The rod catches start from a fairly low base in 1979 and then steadily rise with the sea-trout catches peaking around 2001 – 2003 before steadily falling to 2011 and then more dramatically to 2019. The salmon catches rise more slowly peaking around 2010 – 2013 and then also falling dramatically to 2019. Sea trout catches are about double those of salmon up until 2011 when the catches of both species become more equal and in the second graph you can see that the combined catch peaks between 2010 and 2013.
What is interesting is how the rod catches compare with the fish counts which are of course combined. I have read many times in books and magazine articles that typically around 10% of the fish that run up a river are caught on rod and line. I don’t know how this percentage was arrived at but like many things in angling it just seems to be one of those nuggets that has been accepted as fact. You would therefore expect the graph for the combined rod catch to roughly shadow the graph of the combined fish count but at a 90% lower level, but this is clearly not the case.
The first year of counts was 1994 but there was obviously a problem because the catches were higher than the counts.
I think the simple answer is that the fish counts are inaccurate and cannot be used as an indication of the number of fish running up the river.
Using rod catch data is fraught with danger due to human error/lies. As you touched on, it's hard to rely on peoples honestly, but could certainly be used as an indicator for sure.
1994 fish counts were only ran for 2 months so explain the error. 2004 & 2006 are also incomplete. So removing them shows a steady decline as per previous.
I would disagree that they are inaccurate, they have a margin of error as with everything, but resistivity fish counters have a 96% reliability or you can choose to believe human data on rod catch. Suppose it comes down to your agenda & whether you believe the decline is true or fabricated.