Skip to contentUnited States Department of Transportation - Federal Highway AdministrationFHWA HomeFeedback

Hydraulics Engineering

<< Previous Contents Next >>

Design For Fish Passage at Roadway - Stream Crossings: Synthesis Report

5 Fish Passage Hydrology

How to use this chapter

  • This chapter is used most when using Hydraulic Design methods for culvert retrofits
  • Understand the difference between determining design discharge for flood events and fish passage
  • Learn the importance of timing and seasonality
  • Find common methods used to determine fish passage design discharge

5.1 Hydrologic considerations for fish passage design

Crossings should allow fish passage at a range of flows corresponding to the timing and extent of fish movement within the channel reach. The use of Hydraulic Design techniques tailored to specific fish species and life stages and the need to assess existing culverts for fish passage requires knowledge of fish swimming ability and site hydrology in order to create a passable structure. This process necessitates a more thorough understanding of site flow characteristics than is provided by a typical hydraulic analysis for structure stability. The following discussion details typical design requirements, including state-of-practice hydrology.

5.1.1 Seasonality Timing and Extent of Fish Presence

The timing of fish presence, including migration if important, must be considered when determining appropriate hydrology for fish passage design. Fish presence can vary from watershed to watershed (Scott and Crossman 1973), and in-stream flows may show great disparity with timing of fish migration .

In addition, the presence of multiple fish species can quickly convolute evaluation of fish passage hydrology. Figure 5.1 depicts the general timing of fish spawning migrations for a number of freshwater species in Virginia. Determining species presence and sensitivity within a stream reach requires site-specific knowledge, and consultation with a local fisheries biologist is essential.

Graph showing the months of the year (x-axis) versus peak spawning time (y-axis) for seven freshwater fishes in Virginia. Spawning times cover almost every month of the year.
5.1 Peak spawning periods for a selection of freshwater fish in Virginia, based on biological data from Scott and Crossman (1973) (adapted from Hudy 2006) Species and Life Stage

Timing and movement of regional fish populations will depend on fish species and life stage. In the Pacific Northwest, for example, adult salmon and steelhead migrate in the fall and winter months, while juvenile salmon migrate in the spring as fry and in the fall as fingerlings (Bates et al. 2003). Culverts designers in Maine must consider spawning movement of Atlantic salmon from May to November (Maine Department of Transportation 2004). In addition, resident fish may require movement at any time of the year (Kahler and Quinn 1998; Gowan et al. 1994). Due to variable abilities and periods of migration, each fish species and life stage may necessitate a different set of hydrologic constraints. Representing Seasonal Flows

While predictions of instantaneous peak discharges with return periods of between 25 and 100 years are used for flood conveyance design, fish passage is considered at much smaller flows, less than bankfull discharges. Consequently, flow duration curves (FDC) are useful tools for determining fish flows. For example, a high fish passage discharge may be the discharge exceeded 10% of the time during migration or fish movement season (see Figure 5.1), while a low-flow requirement may be the flow that is equaled or exceeded 90% of the time.

However, the FDC used in analysis of fish passage flows, such as in Figure 5.2, represent averages and fail to account for annual variations in hydrology. A study in Northern California found a culvert using specified low-flow criteria (90% migration period exceedance) created a 1-day migration delay in WY99 (a "wet" year) but a 10-day delay in WY2001 (a "dry" year) (Lang et al. 2004).

Synthetic flow duration curves for May Creek. The x-axis shows the percent of time the flow is equaled or exceeded, and the y-axis shows the value of discharge. One curve represents annual data; the other is for the November to April migration season.
Figure 5.2 Synthetic flow duration curves from May Creek, customary units (Lang et al. 2004)
(Calculations were based on flows occurring from November through April (Migration Season) and October through September (Annual); curves were created using the regional flow duration curve; the annual flow (Qave) for May Creek was estimated to be 5.9 cfs)

