FIRST ESTIMATIONS OF GENETIC PARAMETERS FOR RESISTANCE TO DISEASE AND PRODUCTION TRAITS IN THE MANILA CLAM
Bivalve aquaculture is a rapidly growing sector that already represents a significant part of global aquaculture production, and while open-water grow-out techniques allow producers to benefit from a natural supply of food and oxygen, this method also exposes bivalves to potential pathogens. Diseases in represent a major limitation for bivalve production, namely reduced growth and reproduction, and mass mortality episodes, all of which are exacerbated in the case of shellfish by the difficulty in treating these diseases using the techniques which have been put in place (with variable success) for finfish species.
Genetic selection to improve desirable traits for aquaculture is relatively new in shellfish species, and it represents a sustainable alternative for mitigating the impact of diseases in farmed bivalve species. In order to balance selective breeding across multiple traits, phenotypes for traits of interest are recorded in a large number of related individuals and genetic parameters are calculated across traits, making it possible to estimate heritability for and genetic correlations between growth-related traits and resistance to disease.
A panel of 245 genetic markers (SNPs) was designed and mounted onto an Illumina SNP-chip and used assign parentage of two batches of F1 Manila clams from a single cohort produced from 109 parents using mass spawning. The cohort was sent to two production sites for grow-out in France and Italy for at least one year to assess the genotype by environment (GxE) interaction on production traits. Sampling (>1000 individuals per site) was carried out at approximately 2.5 years of age to record phenotypes (resistance, growth traits, sex) and collect DNA samples for parentage assignment. F1 samples were genotyped and assigned to their parent pairs using AccurAssign software. Despite limited assignment efficiency (54% and 64%), heritability for traits of commercial interest were 0.52 for resistance to Perkinsosis, and 0.23 and 0.46 for total weight in Italy and France, respectively. In addition, genetic correlations between biometric traits and resistance suggest that selecting for growth traits and disease resistance will not negatively affect other traits. The results of this first estimation of genetic parameters in the Manila clam in a mixed family breeding design indicate that developing a selective breeding program for growth traits and/or resistance will be possible without artificial fertilization methods and beneficial to the production industry, though the assignment rate of the current panel should be improved to increase accuracy of the current estimations.
Genetic selection has long since revolutionized the efficiency of terrestrial plant and livestock production, in mollusk aquaculture, however, selection for the improvement of desirable traits remains poorly exploited. Bivalve aquaculture represents a rapidly growing aquaculture sector with high potential for expansion, having already increased its production at a rate of 7% between 2000 and 2014 [1,2].
Large-scale breeding programs for fish weren’t initiated until the 1970s . Since then, a number of breeding programs have been started for fish and shellfish, many of which have obtained high genetic gains due to a combination of high heritability for economically favorable traits, high fecundity, and short generation time of many of the studied species [4,5]. Despite the promising results in genetic gain for aquaculture species, it was estimated that in 2010 only 8.2 % of aquaculture production was based on genetically improved stocks .
Infectious diseases continue to be a major limit to aquaculture production, leading to the development of a number of techniques to control the spread of disease including vaccination, antibiotics, and cleaning methods in fish. However, pathogens continue to be a major limiting factor for aquaculture production, costing over US$ 6 billion per annum in losses alone, and far more in treatment [7,8]. While several pathogens have been reported to infect the economically relevant Manila clam, Ruditapes philippinarum, the protozoan parasite Perkinsus olseni has been identified as a major pathogen responsible for mass mortality of this species in Europe, namely in Italy . The parasite induces the formation of nodules within the clam’s gills that gradually spread to other tissues. In advanced stages of the disease, P. olseni causes lesions throughout various tissues that negatively affect the respiration and other physiological processes of the clam, leading to a reduction in growth, reproduction, and condition index . While this pathogen is consistently present in clam production zones such as the Venice Lagoon in Italy, with disease prevalence regularly reaching 80-100% in clams, uninfected Manila clams are detected in exposed populations during routine monitoring carried out by the Italian health authority and research organization for animal health and food safety (IZSVe). It should also be noted that mortality is not always observed within the two-year grow-out period typically seen in Italy, as the parasite induces a chronic infection leading to varying degrees of infection throughout a given clam population.
Shellfish lack the adaptive immune system necessary for classical vaccination, and the open-water systems in which they are cultured make the use of antibiotics understandably controversial. Genetic selection for disease resistance in fish and shellfish, which is already in place for certain species, represents a feasible and long-term alternative to the current treatment solutions. Recent estimates have shown that resistance could be improved by as much as 15% per generation in mollusks by implementing individual or family-based selection . The few existing breeding programs in bivalve hatcheries are recent and mostly small-scale (few selected lines and limited broodstock size). Another significant bottleneck is the lack of genomic resources for most bivalve species, namely the Manila clam, which make it difficult to implement breeding programs despite having clearly defined objectives. Although Manila clam aquaculture represents roughly a quarter of all cultured bivalve species produced worldwide, European production currently accounts for only about 2% . Most of the studies carried out so far seeking to evaluate the potential for genetic gain in bivalves have focused on traits of commercial interest, such as disease resistance, growth rate, meat yield, and visual aspects such as shape and shell color, and even fewer have tackled the estimation of genetic parameters across all traits simultaneously . Balancing genetic gain for disease resistance with genetic gain for production traits is an essential step in designing rational breeding programs in order to evaluate the potential for genetic improvement across traits and, more importantly, to avoid selecting for a trait that may negatively affect other traits of interest.
