Eighteen Mile Creek Highway 63 Culvert Fish Passage Evaluation

Summary of 2001 and 2002 Electrofishing and Marking Surveys, Scott Toshner - Fisheries Biologist, Douglas and Bayfield Counties - January 2004

Executive Summary

In 2001 and 2002, we used a mark and recapture survey to assess fish passage through a box culvert with baffles on Eighteen Mile Creek, Bayfield County, in northwestern Wisconsin. Electrofishing surveys in May - September were used to collect and mark brown trout (Salmo trutta) and brook trout (Salvelinas fontinalas) from both upstream and downstream of the culvert. Trout collected upstream were then moved (displaced) downstream of the culvert and released. In October, electrofishing surveys were completed upstream of the culvert to discern if trout had moved back through the culvert. Brown trout showed the capability of movement through the culvert. Capture of displaced large brown trout upstream of the culvert > 8.0 in suggested the culvert design was sufficient to allow passage. However, smaller < 8.0 in brown trout may not have either the same ability or motivation to move through the culvert as larger fish. Numbers of brook trout captured were not sufficient to determine their ability to move through the culvert. Future monitoring will help determine culvert water velocities for long-term effects on trout populations and to assess needs for culvert maintenance.

Introduction

In 2000, the Wisconsin Department of Transportation (DOT) replaced a 300-foot concrete box culvert at U.S. Highway 63 on Eighteen Mile Creek, Bayfield County, Wisconsin (Figure 1). Eighteen Mile Creek is a Class I trout stream tributary to the White River with self-sustaining populations of both brown trout (Salmo trutta) and brook trout (Salvelinas fontinalas). Estimated normal flow in Eighteen Mile Creek is 22.3 cfs (Johannes et al. 1971). DOT design of the culvert heeded Wisconsin Department of Natural Resources (DNR) fisheries concerns about fish passage by designing a two-cell structure with a series of baffles in one cell. This design, in theory, would provide suitable depths and velocities for fish passage through that cell under a wide range of flow conditions (Figures 2 and 3). Following initial culvert installation, DOT engineers made some modifications due to concerns raised by DNR fisheries personnel about the ability of fish to migrate through the structure. Modifications included installing smaller baffles on the downstream apron portion of the outfall and adding large rock at the base of the apron to limit the depth of the drop. The objective of this study was to determine if the culvert design on Eighteen Mile Creek was adequate to allow passage of brown and brook trout. More specifically, we intended to monitor movement of trout marked downstream and also of trout captured upstream and displaced downstream as an index of the culvert's utility for passage.

Methods

The project was designed to assess brown and brook trout movement through the Highway 63 culvert on Eighteen Mile Creek in two ways: 1) trout were marked in stations located downstream of the culvert to assess movement through the culvert of trout that were migrating upstream, and 2) trout were marked upstream of the culvert and displaced downstream to assess whether these trout would navigate back through the culvert to locations where they were originally collected.

Electrofishing was used to collect brown and brook trout of all sizes. Electrofishing gear included a towed stream shocker boat equipped with a 220 V DC generator, two anodes, and a cathode of sheet metal (which also protected the bottom of the boat from abrasion). The average operating parameters of the stream shocker were 250 V and 1.5 to 2.0 amps. Sampled trout were measured to the nearest 0.1 in (total length) and marked by either finclip or floy tag. During marking runs, all trout were given identifying clips on the caudal, pelvic, or adipose fins to discern where the trout were originally marked. Brown trout collected larger than 7 in were tagged with an individually numbered anchor type floy tag (Floy Tag Inc.). Brook trout were not given anchor type Floy tags. Numbered tags were also used to monitor and confirm movement of individual fish through the culvert.

In 2001 and 2002 marking runs were made at stations both downstream and upstream of the culvert in May through September (Figure 1). Trout collected upstream in the marking runs were moved downstream of the culvert and released. In October of 2001 and 2002, recapture runs were completed upstream of the culvert to document movement through the culvert. The recapture sampling occurred 42-45 d following displacement of trout from upstream stations, and 16-151 d after sampling occurred downstream.

Movement upstream through the culvert was also compared by trout length. Brown trout that were surveyed upstream of the culvert and displaced downstream in September for both 2001 and 2002 were used. For comparative purposes, brown trout were separated into two length groups: 5.0 - 7.9 in, and 8.0 - 14.0 in. Brown trout that were < 5.0 in were not included do to sampling bias of the gear type used during the survey.

