Abstract The performance of a single cross is determined by the average additive effects of the parents, as well as the interactions between them. These quantities can be estimated using an appropriate genetic design, allowing for the estimation of general (GCA) and specific (SCA) combining abilities. The prediction of GCA for new parents and the total genetic value of unrealized crosses can be made when genome-wide marker information is available. Several studies in crops such as maize and rice have demonstrated the potential of genomic-assisted prediction of single-cross performance in economically important crops. However, no study to date has explored its relevance in perennial ryegrass, an obligate allogamous species that is bred in genetically heterogeneous families. In this study, we aimed to estimate genetic parameters and assess the ability of genomic models to predict the performance of F2 families in terms of dry matter yield and nutritive quality traits. We used data from a large partial diallel involving 104 parents from two distinct subpopulations, as inferred by admixture analysis. F2 families were evaluated in multiple environments and under two nitrogen availability conditions. Genotyping-by-sequencing of the parent plants produced 42,145 variants after quality control, which were used to estimate genomic relationships based on identity-by-state. Variance component estimation revealed limited GCA and SCA interactions with the environment, and particularly with nitrogen management. The predictive abilities of two parental models exceeded 0.60 and often surpassed 0.70 for most traits. However, incorporating non-additive effects into the model did not improve predictive ability. We leveraged the genetic diversity among parents to map genomic regions associated with all recorded traits. Genome-wide association studies (GWAS) by genomic best linear unbiased prediction (GBLUP) identified six quantitative trait loci (QTL) regions, with 45 candidate genes within the linkage disequilibrium range, estimated at approximately 92 kb. Our results demonstrate that genomic prediction of single crosses can be performed with high accuracy, especially when both parents are also progenitors of families in the training set.

Early root traits, particularly total root length, are heritable and show positive genetic correlations with biomass yield in perennial ryegrass; incorporating them into breeding programs can enhance genetic gain. Perennial ryegrass (Lolium perenne L.) is an important forage grass widely used in pastures and lawns, valued for its high nutritive value and environmental benefits. Despite its importance, genetic improvements in biomass yield have been slow, mainly due to its outbreeding nature and the challenges of improving multiple traits simultaneously. This study aims to assess the potential advantages of including early root traits in the perennial ryegrass breeding process. Root traits, including total root length (TRL) and root angle (RA) were phenotyped in a greenhouse using rhizoboxes, and genetic correlations with field yield were estimated across three European locations over two years. Bivariate models estimated significant genetic correlations of 0.40 (SE = 0.14) between TRL and field yield, and a weak but positive correlation to RA of 0.15 (SE = 0.14). Heritability estimates were 0.36 for TRL, 0.39 for RA, and 0.31 for field yield across locations. Incorporating root trait data into selection criteria can improve the efficiency of breeding programs, potentially increasing genetic gain by approximately 10

Background Understanding the mechanisms underlining forage production and its biomass nutritive quality at the omics level is crucial for boosting the output of high-quality dry matter per unit of land. Despite the advent of multiple omics integration for the study of biological systems in major crops, investigations on forage species are still scarce. Results Our results identified substantial changes in gene co-expression and metabolite-metabolite network topologies as a result of genetic perturbation by hybridizing L. perenne with another species within the genus ( L. multiflorum ) relative to across genera ( F. pratensis ). However, conserved hub genes and hub metabolomic features were detected between pedigree classes, some of which were highly heritable and displayed one or more significant edges with agronomic traits in a weighted omics-phenotype network. In spite of tagging relevant biological molecules as, for example, the light-induced rice 1 ( LIR1 ), hub features were not necessarily better explanatory variables for omics-assisted prediction than features stochastically sampled and all available regressors. Conclusions The utilization of computational techniques for the reconstruction of co-expression networks facilitates the identification of key omic features that serve as central nodes and demonstrate correlation with the manifestation of observed traits. Our results also indicate a robust association between early multi-omic traits measured in a greenhouse setting and phenotypic traits evaluated under field conditions.

