Effects of species, culture history, size and residency on relative competitive ability of salmonids

The relative competitive ability of juvenile farm and wild salmonids was investigated to provide insight into the potential effects of introduction of cultured salmon on wild Pacific salmonid ( Oncorhynchus ) species. Aquarium experiments involving equal contests ( i.e. size matched, simultaneously introduced individuals) indicated that two wild coho salmon Oncorhynchus kisutch populations were competitively equal to a farm coho salmon population. In equal contests between farm Atlantic salmon Salmo salar (Mowi strain) and these wild coho salmon populations or coastal cutthroat trout Oncorhynchus clarki clarki , Atlantic salmon were subordinate in all cases. When Atlantic salmon were given a residence advantage, however, they were competitively equal to both wild coho salmon populations, but remained subordinate to coastal cutthroat trout. When Atlantic salmon were given a 10–30% length advantage, they were competitively equal to one wild coho salmon population but remained subordinate to the other. In equal contests in semi-natural stream channels, both wild coho and farm Atlantic salmon grew significantly more in the presence of the other species than when alone. It appears that coho salmon obtain additional food ration by out competing Atlantic salmon, whereas Atlantic salmon were stimulated to feed more in the presence of coho salmon competitors. These results suggest that wild coho salmon and cutthroat trout should out compete farm Atlantic salmon of a similar size in nature. As the relative competitive ability of Atlantic salmon improves when they have a size and residence advantage, should feral populations become established, they may exist on a more equal competitive footing owing to the long freshwater residence of Atlantic salmon.     

Genetic and phenotypic changes in an Atlantic salmon population supplemented with non-local individuals: a longitudinal study over 21 years

While introductions and supplementations using non-native and potentially domesticated individuals may have dramatic evolutionary effects on wild populations, few studies documented the evolution of genetic diversity and life-history traits in supplemented populations. Here, we investigated year-to-year changes from 1989 to 2009 in genetic admixture at 15 microsatellite loci and in phenotypic traits in an Atlantic salmon (Salmo salar) population stocked during the first decade of this period with two genetically and phenotypically distinct source populations. We detected a pattern of temporally increasing introgressive hybridization between the stocked population and both source populations. The proportion of fish returning to the river after a single winter at sea (versus several ones) was higher in fish assigned to the main source population than in local individuals. Moreover, during the first decade of the study, both single-sea-winter and multi-sea-winter (MSW) fish assigned to the main source population were smaller than local fish. During the second decade of the study, MSW fish defined as hybrids were lighter and smaller than fish from parental populations, suggesting outbreeding depression. Overall, this study suggests that supplementation with non-local individuals may alter not only the genetic diversity of wild populations but also life-history traits of adaptive significance.     

A standardized method for quantifying unidirectional genetic introgression

Genetic introgression of domesticated to wild conspecifics is of great concern to the genetic integrity and viability of the wild populations. Therefore, we need tools that can be used for monitoring unidirectional gene flow from domesticated to wild populations. A challenge to quantitation of unidirectional gene flow is that both the donor and the recipient population may be genetically substructured and that the subpopulations are subjected to genetic drift and may exchange migrants between one another. We develop a standardized method for quantifying and monitoring domesticated to wild gene flow and demonstrate its usefulness to farm and wild Atlantic salmon as a model species. The challenge of having several wild and farm populations was circumvented by in silico generating one analytical center point for farm and wild salmon, respectively. Distributions for the probability that an individual is wild were generated from individual‐based analyses of observed wild and farm genotypes using STRUCTURE. We show that estimates of proportions of the genome being of domesticated origin in a particular wild population can be obtained without having a historical reference sample for the same population. The main advantages of the method presented are the standardized way in which genetic processes within and between populations are taken into account, and the individual‐based analyses giving estimates for each individual independent of other individuals. The method makes use of established software, and as long as genetic markers showing generic genetic differences between domesticated and wild populations are available, it can be applied to all species with unidirectional gene flow. Results from our method are easy to interpret and understand, and will serve as a powerful tool for management, especially because there is no need for a specific historical wild reference sample.     

