”EMU“; ' ' " I “I m ' WW I‘ l w-Lg‘éffie 11315“ “V l 3 M .1“ ' iii [1‘ m M: "3": .5 . IVE" 1 1’! II, r.',' | r .J" H I’ ’. I '[‘l|l4.‘llln u ‘1 - I‘ ‘l' ’- I-0‘. r '0 women: 8 ll limlllflllilljll \‘lllllllllllll 3 1293 1093534 THESIS | This is to certify that the thesis entitled TROUT POPULATION RESPONSES TO HABITAT ALTERATIONS BY HALF-LOG COVER DEVICES IN A MICHIGAN STREAM presented by Guy Werner Fleischer has been accepted towards fulfillment of the requirements for M.S . degree in Fisheries Ray J. White Major professor 2,7 >74», #71} 2.. M Date ’ 0-7639 tremor 0 “fishfiflfi P} {33:4‘33 fin .,._.. 55‘: I’“:.I ‘ a- Q . RETURNING MATERIALS: )V153[_) Place in book drop to LIBRARJES remove this checkout from “ your record. FINES will be charged if book is returned after the date stamped below. Wm " _' r- ‘ '.1\i 'K ‘ ' } v.r—'..'\4 v 3230.. Do A. 183?:3 TROUT POPULATION RESPONSES TO HABITAT ALTERATIONS BY HALF-LOG COVER DEVICES IN A MICHIGAN STREAM By Guy Werner Fleischer A THESIS Smeitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1982 ABSTRACT TROUT POPULATION RESPONSES TO HABITAT ALTERATIONS BY HALF-LOG COVER DEVICES IN A MICHIGAN STREAM By Guy Werner Fleischer Thirty-one half-log cover structures were installed in a 200m section of the Pigeon River, Ostego County, Michigan during the summer of 1979. Brook and brown trout standing crops in the treated section and in a surrounding 1.3 km.control (reference) area were compared spring and fall 1979 and 1980. No significant differences (p=.10) in trout standing crop between the reference and treatment sections were found before or after the half-logs were installed. A cover rating method was developed to determine the association of trout standing crop to overhead cover. Length of undercut bank and amount of instream overhead cover meeting minimum.spacing criteria were measured in 1979 and 1980 during low flow in the study area. Under- cut bank, instream overhead cover and their sum, total overhead cover, was correlated to trout standing crop of all sizes, greater than 150mm and 150-399mm. Fall standing crops had greater positive correlations to cover than did spring standing crops. The trout were more associated with Guy Werner Fleischer the instream overhead cover than undercut bank. The best correlations to cover were with the 150-399mm size class standing crops. Cover accounted for less than half the variation of trout standing crop suggesting the cover rating inadequately defined positions chosen by stream trout with other factors being important in determining trout standing crop. ACKNOWLEDGEMENTS This study was made possible by a grant from.Dr. Sibley W. Hoobler. I extend my sincere thanks to Dr. and Mrs. Hoobler for their interest and hospitality, and especially for Mrs. Hoobler‘s home-made lunches. I am.also grateful to James Holburn for coordinating use of the facility. I would like to thank Dr. Ray J. White, my major professor, for this opportunity that furthered my knowledge and whetted my interest in streams. My committee members, Drs. Howard Johnson and Rich.Merritt, provided helpful suggestions in reviewing the manuscript. The following people are thanked for donating their time and much needed services: John Sefcik, Chuck Korson, Elaine Rybak, Mike Enk, John Hamilton, Joe Bohr, Rick Ligman, Andy Raddant, Chris Bennett Mark Ultis and Craig Spencer. I am especially indebted to Greg Curtis for his diligence in electrofishing and for volunteering himself for the tedious portions of the field work; and to Kurt Fausch for not only his help in the field but also for his contagious interest in stream salmonids. And, finally, I acknowledge the moral support provided by my family during this study. ii TABLE OF CONTENTS Page List of Tables ........................................ v List of Figures ....................................... vii INTRODUCTION .......................................... l Cover-Trout Relationship ........................... 1 Cover Habitat Definitions ......... - ................. 5 Study Objectives and Hypothesis .................... 7 STUDY AREA DESCRIPTION ................................ 8 General Basin Characteristics ...................... 3 Study Area Characteristics ......................... 11 Fishes ............................................. 11 METHODS ............................................... 15 Study Area Preparation ............................. 15 Trout Population Investigation ..................... 15 Electrofishing Procedure ........................ 15 Calculation of Population Estimates ............. 19 Overhead Cover Rating Procedure .................... 24 Construction of Half-Log Devices ................... 26 Measurements and Observations of Half-Log 27 Devices ....... ‘ .................................... Water Velocity .................................. 27 Underwater Observations of Fish ................. 27 Invertebrate Colonization ....................... 32 Stream.Mapping Procedure ........................... 32 Analysis of Data ................................... 33 Habitat Alteration Effects ....................... 33 Cover Rating .................................... 33 RESULTS AND DISCUSSION ................................ 34 Trout Population Estimates ......................... 34 Overhead Cover Rating .............................. 42 Physical Evaluation of Half-Log Shelters ........... 46 Accumulation of Silt and Debris ................. 29 Water Velocity Measurements ..................... iii Underwater Observation of Trout Beneath Half-Log Devices ................................ 49 Invertebrate Collection .........; ................ 50 Trout Population Response ........................ 50 Cover-Trout Correlation .......................... 65 CONCLUSIONS ......................................... 79 LITERATURE CITED .................................... 80 APPENDIX ............................................ 84 iv Number 10 11 LIST OF TABLES Station lengths of the study area on the Pigeon River ............................. Permanent fin-clip markings administered during fall population estimates to all young-of-the-year trout on the Pigeon River... Electrofishing dates for trout population estimates in the study area of the Pigeon River ........................................ Spring 1979 trout population estimates for the study area on the Pigeon River ........... Fall 1979 trout pOpulation estimates for the 1.5 km study area on the Pigeon River ........ Spring 1980 trout population estimates for the 1.5 km study area on the Pigeon River .... Fall 1980 trout population estimates for the 1.5 km study area 3/0 In (9. sum) DIAMETER' nsmroncmo Roo_" Figure 1. Side view of a half-log cover device \ a. A!" , 1 “ {g '1‘), «3" V5191"; lfifi 1y ”/./’ 1 ‘ 1v Sléiéfi - . t . I/“l I [”24“ 1" Figure 2. A half-log device situated in a stream. in trout food organisms in the pools formed by the structures. But they stated that more evaluation of the direct effects on fish abundance was needed. Boussou (1954) demonstrated the relation of cover and trout in.Montana streams by removing brush cover and overhanging banks. This resulted in concomitant decreased trout populations. Lewis (1969) found that cover accounted for most of the variation of the brown trout populations in Montana streams. Enk (1977) found that bank cover accounted for large portions of the variation of brook and brown trout abundance in selected sections of tonMichigan streams suggesting that bank cover is a major factor in establishing the standing crop of larger trout in streams. He found 76 percent of the variation in July numbers of trout 150 mm or greater in length, 81 percent of the variation in July trout biomass of fish 150-399 mm in length, and 80 percent of numbers of trout 150 mm or greater in length in October were related to the length of bank which offered overhead cover in the forms of logs, vegetation and undercuts. Lack of suitable cover limits adult brook trout pro- duction in streams having sufficient food supply (O'Conner and Power 1976, Saunders and Smith 1962). Hunt (1971) demonstrated increases of wild brook trout standing crop following man-made modifications in a Wisconsin stream. The trout populations were found to be correlated mostly with pool area and permanent bank cover. The habitat manipulation consisted of series of bank covers and current deflectors placed alternately on opposite streambanks. This increased pool area 289 percent and permanent bank cover 416 percent, resulting in a mean annual brook trout bio- mass increase of 41 percent. This illustrated the need to aim stream improvement at specific limitations to trout production. Suitability of cover for trout is influenced by water velocity (Vincent 1969). Position choice by trout is near areas of sufficient current velocity (Jenkins 1969) with most favorable positions having the greatest velocity difference between the point where the fish is located and the nearby current velocity, termed the "water velocity difference" by Fausch (1978). This would allow minimum energy expenditure to maintain a position, while maximizing food flowing past the position (Fausch and White 1980). Stream-dwelling salmonids display territorial behavior (Kalleberg 1957) or defend a private space termed a ”social force field" (McBride 1964, in Butler 1965). The size and number of defended positions in a stream are strongly in- fluenced by such factors as species and size of fish, current velocity, streambed irregularity, activity and season, but are only slightly influenced by the density of the fish population at "normal" levels (Allen 1969, Butler 1975). This phenomenon leads to optimal use of environmental resources