Isolation and characterization of novel polymorphic microsatellite markers for the white stork , Ciconia ciconia : applications in individual – based and population genetics

Aislamiento y caracterizacion de nuevos marcadores de microsatelites polimorficos para la ciguena blanca, Ciconia ciconia : aplicaciones de la genetica basada en individuos y de poblaciones A pesar de que la ciguena blanca, Ciconia ciconia , es una especie modelo en estudios de migracion y comportamiento, los marcadores moleculares publicados hasta ahora no son lo suficientemente polimorficos para poder realizar estudios geneticos basados en el individuo. Utilizando la secuenciacion de nueva generacion hemos seleccionado 11 marcadores polimorficos y los hemos utilizado en ciguenas de dos localidades de estudio. El numero medio de alelos por locus fue de 5,3 con un minimo de dos y un maximo de 10. La heterocigosidad media observada y esperada fue de 0,519 y 0,565, respectivamente. La P ID (probabilidad de identidad) resulto ser suficientemente sensible para los estudios sobre genetica basada en individuos y genetica de poblaciones. No hemos encontrado evidencias de perdida alelica, alelos nulos ni ningun otro error y ningun locus estaba en desequilibrio de Hardy–Weinberg en ambas localidades a pesar de que dos loci si que lo estuvieran en una unica localidad. Estos marcadores son utiles para dar respuesta a una serie de preguntas relacionadas con la diversidad genetica, el grado de filopatria y las estrategias geneticas de reproduccion.


Introduction
The white stork, Ciconia ciconia, is a model species for studies of bird migration and behavior because of its long life span, suspected monogamy and philopatry, proximity to human settlements, ubiquity, ease of identification, and magnificent migratory journeys.Much is known about the species through observation and biotelemetry tracking, although these approaches might be biased (ring-resightings) or suitable for just a few tens of individuals because of high monetary investments and associated costs (e.g., tracking: data download, GPS trackers).While one population genetics study has been performed (Shephard et al., 2013), the mean number of alleles for the 18 populations considered was very low: 3.01 ± 0.75 (mean ± SD); furthermore, little is known about genetic mating strategies, nest fidelity, and levels of natal philopatry in white storks although these behaviors are relatively easy to study using a powerful set of molecular markers.Therefore, coordination between international research groups at the time of ringing (an annual occurrence at thousands of nests throughout Europe and the Middle East) in order to collect genetic samples (feather collection) and use of a highly polymorphic microsatellite panel can provide vast data useful for a variety of ecological and behavioral studies.
A preliminary test of microsatellite markers published by Shephard et al. (2009), using genetic material from wild individuals, showed even lower polymorphism levels than originally reported.From the initial (published) panel of 13 microsatellites, only seven remained after removing markers where null alleles (n = 2), linkage disequilibrium (n = 1), and amplification issues (n = 3) were found.This reduced panel of largely dimorphic markers is insufficient to elucidate individual behavioral strategies and population dynamics in this model species.The development of new markers, pivotal to future genetic-based studies, is therefore essential.Here, we present 11 polymorphic markers -selected and tested in samples from two wild study sites-that can be used to answer a range of ecological questions on white storks.

