:31“ fifihiw 'l! 1".“ 3% .y NW; 12*; § 1' fiz’é‘i .549"; ‘ ‘ 2 , 3 1 '7 ~ :23“; A . r Ti“? “ ' f .. i. i .. . .1 3*ng - . g? Egg: ‘ . .- : ~fiww. a} 2." I: Q ‘ ' 1%.“ 23;}! . .. I :z! ' 525:3? , 32* .‘ u .u. w'-' ,— “5‘91 . . a; . “3:14 ':_ $3333, ‘ {sighs-,2? “1;;th 215m filmy” 33" if g“ :‘ ‘22; 1' i’ ‘ 2 mi", 42 m 3. S4 -1 - ~ 333.713“ . — h '1 ‘. bfivxnzxs'ttgl "1 2‘2,- _' “Hi”; lupufighun ‘6'? u . fl-2 #3; - 23‘,“ 13“"""C“Y'fii 2. : 292m 2 Hi? 2”- 1%”: 'i "W“ 1‘3" ‘ 3'32: ..§'tlun;2-- :1 31.23"} 3i: n 38153113 . .; 23,2hp“;nT~2n’q... -.' m‘{.]-242§ >. x‘ 'ifigi'333‘332‘: Milaniinm] rm. 333333.331. 93d $3324 “2322:1121: u: up... 2*2 “$51111, 29 %: fikg2 is %§§ yz'gfi‘3f4wg “3332‘ 3‘35?!“ 1." “331333333” 3% ..‘, “.29 0.3516 “53‘3“3 gipiil 2:331; ' I . E’ 5i. :me ' gfigédr' ""7 fit 7::Em 3:". «ffifswfi ’ ,1 2.. £14”: m T"' .2.»- .. .. - . ....2 39 $33- ‘ " g. ‘Efii'-'2'3é.'3’fl‘?3 ‘32:; . “flag!“ 3.1.x; 5,. “:1- 1: ; ”73515;:ka 2234:“ ”2:35 '34: '2? 2. I ‘ .. ‘2 ’ 13.232! W I ' id». .. L3: ,. "— Mir A . "v.2“. " ‘ ' . "fir”: .-- , .2 .. 2 r - . .- v‘" . . . _ I! _ , A..~ 2 ‘ ,_,, a .1_ _ ‘ ”mac-4; . t. n- v.-_ _ 2 J" . .4 ‘ , “$22 3%}; u‘fuwél‘iélgfi 35 {43123: slu- . 3 2"."5‘ .. « v _. 27'; . 5:3; 4'41? ’. m x ;' .2 fl: 1.» u ‘ vh2.15 C increase) impact streams (ione standard error) 43 Figure 13. Influence of changes in mean depth on Sorenson’s similarity index values for fish above and below the dams (R-square=0.59) 44 Figure 14. Influence of mean summer temperature (C) on mean LnBrook density in the six streams containing brook front 49 Figure 15. Logistic curve Showing how the probability of brook trout presence declines with increasing mean summer temperature in all ten study streams ........................... 49 Figure 16. Logistic curve showing how the probability of brook trout presence declines with increasing mean summer temperature at three conductivity values .................... 51 Figure 17. Pattern of presence and absence of brook trout in study streams with increasing mean summer temperature and conductivity ‘1 Figure 18. Influence of mean summer temperature (C) on LnBrown density in all ten study ‘ 51 Figure 19. Logistic curve showing how the probability of brown trout presence declines with increasing mean summer temperatures in all study ‘ 5'4 Figure 20. Influence of mean summer temperature (C) on mean LnSlimy density in all ten study ‘ ss Figure 21. Logistic curve showing how the probability of slimy sculpin presence declines with increasing conductivity in all ten study ‘“ ss Figure 22. Influence of fish species richness on LnMottled density in above and below sections in all ten study ‘ <7 Figure 23. Logistic curve showing how the probability of mottled sculpin presence declines with increasing depth at three conductivity values 59 Figure 24. Pattern of presence and absence of mottled sculpin in study streams with increasing depth and conductivity 59 Figure 25. Regression of brown trout total length on scale radius for back-calculations of length at age 67 Figure 26. Growth of brown trout above and below the dams for the Boardman (Top) and White (Bottom) rivers Figure 27. Growth of brown trout over all study “ 66 Figure 28. Influence of conductivity on incremental growth of brown trout with previous length of 200 mm 62 Figure 29. Influence of temperature on age zero incremental growth of brown trout in study m 69 Figure 30. Relationship between conductivity and temperature for sites where mottled sculpin were present, slimy sculpin were present, neither sculpin species were present, or both sculpin species were present 87 Figure 31. The relationship between incremental growth of brown trout age>1 and conductivity without Prairie Creek’s data (Top) and with Prairie Creek included (Bottom) 88 xiii Introduction There are over 2,000 dams on Michigan’s rivers and streams (Michigan Department of Natural Resources, unpublished data). While these dams return many benefits to society (e.g., hydroelectric power, flood control, water level regulation), they frequently have negative impacts on populations of aquatic organisms, particularly fish. Reductions in fish populations due to impoundments are well documented, and occur through a variety of mechanisms. The best-known mechanism is the reduction in upstream migration that occurs where dams do not have adequate fish passage facilities (Holden 1979; Vogel et al 1988). Effects on fish migration are most obvious for anadromous fishes, and can severely reduce or even extirpate local populations (Brooker 1981, Ward and Stanford 1987). Another mechanism of impact is the change in water quality and habitat conditions that occur below a dam. Creation of a reservoir affects river habitat in many ways, potentially impacting stream-resident fishes as well as migratory species. It has been widely observed that physical/chemical factors such as stream substrate, dissolved oxygen and water temperature can be afi’ected by impoundment (Cummins 1979; Stanford and Ward 1979; Ward and Stanford 1979; Waters 1995; Hayes 1998). Dams afi’ect substrate by acting as a “sediment sink”, holding back finer sediments that normally would be transported downstream (Ward and Stanford 1983a; Ward and Stanford 1987; Waters 1995). This is of particular importance for the region of stream adjacent and immediately upstream of the impoundment, often referred to as the “impacted zone”(Klomp 1998; Mistak 1999). In this area, coarse sediments such as cobble can become covered with sand and silt potentially creating conditions unsuitable for the biota normally occurring in those stream sections (Waters 1995). Downstream reaches are often starved of these finer substrates and therefore become dominated by larger, more stable substrates. Dissolved oxygen decreases occur when deep release dams draw water from the anoxic hypolimnion (Ward and Stanford 1987). It has been assumed that dissolved oxygen decreases are not as much of a problem for smaller facilities, like those that are common in Michigan streams. This is primarily due to the fact that the size of impoundment often precludes anoxia, recovery of oxygen levels is fairly rapid downstream, and releases are often from the epilimnion (Brooker 1981; Ward and Stanford 1987). The characteristics common to dams that make dissolved oxygen changes unimportant, often cause an increase in downstream temperatures, making temperature the primary habitat concern. Reservoirs act as sinks for heat as well as chemicals and sediment (Brooker 1981; Ward and Stanford 1987). Surface waters spilling downstream from shallow or deep, stratified reservoirs, are ofien several degrees warmer than upstream reaches (F raley 1979; Ward and Stanford 1987). Temperature increases downstream of surface release dams are a major habitat concern in Michigan. Increases in stream temperature may shifi the temperature out of the range that a given species is genetically adapted to. Coldwater stenotherrns are genetically adapted to interact, feed, grow and survive better in colder temperatures (Carlander 1969; Allan 1995). They have also evolved to exploit the food sources provided by these colder habitats which are typically cold adapted macroinvertebrates and other fish (Allan 1981; Hubert et al. 1993; Rader 1997). Thus, increases in temperature can not only affect their internal physiology but also their food source. Many of the dams in Michigan are relatively small facilities that once generated hydro-electrical power for local needs. A majority of these facilities were deregulated in the 1950's and so remain as local landmarks, creating lentic fishing opportunities and stabilizing water levels for lake front property. Without regulation, these dams are primarily surface release facilities which have the greatest potential to increase downstream temperatures. Due to concerns over the effects of dams on stream habitat conditions, hydro- power facilities are governed to operate under water temperature and dissolved oxygen standards similar to those for point-sources of pollution. The policy of the Michigan Department of Environmental Quality (MDEQ) stipulates that, for coldwater streams, such facilities cannot raise water temperature by more than 2° C or cause dissolved oxygen to drop below 7 mg/l. The basis for these regulatory standards is not well established and the effects of violations or chronic changes within the standards on fish populations is unknown. This research investigated how changes in stream temperature, downstream of hydro-power facilities impacted coldwater fish and macroinvertebrate communities. My goal was to characterize how temperature increases due to impoundment affected the grth and population density of selected coldwater fish species including: rainbow trout (Oncorhyncus mykiss), brown trout (Salmo trutta), brook trout (Salvelinusfontinalis) slimy sculpin (Cottus cognatus) and mottled sculpin (Cottus bairdi). I also wanted to characterize how the downstream macroinvertebrate community changed with increasing temperatures. I initially hypothesized that population densities of coldwater fish and macroinvertebrates would be lower downstream of dams that increased downstream temperatures, and that this decrease would be magnified as temperatures rose above their physiological optimum. Additionally, coldwater trout species that remained in waters above their optimum would exhibit poorer growth than individuals of the same species that occurred in cooler, upstream waters. Changes in the macroinvertebrate community have the potential to interact with the direct effects of temperature by altering the primary food source for trout and sculpin. Besides the indirect impact that macroinvertebrate changes may have on the selected target fish species, they also ofi’er another way to examine how temperature increases impact coldwater stream communities. The specific objectives for this study were to: (1) (2) (3) (4) (5) Determine and compare population densities of brook, brown and rainbow trout and slimy and mottled sculpin above and below dams in coldwater streams in Michigan. Determine and compare growth rates of brook, brown and rainbow trout above below dams in coldwater streams in Michigan. Examine and compare water quality and habitat differences above and below impoundments. Examine the effects of water temperature differences above and below impoundments on fish growth and population densities of selected species. Characterize and compare macroinvertebrate community composition above and below impoundments. (6) Examine the effects of water temperature differences above and below impoundments on the macroinvertebrate community and its impact on selected fish species. Study Area Three streams were sampled during the summer of 1998 and seven streams were sampled during the summer of 1999 (Table l). The requirements for streams being considered in the study were: 1) they had a dam; 2) they had a resident trout population; 3) there was no operational fish ladder (to allow independent comparisons of upstream and downstream populations); and 4) were wadeable (due to equipment constraints). Streams were chosen with the advice of Michigan Department of Natural Resources personnel and were based on the above criteria and a broad geographic distribution in lower Michigan. Due to the natural distribution of coldwater streams, the arrangement of study streams tended to have a high representation in the northwest portion of the lower peninsula (Figure 1). The majority of streams were sampled at six sites, three upstream and three downstream of the dam (Figure 2). Two streams (Cedar Creek and White River) only had two downstream sites because water depths prevented an additional site. In three of the streams (Cedar Creek, Dowagiac Creek, and Manton Creek), there were two dams in relatively close proximity so an additional site was placed between the two dams. In these streams, upstream refers to sites above both dams and downstream refers to sites below both dams. Site selection was based on access (usually by a road crossing) and distance from the dam. When possible, sites were separated from each other by distances between 1 and 2 river miles. We were careful to select sites upstream that would represent the unregulated river, without going too far upstream to no longer be representative of the river along the normal continuum. We were also careful to not Table 1. Streams sampled in 1998 (designated by *) and 1999. Number of dams are noted in superscript (temperature, width and depth values are means from all six sites). Name County >‘< Temg. (C2 >‘< Width Sm; >‘< Degth (cm) Boardman River1 Grand 16.7 16.9 42.4 Traverse Cedar Creek2 Antrim 15.3 14.3 44.7 *Dowagiac Creek2 Cass 21.1 11.1 33.5 *Fish Creekl Ionia 23.9 18.0 41.3 *Manton Creek2 Wexford 15.8 6.9 17.3 Maple Riverl Emmet 16.3 11.1 38.2 Middle Branch Riverl Osceola 16.3 10.5 28.7 Prairie Creekl Ionia 18.9 12.2 28.7 Sugar Riverl Gladwin 19.4 6.8 23.6 White River1 Newaygo 17.0 10.2 32.0 Figure 1. Location of streams sampled in Michigan’s Lower Peninsula 62953 303 05 9.6% 835 33¢ a .«o Enema—o 3:268 380 .N ounwfi 58m roeoN nouoaqeiestomom" Alll Engagem— + :oaom E3533 38.82 Al :ocoN cocoamcaéoZ: I —m 4 an m an _ — . ‘ F gm m an n so sample too near the reservoir or in the “influence zone”(Figure 2). Below the dams, one site was generally placed approximately 100 m downstream of the plunge pool. The two additional downstream sites were separated by 1 to 2 river miles when possible, in order to determine longitudinal impacts. Methods Field Methods Study reaches (sites) were 60 to 80 m long, generally ending at the boundary between two mesoscale habitat units (e. g. riffle/run, pool/riffle). Each site contained an upstream, downstream and middle transect (Figure 3). The middle transect was located exactly halfway along the length of the site. At each transect, wetted width, depth and substrate were measured. Width was taken to the nearest 0.1 m by extending a tape measure across the stream channel. A depth profile was recorded for each transect by recording the depth in centimeters at one meter intervals across the entire channel. If the stream width was less than 10 m wide, then depth was recorded at shorter intervals so that at least ten depth readings could be recorded. Substrate size composition was determined using the pebble count method (Kondolf and Li 1992). This involved one observer slowly crossing the transect and determining the particle size of the substrate below his/her feet at each step. This process was repeated coming back across the transect. At the middle site above and below each dam (i.e. site 2 and 5) a stream flow profile was recorded at the middle transect using a flow meter at 1 m intervals. Along with the physical habitat measurements described above we measured several water quality parameters at each site. Stream temperature, dissolved oxygen and conductivity readings were recorded on the day of sampling using a Yellow Springs Instruments (YSI) meter. In 1999, Onset® temperature loggers were placed at the upstream and downstream sites proximal to the dam in each river. For those rivers with two impoundments, an additional logger was placed between the two impoundments, in 6882 08 $0855 085 2: 20:3 mezafiefi m 8% Lo 32> pumice ea 5; was 2958 we 5:804 .m 0.53..— m m .7 Eco N 25.". wEmobm >635 coy. :Qmon mEEEmm 12 order for each dam’s individual impact to be documented. Hourly temperature readings were recorded from May to October for all ten streams. Loggers were calibrated in the laboratory before being deployed in the field and were recalibrated in the laboratory after data retrieval. Water samples were taken at each site for laboratory analysis of total phosphorous concentrations. Samples were put on ice in the field, and were frozen until analysis. In 1998, we also collected 1 liter of water from the center of the stream at each site for analysis of seston levels. These samples were kept refrigerated until processing. In addition to habitat conditions, fish and macroinvertebrate populations were sampled at each site. Fish were sampled at sites where the water was deep enough using a barge electroshocking unit and a backpack unit was used at the shallower sites. Block nets were placed at the upstream and downstream transects to keep fish from moving in or out of our site while sampling. A multi-pass removal method was used to estimate the population size of our target species with a minimum of three passes (Van Deventer and Platts 1983). Each pass consisted of shocking the entire stream channel, concentrating on areas with more cover (e.g. weed beds, log jams, overhanging banks etc.), from the downstream transect moving upstream to the top of the site. When the backpack shocker was used it was necessary to shock in a zig-zag motion from one bank to the other along the site. The barge shocker had two anode poles allowing simultaneous shocking of each half of the stream channel. All fish caught in the first pass were measured for total length (up to 25 individuals for each non-target species) and identified to species. Only target species and species not previously caught were identified and measured in the following two passes, and all other species were counted. If there was not enough of a removal of trout in three passes then more passes were done until a good removal was attained. A 13 good removal was defined as a catch of 100 mm in length, above the lateral line between the posterior end of the dorsal fin and the anterior end of the adipose fin (Minard and Dye 1997). Fish <100 mm were recorded as age 0 or young of the year (YOY) trout (Hinz and Wiley, 1998). Three macroinvertebrate samples were collected at each site using a modified Hess sampler (Merritt et al. 1996). When possible, sites were selected that contained at least a portion of stream shallow enough so that the water did not flow over the top of the sampler. The sampler was placed over a cobble or gravel area so that the sampler laid firmly on the stream bottom. The bottom was disturbed for two minutes and surfaces of the rocks were scraped so that macroinvertebrates on and in the substrates would be swept into the 250nm mesh bag attached to the downstream side of the sampler. If no cobble or gravel was available, or if the cobble was too large for the sampler to fit over it, then a representative stable substrate (e.g. logs, roots, boulders) or sofi substrate was sampled. The debris and insects caught in the modified hess sampler were washed into a labeled jar and fixed with ethanol. Samples were transported to the laboratory for processing. Laboratory Memods Total, organic and inorganic seston concenu'ations were determined for the 1998 streams according to the method described by Wallace and Grubaugh (1996). Total phosphorous concentrations were analyzed using persulfate digestion (Valderrama 1981). Unknown fish specimens were keyed out to species (Eddy 1969; Smith 1989). Trout scales were mounted between two glass slides so that they could be read. Scales were aged and annuli were measured (m) using an Optimus imaging system. Macroinvertebrates were dyed with rose bengal and divided into two size categories (greater than 1mm, and less than 1 mm but greater than 45am) using 1mm and 45 um sieves. This helped to reduce the amount of fine substrates in the portion of the sample with the most insects. The greater than 1mm samples were picked, sorted and identified to the family level unless generic level identification was required for functional feeding group classification (Cummins and Wilzbach 1985; Cummins and Merritt 1996). The less than 1mm but greater than 45am samples were further divided into 0.5 or 0.25 subsamples depending on the amount of particulates that were in the sample. These samples were then picked, sorted and identified to the family level. Preliminary analysis indicated that a negligible amount and diversity of insects (e.g. minute midge larvae and zooplankton) were lost from the greater than 1mm samples. Because of this, and indications in the literature that prey <1 mm are not very important components of trout