by the population, allowing maximum survival of the group consistant with the resources (Chapman 1966). The need of larger fish for cover is the basis for stream improvement (White 1979). The addition of suitable cover in a stream should result in an increase of trout abundance by encouraging a more efficient use of the stream resources, assuming cover to be a major limiting factor to trout production. The half-log structure was designed to speci- fically provide overhead cover for larger trout. In a Wisconsin stream this cover enhancement technique increased the April numbers of 250 mm and larger brown trout from an average of 22 per kilometer in pre-treatment years to 283 per kilometer after 3 years after treatment (Hunt 1978). Cover Habitat Definitions Trout production in streams would seem in most cases to be largely a function of the available amount of suitable cover. It follows that stream trout behavior should accomo- date this requirement for suitable cover habitat. A trout cover rating method has been developed by Wesche (1976) using cover, trout standing crop and hydrological data from small, predominately brown trout streams in southeastern wyoming. While his primary goal was to use this rating scheme as a trout habitat quantifier for in— stream.flow needs, wesche defined some minimum criteria for suitable bank and instream cover for trout. Trout 150 mm or longer used overhead bank cover more often (72.8 percent) than instream cover (27.2 percent), whereas, the trout less than 150 mm used instream cover and overhead bank cover about equally (51.6 and 48.4 percent, respectively). The overhead bank cover wesche found most of the trout using had water depth of at least 15 cm beneath it and width of 9 cm or more. The instream rubble-boulder areas that contained fish had water depth of at least 15 cm and sub- strate diameter of at least 7.6 cm. These criteria for instream cover may not be fully applicable to Michigan trout streams where instream cover is composed mostly of logs and brush piles. Therefore, the instream.cover definitions were modified to meet these differing conditions in rating Michigan trout stream cover. DeVore and White (1978) reported that 25-30 cm brown trout show a greater propensity for cover located close overhead. The study fish preferred overhead cover 10 cm rather than 15 or 20 cm above the streambed. This study was conducted in a man~made raceway that afforded no bank cover. I devised a trout cover rating method using the overhead instream and undercut bank criteria patterned after the studies discussed. Enk (1977) showed that the bank cover definition by Wesche is applicable to Michigan trout streams. The cover spacing results of DeVore and White and cover width requirements reported by Wesche directed me to define instream cover as any permanent cover having at least a lO-cm space above the stream bottom and a width of at least 9 cm. Study Objectives and Hypothesis The primary objective of this study was to determine the effect of half-log obver on brook and brown trout standing crop in a section of the Pigeon River, Otsego County, Michigan. Secondly, the relationship of cover to trout was examined by correlating trout standing crop to the cover rating. Overhead cover is assumed to be a major factor deter- mining the standing crop of brook and brown trout in the study .sections of the Pigeon River. I hypothesize the addition of half-log cover will increase the trout standing crop in the study sections. STUDY AREA DESCRIPTION General Basin Characteristics The Pigeon River is located in the north-central part of the southern peninsula of Michigan. It arises north- east of Gaylord, Michigan and flows northerly through Otsego and Cheboygan counties for 70 km (42.5 mi) before emptying into Mullett Lake (Figure 3). The Pigeon River has a drainage basin of 422 km2 (165 miz). With the headwaters at 348 m (1140 ft) above sea level and the mouth at 181 m (594 ft) the gradient is 2.4 m/km, which is rather typical of streams in this area. Emanating on the inner slope of a moraine, the Pigeon River flows primarily through sand and gravel outwash, remnants of the last glacial period. The sand and gravel produce great amounts of groundwater discharge resulting in relatively constant flows. The Pigeon River is mostly wadable throughout its length. Gravel beds predominate the upper reaches while the lower reaches become sandier. Coniferous swamps pre- dominate in the upper portions with a downstream trend to hardwoods in the lower floodplains (Hendrickson and Doonan 1970). Figure 3. Location of the Pigeon Piver and the study area. MILE. ’ an 1' son rssu our: new IIOIIICM 7’3" VMT Luna. “A 3mm 0 Po STUDY AREA um GAYLORD ‘ a 3' R 2' " " Figure 3 . wate‘ tIOU1 towns secti 1.2 1 map) peri: char at t thrc aSpe ER.) Witl 8111 (83] whit 11 The upper reaches of the Pigeon, with good cover, water and streambed characteristics, provide an excellent trout fishery. Study Area Characteristics The study area is located in Section 36 of Corwith township (T32N, R2W), Otsego county (Figure 3). In this section, the Pigeon River flows northwesterly and has 1.2 sinuosity (measured on a 7.5-minute quadrant topographic map). Henderickson and Doonan (1970) report the low-flow period is usually in midsAugust with a mean monthly flow at this time of 1.13 m3/s (40 cfs), with the highest dis- charges normally occurring in.April at a mean monthly flow at this time of 3.68 m3/s (130 cfs). The 1.5-km section chosen for study (Figure 4) flows through brush and coniferous swamp with some birch and aspen. The stream is bordered primarily by alder (Alnus s2.) and northern white cedar (Thuja occidentalis). The Streambed in the study area is sand and gravel, with some cobble, and has moderate amounts of aquatic vegetation. The bank margins of the streambed are heavily silted. Fishes Fishes of the study area are primarily brown trout C§§1gp tggtta), brook trout (Salvenlinus fontinalis), and Vflhite sucker (Catostomus commersoni). Sculpins (Cottusgsp.) 12 Figure 4. The 1.5 km section of the Pigeon Fiver studied. Numbers refer to station markers. 13 :Old Vanderbilt Rd. ' . Downstream limit ' FOOTBRIDGE 6 . 7 . s ‘ 9 IO ‘ FOOTBRIDGE l2 __HALF-Los TREATMENT SECTION I3 . I4 l5 .mLE .l .2 ~ l6 .i 3 L3 17 ' KILOMETER .9 ‘ l9 A O “\ .Figure 4. 14 creek chubs (Semotilus atromaculatus), dace (Rhinichthys sp.) and unidentified lamprey ammocetes were also present. METHODS Study Area Preparation The study area on the Pigeon River was divided into the 15 stations, each approximately lOO-mrlong (Table 1), as close as possible to the stations used by Enk (1977). Old station markers from Enk's study were replaced, where found. The remaining stations were measured to 100 m and also marked. I marked each station on the near-stream vegetation at both sides of the stream with bright colored surveyors flagging. Each station was also marked further away from the stream on more permanent vegetation (large bushes or trees) for future reference in the event that near- stream markers were lost. Each station marker was numbered with an indelible pen so that the station number could be seen from the stream” The number of each marker designates the upstream.border of its study station. Trout Population Investigation Electrofishing Procedure The trout populations in each section were sampled by electrofishing. A 250 VDC, 1.75-kw gasoline-powered 15 16 Table 1. Station lengths of the study area on the Pigeon River. Station Length (m) 5 101 6 98 7 120 8 104 9 106 10 110 ll 98 12 110 13 104 14 115 15 103 16 107 17 78 18 119 19 94 17 generator (Pow-r-Gard brand) was carried in a 2.6 m (5.6 ft) long portable electro-fishing boat. The unit was also equipped with a fish holding tub, a galanized sheet metal bottom cathode and three anodes. Each anode had a 4.8 mm (3/16 inch) diameter stainless steel rod bent in a diamond shape (about 25 cm.long and 15 cm wide) attached to a 1.2 m (4 ft) fiberglass handle, wired to a retractable reel mounted on the front of the electrofishing boat. This allowed each anode to be pulled as much as 6.1 m (20 ft) away from the boat and retracted. The electrofishing unit was operated by a crew of three wading in the stream, each using an anode and a hand net. One crew member towed the boat via a harness as the electro- fishing progressed upstream. Electrofishing requires knowledge of stream fish be- havior and a quick netting hand. The three electrofishermen probed the anodes into the stream cover to draw the fish out of hiding spots. The electrofishing was concentrated in areas with cover where trout were most likely to be. Fish were netted and transferred to the holding tub on the boat. The fish captured in each station were kept separate. At each station marker, a separating "ring net" was placed in the holding tub. After two or three stations were completed, the fish were taken off the boat and placed in the stream in a tub perforated to allow water flow. With a new holding tub in place, the electrofishing continued up- stream. 