Material and methods
We developed microsatellites using a next generation sequencing approach performed by EcoGene NZ: DNA-based diagnostics (Auckland, New Zealand) in 2013 from DNA extracted from blood samples of two white storks, one from a wild population in northern Israel and one from a wild population in northeast Germany.We discovered over 100 potentially polymorphic microsatellite loci with a throughput of 35M bases, an average read length of 441.6 base pairs, and a total of 170,969 reads (64,583 and 106,139 reads per run, respectively); primers were designed by EcoGene NZ using msatcommander (Faircloth, 2008).
Sixty-four microsatellite loci were chosen based on size (optimal length was considered as 100-350 base pairs) and number of bases repeated (loci with tetra-base repeat motifs were preferred).These 64 markers were tested for amplification and polymorphism with PCR using the M13 method for fluorescent labeling (Schuelke, 2000) in a subset of 94 stork samples (see below for sampling information).PCR was performed in 20 µL volumes with 2.7 µL DNA (1:20 dilution of an NaOH extraction; Zhang et al., 1994), 10 µL Taq Plus Master Mix 2x (Lamda Biotech; contains 1.5 mM MgCl 2 ), 0.5 µL (final concentration: 0.25 µM) fluorescent-labeled M13 (either 6FAM or TAMRA), 0.5 µL (0.25 µM) reverse primer, 0.342 µL (0.0175 µM) forward primer, and 5.96 µL double distilled water.PCR conditions were as follows: an initial step at 94°C for five minutes followed by a 'touchdown' cycling program of 16 cycles with 92°C for 30 seconds; annealing for 30 seconds, starting at 60°C and decreasing by 1°C for each of the 16 cycles to 45°C; and 72°C for 30 seconds, followed by 30 cycles continued at an annealing temperature of 45°C, all followed by a final step at 72°C for 10 min.We also applied a similar 'touchdown' cycling from 55ºC to 45ºC with the last 30 cycles at an annealing temperature of 45ºC and/or increased MgCl 2 concentrations to 2.5 mM for those loci that did not amplify (see table 1).
Genotyping was performed using an ABI PRISM™ 3730 xl DNA Analyzer by the Hebrew University Center for Genomic Technologies (Jerusalem, Israel).Allele calling and binning were obtained using GeneMap per 4.0 software (Applied Biosystems, Foster City, CA, USA).Of the 64 loci initially tested, 11 loci were selected based on consistency of amplification and number of alleles per locus (Cc10, Cc15, Cc18, Cc37, Cc42, Cc44, Cc50, Cc58, Cc61, Cc69, and Cc72), and PCR conditions for these loci were then further optimized (see table 1).
Following marker selection, we genotyped 213 individuals using the NaOH extraction method mentioned above.Feathers (five) were collected from juveniles from two sample sites prior to fledging; only one individual per nest was included in this analysis.Samples were collected in 2012 from northeast Germany (n = 152; center point of sampling: 52.7383 o N, 11.6681 o E) and in 2015 from eastern Greece (n = 61; center point of sampling: 41.0520 o N, 25.1223 o E).Following PCR amplification, genotyping, and scoring (as described above), and tests of Hardy-Weinberg equilibrium (Cervus 3.0.3;Kalinowski et al., 2007) were performed as were tests of genotyping error (GIMLET; Valière, 2002), null alleles (FreeNA; Chapuis & Estoup, 2007), allelic drop-out (GIMLET), inbreeding (F IS ; Genetix 4.05.2;Belkhir et al., 2004), linkage disequilibrium (GENEPOP with Bonferroni correction for significance; Raymond & Rousset, 1995;Rousset, 2008), and genetic structure (F ST ; Genetix 4.05.2).Rates of expected and observed heterozygosities, the mean polymorphic information content (PIC) and P ID and P ID-Sib (probability of identity, the likelihood that two unrelated or sibling-related individuals, respectively, will have the same genotype profile by chance) were also calculated (Cervus 3.0.3).

Discussion
We successfully characterized a novel, polymorphic set of microsatellite markers for the white stork.
When comparing mean number of alleles per locus (MNA = 5.3) and mean expected heterozygosity (H Exp = 0.565) from this marker discovery with that from the set of markers originally published (MNA = 3.5; H Exp =0.41; Shephard et al., 2009), this new set of markers is more polymorphic, has greater heterozygosity, and thus heightened power in genetic studies.Furthermore, the PIC is considered at least moderately informative (Hildebrand et al., 1992), the P ID is highly informative for population-based and individual-based studies, and P ID-Sib is moderately to highly informative (Waits et al., 2001) and more so when used in conjunction with markers previously reported (P ID-Sib < 0.0001; Feldman et al., in preparation); alone, P ID-SIB for these markers is sufficient for individual-based sibship studies.Although marker Cc15 and Cc18 have only two alleles and Cc15 has low observed and expected heterozygosities (0.118 and 0.136, respectively), removing these loci from the panel substantially increased the likelihood that two sibling-related individuals would share the same genotype profile by chance (change in P ID-Sib from 0.0007678 to 0.0027647).We thus decided to include them in the final panel due to their utility in individual-based relatedness studies.
Deviations in HWE were not seen in both samples for any markers and in cases where deviation was found, a sample-level explanation was also present (e.g., inbreeding at the given locus was also found).We thus concluded that all the markers are well suited for genetic-based studies and any deviations are likely due to sample site characteristics.The F ST between the German and Greek samples implies that the samples are significantly different, suggesting a weak genetic differentiation between these two regions.This differentiation was not found by Shephard et al. (2013) in comparisons of multiple populations using the previously published panel of markers (overall F ST = 0.005; p-value > 0.05).We believe this difference shows the greater sensitivity of the newly developed panel of markers.
This set of markers, either alone or in combination with markers previously tested in this species, allows researchers to distinguish between genotypes at the individual level, thus providing a tool for relatedness studies in addition to more exact population genetics studies.By adding these markers to the white stork molecular tool kit, questions at both the within-and between-population scales related to reproduction, nest fidelity, migration flyway segregation, and plasticity of philopatry can be much more feasibly and accurately assessed.
ID and P ID-Sib were 0.00000005 and 0.0007678, respectively.Two markers, Cc15 and Cc18, had low polymorphism but had a positive influence on P ID (see Discussion).Removing