diet (Rader 1997), we restricted our analysis to the large category samples. Data Analysis The impact of each dam on mean summer temperature was determined by first computing the mean summer temperature from the sites with the data loggers (sites 3 and 4 in each stream)from June 1 to August 31. The logger data gave values for how temperature changed from directly above the dam to directly below the dam. Mean summer temperature at the sites without data loggers, was computed by calculating the deviation of each point measurement at that site from the logger site in that stream section. This was done by taking the difference between the temperature read by the logger in that section, on the day and time of sampling, and the temperature recorded by the YSI meter. For example, site 2 in the Boardman River was 16°C (i 01°C) at 12:00 pm on July 27, 1999 (the day we sampled) and the temperature read by the logger at site 3 at the same time and day was 15°C (2t 025°C), so the deviation of site 2 from site 3 was 1°C (:1: 0.35 °C). Assuming this deviation remains constant throughout the summer, the best available estimate for mean summer temperature at site 2 was 1°C (:1: 0.35 °C) warmer than site 3. The three additional sites placed between the two impoundments in Dowagiac, Cedar and Manton Creeks were not used to adjust temperatures for any other site, but were used to calculate the impacts of each individual dam. Mean summer temperature above each dam was calculated as the mean of all three upstream sites once they were adjusted for time and date of sampling. The same was done for the downstream sites. The overall impact of each dam was calculated as the difference in adjusted mean summer temperature above the dam minus the adjusted mean summer temperature below the dam. For the purposes of analyzing the streams based on the level of temperature impact, streams were divided into two categories “Low” and “High”. Streams where downstream summer temperatures increased by 2.15 °C or less were considered low impact, and high impact streams were streams where downstream summer temperatures increased by more than 2.15 °C. This value was selected as the cut off point because a division of stream impacts was evident, and because this approximates a change of 2°C which is the maximum temperature increase allowed for Michigan coldwater streams. Sites were categorized as above or below the dam so that analysis could be done on the differences that occurred between stream sections. Sites between dams were dropped from these analyses due to their sporadic occurrence in this study. So stream sections that were compared for those rivers with two dams were the sections above both dams and below both dams, the between section was ignored. An a value (Type I error) of 0.05 was used for all statistical tests. When possible, all stepwise regression analyses were done using the Maximum R2 improvement technique (MaxR) and bounds were put on the significance level required for a parameter to enter and remain in the model (oc=0.1 and 0.05 respectively). MaxR is considered superior to the normal stepwise technique because it does not settle on a single model but gives a series of models starting with the best one parameter model, offering the largest R2, and then the best two parameter model and so on. MaxR is also preferable to normal stepwise regression in that it evaluates all switches for their impact on R2 before any switches are made (Hocking 1976). Habitat (width, depth, substrate, flow, seston, dissolved oxygen, conductivity, phosphorous and temperature) differences above and below the dams for each stream were estimated using a general linear model analysis. Analysis of habitat differences over all study Streams as well as differences in macroinvertebrate family richness, Plecopteran family richness, Ephemeroptera, Plecoptera, Trichoptera (E.P.T.) richness, macroinvertebrate functional feeding group composition (Rosenberg and Resh, 1996) and fish Species richness, comparing above and below sections, was done using a mixed model analysis of variance (AN OVA), where stream was treated as a random effect and position was treated as a fixed effect. y=a+BX+b+e, _ [1] where y is the parameter of interest, or is the intercept and overall mean, [3 is the slope and effect of position, x is an indicator variable for position, b is the effect of stream, e is the error term, and b and e are independent random effects (Littel et a1. 1996). To analyze the relationships between the macroinvertebrate and fish communities above and below the dams and habitat changes, a correlation analysis was run to investigate how each individual habitat parameter was related to macroinvertebrate , Plecopteran, E.P.T. and fish species richness. Next, stepwise regression (MaxR) was used to determine which habitat parameter(s) was the best predictor of each macroinvertebrate and fish community parameter. If significant differences were found in the macroinvertebrate or fish community between upstream and downstream sections, then another stepwise regression (MaxR) was run on these community differences and changes in select habitat parameters (temperature, conductivity, dissolved oxygen, phosphorous, depth and substrate). This was done to see which habitat parameter(s) could be driving changes in community richness. In order to estimate the impact of the dams on fish community size (i.e. average total length of all individuals of all species combined), an analysis similar to that described for changes in richness was performed. Stepwise regression (MaxR) was used to see which habitat parameter changes were most related to changes in fish community size. To investigate the impact of the dams on fish and macroinvertebrate community composition, S¢rensen’s similarity index values (S¢rensen 1948) were computed comparing stream sections using the following equation: QS=2C/(A+B), [2] where QS is the index of community similarity, A is the number of taxa upstream, B is the number of taxa downstream, and C is the number of taxa common in both stream sections. Species level of taxonomic resolution was used for the fish community and family level was used for macroinvertebrates. Stepwise regression (MaxR) was used to see which habitat changes, if any, were related to fish and macroinvertebrate community similarity. Population abundances of the five target species (brown, brook and rainbow trout and slimy and mottled sculpin) were estimated using Microfish (Van Deventer and Platts 1985), a software program for removal methods. Microfish calculates maximum likelihood population estimates based on the pattern of removal for an individual fish species that was obtained during sampling. Population densities (fish/hectare) were calculated for each site by dividing the Microfish abundance estimates of each Species caught by the area sampled (m2), and then converting this estimate to number per hectare. The density data were averaged for each position, to get an idea of how the overall population abundance changed below the dams. Because preliminary analyses indicated that the mean density data were not linear and the variance was not homogeneous a loge transformation on the density estimates (Ln(density+1)) was used to correct these problems. Next, a correlation analysis was run and Pearson r values were calculated for the transformed data and each habitat parameter to see which correlations appeared strongest. Using this analysis as a guide, stepwise regression analysis (MaxR) was run to determine which habitat parameter(s) were most related to the mean Ln(density+l) of each species (referred to as Lnbrook, Lnbrown, Lnrainbow, Lnslimy, and anottled). A predictive model was developed for each species’ density and the most related habitat parameter(s) using a mixed modeling analysis of covariance (ANCOVA), where stream was treated as a random effect and position and habitat covariates were treated as fixed effects. This was done to see if the habitat parameter most correlated with density had an impact that differed according to position when variability across streams was considered. Target species were often caught only in one section of the stream (only above or only below). Therefore, logistic regression was used to model how habitat conditions were related to the presence or absence of our target species. Stepwise logistic regression analysis was used to determine which habitat parameter(s) was most related to the presence of our target species and to develop predictive models. Ageing data were examined to determine the age structure for each species of trout in each study stream. Growth analysis was done using the I-Iile method (modified from the Fraser-Lee method) of linear regression to calculate the length of the fish when scales were first formed (Rum) and backcalculations of the length at age using the following equation (Francis 1990). B.g.=((L-Rm)‘(R.gfiR))+Rm, [3] where Bage is the backcalculated length at age, L is the total length of the fish, Rm is the length of the fish at first scale formation, R“, is the radius (mm) of the scale at the annuli corresponding to the age of interest, and TR is the total scale radius (mm). These backcalculations were then used to calculate the length at age for the previous year and incremental growth (i.e. growth for each age step). An incremental growth model was selected to model the previous year’s growth so that growth rate would not be biased with month of capture (Weisberg 1993; Weisberg and Frie 1987). Since salmonids often show different growth patterns in their first year than in subsequent years due to variability in 20 habitat requirements (Deegan et al. 1999), the analysis was done separately for age 1 fish and all fish ages>1. Age 0 fish were excluded since they only show a partial years growth. Only brown trout were caught in sufficient numbers above and below the dams to characterize and analyze growth robustly. For brown trout ages>l, regression analysis of incremental growth on previous length at age and stream position was done for each stream sampled (Weisberg and Frie 1987). Mixed modeling AN COVA was used to determine differences in growth between dam impacts and stream positions, across all streams, by testing the slopes of the regression lines for homogeneity. Previous length at age, impact and position were treated as fixed effects and steam was again a random effect. Stepwise regression (MaxR) was used to see which factor(s) was most related to incremental growth of brown tout age>l. The factor(s) identified by stepwise regression as important was then put into a mixed model to determine if the factor(s) was still significant once stream variance was considered and also to test for differences between stream positions. For age 1 brown trout, a similar mixed modeling AN COVA was run on incremental growth, to identify what was most related with their growth in the previous year (i.e. age zero growth) across all streams. 21 Results Habitat Characteristics Physical habitat characteristics varied widely among study steams, but several longitudinal tends were apparent (Table 2). The study steams progressively became wider, deeper and had greater discharge below dams compared to sections above the dams (Figure 4). Stream width ranged between 4 and 22 m (>‘<=10.7 In) upstream of the dams and between 8 and 24 m (>‘<=13.2 m) downsteam. Depth ranged between 8 and 53 cm (>‘<=28.1 cm) upstream and between 16 and 69 cm 07:38.4 cm) downstream. Discharge ranged between 0.40 and 20.21 m3/s (i=5.6 m3/s) upstearn and between 1.5 and 21.8 m3/s (>‘<=7.7 m3/s) downstream. Steam substate on average became more coarse downstream of the dams (Figure 5). Most of the study steams were dominated by sand and gravel (3.53) upstream and sand, gravel, cobble (3.84) composites downsteam. Seston concentations for the 1998 steams were not significantly different between above and below sections and so seston was dropped as a parameter of interest in 1999 (Figure 5). l The chemical environment also varied substantially among study steams and among sites within a stream (Table 2). Phosphorous concentrations were highly variable and not significantly different between upstream and downstream sections on average (Figure 6). Analysis of variance (AN OVA) revealed that only the White River contained significantly different total phosphorous concentations between upstream and downstream sections (p=0.0081). Conductivity ranged between 259 to 540 pas/cm across all steams. Average conductivity was slightly higher below the dams (Figure 6) 22 2.... 2.... .2... .8... .3... 5w... .8... 5.2. .88... .583... 88.... 88... R88... 8... 4...... 8... mm... 5.2. 5.3m 2...... n8... 8.... Se 18.. 5. a... 5.2. :82 3.... a... 3.. ma... 8...... 83. 81mg 22.. 8.3.388; S8... 88... :8... a.” 3.... a... an... 22... 55m 2...... «E... @8 am.“ pm: Si. 8.. :2... so: . . . a... 8... as o... 323 2...? on. .8 5. an. 22... 2...? 3... an as 8.: 322. .88 2.. 8n 8. a: 92... .88 8.8 2... e... .8. 322. our... . . . as. E. e... 5.. 26.... are... 88... .8... 8...... . «mm a... 2.. 322. .82.. 0.8.2 .8... m8... .8... . 8m 2.. 3.. 2,2... 5...... 0.8.2 . . . .8 8A a... an. 322. 0...“: fl .8 8” a... we. 92... 2...: .8... «8... e8... 2. .8 n... 2.. 322. 8.52 .8... .8... .8... N... SN 4.: .2. 26.... :25: .8... m8... 8...... Se .8 cs 8.2 30.2. em .8... 8.... a8... 2.. an n... 8.8 26.... as .8... «8... m8... 2. as. .2 EN 322. 8.9%.. m8... m8... 2.... 3n 8.. 3 a...” 92... Basso . . . o... m... . 2 e: 322. .88 8.... 8a a... a... 26% aeoo . 323 :aEvhuOm o>ona 559—60” madami .aamwyi i I 1 l l I II gear. 583 % c.8681. 82:33 2.50on _o>a..wnv veamnm Emun .8..."— nmofim o_o_§m 88.359 .885 858 some E 8.828.. :82... .8 8.8.8.0820 .83.“: .N ofifl. J‘ 6.... .2... ...«.... ...«.... ...«8. ...«8 5.2. .82.... 2:25... 3.. 3... «3 3... ...... «.... 5.2. 5.2m 5.2. =82 viz—l ‘80:“ nag“ & 025a hue—8m 2m 3... ....« 5.. e m... 5.... :82 n. ««... «.m« ...: n 3 5.2. 2...? n... 9.... 28 ...a .. .... 2.2... 2...? .... .... «..« «... n .... 5.2. 2.8 v... «. ... ...... ...... v ...m 5.... 2.8 .... 2.... «.8 .... . .. .... 5.2. 2...... 2 ..«... w: E. .. ...... .5... 2...... ...... S... «....« .... .. ....” 5.2. .82.. 28.2 «.« ..«... .....« ...o m «a 0>...... .22... 22...). .... ...«... ...«w «... .. n... 5.2. 2...: a... ««... ...: ..a n v... 5.... 2...... mm ..«... ...«« m... m «.« 5.2. 8.52 ... ..«... ... . mm m .... 26.... 8.52 ....« ..«... ..S ...: n n... 5.2. .2... «.8 .m... ...2 ...: v m... 5.... .2». .... mm... ..2 ...... « ...... 5.2. 8.2.5.. a... ...... ...«« ...: .. «.« 5.... 8.2.5.. . . «.8 n...« q «... 5.2. .25 .....m «.... « ...« 5.... 22.0 ...? n... v ...n 5.2. 55.8.... ....n «.m. m «.m 5.... SEE... 2.3 .... 2.8... . , s 2...... \lflwa - 5.28.. 38% 22.2.5 €288 N 03...“. Stream Wldth (m) 2 a; a: 9 7 5 I I l I I I 1 2 Dam 5 6 Site 70 A J E 60 3 5 5°“ 5 a 40' H\ i g Soi- 2 !——i a 20- 10 l I I l I I Site 20 18 S, 16 is? 14 3% 13 D— 2% 8 §~ 6 (73 4 2 0| I I I I I l Site Figure 4. Average trends in width (top), depth (middle) and discharge (Bottom) (+- one standard error) above and below the dams. 25 clay=1 silt=2 sand=3 5 gravel=4 I Mean cobble—.5 M d - - I - - o e a 4.5 boulder=6 a bedrock=7 .2 .2 1: G a. o ‘5 ‘6 .n 3 a) 4 5 6 0.02 A 0.018 1 5. g 0.016 - g 0.0141 2 ‘5 0.012 1 8 I: 0.01 - 8 . c 0.008 - 5 o 0.006 " tn 3 0.004 « o " 0.002 - o I I I I A N 0: U m 3 .5 0: Site Figure 5. Average trends in substrate particle size (top) and total seston (bottom) (4— on standard error) above and below the dams. 26 Total Phosphorous (ug/l) 500 .5 8 400- 2, L Conductivity (us/cm) 300: 250 . ’. 12 Site 11.5 4 11 - 10.5 - 8.5 3 Dissolved Oxygen (mg/I) <0 - ‘0 0| 0 8 - 7.5 - 7 l Dam I I 1 2 3 4 5 Site Figure 6. Average trends in total phosphorous (top), conductivity (middle) and dissolved oxygen (bottom) (+- on standard error) above and below the dams. 27 increasing from 342 to 363 its/cm (Table 2). ANOVA on each individual stream showed that three streams (Cedar, Dowagiac, and Manton) had significant (cc=0.05) increases in conductivity downstream and one stream (Middle Branch) had a significant decrease. Dissolved oxygen (D.O.) concentrations were almost always >7.0 mg/l in our study streams (Dowagiac Creek excepted) and were not significantly different below the dams on average. Only Manton Creek had significantly (a=0.05) lower D.O. downstream (p=0.025). Stream summer temperatures ranged from 13.0°C to 23 °C in upstream sections and 15.8 °C to 25 °C in downstream sections (Table 2). Averaging across all streams, downstream temperatures increased significantly from 16.8°C to 195°C. On a site by site basis, mean summer temperature increased below the dams and did not show a return upstream temperatures, even at the furthest downstream site (Figure. 7). The estimated impact of each dam on mean summer temperature is shown in Table 3. Individual dam impacts ranged from a 0.98°C cooling downstream in the Maple River to a 5.46°C warming downstream in Manton Creek. Four streams (Dowagiac, Fish, Maple, and Prairie) fell into the low impact category (<2.15 °C warming) while six streams (Boardman, Cedar, Manton, Middle, Sugar and White) were considered high impact (>2.15 °C). Average temperature in the high impact streams increased from 15 °C to 19°C downstream, while low impact streams increased from 19.7°C to 205°C (Table 4). Macroinvertebrate Community Macroinvertebrate family richness was similar in above and below sections (Table 5). Richness ranged from 16 to 31 taxa (i=22.6 taxa) upstream and 17 to 26 taxa 28 251 23- 21- Mean Summer Temperature (C) «‘5 1 2 3 Dam 4 5 6 Site Figure 7. Trends in mean summer temperature (C) (+— one standard error) above and below the dams. Table 3. The estimated impact of each dam on mean summer temperature (above-below). Stream Temp. Change (C) Impact Manton -5.46 High Boardman -3.98 High Middle -3.90 High Cedar -3.84 ' High White -3.29 High Sugar -3.14 High Dowagiac -2.15 Low Fish -l.99 Low Prairie -0.36 Low Maple 0.98 Low Mean -2.71 29 Table 4. Selected community and habitat attributes ($standard error) averaged across strearns in each level of temperature impact. Level of Temperature Impact Community or Habitat Low High Arline Summer Temperature 19.7 °C ($0.81) 15 °C ($0.4) (Upstream) Summer Temperature 205°C ($1.2) 19°C ($0.3) (Downstream) Conductivity Change -205 ,us/cm ($16.7) -225 its/cm ($14.4) (Above-Below) Depth Change -11.4 cm ($4.3) -95 cm ($6.3) (Above-Below) Macroinvertebrate 0.76 ($0.06) 0.58 ($0.032) Similarity Fish Similarity 0.66 ($0.08) 0.47 ($0.11) Fish Species Richness 19 ($2) 9 ($2) (Upstream) Fish Species Richness 16 ($3) 16($2) (Downstream) 30 me... 2 3 w: 2 S ¢.N N.N EN 3N :82 m3 _- N N _- _ o _- t 2 223 m2 _ NV 2 _ _ N 4 ON 4N 395 23 o : : c _ _ N- 2N 2 035 3o 4 N N_ _ _ N 0 ON 0N can: 33 a : N_ _ N m n 8 a 2%: Bo N 2 N2 m o m _- NN 2N :25: 2.6 m 2 2 _ _ N N a 0N in mg a. N N_ _ o _ N- «N _N season Ed N N 2 m o m m 2 ON 58 :5 _ : N2 N o m N ON NN 553cm Eésa as 3%: 222 ES 3%: 222 £5 3%: 222 58% m.:em=o._Wm mac—Eon— »zfiam Hun—fl mus—Eu.“ man—ah 5239805, fiesta wanna—3323.532, .835 some 5 3350:358an e8 82? x02: bgzfia peomaoSm 93 25% 05 323 was 025“ memes—t DEE—Eco ofiEotoEan—Z .n 032. 