Abstract The growing demand for food and feed crops in the world because of growing population and more extreme weather events requires high‐yielding and resilient crops. Many agriculturally important traits are polygenic, controlled by multiple regulatory layers, and with a strong interaction with the environment. In this study, 120 F2 families of perennial ryegrass (Lolium perenne L.) were grown across a water gradient in a semifield facility with subsoil irrigation. Genomic (single‐nucleotide polymorphism [SNP]), transcriptomic (gene expression [GE]), and DNA methylomic (MET) data were integrated with feed quality trait data collected from control and drought sections in the semifield facility, providing a treatment effect. Deep root length (DRL) below 110 cm was assessed with convolutional neural network image analysis. Bayesian prediction models were used to partition phenotypic variance into its components and evaluated the proportion of phenotypic variance in all traits captured by different regulatory layers (SNP, GE, and MET). The spatial effects and effects of SNP, GE, MET, the interaction between GE and MET (GE × MET) and GE × treatment (GEControl and GEDrought) interaction were investigated. Gene expression explained a substantial part of the genetic and spatial variance for all the investigated phenotypes, whereas MET explained residual variance not accounted for by SNPs or GE. For DRL, MET also contributed to explaining spatial variance. The study provides a statistically elegant analytical paradigm that integrates genomic, transcriptomic, and MET information to understand the regulatory mechanisms of polygenic effects for complex traits.

Abstract Joint modeling of correlated multienvironment and multiharvest data of perennial crop species may offer advantages in prediction schemes and a better understanding of the underlying dynamics in space and time. The goal of the present study was to investigate the relevance of incorporating the longitudinal dimension of within‐season multiple measurements of forage perennial ryegrass (Lolium perenne L.) traits in a reaction‐norm model setup that additionally accounts for genotype × environment (G × E) interactions. Genetic parameters and accuracy of genomic estimated breeding value (gEBV) predictions were investigated by fitting three genomic random regression models (gRRMs) using Legendre polynomial functions to the data. Genomic DNA sequencing of family pools of diploid perennial ryegrass was performed using DNA nanoball‐based technology and yielded 56,645 single‐nucleotide polymorphisms, which were used to calculate the allele frequency‐based genomic relationship matrix. Biomass yield's estimated additive genetic variance and heritability values were higher in later harvests. The additive genetic correlations were moderate to low in early measurements and peaked at intermediates with fairly stable values across the environmental gradient except for the initial harvest data collection. This led to the conclusion that complex (G × E) arises from spatial and temporal dimensions in the early season with lower reranking trends thereafter. In general, modeling the temporal dimension with a second‐order orthogonal polynomial improved the accuracy of gEBV prediction for nutritive quality traits, but no gain in prediction accuracy was detected for dry matter yield (DMY). This study leverages the flexibility and usefulness of gRRM models for perennial ryegrass breeding and can be readily extended to other multiharvest crops.

