Roche, Saaristo, 1998

McCune, Amy R., Bogdanowicz, Steven M., Buston, P., Jackson, J. R. & Harrison, R. G., 2023, Spawning Behavior in a Non-Teleost Actinopterygian: Genetic Evidence for Both Monogamy and Polygamy in Amia, Ichthyology & Herpetology 111 (4), pp. 522-542 : 525-526

Roche /454 sequencing of a bowfin genomic library enriched for microsatellite DNA.— A bowfin genomic DNA library was enriched for a set of dimeric, trimeric, and tetrameric repeats and sequenced with Roche /454 technologies ( Andres and Bogdanowicz, 2011). Briefly, approximately 500 nanograms of genomic DNA was digested with Hinc II and ligated to a double-stranded adapter in the presence of Pme I, to prevent selfligation of the linker. Adapter-ligated DNA was hybridized to 3 0 -biotinylated repeat oligonucleotides (representing two dimers, five trimers, and five tetramers), captured with streptavidin-coated magnetic beads (New England Biolabs [NEB], Beverly, MA), and made double-stranded by PCR with an adapter primer. PCR products were pooled and ligated to a Roche /454 Rapid Library adapter, and small fragments were removed with Ampure beads and a sizing solution (8.4% PEG-8000, 1.2 M NaCl, NEB protocol). Sequence data were collected on a Roche /454 platform with Titanium reagents and software. Reads were trimmed of adapter sequence and high-quality reads were assembled with SeqMan Pro software (DNASTAR, Madison, WI).

Microsatellite primer design and testing loci.— PCR primers for genotyping were designed with PrimerSelect software (DNASTAR). We employed three-primer PCR to determine whether loci amplified cleanly and were polymorphic. One of the locus-specific primers was tagged at the 5 0 end with a 20 bp “long tag” (5 0 –CGAGTTTTCCCAGTCACGAC–3 0 Schuelke, 2000), the other primer was modified with a 5 0 “pig tail” (5 0 –GTTTCTT–3 0; Brownstein et al., 1996). For three-primer PCR, each pair of locus-specific primers was mixed with a third primer identical to the long tag sequence and with a 5 0 -6FAM dye. The PCR concentrations of locus-specific pig-tailed and long tag primers was 0.2 lM and 0.05 lM, respectively, while the concentration of the 6FAM long tag was 0.15 lM. PCR reaction volume was 10 lL, and cycling consisted of 35 cycles at 948C for 50 seconds, 548C for 45 seconds, 728C for 1 minute. The resulting PCR products were analyzed first on 1% agarose gels; clean/robust PCR products were then analyzed on ABI 3100 or 3730xl capillary sequencers to determine levels of polymorphism. Loci used for genotyping had the previously long-tagged primer resynthesized with a 5 0 -labeled fluorescent dye compatible with the G5 dye set (Life Technologies, Carlsbad, CA). The long tag was omitted from these 5 0 dye-labeled primers.

Microsatellite DNA genotyping.— Two primer mixes were constructed for multiplex PCR using a Type-It Microsatellite PCR kit (Qiagen, Valencia, CA). Primer mix one contained six loci (five tetrameric repeats and one trimeric repeat) and mix two contained five loci (two trimeric repeats and three dimeric repeats; Table 1 ). Each adult male or fry DNA sample analyzed was amplified for each primer mix (11 loci total). PCR reaction volume was 10 ll and consisted of 1 lL of DNA, 3 lL of molecular biology grade (MBG) water, 5 lL of 2x Type-It master mix, and 1 lL of primer mix 1 or 2. Reactions were heated at 958C for 5 minutes, then cycled 27 times at 948C for 50 seconds, 558C for 90 seconds, and 728C for 30 seconds, followed by 30 minutes at 608C. PCR products were diluted 1: 30 in MBG water, and 1.5 lL of each diluted PCR product was combined with 15 lL Hi-Di Formamide and 0.12 lL Genscan LIZ-500 ladder (both from Life Technologies). Data were collected on ABI 3100 and 3730xl capillary sequencers. Allele calls were made with Genemapper v. 4.0 software (Life Technologies). We used CERVUS v. 3.0.3 ( Kalinowski et al., 2007) to calculate allele frequencies and probabilities of both identity and parental exclusion for our population sample (excluding fry).

Ascertainment of parentage and polygamy.— We used the program CERVUS v. 3.0.3 ( Kalinowski et al., 2007) to answer the question: is the multilocus microsatellite DNA genotype displayed by the adult male guarding each brood consistent with that male being a parent of each of the genotyped fry he was guarding? To infer the genotype of the primary spawning female, we first assumed that the most abundant fry genotypes in a nest represent the guarding male and inferred primary female. By subtraction, we could infer the primary female’s genotype from the combination of fry genotypes, given the male genotype. We refer to the guarding male and the inferred primary female as the primary pair. We used simple Mendelian inheritance to determine cases of polygamy, in which one or more extra-pair (EP) adults contributed alleles present in some of the offspring in a brood, but these alleles were not attributable to either of the primary pair. We compared the genotypes of each of the sampled fry at a nest and the male guarding them, for each microsatellite locus. Polygamy was invoked when we observed more than two diploid genotypic classes at any microsatellite locus among offspring when the adult male was a homozygote at that locus, when we observed more than four genotypic classes among fry when the adult male was heterozygous at a given locus, and/or if there were fry genotypes inconsistent with the genotypes of the guarding male and inferred primary female. We inferred polyandry when a guarded fry’s genotype did not contain either of the guarding male’s alleles at a locus. Polygyny was indicated when a fry’s genotype did not contain either of the inferred female alleles at a locus and/or when a series of fry from the same nest exhibited variation in size of the mitochondrial D-loop PCR fragment. In the two broods for which we lacked the guarding male genotype, we were only able to infer polygamy without specifying whether polygyny or polyandry was indicated. In two cases, there were single fry at a single locus with a genotype not explained by the parent genotypes, which we regard as insufficient evidence for polygamy given possible explanations for such a rare genotype (including genotyping error, orphan fry, or germline mutation). For the subset of broods with an identified paternal genotype, we estimated the minimum number of additional parents contributing genes to a brood by counting the number of alleles represented in the brood which were not also represented in the primary spawning pair and determining the minimum number of EP adults required to account for these extra alleles. This is a minimum estimate because both alleles of an extra parent may not be represented in the brood and there may be extra adults with alleles identical to those of the primary spawning pair.