Hi team,

Kelly has put together a short one-pager on something we’ve been thinking about for a while: how fish actually get through fast and complex water, and what that means for how we judge culverts.

With respect to water velocity, most fish passage design regulations come down to one number: is the cross-sectional average velocity under a pre-set limit? It’s a fair starting point, but fish don’t experience water velocity as an average. Their lateral line lets them sense the swirls, wakes and slow pockets around them, and they use that structure to work their way upstream.

The worked example in the note shows this clearly. Take a 70mm Galaxiid in a smooth, fully compliant culvert running at 0.3 m/s. It runs out of steam after about 33m. Put the same fish in a “non-compliant” culvert running at 0.6 m/s, but with refuges every 2m or so, and it can burst from one refuge to the next almost indefinitely. One structure meets the number and fails the fish. The other fails the number and passes them.

Those refuges are more than rest stops. They’re also where juvenile fish can drift feed. And because fish can only sense what’s right around them, they have no way of knowing in advance whether a culvert is 20m or 200m long. Regular refuges give them a string of short, sensible decisions instead of one long, blind commitment.

None of this argues against velocity limits. It’s a case for adding refuge spacing and maximum un-refuged distance alongside them, so each structure is judged on what a fish actually experiences.

As always, we’d love to hear your thoughts.
Read the one page note here