Usefulness of Adapted Exotic Maize Lines Developed By Doubled Haploid and Single Seed Descent Methods

Adapted exotic maize ( Zea mays L.) germplasm, such as BS39, provides a unique opportunity for 43 broadening the genetic base of U.S. Corn Belt germplasm. In vivo doubled haploid (DH) technology has been 44 used to efficiently exploit exotic germplasm. It can help to purge deleterious recessive alleles. The objectives of this study were to determine the usefulness of BS39-derived inbred lines using both SSD and DH methods, to 46 determine the impact of spontaneous as compared to artificial haploid genome doubling on genetic variance 47 among BS39-derived DH lines, and to identify SNP markers associated with agronomic traits among BS39 48 inbreds monitored at testcross level. We developed two sets of inbred lines directly from BS39 by DH and SSD 49 methods, named BS39_DH and BS39_SSD. Additionally, two sets were derived from a cross between BS39 50 and A427 (SHGD donor) by DH and SSD methods, named BS39×A427_DH and BS39×A427_SSD, 51 respectively. Grain yield, moisture, plant height, ear height, stalk lodging, and root lodging were measured to 52 estimate genetic parameters. For genome-wide association (GWAS) analysis, inbred lines were genotyped using 53 Genotype-by-Sequencing (GBS) and Diversity Array Technology Sequencing (DArTSeq). Some BS39-derived 54 inbred lines performed better than elite germplasm inbreds and all sets showed significant genetic variance. The 55 presence of spontaneous haploid genome doubling genes did not affect performance of inbred lines. Five SNPs were significant and three of them located within genes related to plant development or abiotic stresses. These 57 results demonstrate the potential of BS39 to add novel alleles to temperate elite germplasm. (ii) determine as compared variance among BS39-derived DH lines, and (iii) to identify SNP markers associated with agronomic traits among BS39 inbreds monitored at testcross level potentially beneficial for 2 nd cycling breeding of improved BS39 germplasm.


60
Genetic improvement in temperate maize breeding programs has been achieved using only 2 to 5% of 61 the available races of maize. Using tropical germplasm is an effective way to broadening the genetic base of 62 maize breeding programs (Goodman 2004). However, assessing the genetic merit of tropical germplasm in 63 temperate environments by phenotypic evaluation is challenging. The maladaptive syndrome when growing 64 non-adapted (exotic) materials masks the value of potentially favorable alleles (Hallauer and Carena 2014),

89
Genetic materials. DH and SSD breeding methods were used to derive inbred lines directly from BS39 and 90 from the cross between BS39 and A427, respectively. BS39 is tropical germplasm adapted to temperate 91 environments, and composed of five exotic accessions of Tusón germplasm, representing South American 92 regions. It is a good source to expand the genetic base of temperate maize germplasm, potentially contributing 93 useful and unique alleles (Hallauer and Carena 2016). A427 is a non-stiff stalk (NSS) public line developed at 94 the University of Minnesota and was used as a source of spontaneous haploid genome doubling alleles (Ren et

117
For spontaneous haploid genome doubling, 648 plants from the BS39 population were crossed with 5 A427 at ISU-AEA. At physiological maturity, 359 F1 ears were harvested, dried, and individually shelled. A 119 balanced bulk was prepared with a total of 750 F1 seeds. The 750 F1 plants were cross-pollinated with the 120 maternal haploid inducer BHI201. At physiological maturity, 700 ears were harvested, dried, and individually 121 shelled. Putative haploid kernels were classified as described before. Approximately 10 haploid kernels were 122 selected from each cob to have a balanced bulk from the cross between BS39 and A427. In total, 7,128 haploid 123 kernels were selected and directly sown in the field at ISU-AEA in summer 2016. Putative DH plants that 124 spontaneously shed pollen were self-pollinated. In total, seed was harvested from 598 DH plants, which were 125 shelled individually, and the seed number was recorded. In summer 2017, ~20 seeds of each DH line were 126 increased at ISU-AEA. DH lines displaying substantial phenotypic variation were discontinued. For 127 homogeneous DH lines all plants were self-pollinated, and the seed bulked. DH lines derived from the cross 128 between BS39 and A427 were named BS39×A427_DH lines ( Figure 1).