Ryegrass is one of the most important forage crops worldwide. It is the basis for 80% of milk production and 70% of meat production and has major economic importance. Breeding programmes for ryegrass started in the 1920s, and breeders have mainly relied on repeated phenotypic and recently genotypic selection of elite individuals. Although this approach has led to significant improvements in several characters including rust resistance, spring growth and aftermath heading, it tends to be laborious, expensive and time-consuming, mainly due to gametophyte self-incompatibility in most ryegrass species (Sampoux et al., 2011). In order to overcome some of the limitations of traditional introgression and selective breeding, modern methods of mutation induction offer attractive opportunities to target specific genes of interest and directly introduce allelic variability. In the last decade, the Clustered Regularly Interspaced Short Palindromic Repeats/CRISPR-associated endonuclease 9 (CRISPR/Cas9) system has been extensively used in most crops and is paving the way to precision trait improvements in factors including yield, quality, biotic- and abiotic stress resistance and breeding rate (Chen et al., 2019; Ran et al., 2017; Wang et al., 2014). However, this powerful tool for genome editing has not yet been used in ryegrass. Meiosis arose early during the evolution of eukaryotes and is vital for sexual reproduction, not only in relation to genomic stability but also to genetic diversity. Meiotic studies of plants in the areas of crop fertility and genetic variation have important potential agronomical applications. DMC1 (DISRUPTED MEIOTIC cDNA1), initially identified in yeast (Bishop et al., 1992) as a homolog of the bacterial strand exchange protein RecA, is a crucial meiotic recombinase. Here, we describe the use of the CRISPR/Cas9 system to introduce mutations in LpDMC1 in two species: Italian ryegrass (Lolium perenne ssp. multiflorum) and perennial ryegrass (Lolium perenne). We succeeded in obtaining both T0 homozygous and heterozygous mutants, and the T0 null mutants of Italian ryegrass exhibited complete male sterility and severely disordered meiosis with univalents and multivalents appearing at diakinesis. To see whether mutations could be introduced into ryegrass using the CRISPR/Cas9 system, we generated a sgRNA (TS-LpDMC1) targeting exon 5 of LpDMC1, with an XceI restriction enzyme site near the protospacer-adjacent motif (PAM) for ease of analysis (Figure 1a). Because plant tissue culture and genetic transformation are time-consuming, we tested the activities of sgRNA in a protoplast transient expression system as described by Shan et al. (2013). TS-LpDMC1 promoted by TaU6 was co-introduced with SpCas9 into ryegrass protoplasts by PEG-mediated transformation. After 40- to 48-h incubation, analysis of genomic DNA using a PCR restriction enzyme digestion assay (PCR/RE) demonstrated the occurrence of indel mutations at the target site (Data not shown). To determine whether the CRISPR/Cas9 method was applicable to other ryegrass genes, we targeted the ryegrass orthologue of centromere-specific histone H3 variant (CENH3). In Arabidopsis thaliana, CENH3 ensures faithful transmission of the genome at cell division, and when cenh3 null mutants producing altered CENH3 proteins are crossed with wild type, many haploid Arabidopsis plants are generated (Ravi and Chan, 2010). When we co-transformed a sgRNA targeting exon 3 of LpCENH3 along with SpCas9 into protoplasts (Figure 1d), PCR/RE analysis revealed frameshift mutations at the target site (Figure 1e). These results show that CRISPR/Cas9 can be used to generate mutations in ryegrass. Next, the sgRNA expression cassette was combined with SpCas9 in a single DNA construct by GIBSON Assembly and introduced along with a hygromycin-resistant plasmid into preconditioned embryogenic callus (EC) lines of ryegrass by gold particles bombardment. To generate these EC lines, seeds of Italian ryegrass cultivar Gepetto and perennial ryegrass cultivar Goyave were de-husked and sterilized, and somatic EC lines were established as described (Ran et al., 2014). Three separate lines designated Gepetto-8, Gepetto-66 and Goyave LMG LDF-Lp3711 (provided by DLF Seeds) with outstanding regeneration ability were selected for transformation. After bombardment, the EC was transferred to hygromycin medium. Surviving calli were obtained after 4 weeks' induction and sub-culture. Thereafter, they were regenerated for 8 weeks; and plantlets with established roots were transferred to potting mix for mutants' identification. The mean transformation efficiencies of these three lines were 3.83%, 4.50% and 2.66%, respectively. The entire experimental cycle took approximately 10 months from target design to mutant identification. Compared with conventional methods, CRISPR/Cas9 provides a rapid and straightforward method for genetic manipulation of ryegrass. T0 generation LpDMC1 knockout mutants were identified by PCR/RE using the same primers as in the protoplast assays (Figure 1b), followed by Sanger sequencing. In total, we obtained eight mutants in two species: seven of Italian ryegrass and one of perennial ryegrass. G8-12 was a homozygous mutant with a 9 bp in-frame deletion in one allele. The other non-wild-type plant G8-16 was a mosaic, with three mutations: −1 bp, −33 bp deletions and −10/+37 bp deletion/insertion. The other six were heterozygotes (Figure 1c). The genome editing efficiencies of LpDMC1 in Gepetto-8, Gepetto-66 and Goyave LMG LDF-Lp3711 were 11.63%, 11.11% and 5.88%, respectively. To detect off-target events, putative off-target sites of TS-LpDMC1 were predicted using the draft genome sequence of the forage grass Lolium perenne (Byrne et al., 2015) (Figure 1f). Two non-mismatch off-target sites, LpOff3 and LpOff4, with highly similar surrounding sequences, which could produce mutations in two different non-coding regions, were selected for further study. PCR products that included LpOff3/LpOff4 from the eight mutants were sequenced, but no mutations were detected (Figure 1g). The next-closest match was LpOff5, which had an imperfect PAM, suggesting that off-target events would be unlikely (Figure 1f). Because Italian and perennial ryegrass are both gametophyte self-incompatibility and have different vernalization requirements, it was difficult and time-consuming to obtain homozygous T1 mutants. Since disruption of DMC1 in Arabidopsis, rice and barley usually leads to reduced fertility, we examined the pollen grains produced by wild type and the T0 Lpdmc1 mutant G8-16. Upon iodine-potassium iodide (I2-KI) staining, the pollen grains from the wild type appeared mostly round, while those from G8-16 were empty and shrunken (Figure 1h,i). This phenotype points to complete male sterility, as in the Osdmc1a Osdmc1b Tos-17 double insertion mutant of rice (Wang et al., 2016). The wild type pollen grains did not stain as strongly as those of rice, probably due to the difference in starch content between Italian ryegrass and rice. To clarify what happened during Lpdmc1 meiosis, we stained meiotic chromosomes with 4',6-diamidino-2-phenylindole (DAPI). The meiotic chromosomes of the Lpdmc1 mutant behaved normally at leptotene (Figure 1j,m). However at diplotene, whereas circular or figure-of-eight shapes of the chromatids were observed in wild type (Figure 1k), in the Lpdmc1 mutant the chromatids were stuck together, mostly due to the existence of multivalents (Figure 1n). At diakinesis, unlike the wild type, which formed 7 bivalents (Figure 1l), condensed univalents and multivalents were scattered throughout the mutant nuclei (Figure 1o). These results point to a meiotic defect in the Lpdmc1 mutant. In summary, we were able to introduce mutations in two genes, LpDMC1 and LpCENH3, of ryegrass in vivo using the CRISPR/Cas9 system. We obtained eight T0 knockout mutants of LpDMC1 in Italian and perennial ryegrass. The Lpdmc1 null mutants were completely male sterile with severely disrupted meiosis, indicating that DMC1, a highly conserved protein, plays a pivotal role in meiosis in many species. We are now using genome editing tools to improve other agronomic traits in ryegrass. We anticipate that this method will come to be used routinely in this economically important forage crop. We thank Professor Zhukuan Cheng's group for helping with meiotic chromosome preparation. This work was supported by grants from the National Transgenic Science and Technology Program of China (2019ZX08010-001, 2019ZX08010-003 and 2018ZX0801002B), the Scientific Program of Beijing Municipal Commission of Science and Technology (Z171100001517001) and Innovation Fund Denmark (6150-00020B). The authors declare no conflict of interest. Y.Z., J.-L.Q., C.S.J. and C.G. designed the experiments; I.N. and T.A. provided genomics and bioinformatics support; Y.R. performed the genetic transformation of ryegrasses; I.L. and C.S.J. provided the ryegrass seeds; Y.Z. performed the molecular and cytologic experiments; C.G supervised the project; and Y.Z., Y.R. and C.G. wrote the manuscript.

