Final Report Summary - ADAPTAWHEAT (Genetics and physiology of wheat development to flowering: tools to breed for improved adaptation and yield potential)
Executive Summary:
ADAPTAWHEAT was established to enhance the knowledge and resources available to wheat breeders to enable them to breed new wheat varieties that were be better adapted to the changing growing environments of Europe and beyond, both now and in the future, and to ensure produce higher yields through the manipulation of floral development. Nine related workpackages, seven science-based and two focused on management, training and dissemination, were set-up achieve this objective.
At project completion, ADAPTAWHEAT has achieved this key objective by advancing the state-of-the-art in relation to knowledge and resources for the wheat breeding community. Key to the success of the project was the development of precise germplasm for gene discovery, achieved by producing five novel bi-parental populations from existing and novel material, and single gene-effect analysis to study how genes control sensitivity to day length, vernalisation requirement & other types of flowering control. The latter was possible through the use of Near Isogenic Lines, which are well characterized wheat varieties with single chromosomal segments replaced with a segment from another genotype. Because of the high value of this germplasm in future studies of the agronomic implications of allelic variation at these important genes, ADAPTAWHEAT has ensured that stocks of this novel germplasm have been properly documented and placed in secure storage.
ADAPTAWHEAT has exploited advanced genotyping technologies to characterize germplasm, discover genes & develop new molecular markers for breeding. Very high throughput genotyping has produce new genetic maps, and from which completely novel new molecular markers have been developed that significantly advance in the state-of-art for wheat breeding, and importantly have been validated by ADAPTAWHEAT partners, including the commercial partners.
ADAPTAWHEAT also developed a comprehensive information database from its physiology and agronomy phenotyping studies characterising how different genetic and environmental traits affect the modulation of floral initiation, heading date, and maturity in the panels of wheat germplasm available to it. Associated with this, the molecular phenotyping studies developed a gene expression toolbox that allowed observed variation, including environmental interactions, to be connected to specific markers of gene expression so that the environmental perception and flowering initiation pathways that mediate specific plant responses could be identified. Finally, efforts to develop models that predict the behaviour of wheat germplasm under specific conditions have led to the establishment of methodologies and that serves as a bridge between genetic and physiological understanding of phenotype behaviour, and a ready-to-use model that captures the actual values of particular alleles.
At its inception it was anticipated that ADAPTAWHEAT would not necessarily generate commercially beneficial material itself but develop much-needed knowledge and resources to underpin future development programmes by researchers and breeders alike. The presence of SMEs and several major breeding companies ensured the project had a long-term commercial focus. Evidence indicates this goal has achieved; proof has been the number of novel research and development programmes that have arisen that depend on ADAPTAWHEAT resources, for example, a successful research bid to the International Wheat Yield Programme (IWYP) by several consortium members. Other avenues national and international avenues are being explored too. Reviewers acknowledged that all ADAPTAWHEAT partners will benefit from this project and will use final results to progress their own research. Ultimately, the long-term legacy of the ADAPTAWHEAT project will be in ensuring the use of its outputs in wheat breeding programmes to lead to varieties better adapted to local environments. By their nature such breeding programmes are long-term ventures; ADAPTAWHEAT has provided a solid foundation for success.
Project Context and Objectives:
Introduction
Wheat is one of the major staple crops worldwide with approximately one-sixth of the total global arable land given over to its cultivation. Wheat is highly adaptable, grown on every continent except Antarctica. Wheat’s food qualities are equally important: its grains have higher protein content than maize and rice, two other major staple foods, and the crop provides approximately 20% of the total calorific needs. Wheat is particularly important within the European Union, where it is the most widely grown arable crop, being grown on over 25 million hectares. Wheat production in the EU is c. 150 million tonnes, while that of maize and barley is around 60 million tonnes each.
Wheat’s importance to the global economy cannot be underestimated with demand expected to increase by 60% to satisfy anticipated global demand. To meet this demand breeding efforts need to be more directed to designing wheat varieties better adapted to local situations. By better understanding the genetic and physiological determinants of wheat development which ultimately underpin wheat yield it will be possible to translate this understanding into more efficient breeding strategies.
Such strategies are urgently needed for two reasons. First, predicted changes in climate are a major risk to the existing levels and stability of crop production. The most deleterious effect of climate change is an increase in temperature as this has a major influence on wheat phenology. In addition, climate change will lead to more severe abiotic stresses, drought and heat, and the increased occurrence of extreme weather events, which in turn will affect wheat development. For example, the most detrimental climatic effects – water and heat stress – on wheat occur during ‘grain filling’ late season. The ability to manipulate flowering time in such instances would allow for greater yields. Generally, the ability to exploit the genetics and physiology of wheat development will be critical for the predictive breeding of climate ready wheat.
The second reason for urgency is the global demand for wheat demand is predicted to exceed the current pace of yield increase through breeding. It is important that yield increase return to levels that characterised the Green Revolution. This need is compounded by the issue of sustainability: future gains in wheat production must be achieved principally through breeding for yield potential and tolerance to stresses; with only minor, if any, increases in the use of water or fertilizer.
Identifying genetic and physiological tools to aid yield improvement has been a major goal of crop science research. Accurate genetic manipulation of developmental traits has the potential to provide higher rates of gain in yield. Wheat lends itself well to this ambition: due to its complex phenological control, wheat has the genetic potential to be adapted to any conditions projected for future EU, and global, climate change, thus improving yields. However, as previous work has shown, there is a lag between the discovery of new traits, the characterization of alleles/allelic combinations needed to deliver them and the deployment of this variation through breeding programmes.
Although there has been much work on wheat phenology, there has been limited research connecting these observations to the physiological and genetic mechanisms that control these differences. For example, there is some empirical knowledge of genotypic differences in heading time, and its value for adaptation, but the physiological and genetic bases of these differences are largely not known. Similarly, the major effects of temperature and photoperiod on time to heading are well described but the effects on the physiological determinants of this process are far less described and not fully understood. Another example is earliness in wheat, which is well-documented but where the physiological determinants of these differences in time to heading beyond photoperiod and vernalisation sensitivities have been relatively ignored.
The ADAPTAWHEAT Project
The purpose of the ADAPTAWHEAT project was bridge the gap between the wealth of information on wheat phenology and paucity of knowledge of the physiological and genetic mechanisms that control these differences. The focus of the project was on the physiological and genetic mechanisms for the control of flowering time and its phenological partitioning in the crop. By generating knowledge and producing resources, including valuable wheat germplasm and predictive models to explore wheat performance under future growing conditions, the project would make a significant contribution to improving wheat adaptation and yield. This in turn would address short-term Global/European food security issues and establish a platform for long-term genetic gain and maintenance breeding.
Why flowering?
Flowering time in wheat is a vital determinant in yield performance as it is key for both crop adaptation and yield potential; it regulates adaptation because it strongly influences life-cycle duration. Crop growth and partitioning around flowering is critical for grain production in wheat. Here, it has been established that yield is far more sensitive to changes in growth during the relatively short period from approximately three weeks before flowering to a few days after than during any other period. This critical period coincides with the period when stems and spikes (spikes are central stem on which grow the tightly packed rows of wheat flowers) grow at the highest rate, resulting in a strong competition for resources between them. Crop yield is particularly sensitive to changes in resource accumulation and allocation in that period. For example, there is a positive relationship between the number of grains that wheat has at maturity and the dry matter allocated into spikes at anthesis (the flowering period of a plant, from the opening of the flower bud), therefore lengthen the duration of the stem elongation phase when spike dry matter before anthesis is accumulated, would bring about direct improvements in yield potential.
The most critical feature of agronomic adaptation is to match, as closely as possible, crop growth demands with suitable environmental conditions, whilst avoiding stresses that affect yield, for example, frost during reproductive periods. In other words, it is a matter of adjusting the developmental window within the crop cycle to match the best possible environmental window in which the crop may grow. The likelihood of hotter drier summers predicted in current climate-change models indicates further adaptation is required if grain quality and quantity is to be maintained, and increased.
Project Objectives
The ADAPTAWHEAT partners looked to test a set of physiological hypotheses, and identify prospective genetic factors and valuable germplasm, for flowering time and its phenological partitioning, in wheat. These hypotheses were related to adaptation and crop performance. Specific objectives were to:
• assemble appropriate wheat germplasm panels (near isogenic lines, segregating populations, collections for association mapping, and specific varieties/breeding lines) representing known genetic variation for developmental traits of interest
• phenotype wheat germplasm panels for agronomic traits that determined adaptation and yield potential.
• characterise the developmental traits that determine flowering time in the genetic material of the germplasm panels (e.g. timing to onset of different phenophases and processes, dynamics of leaf number initiation and phyllochron, dynamics of floret development and grain set)
• relate phenotyped traits across levels of organisation, integrating genotypic and phenotypic information in QTL identification, and testing potentially useful QTLs in realistic breeding conditions
• deliver genetic markers for the physiological components of complex phenological traits, which will allow breeders to build flowering/phenology crop types by marker assisted selection of gene combinations.
• provide a gene expression tool box which will allow the observed variation, including environmental interactions, to be connected to specific markers of gene expression and show through which environmental perception and flowering time pathways genotypic responses are mediated
• explore the actual value of particular developmental alleles (affecting either adaptation or yield potential) across a wide range of growing conditions throughout Europe, to test response to different climates and prepare for climate change and changes in climate variability
To meet the overall objective, the project consortium was composed of a group of leading wheat researchers specialising in molecular biology, genetics, physiology, agronomy, and breeding, to provide the required balance of whole plant physiology and genetic expertise. ADAPTAWHEAT had nine interrelated work packages, focussing on: germplasm development and management; genotyping; molecular analysis; detailed physiological analysis; the agronomic assessment of key traits; integration of data for gene discovery and data validation; high level of modelling; training and dissemination and project management. The consortium was highly collaborative, with many partners involved in multiple work packages.
As a consequence of this effort over the four-year duration of the project, ADAPTAWHEAT generated a broad range of genetic and genomic resources necessary to quantify the agronomic implications of specific variation in crop development. The consortium demonstrated how this variation could be altered to allow the wheat crop to be well-adapted to the growing season, avoiding stresses such as frost, drought, and heat while maximising reproductive growth in good conditions so that the maximum quantity of grain can be produced from the resources available. ADAPTAWHEAT has produced the knowledge, genetic markers, and precisely defined types of wheat required by breeders as tools to fine tune adaptation and performance in this key crop.
Key to the project’s success was data exchange between partners, achieved through regular communication, e.g. the project’s wiki, and meetings at all levels of the project, and through dissemination of information to diverse stakeholders beyond the project including plant breeders, potential scientific collaborators and policy makers. Through these connections the legacy of ADAPTAWHEAT is assured in the future.
Project Results:
There were seven scientific workpackages in the ADAPTAWHEAT project (WP1 – WP7). The main results and outputs from each of these are reported below:
1.3.1 WP1: ESTABLISHMENT OF GENETIC PANEL
Start month 01; End month 36
WP1 leader: 01 JIC; Other WP1 Participants; 05 MTA-ATK; 07 CRI; 08 IPBB; 10 CIMMYT INT; 11 CSIRO; 13 SB; 14 IFVCNS; 19 SELGEN
WP1 was the anchor of the ADAPTAWHEAT project. All other WPs were based on experimentation and the analysis of the germplasm panels selected in WP1. These included varietal panels, Near Isogenic Lines (NILs), segregating populations, association panels, and validation sets. The multiplication, distribution, and quality control of these materials are not research outputs in themselves but they underpinned the success of the other WPs. In addition to background germplasm such as the DH populations Avalon x Cadenza, Buster x Charger, Weebill x Bacanora, WP1 also initiated the production of completely new germplasm resources based on the outputs of other ADAPTAWHEAT WPs. So, in the case segregating populations this includes Paragon x Pamayat Azivo, Paragon x Garcia, Paragon x Magnif 41, Toborzo x Verbunkos.
In addition, some populations that were planned for development under the support of other projects (e.g. UK WGIN project) could be done on a larger scale with better aligned resources as a result of ADAPTAWHEAT. The best example of this is Paragon x Garcia. It should be noted that the UK elite spring wheat Paragon is a common parent in many of these populations. In addition, many of the NILs used in ADAPTAWHEAT are in the Paragon background. Moreover, aligned projects such as UK WISP, and the IWYP phenology project led by Dr Simon Griffiths (JIC) use Paragon as a parent in crosses with CIMMYT germplasm. Together these segregating populations for a nested association mapping (NAM) panel. Similarly, the NILs in a Paragon background provide a benchmark from which the first assessment of breeding value for new traits and trait combinations can be assessed on a level playing field. The intensive efforts of WP1 to distribute these materials, going through the hurdles of phyto-sanitory certification, homogeneity testing and cross site referencing, set up a strong base for the project and its continued influence on others. In addition to segregating populations, new NILs were developed in the WP. These are mostly in the genetic background of Paragon. They include genes for winter habit, capturing an allelic series of increasing vernalization requirement in the former spring wheat but also new QTL such as the EPS effects identifies on chromosome 4A.
As the project progressed and panels were received and multiplied, the focus of the WP moved more to the development of germplasm aspects of validation strategies. This included innovative approaches such as self-fertilising RILs that were heterozygous in QTL regions for testing of single QTL effects. We also focussed on specific association panels for marker validation. Most notably the CIMMYT wheat association mapping initiative (WAMI) panel was established as a highly effective means of showing the value of molecular markers from WP2 in materials far removed from the original genetic dissection populations. JIC made a commitment to maintain all of these materials within its seed store. In addition, the information relating to the germplasm development (crossing books, pedigrees etc.) is maintained.
1.3.2 WP2: GENOTYPING BREEDING & EXPERIMENTAL MATERIAL
Start month 07; End month 48
WP2 leader: 15 TGEN; Other WP2 participants: 1 JIC; 05 MTA-ATK; 08 IPBB; 14 IFVCNS; 19 SELGEN
This WP developed and deployed molecular markers for the ADAPTAWHEAT project. Early in the project the main activity was the generation of genetic linkage maps. The genotying load was carried by the use of the iSelect 90K array, KBiosciece KASP technology, simple sequence repeats and the iSelect 15K array. New maps were produced for Buster x Charger, Weebill x Bacanora, Paragon x Pamayat azivo, Paragon x Glasgow, Toborzo x Verbunkos, Paragon x Magnif 41. These fed into WP6 where QTL analysis was carried out. As QTLs were identified and prioritised, WP2 developed the best possible molecular markers. This included intense position cloning efforts. Successfully isolating EPS-D1, PpdB2, and PpdB3, we developed new diagnostic SNP markers for ELF-3, FT3, and TOE-1 respectively. The importance of the FT gene family led us to develop Taqman copy number variant (CNV) assays for the entire family as well as for ELF3 and GI. The effort on CNV is in recognition of the fact that many of the functional polymorphisms we and others cloned emerged as large scale structural rearrangements involving deletion or various levels of duplication. This approach was fruitful as subsequent Panel screens using the Taqman assays showed CNV association with phenology traits.
In addition to the extensive mapping efforts of ADAPTAWHEAT a large amount of marker assisted selection was conducted. This included the selection of foreground segments in backcross programmes for genes such as EPS4A and Hereward, Claire, and Malacca alleles of VrnA1 into Paragon. In some cases, such as 4A, we also employed background selection using 4 KASP marker per chromosome in BC1 populations >500 to maximise the genomic contribution of the recurrent parent at an early stage of backcrossing. These experiments represent a major commitment of time and resource designed to expedite the programme as a whole.
In addition to backcross selection we self-fertilised heterozygotes from RILs. After phenotyping of single plants in WP6 the individual samples were taken and genotypes to identify the two homozygous and singe heterozygous classes so that single marker regression could be carried out, completing this phase of validation.
A major output of the WP was a booklet (available from project wiki) which described in detail how the various molecular markers produced by the project can be used in research and breeding.
Finally, the WP carried out checks. Random samples from field plots were genotyped using most discriminative KASP marker sets to ensure that packaging or drilling errors had not compromised multi-site experimental analysis.
1.3.3 WP3: MOLECULAR PHENOTYPING
Start month 01; End month 48
WP3 leader: 03 INRA; Other WP3 participants: 01 JIC; 05 MTA-ATK; 11 CSIRO
Initially, WP3 pursued the measurement of developmental gene expression in response to environmental variables. The hope was to extend the state-of-the-art to allow researchers to infer a conceptual framework for ecophysiological models that took explicitly into account gene expression. In pursuit of these goals the WP moved many aspect of phenology transcriptomics forward. This included new qPCR assays for key genes such as Vrn-1, Vrn-2, Vrn-3/FT1, FT2, ELF3, GI, Ppd-1, and HSP70. The data generated helps in our qualitative models of the mechanism by which the QTL we identified are likely to be mediating their effect through the genetic networks controlling phenology. Consequently, we could see that EPS-D1 earliness is associated with reduced ELF-3 expression and increased GI. We saw that the very long vernalisation requirement of the variety Charger was tied in with a resurgence of Vrn2 expression when plants were moved from vernalizing conditions to ambient temperature prematurely. There are many high impact stories such as this. However, half way through the project it was clear that our ambitious goal of predicting phenology outcomes on the basis of expression data was not going to deliver. As a result, for deliverables relating to this objective, a change in direction was agreed with our project coordinator to shift focus to gene based modelling using our QTL data.
In the gene based modelling work we produced new models with the framework of APSIM where the mean prediction error ranged from 2.6 to 4.3 days depending on the environment which constitutes a very reliable estimation of heading date.
The WP made significant steps forward in our understanding of the ambient temperature response of wheat. We showed that ambient temperature as an environmental factor and in association with photoperiod may regulate plant development at various consecutive levels of developmental phases, such as the later stages of floret differentiation and via the process of intensive stem elongation.
The different regulation mechanisms of early apex development and the later plant developmental phases is also underlined by the photoperiod and ambient temperature dependent association patterns of VRN1, VRN3 gene expressions and the apex and plant developmental phases. It is clear from our findings that VRN2 gene has a significant role in determining plant development even after vernalization, which is more pronounced when longer photoperiods occur together with higher temperatures. In these cases, a transient high VRN2 expression peak appearing between 100-200 GDD is significantly associated with the ability of the initiation and the procession of the intensive stem elongation. More generally, we used Panel 1 and 4 screens to show that a wide range of variation for ambient temperature response exists in bread, beyond the passive increase in developmental rate that is often assumed. In collaboration with WP1, this WP has initiated the production of a number of new resources to further dissect these interesting responses.
1.3.4 WP4: PHYSIOLOGICAL PHENOTYPING
Start month 01; End month 48
WP4 leader: 02 UdL; Other WP4 participants: 03 INRA; 04 IPK; 05 MTA-ATK; 07 CRI; 08 IPBB; 09 CONICET; 10 CIMMYT INT; 11 CSIRO; 13 SB; 14 IFVCNS; 18 ARVALIS
The focal aim of WP4 was to characterise developmental traits determining flowering time in the genetic material of the panels studied. Through a number of experiments carried out in a large number of locations and through at least two growing seasons we characterised flowering time and its components in all the genetic panels of the project.
Modern cultivars currently grown in many different countries participating in the project (across three continents) always exhibited variation in phenology, including not only variation in time to anthesis but also in the partitioning of this period into phases occurring before or after the onset of stem elongation. Then, in most of the 21 experiments that were conducted within ADAPTAWHEAT evaluating levels of variation within modern and well adapted cultivars we uncovered that cultivars not varying in flowering time could exhibit differences in partition of that overall duration into component phenophases. Thus, in some genotypes the time until anthesis included a longer phase from sowing to stem elongation whilst in others it included longer phases from stem elongation to anthesis.
Eps alleles affected phenology only marginally, as expected from the literature: in most cases these genes are extremely useful only for fine-tuning phenology. The late alleles normally delayed heading and did so through lengthening both the period from sowing to the onset of stem elongation and the period from then to heading when the Eps gene considered was that of chromosome 3A, whilst the Eps gene in chromosome 1D seemed to have affected time to heading mainly through lengthening the duration of the stem elongation- heading phase. Plastochron, the reciprocal of primordia initiation rate, of leaf primordia was not clearly affected by Eps alleles but that for spikelet primordia tended to be shorter in lines with the early alleles. Therefore, the action of the early alleles in shortening the developmental periods did not result in differences in spikelet number as the spikelet initiation rate was faster in the lines with early alleles resulting in compensation in the number of spikelets per spike. Similarly, there were not clear effects of the Eps alleles on the dynamics of floret development and the rates of generation/degeration of floret primordia were similar for early- and late-NILs. Consequently, there were no significant differences in the number of fertile florets along the spike. Thus these alleles may be used to finetune time to flowering without producing potential penalties in reproductive sink strength.
Photoperiod insensitivity alleles naturally advanced flowering time but in ADAPTAWHEAT we evidenced that even though both phases were responsive the insensitivity alleles in general shortened the duration of the stem elongation-heading phase more than the sowing-stem elongation phase. Besides, the effect of the dosage of insensitivity alleles tended to be stronger on the stem elongation-heading phase than in the preceding period from sowing to the onset of stem elongation. The effects of the insensitivity alleles on the duration of the vegetative phase reduced final leaf number and in some cases also reduced phyllochron. In addition, there was a clear reduction in the number of spikelets initiated in NILs with insensitivity alleles, despite that the rate of spikelet initiation tended to be accelerated by the introgression of Ppd-insensitivity alleles. However, these Ppd alleles did not seem to have affected fertility of the four most proximal florets, even though the time for floret development which was reduced affecting the fertility of the fifth floret. Then, the wild type with strongest sensitivity had a longer floret initiation phase and even though it initiated almost the same number of floret primordia than the NILs with insensitivity alleles, floret mortality tended to be lower.
In terms of quantifying the responses to changes in day-length of the different NILs, the wild type of the panel (with the three ppd sensitivity alleles) showed that the sensitivity to photoperiod was evident during the stem elongation phase, with a direct response to photoperiod manipulated exclusively during this period (when final leaf number had been determined long time before). However, in experiments in which day-length was manipulated throughout the whole period from sowing to anthesis the vegetative phase was more responsive than the reproductive phases, which were responsive and the responsiveness was positively related to the number of sensitivity alleles. These results from experiments in which the different NILs were maintained under contrasting day-length treatments during either the whole period from sowing to anthesis or only from terminal spikelet onwards, were confirmed in experiments with reciprocal transfers between short and long photoperiod conditions. When analysing together the different experiments in which photoperiod was manipulated we concluded that Ppd-1 modified only the photoperiod sensitivity, whilst the threshold photoperiod for saturating the response and the minimum duration of developmental processes at these saturating day-lengths (intrinsic earliness) were largely unaffected.
1.3.5 WP5: AGRONOMIC PHENOTYPING
Start month 06; End month 48
WP5 leader 05 MTA-ATK; Other WP5 participants: 01 JIC; 02 UdL; 07 CRI; 08 IPBB; 09 CONICET; 10 CIMMYT INT; 12 RAGT 2n; 13 SB; 14 IFVCNS; 16 KWS-L; 17 LVH UK; 18 ARVALIS; 19 SELGEN
The main objective of WP5 was to phenotype in the genetic panel of the project the agronomic traits determining adaptation and yield under the diverse field grown conditions. These panels included (i) mapping populations, (ii) collections of wheat varieties, and (iii) near isogenic lines for which we determined yield and its main determinants in a number of field experiments carried out across different countries and growing seasons.
The population of Avalon x Cadenza exhibited variation in time to heading including transgressive segregation, with the exception of the experiments carried out in Kazakhstan, (where only the half of the population with spring habit was grown over spring and summer). In general, yield varied much more than phenology and variation in crop productivity was consistently transgressive. Therefore, in most of the experimental conditions (locations x growing seasons) there were lines out-yielding the highest-yielding parent, which might be considered part of the elite material for crossing in breeding programs focused in those locations. In general, though not always, yield tended to increase with shorter time to heading. Interestingly the relationship seemed stronger if the reduction in time to heading were due to accelerated development in earlier phases, suggesting that lengthening stem elongation phase, even when reducing total time to heading might be beneficial for yield. Variations in yield were strongly related to those in the number of grains per unit land area while they were rather independent of variations in the average weight of the grains.
The population of Toborzó x Verbunkos (both winter wheats, segregating for Ppd-D1) also varied in time to heading, due to variations in both sowing-onset of stem elongation and duration of the stem elongation phase, being the latter more relevant than the former for establishing the genotypic differences.
Regarding the collections of modern cultivars from each location involved in ADAPTAWHEAT WP5, even though all genotypes were modern, well adapted cultivars in each location, there was clear variation in agronomic performance. In general, for this panel there was not a consistent relationship between yield and phenology (when analysing genotypic effects). There was not relationship between yield and the duration of any of the phenophases either. Most differences between cultivars in yield were related to similar differences in the number of grains per unit land area. The average weight of the grains also varied and was also related to yield but the relationship was not only weaker than that with grain number but also less consistent across cases analysed. Thus, across all the ranges of breeding programs analyzed differences in grain number was chiefly responsible for those in yield among modern cultivars with grain weight occasionally contributing as well. The genotypic differences in grain number seemed reflecting differences in fruiting efficiency more than in spike dry weight at anthesis. There was a trend for compensations between fruiting efficiency and spike dry weight at anthesis but not a very close negative relationship, suggesting that when focusing on physiological bases of grain number generation within elite germplasm fruiting efficiency would be more relevant.
Considering the NILs for Ppd alleles, the introgression of insensitivity alleles tended to increased yield, without a clear pattern of in which genome the insensitivity affected more. Overall the experiments conducted; there was a negative relationship between yield and duration of sowing-heading (then, overall environments, the higher number of insensitivity alleles the stronger the reduction in heading time and the higher the yield). Most effects of Ppd alleles on yield were mediated through effects of these alleles on both the number of grains per unit land area and thousand grain weight. Therefore, carrying photoperiod insensitivity allele/alleles increased yield in most environments, in line with the effects it had on both the number of grains/m2 and their average weight; but there was not a clear and consistent ranking between the different insensitivity alleles on the magnitude of their effects on yield and its determinants.
Considering the NILs for Eps alleles introgressed in chromosome 1D and 3A with AxC background out-yielded in general (but not always) lines with Eps early alleles; whilst in all the NIL pairs with Eps late alleles introgressed in chromosome 1D with SxR background, lines with the Eps early alleles yielded more than those with Eps late alleles. Thus, averaging across all the isolines the introgression of Eps early alleles reduced yield respect to Eps late alleles in the chromosome 1D for both backgrounds (but more noticeably in SxR), whilst the effect was opposite with the introgression of Eps early alleles in chromosome 3A. In all cases, changes in yield produced by the introgression of Eps late alleles were much more strongly related to those in grain number than in thousand grain number.
1.3.6 WP6: INTEGRATION ACROSS LEVELS OF ORGANISATION
Start month 13; End month 48
WP6 leader: 10 CIMMYT INT; other WP6 participants: 01 JIC; 02 UdL; 05 MTA-ATK; 08 IPBB; 09 CONICET; 13 SB; 14 IFVCNS; 15 TGEN; 16 KWS-L; 18 ARVALIS
In WP6 the main stream of activity in the first half of the project was to put together the phenotypic data from WP4 and WP5; this was agronomic and physiological data from segregating populations in germplasm Panel 5 and Near Isogenic Lines (NILs) in Panels 2 and 3. This was genetically dissected using the genotypic information from WP2, high density genetic maps (iSelect), medium density (KASP), and association Panels e.g. IPK or ARIHAS panel. The QTL identified were validated and multiple environments and GxE quantified. Certain QTL were prioritised for further genetic dissection to provide the most diagnostic markers possible for SMEs and large breeding companies to test in realistic breeding scenarios. This marks the transition to the second half of the project in which validation methods were fine-tuned together with new marker assays. The ADAPTAWHEAT programme had to align with various stages of breeding programmes at CIMMYT, KWS, RAGT, Limagrain, Semillas Batle, and Selegen to encapsulate a whole breeding cycle into a four-year project. As well as molecular markers, the project was able to help breeders integrate physiological trait selection methods into their in-house programmes. All of this was backed up with training and support.
Highlights of the WP include the identification of independent QTL controlling different stages of stem and inflorescence development in Buster x Charger population giving breeders new tools to control the duration of the early reproductive phase and, hopefully, seed number. We also cloned genes for heading date like EPS-D1, Ppd-B2, and TOE-1 giving excellent molecular markers for selection. These MAS strategies were well validated. They underlie some of the QTL studies in such detail in WP4 and WP5. Putting this information together allows new breeding models to be assembled in which a complex of desirable traits for a particular target environment can be hypothesised, and for future environments aided by the models of WP7. Then, using ADAPTAWHEAT germplasm, molecular markers, and phenotypic selection methods new cultivars can be assembled from these parts.
1.3.7 WP7: QUANTITATIVE VALUE OF DEVELOPMENT ALLELES ACROSS AND THROUGH TIME
Start month 24; End month 48
WP7 leader: 06 UCPH; other WP7 participants: 05 MTA-ATK; 18 ARVALIS; 20 RRes
In WP7 we established the methodology that linked the genetic data assays and crop physiologists’ simulation exercises of crop phenotypes, resulting in a model ready-to-use that captures the actual values of particular alleles. Using representative concentration pathway (RCP) scenarios we were able to model desirable ideotypes for wheat including phenology components. Different modelling methods based on AFRC wheat and Sirius were used. With Sirius, for example, high-yielding wheat ideotypes were designed using the wheat model for climate scenarios based on projections from the Coupled Model Intercomparison Project Phase 5 (CMIP5) multimodel ensemble downscaled by the LARS-WG weather generator. A subsets global climate models (GCMs) were integrated. Climate sensitivity index was used to select a subset of GCMs which preserves the range of uncertainty found in CMIP5. We were able to assess uncertainty in cultivar parameters of high yielding wheat ideotypes resulting from the uncertainty in the CMIP5 ensemble by conducting fewer simulation experiments. Ideotypes were optimised using the Sirius wheat simulation model for future climates in Europe, sampling uncertainty in GCMs, emission scenarios, time periods and European locations with contrasting climates. Two contrasting GCMs were selected for the analysis, “hot” HadGEM2-ES and “cool” GISS-E2-R-CC. Despite large uncertainty in future climate projections, we were able to identify target traits for wheat improvement which may assist breeding for high-yielding wheat cultivars with increased yield stability.
In parallel the use of AFRCWHEAT allowed the incorporation of gene based models. Decision trees were developed and then used to summarize and standardize the information from the genetic assays into model inputs for Ppd1 (top or A) and Vrn1 (bottom or B). AFRCWHEAT2 model was modified to capture the allelic variations of wheat varieties grown in the experimental trials of the project and their effects on heading dates. We facilitated the use of the Bayesian Calibration technique by forcing all parameters vary between 0 and 1 which is convenient to define the limits of their prior probability distributions. The availability of field experimental data from the growth in ADAPTAWHEAT of Near Isogenic Lines (NILs) with several degrees of photoperiod sensitivity (panel 3) provided the opportunity to explore connection between combinations of Ppd-1 alleles and the model parameters to stablish, for the first the time, an equation relating them. In this way we were able to develop the first models that exploited 40 days of heading date variation achieved by the modulation of a single homoeologous gene series- namely Ppd-1. This work show that the genetic variation necessary to produce the ideotypes predicted by the Sirius model is present. We have demonstrated how high-yielding wheat ideotypes were designed using the Sirius wheat model for climate scenarios and how these ideotypes can be constructed. New CMIP5-based scenarios were used in designing wheat ideotypes optimised for future climate conditions in Europe. The in silico design of the aforementioned wheat ideotypes was optimised for future climates in Europe, sampling uncertainty in GCMs, emission scenarios, time periods and European locations with contrasting climates. Despite large uncertainty in future climate projections, we were able to identify target traits for wheat improvement which may assist breeding for high-yielding wheat cultivars with increased yield stability.
Potential Impact:
1.4.1 WP1 ESTABLISHMENT OF GENETIC PANEL
WP1 has set a high bar for standards of germplasm handling, curation, quality control, alignment for international project of this nature. Even in cases where germplasm came to the project as background, the assembly of all materials into panels associated with ADAPTAWHEAT molecular and phenotypic data, together with entry into breeding pedigrees has had the highly synergistic effect of giving new power and significance to these resources. The various panels have become part of the ADAPTAWHEAT vocabulary to the extent that they will outlive the project. Furthermore, the continuity of these materials with series of NILs in the same genetic background and NAM populations adds another level of broader impact for future gene discovery. Altogether the NAM population that ADAPTAWHEAT germplasm is part of comprises 12000 lines and the NIL library 200 lines. These are experimentally unmanageable as a whole. Future project will sample well balanced subsets. The ADAPTAWHEAT portion of these subsets will capture all of the major WP outputs (genes, alleles, traits, model training lines) from ADAPTAWHEAT and so physically carry the best of ADAPTAWHEAT into future projects and breeding programmes. Indeed, some ADAPTAWHEAT germplasm does comprise high value pre-breeding germplasm. The arrangement of the panels and their continued maintenance gives maximum access to breeders to request these prebreeding resources.
1.4.2 WP2: GENOTYPING BREEDING & EXPERIMENTAL MATERIAL
The genetic maps produced by the WP were essential for the identification of ADAPTAWHEAT QTLs. Consequently, the breeding impact of these QTLs explored in the WP6 impact statement are dependent on these activities. Extra power is achieved from WP6 because the project stayed aligned with the state-of-the-art in marker development so that the majority of the work was conducted using genotyping platforms that are the common currency of commercial breeding marker assisted selection and varietal characterisation, namely, iSelect and KASP SNP platforms. This has the result that, in many cases, no new development of ‘breeder friendly’ markers are necessary as the most diagnostic SNP haplotype is simply selected from the genetic map. This signature is then used as a search motif within proprietary germplasm sets. Thus, the transition from ADAPTAWHEAT research activities to application in breeding is seamless. In cases where genes were cloned we have been able to provide diagnostic assays based on the putative functional polymorphism for each QTL. These have utility within any breeding pedigree and represent the gold standard in marker assisted selection. Up to now, very few wheat phenology genes have reached this level of MAS tractability, even less which have relevance in commercial breeding, for example Ppd-1 and Vrn-1. WP2 has delivered a major impact for the project by doubling the length of that list. In addition, we have followed through with detailed protocols, interpretation guides, biological context notes, and trouble-shooting assistance. In summary, the major impact of the WP is its facilitation of QTL discovery and the development of highly efficacious molecular markers for those QTL.
1.4.3 WP3: MOLECULAR PHENOTYPING
The new qPCR assays and knowledge of their expression in a standard set of materials makes a major contribution to the field of phenology in wheat. They are tools that will be widely used by other experts in the field and contribute the body of knowledge for these important traits. They provide insights into the points of contact between ADAPTAWHEAT QTLs and the broader regulatory network for the molecular control of flowering studied by these groups. This opens the possibility of breeding and agronomic intervention to manipulate these traits in the same, favourable, direction as the ADAPTAWHEAT QTL alleles. The gene based models described align very well with the outputs of WP7, allowing researchers to design virtual genotypes adapted to target environments and assembling the gene combinations that are most likely to deliver the required trait combinations. The new understanding of ambient temperature response has added a new element to the way breeding by design might be conducted in terms of adaptation. Up to now the focus has been on vernalizing temperatures, photoperiod, and earliness per se. Here we have shown that EPS may often reflect differential ambient temperature response or even vernalisation response (e.g. from Charger) that models had assumed was fully satisfied. The impact is new models that account for these factors and consequently make more accurate predictions. This will result in better prediction of phenology changes in response to climate change and better choice of tools that can be used (genes, alleles, germplasm, markers, agronomy, and breeding strategies) to mitigate any negative effects and exploit any adaptive opportunities.
1.4.4 WP4: PHYSIOLOGICAL PHENOTYPING
The potential impact of the knowledge produced in WP4 is related to the use of physiological components of time to heading for breeders to be able to fine-tune adaptation with more detail knowledge on which phases will be more affected than others due to the manipulation (depending the allele in question) and to what degree these changes in duration of particular phases would bring about parallel changes in the number of structures being initiated in those phases. The socio-economic impact would be produced more indirectly through the impact that the new cultivars selected using this knowledge will have in productivity/sustainability.
There was a direct dissemination within the consortium to the breeding programs involved (both SME and large companies, which were also running some of the experiments) and to the wider community there are a number of papers being written currently that will be expectedly published during the present year. The nature of the research done, mostly under field conditions, required several growing seasons be carried out before conclusions could be solid enough to warrant publication in the best journals. That is the unique reason why papers on this WP (as well as in WP5) could be published during the experimental period of the project.
1.4.5 WP5: AGRONOMIC PHENOTYPING
Regarding the potential impact and dissemination activities, WP5 is very much like WP4. The main impact is related to the use of lines possessing improved productivity by breeders to further increase yields. The socio-economic impact would be produced more indirectly through the impact that the new cultivars selected using this knowledge will have in productivity/sustainability.
There was a direct dissemination within the consortium to the breeding programs involved (both SME and large companies, which were also running some of the experiments) and to the wider community there are a number of papers being written currently that will be expectedly published during the present year. The nature of the research done, mostly under field conditions, require several growing seasons be carried out before conclusions could be solid enough to warrant publication in the best journals. That is the unique reason why papers on this WP (as well as in WP4) could be published during the experimental period of the project.
1.4.6 WP6: INTEGRATION ACROSS LEVELS OF ORGANISATION
The main impact of WP6 efforts are new breeding tools. The ability of European breeders to breed by design, to assemble trait constellations that result in better adapted wheat for current environments and for future environments is greatly enhanced by the outputs of this WP. This will include response to photoperiod, so for example, the manipulation of FT3 which controls response to short days allows us to produce varieties which are widely adapted using the functional allele or, by stacking the loss of function alleles, to adapt for late frosts so that only vernalisation releases winter varieties for floral induction and SD induction does not occur. For breeding, these approaches will make a major contribution to the recognition that molecular biology makes a vital contribution to modern wheat breeding. It shows that empirical selection alone cannot equip the breeder with all the tools that allow him or her to produce better adapted varieties. For earliness per se ADAPTAWHEAT enjoyed the progress of showing that ELF3 was the causative gene underlying EPS-D1. Multi environment trialling showed that this locus is an excellent target for selection on heading date as our precise Near Isogenic Line analysis showed that it does not carry a yield penalty when early or late, even if the population trend for heading date does. So, the WP gives breeders the most parsimonious route to achieve their goals for a single phenotype with minimal negative trade off (or even benefits) for the other traits they care about. The WP has also demonstrated the limitations of molecular markers and the need for applying physiological selection methods. The take home message from WP6, with high impact for future breeding efforts, is that an eclectic approach to selection is most productive.
As these selection methods are incorporated into real commercial and public breeding programmes we fully expect that they will help to produce better adapted and more productive varieties of wheat. These will help to secure the long term sustainability and profitability of European farming. The interaction of non EU countries such as Australia, Argentina, Kazakhstan, and the CGIAR centre CIMMYT maximise the ‘roll out’ of ADAPTAWHEAT WP6 technologies for wheat, our most global crop. Overall the impacts on farming will help to ensure food security. It provides as exemplar for the way a project can make incremental contributions to breeding which are highly strategic and likely to be adopted in the short term to secure the availability and affordability of wheat for consumers as well as profitability for farmers.
1.4.7 WP7: QUANTITATIVE VALUE OF DEVELOPMENT ALLELES ACROSS AND THROUGH TIME
The mathematical modelling work of WP7 adds a long term dimension to the potential impact and achievements of ADAPTAWHEAT. By taking the best possible information on climate change scenarios we have been able to show what ideotypes of bread wheat are desirable for future climate scenarios. Drawing on the genetic, physiological, and agronomic outputs of the other research WPs we know what variation is available to achieve these ideotypes but also lay down challenges to the wheat community to scope variation, selection tools, and other breeding methodologies that will equip European and global farmers with the best adapted and high performing varieties possible. At the academic level our impact is to show that we have achieved a deeper penetration into the hierarchy of complexity for key breeding traits. We show that continued research and investment in this area will yield models for which in silico wheat crops and their corresponding predicted environments, together will all the variability and uncertainty associated with each can be usefully imagined in platforms such as AFRCWHEAT, SIRIUS, and APSIM to give an informed roadmap of the most productive directions for crop breeding over the next fifty years.
1.4.8 WP9: TRAINING, DISSEMINATION & TECHNOLOGY TRANSFER
The principal objective of WP9 was the provision of training to project partners and the internal and external dissemination of information arising from the project’s activities. This objective remained consistent throughout the project, to assist in the delivery of predictive wheat breeding for phenology traits through the acquisition of new knowledge and methodologies, new molecular markers, and wheat breeding selection protocols.
The ADAPTAWHEAT consortium went a very long way to achieving the objectives of WP9; the open-ended nature of many of the its objectives, some reaching into the post-project period make it difficult to ‘complete’ many of them. Nevertheless, the consortium is satisfied by the outcomes from the WP as the work has addressed the intentions behind the objectives. The project has been active both in training and dissemination during this final reporting period – lists of publications and events are included in section 2 – to ensure the legacy of the project post-closure. The project consortium has taken a pragmatic view on how knowledge and resources generated by the project will be utilised in the future; as such the foreground generated by the ADAPTAWHEAT project has little commercial value but they can be used to support future programmes of development by researchers and breeders alike, both within the public sector and the commercial world, which will ultimately lead to novel varieties of wheat adapted to local environments. This process has already started with the knowledge and materials developed by ADAPTAWHEAT fuelling other successor projects, e.g. successful bids to the recent International Wheat Yield Partnership call.
List of Websites:
The project’s public website can be found at:
http://www.jic.ac.uk/ADAPTAWHEAT/index.htm
The relevant contact details are:
Coordinator: Dr Simon Griffiths
Organisation: John Innes Centre
Tel: +44 1603 450611
Fax: +44 1603 450000
E-mail: simon.griffiths@jic.ac.uk
ADAPTAWHEAT was established to enhance the knowledge and resources available to wheat breeders to enable them to breed new wheat varieties that were be better adapted to the changing growing environments of Europe and beyond, both now and in the future, and to ensure produce higher yields through the manipulation of floral development. Nine related workpackages, seven science-based and two focused on management, training and dissemination, were set-up achieve this objective.
At project completion, ADAPTAWHEAT has achieved this key objective by advancing the state-of-the-art in relation to knowledge and resources for the wheat breeding community. Key to the success of the project was the development of precise germplasm for gene discovery, achieved by producing five novel bi-parental populations from existing and novel material, and single gene-effect analysis to study how genes control sensitivity to day length, vernalisation requirement & other types of flowering control. The latter was possible through the use of Near Isogenic Lines, which are well characterized wheat varieties with single chromosomal segments replaced with a segment from another genotype. Because of the high value of this germplasm in future studies of the agronomic implications of allelic variation at these important genes, ADAPTAWHEAT has ensured that stocks of this novel germplasm have been properly documented and placed in secure storage.
ADAPTAWHEAT has exploited advanced genotyping technologies to characterize germplasm, discover genes & develop new molecular markers for breeding. Very high throughput genotyping has produce new genetic maps, and from which completely novel new molecular markers have been developed that significantly advance in the state-of-art for wheat breeding, and importantly have been validated by ADAPTAWHEAT partners, including the commercial partners.
ADAPTAWHEAT also developed a comprehensive information database from its physiology and agronomy phenotyping studies characterising how different genetic and environmental traits affect the modulation of floral initiation, heading date, and maturity in the panels of wheat germplasm available to it. Associated with this, the molecular phenotyping studies developed a gene expression toolbox that allowed observed variation, including environmental interactions, to be connected to specific markers of gene expression so that the environmental perception and flowering initiation pathways that mediate specific plant responses could be identified. Finally, efforts to develop models that predict the behaviour of wheat germplasm under specific conditions have led to the establishment of methodologies and that serves as a bridge between genetic and physiological understanding of phenotype behaviour, and a ready-to-use model that captures the actual values of particular alleles.
At its inception it was anticipated that ADAPTAWHEAT would not necessarily generate commercially beneficial material itself but develop much-needed knowledge and resources to underpin future development programmes by researchers and breeders alike. The presence of SMEs and several major breeding companies ensured the project had a long-term commercial focus. Evidence indicates this goal has achieved; proof has been the number of novel research and development programmes that have arisen that depend on ADAPTAWHEAT resources, for example, a successful research bid to the International Wheat Yield Programme (IWYP) by several consortium members. Other avenues national and international avenues are being explored too. Reviewers acknowledged that all ADAPTAWHEAT partners will benefit from this project and will use final results to progress their own research. Ultimately, the long-term legacy of the ADAPTAWHEAT project will be in ensuring the use of its outputs in wheat breeding programmes to lead to varieties better adapted to local environments. By their nature such breeding programmes are long-term ventures; ADAPTAWHEAT has provided a solid foundation for success.
Project Context and Objectives:
Introduction
Wheat is one of the major staple crops worldwide with approximately one-sixth of the total global arable land given over to its cultivation. Wheat is highly adaptable, grown on every continent except Antarctica. Wheat’s food qualities are equally important: its grains have higher protein content than maize and rice, two other major staple foods, and the crop provides approximately 20% of the total calorific needs. Wheat is particularly important within the European Union, where it is the most widely grown arable crop, being grown on over 25 million hectares. Wheat production in the EU is c. 150 million tonnes, while that of maize and barley is around 60 million tonnes each.
Wheat’s importance to the global economy cannot be underestimated with demand expected to increase by 60% to satisfy anticipated global demand. To meet this demand breeding efforts need to be more directed to designing wheat varieties better adapted to local situations. By better understanding the genetic and physiological determinants of wheat development which ultimately underpin wheat yield it will be possible to translate this understanding into more efficient breeding strategies.
Such strategies are urgently needed for two reasons. First, predicted changes in climate are a major risk to the existing levels and stability of crop production. The most deleterious effect of climate change is an increase in temperature as this has a major influence on wheat phenology. In addition, climate change will lead to more severe abiotic stresses, drought and heat, and the increased occurrence of extreme weather events, which in turn will affect wheat development. For example, the most detrimental climatic effects – water and heat stress – on wheat occur during ‘grain filling’ late season. The ability to manipulate flowering time in such instances would allow for greater yields. Generally, the ability to exploit the genetics and physiology of wheat development will be critical for the predictive breeding of climate ready wheat.
The second reason for urgency is the global demand for wheat demand is predicted to exceed the current pace of yield increase through breeding. It is important that yield increase return to levels that characterised the Green Revolution. This need is compounded by the issue of sustainability: future gains in wheat production must be achieved principally through breeding for yield potential and tolerance to stresses; with only minor, if any, increases in the use of water or fertilizer.
Identifying genetic and physiological tools to aid yield improvement has been a major goal of crop science research. Accurate genetic manipulation of developmental traits has the potential to provide higher rates of gain in yield. Wheat lends itself well to this ambition: due to its complex phenological control, wheat has the genetic potential to be adapted to any conditions projected for future EU, and global, climate change, thus improving yields. However, as previous work has shown, there is a lag between the discovery of new traits, the characterization of alleles/allelic combinations needed to deliver them and the deployment of this variation through breeding programmes.
Although there has been much work on wheat phenology, there has been limited research connecting these observations to the physiological and genetic mechanisms that control these differences. For example, there is some empirical knowledge of genotypic differences in heading time, and its value for adaptation, but the physiological and genetic bases of these differences are largely not known. Similarly, the major effects of temperature and photoperiod on time to heading are well described but the effects on the physiological determinants of this process are far less described and not fully understood. Another example is earliness in wheat, which is well-documented but where the physiological determinants of these differences in time to heading beyond photoperiod and vernalisation sensitivities have been relatively ignored.
The ADAPTAWHEAT Project
The purpose of the ADAPTAWHEAT project was bridge the gap between the wealth of information on wheat phenology and paucity of knowledge of the physiological and genetic mechanisms that control these differences. The focus of the project was on the physiological and genetic mechanisms for the control of flowering time and its phenological partitioning in the crop. By generating knowledge and producing resources, including valuable wheat germplasm and predictive models to explore wheat performance under future growing conditions, the project would make a significant contribution to improving wheat adaptation and yield. This in turn would address short-term Global/European food security issues and establish a platform for long-term genetic gain and maintenance breeding.
Why flowering?
Flowering time in wheat is a vital determinant in yield performance as it is key for both crop adaptation and yield potential; it regulates adaptation because it strongly influences life-cycle duration. Crop growth and partitioning around flowering is critical for grain production in wheat. Here, it has been established that yield is far more sensitive to changes in growth during the relatively short period from approximately three weeks before flowering to a few days after than during any other period. This critical period coincides with the period when stems and spikes (spikes are central stem on which grow the tightly packed rows of wheat flowers) grow at the highest rate, resulting in a strong competition for resources between them. Crop yield is particularly sensitive to changes in resource accumulation and allocation in that period. For example, there is a positive relationship between the number of grains that wheat has at maturity and the dry matter allocated into spikes at anthesis (the flowering period of a plant, from the opening of the flower bud), therefore lengthen the duration of the stem elongation phase when spike dry matter before anthesis is accumulated, would bring about direct improvements in yield potential.
The most critical feature of agronomic adaptation is to match, as closely as possible, crop growth demands with suitable environmental conditions, whilst avoiding stresses that affect yield, for example, frost during reproductive periods. In other words, it is a matter of adjusting the developmental window within the crop cycle to match the best possible environmental window in which the crop may grow. The likelihood of hotter drier summers predicted in current climate-change models indicates further adaptation is required if grain quality and quantity is to be maintained, and increased.
Project Objectives
The ADAPTAWHEAT partners looked to test a set of physiological hypotheses, and identify prospective genetic factors and valuable germplasm, for flowering time and its phenological partitioning, in wheat. These hypotheses were related to adaptation and crop performance. Specific objectives were to:
• assemble appropriate wheat germplasm panels (near isogenic lines, segregating populations, collections for association mapping, and specific varieties/breeding lines) representing known genetic variation for developmental traits of interest
• phenotype wheat germplasm panels for agronomic traits that determined adaptation and yield potential.
• characterise the developmental traits that determine flowering time in the genetic material of the germplasm panels (e.g. timing to onset of different phenophases and processes, dynamics of leaf number initiation and phyllochron, dynamics of floret development and grain set)
• relate phenotyped traits across levels of organisation, integrating genotypic and phenotypic information in QTL identification, and testing potentially useful QTLs in realistic breeding conditions
• deliver genetic markers for the physiological components of complex phenological traits, which will allow breeders to build flowering/phenology crop types by marker assisted selection of gene combinations.
• provide a gene expression tool box which will allow the observed variation, including environmental interactions, to be connected to specific markers of gene expression and show through which environmental perception and flowering time pathways genotypic responses are mediated
• explore the actual value of particular developmental alleles (affecting either adaptation or yield potential) across a wide range of growing conditions throughout Europe, to test response to different climates and prepare for climate change and changes in climate variability
To meet the overall objective, the project consortium was composed of a group of leading wheat researchers specialising in molecular biology, genetics, physiology, agronomy, and breeding, to provide the required balance of whole plant physiology and genetic expertise. ADAPTAWHEAT had nine interrelated work packages, focussing on: germplasm development and management; genotyping; molecular analysis; detailed physiological analysis; the agronomic assessment of key traits; integration of data for gene discovery and data validation; high level of modelling; training and dissemination and project management. The consortium was highly collaborative, with many partners involved in multiple work packages.
As a consequence of this effort over the four-year duration of the project, ADAPTAWHEAT generated a broad range of genetic and genomic resources necessary to quantify the agronomic implications of specific variation in crop development. The consortium demonstrated how this variation could be altered to allow the wheat crop to be well-adapted to the growing season, avoiding stresses such as frost, drought, and heat while maximising reproductive growth in good conditions so that the maximum quantity of grain can be produced from the resources available. ADAPTAWHEAT has produced the knowledge, genetic markers, and precisely defined types of wheat required by breeders as tools to fine tune adaptation and performance in this key crop.
Key to the project’s success was data exchange between partners, achieved through regular communication, e.g. the project’s wiki, and meetings at all levels of the project, and through dissemination of information to diverse stakeholders beyond the project including plant breeders, potential scientific collaborators and policy makers. Through these connections the legacy of ADAPTAWHEAT is assured in the future.
Project Results:
There were seven scientific workpackages in the ADAPTAWHEAT project (WP1 – WP7). The main results and outputs from each of these are reported below:
1.3.1 WP1: ESTABLISHMENT OF GENETIC PANEL
Start month 01; End month 36
WP1 leader: 01 JIC; Other WP1 Participants; 05 MTA-ATK; 07 CRI; 08 IPBB; 10 CIMMYT INT; 11 CSIRO; 13 SB; 14 IFVCNS; 19 SELGEN
WP1 was the anchor of the ADAPTAWHEAT project. All other WPs were based on experimentation and the analysis of the germplasm panels selected in WP1. These included varietal panels, Near Isogenic Lines (NILs), segregating populations, association panels, and validation sets. The multiplication, distribution, and quality control of these materials are not research outputs in themselves but they underpinned the success of the other WPs. In addition to background germplasm such as the DH populations Avalon x Cadenza, Buster x Charger, Weebill x Bacanora, WP1 also initiated the production of completely new germplasm resources based on the outputs of other ADAPTAWHEAT WPs. So, in the case segregating populations this includes Paragon x Pamayat Azivo, Paragon x Garcia, Paragon x Magnif 41, Toborzo x Verbunkos.
In addition, some populations that were planned for development under the support of other projects (e.g. UK WGIN project) could be done on a larger scale with better aligned resources as a result of ADAPTAWHEAT. The best example of this is Paragon x Garcia. It should be noted that the UK elite spring wheat Paragon is a common parent in many of these populations. In addition, many of the NILs used in ADAPTAWHEAT are in the Paragon background. Moreover, aligned projects such as UK WISP, and the IWYP phenology project led by Dr Simon Griffiths (JIC) use Paragon as a parent in crosses with CIMMYT germplasm. Together these segregating populations for a nested association mapping (NAM) panel. Similarly, the NILs in a Paragon background provide a benchmark from which the first assessment of breeding value for new traits and trait combinations can be assessed on a level playing field. The intensive efforts of WP1 to distribute these materials, going through the hurdles of phyto-sanitory certification, homogeneity testing and cross site referencing, set up a strong base for the project and its continued influence on others. In addition to segregating populations, new NILs were developed in the WP. These are mostly in the genetic background of Paragon. They include genes for winter habit, capturing an allelic series of increasing vernalization requirement in the former spring wheat but also new QTL such as the EPS effects identifies on chromosome 4A.
As the project progressed and panels were received and multiplied, the focus of the WP moved more to the development of germplasm aspects of validation strategies. This included innovative approaches such as self-fertilising RILs that were heterozygous in QTL regions for testing of single QTL effects. We also focussed on specific association panels for marker validation. Most notably the CIMMYT wheat association mapping initiative (WAMI) panel was established as a highly effective means of showing the value of molecular markers from WP2 in materials far removed from the original genetic dissection populations. JIC made a commitment to maintain all of these materials within its seed store. In addition, the information relating to the germplasm development (crossing books, pedigrees etc.) is maintained.
1.3.2 WP2: GENOTYPING BREEDING & EXPERIMENTAL MATERIAL
Start month 07; End month 48
WP2 leader: 15 TGEN; Other WP2 participants: 1 JIC; 05 MTA-ATK; 08 IPBB; 14 IFVCNS; 19 SELGEN
This WP developed and deployed molecular markers for the ADAPTAWHEAT project. Early in the project the main activity was the generation of genetic linkage maps. The genotying load was carried by the use of the iSelect 90K array, KBiosciece KASP technology, simple sequence repeats and the iSelect 15K array. New maps were produced for Buster x Charger, Weebill x Bacanora, Paragon x Pamayat azivo, Paragon x Glasgow, Toborzo x Verbunkos, Paragon x Magnif 41. These fed into WP6 where QTL analysis was carried out. As QTLs were identified and prioritised, WP2 developed the best possible molecular markers. This included intense position cloning efforts. Successfully isolating EPS-D1, PpdB2, and PpdB3, we developed new diagnostic SNP markers for ELF-3, FT3, and TOE-1 respectively. The importance of the FT gene family led us to develop Taqman copy number variant (CNV) assays for the entire family as well as for ELF3 and GI. The effort on CNV is in recognition of the fact that many of the functional polymorphisms we and others cloned emerged as large scale structural rearrangements involving deletion or various levels of duplication. This approach was fruitful as subsequent Panel screens using the Taqman assays showed CNV association with phenology traits.
In addition to the extensive mapping efforts of ADAPTAWHEAT a large amount of marker assisted selection was conducted. This included the selection of foreground segments in backcross programmes for genes such as EPS4A and Hereward, Claire, and Malacca alleles of VrnA1 into Paragon. In some cases, such as 4A, we also employed background selection using 4 KASP marker per chromosome in BC1 populations >500 to maximise the genomic contribution of the recurrent parent at an early stage of backcrossing. These experiments represent a major commitment of time and resource designed to expedite the programme as a whole.
In addition to backcross selection we self-fertilised heterozygotes from RILs. After phenotyping of single plants in WP6 the individual samples were taken and genotypes to identify the two homozygous and singe heterozygous classes so that single marker regression could be carried out, completing this phase of validation.
A major output of the WP was a booklet (available from project wiki) which described in detail how the various molecular markers produced by the project can be used in research and breeding.
Finally, the WP carried out checks. Random samples from field plots were genotyped using most discriminative KASP marker sets to ensure that packaging or drilling errors had not compromised multi-site experimental analysis.
1.3.3 WP3: MOLECULAR PHENOTYPING
Start month 01; End month 48
WP3 leader: 03 INRA; Other WP3 participants: 01 JIC; 05 MTA-ATK; 11 CSIRO
Initially, WP3 pursued the measurement of developmental gene expression in response to environmental variables. The hope was to extend the state-of-the-art to allow researchers to infer a conceptual framework for ecophysiological models that took explicitly into account gene expression. In pursuit of these goals the WP moved many aspect of phenology transcriptomics forward. This included new qPCR assays for key genes such as Vrn-1, Vrn-2, Vrn-3/FT1, FT2, ELF3, GI, Ppd-1, and HSP70. The data generated helps in our qualitative models of the mechanism by which the QTL we identified are likely to be mediating their effect through the genetic networks controlling phenology. Consequently, we could see that EPS-D1 earliness is associated with reduced ELF-3 expression and increased GI. We saw that the very long vernalisation requirement of the variety Charger was tied in with a resurgence of Vrn2 expression when plants were moved from vernalizing conditions to ambient temperature prematurely. There are many high impact stories such as this. However, half way through the project it was clear that our ambitious goal of predicting phenology outcomes on the basis of expression data was not going to deliver. As a result, for deliverables relating to this objective, a change in direction was agreed with our project coordinator to shift focus to gene based modelling using our QTL data.
In the gene based modelling work we produced new models with the framework of APSIM where the mean prediction error ranged from 2.6 to 4.3 days depending on the environment which constitutes a very reliable estimation of heading date.
The WP made significant steps forward in our understanding of the ambient temperature response of wheat. We showed that ambient temperature as an environmental factor and in association with photoperiod may regulate plant development at various consecutive levels of developmental phases, such as the later stages of floret differentiation and via the process of intensive stem elongation.
The different regulation mechanisms of early apex development and the later plant developmental phases is also underlined by the photoperiod and ambient temperature dependent association patterns of VRN1, VRN3 gene expressions and the apex and plant developmental phases. It is clear from our findings that VRN2 gene has a significant role in determining plant development even after vernalization, which is more pronounced when longer photoperiods occur together with higher temperatures. In these cases, a transient high VRN2 expression peak appearing between 100-200 GDD is significantly associated with the ability of the initiation and the procession of the intensive stem elongation. More generally, we used Panel 1 and 4 screens to show that a wide range of variation for ambient temperature response exists in bread, beyond the passive increase in developmental rate that is often assumed. In collaboration with WP1, this WP has initiated the production of a number of new resources to further dissect these interesting responses.
1.3.4 WP4: PHYSIOLOGICAL PHENOTYPING
Start month 01; End month 48
WP4 leader: 02 UdL; Other WP4 participants: 03 INRA; 04 IPK; 05 MTA-ATK; 07 CRI; 08 IPBB; 09 CONICET; 10 CIMMYT INT; 11 CSIRO; 13 SB; 14 IFVCNS; 18 ARVALIS
The focal aim of WP4 was to characterise developmental traits determining flowering time in the genetic material of the panels studied. Through a number of experiments carried out in a large number of locations and through at least two growing seasons we characterised flowering time and its components in all the genetic panels of the project.
Modern cultivars currently grown in many different countries participating in the project (across three continents) always exhibited variation in phenology, including not only variation in time to anthesis but also in the partitioning of this period into phases occurring before or after the onset of stem elongation. Then, in most of the 21 experiments that were conducted within ADAPTAWHEAT evaluating levels of variation within modern and well adapted cultivars we uncovered that cultivars not varying in flowering time could exhibit differences in partition of that overall duration into component phenophases. Thus, in some genotypes the time until anthesis included a longer phase from sowing to stem elongation whilst in others it included longer phases from stem elongation to anthesis.
Eps alleles affected phenology only marginally, as expected from the literature: in most cases these genes are extremely useful only for fine-tuning phenology. The late alleles normally delayed heading and did so through lengthening both the period from sowing to the onset of stem elongation and the period from then to heading when the Eps gene considered was that of chromosome 3A, whilst the Eps gene in chromosome 1D seemed to have affected time to heading mainly through lengthening the duration of the stem elongation- heading phase. Plastochron, the reciprocal of primordia initiation rate, of leaf primordia was not clearly affected by Eps alleles but that for spikelet primordia tended to be shorter in lines with the early alleles. Therefore, the action of the early alleles in shortening the developmental periods did not result in differences in spikelet number as the spikelet initiation rate was faster in the lines with early alleles resulting in compensation in the number of spikelets per spike. Similarly, there were not clear effects of the Eps alleles on the dynamics of floret development and the rates of generation/degeration of floret primordia were similar for early- and late-NILs. Consequently, there were no significant differences in the number of fertile florets along the spike. Thus these alleles may be used to finetune time to flowering without producing potential penalties in reproductive sink strength.
Photoperiod insensitivity alleles naturally advanced flowering time but in ADAPTAWHEAT we evidenced that even though both phases were responsive the insensitivity alleles in general shortened the duration of the stem elongation-heading phase more than the sowing-stem elongation phase. Besides, the effect of the dosage of insensitivity alleles tended to be stronger on the stem elongation-heading phase than in the preceding period from sowing to the onset of stem elongation. The effects of the insensitivity alleles on the duration of the vegetative phase reduced final leaf number and in some cases also reduced phyllochron. In addition, there was a clear reduction in the number of spikelets initiated in NILs with insensitivity alleles, despite that the rate of spikelet initiation tended to be accelerated by the introgression of Ppd-insensitivity alleles. However, these Ppd alleles did not seem to have affected fertility of the four most proximal florets, even though the time for floret development which was reduced affecting the fertility of the fifth floret. Then, the wild type with strongest sensitivity had a longer floret initiation phase and even though it initiated almost the same number of floret primordia than the NILs with insensitivity alleles, floret mortality tended to be lower.
In terms of quantifying the responses to changes in day-length of the different NILs, the wild type of the panel (with the three ppd sensitivity alleles) showed that the sensitivity to photoperiod was evident during the stem elongation phase, with a direct response to photoperiod manipulated exclusively during this period (when final leaf number had been determined long time before). However, in experiments in which day-length was manipulated throughout the whole period from sowing to anthesis the vegetative phase was more responsive than the reproductive phases, which were responsive and the responsiveness was positively related to the number of sensitivity alleles. These results from experiments in which the different NILs were maintained under contrasting day-length treatments during either the whole period from sowing to anthesis or only from terminal spikelet onwards, were confirmed in experiments with reciprocal transfers between short and long photoperiod conditions. When analysing together the different experiments in which photoperiod was manipulated we concluded that Ppd-1 modified only the photoperiod sensitivity, whilst the threshold photoperiod for saturating the response and the minimum duration of developmental processes at these saturating day-lengths (intrinsic earliness) were largely unaffected.
1.3.5 WP5: AGRONOMIC PHENOTYPING
Start month 06; End month 48
WP5 leader 05 MTA-ATK; Other WP5 participants: 01 JIC; 02 UdL; 07 CRI; 08 IPBB; 09 CONICET; 10 CIMMYT INT; 12 RAGT 2n; 13 SB; 14 IFVCNS; 16 KWS-L; 17 LVH UK; 18 ARVALIS; 19 SELGEN
The main objective of WP5 was to phenotype in the genetic panel of the project the agronomic traits determining adaptation and yield under the diverse field grown conditions. These panels included (i) mapping populations, (ii) collections of wheat varieties, and (iii) near isogenic lines for which we determined yield and its main determinants in a number of field experiments carried out across different countries and growing seasons.
The population of Avalon x Cadenza exhibited variation in time to heading including transgressive segregation, with the exception of the experiments carried out in Kazakhstan, (where only the half of the population with spring habit was grown over spring and summer). In general, yield varied much more than phenology and variation in crop productivity was consistently transgressive. Therefore, in most of the experimental conditions (locations x growing seasons) there were lines out-yielding the highest-yielding parent, which might be considered part of the elite material for crossing in breeding programs focused in those locations. In general, though not always, yield tended to increase with shorter time to heading. Interestingly the relationship seemed stronger if the reduction in time to heading were due to accelerated development in earlier phases, suggesting that lengthening stem elongation phase, even when reducing total time to heading might be beneficial for yield. Variations in yield were strongly related to those in the number of grains per unit land area while they were rather independent of variations in the average weight of the grains.
The population of Toborzó x Verbunkos (both winter wheats, segregating for Ppd-D1) also varied in time to heading, due to variations in both sowing-onset of stem elongation and duration of the stem elongation phase, being the latter more relevant than the former for establishing the genotypic differences.
Regarding the collections of modern cultivars from each location involved in ADAPTAWHEAT WP5, even though all genotypes were modern, well adapted cultivars in each location, there was clear variation in agronomic performance. In general, for this panel there was not a consistent relationship between yield and phenology (when analysing genotypic effects). There was not relationship between yield and the duration of any of the phenophases either. Most differences between cultivars in yield were related to similar differences in the number of grains per unit land area. The average weight of the grains also varied and was also related to yield but the relationship was not only weaker than that with grain number but also less consistent across cases analysed. Thus, across all the ranges of breeding programs analyzed differences in grain number was chiefly responsible for those in yield among modern cultivars with grain weight occasionally contributing as well. The genotypic differences in grain number seemed reflecting differences in fruiting efficiency more than in spike dry weight at anthesis. There was a trend for compensations between fruiting efficiency and spike dry weight at anthesis but not a very close negative relationship, suggesting that when focusing on physiological bases of grain number generation within elite germplasm fruiting efficiency would be more relevant.
Considering the NILs for Ppd alleles, the introgression of insensitivity alleles tended to increased yield, without a clear pattern of in which genome the insensitivity affected more. Overall the experiments conducted; there was a negative relationship between yield and duration of sowing-heading (then, overall environments, the higher number of insensitivity alleles the stronger the reduction in heading time and the higher the yield). Most effects of Ppd alleles on yield were mediated through effects of these alleles on both the number of grains per unit land area and thousand grain weight. Therefore, carrying photoperiod insensitivity allele/alleles increased yield in most environments, in line with the effects it had on both the number of grains/m2 and their average weight; but there was not a clear and consistent ranking between the different insensitivity alleles on the magnitude of their effects on yield and its determinants.
Considering the NILs for Eps alleles introgressed in chromosome 1D and 3A with AxC background out-yielded in general (but not always) lines with Eps early alleles; whilst in all the NIL pairs with Eps late alleles introgressed in chromosome 1D with SxR background, lines with the Eps early alleles yielded more than those with Eps late alleles. Thus, averaging across all the isolines the introgression of Eps early alleles reduced yield respect to Eps late alleles in the chromosome 1D for both backgrounds (but more noticeably in SxR), whilst the effect was opposite with the introgression of Eps early alleles in chromosome 3A. In all cases, changes in yield produced by the introgression of Eps late alleles were much more strongly related to those in grain number than in thousand grain number.
1.3.6 WP6: INTEGRATION ACROSS LEVELS OF ORGANISATION
Start month 13; End month 48
WP6 leader: 10 CIMMYT INT; other WP6 participants: 01 JIC; 02 UdL; 05 MTA-ATK; 08 IPBB; 09 CONICET; 13 SB; 14 IFVCNS; 15 TGEN; 16 KWS-L; 18 ARVALIS
In WP6 the main stream of activity in the first half of the project was to put together the phenotypic data from WP4 and WP5; this was agronomic and physiological data from segregating populations in germplasm Panel 5 and Near Isogenic Lines (NILs) in Panels 2 and 3. This was genetically dissected using the genotypic information from WP2, high density genetic maps (iSelect), medium density (KASP), and association Panels e.g. IPK or ARIHAS panel. The QTL identified were validated and multiple environments and GxE quantified. Certain QTL were prioritised for further genetic dissection to provide the most diagnostic markers possible for SMEs and large breeding companies to test in realistic breeding scenarios. This marks the transition to the second half of the project in which validation methods were fine-tuned together with new marker assays. The ADAPTAWHEAT programme had to align with various stages of breeding programmes at CIMMYT, KWS, RAGT, Limagrain, Semillas Batle, and Selegen to encapsulate a whole breeding cycle into a four-year project. As well as molecular markers, the project was able to help breeders integrate physiological trait selection methods into their in-house programmes. All of this was backed up with training and support.
Highlights of the WP include the identification of independent QTL controlling different stages of stem and inflorescence development in Buster x Charger population giving breeders new tools to control the duration of the early reproductive phase and, hopefully, seed number. We also cloned genes for heading date like EPS-D1, Ppd-B2, and TOE-1 giving excellent molecular markers for selection. These MAS strategies were well validated. They underlie some of the QTL studies in such detail in WP4 and WP5. Putting this information together allows new breeding models to be assembled in which a complex of desirable traits for a particular target environment can be hypothesised, and for future environments aided by the models of WP7. Then, using ADAPTAWHEAT germplasm, molecular markers, and phenotypic selection methods new cultivars can be assembled from these parts.
1.3.7 WP7: QUANTITATIVE VALUE OF DEVELOPMENT ALLELES ACROSS AND THROUGH TIME
Start month 24; End month 48
WP7 leader: 06 UCPH; other WP7 participants: 05 MTA-ATK; 18 ARVALIS; 20 RRes
In WP7 we established the methodology that linked the genetic data assays and crop physiologists’ simulation exercises of crop phenotypes, resulting in a model ready-to-use that captures the actual values of particular alleles. Using representative concentration pathway (RCP) scenarios we were able to model desirable ideotypes for wheat including phenology components. Different modelling methods based on AFRC wheat and Sirius were used. With Sirius, for example, high-yielding wheat ideotypes were designed using the wheat model for climate scenarios based on projections from the Coupled Model Intercomparison Project Phase 5 (CMIP5) multimodel ensemble downscaled by the LARS-WG weather generator. A subsets global climate models (GCMs) were integrated. Climate sensitivity index was used to select a subset of GCMs which preserves the range of uncertainty found in CMIP5. We were able to assess uncertainty in cultivar parameters of high yielding wheat ideotypes resulting from the uncertainty in the CMIP5 ensemble by conducting fewer simulation experiments. Ideotypes were optimised using the Sirius wheat simulation model for future climates in Europe, sampling uncertainty in GCMs, emission scenarios, time periods and European locations with contrasting climates. Two contrasting GCMs were selected for the analysis, “hot” HadGEM2-ES and “cool” GISS-E2-R-CC. Despite large uncertainty in future climate projections, we were able to identify target traits for wheat improvement which may assist breeding for high-yielding wheat cultivars with increased yield stability.
In parallel the use of AFRCWHEAT allowed the incorporation of gene based models. Decision trees were developed and then used to summarize and standardize the information from the genetic assays into model inputs for Ppd1 (top or A) and Vrn1 (bottom or B). AFRCWHEAT2 model was modified to capture the allelic variations of wheat varieties grown in the experimental trials of the project and their effects on heading dates. We facilitated the use of the Bayesian Calibration technique by forcing all parameters vary between 0 and 1 which is convenient to define the limits of their prior probability distributions. The availability of field experimental data from the growth in ADAPTAWHEAT of Near Isogenic Lines (NILs) with several degrees of photoperiod sensitivity (panel 3) provided the opportunity to explore connection between combinations of Ppd-1 alleles and the model parameters to stablish, for the first the time, an equation relating them. In this way we were able to develop the first models that exploited 40 days of heading date variation achieved by the modulation of a single homoeologous gene series- namely Ppd-1. This work show that the genetic variation necessary to produce the ideotypes predicted by the Sirius model is present. We have demonstrated how high-yielding wheat ideotypes were designed using the Sirius wheat model for climate scenarios and how these ideotypes can be constructed. New CMIP5-based scenarios were used in designing wheat ideotypes optimised for future climate conditions in Europe. The in silico design of the aforementioned wheat ideotypes was optimised for future climates in Europe, sampling uncertainty in GCMs, emission scenarios, time periods and European locations with contrasting climates. Despite large uncertainty in future climate projections, we were able to identify target traits for wheat improvement which may assist breeding for high-yielding wheat cultivars with increased yield stability.
Potential Impact:
1.4.1 WP1 ESTABLISHMENT OF GENETIC PANEL
WP1 has set a high bar for standards of germplasm handling, curation, quality control, alignment for international project of this nature. Even in cases where germplasm came to the project as background, the assembly of all materials into panels associated with ADAPTAWHEAT molecular and phenotypic data, together with entry into breeding pedigrees has had the highly synergistic effect of giving new power and significance to these resources. The various panels have become part of the ADAPTAWHEAT vocabulary to the extent that they will outlive the project. Furthermore, the continuity of these materials with series of NILs in the same genetic background and NAM populations adds another level of broader impact for future gene discovery. Altogether the NAM population that ADAPTAWHEAT germplasm is part of comprises 12000 lines and the NIL library 200 lines. These are experimentally unmanageable as a whole. Future project will sample well balanced subsets. The ADAPTAWHEAT portion of these subsets will capture all of the major WP outputs (genes, alleles, traits, model training lines) from ADAPTAWHEAT and so physically carry the best of ADAPTAWHEAT into future projects and breeding programmes. Indeed, some ADAPTAWHEAT germplasm does comprise high value pre-breeding germplasm. The arrangement of the panels and their continued maintenance gives maximum access to breeders to request these prebreeding resources.
1.4.2 WP2: GENOTYPING BREEDING & EXPERIMENTAL MATERIAL
The genetic maps produced by the WP were essential for the identification of ADAPTAWHEAT QTLs. Consequently, the breeding impact of these QTLs explored in the WP6 impact statement are dependent on these activities. Extra power is achieved from WP6 because the project stayed aligned with the state-of-the-art in marker development so that the majority of the work was conducted using genotyping platforms that are the common currency of commercial breeding marker assisted selection and varietal characterisation, namely, iSelect and KASP SNP platforms. This has the result that, in many cases, no new development of ‘breeder friendly’ markers are necessary as the most diagnostic SNP haplotype is simply selected from the genetic map. This signature is then used as a search motif within proprietary germplasm sets. Thus, the transition from ADAPTAWHEAT research activities to application in breeding is seamless. In cases where genes were cloned we have been able to provide diagnostic assays based on the putative functional polymorphism for each QTL. These have utility within any breeding pedigree and represent the gold standard in marker assisted selection. Up to now, very few wheat phenology genes have reached this level of MAS tractability, even less which have relevance in commercial breeding, for example Ppd-1 and Vrn-1. WP2 has delivered a major impact for the project by doubling the length of that list. In addition, we have followed through with detailed protocols, interpretation guides, biological context notes, and trouble-shooting assistance. In summary, the major impact of the WP is its facilitation of QTL discovery and the development of highly efficacious molecular markers for those QTL.
1.4.3 WP3: MOLECULAR PHENOTYPING
The new qPCR assays and knowledge of their expression in a standard set of materials makes a major contribution to the field of phenology in wheat. They are tools that will be widely used by other experts in the field and contribute the body of knowledge for these important traits. They provide insights into the points of contact between ADAPTAWHEAT QTLs and the broader regulatory network for the molecular control of flowering studied by these groups. This opens the possibility of breeding and agronomic intervention to manipulate these traits in the same, favourable, direction as the ADAPTAWHEAT QTL alleles. The gene based models described align very well with the outputs of WP7, allowing researchers to design virtual genotypes adapted to target environments and assembling the gene combinations that are most likely to deliver the required trait combinations. The new understanding of ambient temperature response has added a new element to the way breeding by design might be conducted in terms of adaptation. Up to now the focus has been on vernalizing temperatures, photoperiod, and earliness per se. Here we have shown that EPS may often reflect differential ambient temperature response or even vernalisation response (e.g. from Charger) that models had assumed was fully satisfied. The impact is new models that account for these factors and consequently make more accurate predictions. This will result in better prediction of phenology changes in response to climate change and better choice of tools that can be used (genes, alleles, germplasm, markers, agronomy, and breeding strategies) to mitigate any negative effects and exploit any adaptive opportunities.
1.4.4 WP4: PHYSIOLOGICAL PHENOTYPING
The potential impact of the knowledge produced in WP4 is related to the use of physiological components of time to heading for breeders to be able to fine-tune adaptation with more detail knowledge on which phases will be more affected than others due to the manipulation (depending the allele in question) and to what degree these changes in duration of particular phases would bring about parallel changes in the number of structures being initiated in those phases. The socio-economic impact would be produced more indirectly through the impact that the new cultivars selected using this knowledge will have in productivity/sustainability.
There was a direct dissemination within the consortium to the breeding programs involved (both SME and large companies, which were also running some of the experiments) and to the wider community there are a number of papers being written currently that will be expectedly published during the present year. The nature of the research done, mostly under field conditions, required several growing seasons be carried out before conclusions could be solid enough to warrant publication in the best journals. That is the unique reason why papers on this WP (as well as in WP5) could be published during the experimental period of the project.
1.4.5 WP5: AGRONOMIC PHENOTYPING
Regarding the potential impact and dissemination activities, WP5 is very much like WP4. The main impact is related to the use of lines possessing improved productivity by breeders to further increase yields. The socio-economic impact would be produced more indirectly through the impact that the new cultivars selected using this knowledge will have in productivity/sustainability.
There was a direct dissemination within the consortium to the breeding programs involved (both SME and large companies, which were also running some of the experiments) and to the wider community there are a number of papers being written currently that will be expectedly published during the present year. The nature of the research done, mostly under field conditions, require several growing seasons be carried out before conclusions could be solid enough to warrant publication in the best journals. That is the unique reason why papers on this WP (as well as in WP4) could be published during the experimental period of the project.
1.4.6 WP6: INTEGRATION ACROSS LEVELS OF ORGANISATION
The main impact of WP6 efforts are new breeding tools. The ability of European breeders to breed by design, to assemble trait constellations that result in better adapted wheat for current environments and for future environments is greatly enhanced by the outputs of this WP. This will include response to photoperiod, so for example, the manipulation of FT3 which controls response to short days allows us to produce varieties which are widely adapted using the functional allele or, by stacking the loss of function alleles, to adapt for late frosts so that only vernalisation releases winter varieties for floral induction and SD induction does not occur. For breeding, these approaches will make a major contribution to the recognition that molecular biology makes a vital contribution to modern wheat breeding. It shows that empirical selection alone cannot equip the breeder with all the tools that allow him or her to produce better adapted varieties. For earliness per se ADAPTAWHEAT enjoyed the progress of showing that ELF3 was the causative gene underlying EPS-D1. Multi environment trialling showed that this locus is an excellent target for selection on heading date as our precise Near Isogenic Line analysis showed that it does not carry a yield penalty when early or late, even if the population trend for heading date does. So, the WP gives breeders the most parsimonious route to achieve their goals for a single phenotype with minimal negative trade off (or even benefits) for the other traits they care about. The WP has also demonstrated the limitations of molecular markers and the need for applying physiological selection methods. The take home message from WP6, with high impact for future breeding efforts, is that an eclectic approach to selection is most productive.
As these selection methods are incorporated into real commercial and public breeding programmes we fully expect that they will help to produce better adapted and more productive varieties of wheat. These will help to secure the long term sustainability and profitability of European farming. The interaction of non EU countries such as Australia, Argentina, Kazakhstan, and the CGIAR centre CIMMYT maximise the ‘roll out’ of ADAPTAWHEAT WP6 technologies for wheat, our most global crop. Overall the impacts on farming will help to ensure food security. It provides as exemplar for the way a project can make incremental contributions to breeding which are highly strategic and likely to be adopted in the short term to secure the availability and affordability of wheat for consumers as well as profitability for farmers.
1.4.7 WP7: QUANTITATIVE VALUE OF DEVELOPMENT ALLELES ACROSS AND THROUGH TIME
The mathematical modelling work of WP7 adds a long term dimension to the potential impact and achievements of ADAPTAWHEAT. By taking the best possible information on climate change scenarios we have been able to show what ideotypes of bread wheat are desirable for future climate scenarios. Drawing on the genetic, physiological, and agronomic outputs of the other research WPs we know what variation is available to achieve these ideotypes but also lay down challenges to the wheat community to scope variation, selection tools, and other breeding methodologies that will equip European and global farmers with the best adapted and high performing varieties possible. At the academic level our impact is to show that we have achieved a deeper penetration into the hierarchy of complexity for key breeding traits. We show that continued research and investment in this area will yield models for which in silico wheat crops and their corresponding predicted environments, together will all the variability and uncertainty associated with each can be usefully imagined in platforms such as AFRCWHEAT, SIRIUS, and APSIM to give an informed roadmap of the most productive directions for crop breeding over the next fifty years.
1.4.8 WP9: TRAINING, DISSEMINATION & TECHNOLOGY TRANSFER
The principal objective of WP9 was the provision of training to project partners and the internal and external dissemination of information arising from the project’s activities. This objective remained consistent throughout the project, to assist in the delivery of predictive wheat breeding for phenology traits through the acquisition of new knowledge and methodologies, new molecular markers, and wheat breeding selection protocols.
The ADAPTAWHEAT consortium went a very long way to achieving the objectives of WP9; the open-ended nature of many of the its objectives, some reaching into the post-project period make it difficult to ‘complete’ many of them. Nevertheless, the consortium is satisfied by the outcomes from the WP as the work has addressed the intentions behind the objectives. The project has been active both in training and dissemination during this final reporting period – lists of publications and events are included in section 2 – to ensure the legacy of the project post-closure. The project consortium has taken a pragmatic view on how knowledge and resources generated by the project will be utilised in the future; as such the foreground generated by the ADAPTAWHEAT project has little commercial value but they can be used to support future programmes of development by researchers and breeders alike, both within the public sector and the commercial world, which will ultimately lead to novel varieties of wheat adapted to local environments. This process has already started with the knowledge and materials developed by ADAPTAWHEAT fuelling other successor projects, e.g. successful bids to the recent International Wheat Yield Partnership call.
List of Websites:
The project’s public website can be found at:
http://www.jic.ac.uk/ADAPTAWHEAT/index.htm
The relevant contact details are:
Coordinator: Dr Simon Griffiths
Organisation: John Innes Centre
Tel: +44 1603 450611
Fax: +44 1603 450000
E-mail: simon.griffiths@jic.ac.uk