Genetic Linkage Maps and Synteny of Lucania Goodei and L. Parva Reveal a Robertsonian Fusion

G3: Genes|Genomes|Genetics Early Online, published on June 3, 2014 as doi:10.1534/g3.114.012096

Insight into genomic changes accompanying divergence: genetic linkage maps and synteny of Lucania goodei and L. parva reveal a Robertsonian fusion

Emma L. Berdan1§†, Genevieve M. Kozak§†, Ray Ming‡, A. Lane Rayburn*, Ryan Kiehart** Rebecca C. Fuller§

§ Department of Animal Biology, University of Illinois, Champaign, IL, 61820, USA ‡Department of Plant Biology, University of Illinois, Urbana, IL, 61801, USA * Department of Crop Sciences, University of Illinois, Urbana, IL, 61801, USA ** Department of Biology, Ursinus College, PA

1Corresponding author: [email protected]; Current address: Museum für Naturkunde, Leibniz-Institut für Evolutions-und Biodiversitätsforschung, Berlin, Germany

† These authors contributed equally to this work.

Author email addresses Berdan: [email protected] Kozak: [email protected] Ming: [email protected] Rayburn: [email protected] Keihart: [email protected] Fuller: [email protected]

Running Title: Killifish linkage maps

Key words: Synteny, Robertsonian fusion, Chromosomal rearrangement, Linkage map, Speciation, EST-based SNPs, Fundulidae

Corresponding Author:

Emma Berdan Museum für Naturkunde Leibniz-Institut für Evolutions-und Biodiversitätsforschung Invalidenstraße 43 10115 Berlin +49 302093 8564 [email protected]

1 Abstract

2 Linkage maps are important tools in evolutionary genetics and in studies of speciation. We

3 performed a karyotyping study and constructed high-density linkage maps for two closely related

4 killifish species, Lucania parva and L. goodei, that differ in salinity tolerance and still hybridize

5 in their contact zone in Florida. Using SNPs from orthologous EST contigs, we compared

6 synteny between the two species to determine how genomic architecture has shifted with

7 divergence. Karyotyping revealed that L. goodei possesses 24 acrocentric chromosomes (1N)

8 while L. parva possesses 23 chromosomes (1N), one of which is a large metacentric

9 chromosome. Likewise, high-density SNP-based linkage maps indicated 24 linkage groups for

10 L. goodei and 23 linkage groups for L. parva. Synteny mapping revealed two linkage groups in

11 L. goodei that were highly syntenic with the largest linkage group in L. parva. Together, this

12 evidence points to the largest linkage group in L. parva being the result of a chromosomal fusion.

13 We further compared synteny between Lucania with the genome of a more distant teleost

14 relative medaka (Oryzias latipes) and found good conservation of synteny at the chromosomal

15 level. Each Lucania linkage group had a single best match with each medaka chromosome.

16 These results provide the groundwork for future studies on the genetic architecture of

17 reproductive isolation and salinity tolerance in Lucania and other Fundulidae.

20 Introduction

21 Species-specific linkage maps are critical to understanding genomic architecture and how it

22 differs between species. Linkage maps allow the exploration of genotype-phenotype

23 relationships through QTL mapping (Falconer and Mackay 1996; for examples see Peichel et al.

24 2001; Tripathi et al. 2009) and the comparison of genomic architecture between species via

25 synteny mapping (Prince et al. 1993; Backstrom et al. 2008; Muchero et al. 2009; Lucas et al.

26 2011; Mcgraw et al. 2011). However, the majority of previous studies of genome-wide synteny

27 have used pairs of species that are distantly related (i.e. from different orders or families: Stapley

28 et al. 2008; Jaari et al. 2009; Foley et al. 2011; Mcgraw et al. 2011). It is less common for

29 comparisons of synteny to be made within the same genus (but see Ming et al. 1998; Rogers et

30 al. 2007; Lubieniecki et al. 2010; Lee et al. 2011; Timusk et al. 2011; Naish et al. 2013). High

31 density linkage maps are now possible with the advent of high throughput sequencing and have

32 the potential to facilitate fine-scale comparisons of synteny between closely related species.

33 These comparisons are needed to determine how genome structure diverges and potentially

34 contributes to speciation.

35 The problem for maintaining species boundaries in areas of sympatry is that gene flow

36 between species and recombination in hybrids should homogenize species-specific traits and

37 breakdown reproductive isolating barriers. Genomic rearrangements, such as chromosomal

38 fusions, inversions or deletions, can potentially facilitate the maintenance of reproductive

39 isolating barriers because they reduce recombination in portions of the genome (Rieseberg 2001;

40 Butlin 2005; Faria and Navarro 2010). Data from a wide range of taxa support the theory that

41 genes conferring reproductive isolation between sympatric species are often localized to

42 rearranged areas of the genome (Rieseberg et al. 1999; Noor et al. 2001; Coluzzi et al. 2002;

43 Feder et al. 2003; Kitano et al. 2009).

44 One way to study the role of genomic rearrangements in speciation is to compare synteny

45 between species that hybridize at low levels. We do this in Lucania, a genus that is becoming a

46 model system for ecological speciation research. The bluefin killifish (Lucania goodei) and the

47 rainwater killifish (L. parva) are two closely related species that differ radically in their salinity

48 tolerance (Duggins et al. 1983; Whitehead 2010). Lucania goodei is found primarily in

49 freshwater habitats, while L. parva is euryhaline and can be found in fresh, brackish, and marine

50 habitats (Lee et al. 1980). Survival at various life stages differs between Lucania species in

51 different salinities (Fuller et al. 2007; Fuller 2008). Lucania goodei and L. parva also show

52 divergence in sequence and expression of a number of salinity tolerance genes (Berdan and

53 Fuller 2012; Kozak et al. 2014). Despite these ecological differences between species, there is

54 evidence for low levels of ongoing hybridization in sympatric populations. Sympatric

55 populations exist in the Atlantic and Gulf coastal waterways of Florida (Fuller and Noa 2008).

56 Hubbs and colleagues (1943) found hybrids based on morphological characters in one

57 population, and analysis of mtDNA suggests recent gene flow between the two species in

58 multiple river drainages in Florida (R.C. Fuller unpublished data). We hypothesized that

59 differences in genome structure might contribute to speciation in Lucania. This hypothesis was

60 sparked by an unsupported comment made in an older study; Uyeno and Miller (1971) stated that

61 chromosome number differs between L. goodei and L. parva but did not provide any evidence to

62 support this statement.

63 The purpose of this study was to determine whether L. goodei and L. parva differ in

64 genomic architecture. To do this, we (a) karyotyped both species and (b) produced two high-

65 density SNP-based linkage maps. Transcriptome sequencing and high-throughput SNP

66 genotyping was performed in both species to create the linkage maps. These data allowed us to

67 determine (a) the number of chromosomes possessed by each species, (b) whether a

68 fusion/fission event had occurred, and (c) patterns of synteny between the two species. We

69 analyzed synteny at the linkage group level and at the level of marker order to determine whether

70 large or small-scale genomic rearrangements have occurred during divergence (Gale and Devos

71 1998; Ming et al. 1998; Wang et al. 2010). We further compared synteny in Lucania to the most

72 closely related species with a sequenced genome, medaka (Oryzias latipes) to ask how Lucania

73 genomic architecture corresponds to that of other teleost fish (Kasahara et al. 2007). These

74 linkage maps will enable future studies to ask whether the areas of the genome contributing to

75 salinity tolerance are also implicated in reproductive isolation and whether these traits map to

76 genomic rearrangements in Lucania.

78 Materials and Methods

79 Karyotyping

80 Somatic karyotypes were determined for both L. goodei and L. parva using metaphase spreads.

81 For each species, animals of both sexes from multiple populations were used. Details on the

82 karyotyping methods can be found in the Supplemental Methods. Briefly, individuals were

83 injected intraperitoneally at 0h with a phytohemagglutinin solution to stimulate mitosis, injected

84 at 24h with 1% colchicine solution, and then sacrificed at 26h using MS-222. The gills were

85 removed, placed in chilled distilled water to allow the cells to swell (30 minutes), then fixed in a

86 3:1 methanol:glacial acetic acid mixture (30 minutes). The gills were dabbed on the slides to

87 break the cells and release the chromosomes. Slides were cleared, dried, stained with 5%

88 Giemsa solution (Electron Microscopy Sciences, Hatfield, PA), mounted with Permount, and

89 visualized using a compound microscope.

91 Comparative genetic linkage maps - overview

92 Genetic linkage maps were created separately for both L. goodei and L. parva from F2 mapping

93 crosses between two geographically isolated populations. The key to comparing the genetic

94 linkage maps between the two species was to use SNP markers from orthologous ESTs (from

95 454 and Illumina sequencing) that were expressed in both species (Figure 1). F2 offspring and F1

96 parents were genotyped using an Illumina Infinium Bead Chip custom designed for Lucania.

98 Mapping crosses

99 For each species, crosses were set up between geographically and ecologically divergent

100 populations. For Lucania goodei, these populations were Upper Bridge from the spring-fed

101 Wakulla River (Wakulla County, Florida) and a swamp population in the Everglades (Broward

102 County, Florida). For Lucania parva, the populations were Indian River Lagoon, an Atlantic

103 coast population where salinity is typically 35 ppt (Brevard County, Florida) and Pecos River, a

104 freshwater inland river in Texas (Pecos-Crockett County border, Texas). Between population

105 breeding pairs were established in both hybrid cross directions. Cross designs for the creation of

106 the F1 parents are described for L. goodei in Fuller et al. (2010) and for L. parva in Kozak et al.

107 (2012). F1 hybrid offspring were then raised to adulthood and paired with unrelated F1 hybrid

108 individuals to create F2 offspring. For L. parva, we created and genotyped 161 F2 offspring from

109 8 F2 families and 16 F1 parents (parents from 11 F1 families: 4 Indian River female x Pecos male

110 crosses; 7 Pecos female x Indian River male crosses; see Kozak et al. 2012). For L. goodei, we

111 created and genotyped 303 F2 offspring from 14 F2 families and 28 F1 parents (21 F1 families: 11

112 Upper Bridge female x Everglades male crosses; 10 Everglades female x Upper Bridge male

113 crosses).

