2013 NORTHEASTERN NATURALIST 20(1):19–36 Brook Trout (Salvelinus fontinalis) Habitat Use and Dispersal Patterns in New York Adirondack Mountain Headwater Streams Justin Ecret1,* and Timothy B. Mihuc2 Abstract - Minimal research has been conducted involving Salvelinus fontinalis (Brook Trout) habitat use and dispersal patterns within Adirondack Mountain headwater streams. Hence, fishery managers are left with information gaps regarding the specific habitat conditions characteristic of sustainable Brook Trout populations in Adirondack flowing waters. Through the use of single-pass electrofishing and markrecapture techniques, size-class specific microhabitat use and reach-scale movement patterns for Brook Trout were examined within two northern Adirondack streams. Water depth, water velocity, and substrate-size use were observed to be similar among two Brook Trout size classes. Both size classes exhibited use patterns within deeper slowermoving pool habitats; however, larger Brook Trout were found to be associated with smaller-sized substrates within one of our study sites. These habitat-use patterns were also supported by comparison of stream hydrologic condition, including Froude number. Brook Trout movement patterns were found to be dependent on both size class and season. Smaller-sized trout exhibited increased movement during the spring, whereas larger trout were found to be more mobile and move more frequently during early fall. Lastly, we examined the proportion of Brook Trout moving upstream/downstream and found a greater frequency of smaller Brook Trout moving upstream during late summer. Introduction Of the factors that influence the presence, abundance, and distribution of biota in the environment, habitat conditions are known to carry the greatest influence (Reyjol et al. 2001). Fish habitat use is an important component of their biology, especially for stream-resident (non-diadromous) fish, which endure extensive hydrologic variation in stream channels. As a result of deficiencies in literature and scientific research involving Salvelinus fontinalis (Mitchill) (Brook Trout ) in Adirondack streams, it is essential to classify Brook Trout habitat use within these systems. Studies have observed habitat shifts among varying age-classes of streamresident salmonids, and have suggested that younger individuals typically occupy shallow, slow-flowing stream habitats, while the larger individuals reside at greater depths (Mäki-Petäys et al. 1997). Under natural stream conditions, Brook Trout have also been found to predominantly use pool habitats (Hansbarger 2005). The long-term debate over the movement patterns associated with streamresident fish has been closely examined since Gerking’s (1959) seminal research. The early concept of restricted movement was later termed the restricted movement paradigm (RMP) (Gowan et al. 1994). According to the 1US Fish and Wildlife Service Region 5, 3817 Luker Road, Cortland, NY 13045. 2Department of Earth and Environmental Science, State University of New York, 101 Broad Street, Plattsburgh, NY 12901. *Corresponding author - Justin_Ecret@fws.gov. 20 Northeastern Naturalist Vol. 20, No. 1 RMP, stream-resident fish are primarily sedentary, or immobile. As a result, these fish exhibit minimal movement and reside in natal pools. Consequently, these fish remain within, or in proximity to, their initial capture site during mark-recapture and movement studies. In subsequent research, the RMP has been challenged, as a result of findings proposing that stream fish populations are not exclusively homogenous in their movement patterns, but rather are comprised of both sedentary and mobile individuals (Penzcak 2006). Subsequent studies have also provided evidence that individual stream-resident salmonids exhibit dual movement behavior and “switch” between sedentary and mobile movements between seasons (Hilderbrand and Kershner 2000, Morrissey and Ferguson 2011, and Smithson and Johnston 1998). Long-range movement patterns exhibited by salmonids have been well documented (Hilderbrand and Kershener 2000, Peterson and Fausch 2003, Sorensen et al. 1995). Furthermore, riverine fish species populations have been suggested to be comprised of both static (i.e., sedentary) and highly mobile individuals (Penczak 2006). Salmonids have been closely studied by federal and state agencies, as well as among local anglers, as they are of great local conservation importance within the Adirondacks (EBTJV 2006). Recent assessments of Brook Trout populations have been conducted in response to the growing concern of the above-mentioned stakeholders with the rapid decline in native Brook Trout populations. The distribution of self-sustaining Brook Trout populations has been greatly reduced from its natal range, once extending from the Appalachian Mountains to the Carolinas and as far north as Atlantic Canada and west to the Great Lakes regions (Hudy et al. 2008). Extensive research and monitoring has illustrated the physical, chemical, and biological watershed-level changes over the past 200 years and has demonstrated that Brook Trout respond negatively to such changes across the eastern United States (Hudy et al. 2008). Furthermore, the spread of non-native coldwater fishes, as well as habitat fragmentation via dams, road crossings, and extensive channelization, have all contributed to the decline in self-sustaining Brook Trout populations (Hudy et al. 2008). Understanding the habitat use and movement patterns associated with Brook Trout , as well as other stream-resident fish, is crucial in order to provide future protection and commence restoration. Species-specific and