Many efforts have been made to select and isolate naturally occurring animal-friendly Epichloë strains for later reinfection into elite cultivars. Often this process involves large-scale screening of Epichloë-infected wild grass populations where strains are characterized and alkaloids measured. Here, we describe for the first time the use of genotyping-by-sequencing (GBS) on a collection of 217 Epichloë-infected grasses (7 S. arundinaceum, 4 L. perenne, and 206 S. pratensis). This genotyping strategy is cheaper than complete genome sequencing, is suitable for a large number of individuals, and, when applied to endophyte-infected grasses, conveniently genotypes both organisms. In total, 6273 single nucleotide polymorphisms (SNPs) in the endophyte data set and 38 323 SNPs in the host data set were obtained. Our findings reveal a composite structure with three distinct endophyte clusters unrelated to the three main S. pratensis gene pools that have most likely spread from different glacial refugia in Eurasia. All three gene pools can establish symbiosis with E. uncinata. A comparison of the endophyte clusters with microsatellite-based fingerprinting of the same samples allows a quick test to discriminate between these clusters using two simple sequence repeats (SSRs). Concentrations of loline alkaloids and mycelial biomass are correlated and differ significantly among the plant and endophyte subpopulations; one endophyte strain has higher levels of lolines than others, and one specific host genotype is particularly suitable to host E. uncinata. These findings pave the way for targeted artificial inoculations of specific host-endophyte combinations to boost loline production in the symbiota and for genome association studies with the aim of isolating genes involved in the compatibility between meadow fescue and E. uncinata.

