related species of Protozoa (Paramecium) were cultured on a single culture medium i n the laboratory, one species was invariably eliminated i n time by the other. More recent laboratory competition experiments by Park (1962) on the flour beetles Tribolium  castaneum and T^ confusum, have parallelled Gause's findings, although the species of beetle that persisted depended on the environmental conditions: castaneum prevailed i n warm and moist conditions, while confusum did so i n drier and cooler ones; both species, however, pre-ferred warm, moist conditions. 2 The results of the earlier laboratory competition experiments led to the formulation of "Gause's principle" (Lack 1944) or the "competitive exclusion principle" (Hardin 1960), which both in essence reaffirmed Grinnell's concept that no two species can coexist indefinitely on the same limiting resource. Extension of this concept to natural communities rapidly gained popularity when scientists attempted to explain morphological and behavioral differences between species as a means to reducing competition for resources. Since then numerous studies have revealed ecological differences between seemingly similar species in sympatry (e.g. Hartley 1953; Betts 1955; MacArthur 1958). Thus, while some sympatric species may appear to closely resemble one another, no two are exactly alike, and i f one looks hard enough one i s l i k e l y to find some interspecific differences. Juveniles of anadromous populations of coho salmon (Qncorhynchus kisutch) and coastal cutthroat trout (Salmo clarki clarki) frequently occur sympatrically i n freshwater nursery areas throughout their natural geographic range. The nonanadromous forms of coastal cutthroat trout i s also common and frequently occurs i n allopatric populations that are isolated from salmon i n lakes and streams inaccessible from the sea. Like other Pacific salmon species, coho spawn i n the f a l l and in early winter, whereas cut-throat trout typically spawn in late winter - early spring (Hart 1973). Coho emerge earlier and are of a larger size at emergence than trout. The young of both species normally reside i n freshwater for one or more years before going to sea. However, i n some coho populations a large proportion of the young fry move seaward, i f not to i t (e.g. Chapman 1962; Andersen and Narver 1975; Mason 1976) and this i s believed to be the result of 3 intraspecific competition for food and space (Mason and Chapman 1965) . In Bri t i s h Columbia and elsewhere i n Western North America, small coastal streams are important nursery areas for juveniles of coho salmon and coastal cutthroat trout (Hart 1973) . Their seemingly similar external morphology and ecological requirements suggest they may be potential competitors for resources i n streams. Their coexistence may in large part depend on differences i n their behavioral ecology as documented for sympatric populations of juvenile coho and steelhead trout (Hartman 1965b) . Resources between animal populations may be partitioned spatially into horizontal patches or vertical layers, or both (Cody 1968) . The small stream environment provides the three major types of flowing water habitat--riffles, pools and glides or intermediate channels (Mundie 1974), i n which resources may be partitioned horizontally between habitat types or vertically within the deeper water habitats. While the small stream environment may provide.a highly productive nursery area for young salmonids (Egglishaw 1967; LeCren 1969; Mundie 1969; Mason 1976), some may be of extreme temporal and spatial heterogeneity, susceptible to pronounced changes seasonally i n hydrology, physiography (in unstable' streambeds), thermal budget, and i n the ava i l a b i l i t y of cover and food organisms forAfish. In response to environmental i n s t a b i l i t y , stream salmonids have evolved flexible behaviors, permitting opportunistic exploitation of resources, as hypothesized by Larkin (1956) for freshwater fishes i n general. Stream salmonids are usually not 4 characterized by occupying totally different habitats (e.g. Lister and Genoe 1970; Hartman 1965b), or eating distinctly different foods (Griffith 1974), as i n taxonomically distant forms inhabiting relatively stable environments. Rather, stream salmonids are more l i k e l y to exploit broadly overlapping niches during periods when resources are ple n t i f u l , with a more apparent expression of niche differentiation during seasons when resources are scarce. Fish species in tropical stream communities have been reported to show such niche shifts between wet and dry seasons. Zaret and Rand (1971) found that during the dry season, the diets of eleven species of fish i n a small stream i n Central Panama overlapped less "thah^during the wet season. Their findings suggest that when streamflow was reduced, food resources diminished, and the fish adopted more specialized feeding behavior, thereby avoiding intense competition. The tendency towards greater niche p l a s t i c i t y i n species inhabiting fluctuating environments as opposed to those with greater niche special-ization i n more stable ones has been frequently demonstrated i n the literature (e.g. Moldenke 1975; Heinrich 1976; Grant et a l . 1976). This study investigates pattern and mechanism of resource partitioning between juveniles of sympatric populations of coho salmon and coastal cutthroat trout i n small streams on Vancouver Island, Brit i s h Columbia. The general strategy adopted i n this study was to f i r s t investigate the pattern of microhabitat use and the diets of sympatric populations of juvenile coho salmon and coastal cutthroat trout in three small coastal streams. I also investigated microhabitat use of allopatric cutthroat trout populations (upstream of a barrier fa l l s ) i n three small 5 coastal streams to provide 1) an assessment of the spatial niche of cutthroat trout i n the absence of coho, and 2) a comparison of population structure and biomass between allopatric (coho absent) and sympatric (coho present) populations of trout. The f i e l d studies also served to provide the basic framework from which to test pertinent hypotheses and to evaluate the results of laboratory experiments con-ducted on these two salmonids. For these experiments, a laboratory model of a stream section consisting of r i f f l e and pool habitat was used, allowing observation of fish microdistribution and social behavior, and allowing environmental manipulation including that of the food supply. While a number of studies have documented behavioral and spatial inter-actions for various combinations of salmonid species (Lindroth 1955; Newman 1956; Kalleberg 1958; Hartman 1965b; Jenkins 1969; Lister and Genoe 1970; Everest and Chapman 1972; G r i f f i t h 1972), none of these have examined experimentally the role of the food supply i n species inter-actions . Also, as dominance hierarchy i n stream salmonids is size dependent (Jenkins 1969; Mason 1969), and since growth rate may.be'• - :" . positively related to"competitive fitness"(Bagenal 1967; Hall et a l . 1970), a series of laboratory growth experiments were conducted i n this study to investigate the possibility of different seasonal patterns of growth for these two salmonids. The possible importance of interspecific competition for resources i n general, and for food i n particular, between populations of these two salmonids was not investigated directly i n this study. Competition as defined by Birch (1957), Andrewartha (1961) and Milne (1961) continues 'to remain a most elusive phenomenon to demonstrate, and 6 the stream environment i s no exception. Competition for food between fis h species i n streams has been frequently inferred from apparent overlap i n their diets (Hartly 1948, Maitland 1965; Straskraba et a l . 1966; Mann and Orr 1969; Mason and Machidori 1976). However, experimental studies providing hard evidence of interspecific competition i n nature are relatively few and do not include fishes, being namely those by Connell (1961) for barnacles, Grant (1972) and Redfield et a l . (1977) for microtine rodents, Jaeger (1971, 1972) for salamanders, Haven (1973) for limpets and DeBenedictis (1974) for anuran tadpoles. Using the inferential definition of competition proposed by Maitland (1965), I compared biomass of allopatric and sympatric cutthroat trout populations to assess the probable importance of interspecific competition for resources i n general i n the sympatric fish communities. As food may be a limiting factor of juvenile salmonid production i n small streams (Mason 1976), any additional similar species, or species feeding similarly, may reduce the biomass of a species through sharing of a limiting food supply. 7 CHAPTER I I . NATURAL POPULATIONS OF SALMONID AND NONSALMONID FISHES IN THE STUDY STREAMS INTRODUCTION ' Pianka (1973) has emphasized that ecologically similar animal species partition resources i n three basic ways: spatially, trophically and temporally. Schoener (1974) i n his review of the subject presented evidence indicating that animal populations partition resources more often along dimensions of habitat rather than food and even less so by temporal means which may be manifested i n the fact that the distribution of resources are often habitat specific (Werner and Hall 1977) . The concept of interactive segregation as advanced by Nilsson (1956) i s one means by which ecologically similar species might spatially partition the available resource spectrum and reduce interspecific competition. Basically, this concept states that under conditions of intense intra-a specific competition, allopatric populations of species having similar ecological demands, tend to u t i l i z e the f u l l range of their ecological potentials. Conversely, under conditions of intense interspecific competition , sympatric populations of species having similar ecological demands, tend to u t i l i z e those resources to which they are best adapted, or have some competitive advantage over the other. Among salmonid fishes, such segregation has been demonstrated to largely account for resource partitioning between stream populations of juvenile coho salmon and steelhead trout i n coastal Bri t i s h Columbia (Hartman 1965b) . 8 Accordingly, interactive segregation might also largely account for resource partitioning between stream populations of juvenile coho salmon and coastal cutthroat trout. To i n i t i a t e this study,-' I tested two nu l l hypotheses using several small coastal streams on Vancouver Island, during the late summer period of low streamflow: 1) there i s no difference i n the pattern of microhabitat use and i n the diets between juveniles of sympatric populations of coho salmon and coastal cutthroat trout; 2) there i s no difference i n the pattern of microhabitat use and i n the biomass between sympatric .^ebho present) and allopatric (coho absent; upstream of barrier f a l l s ) populations of coastal cutthroat trout. I restricted sampling of fi s h populations to period of low flow, as this presumably is when salmonid-producing streams are at carrying capacity (Burns 1971) and segregation between salmonids i s best defined. Allopatric populations of trout were unique to this study i n that they served as a "control" i n nature. Such populations provided the basis for investigation of the ecological potential of cutthroat trout i n the absence of coho, a competitor species. I assumed that the microhabitat requirements of both trout types and that the physical conditions of their streams were similar. Investigation of the ecological relations between sympatric populations of coho and cutthroat trout i n small coastal streams i s frequently faced with the problem of the presence of sculpins, either Cottus aleuticus or asper, or both. Sculpins are a common cohabitant with anadromous salmonids i n reaches of such streams accessible from 9 the sea (Andersen and Narver 1975; Mason and Machidori 1976) . These two groups of fi s h are distinctly different taxonomically (Hart 1973), but not necessarily so ecologically (Mason and Machidori 1976) . The importance of the ecological role of sculpins i n guilds of these fishes has been demonstrated by 1) their frequently higher biomass levels than those of sympatric salmon and trout combined (LeCren 1965; Mann 1971; Petrosky and Waters 1975; Mason and Machidori 1976), 2) the considerable overlap i n their diets with those of salmonids (Dineen 1951; Andreasson 1971; Mason and Machidori 1976), and 3) by their negative effect on the production of cutthroat trout i n laboratory streams, presumably through interspecific competition for food (Brocksen et a l . 1968) . Based on these premises, I surmised that the pattern of microhabitat use by stream salmonids i s influenced, at least to some degree, by sculpin populations. In this study, sculpins were therefore given similar sampling effort as were coho and cutthroat trout. They have, however, been neglected i n previous studies dealing specifically with the subject of interaction between stream salmonids. DESCRIPTION OF STUDY STREAMS Six small coastal streams on the east and south sides of Vancouver Island, B.C. were selected for study of their fish populations. Of these, Ayum, Bush and Holland creeks (Fig. 1) contained sympatric populations of coho salmon, coastal cutthroat trout and sculpins, primarily Cottus aleuticus. The extent of upstream movements by fishes in 10 Fig. 1. Plan view and stream gradient of each of the-streams containing sympatric populations of coho, cutthroat trout and sculpins. Heavy solid line denotes section of stream sampled; dotted line, perimeter of water-shed; hatched area, ocean; • , barrier f a l l s . 11 each of the three streams are restricted to approximately 2 km above high tide by a waterfall. The others, Bings, French and Shawnigan creeks (Fig. 2) upstream of their barrier f a l l s contained allopatric (nonanadromous) populations of coastal cutthroat trout isolated from salmon. Also present were sparse populations of the three-spined stickleback, Gasterosteus aculeatus, i n Bings Creek and of the coast-range sculpin, Cottus aleuticus, i n Shawnigan Creek (from the lake below) . Coho have not been reported to occur, and none were taken i n this study, upstream of the f a l l s i n each of these three streams. In Bings and French creeks, the f a l l s are situated several km from sea; i n Shawnigan Creek, the f a l l i s situated near the mouth. The length of the stream segments i n which cutthroat trout were abundant was about 3 km i n both Bings and Shawnigan creeks and about 6 km i n French Creek. Physically, the six streams are f a i r l y similar,ranging i n watershed area:> from 17 to 31 knr% average gradient" from 1.0 to 4.5% and minimum summer discharge from 1.2 to 5.7 m^ /min (Table 1) . The forest stand i n a l l i s a mixture of second growth deciduous and coniferous species; the understory, consists mainly of s a l a l , sword fern, salmon-berry, stinkcurrent and devil's club. Vegetation along the streambank i s dense i n most areas, mainly of salmonberry, stinkcurrent and alder. Pools are slightly deeper but not larger i n area i n the streams containing the sympatric populations of fish than in those containing the allopatric populations of cutthroat trout. Streambed materials i n Holland and i n Ayum appeared to contain less gravel and more rubble, boulders and bedrock formation than i n the other four streams. Further, Holland lacks a natural upper estuary due to a large culvert and spillway under-12 Fig. 2. Plan view and stream gradient of each of the three study-streams containing allopatric populations of cutthroat trout. See Fig. 1 for caption details. 13 Table 1. Summary of some phys ica l charac te r i s t i cs of the study streams. ' • -• -MINIMUM • ' WATERSHED STREAM SUMMER MEAN AREA (m2) MEAN DEPTH AREA GRADIENT DISCHARGEa (cm) (km2) (%) (mVmin) POOLS GLIDES RIFFLES POOLS Bush 23 1.3 1.2 39 31 - 28 30 Holland 31 2.6 1.2 27 36 27 36 Ayum 17 4.5 3.5 40 53 56 41 Bings 20 1.5 5.7 46 50 27 36 French 26 1.0 1.6 56 54 20 31 Shawnigan 22 1.0 1.5 71 39 25 31 average weekly readings August 15-September 30 during study years 14 lying the Island Highway/Railroad overpass, but this structure presents no barrier to upstream movements of fi s h during high tide. A complete annual cycle of both stream temperature and discharge is available only for Bings Creek (Fig. 3), with temperature being highest i n July and August and discharge highest from November to January. These data are considered f a i r l y representative i n temporal pattern, a l -though not necessarily i n magnitude, of the other five study streams. In a l l streams, during the f a l l sampling period, the range i n water tempera-ture was from about 11 to 15 °C and i n stream discharge from about 1-2 nvfy min. Hydroxshemical composition (major and minor ions) was monitored monthly over one complete annual cycle in Bush and Holland creeks (Appendix Tables 1, 2); peaks i n total dissolved solids, nitrates, phosphates and other maj or ions occurred at the onset of heavy rains i n the f a l l . These may be " f a i r l y " representative i n seasonal pattern, although not necessarily i n magnitude, i n the other four streams. Turbidity was not measured i n any of the streams, the water being clear i n a l l during the f a l l sampling period. METHODS A. POPULATION ESTIMATES Sampling of f i s h for determination of population estimates took place i n a l l streams approximately fromcimid August- end September. Bush and Holland creeks were sampled i n three consecutive years from 1973 to 1975; the remaining four streams were sampled only one year 15 Fig. 3. Mean monthly temperature (A) and discharge (0) i n Bings Creek; vertical lines represent range (data collected by Water Survey Canada: temperature 1961-1976; discharge 1976). 16 i n 1975-76. The repeated sampling i n Bush and Holland provided replication of the results i n order to assess the v a r i a b i l i t y between years for a portion of the fish/populations studied. Burns (1971) found that variation between years i n salmonid standing crops by natural means alone, was relatively high i n several streams i n northern California. I surmised that significant changes i n species relative abundance between years should they occur, would result i n differences i n species pattern of microhabitat use. Fish population estimates were determined by the removal method (Seber and LeCren 1967). For small streams and catches large relative to the total population, this method has proven satisfactory (Narver and Andersen 1974) . Each stream was sampled progressively i n an upstream direction i n habitat sections of "fairly uniform velocity, selected i f the local physiography appeared suitable to.isolate the inhabiting f i s h with stopnets. At least five of each of the pool, glide and r i f f l e habitats i n the mid-region of each study stream were sampled. Fine mesh minnow seines were stretched across the down- and upstream limits of each chosen section and held snug to the bottom with small rocks when necessary. The blocked off section was then electrofished with a 440-V DC fish shocker (Smith-Rootr Laboratories, Mark V) . .. from the downstream end up, for a minimum of three successive runs, or u n t i l catches declined to zero or nearly so. This sampling technique was similar to that used by Egglishaw (1970) for stream salmonids i n Scotland. The stunned f i s h were collected with dip nets and1 heId i n separate plastic buckets for each successive run u n t i l sampling was completed. A l l f i s h captured were anaesthetized i n MS-222, fork length 17 and species identification recorded for each f i s h and scale samples taken from individuals obviously exceeding the length range of age 0+ fi s h . Upon complete recovery, the fi s h were returned to the section sampled and the stopnets were removed shortly after the fish had distributed themselves within the confined area. Physical data gathered for each sampling site included stream velocity (pool < 8 cm/s; glide 8-20 cm/s; r i f f l e > 20 cm/s), area sampled and water depth. In addition, substrate composition was also recorded i n Bush and Holland creeks i n 1973 . Mean sampling areas i n a l l streams ranged from 27-71 for pools, 31-54 for glides and 20-56 m^  for r i f f l e s . Fish biomass estimates i n each of the three habitat types were computed from the measured mean fork length (X) and mean wet weight (Y) by species for each stream i n the linear regression equation, Y = AXD. The equation was for each population of fi s h , derived from length and weight determination.of live samples of fi s h from each stream, held in the laboratory 1-2 days without food prior to measurement. I determined the distribution and abundance of the f i s h populations i n both Bush and Holland creeks from the mouth to the barrier f a l l s i n 1973, to establish i f the populations were distributed uniformly or clumped. I surmised that their pattern of habitat segregation would be best defined i n the zone of maximum overlap between populations, due to possible accentuation of interspecific competition for resources. Both streams were s t r a t i f i e d longitudinally into lower-, 18 mid- and upper sections based on major changes i n their stream bottom profiles and substrate compositions (Fig. 4). At least five of each of the pool, glide and r i f f l e habitats i n each of the three sections were sampled. Sampling was otherwise restricted to the mid-section of the study reaches i n each of the six streams i n subsequent years. B. DIETS OF SALMONID FRY Underyearling coho and cutthroat trout were subsampled during routine sampling of a selected pool, glide and r i f f l e habitat i n the mid-sections of Bush and Holland creeks, 1973. The fish were preserved i n 10% formalin for diet analysis. No regurgitation of food occurred during this process. In the laboratory, the stomach was extracted and placed i n a labelled v i a l containing 10% formalin. The contents of the fore-stomach (Mundie 1971) were extracted and the identifiable food items were separated and counted into taxonomic categories with the aid of a binocular microscope. C: GROWTH OF SALMONID FRY Coho and cutthroat trout fry were seined biweekly from various habitat types i n Bush and Holland creeks from 3 June - 27 September, 1974. They were anaesthetized i n MS-222 and the species identification and fork length were recorded for each f i s h . Scale samples were taken from in -dividuals i n the upper size range of the fry to ensure no inclusions of age 1+ f i s h . Upon complete recovery, the fish were returned to the section sampled. Stream discharge was estimated for the same time intervals i n both streams using midget-Bentzel pitot tubes (Everest 1967). 19 D. BREADTH AND OVERLAP IN RESOURCE USE I calculated breadth of microhabitat niche for sympatric coho, trout and sculpins i n each of the three streams and breadth of food niche for coho and trout i n Bush and Holland. I used Simpson's index of diversity as employed by MacArthur (1972): 1 B = i where i n the present study, B i s niche breadth and p^ represents the relative proportion of a species i n the ^ habitat or relative consumption of the ^ food-type. Values of p were obtained from average density estimates or percentage composition by number of an item i n their diet i n the relation d where d^ represents grams or numbers of f i s h per unit wetted-area of stream, or frequency of an item i n the total diet of a species. Niche breadth can vary from unity to the number of habitats sampled, or the number of food categories eaten. For comparative purposes, breadth of microhabitat and of food niche were standardized by dividing by the threermicrbhabitatS: (pools,"glides,- . r i f f l e s ) ; or .by the number of food-^types.(seev.Fig. 14), respectively (range 0-1.0). Overlap i n microhabitat use and i n diet between species pairs i n the sympatric populations of f i s h were calculated from Pianka (1973): 20 HC pik P j k ) ° i j = v"vn4 zpjk where i n this study, 0.;. = 0'.... = niche overlap between species i and species j (range 0 to 1.0) ; and P^ are the proportions of the k resource as used by i^ 1 and species. Overlap i n microhabitat niche between species was calculated nby-, both numerical and biomass estimates of f i s h . Confidence limits cannot be calculated for either of these indices, but rather, gross differences or marked similarities between species may be observed. RESULTS SYMPATRIC POPULATIONS Juvenile salmonids and sculpins were not distributed uniformly over the length of the study reaches i n Bush and Holland creeks, 1973 (Fig. 4). Combining the data of pools, glides and r i f f l e s within each of the three zones per stream, average mean abundance of salmonids was highest i n the mid-zone of both streams. This was the area of maximum overlap i n the longitudinal distribution of coho and trout. Sculpin abundance was also highest i n the mid-zone i n Holland, but i n the lower-zone i n Bush, the latter contained numerous young i n shallow r i f f l e s and glides near the mouth. Pooling salmonid and sculpin densities, fish were about twice as abundant i n Bush as i n Holland, throughout the study reaches. My estimated abundances of sculpins are no doubt conservative, due to the d i f f i c u l t y i n adequately sampling age 0+ fi s h , a problem 21 HOLLAND CREEK OJ cn m 0 I 2 km Fig. 4. Density" of coho (solid), cutthroat trout (hatched) and sculpins (open) i n the lower-, mid- and upper regions of Bush and Holland creeks, 1973. The data represent the average mean of pools, glides and r i f f l e s combined. Numbers are total f i s h sampled. 22 reported by others (Krohn 1967; Goodnight and Bjornn 1971) . The pattern of habitat segregation between coho and cutthroat trout was similar i n lower-,mid- and upper zones i n Bush and Holland creeks, 1973 (Figs. 5, 6, bottom) . In both streams, coho made up a greater percentage of the salmonid biomass i n pools (53.1 to 90.8%) than did trout (9.2 to 46.9%), while i n r i f f l e s , trout dominated the biomass, particularly i n Holland (63.4 to 88.0%) ; glides were areas of intermediate biomass for both species, with coho ranging from 51.8 to 80.8% and trout from 25.0 to 48.2%. Total biomass of coho and cutthroat trout combined (Figs. 5, 6, top) was highest i n pools and lowest i n r i f f l e s i n Holland Creek i n a l l three zones; the same pattern occurred i n Bush Creek i n the lower zone but i n mid- and upper zones, glides contained considerably higher salmonid biomass than did pools. In Bush, coho and trout i n pools were of a smaller body size i n the upper two-thirds than i n the lower third of the study reach (Glova and Mason 1974) . Juveniles of coho and cutthroat trout showed a similar pattern of habitat segregation i n Bush and Holland creeks during the late summer period of low streamflow from 1973-75 (Figs. 7, 8). Both relative abundance and biomass of coho were highest i n the slower velocity habitats, while that of trout were highest i n r i f f l e s and lowest i n pools. Glides supported intermediate levels of abundance and biomass of both salmonids, excepting i n Holland, 1974. Sculpins, consisting mainly of 23 4 0 0 CM E O g £ 3 0 0 L CO Qzoo g CD Q I O O z o < o ^ 75 50 CO % 2 5 g m LU > LU or o 25 50 75 5(226) 5(278) 5(556) 5(141) 5(25) Ld - » O b! O — l a. O K L O W E R 4(212) 4(269) 6(101) _ i u O 9 o 0. LU _ l L L Lu o cr M I D i 7(73) o o o. LU Q o LU _ l u. u. U P P E R Fig. 5. Absolute (upper) and relative (lower) biomass of coho (open) and cutthroat trout (hatched) in Bush Creek. The number of samples are indicated and those in parentheses indicate total number of fish sampled. cT «oo E O o E 3 0 0 c n CO ^ 2 0 0 g co Q I O O o < CO 75 50 25 24 CO CO < g CD LU > _ l LU or o 25 50 75 10(384) 13(503) 5(149) 5(67) o o a. UJ a UJ u. • it CO CE O O Q. UJ —1 u. u. 5 cc UJ a L O W E R M I D 10(420) 3(91) 5(80) -I £ o 9 o _i UJ _ l ll_ U . U P P E R Fig. 6. Absolute (upper) and relative (lower) biomass of coho (open) and cutthroat trout (hatched) i n Holland Creek. See Fig. 5 for details. 0 ' « a 1 m/Ui «Ka_J 1 H Z L J « u mrni-i POOL GLIDE RIFFLE POOL GLIDE RIFFLE Fig. 7. Relative abundance of coho (solid), cutthroat trout (hatched) and sculpin (stippled) populations i n Bush and Holland creeks, 1973-75. POOL GLIDE RIFFLE POOL GLIDE RIFFLE Fig. 8. Relative biomass of coho (solid), cutthroat trout (hatched) and sculpin (stippled) population i n Bush and Holland creeks, 1973-75. 