Implementation of Genomic Prediction in Lolium perenne (L.) Breeding Populations

Genomic selection shows improved expected genetic gain over phenotypic selection of agronomic traits in allotetraploid white clover

KEY MESSAGE

Genomic selection using white clover multi-year-multi-site data showed predicted genetic gains through integrating among-half-sibling-family phenotypic selection and within-family genomic selection were up to 89% greater than half-sibling-family phenotypic selection alone. Genomic selection, an effective breeding tool used widely in plants and animals for improving low-heritability traits, has only recently been applied to forages. We explored the feasibility of implementing genomic selection in white clover (Trifolium repens L.), a key forage legume which has shown limited genetic improvement in dry matter yield (DMY) and persistence traits. We used data from a training population comprising 200 half-sibling (HS) families evaluated in a cattle-grazed field trial across three years and two locations. Combining phenotype and genotyping-by-sequencing (GBS) data, we assessed different two-stage genomic prediction models, including KGD-GBLUP developed for low-depth GBS data, on DMY, growth score, leaf size and stolon traits. Predictive abilities were similar among the models, ranging from -0.17 to 0.44 across traits, and remained stable for most traits when reducing model input to 100-120 HS families and 5500 markers, suggesting genomic selection is viable with fewer resources. Incorporating a correlated trait with a primary trait in multi-trait prediction models increased predictive ability by 28-124%. Deterministic modelling showed integrating among-HS-family phenotypic selection and within-family genomic selection at different selection pressures estimated up to 89% DMY genetic gain compared to phenotypic selection alone, despite a modest predictive ability of 0.3. This study demonstrates the potential benefits of combining genomic and phenotypic selection to boost genetic gains in white clover. Using cost-effective GBS paired with a prediction model optimized for low read-depth data, the approach can achieve prediction accuracies comparable to traditional models, providing a viable path for implementing genomic selection in white clover.

Comparison of machine learning methods for genomic prediction of selected Arabidopsis thaliana traits

We present a comparison of machine learning methods for the prediction of four quantitative traits in Arabidopsis thaliana. High prediction accuracies were achieved on individuals grown under standardized laboratory conditions from the 1001 Arabidopsis Genomes Project. An existing body of evidence suggests that linear models may be impeded by their inability to make use of non-additive effects to explain phenotypic variation at the population level. The results presented here use a nested cross-validation approach to confirm that some machine learning methods have the ability to statistically outperform linear prediction models, with the optimal model dependent on availability of training data and genetic architecture of the trait in question. Linear models were competitive in their performance as per previous work, though the neural network class of predictors was observed to be the most accurate and robust for traits with high heritability. The extent to which non-linear models exploit interaction effects will require further investigation of the causal pathways that lay behind their predictions. Future work utilizing more traits and larger sample sizes, combined with an improved understanding of their respective genetic architectures, may lead to improvements in prediction accuracy.

Including marker x environment interactions improves genomic prediction in red clover (Trifolium pratense L.)

Genomic prediction has mostly been used in single environment contexts, largely ignoring genotype x environment interaction, which greatly affects the performance of plants. However, in the last decade, prediction models including marker x environment (MxE) interaction have been developed. We evaluated the potential of genomic prediction in red clover (Trifolium pratense L.) using field trial data from five European locations, obtained in the Horizon 2020 EUCLEG project. Three models were compared: (1) single environment (SingleEnv), (2) across environment (AcrossEnv), (3) marker x environment interaction (MxE). Annual dry matter yield (DMY) gave the highest predictive ability (PA). Joint analyses of DMY from years 1 and 2 from each location varied from 0.87 in Britain and Switzerland in year 1, to 0.40 in Serbia in year 2. Overall, crude protein (CP) was predicted poorly. PAs for date of flowering (DOF), however ranged from 0.87 to 0.67 for Britain and Switzerland, respectively. Across the three traits, the MxE model performed best and the AcrossEnv worst, demonstrating that including marker x environment effects can improve genomic prediction in red clover. Leaving out accessions from specific regions or from specific breeders' material in the cross validation tended to reduce PA, but the magnitude of reduction depended on trait, region and breeders' material, indicating that population structure contributed to the high PAs observed for DMY and DOF. Testing the genomic estimated breeding values on new phenotypic data from Sweden showed that DMY training data from Britain gave high PAs in both years (0.43-0.76), while DMY training data from Switzerland gave high PAs only for year 1 (0.70-0.87). The genomic predictions we report here underline the potential benefits of incorporating MxE interaction in multi-environment trials and could have perspectives for identifying markers with effects that are stable across environments, and markers with environment-specific effects.

Genomic prediction of resistance to Hymenoscyphus fraxineus in common ash (Fraxinus excelsior L.) populations

The increase in introduced insect pests and pathogens due to anthropogenic environmental changes has become a major concern for tree species worldwide. Common ash (Fraxinus excelsior L.) is one of such species facing a significant threat from the invasive fungal pathogen Hymenoscyphus fraxineus. Some studies have indicated that the susceptibility of ash to the pathogen is genetically determined, providing some hope for accelerated breeding programs that are aimed at increasing the resistance of ash populations. To address this challenge, we used a genomic selection strategy to identify potential genetic markers that are associated with resistance to the pathogen causing ash dieback. Through genome-wide association studies (GWAS) of 300 common ash individuals from 30 populations across Poland (ddRAD, dataset A), we identified six significant SNP loci with a p-value ≤1 × 10-4 associated with health status. To further evaluate the effectiveness of GWAS markers in predicting health status, we considered two genomic prediction scenarios. Firstly, we conducted cross-validation on dataset A. Secondly, we trained markers on dataset A and tested them on dataset B, which involved whole-genome sequencing of 20 individuals from two populations. Genomic prediction analysis revealed that the top SNPs identified via GWAS exhibited notably higher prediction accuracies compared to randomly selected SNPs, particularly with a larger number of SNPs. Cross-validation analyses using dataset A showcased high genomic prediction accuracy, predicting tree health status with over 90% accuracy across the top SNP sets ranging from 500 to 10,000 SNPs from the GWAS datasets. However, no significant results emerged for health status when the model trained on dataset A was tested on dataset B. Our findings illuminate potential genetic markers associated with resistance to ash dieback, offering support for future breeding programs in Poland aimed at combating ash dieback and bolstering conservation efforts for this invaluable tree species.

Genomic selection in plant breeding: Key factors shaping two decades of progress

Genomic selection, the application of genomic prediction (GP) models to select candidate individuals, has significantly advanced in the past two decades, effectively accelerating genetic gains in plant breeding. This article provides a holistic overview of key factors that have influenced GP in plant breeding during this period. We delved into the pivotal roles of training population size and genetic diversity, and their relationship with the breeding population, in determining GP accuracy. Special emphasis was placed on optimizing training population size. We explored its benefits and the associated diminishing returns beyond an optimum size. This was done while considering the balance between resource allocation and maximizing prediction accuracy through current optimization algorithms. The density and distribution of single-nucleotide polymorphisms, level of linkage disequilibrium, genetic complexity, trait heritability, statistical machine-learning methods, and non-additive effects are the other vital factors. Using wheat, maize, and potato as examples, we summarize the effect of these factors on the accuracy of GP for various traits. The search for high accuracy in GP-theoretically reaching one when using the Pearson's correlation as a metric-is an active research area as yet far from optimal for various traits. We hypothesize that with ultra-high sizes of genotypic and phenotypic datasets, effective training population optimization methods and support from other omics approaches (transcriptomics, metabolomics and proteomics) coupled with deep-learning algorithms could overcome the boundaries of current limitations to achieve the highest possible prediction accuracy, making genomic selection an effective tool in plant breeding.