114 F2 eggs were collected and raised in freshwater with dilute methylene blue (an anti-

115 fungicide). Fry were fed newly hatched Artemia and raised to 1 month post-fertilization. The

o 116 fry were then euthanized in MS-222, preserved in ethanol, and stored at -80 C. F1 parents and F0

117 grandparents were preserved in ethanol. F2 family sizes ranged from 15 to 24 fry.

118

119 Creation of population-specific EST libraries

120 RNA was extracted from five males and five females from the two L. goodei populations (Upper

121 Bridge and Everglades) and the two L. parva populations (Indian River and Pecos River). Fish

122 were euthanized in MS-222. Tissue samples were taken from the gills (1-2 arches), dorsal fins,

123 eyes, brain, and the gonads (ovaries or testes). RNA was extracted using a protocol modified

124 from Carleton (2011: see Supplemental Methods). A pooled sample containing RNA from all

125 tissues was created for each population that contained equal amounts of RNA from all

126 individuals. The RNA was submitted to the Keck Center for Comparative and Functional

127 Genomics at the University of Illinois for creation of cDNA libraries and sequenced using

128 Illumina HiSeq 2000 (see Supplemental Methods). The two populations from L. goodei were

129 uniquely barcoded and run on a single lane of Illumina HiSeq for 100 bp paired-end sequencing.

130 Indian River L. parva and Pecos River L. parva were run on separate lanes. Average quality

131 scores were above 20 for all cycles. Samples produced from 5 to 10 billion bases of data (reads

132 per end: Upper Bridge = 29,661,140, Everglades = 26,235,855, Pecos River = 55,535,778, Indian

133 River = 53,061,233). The population specific Illumina transcriptome sequences are archived in

134 Genbank (bio-project ID: PRJNA215087).

135 In addition, 454 sequencing was done on pooled samples of RNA from five L. goodei

136 populations across Florida (1- Upper Bridge Wakulla River, Wakulla, Co., FL; 2- St. Mark’s

137 National Wildlife Refuge Gambo Bayou, Wakulla, Co., FL; 3- 26-Mile Bend, Everglades,

138 Broward Co., FL; 4- Rum Island Park, Santa Fe River, Columbia Co., FL; 5- Delk’s Bluff

139 Bridge, Oklawaha River, Marion Co., FL). These sequences were used to construct a reference

140 upon which subsequent assembly was based (see below; reference available in Dryad

141 accompanying this paper). Tissue samples were taken from the gills, fins, eyes, brain, and

142 gonads (ovaries or testes). Additional details on the 454 project can be found in Fuller and

143 Claricoates (2011).

144

145 Assembly and alignment

146 Lucania goodei 454 sequences were used as a reference for the assembly. For the reference,

147 contigs were assembled using Newbler assembler (454 Life Sciences, Branford, CT, USA).

148 Assembly parameters were as follows: the minimum contig length was set at 200 bp; the

149 minimum overlap length was 60 bp; and the minimum overlap identity was 95%. A total of

150 29,838 contigs were generated by the Newbler assembler. Using the L. goodei 454 contigs as a

151 reference anchored the analysis of shorter Illumina sequences and allowed identification of

152 contigs that contained multiple SNPs. The goal was to find contigs containing a SNP that was

153 diagnostic for the two L. goodei populations as well as a SNP that was diagnostic for the two L.

154 parva populations. This method may have missed L. parva contigs that were not expressed in L.

155 goodei, but these contigs are uninformative for the comparison of linkage maps. Illumina

156 sequences were trimmed to 75 bp in length and aligned against the reference using Novoalign

157 (Novocraft Technologies; www.novocraft.com). Alignment parameters were as follows: the

158 maximum alignment score acceptable for a best alignment was set at 45; the gap extend penalty

159 was set at 10; and the number of good quality bases for an acceptable read was set at 50.

160

161 SNP selection

162 The alignments were exported to MAQ (Mapping and Assembling with Quality) software (see

163 Fuller and Claricoates 2011 for details) for SNP detection. Diagnostic SNPs for each population

164 were identified using its population pair as a reference. SNPs were considered to be diagnostic

165 when they were identified unambiguously for both populations. There were many more

166 diagnostic SNPs between the two L. parva populations than there were between two L. goodei

167 populations. In order to increase the number of SNPs for L. goodei, SNPs were used that were

168 fixed in one population but were segregating in the alternate population.

169 Candidate SNPs were submitted to Illumina for initial evaluation for suitability for the

170 Infinium Genotyping Assay. There were three classes of SNPs: L. parva-specific SNPs were

171 SNPs that were segregating within L. parva, L. goodei-specific SNPs were SNPs that were

172 segregating within L. goodei, and between species SNPs that were fixed (or nearly fixed)

173 between the two species. The between species SNPs were designed for another study. The

174 custom Infinium bead chip held probes for 4,545 SNPs. Of these, 1,497 were candidate SNPs

175 for L. goodei, 1,369 were candidate SNPs for L. parva, and 1,679 were candidate between

176 species SNPs. All SNPs were labeled with the ID number of the contig in which they occur.

177 Protein annotations for the linkage map contigs were obtained using blastX searches against

178 teleost reference proteomes (Atlantic killifish: Fundulus heteroclitus, Japanese medaka: Oryzia

179 latipes, three-spined stickleback: Gasterosteus aculeatus, guppy: Poecilia reticulata) and human

180 complete proteome (Uniprot release 2012_8 available at:

181 ftp://ftp.uniprot.org/pub/databases/uniprot/previous_releases/release-2012_08/uniref/). Only

182 contigs that matched a single protein in each species with a blast score >100 were considered

183 high confidence annotations (see Supplemental Table 1). Proteins that matched multiple contigs

184 located on different linkage groups were removed from the annotations.

185

186 DNA extraction and SNP genotyping

187 DNA was extracted using a modified version of the PureGene (Gentra Systems,

188 www.gentra.com) extraction protocol over four days (see Supplemental Methods). Sample

189 concentration and quality were verified using a Nanodrop spectrophotometer (Thermo Fisher

190 Scientific, Waltham, MA). DNA samples were diluted to 75 µg/µL in nuclease-free water and

191 then genotyped at all SNPs using the Lucania Illumina Infinium Bead Chip. Source and probe

192 sequences from the chip are accessible on Dryad

193 (http://datadryad.org/resource/doi:10.5061/dryad.hv75h/9). Bead chips were scanned using the

194 iScan System (Illumina) at the Keck Center for Comparative and Functional Genomics at

195 University of Illinois. Illumina GenomeStudio (v2011.1) was used for genotype calls. Cluster

196 positioning was done separately for L. goodei and L. parva SNPs (no-call threshold was set to

197 0.15). Population-specific SNP alleles were verified from genotypes of 16 Upper Bridge L.

198 goodei (7 males, 9 females), 17 Everglades L. goodei (7 males, 10 females), 6 Indian River L.

199 parva (3 males, 3 females) and 5 Pecos River L. parva (3 males, 2 females). Genotype data

200 were analyzed separately for each family. For each family, SNPs were removed if either (1)

201 genotypes were homozygous in both parents thereby guaranteeing all offspring were

202 homozygotes, or (2) the genotype was a no-call (i.e. the sample did not run) for one or both

203 parents. Not all SNPs were fixed between the two L. goodei populations, so we also genotyped

204 the F0 grandparents to clarify phases.

205

206 Linkage map creation

207 For each species, individual maps were constructed for each individual F2 family (14 for L.

208 goodei; 8 for L. parva) using JoinMap 4.1 (Li et al. 2008). Parental and grandparental genotypes

209 provided phase information for each locus. Grouping thresholds of LOD 4.0 (L. goodei) and 3.5

210 (L. parva) were used and markers significantly out of Hardy-Weinberg equilibrium (p<0.001)

211 were excluded. The Kosambi mapping function (Kosambi 1944) was used to convert

212 recombination frequencies to cM. For each species, a consensus map was constructed using the

213 Map Integration tool in JoinMap from the individual family map inputs. Linkage groups from

214 individual families were joined if they shared two or more markers. MapChart 2.2 (Stam 1993)

215 was used for graphical representation of the consensus map for each species. Lucania parva

216 linkage groups were numbered in descending order based on total length in cM. Lucania goodei

217 linkage groups were numbered based upon synteny with L. parva linkage groups (see below).

218

219 Synteny comparisons

220 To compare synteny between L. parva and L. goodei, SNPs designed from a common contig

221 were considered to be putatively orthologous. Consistent clusters of SNPs from common contigs

222 in both species provided evidence of synteny at the linkage group level (i.e. SNPs from the same

223 contigs clustered together in both species). Within linkage groups, marker order was compared

224 among orthologous SNPs by performing rank order correlations between species. Lucania

225 goodei and L. parva linkage groups were also compared to chromosomes from medaka (Oryzias

226 latipes), the most closely related species with a fully sequenced genome (both are in the

227 superorder Acanthopterygii (Steinke et al. 2006)). Sequences of contigs in the Lucania linkage

228 maps were blasted against medaka sequences using blastn. If a given contig had highly

229 significant blast hits (bit score > 100) against a single medaka linkage group, then its

230 approximate position was estimated in the medaka genome. However, if a given contig had

231 multiple, highly significant blast hits on multiple medaka linkage groups, then the orthologous

232 location in the medaka genome could not be determined. Hence, synteny was examined at the

233 level of linkage group clustering for the L. goodei and L. parva, L. goodei and medaka, and L.

234 parva and medaka comparisons. Synteny was examined at the level marker order for only the L.

235 goodei and L. parva comparison. False discovery rate (FDR) p-values were calculated for rank

236 order correlations using fdrtool in R (Strimmer 2008).

237

238 Flow cytometry

239 The genome size of both species was estimated using flow cytometry in order to determine how

240 much recombination was occurring per megabase. Four L. goodei and four L. parva individuals

241 were collected from the Lower Bridge site on the Wakulla River, Florida. The DNA content of

242 erythrocyte nuclei was measured using flow cytometry at Ursinus College (see Supplemental

243 Methods). Betta splendens was used as a standard.

244

245 Results

246 The two Lucania species differed in chromosome number. Lucania parva had 23 chromosomes

247 (Figure 2a). Twenty-two of these were acrocentric chromosomes of similar size, and one was a

248 metacentric chromosome approximately twice the size of the other chromosomes (denoted by an

249 arrow in Figure 2a). Lucania goodei had 24 acrocentric chromosomes (Figure 2b).

