INTRODUCTION

The spotted rose snapper, Lutjanus guttatus (Steindachner, 1869), is a demersal marine finfish with a wide distribution range along the eastern Pacific Ocean from the Gulf of California in Mexico to Ecuador (^{Fischer et al., 1995}). It is a valuable fishing resource in the region (^{Herrera-Ulloa et al., 2010}; ^{Sarabia-Méndez et al., 2010}; ^{Correa-Herrera & Jiménez-Segura, 2013}), with a high potential for aquaculture (^{Ibarra-Castro et al., 2013}). Lutjanus guttatus is a batch spawner with asynchronous ovarian development during a long reproductive season comprising peak spawning periods in April, August and October (^{Arellano-Martínez et al., 2001}; ^{Sarabia-Méndez et al., 2010}). Little is known regarding the spawning contribution of females and males daily.

Microsatellite genetic markers are a necessary tool for the genetic characterization of wild populations that can be used, for example, to improve their management, in rehabilitation programs, and stock identification (^{Hallerman, 2003}). They are also important in aquaculture as they can be used to determine female spawning frequency. Thus, the effective parental contribution [e.g., red sea bream Pagrus major (^{Perez-Enriquez et al., 1999}), California yellowtail Seriola lalandi (^{Smith et al., 2015}), gilthead seabream Sparus aurata (^{García-Fernández et al., 2018})], which due the disproportion in male to female contributions, large family size variance and null female spawners in the broodstock can lead to the potential accumulation of inbreeding within the hatchery (^{Blonk et al., 2009}; ^{Domingos et al., 2014}). Other markers, such as Single Nucleotide Polymorphisms (SNPs), have shown reliable results for parentage testing in aquaculture species [e.g., shrimp (^{Perez-Enriquez & Max-Aguilar, 2016})]. However, for L. guttatus there is no previous genomic information available.

The present study aimed to obtain a set of microsatellite markers for future population genetic studies of L. guttatus and to test a reduced panel to estimate the temporal parental contribution in a cultivated stock of the species.

MATERIALS AND METHODS

RESULTS

The 18 microsatellite loci showed reliable genotyping patterns in the Lutjanus guttatus broodstock, resulting in high genetic diversity (n_a = 13.4 ± 1.3; Ho = 0.88 ± 0.05; Table 1). Three loci departed from HWE (only one after the Bonferroni correction) (Table 1), which can be explained by the potential presence of null alleles, rather than by stuttering or allele drop-out, as indicated by the Micro-Checker analysis. Fourteen loci are available to assess population genetic structure in wild L. guttatus (N. Diaz-Viloria, unpublish. data), and several loci are potentially useful for the other snapper species (Table 1).

For parentage assignment, 413 embryonated eggs were used. The combination of direct and probabilistic (95% CL) exclusion methods resulted in 95.6% of the progeny (n = 395) assigned to a single parental couple, leaving 4.3% unassigned. Seventeen families (out of 24) were represented in the progeny (Table 2).

The reproductive season within the breeding tank spanned from June to November, with a peak number of spawns occurring in October. Most males (7A62-M, 1170-M, 1538-M and 2924-M) reproduced throughout the season and during consecutive days (Fig. 1a). In contrast, most of the progeny (n = 384; 97.2%) were produced by only two of the four females (4B67-H and 4953-H), and spawning did not occur on consecutive days (Fig. 1b), indicating that females (at least 4B67-H) spawn every other day.

Figure 1 The proportion of breeders contributing to progeny during the spawning events of July-October 2011. a) Males, b) females. 

Unexpected genotypes were observed in several families, resulting in a deviation from the expected Mendelian proportions (Table 3). As it is unlikely that these genotypes come from genotyping errors (the sequencer sizing differences observed in four duplicated samples was between 0-0.7 units in at least four loci), they appear to be a consequence of both null and mutated alleles. Considering a null allele as a mutation event, the mutation rate per locus varied between a maximum of 1.2×10^-1 in Lgut21 to a minimum of 7.7×10^-3 in Lgut34, for a mean of 4.5×10^-2. Null-allele events were three times larger than changes in the number of repeats. While all the mutations in Lgut21 and Lgut39 were due to null alleles, there was a combination of null alleles and base pairs gains in the other loci (Table 3). In Lgut18 and Lgut30, the most common change was a gain in four base pairs (equivalent to one microsatellite repeat). The mutation rates were, on average, almost twice higher for females than males (Table 4).