129
Single-seed descent line development. In winter 2015, 648 plants from BS39 were self-pollinated to produce 130 S1 seeds at Tuniche Seed Services in Graneros, Chile. At physiological maturity, 600 ears were harvested and 131 shelled individually, then shipped to Iowa State University (ISU). Two kernels from each cob were taken to 132 generate a balanced bulk of 1200 S2 seeds. In summer 2016, a balanced bulk of the S2 seeds was planted at ISU-133 AEA. At physiological maturity, 700 self-pollinated ears were harvested and shelled individually. One kernel 134 from each cob was taken to generate an S3 balanced bulk, which was sent to Chile in the winter of 2016. At 135 physiological maturity, 300 ears were harvested and shelled individually, then sent to ISU. In Summer 2017 at 136 ISU-AEA, 10 kernels from each of the 300 S4 ears were planted in an individual row. The first plant of each row 137 was self-pollinated and 120 rows were selected at physiological maturity. Individual ears were harvested and 138 shelled. The same procedure was applied in winter 2017 in Chile with 120 S5 ears. In summer 2018, seeds from 139 96 S6 ears were planted at ISU-AEA for phenotypic evaluation. Inbred lines derived from this process were 140 named BS39_SSD lines (Figure 1).

141
In order to obtain the fourth set of inbred lines 750 F1 plants from the cross of BS39 and A427 were 142 self-pollinated at Tuniche Seed Services during winter 2015. After Winter 2015, the same procedure as 143 described for BS39_SSD was followed, including utilizing the same number of plants and selection intensity. In 144 summer 2018, seeds of the best 96 F6 ears were planted at ISU-AEA for phenotypic evaluation considering the 145 same agronomic traits as BS39_SSD. Inbred lines derived from this process were named BS39×A427_SSD 146 lines ( Figure 1). BS39×A427_DH, BS39_SSD, BS39×A427_SSD), were crossed with the Ex-PVP inbred line LH195, derived 149 from the stiff-stalk synthetic (SSS) heterotic group. In winter 2017, the 384 inbred lines were planted in an 150 isolation plot at Tuniche Seed Services. Additionally, the ex-PVP inbred lines PHG29, PHG83, PHP76, PHN82, 151 PHZ51, and the public line A427 were testcrossed as performance checks. Inbred lines from each set and the 152 checks were sown in a 1-row plot of 25 kernels, to be used as female parents, and the LH195 tester was used as 153 a male parent. Female rows were hand-emasculated, and wind pollinated by the tester to produce testcross seeds 154 for the field trials.

Statistical Analyses.
To compare the mean of the four derivation processes a hierarchical analysis of variance 180 (ANOVA) was carried out for each location according to the following model: Where μ is the overall mean, b i is the effect of block i, S j is the effect of set j, (G/S) k is the effect of genotype k 183 within set j, and ε ijk is the effect of residuals. Coefficients of variation were estimated for each location as Where μ is the overall mean, L i is the fixed effect of location i, (b/L) ij is the block within location nested effect 191 (random), S k is the fixed effect of set k, G k is the random effect of genotype k, GL ik is the interaction between 192 the genotype k and location i, and ε ijk is the random effect of residuals.

193
The estimation of variance components for each set was carried out according to the model for four 194 locations using the same lmerTest package: Where μ is the overall mean, L i is the fixed effect of location i, (b/L) ij is the block within location nested effect 197 (random), G k is the random effect of genotype k, GL ik is the interaction between the genotype k and location i, 198 and ε ijk is the random effect of residuals. Random effects were tested by a Likelihood Ratio Test (LRT) at 5% 199 probability. Trait heritability (H 2 ) was estimated as follows for the four locations: where σ G 2 is the genotypic variance, σ GL 2 is the genotype by environment variance, and σ E 2 is the residual

249
Crawfordsville was the only location where sets did not differ (model 1) (

285
The predicted gain from selection (ΔG) was higher in BS39_DH for all the selection intensity values 286 followed by BS39xA427_DH, BS39_SSD, and BS39xA427_SSD (Table 4). Usefulness (U) criteria were also 287 higher for BS39_DH under 5%, 10% or 40% selection intensity. The lowest U value was observed in 288 BS39xA427_SSD (9.79) at 40% selection intensity. The breeding potential of BS39 germplasm is reflected by 289 the distribution of testcross yields across sets and locations (Figure 2

368
All four sets of inbred lines showed significant genotypic variance. The ratio between variance 369 components (σ g 2 / σ ge 2 ) was, in most cases, higher than 1, which means the genotypic component is more