27 of Cottus aleuticus, showed no distinct pattern of microhabitat use i n either study stream. They were abundant i n a l l habitat types. In Holland, and i n a l l but a few cases i n Bush, the relative abundance and biomass of sculpins surpassed that of coho and trout combined, i n each of the three habitat types. Sculpin biomass has been reported to range from 50-80% i n other salmonid-producing streams on the east coast of Vancouver Island (Mason and Machidori 1976). In Holland, but not i n Bush, major changes i n fi s h standing crops occurred from 1973-75; cutthroat trout abundance and biomass declined significantly (P<0.05) i n 1974, while that of coho and sculpins increased, although nonsignificantly so (Glova and Mason 1976a) . This altered the pattern of segregation between coho and trout slightly; relative abundance and biomass of coho increased i n glides, while that of trout decreased, compared to that i n 1973 (Figs. 7, 8). I compared the pattern of segregation between sympatric cutthroat trout, coho and sculpins between streams, using the average means of the relative biomass of each species i n Bush and Holland creeks from 1973-75 (Fig. 9) . Trout biomass was consistently highest i n r i f f l e s , intermediate i n glides and lowest i n pools, i n a l l three streams. Coho biomass was higher i n both pools and glides than i n r i f f l e s , i n both Bush and Holland; i n Ayum Creek, however, the pattern was reversed, with a high 36% i n r i f f l e s and a low 18.8% i n pools. Sculpins showed no discernible pattern of microhabitat use i n Bush and Holland, but i n Ayum they were distinctly highest i n pools and 28 80 r-70 60 50 40 30 2 0 10 0 CO CO < | 6 0 CD 50 A Y U M C R E E K 1975 x CO li_ li. o LU O or Q . 4 0 30 2 0 10 0 80 70 60 50 4 0 30 20 10 0 B U S H C R E E K 1973-75 H O L L A N D C R E E K 1973 -75 G L I D E R I F F L E Fig. 9. Mean relative biomass ±95 confidence limits of coho (solid), cutthroat trout (open) and sculpin (hatched)populations i n Ayum, Bush and Holland creeks. 29 lowest i n r i f f l e s , the opposite to that of salmonids. Sculpin distribution appeared to be more related to that of trout than that of coho. In cases i n which trout population densities were relatively low, such as i n Bush (1973-75 inclusive) and i n Holland (1974 and 1975) creeks, sculpins i n r i f f l e s comprised an average high of 59.3% as opposed to a low 11.0% i n Ayum Creek, 1975, which contained an abundance of trout. Sculpins l i v e almost entirely on the stream bottom, often beneath rocks, particularly the smaller individuals. I frequently observed trout fry chase and nip young sculpins that emerged from beneath rocks i n shallow glide and r i f f l e areas. Sculpin emergence from beneath rocks was crepuscular and may serve to minimize agonistic interactions with trout during the day. Total biomass of salmonids and sculpins combined showed a negative linear relation with stream water velocity (Fig. 10a) . The negative relationship Y = 7.801-0.137X (Y, biomass; X, velocity) was significantly correlated at P<0.01 with Pearson's correlation coefficient being -0 .922. Fish biomass was highest at lowest water velocities and areas with water velocities exceeding about 60 cm/s are l i k e l y to contain fewtlc or no f i s h . By the same analysis, mean f i s h biomass and pool depth showed no correlation (P>0.05), although biomass was generally higher i n the deeper pools. Both depth and velocity were reported to be important factors affecting the production of juvenile coho i n a r t i f i c i a l rearing channels (Ruggles 1966) . I postulated that individuals of juvenile coho and cutthroat trout compete for space at the head of pools to attain prior i t y i n 30 Y = 7.80l - O.I 37X r = - 0 . 9 2 2 10 20 30 40 MEAN WATER VELOCITY (cm/sec) L O G Y = 1.506 - 0 .729 LOGX * A r = -0 .397 A \ * A -20 30 40 50 60 70 80 90 100 POOL SURFACE AREA (m2) Fig. 10. Relation between a) fish biomass (coho, trout and sculpins) and stream water velocity, and b) salmonid biomass (coho and trout) and pool surface area in Ayum ( A ) , Bush (A) and Holland (A) creeks. 31 exploitation of invertebrate d r i f t . Salmonid biomass-per-unit space would therefore, be higher i n small pools than i n large ones due to relatively low densities of fish toward the downstream ends i n the latter. A l l applicable data were pooled and a logarithmic transformation applied i n linear regression analysis. The negative relationship log Y = 1.506-0.729 log X (Y, biomass; X, pool area) was significantly correlated at P<0.01 and salmonid biomass ranged from 1.1 to 6.0 g/m2, i n pools of 100 and 10 m2, respectively (Fig. 10b) . This phenomenon may--acc6unt 'in-part^i f° r the v a r i a b i l i t y i n salmonid biomass between pools (Appendix, Tables 3-5) . Standardized breadth of microhabitat use for sympatric coho, trout and sculpins i s shown for each of the three streams (Table 2) . Breadth of microhabitat use, either by biomass or numerical densities, was consistently higher for trout than for coho over the 3-yr period i n Bush and Holland and for the single year i n Ayum. For sculpins, breadth of microhabitat niche was similar to that of trout i n Bush and Holland, but substantially less than that of either salmonid i n Ayum. These findings may reflect the ideas of Morse (1974) and Wilson (1975), who postulated that socially subordinate species usually have a broader niche than do dominant ones. Although unquantified, my many casual observations i n the f i e l d suggested cutthroat trout fry are socially subdominant to coho fry, whereas sculpins, at least young-of-year, are subdominant to both. Microhabitat use by coho, cutthroat trout and sculpins over-32 Table 2. Standardized breadth of microhabitat use for populations of (a) sympatric cutthroat trout, coho and sculpins and (b) allopatric cutthroat trout. Values derived from f i s h biomass are unbracketed; those from numerical density are bracketed. a) sympatric Bush Holland Ayum Average Trout Coho Sculpins 0.87(0.90) 0.78(0.77) 0.94(0.96) 0.91(0.95) 0.74(0.71) 0.85(0.97) 0.98(0.85) 0.87(0.91) 0.63(0.64) 0.92(0.90) 0.80(0.80) 0.81(0.86) b) allopatric Trout French 0.92(1.0) Bings 0.84(1.0) Shawnigan 0.99(0.98) Average 0.92(0.99) 33 lapped extensively i n each of the three streams (Table 3) . The analysis consistently showed that whether by biomass or numerical density, spatial overlap was less between coho and trout than i t was for either salmonid with sculpins, i n both Bush and Holland over the 3-yr period of study. Coho and trout segregated into slow and fast velocity habitats, respectively, whereas sculpins were abundant i n a l l habitats. In Ayum, the pattern of interspecific overlap i n microhabitat use was different from that in Bush and Holland creeks; coho broadly overlapped with trout and to some extent also with sculpins, whereas trout and sculpins over-lapped least, particularly i n faster velocity water. COMPARISON OF ALLOPATRIC AND SYMPATRIC POPULATIONS Statistics of the f i s h populations and some related stream physical parameters are summarized for each of the six streams by habitat type (Appendix, Tables 3-8). Here, allopatric populations are those of cutthroat trout upstream of the barrier f a l l s i n Bings, French and Shawnigan creeks; sympatric populations are those of cohabiting coho, cutthroat trout and sculpins i n Bush, Holland and Ayum creeks. I compared the pattern of microhabitat use of these two population types, using estimates of biomass (g/m^ ) rather than density (numbers of fish/m ), as they were more meaningful with the broad range i n size of f i s h present. I examined s t a t i s t i c a l significance of these by parametric one-way analysis of variance, or by the Kruskal-Wallis non-parametric s t a t i s t i c (Siegel 1956) i n cases where Bartlett's test (Sokal and Rolf 1969) indicated lack of homogeneity of variance. 34 Table 3. Overlap i n microhabitat use between cutthroat trout (T), coho (C) and sculpins (S) for (a) by year i n Bush and Holland and (b) by stream, using mean values i n Bush and Holland creeks. Values above the diagonal are calculated from fi s h biomass, those below the diagonal from numerical density. a) 1973 1974 1975 BUSH CREEK I C S T C S T C S T C S 1.0 0.48 0.93 0.78 1.0 0.77 0.92 0.82 1.0 1.0 0.73 0.55 1.0 0.67 0.94 0.76 0.96 1.0 1.0 0.76 0.67 0.82 1.0 0.92 0.97 0.92 1.0 HOLLAND CREEK T C S 1.0 0.64 0.91 0.67 1.0 0.90 0.91 0.92 1.0 1.0 0.88 0.83 1.0 0.98 0.91 0.88 0.83 1.0 1.0 0.67 0.95 0.71 1.0 0.87 0.95 0.89 1.0 b) Bush Holland Ayum T C S 1.0 0.60 0.76 0.78 1.0 0.88 0.89 0.90 1.0 1.0 0.75 0.72 1.0 0.95 0.90 0.91 0.88 1.0 1.0 0.83 0.48 0.94 1.0 0.85 0.48 0.87 1.0 35 The distribution of the biomass of allopatric cutthroat trout by habitat type, resembled that of sympatric coho, trout and sculpins combined (Fig. 11a, b). In both population types, f i s h biomass was consistently highest i n pools, intermediate i n glides and lowest i n r i f f l e s i n a l l six streams. However, the total biomass of the populations i n sympatry were significantly (P<0.05) higher than those i n allopatry, excepting i n Bings Creek. Pooling the data of a l l habitat types by stream, average mean biomass i n Shawnigan and French creeks was only 1.2 and 1.9 g/m2, respectively; those i n Ayum, Bush and Holland ranged from 4.5 to 5.4 g/m2. Sculpins comprised a major portion of the fish biomass i n each of these three streams (Appendix, Tables 3-5) . Other small coastal streams on Vancouver Island with a similar sympatric species composition as i n the present study, have been reported to support similar fish biomass on the west coast (Andersen and Narver 1975), but higher on the east coast, some 100 km north of the present study location (Mason and Machidori 1976). The distribution of the biomass of cutthroat trout by habitat type was markedly different between allopatric and sympatric populations (Fig. 11a, c) . Biomass of allopatric trout wss some two-fold higher i n pools than i n r i f f l e s i n Bings and French creeks, although less so i n Shawnigan Creek; conversely, biomass of sympatric trout \was*j similar between habitat types i n a l l three streams and did not exceed 1 g/m2. Riffles did not support higher biomass of sympatric trout than did pools, as one might expect as an outcome of their segregation from coho. Pooling the data for each of the three habitat types within each of the six streams, trout biomass was significantly (P<0.05) less for the sympatric a ) B I N G S 36 F R E N C H S H A W N I G A N 1 1 1 1 1—1—1—L I 1 1 1 1 1 123 k 1 2 3 M E \ o> CO CO < 10 o CD b ) A Y U M I B U S H +1 H O L L A N D rfl 5 r-C ) A Y U M B U S H H O L L A N D - t^JL^I t l n ^ ^ Fig. 11. Mean biomass ±95% confidence limits i n pools (1), glides (2) and r i f f l e s (3) of a) allopatric populations of cutthroat trout, b) sympatric populations' of coho, cutthroat trout and sculpins combined, and c) sympatric populations of coho (solid) and cutthroat trout (open) 37 (range 0.5 to 0.9 g/m^ ) than for the allopatric (range 1.1 to 3.5 g/m^ ) populations, i n a l l but Shawnigan Creek (1 g/m^ ) . However, the biomass of coho and trout combined i n Ayum and Bush creeks was not significantly different (P<0.05) from that of the trout in Bings and French creeks. Habitat segregation between coho and cutthroat trout showed a size-related pattern. Size frequency histograms by habitat type (Fig. 12) indicated that pools and glides most often contained a broader range i n size of fish than did r i f f l e s , for both salmonids i n Bush, Holland and Ayum creeks. This pattern was more pronounced i n trout than i n coho populations due to higher frequency of age 0+ fish i n the latter. Thus, coho and trout coexisting i n pools were more widely separated by size (more than one age class of trout), whereas i n r i f f l e s they were more closely matched (mostly age 0). Sculpins showed a similar pattern of size distribution between habitat types as did salmonids. Considering a l l three species, the coexistence of a l l age-classes within pools suggests that size segregation within habitat types may be more possible in the deeper water environments. Pooling a l l age-classes within species for a l l three streams i n one-way analysis of variance, only trout showed a significant (P<0.05)* difference i n mean fork length (mm) between habitat types (table below). Pools Glides Riffles F-ratio Coho 51.9 51.4 50.6 0.11 Trout 64.0 51.0 49.1 4.74* Sculpins 69.5 65.9 62.0 0.43 38 Fig. 12. Histograms of fork length frequency of the sympatric populations of coho, cutthroat trout and sculpins by habitat type i n Bush, Holland and Ayum creeks, 1975. 39 Size frequency distribution by habitat for allopatric trout (Fig. 13.) resembled those of sympatric salmonid and sculpin populations. In Bings, French and Shawnigan creeks, trout range i n fork length was broadest i n pools and narrowest i n r i f f l e s ; pools contained trout of a l l age classes whereas r i f f l e s were restricted mostly to age 0+ individuals. In conclusion, microhabitat niche of allopatric trout was broader than that of sympatric trout when compared by numbers; by biomass they were roughly the same (see Table 2). Allopatric trout were about equally distributed between habitat types, whereas sympatric trout were more common i n r i f f l e s than i n pools. The restricted microhabitat use by sympatric trout may i n large part reflect coho social dominance in the slower velocity habitats. DIETS OF SYMPATRIC COHO AND TROUT FRY Segregation was not clearly evident i n the diets between sympatric populationscofcoho and cutthroat trout i n streams (Fig. 14). The food items eaten appeared to largely reflect food availability and f i s h microdistribution, i n contrast to the more selective diet reported for hatchery juvenile rainbow trout i n an experimental stream (Bisson 1978) . Chironomid larvae and pupae were numerically the most important food items i n both coho (43.9-66.4%) and trout (66.5-80.7% i n a l l habitat types i n Bush Creek. Mundie (1969) reported a comparable diet for coho fry i n a small stream on Vancouver Island, similar i n many respects to that of Bush Creek. In Holland, food habits of both salmonids were more variable: chironomid larvae and pupae made up from 32.3-37.7% i n trout and from 5.7-26.4% in coho; BINGS C R E E K FORK LENGTH (mn) Fig. 13. Histograms of fork length frequency of the allopatric populations of cutthroat trout by habitat type i n Bings (1976), French (1976), and Shawnigan (1975) creeks. 41 A V E R A G E P E R C E N T NUMBER IN S T O M A C H S 5 0 0 50 5 0 0 5 0 I I I I I I I I I C O H O adult insects T I I I 1 I I I I I I I I I T R O U T C O H O chironomid g (larvae, ® pupae) a. n ( onchironomid (larvae, pupae) trichopteran ( larvae) ephemeropteran (nymphs) plecopteran (nymphs) Collembola T R O U T — 1 L a miscellaneous B U S H C R E E K H O L L A N D C R E E K Fig. 14. Diet analysis of underyearling coho and cutthroat trout i n Bush and Holland creeks, August, 1973; i n pools (open), glides (hatched) and r i f f l e s (solid). 42 other benthic forms of importance to both salmonids were trichopteran and other dipteran larvae. Plecopteran and ephemeropteran nymphs were relatively unimportant i n the diets of either coho or trout i n Holland. Aerial foods, particularly dipteran and hemipteran forms were more common i n the diet of coho than i n trout i n both streams (Fig. 14), suggesting greater surface feeding by coho. In both streams, however, coho sampled i n r i f f l e s contained more chironomids and less aerial foods than those sampled in pools; this may suggest they forage more on bottom and mid-water sources of food at higher water velocities. Trout showed no such difference between r i f f l e and pool habitats. Coho social dominance i n pools (Hartman 1965b) may have reduced the opportunities for trout to feed at the surface. Breadth of diet for coho and cutthroat trout i n Bush Creek was similar, irrespective of habitat type, but not i n Holland Creek (Table 4) . In Bush, although both salmonids foraged on a variety of food taxa, chironomid larvae constituted a high percentage of their total diets, resulting i n a relatively narrow food niche for either species. In Holland, breadth of diet was greater for coho than for trout i n a l l habitats, but particularly i n pools, due to the higher contribution of aerial foods i n the diet of the former. In general, trout ate a wider array of food items with less variation between taxa in the faster velocity habitats, the reverse being true for coho whose dietary breadth increased i n pools. 43 Table 4. Standardized breadth of diet for sympatric populations of coho and cutthroat trout for 14 food taxa i n two streams, August-September, 1973. Sample size Pools Glides Riffles Average Bush Creek Trout Coho 30 29 0.21 0.29 0.19 0.17 0.16 0.12 0.19 0.19 Holland Creek Trout Coho 25 25 0.33 0.67 0.48 0.54 0.41 0.51 0.41 0.57 Table 5. Overlap i n diet between sympatric populations of coho and cutthroat trout by habitat type i n two streams, August-September, 1973, Bush Creek Holland Creek Pools 0.89 0.41 Glides 0.98 0.67 Riffles 1.0 0.90 Average 0.96 0.66 44 Diets of coho and trout overlapped considerably, but more so i n Bush than i n Holland (Table 5) . In both streams, interspecific overlap i n diet was lowest i n pools and highest i n r i f f l e s , ranging from 0.89-1.0 i n Bush and from 0.41-0.90 i n Holland. In r i f f l e s , the stomachs of both salmonids contained predominantly benthic foods (e.g. chironomid larvae, pupae), whereas i n pools, trout contained more benthic and coho more aerial foods (Fig. 14). GROWTH OF SYMPATRIC COHO AND TROUT Summer growth of salmonid underyearlings was higher for trout than for coho i n both Bush and Holland creeks, 1974 (Figs. 15, 16). In Bush, mean fork length of trout sampled from a l l habitat types i n late June, was about 10 mm less than that of coho; by late September, although the range i n fork length was much greater for coho than for trout, mean length of both species was approximately 47 mm. Similarly, i n Holland, trout mean fork length was about 10 mm less than that of coho i n late June; by late September coho mean fork length was about 51 mm and trout about 45 mm. The growth,•(-insiength):"•<?£.• trout ;;was.similar'iiiT?bth• streams; however, i n Holland Creek, trout i n i t i a l l y were slightly smaller than, and coho grew slightly faster than - their conspecifics i n Bush Creek. These differences appear to account for the pattern of interspecific growth between these two streams. Throughout this three-month period, streamflow was similar and remained consistently low i n both streams. 45 MEAN FORK LENGTH (mm) STREAM DISCHARGE (m3/s) Fig. 15. Mean fork length ± 2 S .E. (horizontal lines) of unde'ryearling coho (®) and cutthroat trout (A) in Bush Creek, 1974. Vertical lines are size ranges: • , denotes stream discharge. 46 MEAN FORK LENGTH (mm) o o J O ) < o r o o i c » r l - ^ - > j o o j c_ c m C > c o c CO H in m H I I I I | I I I I | I ! I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | 11 I 11 _^ h«-|-/ / • 1—•-\ D 1 • I / / / • , . 1 • I I * . » . - L - l - 1 i • . — 1 • i -n>—r-• ; r- s !-•> i • I-II l l l l l III M M I M M I I M i l I I 1 I 1 I I I I I I * I I I I I • • I " O D o o p p p p p p p p — ro oi b i O > ->i oo <o STREAM DISCHARGE (m3/s) Fig. 16. Mean fork length of underyearling coho and cutthroat trout and stream discharge i n Holland Creek, 1974. See Fig. 15 for details. 47 DISCUSSION RESOURCE PARTITIONING Habitat i s an important dimension of resource partitioning among animal populations i n general (Schoener 1974) . Resources are usually not distributed independently of habitat type, but rather the two are. often highly correlated (Werner and Hall 1977) . The present findings indicate that habitat i s the major means of resource partition-ing between sympatric populations of juvenile coho salmon and coastal cutthroat trout i n small streams. During summer, although they over-lap considerably, populations of these two salmonids spatially segregate i n a pattern similar to that reported for sympatric coho and steelhead trout i n larger coastal streams of Briti s h Columbia (Hartman 1965b) . Both Hartman's and my study show that coho dominate i n pools, while steelhead trout and cutthroat trout dominate i n r i f f l e s . Glides or open channels are generally areas of greatest interspecific overlap, with the degree of overlap depending i n part, on relative and absolute densities of populations. Krebs and Wingate (1976) also noted that the habitat use by species of small mammal communities i n the Kluane region, Yukon, changed from year to year with their relative abundance; as population density increased, they moved into more habitat types resulting i n greater interspecific overlap. Resource partitioning pre-dominantly by means of habitat segregation has been documented for other combinations of sympatric salmonids both in streams (Lister and Genoe 1970; Everest and Chapman 1972; G r i f f i t h 1972) and i n lakes (Andrusak and Northcote 1971), and for centrarchids i n lakes 48 ,(Werner. et a l . 19.77),~ and for cyprinids in a stream (Mendelson 1975). Habitat segregation presumably attenuates interspecific competition for resources in general and possibly for food in particular, as shown by Nilsson (1963) for lake dwelling populations of char (Salvelinus alpinus) and trout (Salmo trutta) in Sweden. Like.many stream salmonids, coho and cutthroat trout appear to lack marked or effective interspecific differentiation in their morphology that would allow distinct trophic separation. Their diets show broad over-lap within habitat types in this study and elsewhere (Mason and Machidori 1976). Thus food resources appear to be divided between these two salmonids primarily through partitioning of the available stream space as found i n brook trout and cutthroat trout in some Idaho streams ( G r i f f i t h 1972). However, their diets within habitat types (this study) suggest that coho might be a more specialized feeder of drifting foods and cutthroat trout a more generalized feeder of both benthic and drifting foods. Such interspecific behavioral differences, however, may not be effective i n shallow water habitats due to the lack of adequate vertical separation between species. In other aquatic habitats, partitioning of food resources i s commonly along a combination of resource dimensions rather than a single resource axis, as demonstrated for numerous sympatric species, notable examples being those for intertidal starfishes (Menge and Menge 1974), marine inshore percid fishes (Bray and Ebeling 1975) and lake-dwelling centrarchid fishes (Werner et a l . 1977). Some authors (Morse 1974; Wilson 1975) are of the opinion that 49 subdominant species usually have a broader niche than do socially dominant ones. The possibility of greater niche p l a s t i c i t y among socially subdominant species may be a manifestation of their d i f f i c u l t y i n obtaining an adequate share of available resources. The a b i l i t y of subdominant species to shift to habitats providing optimal foraging yield during periods of superabundant resources, allows them to exploit opportunistically resources from which they are excluded at other times by the socially dominant species. I found that cutthroat trout consistently occupied a broader micro-habitat niche than coho i n each of the three sympatric streams. In addition to u t i l i z i n g the faster velocity habitats, smaller cutthroat trout commonly inhabited shallow waters of stream margins and small side channels (unpublished data). This study and previous studies (Hartman and G i l l 1968; Hall and Lantz 1969; G r i f f i t h 1972; Mason and Machidori 1976) suggest that cutthroat trout may u t i l i z e a variety of microhabitat types, ranging widely i n water velocity, depth and temperature, and i n available cover. Niche breadth of this species appears to assure maintenance of population i n highly fluctuating environments. However, breadth of food niche for cutthroat trout did not exceed that for coho i n two of my study streams, paralleling results by others (Mason and Machidori 1976) . Werner and Hall (1977) also found that the bl u e g i l l sunfish occupies a broader microhabitat niche, but lesser food niche, than the socially dominant green sunfish i n ponds and lakes. Populations of allopatric cutthroat trout u t i l i z e d pools more 50 so than r i f f l e s , possibly due to the absence of coho. I speculate this reflects species behavioral diversity i n habitat response to environmental in s t a b i l i t y , rather than that of genetically based behavioral differences between populations of trout above and below a barrier f a l l s , due to downstream gene flow across such barriers. Habitat shift by cutthroat trout from pools into r i f f l e s when i n sympatry with coho corroborate the interactive segregation concept advanced by Nilsson (1956), and parallel his findings for lake populations of trout and char i n Sweden (1960, 1963). Nilsson ascribes niche shift i n the more plastic, sub-dominant char, to both exploitation and interference competition by the more t e r r i t o r i a l and aggressive trout. Coho may exert a similar behavioral influence on cutthroat trout i n pools and other low velocity habitats. The influence of sculpins on the pattern of microhabitat use by salmonids was not documented i n this study. Spatially, they over-lapped extensively with salmonids and were abundant i n a l l habitat types, although their absolute biomass was generally highest i n pools due to the preponderance there of larger individual sculpins. Sculpins showed no consistent use of microhabitat relative to salmonids between the three study streams; i n Bush and Holland creeks their biomass was high i n a l l habitat types, whereas i n Ayum Creek i t was highest i n pools and lowest i n r i f f l e s . Sculpin breadth of microhabitat niche was slightly less than that of trout i n Holland and Ayum, possibly due to the preponderance of recruits and mature sculpins, respectively. However, the extensive dietary overlap of trout and sculpins (Mason and Machidori 1976) suggest they may be competitors i n small streams for a common food 51 resource - the benthic invertebrate community, as these workers pointed out. LIMITING FACTORS: FOOD OR SPACE? Based on the difference i n biomass between sympatric and allopatric populations of cutthroat trout, I postulate that inter-specific interaction may be limiting sympatric populations of trout, although total fish production may be greater i n multi-species streams. Mean total f i s h biomass levels i n the sympatric populations ranged from about 2-9 g/m2 (av.= 5.1 g/m2), being lowest i n r i f f l e s and highest i n pools, with trout almost exclusively contributing less than 1 g/m2 i n a l l habitat types. In allopatric populations, however, mean biomass density of trout ranged from about 1-6 g/m2 (;av.= 2.2 g/m2) being lowest i n r i f f l e s and highest i n pools. In populations isolated from coho salmon and sculpins, cutthroat trout had up to a tenfold higher biomass in some cases. As to which of the two species, coho or sculpins, may have a more negative impact on cutthroat trout biomass levels i n streams remains unknown. In this context i t i s important to consider whether sympatric trout populations might be food or space limited. Stream production of juvenile coho during summer has been shown to be limited by food rather than space (Mason 1976) . These findings may not be applicable to salmon-trout-sculpin communities, particularly for trout populations due to their much later time of emergence than salmon. In the case of anadromous cutthroat trout, they 52 emerge into a stream environment that may already be f i l l e d to carrying capacity by coho fry, considering the high rates of coho fry emigration and instream mortality (Mason 1975), and loss of rearing habitat with receding streamflows i n summer. Trout are largely restricted to r i f f l e areas during summer-early f a l l , the habitat -that .is usually low i n abundance relative to pools during this period of low streamflow. I surmise that habitat segregation may be lessened further as summer temperatures increase, disadvantageously to trout, as coho use of r i f f l e s may increase due to their higher swimming performance i n warmer water (Glova and Mclnerney 1977). Thus, the availability of l i v i n g space for sympatric trout populations may be reduced considerably by coho during the seasons of best growth, which may i n part explain the low biomass levels of trout i n sympatric streams. Ofiinthe.eothe.r :hand, trout populations in small streams may be l i t t l e affected by direct competition for food with coho, despite their relatively broad overlap i n diets, as the trout are found predominantly i n the food-producing r i f f l e areas. Sculpins are abundant i n a l l habitat types i n streams and often achieve biomass levels much greater than that of sympatric coho and trout combined. Numerous studies, including the present one, have converged on the generalization that sculpin biomass levels are both higher and more variable than those of salmonids, with ranges extending from 25-90% of total f i s h biomass i n any one stream (LeCren 1965; Mann 1971; Petrosky and Waters 1975; Mason and Machidori 1976). If sculpins have any negative impact on production of stream salmonids, I suggest that food, rather than space i s involved. Their benthic and cryptic habits (hiding beneath rocks) appear to minimize interaction with salmonids through vertical 53 separation i n most stream habitats, a -possible exception being that of stream edges. Sculpin distribution showed no consistent evidence of spatial segregation with salmonids i n Bush and Holland creeks, but salmonids may have reduced sculpin abundance i n r i f f l e s of Ayum Creek. In general, sculpins appear to be distributed i n a pattern more related to intra - than interspecific interaction, with larger individuals most common i n deeper, slower velocity areas and juveniles i n shallower, faster velocity habitats. While sculpin predation on stream salmonid fry populations appear to be of minor importance (see review by Moyle 1977), that of possible competition for food may not, and has been frequently suggested by numerous authors (Brocksen et a l . 