31 (>‘<=20.5 taxa) downstream. Correlation analysis, on a site by site basis, indicated that none of the individual habitat parameters measured were significantly (a=0.05) related to macroinvertebrate richness (Table 6). Stepwise regression was unable to form a model with any significant relationships. Plecopteran family richness was significantly lower below the dams and decreased from 2 to 0.70 families on average (Table 5). Plecopteran family richness tended to be lower below dams in high impact streams and was less affected in low impact streams (Figure 8). Stepwise regression indicated that plecopteran family richness was most related to mean summer temperature (p=0.009, R2=0.16). While this relationship was significant it does not explain much of the variability in Plecopteran richness, and even the full 9-parameter model only improved the R2 to 0.32. E.P.T. richness showed a significant (p=0.034) decline downstream. E.P.T. richness upstream and downstream had similar ranges, but different means (7 to 13, >‘<=1 1.1 and 7 to ll,>‘<=9.3 respectively) (Table 5). Stepwise regression indicated that stream size (width and depth) was most related to E.P.T., and that wider, deeper, areas had fewer families in these orders. Again, variability in E.P.T. data reduced the predictive power of this relationship (R2=0.15). Macroinvertebrate family similarity, when comparing upstream and downstream sections, was 0.65 on average (Table 5). Similarity ranged from 0.54 (Cedar Creek) to 0.84 (Maple River). Macroinvertebrate similarity was significantly related to mean summer temperature change (R2 =0.64, p=0.0057) (Figure 9). High impact streams were less similar in terms of the macroinvertebrate community composition than low impact streams. Mean abundance and mean number of taxa calculated for each fimctional 32 m m good 3.0. R @336 mad 5905 33035 Om no _ md m H .o- Nm mmmvd 2 .o 3629/ cm NVde 86 mm ownad 2 .o- sun—0Q 33? cm 2.36 Rd mm 898 86. 52>? 885m cm .NNod 5.0 mm 336 mod- 53: cabana—m on «Sod 86 mm Rand E .o- 252 cabana cm 823 3.6 mm 33.0 m _ .o- omaaoma Emobm 3 n n 86 $3 fl m Km“ .o Had- maceonmmonm Eon. on good wmd wm emumd N— .o- 33:26:00 on 886 wed mm 836 n _ .o- 2382“th Basin :82 Evofim oEEam 2:9»..— .. 959.com Evofim oifiam 2:5..— ._ whey—sum 63:25» «83:: 9.2.55— Eeoam aur— mmoE—ua «can. 0358.853..an . $805.32” “852— :8 ES 3250: nmcbufinugfiobufi e8 mos—3-.“ mEveomwoboo 93 358508 532280 a Macadam .0 03am. 33 3.00 - IHigh GLOW 2.50 - 2.00 ~ 5 1.50- 1.00- 050 ~ /// 0.00 - I Families Above Families Below Figure 8. Average plecopteran family richness (+- one standard error) above and below the dams in high and low impact streams 34 I Fish El Macroinvertebrates l 1 .00 Linear (Fish) I . R2 = 0.64 0.90 " ' Linear (Macroinvertebrates) L'I ‘ ‘U x — ~ ‘ 0.80 o - U a ’ -e .5 D , a” = 0.21 0.70 .5" D l ‘ 7 I E , — 0.60 N p U E ‘ - 0.50 m U) -= — 0.40 i r 0.30 a O m « 0.20 ‘ 0.10 I I I I I I I I fl 0.00 -6.00 -5.00 -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00 ‘— Mean Summer Temperature Change (C) _’ (Warming) (Cooling) Figure 9. Influence of mean summer temperature change (i.e dam impact) on the similarity of macroinvertebrates and fish, above and below the dams. The data, linear functions and corresponding R-square values are shown. 35 feeding group in each stream section (above and below) are shown in tables 7 and 8 respectively. Analysis of changes in functional feeding groups below the dams showed no significant changes in predators, scrapers or shredders. Gathering collectors were similar at upstream sites and declined with increasing distance below the dams. Filtering collectors showed an increase just below the dams and then also decreased further downstream, but these data were highly variable among different streams and sites (Figure 10). Fish Communig Analysis Fish species richness across all streams was higher downstream of the dams than upstream (Table 9). Richness ranged between 4 and 22 species upstream (52:13.1 species) and between 10 and 25 species downstream (>‘<=15.9 species). Across all sites, fish species richness was most correlated with mean summer temperature, conductivity and total phosphorous (Pearson’s r=0.68, 0.58 and 0.44; p=0.0001, 0.0001, and 0.0015 respectively) (Table 6). Stepwise regression indicated that temperature was the best single predictor of fish species richness on a site by site basis (p=0.0001, R2=0.49). Further analysis on the relationship between changes in fish species richness and changes in selected habitat parameters (temperature, conductivity, dissolved oxygen, phosphorous, and depth) below the dams, using stepwise regression, showed that changes in fish richness were most related to changes in mean summer temperature (p=0.003, R2=0.79) (Figure 11). When the streams were separated based on level of impact, high impact streams (>2.15 °C increase downstream) had significantly more species downstream, and low impact streams showed no significant difference between sections (<2.15 °C increase 36 «.«.« «8% .188 «.«m« «.8. ...«« «...«« ..« .«m «8.. «.« .8 «HP...» «..«8. 50.2 «.8 ...«. «.8. «...«« «.«N. «..« «.«8. «..«8 «.88 «...«« «.88 «.8... 53. «.«N. ..«8 ..«.« «.8... «.«m. «..m «.«.«« «8.. «..«.N 8.... .. .8. «6.8 8...... .8. «..«. ...«« «.«.« «..« «.8. «...««. «.«.« «.««« «.««« «.88 «.«8. 5.0a .8. «.«8 .8. ..«8. .8. «.... «.88 .. .«8 ..««8 «.8. «.«.«« ...«8 .38 N..N «.5 m.«.« «.8. ...« «d. «.8. «.«8 «.8.« «.«o.« «.88 «.85 5% «.... «.8. «.u.« «.8. «.8. N.«.. «.88 «.8... «.88 «.88 «.88 «...«« «...-.... «... «.«8 «.N«« «.8 ...« «... «..m «8.. «...««. «...«« ..««.. «.«8. 5.... «.8 «.«8 «.«8 .8. ...... «.8 «..8. «.«8. ..8.. ...«.. «...«« «.88 «6...: «.«8 «..«. ..«8 «.«8 ..«8 .8 «.88 «.«.«N «...... ...««« «.8... «..«8 5.... «.8. . «.«.. ..«8 «.« .« ...«« .8. «.«««« «..«8 «..8. «..«8 «.«o .« ..«8«. 2...: «.8 «.«.N «.«8 88 ...« ..«8 «.88. «.««N. ..«8 «.«.«« «.88 «.88 8....w «.8 «.N«« «.««.. «.«8 «..« N. .8 «...««... «.38 «.... «...«« «.88 «.88. 5.5.2 «... «8 «..«N «.N«« ..««. «..«o ..8.. «.88 «.«8 «..««« «...«. «.«.«« 53... «.8 «.«.. ..«8 «..8. «.«.. ...... ..«««« «.88 «...«« «8.8 «... .8 «.88. ...... «d. «.8. ..«8 N.N.« ...«« «d. N...« «.8... «... ..«... «.8... «.88 E...» «... «.... «.«8 ...««. «... «..m «.«8 «...«« «.«m. «...8 «.8. «.88 8.8.3.. «.«8. «.8. «.«« N.N«« «...«« ...« «...««. «.«8 «.88 ..«.«. «.88 «.88 5.... N.«««. «.N«« «.8. «.8 «.8. «.8 «.88 «.88 «...8 «.98 «8.. «.88. 3x5 ...« N..N «.8« «.8 «.««. «...«« ...««. «...««N «.««.. «...«« «8.. «.««.« 5.... «.8 «.8 «.«««. «.«B «.«8 ...«« «M3. «.«mfi «.88 «.... «.«o.« «.8 .... 528m. 320.— o>ea< 329m «>23. 32.5 95.5. «3.0: 2.2.4 #20.— «>22 .31: 9.2.4 3.8.5 2.8.3.8 £2.28 £8.62. 5.8808885 58:685.... 3o...» . 8.8 05 322. «8 268 «0850:9528... .8 «use.» wince. 3.2.2.8 ..o «08.580 8:... .838... ...... «.80... 2.3825 386:9? 08.03. ... «Bah 37 ..o ..o «.o ..o ..o «.o «.o ..o ..o ..o 8.8 . . ... ... ... «.. ... ..« «.« «.. ... :82 8.. a: 3.9.... 8...... E. .58.... Gem... 6.8.. 3.9.... .88.. 2...... 8.08... ...e... 6.9.. ...««. ...3... 8.8... .«.e«.« 8.3... 8.9... .88 ...«« 6.9.... a: .«d... .«d... 8.8... 8.9... :8 .«.o.«.« 6.8. 6.8. 2...... 8.8.... .8. .....«.« ...««. 8.0.... 8.08.. ...««... .«.o.«.« ...«8. .088 2%... ...««. 8.9... 8.0.... 8.0.... 8.. ...:«.« .«.«.«.« .....«.« .«.o...« ...« 2%: .59.... .9. 8.9... 8.9.... ....v... .08.. .«.«.«.« .....m ....V... 6.9... :95... 6.0: 6.9. 8.08.. 8.8... 69. .«.o...« ...... .«.o.«.« .....«.. 6.08.. ...... 8.08.. .«d... .«.o.«.« ...«.«.« 6.9. a: .«.e«.. :8 .«.o.... 6.3. 3.35: 8.. E. a: .«.«.«.. 3...... 6.9... :8 8.9.... 3.08.. ...o.«.« 5.00 a: .88.. E. 8.9... .88.. .8 a... .«.o.«.« 6.08 .88.. 55.08.. 320.. 0.5.5. 320m 0.5.7.. 320.— 0.5.2 30.0mm 0.5.5. .520.— 0.5.2. E85m «.0:—.0.:m «SA—«Sm «.8360...— «..0.00=00\u£..05«w , «.0.00=oo\w5.0.=.m «5% 2.. 320.. 0:. 025. «08.98.086.00... .8 asew 9.600. 7828:... :80 5 echo 080:8... «om—...... ..o .095: 0m80>< .w 03.... 38 9000 - *FCOLL 8°°° ‘ +GCOLL Mean number/square meter .5 3000 - 2000 ~ ’ 1000 <3 0 n '1‘ 1 2 3 Da'“ Site figure 10. Average trends in functional feeding group abundances above and below the dams (FCOLL=Filtering/Collectors GCOLI:-Gathering/Collectors, PRED=Predators, SCR=Scrpaers SHR=Shreddexs). 39 mm... 3”. Qua flea QN- QmH “.2 :32 $6 Qmm mSw YmE m- 2 3 8E3 906 New- w; S 03 a- mu 2 Swsm cod 3» 06h mdw T m fl 3 0:85 the md Wax mam T 3 3 062—2 mod Qom v. 3 m; 2 S 2 5 2&2 $5 0.0 mSw 30m 2- M: b 5532 cod >4 Eon tab m- mm mm 5E mmd new- :02 3% w E on 023309 cod 02 ado N63 m- S n 330 two wdm- _.mm H 0mg m- 2 v caEanm bin—.55 E5 322— 95.2 ..EG B20“— 0>en< anbm mica—.25 AEEV saw—8‘— :mE :82 $2.53 86on 5E .885 some E can do.“ 32? 59: 35:83 Meowcouom can want 05 323 is 025m mowcano £5888 ii .a 03$. 40 Mean Fish Species Richness Change (Above-Below) 5'» . o d Mean SummerTemperature Change (Above-Below) 15 10 Figure 11. Influence of changes in mean summer temperature (dam impact) on changes in fish species richness (R-square=0.78). 41 downstream) (Figure 12). Changes in conductivity alone were not significantly related to changes in fish species richness below the dams (p=0.62, Adj.R2=-0.09), but when added to the model with temperature changes it was a significant parameter and improved the model fit substantially (p=0.0005, R2=0.98). Fish community similarity, comparing above and below sections, was 0.55 on average and ranged from 0 (Cedar Creek) to 0.90 (Prairie Creek) (Table 9). Fish similarity was not significantly related to changes in mean summer temperature (p=0.28, Adj.R2=0.089) (Figure 9) or conductivity (p=0.16, Adj.R2=O.134). Stepwise regression indicated that changes in mean depth were most related to fish community similarity (p=0.027, Adj.R2=O.59) (Figure 13). Streams that were significantly deeper downstream had lower similarity values. Mean length of the fish community increased from 99 to 103 mm on average, but this was not a significant increase (Table 9). The average length of the fish community ranged from 72.4 to 143.4 mm upstream and 70.7 to 174.7 mm downstream. Changes in mean community length were not related to any of the habitat characteristics used in the regression analysis. Densig of Target Fish Species Across all streams, total trout density averaged 921 fish/ha upstream and 280 fish/ha downstream, a 70% decrease (Table 10). Trout and slimy sculpin, when present, were always found upstream of the dams and were found at varying densities downstream. Brook trout were present in six of the ten study streams and were not found below the dams in three of those streams. Brook trout density decreased below the dams 42 I Above ' Below V. Mean Fish Species Richness -l —L N N 01 O 01 O 01 I l l I J Low Impact High Impact 0. Figure 12. Mean fish species richness above and below the dams for low (dJS C increase in temperature) and high (>2.15 C increase) impact streams (+/- one standard error). 43 ~ 1.00 I r 0.90 ~ 0.80 5 I E - 0.70 >. I :5 . ' - 0.60 5 I E _ $72 I 0.50 _m g - 0.40 2 8 ~ 0.30 e m - 0.20 +- 0.1 0 I I l I j 1' I 0.00 -40 -30 -20 -10 0 10 20 Mean Depth Change (mm) Figure 13. Influence of changes in mean depth on Sorenson's similarity index values for fish above and below the dams (R-square=0.59). 2 ~ #90 WV m2 8— me w v_ no— NhN w OMN toga owN _No 3. [QINM hNH pom. : 0_ ”mm fin N“ 03m :60: ND.” bN— _N of o o o o th an o we toga Nah CNN wnm a2: a c c a Nwlb chm c m— fl 35>) o com“ o VM 0 O o O o com o 0 ...—0va c mlmm c 8 e c c a a amm o e I 5:5 we n.0— wvm :N o c an :. mm mm o o hogan 2: SN 3 «he a e a. 3: im 2 a a II 03:.— NM we 2 v.0— o O o 0 NM NO 9 v flow..." a E a emm o c c c 5. an a v I use: 5% 55m v C oom 5mm aN MN m ~o h?— m we? hue—Bu an new v e :8 film me nu mg a: 3 5. 2%: wh— ~02 NN : q bOm O o No— 00v o— mm: toga Se 3% S = c mice a c g 3m 3 BE :85: o MN 0 Sc 0 0 o o 0 MN O 0 ...—02m : MN : Nm 6 c a e c MN 6 O :3“ 0 50 o ovm O o O o o be c o ...—0va a 2. c E: c c e e e as c c '85..ch 0 mm: o O 0 “mo o o o mm— o E. to?» c 35— c c c §N a c 9 Nb” 9 :3 .530 mm: ow o O v”: mm? o o :— nN~ m— 3 toga new a u w c c bbN VG" u o c Nvm mam an 3N EaE—ENQQ 320m 32: $20: 95.: j Nye—low... «>23. . jj Sim—Em Hafiz—h. Q£ohoa2 00 55.0 —50.0- 3.0- 00 050—..0 00 0000.0 have—o.» 00 00.0 v00.0- 3.0- 00 000F.0 00 0000.0 :32 8953.5 00 00.0 0 5.0- 00.0 00 v0 8.0 00 0900.0 5009 585$ 00 00.0 00 F.0- 00.0- 00 0050.0 00 003.0 53.5 Ego-=0 00 00.0 000.0- 00.0- 00 5500.0 00 002.0 332 8.2.350 00 0 F0 000.0- 00.0- 00 0 50.0 00 05? F0 owns—Ema 5.8.50 0v 00.0 52.0 0V0- 0v v000.0 0v v000.0 39.2—08.3 .590 00 00.0 500.0 00.0- 00 0000.0 00 500.0 365—92 8.800 .....E 00 00.0 000.0 00.0- 00 5000.0 00 500.0 30526.80 03.3.0 .. mien—ac.— 02:32.5 500.0 .. mien-:3.— Z 0295* .. H.523.— Z ...-.0: r 500.0 039»..— ouBanth. 3.55.0 =82 035.5.» .333: $0.50: 3:58:80 00 3:082:30... 93 0.... 8825.80 0.352 :2 .23 860% 50 “0008 280.52.. “we... 58 05 00 «8.3.08 A _+ $203.5 05 8.22 an... Ame—05$ 00 3:55 2 0.... mos—«>0 050503.50 0.... 8560008 cone—oboe 0.8030 .2 03¢. 47 (above or below the dams), mean summer temperature and an interaction term were run on the data from the six brook trout streams. The slopes were not equal to zero but they were equal to each other, meaning that the interaction term was not significant (i.e. above and below sections had equal slopes). Position was also found to be non-significant and so a reduced model was selected. The final mixed model relating Lnbrook to mean summer temperature is: Lnbrook=l7.9-O.915(T), [4] where Lnbrook is the logc transformed mean brook trout density and T is the mean summer temperature. No brook trout were found in sites with mean summer temperatures >l9°C (Figure 14). Due to the large number of sites where no brook trout were caught, we employed logistic regression to model the relationship between mean summer temperature and brook trout presence, using all sites. The logistic curve declines sharply around 17°C, indicating that brook trout are less likely to be found in streams with mean summer temperatures above 17°C (p=0.0001, odds ratio=0.4, 95% CI =O.25(L) and 0.64(U)) (Figure 15). The odds ratio can be interpreted as indicating how quickly the probability of brook trout presence in a stream changes with increasing temperature. An odds ratio of 0.4 indicates that for every increase in temperature of 1°C the odds of having brook trout in a stream are 0.4 of what they were at the cooler temperature. The fact that the 95% CI on the odds ratio does not include one, is further evidence of the significance of this relationship, because an odds ratio of one would indicate no change in the probability of brook trout presence with increasing temperature. The equation for the relationship 48 9 Above D Below 12 14 16 18 20 22 Mean Summer Temperature Figure 14. Influence of mean summer temperature (C) on mean LnBrook density in the six streams containing brook trout. Probability of Brook Trout Presence P .0 .° .0 .0 .0 .0 00 4‘. 01 O) \l 0 (D -l l I l J l l l _J On” .0 O 000 O a O Brook Presence (1 or O) " Predicted Probability '\ ' ' ' Upper CI - - - Lower CI \ \ 0.2 4 0.1 “ .~'-.. “nu..--- 0 $ - H 12 15 18 21 24 MeanSummerTemperature Figure 15. Logistic curve showing how the probability of brook trout presence declines with increasing mean summer temperatures in all ten study streams. 49 between brook trout presence and mean summer temperature is: P(Brook)= exp(15.08-0.91(T))/[1+exp(15.08-0.91(T))], [5] where P(Brook) is the probability of brook trout being present and T is the mean summer temperature. Stepwise logistic regression revealed that conductivity (p=0.0067, odds ratio=0.95, CI=O.92(L) and O.99(U)) and mean summer temperature (p=0.0066, odds ratio=0.41, CI=O.21(L) and 0.79(U)) together were better predictors of brook trout presence in a. stream. The probability of brook trout presence declines with increasing summer temperatures at a faster rate when conductivity values are higher (Figure 16). The equation for this relationship is: P(Brook)= exp(29.7-0.89(T)-0.049(C))/[l+exp(29.7-O.89(T)-0.049(C))], [6] where P(Brook) is the probability of brook trout being present, T is the mean summer temperature, and C is the mean conductivity. Small changes in temperature have more of an influence on brook trout presence than small changes in conductivity, as indicated by the odds ratio values for each habitat parameter. The interaction of the two parameters, however, gives a more specific picture of what type of habitats brook trout prefer. In general, brook trout prefer habitats with mean summer temperatures below 20 °C and conductivity values below 300us/cm (Figure 17). 50 ‘7‘ —I- cond=300 — 4— - cond=350 (mean) —A-— cond=400 12 14 1 6 1 8 20 22 24 Mean Summer Temperature (C) Figure 16. Logistic curve showing how the probability of brook trout presence declines with increasing mean summer temperature at three conductivity values. 600 7 El 550 - D D D €500 ‘ E] U B % @450 + a D n 3' €400 . E! U U U D 8 350 « u u :1 ”mm 5 0,85% Q] U 300 - g fill DNot Present 250 - \Q 0 200 1 r I r —r 1 l T 13 15 17 19 21 23 25 27 29 MeanSummerTempmture(C) Figure 17. Pattern of presence and absence of brook trout in study streams with increasing mean summer temperature and conductivity. 51 A similar approach was used to model Lnbrown trout density. As with brook trout, the slopes relating Lnbrown to temperature were not equal to zero, but were equal to each other. Position was also not a significant parameter for brown trout density and so was dropped from the model. The final mixed model which relates Lnbrown and mean summer temperature is: Lnbrown=14.73-O.602(T), [7] where Lnbrown is the loge transformed mean brown trout density and T is the mean summer temperature. Brown trout tended to persist below the dams more often than brook trout, but were present at very low densities when mean summer temperatures were >20°C (Figure 18). Stepwise logistic regression revealed that mean summer temperature was the most important parameter for predicting brown trout presence in a stream (p=0.0007, odds ratio=0.58, 95% CI=O.42(L) and O.8(U)). The equation relating brown trout presence and temperature is: P(Brown)= exp(l O.9-O.55(T))/[1+exp(lO.9-0.55(T))], [8] where P(Brown) is the probability of brown trout being present and T is the mean summer temperature. Brown trout are less likely to be present in a stream with mean summer temperatures >20°C (Figure 19). Increases in mean summer temperature affect brown trout and brook trout similarly, but brown trout do not decline as sharply and persist at higher temperatures. For one degree increase in summer temperature, the odds of brown trout presence will be 0.58 of what it was at the lower temperature. Conductivity was not an important factor for modeling brown trout presence. Using stepwise regression, the best parameter for predicting natural log of slimy 52 0 Above . n Below Lnbrown 0 - a - a a - ‘5 . 12 1 5 18 21 24 27 Mean Summer Temperature Figure 18. Influence of mean summer temperature (C) on LnBrown density in all ten study streams. 1 - 0000*“ O O 8 O 9 _ -"‘-. 0 Brown Presence (1 am) e ’ I. 3 "-.‘. Predicted Probability E 0.8 - .u‘~. - - . 5 0.7 _( . \ UpperCl ’ 'g 0.6 _ 7 “ha" LowerCl § 0.5 - ‘ P 2 0.4 - 0 g 0.3 1 ii; 0.2 - 3 n. 0.1 - . O - MW" "' 12 15 18 21 24 27 Mean Summer Temperature (C) Figure 19. Logistic curve showing how the probability of brown trout presence declines with increasing mean summer temperatures in all study streams. 53 sculpin abundance (Lnslimy) was also mean summer temperature (p=0.0001, R2=0.45). Including conductivity in the model with mean summer temperature only improved the model slightly (p=0.068 and 0.011 respectively, R2=0.48). The mixed model analysis on the log transformed density data indicated the slopes were not equal to zero and no interaction terms were significant. Position above or below the dams was also not significant and therefore was dropped from the model. The final reduced mixed model relating Lnslirny to mean summer temperature is: Lnslimy=13.89-O.655(T), [Si] where Lnslirny is the loge transformed mean slimy sculpin density and T is the mean summer temperature. Figure 20, shows how this line fits the slimy sculpin density data in the study streams. It appears that other factors are likely to also play a role in determining slimy sculpin abundance. These data appear particularly suited to logistic regression analysis, since slimy sculpin often showed very spotty distributions, but were abundant wherever they occurred. Stepwise logistic regression showed that conductivity was the single parameter most related to slimy sculpin presence (p=0.0048, odds ratio=0.86, 95% CI=0.77(L) and 0.95(U)). Similar to the brook trout, the probability of slimy sculpin being present in a stream declines quickly above conductivity values of 300/.cs/cm (Figure 21). The equation for this relationship is: P(Slimy)= exp(44.6-0.15(C))/[1+exp(44.6-O.15(C))], [10] where P(Slimy) is the probability of slimy sculpin being present and C is the mean conductivity. The natural log of mottled sculpin abundance was most related to fish species 54 0 Above a Below Lnslimy 1 2 14 16 1 8 20 22 24 Mean Summer Temperature (C) Figure 20. Influence of mean summer temperature (C) on mean LnSlimy density in all ten study streams. 