age-class movement data could assist in land-management policies that promote the enhancement of threatened Brook Trout populations. Methods Field-site description Both study streams are within the True Brook watershed in Clinton County, which is located in the Adirondack Park in northern New York. The larger of the two streams, Fall Brook is a total of 24.5 km and drains a catchment area of approximately 10.8 km2, while the smaller un-named stream (TB10) is a total of 5.03 km in length and drains a catchment area of approximately 1.9 km2. The study site (44°39.21'N, 73°49.5'W) was selected on the basis of preliminary examination of potential fish passage impediments by various crossing structures 2013 J. Ecret and T.B. Mihuc 21 within 9 sub-watersheds (including True Brook) in Essex and Clinton counties (Mihuc et al. 2008). Both TB10 and Fall Brook contain a minimum of one crossing structure, all of which were assessed and evaluated and showed varying rates of fish passage (Mihuc et al. 2008). In addition, Salmo trutta L. (Brown Trout ) were observed within the lower reaches of both streams during the sampling portion of this study. The local landscape mostly consists of heavily forested areas that are moderately fragmented by rural residences (Fig. 1). Study design Fish were captured using a DC backpack electrofisher (300–400 volts, 30-cm anode) using a single-pass technique and were individually tagged with a small (1-mm) individually numbered aluminum tag (National Band and Tag Co., Newport, KY) within the two major tributaries. The smaller of the two tributaries, TB10, was sampled in its entirety, and consisted of a total study reach of approximately 1600 m extending from the mainstream of True Brook to its associated headwater. Contrastingly, the larger tributary, Fall Brook, was only sampled from the mainstream of True Brook approximately 900 m upstream, due to restricted stream access. The first fish-tagging event was conducted in June 2010, followed by three subsequent sampling sessions and fish-tagging events during July, August, and October. Selection for sampling sessions was not based on any specific criteria; however, we did expect fish movements to be greatest during early June and July as a result of elevated stream discharge (i.e., snowmelt runoff) and again during October during spawning (Peterson and Fausch 2003). Following Bond and Lake (2003), an anchored float was placed within the stream indicating the fish’s initial location in order to record habitat variables. Throughout all four sampling events, habitat variables were measured and recorded, and fish were identified and tagged within the peduncle region, just anterior to the lower portion of the caudal fin, prior to release at point of capture (Fig. 2). Trout total length (cm) and mass (g) were also recorded, enabling us to provide size-class-specific microhabitat assessments for Brook Trout. Again, following Bond and Lake (2003), fish sampling-point microhabitats (e.g., stream depth [cm], water velocity [m/s], and substrate size [cm]), from the initially marked bobber and sinker, were measured and recorded within an approximate 1-m quadrat around the bobber and sinker. Stream depth was measured from streambed bottom to water surface; substrate size was measured by recording the longest axes of a randomly selected substrate within the quadrat. Additionally, two supplementary depth and velocity measurements were taken within the quadrat. All velocity measurements were recorded at approximately 0.4 x [depth], measured upward from the streambed (Gore 2006). During the first sampling session, researchers delineated and marked with flags 25-m sub-reaches within both streams. The reach location (distance [m] from the main confluence of True Brook) where each tagged trout was captured during each sampling event was recorded. Sub-reach delineation followed that of earlier research conducted on dispersal patterns of stream-dwelling salmonids (Hutchings and Gerber 2002, Morrissey and Ferguson 2011, Wilson et al. 2004). 22 Northeastern Naturalist Vol. 20, No. 1 Figure 2. Photograph depicting placement of individually numbered aluminum tags, placed within the penducle region of successfully captured and tagged Brook Trout. Figure 1. True Brook (Clinton County, NY). The two study streams are Fall Brook and TB10. Each study reach began at the main confluence of the study stream with True Brook and extended a minimum of 0.9 km upstream. 2013 J. Ecret and T.B. Mihuc 23 Three subsequent sampling events allowed us to collect both microhabitat and seasonal movement data for Brook Trout in these streams using point abundance estimates. This method permits both microhabitat variables and fish movement patterns to be accounted for within the system. In order to compare trout habitat variables to overall stream habitat, we also recorded randompoint sampling of each of the microhabitat variables previously mentioned throughout the entire study stream during all sampling events. There were approximately 10 random points to record habitat variables per 25-m sub-reach; these points were randomly selected simply by identifying stream habitats where trout were not observed or captured. In order to determine size-class-specific variation among fish microhabitat use, we used a Mann-Whitney U test to compare microhabitat use between two fish size classes to habitat availability in two headwater streams (SPSS v.16.0) (P = 0.05). We compared the medians of habitat variables among both size classes of Brook Trout, as well as random variables using a non-parametric test. Stream hydraulic habitat conditions (e.g., Froude number) were also