1968; Andreassen 1971; Mason and Machidori 1976). Conceivably, under high population densities and low invertebrate production, sculpins might significantly reduce the availability of food for salmonids. The possible significance of sculpins as a competitor for food with salmonids was not documented i n this study. Range i n mean biomass levels of allopatric trout populations (1.0-5.6 g/m2) approximated those of sympatric trout and coho combined (0.4-4.2 g/m2) . Based on my f i e l d evidence that cutthroat trout have a similar but broader ecological niche than do coho, trout biomass i n allopatry ought to be about the same as the summed biomass of trout and coho i n sympatry, a l l other factors being equal. Further, assuming stream carrying capacity for salmonids to be similar above and below the barrier f a l l s , these biomass comparisons may suggest that sculpin spatial and feeding niches overlap l i t t l e with those of salmonids. The significantly higher ranges of mean fis h biomass i n sympatric populations (2.0-9.0 g/m2) than those i n allopatric trout populations (1.0-5.6 g/m2) may reflect higher productivity i n downstream reaches,.and/or more efficient use of the stream environment, through sculpin exploitation of a niche not u t i l i z e d by salmonids. The ecological role of sculpins i n these simple f i s h communities awaits further definition under more rigorous experimental conditions than those available presently. 5 5 CHAPTER I I I . LABORATORY STUDIES OF STREAM POPULATIONS OF UNDERYEARLING COHO SALMON AND COASTAL CUTTHROAT TROUT INTRODUCTION My f i e l d investigations showed that during the late summer period of low streamflows, sympatric populations of coho salmon and cutthroat trout broadly overlapped i n diet but segregated into habitats of relatively slow and fast water velocity, respectively, with the pattern of separation depending i n large part, on absolute and relative densities of their populations. Hartman (1965b) reported similar seasonal segregation between sympatric populations of coho and steelhead trout, which he experimentally demonstrated to be of the interactive type (Nilsson 1956), occurring i n summer but not i n winter. Such segregation may attenuate interspecific competition for resources i n general, and possibly for food i n particular, during the season of relatively high population density. Distinct spatial segregation but broadly overlapping diets have also been reported for ecologically similar nonsalmonid fishes in streams (Gee and Northcote 1963; Gibbons and Gee 1972). Some advantages of segregation l i k e l y involve higher species overall growth and survival, and i n the case of anadromous salmonids, higher smolt yields to sea. In this study, laboratory populations of underyearling coho and cutthroat trout of natural sympatric origin were tested i n a stream 5.6 simulator to elucidate i n detail, not possible i n nature, pattern and mechanism of partitioning of food and space during summer and winter. Depending on season, I tested the nu l l hypothesis that there was no observable s t a t i s t i c a l difference between their pattern of micro-distribution and of intensity of aggression when exposed to different levels of 1) feeding activity, 2) water velocity, 3) water temperature and 4) relative density of species. Possible adaptive behavioral adjustment i n cutthroat trout, i f any, to li v i n g with coho, was also investigated by comparing trout microhabitat use and aggression for fi s h of sympatric (downstream of barrier f a l l s ) and allopatric (upstream of barrier f a l l s ) origin from two small coastal streams. METHODS A. THE TEST FACILITY The stream simulator used was the one described by Hartman (1965a), now located at the University of Victoria, B. C. Overall dimensions and construction are shown i n Fig. 17a, the volume of the actual experimental section being 5m long x 1.2m wide x 0.75m deep. Two modifications made to the basic f a c i l i t y involved replacing the up- and downstream nylon screens with stainless steel mesh (2.54 x 2.54 x 0.64 mm) and installation of incandescent lighting (12, 25W bulbs) with rheostat control under the ceiling-suspended fluorescent fixtures used by Hartman (1965b) . A r t i f i c i a l light intensity was roughly uniform over the length of the test f a c i l i t y , averaging 250 Lux along the centreline (range 235-DRIVE PULLEY R l PI R 2 >Q o oi o 0 o IO o ° o o l o o ' o°o° Fig. 17. a) Schematic of stream simulator (after Hartman 1965a) with heavy arrows indicating direction of stream flow; b) plan -view of the expermental section showing outline of rocks, logs, and undercut bank (stippled); r i f f l e s - Rl, R2; pools - P l , P2; c) side view showing physiographic profile of the experimental section. ~58 260 Lux) measured with a "Photovolt" model 210 photometer. Natural photo-period was provided through a bank of high windows running the f u l l length and directly opposite the experimental section. Water temperature was maintained to within ±0.5 °C by a refrigeration unit situated at the up-stream end of the test f a c i l i t y . On-off control of water circulation from the simulator through the refrigeration system was maintained by a thermoregulator and solenoid hookup to the recirculating pump, plus a series of gate valves which were manually operated. Water 'that had passed through the refrigeration system re-entered the simulator i n the downstream well, being thoroughly mixed i n the return flume by the drive propellor before entering the experimental section. The water supply was from the City of Victoria, dechlorinated by f a c i l i t i e s at the University of Victoria. Incoming water to the simulator was via a 3-cm PVC line, with ball-valve control, situated above the water surface and running the f u l l width i n the upstream well, with fine jets at 2.5 cm on centre directed downstream. Water i n the simulator was continuously renewed with a turnover cycle of 2 days. The experimental section consisted of four equal units pro-viding a duplicate pair of ri f f l e - p o o l sequences starting from the upstream end (Fig. 17b, c). The foundation for the stream-bottom profile was of f i r plywood prefabricated components, assembled on-site. Running the f u l l length of the right side (facing upstream) of the experimental section was the stream edge, 0.4m wide with a 1 i n 4 ri s e . The underside of this edge i n both pools was covered with a wood lath, and then mottled 5 9 . with an earthen colored mixture of fiberglass resin and sand to simulate texture of undercut bank materials. The outside surface of the glazing panels i n the undercut area of pools was covered with mottle-patterned heavy brown paper. A l l seams between plywood structures and walls of the experimental section were made fish-proof with narrow wood trim and caulking expansion sealant. The substrates of the experimental section consisted of boulders (30 cm and over), rubble (8-29 cm), gravel, sand and inorganic s i l t . The size composition'-.: and distribution of substrates in each of the four microhabitat types i n the simulator were arranged i n a manner resembling those i n the study streams (Fig. 17b, c): r i f f l e s contained rubble i n a staggered pattern, with each of the rocks slightly elevated over a shallow depression i n the streambed; pools contained boulders and gravel f i l l e d i n with fines and inorganic s i l t deposits at the head; undercut bank- and stream-edge areas contained a mixture of gravel and sand, with few rubble i n the latter areas i n r i f f l e s . The substrates on sloped surfaces were held permanently i n place with an earthen-colored fiberglass resin. A log about 0.15m 0 x 1.2m long taken from a stream was obliquely positioned i n each pool (see Fig. 17b) as further cover for f i s h . Two similar logs were also placed longitudinally over the steel frame superstructure of each pool as overhead cover. A darkened observation corridor of black polyethylene from floor to ceiling was provided on the l e f t side (facing upstream) of the simulator. Horizontal s l i t s i n the plastic on the simulator side, along with a wood platform of f u l l length and profile of that of the stream bottom, permitted 60/ observation into the experimental section without disturbing f i s h . A food-dispensing apparatus, one for each r i f f l e , was located i n the refrigeration bath of the test f a c i l i t y , providing controlled simulation of drifting food. Each unit consisted of a 10-1 plastic container, with a water submersible centrifugal-type pump (capacity 45-£ /min) on the bottom, partially submerged in the refrigeration bath. Openings near the bottom of the container covered with fly-screening provided an upwelling of incoming water, keeping food i n suspension. The flows of both pumps were equalized (ball valve and flowmeter) and separately connected to a 1.9-cm PVC pipe buried i n the gravel across the upstream end of each of the two r i f f l e s ; suspended food was released into the stream via fine exit jets directed obliquely upwards and down-stream. Positioned above each plastic container i n the refrigeration bath was a 500-m£ glass beaker with electric agitator and with micro-control (glass flowmeter and diaphragm valve hook-up) of incoming water supply. Food placed into the beaker was released into the stream via the exit jets at a rate dependent on the rate of water overflow from the beaker. B. THE FISH Coho and cutthroat trout fry of sympatric origin were from two small coastal streams situated at the south end of Vancouver Island (see Fig. 1): summer fish were from Craigflower Creek (F.L. range, 35-69 mm); winter f i s h were from Ayum Creek (F.L. range, 43-96 mm). They were 61 • collected with a D.C. f i s h shocker and/or pole seine, and transferred to the laboratory i n fry cans. The summer fis h were collected on the i n i t i a l day of each experiment. The winter fish were bulk-collected i n advance i n late November to avoid possible d i f f i c u l t y i n obtaining adequate numbers of f i s h i n freshets of streams later on. The winter f i s h were held i n laboratory f a c i l i t i e s at the Pacific Biological Station, Nanaimo, B. C. These consisted of a bank of twelve 60 x 60 x 30 cm clear plexiglass tanks with painted plywood covers and a black plastic shroud over the front side to minimize disturbing the f i s h . Incoming fresh water from an overhead mixing manifold was at about 2£/min, maintaining temperatures within 3-5 °C. Photoperiod was natural through a north-facing window with no a r t i f i c i a l lighting provided. Several 15-cm long half-sections of 9-cm diam PVC pipe were scattered on the bottom of each tank as cover for f i s h . The two species were held separately from one another, primarily to reduce handling time when selecting f i s h , each grouped into small- (F.L. range 47-60 mm), medium-(66-72 mm), and large-sized (76-96 mm) individuals per tank. They were fed chopped fresh-frozen euphausiids at least once every 2 days. C. EXPERIMENTAL PROCEDURE In the laboratory the selected fish were individually measured (fork length) and damp weighed under mild anaesthetic (2-phenoxyethanol) . Each test required 40 fi s h comprised of individuals of the large, medium, and small size-classes (6, 14, and 20 animals, respectively), as summarized i n the Appendix (Tables 9, 10). The relative density of each of the three size-classes was approximately proportional to that of the wild sympatric populations of coho and trout i n the three study streams. In tests with coho and trout mixed (in sympatry), the number of fish of each species i n each of the three size-classes was half of that when the species were tested separately (in allopatry) i n order to keep density constant. The selected fish i n each of the winter experiments were transferred i n fry cans at their acclimated test temperature, directly to the test f a c i l i t y at the University of Victoria on the i n i t i a l day of each test. A l l tests were duplicated. The experimental conditions, i n both the summer and winter tests including weight change data for each of the summer experiments, are shown i n the Appendix (Tables 9, 10) . The f i s h were given a minimum of 2-h recovery time from the effects of the anaesthetic and handling, i n well aerated water i n a 90-I dark plastic container with a cover. They were then released i n the centre of the test f a c i l i t y between 1600 and 1800 h under the available natural light i n s t i l l water, followed by i n i t i a t i o n of the low water velocity (25 cm/s) l4-h after their introduction. Each experiment lasted 1 wk. The fish were given 2 days habituation time to the test f a c i l i t y . Thereafter, observations were made at the low test velocity for a period of 2 1/2 days, followed i n .the remaining 2 1/2 days by.the high test velocity,- .-that^was' incrementally, stepped up-oyer-a":3-h period. Both the water temperature and the high water velocity levels differed between the two test seasons: summer fish were tested during the period June 2-September 16 at 13.0±0.5 °C and a high velocity averaging 43.1 cm/s i n r i f f l e s ; winter fish were tested during the period December 2-63. January 27 at both 3 and 5±0.5 °C and a high velocity averaging 50.7 cm/s in r i f f l e s . The increase i n velocity was intended to allow for the larger body size of the winter f i s h , compared to those used i n summer. Absolute velocity at specific sites within each habitat are summarized in previous reports (Glova and Mason 1977a, 1977b). Fish were fed twice daily a ration consisting of chopped fresh-frozen euphausiids, amounting to 5% of their body weight. The food was released as simulated d r i f t i n streams by the apparatus described earlier. Day length was natural with the a r t i f i c i a l lighting superimposed from about 0800-2000 h i n summer and 0800-1700 h i n winter. The timing of the routine daily observations on the positions and aggressive interactions of the f i s h was governed by the imposed feeding cycle: pre-feed period when no food was drifting i n the system; during-feed period began 15 min after i n i t i a t i o n of release of drifting foods; post-feed period began 30 min after the release of any drifting food was stopped. The observation schedule was repeated i n theivmorning and late afternoon, usually extending from 0800- dusk daily. The approximate horizontal and vertical (upper, mid and lower thirds) positions, size-class and species of each fi s h were recorded on outline maps of the stream bottom at each observation period. The aggressive behavior of a l l the fi s h i n each of the r i f f l e and pool sections was recorded for a period of 10 min, with each of the four sections chosen randomly. The behavioral components of aggressive encounters (both intra-and interspecific) of coho and of trout was similarly coded, and quantitatively recorded on a bank of four multiple key laboratory count 64 denominators. The behavioral categories recognized i n this study are described later. In experiments with coho and trout mixed, four possible types of interaction were recorded: coho-coho, coho-trout, trout-trout, trout-coho. These were e l i c i t e d either singly or in a sequence of behavioral events. At the end of each observation period, the information was decoded onto standardized data sheets. Fish mortality i n any one experiment rarely exceeded 5% and most often involved small individuals pinned against the downstream screen at night, particularly during the test period of freshet conditions. Mortalities were accounted for at the beginning of each day, the observations at a l l times reflecting the mean responses of the surviving f i s h . Dead fi s h were removed from the downstream screen during night hours to avoid disturbing the fish unduly. At the end of each experiment the tank was drained with most fis h retreating into the pools. They were dipnetted out and anaesthetized for determination of fork length and weight. D. PROCESSING OF DATA The microdistributions of coho and of trout i n the stream simulator were tested s t a t i s t i c a l l y by factorial analysis of variance. Interaction between test variables was investigated i n each analysis, the maximum number of variables consisting of habitat type, f i s h size, feed period, water velocity, water temperature, and species tested. To standardize the numbers of fis h in each of the three size-classes i n a 65 given habitat, each observation was expressed as a percent of the total fish of each size-class. The data were then transformed by the arcsine transformation (Sokal and Rohlf 1969) . S t a t i s t i c a l analysis was applied to determine i f the difference i n microdistribution was significant 1) between species when tested i n allopatry, 2) between species when tested i n sympatry, and 3) within species between allopatric and sympatric tests. For behavioral analysis, species individual components of aggression i n each observation were summed and divided by the number of fi s h observed i n order to standardize fish density. This provided a comparative measure of species rates of aggression. Size of fish was not considered i n the analyses, such data being available for allopatric but not sympatric t r i a l s , due to the lack of recording equipment necessary to include size i n the latter tests. The data were s t a t i s t i c a l l y tested by student-t and chi-square tests wherever applicable. The Mann-Whitney U-test (Sokal and Rohlf 1969) was used i n cases of nonparametric analysis. RESULTS I. COHO AND TROUT OF SYMPATRIC ORIGIN General In summer, partitioning of the available space in the stream simulator between coho and cutthroat trout fry was rather rapid and similar in pattern to that observed in nature. During the f i r s t day the fish showed a gradual spacing-out from their i n i t i a l aggregations in pools. Trout usually showed a stronger tendency to move upstream than did coho, particularly the smaller individuals. However, both salmonids 6,6-moved temporarily upstream when the water velocity was incrementally accelerated from the low to the high test level. This, i n nature, would seem to counteract downstream displacement during periods of high velocities and fluctuating flows i n streams. In winter, an active spacing-out of the fish through t e r r i t o r i a l i t y and social dominance,as i n summer, was largely limited to feeding period and restricted to low velocity sites for both salmonids. In nonfeeding periods both species tended to aggregate i n the areas of pools with over-head cover and minimal water velocity. Compared to the summer tests, their breadth of microhabitat use, general mobility and social interactions i n winter were low. The scope of these activities was highly influenced by water temperature and velocity. In winter, near-freezing water temperatures and high velocities i n the simulator restricted their exploitation of food and l i v i n g space almost exclusively to slow water areas, which may i n part, reflect their lowered metabolism and poorer swimming a b i l i t y at low temperature. Within the f i r s t two days of any one experiment, fish of either species frequently shifted position within habitat types. Such activity represented the period when individual territories and dominance hierarchies were unstable, and was more noticeable i n summer than i n winter. Pattern of habitat partitioning between species changed l i t t l e after the f i r s t two days, whereas that of their social interactions did, particularly i n summer (Fig. 18). Unlike coho, trout i n i t i a l l y actively defended both r i f f l e s and pools. Subsequently, aggressiveness markedly increased i n coho 67 IN T R ASPECIF IC INTERSPECIFIC X CO Z 10 u. o « 5 UJ CD —I) — \ — • _ l _ 4 I DAYS -4 Fig. 18. Summer aggression (a), and microdistribution (b), of coho (circles) and cutthroat trout (triangles) i n pools (open) and r i f f l e s (closed) during their i n i t i a l 4 days i n the stream simulator. Symbols are means ± S.E. 68. and decreased i n trout, with coho establishing social dominance and showing obvious priorities i n choice of space and feeding opportunities i n both r i f f l e s and pools, at least under the low flow conditions. By the end of the fourth day in each of the experiments, the social structure within each of the r i f f l e and pool habitats was relatively stable. Jenkins (1969) reported stable social structure between individuals of juvenile populations of brown trout and rainbow trout i n streams. Mean levels of aggression may have been slightly higher for trout and lower for coho i f the f i r s t 2 days of each experiment had been included i n routine observations (Fig. 18). As this was generally a period of in s t a b i l i t y i n the process of resource partitioning between the species, observations made during the f i r s t 2 days throughout the experimental series i n both summer and winter, were precluded from further analyses. Simultaneous replicate testing was not possible i n the apparatus used. In order to spread possible time effects evenly, a total of 3 wk lapsed between replicates. With increasing size and/or maturation, certain behaviors and environmental responses of the fish may have changed. Factorial analysis of variance indicated that significant (P<0.01) differences between replicates for either species were invariably that of interaction between habitat type and size of f i s h . Differences i n their mean body size between replicates were restrictedlbut unavoidable in summer, due to rapid growth. 69 MICRODISTRIBUTION Summer The summer microdistributions of coho and cutthroat trout fry tested i n sympatry, butnnot allopatry, showed distinct interspecific differences. Examples of cumulative plots of their actual positions i n the stream simulator for both allopatric and sympatric t r i a l s are shown in Fig. 19. Pooling a l l the data with respect to body size and feed periods (Table 6), the grand means of species numbers per habitat type i n sympatry were significantly (t-test, P<0.01) different at both low-and high test velocities. Through interactive segregation, under low velocity, the numbers of coho i n pools were about double that of trout, while the reverse occurred i n r i f f l e s . Under high velocity, their pattern of segregation was similar, but interspecific differences were less pro-nounced, primarily due to the trout's reduced use of r i f f l e s (Table 6) . In allopatry, species numbers were similar withinhhabitat types with differences ranging only from 1-6%. Expressed on a percent basis (Fig. 20a), i n allopatry, at the low test velocity approximately 40% of either species occupied r i f f l e s and 60% pools; an almost doubling of the velocity reduced their occupancy in r i f f l e s and increased i t i n pools by about 15%. Similarly, i n sympatry under low flow, rif f l e - p o o l percent ratios were about 23:77 for coho and 62:38 for trout. At high flow, trout occupancy decreased i n r i f f l e s and increased i n pools by approximately 12%, while that of coho was essentially unaffected i n both replicates. 70 162% 2 4 8 % 186% 40 4% TROUT a) COHO "0 -o* D 1 . * .0 s , » / s •'0 -0 -TO P om D O ° o-v y ,*Jo •• s ' o o» y o o / y y '<> y ' . • • •> . D 1 t'ry/izy. •y&''} '. . j . • • • D ^ R1 ^ 0 ° o. r 1 n £ r t y^"y': • •• • • ww*y • • »y • • • &*sK->&%'' * ' ' y yKA : • y.j<t.< 120% 32-2% 18-2% 37-6% 37-0% 13-3 % 2 7 0 % 22-7% T R O U T b) COHO ' v o-D 1 O 0 y* •o o-D O °Oo • ' o. o S Op,' ' ' ' o * y^ y^ y** •* D 1 y y y ' y ' n 1 o f o l 1-0 % r l o ' \ y>° n d *o- 0 ° ' o 0 0 13-2% P2 o - v '' y Syr-* y y^ * y ••tt *»: ' • • .. 4 0 - 4 % • / / . • / / ^ • . • -X • . ••*: 45-4% Fig. 19. Cumulative plots of the summer microdistribution (one replicate only) of coho and cutthroat trout during the pre-feed periods' i n a) allopatry and b) sympatry. The percent number of fish per habitat i s shown for each species. The open circles indicate position of f i s h above the undercut bank (i.e. floodbank area):. See Fig. 17 for details of the experimental section. 7 1 ' Table 6. Mean number of coho and cutthroat trout fry in the r i f f l e and pool habitats in summer at the two test velocities. The number of fish shown in sympatry are doubled that of the actual values in order to equalize density with that in allopatry. Cover in riffles refers1 to under rocks.; in- pools, to .undercut, areas. Low velocity High velocity Mean ± S .E % number of fish using cover Mean ± S.E, % number of fish using cover Allopatry Coho Riffle 7.3 + 0.55 0.0 5.1 + 0.34 0.0 Pool 11.8 + 0.51 2.4 14.4 + 0.38 4.8 Trout Riffle 7.4 + 0.43 0.0 5.7 + 0.42 0.0 Pool 11.5 + 0.36 10.1 13.5 + 0.74 10.5 > Sympatry Coho Riffle 4.6 + 0.23 0.0 4.7 + 0.31 0.0 Pool 15.1 + 0.32 7.3 13.8 + 0.45 8.2 Trout Riffle 10.8 + 0.22 1.5 9.0 + 0.31 0.0 Pool 6.3 + 0.23 11.9 8.9 + 0.47 19.7 72 A L L O P A T R Y R I F F L E POOL R I F F L E 100 r 80 r -60 U l o ac ui 0 . 4 0 40 60 80 100 POOL m SYMPATRY R I F F L E POOL o O _ l Iii > o o o —J U J > X C£ X R IFFLE o o —J U J > o o —J U J > X u> X POOL Fig. 20. Relative microdistribution of coho (solid) and cutthroat trout (open) i n allopatry and i n sympatry i n a) summer and b) winter. 73 ' Submerged areas of cover beneath rocks i n r i f f l e s and under-cut banks i n pools were not heavily u t i l i z e d by either coho and cutthroat fry i n summer (Table 6) . Smaller f i s h were the more frequent users of cover sites, often i n escape from aggressive encounters. In both salmonids, sites offering maximization of food-getting rather than over-head cover were generally more directly associated with territories of dominant f i s h . In r i f f l e s , coho were never found, and trout were rarely found, i n areas under cover. In pools, u t i l i z a t i o n of undercut areas ranged from 2.4-8.2% for coho and 10.1-19.71 for trout, both species showing slightly higher u t i l i z a t i o n during periods of accelerated flow and also when tested i n sympatry (Table 6) . Unlike i n the simulator, i n natural streams exploitation of drifting foods by f i s h with territories in undercut areas may be higher due to greater convergent flow at meanders. Factorial analyses of variance were conducted to determine s t a t i s t i c a l significance of each of the test variables in coho and trout microdistributions. The results of these are summarized i n Table 7. Of the possible combinations of interactions between a l l of the five test variables, only that of habitat type interacted significantly (P<0.01) with species and with size of fish i n a l l s t a t i s t i c a l tests, excepting i n the allopatric tests between species. When tested separately, the micro-habitat demands of these two salmonids were very similar for given size-classes . Comparisons of their relative microdistributions i n sympatry for second-order levels of interaction with habitat type (Fig. 21) indicate that 1) size of fish was the most important factor, 2) food supply was of intermediate importance, and 3) acceleration of water velocity was of least importance i n summer. 74 Table 7. Comparison of F-values (P<0.01 underlined) from factorial analyses of variance of the coho and cutthroat trout test series i n summer. Both allopatric and sympatric t r i a l s were tested between and within species. Test variables are H, habitat; Z, size; F, feed-period; V, velocity; S, species; E, experiment type (allopatry or sympatry). Variables dF Between species -. f Within species Allopatry Sympatry Coho Trout H 7 102.78 79.88 131.65 50.69 Z 2 3.32 2.33 2.10 2.60 H Z 14 0.73 9.26 10.30 9.30 F 2 0.79 0.06 0.85 1.02 H F 14 0.39 0.65 1.29 0.28 Z F 4 1.50 1.49 0.13 0.16 H Z F 28 0.67 0.50 0.72 0.74 S/E 1 2.67 3.87 6.47 0.71 H S/E 7 0.09 26.62 9.71 8.27 Z S/E 2 0.42 0.75 1.34 1.20 H Z S/E 14 0.31 2.63 2.37 2.20 F S/E 2 0.38 0.21 0.06 0.56 H F S/E 14 0.30 1.00 0.56 0.27 Z F S/E 4 0.38 0.44 0.74 0.54 • • H : . - Z ; " £ I S / E 28 0.42 0.66 0.45 0.40 V 1 0.46 2.25 0.53 0.75 H V 7 0.99 1.16 2.23 0.85 Z V 2 0.06 2.80 1.43 1.68 H Z V 14 0.31 1.32 0.99 0.94 F V 2 0.09 0.27 0.16 0.13 H F V 14 0.30 0.38 0.49 0.27 Z F V 4 0.34 0.45 0.27 0.38 H Z F V 28 0.08 0.48 0.31 0.40 V S/E 1 0.02 0.44 0.05 0.22 H V S/E 7 0.64 0.72 2.41 1.14 Z V S/E 2 0.70 1.56 0.44 1.10 H Z V S/E 14 0.84 1.18 0.79 0.80 F V S/E 2 0.05 0.09 0.02 0.24 H F V S/E 14 0.28 0.58 0.29 0.22 Z F V S/E 4 0.20 0.39 0.52 0.40 H Z F V S/E 28 0.38 0.44 0.41 0.42 Error 288/96 7 5 HABITAT vs SIZE OF FISH a) o >-O 3 0 Z U J 3 O 4 0 UJ K riffle 2 I I I 3 I I 6 HABITAT vs FEED-PERIOD pool 4 riffle 5 HABITAT vs WATER VELOCITY I KI 2 2 p i 0gg Hi ir m II 11 • •I I K a. b) III I 11 1 11 "1 K1 5 II i | r $ r i 'I II Fig. 21. Relative microdistribution of coho (solid) and cutthroat trout (hatched) in summer in a) sympatry and b) allopatry in-relation to size of fish (1, large; 2, medium; 3, small), feed-period (4, pre-; 5, during-; 6 , post-feed) and test velocity ( 7 , low; 8 , high). Open portion of bars refer to fish in undercut areas in pools. 