3 1 — - ee 5 0.9 _ '. O Slimy Sculpin Presence (1 or O) 0 8 _ \ " Predicted Probability r: ° \.‘ '. - - - Upper Cl “a. 0.7 ‘ I '5 t. - - - Lower Cl a 0.6 ' " E 0.5 .. -_ a 0.4 ~ “‘5 o 3 - g . .5 0.2 - .3 0.1 - é: o 200 300 400 500 600 Conductivity (us/cm) Figure 21. Logistic curve showing how the probability of slimy sculpin presence declines with increasing conductivity in all ten study streams. 55 richness (p=0.015, R2=0. 12), however this relationship was quite weak. Depth and Fish species richness together made the best two parameter model (p=0.035 and 0.013, R2=0.21 respectively). Using the mixed modeling approach to develop a predictive model for mottled sculpin was a bit more difficult. Since mottled sculpin were not strongly related to any of the parameters we measured, as indicated by the correlation analysis (Table 11), I attempted to find the best model using the data available. Based on the stepwise regression results the best two parameter model should include fish species richness and depth. Using a similar approach as in the previous analyses, I first looked to see if the slopes were equal to zero and equal to each other. Depth was not a significant parameter in this analysis and so was dropped from the model. The slopes of the model with only the fish species richness data were not equal to zero but were equal to each other. Position in this case was significant and so there are two models relating mottled sculpin abundance and fish species richness. anottledmvc=1.18+0.245(F), [1 l] anottledbdow=-l .03 +0.245(F), [l 2] where anottledabove and anottledbelow are the loge transformed mean mottled sculpin densities for each position and F is fish species richness. Although these relationships were relatively weak, mottled sculpin densities tended to be higher above the dams and, in both positions, were higher in streams with higher overall fish diversities (Figure 22). 56 0 Above Cl Below Predicted Above "' ' EmfiaedBdow 101 9- 8- 7.. “85* 9 is“ .44- 3- 2- . D . 1‘ x' E] o 4? M .3 o 5 1o 15 20 I r I I 25 30 35 40 Fish Species Richness Figure 22. Influence of fish species richness on LnMottled density in above and below sections in all ten study streams. 57 Stepwise logistic regression showed that conductivity and depth were the two parameters most related to mottled sculpin presence (p=0.02, odds ratio=l .008, 95% CI=1.002 (L) and 1.014(U); p=0.004, odds ratio=0.92, CI=O.87 (L) and 0.98 (U) respectively). The equation for this relationship is: P(Mottled)=exp(-O.44-0.083(D)+0.008(C))/[1+exp(-O.44-0.083(D)+0.008(C))], [1 3] where P(Mottled) is the probability of mottled sculpin being present, D is the depth and C is the conductivity. Conductivity, while significant, did not have a strong effect on mottled sculpin presence. In general, the probability of mottled sculpin presence decreased with increasing depth more slowly at higher conductivities (Figure 23). Mottled sculpin were never present in waters >50 cm deep and tended to be found at conductivity levels >250 us/cm (Figure 24). Trout Age and Growth Analysis Brook trout ages ranged from zero to three over all six streams and age zero (young of the year) was the most abundant age class captured (Table 12). Brown trout ages ranged from zero to seven over all ten streams and age one was the most frequently caught age class (Table 13). Mean age was similar between positions for both species of trout. A regression analysis was used to determine the length of brown trout at first scale formation (Rmo ) (Figure 25). Due to the curvature in the relationship between length and scale radius exhibited by brown trout in the study streams, the Rzero value calculated from the regression line was negative, and so a value of zero was substituted. Regression analysis of incremental growth (fish age>1) on previous length at age and stream position, revealed that only two streams (Boardman and White) had significant differences 58 Probability of Mottled Sculpin Presence —o— cond=250 - - I - - cond=350 (mean) +cond=450 20 40 60 80 100 Depth (cm) Figure 23. Logistic curve showing how the probability of mottled sculpin presence declines with increasing depth at three conductivity values. Conductivity (us/cm) 600 l E] El 550 " . El OPresent s O 500 . ’ Cl El DNot Present 400 - O . Q 350 - éfl 300 - ‘01:: 0 claim El :80; El 250 - n '3. u 200 I I . . O 20 4O 60 80 Depth (cm) Figure 24. Pattern of presence and absence of mottled sculpin in study streams with increasing depth and conductivity. 59 c n o o .38. on. a... c a... . o :32 . o e o o c 2...? o c o o o .25.. 0......E. 3.. b o v m o 2%: 8.. . c o . o 8.5.). . o c o o c 3.30 mm. m o ..-- . N . o 5533 n a . .. 88% om< Ase—ca m a on 8 .58. 2 SN 3 .... 3 we. :82 a... a . v a o 3...? 8.. . o o . o 55... 2.6.2 on. S . 2 m. . 0ass. .2. R o o e. s 8.52 a... o. o m m. o .38 . m . . 553.... ... a . a 58% «mm. .222. .835 some ... .958 o... 9.888 322. ...... act 26.... ....o... x83 .8 own :88 28 owe ... .3552 .2 use... o . o N NN 5 8 ..N .35 N.N N. .N o ..o o N... N.N N... N ...N :52 EN NN o o o o N ..N . o 2...? . o o o o o o o o o 5.5 NN.N . . o o o o N N o o 0.5... 8... . o o o . o o o o 55... 2.5.2 NN.. NN o o o N v N. o. N 0.5.). 8... .... o o o o . N NN N. 8.5.2 . o o o o o o o o o 5.". o o o o o o o o c 5.535 . o o o o o o o o o .58 NN.. .o. o . o N ..N LNuN N. N. 55.5.... 5.. :52 N .afi N N N v N N . .. 52% ou< Ate—om. . o N o. NN NN. .N. ..N .589 o... N.NN ..o o N... . N.N N.N. ..N. N 5...). NN.. NN o o N N N N N o 2...? SN ... o o e o N N. N. o .55 NN.. .N o o o o N N o o 0.5... 5N ... o o . . N N. o o 55... 2.5.2 NN... N. . o o o . N N N o 25:. 8.5 N o o o o N NN NN .... 5.5.2 NN... N o o o o o o N o 5.". N... No o o o o o o N . 5.538 NN.. NN . o o N ... N. NN o 55o 56 NuN o c o .. NN NN .N . 55.5.... em... :52 N .589 N N N a N N . e .585 03. @255 .88.... some ... .55... o... Eczema 322. ...... Rob 98.. .32. 5.65 No own .52.. ...... om“ ... 55:52 .2 03.... 61 owe ... Emce— No 22.23.5333 .8 mac... 258 .8 Swan. .58 Sch 5505 ..o commmemom .mm 2%.... :55 595.. .59.. 8.. 8m 8.. 8... SN 8. o . _ _ _ _ 0 p o - N... 00 S NNouNm o o - . m NNNo... + 58.... u N m m. m - N. m 0 O - N O 62 (p=0.004, and 0.003 respectively) in growth between above and below sections (Table 14). The Boardman River had greater brown trout growth below the dam, while the White River showed poorer growth downstream (Figure 26 ). Mixed modeling ANCOVA of incremental growth on previous length at age and position, where stream was treated as random and previous length and position were treated as fixed, showed no significant difference in brown trout growth between stream sections (p=0. 107). Once position was dropped from the model, the final mixed model relating incremental growth and previous length at age (p<0.001) for brown trout across all streams was: 1:87.8-0. 146(P_len,8, ), [14] where I is the incremental growth for the age step of interest and P_lenage is the length of the fish at the previous age. The data showed a large number of observations below 200 mm, due to the lack of older fish caught in the study streams (Figure 27). Also, variability among study streams, in terms of habitat, food availability etc., creates substantial variability in incremental growth at the same previous length. Since incremental growth decreases as fish become older and larger (Figure 27), I ran stepwise regression analysis on three fish lengths representing the range of lengths caught in the study streams. These length intervals and their corresponding incremental growth labels are 100 mm(incre100), 200 mm (incre200) and 300 mm (incre300). Stepwise regression (MaxR) revealed that incremental growth of brown trout (age>l) was most related to stream conductivity for all three length bins (p=0.017, 0.005, and 0.007; R2=0.53, 0.64 and 0.67 respectively). Mixed modeling ANCOVA, where stream was treated as random and position was treated as a fixed effect, revealed that incremental growth for all three length intervals was significantly related to stream conductivity 63 3.x v93 $.55 oom-mm. mu mood- code 32.5 ov moo vow ooN-Nm. o. mono- coo. 26.2 «0.23 a). a... «E NE NE NN.. 5:. 320m emsm 3.6m v93 mom-wt S End- vmdo. 263.. 59% em. :1 o: 3.3 ow . on. . . . Smo- g...» 325 N. .3. boo- no.3 Nam-Nu. v. .93- flow 33.... 038.. NN..... 93% no.9 mom . wood- NO? 32.5 .m.mm 2.3 SE 373.. .N wood- awe o>3< 06...: 36m 3...:- vnfio owe-mo. mm ammo. Ed: 325 .N.NN . ..NN ...... NNNeN. . . a. . ...- 3.8. 26.... n.52 Sen NN.NN Row arm-om. . o. mto- goo. Nae—om mow-.9 mwodn moo...- oNN-om. mm 2...? $63 26...... 8252 a). 8.. NE NE E: NN.. .8. 323 5.. NE S: 8: S: 5). NN.. 33¢. 8...... NE NN.. ...: 8: NE a). 8.. 32.5 «E NN.. a). «E 8.. 8.. 8.. 26...... 88538 NN.. a). E: a). a). a). ax: 32cm Nov mom in flan-hm. mm on..o- n...» . 265. ...—.00 when 336 Shaw .wv-ow. ac to- www.mo. Nae—om boom Son 5.3. «mm-om. 3 to- Nooo 963.. *Sthom snob:— oouaocu 8.22.— 385 owes— fiuaoa 2 2.2m 33.8:— =o£8m 88.5w .... .8 5.3 .38: 8a 528.. 83.23 80:20.“... 885.8% 5.3 NENobm 8.6... 8... 0... A88 oom ...... oom do: 83 Smac— 025 8 an... .5. 5398 38080.08 “.3030... .83.... :80 ... ..an o... 323 ...... 26.... .8... :30... .8 09.... 59.0. ...... AZV an... ..o .38.... 80.088... .088 5380 .3 use... I an 100 ‘.‘~ 63 ‘ Q ‘~ D - 90 4 7h “‘3 D D 2 ~ g Linear (Above) I 80 ‘ o n‘ ... a . - - 'Linear (Below) E ~‘th CD D 0 Above e 70 - . ~31 :1 Below u ,0 D 3 6° ‘ . ° ° :1 g 50 ‘ . U o g 40 - ° ‘ "‘ o 30 w e 20 I f I I 0 100 200 300 400 Previous Length at Age (mm) o 100 J E 90 fl . —Linear (Above) v 80 ~ ' " 'Linear (Below) g 0 0 Above 2 70 "h. all: u Below 0 60 _ ~§~ U3 g “a. .- 5 50 ' . E 0“ ~~ . § 5 4° “-9 ° "N. . 30 " D D D ~~ ~ 0 20 D ~‘=. O 100 200 300 400 Previous Length at Age (mm) Figure 26. Growth of brown trout above and below the dams for the Boardman (Top) and White (Bottom) rivers. 65 .83... ......» .... .96 80.. :30... ..e 5380 ...N 0...»... A88. 03. ... 59.04 0.8.5.... com com oov com com oo F o (mm) was minimum] =0.055, 0.045, and 0.048 respectively), although marginally for increl, and that there was no significant difference between stream sections. For example, the model relating incre200 and conductivity is: Incre200=130.47-0.25(C), [15] where Incre200 is the incremental growth for a 200 mm brown trout and C is the stream conductivity. Again, the low numbers of large fish makes this relationship more difficult to describe with confidence, but in general incremental growth of brown trout (age>1) decreases with increasing conductivity (Figure 28). Mean summer temperature was not a significant covariate for any of the length bins (p=0.19, 0.26, and 0.46 respectively). Stepwise regression of age zero growth, revealed that mean summer temperature was the most related single parameter (p=0.0001, R2=0.26). Mixed modeling ANCOVA of age zero growth on temperature and position, where stream was treated as random and position was treated as a fixed effect, indicated that there was a significant interaction between temperature and position. The interaction term was ignored and dropped from the analysis because with very few below sites having age one fish, one anomalous site was driving an interaction that should not be interpreted with this data. Once the interaction term was dropped, position was also found to be non-significant and therefore was dropped as well. The final mixed model relating age zero growth and mean summer temperature was: I=32.6+3.18(T), [16] where I is age zero brown trout growth and T is the mean summer temperature (°C). Age zero brown trout growth increased with increasing temperature (Figure 29). 67 .88 oou ..o 580. 0.530... 5.3 80.. :30... .o 539w 88080.0... .8 5.3.8288 ..o 00:03.... .mm 08%... .5003 38.03.50 coo om... com emu 00¢ omm com 0mm com r _ _ _ _ _ _ F o .2. -ON -om -ov -0m -00 -05 1em -om (mm) lime-lg) muamoraul 68 0800.00 .380 8 .00.. 830.: ..0 £30.» 38080.08 0.00 0mm :0 0.80.0880. ..0 00:03.5 .3 0.53m ADV 0.50.3805 008.8% 8002 «0 mm 00 0. 0. v. m. L h b p F . b 0‘ O n ,8 w . m 18 m . m m . I 02 m I‘ m. m I 00. ) . m - 03 320mm. 9 922 o . 8. 69 Discussion General Habitat Stream habitat is expected to change predictably in a downstream direction (V annote et al. 1980). Temperate headwater streams are generally shallower, narrower, and are dominated by coarser substrate than higher order reaches downstream (Cummins 1977; Vannote et a1. 1980). With the widening and deepening of the channel, stream discharge increases as well as temperature fluctuation and daily maximum temperature (Cummins 1977; Vannote et al. 1980). Stream regulation (i.e. damming) alters the natural longitudinal flow of energy and materials, thereby altering the physical and chemical characteristics of downstream reaches and the biological community (V annote et a1. 1980; Ward and Stanford l983a; Ward and Stanford 1983b). The serial discontinuity theory (Ward and Stanford 1983a) offers a hypothesis for the impact that a dam will have on downstream habitat and biota. This theory, however, was developed with the assumption that dams operate via hypolimnetic draw, and so many of the conclusions, especially those involving predictions about water quality and biotic changes, run contrary to what we expected to find in our study streams (Table 15). The study streams showed the expected longitudinal increase in width, depth and discharge as predicted by the River Continuum Concept, and did not appear to be significantly altered by the dams (V annote et al. l980)(Figure 4). There was an increase in substrate particle size downstream, which is consistent with what is expected for dammed middle order streams (Ward and Stanford l983a)(Figure 5). Seston quality (i.e. organic content of seston) is expected to peak below surface release reservoirs in the same way as has been documented for lake outlets, while total seston values often remain 70 09.080 0 Z 0000.00Q 0000.009 0.0.6.8 88.82 0000.08 0000.08 0000.00Q 80.2.5 80....— 0000.08 802.88% 0000.08 0000.02. 0.80.0080... .0885 0000.08 885 0000.08 0000.08 00% 0.08085 t W s . . . .... A t . - h. . , : n -, ,. N. ......ll .. .H ... .. .. , t . 8 0809 .8 80050832— 8 080: ..0 800.0083?— 080: ..0 088.03. 00880 00.0.008O 080:0 080:5...— 800500309 00880 63080.5 8.80: .0 3.8.8800 009.080 0030080 .80 .8080 08. 8 080080 0... .0. 008.080... .8me 0.080% .80 0.026 080080 .080 0:88 0. 32 .0. .9005 38880008 8.00 05 .3 00.080... 00 080.. 30.0.. 009.080 .0880: 8.0 88.8800 0.0080 00.02% .m. 030... 71 unchanged or increase downstream (Herlong and Mallin 1985; Richardson and Mackay 1991). The streams sampled in 1998 showed no significant difference between upstream and downstream sections in both seston quality (measured by total organic seston) and total seston concentration (Table 2). Total phosphorous concentration, is ofien used as an indication of nutrient limitation and overall stream productivity (Wetzel 1975; Allan 1995). Total phosphorous values were highly variable between streams, and generally were at or below that of normal rainwater (Wetzel 1975; Allan 1995). Phosphorous levels in individual streams were not significantly different between upstream and downstream sections, with the exception of the White River. Averaging across all streams, phosphorous showed no longitudinal trends. Conductivity (i.e. total ionic content), which is related to total dissolved solids (TDS) and also is used as a surrogate for productivity (Ryder et al. 1974; Ryder 1982; Allan 1995), was slightly higher in downstream sections when compared to upstream sections. Conductivity increased significantly downstream in three streams (Cedar, Dowagiac and Manton). It was difficult to determine what caused this increase in conductivity, but it could have been due to increased urban and/or agricultural runoff carrying excess salts and other nutrients into downstream reaches. In lakes, total dissolved solid (TDS) concentrations often vary differently from total phosphorous concentrations, especially if phosphorous is limiting (Ryder et al. 1974). Cycling of limiting nutrients does not consume the ion, so fluctuations in phosphorous would not necessarily affect TDS or conductivity (Ryder et al. 1974). Because of this, conductivity (as a correlate to TDS) is sometimes a better indicator of productivity (Ryder et al. 1974). The serial discontinuity theory predicts a decrease in nutrients below dams in middle 72 reaches, because the reservoir often acts as a nutrient sink (Ward and Stanford l983a). My data did not show any difference between sites 3 (last upstream site) and 4 (first downstream site) in terms of total phosphorous concentrations or stream conductivity, but instead a downstream trend of increasing conductivity and variable phosphorous levels. Because of the relatively small changes in average conductivity and lack of significant differences in total phosphorous concentrations, it appeared that upstream and downstream reaches were similar in terms of productivity. Temperature Ward and Stanford (1979) summarized the potential temperature alterations caused by dams into six categories: 1) increased diurnal constancy, 2) increased seasonal constancy, 3) summer depression, 4) summer elevation, 5) winter elevation, and 6) thermal pattern changes. Historically, large hydroelectric operations were of greatest concern, because their mode of operation oflen caused catastrophic changes downstream (Brooker 1981). Hydroelectric facilities often draw water from the hypolimnion, and therefore summer decreases and winter increases have been the primary focus for research on dam induced temperature changes (Brooker 1981; Crisp 1987; Jensen 1987; Ward and Stanford 1987). Temperature increases have been given little attention, and primarily have focused on impacts on the macroinvertebrate community (Ward and Stanford 1979; Fraley 1979). It has been suggested that impacts on temperature are restricted to the area of stream directly below the impoundment and that temperatures quickly equilibrate with the air (Brooker 1981). Downstream reaches with increased summer temperatures, however, will not shed heat during hot summer days, but will continue to warm according to 73 normal River Continuum theory (Boon and Shiers 1976; Vannote et a1 1980). This heat loading has the potential to increase downstream temperatures so that they remain elevated, above the range that cold-stenotherrns can survive in. Webb and Walling (1993) were able to detect 0.5 °C mean summer increase 5 km below an impoundment and F raley (1979) found significant summer increases that never returned to upstream temperatures 56 km downstream of a dam. In this study, I focused on increases in mean summer temperature below dams, and how it related to alterations in downstream communities. While dams can alter temperature in other ways (e. g. reduced diel fluctuation) and in other seasons, it is during the summer that there is maximum potential for physiological heat stress to occur in adult cold stenotherrns. Additionally, bioenergetic models indicate that mean summer temperature is especially important for growth because of the relationship between metabolism and temperature as well as feeding and temperature (Gibbons 1976). Another reason for the focus on mean summer temperature changes versus daily flux was that without continuous loggers at all sites it would be impossible to determine how daily flux changed with distance from the dam. F raley (1979) found that while mean summer temperature changes downstream of a surface release dam were consistently higher than upstream reaches, average diurnal fluctuation was lower directly below the dam and then returned to upstream levels with increasing distance downstream. Impact of the dams on mean summer temperatures was variable, but on average streams were 2.7°C warmer downstream than upstream. Individual impacts among the study streams ranged from a slight cooling to more than 5 °C increase downstream, and were able to be separated into two general categories (high and low impact). Factors 74 influencing temperatures below dams include the size of impoundment (specifically depth and surface area), residence time, whether or not the impoundment stratifies, and the release depth (Ward and Standford 1979; Brooker 1981). All dams in this study were surface release facilities with the greatest potential to increase downstream temperatures. Variability in impact, therefore, most likely was due to differences in surface area, depth and residence time. Another important factor that can contribute to variability is the amount of groundwater coming in downstream of the dam. Groundwater increases the stability of flow and also cools stream summer temperatures (Allan 1995; Giller and Malmqvist 1998). In the Maple River, for example, it appeared that there was sufficient groundwater recharging the system downstream to more than compensate for the impact of the reservoir, resulting in slightly colder downstream temperatures than upstream. The dam on Prairie Creek created a very small impoundment with almost no residence time, therefore there was virtually no difference in mean summer temperature between stream sections. Dowagiac and Fish Creeks both had relatively high temperatures upstream and lower overall temperature impacts, but even a small increase in temperature for a marginal coldwater stream could have dire consequences for cold-stenotherrns downstream. This is especially true if temperatures are raised above 20°C (Allan 1995; Giller and Malmqvist 1998; Taniguchi et a1. 