250 The number of chromosomes corresponded well to the number of linkage groups that

251 were recovered. For L. parva, 23 linkage groups were found, which matched the number of

252 chromosomes observed in the karyotype. Specifically, 766 SNP markers were resolved into 23

253 linkage groups (Table 1; Figure 3). Many of these SNP markers came from an EST that

254 corresponded to a known protein, many of which could be assigned a putative function

255 (Supplemental Table 1). The number of markers per linkage group (LG) ranged from 18 (LG

256 20) to 59 (LG 1). The total length of the map was 605 cM with the average linkage group being

257 26.3 cM (Table 1). Marker density was 1.27 markers per cM on average spanning from 0.67

258 (LG 7) to 2.22 (LG 18). Marker density was relatively consistent with the largest gap being 13

259 cM on LG 8 and with 32 gaps of 4 cM or larger across the map. Average genome size (C-value)

260 in L. parva was 1.423 pg (range: 1.396-1.450). One centimorgan in L. parva is thus

261 approximately 2.3 Mb.

262 Similarly, the number of linkage groups recovered for L. goodei matched the number of

263 chromosomes (1N=24). Significant linkages were found for 915 SNP markers making up 24

264 linkage groups (Table 1; Figure 4). Again, many of these SNP markers came from an EST that

265 corresponded to a known protein (Supplemental Table 1). The number of markers per linkage

266 group ranged from 4 (LG 21) to 66 (LG 2). The linkage map spanned 392 cM with the average

267 linkage group size being 16.33 cM (Table 1). Average marker density was 2.41 markers per cM,

268 ranging from 0.53 (LG 21) to 4.42 (LG 6). Marker density was consistent with the largest gap

269 being 8.7 cM on LG 23 and only 9 gaps of 4 cM or larger across the entire map. Average

270 genome size in L. goodei was 1.349 pg (range: 1.342-1.356). One centimorgan in L. goodei is

271 thus approximately 3.44 Mb.

272

273 Synteny of L. parva and L. goodei maps

274 The linkage groups were highly syntenic between L. goodei and L. parva allowing us to

275 unambiguously assign orthologous linkage groups. Across all linkage groups, we found 368

276 markers shared between the linkage maps that were from putatively orthologous ESTs. Figure 5

277 shows that markers from the putatively orthologous ESTs clustered together in the same linkage

278 groups in L. goodei and L. parva. Specifically, 364 (98.65%) markers (those from a common

279 contig and in both linkage maps) showed a pattern of synteny, while only 4 markers (1.09%)

280 deviated and clustered differently in the two species (Figure 5; Supplemental Figure 1).

281 Two linkage groups from L. goodei were syntenic with the largest linkage group in L.

282 parva, strongly suggesting that linkage group 1 represents the metacentric chromosome observed

283 in the L. parva metaphase spread. Figure 6 shows that linkage group 1A in L. goodei was

284 syntenic with the top portion of LG 1 in L. parva (matching at 24 markers: Figure 5), and LG 1B

285 was syntenic with the bottom portion (matching at 7 markers).

286 Synteny was less conserved at the level of marker order within linkage groups (Table 2).

287 Rank order correlations were high for some linkage groups, but were low and non-significant for

288 others. Across all linkage groups combined, marker order was highly correlated in L. parva and

289 L. goodei (Spearman rank order correlation: n = 364, r = 0.99, P < 0.0001). When rank order

290 correlations were performed on linkage groups separately (and corrected for multiple testing

291 using FDR), marker order was significantly correlated in 11 of 23 syntenic linkage groups (50%;

292 Table 2; note that LG21 lacked sufficient orthologous markers to test marker order). The marker

293 order correlations were not statistically significantly greater than zero for the other 12 linkage

294 groups. In order to determine if non-coding repetitive RNAs or slight differences in the order of

295 densely mapped markers were influencing this result, we repeated these synteny analyses using

296 only syntenic markers > 0.5cM apart in either species that had a single identified location in the

297 medaka genome . Again, we found significant marker order preservation in less than half (38%)

298 of the groups (6 out of 16 with more than 5 syntenic markers).

299

300 Synteny of Lucania and medaka

301 Synteny at the linkage group level was also well preserved between Lucania and medaka

302 (Supplemental Figure 2). Medaka has 24 chromosomes, and all Lucania linkage groups had a

303 single best match to a medaka chromosome (listed at the bottom of Figure 5). Between L.

304 goodei and medaka 559 of 585 (95.6%) orthologous markers were syntenic at the linkage group

305 level. Similarly, L. parva had 532 out of 545 (97.6%) orthologous markers that were found to be

306 syntenic with medaka chromosomes. Linkage group 1A in L. goodei corresponded to

307 chromosome 3 in medaka and linkage group 1B corresponded to medaka chromosome 11.

308

309 Discussion

310 Using high throughput Illumina Infinium genotyping assays, we created two SNP-based linkage

311 maps for Lucania parva and L. goodei with high marker density (1.26 markers/cM in L. parva

312 and 2.41 markers/cM in L. goodei). These linkage maps establish genomic resources for

313 Lucania and provide the groundwork for future linkage disequilibrium studies, QTL mapping,

314 molecular population genetic studies, and further synteny comparisons with other teleost species.

315 The fact that many of these SNPs came from ESTs whose protein functions are known in other

316 groups allows us to estimate the position of functionally important loci in Lucania (see

317 Supplemental Table 1). This is the first linkage map for any member of the Fundulidae family, a

318 group that exhibits an extraordinary ability to tolerate and adapt to physiological extremes

319 (Burnett et al. 2007).

320 By combining our linkage maps with actual chromosome counts, we found strong

321 evidence for a major chromosomal rearrangement between L. parva and L. goodei. Our

322 metaphase spread showed that a large metacentric chromosome was present in L. parva and

323 absent in L. goodei (see arrow in Figure 2). Our linkage maps indicate that the largest L. parva

324 linkage group (LG 1, 46.4 cM) is syntenic with two smaller L. goodei chromosomes (LG 1A and

325 1B). Comparisons of interspecific linkage maps have previously been used in other taxa to

326 identify chromosomal rearrangements between species (Tanksley et al. 1992; Burke et al. 2004).

327 Several pieces of evidence suggest that the metacentric chromosome is due to a Robertsonian

328 fusion of two smaller acrocentric chromosomes in the L. parva lineage rather than a

329 chromosomal fission event in the L. goodei lineage (a metacentric chromosome becoming two

330 acrocentric chromosomes). First, Fundulus parvipinnis is the closest relative to Lucania

331 (Whitehead 2010) and its karyotype is similar to L. goodei with 1N=24 (Chen 1971). Second, L.

332 goodei linkage group 1A is syntenic to chromosome 3 in medaka, and linkage group 1B is

333 syntenic to chromosome 11. The fact that each of these two linkage groups map to a single

334 linkage group in medaka suggests that they are unlikely to reflect the outcome of a fission event.

335 Our evidence for a chromosomal fusion clarifies a previous report (made without any supporting

336 data) that L. parva’s karyotype of 1N=23 deviates from the typicial fundulid karyotype of 1N=24

337 (Uyeno and Miller 1971). However, our evidence for a fusion should be further verified using in

338 situ hybridization (FISH) to determine that LG1 markers are indeed found on the metacentric

339 chromosome in L. parva.

340 With the exception of the fused chromosome, comparisons between our maps reveal that

341 large-scale structure is widely conserved between species. At the level of the linkage group, we

342 found high preservation of synteny between Lucania linkage groups. We also found

343 preservation of synteny between Lucania and medaka chromosomes. Each Lucania

344 chromosome could be assigned unambiguously as syntenic to a single medaka chromosome.

345 This is consistent with findings from other teleost linkage maps and genomes (Schartl et al.

346 2013). This degree of synteny will facilitate future comparisons with other related teleost species

347 for which genomic resources are emerging such as guppies (Tripathi et al. 2009; Fraser et al.

348 2011) and fundulids (Cossins and Crawford 2005; Burnett et al. 2007). Currently, the fusion we

349 document in Lucania (between chromosomes that are syntenic to medaka 3 and 11) is unique

350 among related fish species for which linkage maps exist. Guppies (Poecilia reticulata) also

351 possess a fused chromosome but it has occurred between two chromosomes that are syntenic to

352 medaka 2 and 21 (Tripathi et al. 2009). Tilapia (Oreochromis niloticus) possess two fused

353 chromosomes: one between chromosomes that are syntenic to medaka 2 and 4, another between

354 chromosomes syntenic to medaka 6 and 12 (Liu et al. 2013).

355 Our comparisons of synteny at the marker order level revealed significant marker order

356 preservation in only half of the linkage groups between L. parva and L. goodei. Our estimate of

357 50% collinear markers is much lower than a previous estimate of marker order preservation

358 between hybridizing populations, which found that 83% of markers between incipient species of

359 whitefish (Coregonus clupeaformis) were collinear (Rogers et al. 2007). Comparisons of

360 chinook salmon and rainbow trout (genus Oncorhynchus) also suggest high preservation of

361 marker order between congeners (Naish et al. 2013). Our study differs from these previous ones

362 since we used SNPs derived from ESTs rather than microsatellites and consequently had a much

363 denser distribution of markers on our map. However, further work is needed to determine if the

364 low marker-order synteny in Lucania is genuine or simply the result of (a) low number of

365 orthologous markers on some linkage groups, (b) genotyping errors, or (c) map construction

366 errors due to merging maps from multiple families within each species. If genuine, then the low

367 marker-order synteny found in Lucania would be indicative of small-scale rearrangements,

368 which could potentially contribute to the reduction of gene flow and the evolution of

369 reproductive isolation.

370 Conclusions

371 Our linkage maps for Lucania goodei and L. parva showed that a large-scale genomic

372 rearrangement has occurred between species. This Robertsonian fusion may have aided

373 divergence in these closely related species and helped to maintain species boundaries in zones of

374 contact by suppressing recombination in hybrids. Other fundulid species also differ in

375 karyotype, and our work may give insight into the mechanisms by which these changes occur.

376 These maps will enable the use of numerous genomic techniques to determine how reproductive

377 isolation has evolved in Lucania. The maps will also further a general understanding of the

378 evolution of many other unique traits in this group including vision, color pattern, and salinity

379 tolerance.