5.1.2 Extreme Events Fish Response

Even within a period of fish migration, design is not intended to provide fish passage at all flows. In a natural stream reach, fish respond to high flow events by seeking out shelter until passable conditions resume (Robison et al. 1999). During extreme low flows, shallow depths may cause the channel itself to become impassable (Clarkin et al. 2003; Lang et al. 2004). Generally, upper and lower thresholds bound the flow conditions at which fish passage must be provided. Allowable Delay

Fish may be able to handle a short interruption to upstream migration without negative consequences. The extent of this "allowable delay" depends on the timing and motivations for fish movement. A resident fish may be able to tolerate a short delay without extreme consequences, while a delay of a few days may be detrimental to spawning salmon, whose migrations involve significant physical changes, including a rapid depletion of fat and protein reserves (Groot and Margolis 1991). The delay caused by a single culvert can be compounded by a series of culverts that present short delays, making it imperative to understand a crossing's place in the overall watershed context. Delay has a number of negative consequences including stress and physical damages, susceptibility to disease and predation, and reduction in spawning success (Ashton 1984). Migration Flows

As discussed in Chapter 2, fish movement is triggered by time of year, flow events and a number of environmental factors. For example, the upstream migration of spawning salmon is hypothesized to be in response to maturation, the changing length of days, and temperature regimes (Groot and Margolis 1991). Consultation with local fisheries biologists will help ensure that hydrology is properly matched to requirements of local fish populations.

5.2 Design Requirements

Flood design discharge is estimated for all culverts. For Hydraulic Design and assessment of existing culverts, two additional discharges are required: high and low fish passage flows. These are often compared to bankfull discharge.

5.2.1 High Fish Passage Flows

A high fish passage flow captures the upper bound at which fish are believed to be moving within the stream. Fish passage requirements should be met at all discharges up to and including the high fish passage flow. This may exclude flows falling below a lower threshold, known as the low fish passage flow.

Table 5.1 shows a comparison of available State and agency guidelines for high fish passage flows. Many states use an exceedance flow between 1 and 10% of the annual flow duration curve (a 10% exceedance flow is met or surpassed 10% of the year). It has been suggested that spawning adults should be delayed no more than 3 days during the average annual flood, or 7 days during the 50-yr flood (Ashton 1984).

Table 5.1 State and Agency Guidelines for High Fish Passage Flows, Customary Units (adapted from Clarkin et al. 2003)
(Q2 refers to the 2-year flood)
Alaska Washington Oregon NMFS SW Region California Dept of Fish and Game NMFS NW Region Idaho
Q2d2: the discharge 24 hours before the 2-yr flood. 10% exceedance flow during migration period - species specific 10% exceedance flow during migration period: species specific. Approximate by Q10% = 0.18*(Q2)+36 where Q2 > 44 cfs. Where Q2 < 44 cfs, use Q2. For adult salmon and steelhead 1% annual exceedance flow or 50% Q2. For juveniles, 10% annual exceedance flow. Standards vary from 1-10% exceedance flow for various groups of fish. 5% exceedance flow during period of upstream migration <2 day delay during period of migration

+High flows are for Hydraulic Design Approaches only, with the exception of Alaska and Idaho.

5.2.2 Low Fish Passage Flows

Low fish passage flows define the lower bound at which fish passage is required. This flow condition is used to ensure that depth and velocity barriers are not created within a crossing. Flows below this threshold may cause the channel itself to present a depth barrier to fish movement (Clarkin et al. 2003).

Specific depth requirements vary with the species and life stage of concern. Alaska requires that depth be greater than 2.5 times the depth of a fish's caudal fin (Alaska Department of Fish and Game and Alaska Department of Transportation 2001). For example, a 60 mm (0.2 ft) juvenile Coho Salmon requires a water depth of approximately 48 mm (1.9 in). Washington State specifies a minimum depth of 0.24 m (0.8 ft) for Adult Trout, Pink and Chum Salmon, and a depth of 0.30 m (1.0 ft) for adult Chinook, Coho, Sockeye or Steelhead (Bates et al. 2003).

Table 5.2 depicts available current state guidelines for low flow analysis of fish crossings. Many current design manuals specify design based on a 2-yr 7-day low flow, roughly corresponding to the 95% exceedance flow.