Contrasting patterns of gene diversity between microsatellites and mitochondrial SNPs in farm and wild Atlantic salmon

Levels of genetic variability at 12 microsatellite loci and 19 single nucleotide polymorphisms in mitochondrial DNA were studied in four farm strains and four wild populations of Atlantic salmon. Within populations, the farm strains showed significantly lower allelic richness and expected heterozygosity than wild populations at the 12 microsatellite loci, but a significantly higher genetic variability with respect to observed number of haplotypes and haplotype diversity in mtDNA. Significant differences in allele- and haplotype-frequencies were observed between farm strains and wild populations, as well as between different farm strains and between different wild populations. The large genetic differentiation at mitochondrial DNA between wild populations (F_ST = 0.24), suggests that the farm strains attained a high mitochondrial genetic variability when created from different wild populations seven generations ago. A large proportion of this variability remains despite an expected lower effective population size for mitochondrial than nuclear DNA. This is best explained by the particular mating schemes in the breeding programmes, with 2–4 females per male. Our observations suggest that for some genetic polymorphisms farm populations might currently hold equal or higher genetic variability than wild populations, but lower overall genetic variability. In the short-term, genetic interactions between escaped farm salmon and wild salmon might increase genetic variability in wild populations, for some, but not most, genetic polymorphisms. In the long term, further losses of genetic variability in farm populations are expected for all genetic polymorphisms, and genetic variability in wild populations will be reduced if escapes of farm salmon continue.     

Determinants of hierarchical genetic structure in Atlantic salmon populations: environmental factors vs. anthropogenic influences

Disentangling the effects of natural environmental features and anthropogenic factors on the genetic structure of endangered populations is an important challenge for conservation biology. Here, we investigated the combined influences of major environmental features and stocking with non‐native fish on the genetic structure and local adaptation of Atlantic salmon (Salmo salar) populations. We used 17 microsatellite loci to genotype 975 individuals originating from 34 French rivers. Bayesian analyses revealed a hierarchical genetic structure into five geographically distinct clusters. Coastal distance, geological substrate and river length were strong predictors of population structure. Gene flow was higher among rivers with similar geologies, suggesting local adaptation to geological substrate. The effect of river length was mainly owing to one highly differentiated population that has the farthest spawning grounds off the river mouth (up to 900 km) and the largest fish, suggesting local adaptation to river length. We detected high levels of admixture in stocked populations but also in neighbouring ones, implying large‐scale impacts of stocking through dispersal of non‐native individuals. However, we found relatively few admixed individuals suggesting a lower fitness of stocked fish and/or some reproductive isolation between wild and stocked individuals. When excluding stocked populations, genetic structure increased as did its correlation with environmental factors. This study overall indicates that geological substrate and river length are major environmental factors influencing gene flow and potential local adaptation among Atlantic salmon populations but that stocking with non‐native individuals may ultimately disrupt these natural patterns of gene flow among locally adapted populations.     

Quantifying heritable variation in fitness-related traits of wild,farmed and hybrid Atlantic salmon families in a wild river environment

Farmed fish are typically genetically different from wild conspecifics. Escapees from fish farms may contribute one-way gene flow from farm to wild gene pools, which can depress population productivity, dilute local adaptations and disrupt coadapted gene complexes. Here, we reanalyse data from two experiments (McGinnity et al., 1997, 2003) where performance of Atlantic salmon (Salmo salar) progeny originating from experimental crosses between farm and wild parents (in three different cohorts) were measured in a natural stream under common garden conditions. Previous published analyses focussed on group-level differences but did not account for pedigree structure, as we do here using modern mixed-effect models. Offspring with one or two farm parents exhibited poorer survival in their first and second year of life compared with those with two wild parents and these group-level inferences were robust to excluding outlier families. Variation in performance among farm, hybrid and wild families was generally similar in magnitude. Farm offspring were generally larger at all life stages examined than wild offspring, but the differences were moderate (5–20%) and similar in magnitude in the wild versus hatchery environments. Quantitative genetic analyses conducted using a Bayesian framework revealed moderate heritability in juvenile fork length and mass and positive genetic correlations (>0.85) between these morphological traits. Our study confirms (using more rigorous statistical techniques) previous studies showing that offspring of wild fish invariably have higher fitness and contributes fresh insights into family-level variation in performance of farm, wild and hybrid Atlantic salmon families in the wild. It also adds to a small, but growing, number of studies that estimate key evolutionary parameters in wild salmonid populations. Such information is vital in modelling the impacts of introgression by escaped farm salmon.     