All data were entered into Microsoft Excel spreadsheets and pivot tables were used in manipulation. Length of recaptured trout was compared with the initial marking sample using a Chi-square test. The alpha value significance was 0.05. Proportional stock density (PSD) and relative stock density (RSD) were used as indicies of quality trout in the population; where stock size equals 6 in and quality size 9 in (Anderson and Neumann 1996).

Results

A total of 2,096 brown trout and 82 brook trout were sampled in all the sites surveyed for the culvert assessment in 2001 and 2002 (Table 1). Brown trout composed 96% of all trout sampled. The mean length of brown trout was 7.3 in (S.D. = 2.33, range 1.1 -20.0 in). Brown trout PSD was 38, and RSD-12 was 3. The mean catch per unit effort of brown trout in all stations was 769 trout/mile. The mean length of brook trout was 5.5 in (SD = 1.70, range 1.7 - 9.8 in). Brook trout PSD was 14. The mean catch per unit effort of brook trout in all stations was 28 trout/mile. The length frequencies of brown trout suggested recruitment was occurring on a regular basis (Figure 4).

Trout Passage through Highway 63 Culvert

Only one brook trout was found to have moved through the culvert. However, a total of only 82 brook trout were sample during the study.

Upstream displaced, and downstream brown trout had the ability to move upstream through the culvert. However, there was a greater tendency for brown trout displaced from above the culvert to return to their original area of capture compared to those marked below the culvert migrating upstream (c2 = 308.8, df = 1, P < 0.001). In 2001 and 2002, 75 (39%) of 194 brown trout captured above the culvert and displaced downstream below the culvert had moved back through the culvert after 45 d. In contrast, of 824 brown trout marked downstream of the culvert in 2002, only 5 (<1%) moved through the culvert and were collected again in October of 2002 upstream of the culvert.

Recapture of individually tagged brown trout supported the contention of successful movement through the culvert. While a majority of the tagged brown trout remained in or near the same area in which they were tagged there were exceptions. During both survey years, 49 tagged brown trout moved through the culvert once and 13 moved through the culvert twice. Ten (77%) of the tagged brown trout that moved through the culvert twice were displaced downstream of the culvert twice. Several other brown trout demonstrated movements that also warrant mention. One fish (tag 127; 9.4 in) was tagged in August of 2001 at North Sweden Road (Station 2), then recaptured again at Old Highway 63 (Station 5) on two dates in May of 2002 after having moved upstream 1.7 miles (Figure 1). Another brown trout (tag 322; 10.3 in) was tagged upstream of the culvert and displaced downstream in August of 2001. It then traveled back through the culvert to be sampled again in October 2001 (Station 4). In the spring of 2002, this fish was recaptured again at North Sweden Road (Station 2), indicating it moved upstream through the culvert and then went downstream through the culvert again.

Larger brown trout (8 - 14 in) moved upstream of the culvert at a higher frequency compared to brown trout < 8 in following their displacement downstream 45 d earlier. Return to upstream locations was greater for displaced larger brown trout (8.0-14.0 in) compared to smaller brown trout (5.0 - 7.9 in; c2 = 15.9, df = 1, P < 0.001, Figure 6). The number of brown trout in the 5.0 - 7.9 and the 8.0 - 14.0 in groups that were moved downstream of the culvert were 95 and 99, respectively. In comparison, the numbers of brown trout in the 5.0 - 7.9 and 8.0 - 14.0 in groups that were found upstream of the culvert were 16 and 59, respectively.

Discussion

Original study objectives intended to determine the ability of both brown and brook trout to navigate/pass through the culvert. However, low total numbers of brook trout captured did not allow any analysis for brook trout movement.

One premise of the study was that trout marked downstream of the culvert would move through the culvert to spawn upstream. However, < 1% of the brown trout marked downstream moved through the culvert. Interestingly, nearly 40% of brown trout initially surveyed upstream and displaced downstream of the culvert moved back upstream through the culvert. Such findings suggest that nonreproductive homing of brown trout may have been a mechanism for displaced fish to return to areas near their original capture location. If an upstream spawning migration was occurring, both groups would have been expected to have been captured at similar rates. Low upstream movement of downstream marked brown trout could be due to ample spawning areas observed downstream of the culvert or the timing of the sampling, which may have occurred before their spawning period. In the Brule River, located 25 miles from Eighteen Mile Creek, brown trout typically spawn from mid-October through mid-November (Becker 1983). Nevertheless, the culvert appears to be passable by brown trout thus fulfilling the design objectives in that aspect.

