Last July I wrote about my misgivings over the Avon Roach Project (ARP) on the grounds that there appeared little evidence to suggest roach stocks were shrinking, apart from anglers’ anecdotes, and no reference to monitoring during the restocking years. I can find no reference to the Project on the Environment Agency website.
Recently I discovered a set of fish survey data on the gov.uk website containing observations that date back to the 1970s. So I’ve sifted through to see what it tells us about roach stocks on the Hampshire Avon. Surveys on the river only began in 2001. Counts can usually be taken as proportional to the population in the survey area. The sampling profile is highly variable, perhaps dependent on resources available at the time. In 2001 four sites on the main river were surveyed, but apparently only one in 2002 and 2003. This jumped to 26 in 2004 with a similar number in 2005, dropping back to four in 2006. Presumably the increased activity coincided with reports of poor roach catches from anglers that subsequently led to the ARP’s inception.
The graph below shows counts of roach averaged over the survey sites for that year.

There is a zero or missing reading for roach in 2002 and 2003, though very few fish of any species show in the EA figures for those two years. Roach counts pick up significantly in subsequent years but vary quite considerably over the 19-year span of the EA surveys, as would be expected for fish populations. If these counts are representative of the wider population, it is very unlikely that the ARP restocking project made any substantial difference over its lifetime. Its fish-rearing activity did not begin until 2007. At least two years would be required before any success could be measured. Yet the graph suggests stocks declined from 2008, with an upswing in 2011, then another fall. If 2011 was from the ARP stocking, the improvement was short-lived.
Care has to be taken when interpreting survey data. Possibly the counts during the low-sampling years before 2004 under-represent the stocks. Fish are not evenly spread throughout a river and survey results may be unrepresentative for this reason. But this is less likely with the more thorough sampling from 2004. Unless the survey data are completely unreliable, we should place little faith or money in restocking to restore fish populations that will recover in their own time.