Identification of quantitative genetic components of fitness variation in farmed,hybrid and native salmon in the wild

Feral animals represent an important problem in many ecosystems due to interbreeding with wild conspecifics. Hybrid offspring from wild and domestic parents are often less adapted to local environment and ultimately, can reduce the fitness of the native population. This problem is an important concern in Norway, where each year, hundreds of thousands of farm Atlantic salmon escape from fish farms. Feral fish outnumber wild populations, leading to a possible loss of local adaptive genetic variation and erosion of genetic structure in wild populations. Studying the genetic factors underlying relative performance between wild and domesticated conspecific can help to better understand how domestication modifies the genetic background of populations, and how it may alter their ability to adapt to the natural environment. Here, based upon a large-scale release of wild, farm and wild x farm salmon crosses into a natural river system, a genome-wide quantitative trait locus (QTL) scan was performed on the offspring of 50 full-sib families, for traits related to fitness (length, weight, condition factor and survival). Six QTLs were detected as significant contributors to the phenotypic variation of the first three traits, explaining collectively between 9.8 and 14.8% of the phenotypic variation. The seventh QTL had a significant contribution to the variation in survival, and is regarded as a key factor to understand the fitness variability observed among salmon in the river. Interestingly, strong allelic correlation within one of the QTL regions in farmed salmon might reflect a recent selective sweep due to artificial selection.     

Using Agent-Based Modelling to Predict the Role of Wild Refugia in the Evolution of Resistance of Sea Lice to Chemotherapeutants

A major challenge for Atlantic salmon farming in the northern hemisphere is infestation by the sea louse parasite Lepeophtheirus salmonis. The most frequent method of controlling these sea louse infestations is through the use of chemical treatments. However, most major salmon farming areas have observed resistance to common chemotherapeutants. In terrestrial environments, many strategies employed to manage the evolution of resistance involve the use of refugia, where a portion of the population is left untreated to maintain susceptibility. While refugia have not been deliberately used in Atlantic salmon farming, wild salmon populations that migrate close to salmon farms may act as natural refugia. In this paper we describe an agent-based model that explores the influence of different sizes of wild salmon populations on resistance evolution in sea lice on a salmon farm. Using the model, we demonstrate that wild salmon populations can act as refugia that limit the evolution of resistance in the sea louse populations. Additionally, we demonstrate that an increase in the size of the population of wild salmon results in an increased effect in slowing the evolution of resistance. We explore the effect of a population fitness cost associated with resistance, finding that in some cases it substantially reduces the speed of evolution to chemical treatments.     

Changes in the genetic structure of Atlantic salmon populations over four decades reveal substantial impacts of stocking and potential resiliency

While the stocking of captive‐bred fish has been occurring for decades and has had substantial immediate genetic and evolutionary impacts on wild populations, its long‐term consequences have only been weakly investigated. Here, we conducted a spatiotemporal analysis of 1428 Atlantic salmon sampled from 1965 to 2006 in 25 populations throughout France to investigate the influence of stocking on the neutral genetic structure in wild Atlantic salmon (Salmo salar) populations. On the basis of the analysis of 11 microsatellite loci, we found that the overall genetic structure among populations dramatically decreased over the period studied. Admixture rates among populations were highly variable, ranging from a nearly undetectable contribution from donor stocks to total replacement of the native gene pool, suggesting extremely variable impacts of stocking. Depending on population, admixture rates either increased, remained stable, or decreased in samples collected between 1998 and 2006 compared to samples from 1965 to 1987, suggesting either rising, long‐lasting or short‐term impacts of stocking. We discuss the potential mechanisms contributing to this variability, including the reduced fitness of stocked fish and persistence of wild locally adapted individuals.     

LANDSCAPE GENOMICS IN ATLANTIC SALMON (SALMO SALAR): SEARCHING FOR GENE–ENVIRONMENT INTERACTIONS DRIVING LOCAL ADAPTATION