380

381 Acknowledgements

382 Leslie Noa and Jessie Coonley assisted with F2 crosses, fish rearing, and preservation. Matthew

383 Schrader and Leslie Noa assisted with RNA extractions. Mark Band, Tatsiana Akraiko

384 (Infinium assay) and Alvaro Hernandez (library preparation and sequencing) and other staff at

385 the Keck Center for Comparative and Functional Genomics at the University of Illinois

386 performed the sequencing and genotyping. James Walsh assisted with bioinformatics and SNP

387 calling; Lei-ting Li helped with Joinmap. Sairaghav Nissankula assisted with blast searches.

388 James Urton and Shaun McCann provided invaluable advice and help with the karyotype. Katie

389 Peichel provided helpful comments on the manuscript. This work was approved by the

390 University of Illinois IACUC (Protocol Nos. 08183, 09306, 11143). This work was funded by

391 the University of Illinois (Research Board Award 11153) and the National Science Foundation

392 (IOB 0445127, DEB 0953716, DEB 1110658).

393

394 Data archiving

395 The population specific Illumina transcriptome sequences are archived in Genbank 396 (bio-project ID: PRJNA215087). Source and probe sequences from the Illumina 397 Infinium chip are accessible on Dryad 398 (http://datadryad.org/resource/doi:10.5061/dryad.hv75h/9). 399 The 454 L. goodei reference assembly and the joinmap files are accessible on 400 Dryad with this paper (doi:10.5061/dryad.80822). 401

402

403 References 404 405 Backstrom, N., N. Karaiskou, E. H. Leder, L. Gustafsson, C. R. Primmer et al., 2008 A 406 gene-based genetic linkage map of the collared flycatcher (Ficedula albicollis) 407 reveals extensive synteny and gene-order conservation during 100 million years of 408 avian evolution. Genetics 179: 1479-1495. 409 Berdan, E. L., and R. C. Fuller, 2012 Interspecific divergence of ionoregulatory physiology 410 in killifish: insight into adaptation and speciation. Journal of Zoology 287: 283-291. 411 Burke, J. M., Z. Lai, M. Salmaso, T. Nakazato, S. X. Tang et al., 2004 Comparative 412 mapping and rapid karyotypic evolution in the genus Helianthus. Genetics 167: 449- 413 457. 414 Burnett, K. G., L. J. Bain, W. S. Baldwin, G. V. Callard, S. Cohen et al., 2007 Fundulus as 415 the premier teleost model in environmental biology: Opportunities for new insights 416 using genomics. Comparative Biochemistry and Physiology D-Genomics & 417 Proteomics 2: 257-286. 418 Butlin, R. K., 2005 Recombination and speciation. Molecular Ecology 14: 2621-2635. 419 Carleton, K. L., 2011 Quantification of transcript levels with quantitative RT-PCR, pp. 420 279-295 in Molecular methods for evolutionary genetics edited by V. Orgogozo and 421 M. V. Rockman. Springer, New York. 422 Chen, T. R., 1971 A comparative chromosome study of twenty killifish species of the genus 423 Fundulus (Teleostei: Cyprinodontidae). Chromosoma 32: 436-453. 424 Coluzzi, M., A. Sabatini, A. della Torre, M. A. Di Deco and V. Petrarca, 2002 A polytene 425 chromosome analysis of the Anopheles gambiae species complex. Science 298: 1415- 426 1418. 427 Cossins, A. R., and D. L. Crawford, 2005 Opinion - Fish as models for environmental 428 genomics. Nature Reviews Genetics 6: 324-333. 429 Duggins, C. F., A. A. Karlin and K. G. Relyea, 1983 Electrophoretic variation in the 430 killifish genus Lucania. Copeia: 564-570. 431 Falconer, D. S., and T. F. C. Mackay, 1996 Introduction to Quantitative Genetics. Pearson 432 Education Limited, New York, NY. 433 Faria, R., and A. Navarro, 2010 Chromosomal speciation revisited: rearranging theory 434 with pieces of evidence. Trends in Ecology & Evolution 25: 660-669. 435 Feder, J. L., F. B. Roethele, K. Filchak, J. Niedbalski and J. Romero-Severson, 2003 436 Evidence for inversion polymorphism related to sympatric host race formation in 437 the apple maggot fly, Rhagoletis pomonella. Genetics 163: 939-953. 438 Foley, B. R., C. G. Rose, D. E. Rundle, W. Leong, G. W. Moy et al., 2011 A gene-based SNP 439 resource and linkage map for the copepod Tigriopus californicus. BMC Genomics 440 12. 441 Fraser, B. A., C. J. Weadick, I. Janowitz, F. H. Rodd and K. A. Hughes, 2011 Sequencing 442 and characterization of the guppy (Poecilia reticulata) transcriptome. BMC 443 Genomics 12. 444 Fuller, R. C., 2008 Genetic incompatibilities in killifish and the role of environment. 445 Evolution 62: 3056-3068.

446 Fuller, R. C., and K. M. Claricoates, 2011 Rapid light-induced shifts in opsin expression: 447 finding new opsins, discerning mechanisms of change, and implications for visual 448 sensitivity. Molecular Ecology 20: 3321-3335. 449 Fuller, R. C., K. E. McGhee and M. Schrader, 2007 Speciation in killifish and the role of 450 salt tolerance. Journal of Evolutionary Biology 20: 1962-1975. 451 Fuller, R. C., and L. A. Noa, 2008 Distribution and stability of sympatric populations of 452 Lucania goodei and L. parva across Florida. Copeia: 699-707. 453 Gale, M. D., and K. M. Devos, 1998 Comparative genetics in the grasses. Proceedings of the 454 National Academy of Sciences of the United States of America 95: 1971-1974. 455 Hubbs, C. L., B. W. Walker and R. E. Johnson, 1943 Hybridization in nature between 456 species of American cyprinodont fishes. Contributions from the Laboratory of 457 Vertebrate Biology 23: 1-21. 458 Jaari, S., M. H. Li and J. Merila, 2009 A first-generation microsatellite-based genetic 459 linkage map of the Siberian jay (Perisoreus infaustus): insights into avian genome 460 evolution. BMC Genomics 10: 1. 461 Kasahara, M., K. Naruse, S. Sasaki, Y. Nakatani, W. Qu et al., 2007 The medaka draft 462 genome and insights into vertebrate genome evolution. Nature 447: 714-719. 463 Kitano, J., J. A. Ross, S. Mori, M. Kume, F. C. Jones et al., 2009 A role for a neo-sex 464 chromosome in stickleback speciation. Nature 461: 1079-1083. 465 Kosambi, D. D., 1944 The estimation of map distance from recombination values. Annals of 466 Eugenics 12: 172-175. 467 Kozak, G. M., R. S. Brennan, E. L. Berdan, R. C. Fuller and A. Whitehead, 2014 468 Functional and population genomic divergence within and between two Species of 469 killifish adapted to different osmotic niches. Evolution 68: 63-80. 470 Kozak, G. M., A. B. Rudolph, B. L. Colon and R. C. Fuller, 2012 Postzygotic isolation 471 evolves before prezygotic isolation between fresh and saltwater populations of the 472 rainwater killifish, Lucania parva. International Journal of Evolutionary Biology 473 2012: 523967. 474 Lee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister et al., 1980 Atlas of 475 North American freshwater fishes. North Carolina State Museum of Natural History, 476 Raleigh, N.C. 477 Lee, S. F., L. Rako and A. A. Hoffmann, 2011 Genetic mapping of adaptive wing size 478 variation in Drosophila simulans. Heredity 107: 22-29. 479 Li, H., J. Ruan and R. Durbin, 2008 Mapping short DNA sequencing reads and calling 480 variants using mapping quality scores. Genome Research 18: 1851-1858. 481 Liu, F., F. Sun, J. Li, J. H. Xia, G. Lin et al., 2013 A microsatellite-based linkage map of 482 salt tolerant tilapia (Oreochromis mossambicus x Oreochromis spp.) and mapping of 483 sex-determining loci. BMC Genomics 14. 484 Lubieniecki, K. P., S. L. Jones, E. A. Davidson, J. Park, B. F. Koop et al., 2010 485 Comparative genomic analysis of Atlantic salmon, Salmo salar, from Europe and 486 North America. BMC Genetics 11. 487 Lucas, M. R., N. N. Diop, S. Wanamaker, J. D. Ehlers, P. A. Roberts et al., 2011 Cowpea- 488 Soybean synteny clarified through an improved genetic map. Plant Genome 4: 218- 489 225.

490 McGraw, L. A., J. K. Davis, L. J. Young and J. W. Thomas, 2011 A genetic linkage map 491 and comparative mapping of the prairie vole (Microtus ochrogaster) genome. BMC 492 Genetics 12. 493 Ming, R., S. C. Liu, Y. R. Lin, J. da Silva, W. Wilson et al., 1998 Detailed alignment of 494 Saccharum and Sorghum chromosomes: Comparative organization of closely related 495 diploid and polyploid genomes. Genetics 150: 1663-1682. 496 Muchero, W., N. N. Diop, P. R. Bhat, R. D. Fenton, S. Wanamaker et al., 2009 A consensus 497 genetic map of cowpea [Vigna unguiculata (L) Walp.] and synteny based on EST- 498 derived SNPs. Proceedings of the National Academy of Sciences of the United States 499 of America 106: 18159-18164. 500 Naish, K. A., R. B. Phillips, M. S. O. Brieuc, L. R. Newton, A. E. Elz et al., 2013 501 Comparative genome mapping between chinook salmon (Oncorhynchus 502 tshawytscha) and rainbow trout (O. mykiss) based on homologous microsatellite loci. 503 G3-Genes Genomes Genetics 3: 2281-2288. 504 Noor, M. A. F., K. L. Grams, L. A. Bertucci and J. Reiland, 2001 Chromosomal inversions 505 and the reproductive isolation of species. Proceedings of the National Academy of 506 Sciences of the United States of America 98: 12084-12088. 507 Peichel, C. L., K. S. Nereng, K. A. Ohgi, B. L. E. Cole, P. F. Colosimo et al., 2001 The 508 genetic architecture of divergence between threespine stickleback species. Nature 509 414: 901-905. 510 Prince, J. P., E. Pochard and S. D. Tanksley, 1993 Construction of a molecular linkage map 511 of pepper and a comparison of synteny with tomato. Genome 36: 404-417. 512 Rieseberg, L. H., 2001 Chromosomal rearrangements and speciation. Trends in Ecology & 513 Evolution 16: 351-358. 514 Rieseberg, L. H., J. Whitton and K. Gardner, 1999 Hybrid zones and the genetic 515 architecture of a barrier to gene flow between two sunflower species. Genetics 152: 516 713-727. 517 Rogers, S. M., N. Isabel and L. Bernatchez, 2007 Linkage maps of the dwarf and normal 518 lake whitefish (Coregonus clupeaformis) species complex and their hybrids reveal 519 the genetic architecture of population divergence. Genetics 175: 375-398. 520 Schartl, M., R. B. Walter, Y. J. Shen, T. Garcia, J. Catchen et al., 2013 The genome of the 521 platyfish, Xiphophorus maculatus, provides insights into evolutionary adaptation 522 and several complex traits. Nature Genetics 45: 567-U150. 523 Stam, P., 1993 Construction of integrated genetic linkage maps by means of a new 524 computer package - Joinmap. Plant Journal 3: 739-744. 525 Stapley, J., T. R. Birkhead, T. Burke and J. Slate, 2008 A linkage map of the zebra finch 526 Taeniopygia guttata provides new insights into avian genome evolution. Genetics 527 179: 651-667. 528 Steinke, D., W. Salzburger and A. Meyer, 2006 Novel relationships among ten fish model 529 species revealed based on a phylogenomic analysis using ESTs. Journal of 530 Molecular Evolution 62: 772-784. 531 Strimmer, K., 2008 fdrtool: a versatile R package for estimating local and tail area-based 532 false discovery rates. Bioinformatics 24: 1461-1462. 533 Tanksley, S. D., M. W. Ganal, J. P. Prince, M. C. de Vicente, M. W. Bonierbale et al., 1992 534 High density molecular linkage maps of the tomato and potato genomes. Genetics 535 132: 1141-1160.