compared using the same analysis techniques among fish size classes and random-point sampling. All variables were tested for correlation using Pearson’s correlation (2 -tailed), and positive correlations existed between random depth, Froude number, and substrate size (P less than 0.001). Random velocity did not correlate with any of the other random variables or any of the measured fish variables in either fish size class (SC1, SC2). Hydraulic habitat assessments facilitate proficient habitat representation in that they incorporate multiple habitat use. Froude number (v/[dg]0.5) is a dimensionless value that characterizes open channel flow based on a ratio of velocity (v) to hydraulic depth (d) and the acceleration due to gravity (g = 9.81m/s2). A Froude number greater than 1.0 represents fast rapid flow conditions (supercritical flow), whereas a Froude number less than 1.0 represents slow or tranquil hydraulic conditions (subcritical flow). For the purpose of this study, we substituted stream depth for hydraulic depth in order to calculate Froude number for both fish and random sampling points. Additionally the proportion of Brook Trout found in pool and riffle habitats was also determined. Jowett (1993) classifies a Froude number less than 0.18 as pools and greater than 0.41 as riffles. This study used a modification of Jowett (1993) classification of pool and riffles, in that pools were determined by a Froude number less than 0.18 and riffles by a Froude number greater than 0.18 due to the low sample size of riffle features. The location within the reach (number of meters from stream outlet) of each successfully recaptured Brook Trout was recorded in order to determine approximate distance and direction moved between sampling events. Habitat use and movement data was only collected for Brook Trout; however, total length and weight for successfully captured Brown Trout was also recorded. Results During each sampling event, stream and fish sampling consistently began at TB10 and continued until the stream was sampled in its entirety (approximately 24 Northeastern Naturalist Vol. 20, No. 1 1600-m study reach), followed by the same sampling protocols at Fall Brook (approximately 900 m). The entire research study occurred over a total of 28 sampling days (18 and 10 days at TB10 and Fall Brook, respectively). The mean and median water depth, velocity, and substrate size was calculated from all fish, random, and supplementary locations for both TB10 and Fall Brook for each sampling session. Overall, there was a greater number of successfully captured Brook Trout at TB10 (n = 736) compared to Fall Brook (n = 273), both sites combined constituted a total of 173 recaptured marked Brook Trout upon study completion within the two major tributaries (Table 1). Overall, there was consistently a greater proportion of Brook Trout captured and tagged among the smaller size class within both study streams. However, there was a greater number of larger trout Table 1. Capture and tagging data per sampling event for Brook Trout size class within TB10 (top) and Fall Brook (bottom). Brook Trout length is expressed as the mean ± SD. SC1 SC2 Total TB10 June NCaptured 239 84 323 NTagged 211 80 291 NRecapture 0 0 0 Mean length (cm) 12.1 ± 1.3 17.1 ± 3.2 July NCaptured 144 64 208 NTagged 23 7 30 NRecapture 17 24 41 Mean length (cm) 12.4 ± 1.1 17.4 ± 4.2 August NCaptured 94 35 129 NTagged 9 5 14 NRecapture 37 22 59 Mean length (cm) 12.4 ± 1.0 17.0 ± 4.0 October NCaptured 60 16 76 NTagged 0 0 0 NRecapture 12 15 27 Mean length (cm) 10.0 ± 2.3 16.7 ± 2.3 Fall Brook June NCaptured 50 25 75 NTagged 49 25 74 NRecapture 0 0 0 Mean length (cm) 12.4 ± 1.4 17.0 ± 2.3 July NCaptured 51 38 89 NTagged 21 35 56 NRecapture 6 9 15 Mean length (cm) 12.5 ± 1.1 15.5 ± 2.5 August NCaptured 29 21 50 NTagged 4 20 24 NRecapture 13 14 27 Mean length (cm) 13.0 ± 0.7 17.7 ± 2.9 October NCaptured 46 13 59 NTagged 0 0 0 NRecapture 3 1 4 Mean length (cm) 9.6 ± 3.2 18.7 ± 7.1 2013 J. Ecret and T.B. Mihuc 25 recaptured compared to smaller trout as the study progressed (Table 1). There was a total of 43 Brook Trout that exhibited tag loss with both study streams, and smaller-size Brook Trout had a higher percentage of tag loss (SC1 = 69.8%, SC2 = 30.2%) within both study streams. However, tagged Brook Trout did not exhibit any noticeable impairment of swimming abilities, nor was there any immediate mortality observed. Other observed fish species consisted of Brown Trout , Cottus cognatus Richardson (Slimy Sculpin), Rhinichthys atratulus (Hermann) (Blacknose Dace), and Catostomus commersoni (Lacepède) (White Sucker). Although Brown Trout were observed within both study streams, habitat and movement data were not collected. The frequency and size of both Brook and Brown Trout was collected and showed that theses streams were predominantly inhabited by Brook Trout among the 13.5– 14.5-cm size range, with a fewer number of Brown Trout (Table 2). Habitat data In order to evaluate Brook Trout microhabitat use, all measured Brook Trout were placed into two major size classes (SC), where SC1 and SC2 represented Brook Trout less than and greater than 14.5 cm in total length, respectively. Length-frequency data were compiled for both study sites; the smaller stream, TB10, was composed of predominately Brook Trout within the 11–12 cm length range, while Fall Brook consisted mostly of Brook Trout in the 13–14 cm range. The median seasonal water depth, velocity, and substrate size were compared between Brook Trout size classes, as well as between random sampling points, in order to determine fish size-class-specific microhabitat use. Overall, water depth