Genome-wide family prediction unveils molecular mechanisms underlying the regulation of agronomic traits in Urochloa ruziziensis

Tropical forage grasses, particularly those belonging to the Urochloa genus, play a crucial role in cattle production and serve as the main food source for animals in tropical and subtropical regions. The majority of these species are apomictic and tetraploid, highlighting the significance of U. ruziziensis, a sexual diploid species that can be tetraploidized for use in interspecific crosses with apomictic species. As a means to support breeding programs, our study investigates the feasibility of genome-wide family prediction in U. ruziziensis families to predict agronomic traits. Fifty half-sibling families were assessed for green matter yield, dry matter yield, regrowth capacity, leaf dry matter, and stem dry matter across different clippings established in contrasting seasons with varying available water capacity. Genotyping was performed using a genotyping-by-sequencing approach based on DNA samples from family pools. In addition to conventional genomic prediction methods, machine learning and feature selection algorithms were employed to reduce the necessary number of markers for prediction and enhance predictive accuracy across phenotypes. To explore the regulation of agronomic traits, our study evaluated the significance of selected markers for prediction using a tree-based approach, potentially linking these regions to quantitative trait loci (QTLs). In a multiomic approach, genes from the species transcriptome were mapped and correlated to those markers. A gene coexpression network was modeled with gene expression estimates from a diverse set of U. ruziziensis genotypes, enabling a comprehensive investigation of molecular mechanisms associated with these regions. The heritabilities of the evaluated traits ranged from 0.44 to 0.92. A total of 28,106 filtered SNPs were used to predict phenotypic measurements, achieving a mean predictive ability of 0.762. By employing feature selection techniques, we could reduce the dimensionality of SNP datasets, revealing potential genotype-phenotype associations. The functional annotation of genes near these markers revealed associations with auxin transport and biosynthesis of lignin, flavonol, and folic acid. Further exploration with the gene coexpression network uncovered associations with DNA metabolism, stress response, and circadian rhythm. These genes and regions represent important targets for expanding our understanding of the metabolic regulation of agronomic traits and offer valuable insights applicable to species breeding. Our work represents an innovative contribution to molecular breeding techniques for tropical forages, presenting a viable marker-assisted breeding approach and identifying target regions for future molecular studies on these agronomic traits.

Developing an integrated genomic selection approach beyond biomass for varietal protection and nutritive traits in perennial ryegrass (Lolium perenne L.)

Breeding target traits can be broadened to include nutritive value and plant breeder's rights traits in perennial ryegrass by using in-field regression-based spectroscopy phenotyping and genomic selection. Perennial ryegrass breeding has focused on biomass yield, but expansion into a broader set of traits is needed to benefit livestock industries whilst also providing support for intellectual property protection of cultivars. Numerous breeding objectives can be targeted simultaneously with the development of sensor-based phenomics and genomic selection (GS). Of particular interest are nutritive value (NV), which has been difficult and expensive to measure using traditional phenotyping methods, resulting in limited genetic improvement to date, and traits required to obtain varietal protection, known as plant breeder's rights (PBR) traits. In order to assess phenotyping requirements for NV improvement and potential for genetic improvement, in-field reflectance-based spectroscopy was assessed and GS evaluated in a single population for three key NV traits, captured across four timepoints. Using three prediction approaches, the possibility of targeting PBR traits using GS was evaluated for five traits recorded across three years of a breeding program. Prediction accuracy was generally low to moderate for NV traits and moderate to high for PBR traits, with heritability highly correlated with GS accuracy. NV did not show significant or consistent correlation between timepoints highlighting the need to incorporate seasonal NV into selection indexes and the value of being able to regularly monitor NV across seasons. This study has demonstrated the ability to implement GS for both NV and PBR traits in perennial ryegrass, facilitating the expansion of ryegrass breeding targets to agronomically relevant traits while ensuring necessary varietal protection is achieved.

Agrobacterium- and a single Cas9-sgRNA transcript system-mediated high efficiency gene editing in perennial ryegrass

Genome editing technologies provide a powerful tool for genetic improvement of perennial ryegrass, an important forage and turfgrass species worldwide. The sole publication for gene editing in perennial ryegrass used gene-gun for plant transformation and a dual promoter based CRISPR/Cas9 system for editing. However, their editing efficiency was low (5.9% or only one gene-edited plant produced). To test the suitability of the maize Ubiquitin 1 (ZmUbi1) promoter in gene editing of perennial ryegrass, we produced ZmUbi1 promoter:RUBY transgenic plants. We observed that ZmUbi1 promoter was active in callus tissue prior to shoot regeneration, suggesting that the promoter is suitable for Cas9 and sgRNA expression in perennial ryegrass for high-efficiency production of bi-allelic mutant plants. We then used the ZmUbi1 promoter for controlling Cas9 and sgRNA expression in perennial ryegrass. A ribozyme cleavage target site between the Cas9 and sgRNA sequences allowed production of functional Cas9 mRNA and sgRNA after transcription. Using Agrobacterium for genetic transformation, we observed a 29% efficiency for editing the PHYTOENE DESATURASE gene in perennial ryegrass. DNA sequencing analyses revealed that most pds plants contained bi-allelic mutations. These results demonstrate that the expression of a single Cas9 and sgRNA transcript unit controlled by the ZmUbi1 promoter provides a highly efficient system for production of bi-allelic mutants of perennial ryegrass and should also be applicable in other related grass species.

Sparse bayesian learning for genomic selection in yeast

Genomic selection, which predicts phenotypes such as yield and drought resistance in crops from high-density markers positioned throughout the genome of the varieties, is moving towards machine learning techniques to make predictions on complex traits that are controlled by several genes. In this paper, we consider sparse Bayesian learning and ensemble learning as a technique for genomic selection and ranking markers based on their relevance to a trait. We define and explore two different forms of the sparse Bayesian learning for predicting phenotypes and identifying the most influential markers of a trait, respectively. We apply our methods on a Saccharomyces cerevisiae dataset, and analyse our results with respect to existing related works, trait heritability, as well as the accuracies obtained from linear and Gaussian kernel functions. We find that sparse Bayesian methods are not only competitive with other machine learning methods in predicting yeast growth in different environments, but are also capable of identifying the most important markers, including both positive and negative effects on the growth, from which biologists can get insight. This attribute can make our proposed ensemble of sparse Bayesian learners favourable in ranking markers based on their relevance to a trait.

Genome-Enabled Prediction Methods Based on Machine Learning

Growth of artificial intelligence and machine learning (ML) methodology has been explosive in recent years. In this class of procedures, computers get knowledge from sets of experiences and provide forecasts or classification. In genome-wide based prediction (GWP), many ML studies have been carried out. This chapter provides a description of main semiparametric and nonparametric algorithms used in GWP in animals and plants. Thirty-four ML comparative studies conducted in the last decade were used to develop a meta-analysis through a Thurstonian model, to evaluate algorithms with the best predictive qualities. It was found that some kernel, Bayesian, and ensemble methods displayed greater robustness and predictive ability. However, the type of study and data distribution must be considered in order to choose the most appropriate model for a given problem.

Buffer Green Patches around Urban Road Network as a Tool for Sustainable Soil Management

Urban areas are facing a range of environmental challenges including air, water and soil pollution as a result of industrial, domestic and traffic emissions. In addition, global climate change is likely to aggravate certain urban problems and disturb the urban ecology by increasing the frequency and severity of extreme weather events. In the context of urbanization growth and the consequent impact on the environment, there is a growing interest in maintaining urban soil quality and functions as they are the medium for green infrastructure development. Furthermore, urban soils are becoming one of the key factors in the delivery of many ecosystem services such as carbon storage, climate regulation, water flow regulation, etc. On the other hand, urban soils are well-known to be a major sink of air pollutants due to the wet and dry atmospheric deposition and recirculation. Soil has the ability to degrade some chemical contaminants but when the levels are high, urban soils could hold on large amounts and pose a risk to human health. A cost-effective technological solution is to use the ability of some plant species to metabolize, accumulate and detoxify heavy metals or other harmful organic or inorganic compounds from the soil layer. The establishment of urban lawns (grass covered surfaces) is a helpful, environmentally friendly, economically sustainable and cost-effective approach to remove contaminants from polluted soils (terrains), which also has some aesthetic benefits. In this paper, an overview of the benefits and limitations of urban lawn construction is presented. The focus is on the perspectives for sustainable management of urban lawns, especially as buffer green patches in the road network surroundings, that can represent strategies to provide ecological and social multifunctionality of urban soils, and thus, increasing their ecosystem services capacity. Specifically, the paper highlights (i) the possibilities for phytoremediation of urban soils, (ii) potential of some perennial grasses and (iii) key issues that should be considered in the planning and design of urban lawns.