536 Timusk, E. R., M. M. Ferguson, H. K. Moghadam, J. D. Norman, C. C. Wilson et al., 2011 537 Genome evolution in the fish family salmonidae: generation of a brook charr genetic 538 map and comparisons among charrs (Arctic charr and brook charr) with rainbow 539 trout. BMC Genetics 12. 540 Tripathi, N., M. Hoffmann, E. M. Willing, C. Lanz, D. Weigel et al., 2009 Genetic linkage 541 map of the guppy, Poecilia reticulata, and quantitative trait loci analysis of male size 542 and colour variation. Proceedings of the Royal Society B-Biological Sciences 276: 543 2195-2208. 544 Uyeno, T., and R. R. Miller, 1971 Multiple sex chromosomes in a Mexican cyprinodontid 545 fish. Nature 231: 452-453. 546 Wang, J., B. Roe, S. Macmil, Q. Yu, J. E. Murray et al., 2010 Microcollinearity between 547 autopolyploid sugarcane and diploid sorghum genomes. BMC Genomics 11: 261. 548 Whitehead, A., 2010 The evolutionary radiation of diverse osmotolerant physiologies in 549 killifish (Fundulus Sp.). Evolution 64: 2070-2085. 550 551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

Table 1. Summary of integrated linkage maps for L. parva and L. goodei. L. parva L. goodei # of chromosomes 23 24 # of linkage groups 23 24 Map size (cM) 605 392 Average linkage group size (cM) 26.3 16.3 Markers per cM 1.26 2.41 Total # of markers 766 915 # individuals 161 303

Table 2. Marker order correlations for syntenic markers on each linkage group in Lucania parva and L. goodei.

Size L. Size L. Number of FDR Spearman LG parva goodei syntenic P –value corrected r (cM) (cM) markers p-value A:17.1 24 0.8183 0.000001 0.000084 1 46.4 B:13.9 7 0.607 0.1362 0.251906 2 45.5 22 27 0.835 0.000001 0.000084 3 38.7 28.6 13 0.4341 0.137904 0.251906 4 34.8 27.9 22 0.9198 0.000001 0.000084 5 33.5 26 28 0.7225 0.000014 0.007719 6 30.2 11.3 14 0.033 0.914228 1.000000 7 28.5 7.2 7 0.785714 0.0536 0.251906 8 27.9 17.5 11 0.6818 0.02071 0.030374 9 26.7 8.3 7 0.214286 0.5962 1.000000 10 26.5 11.6 14 0.3626 0.20193 1.000000 11 25.5 14.6 13 0.1923 0.529033 1.000000 12 25.2 22.2 25 0.5354 0.005817 0.030374 13 22.7 13.8 17 0.348 0.170412 0.251906 14 22.6 18.4 14 0.5956 0.02454 0.094289 15 21 13.2 21 0.441 0.039729 0.094289 16 20.6 15.3 18 0.5542 0.017116 0.030374 17 20.4 14.1 7 0.142857 0.7264 1.000000 18 20.3 30 22 0.8701 0.000001 0.000084 19 20.2 17.9 15 0.7214 0.002382 0.007719 20 19 9.5 10 0.3455 0.328748 1.000000 21 18.8 7.5 2 N/A N/A N/A 22 17.9 22.1 13 0.7637 0.002393 0.021931 23 13.1 21.4 12 0.6783 0.0153 0.030374

568

569

570

571

572 Figures

573 Figure 1. Procedures used to generate SNPs from orthologous ESTs for linkage maps. 574 575 Figure 2. Somatic metaphase/anaphase spread of (a) Lucania parva and (b) L. goodei. The 576 arrow in (a) indicates the fused chromosome. 577 578 Figure 3. Lucania parva linkage map. Numbers on the right of each linkage group indicate the 579 marker name and numbers on the left indicate the position in centimorgans (cM). 580 581 Figure 4. Lucania goodei linkage map. Numbers on the right of each linkage group indicate 582 the marker name and numbers on the left indicate the position in centimorgans (cM). 583 584 Figure 5. Summary of synteny comparisons between L. parva and L. goodei linkage groups. 585 Bolded numbers along the diagonal show the number of orthologous SNPs on the linkage 586 groups. Numbers off the diagonal are non-syntenic markers. Identity of syntenic medaka 587 chromosomes listed along the bottom row. 588 589 Figure 6. Synteny of linkage group 1 in L. parva and 1A,B in L. goodei. Orthologous SNPs 590 between L. parva linkage group 1 (P: middle) and L. goodei (G) linkage groups 1A (far left) and 591 1B (far right) are shown connected by lines. SNPs syntenic with L. goodei 1A are highlighted in 592 red; SNPs syntenic with L. goodei 1B are highlighted in blue. 593 594 Supplemental Figure 1. Synteny between linkage groups 2-23 in L. parva and L. goodei. For 595 each linkage group, the L. parva group is on the left, L. goodei on the right. Orthologous SNPs 596 between species are highlighted in red and connected by a line. 597 598 Supplemental Figure 2. Summary of synteny comparisons between Lucania and medaka 599 linkage groups. 600 Bolded numbers along the diagonal show the number of orthologous SNPs on the linkage 601 groups. Numbers off the diagonal are non-syntenic markers. (a) synteny between L. parva and 602 medaka, (b) synteny between L. goodei and medaka. 603 604 Supplemental Table 1. Protein coding matches for linkage map markers and their location on 605 the L. goodei and L. parva linkage groups. 606

607

608

609

610

611

Figure 1

Step 1. Isolate RNA from gills, fins, eyes, brains, ovaries, and testes for males and females from 4 populations (2 L. goodei populations and 2 L. parva populations) using 10 individuals from each population. Step 2. Create pooled samples for each population using the same amount of RNA from each individual. Step 3. Prepare cDNA libraries uniquely barcoded by population. Step 4. Sequence 100-bp paired-end libraries with Illumina. Step 5. Trim ends and check sequence quality. Step 6. Use contigs from previous 454 sequencing of L. goodei (Fuller and Claricoates 2011) as a reference for assembling all Illumina sequences. Step 7. Assemble each of 4 populations separately (L. parva: Pecos River and Indian River; L. goodei: Upper Bridge and Everglades) using Novoalign. Step 8. Export the assemblies to MAQ software. Find diagnostic, population-specific SNPs using MAQ's SNP caller. Step 9. Assess SNP quality and choose SNPs from contigs that have all types of SNPs (L. goodei population-specific, L. parva population-specific, between species). Step 10. Genotype mapping families at all SNPs using custom designed Iillumina Infinium Bead Chip. Step 11. Create linkage maps for each species in JoinMap. Figure 2 A.

B. Figure 3 LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8 LG9 LG10 LG11 P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14