18 Two other crew members then processed the fish left behind in the tub. The trout were anesthetized with MS-222 before being measured to the nearest millimeter (total length) and weighed to the nearest gram. Any previous markings were noted or markings administered for age or population estimation. All data were recorded on a standard form. After the fish were weighed and measured, their lower caudal fin lobe was clipped. This marking was not permanent, but lasted the duration of the population estimate. All fish were revived, then released downstream of the station where they were captured, to allow them to re-orient to their home stations in the stream.by smell. Trout population size was estimated by a mark and recapture method. This estimation technique required a marking run followed by a recapture run. A period of one week separated the two runs to allow the marked trout to become distributed into the population. Opercule punch markings were used for the first es- timate. This technique was abandoned, as the smaller marked fish began developing eroded gill opercules. On the recapture run, lengths were measured on all captured trout, while only the unmarked fish were weighed. All fish.were marked by clipping the top of the caudal fin. This prevented multiple counting on the following day in the event that the whole study area could not be electrofished in one day. F nation the-ye year a analy: trout and l in 19 and enha and Spri Calc \ mod 197 Whe lati 19 Fin-clip marks were also used for future age determi- nation. During the fall population estimates, all young-of- the-year age were given permanent fin-clips. Young-of—the- year age designation was determined by a length-frequency analysis, whereby 110 mm was chosen as the upper limit. All trout less than or equal to 110 mm.were marked with adipose and left ventral fin clips in 1979 and adipose fin clips in 1980 (Table 2). Trout population estimates were conducted each spring and fall during the study (Table 3). This yielded a pre— enhancement population estimate for the spring of 1979 and three subsequent estimates in the fall of 1979 and spring and fall of 1980. Calculation of Population Estimates Trout numbers for each species were calculated by the modified Petersen estimate formula (Seber 1973, Ricker 1975): we a: [$9-I-1-2E$£-'{-9-l‘-lil_ 1 where; N* - population estimate (stock density) m - number of fish marked during first run r = number of fish recaptured during second run u = number of unmarked fish captured during the second run. Ninety-five percent confidence intervals for the popu- lation estimates were approximated by the Poisson, bionmmial 20 Table 2. Permanent fin-clip markings administered during fall population estimates to all young-of-the- year trout on the Pigeon River. Year Fin Clip 1979 *Adipose - left ventral 1980 . Adipose * note: some trout received adipose - both ventral fin clips. Table 3. Electrofishing dates for trout population estimates in the study area of the Pigeon River. Season and Year Marking Run Recapture Run Spring 1979 June 9 & 15 June 23 Fall 1979 September 2 & 3 September 7 & 8 Spring 1980 June 8 June 15 Fall 1980 September 11 September 19 21 or normal distributions. The ratio of recaptured fish to the total number of fish captured on the second run (r/r + u), determined which distribution is applicable to an estimate as recommended by Seber (1973). The Poisson approximation was used when this ratio was less than 0.1 and r/m was less than 0.1. The Poisson confidence interval of r was read off a table in Ricker (1975). The numerator of the estimate formula was divided by the upper and lower r values (with 1 added) to find the upper and lower 95 percent confidence limits for N*. If r/r + u was less than 0.1 and r greater than 50, a normal approximation to obtain the 95 percent confidence interval for this ratio was used. This interval was calculated by r/r+u : 1.96 *El - r/m)(r/r+u)(1-r/r+u) / (r+u-1) 35 where (l - r/m) can be neglected if r/m is less than 0.1. The upper and lower r/rwtu.values were inverted, multiplied by m.+ l and subtracted by l to give the confidence interval for N*. When r/ri-u was greater than 0.1, either the bionomial or normal approximations were used. The smallest values of the sum.r + u for which the normal approximation is applicable is given by the following table from Seber (1973: 64): 22 p(or 1 - p) 0.5 0.4 0.3 0.2 0.1 r + u 30 50 80 200 600 If the normal approximation was applicable, confidence limits were calculated by the aforementioned method. When not, the bionomial approximation for the 95 percent con- fidence interval for r/r-+11 was used. These values, read off a chart from Pearson and Hartly (1966), were multipled by r + u to give upper and lower r limits. These r limits were then applied to the Petersen estimate formula to get the upper and lower 95 percent confidence limits for N*. In some instances, the calculated lower limit for an estimate was less than the m.+ u-value for that estimate. In these cases, the m + u-value was used as the lower limit. To correct for the size-related bias displayed by stream trout when captured by electrofishing (Cooper and Lagler 1956, Vincent 1971, Bohlin and Stundtrbm 1977), especially for the smaller fish, separate population estimates were made by 25-mm length groups. When the number of fish re- captured or the number of total fish captured in a particular length group was so small as to produce an invalid estimate, length groups were combined to decrease the variance of the estimate while avoiding the systematic bias. Estimates for fishes 200 mm and greater were usually grouped to a greater degree, since their susceptability to capture is more or less equal. 23 Population estimates for the study area were prorated into estimates for individual stations according to the number of fish marked in the marking run (m) plus the unmarked fish caught in the recapture run (u). Proration factors were cal- culated as the population estimate divided by the total number of m + u fish for each length group. These factors were then applied to the number of m,+ u fish per length group in each station. In spring 1979, trout populations were estimated by proration iJI only 10 of the 15 stations. Lengths of re- captured trout had not been recorded in the five downstream stations, preventing population estimation by length class for these stations. Fish captured from adjacent stations were sometimes mis- takenly mixed during electrofishing. When this occurred, the data from the stations were combined and treated as one station. This was the case for stations 8 and 9 in fall 1979 and for stations 6 and 7 in spring 1980. Trout biomass was estimated for each station. The mean weight of all m,+ u fish in each length group was multiplied by the population estimate of its length group. The biomass estimates of the length groups within each station were summed to obtain total biomass for the station. Trout standing crop was determined by dividing total biomass of each station by the length of stream in that station. This procedure, resulting in grams of trout per mete allc leng ‘ra til W8 ,4 24 meter of stream, Standardized the sectional trout abundance, allowing direct comparison between stations of differing length. Upper and lower confidence limits were estimated for the trout standing crop in each length group. The 95 percent confidence upper and lower bounds for the population estimate of each section was multiplied by the mean biomass of the respective length group. Overhead-Cover-Rating Procedure The trout overhead cover habitat was rated by the cover rating method adapted for Michigan streams I developed for this study (see page 6). The overhead cover in each station was measured in mid- to late-summer during the period of low flow (July 9, 10, 17 and 18 in 1979 and July 31 to August 2 in 1980). The limitation on trout standing crop by cover is assumed to be expressed to a large degree at this time (White 1975b). The cover ratings for each year were performed at equal flows. No ratings were performed subsequent to rainstorms. The water depth at a culvert, acting as a control wier, just downstream of the study section at the Old Vanderbilt Road was measured each day during the cover rating periods to insure equal flows. A gauge, adjusted to the minimum depth and width re- quirements of the particular cover being rated, was used to determine the amount of overhead cover in each section. The gauge was made of two lengths of wood dowel attached to each 25 other perpendicularly by a two-way clamp. The length of cover beneath which the gauge fit was measured with a re- tractable steel tape measure. Undercut bank cover was measured along the length of bank meeting the defined minimum criteria; if the bank was curved, the tape was contoured to the bank. Instream cover was measured along the length of the instream cover that met the required dimensions no matter what orientation to the current. The instream and bank cover was summed to give total cover. I rated instream log jams and fallen trees differently in 1980 than in 1979. High 1979 cover ratings without con- current high trout populations in sections with much in- stream log jams and fallen trees led me to believe that these types of instream cover were not of the same value for suitable instream overhead cover. I feel that this cover is of some value as feeding stations and current shelter, but it is not of such value for hiding and resting as is cover situated closer to the streambed. In 1980, I measured the log jams only along such length as met the minimum spacing requirements, as opposed to along their entire perimeters as done in 1979. The response of brown trout to close overhead cover re- ported by DeVore and White (1978) may also be applicable to overhead bank cover. A lO-cm depth overhead bank criterion was compared to wesche's 15-cm criterion. The amounts of 10 cm and 15 cm.high bank cover and in- stream cover in each study section were standardized by di th ra de 26 dividing each amount of cover by the respective length of the study section. This allowed comparison of the cover ratings of the different length sections. The amount of instream cover provided by the half-log devices in both years was measured separately. Construction of Half-Log Devices Thirty-one half-log structures were installed in stations 12 and 13 during July 6 to 16, 1979. Each half-log cover structure was constructed from.white cedar and 9.5 mm (3/8 inch) diameter metal reinforcing rod. White cedar was used because of its abundance in the study site and because of its durability (Tarzwell 1937, Behr 1978). Trees of 20-30 cm (8-12 inch) diameter were cut into 2.5 m (8 ft) long logs, then split longitudinally. This produced fairly flat bottomed half logs, but any irregularities were trimmed with an axe. Holes (13 mm diameter) were drilled through each half-log approximately 20 cm (8 inches) from each end. Each half-log was supported 10 cm (4 inches) off the stream bottom with two cedar spacer blocks. These were 15 cm (6 inch) diameter logs cut into 10 cm (4 inch) lengths. Holes (13 mm.diameter) were bored through the center of each spacer. The preparation phase described above was completed at a site upstream of the study area. After the half-logs and spacers were cut and drilled, they were floated downstream to stations 12 and 13. 