Genomic Prediction of Complex Traits in Forage Plants Species: Perennial Grasses Case

The majority of forage grass species are obligate outbreeders. Their breeding classically consists of an initial selection on spaced plants for highly heritable traits such as disease resistances and heading date, followed by familial selection on swards for forage yield and quality traits. The high level of diversity and heterozygosity, and associated decay of linkage disequilibrium (LD) over very short genomic distances, has hampered the implementation of genomic selection (GS) in these species. However, next generation sequencing technologies in combination with the development of genomic resources have recently facilitated implementation of GS in forage grass species such as perennial ryegrass (Lolium perenne L.), switchgrass (Panicum virgatum L.), and timothy (Phleum pratense L.). Experimental work and simulations have shown that GS can increase significantly the genetic gain per unit of time for traits with different levels of heritability. The main reasons are (1) the possibility to select single plants based on their genomic estimated breeding values (GEBV) for traits measured at sward level, (2) a reduction in the duration of selection cycles, and less importantly (3) an increase in the selection intensity associated with an increase in the genetic variance used for selection. Nevertheless, several factors should be taken into account for the successful implementation of GS in forage grasses. For example, it has been shown that the level of relatedness between the training and the selection population is particularly critical when working with highly structured meta-populations consisting of several genetic groups. A sufficient number of markers should be used to estimate properly the kinship between individuals and to reflect the variability of major QTLs. It is also important that the prediction models are trained for relevant environments when dealing with traits with high genotype × environment interaction (G × E). Finally, in these outbreeding species, measures to reduce inbreeding should be used to counterbalance the high selection intensity that can be achieved in GS.

Prediction of heterosis in the recent rapeseed (Brassica napus) polyploid by pairing parental nucleotide sequences

The utilization of heterosis is a successful strategy in increasing yield for many crops. However, it consumes tremendous manpower to test the combining ability of the parents in fields. Here, we applied the genomic-selection (GS) strategy and developed models that significantly increase the predictability of heterosis by introducing the concept of a regional parental genetic-similarity index (PGSI) and reducing dimension in the calculation matrix in a machine-learning approach. Overall, PGSI negatively affected grain yield and several other traits but positively influenced the thousand-seed weight of the hybrids. It was found that the C subgenome of rapeseed had a greater impact on heterosis than the A subgenome. We drew maps with overviews of quantitative-trait loci that were responsible for the heterosis (h-QTLs) of various agronomic traits. Identifications and annotations of genes underlying high impacting h-QTLs were provided. Using models that we elaborated, combining abilities between an Ogu-CMS-pool member and a potential restorer can be simulated in silico, sidestepping laborious work, such as testing crosses in fields. The achievements here provide a case of heterosis prediction in polyploid genomes with relatively large genome sizes.

Oilseed rape (Brassica napus) is of significant economic interest worldwide, providing high-quality oil with excellent health-promoting properties. It represents an excellent model of a successful recent polyploid that rapidly became an important crop worldwide. The utilization of hybridization, leading to hybrid vigor, or heterosis, is a successful strategy in increasing yield and vigor for many field crops including rapeseed (Brassica napus). However, the procedure of using classical breeding methods remains slow and laborious, illustrating the need for predictive and innovative methods. Here, we have achieved a significant breakthrough by using genome selection and significantly advanced models to predict the heterosis by pairing genome-wide nucleotides of parents. We provided maps with overviews of quantitative trait loci that were responsible for the heterosis of various agronomic traits. The research used deep resequencing (>30x) data of the entire polyploidy rapeseed genome, providing a successful case for the prediction of heterosis in polyploid genomes with relatively large genome sizes. Moreover, we provided the genetic information (SNPs) of 1007 core accessions of this species in the public domain for testing combinations with high heterosis using our predicting model for rapeseed breeders all over the world.

Characterization and practical use of self-compatibility in outcrossing grass species

BACKGROUND

Self-incompatibility (SI) systems prevent self-fertilization in several species of Poaceae, many of which are economically important forage, bioenergy and turf grasses. Self-incompatibility ensures cross-pollination and genetic diversity but restricts the ability to fix useful genetic variation. In most inbred crops it is possible to develop high-performing homozygous parental lines by self-pollination, which then enables the creation of F1 hybrid varieties with higher performance, a phenomenon known as heterosis. The inability to fully exploit heterosis in outcrossing grasses is partially responsible for lower levels of improvement in breeding programmes compared with inbred crops. However, SI can be overcome in forage grasses to create self-compatible populations. This is generating interest in understanding the genetical basis of self-compatibility (SC), its significance for reproductive strategies and its exploitation for crop improvement, especially in the context of F1 hybrid breeding.

SCOPE

We review the literature on SI and SC in outcrossing grass species. We review the currently available genomic tools and approaches used to discover and characterize novel SC sources. We discuss opportunities barely explored for outcrossing grasses that SC facilitates. Specifically, we discuss strategies for wide SC introgression in the context of the Lolium-Festuca complex and the use of SC to develop immortalized mapping populations for the dissection of a wide range of agronomically important traits. The germplasm available is a valuable practical resource and will aid understanding the basis of inbreeding depression and hybrid vigour in key temperate forage grass species.

CONCLUSIONS

A better understanding of the genetic control of additional SC loci offers new insight into SI systems, their evolutionary origins and their reproductive significance. Heterozygous outcrossing grass species that can be readily selfed facilitate studies of heterosis. Moreover, SC introduction into a range of grass species will enable heterosis to be exploited in innovative ways in genetic improvement programmes.

Deterministic and stochastic modelling of impacts from genomic selection and phenomics on genetic gain for perennial ryegrass dry matter yield

Increasing the efficiency of current forage breeding programs through adoption of new technologies, such as genomic selection (GS) and phenomics (Ph), is challenging without proof of concept demonstrating cost effective genetic gain (∆G). This paper uses decision support software DeltaGen (tactical tool) and QU-GENE (strategic tool), to model and assess relative efficiency of five breeding methods. The effect on ∆G and cost ($) of integrating GS and Ph into an among half-sib (HS) family phenotypic selection breeding strategy was investigated. Deterministic and stochastic modelling were conducted using mock data sets of 200 and 1000 perennial ryegrass HS families using year-by-season-by-location dry matter (DM) yield data and in silico generated data, respectively. Results demonstrated short (deterministic)- and long-term (stochastic) impacts of breeding strategy and integration of key technologies, GS and Ph, on ∆G. These technologies offer substantial improvements in the rate of ∆G, and in some cases improved cost-efficiency. Applying 1% within HS family GS, predicted a 6.35 and 8.10% ∆G per cycle for DM yield from the 200 HS and 1000 HS, respectively. The application of GS in both among and within HS selection provided a significant boost to total annual ∆G, even at low GS accuracy r_A of 0.12. Despite some reduction in ∆G, using Ph to assess seasonal DM yield clearly demonstrated its impact by reducing cost per percentage ∆G relative to standard DM cuts. Open-source software tools, DeltaGen and QuLinePlus/QU-GENE, offer ways to model the impact of breeding methodology and technology integration under a range of breeding scenarios.