A growing number of studies are examining the factors driving historical and contemporary evolution in wild populations. By combining surveys of genomic variation with a comprehensive assessment of environmental parameters, such studies can increase our understanding of the genomic and geographical extent of local adaptation in wild populations. We used a large‐scale landscape genomics approach to examine adaptive and neutral differentiation across 54 North American populations of Atlantic salmon representing seven previously defined genetically distinct regional groups. Over 5500 genome‐wide single nucleotide polymorphisms were genotyped in 641 individuals and 28 bulk assays of 25 pooled individuals each. Genome scans, linkage map, and 49 environmental variables were combined to conduct an innovative landscape genomic analysis. Our results provide valuable insight into the links between environmental variation and both neutral and potentially adaptive genetic divergence. In particular, we identified markers potentially under divergent selection, as well as associated selective environmental factors and biological functions with the observed adaptive divergence. Multivariate landscape genetic analysis revealed strong associations of both genetic and environmental structures. We found an enrichment of growth‐related functions among outlier markers. Climate (temperature–precipitation) and geological characteristics were significantly associated with both potentially adaptive and neutral genetic divergence and should be considered as candidate loci involved in adaptation at the regional scale in Atlantic salmon. Hence, this study significantly contributes to the improvement of tools used in modern conservation and management schemes of Atlantic salmon wild populations.     

Lifetime success and interactions of farm salmon invading a native population

Farm Atlantic salmon escape and invade rivers throughout the North Atlantic annually, which has generated growing concern about their impacts on native salmon populations. A large-scale experiment was therefore undertaken in order to quantify the lifetime success and interactions of farm salmon invading a Norwegian river. Sexually mature farm and native salmon were genetically screened, radio tagged and released into the River Imsa where no other salmon had been allowed to ascend. The farm fishes were competitively and reproductively inferior, achieving less than one-third the breeding success of the native fishes. Moreover, this inferiority was sex biased, being more pronounced in farm males than females, resulting in the principal route of gene flow involving native males mating with farm females. There were also indications of selection against farm genotypes during early survival but not thereafter. However, evidence of resource competition and competitive displacement existed as the productivity of the native population was depressed by more than 30%. Ultimately, the lifetime reproductive success (adult to adult) of the farm fishes was 16% that of the native salmon. Our results indicate that such annual invasions have the potential for impacting on population productivity, disrupting local adaptations and reducing the genetic diversity of wild salmon populations.     

Fitness reduction and potential extinction of wild populations of Atlantic salmon, Salmo salar, as a result of interactions with escaped farm salmon

The high level of escapes from Atlantic salmon farms, up to two million fishes per year in the North Atlantic, has raised concern about the potential impact on wild populations. We report on a two-generation experiment examining the estimated lifetime successes, relative to wild natives, of farm, F(1) and F(2) hybrids and BC(1) backcrosses to wild and farm salmon. Offspring of farm and "hybrids" (i.e. all F(1), F(2) and BC(1) groups) showed reduced survival compared with wild salmon but grew faster as juveniles and displaced wild parr, which as a group were significantly smaller. Where suitable habitat for these emigrant parr is absent, this competition would result in reduced wild smolt production. In the experimental conditions, where emigrants survived downstream, the relative estimated lifetime success ranged from 2% (farm) to 89% (BC(1) wild) of that of wild salmon, indicating additive genetic variation for survival. Wild salmon primarily returned to fresh water after one sea winter (1SW) but farm and 'hybrids' produced proportionately more 2SW salmon. However, lower overall survival means that this would result in reduced recruitment despite increased 2SW fecundity. We thus demonstrate that interaction of farm with wild salmon results in lowered fitness, with repeated escapes causing cumulative fitness depression and potentially an extinction vortex in vulnerable populations.     

Why do some fish do it younger than others? learning from experiments

Atlantic salmon in Maine were once abundant but have become depleted, and are listed as endangered under the federal Endangered Species Act. Historically, salmon numbers in Maine may have been as high as 100 000 adults, but habitat loss, pollution and overfishing have contributed to the decline of the species. In 2000, only 110 adults returned to spawn in Maine rivers. Maine produces c. 15 000 metric tons/year of aquacultured Atlantic salmon from a total of nearly 600 coastal net pens. Escapees from these pens may interact with the wild salmon. The dynamics of salmon populations under such conditions are poorly understood. In order to illuminate the role aquaculture may play in such a system, we have developed a model for simulating population trajectories for both wild salmon and competing populations derived from aquaculture escapes. The model simulates a small population of wild salmon based in a stream/estuary system, into which an aquaculture facility is losing fish to escapes. Biological parameters in the model were estimated as much as possible from data in the USFWS report on Maine salmon. We used the model to investigate the consequences of a variety of ecological interactions between the wild and cultured fish including competitive, genetic and disease effects. Initial results indicate that many of these effects allow the aquaculture‐derived population to supplant the wild fish, but that wild populations may still persist under some conditions.     