Brachypodium distachyon is an established model for drought tolerance. We previously identified accessions exhibiting high tolerance, susceptibility and intermediate tolerance to drought; respectively, ABR8, KOZ1 and ABR4. Transcriptomics and metabolomic approaches were used to define tolerance mechanisms. Transcriptional analyses suggested relatively few drought responsive genes in ABR8 compared to KOZ1. Linking these to gene ontology (GO) terms indicated enrichment for "regulated stress response", "plant cell wall" and "oxidative stress" associated genes. Further, tolerance correlated with pre-existing differences in cell wall-associated gene expression including glycoside hydrolases, pectin methylesterases, expansins and a pectin acetylesterase. Metabolomic assessments of the same samples also indicated few significant changes in ABR8 with drought. Instead, pre-existing differences in the cell wall-associated metabolites correlated with drought tolerance. Although other features, e.g., jasmonate signaling were suggested in our study, cell wall-focused events appeared to be predominant. Our data suggests two different modes through which the cell wall could confer drought tolerance: (i) An active response mode linked to stress induced changes in cell wall features, and (ii) an intrinsic mode where innate differences in cell wall composition and architecture are important. Both modes seem to contribute to ABR8 drought tolerance. Identification of the exact mechanisms through which the cell wall confers drought tolerance will be important in order to inform development of drought tolerant crops.

Ryegrass single plants, bi-parental family pools, and multi-parental family pools are often genotyped, based on allele-frequencies using genotyping-by-sequencing (GBS) assays. GBS assays can be performed at low-coverage depth to reduce costs. However, reducing the coverage depth leads to a higher proportion of missing data, and leads to a reduction in accuracy when identifying the allele-frequency at each locus. As a consequence of the latter, genomic relationship matrices (GRMs) will be biased. This bias in GRMs affects variance estimates and the accuracy of GBLUP for genomic prediction (GBLUP-GP). We derived equations that describe the bias from low-coverage sequencing as an effect of binomial sampling of sequence reads, and allowed for any ploidy level of the sample considered. This allowed us to combine individual and pool genotypes in one GRM, treating pool-genotypes as a polyploid genotype, equal to the total ploidy-level of the parents of the pool. Using simulated data, we verified the magnitude of the GRM bias at different coverage depths for three different kinds of ryegrass breeding material: individual genotypes from single plants, pool-genotypes from F2 families, and pool-genotypes from synthetic varieties. To better handle missing data, we also tested imputation procedures, which are suited for analyzing allele-frequency genomic data. The relative advantages of the bias-correction and the imputation of missing data were evaluated using real data. We examined a large dataset, including single plants, F2 families, and synthetic varieties genotyped in three GBS assays, each with a different coverage depth, and evaluated them for heading date, crown rust resistance, and seed yield. Cross validations were used to test the accuracy using GBLUP approaches, demonstrating the feasibility of predicting among different breeding material. Bias-corrected GRMs proved to increase predictive accuracies when compared with standard approaches to construct GRMs. Among the imputation methods we tested, the random forest method yielded the highest predictive accuracy. The combinations of these two methods resulted in a meaningful increase of predictive ability (up to 0.09). The possibility of predicting across individuals and pools provides new opportunities for improving ryegrass breeding schemes.