1998). For those streams that did have increased summer temperatures downstream, this impact remained with increasing distance from the dam and in most cases increased further. Macroinvertebrate Commqu Studies comparing rivers across the globe have shown macroinvertebrate family richness to be linearly related with stream temperature, with diversity increasing with 75 increasing temperature (J acobsen et al. 1997). In this study, macroinvertebrate family richness was not significantly different between stream sections and was not related to any of the habitat parameters we measured. Macroinvertebrates often show shifis in habitat preferences along environmental gradients within a family, and so identification to the species level may have shown some different relationships (Fraley 1979; Hauer et al. 2000). Fraley (1979) found a decrease in macroinvertebrate diversity directly below a surface release dam, but then similar levels of diversity between upstream sites and sites further downstream. Ward and Stanford (1979) summarized the results of several studies on dam impacts and showed that regardless of how temperature was altered (increased, decreased, daily flux etc. ) plecopterans and ephemeropterans tended to decrease below dams and trichopterans showed variable results. Plecopterans, on the whole, are considered cold- stenotherms, although certain families (e.g. Perlidae) are found in warmer habitats as well (Ward 1992). In my study streams, plecopteran richness showed a significant, negative correlation with mean summer temperature. When streams were separated according to level of impact, only high impact streams showed significant decreases in plecopteran richness downstream (Figure 8). In high impact streams, plecopterans were often eliminated below the dams and when plecopterans were found, they were primarily Perlidae. From these data, it appears that the type of temperature alteration has an effect on how plecopteran richness changes below dams. E.P.T. richness values also were lower downstream, but had higher variability and were not strongly related to any habitat parameters, although they tended to be higher in shallower, narrower reaches. Functional feeding group abundances tended to be similar between upstream and 76 downstream sections with the exception of an increase in filtering collectors directly below the dam. This response by filtering collectors below dams has been noted elsewhere and is most likely due to an increase in suspended food particles coming out of the reservoir (Fraley 1979; Herlong and Mallin 1985; Richardson and Mackay 1991). Although our data did not detect a difference in seston in the four study streams where seston was examined, it’s possible that water samples for seston analysis were taken too far downstream to detect a difference. Studies have shown that below reservoirs and lake outlets there is often rapid use of high quality seston by high densities of filtering/collector macroinvertebrates, reducing the concentration of these suspended particulates longitudinally downstream (Wallace and Merritt 1980; Herlong and Mallin 1985; Richardson and Mackay 1991). This may have been a factor in this study, because of the observed peak in filtering collectors at the first site downstream when compared to the last site upstream, across all streams. Following this peak there was a general decline in filterers moving downstream, indicating some orientation with respect to the dam (Figure 10). Macroinvertebrate community similarity was significantly related to differences in mean summer temperature, and level of impact was a good gauge of how similar stream sections above and below dams are likely to be. In general, when a dam increases downstream temperatures 3-5 °C, then similarity would be expected to be around 0.56 (3F 0.032), and if a dam increases downstream temperatures 2°C or less, similarity would be higher, around 0.76 (x 0.06) (Table 4). The fact that macroinvertebrate richness was not significantly different between stream sections indicates that it is not a change in diversity downstream causing this relationship, but rather a replacement of taxa. The 77 relationship between similarity and level of impact points to temperature as the reason for this replacement. Fish Communig Species diversity is expected to increase as you move from headwater areas to middle reaches and then decrease again toward the mouth (V annote et al. 1980). Increased environmental heterogeneity through normal stream processes or moderate levels of human disturbance is predicted to increase habitat availability and allow for more species to co-exist (Gorman and Karr 1978; Vannote et al. 1980; Ward and Stanford 1983b). Ward and Stanford (l983a) postulated in the serial discontinuity theory that biotic diversity below a dam would decrease either from reduced detrital transport in headwater streams or from reduced temperature fluctuation in middle reaches. In my study streams, fish species richness was almost always highest directly below the dam, particularly in high impact streams. High impact streams had significantly lower fish species richness upstream when compared to downstream sections of high impact streams and upstream sections of low impact streams (Table 4). High and low impact streams had similar levels of fish species richness downstream, most likely due to their similar average temperatures downstream (Table 4). Coldwater habitats tend to have fewer species due to their harsh nature as well as more constant temperatures (Allan 1995). High impact dams increased summer temperatures from a mean of 15 °C to a more moderate mean of 19°C, which was accompanied by a significant increase in fish species richness from 9 to 16 (Table 4). If we think of increased temperature as a disturbance, as it is defined in the intermediate disturbance hypothesis (Ward and Stanford 1983b), then it should be considered a “moderate 78 disturbance”, since temperatures did not rise into the lethal range for most cool- and warm-water fish species. As such, there would be an expected increase in diversity downstream, even if diel fluctuation was reduced directly below the dams. In the low impact streams, where temperature on average only increased from 19.7°C to 20.5 °C, fish species richness was not significantly different (Table 4). Lyons (1996) found cold, headwater streams to be dominated by a few species of salmonids and sculpins, and noted an increase in species diversity as well as a shift in dominance to more eurythermal, warm-water species in streams with higher summer temperatures. Wehrly et al. (1998) showed that warmer streams consistently contained more species than cold streams across and within three different levels of July temperature fluctuation. Other studies of differences in fish species richness below dams also indicate that temperature change is important, and show decreasing richness with decreasing temperatures (Brooker 1981). A study of low-head barrier effects on fish species richness noted a peak in richness just below dams and then a decline in richness with increasing distance downstream, toward what was observed for uninterrupted streams (Dodd 1999). These barriers did not alter temperature and so the peak below dams was attributed to a pooling of fish species at the barrier (Dodd 1999). Aggregation of fish species attempting to move upstream may have played a role in my study streams as well, but the fact that fish species richness remained elevated with increasing distance downstream in both high and low impact streams indicates that temperature was probably more important. Sheldon (1968) found stream depth to be the most important factor in determining diversity within a stream. In this study, both high and low impact streams showed similar longitudinal increases in depth moving downstream. When all the habitat 79 parameters were considered, the best predictor of fish species richness, as well as changes in richness with respect to the dams, was temperature. It seems reasonable, therefore, to conclude that changes in mean summer temperature below these types of dams are primary in determining how diverse the fish community downstream will be. It is also likely that the reduction in fish species richness seen below deep-release dams, is probably more related to the reduction in temperature below the physiological optimal for most warm—water fish than previous studies have indicated. Fish community similarity was not related to differences in mean summer temperature, but instead was negatively related with stream depth changes. This result was surprising since low and high impact streams had similar mean differences in depth between stream sections, but different mean similarities. Studies on the relationship between depth and fish community richness showed that increased habitat complexity from expanded vertical habitat allowed for more species to co-exist in deeper waters (Sheldon 1968; Gonnan and Karr 1978). In these streams, depth was not related to fish species richness, and therefore it seems unlikely that depth changes would cause a replacement of species downstream since fish are not usually excluded from habitat based on depth alone. High variability in stream fish similarities made it difficult to attribute any strong relationships between habitat and similarity in these streams. Low impact streams, however, were more similar on average than high impact streams (Table 4). The average similarity of the low impact streams (0.66) was almost equal to what was calculated by Dodd (1999), comparing upper and lower reaches of unregulated streams in Michigan (0.68). This indicates that the presence of a dam alone does not necessarily impact the fish community in terms of similarity, but requires some habitat alteration as 80 well. At this time it is difficult to point to one habitat parameter as the reason for the level of fish community similarity that was demonstrated in these streams. Densig of Target Smcies Brook and brown trout are coldwater stenotherrns that have positive growth between 4 and 20°C, with upper lethal limits around 25 °C (Elliott 1994; Allan 1995; Marod 1995). Brook trout catch and movement patterns have been related to the number of days temperature exceeded 16°C (Marod 1995), and in my study streams brook trout presence declined sharply above 17 °C, Brook trout were more often eliminated below the dams than brown trout. The former showed a 96% downstream decrease across all streams, while the latter were 54% lower downstream. For both of these species, mean summer temperature was the most important predictor of presence and density. Both slimy and mottled sculpin are typically found in cold, shallow stream habitats and are usually associated with brook trout, although slimy sculpin are more consistently considered a headwater species (Scott and Crossman 1973). Stream substrate and temperature together are considered the primary determinants for sculpin habitat suitability (Scott and Crossman 1973). Both sculpin species were 78% lower below the dams on average, but slimy sculpin were in fewer streams. There were only a few instances where both species were found together and in general slimy sculpin were in the coldest streams with lower conductivities and mottled sculpin were in the warmer streams with higher conductivities (Figure 30). This may have been a coincidence since the natural distribution of slimy sculpin only dips into the northern portion of Michigan’s lower peninsula and mottled sculpin distribution issaid to be more spotty in the Great Lakes basin (Scott and Crossman 1973). The coldest streams in this study tended to be 81 8000.0 0.03 00.0000 800.00 .000 .0 8000.0 0.03 00.0000 800.00 .280: 8000.0 0.03 80800 .880 8000.0 0.03 80800 000.08 0.083 0080 .00 0.30.0080. 000 88000000 0003.00 080000800 .8 0.080 ADV 0.8800809 om mN ON mr OP f . . . GDP .0000 .08on . com Swag.” 0 0 0 .0 .w MR0 .... 8=m .... 0 9 0 OLWOO C - com m o 4 0 4 4 m 4 4 4 1 D. 4 - 8.. m. o o co . m 4 « 4 4 r 000 m. 0 4 o o - coo t con 82 more northern, due to the distribution of trout streams in this state. However, other research of sculpin distribution in northern Michigan streams found both species present, but slimy sculpin showed a more narrow distribution with respect to temperature and mottled sculpin were found in both cold and warmer habitats (Wehrly et al.1998). In my streams, slimy sculpin density was most related to mean summer temperature and mottled sculpin density was not strongly related to any habitat parameters. Other research relating mottled sculpin abundance and habitat in Michigan was also unable to detect any significant relationships (Newman 1999). Position with respect to the dam was not an important factor in determining trout or slimy sculpin abundance, indicating that the presence of a dam alone does not negatively impact these species. Instead, it is the alteration of downstream habitat away from what is preferred, that is detrimental. Hayes et al. (1998) have indicated the value of cold, headwater streams for providing thermal refugia for brook trout during the summer. Brook trout under suboptimal thermal conditions will move into colder, headwaters or tributaries and then return to middle reaches in the fall and winter (Scott and Crossman 1973; Marod 1995). Dams that prevent this migration and increase downstream summer temperatures above 20°C limit the ability of brook trout to escape or survive until fall. Brown trout also show behavioral movements in response to temperature (Elliott 1994), and are probably impacted by increased temperatures below dams in a similar way as brook trout. Brown trout, however, are able to remain competitive at warmer temperatures than brook trout (Taniguchi et al. 1998), and therefore persist in downstream reaches more often. Considering my results and those from other studies, it appears that a dam which increases summer temperatures above 17 °C can be expected to 83 hinder brook trout survival, and if temperatures are increased above 20°C all three species (brook trout, brown trout and slimy sculpin) are likely to be reduced. Conductivity was related to the probability that brook trout, slimy sculpin and mottled sculpin would be present in the study streams. In general, brook trout and slimy sculpin were more likely to be present in stream sections with conductivity values below 300us/cm and mottled sculpin were more likely to be present at higher conductivities. At present it is difficult to determine what could have caused these relationships. The conductivity values documented for these streams were well within the normal ranges for streams in Michigan (Michigan Department of Environmental Quality, unpublished data). There appears to be a geographical shift in conductivities in Michigan due to differences in soils and geology, with northern streams usually having lower conductivities than southern streams (Michigan Department of Environmental Quality, unpublished data). This may have been a factor for the two sculpin species which seem to show a geographic separation. Modde et al. (1986) found brook trout to be tolerant of moderate watershed alterations including changes in pH, turbidity and nutrient loading at temperatures between 9°C and 12°C. In other studies, brook trout densities and growth have been shown to be positively correlated with stream conductivities ranging between 27us/cm and 869us/cm (Cooper and Sherer 1967; Scameccia and Bergersen 1987;Kwak and Waters 1997). I can only speculate on why brook trout presence showed a negative relationship with conductivity in these streams. This could be a spurious correlation, or possibly something that we did not measure is also correlated with conductivity and is impacting the brook trout, but in any case it should receive further research. 