DISCUSSION

The usefulness of microsatellites as genetic markers for parentage assignment has been demonstrated in more than 20 cultivated fish species (^{Yue & Xia, 2014}). The capability of correct assignment is dependent on several characteristics of the genetic markers, of which their variability is one of the most relevant (^{Vandeputte & Haffray, 2014}). The five high-variable microsatellites selected (with a combined probability of non-exclusion in the order of 10^–6) were enough to confidently determine, by both exclusion methods, the parentage of 92% of the progeny from a relatively small Lutjanus guttatus broodstock, supporting the usefulness of this reduced panel, for the assessment of multiple spawning events. The use of new genetic markers, such as SNPs, is an alternative for parentage testing in relevant aquaculture such as shrimp (^{Perez-Enriquez & MaxAguilar, 2016}), and oysters (^{Lapègue et al., 2014}). Routine genotyping platforms are available; however, these types of platforms are not yet available for L. guttatus. Other techniques (e.g., KASP, Taqman, HRM) are not economically feasible for more than 50 SNPs.

The reproductive pattern of males, most of them reproducing throughout the season and during consecutive days, has also been observed in the California yellowtail Seriola lalandi Valenciennes, 1833 (^{Smith et al., 2015}). In wild L. guttatus females, asynchronous development of the gonads and partial spawning behavior has been described (^{Arellano-Martínez et al., 2001}). The overrepresentation of females should be taken into account for hatchery management as an unbalanced family size that can lead to an increased inbreeding rate (^{Perez-Enriquez et al., 1999}; ^{García-Fernández et al., 2018}).

Parentage assessment within a day of a spawning event by using the DNA extracted from fish embryonated eggs is recommended using a mechanical method rather than a chemical method as in other fish species [e.g., gilthead seabream Sparus aurata (^{García-Fernández et al., 2018}); zebrafish Danio rerio (^{Westerfield, 2007})]. However, the selection of embryonated eggs posterior to gastrula for DNA analysis is critical for PCR success, as similar results were reported for the gilthead seabream (^{García-Fernández et al., 2018}).

Mutations and null alleles in microsatellites are a common phenomenon resulting in failed assignments (^{Ellegren, 2000}). The mean mutation rate obtained in our study (10^–2 per locus per generation) is higher than other fish species, such as the carp Cyprinus carpio with 10^–4 (^{Yue et al., 2007}), or various salmonids with 10^–2-10^–5 (^{Shaikhaev & Zhivotovsky, 2014}). Despite the high mutation rate, five high-variable microsatellites were enough to confidently determine, by direct exclusion, the parentage of progeny from a relatively small broodstock of the spotted rose snapper. For a larger broodstock, the number of genetic markers can be increased to minimize the non-exclusion probability (in the order of magnitude of 10^–6 in the present study), using the remaining markers developed for the species (Table 1).

The estimation of the contribution of males and females of broodstocks kept in communal tanks is relevant for the implementation of selective breeding programs (^{García-Fernández et al., 2018}). A more intensive and extended in time genotyping that gives a better genetic representation of the gene pool of the selected broodstock has been suggested for the red sea bream Pagrus major (^{Nugrohoa & Taniguchi, 2004}) and the barramundi Lates calcarifer (^{Domingos et al., 2014}). This information will also be important for the definition of the breeding goal, not only if the plan is focused on the improvement of reproductive traits, but also for other characteristics (growth, stress resistance, meat quality, others) (^{Gjedrem, 2012}).

As an additional contribution, the genetic markers panel will also be useful for genetic studies in wild populations focused on their management in other lutjanid species.