76 Relative size of fish largely determined their p r i o r i t y of access to food and space. Fish i n r i f f l e s and at the heads of pools had feeding advantages over individuals in other areas as they were nearer to the incoming food source. In both coho and trout the pattern of habitat segregation into pools and r i f f l e s was influenced by size effects. For both species mean percent frequencies of fish in r i f f l e s were higher for the larger than for the small-sized individuals, the reverse occurring i n pools (Fig. 21) . Size effects on f i s h micro-distribution pattern i n allopatry were generally similar, but of lesser magnitude than i n sympatry. Further, their vertical distribution i n pools differed between size-classes. There was a significantly (P<0.05) higher number of small than of medium-large size fish for both species i n the lower level of pools i n both allopatric and sympatric t r i a l s , excepting i n the latter case for trout (Table 8) . The bottom and undercut areas of pools were common refuge sites for small f i s h , actively contained there by larger, socially dominant individuals. The food supply influenced pattern of microdistribution and revealed certain interspecific differences i n strategies of food exploitation. In sympatry, unlike trout, coho microdistribution showed a more pronounced association to the food supply: during feeding periods the numbers of coho i n r i f f l e s increased (Fig. 21a) with many establishing transient territories superimposed on trout from above (Fig. 22a); i n post-feeding periods there was a net return of coho from r i f f l e s into the pools. However, when pooled together, these movements between habitat t" types were generally nonsignificant (P>0.05). In allopatry, both species 77 Table 8. Vertical microdistribution of coho and cutthroat trout by size class (S, small; M, medium; L, large) i n pools, combining the data of pre-, during- and post-feeding periods i n summer at both test velocities. Differences (* P < 0 .05) i n the data between upper and lower levels were tested by factorial analysis of variance ( 2 X 2 design with a minimum of 25 observations per c e l l ) . The mean number of fish shown i n sympatry are twice that of the actual values i n order to equalize density with that i n allopatry. Coho Trout Upper Mid Lower Upper Mid Lower Allopatry S M § L 3.3* 1.7 3.7 4.5 8.0* 2.2 1.3* 0.9 3.1 6.4 8.3* 2.8 Sympatry S 1.4* 6.6 6.6* 1.8 2.4 1.8 M § L 1.2 7.2 4.8 0.9 2.8 1.4 78 b) Fig. 22. a) Common pattern of body alignment and vertical positioning of coho (upper) and cutthroat trout (lower) i n the stream simulator; b) cutthroat trout i n high intensity lateral threat posture. 79 showed similar microhabitat responses while exploiting the food supply as they did i n sympatry (Fig. 21b) . Feeding also affected f i s h microdistribution longitudinally within habitat types. In allopatry, during-feeding, medium-large size coho were present i n significantly (P<0.05) higher numbers than were small fish i n the upstream half of r i f f l e s (Table 9) . Trout showed a similar pattern, although one not s t a t i s t i c a l l y significant. In sympatry, similar size effects occurred during-feeding but trout were more numerous than were coho i n the upstream half of r i f f l e s (Table 10) . Analysis of the microdistribution plots i n allopatric and sympatric t r i a l s indicated that feeding territories of both salmonids i n r i f f l e s were most commonly situated upstream of rocks and least so alongside them (Table 11). An almost doubling of the water velocity did not significantly alter either species overall microdistribution pattern (Fig. 21): i n allopatry coho occupancy i n r i f f l e s was reduced by 31%, trout by 23%; i n sympatry, unlike trout, coho occupancy i n r i f f l e s actually increased sl i g h t l y , probably i n response to increased levels of aggression i n pools under the accelerated velocity conditions. Unlike trout, small sub-ordinate coho did not associate closely with bottom cover and frequently were actively chased out of pools by larger f i s h , particularly during post-feed periods. Winter The winter microdistributions of coho and cutthroat trout fry at 3 °C were f a i r l y similar i n both allopatric and sympatric tests, as 80 Table 9. Horizontal microdistribution (mean ± S.E.) o£ coho and cutthroat trout i n allopatry by size class (see Table 8) i n the upstream and downstream halves i n r i f f l e s and pools (upper third of water column) during-feeding period i n summer. The differences between size classes were tested by student t-distribution; - P< 0.05 .is underlined. R i f f l e segment Pool segment Upstream Downstream Upstream Downstream S M § L . S M § L S M U S M $ L Coho 1.3±.2 4.2±.5: 0.4±.1J: 1.5±.3 1.2±.3 ' Trout 1.5±.3 2.3±.3 1.5±.2 1.3±.l 0.8±.l 1.5±.l 2.1±.4^  0.8±.2 0.8±.2 0.8±.2 0.5±.2 .81 Table 10. Horizontal microdistribution (mean ± S.E.) of^sympatric coho (C) and cutthroat trout (T) for a l l size, classes combined, i n the upstream and downstream halves i n r i f f l e s and i n pools (upper third of water column) during-feeding period i n summer; P < 0.05, is underlined. R i f f l e segment Pool segment upstream Downstream Upstream Downstream Replicate 1 2.7±.5 5.2±.2 1.2±.3 1.6±.3 1.6±.3 0.9±.2 0.7±.2 0.7±.2 Replicate 2 1.8±.3 3.0±.2 0.6±.l 2.3±.l 0.5±.l 1.0±.2 0.6±.2 0.8±.2 2.3 3.2 0.9' 2.0 1.1 1.0 0.7 0.8 82 Table 11. Percent number of coho and cutthroat trout with respect to their position to rocks i n r i f f l e s (DPS, upstream of; DNS, downstream of, and ALS, alongside of rocks) during-feeding period i n summer. Coho Trout UPS DNS ALS UPS DNS ALS Allopatry 79.0 12.6 8.4 60.4 22.9 16.7 . v. Sympatry 72.1 19.2 8.7 58.8 31.5 9.7 83 shown by examples of cumulative plots of their actual positions (Fig. 23). There was no obvious species interactive effect on their distribution i n the ri f f l e - p o o l space, other than that total number of f i s h i n pools was slightly higher i n sympatry than, i n allopatry, and that trout were more numerous than coho i n undercut areas i n pools when tested together. Mixing the species had the effect of slightly increasing the number of fi s h i n pools. Pooling the data with respect to body size and feed-periods (Table 12), the grand means of species numbers per habitat type showed no significant (t-test, P>0.05) interspecific differences i n both allopatry and sympatry at either test velocity. Mean numbers of coho i n pools were slightly higher than those of trout, ranging from 19.6-20.0 and from 18.0-19.2, respectively. On a percent basis (Fig. 20b), more than 981 of the coho occupied pools whereas trout ranged from 88-97%, being highest under the accelerated velocity conditions when the species were mixed. Increased velocity had vi r t u a l l y no effect on the coho's overall pattern of microdistribution. Six-way factorial analyses of variance were computed on the block of winter data to determine s t a t i s t i c a l significance of interactions between the test variables.T/. The results of these are summarized i n Table 13. Overall, habitat type interacted significantly (P<0.01) with a l l test variables i n either intra- ior interspecific cases, excepting that of feed-period. Relative differences i n the quality of cover between r i f f l e s and pools (i.e. under rocks or within undercut bank areas, respectively) were probably the main determinants of f i s h microhabitat use i n winter, which appeared to be interrelated to a number of other factors. In pools, the use of cover by both coho and trout decreased 84 11-4% 50-3 % 1-3% 370 % TROUT a) COHO ° O O 0 o - o o • •• • /• A A- A' « X A ' ' X A • mm •• • , Rl PI R2 , P2 ° o ~o ' * • ' • .•* V . * ^ v ° o ° o 0 o n o 0 ° o o d / • A. . •PsVU . • • • • •' X * X / • X^ x^ * x 54-3% 0-3% 45-4% 0-5% 50-8 % 0-3% 48-7% O Q O DX"O O <V R1 ; • .4 * V ' * .v X>' X ft °o ° o 0 o n o 0 O <D • • • •• -* • " / A X . < D 1 A A •• r\ i ° o o r l •• .y.lA R2 ° O 0 0 0 o n o o o «o P2 . x n. ' / . ' V • • / . / * X % * ' V X • / • x * • *x ^x • yA A TROUT b) C0H0 58-6 % 41-4 % Fig. 23. Cumulative plots, of the winter microdistribution (of one replicate only) of coho and cutthroat trout at 3 °C during the pre-feed periods i n a) allopatry and b) sympatry. See Fig. 19 for caption details. :85 Table 12. Mean number of coho and cutthroat trout fry i n the r i f f l e and pool habitats.in winter at the two: test velocities. The number of fis h shown i n sympatry are doubled those of the actual values i n order to equalize density with those i n allopatry. Low velocity High velocity % number % number of fish of fish using using Mean ± S.E. cover Mean ± S.E. cover Tested at 3 °C Allopatry Coho Ri f f l e 0.5 + 0.1 4.4 0.4 + 0.1 11.5 Pool 19.7 + 0.6 30.9 19.6 + 0.8 39.3 Trout Ri f f l e 2.3 + 0.2 69.2 2.0 + 0.2 54.8 Pool 18.0 + 0.6 49.2 18.0 + 0.7 51.2 Sympatry Coho Ri f f l e 0.2 + 0.0 0.0 0.2 + 0.0 0.0 Pool 20.0 + 0.3 41.6 20.0 + 0.2 47.5 Trout Ri f f l e 1.3 + 0.1 48.9 0.6 + 0.1 28.3 Pool 19.0 + 0.5 72.4 19.2 + 0.4 69.8 Tested at 5 °C Sympatry Coho Rif f l e 4.0 + 0.3 0.0 2.2 + 0.3 0.0 Pool 15.2 + 0.4 15.7 18.4 + 0.4 23.4 Trout Ri f f l e 5.5 + 0.3 42.4 5.0 + 0.3 28.1 Pool 15.2 + 0.4 54.4 15.0 + 0.6 63.7 "86 Table 13. Comparison of F-values (P<0.01 underlined) from factorial analyses of variance of the coho and cutthroat trout test series i n winter. Both allopatric and sympatric t r i a l s were tested between and within species. Test variables are H, habitat; Z, size, F, feed-period, V, velocity; T, temperature; S, species; E, experiment type (allopatry or sympatry). Variables dF Between species Within species Allopatry Sympatry Coho Trout H T 293.69 311.29 364.24 242.43 Z • "2 • • 1.74 8.67 1.72 6.52 H Z 14 31.03 28.83 38.63 19.04 F 2 0.14 2.01 0.08 0.78 H F 14 3.67 7.90 6.35 4.02 Z F 4 0.06 0.63 0.07 0.05 H Z F 28 0.23 1.87 0.83 0.47 S/E 1 6.45 2.49 1.48 9.14 H S/E 7 10.44 81.53 6.02 19.88 Z S/E 2 2.05 0.45 0.91 0.59 H Z S/E 14 7.52 3.93 6.51 2.25 F S/E 2 0.01 0.95 0.06 0.29 H F S/E 14 0.34 1.23 0.31 0.80 Z F S/E 4 0.03 0.43 0.07 0.15 H Z F S/E 28 0.20 1.23 0.46 0.65 T 1 6.13 H T 7 36.33 Z T 2 0.24 H Z T 14 5.29 F T 2 0.33 H F T 14 1.40 Z F T 4 0.41 H Z F T 28 0.79 S T 1 1.18 H S T 7 6.48 Z S T 2 0.02 H Z S T 14 2.63 F S T 2 0.04 H F S T 14 1.21 Z F S T 4 0.20 HZ F S T 28 1.09 V 1 1.15 6.26 2.17 0.82 H V 7 18.44 21.72 24.91 19.36 Z V 2 0.55 0.10 0.53 0.52 H Z V 14 1.77 1.97 1.23 3.22 F V 2 0.41 0.47 0.09 0.32 87 Table 13 (cont'd) Variables dF Between species Within species Allopatry Sympatry Coho Trout H F V 14 0.59 1.21 0.74 0.31 Z F V 4 0.13 0.34 0.15 0.27 H Z F V 28 0.19 0.33 0.27 0.39 V S/E 1 0.0 0.95 0.16 0.01 H V S/E 7 0.65 10.84 3.65 3.61 Z V S/E 2 0.06 0.64 0.06 0.05 H Z V S/E 14 1.82 1.73 2.50 2.08 F V S/E 2 0.07 0.54 0.07 0.07 H F V S/E 14 0.14 0.58 0.15 0.49 Z F V S/E 4 0.24 0.34 0.09 0.06 H Z F V S/E 28 0.24 0.58 0.37 0.20 T V 1 0.87 H T V 7 1.14 Z T V 2 0.32 H Z T V 14 2.49 F T V 2 0.15 H F T V 14 0.36 Z F T V 4 0.20 H Z F T V 28 0.60 S T V 1 0.19 H S T V 7 5.61 Z S T V 2 0.65 H Z S T V 14 1.31 F S T V 2 0.63 H F S T V 14 0.60 Z F S T V 4 0.08 H Z F S T V 28 0.43 Error 576/288 88. with decreasing size of f i s h , decreasing water velocity, increasing water temperature and when food was drifting i n the system (Fig. 24) . The use of cover i n r i f f l e s (under rocks) was almost exclusively by that of small and medium size trout. Comparisons of the relative micro-distributions of coho and trout i n sympatry for second-order levels of interaction (Fig. 24a) indicated that temperature was the primary factor controlling salmonid use of space i n winter. Relative size of f i s h , i n -creased water velocity within limits and the food supply were of lesser importance, ranked i n that order. Temperature was the major determinant of coho and trout breadth of microhabitat niche i n winter. A relatively small elevation i n temperature resulted i n pronounced differences i n f i s h microdistribution (Table 12; Fig. 24a) . At 3 °C both coho and trout fry did not u t i l i z e r i f f l e s to any great extent but remained i n pools under cover. At 5 °C their use of r i f f l e s increased (P<0.05) i n a l l tests excepting that by coho at the high velocity. With the 2 °C rise in water temperature, mean numbers of fish i n r i f f l e s increased by about 20% for coho and 15% for trout under low flow conditions; rif f l e : p o o l percent ratios of the means were about 21:79 for coho and 27:73 for trout. At high flow, coho occupancy decreased i n r i f f l e s by approximately 10% relative to that at low flow, whereas that of trout essentially remained unchanged. Interspecifically, mean numbers of fish per habitat type differed significantly (P<0.05) only at the high velocity conditions, with numbers of trout i n r i f f l e s being more than twice as high as for coho. Fig. 24. Relative microdistribution of coho (solid) and cutthroat trout (hatched) i n winter i n a) sympatry and b) allopatry i n relation to size of f i s h (1, large; 2, medium; 3, small), feed-period (4, pre-; 5, during-; 6, post-feed), water velocity (7, low; 8, high) and water temperature (9, at 3 °C; 10, at 5 °C). See Fig. 21 for additional details. 90 Priority of access to food and space was largely determined by an individual's relative body size. For both salmonids size of fish interacted significantly (P<0.01) and similarly with habitat type i n both allopatry and sympatry. The larger-sized coho and trout were rarely found i n r i f f l e s but preferred the deeper water with overhead cover i n pools (Fig. 24) . Cover response also differed slightly between species, being more pronounced i n trout than i n coho i n both r i f f l e s and pools (Table 12) . In allopatry, the percentage of coho using cover in r i f f l e s and pools, respectively, ranged from 4.4-11.5 and from 30.9-39.3, whereas that of trout ranged from 54.8-69.2 and from 49.2-51.2. Within the depressions beneath each of the rocks i n r i f f l e s , there was almost never more than one fish per rock. In sympatry, the use of cover increased slightly i n pools but not i n r i f f l e s . When the species were mixed, coho use of cover i n pools ranged from 41.6-47.5% and trout from 69.8-72.4%. The microdistribution of both coho and trout i n pools differed significantly (P<0.05) with size of fish and with vertical strata i n both allopatric and sympatric t r i a l s (Table 14). In the lower third of pools, small f i s h were more numerous than were medium and large size individuals. Also, fish of a l l size-classes were most abundant i n the lower two-thirds of pools. Their relative abundance i n these lower strata depended on water velocity. From the low to the high test velocity, the numbers of both species for a l l three size-classes increased significantly (P<0.05) i n the lower third of pools. Both coho and trout fry showed a reduced feeding response at 3 °C. 91 Table 14. Vertical microdistribution of coho and cutthroat trout by size class (see Table 8) in pools, combining the data of pre-, during -and post-feeding periods i n winter. Differences in the means between a l l upper and lower levels by size class were significant (P < 0.05) when tested by factorial analysis of variance. The mean number of fish shown i n sympatry are twice those of the actual values in order to equalize density with those in allopatry. Coho Trout Upper Mid Lower Upper Mid Lower r) Allopatry low velocity s 0.5 8.8 9.6 0.7 9.7 4.5 M t\ L 0.1 13.1 6.8 0.1 13.0 6.5 Total 0.6 21.9 16.4 0.8 22.7 11.0 high velocity S 0.0 4.3 14.7 0.4 5.5 11.1 M t\ L 0.1 11.8 8.0 0.1 9.6 9.5 Total 0.1 16.1 22.7 0.5 15.1 20.6 Sympatry low velocity S 0.9 7.0 12.6 0.3 15.0 8.8 M § L 0.0 13.4 5.7 0.1 14.3 5.3 Total 0.9 20.4 18.3 0.4 29.3 14.1 high velocity S 0.0 4.4 15.4 o\o 8.1 10.5 M t\ L 0.0 12.4 7.4 0.0 11.6 7.9 Total 0.0 16.8 22.8 0.0 19.7 18.4 92 Accordingly, the imposed cycle of food availability had minimal impact on species microdistribution patterns, both i n allopatry and sympatry. However, a 2 0 increase i n temperature (3 to 5 °C) altered behavioral responses to food and space appreciably. At 3 °C, neither species exploited the food supply i n r i f f l e s but remained primarily i n pools. At 5 °C, the relative abundance of fish i n r i f f l e s was higher (Fig. 24), primarily due to their movement into this habitat during feeding period. However, none of these behavioral differences showed any significant interactions between feed periods and patterns of space u t i l i z a t i o n when tested by factorial analysis of variance. Under c r i t i c a l l y low temperature and high velocity conditions, trout appear better adapted to feed i n r i f f l e s than do coho possibly due to hydromechanical advantages gained by their closer association with the stream bottom. Coho feeding i n r i f f l e s under severe physical conditions consisted of short-term invasion of choice feeding sites . During winter freshets, stream salmonids may experience a net downstream displacement. To counteract this phenomenon, fish may instinctively move upstream. I compared the numbers of coho and trout i n the upstream- and i n the downstream halves of the experimental section for tests conducted at 3 °C (Table 15) . The numbers of f i s h for both species for a l l three size-classes, increased significantly (P<0.05) i n the upstream half when the velocity was increased, i n both allopatric and sympatric t r i a l s . In addition, the small f i s h of both species were more abundant i n the upstream half than were the medium and large f i s h combined. In nature, such longitudinal distribution may .permit a size segregation, with smaller f i s h i n upstream reaches and larger f i s h i n .93 Table 15. Horizontal microdistribution (mean ± S.E.) of coho and cutthroat trout by size class (see Table 8) in the upstream half (Rl + PI) and the downstream half (R2 + P2) during-feeding period i n winter. The number i n sympatry are twice that of the actual values to equalize density with that i n allopatry. upstream half Downstream half Coho Trout Coho Trout M § L S M § L S M § L S M U Allopatry  low velocity 11.0±0.4 8.1+0.7 12.2±0.3 10.9±0.4 8.4±0.4 13.1±0.3 7.6+.0.4 11.4±0.5 high velocity 16.5±0.5 11.3±0.6 15.5±0.4 13.7+0.5 3.1±0.4 8.7±0.7 4.4±0.4 6.3±0.5 V Sympatry low velocity 10.4±0.3 8.8±0.2 5.6±0.3 6.6+0.2 9.6±0.3 11.0±0.2 14.4±0.3 10.9±0.5 high velocity 16.5±0.2. 9.0±0.2 14.6±0.3 13.4±0.2 3.5±0.2 11.2±0.2 5.5±0.3 9.2±0.5 94 downstream reaches. AGGRESSIVE BEHAVIOR General Coho and cutthroat trout fry used similar body postures and movements in social interactions, as previously described by other workers for stream-dwelling Salmonidae; lateral and frontal threat dis-plays (Fabricius 1953; Kalleberg 1958; Chapman 1962); intention movement, chasing, threat and contact nips, and wig-wag threat display (Hartman 1965b; Mason 1969); parallel-swimming, c i r c l i n g and biting (Mason 1969). Of these, only the lateral threat display showed apparent differences between species. F i r s t l y , duration of intraspecific displays was generally longer i n trout than in coho and frequently involved either singly or i n sequence, parallel swimming, c i r c l i n g , intense nipping and biting of the opponent's peduncular region. Secondly, cutthroat trout possess a bright orange-colored hyoid slash, which i s exposed when the basihyal apparatus i s lowered in bouts of high intensity: 1 lateral threat aggression (see Fig. 22b), and i s accompanied by rapid quivering of the caudal region. Its adaptive significance is uncertain but i t appears to function as an intraspecific signal between contesting f i s h . Size and color intensity of the hyoid slash may also.be important. Particularly i n summer, intra-specific lateral threats between closely matched trout often led to pro-longed bouts of butting and biting usually near the bottom of pools, occasionally to a state of physical exhaustion. In an extreme case, a total of 530 aggressive acts over a period of 12 min, mostly intense nipping and biting, was exchanged between two trout i n a t e r r i t o r i a l dispute. In 95 contrast, interspecific lateral threat encounters were brief even i n cases when the fi s h were closely matched. Individual components of aggressive behavior were expressed on a percent basis of the pooled data for either species (see Fig. 26-27, bottom). In both summer and winter, the most frequently used components of aggression by both coho and cutthroat trout were those of chase, nip and lateral display; together these comprised more than 80% of the total aggressive acts for each species. In summer, the more elaborate threat displays and non-contact behaviors were more frequently used by coho as opposed to the predominant nipping behavior by trout. Of their total aggressive acts, nipping activity made up 45% for trout and 33% for coho. Neither species showed obvious differences i n frequencies of display and non-display forms of aggressive activity between the r i f f l e and pool environments when tested under the low and high velocity con-ditions . Hartman (1963) reported that young brown trout used less display than non-display forms of aggression when i n faster water. In winter, fi s h activity was relatively low and confined almost entirely to pools. Accordingly, they showed a proportional reduction in relative frequency of chases and an increase i n threat nips and displays compared to those i n summer. Lateral displays and threat nips combined made up almost 60% of the total aggressive encounters for both coho and trout. Rate of aggressive activity in both coho and trout was roughly proportional to body size at either test velocity (Table 16) . Aggression was least for small fish and highest for large fi s h , the differences ranging from three to -six fold i n magnitude. During high water velocity, 96 Table 16. Mean level of aggression (number of encounters per f i s h per 100 min) of coho and cutthroat trout i n allopatry by size class (see Table 8) of f i s h i n pools i n winter at 3 °C, combining the data for pre-, during- and post-feeding periods. P l and P2 as i n Fig. 17. Coho Trout M L S M Low velocity P l 6 31 57 22 45 49 P2 11 21 47 10 37 54 Average 8.5 26 52 16 41 52 High velocity Pl 8 19 50 13 31 81 P2 _8 29_ 47 13 39 63 Average 8 24 49 13 35 72 97 large trout were considerably more aggressive i n the upstream pool, which may be a response to counteract downstream displacement i n streams. Summer Habitat had greater effects on species levels of aggressive-ness i n sympatry than i n allopatry (see Figs... 25,. 26)-.. : Mixing the . \. two species i n a r i f f l e and pool environment had the overall effect of reducing coho aggressiveness i n r i f f l e s and trout i n pools. In sympatry, intraspecific aggression i n coho was significantly (P<0.01) higher i n pools than i n r i f f l e s , the pattern being reversed i n trout but significant (P<0.01) only under the low test velocity conditions. When the data were pooled for^both-te'st velocities. /(to'taUi observation .period. 2400 rnin), intra-specific aggressive activity i n pools and r i f f l e s , respectively, was 2152 and 515 for coho, and 242 and 703 for trout. However, their inter-specific aggression between pool and r i f f l e environments was similar for either species totalling 677 and 477 for coho, and 435 and 618 encounters for trout. In to t a l , coho aggressive activity was some 30% higher than that of trout. In allopatry, at low velocity overall levels of aggression i n either salmonid were similar between pools and r i f f l e s ; at high velocity both species were socially less active i n r i f f l e s than i n pools. Their aggressive activity i n pools and r i f f l e s , respectively, amounted to 3225 and 2022 encounters for coho, and 2326 and 2054 for trout. Rate of aggression i n both salmonids showed a definite relation to the feeding cycle, i n both allopatry and sympatry (Figs. 25, 26) . Typically, mean levels of aggressiveness peaked i n both r i f f l e and pool environments when food was drifting i n the system. However, chi-square "98 Fig. 25. Mean aggression ±S.E. of allopatric coho (solid) and cutthroat trout (open) i n a) summer and b) winter (3 °C) i n relation to the feeding cycle (1, pre-; 2, during-; 3, post-feed period) and water velocity. ,99. Fig. 26. Upper: Mean summer aggression of sympatric coho (solid) and cutthroat trout (open) i n r i f f l e s and pools. Numbers relate to the feed cycle as i n Fig. 25. Lower: Relative frequency of the components of aggression i n intra- and interspecific cases for coho (solid) and trout (open). Symbols are: IM, intention movement; DT, drive toward; CH, chase; TN, threat nip; CN, contact nip; L, lateral display, WW, wig-wag display; F, frontal display; PS, parallel swimming; C, ci r c l i n g ; B, biting. 100 I N T R A S P E C I F I C I N T E R S P E C I F I C I— R I F F L E — , |— P O O L — ) I— R I F F L E —| f—- POOL —i 2 101 tests were not significant i n a l l cases, when combining the data of both replicates. In allopatry, only the aggressiveness of trout i n pools rose significantly (P<0.01) i n relation to feeding at both test velocities (Fig. 25a). In sympatry, interspecific levels of aggression for both coho and trout showed a significant (P<0.01) increase when feeding i n pools only at high velocity (Fig. 26, upper). In r i f f l e s , rate of coho intra- and interspecific aggression peaked significantly (P<0.01) when feeding at both test velocities; trout aggressiveness increased significantly (P < 0.01) only against coho for the high velocity conditions. With the onset of feeding, aggressiveness was more rapidly elevated i n coho than in trout; coho actively penetrated r i f f l e s , exerting social dominance and largely displaced trout from the better feeding territories. Unlike i n r i f f l e s , i n pools trout appeared to be less vigorous competitors against coho, as suggested by the latter's significantly (P < 0.01) higher aggressive activity against members of their own species than against trout at both test velocities. The nearly twofold acceleration of water velocity did not appreciably affect species levels of aggression. In both allopatry and sympatry, velocity effects on fish aggression were similar i n r i f f l e s , as level of aggression i n both species decreased although non-significantly (P>0.05); i n pools, aggressiveness i n coho but not trout, increased significantly (P<0.05) when the velocity was accelerated, paralleling results reported for Atlantic salmon (Kalleberg 1958) . Winter The winter test conditions showed marked but similar .102 environmental effects on both coho and cutthroat trout patterns of aggressive behavior. Neither salmonid defended r i f f l e space at 3 °C, irrespective of the test conditions, with the exception that trout occasionally interacted for cover sites beneath rocks. Mixing the species showed no interactive effect on their levels of aggression. Overall, total aggressive encounters for each species i n sympatry was proportionately halved of that i n allopatry with respect to actual f i s h densities used, being 3410 and 6409 for coho and 1988 and 3892 for trout. However, rate of aggressive activity differed significantly (P<0.05) between species i n sympatry but not i n allopatry, but only during periods when food was drifting i n the system. When mixed with coho, trout defended feeding stations less and usually remained more i n areas of cover. Accordingly, coho directed a near two-fold greater amount of their total aggressive activity against conspecifics than against trout, with total encounters amounting to 2243 and 1167, respectively. Trout total aggressive activity was more evenly distributed with a total of 944 encounters against conspecifics and 1044 against coho. As i n summer, i n winter, rate of aggression i n both salmonids showed a definite relation to the feeding cycle i n both allopatry (Fig. 25b) and sympatry (Fig. 27). In general, aggression was low for both species but pulsed i n synchrony with the feeding cycle. Despite the rigorous test conditions, f i s h actively competed for food i n pools as portrayed by their significant (P<0.01) increase i n rate of aggression when food was drifting i n the system. Typically, aggression i n both species was lowest i n pre-, highest i n during-, and intermediate i n 103 so i 40 SO CO Ii . 20 I CO oz io | 10 I I N T R A S P E C I F I C •RIFFLE—, I POOL 2 12 3 12 3 I I T o o o -I o o Ul > I N T E R S P E C IFIC •RIFFLE—i I POOL 12 3 12 3 I I I I 40 T N Fig. 27. Aggression of sympatric coho and cutthroat trout i n winter at 3 °C (upper) and relative frequency of the components of aggression (lower) . Symbols as i n Fig. 26. 