There is a possibility that larger brown trout exhibited greater home range fidelity, but this could be a size/age effect or suggests reduced ability of small brown trout to navigate the culvert. Smaller brown trout may not be able to move through the culvert due to higher than passable velocities encountered in the culvert. Barber and Downs (1996) found that culverts often create barriers to the upstream migration of juvenile fish. In addition, the National Marine Fisheries Service concluded that the most representative flows that juvenile salmonids migrate upstream are up to the 10% annual exceedance flow. The maximum average water velocity at this flow should not exceed one foot per second. In some cases over short distances two feet per second may be adequate (NMFS 2001). Measuring the velocities in the culvert at base flow conditions could determine if the velocities are higher than two feet per second. A more intensive survey is required to answer the question about juvenile brown trout passage. The present survey did not adequately acquire information on juvenile trout with regard to flow velocities.

Summary and Recommendations

The 2001 and 2002 surveys indicate the Highway 63 culvert allowed passage of brown trout. However, there was a significant bias with regard to the size of brown trout that move through the culvert. Larger brown trout (> 8 in) moved upstream through the culvert more readily than smaller brown trout (5.0 - 7.9 in). The mechanism causing this occurrence is uncertain but is likely related to either flow velocities affecting movement by smaller fish, or stronger homing by larger fish. Regardless, this may lead to a bottle neck situation below the culvert of small fish, resulting in some long-term affects of the brown trout population upstream of the culvert. This issue warrants future study.

Baffled culverts are known to have high maintenance requirements, especially in high-energy stream systems, where large bed-load movements tend to fill the spaces between baffles rendering them ineffective (BCMTH 2000). This points to the need to evaluate flow rates through the culvert and assess whether accumulations of bed-load materials are having an effect on velocities and thus fish passage. Once this question is answered, a maintenance plan should be written that provides a schedule for monitoring of velocities and subsequent maintenance. During these maintenance inspections, large debris should be removed if presenting impedance to velocities; smaller material both upstream and downstream should be left in place to provide habitat as trout enter and leave the culvert.

If flow monitoring indicates that flow exceeds 2 ft/sec during low flow periods even in the absence of accumulations of bed-load materials in the culvert baffles, altering the design of the culvert to accommodate passage of smaller brown trout should be considered. Such design alterations might include diversion of more of the flow through the non-baffled side of the culvert.

Literature Cited

Anderson, R.O., and R.M. Neumann. 1996. Length, weight, and associated structural indices. Pages 447-481 in B.R. Murphy and D.W. Willis, editors. Fisheries techniques, second edition. American Fisheries Society, Bethesda, Maryland.

Barber, M.E. and Downs, R.C., 1996. Investigation of culvert hydraulics related to juvenile fish passage. Pullman, Washington: Washington State Transportation Center (TRAC), Washington State University. WA-RD-388.2:1-172. Final Technical Report.

Becker, G.C., 1983. Fishes of Wisconsin. The University of Wisconsin Press, Madison, Wisconsin.

BCMTH (British Columbia Ministry of Transportation and Highways, Environmental Management Section). 2000. Culverts and Fish Passage - Fact Sheet.

Johannes, S.I., L.M. Sather, and C.W. Threinen. 1970. Surface water resources of Bayfield County. Wisconsin Department of Natural Resources, Madison.

NMFS (National Marine Fisheries Service - Southwest Region). 2001. Guidelines for salmonid passage at stream crossings.

Table 1. Number and length ranges of trout from stations on Eighteen Mile Creek, Bayfield County, Wisconsin

Figure 1. U.S. Highway 63 crossing of Eighteen Mile Creek

Locations of electrofishing and marking/tagging in 2001 and 2002. 1 = Upstream of Taylor Lane, 2 = Upstream of North Sweden Road, 3 = Downstream of Hwy. 63, 4 = Upstream of Hwy. 63, and 5 = Upstream of Old Hwy. 63.

Figure 2. Upstream section of culvert showing baffle, Eighteen Mile Creek

Figure 3. Downstream section of culvert showing baffle, Eighteen Mile Creek

Figure 4. Brown trout length frequencies, 2001 and 2002, Eighteen Mile Creek

Figure 5. Length frequency of brown trout >5.0 in.

The chart below shows the length frequency of brown trout >5.0 in. displaced downstream, and those that returned upstream of the culvert, Eighteen Mile Creek, Bayfield, Wisconsin.