84 Depth was related to mottled sculpin presence and density, although during mixed modeling analysis it was dropped from the density model. Mottled sculpin tended to be found in depths below 50cm. This type of specialization is surprising since mottled sculpin occupy lake habitat as well as streams (Scott and Crossman 1973). Van Snik and Stauffer (1999) found both slimy and mottled sculpin to be generalists when it comes to depth selection. Experience in these rivers has shown catchability of sculpin to vary with flow velocity, substrate and depth. This could have been a factor in the deeper, downstream areas of the rivers where mottled sculpin were more difficult to catch. Trout Growth Growth of fishes is influenced by many environmental factors, but among the most important are food quantity and quality, and temperature (Elliott 1994). How these relate to growth of individual species depends on the endogenous physiology of each species and body size (Elliott 1994; Diana 1995). Metabolic maintenance increases with increasing temperature, and so if food is limited, fish may seek colder temperatures to maximize growth between food availability and metabolic costs (Diana 1995). Fish size also is of importance, because larger trout are less susceptible to changes in water temperature than small trout (Elliott 1994). Brook and brown trout growth is generally positive between 4°C and 20°C (Baldwin 1956; Elliott 1994; Marod 1995). Thermal stress in most trout is expected to begin at temperatures above 20°C, which means growth would stop and body size could decrease if metabolic costs exceed what the fish can consume (Elliott 1994). Jensen (1987), predicted growth of brown trout below a dam that decreased summer temperature ranges from 12—15°C down to 9.5-12.5°C, to be 19.7 mm shorter than trout in 85 unregulated sections. In a study conducted on the relationship between brook trout growth and feeding at three temperatures, the highest rate of feeding and weight gain was at 13 °C while lower rates were found at both 9°C and 17°C (Baldwin 1956). The age distribution of brook trout and brown trout in these streams was similar to what other investigators have found in Michigan trout streams (Gowing and Alexander 1980; Mistak 1999). Brook trout densities below the dams were too small to do a meaningful comparison of growth rates between stream sections. In general, brown trout (age>l) growth was not significantly different between stream sections. The Boardman and White Rivers were exceptions, however, and showed opposite trends, with higher growth downstream in the Boardman and lower growth downstream in the White. Given the differences in temperature between streams and sections within a stream and the strong relationship between brown trout density and temperature, it was surprising that conductivity was most related to incremental growth of brown trout (age>1) and that temperature was not a significant factor. Another surprising result was that the relationship between conductivity and growth of brown trout was negative. Significant variation among the streams in terms of productivity (using conductivity as a surrogate) would be expected to lead to variation in growth, with more productive streams having higher growth as long as the temperature was within the limits for trout. But in these streams, conductivity was not substantially different and there was a negative trend between productivity and growth. I suspect that the uneven distribution of length and age data for brown trout in these streams could be indicating a false correlation, and that if we had caught more fish of larger sizes, the relationships shown here would change. Further indication of this is that Prairie Creek, which had relatively high conductivity and 86 appeared to provide good trout habitat, showed poor brown trout growth. Prairie Creek is one of the few rivers in the southern portion of Michigan’s lower peninsula that maintains cool temperatures and is known for relatively good trout fishing. Because of this, and its proximity to urban areas, this river appears to get high fishing pressure. Michigan’s catch regulations only allow fish to be kept if their length is at least 203 mm. For a heavily fished population, this means that the older fish that remain in the system are often the slow growers (i.e. fish of the same age but smaller than 203 mm). Prairie Creek is really driving the regression line relating brown trout growth and conductivity, and without it this relationship would not be significant (Figure 31). Obviously, what is determining growth of brown trout (age>1) in these streams needs to be looked at further. Age zero brown trout growth was significantly related to mean summer temperature. Age zero brown trout incremental growth increased with increasing temperature. This reinforces the notion that smaller fish are more sensitive to temperature than older, larger fish. These data are also more robust since fishing pressure does not directly alter natural variability in growth of young of the year (YOY) fish. As with density, growth of trout below dams would not be expected to be different from upstream reaches unless the dam altered the downstream habitat in a way that was significant to growth. Significant changes would include alterations of the downstream thermal regime and changes in the macroinvertebrate community. Another study of trout grth above and below a dam that did not alter temperature found no differences in growth of brook, brown or rainbow trout (Klomp 1998). Temperature plays a dual role in that it directly impacts growth, as I have described earlier, and indirectly impacts growth by directly influencing primary and secondary production. 87 o 0 . . 0 . 250 300 350 400 450 500 Conductivity (us/cm) o T T 1 1 I 250 300 350 400 450 500 Conductivity (us/cm) Figure 31. The relationship between incremental growth of brown trout age>1 and conductivity without Prairie Creek's data (Top) and with Prairie Creek included (Bottom). 88 Stream macroinvertebrates, which make up a large portion of the secondary production in streams, are a primary food source for trout and can be strongly altered by temperature (Fraley 1979; Ward and Stanford 1979; Hawkins et al. 1983; Rader 1997). Models predicting growth of brown trout fed on maximum rations were not appreciably higher than the growth rates of brown trout measured in field studies, suggesting that food supply in nature is often adequate and therefore temperature may be the primary determinant of growth rates (Edwards et a1. 1979). In this study, the macroinvertebrate community showed shifts in certain taxa according to temperature increases below the dams, but there were no significant decreases in overall abundance of macroinvertebrates (Dowagiac Creek excepted). Therefore, food for trout in these streams does not appear to have been a major factor in determining their density or growth rates. Temperature was still most likely the primary factor determining growth rates of both YOY and older brown trout. A dam can improve growth by increasing summer temperatures to a point, but once the physiological optimum of trout is exceeded growth would be expected to decline. In these streams, the trout populations tended to respond with density changes rather than large growth differences. It may be that a temperature stressed fish will choose to escape to cooler downstream reaches, if they are available, but if they can find no refugia they may die before showing poor growth. Temperature alterations from these dams in other seasons may also impact egg, larval and juvenile mortality rates which would eventually alter overall density estimates and not necessarily growth. 89 Conclusions Stream habitats change from headwater reaches to the mouth, and dams are known to alter certain habitat parameters so that the normal continuum is disrupted. In this study I compared several habitat parameters in stream sections above and below ten dams in Michigan. The dams on these rivers did not appear to significantly alter the physical attributes (e. g. depth, width, substrate size) of these streams beyond the natural progression that would be expected for downstream reaches when compared to upstream reaches of unregulated streams. Likewise, chemical attributes of these streams (e. g. dissolved oxygen, nutrient levels) also were not significantly altered downstream, with the exception of mean summer temperature. In general, mean summer temperature was substantially increased downstream by these small, surface release facilities. Downstream macroinvertebrate and fish communities responded differently to increased mean summer temperatures below dams. The macroinvertebrate community showed no significant change below dams in terms of family richness, but did respond with decreases in certain taxa that are sensitive to increased temperatures, particularly plecopterans. Dams that significantly increased summer temperatures (i.e. high impact streams) showed relatively large changes in community composition. This resulted in a strong relationship between macroinvertebrate community similarity and temperature change below the dams. It appeared, at the family level, that macroinvertebrates may have been responding to temperature increases by replacing taxa based on temperature preferences rather than changes in richness. Fish community composition also showed a general shift from more cold- stenothermic species upstream to more cool-water, eurythermic species below high 90 impact dams. However, the relationship between temperature increases (below dams) and fish community similarity was not as strong as for macroinvertebrates. Instead, the fish community below high impact dams showed increases in species richness rather than a consistent pattern of species replacement. Although the macroinvertebrate and fish communities changed differently below high impact dams, the discontinuity in both cormnunities below these dams indicates that dams that increase temperature may have an important impact on downstream communities. Investigations of dam impacts on the presence and density of coldwater fish species including brook trout, brown trout, slimy sculpin and mottled sculpin revealed differing relationships with temperature. Both trout species showed significant, negative relationships with increasing summer temperature, resulting in lower population densities below dams where downstream temperature exceeded 20°C . Probability of trout occurring in the study streams also declined with increasing temperature. Brook trout were more sensitive to warming, which resulted in their elimination below many of the dams in this study. Slimy sculpin presence and density were also negatively related with increasing summer temperature, but not as strongly as the two trout species. Mottled sculpin presence and density were not significantly related to temperature or any of the habitat parameters measured in this study. The relationship that the two trout species have with temperature as well as the dramatic reductions noted below high impact dams, leads me to conclude that these types of dams are having important, detrimental effects on these coldwater species. Growth of brown trout agel and older was not related to temperature, but instead showed a significant, negative relationship with stream conductivity. These results were 91 surprising and are most likely due to the uneven distribution of fish caught above age one. Determinants of brown trout growth for older fish in these streams needs further investigation. Age zero brown trout growth (i.e. growth of age 1 fish in the previous year) was significantly, positively related to mean summer temperature. This relationship was not surprising since temperatures, where brown trout were present, were within the tolerance range for this species. It appears that trout populations respond to temperatures above their optimum with decreases in abundance rather than reductions in growth rates. At this time I can only speculate as to the mechanism leading to these reductions in trout density below high impact dams, but it may be that these dams also affect these species at other life stages. Increased mortality at the egg, larval or juvenile stage below high impact dams would lead to decreases in population densities downstream without necessarily leading to reductions in growth. It may be that individuals that do live to be adults feed and grow well downstream of dams, especially with reduced intra-specific competition. Once temperatures become too warm in the summer, trout species most likely migrate downstream in search of thermal refugia, and perish if none is available. My results indicate that small, surface release dams often significantly increase downstream summer temperatures. These increases in temperature were maintained at least 2-3 miles below the dams. When this occurs, downstream communities often respond to warming below darns with reductions in plecopterans, shifts in the macroinvertebrate community, increased fish species richness, reductions in brown trout, brook trout and slimy sculpin population densities and increased growth rates in age] brown trout. If summer temperatures below a darn exceeds 20°C, it is very likely that 92 brook trout will be eliminated entirely from downstream waters. In order to retain viable coldwater fisheries downstream of these types of dams, alteration of the dam operation appears necessary. Releasing water from a deeper, colder level, in stratified reservoirs, is one option. Many of these dams, however, are small facilities, with reservoirs that donot stratify and no longer produce power. Another option, which may be viable for certain communities, is to construct a channel to bypass the dam which would allow communities to keep their reservoirs and potentially also a downstream coldwater fishery. One complication though is that dams require frequent upkeep to retain their physical integrity and public safety. For this reason, the removal of dams should be seriously considered. In conclusion, communities need to consider their mutual interests and do a cost/benefit analysis when making decisions about dam maintenance, fishery management, and dam removal. 93 APPENDICES 94 mm... mm... Mn... 0... 0.0 m0... 0.00.0... .0002 000... N00... .000... No.0 .08... .000... 00.0>-0 ...000 0008.8. ..0... 080.0080. 0008.0... ..0... 0.80.0080. 0.80.0080. 0.0.8.030 2+ =30...ch .0... .0... .0... on... 00... ..md 0.0000-.. .0002 ..o . ..o «00.0 m . .o .0000 .000... 00.0>-0 0090.0 0030080 880000000 0.80.0080. 00080.. .000... 080.0080. 080.0080. 0.0.8.030 2+ 0.00.88. w»... w»... mm... 00.0 00.0 mm... 0.00.0-0 .0002 .0... m . .o 00.0 ...o 00.0 08.0 00.0>-0 09.0...0I0... 09.0..0I...000 808.030... 09.0..0I0... 09.801500... 09.0..0I...000 0.0.8.800 08.0.8.5 80... 00.0 00.0 00.0 mad .3... 00.0 0.0000... .0002 00... wood .08... woos... .000... m8... 00.0>-0 30.00-0800. 09.080 09.804000... 09.8.01008 09.0..0I080. 09.0..0I0..00 09.0..0I080. 09.0..0I080. 0.0.8.0200 000.80.». 00.000m an... hm... hm... 0m... and mm... 00.0 0.0000-.. .0002 90.0.0 mac... .000... we... Soc... .000... 00.0>-0 0.080000 0008 0.000 0.0.0.0080. ...000 080.0080. 0.0.0.0080. 0.0.8.050 0008.0... 00.000w 00.... w0.o .0... .0... m0... 00... an... 0.0000-.. .0002 ..m... 00.0 wmod v.0 w~o.o 0N0... 00.0>-0 08.0.8.8 09.8.0J00 09.800000 09.801080. 0w0080I0000 09.0..0I080. 09.8.0080. 0.0.8.030 0.0.00..0>80.002 o. .o 0 . .o 0. .o m . .o m . .o 08.0 0.0000... .0002 No.0 0 . .o 80.0 000.0 30.0 50.0 00.0>-0 8.0.3 0008000000 ...000 0.0.3 0.000 ...000 0.0.8.050 000.0> x000. .....0.m. R... mm... mm... mm... mm... 0.... 0.000050 .0002 So... 0 . .o 2.0... NE... .8... 000.0 00.0>-0 0000003. 009.08 003000... 000.80.. ..0... 0.80.0080. 0093.0 002000... 0.80.0080. 0.80.0080. 0.0.8.050 3.800 00.0.0000... 000.0 000.0 000... mg... who... 00.0 0.0000-.. .0002 0mm... m0~d m-d um... 00. .0 0m. .0 00.0>-0 0008.80. 0.000000 0008 00938 0020080 8.800.800 0.080000 0008 0008000080 0080800080 0.0.8.080 0.0.00..0>80.002 8.080.004” .0.080.00- 5.08080. 0.08....» 8000000.. 0.03000 m.0 8.30% 0.080000 8008 0080800080 8.0800 .00.. 8.000 0.0.0.0080. 0.80.0080. 0.0.8.050 88080.08. 0.00 03. 8.0 3.0 00.0 «0.0 N00 00.0 0.00.0050 .0002 .0000 .0000 «00.0 3.0 N00 000.0 00.0>.0 0080800080 ...>..000800 0.80.0080. 0080800080 8.8000800 830000800 0.0.8.050 00.0.0... 000 00.0 00.0 00.0 00.0 00.0 0.00.0... .0002 0000.0 300.0 .000 00.0 .000 000.0 00.0>-0 0080800080 83.000800 0.0.0.0080. 0080800080 ....>..000800 ....>..000800 0.0.8.050 000.08. 00.0 00.0 mm... 3.0 3.0 «.00 0.00.002”. .0002 000.0 N00 .00 000.0 800 No.0 00.0>-0 0.080800 0008 0.080800 8008. 8.8000800 0.000800 0008 0.080800 8008 8.8000800 0.0.8.850 00.0.08. Nm.0 de N020 .~.0 .~.0 N. .0 0.00.005”. .0002 000.0 80.0 N000 0.0.0 20.0 0.0.0 00.0.70 0.8000800 000880.. 80... 0.0.0.0080. 8.000 000880.. 80... 000880.. 80... 0.0.8.860 ..+ 00808.8... 000 mm .0 mm .0 00.0 00.0 00.0 0.0000-.. .0002 00.0 ~80 000.0 30.0 .0000 .0000 00.0>.0 8.0.3 830000800 0.0.0.0080. 000880.. .0008. 0.0.0.0080. 0.0.0.0080. 0.0.8.0300 0+ 88.0.8. 0.080.000-..” .0.080.00-N 0.08800. 0.88....» 8008000.. .008. .< 0082...... 96 APPENDIX B: Total catch of each fish species (all passes combined) for all sites on all ten study streams. =§tream Site Species Total BOARDMAN l Brook Trout l7 BOARDMAN 1 Brown Trout 62 BOARDMAN l Slimy Sculpin 28 BOARDMAN 2 Brook Trout 20 BOARDMAN 2 Brown Trout 27 BOARDMAN 2 Slimy Sculpin 93 BOARDMAN 2 White Sucker 3 BOARDMAN 3 Brook Trout l6 BOARDMAN 3 Brown Trout 56 BOARDMAN 3 Slimy Sculpin 58 BOARDMAN 4 Blacknose Dace 3 BOARDMAN 4 Brown Trout 38 BOARDMAN 4 Common Shiner l8 BOARDMAN 4 Creek Chub 2 BOARDMAN 4 Lamprey 2 BOARDMAN 4 Largemouth Bass 3 BOARDMAN 4 Rock Bass 8 BOARDMAN 4 Slimy Sculpin 14 BOARDMAN 4 Smallmouth Bass 1 BOARDMAN 4 Warmouth 2 BOARDMAN 4 White Sucker 51 BOARDMAN 5 Blacknose Dace 8 BOARDMAN 5 Brook Trout 1 BOARDMAN 5 Brown Trout 53 BOARDMAN 5 Common Shiner 2 BOARDMAN 5 Green Sunfish 1 BOARDMAN 5 Slimy Sculpin 13 BOARDMAN 5 Warmouth 1 BOARDMAN 5 White Sucker 2 BOARDMAN 6 Blacknose Dace 3O BOARDMAN 6 Brook Trout 5 BOARDMAN 6 Brown Trout 26 BOARDMAN 6 Common Shiner 4 BOARDMAN 6 Slimy Sculpin 36 BOARDMAN 6 White Sucker 4 CEDAR l Brook Trout 9 CEDAR 1 Brown Trout 18 CEDAR 1 Central Stoneroller 9 CEDAR l Slimy Sculpin 95 CEDAR 2 Brook Trout 2 CEDAR 2 Brown Trout 54 CEDAR 2 Slimy Sculpin 51 CEDAR 3 Brook Trout l9 CEDAR 3 Brown Trout 6O CEDAR 3 Slirny Sculpin 71 97 APPENDIX B cont'd Stream Site S lea Total CEDAR 3 White Sucker l CEDAR 4 Brook Trout 2 CEDAR 4 Brown Trout 10 CEDAR 4 Central Stoneroller l CEDAR 4 Largemouth Bass 3 CEDAR 4 Rock Bass 1 CEDAR 4 Slimy Sculpin l3 CEDAR 4 White Sucker 4 CEDAR 5 Bluegill 3 CEDAR 5 Common Shiner 1 CEDAR 5 Green Sunfish 3 CEDAR 5 Iowa Darter 2 CEDAR 5 Largemouth Bass 2 CEDAR 5 Logperch 2 CEDAR 5 Rock Bass 13 CEDAR 6 Common Shiner 1 CEDAR 6 Grass Pickeral l CEDAR 6 Green Sunfish 4 CEDAR 6 Largemouth Bass 4 CEDAR 6 Logperch 4 CEDAR 6 Rock Bass 3 CEDAR 6 Smallmouth Bass 3 CEDAR 6 Yellow Perch 12 DOWAGIAC 1 Blacknose Dace 5 DOWAGIAC l Bluntnose Minnow l6 DOWAGIAC 1 Common Shiner 49 DOWAGIAC 1 Creek Chub 13 DOWAGIAC 1 Green Sunfish ll DOWAGIAC 1 Johnny Darter 14 DOWAGIAC l Largemouth Bass 3 DOWAGIAC l Mottled Sculpin l DOWAGIAC 1 Rainbow Darter 5 DOWAGIAC 1 River Chub 10 DOWAGIAC 1 Rock Bass 3 DOWAGIAC 1 Yellow Bullhead 1 DOWAGIAC 2 Blacknose Dace 96 DOWAGIAC 2 Bluegill 4 DOWAGIAC 2 Bluntnose Minnow 11 DOWAGIAC 2 Brown Trout 8 DOWAGIAC 2 Central Stoneroller 9 DOWAGIAC 2 Common Shiner 24 DOWAGIAC 2 Creek Chub 44 DOWAGIAC 2 Grass Pickeral 1 DOWAGIAC 2 Green Sunfish 15 DOWAGIAC 2 Johnny Darter 4 DOWAGIAC 2 Lamprey 11 98 APPENDIX B cont'd Stream Site S ies Total DOWAGIAC 2 Largemouth Bass 1 DOWAGIAC 2 Mottled Sculpin 76 DOWAGIAC 2 Pumpkinseed 1 DOWAGIAC 2 Rock Bass 2 DOWAGIAC 2 White Sucker 48 DOWAGIAC 2 Yellow Bullhead 2 DOWAGIAC 3 Blacknose Dace 47 DOWAGIAC 3 Bluegill 6 DOWAGIAC 3 Brown Trout l DOWAGIAC 3 Central Stoneroller 2 DOWAGIAC 3 Green Sunfish 1 DOWAGIAC 3 Johnny Darter 1 DOWAGIAC 3 Lamprey 5 DOWAGIAC 3 Mottled Sculpin 108 DOWAGIAC 4 Bowfin l DOWAGIAC 4 Common Shiner 1 DOWAGIAC 5 Bowfin 1 DOWAGIAC 5 Grass Pickeral 1 DOWAGIAC 5 Rainbow Darter 1 DOWAGIAC 5 Rock Bass 1 DOWAGIAC 5 Smallmouth Bass 2 DOWAGIAC 5 Stonecat 1 DOWAGIAC 5 Yellow Bullhead 3 DOWAGIAC 6 Common Shiner 1 DOWAGIAC 6 Creek Chub 5 DOWAGIAC 6 Green Sunfish l DOWAGIAC 6 Northern Hogsucker 3 DOWAGIAC 6 Rock Bass 1 DOWAGIAC 6 Stonecat 2 DOWAGIAC 7 Carp l DOWAGIAC 7 Johnny Darter 1 FISH 1 Blacknose Dace 51 FISH 1 Blacksided Darter 2 FISH 1 Bluegill 2 FISH l Bluntnose Minnow l FISH l Brook Stickleback 5 FISH 1 Brown Trout 6 FISH 1 Central Stoneroller 3 FISH 1 Common Shiner 86 FISH l Creek Chub 102 FISH l Homyhead Chub 6 FISH 1 Johnny Darter 23 FISH 1 Lamprey 4 FISH 1 Largemouth Bass 4 FISH l Mottled Sculpin l3 FISH 1 Northern Hogsucker 2 99 APPENDIX B cont'd Stream Site Species Total FISH 1 Rainbow Darter ' 94 FISH 1 River Chub 9 FISH 1 Rock Bass 3 FISH l Rosyface Shiner l FISH 1 White Sucker 24 FISH 2 Blacknose Dace 49 FISH 2 Blacksided Darter 4 FISH 2 Bluegill 15 FISH 2 Brook Stickleback 13 FISH 2 Common Shiner 106 FISH 2 Creek Chub 29 FISH 2 Hornyhead Chub 9 FISH 2 Johnny Darter 12 FISH 2 Lamprey 2 FISH 2 Mottled Sculpin l FISH 2 Rainbow Darter 49 FISH 2 Rosyface Shiner 4 FISH 2 White Sucker l FISH 5 Blacknose Dace 78 FISH 5 Blacksided Darter 27 FISH 5 Bluegill 3 FISH 5 Carp 2 FISH 5 Central Stoneroller l FISH 5 Common Shiner 149 FISH 5 Creek Chub 97 FISH 5 Green Sunfish 4 FISH 5 Johnny Darter 106 FISH 5 Lamprey 10 FISH 5 Logperch 6 FISH 5 Northern Hogsucker 9 FISH 5 Northern Pike 1 FISH 5 Pumpkinseed 3 FISH 5 Rainbow Darter l3 FISH 5 Rock Bass 13 FISH 5 Shorthead Redhorse 28 FISH 5 Smallmouth Bass 8 FISH 5 White Sucker 115 FISH 6 Blacknose Dace 63 FISH 6 Blacksided Darter 29 FISH 6 Bluegill I FISH 6 Bluntnose Minnow 2 FISH 6 Burbot 1 FISH 6 Common Shiner 183 FISH 6 Creek Chub 79 FISH 6 Emerald Shiner 1 FISH 6 Green Sunfish l 100 APPENDIX B cont'd Stream Site S lee Total FISH 6 Johnny Darter 33 FISH 6 Lamprey 19 FISH 6 Logperch 2 FISH 6 Northern Hogsucker l FISH 6 Pumpkinseed l FISH 6 Rock Bass 17 FISH 6 Rosyface Shiner 4 FISH 6 Shorthead Redhorse l7 FISH 6 Smallmouth Bass 17 FISH 6 White Sucker 44 FISH 6 Yellow Bullhead 1 MANTON l Brook Trout 42 MANTON 1 Brown Trout 39 MANTON 1 Central Stoneroller 1 MANTON l Slimy Sculpin 19 MANTON 2 Brook Trout 28 MANTON 2 Brown Trout 34 MANTON 2 Cenn'aI Stoneroller 1 MANTON 2 Slimy Sculpin 20 MANTON 3 Brook Trout 27 MANTON 3 Brown Trout 70 MANTON 3 Lamprey 2 MANTON 3 Mottled Sculpin 1 MANTON 3 Slirny Sculpin 26 MANTON 3 White Sucker l MANTON 4 Blacknose Dace 123 MANTON 4 Brook Stickleback l MANTON 4 Brown Trout 42 MANTON 4 Central Stoneroller l MANTON 4 Creek Chub 25 MANTON 4 Mottled Sculpin 14 MANTON 4 White Sucker 22 MANTON 5 Blacknose Dace 39 MANTON 5 Brown Trout 9 MANTON 5 Central Stoneroller 1 MANTON 5 Common Shiner 48 MANTON 5 Creek Chub 42 MANTON 5 Homyhead