104 post-feed periods. However, coho maintained a higher level of aggression i n post-feed periods than did trout, but significantly (P<0.01) only intraspecifically. Aggressive response i n pools i n relation to the feeding cycle differed markedly between species (Fig. 28). When mixed, unlike trout, coho total aggressive activity showed a rapid i n i t i a l increase, reaching peak levels shortly after i n i t i a t i o n of the simulated d r i f t . Trout response to food was slower and less intense, with peak levels of aggressiveness being less than half of that for coho and lagging behind by some 15 min. Coho feeding strategy showed obvious advantages over that of trout. Theiriiimore rapid response gave them priorit y to choice sites, permitting a greater take of the limited food supply. The overall effect of increased aggressiveness i n both species when feeding, tended to disrupt aggregations i n the preferred cover sites and led to a size-related longitudinal and vertical partitioning of open pool space, with only slight increase i n numbers i n r i f f l e s . Typically, the larger f i s h were positioned near the head and i n the upper level of the pools, with coho most often i n front of and above trout. Increasing the water velocity had insignificant (P>0.05) effects on species levels of aggressiveness at 3 °C i n both allopatric and sympatric t r i a l s , as fi s h remained predominantly i n pools close to cover. With a doubling of the water velocity, rates of aggression i n coho, but not trout, decreased slightly for both intra- and interspecific cases, particularly i n the latter (Fig. 27, upper). Trout aggression intraspecifically,. was unaffected by the acceleration of the water JL05 75 70 CO c c Q: 6 5 ui ^ 60 ° 55 U J 50 Ul > 45 CO CO UJ 40 t r CD © 35 U- 30 o Ul CD z z < Ul 25 20 15 h-10 \ - 4 \ _L PRE-4 5 DURING-9 10 II 12 P O S T - F E E D T I M E P E R I O D ( 1 0 min in terva l ) Fig. 28. Winter aggression (mean number of encounters per fish per 100 min) of coho (0) and cutthroat trout ( • ) i n pools i n relation to the feed cycle for the high test velocity at 5 °C. Duration of each observation period was 5 min at each successive 10 min intervals. Vertical lines indicate range. 106 velocity but increased considerably interspecifically when food was drifting i n the system. The latter was mostly the response of highly t e r r i t o r i a l and aggressively active large-sized trout during feeding. Behavioral interactions increased when water temperature was elevated by 2 °C (Fig. 29, upper) . Both species at 5 °C actively defended r i f f l e s at least during feeding. With this relatively small rise i n temperature, their use of cover decreased while that of feeding increased. Total aggressive activity i n both pools and r i f f l e s combined, increased by 14% for coho and 50% for trout i n response to the 2 0 increase i n temperature. However, levels of aggression did not differ significantly between species for any of the test conditions, excepting a significantly (P<0.01) higher rate of intraspecific aggression for coho than for trout i n pools at the low test velocity. At least i n pools, species patterns of aggressiveness to the various test conditions at 5 °C was i n general similar to that at 3 °C, but more pronounced. Coho but not trout, showed a significant (P<0.05) decline i n both intra- and interspecific rates of aggression under the accelerated velocity conditions i n r i f f l e s , as well as pools. Under the low velocity conditions, i n pools but not r i f f l e s , coho directed a significantly (P<0.05) greater portion of their aggressive activity against con-specifics than against trout, respectively, totalling to 2184 and 1059 encounters. Trout offensive activity was more evenly distributed within and between species i n both r i f f l e s and pools. 10? INTRASPECIFIC INTERSPECIFIC I RIFFLE—, , POOL 1 ( — RIFFLE—, , POOL , Fig. 29. Aggression of sympatric coho and cutthroat trout i n winter at 5 oc (upper) and relative frequency of the components of aggression (lower). Symbols as i n Fig. 26. .10.8 RELATIVE DENSITY OF SPECIES Sympatric populations of juvenile coho salmon and cutthroat trout i n streams spatially segregate into pools and r i f f l e s , respectively, during the seasons of rapid growth. The degree of overlap i n micro-habitat use between populations of these salmonids may i n part reflect species relative density effects. The possibility of greater intra-specific competition for food and space under relatively high population density may force a species to exploit a broader microhabitat niche. I tested this possibility for summer fry of coho salmon and cutthroat trout i n the stream simulator at 14 °C, looking specifically at patterns of microhabitat use and rates of aggressive activity (see Appendix, Table 9, for the relative numbers and size of fish used i n each experiment). Habitat segregation was less distinct when species relative densities were grossly different from 1:1. In the 1:3 coho:trout density situation, trout overlapped considerably with coho i n pools, i n the reverse experiment, coho overlapped with trout i n r i f f l e s more so than i n 1:1 situations (Fig. 30, bottom) . Rates of intraspecific aggression i n either species were similar, being positively proportional to their relative density and probably served to increase dispersal between habitats (Fig. 30). Intra-specific aggression increased at least three-fold for the relatively high density test i n either species. Coho and trout differed markedly in level of interspecific aggressions accompanying density changes. 109 a ) b) I N T R A S P E C I F I C _ l I 1 L Rl PI R2 P2 Rl PI H A B I T A T T Y P E R2 P2 Fig. 30. Mean ±S.E. of aggression and of microdistribution of sympatric coho (circles) and trout (triangles) i n summer i n pools (Pl, P2) and r i f f l e s (Rl, R2) for the low (open) and high (closed) water velocity at two different "species relative densities: a) 10 coho, 30 trout; b) 30 coho, 10 trout.. -110. Unlike trout, coho upheld social dominance at both high and low densities, and their aggressive activity was at least 6 times higher i n the case i n which trout was high rather than low i n numbers. In coho, but not i n trout, rate of interspecific aggressive activity was related to probability of encounter. I I . COMPARISON OF ALLOPATRIC AND SYMPATRIC TROUT TYPES Comparisons were made of the summer microdistribution and aggression of allopatric and sympatric cutthroat trout population types when tested i n the stream simulator. Trout of allopatric origin (F.L. range 35-60 mm) li v i n g isolated from coho were from.the. area upstream of the barrier f a l l s i n Shawnigan Creek (inlet); trout of sympatric origin (F.L. range 35-53 mm) li v i n g together with coho were from Craigflower Creek. I postulated that the microhabitat niche and mode of feeding of these two populations might differ, due to the influence of coho i n one and not the other. Each trout type was tested separately i n two replicates, using the routine experimental procedures previously described to test the n u l l hypothesis: There was no observable difference between their rates and quality of aggressive behaviors and their microhabitat use when exposed to routine levels of (1) feeding activity, and (2) water velocity, as described earlier. Microdistribution Allopatric and sympatric trout population types showed similar - I l l inicrodistributions i n the stream simulator. Pooling the data with respect to body size and feed-periods, either population type showed similar densities i n r i f f l e and pool habitats (Table 17). At the low test velocity approximately 40% of the f i s h occupied r i f f l e s and 60% pools; with an almost doubling of the velocity, their r i f f l e occupancy decreased i n favour of pools by some 25%. In five-way factorial analysis of variance only habitat interacted significantly (P<0.01) with fish size for both population types. The simulated food supply and water velocity showed no significant interaction with f i s h micro-habitat use. The relative microdistributions for both trout population types indicated that (1) size of fi s h was the most important factor, (2) simulated food supply was of secondary importance, and (3) acceleration of the water velocity was of least importance i n summer. Relative size 'seemed; t6.~de to food and space for both types of trout. Trout positioned i n r i f f l e s and at the heads of pools had feeding advantages over individuals i n other areas of the simulator. For both population types, mean percent frequencies of f i s h i n r i f f l e s were slightly higher, although non-significantly so (P>0.05), for the larger than for the small-sized trout, whereas i n pools there was a preponderance of small fish at the bottom and in undercut areas. Similarly, the food supply stimulated comparable feeding responses i n both types of trout. During feeding periods, many actively penetrated into r i f f l e s and either established transient feeding territories superimposed on territories of resident trout, or displaced .112 Table 17. Mean number of cutthroat trout fry of allopatric and of sympatric origin, tested separately i n the r i f f l e and pool habitats i n summer at the two test velocities. Cover i n r i f f l e s refer to areas beneath rocks; i n pools to undercut areas. Low velocity High velocity Mean ± S.E. % number of fish using cover Mean ± S.E. % number of fi s h using cover ; Allopatric R i f f l e 7.7±0.38 2.0 5.7±0.48 4.2 Pool 11.5±0.42 7.3 13.6±0.84 13.0 ; Sympatric R i f f l e 7.4±0.43 0.0 5.7±0.42 0.0 Pool 11.5±0.36 10.1 13.5±0.74 10.5 113 some residents into pools. In post-feed periods there was typically an influx of transient riffle-dwellers back into pools, causing a net out-movement of previously displaced trout back into r i f f l e s . None of these movements between habitat types i n food exploitation were s t a t i s t i c a l l y significant (P>0.05) when pooled in each experiment. Submerged areas of cover beneath rocks in r i f f l e s and undercut banks i n pools were seldom used by either trout type. Small fish were the most frequent users of cover, often to escape aggression from larger f i s h . In both allopatric and sympatric trout, sites most opportune for feeding rather than for cover were more directly associated with territories of dominant f i s h . In r i f f l e s , u t i l i z a t i o n of cover was rare, not exceeding 4%. In pools, the use of cover was slightly higher and similar for both trouts, ranging from 7-13% with the higher levels of use occurring during periods of high velocity (Table 17). Aggressive behavior Allopatric and sympatric trout types used the same signal set of social interactions described earlier. The most frequently used behavioral elements were those of chases, nips and lateral displays, which comprised about 85% of their total aggressive activity in r i f f l e s , with the same i n pools for sympatric trout but slightly less for allopatric trout (Fig. 31). However, allopatric trout chased and threat-nipped less, but used lateral threat, c i r c l i n g and biting more than did sympatric trout. Total aggressive activity in both the r i f f l e and pool environments combined was similar for both trouts, amounting to 4298 acts for allopatric and 4380 acts for sympatric trout over a period of >-o z UJ Z> a m ac u. UJ o cc UJ 0_ 30 i -20 h-20 h 30 L Fig. 31. Upper: summer aggression (mean ±S.E.) of cutthroat trout of allopatric (heavy line) and of sympatric (light line) origin i n r i f f l e s and pools, tested separately i n the stream simulator. Lower: relative frequency of the components of aggression for trout of allopatric (solid) and sympatric (open) origin. Symbols are as i n Fig. 25, 26. 115 observation totalling 2400 min for each. Habitat had similar but greater effects on levels of aggression i n allopatric than i n sympatric trout: total aggression for allopatric trout between pools and r i f f l e s differed significantly (P<0.05) from those expected being 2602 and 1696, but not for sympatric trout, being 2326 and 2054. In pools, total aggression was about 12% higher for allopatric than for sympatric trout, whereas i n r i f f l e s total aggression was about 21% higher for sympatric than for allopatric trout. Rates of aggression i n sympatric trout showed a more definite relation to the feeding cycle (Fig. 31, upper) than i n allopatric trout. Aggression of sympatric trout was highest i n both r i f f l e s and pools when food was present, although significant (P<0.05) only i n the latter. Aggression of allopatric trout was inconsistent i n relation to the feeding cycle, peaking as often when food was present as when food was absent. Aggression decreased for both types of trout when water velocity was increased, except for the significant (P<0.05) increase by allopatric trout i n pools. However, the latter may not be representative of the population per se, as the data include an atypical case of intensive and extended aggression between two closely matched individuals. For either of the two trout types, the total number of aggressive acts was con-siderably less i n both r i f f l e s and pools during the high test velocity, with a maximum threefold reduction for sympatric trout i n r i f f l e s . 116 DISCUSSION BEHAVIORAL ECOLOGY OF COHO AND CUTTHROAT TROUT Juvenile coho salmon and coastal cutthroat trout are potential competitors for food and space during the summer season of low stream flows. Segregation, either selective (Brian 1956) or interactive (Nilsson 1956), i s one means by which competition between species might be attenuated. My laboratory findings confirm f i e l d observations. Sympatric coho and cutthroat trout segregate into pools and r i f f l e s through social interaction, the degree of overlap depending on relative and absolute densities of population. These findings parallel the pattern and mechansm of segregation between sympatric populations of juvenile coho and steelhead i n spring and summer (Hartman 1965b) . Segregation appears to be primarily the outcome of interspecific differences i n behavior. When together, the socially dominant coho more frequently defend pools, whereas trout more frequently defend r i f f l e s . In addition, habitat shift by either cutthroat or steelhead trout from pools (their preferred space) into r i f f l e s , when i n sympatry with coho, may il l u s t r a t e exploitation and/or interference competition by coho, a phenomena that has been demonstrated for sympatric populations of salmonids (Nilsson I960;- 1963; Andrusak and Northcote 1971) and of centrarGhdlds(Werner and Hall 1977) . In a competitive context, coho may feed more effic i e n t l y i n pools and trout i n r i f f l e s . Quantitative evidence for this i s lacking, but the more frequent occurrence of slightly higher growth for coho than trout i n the stream simulation experiments i n summer, suggest that coho may have taken a greater share 117 of the food supply than did trout (see Appendix Table 9). Underyearlings i n sympatric populations of coho salmon and coastal cutthroat trout did not segregate distinctly i n winter as i n summer, but rather coexisted i n pools. Sympatric coho and steelhead trout have also been found.to remain i n the unsegregated state i n pools in winter (Hartman 1965b). Both Hartman's and my findings suggest that relatively low level of aggression and slight interspecific differences i n their microhabitat demands in winter are important factors in the co-existence of coho and trout i n pools. In particular, trout associate more closely with bottom cover than do coho i n pools. In streams, however, this appears to be the niche of older trout: age 1+ rather than age 0+ trout are the more common cohabitants with coho i n pools; age 0+ trout are more frequently found near the edge i n the shallower, faster waters, containing an abundance of both instream and overhead cover (Glova and Mason 1977b) . The apparent lack of segregation between coho and trout i n the stream simulator i n winter may reflect the age 0+ trout's occupancy of a vacant niche, otherwise occupied by their age^1+ conspecifics i n nature. In the simulator, age 0+ trout were behaviorally flexible, and ut i l i z e d cover i n both r i f f l e s and pools, but "preferred" the latter. Unlike i n summer, in winter, quality of space rather than food supply appears to be of greater importance to juvenile coho salmon and cutthroat trout in small streams. Sufficient cover may be a key require-ment for salmonid residency i n streams in winter (Bjomn 1971; Bustard and Narver 1975a; Mason 1976) . In the stream simulator, the need for 118 cover and deeper water i n winter of both coho and cutthroat trout was illustrated by their predominant use of the undercut areas i n pools. Such knowledge of seasonal spatial preferences by stream salmonids has been known for some time. Some three centuries ago, Isaak Walton (1676) said of brown trout to his good friend Charles Cotton, while fishing i n one of his favorite trout streams i n England: "- - - and you are to take notice, that the f i s h lies or swims nearer the bottom, and i n deeper water, i n winter than i n summer; and also nearer the bottom i n any cold day, and then gets nearest the lee side of the water". It has been shown experimentally that both coho and trout prefer areas of cover rather than areas of no-cover when tested under semi-natural conditions (Bustard and Narver 1975b) . Low water temperatures and high stream flows i n winter create sufficiently adverse conditions that f i s h tend to exploit areas of shelter and rest. Possible adaptive significance of this behavior could involve reduction of downstream displacement and predation (Hartman 1965b) during a period of lowered metabolism, reduced food requirements and poor swimming a b i l i t y . Quality and quantity of cover may be the regulatory factor i n overwintering stream salmonid populations as suggested by Mason (1976) for juvenile coho. While adverse conditions i n natural environments may be tempered by greater spatial complexity, cover i n the simulator was intentionally kept simple to f a c i l i t a t e observation of the f i s h . My laboratory findings, however, agree with those of natural populations of sympatric coho and cutthroat trout. Coho and cutthroat trout interact minimally over space per se in winter. Typically, rates of interaction were positively related to 119 temperature but inversely to velocity. Hartman (1966) observed similar responses to temperature i n juvenile coho and steelhead. Food drifting i n the system at periods of dawn and dusk i n the present study markedly elevated species levels of interaction i n pools, but e l i c i t e d only minor dispersal from their preferred winter habitat. Intense inter-action was typically short term, and waned rapidly to relatively low levels i n post-feed periods, paralleling results reported for Atlantic salmon (Keenleyside and Yamamoto 1962) . Level of aggression and scope i n microhabitat use for both coho and cutthroat trout were more strongly influenced by change i n temperature than that of water velocity. The pronounced species responses obtained i n raising the temperature from 3 to 5 °C, relative to that arising from a doubling of the water velocity from 25-50 cm/s, suggest that minor differences i n temperature at the lower end of the temperature scale can be of significant ecological importance. The role of low winter temperature as a factor controlling microdistribution of juvenile coho and steelhead i n streams was demon-strated by Bustard and Narver (1975a). As both thermal and hydrological conditions i n streams are commonly severe i n winter, and drifting foods may be sparse, I infer from the present findings that wild sympatric populations of coho and cutthroat trout interact minimally during winter, despite their similar microhabitat demands. However, i n streams with restricted overwintering cover they may compete for preferred spaces through mere physical occupancy of specific sites. Factors affecting the microdistribution of coho and cutthroat trout i n the stream simulator differed seasonally i n rank, of their importance. In summer, size of f i s h , the simulated food supply and 120 acceleration of the water velocity within limits, were ranked i n decreasing order of importance as affecting fish microdistribution. In winter, temperature appeared to be the primary factor controlling u t i l i z a t i o n of space; relative size of fi s h , acceleration of the water velocity within the imposed limits and the food supply were of lesser importance, ranked i n that order. Size-related differences i n use of space minimize potential for social interactions, both intra- and interspecific (Everest and Chapman 1972). In summer, space i n the stream simulator was partitioned longitudinally in r i f f l e s and both longitudinally and vertically i n pools, with the larger fish being more common at the upstream half i n each habitat type. In winter, partitioning of space was vertical and confined to pools, with both species aggregated i n the lower two-thirds and small f i s h being more common on the bottom. Laboratory studies (this and Hartman 1965b) have shown that the r i f f l e environment i s a refuge from interspecific competition for underyearlings of either cutthroat or steelhead trout in summer when tested i n sympatry with coho salmon. Behavioral differences between these salmonids appear to account for resource partitioning in streams. In contrast, genetically-based differences rather than interspecific interaction appear to account for resource partitioning between sympatric populations of coho and chinook (Lister and Genoe 1970; Stein et a l . 1972) and of steelhead and chinook (Everest and Chapman 1972) in streams. Populations of coho, cutthroat and steelhead trout occurring sympatrically, are common in coastal streams i n Brit i s h Columbia. I speculate that for such 121 populations coho would maintain social dominance i n pools and that the trout would partition the available r i f f l e habitats longitudinally, overlap being greatest i n the mid-region of the system: steelhead i n the deeper, lower reaches; cutthroat i n the shallower, upper reaches, small tributaries being common. Such distribution patterns for sympatric populations of steelhead and cutthroat have been reported to occur i n numerous southwestern Brit i s h Columbia streams (Hartman and G i l l 1968). Exploitation by cutthroat trout of the more marginal habitats i n streams also containing steelhead suggest the former are li k e l y to be socially subdominant. IMPLICATIONS FOR MANAGEMENT OF STREAM SALMONIDS Habitat diversification The findings of the present study point to the importance of maintaining adequate habitat diversity i n streams i n summer when managing for sympatric populations of salmon and trout. Low habitat diversity may favour one species over">thefothe'r. ' During this season streams typically have reduced riffle/pool ratios and elevated stream-water temperatures, conditions which tend to minimize effective spatial segregation between salmon and trout. Low summer stream-flows offer competitive advantages to salmon over trout, despite the broader spatial and feeding niches of the latter. The cutthroat trout i s a polytypic species potentially capable of survival over a relatively broad range i n stream water temperatures and velocities (Hall and Lantz 1969; Glova and Mason 1977c) . Juvenile coho salmon are considered equally flexible to a variety of stream conditions but make restricted 1-22 use of higher velocity habitats, particularly at lower temperatures. When such habitat is sparse, coho may socially control access to marginal stream space and trout survival may be reduced. Certain velocity and substrate-oriented instream engineering (Parkinson and Slaney 1975) implemented i n specific streams would encourage habitat segregation between sympatric salmon and trout populations, simultaneously improve their food supply, and probably enhance their production and smolt yield to sea. Winter cover Winter carrying capacity of some salmonid-producing streams may be limited by the level of appropriate cover available to fis h (BustardSand. Narver -1975a;feMaTS0h"i976-}. The findings of the present study emphasize the importance of cover to stream salmonids at temperatures below 3 °C, a range that i s common i n streams of coastal Br i t i s h Columbia during winter. Streams managed for production of sympatric populations of coho salmon and coastal cutthroat trout should provide sufficient optimal cover types appropriate to age 0+ trout and also to coho and age 1+ trout. Age 0+ trout are most frequently found near the edge i n the shallower, faster waters, containing an abundance of large boulders and thick, low, overhanging shrubs at the streambank. They are almost never present i n Whitewater areas lacking such cover types (Glova and Mason 1977b) . Enhancement of age 0 trout may be most effectively achieved by manipulating both instream and overhead cover, particularly those of the larger sub-strates and streambank vegetation (see Parkinson and Slaney 1975) . 123 Analysis of overwintering requirements of coho and age 1+ trout is confounded by broad overlap and diversity i n cover types used: coho u t i l i z e a variety of cover types i n both main channel and side channel habitats (Bustard and Narver 1975a), whereas trout remain mostly i n main channel areas. Sites.common' to both species are the deeper waters con-taining upturned or undercut root masses and log accumulations at meanders. Trout, but not coho, are also found i n close association with large boulders. Strategies to improve winter cover for these salmonids should consider the entire river course and include existing hydrological and physiographical features (sidepools, side channel, backwaters at meanders, etc.) Such an approach may well reduce installation and maintenance costs and increase u t i l i z a t i o n of existing sites by f i s h . Superimposition of coho on trout The behavioral similarity of allopatric and sympatric trout types, may reflect a general environmental similarity above and below barrier f a l l s . Within the populations investigated in this study, inter-action with coho salmon has not produced any "apparent" evolutionary changes i n feeding behavior and microhabitat responses of trout. However, sympatric trout defended r i f f l e territories more vigorously, showed a more synchronous response to the feeding cycle, and used aggressive display components more suited hydro-mechanicaliy to faster velocity habitats than did allopatric trout. These differences could be inter-preted as adaptive responses to sympatry with coho. Any evolutionary changes i n sympatric trout populations attributable to their interaction with coho would face dilution from downstream gene flow from allopatric populations above barrier f a l l s . Until the magnitude of downstream 124 displacement from isolated trout populations, relative to size of the sympatric receiver population has been documented, especially as instigated by severe winter freshets, the potential importance of this genetic dilution factor w i l l remain unknown. In contrived sympatry such as that which would be produced from superimposition of hatchery-reared echo fry on wild allopatric trout populations, the interactive outcome may not differ appreciably from that for natural sympatry. The polytypic nature of trout populations i n general (Trojnar and Behnke 1974) would no doubt induce appropriate shifts i n feeding and microhabitat responses to allow for coho social dominance i n pools and other low velocity habitats. Assuming that coho and trout populations brought together i n an unnatural sympatry above a barrier falls# would segregate into pools and r i f f l e s , respectively, as i n natural sympatry, we might expect the biomass levels of such trout populations to decline to below 1 g/m2, or be approximately halved. Low summer flows disproportionately reduce r i f f l e areas relative to pools, and thus further extend space limitations to trout sympatric with coho, through habitat segregation. The present results suggest that superimposition on wild cutthroat trout stocks of cultured coho fry surplus to hatchery needs requires additional testing i n coastal streams under a variety of experimental conditions before recommendations can be made as to the s u i t a b i l i t y of such stocking as a strategy to enhance salmon stocks consistent with conservation of sympatric populations of trout. 