We use a short range DIDSON in Redwood Cr, and a long range ARIS in the Mad R. Redd surveys are independently counted in Redwood Cr. To date (3 yrs of data), we are not seeing any agreement with sonar counts and redd counts for Chinook salmon and steelhead trout. The redd counts are negatively biased, even if you take into account that not all salmon passing the sonar beam will spawn (pre-spawn mortality). I know there is similar work on the Secesh R in Idaho, and I am wondering if any other place has a similar study going on and what their results might be.