Genomic survey of edible cockle (Cerastoderma edule) in the Northeast Atlantic: a baseline for sustainable management of its wild resources

Knowledge on how environmental factors shape the genome of marine species is crucial for sustainable management of fisheries and wild populations. The edible cockle (Cerastoderma edule) is a marine bivalve distributed along the Northeast Atlantic coast of Europe and is an important resource from both commercial and ecological perspectives. We performed a population genomics screening using 2b-RAD genotyping on 9,309 SNPs localised in the cockle's genome on a sample of 536 specimens pertaining to 14 beds in the Northeast Atlantic Ocean to determine the genetic structure with regard to environmental variables. Larval dispersal modelling considering species behaviour and interannual / interseasonal variation in ocean conditions was carried out as an essential background to which compare genetic information. Cockle populations in the Northeast Atlantic displayed low but significant geographical differentiation between populations (FST = 0.0240; P < 0.001), albeit not across generations. We identified 742 and 36 outlier SNPs related to divergent and balancing selection in all the geographical scenarios inspected, and sea temperature and salinity were the main environmental drivers suggested. Highly significant linkage disequilibrium was detected at specific genomic regions against the very low values observed across the whole genome, suggestive of selective sweeps. Two main genetic groups were identified, northwards and southwards of French Brittany, in accordance with the larval dispersal modelling, which suggested a barrier for larval dispersal linked to the Ushant front. Further genetic subdivision was observed using outlier loci and considering larval behaviour. The northern group was divided into the Irish/Celtic Seas and the English Channel/North Sea, while the southern group was divided into three subgroups. This information represents the baseline for management of cockles, designing conservation strategies, founding broodstock for depleted beds, and producing suitable seed for aquaculture production.

Methods
Single Nucleotide Polymorphism (SNP) genotyping

Total DNA was extracted from gill tissue samples using the e.Z.N.A. E-96 mollusc DNA kit (OMEGA Bio-tech), following manufacturer recommendations. SNP identification and selection, as well as genotyping and validation protocols followed those described by Maroso et al. (2019). Briefly, AlfI IIb restriction enzyme (RE) was used to construct the 2b-RAD libraries, which were evenly pooled for sequencing in Illumina Next-seq including 90 individuals per run. The recently assembled cockle's genome (794 Mb; Bruzos et al., unpublished data) was used to align reads from each individual using Bowtie 1.1.2 (Langmead et al., 2009), allowing a maximum of three mismatches and a unique valid alignment (-v 3 -m 1). Individuals with < 250,000 reads were discarded. STACKS 2.0 (Catchen et al. 2013) was then used to call SNPs and genotype a common set of markers in the sample set, applying the marukilow model with default parameters in the gstacks module of Stacks 2.0. This SNP panel was further filtered by applying the following criteria: i) genotyped in > 60% individuals in the total sample; ii) minimum allele count (MAC) ≥ 3 in the total sample; iii) conformance to Hardy-Weinberg equilibrium within each sample (HWE) across the whole collection; i.e. loci with significant deviation from HWE (P < 0.05) in more than 25% of samples were removed; and iv) selection of the most polymorphic SNP in each RAD-tag.