Harnessing Crop Wild Diversity for Climate Change Adaptation

Warming and drought are reducing global crop production with a potential to substantially worsen global malnutrition. As with the green revolution in the last century, plant genetics may offer concrete opportunities to increase yield and crop adaptability. However, the rate at which the threat is happening requires powering new strategies in order to meet the global food demand. In this review, we highlight major recent 'big data' developments from both empirical and theoretical genomics that may speed up the identification, conservation, and breeding of exotic and elite crop varieties with the potential to feed humans. We first emphasize the major bottlenecks to capture and utilize novel sources of variation in abiotic stress (i.e., heat and drought) tolerance. We argue that adaptation of crop wild relatives to dry environments could be informative on how plant phenotypes may react to a drier climate because natural selection has already tested more options than humans ever will. Because isolated pockets of cryptic diversity may still persist in remote semi-arid regions, we encourage new habitat-based population-guided collections for genebanks. We continue discussing how to systematically study abiotic stress tolerance in these crop collections of wild and landraces using geo-referencing and extensive environmental data. By uncovering the genes that underlie the tolerance adaptive trait, natural variation has the potential to be introgressed into elite cultivars. However, unlocking adaptive genetic variation hidden in related wild species and early landraces remains a major challenge for complex traits that, as abiotic stress tolerance, are polygenic (i.e., regulated by many low-effect genes). Therefore, we finish prospecting modern analytical approaches that will serve to overcome this issue. Concretely, genomic prediction, machine learning, and multi-trait gene editing, all offer innovative alternatives to speed up more accurate pre- and breeding efforts toward the increase in crop adaptability and yield, while matching future global food demands in the face of increased heat and drought. In order for these 'big data' approaches to succeed, we advocate for a trans-disciplinary approach with open-source data and long-term funding. The recent developments and perspectives discussed throughout this review ultimately aim to contribute to increased crop adaptability and yield in the face of heat waves and drought events.

A comparison of shared patterns of differential gene expression and gene ontologies in response to water-stress in roots and leaves of four diverse genotypes of Lolium and Festuca spp. temperate pasture grasses

Ryegrasses (Lolium spp.) and fescues (Festuca spp.) are closely related and widely cultivated perennial forage grasses. As such, resilience in the face of abiotic stresses is an important component of their traits. We have compared patterns of differentially expressed genes (DEGs) in roots and leaves of two perennial ryegrass genotypes and a single genotype of each of a festulolium (predominantly Italian ryegrass) and meadow fescue with the onset of water stress, focussing on overall patterns of DEGs and gene ontology terms (GOs) shared by all four genotypes. Plants were established in a growing medium of vermiculite watered with nutrient solution. Leaf and root material were sampled at 35% (saturation) and, as the medium dried, at 15%, 5% and 1% estimated water contents (EWCs) and RNA extracted. Differential gene expression was evaluated comparing the EWC sampling points from RNAseq data using a combination of analysis methods. For all genotypes, the greatest numbers of DEGs were identified in the 35/1 and 5/1 comparisons in both leaves and roots. In total, 566 leaf and 643 root DEGs were common to all 4 genotypes, though a third of these leaf DEGs were not regulated in the same up/down direction in all 4 genotypes. For roots, the equivalent figure was 1% of the DEGs. GO terms shared by all four genotypes were often enriched by both up- and down-regulated DEGs in the leaf, whereas generally, only by either up- or down-regulated DEGs in the root. Overall, up-regulated leaf DEGs tended to be more genotype-specific than down-regulated leaf DEGs or root DEGs and were also associated with fewer GOs. On average, only 5–15% of the DEGs enriching common GO terms were shared by all 4 genotypes, suggesting considerable variation in DEGs between related genotypes in enacting similar biological processes.

Machine learning approaches reveal genomic regions associated with sugarcane brown rust resistance

Sugarcane is an economically important crop, but its genomic complexity has hindered advances in molecular approaches for genetic breeding. New cultivars are released based on the identification of interesting traits, and for sugarcane, brown rust resistance is a desirable characteristic due to the large economic impact of the disease. Although marker-assisted selection for rust resistance has been successful, the genes involved are still unknown, and the associated regions vary among cultivars, thus restricting methodological generalization. We used genotyping by sequencing of full-sib progeny to relate genomic regions with brown rust phenotypes. We established a pipeline to identify reliable SNPs in complex polyploid data, which were used for phenotypic prediction via machine learning. We identified 14,540 SNPs, which led to a mean prediction accuracy of 50% when using different models. We also tested feature selection algorithms to increase predictive accuracy, resulting in a reduced dataset with more explanatory power for rust phenotypes. As a result of this approach, we achieved an accuracy of up to 95% with a dataset of 131 SNPs related to brown rust QTL regions and auxiliary genes. Therefore, our novel strategy has the potential to assist studies of the genomic organization of brown rust resistance in sugarcane.

Natural variation and genomic prediction of growth, physiological traits, and nitrogen-use efficiency in perennial ryegrass under low-nitrogen stress

Genomic prediction of nitrogen-use efficiency (NUE) has not previously been studied in perennial grass species exposed to low-N stress. Here, we conducted a genomic prediction of physiological traits and NUE in 184 global accessions of perennial ryegrass (Lolium perenne) in response to a normal (7.5 mM) and low (0.75 mM) supply of N. After 21 d of treatment under greenhouse conditions, significant variations in plant height increment (ΔHT), leaf fresh weight (LFW), leaf dry weight (LDW), chlorophyll index (Chl), chlorophyll fluorescence, leaf N and carbon (C) contents, C/N ratio, and NUE were observed in accessions , but to a greater extent under low-N stress. Six genomic prediction models were applied to the data, namely the Bayesian method Bayes C, Bayesian LASSO, Bayesian Ridge Regression, Ridge Regression-Best Linear Unbiased Prediction, Reproducing Kernel Hilbert Spaces, and randomForest. These models produced similar prediction accuracy of traits within the normal or low-N treatments, but the accuracy differed between the two treatments. ΔHT, LFW, LDW, and C were predicted slightly better under normal N with a mean Pearson r-value of 0.26, compared with r=0.22 under low N, while the prediction accuracies for Chl, N, C/N, and NUE were significantly improved under low-N stress with a mean r=0.45, compared with r=0.26 under normal N. The population panel contained three population structures, which generally had no effect on prediction accuracy. The moderate prediction accuracies obtained for N, C, and NUE under low-N stress are promising, and suggest a feasible means by which germplasm might be initially assessed for further detailed studies in breeding programs.

Advancing crested wheatgrass [Agropyron cristatum (L.) Gaertn.] breeding through genotyping-by-sequencing and genomic selection

Crested wheatgrass [Agropyron cristatum (L.) Gaertn.] provides high quality, highly palatable forage for early season grazing. Genetic improvement of crested wheatgrass has been challenged by its complex genome, outcrossing nature, long breeding cycle, and lack of informative molecular markers. Genomic selection (GS) has potential for improving traits of perennial forage species, and genotyping-by-sequencing (GBS) has enabled the development of genome-wide markers in non-model polyploid plants. An attempt was made to explore the utility of GBS and GS in crested wheatgrass breeding. Sequencing and phenotyping 325 genotypes representing 10 diverse breeding lines were performed. Bioinformatics analysis identified 827, 3,616, 14,090 and 46,136 single nucleotide polymorphism markers at 20%, 30%, 40% and 50% missing marker levels, respectively. Four GS models (BayesA, BayesB, BayesCπ, and rrBLUP) were examined for the accuracy of predicting nine agro-morphological and three nutritive value traits. Moderate accuracy (0.20 to 0.32) was obtained for the prediction of heading days, leaf width, plant height, clump diameter, tillers per plant and early spring vigor for genotypes evaluated at Saskatoon, Canada. Similar accuracy (0.29 to 0.35) was obtained for predicting fall regrowth and plant height for genotypes evaluated at Swift Current, Canada. The Bayesian models displayed similar or higher accuracy than rrBLUP. These findings show the feasibility of GS application for a non-model species to advance plant breeding.