appeared to be the most definitive microhabitat variable influencing Brook Trout of both size classes within both study streams. A greater proportion of Brook Trout among both size classes was found in pool habitats (Fr < 0.18), as compared to riffle habitats (Fr > 0.18) in both TB10 and Fall Brook (Fig. 3). Microhabitat use With the exception of median substrate size, microhabitat use did not significantly differ between the two Brook Trout size classes in TB10 (Table 3). The median substrate size associated with Brook Trout of SC1 (15.0 cm) was significantly larger than the median substrate size of SC2 (12.0 cm) (Table 3, Fig. 4B). Both the median water depth and water velocity use did not significantly differ between SC1 and SC2 (Table 3; Fig. 4A, C). However, both size classes greatly differed for multiple microhabitat use in relation to overall habitat availability (random points). Median water depth for SC1 (20.0 cm) and SC2 (21.3 cm) were Table 2. Mean length ± SD for brook and Brown Trout in TB10 and Fall Brook. TB10 Fall Brook Species n Length (cm) n Length (cm) Brook Trout 736 13.4 ± 2.5 273 14.4 ± 3.3 Brown Trout 8 13.4 ± 2.6 1 19.5 26 Northeastern Naturalist Vol. 20, No. 1 significantly greater than the median water depth at random stream locations (14.0 cm) (Table 3, Fig. 4A). Additionally, both size classes exhibited different microhabitat use within regions of stream with lower water velocities, where SC1 (0.10 m/s), and SC2 (0.09 m/s) were more prevalent in stream habitats with significantly lower water velocities in relation to their habitat availability (0.24 m/s) (Table 3, Fig. 4C). Lastly, the median substrate size of SC2 (10.0 cm) within TB10 was significantly smaller than the median substrate size at random stream locations (16.0 cm) (Table 3, Fig. 4B). The only microhabitat variable that significantly differed between Brook Trout size classes within Fall Brook was median water depth, where the median water depth for SC2 (26.0 cm) was significantly deeper compared to the median depths of SC1 (22.0 cm) (Table 4, Fig. 5A). There was no statistical difference in median water velocity and median substrate-size use between size classes within Fall Brook (Table 4; Fig. 5B, C). Similar to TB10, both size classes differed in their median water-depth use in relation to their overall habitat availability. The median depth for SC1 (22.0 cm) and SC2 (26.0 cm) were both significantly deeper compared to median random depth locations (18.0 cm) (Table 4, Fig. 5A). There were also statistical differences between Figure 3. The number of Brook Trout that were found in pools and riffles at (A) TB10 and (B) Fall Brook throughout the study duration. Pools and riffle classifications derivation after Jowett (1993). Table 3. Mann Whitney U test for the effects of Brook Trout size-class on the habitat use among three microhabitat characteristics at study site TB10. Ns = not significant (P > .05). Size class Depth Velocity Sub Froude No SC1 vs. SC2 ns ns 0.01* ns SC1 vs. Random <0.001** <0.001** 0.05* <0.001** SC2 vs. Random <0.001** <0.001** <0.001** <0.001** 2013 J. Ecret and T.B. Mihuc 27 Figure 4. The median water depth (A), substrate size (B), water velocity (C), , and Froude number (D) at TB10. Error bars display standard error. size classes and random points within Fall Brook in terms of median water velocity, where the median water velocity for SC1 (0.20 m/s) and SC2 (0.16 m/s) were significantly lower than the median water velocity at random points (0.30 m/s) (Table 4, Fig. 5C). There was no statistical difference in regards to median substrate size within study site Fall Brook (Table 4, Fig. 5B). Table 4. Mann Whitney U test for the effects of Brook Trout size-class on the habitat use among three microhabitat characteristics at study site Fall Brook. Ns = not significant (P > .05). Size class Depth Velocity Sub Froude No SC1 vs. SC2 <0.001** ns ns <0.05* SC1 vs. Random <0.001** <0.001** ns <0.001** SC2 vs. Random <0.001** <0.001** ns <0.001** 28 Northeastern Naturalist Vol. 20, No. 1 Hydrologic habitat classification Brook Trout habitat was also evaluated among Brook Trout size classes and random points by comparing the median Froude number. Both study sites showed similar trends, such that Brook Trout among both size classes were observed in stream conditions with lower than median Froude number as compared to that of random stream locations (Tables 3, 4; Figs. 4D, 5D). There was no statistical difference in mean Froude number between Brook Trout size classes; however, microhabitat use for both size classes showed significant differences in comparison to random stream habitats. The median Froude number for SC1 (0.08) was significantly lower than the median Froude number (0.21) associated with random stream locations. Additionally, the median Froude number of SC2 (0.12) was significantly lower than the median Froude number at random locations (0.23) (Table 3, Fig. 4D). There were greater differences between Brook Trout size classes in Fall Brook as compared to those of TB10. Specifically, Brook Trout of SC1 were observed more often in stream habitats with higher median Froude number as compared Figure 5. The median water depth (A), substrate size (B), water velocity (C), and Froude number (D) at Fall Creek. Error bars display standard error. 