125 Escapement control Stocks of anadromous cutthroat trout are a potentially valuable recreational resource (Giger 1972; Johnston and Mercer 1976). Enhancement of natural production of cutthroat trout smolts for stream populations sympatric with coho salmonv are l i k e l y to be limited by the trout's socially subdominant role i n allocation of stream resources. Coho escapement could be controlled i n selected sympatric streams particularly productive of cutthroat trout. Since coho spawn i n the f a l l and cutthroat trout spawn i n the spring, partial or complete ex-clusion of coho i n specific streams by escapement control i s feasible, using temporary stream barriers. Comparison of biomass density of allopatric and sympatric trout populations i n this study, suggest that the above management strategy would encourage the behaviorally flexible trout to approach production levels similar to those of sympatric coho and trout combined. The exclusion of sculpins i n such streams also merits consideration i n the context discussed by Mason and Machidori (1976) . -126 CHAPTER IV. LABORATORY GROWTH OF UNDERYEARLING COHO SALMON AND COASTAL OJTTHROAT TROUT INTRODUCTION Relative body size plays an important role i n competitive interactions between stream salmonids (Hartman 1965b; Mason 1965; Jenkins 1969). Larger individuals by way of social dominance have access to sites providing locally superior feeding and shelter conditions, which i n turn may better their probability of survival compared to that of smaller f i s h . Size-related competitive advantages may therefore be associated with accelerated growth. In the case of two closely related and ecologically similar species l i v i n g sympatrically, the pattern of interspecific growth may be clearly dominated by one of the species or the growth rates of both species may be similar. Social dominance is one factor that may influence the rate of growth of sympatric populations, particularly i n situations where food shortages occur. Stream populations of cutthroat trout fry face marked size disadvantages i n partitioning of resources with coho salmon. Size disparity between these two salmonids i s assured each generation by the trout's later emergence, and smaller size at emergence. Reduction i n interspecific size differences would increase the trout's competitive a b i l i t y and provide them a more equitable share of the available resources. 127 In Bush, and Holland creeks I observed that fry of cutthroat trout grew faster than fry of coho (see Figs. 15, 16). Faster growth for trout may have accrued from behavioral (e.g. lesser social dissipation of energy, greater food intake) and physiological (e.g. higher food conversion efficiency) strategies, or both. Laboratory growth studies were conducted to investigate the possibility of differences in pattern of growth between these two salmonids that might be of consequence i n their interspecific social interactions and partitioning of stream resources. In addition, my many casual observations of their distribution and abundance i n larger streams (coho being more common in the lower, warmer reaches; trout in the upper, cooler reaches), suggested coho may "do better" in warmer water, while trout may "do better" in colder water. I tested the n u l l hypothesis that there was no s t a t i s t i c a l difference in growth and body condition (Brown 1957) between underyearlings of coho and cutthroat trout for a seasonally representative range of combinations of temperature and photoperiod. Photoperiod was included in the design of the test space, as several works give evidence to suggest i t has important underlying interactions with temperature on,growth of f i s h (Brown 1957; Gross et a l . 1965; Huh et a l . 1976) and on f i s h movements i n streams (Northcote 1958). Hierarchical and density effects on growth were also investigated to a limited degree as these phenomena have important impact on relative growth i n laboratory studies (e.g. Brown 1946, juvenile brown trout). 128 METHODS A. THE FACILITIES Fa c i l i t i e s at the Pacific Biological Station, Nanaimo, B.C. were used. A total of nine troughs, each 2.42 x 0.89 x 0.14 m, were partitioned lengthwise into three identical compartments constructed entirely of 1-cm thick polyvinylchloride '5(PyC)'' • Each compartment was further subdivided across with removable nylon f l y screens into two equal experimental cells (total of 54 cells i n a l l ) each 1.09 x 0.29 x 0.14 m, and a small outflow chamber at the downstream end. These were placed on Dexion 225 steel angle frame and plywood shelving arranged to provide a 3 x 3 factor space having a total of nine test combinations of temperature (rows) and photoperiod (columns) .(Fig. 32). Each c e l l was equipped with a removable clear plexiglass cover with a 10-cm diam Nalgene feeding funnel at the upstream end, an airstone, and a 15-cm halfsection of 9-cm diam PVC pipe, placed centrally on the tank bottom as cover. A movable shroud of black drapery and black poly-ethylene sheeting was placed over the entry side of each of the three stacks to provide access and complete photoperiod isolation to each stack. Test temperatures of 5, 10 and 15 °C were manually maintained to within ±0.5 °C i n a l l but the summer 5 °C level (+1 °C) by manipulating b a l l valves controlling the delivery of chilled, normal and heated fresh water from header tanks (each 0.91 x 0.46 x 0.44 m) to three mixing manifolds. Water temperatures i n each manifold were continuously 129 a) I5k o o UJ or £ it>H or UJ a. UJ + HOURS OF LIGHT 11 b) 0 1 1 1 1 1 0 i i 1 1 1 i i i — i — i — —L_l—I ° 1 1 1 —1—1—1 1 o 1 M - t -i i - l - l - i .09 m HEADER TANK LIGHT SOURCE EXPERIMENTAL TROUGHS BLACK POLY AND BLACK CURTAIN 0.29 m [ • -SCREEN 1 !° ^ W A T E R INLET f [ OVERFLOW— 1 ! 1 ' O 1 c) Fig. 32. Details of the test f a c i l i t y used i n laboratory growth experiments: a) diagrammatic representation of the 9 test combinations of temperature and photoperiod including the 3-diagonal test points (along broken line) and 2 points tested i n sympatry i n winter ( - f - ) , b) cross-section of the entire test f a c i l i t y and c) plan detail of one test point unit. 130 monitored by Taylor manual recorders. The inflow to each c e l l from the respective manifolds was held at about 2 £/min (renewal rate once/ 45 min) and the water level constant at 0.13 m with a standpipe i n the outflow chamber. Test photoperiods of 8, 12 and 16 hf light at an intensity of 1 lux were obtained through time-switch control of incandescent overhead lights i n each test space. The f a c i l i t y provided a maximum of nine test combinations of temperature and photoperiod with three replicates per point for each of the two species tested,(Fig. 32a). B. THE FISH Comparable growth studies were done during both summer and winter; summer fis h were early fry (coho: FL' 42.9±1.3, range 38-55 mm; trout: FL 44.2±1.2, range 38-51 mm) from Craigflower Creek and winter fish were advanced fry (coho: FIT 56.7±1.1, range 51-62 mm; trout: FL 56.4±1.4, range 48-63 mm) from Holland and Bush creeks (see Fig. 1 for stream locations). Their mean i n i t i a l weights ranged from 0.7-0.8 g in summer and from 1.7-2.0 g i n winter. They were collected with a D.C. fish shocker and ;pole3 seine, transferred to the laboratory, and held i n fresh water at the respective ambient stream-water temperatures at time of collecting. C. EXPERIMENTAL PROCEDURE In the laboratory a l l individuals were measured and damp weighed 131 under mild anaesthetic (2-phenoxyethanol) and simultaneously sorted into 1-mm fork-length classes. Ten f i s h of each species for each experimental c e l l (0.25 fish/£ were then dipnetted from tanks containing individual length class i n systematic order to equalize mean sizes over the entire test space (see Appendix Glova and Mason 1976c). Each species was assigned alternate cells i n the test f a c i l i t y to account for any possible biases inherent i n the apparatus. They were then acclimated to their respective test temperatures at the rate of 1 °C change/day under natural photoperiod. When acclimation was completed i n a l l c e l l s , test temperatures and photoperiods were then imposed for the duration of the experiments. As a precautionary measure against possible diseases, a l l f i s h were i n i t i a l l y exposed to 1:4000 formalin solution for 20 min. The summer experiment lasted from 17 June to 8 September, 1975. Growth rates for both coho and trout i n allopatry were determined at a l l nine test points i n the factor space (see Fig. 32a) with three replicates/ point. The fish were individually measured to the nearest mm i n fork length but batch damp-weighed (10 fish of each species/cell) to the nearest 0.1 g at each 3-wk interval. Fish mortalities that occurred during the f i r s t week of the experimental period were replaced with comparable-sized fish to account for the relatively high i n i t i a l losses i n some tests. The winter experiment lasted from 16 December, 1974 to 27 February, 1975. Growth rates i n allopatry were determined at a l l nine test points i n the factor space for coho, but for trout only the diagonal points (see Fig. 32a) were done due to the d i f f i c u l t y i n obtaining the 132 necessary numbers of winter fry. Growth i n sympatry, five fish of each species per c e l l , was also determined at test points 5C--12-hr light and IOC—16-hr light (see Fig. 32a). At the end of the f i r s t month, the fi s h were individually measured to the nearest mm in fork length and to the nearest 0.01 g damp-weight while under mild anaesthetic. Subsequent measurements were done at each 2-wk interval. As i n the summer experiment, the fi s h were starved for 24 hr prior to being measured. Mortalities i n the i n i t i a l week were replaced with comparable-sized f i s h . The f i s h were hand-fed twice daily at regular hours with chopped fresh-frozen zooplankton consisting almost entirely of Euphausia  pacifica. This diet was chosen over commercial feeds i n the light of favourable reports on growth and stimulation of appetite i n the supple-mentary feeding of juvenile salmonids i n streams by Mason (1976) . Each quantum ration was deposited i n the feeding funnel fixed i n the plexi-glass cover of each c e l l , washed down with water. Daily rations were determined on a wet-weight basis and were related to test temperatures: 15 °C level received 8%, 10 °C level received 5% and the 5 °C level received 2.5% of the mean body weight of 10 fi s h . These rations were slightly i n excess of those required for maximum growth at each of the three test temperatures and were adjusted accordingly following each measurement period to account for.growth. As part of routine maintenance, a l l cells were cleaned free of wastes every 3 days. 133 RESULTS MORTALITY AND GROWTH Fish losses appeared to be primarily of behavioral rather than physiological origin, regardless of season. Mortality of either species was mostly the outcome of caudal infections from Saprolegnia sp., which appeared to be init i a t e d by the aggressive nipping of f i s h . Total mortality i n the summer and winter period, respectively, amounted to 8.1% and 10.7% for coho, and 5.2% and 35.5% for trout. In both species most mortalities involved f i s h that were about 20% less than the mean fork length i n any given test group. The much greater percent loss of trout than coho i n winter, may be due i n part to the bias i n testing trout at the diagonal points only of the test space. Chi-square tests for goodness of f i t indicated fish losses in most cases were significantly (P<0.01) higher than those expected at the higher temperatures and longer photoperiods. Both salmonids were approximately three times as aggressive at 15 than at 5 °C. Presumably, subdominant fi s h exposed to combinations of higher temperature and longer photoperiod were sub-jected to greater social stress, which may have increased their susceptib-i l i t y to infectious diseases. Mean growth rates for both coho and trout were calculated for each of the test points at the end of each time interval, according to Brown (1957) . log e Y T - log eY t G = X 100, T 134 where G designates specific- rate of growth expressed as percent change in weight of fi s h per day; Y^, Y represents weight of fi s h at end and beginning of each time interval, respectively, and T i s the time inter-val between measurements i n days. The measure of specific growth was useful i n comparing growth of fi s h of different sizes i n the summer and winter experiments. The results are summarized graphically i n three-dimensional plots only for those experiments in which a l l points of the factor space were tested (Fig. 33) . Summer growth of allopatric coho and cutthroat trout fry were f a i r l y similar (Fig. 33). Typically, they grew faster at combinations of higher temperature and longer photoperiod. An exception was the i n i t i a l reverse response to photoperiod at 15 °C for coho, which i s suspect of being of behavioral rather than physiological origin. Shorter day length may have lessened stress imposed by the test f a c i l i t y and/or by social interactions, resulting i n better growth. Outstanding were the more rapid and more pronounced effects of temperature than those of photoperiod on growth of both salmonids. By the end of the f i r s t three weeks, their growth was an approximate two-four fold higher at 15 than at 5 °C, irrespective of photoperiod. Definite trends i n photoperiod effects were delayed u n t i l after the i n i t i a l six weeks. Four-way factorial analysis of variance indicated that the interaction between temperature and photoperiod i n summer was in-significant (P>0.05), although photoperiod effects appeared earlier and were more influential at higher temperatures. The only significant 135 COHO winter summer TROUT summer Fig. 33. Mean ±95% confidence limits (three replicates) of specific growth rates of coho and cutthroat trout fry tested separately for the 9 test combinations of temperature and photoperiod. The growth rates are chronologically shown: i n summer at each 3-wk intervals (a-d); i n winter at 4-wk after i n i t i a l time (a) and at each subsequent 2-wk inter-vals (b-d) . 136 (P<0.01) interaction was that between temperature, time and species: growth rates decreased over time, at a faster rate at higher temperatures and differently between species. From the beginning to the end of the experiment maximum growth decreased from 3.4 to 1.2%/day i n coho, and from 2.2 to 1.4%/day in trout. Deceleration of growth was probably related to increasing size and/or age of fish and parallels findings by Brown (1946) and Laarman (1969) . Final acceleration of summer growth for both coho and trout was highest at 15 °C--16-hr-light and lowest at 5 °C--8-hr-light. Winter growth of allopatric coho and cutthroat trout fry, for the diagonal points of the test space, were similar i n pattern but differed strikingly i n magnitude (Table 18). As i n summer, growth of --both - species was strongly accelerated by temperature compared to that of photoperiod, but coho grew almost twice as fast as trout. Final optimum conditions for acceleration of coho growth coincided at test combinations that were seasonally i n phase, being best at those of lower temperature and shorter photoperiod (Fig. 33). Photoperiod seasonally out of phase i n winter inhibited coho growth at low temperatures. Minimum growth for coho at the 15 °C level i n the f i n a l stage was largely attributable to the negative effects of smoltification on growth (Wagner 1974) . As i n summer, i n winter the interactive effects between temperature and photoperiod on growth were not significant (P>0.05), whereas those of temperature and time were significant (P < 0.01); both salmonids showed a near twofold decrease between their i n i t i a l and f i n a l maximum rate of growth. 0 3 7 Table 18. Coho and cutthroat trout fry mean rate of growth ± 95 confidence limits for a 3 X 3 temperature and photoperiod test space. Photoperiod Temperature (hr of (C) light) Growth (% / day) Allopatry Coho Trout Sympatry CG'oho Trout a) In summer 5 5 5 10 10 10 15 15 15 12 16 8 12 16 8 12 16 0.85 ±0.24 0.94±0.06 1.10±0.13 1.48±0.31 1.59±0.09 1.59±0.12 1.84±0.60 1.80±0.43 1.69±0.39 0.80±0.14 0.71±0.10 0.92±0.07 1.06±0.17 1.33±0.20 1.36±0.20 1.34±0.27 1.66±0.26 1.69±0.20 b) In winter 5 5 5 10 10 10 15 15 15 12 16 8 12 16 8 12 16 1, 1. ,12±1.02 .09±1.92 0.85±0.97 2.00±1.54 1.86±1.84 1.85±1.38 2.00±3.01 2.17±3.58 1.90±2.89 0.65±0.22 1.26±0.29 1.49±0.29 1.19±0.17 0.57±0.06 1.94±0.45 1.14±0.18 138 In both summer and winter, the average rate of growth of coho and cutthroat trout was lowest at 5 °, intermediate at 10 0 and highest at 15 °C (Table 18). In summer and winter, respectively, daily acceleration of growth for coho i n allopatry ranged from 0.85-1.841 and from 0.85-2.17%; those for trout ranged from 0.80-1.69% and from 0.65-1.49%. Overall, photoperiod effects were some three-fourfold less influential than those of. temperature, regardless of season. Con-sidering both salmonids, average photoperiod effects on growth acceleration ranged from 0.17-0.26%/day, whereas those of temperature ranged from 0.75-1.04%/day. Sympatrically, daily growth rates for the experimental period i n winter were slightly more for coho and less for trout than their respective growth i n allopatry (Table 18). Ranking of growth i n sympatric t r i a l s also remained unchanged with coho growing about twice as fast as trout at both the 5 and 10 °C test combinations. CONDITION FACTOR Patterns of summer condition factor over the test space resembled those of specific growth (Fig. 34). In both species fish were in better condition at test combinations of higher temperature and longer photoperiod, with maximum condition at 15 °C and 16-hr light. More pro-nounced i n coho than in trout was a second peak i n body condition at 10 °C and 8-hr light which probably reflects the influence of short photo-period on fish social interactions. Aggression was particularly high i n coho and interaction with photoperiod on condition factor is suspected. -139 winter C O H O summer T R O U T summer Fig. 34. Mean condition factor of coho and cutthroat trout fry tested separately, for the 9 test combinations of temperature and photoperiod. I n i t i a l f i s h condition i s shown at a). Subsequent time intervals (b-e) are as i n Fig. 33 (a-d). 140 Changes i n condition over time were consistently positive i n both salmonids but considerably higher i n coho. Final average condition factor over the test space was 1.162 for coho and 0.967 for trout, their i n i t i a l differences i n condition being negligible. Ultimately, condition factor ranged from 1.08 to 1.25 i n coho and from 0.87 to 1.06 i n trout, being minimal i n both cases at the low temperature combinations. rCondition factOr^ o£\coho.;ahdi-trbut i;nn:summer:'was analysed by four-way factorial analysis of variance. Like growth rate, salmonid condition showed a highly significant difference between species (P<0.01) and pronounced effects of both temperature and time (P<0.01) on condition factor relative to those of photoperiod (P>0.05). Temperature interactions with both time and photoperiod were also significant (P<0.01) due to condition factor increasing more rapidly at the higher test combinations. Variance i n condition factor over the entire test space was approximately two times greater i n coho than i n trout, but i n both csises variances increased with time and at increasing temperature and photoperiod. In winter, coho i n allopatry were i n i t i a l l y i n slightly better condition than were trout and this difference was magnified approximately twofeldd within the f i r s t 6 wk; average condition factor for coho in -creased from 1.050 to 1.172; for trout from 0.956 to 1.022. Coho condition factor increased significantly at rising temperature (P<0.01) and at shortening photoperiod (P<0.01), showing a maximum at 15 °C and below 16-hr l i g h t . This coincided with the onset of smoltification of the 141 larger f i s h and parallels the decrease i n condition factor reported by Wagner (1974) i n steelhead trout. As trout i n allopatry were tested only at the diagonal points of the test space, temperature and photo-period effects could not be separated. Trout condition factor was lowest at 5 °C and 8-hr light. Comparisons of coho and trout condition factors i n sympatry were limited to t r i a l s at 5 °C--12-hr light and at 10 °C--.16-hr light. Condition of both species was slightly less i n sympatry than i n allopatry, but interspecific differences were similar i n both experimental groups; coho were i n much better body condition than were trout. HIERARCHICAL EFFECTS ON GROWTH Relative size within each t r i a l was' probably the most important factor influencing the growth of individual f i s h . Although the range i n size of the test f i s h was i n i t i a l l y restricted, a maximum range of 17 mm was unavoidable i n some t r i a l s due to the large numbers of fish needed. Size disparity probably accelerated establishment of size-dependent social hierarchies and an individual's growth rate presumably was largely determined by i t s rank. In a l l allopatric t r i a l s , the larger fish consistently grew faster than the smaller ones during both summer and winter. By length, mean summer growth of the largest fish (presumably the social dominant) i n allopatry was similar for coho and trout as exemplified i n t r i a l s at the diagonal points of the test space (Fig. 35). However, i n winter the 1'42 SUMMER 15 C -- 16 h J 1 I I I I L_ 3 4 5 6 7 9 10 II 12 WINTER 15 C - - 16 h I I I I I I I I I I I I 1 8 5 I I I I I I I I 1 1 L. I I I I I I 1 I 1 I 1 75 5 C - - 8 h 35 I I 1 1 1 1 L _l I I L. 0 1 2 3 4 5 6 7 W E E K S Fig. 35. Growth comparisons by length of the largest coho (•), the largest cutthroat trout (A) and the group mean of coho (o) and of cutthroat trout ( A ) fry,: in allopatry, at each of the three diagonal points of the test space in summer and winter. 143 largest coho i n any one t r i a l grew more rapidly than did trout particularly at 10 and 15 °C, at which coho increase in length doubled that obtained by trout. For both species i n most t r i a l s , the f i n a l differences i n size between the largest fish and the mean size of a l l test f i s h per t r i a l was about double that of their i n i t i a l size difference i n both the summer and winter experiments. In sympatry, the absolute gain i n weight of the largest f i s h i n both species was approximately twofoldd less than that i n allopatry for the same test conditions. However, ranking of interspecific growth i n sympatry remained unchanged from that i n allopatry, with coho growing faster than trout at both 5 and 10 °C (Fig. 36). In most t r i a l s , the f i n a l mean weight of the largest trout was less than that of the smallest coho, although their i n i t i a l weights were comparable. The prevailing hierarchical growth pattern i n sympatry was coho at the top and trout at the bottom. DENSITY EFFECTS ON WINTER GROWTH In winter, densities of stream populations of juvenile salmonids i n specific habitat sites (log jams at meanders, undercut root mats of trees, backwaters, etc.) may be relatively high (Hartman 1965b, Bustard and Narver 1975a, Glova and Mason 1977b) . Under such conditions growth of fi s h may be lessened, particularly through greater competition for a limited winter food supply. I tested the nul l hypothesis that there i s no difference i n winter growth of coho and cutthroat trout fry 14.4 7 - 0 r 6 0 5 0 o>4 0 h -X 3 0 CD Ixl ^ 2 0 L d 10 10 C - - 16 h J L J I L < 4 0 Ixl 3 0 2 0 5 C - - I 2 h J I I I I L 0 I 2 3 4 5 6 W E E K S 7 8 9 10 Fig. 36. Winter growth comparisons of the largest coho (•), the largest cutthroat trout ( A ) , and 'the group mean of coho (o) and of cutthroat trout ( A ) fry„ i n sympatry, at two temperature and photoperiod combinations i n the test space. 145 both i n allopatry and i n sympatry at low, medium and high density levels (14, 7 and 3.5 £ l i v i n g space per f i s h , respectively), fed a minimum food ration. Daily feeding rates were well below optimum and consisted of chopped fresh-frozen euphausiids at 2.0% of their wet body weight. Three replicates per each treatment were done i n the period 30 January-19 March, 1975, i n clear plexiglass tanks (58 x 48 x 20 cm), each partitioned into three compartments with flow-through screens for allopatric and sympatric t r i a l s . Water temperature was maintained at 10 °C and photoperiod at 12-hr light. Mean starting weights ranged from 3.1-3.8 g for coho and from 2.8-3.5 g for trout. Their daily weight gain showed a density-dependent relation-ship; relative growth i n both species was approximately two-fold greater at low than at high densities i n both al lopatric.-and sympatric t r i a l s (Table 19) . Lessened growth at higher densities may have re-sulted from an increase i n their maintenance requirements primarily i n response to greater social dissipation of energy. Species aggressive interactions were not quantified but were noted to be intense during-feeding periods; these appear to have had the greatest influence on growth at the medium density level, i n which both species gained approxi-mately 15% less weight i n sympatry than i n allopatry. Further, when fed a minimum food supply, weight gained by trout was generally higher than that of coho, particularly at the higher density levels i n sympatry. During feeding, coho became highly aggressive while trout were more inclined to scramble for food. 146 Table 19. Mean winter growth of coho and cutthroat trout fry at three different densities, fed a submaintenance ration for seven weeks: C, coho; T, trout. Liters of space per fish % weight gain ± S.E. Allopatry Sympatry 3.5 C 24.5 ± 12.2 19.7 ± 8.1 T 39.1 ± 7.5 43.4 ± 5.3 7 C 56.1 ± 14.8 41.4 ± 13.6 T 55.7 ± 7.1 40.2 ± 5.7 14 C 59.6 ±18.7 64.7 ± 47.5 T 52.6 ± 5.1 74.3 ± 3.2 -147 DISCUSSION TEMPERATURE AND PHOTOPERIOD EFFECTS ON GROWTH Growth of underyearling coho salmon and coastal cutthroat trout over the range of test combinations of temperature and photo-period, was similar i n pattern but not i n magnitude. Coho grew faster than trout i n most of the test combinations, but particularly at those above 5 °C. The higher growth rates observed for coho may reflect the summation effects of a number of underlying physiological and behavioral interspecific differences. Possible differences i n food consumption rates may have been an important factor. Cutthroat trout fry i n general show a much slower conditioning response to a r t i f i c i a l environments than do coho. As maintenance requirements are positively related to temperature up to a certain limit (Brown 1957), i n i t i a l l y , food intake by trout might have been below that required for maximum growth, particularly at the higher temperatures. Further, better growth by coho might also be coupled to a higher food conversion efficiency than i n trout. No data are available forfurther qualify this statement. Also a reduction i n swimming activity at rising temperatures has been suggested to compensate for increases i n basal metabolism (Kelso 1972; Brown 1946), keeping energy budgets i n check. However, i n this study, a lesser energy expenditure for coho than for trout i n the test f a c i l i t y i s doubted due to the more intense aggressive activity of the former. Aggression i n coho was found to be at least, threefold'; higher than i n trout at a l l three test temperatures (Glova and Mason 1976c) . Thus, social dissipation of energy through aggressive activity alone would seem to be higher for coho than for trout. However, trout might have dissipated more energy than coho through general stress imposed by the test f a c i l i t y . Temperature effects on growth of coho and cutthroat trout fry were far more pronounced and more rapid than those of static photo-periods during summer and winter. Within a ininimum period of 3 wk, "I obtained-maximum ^ temperature'acceleration-of specific growth rates ,r which were at least twofoi'dli higher at 15 than at 5 °C. The relative effects of temperature and photoperiod i n this study agree with those of Huh et a l . (1976) for age 0+ walleye, Stizostedium vitreum, one of the few comparable experiments i n the literature. Clarke^ recently 'obtaineH^a s i m i i a r i p a t t e ^ and -photoperiod on growth of coho fry i n spring, but higher growth rates than those of the present study. I obtained average growth rates for coho of 0.96, 1.55 and 1.78%/day for 12 wk at 5, 10 and 15 °C, respectively, i n summer. Clarke obtained growth rates of 1.85, 2.58 and 3.01%/day for 13 wk at 8, 11.5 and 15 °C. Both of these experiments used fis h of comparable i n i t i a l weights (0.6-0.8 g) but differed i n other respects; Clarke used a diet of Oregon Moist Pellet, large tanks with circular flow and dynamic photoperiod (similar to that of natural photo-period but accelerated i n time), whereas I used a diet of zooplankton, small tanks with thru flow and static photoperiod. In my experiments, tank size, shape and flow pattern may have permitted important behavioral influences on growth. b Clarke, C. Letter dated 24 August, 1977. Pacific Biological Station, Nanaimo, B.C.-149 Unlike the direct effects of temperature on metabolic activity (Fry and Hart 1948; Brown 1946; Schaeperclaus 1933), photo-period regulation has been postulated to act indirectly on fishes via the endocrine system (Saunders and Henderson 1970; Gross et a l . 