Fish farms underreporting sea lice counts

Roy L Hales/ Cortes Currents - “Our results suggest that sea-louse counts reported by the salmon farming industry are lower than the true abundance of parasites on their fish. When the federal government audited a sea-louse count, the industry’s mean counts for that month increased by a factor of 1.18 for L. salmonis and by 1.95 for C. clemensi.” - Sean Godwin et al, Bias in self-reported parasite data from the salmon farming industry.

Dr Sean Godwin is the lead author of eight of the fifteen scientific papers listed on his website. The bulk of these explore interactions between wild and farmed salmon. One of the most troubling, published in the journal Ecological Application, is a survey that showed Fish farms underreporting sea lice during the months they are not audited by Fisheries and Oceans Canada.

“The period of study in that paper was 2011 to 2016. It involved every salmon farm that operated in British Columbia during that time. What we found was that when DFO Auditors were there to check in on the salmon farm counts, the industry counters were higher …,” he said.

Godwin said that while this is not true in every single case, it is the overall trend.

“Once we took into account all these other things that upset sea lice numbers - like temperature, salinity, location and all these things - there was this really obvious effect across the board. When DFO Auditors were there, the accounts were higher and when DFO Auditors were not here, the counts were lower.”

The BC Salmon farmer’s Association emailed a previously published statement, “ …The report by Sean Godwin et. al is presented as a direct comparison of sea lice data collected by salmon farmers and submitted to the DFO with audits conducted by the DFO. However, that’s not what it actually does. Rather, the study’s authors created a complex model themselves to try and estimate what sea lice counts should have been in the past, and report on variance from their own model to actual sea lice counts. The model feels rushed, and is based on assumptions that aren’t clear in the study and don’t reflect a number of complex variables including the reality of ocean conditions.”

To which Godwin responded, also by email,“The same trend we identified is present in the raw, publicly-available count data (i.e. -absent any modelling). The average industry louse counts are still ~15% less for L. salmonis and ~50% for C. clemensi when auditors are not present.”

“ … This analysis began in 2017 as part of my PhD thesis, and has taken the better part of 3.5 years to complete to our satisfaction and go through the thorough peer-review process in a leading scientific journal (to which we first submitted the manuscript last October).”

“ … Our study does not compare industry counts during audits to DFO audit counts, nor do we ever present it as doing so – we know DFO already does this … We were instead curious about what happens to industry's own counts when auditors are not present. Our use of a statistical model (fit to real industry counts) was designed to account for factors known to affect sea lice (e.g., treatments, seasonality, etc). Our use of a statistical model was in-part an honest attempt to find explanations other than industry bias.”

There is much more in the podcast

Photo credit: Tavish Campbell

Photo