Chub 2 MANTON 5 Lamprey 3 MANTON 5 Mottled Sculpin 2 MANTON 5 Northern Pike 2 MANTON 5 Rock Bass 3 MANTON 5 Rosyface Shiner l MANTON 5 Spottail Shiner 5 MANTON 5 White Sucker 15 MANTON 6 Blacknose Dace 55 101 APPENDIX B cont'd Stream Site Species Total MANTON 6 Brown Trout 20 MANTON 6 Common Shiner 18 MANTON 6 Creek Chub 54 MANTON 6 Northern Pike 46 MANTON 6 Pumpkinseed l MANTON 6 White Sucker 20 MANTON 7 Blacknose Dace 40 MANTON 7 Bluegill 3 MANT ON 7 Brook Trout 2 MANTON 7 Brown Trout l9 MANTON 7 Creek Chub 27 MANTON 7 Lamprey 2 MANTON 7 Logperch 2 MANTON 7 White Sucker 4 MAPLE 1 Emerald Shiner 6 MAPLE I Rock Bass 10 MAPLE l Slimy Sculpin 9 MAPLE 1 White Sucker 3 MAPLE 2 Brook Stickleback l MAPLE 2 Brook Trout 55 MAPLE 2 Central Stoneroller 3 MAPLE 2 Lamprey l MAPLE 2 Slimy Sculpin 40 MAPLE 2 White Sucker 1 MAPLE 3 Brook Trout 8 MAPLE 3 Brown Trout 17 MAPLE 3 Creek Chub 1 MAPLE 3 Rainbow Trout 3 MAPLE 3 Slimy Sculpin 30 MAPLE 3 White Sucker 4 MAPLE 4 Banded Killifish 1 MAPLE 4 Blacknose Dace 1 MAPLE 4 Bowfm l MAPLE 4 Central Stoneroller l MAPLE 4 Creek Chub 39 MAPLE 4 Emerald Shiner ll MAPLE 4 Golden Shiner 7 MAPLE 4 Johnny Darter 1 MAPLE 4 Largemouth Bass 1 MAPLE 4 Northern Pike 3 MAPLE 4 White Sucker 3 MAPLE 4 Yellow Perch 1 MAPLE 5 Blacknose Dace 1 MAPLE 5 Brook Trout 4 MAPLE 5 Central Stoneroller 4 MAPLE 5 Common Shiner 13 102 APPENDIX B cont'd Stream Site S ies Total MAPLE 5 Creek Chub 33 MAPLE 5 Lamprey 2 MAPLE 5 Mottled Sculpin 9 MAPLE 5 Northern Pike 1 MAPLE 5 Slimy Sculpin 50 MAPLE 5 White Sucker 27 MAPLE 6 Brook Trout 4 MAPLE 6 Brown Trout 7 MAPLE 6 Creek Chub 4 MAPLE 6 Rainbow Trout 3 MAPLE 6 Rock Bass 6 MAPLE 6 Slirny Sculpin 98 MAPLE 6 Yellow Perch 1 MAPLE 7 Brook Trout 4 MAPLE 7 Brown Trout 28 MAPLE 7 Rainbow Trout 5 MAPLE 7 Slimy Sculpin 36 MAPLE 7 Spottail Shiner 1 MAPLE 7 White Sucker 3 MAPLE 7 Yellow Perch l MAPLE 8 Brook Trout 4 MAPLE 8 Brown Trout 12 MAPLE 8 Largemouth Bass 1 MAPLE 8 Mottled Sculpin 1 MAPLE 8 Rainbow Trout 2 MAPLE 8 Slimy Sculpin 33 MAPLE 8 Yellow Perch 1 MIDDLE 1 Blacknose Dace 8 MIDDLE 1 Brown Trout 4 MIDDLE 1 Central Stoneroller 4 MIDDLE l Creek Chub 4 MIDDLE 1 Green Sunfish 1 MIDDLE 1 Johnny Darter 1 MIDDLE 1 Lamprey l MIDDLE l Longnose Dace 1 MIDDLE 1 Mottled Sculpin 19 MIDDLE 1 Warmouth 4 MIDDLE 2 Black Bullhead 3 MIDDLE 2 Blacknose Dace 7 MIDDLE 2 Bluegill 4 MIDDLE 2 Brook Trout l MIDDLE 2 Brown Trout 23 MIDDLE 2 Central Stoneroller 1 MIDDLE 2 Creek Chub 10 MIDDLE 2 Longnose Dace l MIDDLE 2 Mottled Sculpin 50 103 APPENDIX B cont'd Stream MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE MIDDLE Site N amooo‘o‘ao‘ummmummmmum-hAvbbAAAJstthJs-hhwuuuwuuwwuwto S es Warmouth White Sucker Black Bullhead Blacknose Dace Blacksided Darter Brown Trout Central Stoneroller Creek Chub Green Sunfish Johnny Darter Mottled Sculpin Warmouth White Sucker Black Bullhead Blacknose Dace Blacksided Darter Bowfin Brown Trout Common Shiner Creek Chub Johnny Darter Lamprey Mottled Sculpin Northern Hogsucker Northern Pike Warmouth White Sucker Blacknose Dace Blacksided Darter Bowfin Central Stoneroller Common Shiner Creek Chub Johnny Darter Lamprey Mottled Sculpin Warmouth White Sucker Blacknose Dace Blacksided Darter Central Stoneroller Common Shiner Creek Chub Johnny Darter Lamprey Longnose Dace 104 APPENDIX B cont'd Stream Site Species Total MIDDLE 6 Warmouth 2 MIDDLE 6 White Sucker 2 PRAIRIE 1 Blacknose Dace 55 PRAIRIE 1 Brown Trout 6 PRAIRIE 1 Central Stoneroller 8 PRAIRIE l Creek Chub 63 PRAIRIE 1 Green Sunfish 2 PRAIRIE 1 Johnny Darter l8 PRAIRIE 1 Largemouth Bass 1 PRAIRIE 1 Mottled Sculpin 78 PRAIRIE 1 Pumpkinseed 1 PRAIRIE 1 Rainbow Trout 6 PRAIRIE 1 Rainbow Darter 7 PRAIRIE 1 White Sucker 34 PRAIRIE 2 Blacknose Dace l 10 PRAIRIE 2 Brown Trout 9 PRAIRIE 2 Central Stoneroller l PRAIRIE 2 Creek Chub 50 PRAIRIE 2 Johnny Darter 5 PRAIRIE 2 Lamprey 1 PRAIRIE 2 Mottled Sculpin 2 PRAIRIE 2 Rainbow Trout 15 PRAIRIE 2 Rainbow Darter 1 PRAIRIE 3 Blacknose Dace 26 PRAIRIE 3 Bluegill l PRAIRIE 3 Brown Trout 1 PRAIRIE 3 Creek Chub 31 PRAIRIE 3 Green Sunfish l PRAIRIE 3 Johnny Darter 20 PRAIRIE 3 Mottled Sculpin 20 PRAIRIE 3 Rainbow Trout 4 PRAIRIE 3 Rainbow Darter 4 PRAIRIE 3 White Sucker 1 PRAIRIE 4 Blacknose Dace 40 PRAIRIE 4 Brown Trout 10 PRAIRIE 4 Central Stoneroller 2 PRAIRIE 4 Creek Chub 90 PRAIRIE 4 Green Sunfish 16 PRAIRIE 4 Johnny Darter 15 PRAIRIE 4 Lamprey 5 PRAIRIE 4 Largemouth Bass 1 PRAIRIE 4 Mottled Sculpin 49 PRAIRIE 4 Pumpkinseed 1 PRAIRIE 4 Rainbow Trout ll PRAIRIE 4 Rainbow Darter 32 PRAIRIE 4 Warmouth 4 105 APPENDIX B cont'd 106 Stream Site S . ' lee Total PRAIRIE 4 White Crappie 2 PRAIRIE 4 White Sucker ll PRAIRIE 5 Blacknose Dace 42 PRAIRIE 5 Brown Trout l PRAIRIE 5 Creek Chub 25 PRAIRIE 5 J ohnny Darter 6 PRAIRIE 5 Mottled Sculpin 10 PRAIRIE 5 Rainbow Trout l PRAIRIE 5 Rainbow Darter 17 PRAIRIE 5 White Sucker 2 PRAIRIE 6 Blacknose Dace 12 PRAIRIE 6 Brown Trout 2 PRAIRIE 6 Central Stoneroller 1 PRAIRIE 6 Creek Chub 36 PRAIRIE 6 Johnny Darter 10 PRAIRIE 6 Lamprey l PRAIRIE 6 Mottled Sculpin 2 SUGAR 1 Blacknose Dace 2 SUGAR 1 Green Sunfish 1 SUGAR 1 Northern Pike 3 SUGAR l Warmouth 1 SUGAR 1 White Sucker 1 SUGAR 2 Blacknose Dace l9 SUGAR 2 Bluegill 1 SUGAR 2 Brown Trout 16 SUGAR 2 Central Stoneroller l SUGAR 2 Common Shiner 1 SUGAR 2 Largemouth Bass 3 SUGAR 2 Mottled Sculpin 1 SUGAR 2 Northern Pike 5 SUGAR 2 Pumpkinseed 1 SUGAR 3 Bluegill 10 SUGAR 3 Brown Trout l6 SUGAR 3 Green Sunfish 1 SUGAR 3 Largemouth Bass 3 SUGAR 3 Mottled Sculpin 4 SUGAR 3 Northern Pike 1 SUGAR 3 Rock Bass 1 SUGAR 4 Blacksided Darter 1 SUGAR 4 Bluegill 3 SUGAR 4 Common Shiner 2 SUGAR 4 Creek Chub 6 SUGAR 4 Green Sunfish 5 SUGAR 4 Logperch 1 SUGAR 4 Longnose Dace ‘ 1 SUGAR 4 Rainbow Darter 1 APPENDIX B cont'd Stream Site S - SUGAR 4 Rock Bass 2 SUGAR 4 Warmouth 1 SUGAR 4 White Crappie 5 SUGAR 4 White Sucker 7 SUGAR 4 Yellow Bullhead 5 SUGAR 5 Blacknose Dace l SUGAR 5 Common Shiner 20 SUGAR 5 Creek Chub 3 SUGAR 5 Green Sunfish 1 SUGAR 5 Homyhead Chub 6 SUGAR 5 Lamprey l SUGAR 5 Logperch 2 SUGAR 5 Northern Hogsucker 1 SUGAR 5 Rainbow Darter 3 SUGAR 5 White Crappie l SUGAR 5 White Sucker 22 SUGAR 6 Blacknose Dace 2 SUGAR 6 Blacksided Darter 6 SUGAR 6 Bluegill l SUGAR 6 Brown bu 2 SUGAR 6 Central Stoneroller 4 SUGAR 6 Common Shiner 7 SUGAR 6 Creek Chub 1 SUGAR 6 Green Sunfish l SUGAR 6 Homyhead Chub 1 SUGAR 6 Iowa Darter 2 SUGAR 6 Lamprey 2 SUGAR 6 Logperch 3 SUGAR 6 Longnose Dace 2 SUGAR 6 Rainbow Darter 6 SUGAR 6 Rock Bass 2 SUGAR 6 White Crappie l SUGAR 6 Yellow Bullhead 1 WHITE 1 Blacknose Dace 2 WHITE 1 Brook Trout 5 WHITE 1 Brown Trout 7 WHITE 1 Creek Chub 1 WHITE 1 Green Sunfish 1 WHITE 1 Mottled Sculpin 26 WHITE 1 White Sucker 16 WHITE 2 Brook Trout 7 WHITE 2 Brown Trout 13 WHITE 2 Mottled Sculpin 44 WHITE 2 White Sucker 31 WHITE 3 Blacknose Dace 6 WHITE 3 Brown Trout 6 107 APPENDIX B cont'd Stream Site Species Total WHITE 3 Central Stonero=ller 1 WHITE 3 Creek Chub 5 WHITE 3 Longnose Dace 6 WHITE 3 Mottled Sculpin 29 WHITE 3 Northern Pike 1 WHITE 3 White Sucker 7 WHITE 4 Blacknose Dace 72 WHITE 4 Brown Trout 17 WHITE 4 Common Shiner 20 WHITE 4 Creek Chub 10 WHITE 4 Green Sunfish 1 WHITE 4 Johnny Darter 2 WHITE 4 Lamprey 2 WHITE 4 Longnose Dace 2 WHITE 4 Mottled Sculpin 11 WHITE 4 Rainbow Darter 2 WHITE 4 White Sucker 52 WHITE 5 Blacknose Dace 20 WHITE 5 Brown Trout 10 WHITE 5 Central Stoneroller 5 WHITE 5 Common Shiner 12 WHITE 5 Creek Chub 10 WHITE 5 Homyhead Chub 1 WHITE 5 Iowa Darter 1 WHITE 5 Johnny Darter l 1 WHITE 5 Lamprey 2 WHITE 5 Longnose Dace 1 WHITE 5 Mottled Sculpin 25 WHITE 5 Pumpkinseed 2 WHITE 5 White Sucker 24 108 APPENDIX C: Total number of macroinvertebrates per sample and per square meter in each site for all ten study streams. Stream BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN BOARDMAN Site — «h-h&#UJUWWMWWWWWNNNNNNNNNNNNNNH—Ht-‘t-‘F‘D-‘t-‘t-‘F‘D-‘e-‘e-‘I-‘F‘h‘ Taxa Total NoJSqM Tipulidae (Antocha) 2 0.065 Athericidae 11 0.356 Baetidae 39 1.264 Brachycentridae 1 0.032 Chironomidae 320 10.368 Elmidae 56 1.814 Empididae 17 0.55 l Ephemeridae 5 0.162 Glossosomatidae 5 0.162 Gomphidae 1 0.032 Hydropsychidae 10 0.324 Limnephilidae 4 0.130 Oligocheates 1 1 0.356 Perlodidae 1 0.032 Philipotamidae 1 0.032 Simulidae 53 1.717 Tricorythidae 10 0.324 Baetidae 18 0.583 Chironomidae 190 6.156 Elmidae 12 0.389 Empididae 1 0.032 Ephemeridae 1 0.032 Gomphidae 2 0.065 Hydropsychidae 10 0.324 Isopoda 1 0.032 Oligocheates 25 0.810 Perlodidae 1 0.032 Pteronarcyidae 2 0.065 Simulidae 29 0.940 Tabanidae 2 0.065 Tricorythidae 16 0.518 Tipulidae (Antocha) 1 0.032 Athericidae 6 0.194 Baetidae 3 0.097 Chironomidae 92 2.981 Elmidae 5 0.162 Ephemeridae 2 0.065 Heptageniidae 1 0.032 Nemouridae 1 0.032 Oligocheates 9 0.292 Simulidae 5 0.162 Tipulidae (Antocha) 3 0.097 Baetidae 10 0.324 Chironomidae 38 1.231 Elmidae 52 1.685 109 APPENDIX C cont'd Stream Site Taxa Total NoJSq.M BOARDMAN 4 Ephemeridae 5 0. 162 BOARDMAN 4 Hydropsychidae 138 4.471 BOARDMAN 4 Hirudinea 19 0.616 BOARDMAN 4 Lepidstomatidae 5 0.162 BOARDMAN 4 Oligocheates 60 1.944 BOARDMAN 4 Simulidae 13 0.421 BOARDMAN 4 T ricorythidae 1 0.032 BOARDMAN 5 Tipulidae (Antocha) 10 0.324 BOARDMAN 5 Athericidae 3 0.097 BOARDMAN 5 Baetidae 6 0. 194 BOARDMAN 5 Brachycentridae 1 1 0.356 BOARDMAN 5 Chironomidae 20 0.648 BOARDMAN 5 Elmidae 93 3.013 BOARDMAN 5 Ephemeridae 8 0.259 BOARDMAN 5 Hydropsychidae 2 1 0.680 BOARDMAN 5 Limnephilidae 1 0.032 BOARDMAN 5 Oligocheates 7 0.227 BOARDMAN 5 Simulidae 1 0.032 BOARDMAN 5 Tricorythidae 2 0.065 BOARDMAN 6 Tipulidae (Antocha) 13 0.421 BOARDMAN 6 Athericidae 4 0.130 BOARDMAN 6 Baetidae 1 1 0.356 BOARDMAN 6 Brachycentridae 12 0.3 89 BOARDMAN 6 Chironomidae 1 50 4. 860 BOARDMAN 6 Elmidae 3 8 1.23 1 BOARDMAN 6 Ephemeridae 9 0.292 BOARDMAN 6 Glossosomatidae 1 0.032 BOARDMAN 6 Helicopsychidae 1 0.032 BOARDMAN 6 Heptageniidae 4 0.130 BOARDMAN 6 Hydropsychidae 10 0.324 BOARDMAN 6 Hirudinea 4 0.130 BOARDMAN 6 Philipotamidae 1 2 0.3 89 BOARDMAN 6 Chironomidae (Rheotanytarsus) 80 2.592 BOARDMAN 6 Rhyacophilidae 2 0.065 BOARDMAN 6 Simulidae 5 0.162 BOARDMAN 6 Tricorythidae 2 0.065 CEDAR l Baetidae 25 0.810 CEDAR 1 Brachycentridae 4 0.130 CEDAR 1 Ceratopogonidae 1 0.032 CEDAR 1 Chironomidae 94 3 .046 CEDAR 1 Elmidae 6 0.194 CEDAR 1 Ephemeridae 5 0.162 CEDAR 1 Hydropsychidae 43 1 .393 CEDAR l Nemouridae 5 0.162 CEDAR 1 Perlodidae 1 0.03 2 CEDAR 1 Philipotamidae 3 0.097 110 APPENDIX C cont'd Stream CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR CEDAR Site ~ MMMMMMMMAAIhbb#bAwwuwwwwwwwwwUWNNNNNNNNNNNNNM!— Taxa Rhyacophilidae Simulidae Amphipoda Tipulidae (Antocha) Baetidae Brachycentridae Ceratopogonidae Chironomidae Elmidae Ephemeridae Hydropsychidae Oligocheates Philipotamidae Rhyacophilidae Simulidae Tabanidae Amphipoda Baetidae Chironomidae Elmidae Ephemeridae Glossosomatidae Hydropsychidae Isopoda Nemouridae Oligocheates Perlodidae Pteronarcyidae Rhyacophilidae Simulidae Amphipoda Baetidae Chironomidae Elmidae Hydropsychidae Isopoda Philipotamidae Simulidae Aeshnidae Amphipoda Baetidae Brachycentridae Ceanidae Chironomidae Coenagrionidae Elmidae 111 Total NoJSrkM 1 0.032 2 0.065 12 0.389 1 0.032 54 1.750 1 0.032 2 0.065 105 3.402 19 0.616 7 0.227 20 0.648 1 0.032 15 0.486 3 0.097 94 3.046 4 0.130 8 0.259 102 3.305 73 2.365 23 0.745 2 0.065 36 1.166 66 2.138 2 0.065 7 0.227 3 0.097 1 0.032 2 0.065 3 0.097 145 4.698 1 0.032 10 0.324 1 15 3.726 3 0.097 1 1 1 3.596 8 0.259 28 0.907 58 1.879 1 0.032 89 2.884 1 0.032 1 0.032 1 0.032 13 0.421 13 0.421 6 0.194 APPENDIX C cont'd Stream Site Taxa Total NoJSchM CEDAR 5 Ephemeridae 3 0.097 CEDAR 5 Isopoda 2 0.065 CEDAR 5 Mesoveliidae 1 0.032 CEDAR 5 Naucon'dae 1 0.032 CEDAR 5 Nepidae 1 0.032 CEDAR 5 Polycentropodidae 1 0.032 DOWAGIAC 1 Amphipoda 1 8 0. 583 DOWAGIAC 1 Baetidae 1 0.032 DOWAGIAC 1 Brachycentridae 1 0.032 DOWAGIAC l Ceanidae 14 0.454 DOWAGIAC 1 Ceratopogonidae 1 0.032 DOWAGIAC l Chironomidae 25 0.810 DOWAGIAC 1 Bivalvia 20 0.648 DOWAGIAC 1 Elmidae l 3 0.421 DOWAGIAC 1 Ephemeridae 3 0.097 DOWAGIAC 1 Heptageniidae 12 0.389 DOWAGIAC l Hydropsychidae 10 0.324 DOWAGIAC 1 Lepidstomatidae 1 0.032 DOWAGIAC 2 Amphipoda 4 0. 130 DOWAGIAC 2 Tipulidae (Antocha) 2 0.065 DOWAGIAC 2 Baetidae 4 0.130 DOWAGIAC 2 Brachycentridae 2 0.065 DOWAGIAC 2 Ceanidae 24 0.778 DOWAGIAC 2 Chironomidae 276 8.942 DOWAGIAC 2 Elmidae 1 6 0.5 1 8 DOWAGIAC 2 Ephemeridae 12 0.389 DOWAGIAC 2 Glossosomatidae 2 0.065 DOWAGIAC 2 Helicopsychidae 14 0.454 DOWAGIAC 2 Hydropsychidae 38 1 .23 1 DOWAGIAC 2 Limnephilidae 2 0.065 DOWAGIAC 2 Oligocheates 2 0.065 DOWAGIAC 2 Gastropoda 12 0.389 DOWAGIAC 2 Tabanidae 2 0.065 DOWAGIAC 2 Tricorythidae 4 0.130 DOWAGIAC 3 Amphipoda 2 0.065 DOWAGIAC 3 Baetidae 12 0.389 DOWAGIAC 3 Ceanidae 2 0.065 DOWAGIAC 3 Chironomidae 58 1 .879 DOWAGLAC 3 Elmidae 32 1 .037 DOWAGIAC 3 Glossosomatidae 20 0.648 DOWAGIAC 3 Helicopsychidae 6 0. 194 DOWAGIAC 3 Heptageniidae 2 0.065 DOWAGIAC 3 Hydropsychidae 144 4.666 DOWAGIAC 3 Oligocheates 6 0.194 DOWAGIAC 3 Perlidae 2 0.065 DOWAGIAC 4 Amphipoda 2 0.065 112 APPENDIX C cont'd Stream Site Taxa Total No./ o.M DOWAGIAC 4 Calopterygidae 3 0.097 DOWAGIAC 4 Chironomidae 8 0.259 DOWAGIAC 4 Bivalvia 2 0.065 DOWAGIAC 4 Culicidae 1 0.032 DOWAGIAC 4 Elmidae 27 0.875 DOWAGIAC 4 Helicopsychidae 2 0.065 DOWAGIAC 4 Heptageniidae 3 0.097 DOWAGIAC 4 Hydropsychidae 1 0.032 DOWAGIAC 4 Isopoda 1 0.032 DOWAGIAC 4 Uenoidae (Neophylax) 3 0.097 DOWAGIAC 4 Oligocheates 2 0.065 DOWAGIAC 4 Psphenidae 1 0.032 DOWAGIAC 4 Gastropoda 2 0.065 DOWAGIAC 4 Tabanidae 2 0.065 DOWAGIAC 5 Amphipoda 2 0.065 DOWAGIAC 5 Baetidae 2 0.065 DOWAGIAC 5 Elmidae 7 0.227 DOWAGIAC 5 Hydropsychidae 2 0.065 DOWAGIAC 5 Philipotamidae 1 0.032 DOWAGIAC 5 Polycentropodidae 2 0.065 DOWAGIAC 5 Psphenidae 3 0.097 DOWAGIAC 6 Brachycentridae 3 0.097 DOWAGIAC 6 Chironomidae 42 1.361 DOWAGIAC 6 Elmidae 25 0.810 DOWAGIAC 6 Helicopsychidae 1 0.032 DOWAGIAC 6 Hydropsychidae 3 0.097 DOWAGIAC 6 Limnephilidae 2 0.065 DOWAGIAC 6 Gastropoda 7 0.227 FISH 1 Baetidae 102 3.305 FISH 1 Brachycentridae 8 0.259 FISH 1 Chironomidae 166 5.378 FISH 1 Elmidae 44 1.426 FISH 1 Ephemeridae 40 1 .296 FISH 1 Heptageniidae 4 0. 130 FISH 1 Hydropsychidae 98 3.175 FISH 1 Limnephilidae 8 0.259 FISH 1 Perlidae 18 0.583 FISH 1 Psphenidae 2 0.065 FISH 1 Pteronarcyidae 2 0.065 FISH 1 Sialidae 2 0.065 FISH l Tricorythidae 2 0.065 FISH 2 Tipulidae (Antocha) 1 0.032 FISH 2 Athericidae 52 1 .685 FISH 2 Baetidae 38 1.231 FISH 2 Brachycentridae 1 0.032 FISH 2 Ceratopogonidae 5 0.162 113 APPENDIX C cont'd Site N Osaaxao‘osaamuma-AAAAAAA-bnwwwwwuuwuwwwwwnmnuwNNNNN Taxa Chironomidae Elmidae Ephemeridae Helicopsychidae Heptageniidae Hydropsychidae Isonychidae Limnephilidae Perlidae Simulidae Tricorythidae Amphipoda Athericidae Ceanidae Chironomidae Elmidae Ephemeridae Heptageniidae Hydropsychidae Limnephilidae Perlidae Chironomidae (Rheotanytarsus) Sialidae Tipulidae Tricorythidae Amphipoda Baetidae Brachycentridae Chironomidae Elmidae Ephemeridae Hydropsychidae Isonychidae Perlidae Simulidae Chironomidae Elmidae Tricorythidae Amphipoda Tipulidae (Antocha) Athericidae Ceanidae Chironomidae Bivalvia Elmidae Ephemeridae 114 189 3 1 2 1 7 APPENDD( C cont'd Stream Site . FISH 6 Heptageniidae 10 0.324 FISH 6 Hydropsychidae 9 0.292 FISH 6 Isonychidae 7 0.227 FISH 6 Isopoda 1 0.032 FISH 6 Sialidae 1 0.032 FISH 6 Tricorythidae 1 0.032 MANTON l Amphipoda 3 0.097 MANTON 1 Tipulidae (Antocha) 1 0.032 MANTON 1 Athericidae 1 0.032 MANTON 1 Baetidae 18 0.583 MANTON 1 Brachycentridae 2 0.065 MANTON 1 Chironomidae 23 7 7.679 MANT ON 1 Chloroperlidae 1 0.032 MANTON 1 Nemouridae 4 0.130 MANTON 1 Perlidae 1 0.032 MANTON 2 Amphipoda 2 0.065 MANTON 2 Athericidae 20 0.648 MANTON 2 Baetidae 14 0.454 MANTON 2 Brachycentridae 14 0.454 MANTON 2 Ceanidae 8 0.259 MANTON 2 Chironomidae 74 2 .398 MANTON 2 Elmidae 5 2 l .685 MANTON 2 Ephemeridae 16 0.518 MANTON 2 Hydropsychidae 366 1 1 .85 8 MANTON 2 Leptoceridae 2 0.065 MANTON 2 Perlidae 10 0.324 MANTON 2 Philipotamidae 2 0.065 MANTON 2 Pteronarcyidae 2 0.065 MANTON 2 Chironomidae (Rheotanytarsus) 2 0.065 MANT ON 2 Simulidae 2 0.065 MANTON 2 Tricorythidae 44 1 .426 MANTON 3 Amphipoda 3 l 1 .004 MANTON 3 Tipulidae (Antocha) 12 0.389 MANT ON 3 Baetidae 38 1.231 MANTON 3 Brachycentridae 1 0.032 MANTON 3 Chironomidae 65 2.106 MANTON 3 Elmidae 3 0.097 MANTON 3 Ephemeridae 1 0.032 MANTON 3 Hydropsychidae 2 0.065 MANTON 3 Oligocheates 4 0.130 MANTON 3 Perlidae 6 0. 194 MANTON 3 Philipotamidae 1 0.032 MANTON 3 Tipulidae 1 0.032 MANT ON 4 Baetidae 8 0.259 MANTON 4 Chironomidae 500 16.200 MANTON 4 Corydalidae 1 0.032 115 APPENDD( C cont'd Stream Site Taxa Total 1‘1ng .M MANTON 4 Elmidae 79 2.560 MANTON 4 Heptageniidae 1 0.032 MANTON 4 Hydropsychidae 9 0.292 MANTON 4 Gastropoda 1 0.032 MANTON 4 Tricorythidae 4 0.130 MANTON 5 Tipulidae (Antocha) 3 0.097 MANTON 5 Baetidae 1 0.032 MANTON 5 Chironomidae 43 1 .393 MANTON 5 Bivalvia 2 0.065 MANTON 5 Elmidae 18 0.583 MANTON 5 Ephemeridae 2 0.065 MANTON S Glossosomatidae 5 0.162 MANTON 5 Heptageniidae 1 0.032 MANTON 5 Hydropsychidae 50 1 .620 MANTON 5 Isonychidae 1 0.032 MANTON 5 Philipotamidae 1 0.032 MANTON 5 Simulidae 7 0.227 MANTON 6 Amphipoda 2 0.065 MANT ON 6 Baetidae 6 0.194 MANTON 6 Ceanidae 1 0.032 MANTON 6 Ceratopogonidae 1 0.032 MANT ON 6 Chironomidae 10 0.324 MANTON 6 Bivalvia 3 0.097 MANTON 6 Tipulidae (Dicranota) 1 0.032 MANTON 6 Elmidae 64 2.074 MANTON 6 Empididae 1 0.032 MANTON 6 Glossosomatidae 1 0.032 MANTON 6 Gomphidae 1 0.032 MANTON 6 Hydropsychidae 2 0.065 MANTON 6 Linmephilidae 1 0.032 MAPLE l Amphipoda 1 1 0.356 MAPLE 1 Baetidae 20 0.648 MAPLE 1 Calopterygidae 4 0.130 MAPLE 1 Chironomidae 17 0.551 MAPLE 1 Hydropsychidae 4 0.130 MAPLE 1 Limnephilidae 2 0.065 MAPLE 1 Rhyacophilidae 1 0.032 MAPLE 1 Simulidae 1 1 0.356 MAPLE 1 Gastropoda 1 0.032 MAPLE 2 Tipulidae (Antocha) 1 0.032 MAPLE 2 Athericidae 1 0.032 MAPLE 2 Baetidae 14 0.454 MAPLE 2 Brachycentridae 40 1 .296 MAPLE 2 Chironomidae 122 3.953 MAPLE 2 Elmidae 6 0.194 MAPLE 2 Ephemeridae 1 0.032 116 APPENDIX C cont'd Stream Site Taxa Total N oJScpM MAPLE 2 Heptageniidae 4 0.130 MAPLE 2 Hydropsychidae 124 4.018 MAPLE 2 Hirudinea 4 0.130 MAPLE 2 Perlidae 1 0.032 MAPLE 2 Perlodidae 3 0.097 MAPLE 2 Philipotamidae 1 0.032 MAPLE 2 Polycentropodidae 7 0.227 MAPLE 2 Simulidae 210 6.804 MAPLE 2 Tricorythidae 13 0.421 MAPLE 3 Tipulidae (Antocha) 3 0.097 MAPLE 3 Baetidae 45 1 .458 MAPLE 3 Brachycentridae 1 l 1 3.596 MAPLE 3 Calopterygidae 2 0.065 MAPLE 3 Chironomidae 236 7.646 MAPLE 3 Corydalidae 2 0.065 MAPLE 3 Elmidae 31 1.004 MAPLE 3 Empididae 1 0.032 MAPLE 3 Heptageniidae 8 0.259 MAPLE 3 Hydropsychidae 49 1.588 MAPLE 3 Isopoda 2 0.065 MAPLE 3 Oligocheates 1 0.032 MAPLE 3 Perlidae 2 0.065 MAPLE 3 Perlodidae 8 0.259 MAPLE 3 Philipotamidae 9 0.292 MAPLE 3 Polycentropodidae 1 0.032 MAPLE 3 Pteronarcyidae 1 0.032 MAPLE 3 Chironomidae (Rheotanytarsus) 35 1.134 MAPLE 3 Simulidae 85 2.754 MAPLE 3 Tricorythidae 40 1 .296 MAPLE 4 Amphipoda 75 2.430 MAPLE 4 Brachycentridae 25 0.810 MAPLE 4 Chironomidae 270 8.748 MAPLE 4 Corixidae 1 0.032 MAPLE 4 Elmidae 17 0.55 1 MAPLE 4 Hydropsychidae 27 0.875 MAPLE 4 Hydroptilidae 1 0.032 MAPLE 4 Isopoda 18 0.583 MAPLE 4 Hirudinea 36 1.166 MAPLE 4 Oligocheates 5 0.162 MAPLE 4 Chironomidae (Rheotanytarsus) 27 0.875 MAPLE 4 Simulidae 3 0.097 MAPLE 4 Gastropoda 22 0.713 MAPLE 4 Tipulidae 1 0.032 MAPLE 5 Tipulidae (Antocha) 1 0.032 MAPLE 5 Baetidae 47 1 .523 MAPLE 5 Brachycentridae 8 0.259 117 APPENDIX C cont'd 118 Stream Site . MAPLE 5 Chironomidae 35 1 . 134 MAPLE 5 Elmidae 3 0.097 MAPLE 5 Heptageniidae 2 0.065 MAPLE 5 Hydropsychidae 3 5 1 .134 MAPLE 5 Isopoda 1 0.032 MAPLE 5 Oligocheates 1 0.032 MAPLE 5 Perlidae 1 0.032 MAPLE 5 Perlodidae 1 0.032 MAPLE 5 Polycentropodidae 1 0.032 MAPLE 5 Chironomidae (Rheotanytarsus) 3 0.097 MAPLE 5 Simulidae 104 3 .370 MAPLE 5 Gastropoda 1 0.032 MAPLE 5 Tricorythidae 1 0.032 MAPLE 6 Baetidae 5 0.162 MAPLE 6 Brachycentridae 3 0.097 MAPLE 6 Calopterygidae 1 0.032 MAPLE 6 Chironomidae 50 1 .620 MAPLE 6 Ephemeridae 4 0.130 MAPLE 6 Heptageniidae 4 0.130 MAPLE 6 Hydropsychidae 17 0.55 1 MAPLE 6 Isopoda 6 0.194 MAPLE 6 Limnephilidae 3 0.097 MAPLE 6 Perlodidae 2 0.065 MAPLE 6 Polycentropodidae 8 0.259 MAPLE 6 Pteronarcyidae 3 0.097 MAPLE 6 Simulidae 3 0.097 MAPLE 6 Gastropoda 1 0.032 MAPLE 6 Tricorythidae 2 0.065 MIDDLE 1 Amphipoda 22 0.713 MIDDLE 1 Tipulidae (Antocha) 3 0.097 MIDDLE 1 Baetidae 8 0.259 MIDDLE 1 Chironomidae 181 5.864 MIDDLE 1 Bivalvia 13 0.421 MIDDLE 1 Elmidae 8 0.259 MIDDLE 1 Hydropsychidae 8 0.259 MIDDLE 1 Psychomyiidae 7 0.227 MIDDLE 1 Tabanidae 2 0.065 MIDDLE 1 Tipulidae 1 0.032 MIDDLE l Tricorythidae 2 0.065 MIDDLE 2 Amphipoda 2 0.065 MIDDLE 2 Tipulidae (Antocha) 34 1 . 