27 The half-logs were placed in areas of the stream with stable substrate and along the thalweg to take advantage of the greatest water depth. The logs were spaced an average of one per 6 meters (1 per 20 ft), after Hunt (1978). The half-log devices were pegged to the stream bed as in Figure 1. Each was oriented at a slight angle to the current to provide some flow underneath for reducing silt accumulation. The distributions of half-logs in the treatment sections are shown in Figures 5 and 6. The half-log devices were numbered 1(downstream) through 31 (upstream). Measurements and Observations of Half-Log Devices Water Velocity The water velocities around and beneath selected half- 1og devices were measured on July 25, 1980. The impeller from a Swoffert flow meter was removed from its stand and hand-held for the various water velocity readings. Water velocities were measured (1) 10 cm above the bottom just upstream.from the half-logs, (2) beneath the half-logs behind the upstream spacer, (3) beneath in the middle and (4) beneath just in front of the downstream spacer. Surrounding water velocities on both sides and above each half-log were also measured. Underwater Observations of Fish Wearing a wet-suit, mask and snorkel, I observed the half-log devices on July 26 and 30, 1980. I crawled and 28 Figure 5. Study section 12 of the treatment section. Numbers in parantheses denote the thalweg depth in centimeters. Depths were measured when stream discharge was 1.05 m3/sec (37 cfs). 29 I 2 24? ® 3 :1, l/ me an: N 4 h '1’? / 5 ‘1‘ ISLAND c— U '7 0 .44 'm) 9— \ O“. \O 1. IO NO" X x, . I .1. 0‘ ‘.\ ll mf\ ‘\‘\ l2 U.“ \ l3 motors ‘ in: 3 0 IO "'i‘ ’ \ I4 ..... Tholvoo ' 1:3 Hon-Io! " Devin '- ® Station ‘ ' \" i5 Figure 5. 30 Figure 6. Study section 13 of the treatment section. Numbers in parantheses denote the thalweg depth in centimeters. Depghs were measured when stream discharge was 1.05 m /Sec (37 cfs). 31 MO" 0 i0 .......- Thai-cg =3 Halts-log Device ® Station @ l7 Figure 6. 32 swam upstream approaching each half-log device cautiously to determine if it was being used by fish. I recorded the number, species and sizes of fish using the device. The fishes' positions under the device were recorded. I also characterized the devices by amount of cover provided under the half-logs one year after installation. Invertebrate Colonization Stream sections with high invertebrate drift rates might be expected to have higher fish carrying capacities (Waters 1972). To determine the degree to which the half- 1og devices were functioning as substrate for fish food organisms, qualitative samples of the invertebrate popu- lations on selected half-logs were taken on July 30, 1980. Five half—log devices from different areas in the treatment sections were scrubbed with a brush while a wire-mesh screen was held immediately downstream to collect dis- lodged organisms. Sections of bark were also removed and collected. The captured invertebrates from the screen and the pieces of bark were preserved in alcohol. The insect larvae were identified to genera using Hilsenhoff (1975), and all other organisms were identified using Pennak (1978). Stream Mapping Procedure The treatment sections were mapped on August 1 and 2, l980_using a compass traverse method (Compton 1962). 33 This technique allows a meandering course to be plotted by measuring the distances and the.bearings of several small, straight transects along the winding course. Analysis of Data Habitat Alteration Effects Trout standing crop estimates for the treatment section were compared to those of the reference sections before and after the half-log devices were installed. Pre- and post-treatment estimated brown trout standing crop, brook trout standing crop and a total trout standing crop were compared for all size classes and for trout greater than or equal to 150 mm in length. Cover Rating The trout standing crop association with overhead cover rating was analyzed by simple regression. The spring and fall trout standing crop estimates were regressed against the standardized instream.overhead cover, 15-cm and lO-cm undercut bank cover, and total overhead cover ratings for each section during both years of the study. The specific associations of trout standing crop with the cover ratings were analyzed by comparing standing crops for the trout of different size classes. Standing crops of trout greater than or equal to 150 mm, as well as of the 150-399-mm length group were regressed against the cover rating. RESULTS AND DISCUSSION Trout Population Estimates Length class groupings for the population estimates re- sulted in species—specific patterns (Tables 4-7). Brown trout attained greater lengths than did the brook trout in each estimate. Brook and brown trout previously marked by Enk (1977) as young-of—the-year were encountered during the study (Table 8). The scarce number of these fish encountered, especially for the brook trout, disallowed definitive con- clusions of age and growth for either species. Seasonal patterns also characterize the length class groupings for the population estimates (Tables 4-7). Spring population estimates for both species were calculated with fewer length classes than the fall population estimates. Higher streamflow discharge during spring created poor electro- fishing conditions, such that smaller portions of the popu- lation were sampled in the spring than in fall. A less complete sample required combining into larger length groups to provide valid estimates. Lower recapture efficiencies for the spring, compared to fall estimates, reflect this seasonal trend. ll‘ FT. 1‘ I. I‘ll. lilt 11:01.15] If‘ll I‘ AvN Ru h .-.r\. Nmm omH cum HmuOH o.o m o o n cos M o o o ammumnm o H H qanomm H o N mqmummm m.m~ ow Hm as o o H «Nmuoom q o o mmmnmmm m 0 HH enmnomm . w a m mqmummm m mm Ho mm mm m m m Mmmummm . N H. H c om mMH om mm 0H 3 m quuomH . NH H 0H quumNH m m wNN mm mNH %.o o H qNHTOOH uoouu sebum ”w Hoe emN mam Hmuoe N o H Summonu H H m mqwnmmm H H m «NNToow «.mm mom mmH mHN HN oH MH mmHnmmH mm oH we qunomH . . om N oH quumNH 0 ma «cm on ems m H H a s--ooa uoouu xooum HNV hocoHonwo Home: HoSOH oumEHumo 5 H E AEEV mooum ousuamoom muHEHH EOHumHomom nuwcmq mommeamcoo ems o£u co mono Senna map How moumEHumo GOHumHsmom udouu man mcHHmm .Ho>Hm aoome .q oHan . V 0‘ sf...- .v- Vnt.‘ - vs... - 5.... DIN. ‘ .i luv H I-h.‘- -.h s... \.a\\ ‘w L. ~ \ s I i l ' ~ .\~ \ 36 H m a mamumNm m H H quuoom N.qn oHH we we a e w mmNumNN m N N SNNnomN q n c quumNN o.o no He mm NH NH «N qNNIOON o.N¢ HwH NOH mmH Hm 0N Ho mmHnmNH m.mm NNH mm «m 0H 0 NH «NHuomH m o H N qunmNH m.mm. o o H qNHnHHH m.NN Noe HnH me Ho a OS OHH W usouu zzoum HHN on mme Hmuoe H H H moNummN w.Nn MN 0H NH o N c «NNuomN N q q quumNN N.H¢ on He on N cH em cNNTOON «.mm HNH moH omH He mm ow mmHnnNH H.wm mmH mm ONH mm on No quuomH . NH MN mcHumNH 0 we mm mN mm H o N «NHTHHH 1». mH mNN mm mm 0H N NH OHH V: uaouu xooum ANV SocoHonwm Home: Hoon oumEHumo o u E AEEV enouw manuamoom muHEHH COHumHamom numcoH mocovacoo Nmm .Ho>Hm coome osu mo mono monum Ex m.H map you moumEHumo EOHDoHsmoa usouu aNmH HHom magma 37 mHOH mum me HouOH o.o H o o H 2: N fie o H ammumnm H o o «mmuomm unouu czoum ANV mocoHonwo Home: HoSOH oumEHumo a H E AEEV macaw ousuamoom muHEHH EOHuwHamom . Summoq monopHmsou Nmm A.e.ucoov .m wanes p... nerr-¢~ \fl-vu..u.,~ .: v. H..- . q, Psi-N s T..v — n~.v .71... u.._.._.. ...u N s... ~ -.~..A~ ..-.A-..~ u «\uxxvfi nx-\,~.\..l.. rs nyx .\u..\. 38 NHm Non New kuoe 0.0 q o o 3 cos M o o H mmmamNm H o m quuomm H o n aqmumNm m.o¢ qu HN om N a w «Nmnoom o m w mmNumNN mH HH 3H qNNuomN , .. ON w 0N mQNnmNN a.ee sea as as Ha a N a--oo~ ..-. q H H mmHumNH o mN ma me an NN m Ha «Na-oma m. a. a as usouw ssoum one NoN mos HEDOH o m e qNNnomN mH m mH qunmNN ¢.oq oNq wwH NoN mH 0 HH qNNTOON aH H OH mmHumNH «m m NH «NHuomH Q.os emu as ems Mm m MN wwwummw unopu xooum HNV NocoHonmo Home: uoaoH oumEHumo s H E HEBV macaw ousummomm muHEHH SOHumHsaom numcoH NUSGUHHCOU Nmm .Ho>Hm coome ofiu co mono Spoon Ex n.H onu How moumEHumo COHuwHDmom unouu ome wcHumm .o oHan hf. .. u u . F-AU .ul 1...... > A U . .—-‘ Uhkoi -Av‘ ' v v F fink P w A. V ‘ .- .— . ~. A ’\‘U ‘o g 5 l ‘T .. ~ ~ha'.. .- - - v-r - 5- Av . . . . . Q.- -uh~fi AN Q . , T . ~ ~ .~ . .\ 5.5 N \ .»