NEW INFORMATION UNDERPINS ASC’S INNOVATIVE APPROACH TO REDUCING SEA LICE TREATMENTS

When the salmon standard was originally drafted there was limited information available on how often salmon were treated to remove sea lice. ASC has more recently been able to collect a wide range of treatment information and it has been possible to design new ASC requirements based on the statistical analysis of this data. Records submitted to ASC as a requirement of farm certification, provided by the members of the Global Salmon Initiative and those available through government agencies were all used. To lead this work a multi-stakeholder Technical Working Group was established to provide advice into the ASC’s governance system, including on the feedback received from two public consultation rounds conducted in December 2015 and October 2017.

In the absence of information about current industry practice, the Parasiticide Treatment Index (PTI)  was devised and embedded in the original salmon standard. The resulting PTI calculation involved a number of assumptions related to the frequency of use, impact and toxicity of treatments out of necessity. The implications of these assumptions are explored (Payne, 2015) and the paper formed the basis of the first public consultation on the topic. The analysis of a comprehensive data set on the frequency of treatment use revealed three important contributions to the design of the revised PTI requirements.

First, there is a large variation in current farm performance between the various production countries with PTI-scores ranging from 0 in all countries to 400 in Scotland (p.4) while the current metric limit is set at 13. Eighty-two percent  of the observed variance can be attributed to the environmental differences between countries (p.6). This highlights the difficulty of devising a single global metric that would consistently incentivise the performance improvements required in the frequency of use of paraciticide treatments..

Secondly, PTI-performance also differs significantly between years (P<0.001; p.7) and this variance is attributed not to prior company performance but yearly variation in sea lice abundance.

Finally, the current PTI-calculation is silent on the use of treatments that target only infected pens within a farm; a practice increasingly used to target very early infestations of lice. This reduces intra-site amplification of lice loads, and reduces the overall need for subsequent treatments and the risk of developing treatment resistant sea lice. In the original PTI-calculation a partial treatment is considered equal to a farm-wide treatment (p.8), thereby overstating incorrectly the ‘calculated’ environmental impact of partial treatments and potentially discouraging their use.

The second analysis (Revie, 2018) reviewed country specific distributions of treatment frequency over 896 farm production cycles. Within this analysis details on the characteristics of the data used (p.4), the global and regional distribution curves (p.8-16) as well as on the availability and suitability of publicly available information (p.23-33) are presented. The key observations were:

1) Sea lice treatments differ significantly between the various production regions (p.12-15);

2) Identical percentiles (e.g. 20^th, 33^rd, 50^th) represent different number of treatments applied per region (p.12-15); and

3) Publicly available data sets have not yet reached the point of maturity at which they could operate as useful substitutes to the types of data that have been assembled as part of the TWG work (p.23-33).

In conclusion the TWG and the TAG concluded the need for the ASC to adopt a regional approach when establishing PTI limits.

On the basis of these analyses and recommendations the ASC board approved a number of changes to indicators 5.2.5 and 5.2.6 related to the PTI in the salmon standard.

Measurement of parasiticide load is calculated using the Weighted Number of Medicinal Treatments (WNMT);

Partial treatments are counted as proportional to the ratio of treated pens to the total number of pens within the farm;

The Global Level is set at 3 WNMT, except for cases when a double bath treatment is applied. In which case the level would be set at 4 WNMT;

A Regional Entry Level is based on the 50^th percentile of the relevant regional distribution curve obtained from the second statistical study;

A 25% reduction in WNMT is required by farms over a two-year period until the GL is achieved; and

That farms shall be required to make an IPM-plan publicly available, and that the plan should be signed off by a relevant professional (e.g. farm veterinarian).

Notes:

1. WNMT includes treatments with hydrogen peroxide. These were not previously counted within the PTI index.

2. The previous PTI metric established a global level of approximately 2 treatments, which is less than the 20^th percentile observed for all regional distributions other than Canada. The proposed Global Level falls between the 20^th – 33^rd percentile for the global distribution.

3. Double bath treatments are necessary for some chemical to maintain the efficacy of the treatment. The second bath treatment is undertaken a few days after the first bath.

4. The 50^th percentile for each regional distribution curve results in the following Entry Level: Canada (1), Chile (9), Faroe Islands (6), Ireland (3), Norway (5), Scotland (9).

5. Given the skewed treatment frequency distributions observed an entry gate extends the reach of the incentives created by the ASC programme and will engage more farms within a defined improvement journey.