0.0 02752 0.0 13959 11480 11075 10824 0.0 02482 0.0 6.3 11787 0.3 13911 12613 13471 28526 0.0 10795 12427 8.9 14008 08900 00091 0.0 15982 0.4 0.5 11778 0.8 12438 1.1 05740 10.1 04303 14034 2.8 10924 0.0 11493 1.4 09794 4.6 13091 2.7 14191 12.9 09168 1.8 13938 01430 10.8 13005 2.6 11384 00047 1.9 16681 6.3 13631 4.6 11749 13.4 10797 10940 16266 2.0 12306 12.8 11211 0.0 12943 0.0 13050 0.0 14734 0.0 11856 06307 5.7 12464 09049 0.0 14977 0.0 05705 0.0 2.2 15163 7.0 12059 09478 5.2 16794 13.9 12490 12588 4.1 13869 14.6 11521 7.9 22458 0.6 12007 6.4 11968 6.0 13928 14.7 12840 12063 8.4 16823 16.0 12358 01319 20483 02552 5.4 2.4 08984 2.4 15370 6.2 14083 10909 14.9 17109 8.2 1.8 10.3 20379 10.3 11307 17.8 12194 3.0 19419 17433 7.1 12496 6.4 11087 9.3 05074 15.0 19878 10.6 12489 10.9 10980 18.2 15345 7.1 14891 10.7 14663 11.9 13441 4.9 19075 9.7 13658 15.1 12040 8.4 10838 5.3 12036 10.7 14432 12.3 05651 20.4 11248 9.3 13068 7.6 11774 5.4 12888 11.0 12049 16.7 12170 10.4 12894 15.2 11207 09418 09969 13245 7.5 ID_48342 13.1 14758 13.0 12638 20.8 11321 16663 9.8 16032 8.3 6.0 00893 7.8 11361 11.3 07303 17.2 02823 00038 13135 16758 07283 11015 09980 8.5 02735 05674 13.4 12530 14.0 12090 21.0 11801 7.6 11120 11.1 15.3 14.8 13073 8.5 9.2 12373 6.4 14607 15393 12.9 11374 17.5 11887 12458 11483 12206 10894 10.7 05824 15.2 10994 09016 21.2 12893 15.2 11002 8.6 12772 13884 9.6 13676 13.7 05021 14.7 18.1 11267 11.9 12844 15.4 27975 9.0 11305 8.7 13284 11511 10.8 11904 10991 21.3 12704 15.5 07072 10.1 15299 12.4 10899 15.3 14.9 13681 18.5 11284 12.9 12491 14845 09523 9.2 09172 9.5 23094 12.3 11446 14.5 15717 21.7 09312 15.5 17.2 11435 10.2 00141 11408 13.1 15387 16.0 11802 15.1 19824 18.8 15343 13374 13.1 11253 05010 06150 13627 13089 12.6 10958 10975 14.7 12472 22.0 08901 11.9 19868 18.2 10765 10.3 05769 9.7 15.1 02595 11.2 16494 16.1 23772 11846 11892 19.5 03801 03840 13475 15.6 16735 05767 13.0 09013 19589 12152 15.3 22.2 11673 14.4 18.7 20489 11.2 10.0 03705 08909 10917 14.8 11614 12128 04565 19.6 16492 06235 15.7 11182 09967 15.7 13.3 10844 11880 22.6 15908 19.1 06034 11.6 10.1 11215 11104 14.3 12701 16.2 09080 15093 15.9 11542 20.0 03425 14.5 11371 15.8 09306 12361 16488 10896 12492 24.2 10875 18.3 03541 20.2 16973 04516 11.9 22399 10.4 13027 15.3 06375 12134 13.4 15.3 02557 16.0 14271 20.2 03807 14.7 11620 15.9 10949 15.8 06554 18043 09040 25.0 18463 19.0 14664 23.0 15336 10.7 11522 15.5 11407 05998 13.5 05912 16.7 12239 12461 13486 20.3 00466 14345 14.8 13378 16.1 16263 16.5 06876 15.6 13179 16.3 26.1 11472 05972 19.4 09090 23.3 21373 07328 12.4 07271 15.6 23542 14.0 12218 16.9 11932 20.7 12202 15.1 15733 16.3 05885 17.3 18617 12400 17.3 11625 26.5 14739 20.6 18112 16912 23.5 11255 14.9 12800 15281 15.6 18255 16.3 21.5 15856 15.5 11696 11070 16.4 11812 17.8 12997 17.2 17.2 11077 17.9 11688 26.8 11004 20.8 09531 23.6 16476 16.8 09348 21.6 11355 04467 16.0 11335 16.6 05713 18120 18.3 10986 18.4 11784 17.4 10883 18.0 11224 27.4 11122 20.9 10291 10764 23.7 09326 17.1 15226 22.0 17258 02448 17.3 06026 12955 18.6 13607 19.8 06302 17.7 15178 18.9 06130 13326 11468 21.1 11116 12637 16.3 24.1 11661 19.0 17.9 17367 27.6 22.1 10854 12111 17.6 13695 19.5 10865 18.9 15434 19.0 15126 21053 21.2 11092 16.9 24.2 19877 20.8 13980 11023 02149 22.8 14450 17.0 13969 17.9 10835 15619 12927 14594 20.3 19903 22.1 11869 18.2 19.0 11094 19.6 11621 27.8 16083 21.3 10789 10974 14316 24.1 21.5 11840 11844 12407 23.2 11291 17.2 06346 19.5 12447 24.4 05334 21.9 12699 22.6 12720 19.4 06866 24294 15563 12256 01687 12035 06697 18.8 11272 19.7 28.1 23.3 10944 17.3 18999 12265 19.6 12694 20.0 15839 14210 21.4 12196 09140 24.6 12283 14802 12898 20.4 03455 23.5 09038 18.9 11564 19.7 03271 08954 25.2 10926 19.0 28.3 12852 09384 24.8 10919 25.5 14917 21.1 09649 20.6 11258 11095 13729 19.0 14404 20.8 12192 19.7 26257 28.8 11850 23.6 21.7 08990 25.3 09686 12541 28.5 12378 26.7 17870 26.5 12723 21.3 10939 22.1 19322 11254 09333 20.3 13823 12590 22.3 21946 20.4 15121 11414 28.9 09707 03564 26.4 01326 04225 11213 13217 21.9 15799 22.6 11001 24.1 13694 21.2 13322 22.4 10816 25.5 27.9 20.6 11629 11303 29.7 11691 26.6 12021 12841 11127 09024 22.9 12415 22.6 00230 19932 22.4 17547 29.8 12962 27.6 15036 26.4 11453 25.2 10833 12023 23.0 12789 22.7 10823 25.2 11527 30.0 13507 29.1 20951 27.8 18260 19221 11571 23.3 14930 20795 11108 31.2 11183 23.7 30.2 16872 25.3 06983 28.8 25372 16114 25.3 14229 11352 12158 35.4 06383 16224 30.3 25.7 15948 14036 28.6 05731 00597 09467 29.8 27.1 17519 11398 30.0 11474 11061 30.3 28.4 11315 14732 14336 31.0 12851 30.5 09278 11915 38.7 12735 34.8 28.9 06908 02380 31.8 06421 31.1 13716 31.2 13976 10995 09144 31.5 10904 32.1 06709 33.1 09003 12051 33.2 13257 34.1 11167 10981 34.0 06479 36.7 16141 13263 13463 35.3 11533 08925 38.9 10955 43.9 15026 11820 17015 39.6 01889 40.7 14093 45.5 10971 41.3 02050 46.4 11593 LG12 LG13 LG14 LG15 LG16 LG17 LG18 LG19 LG20 LG21 LG22 LG23

P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 P18 P19 P20 P21 P22 P23

0.0 04023 0.0 13959 11480 11075 10824 0.0 11574 11031 0.0 02482 0.0 0.4 15543 0.3 13911 12613 13471 28526 4.7 11853 0.0 17583 08900 00091 0.0 04077 1.0 15704 0.4 0.5 11778 0.8 12438 5.4 15419 0.0 14744 0.5 05942 14034 1.0 13007 13468 15224 10818 0.0 11493 1.4 09794 4.6 13091 5.6 23861 1.1 0.0 00469 0.3 11150 1.8 12162 1.8 13938 01430 3.4 22018 14084 14353 0.0 10829 05265 2.6 11384 00047 1.9 16681 6.3 13631 6.0 11134 0.0 0.0 0.4 10884 18411 2.4 05931 13663 15470 17179 10940 16266 2.0 12306 3.6 06641 18960 00137 11504 06893 15110 3.7 09916 06625 4.7 5.7 12464 09049 0.0 14977 0.0 05705 0.0 2.2 15163 7.0 12059 09478 6.1 11401 3.5 1.2 0.5 0.9 07394 2.7 18807 14384 12588 4.1 13869 11029 06020 12401 17978 6.5 11117 4.8 05762 7.9 22458 0.6 12007 6.4 11968 8.4 16823 6.7 13317 3.9 0.8 09114 1.8 23270 2.8 21091 5.4 12063 01441 12538 1.3 12185 7.7 12544 6.2 14134 17675 01319 20483 1.8 02552 10.3 20379 10.3 11307 8.8 18934 5.2 15829 1.5 04279 2.2 03556 2.9 12539 8.2 7.1 12496 13997 09593 1.9 11562 9.3 16166 16472 01218 10836 17433 10.6 12489 10.9 10980 9.0 17027 12219 5.8 18723 2.2 16925 2.3 11466 3.0 11889 11581 6.4 10.7 14663 4.3 2.4 14464 11794 11259 09436 8.4 10838 10.7 14432 05651 11701 13902 9.8 04992 6.0 12859 12242 10.1 2.8 12082 2.4 13169 11035 5.3 12036 5.4 12888 11.0 12049 12.3 15035 11300 3.4 7.0 07138 09969 13245 14758 11436 16693 6.2 12396 4.1 13022 13057 7.5 ID_48342 00893 13.1 13.0 12638 22194 10950 05745 11336 3.2 3.3 03326 3.5 15719 16922 02735 05674 6.0 7.8 11361 11.3 07303 10964 6.5 22393 11051 8.4 07220 10.8 21039 8.6 09070 8.5 9.2 12373 13.4 12530 14.0 12090 10.0 10886 20448 11695 4.4 15785 3.7 03548 6.4 14607 15393 10.7 05824 12.9 11374 4.7 13662 6.7 14775 10.1 15666 4.6 11154 9.9 13543 12206 10894 9.6 13676 15.2 10994 14.7 09016 11076 5.9 09600 12392 10.8 11904 13.7 05021 6.8 08929 09320 10.5 14690 11.0 11651 06623 6.3 10.0 09694 9.0 11305 10899 8.7 13284 11511 15.3 10991 14.9 13681 10.2 12642 05133 6.9 6.7 6.4 11877 12.4 11446 14.5 15717 11.4 25127 11.2 12180 18101 6.6 16360 09784 9.2 09172 9.5 23094 12.3 16.0 11802 15.1 19824 7.0 08930 10.8 11235 9.2 12847 8.1 10782 10.4 13.1 15387 14.7 12472 11.6 09448 11.3 09709 6.8 11097 9.7 13089 12.6 10958 10975 16.1 23772 11846 11892 9.0 14075 11014 12364 9.7 10776 11.4 16885 10.5 16039 15.1 02595 11.2 16494 19589 12152 11.0 13992 13.0 18519 7.0 06295 10.0 03705 13.0 09013 14.8 11614 12128 15.3 9.1 04117 03317 01506 10.0 14398 12.1 11.6 11576 10.7 09244 08909 10917 11880 04565 13.1 13195 11.5 06712 10825 10.1 11215 15.7 12701 13.3 10844 9.2 11049 12200 12.4 10963 12.7 12259 7.8 10.9 11973 03868 11104 14.3 16.2 09080 15093 15.9 11542 11.2 13600 13.5 11779 11.6 16793 10.4 13027 13.4 16488 10896 12492 9.3 09043 14106 12.8 13025 9.3 16976 11.0 10808 09136 15.8 06554 15.3 06375 12134 15.3 02557 16.0 14271 11.3 11722 13536 11.8 06901 11522 13.5 05912 18043 09040 9.4 06037 13.3 11081 14.8 9.7 12583 10.7 16.5 06876 15.5 11407 05998 16.7 12239 16.3 12461 13486 12.6 05649 15.5 12208 11.1 13798 14.0 12218 15.6 13179 13.9 06271 13.9 12937 11791 15.4 12084 11.6 18177 12.4 07271 17.3 18617 15.6 23542 16.9 11932 17.3 11625 03341 13310 11.2 26404 13748 15.6 18255 16.3 12400 14.7 13282 13.9 11732 10798 16.2 12504 18.6 13132 13.5 12030 14.9 12800 17.8 12997 17.2 15281 17.2 11077 17.9 11688 06263 12.2 10873 16.8 09348 15.2 18385 14.5 12621 19282 17.0 12144 14098 11373 17.4 12047 18.3 10986 18.4 11784 17.4 10883 18.0 11224 15.2 09028 18.7 18.8 23117 09424 13.1 04085 17.1 15226 17.5 18302 13872 13455 17.2 17320 13220 11588 11866 19.0 12955 18.6 13607 19.8 06302 17.7 15178 18.9 06130 15.5 11356 17.9 17367 14069 20.4 18587 11519 14.6 15095 19.1 12262 06223 17.6 13980 19.5 10865 18.9 15434 19.0 15126 18.1 15.7 21126 19.0 10800 20.8 11023 02149 15.1 24372 19.8 10900 11840 20.3 19903 22.1 11869 18.2 19.0 11094 19.6 11621 19.8 21162 16.0 11538 17.7 14246 21.5 11844 12407 15.2 11084 11619 20.2 09583 21.9 12699 22.6 12720 19.4 06866 19.7 24294 15339 15926 16.3 10999 17.9 12528 18.8 11272 20.4 15.3 13011 20.0 15839 20.4 03455 05953 10877 16.7 19046 19.0 14802 12898 15.4 11905 13348 25.2 10926 25.5 14917 21.1 09649 20.6 11258 21.0 12096 03545 16.8 07450 19.7 26257 15.5 11683 17870 26.5 12723 21.3 10939 22.1 19322 11254 20.6 19780 26.7 20.4 15121 11414 11971 21.9 15799 22.6 11001 15.7 27.9 13217 20.6 11629 11303 16.0 13953 13583 22.4 17547 16.1 15496 25.2 11527 18.2 02361 19.0 20499 20.3 12756 Figure 4 LG1A LG1B LG2 LG3 LG4 LG5 LG6 LG7 LG8 LG9 LG10 LG11

G1A G1B G2 G3 G4 G5 G6 G7 G8 G9 G10 G11 G12 G13

0.0 10924 0.0 13441 0.0 07153 0.0 14375 10787 13702 0.0 17442 0.9 04217 1.4 02823 0.6 25230 05815 09294 0.0 11954 0.0 13038 0.0 0.6 06742 0.8 11023 1.3 13121 3.5 11908 2.5 04303 15487 2.7 10847 1.1 11511 0.6 1.7 08959 0.0 14643 4.1 22875 2.9 14008 0.7 07328 16872 0.8 05912 1.2 11880 3.0 05853 0.0 13591 10926 1.5 14607 2.2 12358 2.8 13643 6.0 13374 4.6 06361 0.8 06697 12424 05812 1.7 11272 3.5 09433 1.9 12997 2.8 15393 1.4 0.0 16681 2.4 14413 3.7 11506 6.3 11284 05025 4.8 08895 0.9 16476 0.0 14891 10940 2.6 10349 11116 10795 10894 12464 2.5 27933 2.9 12701 11778 11480 2.9 11211 3.8 14248 6.4 11887 11209 0.0 0.0 5.2 10797 13749 29135 0.5 14667 4.4 1.5 16488 10339 2.9 12407 2.3 1.1 11522 13450 3.4 11104 3.0 14454 13911 4.1 03576 4.2 26343 7.1 14340 6.3 18120 11739 0.9 00061 14469 3.2 03685 5.1 09049 3.3 00893 1.7 3.1 08923 4.3 12876 6.2 09467 8.0 11267 7.5 17994 13397 10807 1.4 09123 3.5 10917 3.5 09388 6.0 11042 11686 1.8 08979 15799 10939 03667 6.5 19517 15958 11812 05126 11458 2.0 06307 11610 3.7 16817 3.4 12888 3.7 08900 4.0 5.2 9.3 9.2 1.2 8.1 10838 1.9 14917 12588 7.3 10851 2.6 09418 4.1 11327 3.8 12007 06305 13869 13555 4.5 09794 5.3 11482 9.8 03807 9.3 11038 00146 15094 2.0 03623 13435 3.8 16896 11202 16956 8.4 4.3 09984 3.9 05031 4.6 11077 5.4 12194 8.1 9.9 25105 11.4 05885 26509 13415 2.7 2.1 16266 14530 4.3 12306 11061 1.3 9.6 20483 4.5 11713 4.0 14647 4.8 17459 5.5 12704 10.9 06748 6.8 10909 11.8 13949 10919 03467 11408 12552 2.2 12218 4.4 02482 8.6 11915 3.2 11.4 10842 02595 10956 4.2 05705 23094 5.0 17741 5.6 11549 11.3 11355 12.2 05076 12.2 17591 1.4 12324 13844 4.9 2.8 10844 4.5 22399 8.8 13403 14093 7.7 20284 11.5 07271 16009 4.4 02552 5.1 02161 08913 6.3 15908 11.4 10854 14.9 17639 12.3 06026 1.6 11661 5.4 05764 11814 3.7 12286 11744 23186 11673 8.9 24194 09333 10.1 10766 10949 19462 11.7 13089 6.1 12446 4.6 22106 4.8 5.2 12006 6.7 13.9 15.0 12458 13.0 1.7 5.8 09967 3.9 10832 12027 16987 9.2 11593 16562 10.4 05863 22954 12288 11453 12.0 13429 6.3 14605 4.7 11039 5.3 06349 12470 7.4 14.6 16.2 06992 13.3 1.8 6.1 15299 4.6 10938 4.9 00091 7.6 11175 9.4 15228 14.8 11803 10.5 12637 13.4 05713 10848 2.1 11312 12.4 11044 6.4 01429 5.4 12134 5.5 14429 10991 16.4 11609 13138 6.3 05769 4.7 14038 5.0 12496 8.1 15345 9.7 11108 14.9 13999 06200 10.6 12125 12907 03779 12726 12361 2.2 15992 12.6 13399 6.7 11662 6.3 13284 5.6 10883 17.2 13.7 7.1 14308 4.8 10975 5.3 12152 8.4 12332 10.0 03504 15.1 12589 11.4 03541 05635 18.3 15125 09946 09804 09306 2.6 14623 12.8 11230 6.8 09234 7.0 21090 11614 10987 7.2 06790 5.5 11446 5.4 11238 5.7 8.5 11468 10.2 13716 12821 15985 11.8 15697 19.6 08898 14.0 12040 11182 4.3 05957 13.4 12014 7.8 15574 7.3 13607 06370 15.4 6.3 03057 6.0 11882 8.7 11472 12122 10.9 16474 11917 11.9 15627 19.7 11890 11371 14.1 07283 17066 4.4 09776 10765 13.8 10943 8.2 18617 7.9 ID_203792 5.8 11932 05734 6.4 11361 6.2 10056 9.0 11004 05972 11.0 17337 15.6 10833 12.1 09090 19.8 13120 14.2 12563 4.6 10783 14.6 14869 8.3 12955 8.4 07780 5.9 05834 10828 05824 13778 27198 9.1 13243 11.6 10870 15.7 06439 12.4 10973 19.9 13378 14.3 15567 5.2 12748 00567 11305 8.6 02034 6.5 6.3 6.1 12239 14.9 02495 13537 9.2 10875 11.7 11119 15.9 10944 12.5 08990 20.3 11070 14.4 12630 5.4 19942 12541 12052 27854 9.6 13409 6.7 6.6 11802 11968 12828 6.4 13938 9.5 21053 12.2 19951 16.0 11135 13.3 09140 20.4 06096 14.6 14845 5.5 00135 15.0 09929 9.8 16243 7.3 7.8 12749 00005 11904 6.7 12472 9.6 14210 12.9 16910 17258 13389 13.4 20.5 11885 14.8 15765 5.6 04981 11322 15.3 14053 10.0 12425 9.2 9.5 11317 16.2 03290 11138 6.9 11558 09348 9.7 10961 13.0 15150 13830 14.2 20.6 18735 15.6 19340 5.7 09686 15.4 15429 10.1 17745 9.7 10.2 09649 26475 12018 7.0 13179 10.6 18463 13.8 17089 16.3 06983 18380 14.6 20.8 13928 16.4 11161 5.8 04982 15.5 12800 12699 14758 10.2 14.6 12244 06580 13461 11.1 10.7 03579 12852 13.9 11382 11291 06709 14.7 18426 00024 20.9 12894 16.8 19317 5.9 10974 15.6 06256 00036 7.1 11.6 14219 13545 13.6 15206 11.0 05688 15026 28398 14.9 06052 21.1 12480 03016 03700 6.1 20287 15.7 12073 09475 16.4 17.3 7.2 14941 11.1 11122 12714 11071 15.1 08992 11324 21.2 11755 00152 20340 6.6 12520 15.8 10935 09216 10789 7.3 09040 11.4 09707 12893 11901 11533 15.2 07510 13135 17.5 03271 6.7 12062 15.9 13217 13099 15.3 12196 12316 21.3 7.5 16477 15386 11.5 12728 15973 10815 09038 11564 12265 18.7 07495 6.8 12283 16.0 12676 16.5 11974 10880 7.6 11499 11.7 11691 03564 10971 13976 15.5 21.5 14521 09602 19.2 21946 7.0 11213 16.1 13061 12269 7.8 10896 11.8 03260 04976 13447 21.6 11335 19.3 00230 04019 8.0 06034 16.2 06872 16.6 15.6 10771 7.9 14549 12.1 03655 15563 09024 21.7 02448 19.5 09743 9.3 12198 16.4 21585 12.4 16419 11315 08993 15.7 11396 06283 15692 05995 19.6 10995 9.8 11681 8.0 11288 16.7 22.0 16.5 09066 12.7 06511 11127 01889 15.8 16588 12111 05731 12192 11.3 01640 8.6 26257 22.1 19.7 17.5 11781 12863 13.2 ID_11485 16.8 17916 09513 15.9 06812 23.5 09841 11483 9.7 13879 13.8 11850 16.9 04597 15328 16.2 09222 23.6 03010 19.8 08954 19932 10.6 11303 14.5 02720 17.0 11125 18.8 12777 23.8 00138 20.0 11474 11.3 11629 16.3 14230 17.1 11167 14356 18.9 11366 11183 25.4 20225 25372 20.8 09228 11.9 12898 14804 17.1 13287 17.4 06683 19.2 12735 25.5 13823 21.2 11526 12.0 11050 17.9 24860 20.8 13460 26.0 12590 21.3 11671 12.7 11414 18.1 10821 21.4 15036 26.3 11571 21.7 10981 13.0 14718 11844 11354 18.3 17519 22.2 26.7 00172 21.8 09144 13.2 17367 10932 18.8 11214 23.0 27.4 16092 13322 21.9 15119 16.2 14802 23.7 11454 27.6 14336 16114 22.2 11310 22.2 09415 24.0 11225 27.7 14036 22.3 05153 24.2 12021 27.9 18857 24.1 14229 28.6 11120 LG12 LG13 LG14 LG15 LG16 LG17 LG18 LG19 