Effects of Different Strategies for Exploiting Genomic Selection in Perennial Ryegrass Breeding Programs

Genomic selection (GS) is a potential pathway to accelerate genetic gain for perennial ryegrass (Lolium perenne L.). The main objectives of the present study were to investigate the level of genetic gain and accuracy by applying GS in commercial perennial ryegrass breeding programs. Different scenarios were compared to a conventional breeding program. Simulated scenarios differed in the method of selection and structure of the breeding program. Two scenarios (Phen-Y12 and Phen) for phenotypic selection and three scenarios (GS-Y12, GS and GS-SP) were considered for genomic breeding schemes. All breeding schemes were simulated for 25 cycles. The amount of genetic gain achieved was different across scenarios. Compared to phenotypic scenarios, GS scenarios resulted in substantially larger genetic gain for the simulated traits. This was mainly due to more efficient selection of plots and single plants based on genomic estimated breeding values. Also, GS allows for reduction in waiting time for the availability of the superior genetic materials from previous cycles, which led to at least a doubling or a trebling of genetic gain compared to the traditional program. Reduction in additive genetic variance levels were higher with GS scenarios than with phenotypic selection. The results demonstrated that implementation of GS in ryegrass breeding is possible and presents an opportunity to make very significant improvements in genetic gains.

High-Throughput Genome-Wide Genotyping To Optimize the Use of Natural Genetic Resources in the Grassland Species Perennial Ryegrass (Lolium perenne L.)

The natural genetic diversity of agricultural species is an essential genetic resource for breeding programs aiming to improve their ecosystem and production services. A large natural ecotype diversity is usually available for most grassland species. This could be used to recombine natural climatic adaptations and agronomic value to create improved populations of grassland species adapted to future regional climates. However describing natural genetic resources can be long and costly. Molecular markers may provide useful information to help this task. This opportunity was investigated for Lolium perenne L., using a set of 385 accessions from the natural diversity of this species collected right across Europe and provided by genebanks of several countries. For each of these populations, genotyping provided the allele frequencies of 189,781 SNP markers. GWAS were implemented for over 30 agronomic and/or putatively adaptive traits recorded in three climatically contrasted locations (France, Belgium, Germany). Significant associations were detected for hundreds of markers despite a strong confounding effect of the genetic background; most of them pertained to phenology traits. It is likely that genetic variability in these traits has had an important contribution to environmental adaptation and ecotype differentiation. Genomic prediction models calibrated using natural diversity were found to be highly effective to describe natural populations for almost all traits as well as commercial synthetic populations for some important traits such as disease resistance, spring growth or phenological traits. These results will certainly be valuable information to help the use of natural genetic resources of other species.

Predicting rice phenotypes with meta and multi-target learning

The features in some machine learning datasets can naturally be divided into groups. This is the case with genomic data, where features can be grouped by chromosome. In many applications it is common for these groupings to be ignored, as interactions may exist between features belonging to different groups. However, including a group that does not influence a response introduces noise when fitting a model, leading to suboptimal predictive accuracy. Here we present two general frameworks for the generation and combination of meta-features when feature groupings are present. Furthermore, we make comparisons to multi-target learning, given that one is typically interested in predicting multiple phenotypes. We evaluated the frameworks and multi-target learning approaches on a genomic rice dataset where the regression task is to predict plant phenotype. Our results demonstrate that there are use cases for both the meta and multi-target approaches, given that overall, they significantly outperform the base case.

Population structure and genetic diversity in red clover (Trifolium pratense L.) germplasm

Red clover (Trifolium pratense L.) is a highly adaptable forage crop for temperate livestock agriculture. Genetic variation can be identified, via molecular techniques, and used to assess diversity among populations that may otherwise be indistinguishable. Here we have used genotyping by sequencing (GBS) to determine the genetic variation and population structure in red clover natural populations from Europe and Asia, and varieties or synthetic populations. Cluster analysis differentiated the collection into four large regional groups: Asia, Iberia, UK, and Central Europe. The five varieties clustered with the geographical area from which they were derived. Two methods (BayeScan and Samβada) were used to search for outlier loci indicating signatures of selection. A total of 60 loci were identified by both methods, but no specific genomic region was highlighted. The rate of decay in linkage disequilibrium was fast, and no significant evidence of any bottlenecks was found. Phenotypic analysis showed that a more prostrate and spreading growth habit was predominantly found among populations from Iberia and the UK. A genome wide association study identified a single nucleotide polymorphism (SNP) located in a homologue of the VEG2 gene from pea, associated with flowering time. The identification of genetic variation within the natural populations is likely to be useful for enhancing the breeding of red clover in the future.

Genomic Predictive Ability for Foliar Nutritive Traits in Perennial Ryegrass

Forage nutritive value impacts animal nutrition, which underpins livestock productivity, reproduction and health. Genetic improvement for nutritive traits in perennial ryegrass has been limited, as they are typically expensive and time-consuming to measure through conventional methods. Genomic selection is appropriate for such complex and expensive traits, enabling cost-effective prediction of breeding values using genome-wide markers. The aims of the present study were to assess the potential of genomic selection for a range of nutritive traits in a multi-population training set, and to quantify contributions of family, location and family-by-location variance components to trait variation and heritability for nutritive traits. The training set consisted of a total of 517 half-sibling (half-sib) families, from five advanced breeding populations, evaluated in two distinct New Zealand grazing environments. Autumn-harvested samples were analyzed for 18 nutritive traits and maternal parents of the half-sib families were genotyped using genotyping-by-sequencing. Significant (P < 0.05) family variance was detected for all nutritive traits and genomic heritability (h2g ) was moderate to high (0.20 to 0.74). Family-by-location interactions were significant and particularly large for water soluble carbohydrate (WSC), crude fat, phosphorus (P) and crude protein. GBLUP, KGD-GBLUP and BayesCπ genomic prediction models displayed similar predictive ability, estimated by 10-fold cross validation, for all nutritive traits with values ranging from r = 0.16 to 0.45 using phenotypes from across two locations. High predictive ability was observed for the mineral traits sulfur (0.44), sodium (0.45) and magnesium (0.45) and the lowest values were observed for P (0.16), digestibility (0.22) and high molecular weight WSC (0.23). Predictive ability estimates for most nutritive traits were retained when marker number was reduced from one million to as few as 50,000. The moderate to high predictive abilities observed suggests implementation of genomic selection is feasible for most of the nutritive traits examined.

Advancing Bromegrass Breeding Through Imaging Phenotyping and Genomic Selection: A Review

Breeding forage crops for high yields of digestible biomass along with improved resource-use efficiency and wide adaptation is essential to meet future challenges in forage production imposed by growing demand, declining resources, and changing climate. Bromegrasses (Bromus spp.) are economically important forage species in the temperate regions of world, but genetic gain in forage yield of bromegrass is relatively low. In particular, limited breeding efforts have been made in improving abiotic stress tolerance and resource-use efficiency. We conducted a literature review on bromegrass breeding achievements and challenges, global climate change impacts on bromegrass species, and explored the feasibility of applying high-throughput imaging phenotyping techniques and genomic selection for further advances in forage yield and quality selection. Overall genetic gain in forage yield of bromegrass has been low, but genetic improvement in forage yield of smooth bromegrass (Bromus inermis Leyss) is somewhat higher than that of meadow bromegrass (Bromus riparius Rehm). This low genetic gain in bromegrass yield is due to a few factors such as its genetic complexity, lack of long-term breeding effort, and inadequate plant adaptation to changing climate. Studies examining the impacts of global climate change on bromegrass species show that global warming, heat stress, and drought have negative effects on forage yield. A number of useful physiological traits have been identified for genetic improvement to minimize yield loss. Available reports suggest that high-throughput imaging phenotyping techniques, including visual and infrared thermal imaging, imaging hyperspectral spectroscopy, and imaging chlorophyll fluorescence, are capable of capturing images of morphological, physiological, and biochemical traits related to plant growth, yield, and adaptation to abiotic stresses at different scales of organization. The more complex traits such as photosynthetic radiation-use efficiency, water-use efficiency, and nitrogen-use efficiency can be effectively assessed by utilizing combinations of imaging hyperspectral spectroscopy, infrared thermal imaging, and imaging chlorophyll fluorescence techniques in a breeding program. Genomic selection has been applied in the breeding of forage species and the applications show its potential in high ploidy, outcrossing species like bromegrass to improve the accuracy of parental selection and improve genetic gain. Together, these new technologies hold promise for improved genetic gain and wide adaptation in future bromegrass breeding.