Ecological model of interactions between escaped and wild Atlantic salmon Salmo salar

Atlantic salmon in Maine were once abundant but have become depleted, and are listed as endangered under the federal Endangered Species Act. Historically, salmon numbers in Maine may have been as high as 100 000 adults, but habitat loss, pollution and overfishing have contributed to the decline of the species. In 2000, only 110 adults returned to spawn in Maine rivers. Maine produces c. 15 000 metric tons/year of aquacultured Atlantic salmon from a total of nearly 600 coastal net pens. Escapees from these pens may interact with the wild salmon. The dynamics of salmon populations under such conditions are poorly understood. In order to illuminate the role aquaculture may play in such a system, we have developed a model for simulating population trajectories for both wild salmon and competing populations derived from aquaculture escapes. The model simulates a small population of wild salmon based in a stream/estuary system, into which an aquaculture facility is losing fish to escapes. Biological parameters in the model were estimated as much as possible from data in the USFWS report on Maine salmon. We used the model to investigate the consequences of a variety of ecological interactions between the wild and cultured fish including competitive, genetic and disease effects. Initial results indicate that many of these effects allow the aquaculture‐derived population to supplant the wild fish, but that wild populations may still persist under some conditions.     

Dissecting the loci underlying maturation timing in Atlantic salmon using haplotype and multi-SNP based association methods

Characterizing the role of different mutational effect sizes in the evolution of fitness-related traits has been a major goal in evolutionary biology for a century. Such characterization in a diversity of systems, both model and non-model, will help to understand the genetic processes underlying fitness variation. However, well-characterized genetic architectures of such traits in wild populations remain uncommon. In this study, we used haplotype-based and multi-SNP Bayesian association methods with sequencing data for 313 individuals from wild populations to test the mutational composition of known candidate regions for sea age at maturation in Atlantic salmon (Salmo salar). We detected an association at five loci out of 116 candidates previously identified in an aquaculture strain with maturation timing in wild Atlantic salmon. We found that at four of these five loci, variation explained by the locus was predominantly driven by a single SNP suggesting the genetic architecture of this trait includes multiple loci with simple, non-clustered alleles and a locus with potentially more complex alleles. This highlights the diversity of genetic architectures that can exist for fitness-related traits. Furthermore, this study provides a useful multi-SNP framework for future work using sequencing data to characterize genetic variation underlying phenotypes in wild populations.Subject terms: Evolutionary genetics, Genetic association study     

Molecular Genetic Analysis of Stomach Contents Reveals Wild Atlantic Cod Feeding on Piscine Reovirus (PRV) Infected Atlantic Salmon Originating from a Commercial Fish Farm

In March 2012, fishermen operating in a fjord in Northern Norway reported catching Atlantic cod, a native fish forming an economically important marine fishery in this region, with unusual prey in their stomachs. It was speculated that these could be Atlantic salmon, which is not typical prey for cod at this time of the year in the coastal zone. These observations were therefore reported to the Norwegian Directorate of Fisheries as a suspected interaction between a local fish farm and this commercial fishery. Statistical analyses of genetic data from 17 microsatellite markers genotyped on 36 partially-degraded prey, samples of salmon from a local fish farm, and samples from the nearest wild population permitted the following conclusions: 1. The prey were Atlantic salmon, 2. These salmon did not originate from the local wild population, and 3. The local farm was the most probable source of these prey. Additional tests demonstrated that 21 of the 36 prey were infected with piscine reovirus. While the potential link between piscine reovirus and the disease heart and skeletal muscle inflammation is still under scientific debate, this disease had caused mortality of large numbers of salmon in the farm in the month prior to the fishermen''s observations. These analyses provide new insights into interactions between domesticated and wild fish.     

Alternative male life-history tactics as potential vehicles for speeding introgression of farm salmon traits into wild populations

Releases of cultured organisms, such as farm Atlantic salmon (Salmo salar L.), threaten native biodiversity and the integrity of natural communities. Salmon escaping from sea farms, however, have relatively poor reproductive success, suggesting that the rate of spread of domesticated traits may be reduced. We now compare the relative reproductive success of males that mature precociously in freshwater (parr) and find that those of farm origin have higher breeding and fertilization success than wild and hybrid individuals. Specifically, hybrid parr had 57% and wild parr 25% the success of farm parr. Early maturing males could thus be important vehicles promoting introgression of domesticated and/or non‐native traits into wild populations and ultimately have long‐term impact on the genetic integrity of native populations.     

