Genomic predictive ability for foliar nutritive traits in perennial ryegrass

Abstract

11Forage nutritive value impacts animal nutrition, which underpins livestock productivity, reproduction 12 and health. Genetic improvement for nutritive traits has been limited, as they are typically expensive 13 and time-consuming to measure through conventional methods. Genomic selection is appropriate for 14 such complex and expensive traits, enabling cost-effective prediction of breeding values using genome-15 wide markers. The aims of the present study were to assess the potential of genomic selection for a 16 range of nutritive traits in a multi-population training set, and to quantify contributions of genotypic, Genomic selection for nutritive traits 2 39 Genomic selection for nutritive traits 4 program. From each population, 102 to 117 plants that tested positive for endophyte infection (Epichloё 126 festucae var lolli) by immunoblotting (HAHN et al. 2003), were polycrossed in isolation during spring 127 2012 in Palmerston North, New Zealand (FAVILLE et al. 2018). Polycrosses were performed separately 128 for each population, without admixing, and seeds from the maternal parents were harvested and 129 cleaned. In total 543 half-sib families were harvested for seed, however only 517 families had sufficient 130 seed (≥ 3.6g) for sowing field trials. 131 A total of six trials were sown (FAVILLE et al. 2018), of which two were used for the current study. 132 These were trials established at Lincoln (Canterbury region, southern New Zealand, 43.38°S 172.62°E; 133 Wakanui silt loam) and Aorangi (Manawatu region, central New Zealand, 40.34°S 175.46°E; Kairanga 134 sandy loam), during the autumn of 2013. The experimental design at each site was row-column with 135 three replicates. Within each replicate, populations were blocked, and families randomized within 136 blocks. Multiple repeated checks (clonal replicates) were also randomly allocated within and across 137 the replicated blocks. Half-sib families were evaluated as a 1m row of plants (referred to from now as 138 plots), by sowing 0.2 g of seed (which is equivalent to 14 kg ha -1 , if a sward was sown at 7 rows m -1 ). 139 Nitrogen and phosphate fertilizer was applied at the rate of 15-30 kg N ha -1 and 8.8 kg P ha -1 , in late 140 autumn each year (FAVILLE et al. 2018). 141 2.2 Phenotypic measurements 142

Multi-Trait Genomic Prediction Improves Predictive Ability for Dry Matter Yield and Water-Soluble Carbohydrates in Perennial Ryegrass

“…Genomic heritability and variance components for nutritive traits phenotyped in TP2 were previously reported by Arojju et al (2020). Genomic heritability for the seven nutritive traits included in the current study were low to high (0.21 to 0.48) and the family variance was significant for all the seven traits ( Supplementary Table 4).…”

“…TP2 is the same training population described previously by Faville et al (2018) and Arojju et al (2020). This was used in the current study to investigate MT genomic prediction for nutritive traits.…”

“…From the original set of 18 nutritive quality traits originally assessed in the trial (Arojju et al, 2020) a subset of seven traits ( Supplementary Table 4) measured from half-sib families were selected for the current study to test the potential of MT genomic prediction approaches. As described in detail by Arojju et al (2020) half-sib families were harvested on 14 th April 2014 at Lincoln and 29 th April 2014 at Aorangi. To reduce the influence of diurnal variation on the measured components, harvests were undertaken at each site over three consecutive days, between 10:30 AM and 3:00 PM each day.…”

“…Because of similar experimental design, TP2 was analyzed in a manner similar to TP1, but without family-by-measure and measure components in the model. The linear mixed model used to analyze TP2 data was described in detail by Arojju et al (2020).…”

“…While trait heritability and the extent of LD cannot be easily manipulated, previous studies in perennial ryegrass have developed new knowledge on the effect of relatedness between training and selection population, marker density, and training population size on predictive ability (Fè et al, 2015;Fè et al, 2016;Grinberg et al, 2016;Byrne et al, 2017;Arojju et al, 2018;Faville et al, 2018;Arojju et al, 2020). In perennial ryegrass, genomic prediction models have been developed and validated for many quantitative traits, including DMY (Arojju, 2017;Faville et al, 2018;Guo et al, 2018;Pembleton et al, 2018), nutritive quality traits (Grinberg et al, 2016;Arojju et al, 2020), heading date (Fè et al, 2015;Byrne et al, 2017;Faville et al, 2018;Guo et al, 2018) and crown rust resistance (Fè et al, 2016;Arojju et al, 2018;Guo et al, 2018). In addition, different statistical methods for prediction, with different assumptions regarding the trait inheritance pattern, have been assessed for genomic prediction in this species but little difference in predictive ability was observed (Grinberg et al, 2016;Byrne et al, 2017;Faville et al, 2018).…”