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

In perennial ryegrass (Lolium perenne L), annual and seasonal dry matter yield (DMY) and nutritive quality of herbage are high-priority traits targeted for improvement through selective breeding. Genomic prediction (GP) has proven to be a valuable tool for improving complex traits and may be further enhanced through the use of multi-trait (MT) prediction models. In this study, we evaluated the relative performance of MT prediction models to improve predictive ability for DMY and key nutritive quality traits, using two different training populations (TP1, n = 463 and TP2, n = 517) phenotyped at multiple locations. MT models outperformed single-trait (ST) models by 24% to 59% for DMY and 67% to 105% for nutritive quality traits, such as low, high, and total WSC, when a correlated secondary trait was included in both the training and test set (MT-CV2) or in the test set alone (MT-CV3) (trait-assisted genomic selection). However, when a secondary trait was included in training set and not the test set (MT-CV1), the predictive ability was not statistically significant (p > 0.05) compared to the ST model. We evaluated the impact of training set size when using a MT-CV2 model. Using a highly correlated trait (r_g = 0.88) as the secondary trait in the MT-CV2 model, there was no loss in predictive ability for DMY even when the training set was reduced to 50% of its original size. In contrast, using a weakly correlated secondary trait (r_g = 0.56) in the MT-CV2 model, predictive ability began to decline when the training set size was reduced by only 11% from its original size. Using a ST model, genomic predictive ability in a population unrelated to the training set was poor (r_p = -0.06). However, when using an MT-CV2 model, the predictive ability was positive and high (r_p = 0.76) for the same population. Our results demonstrate the first assessment of MT models in forage species and illustrate the prospects of using MT genomic selection in forages, and other outcrossing plant species, to accelerate genetic gains for complex agronomical traits, such as DMY and nutritive quality characteristics.

Genomic basis of European ash tree resistance to ash dieback fungus

Populations of European ash trees (Fraxinus excelsior) are being devastated by the invasive alien fungus Hymenoscyphus fraxineus, which causes ash dieback. We sequenced whole genomic DNA from 1,250 ash trees in 31 DNA pools, each pool containing trees with the same ash dieback damage status in a screening trial and from the same seed-source zone. A genome-wide association study identified 3,149 single nucleotide polymorphisms (SNPs) associated with low versus high ash dieback damage. Sixty-one of the 192 most significant SNPs were in, or close to, genes with putative homologues already known to be involved in pathogen responses in other plant species. We also used the pooled sequence data to train a genomic prediction model, cross-validated using individual whole genome sequence data generated for 75 healthy and 75 damaged trees from a single seed source. The model's genomic estimated breeding values (GEBVs) allocated these 150 trees to their observed health statuses with 67% accuracy using 10,000 SNPs. Using the top 20% of GEBVs from just 200 SNPs, we could predict observed tree health with over 90% accuracy. We infer that ash dieback resistance in F. excelsior is a polygenic trait that should respond well to both natural selection and breeding, which could be accelerated using genomic prediction.

Integrating a newly developed BAC-based physical mapping resource for Lolium perenne with a genome-wide association study across a L. perenne European ecotype collection identifies genomic contexts associated with agriculturally important traits

BACKGROUND AND AIMS

Lolium perenne (perennial ryegrass) is the most widely cultivated forage and amenity grass species in temperate areas worldwide and there is a need to understand the genetic architectures of key agricultural traits and crop characteristics that deliver wider environmental services. Our aim was to identify genomic regions associated with agriculturally important traits by integrating a bacterial artificial chromosome (BAC)-based physical map with a genome-wide association study (GWAS).

METHODS

BAC-based physical maps for L. perenne were constructed from ~212 000 high-information-content fingerprints using Fingerprint Contig and Linear Topology Contig software. BAC clones were associated with both BAC-end sequences and a partial minimum tiling path sequence. A panel of 716 L. perenne diploid genotypes from 90 European accessions was assessed in the field over 2 years, and genotyped using a Lolium Infinium SNP array. The GWAS was carried out using a linear mixed model implemented in TASSEL, and extended genomic regions associated with significant markers were identified through integration with the physical map.

KEY RESULTS

Between ~3600 and 7500 physical map contigs were derived, depending on the software and probability thresholds used, and integrated with ~35 k sequenced BAC clones to develop a resource predicted to span the majority of the L. perenne genome. From the GWAS, eight different loci were significantly associated with heading date, plant width, plant biomass and water-soluble carbohydrate accumulation, seven of which could be associated with physical map contigs. This allowed the identification of a number of candidate genes.

CONCLUSIONS

Combining the physical mapping resource with the GWAS has allowed us to extend the search for candidate genes across larger regions of the L. perenne genome and identified a number of interesting gene model annotations. These physical maps will aid in validating future sequence-based assemblies of the L. perenne genome.

Use of identifiability analysis in designing phenotyping experiments for modelling forage production and quality

Agricultural systems models are complex and tend to be over-parameterized with respect to observational datasets. Practical identifiability analysis based on local sensitivity analysis has proved effective in investigating identifiable parameter sets in environmental models, but has not been applied to agricultural systems models. Here, we demonstrate that identifiability analysis improves experimental design to ensure independent parameter estimation for yield and quality outputs of a complex grassland model. The Pasture Simulation model (PaSim) was used to demonstrate the effectiveness of practical identifiability analysis in designing experiments and measurement protocols within phenotyping experiments with perennial ryegrass. Virtual experiments were designed combining three factors: frequency of measurements, duration of the experiment. and location of trials. Our results demonstrate that (i) PaSim provides sufficient detail in terms of simulating biomass yield and quality of perennial ryegrass for use in breeding, (ii) typical breeding trials are insufficient to parameterize all influential parameters, (iii) the frequency of measurements is more important than the number of growing seasons to improve the identifiability of PaSim parameters, and (iv) identifiability analysis provides a sound approach for optimizing the design of multi-location trials. Practical identifiability analysis can play an important role in ensuring proper exploitation of phenotypic data and cost-effective multi-location experimental designs. Considering the growing importance of simulation models, this study supports the design of experiments and measurement protocols in the phenotyping networks that have recently been organized.

Boosting Genetic Gain in Allogamous Crops via Speed Breeding and Genomic Selection

Breeding schemes that utilize modern breeding methods like genomic selection (GS) and speed breeding (SB) have the potential to accelerate genetic gain for different crops. We investigated through stochastic computer simulation the advantages and disadvantages of adopting both GS and SB (SpeedGS) into commercial breeding programs for allogamous crops. In addition, we studied the effect of omitting one or two selection stages from the conventional phenotypic scheme on GS accuracy, genetic gain, and inbreeding. As an example, we simulated GS and SB for five traits (heading date, forage yield, seed yield, persistency, and quality) with different genetic architectures and heritabilities (0.7, 0.3, 0.4, 0.1, and 0.3; respectively) for a tall fescue breeding program. We developed a new method to simulate correlated traits with complex architectures of which effects can be sampled from multiple distributions, e.g. to simulate the presence of both minor and major genes. The phenotypic selection scheme required 11 years, while the proposed SpeedGS schemes required four to nine years per cycle. Generally, SpeedGS schemes resulted in higher genetic gain per year for all traits especially for traits with low heritability such as persistency. Our results showed that running more SB rounds resulted in higher genetic gain per cycle when compared to phenotypic or GS only schemes and this increase was more pronounced per year when cycle time was shortened by omitting cycle stages. While GS accuracy declined with additional SB rounds, the decline was less in round three than in round two, and it stabilized after the fourth SB round. However, more SB rounds resulted in higher inbreeding rate, which could limit long-term genetic gain. The inbreeding rate was reduced by approximately 30% when generating the initial population for each cycle through random crosses instead of generating half-sib families. Our study demonstrated a large potential for additional genetic gain from combining GS and SB. Nevertheless, methods to mitigate inbreeding should be considered for optimal utilization of these highly accelerated breeding programs.

