AEI lay summary of:
Daniel E. Holt, Carol E. Johnston. Traffic noise masks acoustic signals of freshwater stream fish. Biological Conservation 187 (2015) 27-33 (ScienceDirect link)
With each passing year, we learn more about the ways that animals use sound—and so also how human noise interferes with their lives.  A new paper looks at how traffic noise from bridges may impinge on the mating calls of freshwater fish; this is the first study to use some of the new metrics of “communication space” in these important and widespread habitats. The species studied was the blacktail shiner, a member of the largest family of fishes (including carps and minnows), with the study sites being small streams passing under bridges on I-85 in Alabama (image shows one of six sites).

Male shiners make two sounds during mating: loud “knocks” used to challenge other males who are intruding, and softer “growls” used to court females.  Streams are naturally loud environments, with noise from wind, rain, and turbulence; shiners take advantage of a relatively quiet “window” in the broadband noise, between 172 and 366Hz (like many other animals that vocalize in frequency ranges less cluttered by local sounds or other species).  While the traffic noise is not much louder than the natural stream sounds at frequencies above 700Hz, unfortunately for the shiners, in this key quiet window it is significantly louder than the stream noise—and also the seductive growls of male shiners.  The graph shows natural ambient noise (green), road noise (red), and growls (black dotted line).  The two peaks in the growl acoustic spectrum are particularly important; the lower peak in particular is dramatically drowned out by traffic noise.

The bottom line for the fish is that their knocks, which can be heard above the natural sounds of the creek out to about a half meter, are just slightly masked—only within three meters of the bridge are they lost in road noise (3m is the mean; maximum modeled range of effect is 22m).  So these calls of challenge and defense among males, which may also show females who’s the most fit, can serve their purpose unless the action is taking place right under a bridge.  The subtler sounds of the growls, however, are much more impacted.  These sounds, being quieter, are meant to be heard at very close range (generally just a few inches from the nest sites); yet the lower peak in the growl sound spectrum will be effectively inaudible in areas out to 640m (almost a half mile) from a bridge, and the second peak will be similarly masked out to 40m (both distances are means; maximum ranges are, respectively, 12km/7 miles and 1600m/1 mile). Adding insult to injury, peak spawning time is morning, before water temperatures rise, which may coincide with peak morning traffic.  Of course, only a small portion of most spawning streams is near heavily travelled interstates or secondary roads; those near more sporadically-travelled local roads are likely to be less affected.  Still, if the effect extends a half mile or more, large stretches of many streams could have some degradation of their natural and necessary acoustic habitat.

The authors’ conclusion neatly sums up what all this means going forward: