Method for analysis of male pollinator plants

Abstract

The invention relates to a method for identifying male pollinator plants, through analysis of specific F1 plants from such male pollinator plants.

Description

[0001] METHOD FOR ANALYSIS OF MALE POLLINATOR PLANTS

[0002] The invention relates to a method for identifying male pollinator plants, through analysis of specific F1 plants from such male pollinator plants.

[0003] The exploitation of heterosis, obtained by breeding male lines with female lines, in wheat and other small-grain cereals like barley holds great potential to increase grain yield and yield stability. Hybrid breeding also offers the opportunity to split major genes with dominant gene action between the two parental pools. This leads to reduced gene stacking efforts compared to line breeding and enables the combination of gametes in the F1 which are linked in repulsion in the parental components. However, many small-grain cereals like wheat or barley possess a cleistogamic nature and it is necessary to be able to select plants for outcrossing ability, which is required to enable a cost-efficient hybrid seed production and to reduce cost of goods.

[0004] Good male floral characteristics like extruded anthers releasing enough viable pollen over a long period of time at anthesis are required irrespectively of the used hybrid mechanism that makes the female male sterile: chemical hybridization agents (CHA), cytoplasmic male sterility (CMS), or nuclear genetic male sterility (NGMS). Several indirect traits (anther extrusion, anther filament length, size of anthers, number of pollen grains, viability of pollen grains, plant height, flowering time and time-period of pollen release, etc.) have been evaluated for the selection of male candidates. There is a large variation for trait heritability, for the feasibility of scoring each trait in high-throughput screening.

[0005] There is a need to identify male plants that would have good pollinating ability (or good “ability to give seed set”, or pollination ability), that would be easy to implement and provide robust and reliable results. Such “ability to give seed set” correspond to the ability to adequately pollinate females and obtain seeds from the female. It can also be named “pollinating ability” in the present application. A male 1 shall be considered as having a higher ability to give seed set than a male 2, when more seeds are obtained after pollination of the same female (or of a set of females) in the same conditions. Currently, for comparing males in their pollinating ability, one would produce each potential hybrid combination in a crossing block and then measure how many F1 seed can be harvested from each crossing block. However,such a method is expensive. The method described herein makes it possible to use the ability to analyze F1 seeds with markers to precisely determine their parental origin, using F1 seeds that can be considered as obtained from a simulation of a crossing block on a very small scale.

[0006] Kherde et al (Crop. Sci. 1967, 7, 389-394) investigate the effectiveness of 45 wheat pollinators in fertilizing male-sterile wheat lines carrying one of three cytoplasmic types: Triticum timopheevi, Aegilops caudata, and Aegilops ovata. Seed set varied widely depending on both pollinator and cytoplasm male sterility system, with A. ovata generally resulting in the lowest fertility. While certain pollinators (e.g., Supremo, Denton) showed high seed set with specific CMS lines, no single pollinator performed best across all cytoplasmic types.

[0007] D’Souza et al (ZEITSCHRIFT FUER PFLANZENZUECHTUNG., 1970, 63, 246-269,) investigate the potential of wheat as a pollen donor for cross-pollination, particularly in the context of hybrid wheat breeding. This document shows significant variation among wheat varieties, with some (e.g., Diamant, Frontana Thatcher) showing strong potential due to high anther emergence and adequate pollen release. The males are not used in pools and this document does not use genetic markers.

[0008] Langer et al (2014, Plant Breed, 133: 433-441) focus on identifying phenotypic traits in male wheat plants that enhance their effectiveness as pollen donors for hybrid wheat breeding. The authors found significant genetic variation and high heritability for key traits like pollen mass and anther extrusion, indicating strong potential for selection. The authors only focus on optimizing male traits.

[0009] Boeven et al. (2018, Euphytica, 214:110) observed that hybrid seed set in wheat shows a large genetic variance and high heritability. The authors used 31 male lines and two female testers. Synchronization of flowering time and plant height influenced seed set and the indirect trait anther extrusion was only moderately correlated with the target trait seed set.

[0010] Schneider et al (2021, Front Plant Sci. 12:689825) focus on female elite lines and their suitability for hybrid seed production and as testers in wide crosses. The male lines used in this study were pre-selected for their good pollination performance. Hence, their pollinating ability was not evaluated in this study Miedaneret al (2022, PLANTS, 11(9), 1-17) discuss how effective pollen-fertility restoration, via restorer genes (Rf), is important for successful hybrid rye production and for reducing ergot contamination caused by Claviceps purpurea. The authors discuss advances in genetic mapping, marker-assisted selection, and fine mapping(notably of the Rfp1 gene on chromosome 4RL) and envisage future breeding strategies.

[0011] De Vries (1974, EUPHYTICA, 23 (3), 601-622,) investigates how seed set on male sterile wheat plants is affected by distance from the pollen source, the pollinator-to-sterile plant ratio, and the width of female strips. The study also indicates that pollinator varietal traits, such as flowering time synchronization and pollen-shedding ability, significantly influence cross-pollination success.

[0012] Whitford et al (2013, Journal of Experimental Botany, 64(18), 5411-5428) review the potential of hybrid wheat breeding to boost wheat yields and enhance food security amid environmental and resource constraints. It discusses the limitations of current hybridization methods and mentions genetic, genomic, and biotechnological approaches to improve hybrid seed production.

[0013] The Applicant thus presents a new method for assessing the suitability of cereal lines as pollinators (their pollinating ability). The method is based on the analysis of F1 seeds that have been obtained after pollination of male-sterile female tester lines by a pool of male lines. Many male candidates can thus be tested at once for their cross-pollination ability in one single crossing block. Furthermore, pooling the males would lead to competition between their pollinating ability, thus making it possible to intensify any difference between them in their “ability to give seed set” in the context of implementation of the method. The produced F1 kernels are analyzed with the same set of molecular markers that has been designed to individually discriminate the parental lines (male and female lines), enabling the inference of F1 pedigrees produced by a pollen cloud of bulked males. The relative contribution of each male corresponds to their relative “ability to give seed set” (the more F1 seeds originate from a given male parent, the more the male parent has the “ability to give seed set”).

[0014] Implementation of this method thus does not necessitate much space in the field and it can be performed with a relatively low number of seeds from each male line (present in the pool) making it possible to be applied in early breeding stages. It can also be repeated with different male-sterile females, thereby strengthening the “ability to give seed set” qualification that can be assigned to a given male.

[0015] The measured trait “ability to give seed” shows very high heritability and the data generated with this method can be directly used for the selection of genotypes with superior cross-pollination ability.Depending on the number and relatedness of male components (diversity set with sufficient number of males, bi-parental or multi-parent populations of males) and with an appropriate applied field design, it is possible to use the results obtained by implementation of the method (identification of the best pollinating lines) for analyzing the genetic architecture of this trait “ability to give seed set” by genomewide association studies or by linkage mapping. The data can be also used to train genomic prediction models. Depending on the used hybrid system, the harvested F1 kernels can be multiplied and hybrid performance and combining ability can be assessed in F2 stage.

[0016] The invention relates to a method for assigning a pollinating profile to male plant lines, the method comprising:

[0017] a) providing a plurality of biological samples obtained from individual plants of a mixed-paternity plant population, wherein each plant originates from fertilization of a male-sterile female line by a plurality (pool) of distinct homozygous male plant lines, and wherein the male lines of the plurality of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,

[0018] b) determining, for each biological sample, the paternal parent, using the set of markers,

[0019] c) attributing, for each biological sample, a male parent within the plurality of male plant lines based on the determination of b),

[0020] d) determining, for each male plant line, the proportion of plants in the population that have the male plant line as the male parent, and

[0021] e) generating, for each male plant line, a pollinating profile based on the determined proportion.

[0022] The invention relates to a method for characterizing and / or classifying male plant lines of a pool of male plant lines for their pollination ability, comprising a) Providing samples from individual plants of a population of F1 plants obtained by fertilizing a male-sterile female plant line by the pool of homozygous male plant lines, wherein the male lines of the pool of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,

[0023] b) Determining, for each plant of the population of F1 plants, the paternal parent, using the set of markers,c) Determining, for each male plant in the pool of male plant lines, the proportion of offspring plants in the population of F1 plants,

[0024] e) Characterizing and / or classifying the male plant lines of the pool of male plant lines for their pollination ability according to the proportion hereby determined.

[0025] The invention relates to a method for determining the pollinating ability of a given male plant line, comprising

[0026] a) Providing samples from individual plants of a population of F1 plants obtained by fertilizing a male-sterile female plant line by a pool of homozygous male plant lines, comprising the given male plant line, wherein the male lines of the pool of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,

[0027] b) Determining, for each plant of the population of F1 plants, the paternal parent, using the set of markers,

[0028] c) Determining, for each male plant in the pool of male plant lines, the proportion of offspring plants in the population of F1 plants,

[0029] d) Comparing the pollinating ability of each male plant line of the pool of male plant lines according to the proportion hereby determined, and

[0030] e) determining the pollinating ability of the given male plant line as compared to the other lines of the pool of male plant lines.

[0031] The invention also comprises a method for generating a dataset characterizing pollinating ability of male plant lines, comprising:

[0032] a) using a set of markers, determining genotype information for individual plants of a mixed-paternity plant population, wherein each plant originates from fertilization of a male-sterile female line by a plurality (pool) of distinct homozygous male plant lines, and wherein the male lines of the plurality of male plant lines and the male-sterile female plant line are individually distinguishable by the set of markers, b) attributing each plant of the population to a paternal plant line of the plurality of male plant lines, based on the genotype information,

[0033] c) calculating, for each paternal plant line, the proportion of plants in the population that have the male plant line as the male parent, and

[0034] d) storing the proportion values in an electronic dataset associated with the respective paternal plant line.

[0035] This method may be computer-implemented.The invention thus relates to a method for classifying male plant lines of a pool of male plant lines for their pollination ability, comprising

[0036] a) Providing a population of F1 plants obtained by pollinating a male-sterile female plant line by the pool of homozygous male plant lines, wherein the male lines of the pool of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,

[0037] b) Obtaining material samples from the plants in said population of F1 plants c) Determining, for each plant of the population of F1 plants, the paternal parent, using the set of markers,

[0038] d) Determining, for each male plant in the pool of male plant lines, the proportion of offspring plants in the population of F1 plants,

[0039] e) Classifying the male plant lines of the pool of male plant lines for their pollination ability according to the proportion hereby determined.

[0040] In some embodiments, the markers of the set of markers are molecular markers (epigenetic markers such as DNA methylation or histone modifications (that are associated with gene expression differences) are included in the molecular markers). These markers are very preferred. However, it may also be possible to use other inheritable markers such as isozymes (presence of variants of proteins detected by electrophoresis), presence or amount of specific metabolites,

[0041] In some embodiments, the material samples from the plants of the population of F1 plants are DNA samples.

[0042] In preferred embodiments, the plants are autogamous plants, in particular as listed below.

[0043] The population of F1 seeds on which the marker analysis is performed may be a subpopulation of the whole population of F1 seeds harvested after the fertilization of the male-sterile female plants. The number of F1 seeds that are analyzed can be determined by the person of skill in the art to allow statistically significant analysis of the proportions of the male parents in the F1 population. It would thus depend notably on the number of different male lines included in the pool of male lines. When using 10 to 30 different males, it may be sufficient to analyze between 150 F1 and a few thousands (such as 4500-5000 F1 seeds), although there is not real upper limit. Testing at least 150 F1 seeds, at least 180 seeds, at least 250 seeds may be sufficient.The seeds on which the analysis is performed are called F1 seeds or offspring seeds. They are harvested on the female lines, and result from the pollination of such male-sterile female lines by the pool of homozygous male plant lines.

[0044] After determining the parental origin for each F1 seeds, it is possible to determine the proportion of F1 seeds having a given male parent, in the population of F1 seeds. Thus, one can obtain, for each male plant in the pool of male plant lines, the proportion of offspring (seeds having this male as a parent) in the population of F1 seed. One can thus classify the male autogamous plant lines of the pool of male wheat autogamous plant lines for their pollination ability according to the proportion hereby determined, the higher the proportion, the higher the pollination ability.

[0045] The process is preferably performed with autogamous plant lines, although it can also be performed on semi-autogamous plant lines or allogamous plant lines. The description below focuses on the autogamous plant lines, but every aspect can be implemented with semi-autogamous plant lines or allogamous plant lines.

[0046] The process herein disclosed thus makes it possible to select the best pollinating lines from the pool of male autogamous plant lines. These best lines can then be used for further characterization (notably genetic characterization), or for implementing breeding schemes.

[0047] The pool of male autogamous plant lines thus comprises

[0048] multiple individuals of

[0049] multiple different male autogamous plant lines. Generally, the pool comprises at least 5, 10, 15, 20, 25, 50, 75, 100, 150, 200, 250 different male autogamous plant lines. However, it is conceivable that the pool comprises higher number of different male autogamous plant lines.

[0050] The number of plants of each different male present in the bulk of male plants that is used for pollinating the male-sterile female plants is generally at least 10, at least 20, at least 25, at least 50.

[0051] An “autogamous” plant is a plant that is able to perform self-pollination (autogamy), i.e. pollination in which pollen from the anthers of a flower is transmitted to the stigma of the same flower, and that would predominantly self-pollinate. The majority (generally more than 90%) of ovules are fertilized through self-pollination,in controlled settings (in the presence of other pollinating plants that would compete with self-pollination).

[0052] As autogamous plants, one can cite autogamous cereals, in particular oats, wheat, barley, rice, spelt, triticale, millet, rye, or rape. One can specifically mention oat (Avena sativa comprising byzantine; Avena nuda Avena strigosa), barley (Hordeum vulgare), rice (Oryza sativa), wheat (Triticum aestivum), durum wheat (Triticum durum), spelt (Triticum spelta) and triticale (Triticosecale).

[0053] One can also cite other plants such as pea (Pisum sativum), soybean (Glycine max), common bean (Phaseolus vulgaris), chickpea (Cicer arietinum), lentil (Lens culinaris), or tomato (Solanum lycopersicum) or eggplant (Solanum melongena).

[0054] In contrast, allogamous plants rely on cross-pollination (allogamy), where fertilization occurs between the gametes of two different plants of the same species. This can also be determined in controlled settings, where plants would produce significantly fewer seeds or fruits in the absence of external pollinators than in the presence of external pollinators. As examples of allogamous plants, which predominantly rely on cross-pollination for reproduction, and may exhibit traits such as dioecy, self-incompatibility, or reliance on external pollinators, one can cite maize (Zea mays), rye (Secale cereale) or carrot (Daucus carota).

[0055] One can note that some plants are partially autogamous plants, and rely on a combination of self-pollination (autogamy) and cross-pollination (allogamy). As partially autogamous plants, one can cite sunflower (Helianthus annuus, essentially allogamous, but autogamous under specific conditions), cotton (Gossypium spp. essentially autogamous, with cross-pollination performed by insects), tomato (Solanum lycopersicum) (allogamous under specific conditions), and Brassica species (essentially autogamous, with cross-pollination performed by insects).

[0056] One can note that each plant species would have both autogamous and allogamous ability, but would have predominantly one property. Allogamous ability makes it possible to ensure genetic diversity, where autogamous ability makes it possible to have offspring even in the absence of exogenous pollinator plants. The classification of plants as “autogamous” or “allogamous” is known by the person skilled in the art.

[0057] The method is preferably performed on autogamous or partially autogamous plants. Indeed, these plants generally have little ability to pollinate other plants (as the pollen would be used to fertilize the ovules of the same plant). These plants generally have anthers and stigma located in close proximity, often within the sameflower, facilitating self-pollination. In the context of making hybrids from these plants, it is thus needed to identify these autogamous plants that present the best ability to produce pollen with fertilizing ability. The method is thus perfectly adapted for detecting the ability of a given male plant to properly fertilize other plants.

[0058] As indicated above, the method can be performed on autogamous or partially autogamous plants. It is however not discarded to perform this method on allogamous plants, in order to identify and select best pollinator for hybrid breeding programs. In particular, the method could be performed with maize plants.

[0059] A “plant line” refers to a genetically distinct and homozygous group of plants of a given species that share specific, stable, and heritable characteristics. The homozygosity of the line ensures uniformity and consistency in the expression of its traits across successive generations. Typically developed through breeding, selection, or genetic engineering, a plant line is defined by its genetic stability and reproducibility, the plants exhibiting at least one specific trait, such as agronomic performance, disease resistance, or quality attributes.

[0060] A “male plant” is a plant that can pollinate female plants of the same species, and that thus produces pollen.

[0061] A “female plant” is a plant that can be pollinated and fertilized by the pollen of a male plant. After fertilization by male pollen, the plant will produce seeds or fruits (depending on the plant species).

[0062] A “male-sterile female plant” is a plant that is a female plant and that has lost its ability to produce pollen. Consequently, a male-sterile female plant would be fertilized by the pollen of another male plant. Various methods exist to induce male sterility. One can cite:

[0063] Cytoplasmic Male Sterility (CMS), which is caused by specific interactions between nuclear and mutated mitochondrial genomes, resulting in inability to produce functional pollen. Such sterility is stably inherited due to its origin in the maternal cytoplasm.

[0064] Genetic Male Sterility (GMS), which is controlled by appropriate nuclear genes, and is thus generally recessive. It may be called Nuclear Genetic Male Sterility (NGMS). In particular, fertility can be restored if the malesterility gene is complemented by a fertile allele. Several nuclear genes responsible for male sterility have been identified in wheat. Some notable examples include ms1, which leads to defective pollen development, ms2,where pollen wall formation is disrupted, ms5 or ms6, additional loci linked to male sterility. Other genes (ms45 or the like) are also known.

[0065] Male Sterility induced by chemical hybridizing agents (CHAs), where chemical agents can

[0066] o inhibit microsporogenesis and prevent the formation of viable pollen by disrupting meiosis in microspore cells (Methyl Benzimidazole Carbamate (MBC), Trifluoromethanesulfonamide (TFMS)) o Disrupt pollen wall formation, rendering pollen non-functional (such as 2-Hydroxy-5-nitrobenzyl Bromide or cycloheximide) o alter hormonal pathways (such as gibberellin, auxin, or cytokinin levels) required for normal anther and pollen development (one can use Ethephon, gibberellin synthesis inhibitors (such as paclobutrazol)

[0067] o have cytotoxic effects such as causing cell death in developing anthers (one can cite sodium methyl arsenate or maleic hydrazide) Environment-Sensitive Male Sterility (EGMS), where fertility is influenced by specific environmental factors. Genes involved in EGMS in wheat include tms5, which is linked to Thermo-Sensitive Male Sterility where male sterility is induced by temperature.

[0068] Although the CMS and NGMS methods would generally provide completely sterile plants, there may be some leakage, notably with the CHA or EGMS methods, but also with these CS and NGMS methods: in other words, some plants may produce pollen, although generally in a smaller amount than non-male-sterile plants. When the F1 seeds are analyzed, such leakage would be detected and the F1 seeds resulting from the self-pollination of leaking male-sterile plants would be discarded.

[0069] The method thus makes use of F1 plants or seeds that were harvested from the male-sterile female plants that had been exposed to the pollen of the pool of the male plants that are to be tested. The method thus comprises the technical steps of characterizing the F1 plants or seeds, by use of molecular markers (through DNA amplification, markers determination and genome analysis and comparison, usually computer-implemented). The method does not comprise any step of crossing plants that would take place before or after the steps specifically recited above.“Molecular markers” are segments of DNA associated with specific regions of the genome that can be used to identify genetic variation within and between plant lines. Common types of molecular markers include microsatellites (SSRs), single nucleotide polymorphisms (SNPs), and amplified fragment length polymorphisms (AFLPs). SNPs are currently generally used in modern studies for their abundance and high resolution. Epigenetic markers (such as the presence of methylation at specific locations of the genome) are also envisaged.

[0070] In the present methods, the male lines are generally not closely related, and the number of polymorphic molecular markers that are needed to be able to differentiate them does not need to be very high. It is thus generally possible to use between 13 and 20 polymorphic molecular markers to individually characterize each male parent line. When the genome of the male-sterile females is also known (this is preferred), when performing the method herein disclosed, the number of markers is appropriate when it makes it possible to discriminate the male and female lines. This would thus allow to determine, from the genome of the F1 seeds or plants, the paternal chromosomes from the maternal chromosomes.

[0071] If the genome of the male-sterile females is not known (such embodiment not being preferred), the number of polymorphic markers to be selected may need to be higher to be able to assign the paternal origin of the F1 seeds in the absence of knowledge of the maternal genome.

[0072] To characterize a plant line, one can follow the following protocol: DNA is first extracted from the plant tissue (which may be the F1 seed, F1 half-grain, or a tissue from a plant or plantlet grown from the seed). It is used as a template for PCR amplification of the genomic regions containing the selected markers. The amplified DNA fragments are then characterized using any technique appropriate for the marker type (one can cite gel electrophoresis for SSRs or sequencing, including sequencing, notably next-generation sequencing (NGS) for SNPs; one can also use a DNA chip, such as a DNA chip that is able to differentiate SNPs at various locations of the genome of the plant). The resulting genetic profiles are then analyzed using a computer to create the marker map of the plant, which is then compared to the ones of the male parent lines, or to compute conservative genetic distance between F1 and male parent, thereby making it possible to assign a male parent to the F1 seed or plant.Although it is possible to use the pool of male plants to pollinate heterozygous male-sterile female plants, it is preferred when the male-sterile female plants are male-sterile female lines ( / .e. are homozygous).

[0073] In some embodiments, the pool of male plants was used to pollinate a pool of male-sterile female plants. In this case, a higher number of molecular markers (several hundreds or thousands of markers) can be used to infer the conservative genetic distance between F1 and males. The female marker profile is not needed in this case. However, it is preferred when the male plants were used to pollinate a single line of male-sterile female plants. The resulting F1 seeds or plants thus all share the same maternal genome and only differ by the pollinating male genome. This makes the genome analysis easier, and allows to obtain specific information as to the effect of the female and flowering window on the pollination ability of males. Furthermore, the conclusion as to the seed-giving ability of the males would only depend on the male genome (the female plants all have the same genome) so that it makes it easier to conclude on this character. It is possible to perform the method with different male-sterile female lines, for instance male-sterile lines from different genetic backgrounds, or presenting different flowering time (nicking), so that the “ability to give seeds” of the males can be assessed depending on the nature of the female line. Indeed, some males can be very efficient in pollinating early-flowering females, while not appropriate for late-flowering females. Using such different female lines would make it possible to refine the knowledge of the ability to give seeds for the males of the pools. In this embodiment, it is possible to classify the male lines’ pollinating ability according to each flowering date.

[0074] In some embodiments, the pool of male lines is used to pollinate plants of a single male-sterile female line.

[0075] In some embodiments, the pool of male lines is used to pollinate plants of a pool of multiple male-sterile female lines (generally from 2, 3, 5 or 10 different male-sterile female lines, rarely more).

[0076] In this case, a SNP chip (containing few hundreds or thousands of molecular markers) can be used and analysis would measure conservative distance to the males. This allows to assign F1 profiles to males.

[0077] One can thus summarize a preferred process as follows:Provide F1 seeds (or plants) being offspring of a male-sterile female line fertilized by a pool of male lines that one wishes to characterize, wherein each male line and the male-sterile female line can be differentiated (individualized) from one another, using a set of markers (notably molecular markers)

[0078] Isolate sample (notably DNA) from the F1 seeds or plants, and use the set of markers on the isolated sample to obtain a profile of each F 1 seed or plant For each F1 seed or plant, determine the male parent, by comparing the profile of the F1 seed of plant to the profiles of the male plants. In some cases, it is not possible to determine a male parent, and the data for the F1 seed or plant will be discarded (or noted as “not usable”).

[0079] For each male parent, counting the number of F1 seeds or plants that originate from this male parent.

[0080] Calculating the frequency of the F1 seeds or plants that originate from each male parent among the F1 seeds or plants that have not been considered as “not usable”.

[0081] The male parent that has the highest frequency is considered to have the best “ability to give seeds” as compared to the other male plants from the pool. Although one could believe that the “ability to give seeds” may depend on the field conditions, the examples show that this property is conserved between experiments. The trait “ability to give seeds” is thus heritable and recurrent selection for it is feasible.

[0082] To improve the quality of the data, it is preferred to perform replication of the pollination of the females by the males in the male pool. Each replication can thus be individually analyzed, and the various replication or environment (year x location) data would be used for obtaining quantitative genetic parameters such as heritability. Repeating the methods with F1 seeds obtained from different locations, and on different years would allow identification of good pollinator lines, that maintain the ability in different environments and remain such overtime.

[0083] In one embodiment, the male-sterile female plants present Cytoplasmic Male Sterility (CMS). In another embodiment, the male-sterile female wheat plants present Nuclear Genetic Male Sterility (NGMS). In yet one embodiment male-sterility is induced by use of CHA in the female plants. In another embodiment, the male-sterile female plants present Environment-Sensitive Male Sterility and the pollination is performed under environmental conditions that trigger male-sterility.As indicated above, the male parents from the pools can be individually distinguished from each other by molecular (or genetic) markers. Furthermore, it is preferred when the set of markers also makes it possible to distinguish the male lines from the male females. This can ease the analysis of the F1 seeds to obtain the parents.

[0084] This embodiment is specifically useful when one uses a pool of different male-sterile female plants. In this case, the female plants of the pool (the population) of female plants are individually distinguishable by molecular markers.

[0085] In this embodiment, it may be interesting to determine both the paternal patent of the F1 seeds or plants and the maternal parent.

[0086] To obtain appropriate statistics, it is advisable that the pool of male plants contains the same number of seeds adjusted for germination rate for each male line. The method summarized above implied such embodiment. In other embodiments, the number of seeds of each line may be different, but should be known and taken into consideration to restate the statistics (the frequency calculating above should be adjusted, taking into account the frequency (relative weight) of each male line in the male pool).

[0087] The design of the pollinating experiment will be performed by the person of skill in the art, taking into account the following elements:

[0088] The males’ lines in the male pool should be classified in the same flowering date group (i.e. that the flowering time for the male and female plants should be within 0-5 days). In wheat, flowering time typically refers to the stage when 50% of the plants in a population have reached anthesis. In wheat, one can define an early-flowering group (plants flower earlier in the growing season), an intermediate-flowering group (plants flower midway between early and late groups) and a late late-flowering group (plants flower later in the season).

[0089] The female line (or lines) should be in the same flowering date group than the males (in order to favor that there are stigmas able to receive the pollen of the malesThe size of the female line (or lines) should be a little shorter than the size of the male lines (the pollen being transported by the wind, it would fall by gravity during before reaching the stigmas)

[0090] The orientation of the blocks or rows will preferably take the orientation of the dominant wind into consideration so that the male pollen will be transported to the female lines

[0091] The distance between the males and the female lines should be adapted to the pollen weight, so that the pollen is actually able to reach the female plants.

[0092] One of skill in the art is aware of the flowering group of the male or female lines, or of the size of the plants. Knowledge of this information could be a prerequisite before performing the method herein disclosed. Broad knowledge of the plants that are used to obtain the F1 population would indeed be important to make sure that the results are reliable and informative.

[0093] Synchronization of flowering would play a role in the interpretation of the results. More robust results may be obtained using several female testers with different flowering times (although they need to be able to be pollinated by the males so that their flowering time should be adapted to the one of the males). Indeed, a male plant may be considered as not having “ability to provide a seed set” if used with female plants with delayed flowering time.

[0094] To perform the pollination of the male-sterile females with the males of the pools, so that the F1 population needed for implementing the analysis method is obtained, one can thus envisage sowing one or more blocks of female plants, surrounded on the right and the left by male plants. The number of seeds that are sowed in the bulk of male plants and in the bock of female plants depends on the design of the field experiment and on the number of seeds of male-sterile female plants.

[0095] The width of the female plants’ blocks should preferably be between 20 cm and around 1 m, so as to make sure that the females in the middle of the blocks can be fertilized. However, width can be larger. Notably, width would be 1.55m if a standard yield plot drilling machine is used. One can note that, if the pollen cannot reach the middle of the female block, the female plants won’t be pollinated and no seed will be harvested. The width of the female block is thus not limited and can thus be several meters. Such large widths does not harm the experiment and could enable an easier field design using appropriate existing machines.The male plants blocks would be around the same size and should be no more than 1-1.5m from the female block. The distance of male plants to female plants could be much shorter (such as 20cm) if CMS or NGMS is used an no gametocide application is required. It is also to be noted that when the width of the male plants is too large, this may impair the ability of the most distant plants to pollinate the female plants. However, since the male seeds are all mixed before sewing, a large width does not modify the odds of the plants to pollinate the female plants.

[0096] The length of the block can be several meters long.

[0097] The density of the female and male plants can be adapted by one of skill in the art. A density between 150 and 400 plants I m2would be adapted for wheat. Density of the female tester may be higher than the density of male pool. This would result in less tillering of female plants and thus in a slightly earlier flowering which may be favorable.

[0098] It is preferred when the design of the pollination scheme of the population of male sterile female wheat lines by the male wheat lines is such as to allow potential pollination of each female line by each male line. Thus the width of the female rows is not too wide, and the male plants are close enough to allow pollen to be transported to the plants. Thus, the female plants in the population of the female wheat plants are exposed to pollination

[0099] In some embodiments, the male and female plants are sown together. This is not preferred as this would need more analysis of F1 kernels to identify the ones that are offspring of males and male-sterile females.

[0100] In some embodiments, each female line was planted in an individual (a specific) plot and wherein the plants of the pool of male lines were planted in rows adjacent (preferably on each side) to the female plots.

[0101] In summary, one can cite a method for classifying male plant lines of a pool of male plant lines for their pollination ability, comprising

[0102] a) Providing a population of F1 plants obtained by pollinating a male-sterile female plant line by the pool of homozygous male plant lines, wherein the male lines of the pool of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,

[0103] b) Obtaining material samples from the plants in said population of F1 plants c) Determining, for each plant of the population of F1 plants, the paternal parent, using the set of markers,d) Determining, for each male plant in the pool of male plant lines, the proportion of offspring plants in the population of F1 plants,

[0104] e) Classifying the male plant lines of the pool of male plant lines for their pollination ability according to the proportion hereby determined.

[0105] In one embodiment, the markers are genetic markers.

[0106] In one embodiment, the parental parent of the F1 plants is determined by sequencing or by analysis of the DNA genome of the F1 plants on a DNA chip for determining SNPs.

[0107] In one embodiment, the plant is an autogamous plant.

[0108] In one embodiment, the autogamous plant is an autogamous cereal selected from oat, barley, rice, wheat, durum wheat, spelt and triticale.

[0109] In one embodiment, wherein the male-sterile female plants are male-sterile female lines, in particular presenting Cytoplasmic Male Sterility (CMS), Genetic Male Sterility (GMS), or the male sterility of the female plants being induced by chemical hybridizing agents.

[0110] In one embodiment, the set of markers allows to individually distinguish the female plants and to distinguish them from the male plants.

[0111] In one embodiment, the pool of male plants contains at least 10 different male lines. In one embodiment, the male-sterile female wheat plants contain plants from between 1 and 5 different lines.

[0112] In one embodiment, the male sterile female plants and the male plants have synchronous flowering times.

[0113] in one embodiment, the pollination scheme of the population of male sterile female wheat lines by the male wheat lines was designed to allow potential pollination of each female line by each male line.

[0114] In some embodiments, the method may also be used to determine the general combining ability (GCA) of the male lines that are present in the pool of male lines that are used to pollinate the female lines.

[0115] In this embodiment, it is preferred to perform the method as disclosed above (pollinating female lines with the pool of male lines, obtaining the F1 seeds on the pollinated females) with a large number of females lines, in order to increase the value of the results that are obtained and reduce the risk of identifying specific combining ability of a given male with a female line used in the method. It is thus preferred when the method as described above is performed with at least 5 femalelines, more preferably at least 10 female lines, more preferably at least 15 female lines. These female lines have preferably the same flowering date window.

[0116] In short, the GCA of a given male of the pool is performed by:

[0117] Performing the method as described above, by fertilizing “n” females. It is reminded that the method uses the F1 seeds obtained from the female lines. Each seed has the female line as the maternal chromosome supplier and one of the male lines of the pool as the paternal chromosome supplier. A sample of the seed is used to perform the marker analysis and determining the pollinating ability of the male, according to the method herein disclosed (percentage of the male being the pollinator in a large set of F1 seeds). However, the sample is performed in such a way that the seeds retain their germinating ability.

[0118] For each male parent, a few seeds are thus sowed, for each of the “n” females. The F1 plants are grown, self-fertilized, and the F2 seeds (also called F1:2 seeds) are harvested and sowed.

[0119] The F1:2 seeds are grown and an agronomic trait is determined for the grown F1:2 seeds. Using this data, it is possible to determine the GCA of the initial parent male line (it is reminded that the initial parent male line is determined and known for each F1:2 plant), for instance using the method described in Technow (G3 Genes|Genomes|Genetics, 9 (4) 2019, 1557- 1569).

[0120] In more details, it is first reminded that “general combining ability” (GCA) refers to the average performance contribution of a given parental line (here a male line used in the pool) when crossed with a plurality of other lines (hence the need to use multiple male-sterile female, when the method is to be extended beyond the sole determination of the pollinating ability of the male line), measured on the performance of the resulting progeny (typically hybrids) relative to a population mean. In genetics terms, GCA is commonly used to represent the additive (and additive-related) component of a parent’s breeding value expressed in hybrid combinations. If not sufficient females are used, the GCA will be confounded with the “specific combining ability” (SCA), which represents cross-specific deviations attributable primarily to non-additive effects (e.g., dominance and epistasis).

[0121] Generally, GCA of a male parent line is determined by evaluating phenotypes of F1 progeny obtained from that male parent line crossed with multiple female lines and / or in various environments and estimating the male parent’s average effect onthe progeny performance, for example by analysis of variance, mixed models, or other statistical estimators that separate parental main effects from cross-specific interaction effects.

[0122] As indicated above, in the context of the present invention and disclosure, the male parent line (“Male Parent”) is characterized for its GCA with respect to one or more agronomically relevant traits by producing an F1 plant population from crosses between the Male Parent (in the pool) and male-sterile female lines (“Female Parent”), generating segregating progeny derived from said F1 population (F1:2 populations), phenotyping the F1:2 populations in field or controlled trials for the desired agronomically relevant traits, and estimating the Male Parent’s GCA for the target trait(s) based on performance measures of said F1:2 populations. The GCA estimate would be computed as the Male Parent’s average contribution to progeny performance across multiple Female Parent genotypes, with statistical adjustment for environmental, replication, and experimental design factors.

[0123] F1 populations are produced by crossing the Male Parent line as a pollen donor to a Female Parent line. As indicated above, the Female Parent line is male-sterile at the time of pollination. It is recalled that the Female Parent may be male-sterile by any suitable system, including (i) nuclear genic / genomic male sterility (NGMS), such as recessive or dominant nuclear-encoded male sterility alleles, optionally maintained using a maintainer line; (ii) chemically induced male sterility, such as application of a chemical hybridization agent (CHA) at an effective dose and timing to suppress pollen formation or viability; and / or (iii) cytoplasmic male sterility (CMS), including CMS systems where male sterility is conferred by cytoplasmic factors (e.g., mitochondrial) and optionally modulated by nuclear restorer-of-fertility (Rf) loci.

[0124] The pollination is performed as disclosed above, where the Male Parent is present in a pool of male lines.

[0125] Following production of the F1 population, each F1 seed is analyzed for identifying the Male Parent, for instance using part of the endosperm tissue, and F1:2 populations are generated by sowing, growing and selfing F1 plants (i.e., allowing or inducing self-pollination of individual F1 plants) for a the given Male Parent. In one embodiment, each F1 plant is self-fertilized to generate an F2 seed lot, and the F2 seed from a given F1 plant is maintained as a distinct “F1:2 family bulk” corresponding to that F1 plant (where both the Female Parent and Male Parent are known). In another embodiment, F2 seed lots from multiple F1 plants having thesame parental origin (same Male Parent and same Female Parent) are combined to form a composite F1:2 bulk for phenotyping. In another embodiment, F2 seed lots from multiple F1 plants having different female origin (same Male Parent and different Female Parents) are combined to form a composite F1:2 bulk for phenotyping. The number of F1 plants self-fertilized per cross, the number of F2 seeds harvested per F1 plant, and the bulking strategy are chosen to provide sufficient seed for replicated multi-environment evaluation and to reduce sampling error of the bulk mean. It is preferred when at least 20, at least 50, at least 100, or at least 200 F2 individuals are represented within an F1:2 bulk used for phenotyping, as this would makes the bulk mean more reliable and improve repeatability.

[0126] Because the Female Parent is male-sterile, the type of male sterility may influence pollen production in the F1 plants and, consequently, the ability and efficiency of the production of F1:2 seed when the F1 plants are self-fertilized. In some embodiments, one shall verify and / or ensure that the F1 plants are sufficiently fertile to produce F2 seed.

[0127] For genic or genomic male sterility, the F1 genotype at the sterility locus (or loci) determines whether the F1 plants are male-fertile, partially fertile, or male-sterile. For example, when male sterility is recessive and the Female Parent is homozygous sterile, an F1 produced using a fertile Male Parent typically becomes heterozygous and would be male-fertile, enabling F1 self-fertilization to generate the F1:2 seeds. When sterility is dominant or involves complex nuclear interactions, fertility of the F1 plant may be reduced, and alternative approaches (such as use of fertile counterparts) can be envisaged to ensure generation of segregating progeny for phenotyping.

[0128] For chemically induced male sterility, since the sterility is induced by treatment conditions and is generally not inherited genetically, the resulting F1 plants are typically fertile, absent continued treatment, and can be self-fertilized to produce F1:2 seed. In some embodiments, residual effects of the chemical treatment could be verified and lessened by selecting appropriate treatment of the plants, or seeds (such as washing) prior to self-fertilization.

[0129] For CMS-based systems, the Female Parent cytoplasm conferring sterility is inherited maternally and present in the F1 plants. In such embodiments, F1 fertility depends on whether the Male Parent genetic material present in a given F1 seed provides effective nuclear restorer allele(s) (Rf) that restore pollen fertility in thepresence of the CMS cytoplasm. F1:2 segregation may be hampered in this embodiment, and a larger number of seeds may be needed.

[0130] In all embodiments, fertility of the F1 plants intended for producing F1:2 seed may be assessed by any method available in the art, in particular pollen shed scoring, anther extrusion scoring, pollen viability staining, or presence of molecular markers linked to sterility / restorer loci.

[0131] In order to determine GCA of the Male Parent, the F1:2 populations are phenotyped for one or more agronomically relevant traits under conditions appropriate to the crop, including multi-location and / or multi-year field trials, optionally with replicated plots and randomized complete block designs, alpha-lattice designs, row-column designs, or other designs known to reduce environmental variance. Plot size, plant density, and management practices are selected to allow meaningful measurement of yield and yield components.

[0132] In some embodiments, each F1:2 family bulk is evaluated in at least two environments, at least three environments, or at least five environments, and with at least two replications per environment, to obtain stable estimates of family performance.

[0133] In one embodiment, the Male Parent’s GCA for a target trait is estimated by comparing the mean performance of multiple F1 :2 families that share the same Male Parent but differ in Female Parent, environment, or both. For example, where the Male Parent is crossed to multiple male-sterile Female Parents, separate F1 populations are generated and corresponding F1:2 families are produced and evaluated. The Male Parent’s GCA can be estimated as the average male effect across those crosses, optionally taking the female effects and environment into account. In another embodiment, when a limited number of Female Parent lines is used (or even for only one Female Parent line), the Male Parent’s GCA can be estimated relative to the other male lines of the pool by evaluating multiple F1:2 families derived from the different male lines of the pool crossed to the same male-sterile female background. This makes it possible to estimate the Male Parent’s relative general combining performance for the trait in that female background and across tested environments. However, the result may also highlight a specific combining ability of the Male Parent with this specific Female Parent genotype.Statistical estimation are performed using any method used in the art, notably linear mixed models where genotype (e.g., male, female, and / or family) is treated as a fixed or random effect, and environment and replication are treated as fixed or random effects as appropriate. In some embodiments, the model includes at least: an overall mean; an effect for Male Parent; an effect for Female Parent (when multiple females are used); an effect for F1:2 family nested within cross; and environment and / or environment-by-family interaction terms.

[0134] The Male Parent’s GCA may be reported as an estimated breeding value, a best linear unbiased predictor (BLLIP), a best linear unbiased estimate (BLUE), a genomic estimated breeding value (GEBV), or another estimator, together with confidence intervals or standard errors. In one embodiment, the Male Parent is classified as having a favorable GCA when its estimated effect exceeds a defined threshold relative to a control male line set, a population mean, or a breeding program benchmark. Using genome with markers and kinship matrices enable the training of genomic selection models for combining ability of the trait “ability to give seeds”.

[0135] The use of F1:2 family bulks makes it possible to estimate the Male Parent’s combining performance based on the bulk mean of segregating progeny, which captures the Male Parent’s average contribution to the phenotype across the various F1:2 genotypes derived from the same F1 genetic constitution. In some embodiments, within-family variance measures (such as variance of plant height, flowering time, or other segregating traits) are recorded as auxiliary information. However, the GCA estimate still remains primarily based on the family mean for the agronomically relevant traits.

[0136] This method can be applied to any agronomically relevant phenotype for which combining ability is informative, including yield and yield-related traits, stress tolerance, quality traits, and phenology traits. Non-limiting examples include:

[0137] For cereals, such as wheat, or barley: grain yield per unit area (e.g., t / ha), thousand-kernel weight, grains per spike, spikes per unit area, harvest index, lodging resistance, heading date, maturity date, plant height, disease resistance (e.g., rusts, powdery mildew), abiotic stress tolerance (e.g., drought, heat, cold), grain protein content, test weight, milling yield, and end-use quality parameters (e.g., gluten strength in wheat, malting quality traits in barley such as diastatic power and betaglucan content).For maize, grain yield per unit area, silage yield and digestibility, kernel number, kernel weight, ear traits (ear length, ear diameter), standability (root and stalk lodging), flowering traits (anthesis-silking interval, days to anthesis), plant and ear height, drought tolerance indices, nitrogen use efficiency, disease resistance (e.g., northern corn leaf blight), and grain quality traits (oil, protein, starch; and specialty traits where applicable).

[0138] For oilseed crops such as rapeseed / canola or sunflower, seed yield, oil content, fatty acid profile, glucosinolate content (canola), seed weight, flowering time, lodging resistance, pod shatter resistance (rapeseed), head size and seed set (sunflower), and resistance to relevant pathogens.

[0139] For other crops, seed yield, biomass, maturity, stand persistence, disease resistance, and quality traits relevant to end use (e.g., protein content, fiber traits, or processing characteristics).

[0140] In some embodiments, the F1 population and corresponding F1:2 populations used for estimation of the general combining ability (GCA) of the Male Parent are additionally used to analyze the genetic architecture of an interesting phenotype of interest (such as the “ability to give seeds” phenotype) by quantitative trait locus (QTL) mapping. The phenotype of interest may be linked to yield, flowering time, height, disease resistance, quality, or stress tolerance.

[0141] Such embodiment is performed on F1 population obtained from a single Male Parent line crossed to a single Female Parent line, resulting in a defined genetic background in which allelic segregation originates from the two specific parental genomes.

[0142] In these embodiments, the F1:2 populations constitute segregating progeny suitable for classical QTL analysis, as they comprise a plurality of F2 genotypes derived from self-fertilization of the F1 plants, and segregating for alleles contributed by the Male Parent and the Female Parent. The F1:2 populations are phenotyped for the phenotype of interest, and individual plants, sub-samples, or the F1:2 family bulks themselves are genotyped using molecular markers distributed across the genome, in particular single nucleotide polymorphisms (SNPs), insertions / deletions (InDeis), simple sequence repeats (SSRs), or any other polymorphic markers known in the art to perform QTL mapping.

[0143] In particular, individual F2 plants within an F1:2 family are genotyped and phenotyped, which allows construction of a genetic linkage map and mapping of QTLassociated with the phenotype of interest using interval mapping, composite interval mapping, multiple-QTL models, or equivalent statistical approaches. In another embodiment, one will genotype F1:2 family bulks, for example by bulked DNA sampling or low-pass sequencing approaches, and the mean phenotypic value of each F1:2 family is associated with marker segregation patterns to identify genomic regions contributing to variation in the phenotype of interest. In both cases, the resulting QTL represent chromosomal regions in which alleles derived from the Male Parent or the Female Parent contribute additively or partially additively to the expression of the phenotype of interest.

[0144] The identified QTL may be characterized by its or their genomic position, confidence interval, effect size, direction of effect, and interaction with environmental conditions. In some embodiments, QTL alleles contributed by the Male Parent that are associated with increased or otherwise favorable expression of the phenotype of interest are identified, thereby providing a genetic explanation for the Male Parent’s favorable GCA observed in the F1 and / or F1:2 evaluations. Such QTL information may be used to support marker-assisted selection, validation of the Male Parent’s breeding value, or further breeding decisions.

[0145] As the Female Parent used for generating the F1 population is male-sterile, it is important that segregating progeny can be obtained, as disclosed above, so that the contribution of the Male Parent to the segregating population can be determined.

[0146] In this embodiment, QTL mapping can be used as a complementary analytical step to GCA estimation, enabling identification of the genetic basis of the phenotype of interest observed in the F1 and F1:2 populations and providing additional characterization of the Male Parent’s contribution to said phenotype. The use of QTL mapping provides further insight into the loci underlying the additive genetic effects observed according to the Male Parent’s GCA.

[0147] In further embodiments, the method comprises applying a genome-wide association study (GWAS) and / or linkage mapping (QTL mapping) to dissect the genetic architecture underlying the trait “ability to give seeds.” In certain embodiments, a GWAS is performed using a genetically diverse population of male lines. The phenotype associated with the “ability to give seeds” for each male line is inferred from the proportion of offspring plants obtained in one or more populations of F1 plants produced using one or several female tester lines. The phenotypic values for the trait are estimated using a mixed-model approach to obtain best linearunbiased estimators (BLUEs). Subsequent genotyping of the male lines with genome-wide molecular markers, for example via a SNP genotyping array, enables the identification of marker-trait associations associated with the ability to give seeds.

[0148] In another embodiment, progeny derived from a bi-parental population are used as male lines for pollination of one or several female testers. The bi-parental population may be developed from two parental lines that exhibit contrasting phenotypic values for the ability to give seeds. The contrasting parents are crossed to produce F1 kernels, which are subsequently used to generate a doubled-haploid (DH) population. The resulting DH population is genotyped with genome-wide markers and employed as male lines in testcrosses with one or several female testers. This design enables classical QTL mapping to identify genomic regions contributing to the genetic architecture of the trait “ability to give seeds.”

[0149] DESCRIPTION OF THE FIGURES

[0150] Figure 1: Illustration of the design of a field experiment. Two rows of female plants are surrounded by two rows of a pool of male plants.

[0151] Figure 2: Contribution and 95% confidence intervals of each male in each set of analyzed F1 kernels. A-C Mid-early CMS female 1 line. Left and right panels represent two repetitions. D-E Mid-late CMS female 2 line. Left and right panels represent two repetitions.

[0152] Figure 3: Illustration of a simulation experiment. Expected least significant ratio based on the varying number of marker analyzed F 1 kernels (384, 768, 1152 or 1536 progenies) and number of males used in a mix (2 to 200).

[0153] Figure 4: Ability to give seed set in a trial. Ability to give seed set score for each female (light or darker grey) and a mean across females (darker grey).

[0154] Figure 5: Scatter plot showing the correlation between ability to give seed set and anther extrusion.

[0155] Figure 6: Series of 1.5m blocks sown in 6 headrows from the early group. The pollinator mix is sown in headrows 1, 2, 5 and 6. The females are sown in headrows 3 and 4.

[0156] Figure 7: percentage of F1 obtained after analysis of F1 harvested after pollination with different males. A result for Group 1. B. Result for Group 3. C Result for Group 7.

[0157] Figure 8: A. representation of a design according to the invention. Columns “201” represent blocs containing pools of male lines (334 lines per pool). Columns “202”and “203) represent a bloc of male-sterile female lines. Using 99 blocks thus make it possible to assess 334 male lines. B. representation of a regular design: each male line is assessed against one or the other female line on a given row. 334 rows and 5 columns are needed.

[0158] EXAMPLES

[0159] All experiments and field trials were performed in private premises.

[0160] Example 1. Measure of the ability to pollinize of multiple males.

[0161] Thirty (30) winter wheat lines from the Central European germplasm were selected to establish three male mixes containing 10, 20 or 30 males

[0162] Male mix 1 contains 10 genotypes,

[0163] male mix 2 contains all male mix 1 genotypes plus 10 further genotypes, male mix 3 contains all genotypes from male mix 1 and 2 plus 10 further genotypes.

[0164] Each male mix contained the cultivar PI KO which is known for its good crosspollination ability.

[0165] A seed counting device was used to ensure equal contribution of each male in the respective male mix.

[0166] Two CMS female lines were used as female testers to ensure a broad nicking window.

[0167] CMS female 1 has a mid-early flowering time,

[0168] CMS female 2 has a mid-late flowering time.

[0169] In total, 12 crossing blocks were drilled: Each male mix was crossed with each CMS A-line in two replications (3 mixes x 2 females x 2 replications = 12 crossing blocks).

[0170] Replications within one block were separated by rye to have an isolation and to facilitate easier visual identification in the field.

[0171] A standard single row seeder was used for drilling the crossing blocks, each of which contained six rows. 10g of seeds were used for each row. The two middle rows contained the CMS A-line. The two rows left and right of the two middle rows contained the male mixes.

[0172] Each row had a length of 75cm, and the distance between each row was 19cm. Distance between each plot in a block was 50cm.Nine plots represented one replication and two replications were isolated using two plots of rye. Therefore, one block comprised 20 plots (Figure 1).

[0173] The 12 blocks crossing blocks were surrounded by two known wheat cultivars which were not part of the experiment.

[0174] Before anthesis, five female ears within each replication were isolated by bags to check sterility. Male rows were removed after anthesis to avoid contamination with female rows during harvest.

[0175] Female rows of each replication were harvested separately using a sickle. A random sample of 372 F1 kernels from each replication was then sent for genotyping using 16 highly polymorphic SNP markers covering the following wheat genomes: 1A, 1B, 1D, 2A, 2B, 2D, 3A, 3B, 3D, 4A, 5A, 5B, 6A, 6B, 7A, and 7B. All 30 males and the two CMS females were genotyped with the same set of markers. All potential F1 marker profiles were calculated based on male and female marker information. The marker data of the genotyped F1 kernels was then used to assign the respective F1 pedigrees. F1 profiles with no clear match due to the two additional surrounding wheat cultivars were discarded from the analysis.

[0176] The contribution of each male within each replication was calculated.

[0177] Figure 2 shows the contribution of each male in each crossing block. The visual assessment of two replications for a respective CMS female and male mix show a consistent pattern and clear variation between the different male candidates.

[0178] This is supported by the analysis of repeatability (heritability of a single location) of the trait “ability to give seed set” for a respective female and male mix combination which were in a range of 0.82 to 0.91 (Table 1). Repeatability was calculated as h2= 1 -BLU2Pwhere i9B Pis the mean variance of a difference of two BLUPs and2C7G

[0179] OQ is the genotypic variance (Cullis et al. 2006).

[0180]

[0181]

[0182] Tab e 1: Repeatability of the trait “ability to give seed set” in first trial.

[0183] In total, 4352 F1 kernels were genotyped and 54% of these could be assigned to one of the potential male lines. The remaining 46% belong to one of the two surrounding wheat cultivars which were not part of the experiment.

[0184] There is a clear effect between the mid-early and mid-late CMS female lines. The cultivar PI KO confirmed its very high cross-pollination potential with both female testers, but some male perform differently depending on the flowering date.

[0185] Another trial was performed the following year.

[0186] 161 male lines were tested in front of two CMS female tester.

[0187] The trial was allocated in strips using yield plot sizes of male mixes and female testers

[0188] Each cell is representing a standard yield plot (7.2m2) instead of using a Hege rack drilling machine. Female plots were harvested with a combine and a 200g sample of each plot was taken.

[0189] Afterwards, a random sample of 4650 seeds of each female was genotyped using again the same 16 polymorphic markers. Due to the higher number of male lines and the limited number of markers, not all F1 seed could be assigned without error. 2889 and 3183 F1 kernels could be assigned correctly, respectively. The estimated repeatability was on a similar level compared to the previous trial (Table 2).

[0190]

[0191] Tab e 2 Repeatability of the trait “ability to give seed set” in another trial

[0192] Example 2. Simulation experiment

[0193] A simulation and analytical estimation of the experiment was conducted to assess the power of the method described in example 1.

[0194] Considering a male with an average True Pollination Rate of 1 / n (n the number of male in the experiment), its Observed Pollination rate will follow a binomial distribution X0 ~B(m,1 / n) where m is the number of F1s (and n the number of males). Another male significantly better than the average male should have a TruePollination Rate of k / n (k>0) and follow a binomial distribution X1 ~B(m,k / n). It is possible to find, depending on m and n, the least significant k values so that the OPR of the second male is higher that the OPR of an average male 95% of the time.

[0195] This least significant ratio has been calculated in dependence of different numbers of F1 progenies (384, 768, 1152 or 1536) and males used as a mix (2 -200). For instance, with 50 males and 384 F1s, each male may appear 1 / 50th = 2% in the progenies, only males at least 2.5 times more productive will have higher OPR 95% compared to an average male (Figure 3). This example provides information can be used to determine the number of F1 that need to be tested.

[0196] Example 3. Field trial

[0197] Another trial using 200 males in a mix and two females was implemented. A standard wheat yield plot drilling machine (7.2m2per plot) was used and the trial design was arranged as shown in Table 3.

[0198]

[0199] Table 3. Trial design using yield plot size

[0200] This design allowed a harvest with a combine with sampling. From each female strip, a total of 4650 seeds were randomly sampled and genotyped.

[0201] For sake of higher precision, all seeds were genotyped with a SNP chip containing 2681 markers (midplex).Based on a conservative genetic distance (D=1-N / M, where M=total number of markers, N=number of markers with at least one allele in common between the two individuals) F1 profiles and the ability to give seed set of each male was deduced.

[0202] A conservative threshold for genetic distance measure was set to 0.002. F1 kernels with a higher distance were not considered for analysis. This led to 5376 declared F1 profiles out of 9300 samples. For sake of simplicity, the frequency of each male was transformed linearly to a 1 to 9 APS score (1=low, 9=high).

[0203] Figure 4 is showing the results for each female tester individually and a mean across females. In other trials, male lines were also evaluated for their anther extrusion and genomic estimated breeding values for assessed. A scatter plot showing the correlation between anther extrusion and ability to give seed set can be drawn (figure 5). This data is revealing just a moderate correlation of r=0.37 and is supporting the hypothesis that anther extrusion is required but is not sufficient to select for the target trait seed set.

[0204] Example 4. Other trials

[0205] 54 hybrid production plots each made of 6 headrows and being 1.5m long were sown in Autumn 2021 with a nursery seed sowing machine.

[0206] 27 different combinations each made of single CMS females (headrows 3 and 4) and a mix of 10, 20 respectively 30 pollinators (headrows 1 , 2, 5 and 6) were sown in 2 repetitions.

[0207] The hybrid production was organized in 3 sub-blocks following 3 earliness groups: early, mid-early and mid-late. The 3 sub-blocs were separated with barley plots which pollen production occurs before the start of the wheat flowering and did not consequently compete with the production of pollen from the wheat pollinators.

[0208] The 9 females selected were CMS lines carrying the cytoplasm of Triticum timopheevii and thus fully sterile:

[0209] • Early: E1 (Early 1), E2 (Early 2) and E3 (Early 3)

[0210] • Mid-Early: ME1, ME2 and ME3

[0211] • Mid-late: ML1, ML2 and ML3

[0212] One of the 90 males selected wad a restorer line.

[0213] For each earliness category, 3 male sub-groups were constituted:

[0214] • first sub-group with 10 males chosen randomly.• first sub-group with 20 males chosen randomly.

[0215] • first sub-group with 30 males chosen randomly.

[0216] In every of the male groups, a pollinator line with a known high anther extrusion (rated 9 in a scale from 0=no extrusion to 9=full extrusion) was added for each earliness category (P-E, P-ME and P-ML).

[0217] Molecular analysis and results

[0218] The females of 48 plots were harvested by hand (with a sickel) at crop maturity stage (from 170g to 1 ,3kg per plot) and one seed sample per every plot was prepared for molecular analysis. A total of 8 samples was selected for molecular analysis.

[0219] • MIX 1 = E1 x 10 males

[0220] • MIX 2 = E1 x 20 males

[0221] • MIX 3 = E1 x 30 males

[0222] • MIX 4 = ME2 x 10 males

[0223] • MIX 5 = ME2 x 20 males

[0224] • MIX 6 = ME2 x 30 males

[0225] • MIX 7 = ML1 x 10 males

[0226] • MIX 8 = ML1 x 20 males

[0227] For each of these 8 samples, 186 seeds were individually analyzed with a set of 43 markers (PLI42 and the marker for Rf1). Not all analyzed seeds could be assigned to a given male, as explained above.

[0228] From the analysis, the male lines can be sorted (Tables 4-11).

[0229] Line with superior pollinating ability can be easily identified. For example, the line 9 in mixes 7 and 8 always show higher pollinating ability. It is not surprising that some male lines show good pollinating ability in some trials and less good ability in others. This is because the sampling of F1 is low and the standard deviation remains high. However, this example shows that the it is possible to classify the pollinating ability (or ability to gibe seed sets) of males by analyzing the F1 progeny.

[0230] Tables 4-6 show the results observed for Mixes 1-3 (early subgroups).

[0231]

[0232]

[0233] Table 4. Results for Mix 1. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 10%

[0234]

[0235] Table 5. Results for Mix 2. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 5%

[0236]

[0237]

[0238] Table 6. Results for Mix 3. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 3.33%

[0239] Tables 7-9 show the results observed for Mixes 4-6 (mid-early subgroups).

[0240]

[0241] Table 7. Results for Mix 4. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 10%

[0242]

[0243]

[0244] Table 8. Results for Mix 5. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 5%

[0245]

[0246]

[0247] Table 9. Results for Mix 6. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 3.33%

[0248] Tables 10-11 show the results observed for Mixes 7-8 (mid-late subgroups).

[0249]

[0250] Table 10. Results for Mix 7. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 10%

[0251]

[0252]

[0253] Table 11. Results for Mix 8. Count corresponds to the number of observed F1 for each parent. Expected frequency if the pollinating ability of all lines was identical: 5%

[0254] Example 5. Other trials

[0255] Field experiment

[0256] 14 production blocks of each 3 yield plots of 6m2each were sown in autumn 23. In every block, the central plot was sown with a mix of 3 CMS females and the 2 surrounding plots with a mix of 17 to 21 males.

[0257] 2 groups of 3 CMS females each were formed, the CMS females being in equal quantity in their respective groups:

[0258] Mid early group: ME-F1 + ME-F2 + ME-F3

[0259] Mid-late group: ML-F1 + ML-F2 + ML-F37 groups of 17 to 20 pollinators were constituted. The lines forming these groups are lines that possess an anther extrusion rated from 8 to 9 (scale from 0=no extrusion to 9=full extrusion).

[0260] - Mid-early groups: GROLIP1 to GROLIP5

[0261] - Mid-late group: GROLIP6 and GROLIP7

[0262] For each of the 9 blocks, 2 repetitions were sown.

[0263] At maturity the 18 female blocks were harvested with a plot harvester. 3.7 to 4.9kg were harvested from every single plot.

[0264] Molecular analysis and results

[0265] A sample of 186 seeds from each of the 14 harvested female plots was analyzed with a panel of 16 divergent SNPs (PLI16). The molecular data set of each of the 2 repetitions were merged. In every of the 7 combinations, it was possible to identify better pollinators, which gave the highest number of seeds in the harvested samples.

[0266] Figure 7 shows representative results of the percentage of F1 obtained after analysis of F1 harvested after pollination with different males for Group 1 (Fig. 7.A), Group 3 (Fig. 7.B) and Group 7 (Fig. 7.C). The highest percentage show a higher ability to give seed set. The same kinds of figures could be obtained for the other groups (not shown).

[0267] The method shows high potential to select for genotypes with good crosspollination ability and shows high flexibility as only needed genotyping of a few hundred or thousand F1 kernels. When using a CMS or NGMS female line, the method allows small-scale crossing blocks combined with manual harvest.

[0268] If a larger number of males should be tested, the method can be implemented using yield plots or long strips for males and females, as there is no limitation regarding crossing block size. If half-grains are genotyped, there is further the possibility to multiply the analyzed F1 kernels. The trait “ability to give seed set” showed very high repeatability when detected by the method herein disclosed.

[0269] Example 6.The execution of the method herein described makes it possible to perform the precise pollination screening of a significant number of males in a reduced land surface, hence reducing labor, costs, and land needs.

[0270] Without this approach, to get to the same results (classifying male plants), a conventional approach would require 10 times more plots. Indeed, each of the 334 males would be tested in isolated blocks made of 3 male plots and 2 female plots (male / female1 / male / female2 / male) so 5 plots total, plus a full 5-plot isolation row, and this for every single male instead of only 99 plots in the current design. This would bring the total amount of surface required to several hectares and significantly higher expenses. In addition, a harvest by combine or by hand result in more noise as there could be always mixtures.

[0271] Thus, the proposed design with internal competition in a minimal number of plots, is a much more time-efficient and cost-effective design and brings much higher precision due to the application of molecular markers for all observed F1 kernels.

Claims

1. CLAIMS1. Method for classifying male plant lines of a pool of male plant lines for their pollination ability, comprisinga) Providing a population of F1 plants obtained by pollinating a male-sterile female plant line by the pool of homozygous male plant lines, wherein the male lines of the pool of male plant lines and the male-sterile female plant line are individually distinguishable by a set of markers,b) Obtaining material samples from the plants in said population of F1 plants c) Determining, for each plant of the population of F1 plants, the paternal parent, using the set of markers,d) Determining, for each male plant in the pool of male plant lines, the proportion of offspring plants in the population of F1 plants,e) Classifying the male plant lines of the pool of male plant lines for their pollination ability according to the proportion hereby determined.

2. The method of claim 1, wherein the markers are genetic markers.

3. The method of claim 1, wherein the parental parent of the F1 plants is determined by sequencing or by analysis of the DNA genome of the F 1 plants on a DNA chip for determining SNPs.

4. The method of claim 1 or 2, wherein the plant is an autogamous plant.

5. The method of any one of claims 1 to 4, wherein the male-sterile female plants are male-sterile female lines.

6. The method of any one of claims 1 to 5, wherein the male sterile female plants present Cytoplasmic Male Sterility (CMS).

7. The method of any one of claims 1 to 5, wherein the male sterile female plants present Nuclear Genetic Male Sterility (NGMS).

8. The method of any one of claims 1 to 5, wherein the male sterility of the female plants was induced by chemical hybridizing agents.

9. The method of any one of claims 1 to 8, wherein the set of markers allows to individually distinguish the female plants and to distinguish them from the male plants.

10. The method of any one of claims 1 to 9, wherein the pool of male plants contains at least 10 different male lines.

11. The method of any one of claims 1 to 10, wherein the male-sterile female wheat plants contain plants from between 1 and 5 different lines.

12. The method of any one of claims 1 to 11 , wherein the male sterile female plants and the male plants have synchronous flowering times.

13. The method of any one of claims 4 to 12, wherein the autogamous plant is an autogamous cereal selected from oat, barley, rice, wheat, durum wheat, spelt and triticale.

14. The method of any one of claims 1 to 13, wherein the pollination scheme of the population of male sterile female wheat lines by the male wheat lines was designed to allow potential pollination of each female line by each male line.

Images