2013 J. Ecret and T.B. Mihuc 29 to SC2. The median Froude number for SC1 (0.12) showed no statistical significance when compared to the median Froude number for SC2 (0.12) (Table 4, Fig. 5D). Finally, there was no statistical distinction between the median Froude number of Brook Trout size classes and random stream locations within Fall Brook (Table 4, Fig. 5D). Fish Movement It was assumed that fish immigration derived exclusively from the mainstem of True Brook in TB10 due to it being sampled in its entirety (1600 m), whereas fish immigration could derive from both the mainstream of True Brook as well as upstream reaches in Fall Brook because Fall Brook was not sampled to its headwater. The average distance moved, directional movement patterns (i.e., upstream or downstream), and percent of movement were determined for both size classes within both TB10 and Fall Brook. There was a greater number of recaptured Brook Trout in TB10; however, both sites were composed of both sedentary and mobile individuals. The majority of the recaptured Brook Trout made short distance movements (<100 m) in both TB10 and Fall Brook, while 28% and 24% of recaptured Brook Trout, in TB10 and Fall Creek respectively, exhibited no movement and remained at their initial point of capture throughout the duration of the study (Table 5). SC1 (n = 72) in TB10 and SC1 (n = 25) in Fall Creek moved a greater mean distance during July compared to SC2; furthermore, SC2 (n = 51) in TB10 and SC2 (n = 23) in Fall Creek moved a greater distance during August and October compared to SC1 (Fig. 6A, B) However, both size classes in TB10 exhibited the most movement during October (Fig. 6A). In Fall Brook, the longest average distance moved by Brook Trout was observed by SC1 in July (Fig. 6B). In addition, smaller sized Brook Trout within both study streams exhibited a greater mean distance moved upstream during July, and a greater mean distance moved downstream during August and October (Fig. 7). Both TB10 and Fall Brook were composed of sedentary and mobile Brook Trout. Overall, the majority of Brook Trout in both streams showed small-scale movement patterns and remained close to or at their initial capture location. However, individual Brook Trout did move distances greater than 600 m. In both study streams, SC1 Brook Trout displayed increased movement patterns (>200 m) compared to SC2 Brook Trout both upstream and downstream (Fig. 7). Despite moving distances greater than 100 m, a greater proportion of SC2 Brook Trout displayed shorter movement patterns (<100 m) (Table 5) compared to SC1 Table 5. Percent of movement for Brook Trout size classes at TB10 and Fall Brook. (Percents are based on the number of recaptured trout upon study completion.) SC1 SC2 % short % long % short % long % no movement movement % no movement movement Stream movement (<100 m) (>100 m) movement (<100 m) (>100 m) TB10 28% 54% 18% 26% 62% 12% Fall Brook 24% 68% 8% 26% 61% 13% 30 Northeastern Naturalist Vol. 20, No. 1 Figure 7. Seasonal mean distance moved for Brook Trout of SC1 at TB10 (A), SC2 at TB10 (B), SC1 at Fall Brook (C), and SC2 at Fall Brook(D). Distance moved was between previous point of capture and final recapture location. Negative values indicate Brook Trout that moved downstream. Figure 6. Mean distance moved for Brook Trout at TB10 (A) and Fall Brook (B). 2013 J. Ecret and T.B. Mihuc 31 Brook Trout in both TB10 and Fall Creek. Brook Trout in SC1, in both TB10 and Fall Brook, exhibited greatest movement downstream (>200 m) during October (Fig. 7A, C). Discussion The initial aim of this research study was to define and quantify Brook Trout habitat use within Adirondack headwater streams, as well as to examine the seasonal movement patterns among different age-classes of Brook Trout. We hypothesized that habitat use would vary between Brook Trout size classes, where smaller-sized Brook Trout would occupy shallow, slow-moving stream areas and larger Brook Trout would occupy deeper stream regions. We also theorized that smaller Brook Trout would exhibit greater movement patterns, and move larger distances, in contrast to larger Brook Trout, which would remain sedentary. Research has shown that salmonid fishes are not uniform in their habitat selection; rather they vary habitat selection based on seasonal variation, fish size, diel variation, food availability, and presence of intra-and interspecfic competitors (Huusko and Yrjana 1997, Mäki-Petäys et al. 1997). Water depth was the most prominent habitat variable exhibited for Brook Trout habitat use. Both size classes were more prevalent in pool habitats (Fig. 3) and were also found to occupy significantly deeper stream areas when compared to overall habitat availability. Within both study streams, largersized Brook Trout were observed in significantly deeper stream habitats than younger fish demonstrating varying habitat use among different size-classes of Brook Trout (Tables 3, 4; Figs. 4A, 5A). Fausch and White (1986) suggest that stream-resident salmonids choose focal points in streams with low velocities in order to minimize energy expenditure from actively swimming; furthermore, stream-resident salmonids remain close to currents to maximize energetic gains by feeding on invertebrate drift. This finding could suggest that larger-sized Brook Trout demonstrate competitive dominance for favorable habitat conditions. These results provide support for Mäki-Petäys et al.’s (1997) findings that showed statistical differences in habitat use among trout age classes in northern Finland. Mäki-Petäys et al. (1997) illustrated preference for deeper stream habitats among larger-sized Brown Trout, as compared to smaller trout. These results could in part be attributed to intercohort competitive interactions, whereby larger trout exclude smaller trout from more favorable stream habitat areas (Fausch and White 1986). Water velocity has been shown to be a strong influential factor contributing to habitat selection for stream-resident