Genetic interactions between marine finfish species in European aquaculture and wild conspecifics

The principal species of marine aquaculture in Europe are Atlantic salmon (Salmo salar), sea bass (Dicentrarchus labrax) and sea bream (Sparus auratus). For Atlantic salmon and sea bass, a substantial part of total genetic variation is partitioned at the geographical population level. In the case of sea bream, gene flow across the Azores/Mediterranean scale appears to be extensive and population structuring is not detected. For Atlantic salmon and sea bass, natural population structure is at risk from genetic interaction with escaped aquaculture conspecifics. The locally adaptive features of populations are at risk from interbreeding with non‐local aquaculture fish. Wild populations, generally, are at risk from interactions with aquaculture fish that have been subject to artificial selection or domestication. Atlantic salmon is the main European aquaculture species and its population genetics and ecology have been well‐studied. A general case regarding genetic interactions can be based on the information available for salmon and extended to cover other species, in the appropriate context. A generalized flow chart for interactions is presented. Salmon escape from aquaculture at all life stages, and some survive to breed among wild salmon. Reproductive fitness in the escaped fish is lower than in native, wild fish because of behavioural deficiencies at spawning. However, as the number of salmon in aquaculture greatly exceeds the number of wild fish, even small fractional rates of escape may result in the local presence of large numbers, and high frequencies, of escaped fish. At present, policy and legislation in relation to minimizing genetic interactions between wild and aquaculture fish is best developed for Atlantic salmon, through the recommendations of the Oslo Agreement developed by the North Atlantic Salmon Conservation Organization and subsequent agreements on their implementation. In future, the potential use of genetically modified fish in aquaculture will make additional policy development necessary. Improved containment is recommended as the key to minimizing the numbers and therefore the effects of escaped fish. Emergency recovery procedures are recommended as a back‐up measure in the case of containment failure. Reproductive sterility is recommended as a future key to eliminating the genetic potential of escaped fish. The maintenance of robust populations of wild fish is recommended as a key to minimizing the effects of escaped fish on wild populations.     

Restitution and genetic differentiation of salmon populations in the southern Baltic genotyped with the Atlantic salmon 7K SNP array

Background

Native populations of Atlantic salmon in Poland, from the southern Baltic region, became extinct in the 1980s. Attempts to restitute salmon populations in Poland have been based on a Latvian salmon population from the Daugava river. Releases of hatchery reared smolts started in 1986, but to date, only one population with confirmed natural reproduction has been observed in the Slupia river. Our aim was to investigate the genetic differentiation of salmon populations in the southern Baltic using a 7K SNP (single nucleotide polymorphism) array in order to assess the impact of salmon restitution in Poland.

Methods

One hundred and forty salmon samples were collected from: the Polish Slupia river including wild salmon and individuals from two hatcheries, the Swedish Morrum river and the Lithuanian Neman river. All samples were genotyped using an Atlantic salmon 7K SNP array. A set of 3218 diagnostic SNPs was used for genetic analyses.

Results

Genetic structure analyses indicated that the individuals from the investigated populations were clustered into three groups i.e. one clade that included individuals from both hatcheries and the wild population from the Polish Slupia river, which was clearly separated from the other clades. An assignment test showed that there were no stray fish from the Morrum or Neman rivers in the sample analyzed from the Slupia river. Global F_ST over polymorphic loci was high (0.177). A strong genetic differentiation was observed between the Lithuanian and Swedish populations (F_ST = 0.28).

Conclusions

Wild juvenile salmon specimens that were sampled from the Slupia river were the progeny of fish released from hatcheries and, most likely, were not progeny of stray fish from Sweden or Lithuania. Strong genetic differences were observed between the salmon populations from the three studied locations. Our recommendation is that future stocking activities that aim at restituting salmon populations in Poland include stocking material from the Lithuanian Neman river because of its closer geographic proximity.

Electronic supplementary material

The online version of this article (doi:10.1186/s12711-015-0121-9) contains supplementary material, which is available to authorized users.