1965; Hoar 1957), thus requiring longer exposure time for measurable effects to occur. The rate of photoperiod response i n this study was temperature-related, being more rapid at the higher temperatures. Minimum exposure time for definite trends i n photoperiod effects on patterns of salmonid growth was about 8 wk. Gross et a l . (1965) obtained more marked photo-period effects on growth i n 2- to 4-yr-old green sunfish within 6 wk, but their experimental temperatures went to a high of 25.5 °C. These authors suggest that dynamic photoperiod, rather than static, might obtain more pronounced stimulative, or inhibitive effects on growth. In my. study, prior acclimation to test photoperiods might have accelerated and possibly magnified the photoperiod effects. My findings demonstrate definite shifts i n the optimum con-ditions for coho growth, and possibly the same for trout, between the summer and winter periods. In summer, optimum conditions for growth occurred at combinations of high temperatures -- longer photoperiods, whereas i n winter there was a shift toward combinations of intermediate temperatures - - shorter photoperiods. Such shifts i n growth responses parallel those of seasonal growth cycles of fi s h i n north temperate regions. Further evaluation requires consideration of the possible i n -fluence of prior temperature-photoperiod history on the experimental f i s h . The summer stock had prior exposure to both naturally increasing water temperature and daylength; the winter stock had prior exposure to 150 reverse conditions. The observed stimulative and inhibitive effects of longer photoperiod on growth i n the summer and winter periods, respectively, suggest prior phototropic exposure may be an important consideration i n growth studies, as suggested by P h i l l i p s (1969) . I conclude, at least for static photoperiods, that growth rate i s l i k e l y to be less for fish exposed to photoperiods seasonally out of the phase, a l l other factors being equal. The precise role of photoperiod as i t may affect growth remains unknown, but activation of the pituitary gland via light manipulation (Pickford and Atz 1957), and occurrence of specific growth hormone i n f i s h pituitary gland (Hoar 1957) have been demonstrated, and shown to vary seasonally i n quantity i n accord with higher growth rates i n spring and summer (Swift and Pickford 1962). Photoperiod control of growth hormone production may i n part regulate protein and/or fat synthesis i n fish as speculated by Gross et al.(1965). ECOLOGICAL IMPLICATIONS Growth comparisons between underyearling coho salmon and coastal cutthroat trout i n the laboratory differed from those observed,in natural sympatric populations i n small streams. In the laboratory, when tested separately and fed an excess ration, coho grew faster than trout at most of the temperature/photoperiod test combinations. In streams, growth of trout appears to be less affected by being sympatric than i s coho. Despite possible interspecific competitive disadvantages to trout, associated with their later emergence, and smaller body size at emergence, growth of trout was higher, than that of coho (Glova and Mason 1974, 1976b). 151 Better growth by trout i n streams, may at least i n part, be due to their greater behavioral diversity i n feeding and microhabitat use, permitting greater potential exploitation of foods than by coho. For example, the trout's a b i l i t y to forage on both d r i f t and benthos (Mason and Machidori 1976) offers an obvious feeding advantage when the food supply is limited. Further, this study and other studies (Chapman 1962; Mason and Chapman 1965) have shown that coho fry are highly aggressive. The dissipation of energy i n social interactions i n my two study streams!might have been considerably higher by coho than by trout, depressing the growth of the former. Such interspecif differences may have assisted i n closing the size gap between the young-of-the-year among populations of these two salmonids. 152 CHAPTER V. GENERAL DISCUSSION AND CONCLUSIONS The niche of juvenile coho salmon and of coastal cutthroat trout i n small streams broadly overlap. These two salmonids are generalized exploiters of stream resources, trout slightly more so than coho, possibly due to their social subdominance i n interspecific competitive interactions. The behaviorally flexible feeding and microhabitat responses of these salmonids, may i n large part, be manifested i n the in s t a b i l i t y and heterogeneity, both spatially (diversity of habitat) and temporally (within the same space),of small streams i n general. Sanders (1968) suggested that i n severe and unpredictable environments, species adaptations are primarily i n response to the physical environment, resulting i n broadly overlapping niche development. On the other hand, i n benign and predictable environments, adaptations are primarily to other organisms, yielding narrowly over-lapping niches of populations. It i s commonly accepted that generalists are favored i n fluctuating environments, while specialists are favored i n stable ones (Schoener 1969) . The small stream environment seems to offer few discrete spatial and trophic choices to fish -- they are either r i f f l e - or pool-dwellers, feeding on d r i f t or benthos, or both. Accordingly, the niches of sympatric species broadly overlap. This study and others (Andersen and Narvers 1975; Mason and Machidori 1976) have adequately documented that fish biomass i n small coastal streams of Briti s h Columbia is,"' typically dominated by two salmonid species (a d r i f t feeder, and a d r i f t 153-and benthic feeder) and sculpins (benthic feeder). The food niche of the more generalized salmonid (the d r i f t and benthic feeder) broadly over-laps with those of i t s cohabitants as shown by Mason and Machidori (1976). In small streams, competition between fish species may reduce fis h diversity through lack of opportunities for niche specialization. Menge and Sutherland (1976) suggest that competition may be the dominant organizing interaction i n trophically simple communities, whereas predation may be dominant i n trophically complex ones. During the late summer period of low streamflows, competition between species continues to remain an elusive phenomenon to demonstrate experimentally on the basis of accepted concepts (Birch 1957, Milne 1961) . Its presence i n nature, however, may be implicated by, for example, the frequently low condition factor for salmonids (Glova and Mason 1976a), the substantial increase i n coho growth and survival with supplementary feeding i n streams (Mason 1976), and by the up to ten-fold higher biomass for allopatric than for sympatric populations of cutthroat trout (Glova and Mason 1977c) . Habitat segregation between stream populations of coho and cutthroat trout during the seasons of rapid growth presumably functions to reduce interspecific competition for resources. Competition i s seen to be of the exploitative or interference type (Brian 1956; Case and Gilpin 1974). Segregation may stem from interference when one species learns from experience that resources are less easily secured i n habitats frequented by the other species, or of the offensive'nature of the other species (e.g. aggression1)-.odor).. For. example', "Randall. (1978) 154 reported that although sympatric populations of Microtus montanus and M. longicaudus i n eastern Washington rarely showed strong aggressive encounters, the latter tended to be excluded from i t s preferred grass habitat by i t s greater propensity to withdraw from the larger and socially dominant, montanus. Alternatively, segregation may occur when one species i s more efficient than another i n exploitation of a specific resource (e.g. food, space) as illustrated for example, by Nilsson (1967) for salmonids and by Heinrich (1976) for social insects. Of these two types of competition, I consider the exploitative strategy to be of lesser importance i n the segregation process between sympatric populations of coho and cutthroat trout. Habitat shift.,1 by trout from their preferred pool space to r i f f l e s when in sympatry with coho, do-not appear to be due to their lesser efficiency than coho i n resource exploitation i n pools, but rather to social subdominance. Trout i n allopatry appear equally adept as coho i n feeding and i n u t i l i z i n g cover i n pools. In r i f f l e s , however, trout might be considered a more efficient exploitor of resources than coho, as reflected i n their a b i l i t y to u t i l i z e both passive and active foraging, and to also use submerged cover. Coho use of r i f f l e s involves short-term exploitation of food resources more limited by decreasing water temperature and increasing velocity than i n the case of trout. Mutual^agQnistic interference between coho and cutthroat trout i n this study appears i n large part to account for partitioning of resources (e.g. food, habitat), as documented by Hartman (1965b) for fry of sympatric coho and steelhead trout. The highly aggressive and 155 socially dominant coho i s an effective interference competitor against either trout species i n pools and other slow-water habitats. Con-versely, the equally aggressive but socially subdominant cutthroat and steelhead trout appear to exert a similar interference against coho i n r i f f l e s and other fast-water habitats. These reciprocal outcomes of interspecific interaction render unprofitable u t i l i z a t i o n of resources i n habitats i n which the other species has adaptive competitive advantages. Hydromechanically, the predominant aggressive behaviors and relative body positioning with respect to the streambed for trout (nipping, positioning nearer to streambed) are more suited to faster velocity habitats, whereas those for coho (threat displays, positioning nearer to water surface) are more suited to slower velocity habitats. Additionally, species rates of aggression differ within habitat types, coho being more inclined to defend pools and trout to defend r i f f l e s when i n sympatry. Werner and Hall (1977) concluded that agonstic inter-ference largely accounts for habitat segregation between sympatric populations of centrarchids i n lakes and ponds. Southwood (1977) i n a thorough review on the subject of ecological strategies i n nature, concluded that each arises from the evolutionary "trade-offs" of costs versus benefits i n the process of adaptation to habitats. Interference between sympatric coho and cut-throat trout, may i n part, be energetically governed; i.e. trout may be restricted to microhabitats i n which interference is energetically unprofitable to coho. The cost of maintaining social dominance over trout i n fast-water and marginal stream habitats may exceed the benefits (food, shelter) to coho. Structurally complex environments (e.g. r i f f l e s ) 156 might also decrease the foraging efficiency of a predator as shown for juvenile rainbow trout (Ware 1972). In this context, pools might permit more efficient feeding by coho and by salmonids i n general, than do r i f f l e s . As invertebrate d r i f t comprises a ;major portion of the diet of juvenile coho (Mundie 1969, 1971), the more complex array of submerged cover and of higher velocities i n r i f f l e s than i n pools, might reduce their foraging efficiency on d r i f t . Moreover, Case and Gilpin (1974) emphasize that i f the interference competitor i s to be able to dominate or exclude the exploitation competitor, i t must do so i n those habitats i n which the carrying capacity i s highest for populations of both species. This argument is consistent with my findings; coho socially minimize the cutthroat trout's use of pools, the habitat i n which salmonid carrying capacity i s typically some three-fold higher than i n r i f f l e s . In streams, the pool environment appears to be the most pre-dictable of habitats for rearing of f i s h . The species having priority of access to stream resources, w i l l i n a l l probability, maximize i t s competitive fitness i n pool habitats. The earlier emerging and socially dominant coho are competitively oriented to habitats (e.g. pools, glides) at the rich end of the resource gradient, while the later emerging and socially subdominant trout are so at the impoverished end (submarginal rearing habitats). Coho i s a "sit-and-wait" predator on invertebrate d r i f t i n pools, and may face food shortages due to diel patterns of d r i f t abundance (Waters 1969; Mundie 1971) and low d r i f t rates during summer minimum streamflows. Cutthroat trout i s a more generalized "searching" predator, capable of cropping the d r i f t and grazing the benthos i n both 157 fast and slow water habitats. The greater breadth of food and micro-habitat niche of trout than.of coho may allow them greater opportunism in exploitation of resources. Heinrich (1976) suggested that the optimal foraging strategy for any species w i l l ultimately be tested during times of lowest, not highest resource avail a b i l i t y . In streams, this occurs during the summer period of low streamflow, the season when densities, metabolism and food requirements of fish populations may a l l be high relative to the available food supply. Under such conditions, fish i n r i f f l e s may show more rapid growth than those i n pools. Food shortages for salmonids i n some streams i s seen to be wrought through summer low flows as was emphasized more than a decade ago by Roderick Haig-Brown (1964) -- " s t a b i l i t y of flow, especially enough flow in the summer months, is probably the most c r i t i c a l factor, since the young fi s h spend a f u l l year or more feeding i n a stream before migrating to sea; a very low summer flow w i l l mean losses through starvation". Predation and cannibalism, although not investigated in this study, may significantly influence the pattern of segregation between stream populations of juvenile salmonids. Lindstrom (1962) emphasized that the differential a b i l i t y of fish to avoid predators i n different situations, may reinforce segregation of sympatric prey species. In stream simulator studies, young salmonid fry are readily consumed by adult sculpins (Patten 1977) and by presmolt salmonids (Glova, unpublished data); both predators are commonest i n sites of deeper water with adequate cover i n natural streams. Both.Patten's and my work identified relative body size and learned avoidance behavior as important factors counter-acting predation. The possibility that size-related patterns of micro-158 habitat use of coho and of cutthroat trout fry i n streams, may i n part, represent a predator avoidance behavior cannot be precluded i n this study. Habitat segregation,while due to the outcome of social inter-actions i n the stream simulator, might i n nature be reinforced through size selective predation. Unlike coho, the smaller and more generalized trout may find the shallower water habitats an effective refuge from possible predators as well as one from the socially dominant coho i n pools. Symons and Heland (1978) found that i n some Atlantic salmon (Salmo salar) populations o£ NewnBrunswrck,, age 1+ fish (fork length > 10 cm) reduced the number of underyearlings i n the deeper r i f f l e habitats, by chasing them, and occasionally by catching and eating them. Bohlin (1977) showed a similar size segregation for a population of juvenile sea trout (Salmo trutta) i n a closed area of a small stream, with the larger age 1+ f i s h being most common i n deeper water areas and age 0+ fish i n shallow, smooth-bottom r i f f l e s . In summary, juveniles of coho salmon and coastal cutthroat trout are viewed as generalized exploiters of stream resources, possibly in response to spatial and temporal heterogeneity and unpredictability of small stream environments i n general. Accordingly, sympatric popu-lations of these two salmonids show broad overlap i n resource use. Partitioning of resources i s predominantly through physical habitat rather than trophic dimensions, and varies seasonally. Overlap i n microhabitat use is least during the late summer period of low streamflow when coho concentrate i n pools and trout i n r i f f l e s , the season when water tempera-tures are warm and populations of these salmonids are typified by elevated levels of metabolism, food requirement, aggression and density. 159 Conversely, overlap i n microhabitat use i s greatest during the winter period of high streamflow and cold water temperatures, with both salmonids seeking cover predominantly i n pools, the season when popu-lations show reversed levels to the above phenomena i n summer. Allopatrically, both species prefer pools irrespective of season. While a number of mechanisms may contribute to their segregation seasonally, that of mutual agonistic interference within habitat types appears to be most important. Laboratory growth of coho surpasses that of trout, regardless of season and of temperature/photoperiod conditions. In nature, the reverse pattern of interspecific growth i n summer might be attributable to higher social dissipation of energy and to lesser food- ava i l a b i l i t y for coho i n pools than for trout i n r i f f l e s . Comparison of body size and biomass of allopatric and sympatric cutthroat trout populations suggest that populations of the latter type might be resource-limited through interspecific competition from coho or sculpins, or both. 160 REFERENCES Andersen, B. C, and D. W. Narver. 1975. Fish populations of Carnation Creek and other Barkley Sound streams - 1974: Data Record and Progress Report. Fish. Res. Board Can. MS Rep. 1351: 73 p. Andreasson, S. 1971. Feeding habits of a sculpin (Cottus gobio L. Pisces) population. Inst. Freshwater Res. Drottningholm Ann. Rep. 51: 5-30. Andrewartha, H. G. 1961. Introduction to the study of animal populations. London. Andrusak, H., and T. G. Northcote. 1971. Segregation between adult cutthroat trout (Salmo clarki) and dolly varden (Salvelinus  alpinus) i n small coastal Bri t i s h Columbia lakes. J. Fish. Res.. Board Can. 28: 1259-1268. Bagenal, T. B. 1967. A short review of f i s h fecundity, In S.D. Gerking [edj The biological basis of freshwater fish production. Wiley and Sons Inc., New York. p. 89-111. Betts, M. M. 1955. The food of titmice i n oak woodlands. J. Anim. Ecol. 24: 282-323. Bjornn, T. C. 1971. Trout and salmon movements i n two Idaho streams as related to temperature, food, stream flow, cover, and population density. Trans. Am. Fish. Soc. 100(3): 423-438. Birch, L. C. 1957. The meanings of competition. Am. Nat. 91: 5-18. Bisson, P. A. 1978. Diel food,selection by two sizes of rainbow trout (Salmo gairdneri) i n an experimental stream. J. Fish. Res. Board Can. 35: 971-975. Bohlin, T. 1977. Habitat selection and intercohort competition of juvenile sea-trout (Salmo trutta). Oikos 29(1): 112-117 Bray, R. N., and A. W. Ebeling. 1975. Food, activity and habitat of three "picker-type" microcarnivorous fishes i n the kelp forests off Santa Barbara, California. Fish. B u l l . 73(4): 815-829. Brian, M. V. 1956. Segregation of species of the ant genus Myrmica. J. Anim. Ecol. 23: 319-337. Brocksen, R. W., G. E. Davis, and C. E. Warren. 1968. Competition, food consumption and production of sculpins and trout i n laboratory stream communities. J. Wildl. Manage. 32(1): 51-75. Brown, M. E. 1946. The growth of brown trout (Salmo trutta Linn.) I. Factors influencing the growth of trout fry. J. Exp. B i o l . 22: 118-129. 161 . 1957. Experimental studies on growth, p. 361-398. In M. E. Brown [ed.] The physiology of fishes. Vol. 1. Academic Press, New York and London. Burns, J. W. 1971. The carrying capacity for juvenile salmonids in some northern California streams. Calif. Fish and Game 57(1): 44-57. Bustard, D. R., and D. W. Narver. 1975a. Aspects of the winter ecology of juvenile coho salmon (Cncorhynchus kisutch) and steelhead trout (Salmo gairdneri). J. Fish. Res. Board Can. 32: 667-680. _ _ . 1975b. Preferences of juvenile coho salmon (Oncorhynchus kisutch) and cutthroat trout (Salmo clarki) relative to simulated alteration of winter habitat. J. Fish. Res. Board Can. 32: 681-687. Case, T. J., and M. E. Gilpin. 1974. Interference competition and niche theory. Proc. Nat. Acad. Sci. U.S.A. 71(8): 3073-3077. Chapman, D. W. 1962. Aggressive behavior in juvenile coho salmon as a cause of emigration. J. Fish. Res. Board Can. 19: 1047-1080. Cody, M. L. 1968. On the methods of resource division in grassland bird communities. Am. Nat. 102 (924): 107-147. Connell, J . H. 1961. The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus  stellatus. Ecol. 42: 710-723. DeBenedictis, P. A. 1974. Interspecific competition between tadpoles of Rana pipiens and Rana,sylvatica: an experimental f i e l d study. Ecol. Monogr. 44: 129-151. Dineen, C. F. 1951. A comparative study of the food habits of Cottus  b a i r d i i and associated species of Salmonidae. Am. Midi. Nat. 46: 640-645. Egglishaw, H. J. 1967. The food, growth and population structure of salmon and trout i n two streams i n the Scottish Highlands. Freshwater Salmon Fish. Res. 38: 1-32. . 1970. Production of salmon and trout in a stream i n Scotland. J. Fish. B i o l . 2: 117-136. Everest, F. H. 1967. Midget Bentzel current speed tube for ecological investigations. Limnol. Oceanogr. 12: 179-180. 162 Everest, F. H., and D. W. Chapman, 1972. Habitat selection and spatial interaction by juvenile chinook salmon and steelhead trout i n two Idaho streams. J . Fish. Res. Board Can. 29: 91-100. Fabricius, E. 1953. Aquarium observations on the spawning behavior of the char, Salvelinus alpinus. Rep. Int. Freshwater Res. Drottningholm, 34: 14-48. Fry, F. E. J., and S. J. Hart. 1948. The relation of temperature to oxygen consumption i n goldfish. B i o l . B u l l . 94: 66-77. Gause, G. F. 1934. The struggle for existence. The Williams and Wilkins Co., Baltimore. ":163 p. Gee, J. H., and T. G. Northcote. 1963. Comparative ecology of two sympatric species of dace (Rhinichthys) i n the Fraser River system, B r i t i s h Columbia. J . Fish. Res. Board Can. 20(1): 105-118. Gibbons, J . R. H., and J. H. Gee. 1972. Ecological segregation between Longnose and Blacknose dace (Genus:Rhinichthys) i n the Mink River, Manitoba. J . Fish. Res. Board Can. 29: 1245-1252. Giger, R. D. 1972. Ecology and management of coastal cutthroat trout i n Oregon. Fish. Res. Rep. 6, Res . Div. Oregon State Game Comm. 61 p. Glova, G. J., and J. E. Mclnerney. 1977. C r i t i c a l swimming speeds of coho salmon (Oncorhynchus kisutch) fry to smolt stages i n relation to sa l i n i t y and temperature. J. Fish. Res. Board Can. 34(1): 151-154. Glova, G. J., and J. C. Mason. 1974. Interactive ecology of juvenile salmon and trout i n streams. I. Progress during 1973. Fish. Res. Board Can. MS Rep. 1300: 35 p. . 1976a. Interactive ecology of juvenile salmon and trout i n streams. I I . Progress during 1974. Fish. Res. Board Can. MS Rep. 1391: 24 p. . 1976b. Interactive ecology of juvenile salmon and trout i n streams. I I I . Progress during 1975. Fish. Res. Board Can. MS Rep. 1400: 38 p. . 1976c. Comparative effects of temperature and photoperiod on the laboratory growth of underyearling coho salmon and coastal cutthroat trout. Fish. Res. Board Can. MS Rep. 1414: 50 p. . 1977a. Interactions for food and space between sympatric populations of underyearling coho salmon and coastal cutthroat trout in a stream simulator during summer. Fish. Mar. Serv. MS Rep. 1428, 36 p. 163 . 1977b. Interactions for food and space between sympatric populations of underyearling coho salmon and coastal cutthroat trout i n a stream simulator during winter and spring. Fish. Mar. Serv. MS Rep. 1429, 31 p. . 1977c. Comparison of coastal cutthroat trout populations i n allopatry and those sympatric with coho salmon and sculpins i n several small coastal streams on Vancouver Island, B.C. Fish. Res. Board Can. MS Rep. 1434: 35 p. Goodnight, W. H., and T. C. Bjornn. 1971. Fish production i n two Idaho streams. Trans. Am. Fish. Soc. 100(4): 769-780. Grant, P. R. 1972. Interspecific competition among rodents. Ann. Rev. Ecol. Syst. 3: 79-106. Grant, P. R., B. R. Grant, J . M. M. Smith, I. J . Smith, and L. K. Abbott. 1976. Darwin's finches: population variation and natural selection. Proc. Nat. Acad. Sci. U.S.A. 73: 257-261. G r i f f i t h , J. S., J r . 1972. Comparative behavior and habitat u t i l i z a t i o n of brook trout (Salvelinus fontinalis) and cutthroat trout (Salmo clarki) i n small streams m northern Idaho. J . Fish. Res. Board Can. 29: 265-273. . 1974. Utili z a t i o n of invertebrate d r i f t by brook trout (Salvelinus fontinalis) and cutthroat trout (Salmo clarki) i n small streams i n Idaho. Trans. Am. Fish. Soc. (3): 449-447. Grinnell, J . 1904. The origin and distribution of the chestnut-backed chickadee. Auk. 21: 364-382. Gross, W. L., E. W. Roelofs, and P. 0. Fromm. 1965. Influence of photo-period on growth of green sunfish, Lepomis cyanellus. J. Fish. Res. Board Can. 22: 1379-1386. Haig-Brown, R. 1964. Fisherman's F a l l . William Collins Sons ^ Co., Toronto. 279 p. Hall, J . D., and R. L. Lantz. 1969. Effects of logging on the habitat of coho salmon and cutthroat trout i n coastal streams, p. 355-375. In T. G. Northcote [ed.] Symposium on salmon and trout in streams, H. R. MacMillan Lectures i n Fisheries, University of Britis h Columbia (1968), 388 p. Hall, D. J., W. E. Cooper and E. E. Werner. 1970. An experimental approach to the production dynamics and structure of fresh-water.animal communities. Limnol. Oceanogr. 15: 839-928. Hardin, G. 1960. The competitive exclusion principle. Science 131: 1292-1297. -164 Hart, J. L. 1973. Pacific Fishes of Canada. Fish. Res. Board Can. Bu l l . 180, 740,p. Hartley, P. H. T. 1948. Food and feeding relationships i n a community of freshwater fishes. J . Anim. Ecol. 17: 1-14. . 1953. An ecological study of the feeding habits of the English titmice. J. Anim. Ecol. 22: 261-268. Hartman, G. F. 1963. Observations on behavior of juvenile brown trout i n a stream aquarium during winter and spring. J . Fish. Res. Board Can. 20 (3): 769-787. . 1965a. An aquarium with simulated flow. Trans. Am. Fish. Soc. 94(3): 274-276. . 1965b. The role of behavior i n the ecology and interaction of underyearling coho salmon (Oncorhynchus kisutch) and steel-head trout (Salmo gairdneri). J . Fish. Res. Board Can. 22(4): 1035-1081. . 1966. Some effects of temperature on the behavior of underyearling coho and steelhead. B r i t i s h Columbia Fish Wildl. Branch Mgmt. Rep. 51: 15 p. Hartman, G. F., and C. A. G i l l . 1968. Distributions of juvenile steel-head and cutthroat trout (Salmo gairdneri and S^ clar k i clarki) within streams i n southwestern B r i t i s h .Columbia. J. Fish. Res. Board Can. 25(1): 33-48. Haven, S. B. 1973. Competition for food between the intertidal gastropods Acamea scabra and Acamea d i g i t a l i s . Ecol. 54: 143-151. Heinrich, B. 1976. Resource partitioning among some eusocial insects: bumblebees. Ecol. 57: 874-889. Hoar, W. S. 1957. Endocrine Organs, p. 245-285. In M. E. Brown Led.l The physiology of fishes. Vol. 1. Academic Press, New York and London. Huh, H. T., H. E. Calbert, and D. A. Stuiber. 1976. Effects of temperature and light on growth of yellow perch and walleye using formulated feed. Trans. Am. Fish. Soc. 2: 254-258. Hutchinson, G. E. 1959. Homage to Santa Rosalia or why are there so many kinds of animals? Am. Nat. 93 (145): 145-159. Jaeger, R. G. 1971. Competitive exclusion as a factor influencing the distributions of two species of ter r e s t r i a l salamanders. Ecol. 52: 632-637. .165 . 1972. Food as a limited resource i n competition between two species of terrestrial salamanders. Ecol. 53: 535-546. Jenkins, T. M. 1969. Social structure, position choice, and micro-distribution of two trout species (Salmo trutta and Salmo  gairdneri) resident i n mountain streams. Anim. Behav. Monogr. 2(2): 57-123. Johnston, J. M., and S. P. Mercer. 1976. Sea-run cutthroat i n salt-water pens: Broodstock development and extended juvenile rearing (with a l i f e history compendium). Wash. State Game Dept. Fish. Res. Rep. 92 p. Kalleberg, H. 1958. Observations i n a stream tank of t e r r i t o r i a l i t y and competition on juvenile salmon and trout (Salmo salar L. and S. trutta L.). Rep. Inst. Freshwater Res. Drottningholm 39: 55-98. Keenleyside, M. H., and F. T. Yamamoto. 1962. Territorial behavior of Atlantic salmon (Salmo salar) . Behav. 19: 139-169. Kelso, J . R. M. 1972. Conversion, maintenance, and assimilation for walleye, Stizostedium vitreum vitreum, as affected by size, diet, and temperature. J . Fish. Res. Board Can. 29: 1181-1192 Krebs, C. J., and I. Wingate. 1976. Small mammal communities of the Kluane Region, Yukon Territory. The Can. Field Nat. 90(4): 379-389. Krohn, D. C. 1967. Production of the reticulate sculpin (Cottus perplexus) and i t s predation on salmon ."fry in three"Oregon streams. M.Sc. thesis, 78 p., Oregon State University, Corvallis. Laarman, P. W. 1969. Effects of a limited food supply on growth rates of coho salmon and steelhead trout. Trans. Am. Fish. Soc. 3: 393-397. Lack, D. 1944. Ecological aspects of species formation i n passerine birds. Ibis 86: 260-286. Larkin, P. A. 1956. Interspecific competition and population control i n freshwater f i s h . J . Fish. Res. Board Can 13(3): 327-342. LeCren, E. D. 1965. Some factors regulating the size of populations of freshwater f i s h . Mitt. Int. Ver. Theor. Angew Limnol. 13: 88-105. _. 1969. Estimates of fis h populations and production in small streams i n England, P. 269-280. In T. G. Northcote [ed.] Symposium on salmon and trout i n streams, H. R. MacMillan Lectures i n Fisheries, Univ. Brit i s h Columbia (1968). 