~ 39 . N a wH as~-nN~ 0 ON ms Hm om OH mH oH eN~-omN . . H H a aaN-m- m an maH mm mm AH w mH s--oo~ N.eo SNH NOH mHH mm as so aaH-mNH m.He aoH Hm mm mu «m mm «NH-omH . m m o meH-m~H o om ow w HH o o o «NH-HHH m.om NNH mm as ON N om OHH w usouu GBOHm mNN mmm was Hmooe 3... s a z m M w ”mam . N a n an-mN~ 0 ea om wH mH Mm N w emm-oom .. m NH «N SSH-mNH n as SSH «SH ASH Maw om HN «HH-omH .. A .. a H. H. mm HAN m? 3H N2 S NH 2 SH m. usouu xOOHm HNV NocoHonmo moan: Ho3oH ouoEHumo D H E HEEV moouw ouaummoom muHEHH GOHuoHaaom nuNCoH mesmeHHcoo ems .Ho>Hm coome osu no moum Nesum Ex m.H onu pom moumEHumo GOHDoHDmoa usouu owaH HHom .N oHan 40 noN omm NHm Hmuoy 0.0 o o m N o H 23 M H m a mmmInNm q o H onuomm N.mo mm me mm o N m mqmumNm m O NH «Nmuoom usouu a3oum ANV NocoHonwo momma Ho3om oumEHumo a u E AEEV moouw ououamuom muHEHH GOHuoHdmom nuwcoH oocmpHmcoo Nmm 1.383 .N 33 Fh IlfIL I - I. FIFTW ah I I Iflilh F-h1bfhvfi f h f. 1H! 11 F..\Cl.- f. ‘1‘. I... IngFI ll pic IITN ll (10.1.0 I..\ .7 l‘ I! ‘0: lac: I..I-nl I III; (9‘ .. 1:! I I... a L _ _ — H u \— r u q H - H < 3 H N L. 41 commHHo CHM Hmuuso> umoH >4 pommHHo ch Hmuuco> uanH u >m k. m ooquoNN SNm .HHH >m usouu caoum ome HHmm H was >H >H uoouu :30Hm N moN-HoN mmN HHH >m “sous sooom ommH weHuam H oqN. HH >m uaouu xooum m ommanN NoN HH >m unouu crown NNmH HHmm N ANN-oON NmN HH >a “sous agoum NNNH NEHHNm Hoasbz AEEV owcmm ABEV zuwcoH ow< «mHHocHh moHoon mums coo: H: monouemo usouu czoun mam xooun poxums NHm50H>on mo nuwcmH some paw ow< .cmenon .Nucboo owomuo .Ho>Hm coomeHEfiH .N oHHme 42 The smaller size classes had lower recapture efficien- cies. The recapture efficiencies of very large brown trout greater than 399 mm) were usually near zero, but this size class was not grouped with the next smaller class, as it was necessary to keep these length groups separate for an- alysis of length-class biomass with cover. The estimated standing crops for the study area were 47 gm/m and 36 gm/m for the spring and fall of 1979 re- spectively, and 47 gm/m and 41 gm/m for the spring and fall of 1980 respectively. These figures are lower than the trout standing crop in this area reported by Enk (1977). He estimated the standing crop at 53 gm/m and 57 gm/m for July and October of 1977 respectively. Coopes (1974) reported the trout standing crop in the Pigeon River at 40 gm/m downstream in the trout research area. The Spring standing crop estimates for both years of the study were greater than those for fall. The higher spring estimates may have resulted from (1) inflated popu- lation estimates from the lower recapture efficiency in the spring estimates or (2) angling during summer. Prorated trout standing crop estimates per station for spring and fall of each year are shown in Appendix B. Overhead Cover Rating The measurement of 15-cm- and lO-cmrdeep overhead bank cover resulted in differing amounts of overhead bank cover in the study sections (Tables 9 and 10). In most cases, there was more lO-cm-deep than 15-cm-deep bank cover. 43 .uooH>ov uo>oo noHuuHon no aoHuoHHou-EH on» up you voussoooo uo>oo no unseen ouoochH nonozusouon :H uuoaauco Na. No. on. nN. oH. N.oo . a.Ho o.co o.HN c.NH co mH na. na. NN. nNN. HN. o.NHH n.oHH N.mo c.0N H.nN mHH oH Nc.H on. Na. co. Hn. N.no e.oN c.Nn N.oc e.an oN NH Hn.H aN.H co. No. no. n.ocH n.oNH o.uo a.HN N.mo NoH oH No. No. Nn. mm. mm. n.co N.ca o.cn m.nn H.on noH mH nc. cc. NN. NH. nH. o.Nn e.ne o.Hn c.HN e.eH nHH «H 3. 3. «SNJan. S. 3. 3.3 T? «363043 n.N n.n 3H NH no. we. chN.voN. oN. NH. n.nn o.oc «an.HNvaaon a.NN n.NH OHH NH Nn. oH. NH. ON. No. o.nn N.oH n.0H n.mH ¢.N no HH me. an. NH. Nn. HN. o.o¢ H.on c.NH N.nn N.NN OHH oH No. «a. on. NN. «H. c.nm m.oe N.Hn n.n~ N.nH oOH 0 NH. 0H. 9H. No. co. N.NH n.0H n.cH e.N «.0 «OH o me. on. nH. NN. «N. °.Hn o.oc c.NH o.nn o.oN oNH N an. Na. Na. nNc. no. n.nn n.nn o.Hn n.N n.c mm 0 Nn. on. Na. no. No. n.on «.00 H.nn ¢.n H.N HoH n scan BUTOH scan IoTnH noosuo>o sumov sueov soup ao-oH Jean lounH voonuo>o sumac canon Hav EOHunum waHon mcHo: soapy-EH EOTOH. abunH wcHo: mcHo: aiouuocu Bulb“ BoumH camcoH .uo>oo Houoh uo>oo HauOH “can vooauo>o uo>ou HouOH uo>ou Hogan mmmm moocuo>o anouuu no nouns noQTuo>ou mo «Noun: . uo>oo no cacao: .HoHuo hHsn vouaoooav uo>Ha cooNHm can no coHuou- nuoo now oucHuou uo>oo oNoH .a oHaoH 44 .no00>ov uoHnuHos osu up now voucsoouo uo>oo uo ucaoao vow-ochH coconucouon cH ouoaasco On. O0. O0. OH. OH. H.N¢ H.N0 o.Nn n.O ,n.O 0O OH an. O0. .Hc. NH. OO. n.n0 0.0n N.O¢ 0.0H H.HH OHH 0H O0. Nn. OH. HN. NH. 0.Hn n.0N N.nH N.0H n.OH ON NH nn. O0. ON. 0n. On. 0.00 N.N0 H.Hn N.Hn 0.Hn NOH 0H Hn. N0. ON. ON. ON. n.Nn N.Nn 0.0N O.nN n.0N nOH OH «O. no. 0N. OH. NH. 0.0n 0.0c 0.0N 0.0N 0.0H nHH OH HO. . Nn. «HOH.0On. HO. NO. O.Nn N.nn «H0.0H00.Hn n.H N.H OOH nH 00. n0. oHOH.0Hn. nH. NH. n.00 0.00 «AN.OH00.nn N.0H N.nH OHH NH ON. ON. OH. OH. 0O. 0.NN 0.0N 0.0H 0.0 0.0 0O HH 0n. an. NH. NN. ON. 0.0n 0.0m O.NH H.0N O.NN OHH OH O0. N0. 0N. 0N. NN. H.N0 0.00 n.0N 0.0N n.nN 0OH O 0H. OH. OH. 0O. 0O. 0.0H 0.0H 0.0H 0.0 0.0 OOH O N0. Ho. 0H. ON. NN. 0.0¢ 0.0e N.0H H.nn N.Nn ONH N NN. HN. OH. no. NO. O.HN 0.0N c.0H 0.N c.N OO 0 00. OH. 00. . 0O. 0O. 0.0c H.O0 n.00 n.e O.n HOH n soon IOTOH scan lunnH voonuo>o canon guano scan nouOH Hana lannH voozuo>o sunov suOoO Hav EOHuuum ucH-a . ncHos loouuocH auTOH. BOTnH NEH-s mcH-s loouuucu aUTOH aoumH suwcoH uo>oo Hooch uo>ou Hooch unanimoocuo>o uo>oo Hench uo>ou Honey scan moocuo>o aoouun uo wouoa hon uo>oo mo ououoz uo>ou 00 gauge: .HN unamsH¢ coome osu co ooHuouo soon now omcHuou uo>ou OOOH .OH oHnoH .— .0.-.-....- 45 Underwater observation revealed that areas of bank undercut by stream erosion were deep labrinthys held together by bank vegetation. Using 10 cm usually differentiated marginal undercut bank on the distal ends of larger undercut bank sections from lS-cm-deep bank. Much of the overhead bank cover in the upper stations of the study area was a result of excavations by bank beaver. This undercut bank, though meeting the spacing requirements and being quite extensive, was not a product of stream erosion. Most of the bank undercut by beaver did have flow of water along it, but not all of the bank undercut by beaver may have had water velocity attributes necessary for trout positions. The inclusion of beaver-excavated bank undercuts in the bank overhead cover measurements may have biased the cover rating. The crenulate stream bank associated with bank beaver was measured only along those portions bordering the main channel of the stream in 1980 and not along its entire lengths, no matter how far indented from the bank margin, as done in 1979. This is revealed in the lower undercut bank measurements, especially for the upper stations, reported for 1980 (Tables 9 and 10). Also, undercut bank not in erosional areas that was definitely a result of bank beaver was not included in the 1980 under- cut bank measurements. Instream overhead cover included log jams and fallen trees, solidly secured to the stream bank, that protruded into the stream and met the defined minimum spacing 46 requirements for the cover rating method. The different measurement of these instream cover types in 1980 resulted in lower instream cover ratings and lower total cover ratings than in 1979, primarily in the upper sections in the study area which contained relatively more deadfall trees and log jams (Tables 9 and 10). Installation of half-log devices initially increased the instream cover rating in the treatment sections 250 percent and the total cover rating 56 percent. The half- log devices provided 84 percent and 77 percent of their potential instream cover in 1979 and 1980, respectively. This is based on each half-log device potentially providing 1.63 m of instream overhead cover, the distance of overhead cover between the spacers. The decrease in cover provided in 1980 to that provided in 1979 is a result of silt and debris collecting under the half-logs. Small amounts of re- cently beaver-cut alder were removed from beneath the structures before the 1980 cover ratings. Gross silting or more permanent debris collected by the half-log devices were not removed for the 1980 cover rating. Physical Evaluation of Half-Log Shelters Accumulation of Silt and Debris The majority of the structures showed little Sign of sediment accumulation a year after installation. Stream- bed irregularity caused some of the space under the half- logs to be less than 10 cm high. 47 Three of the 31 half-log devices were rendered par- tially or totally useless for overhead cover by silt and debris, each case a result of poor placement. Half-log number 10 (Figure 5) was in a shallow area where the thalweg was crossing-over from one side to the other, which I feel caused debris to collect under the log. Half-log number 6 was on the outside of a deep pool in a meander bend. Erosion along the outside of the bend evidently caused some bank to slough in and partially cover this de- vice. The third structure with silt accumulations, number 26 (Figure 6), was downstream of a Llog jam that was moved either by ice or high flow. The thalweg moved as a result, placing the log in a depositional area. Half-log devices require maintenance to rectify place- ment problems. The dynamic properties of a stream result in changing areas of depositions and erosion during differing flows (Leopold and Langbein 1966). Over time, some half- logs are undoubtedly going to be silted in. Also, any up- stream activity that introduces debris or sediment into the stream will necessitate renewed clearance of debris from the half-logs. Water Velocity Measurements The water velocities found under the half-log devices were much less than the velocities in front and beside the structures. The lowest water velocities under the half- 1ogs, in most instances, were in the upstream area just behind the front spacer (Table 11). Wesche (1976) reported 48 ¢.q¢ m.HH o.wN O.w m.N «.0 0.HN amoE H0 NH as m HH 0 NN Hm oa NH NH H m H m cm as «N NN o NH N NH N Na NN NH N a e NH N «H o N o o m N n ma mN Nm 0 «H NH em a N0 NH mm He 0 c an N m>oma umem HHmH name wHeeHz snows was on“ on-HHm= onTMHmn wchasouunm slo ouauoduum w0Hanmn nuooaom owmmwmflmmH mo awesome: .mucoEouammoE huHooHo> HoumB o5» waHHDO Ammo NmV oom\m8 no.H on On poGHEhouop mm3 owumaomHn .mooH>oO onumen wouooHOm oEOm Ocaouw paw snowshoes: Hoom\Eov mmHuHUOHo> Houo3 .HH oHHMH 49 that trout prefer velocities less than 15 cm/sec, and the half-log structures provide these. If trout chose positions in areas of low velocity to minimize energy expenditure, positions under the half-log structure near the upstream spacer would be most favorable. Underwater Observation of Trout Beneath Half-Log Devices A total of nine brook and brown trout were observed beneath half-log cover devices during the two underwater observation sessions. The trout seen were approximately 15 to 20 cm.