salmonids (Reyjol et al. 2001). However, we observed water velocity and substrate size to be secondary factors contributing to Brook Trout habitat use. There were no statistical differences between fish size classes in either study stream in regards to water velocity. However, in both study streams, both fish size classes were found in significantly slower habitats as compared to overall habitat availability (Tables 3, 4; Figs. 4C, 5C). Likewise, there was no significant difference in substratesize selection between trout size classes at Fall Brook; however, Brook Trout 32 Northeastern Naturalist Vol. 20, No. 1 of SC2 within TB10 were found at stream regions with significantly smaller substrates compared to SC1 (Table 3, Fig. 4B). Witzel and MacCrimmon (1983) showed that Brook Trout selected habitats with significantly smaller-sized substrates, as compared to Brown Trout, in a series of southwestern Ontario stream systems, successfully demonstrating the varying habitat use among salmonid species. Water velocity has also been shown to be interdependent with substrate size because larger substrates typically are associated with higher velocities (Armstrong et al. 2003). These two coupled habitat parameters may play a more focal role for stream-resident salmonids during spawning seasons (Shirvell and Dungey 1983). Larger reproductive Brook Trout may select for stream areas with higher velocities in order to adequately ventilate redds during embryo development, or may choose to spawn in stream regions with larger substrates to improve redd structure and stability, or a combination of both. Shirvell and Dungey (1983) suggest that stream-resident salmonids select for water velocity as a surrogate for substrates during spawning. Froude number is a useful descriptor of the hydraulic habitat conditions at stream regions because it is a dimensionless value that can be used to compare habitat conditions between small-scale streams and larger river systems as well as among different fish species (Armstrong et al. 2003). Similarly with water velocity and substrate size, both Brook Trout size classes within TB10 were found in stream habitats with significantly lower Froude numbers as compared to that of random locations (Table 3, Fig. 4D). In-stream enhancement structures were found to increase habitat availability and trout density by creating a more spatially complex microhabitat with reduced water velocity and Froude number values (Huusko and Yrjana 1997). Overall, Brook Trout populations in our study were composed of both immobile and mobile individuals. Brook Trout of SC1 within both TB10 and Fall Creek showed increased upstream movement (>200 m) during July and increased downstream movement during August and October. The only increased movement displayed by SC2 was during October, in which larger Brook Trout moved larger distances upstream (> 100 m) compared to the other two sampling events during July and August. (Fig. 7). These Brook Trout movement patterns have been well documented, showing an increase in movements during early spring and early fall due to increased flow events (Gowan and Fausch 1996, Jackson and Zydlewski 2009, and Peterson and Fausch 2003). The reduction in movement during summer months has been attributed to Brook Trout seeking thermal refuge in deeper pool habitats. Moreover, we also observed minimal trout movements (<50 m) during the summer. Our stream and fish sampling methods could have favored capturing Brook Trout in pool habitats due to their reduction in movement during summer months. Brook Trout have been shown to migrate to deeper pond or lake areas, and remain in these areas and exhibit little movement (Jackson and Zydlewski 2009). Peterson and Fausch (2003) suggested that during the summer, Brook Trout move more frequently upstream seeking colder water conditions. Movement and habitat selection may work in conjunction, causing smaller Brook Trout to move to more suitable foraging sites. Riffle habitats have been 2013 J. Ecret and T.B. Mihuc 33 shown to have higher densities of some aquatic macoinvertebrates that occupy the interstitial spaces within the substratum; these habitats provide both food resources and protection from fish predators. Also, the larvae of other suspension-feeding macroinvertebrates are more common in stream habitats with reduced flow. Likewise, riffle habitats serve to assist in dispersal for aquatic macroinvertebrates (Malmqvist 2002). Mäki-Petäys et al. (1997) suggest that riffle margins may constitute the more profitable habitat conditions for stream-resident Brown Trout. This study shows that smaller Brook Trout utilize riffle habitats more frequently than pool habitats, as compared to larger trout in both study streams (Fig. 3). This finding may suggest that younger Brook Trout utilize riffle habitats as foraging sites more often, while larger trout inhabit pools to forage. Differential feeding behavior exhibited by Brook Trout may constitute varying habitat selection based on foraging efficacy. In addition, preference for deeper habitats could be related to stream cover, as these two habitat variables are highly correlated. Canopy cover has been well recognized as a crucial habitat requirement for Brook Trout as a means to control stream temperature, as well as providing allochthonous inputs (Raleigh 1982). Selection for deeper habitats with increased canopy cover may also be attributed to predator avoidance by stream fishes (Power 1987). However, the absence of canopy cover as a habitat parameter within this study may be limiting our interpretation of differential habitat use among difference size classes of Brook