388 p. .166 Lindroth, A. 1955. Distribution, t e r r i t o r i a l behavior and movements of sea trout fry i n the River Indalsalven. Rep. Inst. Freshwater Res. Drottningholm 36: 104-119. Lindstrom, T. 1962. Life history of whitefish young i n two lake reservoirs. Rep. Inst. Freshwater Res. Drottningholm 44: 113-144. Lister, D. B., and H. S. Genoe. 1970. Stream habitat u t i l i z a t i o n by cohabiting underyearlings of chinook (Oncorhynchus tshawytscha) and coho (G\_ kisutch) salmon i n the Big Qualicum River, Briti s h Columbia. J . Fish. Res. Board Can. 27: 1215-1224. Lotka, A. J . 1925. Elements of Physical Biology. Baltimore: Williams § Wilkins. (Reprinted 1956 by Dover Publ., New York as Elements of Mathematical Biology). MacArthur, R. H. 1958. Population ecology of some warblers of north-eastern coniferous forests. Ecol. 39: 599-619. . 1972. Geographical ecology. Harper and Row, New York. 269 p. Maitland, P. S. 1965. The feeding relationships of salmon, trout, minnow, stone loach and three-spined sticklebacks i n the River Endrick, Scotland. J. Anim. Ecol. 34: 109-133. Mann, R. H. K. 1971. The populations, growth and production of fish i n four small streams i n southern England. J. Anim. Ecol. 40: 155-190. Mann, R. H. K., and D. R. 0. Orr. 1969. A preliminary study of the feeding relationships of fi s h i n a hard-water and a soft-water stream i n southern England. J. Fish. B i o l . 1: 31-44. Mason, J . C. 1969. Hypoxial stress prior to emergence and competition among coho salmon fry. J . Fish. Res. Board Can. 26: 63-91. . 1975. Seaward movement of juvenile fishes, including lunar periodicity i n the movement of coho salmon (Oncorhynchus  kisutch) fry. J. Fish. Res. Board Can. 32(12): 2542-2547. . 1976. Response of underyearling coho salmon to supplemental feeding i n a natural stream. J. Wildl. Manage. 40(4): 775-788. Mason, J . C, and D. W. Chapman. 1965. Significance of early emergence, environmental rearing capacity, and behavioral ecology of juvenile coho salmon i n stream channels. J. Fish. Res. Board Can. 22: 173-190. Mason, J. C, and S. Machidori. 1976. Populations of sympatric sculpins, .Cottus aleuticus and Cottus asper, i n four adjacent salmon-producing coastal streams on Vancouver Island, B.C. Fish. B u l l . 74(1): 131-141. 167 Mendelson, J . 1975. Feeding relationships among species of Notropis (Pisces:Cyprinidae) in a Wisconsin stream. Ecol. Monogr. 45(3): 199-230. Menge, J. L., and B. A. Menge. 1974. Role of resource allocation, aggression and spatial heterogeneity i n coexistence of two competing intertidal starfish. Ecol. Monogr. 44: 189-209. Menge, B. A., and J. P. Sutherland. 1976. Species diversity gradients: Synthesis of the role of predation, competition, and temporal heterogeneity. Am. Nat. 110(973): 351-369. Milne, A. 1961. Definition of competition among animals. Symp. Soc. Exptl. B i o l . 15: 40-61. Moldenke, A. R. 1975. Niche specialization and species diversity along a California transect. Oecologia 21(3): 219-242. Morse, D. H. 1974. Niche breadth as a function of social dominance. Am. Nat. 108(964): 818-830. Moyle, P. B. 1977. In defense of sculpins. Fisheries, a bulletin of the Am. Fish. Soc. 2(1): 20-23. Mundie, .-'J.. H. 1969. Ecological implications of the diet of juvenile coho in streams, p. 135-152. In T. G. Northcote [ed.] Symposium on Salmon and Trout in streams, H. R. MacMillan Lectures in Fisheries, Univ. Brit i s h Columbia. (1968) . 388 p. . 1971. The di e l d r i f t of Chironomidae in an a r t i f i c i a l stream and i t s relation to the diet of coho salmon fry, Oncorhynchus kisutch (Wabaum) Can. Entom. 103(3): 289-297. . 1974. Optimization of the salmonid nursery stream. J. Fish. Res. Board Can. 31: 1827-1837. Narver, D. W., and B. C. Andersen. 1974. Fish populations of Carnation Creek and other Barkley Sound streams -- 1970-1973: Data Record and Progress Report. Fish. Res. Board Can. MS Rep. 1303: 115 p. Newman, M. A. 1956. Social behavior and interspecific competition in two trout species. Physiol. Zool. 29(1): 64-81. Nilsson, N. A. 1956. Om konkurrensen i naturen. Zoologisk Reny: 40-47. . 1960. Seasonal fluctuations i n the food segregation of trout, char and whitefish in 14 North-Swedish lakes. Rep. Inst. Freshwater Res. Drottningholm 41: 185-205. . 1963. Interaction between trout and char i n Scandinavia. Trans. Am. Fish. Soc. 92: 276-285. . 1967. Interactive segregation between fish species, p. 295-313. In S. D. Gerking [ed.] The Biological Basis of Freshwater Fish Production. Blackwell Sci. Publ. Oxford and Edingburgh. 168 Northcote, T. G. 1958. Effect of photoperiodism on response of juvenile trout to water currents. Nature, Lond'., 181: 1283-1284. Park, T. 1962. Beetles, competition, and populations. Science 138: 1369-1375. Parkinson, E. A., and P. A. Slaney. 1975. A review of enhancement techniques applicable to anadromous gamefishes. Brit i s h Columbia Fish Wildl. Fish Mgmt. Rep. 66: 100 p. Patten, B. G. 1977. Body size and learned avoidance as factors affecting predation on coho salmon fry, Oncorhynchus kisutch, by torrent sculpin, Cottus rhotheus. Fish. B u l l . 75(2): 457-459. Petrosky, C. E., and T. F. Waters. 1975. Annual production by the slimy sculpin population of a small Minnesota trout stream. Trans. Am. Fish. Soc. 104 (2): 237-244. P h i l l i p s , A. M., Jr. 1969. Nutrition, digestion and energy u t i l i z a t i o n . P. 391-432. In W. S. Hoar and D. J. Randall [ed.] Fish /Physiology. Vol. 1. Academic Press, New York and London. Pianka, E. R. 1973. The structure of lizard communities. Ann. Rev. Ecol. Syst. 4: 53-74. Pickford, G. E., and J. W. Atz. 1957. The physiology of the pituitary gland of fishes. New York Zool. Soc, New York. 613 p. Randall, J. A. 1978. Behavioral mechanisms of habitat segregation between sympatric species of Microtus: habitat preference and interspecific dominance. Behav. Ecol. Sociobiol. 3: 187-202. Redfield, J. A., C. J. Krebs, and M. J. Taitt. 1977. Competition between Peromyscus maniculatus and Microtus towsendii in grasslands of coastal Bri t i s h Columbia. J. Anim. Ecol. 46: 607-616. Ruggles, C. P. 1966. Depth and velocity as a factor i n stream rearing and production of juvenile coho salmon. Can. Fish-Cult. 38: 35-53. Sanders, H. L. 1968. Marine benthic diversity: a comparative study. Am. Nat. 102: 243-282. Saunders, R. L., and E. B. Henderson. 1970. Influence of photoperiod on smolt development and growth of Atlantic salmon (Salmo  salar). J. Fish. Res. Board Can. 27: 1295-1311. Schaeperclaus, W. 1933. Textbook of pond culture. U. S. Fish Wildl. Serv., Fishery Leaflet 311: 1-240. 169 Schoener, T. W. 1969. Optimal size and specialization in constant and fluctuating environments: An energy-time approach, p. 103-114. In G. M. Woodwell and H. H. Smith [ed.] Diversity and s t a b i l i t y i n ecological system. Brookhaven Symp. B i o l . 22: 103-114. . 1974. Resource partitioning i n ecological communities. Science 185: 27-39. Seber, G. A. F., and E. D. LeCren. 1967. Estimating population parameters from catches large relative to the population. J. Anim. Ecol. 36: 631-643. Siegel, S. 1956. Nonparametric Statistics for the behavioral sciences. McGraw-Hill Book Co., New York, 312 p. Sokal, R. R., and F. J. Rohlf. 1969. BIOMETRY - The principles and practice of statistics i n biological research. W. H. Freeman and Company. 776 p. Southwood, T. R. E. 1977. Habitat, the templet for ecological strategies? J. Anim. Ecol. 46: 337-365. Stein, R. A., J. D. Hall and P. E. Reimers. 1972. Social interaction between juvenile coho (Oncorhynchus kisutch) and f a l l chinook salmon (G\_ tshawytscha) i n Sixes River, Oregon. J. Fish. Res. Board Can. 29(12): 1737-1748. Straskraba, M., J. Char, S. Frank, and V. Hruska. 1966. Contribution to the problem of food competition among the sculpin, minnow and brown trout. J. Anim. Ecol. 37(2): 303-311. Swift, D. R., and G. E. Pickford. 1962. Seasonal variations i n the growth hormone content of pituitary glands of the perch, Perca f l u v i a t i l i s L. Am. Zool. (Abstracts) p. 451. Symons, P. E. K. and M. Heland. 1978. Stream habitats and behavioral interactions of underyearling and yearling Atlantic salmon (Salmo salar). J. Fish. Res. Board Can. 35(2): 175-183. Trojnar, J. R., and R. J. Behnke. 1974. Management implications of ecological segregation between two introduced populations of cutthroat trout in a small Colorado lake. Trans. Am. Fish. Soc. 103(3): 423-429. Volterra;, V. 1926. Variations and fluctuations of the number of individuals of animal species l i v i n g together, p. 409-448. In R. N. Chapman [ed.] Animal Ecology, New York. McGraw-Hill Book Co. Wagner, H. H. 1974. Seawater adaptation independent of photoperiod i n steelhead trout (Salmo gairdneri). Can. J. Zool. 52: 805-812. Walton, I., and C. Cotton. 1676. The Compleat Angler. 5th edition (facsimile of original edition). Cassell and Company Ltd., London. 344 p. 170" Ware, D. M. 1972. Predation by rainbow trout (Salmo gairdneri): the influence of hunger, prey,density, and prey size. J. Fish. Res. Board Can. 29: 1193-1201. Waters, T. F. 1969. Invertebrate d r i f t - ecology and significance to stream fishes, p. 121-134. In T. G. Northcote [ed.] Symposium on Salmon and Trout i n Streams, H. R. MacMillan Lectures i n Fisheries, Univ. British Columbia (1968). 388 p. Werner, E. E., and D. J. Hall. 1977. Competition and habitat shift i n two sunfishes (Centrarchidae) Ecol. 58: 869-876. Werner, E. E., D. J. Hall, D. R. Laughlin, D. J. Wagner, L. A. Wilsmann, and F. C. Funk. 1977. Habitat partitioning i n a freshwater fish community. J. Fish. Res. Board Can. 34: 360-370. Wilson, E. 0. 1975. Sociobiology. The new synthesis. The Belknap Press of Harvard University Press, Cambridge, Mass. 697 p. Zaret, T. M., and A. S. Rand. 1971. Competition in tropical stream fishes; support for the competitive exclusion principle. Ecol. 52: 336-342. Appendix Table 1. Monthly water chemistry of Bush Creek over a 1-yr period, 1973-74 O N D J F ' M A M J J A S 0 Chlorides 210.0 3.6 1.9 13.6 6.8 4.2 4.2 9.6 12.8 31.8 33.6 63.5 57.8 Calcium - 2.1 1.8 5.6 2.9 2.3 3.1 4.8 5.8 - 16.5 20.0 25.0 Magnesium 0.4 0.4 1.0 0.3 0.5 0.2 0.9 1.0 - 3.5 3.2 4.3 Sodium 97.0 1.7 1.6 5.6 3.0 2.2 2.8 6.2 5.8 - 18.0 23.0 22.0 Potassium '3.4 0.1 0.2 0.1 0.1 0.1 0.2 0.2 0.2 - 0.4 0.6 0 .2 Diss. Iron 0.1 - 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 T.D.S. 523.4 28.2 37.5 59.0 32.7 37.4 31.9 49.7 53.0 21.2 - 210.3 228.3 Sulphates 27.0 4.1 2.8 5.1 4.8 5.8 7.0 7.4 3.0 3.0 3.2 5.6 4.8 Nitrates 11.2 8.9 7.6 9.2 5.2 3.1 1.5 0.3 3.6 10.6 10.0 1.1 Phosphates 0.8 0.3 0.2 0.2 0.1 0.4 0.1 0.1 0.2 - 0.1 0.1 0.2 Ammonia 7.1 0.5 0.5 1.8 0.6 0.3 0.1 0.3 0.3 0.6 0.1 - 0.4 Silicates 107.9 30.0 34.1 142.6 95.0 71.1 75.0 107.9 87.6 - 111.2 200.9 203.4 Appendix Table 2. Monthly water chemistry of Holland Creek over a 1-yr period, 1973-74. 0 N D J F M A M J J A S 0 Chlorides 40.0 2.6 1.5 5.4 2.2 Calcium 9.1 2.0 1.3 3.8 2.2 Magnesium 1.3 0.4 0.3 0.6 0.4 Sodium 8.0 1.1 0.9 2.6 1.1 Potassium 0.3 0.1 0.1 0.2 0.1 Diss. Iron 0.1 - 0.1 0.1 0.1 T.D.S. 99.5 13.7 28.1 40.6 24.0 Sulphates 7.0 4.1 2.7 5.2 4.3 Nitrates 11.1 9.4 4.1 7.1 4.2 Phosphates 0.9 0.3 0.2 0.1 0.1 Ammonia 5.9 0.3 0.3 1.2 0.4 Silicates 112.9 22.8 14.1 110.6 69.0 2.0 1.6 3.0 3.0 13.2 ' 37.8 31.5 36 .2 1.8 2.0 2.3 2.4 - 16.5 14.0 19.5 0.4 0.1 0.4 0.4 - 3.2 2.2 3.4 1.2 1.2 1.6 1.6 - 11.5 11.0 14.5 0.1 0.2 0.1 0.1 - 0.5 0.5 1.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 28.7 21.5 24.5 22.9 15.1 - 136.7 164.8 5.2 6.8 6.8 2.6 1.8 4.3 7.4 5.7 2.4 0.6 0.3 1.0 13.0 11.9 0.3 0.1 0.1 0.1 - 0.1 0.1 0.1 0.2 0.1 0.1 0.3 0.1 0.1 - 1.1 52.4 74.5 93.1 36.6 - 101.0 175.4 96.3 Table 3. Summary of. sta t i s t i c s for fish populations and related stream physical parameters in Bush Creek, September-October 1974 and 1975. Trout Total fish Water depth Vel. cm cm/s % Biomass N/i r Trout Coho Sculpins Age 0 (mm) A l l ages combined (mm) N N/ms g/m2 % age 0 % > age 1+ Mean F.L. + S.E. Range Mean F.L. + S.E. . POOLS 19 74 27 40 6 7 192 7 1 7.4 10 7 72 0 17 3 14 0 5 0 8 '. 79 21 45 1 + 0 93 43-52 55 7 + 1 02 29 14 9 5 211 7 1 8.6 8 5 58 9 32 6 17 0 6 0 1 100 - 44 6 + 1 02 32-49 44 6 + 1 02 57 19 8 9 183 3 2 4.3 4 2 38 4 57 ^ 8 0 1 0 2 100 - 44 6 + 0 82 36-55 44 6 + 0 82 24 18 7 1 174 7' 3 9.5 2 2 44 8 55 0 4 0 2 0 2 .75 25 45 3 + 1 03 40-50 47 0 + 1 05 22 21 5 0 101 4 5 6.2 10 5 33 3 56 2 9 0 4 0 7 100 - 41 3 + 0 40 35-46 41 3 + 0 40 X, !' 32 22 7.4 861 5.8 7.2 7.2 49.5 43.7 52 0.4 0.5 91 9 46.2 + 0.84 32-55 46.6 + 0.86 19 75' 37 36 5 4 190 5 0 9 2 3 7 26 4 69.9 8 0 2 0 3 75 25 48 1 + 1 56 3 7-64 57 3 + 6 56 23 15 5 0 159 6 6 10 0 9 2 29 8 61.0 . 12 0 5 0 9 83 17 45 1 + 0 99 '35-55 59 8 + 6 99 59 40 3 8 218 3 7 5 8 18 3 33 1 48.6 28 0 5 1 1 79 21 46 5 + 1 09 33-63 64 2 + 8 09 54 54 4 5 123 2 3 3 3 13 0 25 2 61.8 18 0 3 0 4 89 11 47 8 + 0 96 38-64 53 0 + 3 96 53 47 3 5 171 3 2 4 8 7 7 39 6 52. 7 12 0 2 0 4 85 15 49 8 + 1 30 38-61 57 9 + 6 30 45 38 4 4 861 4 2 6 6 10 30 8 58.8 78 0 3 0 6 82 18 47 5 + 1 18 33-64 . 58 4 + 6 38 Table 3 (cont'd) Trout Total fish Water 7- Biomass Area depth Vel. m s c m cm/s N N/ms g/m^  Trout Coho Sculpins Age 0 (mm) A l l ages c omb i ne d (mm) N N/ms g/ms % age 0 7, > age 1+ Mean F.L. + S.E. Range Mean F.L. + S.E. GLIDES 1974 X. I 29 . 11 10. .3 229 7. , 7 10. . 1 5. . 7 24. .5 69. .8 11 0, , 4 0. 6 100 - 41. 8 + 1, .30 33-50 41. .8 + 1, ,30 43 12 16. 4 237 5 .5 6. .9 9. .2 40. ,4 50. 4 22 0, .5 0. 6 86 • 14 52. ,1 + 1. .06 40-55 . 53. .0 + 3. .50 24 12 16. . 1 80 3 .4 4, .9 5 .5 13, 81. .3 6 0, .3 0, ,3 85 15 49. .8 + 0. 78 42-55 50. .2 + 1, .20 2 7 14 15 .9 81 3. .0 3 .4 13. . 7 49 .3 37, ,0 12 0, ;4 0. ,5 100 - 49. .4 + 0, ,42 39-55 49. .4 0, .42 36 12 9 , 2 136 3 .8 4 . 7. 7 .2 42 .5 50, .3 11 0. .3 0, ,3 91 9. 50, . 1 + 0. .56 38-54 53, .0 + 1, . 71 32 12 13 .6 763 4 . 7 6 .0 8 .3 .34 .0 57, .8 62 0 .4 0, .5' 92 8 48. .6 + 0, .82 33-55 49 .5 + 1 .63 19 75 X. 21 17 10. . 1 161 7. ,6 13. 1 9.8 32. . 7 57, .5 11 0. ,5 1. .3 62 38 49. .9 + 1, .78 33-59 66, ,2 + 7, ,78 37 10 14. .9 . 116 3 , 1 4, ,3 6.7 22 , 6 70, , 7 ' 9 0. 2 0, .3 78 22 43, .9 + 1. .57 3 7-54 . 51, ,8 + 5 , .57 25 13 11. .2 79 3, , 1 3. ,3 8.8 37. .9 53, ,3 10 0, 4 0. ,3 100 - 44. .0 + 1, .45 38-52 *44. ,0 + 1, ,45 33 9 8 .3 12 7 ' 3, .8 4, , 1 7.8 • 32, ,2 60, .0 14 0, /. 0, .3 100 - 44, ,2 + 1, . 79 36-60 44, .2 + 1 .79 32 10 8, .9 97 2, .9 3, .5 • 10.3 40. .0 49 . 7 12 0, /, 0. ,3 100 - 45, .8 + 1 .84 38-5 7 45 .8 + 1 .84 30 12 10 . 7 580 4 . 1 5 , 7 3.7 33, . 1 58 . 2 56 0, , 4 0 .5 88 12 45 .6 + 1 .69 33-60 50 .4 + 3 .69 175 f-t o m f-t o t-» r*. \Q O f-st M m vt H iri in u*\ I A i i i i j H Nst O J O f i n n <}• <f o o o O o m r-C O rt o m O m O 00 <r -ct- m <r •CN -JD 00 CN vD CN ' i—I <—I O O CO O I S \0 !-< —' ~< O O + 1 o <r ^  <t o> 00 o o O oo o> m m u-i rg r~* rt rt rt + 1 + +1+1+1 + 1 o <t vO <t Oi -> CO in H n o <f <f <f m vt 1 A vt in <T o o o o o <t vo N ^ m o o o o o •—1 co CN GO <r o^ CN o m in •-J u*> <t o> I A n n i N co o> n co \D n f-t m ->t vo <t CO O O M O CN 00 vO 1 m m so >—* i cr> vO n m m CN N M N m m vt <—< <r CN m — i m m O 1 CN CN — i fN m O O O in vD O CO CD 3 T3 O m •— CN CO o> O " I Table! 4. Summary of s t a t i s t i c s for fish populations and related stream physical parameters in Holland Creek, September-October 1974 and 1975. Area Water depth Vel. cm cm/s Total fish % Biomass N/m2 g/m2 Trout Coho Sculpins Trout Age 0 (mm) A l l ages combined (mm) N N7m2 g/m2 7, % > age 0 age 1+ Mean F.L. + S.E. Range Mean F.L. + S.E. 24 1 39 6. 1 89 . 3, ,6 7. .5 10. 9 24. 3 64.8 18 33 9. .8 58 3, .2 7, .0 11. . 7 13. .6 74.7 30 38 7, .8 74 2 .4 5 .4 5. . 1 26. .5 68.4 24 25 6, .2 57 2 .3 5 .2 10. .0 19, ,6 70.4 21 33 8. .3 65 3 .0 7 .2 6. .4 16, .6 77.0 X, T 23 34 7, .6 343 2 .9 6 .5 8 .8 20. . 1 71.1 POOLS 1974 10 10 6 7 39 0.4 0.6 0.2 0.3 0.3 0.'-0.8 0.8 0.3 0.5 0.5 0.6 100 . 100 100 100 100 38.6 + 1.30 42.4 + 1.41 43.7 + 2.50 40.0 + 2.42 41.0 + 1.30 33-46 3 7-51 35-50 31-50 35-43 100 41.1 + 1.79 31-51 38.6 + 42.4 + 43.7 + 40.0 + 41.0 + 1.30 1.41 2.50 2.42 1.30 41.1 + 1.79 1975 X, T 22 14 16 45 56 31 18.5 69 36 35 39 37 12.7 57 12.1 65 7.8 121 7.3 167 11.7 479 2.8 3.2 2 . 7 2.6 3.0 2.9 5.5 6.0 8.0 6.9 8.5 . 7.0 1. ,6 28. .4 ' 70. ,0 5 0. .2 0. 1 100 37. .6 + 1, .50 34-38 37. 6 +. 1. .50 1, .3 22. .2 76, .5 3 0. 2 0. 1 100 - 37. ,7 + 1. .20 36-40 37. . 7 + 1. 20 26 .8 20, ,4 52 .8 5 0. .2 2. 1 21 79 40. ,0 + 0, .00 - 103. ,6 + 19. .99 3 .5 22. .6 78. .9 3 0. ,1 0. 2 100 - 37. .5 + 1. .20 36-39 37. ,5 + • 1. ,20 1 .4 9. .0 89 .6 6 0. , 1 0. ,1 83 17 40. .0 + 1. .30 36-44 49 .7 + 9, .73 6 .9 20 .5 72 .6 22 0, .2 0. ,5 81 19 38 • 7 + 1 .04 34-44 53 .2 + 6 .72 177. . 3 +»l x H O H M N + I + I+I+I+I vo M <t n s j H H H N rg + I+I+I+I+I vo C N <r n <f CO I O <f u-l n <*i \t ci n f~- <r <r co o —< o o o v t vO CM N n o o o o o <t o m +i O CN O vO f-< + 1+1+1 +1+1 n O N co N CO H 00 <t o> n <f n <f n vO O vO <t O s J i O O t N CN O CN O f t r-4 + + I+I+I+I + 1 CO 0> CO CN 00 <T CO ro H CO CO o <j" m co co o> ro H H M n H o o o o o 1—• CN vt m —i o o o o o ON CO i—1 00 co r-- C N <r o H H <t n O m 0> O i CN M n u*i u-i <t o> m Ix" C N <f CN] C N <T m M st \ 0 C N m v£> O vo o ro <f <r <r m o> o> in C N o> —i in <r <r C N tn m r^ . <t vrj C N m C N m cn CN [*"- in v£> CN —> CN CN CN CN i (--. r-. o O o> n C N o> o co —* C N m cn m KO O -J-Table 4 (cone'd) Trout Area Water depth Vel. cm cm/s N/n Total fish % Biomass g/m2 Trout Coho Sculpins Age 0 (mm) A l l ages combined (mm) N N/m2 g/m2 age 0 age 1+ Mean F.L. + S.E. Range Mean F.L; + S.E. RIFFLES 1974 22 9 36 6 50 2 2 3 5 11 4 23 6 65 0 6 0 3 0 4 67 33 56 0 + 2 50 34-45 57 0 + 2 56 20 11 40 7 20 1 0 1 7 3 7 4 1 92 2 . 1 0 1 0 1 100 - 45 0 + 0 00 32-45 45 0 + 0 00 13 6 41 2 21 1 6 2 2 46 7 13 9 39 4 ' 7 0 5 1 0 100 - 38 7 + 2 61 32-51 38 7 + 2 61 13 5 40 0 29 2 1 3 0 28 9 24 8 46 3 7 0 5 0 9 100 - 40 4 + 1 43 34-45 40 4 + 1 43 24 22 40 5 39 1 6 2 3 26 5 20 2 53 3 11 0 4 0 6 100 - 43 5 + 1 70 35-50 43 5 + 1 70 18 11 39 8 159 1 7 2 5 23 4 17 4 ' 59 2 32 0 4 0 6 93 7 44 7 + 1 65 32-51 44 9 + 1 66 GO 1975 15 15 38 5 50 3 1 5 6 2 9 10 5 86 6 47 12 48 3 43 0 9 1 4 8 6 7 1 84 3 51 13 40 5 40 1 1 1 3 18 5 7 7. 73 8 23 15 36 1 - 34 1 4 3 9 4 6 7 2 88 2 37 6 44 1 103 2 8 1 9 16 8 1 6 81 6 35 12 41 5 2 70 1 9 2 8 10 2 6 8 83 0 5 0 3 0 2 100 - 38 4 + 2 54 34-48 38 4 + 2 54 8 0 2 0 1 100 - 42 6 + 1 94 36-53 42 6 + 1 94 12 0 2 0 2 92 8 40 3 + 1 58 34-53 48 8 + 8 58 6 0 3" 0 2 100 - 43 3 + 1 20 41-48 43 3 + 1 20 16 0 4 0 3 100 43 6 + 1 46 34-56 43 6 + 1 46 47 0.1 0.2 98 - 41.6 + 1.74 34-56 43.3 + 3.14 Table! 5. Summary of s t a t i s t i c s for fish populations and related stream physical parameters in Ayum Creek, October 1975. Area m 2 Water depth Vel. cm cm/s Total fish % Biomass N/m2 g/m3 Trout Coho Sculpins Trout Age 0 (mm) A l l ages combined (mm) N N/m2 g/m2 • % % > age 0 age 1+ Mean F.L. + S.E. Range Mean F.L. + S.E. POOLS 60 55 6.2 208 3, .5 21. ,3 0. 7 11.9 87.4 7 0. .1 0. ,1 100 - 51. 7 + 1,87 45-5 7 51. , 7 + 1, .87 21 37 5.1 41 2 .0 9, .7 22. .7 26.6 50. 7 4 0. .2 2. .2 25 75 57. .0 + 0.00 - 108. .3 + 20 .50 45 44 4.7 53 1 .2 4. .6 10. , 1 20.4 69.5 6 0. .1 0. .5- 84 16 46. ,6 + 1.47 42-55 73. .0 + 26. .47 41 44 3.8 84 2 .0 6 .8 25 , 3 21.3 53.4 8 0. .2 1, .7 38 62 44. . 7 + 0.50 41-47 99. ,9 + 17. .50 29 24 8.1 31 1 . 1 2 .6 5. .7 35.6 58.7 " 4 0. , 1 0, ,2 100 - 49. .8 + 0.48 49-51 49. ,8 + 0 .48 X, r 40 41 5.6 417 1 .9 9 .0 12 .9 23.2 63.9 29 0 . 1 0 .9 69 31 49. .9 + 1.08 41-57 76, .5 + 13 .36 -•J GLIDES 52 20 20. .0 117 2, .3 4. ,9 25, ,0 50.3 24. . 7 21 0, ,4 1.2 81 19 54. .1 + 1, ,23 43-65 69 .5 + 8 .23 36 19 17. .3 84 2 .3 5. , 7 25. ,3 49.4 25. ,3 30 0, .8 1.4 90 10 51. ,4 + 0. .93 37-66 57. .8 + 3 .93 75 25 19. .5 115 1 .5 5. .0 20, .9 31.4 47. , 7 39 0 .5. 1.1 82 18 51. .9 + 0. ,98 41- 76 61, . 1 + 3. .98 37 17 19 ,5 50 1 .3 7 .2 9 .8 9.2 81. .0 8 0, .2 0.7 63 37 47. ,2 + 1. , 19 38-5 7 71. .8 + 13. .19 63 14 17 .5 69 1 .1 • 2, .9 17 .9 . 46.3 35. .8 27 0, .4 0.5 100 - 50. .9 + 1. .25 39-66 50, .9 + 1, .25 53 19 18 .8 435 1 . 7 5 .1 • 19 .8 37.3 42 .9 125 0 .5 1.0 83 17 51. • 1 ± 1. .12 3 7- 76 62, .2 + 6. .12 RIFFLES 29 13 46, .9 39 1 .4 2 .6 48. . i 35, .9 . 16. ,0 20 . 0. 7 1.3 95 5 55 .4 + 1. , 10 40-82 58. .9 + 4. .10 61 15 41, .2 106 1. .6 3, .2 37, , 7 43, .0. 19. .3 48 0. ,8 1.2 96 4 52, .8 + 0, ,66 42-78 55, .4 + 2. ,66 29 19 21 .2 23 0 .8 2 . 7 7, .9 29. . 7 62. .4 5 0. 2 0.2 100 - 52, .0 + 5, ,17 3 7-64 52 .4 + 5 .17 28 13 23 .8 28 1 .0 1 .2 45, ,0 31 . 7 23, .3 15 0. .5 0.5 100 - 48 .4 + 1, ,64 39-62 48 .4 + 1. .64 131 12 23 .9 42 0 .3 0 .5 54, ,0 38 .0 8 .0 27 0. ,2 0.3 100 - 52 .2 + 0. .89 44-62 52. .2 + 0, .89 56 15 • 31 .4 238 1 .0 2 .0 38, ,5 35 . 7 25, .8 115 0. ,5 0.7 98 2 52 .2 + 1 ,90 3 7-82 .53 .4 + 2 .89 Table 6- Summary of statistics for the resident population of cutthroat trout and related stream physical parameters upstream of the barrier f a l l s in French Creek, September 1976. Water depth Vel. cm cm/s Total fish N/m2 70 Biomass Trout Others Trout A l l ages combined Age 0 (mm) (mm) % % > Mean F.L. Mean F.L. age 0 age 1+ + S.E. Range + S.E. POOLS X, I 57 22 - 17 . 0 3 2 2 100 71 29 6 2 43 0 6 1 4 100 68 32 7 8 25 0 4 1 9 100 66 26 10 0 41 0 6 1 8 100 52 22 10 0 33 0 6 1 7 100 22 53 3 3 51 2 3 6 0 100 56 31 7 5 210 0 8 2 5 100 18 82 60 0 + 1 53 63-68 94 0 + 24 95 5 .60 6 + 1 16 50-84 64 3 + 2 78 76 24 66 5 + 1 54 51- 75 83 2 + 6 65 88 12 63 5 + 1 21 47- 76 69 2 + 2 90 94 6 64 3 + 1 26 53-80 67 1 + 2 66 78 22 52 9 + 1 88 36- 71 66 9 + 4 41 75 25 62 3 + 1 43 36- 84 74 1 + 3 94 CO o GLIDES X, I 51 11 - 8 • 0 2 1 3 100 55 14 22 2 15 0 3 0 8 100 67 19 14 9 31 0 5 1 3 100 53 13 19 2 105 2 0 2 3 . 100 43 21 18 9 3 7 0 9 3 0 100 54 16 18 8 196 0 8 1 7 100 25 75 61 5 + 5 51 56-67 97 6 + 12 14 93 7 64 3 + 1 73 55- 73 68 9 + 4 89 93 7 64 0 + 1 52 51-80 69 1 .+ 3 86 96 4 48 6 + 0 86 29-67 50 6 + 1 31 81 19 59 8 + 1 34 49- 75 73 0 + 5 09 78 22 59 6 + 2 16 29-80 71 8 + 3 03 RIFFLES x, r-32 11 33 0 15 0.5 0 9 100 11 13 40 0 9 0.8 * 1 2 100 17 8 40 0 23 •1.3 1 6 100 20 11 36 0 14 .0.7 1 2 100 20 11 37 3 61 .0.8 1 2 100 100 - 60 3 + 2 57 45- 77 60 3 + 2 57 90 10 52 4 + 4 00 36- 68 55 8 + 4 90 100 _ 46 0 + 1 28 36- 60 46 0 + 1 28 100 - 5 7 7 + 2 12 49- 71 57 7 + 2 12 98 2 54 1 + 2 50 36- 77 . 54 9 + 2 •72 Table 7. Summary of stat i s t i c s for the resident fish populations and related stream physical parameters upstream of the barrier f a l l s in Bings Creek, October 1976. Trout Total fish Water % Biomass Area depth Vel. : m2 cm cm/s N N/m2 g/m2 Trout Stbk* Age 0 (mm) A l l ages combined (mm) N N/m2 g/m2 % % > age 0 age.1+ Mean F.L. + S.E. Range Mean F.L. + S.E. POOLS 41 33 6. .4 32 0. .8 3. . 1 97, ,4 2 .6 31 0, ,8 . 3 .0 68 32 59. .2 + 1 .03 50- 70 76. .0 + . .99 76 35 9. .6 61 0. .8 3, .6 100 - 61 0, ,8 3, .6 .64 . 36 59. . 5 + 2 .08 48- 70 79, .6 + 4 . 14 39 40 5, .0 42 1. . 1 3, .3 97. .8 2, .2 41 1. .0 3 .2 78 22 57. .6 + 0 .92 48- 71 70. .3 + 4, . 12 28 30 8. .6 37 1 .3 5 . 7 100 - 37 1, ,3 5 . 7 54 46 ' 58. 2 + 2 .39 46- 72 78. 7 + 4, .84 38 28 ' 4, .0 70 1. .9 4. .0 82 , .0 18 .0 58 1, ,6 3. .3 90 10 57. .8 + 1 .00 45- 75 62. .2 + 2, .03 53 50 3, .2 113 2, .2 14. .0 97. ,6 2, ,4 96 1, 8 13 .7 44 56 63 . 2 + 1, .72- 49-88 94. 4 + 3, .89 X, I 46 36 6. . 1 355 1. ,4 5. .6 95. ,8 4. ,2 324 1. ,2 5. .4 66 34 • 59. .3 + 1 .52 45-88 76. 9 + 4, .00. 1 GLIDES 64 14 17 . 2 33 0, .5 1. ,2 100 - 33 0 .5 1 .2 88 12 59 . 1 + 0, ,83 51- 73 64 .0 + 2 .45 49 18 13 .3 26 0. .5 1, . 1 . 92. . 7 7, ,3 23 0, .5 1 .0 83 17 58, . 7 + 1, .02 51-67 62 .6 + 2, ,06 29 25 14, .9 60 2 . 1 4. .8 97. .3 2, . 7 58 2, .0 4, . 7 78 22 54, .5 + 0, ,69 4 7-69 64. .1 + 2 , .26 61 19 23, .8 58 1. .0 2, . 1 98. .0 2, .0 56 0, .9 2, .0 89 11 59, ,0 + 0, 92 4 7- 72 63. ,0 + 1. .81 49 16 22, .7 69 1. .4 4. ,0 99. .3 0, . 7 68 1 ,4 4, .0 86 14 63 , .0 + 0. 84 53-83 68. , 7 + 1. .84 50 18 18. , 4 246 1. . 1 2, .6 97, ,5 2. 5 238 1. . 1 2, .6 85 15 58, ,9 + 0. .86 4 7-83 64, ,5 + 2 , .08 RIFFLES 53 10 29. ;4 ' 13 • 0. .3 0. ,6 100 13 o'. .3 0. .6 92 8 61 . 2 0 .95 55-67 63, , 7 + 2 . 10 14 14 20. ,8 • 47 3, .3 4, .9 100 47 3 .3 4, ,9 • 94 6 52 .8 + 1. .00 45- 73 55, . 1 + 1 .67 23 10 40. .0 14 0, .6 1. ,4 100 14 0, .6 1, ,4 86 ' 14 ' 52 . 4 + 1 .23 45-63 63. ,9 + 8 .50 20 9 29. 21 1 . 1 1, . 7 100 21 1. . 1 1, , 7 . 90 10 54 .3 -i- 1 .65 41-68 57 .0 + 2 .41 25 13 33. .3 24 1. .0 2, , 1 100 24 1, .0 2. 1 92 8 60. .6 1 .08 51-69 63. .0 + 1 .99 27 11 30. ,6 119 1, .3 2 , 1 100 119 1. ,3 2. 1 91 9 56. .3 + 1 , . 18 41- 73 60. .5 + 3 .33 — 'Denotes stickleback. Table 1 8. Summary of sta t i s t i c s for the resident fish populations and related stream physical parameters in Shawnigan Creek (.inlet), October 19 75. Trout A l l ages c omb i ne d Age 0 (mm) (mm) 'I- > Mean F.L. Mean F.L. N N/m2 g/m2 age 0 age 1+ + S.E. Range + S.E. POOLS X, I 63 32 3 9 15 0 2 0 6 96 4 3 6 13 0 2 0 5 92 8 61 4 + 2 28 49-95 62 2 + 5 28 45 42 3 4 34 0 8 2 0 78 0 22 0 30 0 7 1 6 100 - 63 9 + 1 24 50-78 63 9 + 1 24 82 54 4 8 49 0 6 1 2 100 - 49 0 6 1 2 90 10 54 1 + 1 17 42-74 60 8 + 3 17 70 22 2 5 58 0 8 1 7 82 6 17 4 48 0 7 1 4 94 6 58 3 + 1 31 42-92 61 4 + 2 31 71 22 . 3 5 56 0 8 1 4 90 0 10 0 53 0 8 1 3 96 4 52 3 + 1 17 39-98 57 4 + 2 17 97 16 3 8 80 0 8 1 5 89 7 10 3 76 0 8 1 3 95 5 50 3 + 1 24 39- 71 57 2 + 2 24 71 31 3 7 292 0 7 1 4 89 5 10 5 269 0 6 1 2 94 6 56 7 + 1 40 39-98 60 5 + 2 74 Total fish Water 7- Biomass depth Vel. cm cm/s N N/m2 g/m2 Trout Sculpins GLIDES 44 .10 9 6 27 0 6 0 6 100 - 27 0 6 0 6 100 48 7 + 0 81 42-59 53 23 9 9 77 1 4 2 3 79.5 20.5 74 1 4 1 8 100 52 8 + 1 04 38-98 40 11 14 9 32 0 8 0 9 100 - 32 0 8 0 9 100 51 0 + 1 41 38- 71 22 16 13 8 23 1. 0 1 2 100 - 23 1 0 1 2 100, 51 7 + 1 77 39- 72 '34 5 16 6 16 0 5 0 6 80 20 14 0 4 0 4 100 49 6 + 1 70 39-61 X. I 39 13. 13.0 175 0.9 1.3 91.9 8.1 170 0.8 1.0 100 - 50.8 + 1.35 38-98 RIFFLES 19 8 . 32 3 8. 0 4 0 5 100 - 8 0 4 0 5 100 52 1 + 2 68 42-67 45 15 18 6 30 0 7 1 4 95.6 4.4 28 0 6 1 3 100 61 4 + 1 54 4 7- 76 37 7 24 4 . 28 1 5 • 1 5 88 12 26 1 4 1 3 100 47 2 + 1 15 38-5 7 14 12 18 2 12 0 8 0 8 100 - 12 0 8 0 8 100 48 7 + 1 92 38-62 10 5 34 5 9 0 8 0 7 100 - 9 0 8 0 8 100 45 9 + 1 66 38-52 X. r 25 9 25.6 87 0.8 1.0 96.7 3.3 83 0.8 0.9 100 51.1 + 1.79 38- 76 Appendix Table 9. Summary of experiments conducted i n the stream simulator i n summer, showing fis h fork length and range, and weight change data i n each of the three size-classes. Coho Trout Mean F.L. F.L .range Mean F.L. F.L.range Experiment Time period ±S .E. (mm) (mm) % Awt ±S .E. (mm) (mm), % A wt ' :a \ Main experiments Allopatry (1) Jun. 2 -16 36.3 + 0.19 35 .38 +15.5 37.3 + 0.38 35--40 +12.8 40.2 + 0.26 39 -42 +16.0 40.7 + 0.40 39--43 +14.3 47.0 + 1.29 43--53 +21.8 48.7 + 1.74 43--53 +15 .8 Sympatry (1) Jun.16 -23 38.0 + 0.26 37--39 + 7.7 37.3 + 0.43 34--39 - 3.8 44.4 + 0.20 44--45 + 8.9 44.1 + 0.74 40--46 + 3.6 53.7 + 0.88 52 -55 +14.5 54.3 + 1.67 51--56 + 3.6 Allopatry (2) Jul. 7--22 38.8 + 0.27 37--41 + 3.8 38.7 + 0.33 35--41 +28.1 48.1 + 0.40 45--50 + 6.2 44.3 + 0.28 43--46 +12.2 57.7 + 1.20 53--60 +12.7 51.6 + 0.62 49 -53 + 8.5 Sympatry (2) Jul.22 -28 38.6 + 0.22 38--40 + 4.1 38.6 + 0.22 37--39 + 7.7 46.7 + 0.42 45--48 + 8.0 45.8 + 0.74 43--48 + 4.2 56.3 + 1.20 54--58 + 6.8 55.3 + 2.68 50--59 + 6.0 No food i n system Sympatry (1) Aug.11--18 41.7 + 0.27 40--43 - 9.1 41.2 + 0.49 39--44 -13.1 51.3 + 0.64 49--53 -15.1 50.6 + 0.42 50--53 -13.3 61.2 + 0.44 61--62 - 9.6 61.5 + 1.32 59--64 - 9.7 Sympatry (2) Aug.18--24 40.6 + 0.40 39--43 - 4.1 39.7 + 0.56 37--42 - 6.2 51.7 + 0.52 50--53 - 6.0 48.0 + 1.38 43--53 - 5.7 63.7 + 0.88 62--65 - 4.5 65.7 + 2.03 62--69 - 7.1 Appendix Table 9 (cont'd) Coho - Trout Mean F.L. F.L.range Mean F.L. F.L.range Experiment Time period ±S .E. (mm) (mm) % Awt ±S .E. (mm) (mm) % A wt I n i t i a l 2 days of test period Sympatry Aug.27-Sep.l 42.2 + 0.30 41-43 40.7 + 0.68 38-45 49.4 + 0.90 47-54 50.4 + 0.72 47-53 60.7 + 0.33 60-61 58.0 + 0.58 57-59 £ 1' Relative density Sep. 1-7 42.0 + 0.00 42 - 3.5 39.5 + 0.23 37-41 • - 7.9 ) Total coho: 10 46.0 + 0.00 46 + 1.9 43.9 + 0.46 42-46 - 1.2 Total trout: 30 51.5 + 0.50 51-52 - 5.7 50.2 + 0.58 49-52 + 4.2 Sep. 7-16 42.9 + 0.47 40-46 - 4.3 43.6 + 1.03 40-46 -11.7 ) Total coho: 30 50.1 + 0.53 47-53 - 2.7 50.7 + 1.76 48-54 - ,8.8 Total trout: 10 58.2 + 0.20 58-59 + 1.8 57.5 + 0.50 57-58 -12.8 185 Appendix Table 10. Summary of experiments conducted i n the stream simulator i n winter showing fi s h fork length and range i n each of the three size-classes. Coho Trout Mean ± S.E. Range Mean ± S.E. Range (mm) (mm) (mm) (mm) f]• Winter Tested at 3 °C Allopatry Dec .2-18, 1975 50.5 + 0.60 45 -55 47.7 + 0.55 43--53 66.9 + 0.45 65 -70 59.9 + 0.86 56--67 77.7 + 0.49 76 -79 73.8 + 1.56 71--81 Sympatry Dec .18-25 55.8 + 0.58 52--58 53.9 + 0.48 51--57 68.6 + 0.87 65 -72 65.4 + 1.36 61--71 83.7 + 1.33 81 -85 82.7 + 0.88 81--84 Allopatry Dec .25-Jan.8 58.1 + 0.48 51 -60 53.4 + 0.56 50--58 72.9 + 0.41 70--75 63.0 + 0.62 60--67 88.7 + 0.61 86 -90 76.8 + 1.08 74--80 Sympatry Jan .8-14 60.1 + 0.40 57--62 57.3 + 0.54 54--61 70.1 + 0.70 69--74 70.0 + 0.44 68--71 90.7 + 3.33 84--94 92.0 + 3.05 86--96 Tested at 5 °C Sympatry Jan .14-21 56.0 + 0.83 51--60 55.1 + 0.60 51--59 67.3 + 0.81 64--70 65.9 + 1.08 62--70 83.3 + 3.18 77--87 81.3 + 3.18 75--85 Sympatry Jan, .21-27 57.4 + 0.47 55--60 55.9 + 0 .54 53--59 66.4 + 0.65 64--69 . 64.7 + 0.99 63-•69 80.3 + 1.20 78--82 81.0 + 1.53 78-•83 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0094633/manifest

Comment

Related Items