Exploitation of data from breeding programs supports rapid implementation of genomic selection for key agronomic traits in perennial ryegrass

Exploitation of data from a ryegrass breeding program has enabled rapid development and implementation of genomic selection for sward-based biomass yield with a twofold-to-threefold increase in genetic gain. Genomic selection, which uses genome-wide sequence polymorphism data and quantitative genetics techniques to predict plant performance, has large potential for the improvement in pasture plants. Major factors influencing the accuracy of genomic selection include the size of reference populations, trait heritability values and the genetic diversity of breeding populations. Global diversity of the important forage species perennial ryegrass is high and so would require a large reference population in order to achieve moderate accuracies of genomic selection. However, diversity of germplasm within a breeding program is likely to be lower. In addition, de novo construction and characterisation of reference populations are a logistically complex process. Consequently, historical phenotypic records for seasonal biomass yield and heading date over a 18-year period within a commercial perennial ryegrass breeding program have been accessed, and target populations have been characterised with a high-density transcriptome-based genotyping-by-sequencing assay. Ability to predict observed phenotypic performance in each successive year was assessed by using all synthetic populations from previous years as a reference population. Moderate and high accuracies were achieved for the two traits, respectively, consistent with broad-sense heritability values. The present study represents the first demonstration and validation of genomic selection for seasonal biomass yield within a diverse commercial breeding program across multiple years. These results, supported by previous simulation studies, demonstrate the ability to predict sward-based phenotypic performance early in the process of individual plant selection, so shortening the breeding cycle, increasing the rate of genetic gain and allowing rapid adoption in ryegrass improvement programs.

Genomic prediction of crown rust resistance in Lolium perenne

BACKGROUND

Genomic selection (GS) can accelerate genetic gains in breeding programmes by reducing the time it takes to complete a cycle of selection. Puccinia coronata f. sp lolli (crown rust) is one of the most widespread diseases of perennial ryegrass and can lead to reductions in yield, persistency and nutritional value. Here, we used a large perennial ryegrass population to assess the accuracy of using genome wide markers to predict crown rust resistance and to investigate the factors affecting predictive ability.

RESULTS

Using these data, predictive ability for crown rust resistance in the complete population reached a maximum of 0.52. Much of the predictive ability resulted from the ability of markers to capture genetic relationships among families within the training set, and reducing the marker density had little impact on predictive ability. Using permutation based variable importance measure and genome wide association studies (GWAS) to identify and rank markers enabled the identification of a small subset of SNPs that could achieve predictive abilities close to those achieved using the complete marker set.

CONCLUSION

Using a GWAS to identify and rank markers enabled a small panel of markers to be identified that could achieve higher predictive ability than the same number of randomly selected markers, and predictive abilities close to those achieved with the entire marker set. This was particularly evident in a sub-population characterised by having on-average higher genome-wide linkage disequilibirum (LD). Higher predictive abilities with selected markers over random markers suggests they are in LD with QTL. Accuracy due to genetic relationships will decay rapidly over generations whereas accuracy due to LD will persist, which is advantageous for practical breeding applications.

Predictive ability of genomic selection models in a multi-population perennial ryegrass training set using genotyping-by-sequencing

KEY MESSAGE

Genomic prediction models for multi-year dry matter yield, via genotyping-by-sequencing in a composite training set, demonstrate potential for genetic gain improvement through within-half sibling family selection. Perennial ryegrass (Lolium perenne L.) is a key source of nutrition for ruminant livestock in temperate environments worldwide. Higher seasonal and annual yield of herbage dry matter (DMY) is a principal breeding objective but the historical realised rate of genetic gain for DMY is modest. Genomic selection was investigated as a tool to enhance the rate of genetic gain. Genotyping-by-sequencing (GBS) was undertaken in a multi-population (MP) training set of five populations, phenotyped as half-sibling (HS) families in five environments over 2 years for mean herbage accumulation (HA), a measure of DMY potential. GBS using the ApeKI enzyme yielded 1.02 million single-nucleotide polymorphism (SNP) markers from a training set of n = 517. MP-based genomic prediction models for HA were effective in all five populations, cross-validation-predictive ability (PA) ranging from 0.07 to 0.43, by trait and target population, and 0.40-0.52 for days-to-heading. Best linear unbiased predictor (BLUP)-based prediction methods, including GBLUP with either a standard or a recently developed (KGD) relatedness estimation, were marginally superior or equal to ridge regression and random forest computational approaches. PA was principally an outcome of SNP modelling genetic relationships between training and validation sets, which may limit application for long-term genomic selection, due to PA decay. However, simulation using data from the training experiment indicated a twofold increase in genetic gain for HA, when applying a prediction model with moderate PA in a single selection cycle, by combining among-HS family selection, based on phenotype, with within-HS family selection using genomic prediction.

Effect of Co-segregating Markers on High-Density Genetic Maps and Prediction of Map Expansion Using Machine Learning Algorithms

Advances in sequencing and genotyping methods have enable cost-effective production of high throughput single nucleotide polymorphism (SNP) markers, making them the choice for linkage mapping. As a result, many laboratories have developed high-throughput SNP assays and built high-density genetic maps. However, the number of markers may, by orders of magnitude, exceed the resolution of recombination for a given population size so that only a minority of markers can accurately be ordered. Another issue attached to the so-called 'large p, small n' problem is that high-density genetic maps inevitably result in many markers clustering at the same position (co-segregating markers). While there are a number of related papers, none have addressed the impact of co-segregating markers on genetic maps. In the present study, we investigated the effects of co-segregating markers on high-density genetic map length and marker order using empirical data from two populations of wheat, Mohawk × Cocorit (durum wheat) and Norstar × Cappelle Desprez (bread wheat). The maps of both populations consisted of 85% co-segregating markers. Our study clearly showed that excess of co-segregating markers can lead to map expansion, but has little effect on markers order. To estimate the inflation factor (IF), we generated a total of 24,473 linkage maps (8,203 maps for Mohawk × Cocorit and 16,270 maps for Norstar × Cappelle Desprez). Using seven machine learning algorithms, we were able to predict with an accuracy of 0.7 the map expansion due to the proportion of co-segregating markers. For example in Mohawk × Cocorit, with 10 and 80% co-segregating markers the length of the map inflated by 4.5 and 16.6%, respectively. Similarly, the map of Norstar × Cappelle Desprez expanded by 3.8 and 11.7% with 10 and 80% co-segregating markers. With the increasing number of markers on SNP-chips, the proportion of co-segregating markers in high-density maps will continue to increase making map expansion unavoidable. Therefore, we suggest developers improve linkage mapping algorithms for efficient analysis of high-throughput data. This study outlines a practical strategy to estimate the IF due to the proportion of co-segregating markers and outlines a method to scale the length of the map accordingly.

Toward Genomics-Based Breeding in C3 Cool-Season Perennial Grasses

Most important food and feed crops in the world belong to the C3 grass family. The future of food security is highly reliant on achieving genetic gains of those grasses. Conventional breeding methods have already reached a plateau for improving major crops. Genomics tools and resources have opened an avenue to explore genome-wide variability and make use of the variation for enhancing genetic gains in breeding programs. Major C3 annual cereal breeding programs are well equipped with genomic tools; however, genomic research of C3 cool-season perennial grasses is lagging behind. In this review, we discuss the currently available genomics tools and approaches useful for C3 cool-season perennial grass breeding. Along with a general review, we emphasize the discussion focusing on forage grasses that were considered orphan and have little or no genetic information available. Transcriptome sequencing and genotype-by-sequencing technology for genome-wide marker detection using next-generation sequencing (NGS) are very promising as genomics tools. Most C3 cool-season perennial grass members have no prior genetic information; thus NGS technology will enhance collinear study with other C3 model grasses like Brachypodium and rice. Transcriptomics data can be used for identification of functional genes and molecular markers, i.e., polymorphism markers and simple sequence repeats (SSRs). Genome-wide association study with NGS-based markers will facilitate marker identification for marker-assisted selection. With limited genetic information, genomic selection holds great promise to breeders for attaining maximum genetic gain of the cool-season C3 perennial grasses. Application of all these tools can ensure better genetic gains, reduce length of selection cycles, and facilitate cultivar development to meet the future demand for food and fodder.