Trout within our study streams. The absence of Brown Trout habitat use and movement data within this study may also be limiting the validity and understanding of Brook Trout habitat selection, as well as their seasonal dispersal patterns within these stream systems. The European Brown Trout has been widely introduced into stream/ river systems in the eastern United States and has been partially attributed to the reduction and alteration of Brook Trout populations (Grant et al. 2002). Although these two salmonid species differ in their natural habitat range, they have been shown to share several biological and ecological characteristics, which facilitate their overlap within stream habitats. Both sympatric Brook and Brown Trout share similarities in life-history events, including fall spawning. However, advantages in growth and foraging behavior exhibited by Brown Trout may contribute to their competitive advantage and potential displacement of Brook Trout (Dewald and Wilzback 1992). The reproductive and behavioral activities of sympatric Brook and Brown Trout have been examined in natural and laboratory settings, both of which have suggested a negative response of Brook Trout when interacting with the nonnative Brown Trout (DeWald and Wilzback 1992, Grant et al. 2002, Sorensen et al. 1995). The sampling methods utilized in our study allowed us to measure Brook Trout movement, but not without bias. Some factors of our study design, including timing and spacing between sampling sessions, as well as tag loss, may have influenced results. We did not observe any impairment of fish movement or mortality, but external tags have been shown to significantly influence movement by penetrating and impairing swimming muscles when inserted (Gowan 34 Northeastern Naturalist Vol. 20, No. 1 and Fausch 1996). Electrofishing may have also influenced Brook Trout habitat data by causing Brook Trout to move further upstream during stream sampling sessions. Studies have examined the potential impacts of electrofishing on rates of fish injury and also the influence on fish movement (Gowan and Fausch 1996). Although we marked Brook Trout habitat at first detection, fish may have shifted their habitat use in response to our sampling methods. Consequences for management The influence and underlying dynamic nature of lotic systems has led authors to overgeneralize stream-resident fish habitat-use patterns. Habitat suitability and availability are often regarded as the most significant factors affecting populations; thus, characterizing fish habitat is fundamental in restoring threatened populations (Armstrong et al. 2003, Bond and Lake 2003). Factors that influence habitat availability include dispersal barriers (i.e., road crossings) and introduced species that restrict target species’ response to restoration efforts (Bond and Lake 2003). Managers have utilized in-stream structures, including road crossings and barriers, to isolate and prevent threatened salmonid communities from non-native competitors. However, modifications to a stream or river channel, such as a road crossing, change physical habitat components (e.g., depth, velocity, and substrate) and greatly alter fish populations. Research on the ecological impacts of road crossings on stream-resident fish demonstrates that many species are negatively impacted; moreover, dispersal barriers such as road crossings have been shown to be major threats to both stream fish abundance and diversity; furthermore, habitat fragmentation has been shown to drastically increase the rate of extinction in stream networks (Letcher et al. 2007). However, understanding individual salmonid species behavioral movement patterns, based on varying life-history stages, must be considered prior to barrier construction. Land-management agencies may be inadvertently trading the risk of extirpation from invading fish species for other risks including restricted movement to available stream habitats, as well as overall habitat fragmentation (Gowan et al. 1994, Gowan and Fausch (1996). Our study demonstrates that native Brook Trout share stream habitats with non-native Brown Trout, and preceding studies have successfully illustrated behavioral and reproductive interactions between these sympatric species (DeWald and Wilzback 1992, Grant et al. 2002, Sorensen et al. 1995). Additionally, these results show that Brook Trout populations are not homogenous in their movement patterns within Adirondack headwater streams; rather, they are comprised of both sedentary and highly mobile individuals. Understanding individual fish species movements and habitat-use patterns will ultimately lead to a better comprehension of the dynamic behavior of stream fish ecology allowing us to preserve and enhance threatened fish populations. Acknowledgments A great deal of thanks to my graduate committee, Dr. Timothy Mihuc, Dr. Danielle Garneau, and Dr. Edwin Romanowicz. I would also like to thank the faculty and staff 2013 J. Ecret and T.B. Mihuc 35 members of SUNY Plattsburgh and The Lake Champlain Research Institute for their time and effort in completing this research. This work was funded by the United States Fish and Wildlife Service, Trout Unlimited, and The Lake Champlain Research Institute. Literature Cited Armstrong, J.D., P.S. Kemp, G.J.A. Kennedy, and M. Ladle. 2003. Habitat use of Atlantic Salmon and Brown Trout in rivers and streams. Fisheries Research 62:143–170. Bond, N.R., and P.S. Lake. 2003. Characterizing fish-habitat associations in streams as the first step in ecological restoration. Austral Ecology 28:611–621. DeWald, L., and M.A. Wilzback. 