(6 to 8 in) in length. They were positioned just behind the upstream spacer block, using the forward water current refuge area. One brown trout was in the up- stream area, but facing downstream, apparently orienting into a back eddie under the half-log. No more than one fish at any time was seen using a half-log device. Kalleberg (1958) demonstrated that visual isolation is important in determining defended territory by stream trout. The agonistic nature of stream salmonids is also dependent on activity. Feeding positions are more vigorously defended than resting positions (Butler 1975). Bassett (1978) determined that a length of cover habitat could accomodate 7 to 10 times the number of adult brown trout than the same length of feeding habitat. The use of half-log structures by trout as feeding stations may limit their capacity only to one fish. This may not be the case during other times (i.e. winter or high flows) when the 50 half-log structures may be used for hiding cover, allowing more than one occupant. Intervebrate Collection The half-log structures provided substrate for an assemblage of chironimids, other dipterans, trichopterans, ephemerids and water mites (Table 12). The identified groups of invertebrates reveal an array of filter feeders, scrapers, tube builders, and predators. The generally mid- current position of the half-log devices supported the community of filter-feeding invertebrates. The bark on the structures provided substrate for populations on the half- log devices, evidenced by a covering of travertine, provided food for the scrapers. This invertebrate assemblage was preyed on by dipterans and water mites. The attractiveness of half-log devices to fish may also be a function of increased food organisms. This would be a major factor in streams with meager food production. Although no comparative drift samples were collected from the treatment sections, it is doubtful that the inverte- brates produced by the half-log enhancement would add sig- nificantly to the ambient production of organisms in the study area. Trout Population Response The trout standing crop point estimates indicate a greater standing crop in the reference sections than in 51 Table 12. Invertebrates inhabiting half-log devices. Taxa Habit* Functional Group* Diptera Chironomidae Rheotanytarsus , Clinger Tube and net builder Micropsectra Climber Sprawler Chironomus Burrower Tube builder Crictopus Clinger Tube builder and miner Pseudodiamesa 1e erei a Sprawler Predators on chironomid eggs and larvae Rhagionidae Atherix Sprawler- Piercer-predator burrower Tipulidae » Antocha Clinger Collector-gather (in silk tube) Simulidae Cnephia Clinger Collector-filterer Empidaidae Sprawler- Generally engulfers burrower Trichoptera Hydorsychidae Symphitopsyche Clinger Collector-filterer (net Spinner) Brachycentridae Brachycentrus Clinger Collector-filterer Glossosomidae Glossosoma Clinger Scraper 52 Table 12. (cont'd.) Taxa Habit* Functional Group* Ephemeroptera Ephemeridae Ephemerella Clinger, Collector-gatherer Sprawler, swimmer Caenidae Caenis Sprawler Collector-gatherer (scraper) Baetidae Baetis Swimmer, Collector-gatherer climber, clinger Centroptilum Swimmer, Collector-gatherer climber, clinger Hydracarina Clinger Predator-detritivor *Taken from Merrit and Cummins (1978) except Hydracarina which is modified from.Pennak (1978). 53 the treatment section, except for the pre-treatment estimate where the standing crop estimate was greater in the treatment section (Figures 7 and 8). A possible explanation for this unexpected result would be that the upper portions of station 13 may have been influenced by station 14. The reach in station 14 immediately upstream from station 13 contains deep water, good cover, and a relatively high trout density. Another factor that may have biased the spring 1979 population estimate in the treatment section was the two- week period required to attain a suitable mark-and-recapture estimate. High water during the recapture run resulted in poor electrofishing conditions and a second recapture run was needed. The extra week possibly allowed the movement of marked trout out of the treatment section, as evidenced by the smaller portion of unmarked fish in station 13 (10 trout) as compared to station 14 (23 trout) given similar number of marked fish in the two stations. The spring 1979 estimate for station 13 was greater than in subsequent estimates (Appendix B). A gross violation of the closed population assumption as implied above would lead to an over estimate. Interestingly, the standing crop estimates for trout in stations 12 and 13 reported by Enk (1977) were greater than the estimated standing crop for the remainder of the stream (Enk's stations 11 and 12 correspond to stations 54 .mome HHo Mom mHo>uoucH mocovaaoo NmO Ono mommaHumo Oouo wchcoum uaouu Hmuoe .N oustm om. om. ON. ON. .3... 8:8 .3... 2...... . _ . _ H H o 23:30 8:23.: 23.38 2253:. ....:. s 1 :3 w. N . m f u w my .. :2. m _ o .d D .. loo / m 1 OO 55 .numaoH EH Monmouw use as OmH unouu mom mHm>HouaH monopHmcoo NnO paw moumaHumo mono wcHanum usouu Hmuoy .w ouame oo. 00. ON. ON. :6... 3:3 .3... Sta _ _ _ H o 2283 .2283. 23:3...“ 22:82... s u i I 8 m m m m ... . . 9 u coco... ooooooou u a O 1 m ...: ‘3...- m .l CV .d . H .... ... m . n .. .. co m oo 56 12 and 13 in this study). He estimated the standing crop in stations 12 and 13 at 64 gm/m and the other stations at 52 gm/m in July 1976; and at 60 gm/m in stations 12 and 13 and 57 gm/m in the others in October 1976. These results indicate that stations 12 and 13 were used by trout before half-log cover addition. The standing crops of each species show a similar pattern as the combined standing crops (Figures 9-12). Both the brook trout and brown trout standing crop estimates were at higher values in the treatement section than those of the reference sections during the spring 1979 estimate and at lower levels than in the reference sections during the post-treatment period. No significant differences (p = .10) between the trout standing crops of the reference sections and the treatment section could be detected (Table 13). There is no evidence to show that the treatment section supported a lower standing crop, nor that there was any change in trout standing crop after the addition of the half-log devices. This apparent lack of response of trout to the installation of half—log cover in the present study is contrary to the positive results from half-log enhancement described by Hunt (1978). For comparison, differences in the two studies are discussed below. Hunt enhanced a much longer section of stream. He installed half-log devices in the headwaters of a stream 57 Figure 9. Brown trout standing crop estimates and 95% confidence intervals for all sizes. .O ousth 58 oo. 00. ON. ON. ..o... 3.30 :6... as... H H _ 0 23.85 02.23:. l 33.88 .5688... n... lo. W m N 9 ron 0 d Ion.” m . 13 Ion om 59 Figure 10. Brown trout standing crop estimates and 95% confidence intervals for trout 150 mm and greater in length. (_’ 4r — — H (Ti .OH mustm 60 0O. 00. ON. ON. :6“. 3:3 :0... 3:8 . _ _ _ H0 oEEouO 02.233... 23:00” “cos—«Dot. ...... .......... . Ha . 1o. u ooooo o W m ooooooooo . loco INI m .x:; m a as H m m 8 m u . H HI .. ...... w M lon/ m .Iva .IHyn va 61 .mmnam Ham mom mam>pmucw mocmvwmcou Nmm vcm mmumawumm mono wcfiwcmum usouu xooum .HH muawwm om. om. ms. as. :o... as... :0... 953m _ _ — — Av 22.8w 8:23.: ocozoom 22502... s 4 lo. w. . N m m m . N u m m 9 4 m ........ m 18 m u 0 fl . d J m II / on w Ow 62 .nuwcmH ca “mummuw vcw EB omH uaouu How mHm>HmuCa muamvfiwsoo Nmm van mmuwafiumm mono waavamum unouu xooum .NH ouamwm on. 8. 2.. 2. :0... 2:3 :0... 2......» P b _ _ 1 l CD CD I) an (w/ 5) doao smauus 23:93 8:23.: acozoom 2.2502... ............ CD ' 63 Table 13. Mann-Whitney U-test comparing trout standing crops of the reference and treatment sections. The null hypothesis is that the standing crop in the reference sections is the same as the standing crop in the treatment section. Test Statistic Critical Value (Us) (Us [‘11, “21) Result Spring 1979 8.5 U = 14 fail to reject lO(8,2) (P = .10) Fall 1979 18 U = fail to reject 10(12,2) 20 (P = .10) Spring 1980 15 U = 20 fail to reject 10(12,2) (p a .10) Fall 1980 21 U10(13,2) = 22 fail to reject (p = .10) 64 section that was shallow and flat as a result of channel straightening many years ago below a dam (R. L. Hunt, pers. comm. 1981). Cover abundance in my study area probably was not as low and so critical to larger trout survival as in Hunt's study area. Also, the lesser length of stream treated with the half—log structures in this study was possibly not enough to elicit a significant increase in trout abundance. This will be discussed further in the next section dealing with cover and trout correlation. The half-log structures used by Hunt were spaced 15 cm (6 inches) above the streambed as opposed to the lO-cm (A-inch) spacing used in this study. My smaller spacing may have restricted use of the half-log cover to smaller trout. The underwater observations revealed trout using the the structures, but the fish were of small size. Added smaller trout using the logs may not have appreciably in- creased the standing crop. Ecological adjustments, such as migratory movements or the displacement of subordinate fish that would encourage colonization of the structures, may take longer than the time covered by this study (White 1975b, Elser 1968). The section of stream Hunt studied experienced two-fold increases in October standing crops compared to April standing crops during the 3 pre-treatment years, but showed very little differences between April and October standing crops in the 3 years subsequent to treatment (R. L. Hunt, pers. com. 1981). 