Table 5.2 State and Agency Guidelines for Low Fish Passage Flows, Customary Units (adapted from Clarkin et al. 2003)
Alaska Washington Oregon NMFS SW Region California Dept of Fish and Game NMFS NW Region
None 2-yr, 7-day low flow (WAC 220-110-070) Natural bed culverts must be maintained to ensure low flow channels are ok 2-yr, 7-day low flow or 95% exceedance flow for migration period: species specific Adult Salmon - Greater of 3 cfs or 50% exceedance flow Juveniles - Greater of 1 cfs or 95% annual exceedance flow Standards vary from 50-95% exceedance flow for various groups of fish. 95% exceedance flow during months of upstream migration

+ Low flows are for Hydraulic Design approaches only, with the exception of Alaska.

5.2.3 Bankfull Flow

Bankfull flow is the discharge at which flow from the main channel begins to spill over into the floodplain (see Glossary). Generally, this discharge is referenced as the 1- to 2-yr flood event (Leopold and Wolman 1957), although this does not always correspond to field observations (Mussetter 1989). Bankfull is an important parameter in alluvial channels, as it is the discharge that effectively transports the most sediment, impacting long-term channel form, function, and stability (Harrelson et al. 1994). Although bankfull flow is rarely calculated for fish passage analysis, the concept of bankfull width is an important design parameter for fish passable structures. This concept will be discussed further in Chapter 6.

5.2.4 Streambed Stability and Crossing Capacity

Although design for fish passage will generally control structure size, culverts must still comply with flood flow conveyance requirements. At any road crossing, structure stability must be maintained up to and including a design flood (Norman et al. 2005). An outline of the hydrologic cycle, and methods for determining extreme flows are included in HDS-2 (Federal Highway Administration 2002).

Occasionally, design methods will also require that streambed material be sized for stability during a specific design flood (e.g. Alaska Department of Fish and Game and Alaska Department of Transportation 2001; Bates et al. 2003). Generally, this stability analysis corresponds to the discharge used to check culvert capacity - on the order of a 50-year event. Table 5.3 includes design flows used for streambed stability in fish culverts.

Table 5.3 Flows Used in Determining Adequate Streambed Stability
(adapted from Clarkin et al. 2003)
(Q50 and Q100 refer to the 50-year and 100-year floods, respectively)
Alaska Washington Oregon NMFS SW Region California Dept of Fish and Game
Q50 or Q100 * Q100 with Debris * Q100 Q100 at headwater/rise = 1 Q100 at headwater/rise = 1.5

* Streambed stability check required

5.2.5 Tidal Influence

The hydrology of culverts in tidal areas requires consideration of both upland flow and tidal impact (Zevenbergen et al. 2004). Methods for determining culvert outflow with changes in tidal elevation must account for stream flow as well as tidal outflow as an ebbing tide causes water to return to the ocean. Successfully meeting fish passage provisions may require tidal data in appropriate time increments and a continuous hydrologic simulation model for tidal elevations and stream flow. Examples include the U.S. Environmental Protection Agency's Hydrological Simulation Program - Fortran (HSPF) or Storm Water Management Model (SWMM) (Bates et al. 2003). Observed and predicted tidal elevations, including information on benchmarks for tidal stations, are available on NOAA's internet site at

A detailed discussion of tidal patterns, influence, and references are provide in Hydraulic Engineering Circular 25, available at, and the Army Corps of Engineers has a number of publications on construction in coastal areas, available at

5.3 Hydrologic Procedures

Procedures for determining flood conveyance and streambed stability discharges are well defined and can be found in State drainage design manuals.

Procedures for determining high and low discharges for fish passage are not as well established. For culvert locations at or near long-term stream gages, statistical analyses will yield estimates for (1) 7-day low flows for return periods of interest and (2) discharges associated with the percent time exceeded during the year, based on the flow duration curve.

Most culverts are located at ungaged sites. Determination of high and low fish passage flows will likely be based on statistical regression methods using local or regional stream gages (Rowland et al. 2003; Powers and Saunders 1996)

<< Previous Contents Next >>

Updated: 09/22/2014

United States Department of Transportation - Federal Highway Administration