1992. Interactions between native Brook Trout and hatchery Brown Trout: Effects on habitat use, feeding, and growth. Transactions of the American Fisheries Society 121:287–296. Eastern Brook Trout Joint Venture (EBTJV). 2006. Eastern Brook Trout : Status and threats. Prepared by Trout Unlimited, Arlington, VA, for EBTJV. 36 pp. Fausch, K.D., and R.J. White. 1986. Competition among juveniles of Coho Salmon, Brook Trout, and Brown Trout in a laboratory stream, and implications for Great Lakes tributaries. Transactions of the American Fisheries Society 115:363–381. Gerking, S.D. 1959. The restricted movement of fish populations. Biological Review 34:221–242. Gore, J.A. 2006. Dishcharge measurements and streamflow analysis. Pp. 51–77, In F.R. Hauer and G.A. Lamberti (Eds.). Methods in Stream Ecology, 2nd Edition. Academic Press, London, UK. Gowan, C., and K.D. Fausch. 1996. Mobile Brook Trout in two high-elevation Colorado streams: Re-evaluating the concept of restricted movement. Canadian Journal of Fisheries and Aquatic Sciences 53:1370–1381. Gowan, C., M.K. Young, K.D. Fausch, and S.C. Riley. 1994. Restricted paradigm in resident stream salmonids: A paradigm lost? Canadian Journal of Fisheries and Aquatic Sciences 51:2626–2637. Grant, G.C., B. Vondracek, and P. Sorensen. 2002. Spawning interactions between sympatric brown and Brook Trout may contribute to species replacement. Transactions of the American Fisheries Society 131:569–576. Hansbarger, J.L. 2005. Trout movement and habitat use in the Upper Shavers Fork of the Cheat River, West Virginia. M.Sc. Thesis. West Virginia University, Morgantown, WV. Hilderbrand, R.H., and Kershner, J.L. 2000. Movements patterns of stream-resident Cutthroat Trout in Beaver Creek, Idaho–Utah. Transactions of the American Fisheries Society 129:1160–1170. Hudy, M., T.M. Thieling, N. Gillispie, and E.P. Smith. 2008. Distribution, status, and land-use characteristics of subwatersheds within the native range of Brook Trout in the eastern United States. North American Journal of Fisheries Management 28:1069–1085. Hutchings J.A., and L. Gerber. 2002. Sex-biased dispersal in salmonid fish. Biological Sciences 269:2487–2493. Huusko, A., and T. Yrjana. 1997. Effects of instream enhancement structures on Brown Trout, Salmo trutta L., habitat availability in a channelized boreal river: A PHABSIM approach. Fisheries Management and Ecology 4:453–466. Jackson, C.A.L., and J. Zydlewski. 2009. Summer movements of sub-adult Brook Trout , landlocked Atlantic Salmon, and Smallmouth Bass in the Rapid River, Maine. Journal of Freshwater Ecology 24:567–580. 36 Northeastern Naturalist Vol. 20, No. 1 Jowett, I. 1993. A method for objectively identifying pool, run, and riffle habitats from physical measurements. New Zealand Journal of Marine and Freshwater Research. 27:241–248. Letcher, B.H., K.H. Nislow, and J.B. Coombs. 2007. Population response to habitat fragmentation in a stream-dwelling Brook Trout population. PlosOne 11:e1139. Mäki-Petäys, A., T. Muotka, A. Huusko, P. Tikkanen, and P. Kreivi. 1997. Seasonal changes in habitat use and preferences by juvenile Brown Trout, Salmo trutta, in a northern boreal river. Canadian Journal of Fisheries and Aquatic Sciences 54:520–530. Malmqvist, B. 2002. Aquatic invertebrates in riverine landscapes. Freshwater Ecology 47:679–694. Mihuc, T.B., E. Allen, and E. Romanowicz. 2008. Lake Champlain basin fish passage initiative. Unpublished report submitted to Lake Champlain Sea Grant Program, Burlington, VT. Morrissey, M.B., and M.M. Ferguson. 2011. Individual variation in movement throughout the life cycle of stream-dwelling salmonid fish. Molecular E cology 20:235–248. Penczak, T. 2006. Restricted-movement paradigm: Fish displacements in a small lowland streamlet. Polish Journal of Ecology 54:145–149. Peterson, D.P., and K.D. Fausch. 2003. Upstream movement by nonnative Brook Trout (Salvelinus fontinalis) promotes invasion of native Cutthroat Trout (Oncorhynchus clarki) habitat. Canadian Journal of Fisheries and Aquatic Sciences 60:1502–1516. Power, M.E. 1987. Predator avoidance in grazing stream fishes in temperate and tropical streams: Importance of stream depth and prey size. Pp. 333–351, In W.C. Kerfoot and A. Sih (Eds.). Predation: Direct and Indirect Impacts in Aquatic Communities. University of New England Press, Hanover, NH. Raleigh, R.F. 1982. Habitat-suitability index models: Brook Trout. US Department of Interior, Fish and Wildlife Service, Biological Services Program, Lafayette, LA. FWS/ OBS-82/10.24. 42 pp. Reyjol, Y., L. Puy, B. Alain, and L. Sovan. 2001. Modelling of microhabitat used by fish in natural and regulated flows in the river Garonne France. Ecological Modelling 146:131–142. Shirvell, C.S., and R.G. Dungey. 1983. Microhabitats chosen by Brown Trout for feeding and spawning in rivers. Transactions of the American Fisheries Society 112:355–367. Smithson, E.B., and C.E. Johnston. 1998. Movement patterns of stream fishes in an Ouchita highlands stream: An examination of the restricted movement paradigm. Transactions of the American Fisheries Society 128:847–853. Sorensen, P.W., J.R. Cardwell, T. Essington, and D.E. Weigel. 1995. Reproductive interactions between sympatric Brook and Brown Trout in a small Minnesota stream. Canadian Journal of Fisheries and Aquatic Sciences 52:1958–1965. Wilson, A.J., J.A. Hutchings, and M.M. Ferguson. 2004. Dispersal in a stream-dwelling salmonid: Inferences from tagging and microsatellite studies. Conservation Genetics 5:25–37. Witzel, L.D., and H.R. MacCrimmon. 1983. Redd-site selection by Brook Trout and Brown Trout in southwestern Ontario streams. Transactions of the American Fisheries Society 112:760–771.WV. 155 pp.