65 the fast response of trout to half-log enhancement in Hunt's study was probably a result of this large number of trout moving into this section each fall. Even with the initial increase, the trout population response to half-log enhancement reported by Hunt (1978) was still developing after 3 years. Cover-Trout Correlation The correlations of standing crop with the overhead cover ratings are shown in Tables 14 through 17. Corre- lation coefficients and coefficients of determination are given for each size group standing crop estimate during spring and fall of each year. The 1979 total cover ratings correlated negatively with the spring standing crop for all sizes and for trout 150 mm or greater in length. The spring standing crop for the 150 mm to 399 mm size class correlated positively with the total cover ratings, but at a low level to suggest little relationship. Instream overhead cover and the bank overhead cover ratings followed the same pattern, correlating to a lesser degree than the total cover rating. The negative correlation of bank cover to trout population is probably a result of electrofishing biases. During the high spring flows, the deep undercut banks are harder to wade and capture fish from. Fall standing crop estimates correlated positively with the 1979 cover ratings. Variation in 150-399 mm size 66 Table 14. Spring 1979 intercept and slope values, correlation coefficiegts (r) and coefficients of determi- nation (r ) for the trout standing crop-cover rating correlations on the Pigeon River. Cover Rating Intercept Slope r r‘ All Sizes Instream 50.05 -20.3 -.51* .26 15-cm bank 47.6 -22.2 -.44* .19 Total 50.9 -l3.0 -.54** .30 lO-cm bank 50.7 -28.8 -.51 .26 Total 53.5 -l6.5 -.59** .35 150 mm and Greater Instream 41.9 -15.9 -.42 .18 15-cm bank 40.1 -17.9 -.37 .14 Total 42.7 -10.9 -.45* .21 lO-cm bank 42.9 -24.1 -.45* .20 Total 44.8 -13.3 -.49* .24 150-399 mm Instream 29.3 3.5 .11 .01 lS-cm bank 31.6 -3.5 -.09 .01 Total 30.4 .5 .03 .00 lO-cm.bank 32.7 -6.9 -.15 .03 Total 30.7 0.0 .015 .00 * Significant at the 10% level ** Significant at the 5% level 67 Table 15. Fall 1979 interceptanuilepe values, correlation coefficients (r) and coefficients of determi— nation (rz) for the trout standing crop-cover rating correlations on the Pigeon River. Cover Rating Intercept Slope r r2 All Sizes Instream 24.5 39.2 .64** .41 lS-cm bank 35.5 15.8 .21 .04 Total 26.9 20.6 .52** .27 lO-cm bank 38.4 1.7 ’ .02 .00 Total 27.7 17.9 .44* .20 150 mm and Greater Instream 23.4 35.4 .65 .42 15-cm bank 32.8 16.9 .24 .06 Total 25.2 19.4 .55** .30 lO-cm.bank 35.2 4.4 .07 .00 Total 25.8 17.1 .47** .22 150-399 mm Instream 22.3 37.3 .66 .44 lS-cm.bank 32.5 16.5 .23 .05 Total 24.4 20.0 .55** .30 lO-cm bank 34.9 4.1 .06 .00 Total 24.9 17.8 .46**. .23 *Significant at the 10% level ** Significant at the 5% level 68 Table 16. Spring 1980 intercept and slope values, corre- lation coefficients (r) and coefficients of determination (r2) for the trout standing crop- cover rating correlations on the Pigeon River. Cover Rating Intercept Slope r r2 All Sizes Instream 29.7 75.1 .43* .18 lS-cm bank 57.6 -60.4 —.26 .07 Total 36.9 31.3 .20 .04 lO-cm bank 56.1 -43.5 -.21 .05 Total 37.1 29.4 .20 .04 150 mm and Greater Instream 26.1 64.2 .40* .16 lS-cm bank 49.8 -50.4 -.24 .06 Total 32.0 27.2 .19 .04 lO-cm.bank 48.5 -36.2 —.19 .04 Total 32.3 25.7 .19 .04 150-399 mm Instream 22.6 70.5 .46** .21 15-cm bank 49.7 -64.5 -.31 .10 Total 30.6 26.1 .20 .04 lO-cm bank 49.5 -54.9 -.31 .10 Total 32.6 20.1 .16 .02 * Significant at the 10% level ** Significant at the 5% level 69 Table 17. Fall 1980 intercept and slope values, correlation coefficients (r) and coefficients of determi- nation (r ) for the trout standing crop-cover rating correlations on the Pigeon River. Cover Rating Intercept Slope r r2 All Sizes Instream 32.8 45.8 _ .36 .13 lS-cm bank 44.0 3.5 .02 .00 Total 30.9 34.5 .32 .10 lO-cm bank 42.2 15.4 - .12 .01 Total 28.3 39.5 .38* .15 150 mm and Greater Instream 31.3 43.3 .34 .12 15-cm bank 41.5 5.9 .03 .00 Total 28.9 33.9 .32 .10 lO-cm bank 39.5 18.2 .14 .02 Total 26.1 39.6 .39* .15 150-399 mm Instream 24.6 55.5 .47 .22 lS-cm bank 36.7 15.3 .11 .01 Total 20.1 47.6 .48** .23 10-cm bank 35.7 19.8 .16 .03 Total 18.9 48.6 .51** .26 * Significant at the 10% level **Significant at the 5% level 70 class standing crop was accounted for by the cover ratings to the highest degree. The instream cover ratings are correlated to standing crop better than the total and bank cover ratings. This implies instream cover to be a greater factor to trout distribution than bank cover, contrary to other studies (Hunt 1971, Wesche 1976, Enk 1977). The 1980 total cover ratings were correlated positively with the spring standing crOp estimates. The standing crops showed similar correlations to total cover for the different size classes. Instream cover was positively cor- related to spring trout standing crOp while the bank cover was negatively correlated, perhaps as result again of electrofishing bias. Instream cover is implied to be im- portant in determining spring trout distributions, particularly for the 150-399mm size class. Fall standing crop estimates were positively correlated with the 1980 cover ratings. The 150-399mm size class standing crops correlated the best with the cover ratings. Variation in total cover accounted for standing crop variation to the greatest degree, with the 10 cm bank criterion demon- strating a slightly stronger relationship to trout bank use than did the lS-cm bank criterion. The low spring standing crop correlations with the cover ratings indicate that trout populations at that time may be using overhead cover differently than in the fall. A separate cover rating during spring may be more applicable than the annual cover rating at low-flow used in this study. 71 The instream overhead cover rating was a significant factor in explaining the distribution of trout in this study. Instream cover appears to have been more important to trout than undercut bank, but this is probably an erroneous conclusion for the reason that much of the instream cover was associated with undercut bank cover. Many trees, logs or alder roots that protruded into the stream and provided measurable instream cover also allowed the undercutting of bank and provided bank cover. The inter- dependence of instream and bank cover in.mmny cases suggests that the importance of instream cover is influenced by the bank cover. This further suggests that instream cover associated with bank cover may be more attractive to trout because both cover and feeding habitats are readily available. The greatest significance of the cover rating cor- relations to fall trout standing crops occurred with total cover, using the 10- or 15—cm.undercut bank criterion, to the 150-399 mm size class (Tables 15 and 17). This implies cover is more important to the standing crop of the 150-399 mm size class of trout. It should be noted that though the cover correlations with lO-cm undercut bank are greater than those with 15-cm.undercut bank, these criteria were significant to the same levels (Tables 15 and 17). The undercut bank cover criterion of 10 cm does not seem to define undercut bank better than 15 cm. 72 The revised method of rating cover in 1980 described the relation of fall trout standing crop to cover more as was expected. Variation in the total cover rating accounted for the distribution of trout more strongly than either the instream or bank did singly. Even with the interdependence of some bank and instream.cover, the total cover rating should be indicative of the amount of cover available for trout use. The revised rating also resulted in lower y-intercept values in the regression line equations, i.e. the predicted trout standing crop value at a zero cover rating (Figures 13 and 14). With all other factors not critical to trout, stream areas with little or no cover should have lower trout standing crops. The lower correlations of trout to cover in 1980 than in 1979 may be also a result of year-to-year differences. Other variables such as winter survival or fishing mor- tality that were not considered in this study could confound the trout-cover relationship. The correlation of cover rating to trout standing crop for the two different years are not directly comparable owing to the possible inter- actions of natural trout population fluctuations. Problems with the cover rating method are demonstrated in considering the upper stations of the study area. The log jams and bank-beaver undercut bank that were prevalent in this area met the physical criteria, but evidently were 73 were not properly suited for trout use. For example, the correlations of fall standing crop (150-399 mm size class) with the cover ratings sans stations 15 through 19 are much greater than for the total area (Figures 13 and 14). The coefficients of determination for the total cover ratings indicate a better relationship between