Perennial ryegrass is an outbreeding forage species and is one of the most widely used forage grasses in temperate regions. The aim of this study was to investigate the possibility of implementing genomic prediction in tetraploid perennial ryegrass, to study the effects of different sequencing depth when using genotyping-by-sequencing (GBS), and to determine optimal number of single-nucleotide polymorphism (SNP) markers and sequencing depth for GBS data when applied in tetraploids. A total of 1,515 F-2 tetraploid ryegrass families were included in the study and phenotypes and genotypes were scored on family-pools. The traits considered were dry matter yield (DM), rust resistance (RUST), and heading date (HD). The genomic information was obtained in the form of allele frequencies of pooled family samples using GBS. Different SNP filtering strategies were designed. The strategies included filtering out SNPs having low average depth (FILTLOW), having high average depth (FILTHIGH), and having both low average and high average depth (FILTBOTH). In addition, SNPs were kept randomly with different data sizes (RAN). The accuracy of genomic prediction was evaluated by using a "leave single F-2 family out" cross validation scheme, and the predictive ability and bias were assessed by correlating phenotypes corrected for fixed effects with predicted additive breeding values. Among all the filtering scenarios, the highest estimates for genomic heritability of family means were 0.45, 0.74, and 0.73 for DM, HD and RUST, respectively. The predictive ability generally increased as the number of SNPs included in the analysis increased. The highest predictive ability for DM was 0.34 (137,191 SNPs having average depth higher than 10), for HD was 0.77 (185,297 SNPs having average depth lower than 60), and for RUST was 0.55 (188,832 SNPs having average depth higher than 1). Genomic prediction can help to optimize the breeding of tetraploid ryegrass. GBS data including about 80-100 K SNPs are needed for accurate prediction of additive breeding values in tetraploid ryegrass. Using only SNPs with sequencing depth between 10 and 20 gave highest predictive ability, and showed the potential to obtain accurate prediction from medium-low coverage GBS in tetraploids.

Silicon (Si) is taken up from the soil as monosilicic acid by plant roots, transported to leaves and deposited as phytoliths, amorphous silica (SiO2) bodies, which are a key component of anti-herbivore defense in grasses. Silicon transporters have been identified in many plant species, but the mechanisms underpinning Si transport remain poorly understood. Specifically, the extent to which Si uptake is a passive process, driven primarily by transpiration, or has both passive and active components remains disputed. Increases in foliar Si concentration following herbivory suggest plants may exercise some control over Si uptake and distribution. In order to investigate passive and active controls on Si accumulation, we examined both genetic and environmental influences on Si accumulation in the forage grass Festuca arundinacea. We studied three F. arundinacea varieties that differ in the levels of Si they accumulate. Varieties not only differed in Si concentration, but also in increases in Si accumulation in response to leaf damage. The varietal differences in Si concentration generally reflected differences in stomatal density and stomatal conductance, suggesting passive, transpirationmediated mechanisms underpin these differences. Bagging plants after damage was employed to minimize differences in stomatal conductance between varieties and in response to damage. This treatment eliminated constitutive differences in leaf Si levels, but did not impair the damage-induced increases in Si uptake: damaged, bagged plants still had more leaf Si than undamaged, bagged plants in all three varieties. Preliminary differential gene expression analysis revealed that the active Si transporter Lsi2 was highly expressed in damaged unbagged plants compared with undamaged unbagged plants, suggesting damage-induced Si defenses are regulated at gene level. Our findings suggest that although differences in transpiration may be partially responsible for varietal differences in Si uptake, they cannot explain damage-induced increases in Si uptake and deposition, suggesting that wounding causes changes in Si uptake, distribution and deposition that likely involve active processes and changes in gene expression.