102 MIDDLE 2 Athericidae 1 0.032 MIDDLE 2 Baetidae 3 0.097 MIDDLE 2 Chironomidae 106 3.434 MIDDLE 2 Elmidae 7 0.227 MIDDLE 2 Ephemeridae 1 0.032 APPENDDI C cont'd uuumumuummuuuAAbaAhwuwuuuuwwwuwuwuwNNNNNNNNNN Taxa Glossosomatidae Heptageniidae Hydropsychidae Hirudinea Perlidae Philipotamidae Psychomyiidae Rhyacophilidae Simulidae Stratiomyidae Tricorythidae Amphipoda Tipulidae (Antocha) Athericidae Baetidae Brachycentridae Chironomidae Elmidae Heptageniidae Hydropsychidae Hydroptilidae Isonychidae Limnephilidae Perlodidae Psychomyiidae Simulidae Tricorythidae Amphipoda Athericidae Ceanidae Chironomidae Elmidae Tabanidae Amphipoda Tipulidae (Antocha) Athericidae Baetidae Brachycentridae Chironomidae Elmidae Glossosomatidae Heptageniidae Hydropsychidae Hydroptilidae Perlidae Psphenidae 119 275 3 5 1 2 2 1 1 .7 1 N \O APPENDIX C cont'd Stream Site MIDDLE 5 Psychomyiidae 1 0.032 MIDDLE 6 Tipulidae (Antocha) 7 0.227 MIDDLE 6 Baetidae 5 0. 162 MIDDLE 6 Ceanidae 2 0.065 MIDDLE 6 Chironomidae 49 1 .588 MIDDLE 6 Corydalidae 1 0.032 MIDDLE 6 Elmidae 12 0.389 MIDDLE 6 Heptageniidae 3 0.097 MIDDLE 6 Hydropsychidae 30 0.972 MIDDLE 6 Hydroptilidae 3 0.097 MIDDLE 6 Isonychidae 1 0.032 MIDDLE 6 Psphenidae 1 0.032 MIDDLE 6 Tabanidae 3 0.097 PRAIRIE 1 Baetidae 25 0.8 10 PRAIRIE 1 Chironomidae 70 2.268 PRAIRIE 1 Corydalidae 3 0.097 PRAIRIE 1 Elmidae 3 0.097 PRAIRIE 1 Ephemeridae 6 0. 194 PRAIRIE 1 Helicopsychidae 13 0.421 PRAIRIE 1 Heptageniidae 9 0.292 PRAIRIE 1 Hydropsychidae 1 0.032 PRAIRIE 1 Hydroptilidae 2 0.065 PRAIRIE 1 Limnephilidae 12 0.389 PRAIRIE 1 Perlidae 1 0.032 PRAIRIE l Polycentropodidae 1 0.032 PRAIRIE 2 Athericidae 2 0.065 PRAIRIE 2 Baetidae 28 0.907 PRAIRIE 2 Chironomidae 38 1.231 PRAIRIE 2 Elmidae 2 0.065 PRAIRIE 2 Ephemeridae 5 1 1 .652 PRAIRIE 2 Helicopsychidae 1 0.032 PRAIRIE 2 Heptageniidae 9 0.292 PRAIRIE 2 Hydropsychidae 193 6.253 PRAIRIE 2 Limnephilidae 5 0. 162 PRAIRIE 2 Perlidae 1 0.032 PRAIRIE 2 Philipotamidae 1 0.032 PRAIRIE 2 Simulidae 1 5 0.486 PRAIRIE 3 Tipulidae (Antocha) 1 0.032 PRAIRIE 3 Ceanidae 7 0.227 PRAIRIE 3 Chironomidae 83 2.689 PRAIRIE 3 Empididae 4 0.130 PRAIRIE 3 Heptageniidae 32 1 .03 7 PRAIRIE 3 Hydropsychidae 7 0.227 PRAIRIE 3 Limnephilidae 2 0.065 PRAIRIE 3 Tricorythidae 1 0.032 PRAIRIE 4 Tipulidae (Antocha) 3 0.097 120 APPENDIX C cont'd Stream Site Taxa Total No./ .M PRAIRIE 4 Athericidae 1 0.032 PRAIRIE 4 Baetidae 12 0.389 PRAIRIE 4 Chironomidae 82 2.657 PRAIRIE 4 Elmidae 36 1.166 PRAIRIE 4 Empididae 1 0.032 PRAIRIE 4 Ephemeridae 24 0.778 PRAIRIE 4 Heptageniidae 14 0.454 PRAIRIE 4 Hydropsychidae 416 13.478 PRAIRIE 4 Isonychidae 1 0.032 PRAIRIE 4 Perlidae 2 0.065 PRAIRIE 4 Simulidae 26 0.842 PRAIRIE 4 Siphlonouridae 1 0.032 PRAIRIE 4 Tipulidae 1 0.032 PRAIRIE 4 Tricorythidae 1 0.032 PRAIRIE 5 Tipulidae (Antocha) 12 0.389 PRAIRIE 5 Baetidae 8 0.259 PRAIRIE 5 Ceanidae 2 0.065 PRAIRIE 5 Chironomidae 50 1 .620 PRAIRIE 5 Corydalidae 1 0.032 PRAIRIE 5 Elmidae 8 0.259 PRAIRIE 5 Empididae 1 0.032 PRAIRIE 5 Ephemeridae 7 0.227 PRAIRIE 5 Gomphidae 1 0.032 PRAIRIE 5 Helicopsychidae 2 0.065 PRAIRIE 5 Heptageniidae 8 0.259 PRAIRIE 5 Hydropsychidae 94 3.046 PRAIRIE 5 Linmephilidae 3 0.097 PRAIRIE 5 Perlidae 2 0.065 PRAIRIE 5 Polycentropodidae 1 0.032 PRAIRIE 5 Siphlonouridae 2 0.065 PRAIRIE 6 Tipulidae (Antocha) 3 0.097 PRAIRIE 6 Baetidae 8 0.259 PRAIRIE 6 Chironomidae 24 0.778 PRAIRIE 6 Elmidae 1 0.032 PRAIRIE 6 Helicopsychidae 2 0.065 PRAIRIE 6 Heptageniidae 8 0.259 PRAIRIE 6 Hydropsychidae 7 0.227 PRAIRIE 6 Perlidae 1 0.032 SUGAR l Aeshnidae 1 0.032 SUGAR 1 Amphipoda 15 0.486 SUGAR 1 Baetidae 24 0.778 SUGAR 1 Chironomidae 43 1 .393 SUGAR 1 Elmidae 14 0.454 SUGAR 1 Empididae 1 0.032 SUGAR 1 Hydropsychidae 4 0.130 SUGAR 1 Hydroptilidae 1 0.032 121 APPENDIX C cont'd SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR SUGAR Stream MUM&&&-hbkhbb3AuwwwwwwwwMMWU-JWNNNNNNNNNNNNNNNNH Taxa Leptophlebiidae Simulidae Athericidae Brachycentridae Chironomidae Corydalidae Elmidae Glossosomatidae Heptageniidae Hydropsychidae Isopoda Leptophlebiidae Leuctridae Odontoceridae Oligocheates Perlidae Psychomyiidae Simulidae Athericidae Baetidae Brachycentridae Chironomidae Elmidae Ephemeridae Glossosomatidae Heptageniidae Hydropsychidae Hydroptilidae Odontoceridae Polycentropodidae Simulidae Tricorythidae Baetidae Chironomidae Elmidae Empididae Heptageniidae Hydropsychidae Hirudinea Perlidae Philipotamidae Pyralidae Simulidae Tipulidae (Antocha) Baetidae Chironomidae 122 61 12 70 APPENDIX C cont'd Stream Site Taxa SUGAR 5 Elmidae 15 SUGAR 5 Empididae 1 0.032 SUGAR 5 Ephemeridae 20 0.648 SUGAR 5 Heptageniidae 29 0.940 SUGAR 5 Hydropsychidae 103 3.337 SUGAR 5 Oligocheates 9 0.292 SUGAR 5 Philipotamidae 1 0.032 SUGAR 5 Psphenidae 1 0.032 SUGAR 5 Simulidae 3 0.097 SUGAR 5 Gastropoda 2 0.065 SUGAR 6 Baetidae 14 0.454 SUGAR 6 Chironomidae 27 0.875 SUGAR 6 Elmidae 1 0.032 SUGAR 6 Ephemeridae 3 0.097 SUGAR 6 Glossosomatidae 1 0.032 SUGAR 6 Helicopsychidae 5 0.162 SUGAR 6 Heptageniidae 2 0.065 SUGAR 6 Hydropsychidae 69 2.236 SUGAR 6 Limnephilidae 13 0.421 SUGAR 6 Perlidae 6 0.194 SUGAR 6 Psphenidae 1 0.032 SUGAR 6 Psychomyiidae 1 0.032 SUGAR 6 Simulidae 4 0.130 WHITE 1 Amphipoda 1 0.032 WHITE 1 Baetidae 32 1 .037 WHITE 1 Chironomidae 90 2.916 WHITE 1 Simulidae 23 0.745 WHITE 1 Tricorythidae 1 0.032 WHITE 2 Amphipoda 19 0.616 WHITE 2 Tipulidae (Antocha) 17 0.55 1 WHITE 2 Baetidae 4 0.130 WHITE 2 Brachycentridae 7 0.227 WHITE 2 Chironomidae 177 5.735 WHITE 2 Elmidae 31 1.004 WHITE 2 Empididae 4 0.130 WHITE 2 Ephemeridae~ 1 0.032 WHITE 2 Glossosomatidae 4 0.130 WHITE 2 Heptageniidae 1 0.032 WHITE 2 Tipulidae (Hexatoma) 1 0.032 WHITE 2 Hydropsychidae 19 0.616 WHITE 2 Hydroptilidae 2 0.065 WHITE 2 Simulidae 95 3 .078 WHITE 2 Tricorythidae 1 0.032 WHITE 3 Amphipoda 8 0.259 WHITE 3 Tipulidae (Antocha) 7 0.227 WHITE 3 Baetidae 4 0.130 123 APPENDIX C cont'd WHITE MMMMMMMMMA§¥¥$¥§A§¥AA##wawwwwu " Brahcycerm'le Taxa Chironomidae Elmidae Ephemeridae Heptageniidae Hydropsychidae Hydroptilidae Simulidae Gastropoda Amphipoda Chironomidae Elmidae Empididae Ephemeridae Heptageniidae Hydropsychidae Hirudinea Leptoceridae Odontoceridae Perlidae Philipotamidae Psychomyiidae Gastropoda Ameletidae Amphipoda Chironomidae Ephemeridae Heptageniidae Hydropsychidae Isopoda Perlidae Tricorythidae 124 ~ fl 0 O Nun‘s—ab.)— LITERATURE CITED 125 Literature Cited Allan, J .D. 1981. Determinants of diet of brook trout (Salvelinusfontinalis) in a mountain stream. Can. J. Fish. Aquat. Sci. 38:184-192. Allan, J .D. 1995. Stream ecology: Structure and function of nmning waters. Chapman and Hall, London. Baldwin, NS. 1956. Food consumption and grth of brook trout at different temperatures. Transactions of the American Fisheries Society 86:323-328. Boon, P.J., and Shiers, S.W.. 1976. Temperature studies on a river system in north-east England. Freshwater Biology 6:23-32. Brooker, MP. 1981. The impact of impoundments on the downstream fisheries and general ecology of rivers. Advances in Applied Biology 6291-152. Carlander, K.D. 1969. Handbook of freshwater fishery biology, volume 1. The Iowa State University Press, Ames, Iowa. Cooper, E.L., and Scherer, RC. 1967. Annual production of brook trout (Salvelinus fontinalis) in fertile and infertile streams of Pennsylvania. Proceedings of the Pennsylvania Academy of Science 41:65-70. Crisp, D.T. 1987. Thermal “resetting” of streams by reservoir releases with special reference to effects on salmonid fishes. In Regulated Streams Advances in Ecology. Edited by J .F . Craig and J .B. Kemper. Plenum Press, New York pp. 1 63-1 82. Cummins, K.W. 1977. From headwater streams to rivers. The American Biology Teacher. May:305-312. Cummins, K.W. 1979. The natural stream ecosystem. In The ecology of regulated streams. Edited by J .V. Ward and J .A. Stanford. Plenum, New York. pp.7-24. Cummins, K.W., and Wilzbach, M.A.. 1985. Field procedures for analysis of functional feeding groups of stream macroinvertebrates. Field Manual. Pymantuning Laboratory of Ecology, University of Pittsburg. Linesville, PA. 16424. Cummins, K.W., and Merritt, R.W.. 1996. Ecology and distribution of aquatic insects. In An introduction to the aquatic insects of North America, 2nd edn. Edited by R.W. Merritt and K.W. Cummins Kendall/Hunt, Iowa pp.74-86. Deegan, L.A., Golden, H.E., Harvey, C.J., and Peterson, BJ. 1999. Influence of environmental variability on the growth of age-0 and adult arctic grayling. Transactions of the American Fisheries Society 128:1163-1175. 126 Diana, IS 1995. Biology and ecology of fishes. Biological Sciences Press, Cooper Publishing Group, Carmel, IN. Dodd, HR. 1999. The effects of low-head lamprey barrier dams on stream habitat and -’/ fish communities in tributaries of the Great Lakes. Master’s Thesis, Michigan State University, East Lansing, MI. USA Eddy, S. 1969. The Freshwater Fishes. WM. C. Brown Company Publishers, Dubuque, Iowa. Edwards, R.W., Densem, J .W., and Russell, RA. 1979. An assessment of the importance of temperature as a factor controlling the growth rate of brown trout in streams. J. Anim. Ecol. 48:501-507. Elliot, J .M. 1994. Quantitative Ecology and the Brown Trout. Oxford University Press, New York. Fraley, J .J . 1979. Effects of elevated stream temperatures below a shallow reservoir on coldwater macroinvertebrate fauna. In The ecology of regulated streams. Edited by J .V. Ward and J .A. Stanford. Plenum Press, New York. pp.257-272. Francis, R.I.C.C. 1990. Back-calculation of fish length: a critical review. J. Fish Biol. 36:883-902. Gibbons, J .W. 1976. Thermal alteration and the enhancement of species populations. In Thermal ecology 11. Edited by G.W. Esch and R.W. McFarlane. ERDA Symposium Series (CONF-750425) pp.27-31. Giller, PS, and Malmqvist, B. 1998. The Biology of Streams and Rivers. Oxford Press, New York. Biology of Habitats No.4. Go/rmgp. T., and Karr, J. R. 1978. Habitat structure and stream fish communities. Ecology 59(3): 507- 515. -e Gowing, H., and Alexander, GR. 1980. Population dynamics of trout in some streams of the northern lower peninsula of Michigan. Fisheries Research Report No. 1 877. Hauer, F.R., Stanford, J .A., Giersch, J .J ., and Lowe, W.H. 2000. Distribution and abundance patterns of macroinvertebrates in a mountain stream: an analysis along multiple environmental gradients. Verh. Intemat. Verein. Limnol. (In Press). Hawkins, C.P., Murphy, M.L., Anderson, NH, and Wilzbach, M..A. 1983. Density of fish and salamanders in relation to riparian canopy and physical habitat in streams of the northwestern United States. Can. J. Fish. Aquat. Sci. 40:1173-1185. 127 Hayes, D. B. 1998. Issues affecting fish habitat in the Great Lakes Basin. In Great Lakes fishery policy and management: a binational perspective. Edited by W. W. Taylor and C. P. Ferreri. in press. Hayes, D.B., Taylor, Drake, W.W., M.T., Marod, S.M., and Whelan, GE. 1998. The value of headwaters to brook trout (Salvelinus fontinalis) in the Ford River. In Headwaters: Water resources and soil conservation; Proceedings of Headwater ‘98, the fourth international conference on headwater control, Merano, Italy. Edited by M.J. Haigh, J. Krecek, G.S. Rajwar and MP. Kilmartin. Oxford and IBH Publishing Co., New Delhi, pp.175-185. Herlong, DD, and Mallin, M.A. 1985. The benthos-plankton relationship upstream and downstream of a blackwater impoundment. Journal of Freshwater Ecology 3(1):47-59. Hinz, L.C. Jr., and M.J. Wiley. 1998. Growth and production of juvenile trout in Michigan streams: influence of potential ration and temperature. Fisheries Division Research Report No. 2042. Hocking, RR. 1976. The analysis and selection of variables in linear regression. Biometrics 32:1-50. Holden, PB. 1979. Ecology of riverine fishes in regulated stream systems with emphasis on the Colorado River. In The ecology of regulated streams. Edited by J .V. Ward and J .A. Stanford. Plenum, New York pp.57-74. Hubert, W.A., Douglas, DD, and Rhodes, HA. 1993. Variation in the summer diet of age-0 brown trout in a regulated mountain stream. Hydrobiologia 259:179-185. Jacobsen, D., Schultz, R., and Encalada, A. 1997. Structure and diversity of stream invertebrate assemblages: the influence of temperature with altitude and latitude. Freshwater Biology 38:247-261. Jensen, A.J. 1987. Hydropower development of salmon rivers: Effect of changes in water temperature on growth of brown trout (Salmo trutta) presmolts. In Regulated Streams Advances in Ecology. Edited by J.F. Craig and J .B. Kemper. Plenum Press, New York pp.207-218. Klomp, K.D. 1998. An initial evaluation of the habitat and fisheries resources associated with a Wemoval in a Michigan coldwater stream. Master’s Thesis, Michigan State University, East Lansing, MI. USA. Kondolf, G. M., and Li, S. 1992. The pebble count technique for quantifying surface bed material size in instream flow studies. Rivers 3:80-87. 128 Kwak, T.J., and Waters, T.F. 1997. Trout production dynamics and water quality in Minnesota streams. Transactions of the American Fisheries Society 126:35-48. Littell, R.C., Milliken, G.A., Stroup, W.W., and Wolfinger, RD. 1996. SAS® system for mixed models. SAS Institute Inc. Cary, NC. Lyons, J. 1996. Patterns in the species composition of fish assemblages among f,” Wisconsin streams. Environmental Biology of Fishes 45:329-341. Marod, SM. 1995. The influence of temperature and discharge on movement patterns of brook trout, Salvelinusfontinalis, in the Ford River, Dickinson County, Michigan. Master’s Thesis, Michigan State University, East Lansing, MI. USA. Merritt, R.W., Resh, V.H., and Cummins, K.W. 1996. Design of aquatic insect studies: Collecting, sampling and rearing procedures. In An introduction to the aquatic insects of North America, 2nd edn. Edited by R.W. Merritt and K.W. Cummins Kendall/Hunt, Iowa. pp. 12-28. Minard, RE, and Dye, J.B. 1997. Rainbow trout sampling and ageing protocol. Alaska Department of Fish and Game, Special Publication No.98-2. Mistak, J .L. 2000. Dam removal effects on fisheries resources, habitat, and summer diet of trout in the Pine River, Manistee county, Michigan. Master’s Thesis. Michigan State University, East Lansing, MI. USA. Modde, T., Drewes, H.G., and Rumble, M.A. 1986. Effects of watershed alteration on the brook trout population of a small black hills stream. Great Basin Naturalist 46(1):39-45. Newman, KR. 1999. Effects of changing land cover and human development on the fish and hydrology of the Huron River watershed. Ph.D. Dissertation. Michigan State University, East Lansing, MI. USA. Rader, RB. 1997. A functional classification of the drift: traits that influence invertebrate availability to salmonids. Can. J. Fish. Aquat. Sci. 54:1211-1234. Richardson, J .S. and Mackay, R.J. 1991. Lake outlets and the distribution of filter feeders: an assessment of hypotheses. Oikos 62:370-380. Rosenberg, D. M., and Resh, V.H. 1996. Use of aquatic insects in biomonitoring. In An introduction to the aquatic insects of North America, 2nd edn. Edited by R.W. Merritt and K.W. Cummins Kendall/Hunt, Iowa pp.87-97. Ryder, RA. 1982. The morphoedaphic index-use, abuse, and fundamental concepts. Transactions of the American Fisheries Society 111:154-164. 129 Ryder, R.A., Kerr, S.R., Loftus, K.H., and Regier, HA. 1974. The morphoedaphic index, a fish yield estimator-review and evaluation. J .F ish. Res. Board Can. 31:663-688. Scamecchia, D.L., and Bergersen, ER 1987. Trout production and standing crop in Colorado’s small streams, as related to environmental features. North American Journal of Fisheries Management 7(3):315-330. Scott, W.B., and Crossman, EJ. 1973. Freshwater fishes of Canada. Bull. Fish. Res. Board Can. No. 184. Sheldon, AL. 1968. Species diversity and longitudinal succession in stream fishes. Ecology 49:193-198. Smith, G. 1989. Guide to Nongame Fishes of Michigan. Museum of Zoology, Ann Arbor, MI. S¢renson, T. 1948. A method of establishing groups of equal amplitude in plant sociology based on similarity of species content. Kong. Dan. Vidensk. Selak. Biol. Skr. 5:1-34. Stanford, J .A., and Ward, J .V. 1979. Darnmed rivers of the world: Symposium rationale. “In The ecology of regulated streams. Edited by J .V. Ward and J .A. Stanford. Plenum, New York pp.1-6. Taniguchi, Y., Rahel, F. J ., Novinger, D. C., and Gerow, K.G. 1998. Temperature mediation of competitive interactions among three fish species that replace each other along longitudinal stream gradients. Can. J. Fish. Aquat. Sci. 55:1894-1901. Valderrama, J .C. 1981. The simultaneous analysis of total nitrogen and total phosphorous in natural waters. Marine Chemistry 10:109-122. Van Deventer, J .S., and Platts, W.S. 1983. Sampling and estimating fish populations from streams. Transactions of the North American Wildlife and Natural Resources Conference 48:349-354. Van Deventer, J .S., and Platts, W.S. 1985. A computer software system for entering, managing, and analyzing fish capture data from streams. US. Department of Agriculture, Forest Service. Intermountain Forest and Range Exp. Sta. Research Note INT -3 52. Ogden, UT. jamefiRL, Minshall, G.W., Cummins, K.W., Sedell, J.R., and Cushing, CE. 1980. e river continuum concept. Can. J. Fish. Aquat. Sci. 37 :130-1 37. Van Snik, GE, and Stauffer, J .R. 1999. Comparative microhabitat use of ecologically similar benthic fishes. Environmental Biology of Fishes 56(4):443-453. 130 Vogel, D. A., Marine, K. R., and Smith, J. G. 1988. Fish passage action program for Red Bluff Diversion Darn. U. S. Fish and Wildlife Service Report FRI/F AO-88-19. Wallace, J .B., and Merritt, R.W. 1980. F ilter-feeding ecology of aquatic insects. Annual Review of Entomology 25:103-32. Wallace, J.B., and Grubaugh, J .W. 1996. Transport and storage of FPOM. In Methods in stream ecology. Edited by F .R. Hauer and GA. Lamberti. Academic Press, San Diego, CA. pp. 191-216. Ward, J .V. 1992. Aquatic Insect Ecology, 1. Biology and Habitat. John Wiley and Sons Inc. No.1. Ward, J .V., and Stanford, J.A. 1979. Ecological factors controlling stream zoobenthos with emphasis on thermal modification of regulated streams. In The ecology of regulated streams. Edited by J .V. Ward and J .A. Stanford. Plenum Press, New York pp35-56. Ward, J .V., and Stanford, J .A. 1983a. The serial discontinuity concept of lotic ecosystems. In Dynamics of Lotic Ecosystems. Edited by TD. Fontaine and SM. Bartell Ann Arbor Science, Ann Arbor, MI. pp.29-42. Ward, J .V., and Stanford, J.A. l983b. Tire intermediate-disturbance hypothesis: An explanation for biotic diversity patterns in lotic ecosystems. In Dynamics of Lotic Ecosystems. Edited by TD. Fontaine and SM. Bartell. Ann Arbor Science, Ann Arbor, MI. pp.347-356. Ward, J .V., and Stanford, J .A. 1987. The ecology of regulated streams: Past accomplishments and directions for future research. In Regulated Streams Advances in Ecology. Edited by J .F. Craig and J .B. Kemper. Plenum Press, New York pp.391-409. Waters, TE. 1995. Sediment in streams: Sources, biological effects and control. American Fisheries Society, Bethesda, Maryland. Webb, B.W., and Walling, DE. 1993. Temporal variability in the impact of river regulation on thermal regime and some biological implications. Freshwater Biology 29:167-182. Weisberg, S. 1993. Using hard-part increment data to estimate age and environmental effects. Can. J. Fish. Aquat. Sci. 50:1229-1237. Weisberg, S., and Frie, R.V. 1987. Linear models for the growth of fish. Pages 127-143 in RC. Summerfelt and GE. Hall (Eds) The Age and Growth of Fish. Iowa State University Press, Ames,Iiowa. 131 Wetzel, KG. 1975. Lirnnology. Saunders College Publishing, Philadelphia, PA. Wehrly, K.E., Wiley, M.J., and Seelbach, P.W. 1998. A thermal habitat classification for «’ lower Michigan rivers. Fisheries Research Report No.2038. 132 "Ililllllllillilillf