LG20 LG21G21 LG22G22 LG23G23 0.0 10989 G7 G8 G9 G10 G11 G12 G13 G14 G15 G16 G17 G18 G19 G20 0.1 16952 1.8 10881 0.0 17442 0.0 ID_15917 2.4 12162 0.0 11954 0.0 13702 0.0 13038 0.8 11023 0.6 20361 2.6 06906 2.7 10847 0.6 06742 1.1 11511 1.2 11880 1.7 08936 02462 05670 2.7 05122 3.0 05853 0.8 05912 0.0 ID_134768 11154 0.0 11889 0.0 13591 10926 1.5 14607 1.7 11272 1.9 06641 0.0 11031 11574 0.0 04023 0.0 0.0 11877 3.3 04653 12917 3.5 09433 12424 05812 16681 0.1 10968 06625 0.8 09048 0.0 14891 1.9 12997 2.8 15393 1.4 2.6 10349 0.0 2.1 09274 25886 1.3 06893 2.0 10894 12464 10940 1.1 14993 0.0 16313 04279 3.4 05942 0.5 14667 4.4 2.5 27933 2.9 12701 12407 11778 11480 2.4 11430 0.2 19481 1.4 12185 2.1 0.9 11581 14384 11522 13450 1.5 16488 10339 2.9 2.3 2.3 23810 0.7 11117 12544 3.5 11616 00061 3.4 11104 3.0 14454 13911 0.0 09669 2.5 13677 17986 10999 1.7 14690 2.4 12767 1.7 02830 0.9 1.7 14469 3.2 03685 0.3 3.1 11756 11286 17868 5.1 09049 3.5 10917 3.3 00893 3.1 08923 0.7 11075 11608 2.0 13992 11033 0.8 2.8 13022 3.6 07138 1.4 09123 1.8 08979 3.5 09388 2.7 3.1 11401 10892 16417 6.0 11042 11686 3.7 16817 3.4 12888 4.0 15799 10939 3.2 10818 3.1 01918 3.8 11531 2.0 06307 11610 14917 12588 3.7 08900 3.5 12086 3.2 11853 2.1 15879 12534 8.1 10838 3.8 12007 06305 1.9 4.5 09794 3.3 17942 1.0 3.4 17817 3.7 11577 27019 2.6 09418 4.1 11327 3.8 13869 13555 11621 3.6 04077 3.6 15419 2.8 09448 4.7 8.4 15094 2.0 03623 13435 11077 4.4 3.4 13400 01286 14744 2.7 16956 4.3 09984 3.9 05031 4.3 12306 4.6 4.2 13010 11903 3.9 09411 23861 2.9 06808 3.8 05265 3.8 4.9 17675 05936 9.6 20483 2.1 16266 14530 17459 5.0 12001 3.5 10776 11408 12552 4.5 11713 4.0 14647 4.4 02482 4.8 4.5 12365 4.7 16048 4.4 10798 3.6 10829 3.9 13219 4.7 03548 5.2 20857 3.2 11.4 10842 2.2 12218 17741 5.7 08918 3.7 10988 5.2 22584 13844 02595 10956 4.2 05705 23094 4.5 22399 5.0 4.9 09593 6.0 11538 4.7 10963 4.0 11794 11016 4.6 21039 15470 11.5 07271 4.9 2.8 10844 7.8 09478 3.8 11504 7.7 17843 5.3 5.4 05764 11814 16009 4.4 02552 11744 23186 5.1 02161 08913 5.6 13942 05920 11134 5.1 11471 4.7 11576 13220 6.1 23117 13089 3.7 12286 8.1 00127 6.2 4.2 06850 9.2 09503 5.7 20311 09967 11.7 6.1 12446 4.6 22106 4.8 5.2 12006 03744 6.9 16534 11373 15977 5.8 3.9 10832 12027 8.5 17568 5.7 6.5 13269 11695 12180 4.8 6.0 12411 12.0 13429 6.3 14605 4.7 11039 5.3 06349 12470 4.7 07205 11905 6.0 06962 06961 11.0 16976 6.1 15299 4.6 10938 4.9 00091 8.8 10824 6.3 14175 13727 7.0 11014 7.4 7.5 15218 12.4 11044 6.4 01429 5.4 12134 5.5 14429 10991 4.8 ID_295974 6.9 11651 5.5 09114 11.1 12249 6.1 11505 09805 6.3 05769 4.7 14038 5.0 12496 9.0 05828 6.8 ID_28427 7.8 13923 7.8 10862 12.6 13399 6.7 11662 6.3 13284 5.6 10883 6.3 19357 11051 04169 5.6 14113 11.8 08943 6.6 00087 7.1 14308 4.8 10975 5.3 12152 9.2 11307 28526 6.9 11350 8.2 12124 11172 8.2 11971 8.8 12.8 11230 6.8 09234 7.0 21090 11614 10987 7.6 14677 17803 12.1 12304 7.1 09436 09340 7.2 06790 5.5 11446 5.4 11238 5.7 9.4 05651 11883 7.1 11032 01441 03658 06263 9.5 11619 9.6 13.4 12014 7.8 15574 7.3 13607 06370 9.4 8.3 12859 11679 13621 13740 16039 6.3 03057 6.0 11882 05749 7.6 13902 03317 9.6 13348 13011 10.4 12.8 14246 13.8 10943 8.2 18617 7.9 ID_203792 5.8 11932 05734 9.7 8.7 12344 7.2 6.4 11361 6.2 10056 11331 8.0 19475 9.6 11235 9.7 11084 11.1 03647 13.0 16922 06592 14.6 14869 8.3 12955 8.4 07780 5.9 05834 10828 9.8 9.4 12847 05824 13778 27198 8.2 11894 9.7 18934 9.8 08952 11.7 11779 13.6 12528 7.4 03635 00567 11305 8.6 02034 6.5 6.3 6.1 12239 9.9 05145 10.3 20229 14.9 02495 13537 8.3 12481 18385 9.9 11684 15144 14193 11.9 12262 7.6 03637 12052 27854 9.6 13409 6.7 6.6 11802 11968 10.2 11584 10.5 13098 10.0 13.9 11543 12828 6.4 13938 9.3 03545 23557 10.7 11806 14971 12.1 15789 15.0 09929 9.8 16243 7.3 7.8 12749 12.4 19432 11.9 14795 14.0 18177 7.8 11103 12636 11904 6.7 12472 9.4 10877 10.8 11356 11.0 13953 12.2 12144 15.3 14053 10.0 12425 9.2 9.5 11317 12.8 04565 08947 12.0 18338 14398 14.2 07394 8.0 13798 15385 11138 6.9 11558 09348 9.6 11701 12.1 21581 11.6 24372 12.3 13536 15.4 15429 10.1 17745 9.7 10.2 09649 26475 13.3 12886 12.1 20548 14.5 12030 8.2 13748 7.0 13179 9.7 11029 11076 12.9 23589 12.8 12598 15.5 12800 10.2 12699 11.1 14758 13.5 11625 12.7 12.2 12396 12287 03868 14.6 12244 06580 13461 08930 14069 13.0 15496 13.1 11300 14.6 12047 20319 15.6 06256 00036 7.1 15206 11892 07768 13.6 06286 8.6 11.6 14219 13545 13.6 13.6 9.8 12.7 19087 13872 13.4 13195 14.8 00725 07657 15.7 12073 09475 03836 12096 13.9 10902 19780 14.0 7.2 14941 13.0 09156 12084 12670 9.4 09136 14107 15.8 10935 09216 14.1 19824 9.9 11561 14.2 12364 14.3 14844 13.5 15.1 7.3 09040 14.1 13839 13182 9.8 26966 15.9 13217 13099 14.3 11688 10.1 12200 14.6 02361 14.2 15.2 20476 00469 7.5 16477 15386 10285 16.0 12676 16.0 11224 10.2 11049 14.8 15938 14.7 15.3 12408 10.2 ID_322506 7.6 11499 12504 16.1 13061 16.1 06130 11636 10.4 21162 12756 12282 16.0 02541 06331 10.6 03320 7.8 10896 15.2 ID_208640 16.1 16.2 06872 17.3 12727 10.6 11921 13478 11623 17.4 06333 11.1 15098 12024 7.9 14549 16.4 21585 06292 10.8 04117 17.4 11612 12.6 21657 8.0 11288 17.6 16.2 12914 16.5 09066 10.9 14106 8.6 26257 17.0 13169 17.5 11781 12863 11.0 06037 9.7 13879 09043 18.9 10800 11.4 10836 10.6 11303 11.6 18302 14075 21.4 11.3 11629 11.9 09019 11165 22.9 13522 11.9 12898 14804 13.2 14566 12.0 11050 12.7 11414 13.0 14718 11844 25.2 20499 13.2 17367 16.2 14802 22.2 09415 Figure 5

L. goodei LG 1A 1B 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 L. parva LG 1 24 7 2 27 3 13 4 22 5 28 6 14 7 7 1 8 11 9 7 10 14 11 13 12 25 13 17 14 14 15 22 16 18 17 7 18 22 19 15 20 10 21 2 3 22 13 23 12 medaka # 3 11 9 5 8 16 4 20 17 15 12 21 7 6 24 14 10 13 19 23 18 2 22 1 Figure 6