By using the genotyping-by-sequencing method, it is feasible to characterize genomic relationships directly at the level of family pools and to estimate genomic heritabilities from phenotypes scored on family-pools in outbreeding species.

The implementation of genomic selection (GS) in plant breeding, so far, has been mainly evaluated in crops farmed as homogeneous varieties, and the results have been generally positive. Fewer results are available for species, such as forage grasses, that are grown as heterogenous families (developed from multiparent crosses) in which the control of the genetic variation is far more complex. Here we test the potential for implementing GS in the breeding of perennial ryegrass ( L.) using empirical data from a commercial forage breeding program. Biparental F and multiparental synthetic (SYN) families of diploid perennial ryegrass were genotyped using genotyping-by-sequencing, and phenotypes for five different traits were analyzed. Genotypes were expressed as family allele frequencies, and phenotypes were recorded as family means. Different models for genomic prediction were compared by using practically relevant cross-validation strategies. All traits showed a highly significant level of genetic variance, which could be traced using the genotyping assay. While there was significant genotype × environment (G × E) interaction for some traits, accuracies were high among F families and between biparental F and multiparental SYN families. We have demonstrated that the implementation of GS in grass breeding is now possible and presents an opportunity to make significant gains for various traits.

Background Genomic selection (GS) has become a commonly used technology in animal breeding. In crops, it is expected to significantly improve the genetic gains per unit of time. So far, its implementation in plant breeding has been mainly investigated in species farmed as homogeneous varieties. Concerning crops farmed in family pools, only a few theoretical studies are currently available. Here, we test the opportunity to implement GS in breeding of perennial ryegrass, using real data from a forage breeding program. Heading date was chosen as a model trait, due to its high heritability and ease of assessment. Genome Wide Association analysis was performed to uncover the genetic architecture of the trait. Then, Genomic Prediction (GP) models were tested and prediction accuracy was compared to the one obtained in traditional Marker Assisted Selection (MAS) methods. Results Several markers were significantly associated with heading date, some locating within or proximal to genes with a well-established role in floral regulation. GP models gave very high accuracies, which were significantly better than those obtained through traditional MAS. Accuracies were higher when predictions were made from related families and from larger training populations, whereas predicting from unrelated families caused the variance of the estimated breeding values to be biased downwards. Conclusions We have demonstrated that there are good perspectives for GS implementation in perennial ryegrass breeding, and that problems resulting from low linkage disequilibrium (LD) can be reduced by the presence of structure and related families in the breeding population. While comprehensive Genome Wide Association analysis is difficult in species with extremely low LD, we did identify variants proximal to genes with a known role in flowering time (e.g. CONSTANS and Phytochrome C).

A variety of selection systems have been developed for transformation of forage crops. To compare the most frequently used systems, we tested three selectable marker genes for their selection efficiency under four selection procedures for the production of transgenic tall fescue. Embryogenic calluses initiated from mature embryos were bombarded with three constructs containing either the phosphinothricin acetyltransferase ( bar ) gene, the hygromycin phosphotransferase ( hpt ) gene or the neomycin phosphotransferase II ( nptII ) gene. Transformation efficiency was strongly influenced by the selectable marker gene, selection procedure and genotype. The highest transformation efficiency was observed using the bar gene in combination with bialaphos. Average transformation efficiencies with bialaphos, phosphinothricin (glufosinate), hygromycin and paromomycin selection across the two callus lines used in the experiments were 9.4%, 4.4%, 5.2% and 1.6%, respectively. Southern blot analysis revealed the independent nature of the tested transgenic plants and a complex transgene integration pattern with multiple insertions.