Using variable importance measures to identify a small set of SNPs to predict heading date in perennial ryegrass

Prior knowledge on heading date enables the selection of parents of synthetic cultivars that are well matched with respect to time of heading, which is essential to ensure plants put together will cross pollinate. Heading date of individual plants can be determined via direct phenotyping, which has a time and labour cost. It can also be inferred from family means, although the spread in days to heading within families demands roguing in first generation synthetics. Another option is to predict heading date from molecular markers. In this study we used a large training population consisting of individual plants to develop equations to predict heading date from marker genotypes. Using permutation-based variable selection measures we reduced the marker set from 217,563 to 50 without impacting the predictive ability. Opportunities exist to develop a cheap assay to sequence a small number of regions in linkage disequilibrium with heading date QTL in thousands of samples. Simultaneous use of these markers in non-linkage based marker-assisted selection approaches, such as paternity testing, should enhance the utility of such an approach.

A Novel Multivariate Approach to Phenotyping and Association Mapping of Multi-Locus Gametophytic Self-Incompatibility Reveals S, Z, and Other Loci in a Perennial Ryegrass (Poaceae) Population

Self-incompatibility (SI) is a mechanism that many flowering plants employ to prevent fertilisation by self- and self-like pollen ensuring heterozygosity and hybrid vigour. Although a number of single locus mechanisms have been characterised in detail, no multi-locus systems have been fully elucidated. Historically, examples of the genetic analysis of multi-locus SI, to make analysis tractable, are either made on the progeny of bi-parental crosses, where the number of alleles at each locus is restricted, or on crosses prepared in such a way that only one of the SI loci segregates. Perennial ryegrass (Lolium perenne L.) possesses a well-documented two locus (S and Z) gametophytic incompatibility system. A more universal, realistic proof of principle study was conducted in a perennial ryegrass population in which allelic and non-allelic diversity was not artificially restricted. A complex pattern of pollinations from a diallel cross was revealed which could not possibly be interpreted easily per se, even with an already established genetic model. Instead, pollination scores were distilled into principal component scores described as Compatibility Components (CC1-CC3). These were then subjected to a conventional genome-wide association analysis. CC1 associated with markers on linkage groups (LGs) 1, 2, 3, and 6, CC2 exclusively with markers in a genomic region on LG 2, and CC3 with markers on LG 1. BLAST alignment with the Brachypodium physical map revealed highly significantly associated markers with peak associations with genes adjacent and four genes away from the chromosomal locations of candidate SI genes, S- and Z-DUF247, respectively. Further significant associations were found in a Brachypodium distachyon chromosome 3 region, having shared synteny with Lolium LG 1, suggesting further SI loci linked to S or extensive micro-re-arrangement of the genome between B. distachyon and L. perenne. Significant associations with gene sequences aligning with marker sequences on Lolium LGs 3 and 6 were also identified. We therefore demonstrate the power of a novel association genetics approach to identify the genes controlling multi-locus gametophytic SI systems and to identify novel loci potentially involved in already established SI systems.

Transcriptome Analysis Reveals Differential Gene Expression and a Possible Role of Gibberellins in a Shade-Tolerant Mutant of Perennial Ryegrass

The molecular basis behind shade tolerance in plants is not fully understood. Previously, we have shown that a connection may exist between shade tolerance and dwarfism, however, the mechanism connecting these phenotypes is not well understood. In order to clarify this connection, we analyzed the transcriptome of a previously identified shade-tolerant mutant of perennial ryegrass (Lolium perenne L.) called shadow-1. shadow-1 mutant plants are dwarf, and are significantly tolerant to shade in a number of environments compared to wild-type controls. In this study, we treated shadow-1 and wild-type plants with 95% shade for 2 weeks and compared the transcriptomes of these shade-treated individuals with both genotypes exposed to full light. We identified 2,200 differentially expressed genes (DEGs) (1,096 up-regulated and 1,104 down-regulated) in shadow-1 mutants, compared to wild type, following exposure to shade stress. Of these DEGs, 329 were unique to shadow-1 plants kept under shade and were not found in any other comparisons that we made. We found 2,245 DEGs (1,153 up-regulated and 1,092 down-regulated) in shadow-1 plants, compared to wild-type, under light, with 485 DEGs unique to shadow-1 plants under light. We examined the expression of gibberellin (GA) biosynthesis genes and found that they were down-regulated in shadow-1 plants compared to wild type, notably gibberellin 20 oxidase (GA20ox), which was down-regulated to 3.3% (96.7% reduction) of the wild-type expression level under shade conditions. One GA response gene, lipid transfer protein 3 (LTP3), was also down-regulated to 41.5% in shadow-1 plants under shade conditions when compared to the expression level in the wild type. These data provide valuable insight into a role that GA plays in dwarfism and shade tolerance, as exemplified by shadow-1 plants, and could serve as a guide for plant breeders interested in developing new cultivars with either of these traits.

Limitation of Grassland Productivity by Low Temperature and Seasonality of Growth

The productivity of temperate grassland is limited by the response of plants to low temperature, affecting winter persistence and seasonal growth rates. During the winter, the growth of perennial grasses is restricted by a combination of low temperature and the lack of available light, but during early spring low ground temperature is the main limiting factor. Once temperature increases, growth is stimulated, resulting in a peak in growth in spring before growth rates decline later in the season. Growth is not primarily limited by the ability to photosynthesize, but controlled by active regulatory processes that, e.g., enable plants to restrict growth and conserve resources for cold acclimation and winter survival. An insufficient ability to cold acclimate can affect winter persistence, thereby also reducing grassland productivity. While some mechanistic knowledge is available that explains how low temperature limits plant growth, the seasonal mechanisms that promote growth in response to increasing spring temperatures but restrict growth later in the season are only partially understood. Here, we assess the available knowledge of the physiological and signaling processes that determine growth, including hormonal effects, on cellular growth and on carbohydrate metabolism. Using data for grass growth in Ireland, we identify environmental factors that limit growth at different times of the year. Ideas are proposed how developmental factors, e.g., epigenetic changes, can lead to seasonality of the growth response to temperature. We also discuss perspectives for modeling grass growth and breeding to improve grassland productivity in a changing climate.

A Novel Two-Step Method for Screening Shade Tolerant Mutant Plants via Dwarfism

When subjected to shade, plants undergo rapid shoot elongation, which often makes them more prone to disease and mechanical damage. Shade-tolerant plants can be difficult to breed; however, they offer a substantial benefit over other varieties in low-light areas. Although perennial ryegrass (Lolium perenne L.) is a popular species of turf grasses because of their good appearance and fast establishment, the plant normally does not perform well under shade conditions. It has been reported that, in turfgrass, induced dwarfism can enhance shade tolerance. Here we describe a two-step procedure for isolating shade tolerant mutants of perennial ryegrass by first screening for dominant dwarf mutants, and then screening dwarf plants for shade tolerance. The two-step screening process to isolate shade tolerant mutants can be done efficiently with limited space at early seedling stages, which enables quick and efficient isolation of shade tolerant mutants, and thus facilitates development of shade tolerant new cultivars of turfgrasses. Using the method, we isolated 136 dwarf mutants from 300,000 mutagenized seeds, with 65 being shade tolerant (0.022%). When screened directly for shade tolerance, we recovered only four mutants from a population of 150,000 (0.003%) mutagenized seeds. One shade tolerant mutant, shadow-1, was characterized in detail. In addition to dwarfism, shadow-1 and its sexual progeny displayed high degrees of tolerance to both natural and artificial shade. We showed that endogenous gibberellin (GA) content in shadow-1 was higher than wild-type controls, and shadow-1 was also partially GA insensitive. Our novel, simple and effective two-step screening method should be applicable to breeding shade tolerant cultivars of turfgrasses, ground covers, and other economically important crop plants that can be used under canopies of existing vegetation to increase productivity per unit area of land.