Evaluation of the shucking of certain species of scallops contaminated with domoic acid with a view to the production of edible parts meeting the safety requirements foreseen in the Union legislation

Abstract

EFSA was asked by the European Commission to provide information on the levels of domoic acid (DA) in whole scallops that would ensure that levels in edible parts are below the regulatory limit after shucking. This should include five species of scallops. In addition, EFSA was asked to recommend the number of scallops to be used in an analytical sample. To address these questions, EFSA received suitable data on DA for only one scallop species, Pecten maximus, i.e. data on pooled samples of edible and non-edible parts. A large part of the concentration levels was above the limit of quantification (LOQ) and only these data were used for the assessment. Shucking in most cases resulted in a strong decrease in the toxin levels. Statistical analysis of the data showed that levels in whole scallops should not exceed 24 mg DA/kg, 59 mg DA/kg and 127 mg DA/kg to ensure that levels in, respectively, gonads, muscle and muscle plus gonads are below the regulatory limit of 20 mg DA/kg with 99% certainty. Such an analysis was not possible for the other scallop species. In the absence of data from member states, published data of variations between scallops were used to calculate the sample size to ensure a 95% correct prediction on whether the level in scallops in an area or lot is correctly predicted to be compliant/non-compliant. It was shown that 10 scallops per sample would be sufficient to predict with 95% certainty if DA levels in the area/lot were twofold below or above the regulatory limit for the highest reported coefficient of variance (CV) of 1.06. To predict with 95% certainty for levels between 15 and 27 mg DA/kg, a pooled sample of more than 30 scallops would have to be tested.

Keywords: Domoic acid; Pecten maximus; sample size; scallops; shucking.

Figure 1

Anatomy of a scallop (simplified)


1. The term ‘gonad’ is used to refer to the anatomically distinct reproductive organ sometimes also referred to as ‘roe’. Physiologically, this structure is involved in reproduction and also contains a loop of the intestine, which is significant in terms of toxin concentration for the structure. The term ‘gonad’ is used in this Opinion to refer to the structure as recognised macroscopically and commercially harvested, therefore also including some of the intestine.

Figure 2

Structure of domoic acid and epi‐domoic acid

Figure 3

Scatter plot of concentration of domoic acid at the lower bound in (a) ‘gonad’ and (b) ‘adductor muscle’ and (c) ‘muscle plus gonad’ vs. concentration in entire body as reported by Ireland for Pecten maximus (pooled)

Figure 4

Scatter plot of concentrations of domoic acid in ‘adductor muscle and gonad’ versus entire animal, in animals belonging to the same lot, as submitted by France

Figure 5

Scatter plot for the lower bound data submitted by France and Ireland together with their corresponding robust regression fits and confidence interval

Figure 6

Number of samples required for a pooled analytical sample to ensure with 95% certainty a correct conclusion on either compliance (if level below the regulatory limit) or noncompliance (if level above the regulatory limit), at a given concentration in the area or lot


1. Numbers above 30 were not estimated and are presented as 30. Coefficients of variance (CVs) are based on data reported in the literature for domoic acid levels in hepatopancreas (0.40, 0.70 and 0.83) and adductor muscle (1.06) of Pecten maximus.

Figure A.1

Distribution of the ratios between the levels in edible parts (muscle plus gonad) and whole scallop for pooled scallops from the same lot submitted by France, for each sample ranging between the ratio based on lower bound and those based on upper bound levels


1. CDF: cumulative distribution function.

Figure A.2

Distribution of the ratios between the levels in edible parts (muscle plus gonad) and whole scallop for pooled scallops from the same lot submitted by France, for each sample ranging between the ratio based on LB levels and those based on UB levels


1. CDF: cumulative distribution function.

Figure A.3

Distribution of the ratios between the levels in edible parts (muscle plus gonad) and whole scallop for pooled scallops from Ireland, for each sample ranging between the ratio based on lower bound levels and those based on upper bound levels


1. CDF: cumulative distribution function.

Figure A.4

Distribution of the ratios between the levels in edible parts (muscle plus gonad) and whole scallop for pooled scallops from Ireland, for each sample ranging between the ratio based on lower bound and those based on upper bound levels


1. CDF: cumulative distribution function.

Figure A.5

Distribution of the ratios for data from France and Ireland

Figure A6

Scatter plot for the lower bound data submitted by France and Ireland together with their corresponding robust regression fits and confidence interval

Figure A.7

Distribution of the ratios between the levels in muscle and the whole scallop for pooled scallops obtained from Ireland, for each sample ranging between the ratio based on LB levels and those based on UB levels


1. CDF: cumulative distribution function.

Figure A.8

Distribution of the ratios between the levels in muscle and whole scallop for pooled scallops from Ireland, for those samples in which the LB levels are equal to the UB levels


1. CDF: cumulative distribution function.

Figure A.9

Distribution of the ratios for data from Ireland considering all observations (ALL) and only those where the lower bound is larger than zero (Reduced)

Figure A.10

Distribution of the ratios between the levels in gonad and whole scallop for pooled scallops obtained from Ireland, for each sample ranging between the ratio based on lower bound and those based on upper bound levels


1. CDF: cumulative distribution function.

Figure A.11

Distribution of the ratios between the levels in gonad and whole scallop for pooled scallops from Ireland, for samples where the LB levels are equal to the UB levels


1. CDF: cumulative distribution function.

Figure A.12

Distribution of the ratios for the data from Ireland considering all observations (ALL) and only those where the lower bound is larger than zero (Reduced)