cover and trout in the lower sections for both years. There- fore, purely physical space may have improperly defined the requirement for suitable cover for trout. Other attributes such as flow velocity, light intensity, and food availability need to be considered for defining suitability of trout overhead cover. In view of the lack of effect by the addition of half- log cover, correlations of total cover, using lO-cm.under- cut bank, without the cover afforded by the half-log devices, to fall standing crops of 150-399 mm trout re- sulted in greater correlation coefficients of .48 and .63 for fall 1979 and 1980, respectively. These higher correlations without the half-log devices further indicate that the half-log cover was not used to any great degree by trout during this study. Regression of standing crop against total cover can explain at most less than half of the variation of trout standing crop. This would suggest that other factors important in determining trout standing crop were not considered in this study. If this is the case, the addition of cover may not be an effective method to increase trout 74 standing crop because of its partial role in determining trout standing crop. Figure 13. Relationship between trout standing crop for the 150-399 mm size class and total cover ratings in 1979. The solid line describes the linear relationship for all stations in the study area. The dotted line describes the relationship, sans stations 15-19. 76 63.5 .o .322 com :22. 5-..... .260 .o «:3: 25.. .86: . .2 3am: nu; 0.. as. n. DN. 0.0 n~.. 0.. as. 0. mm. 0.0 H q . q A 4 q u q 4 O N .I b ...-on. .... .I III.- N 2. N. o I b I a... — 0| h ..IIIII mm a .... .. u u 4. :- w. ..ON I O Q’ I O 0 Oh “”6 (muses-om) 60:: buspums Figure 14. 77 Relationship between trout standing cr0p for the 150-399 mm size class and total cover ratings in 1980. The solid line describes the linear relationship for all stations in the study area. The dotted line describes the relationship, sans stations 15-19. ms. :55 599. 2:3. 3,8 .23 n. 3. m». .4H spawns 3.3.. 5-0: 2:3. 58 3:: .... 8. O .0 O. O Q ......J: (mm 662 '09!) due buwums CONCLUSIONS Half-log structures require periodic maintenance. Half-log cover devices provide current velocity shelter; trout using half-log devices for cover are found in the upstream areas of lesser flow. Fish-food macroinvertebrates colonize half-log structures. Cover rating during low flow relates to cover use by trout better in fall than in spring. A lO-cm deep bank criterion does not define overhead under-cut bank better than a lS-cm deep bank criterion. Overhead cover is important to trout in the lSO-to- 399-mm size class. The lack of trout population response to the lO-cm— high half-log cover installation does not support the hypotheses that the installation of half-log structures would increase trout standing crop. Overhead cover explained less than half the variation of trout standing crop in the study sections. This suggests that factors besides cover spacing were important in determining trout distribution. 79 LITERATURE C ITED 80 Allen, K. R. 1969. Limitations on production in salmonid populations in streams. pp. 3- 18 Ln T. G. . Northcote, ed. Salmon and trout in —streams. Univ. Brit. Col., Vancouver. Bassett, C. E. 1978. Effect of cover and social structure on position choice by brown trout (Salmo trutta) in a stream. M.S. Thesis, Michigan State University. Behr, E. A. 1977. How durable is Northern White Cedar? Extension Bull. E-929, Coop. Exten. Service, Michigan State University Bohlin, T. and B. Sundster. 1977. Influence of unequal catchability on population estimates using the Lincoln index and the removal method applied to electro-fishing. Oikos, 28: 123-129. Boussu, F. 1954. Relationship between trout populations and cover on a small stream. J. Wildl. Mgmt. 18 (2): 229-239. Butler, R. L. 1975. Some thoughts on the effects of stocking hatchery trout on wild trout populations. Pages 83- 89 Ln W. King, ed. Proc. Wild Trout Management Symposium. Trout Unlimited, Inc. Chapman, D. W. 1966. Food and space as regulators of sal- monid in streams. Amer. Nat. 100: 345-357. Compton, R. R. 1962. Manual of field geology. John Wiley and Sons, New York. 378 pp. Cooper, G. P. and K. F. Lagler. 1956. The measurement of fish population size. Trans. N. Am. Wildl. Conf. 21: 281-297. Coopes, G. F. 1974. Au Sable River Watershed Project bio- logical report (1971-1973). Fisheries M Report No. 7, Mich. Dept. Nat. Res., Lansing. 2 6t pp. Cited Ln A. Nuhfer. 1979. Use of artifical in- stream trout shelters by trout in the Au Sable River, Michigan. ‘M. S. Thesis, Michigan State University. DeVore, P. W; and R. J. White. 1978. Daytime responses of brown trout (Salmo trutta) to cover stimuli in stream channels. Trans. Am. Fish. Soc. 107: (6): 763-771. Elser, A. A. 1968. Fish populations of a trout stream in relation to major habitat zones and channel alterations. Trans. Am. Fish. Soc. 97: 389-397. 81- Enk, M. D. 1977. Instream overhead bank cover and trout abundance in two Michigan streams. M.S. Thesis, Michigan State University. Fausch, K. D. 1978. Competition between brook and brown trout for positions in a Michigan stream. M.S. Thesis, Michigan State University. Fausch, K. D. and R. J. White. 1980. Competition between brook and brown trout for positions in a Michigan stream. (Unpub. Manuscript). Hendrickson, G. E. and C. J. Doonan. 1970. Reconnaissance of the Pigeon River, a cold-water river in the northcentral part of Michigan's Southern Penninsula. U.S.G.S. Hydrologic Investigation, Atlas HA-333. Hilsenhoff, W. L. 1975. Aquatic insects of Wisconsin. Wisc. Dept. of Nat. Res Tech. Bull. 39. Hunt, R. L. 1971. Responses of a brook trout population to habitat development in Lawrence Creek. Wisc. Dept. of Nat. Res. Tech. Bull. No. 48. 35 pp. Hunt, R. L. 1978. Instream enhancement of trout habitat pp. 19-27 in A national symposium.on wild trout management. California Trout, Inc. Kalleberg, H. 1958. Observations in a stream tank of ter- ritoriality and competition in juvenile salmon and trout (Salmo salar L. and Salmo trutta L.). Rep. Inst. FreshwatergRes. Brottingholm. 39: 55-78. Jenkins, T. M. 1969. Social structure, position and micro- distribution of two trout species (Salmo trutta and Salmo gairdneri) resident in mountain streams. Anim. Behav. Monogr. 2: 57-123. Leopold, L. B. and W. B. Langbein. 1966. River meanders. Sci. Am. 214: 60-70. Lewis, S. L. 1969. Physical factors influencing fish popu- lations in pools of a trout stream. Trans. Am. Fish. Soc. 98: 14-19. Merritt, R. W. and K. W. Commers, eds. 1978. An intro- duction to the aquatic insects of North America. Kendall/Hunt, Dubuque, Iowa. McFadden, J. T. 1961. A population study of the brook trout (Salvelinus fontinalis), J. Wildl. Monogr. No. 7. 73 pp. 82 O'Conner, J. F. and G. Power. 1976. Production by brook trout in four streams in the Matamek Watershed, Quebec. J. Fish. Res. Bd. Canada. 33: 6-18. Pearson, E. S. and H. O. Hartley. 1966. Biometrika tables for statisticians. Volume I, Cambridge University Press, London. Pennak, R. W. 1978. Fresh-water invertebrates of the United States. John Wiley and Sons, New York. 803 pp. Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Bd. Canada. Saunders, J. W. and M. W. Smith. 1962. Physical alteration of stream habitat to improve brook trout production. Trans. Am. Fish. Soc. 91(2): 185-188. Seber, G. A. 1973. The estimation of animal abundance and related parameters. Hafner Press, New York. 506 pp. Shetter, D. S., O. H. Clark.and A. S. Hazzard. 1946. The effects of deflectors in a section of a Michigan trout stream. Trans. Am. Fish. Soc. 76: 248-278. Tarzwell, C. M. 1937. Experimental evidence on the value of trout stream improvement in MiChigan. Trans. Am. Fish. Soc. 66: 177-187. Vincent, R. 1971. River electrofishing and fish population estimates. Prog. Fish. Cult. 33(b): 163-169. Vincent, R. E. 1969. The tolerance of water velocity by trout as a basis for enhancement of the stream fishery. Proc. West. Assoc. Game and Fish. Comm. 1969: 189-190. Waters, T. F. 1972. The drift of stream insects. Am. Rev. Ent. 17: 253-272. Wesche, T. A. 1976. Development and application of a trout cover rating system for IFN determinations. pp. 224- 234 in J. F. Osborn and C. H. Allman, eds. Pro- ceedings of the symposium and specialty conference on instream flow needs. Volume II. Amer. Fish. Soc., Bethesda, MD. 83 White, R. J. 1975a. Trout population responses to stream flow and habitat management in Big Roche-a-Cri Creek, Wisconsin. Vern. Internat. Verein. Limnol. 19:2469-2477. White, R. J. 1975b. In-stream management for wild trout. Pages 48-58 £2_W. King, ed. Proc. Wild Trout Management Symposium. Trout Unlimited, Inc. White, R. J. 1979. Stream habitat management. Pages 241-250 in R. D. Teague and E. Decker, eds. Wildlife—Conservation. The Wildlife Society, Washington, D.C. APPENDIX 84 m« m« «« «m on m« N« N« m« NN NN «n ma «« ow mm m« m« m« N« N« «« mm mm Nm NH mm mm mm m« Nm No N« N« m« NN NN mm NH Nm Nm «m NN om «m N« N« «« an an N« 0N on «m NN «« «« on Nm Nm «m mN mN Hm n. am am o« no as «N «N «N oN om mm mm «a NN NN «N NN NN mm NN NN on N« om on m. NN Nm N« N« N« N« mN mN om NN on o« Na a. mm mm mm mm N« NN NN mN Nm Nm N« N. 0N «« «« Nm Nm mm mm mm «m ma om mm ON am am mm NN NN «m a «N «N NN AN AN Nm ea c. N. m N« N« am «N om an N .« N« N« «m o« om Nm Nm mm o m« m« a« NN NN moa am an mm m H _> W H _> W n _> W H _> W. :3 $3 0 ..L ..L 0 7L T. 0 .L I 0 ..L T. . C... _ C.— . C.- . C.- E O S CD 0 3 CD 0 S Ch 0 S 6 T.. 6 T. 6 ...: 6 T. ,6 m 2 Au m z ,6 m z ,6 m z m m m M” m m m a. owaa Name omma wcaumm mNaa Name ANN. maauam cowuowm upSum pom AE\wv moquHumm mmouu wcflpcmum udouh Hmuoe