Pedigree information is essential to evaluate the expected breeding values for measurable traits and develop sustainable breeding programs, however most aquatic species are impossible to physically tag at the larval stage due to their small size, which makes pedigree information difficult and costly to obtain. In order to distinguish between families, it was until recently unavoidable to raise large numbers of them separately, a logistical issue has long been among the many limits for the development of optimized breeding schemes in aquatic species. Individual tagging and separate family rearing also come with major disadvantages including high costs, extensive infrastructure, and the risk of confounding factors such as a “batch effect” which can significantly skew the accuracy of the estimated breeding values.
The use of genetic markers to trace the relationships between individuals within a large number of families greatly enhances the efficiency of family-based breeding schemes . Furthermore, as genotyping costs continue to decrease it is becoming increasingly feasible to carry out large-scale studies with dense genotypic marker panels, allowing for more an accurate prediction of breeding values . Bivalve genomes are notoriously difficult to assemble due to particularly high rates of heterozygosity, large insertions, and complex structural variation . Elevated levels of polymorphism, namely de novo mutations, can lead to Mendelian errors and null alleles which make it difficult to reconstruct pedigree through genetic analyses [17– 19]. Notwithstanding, several recent studies have successfully applied SNP panels for parentage assignment in order to estimate genetic parameters of traits of commercial interest in mollusks [20–23]. A 245-SNP panel was conceived for parentage assignment in the Manila clam. In addition, a number of programs have been developed to be able to integrate phenotypic, pedigree and genomic information in prediction of breeding values, such as the BLUPF90 program family, under which fall several programs for variance component estimation such as GIBBSfF90 and THRGIBBSxf90 [24,25].
The objective of this study is to carry out a first estimation of the genetic parameters of production traits and resistance to infection with P. olseni in the commercially important Manila clam, R. philippinarum, in order to evaluate the feasibility of carrying out rational genetic selection of multiple production traits and lay the basis for designing balanced breeding programs in hatcheries.
MATERIALS AND METHODS
An experimental group of Manila clam R. philippinarum was produced by mass spawning in May 2016 at the French hatchery SATMAR (Barfleur, France) with 109 adult clams of hatchery broodstock, according to a two-factorial mating design: 31 dams x 25 sires (775 putative families) and 22 dams x 32 sires (704 putative families), providing potentially 1479 families. A piece of mantle from each of the 109 parent clams was kept in 99% ethanol for genotyping.
Sampling strategy and data collection
In Chioggia, prevalence and intensity of P. olseni in the gill tissues of clams was evaluated monthly from the time of seeding during one year. Each month, about 30 clams were sampled for whole weight, shell weight, and shell length, and the gills were dissected and frozen in liquid nitrogen for subsequent DNA extraction. Biometric measurements allowed us to follow growth parameters of the clams. Disease intensity was measured based on quantification of parasite DNA in the total DNA from the gill sample (protocol described below), and prevalence was considered as the percentage of quantifiably infected clams. This monthly sampling strategy was employed in order to monitor disease prevalence, bearing in mind the annual variability of disease intensity and the historical risk of mass mortality in the area. Mass sampling was carried out at commercial size, when about 50% of the group showed quantifiable levels of infection. In Chioggia, the mass sampling event was carried out in September 2018, and then in November 2018 for the Marennes batch. Mass sampling entailed gathering the measurements for biometric traits in both sites (Table 2). The number of samples was set at n > 1 000 in order to expect a mean of 18-20 progenies per parent, for a relevant estimation of genetic parameters in a mixed family design which accounts for a variable number of sibs per parent . based on the estimations described by Dupont-Nivet et al. (2006) who demonstrated effective population size (Ne) and genetic variability to be maintained in a total population size of n = 1 000 for full factorial mating designs considering 50 x 50 parents . Over 1 000 clams at each site were individually weighed, and length, width and height were recorded, as depicted in Figure 1 . Then, meat and shell were separated and weighted after draining. Sex was determined by visualization under a microscope of a gonad smear during Chioggia sampling. For Marennes clams it was not possible to determine sex as individuals were not mature due to lower water temperatures at the time of sampling.
Both gills were collected from each of the Chioggia clams and a piece of mantle tissue was collected from the Marennes clams. Gills were selected for the samples from Chioggia as they are the first tissue to become infected with P. olseni, and thus it was possible to carry out total DNA extraction from the same sample for both parasite quantification and host genotyping (see below: Perkinsus olseni quantification). Mantle tissue was preferred for Marennes clams due to ease of sampling. Gill samples were stored in 70% ethanol, and mantle sample in 99% ethanol until DNA extraction (see below: DNA extraction and genotyping).
Table of contents :
ORIGIN AND TRANSFER OF THE MANILA CLAM
AQUACULTURE AND GENETIC RESOURCES
GENETIC BACKGROUND OF MANILA CLAM POPULATIONS
TRANSMISSIBLE DISEASES IN THE MANILA CLAM
THE PROTOZOAN PARASITE, PERKINSUS OLSENI
THE BACTERIAL AGENT OF BROWN RING DISEASE, VIBRIO TAPETIS
SELECTION FOR RESISTANCE TO DISEASE IN BIVALVES
OVERALL OBJECTIVES AND APPROACH
CHAPTER 1: DEVELOPMENT OF A SNP PANEL FOR PARENTAGE ASSIGNMENT IN THE MANILA CLAM
MATERIALS AND METHODS
CHAPTER 2: FIRST ESTIMATIONS OF GENETIC PARAMETERS FOR RESISTANCE TO DISEASE AND PRODUCTION TRAITS IN THE MANILA CLAM
MATERIALS AND METHODS
CHAPTER 3: A PROTEOMIC STUDY OF RESISTANCE TO BROWN RING DISEASE IN THE MANILA CLAM, RUDITAPES PHILIPPINARUM.
MATERIALS AND METHODS
RESULTS AND DISCUSSION
GENERAL DISCUSSION AND CONCLUSIONS