Evidence Project Final Report

Fruit samples were made available to users for a number of apple day events as well as for research. In general the material was provided from outside of the project resources via the packhouse but where relevant these were made available under the sMTA (and are also listed above). Samples were supplied to the following organisations, in addition to a number of private individuals:

Research visitors were received from the following organisations/institutes in addition to a number of private individual researchers:

Brogdale Collections hosted numerous further professional visitors with a less specific research focus, including: filming teams from the BBC (Countryfile, Great British Food Revival, Apples: British to the Core, The Hairy Bikers) and BBC Weather; Beechgrove Garden and James Wong. Other professional visitors hosted by Brogdale Collections included:

• 
  Morphological verification was continued, using the descriptors established by the previous curators (in line with both the European Co-operative Program on Plant Genetic Resources [ECPGR] and the International Union for the Protection of New Varieties of Plants [UPOV]). 20 gooseberry accessions, 11 blackcurrant accessions and 13 cherry accessions were compared to published descriptions to complete the morphological verification of each of these collections. 176 pear accessions were either verified against published descriptions or described (where published descriptions were not available) and an additional 50 pear accessions were checked morphologically as part of the verification of the new collection (specifically where accessions, such as sports, were expected to be indistinguishable by genetic markers [as detailed in Appendix 3]);
  
• 
  Additional morphological characters were scored for fruit quality across a subset of 300 accessions of the apple collection as part of the FP7 FruitBreedomics project and further ECPGR standardised fruit quality characters were scored across 100 accessions from the cherry collection. Further morphological assessment was carried out in cases where suspected duplicates were identified in the apples and pears (see for example, Defra project GC0139/40) and an assessment of the old bush fruit collection was made such that the old plants in collection NFC 3 could be signed off;
  
• 
  Diversity Arrays Technology (DArT) was utilised to carry out a genomic screen of the apple collection. In line with an amendment to the original work plan, this allowed us to make use of the newly developed DArT technology and allowed us to include the majority of the collection in the analysis (originally planned for a subset of approximately 1,000 accessions). DNA was extracted from all accessions in the apple collection (Milestone 3) and samples were sent to DArT P/L for analysis. The data were produced in three phases, an original data set was produced for a selection of 188 samples (Milestone 7) and these reported data for 1,256 markers; a second data set was produced for the majority of the remainder of the collection, reporting 1,423 markers across 1,886 samples (Milestone 11), and a third phase of analysis reported 1371 markers across 146 samples (Milestone 18). 959 of the markers were consistent across all data sets and of these, 562 were identified in a non-redundant set from additional data kindly supplied by Henk Schouten (pers comm.) The data were brought together and analysed with a view to creating an objective measure of the structure of genetic diversity held in the collection. To summarise findings: we found that the collection contains relatively limited genetic structure when allowing for the known clonal replicates (a finding in keeping with other, more recent analyses of similar collections [FruitBreedomics project, unpublished, Myles et al., 2011, PNAS, 108:3530-3535). Large amounts of admixture could be seen in attempts to cluster accessions and these resulted in a clustering pattern of genotypes which was, for the most part, unable to clearly distinguish distinct clusters (and/or distinct gaps in the diversity of the collection as was originally postulated). This was assumed to be reflective of the outbreeding and heterozygous nature of the crop. Further discussion and detailed analysis is currently being prepared for presentation in a scientific publication. DArT data were also used as an independent replication of the original molecular fingerprinting work (Defra project GC0140) to confirm duplication in the collection. Following publication these data will be made available for further exploitation by users of the collection.
  
• 
  Our initial plan was to score microsatellite (SSR) alleles linked to previously established quantitative trait loci (QTL) across the collections with a view to typing the collection for traits of use in breeding (Milestone 6). It was, however, generally accepted that many of the QTL markers previously established in apple were not robust or accurate enough to be usefully scored across wider germplasm collections (for reasons discussed by Collard and Mackill, 2008, Phil. Trans. R. Soc. B, 363:557-572). Consequently, efforts were focussed on a smaller number of gene linked markers with established linkage to important fruit quality traits: scores were generated for alleles of 1-aminocyclopropane-1-carboxylate synthase (ACS) and 1-aminocyclopropane-1-carboxylate oxidase (ACO), two key enzymes in the ethylene synthesis pathway with genetically characterised alleles associated with high and low ethylene production (Costa et al., 2005, Euphytica, 141:181-190); the Md-PG1 gene for polygalacturonase has also been identified as a key regulator of fruit ripening and associated quality (Wakasa, 2006, Postharvest Biol. Technol., 56:193-198) and a gene linked marker system has also been developed (Broothaerts et al., unpublished, and more recently Nybom et al., 2013, Tree Genetics & Genomes, 9:279-290) and this was also scored across a subset of material (this component of work was delayed in the project and further scoring is expected to be completed in the coming months) [Milestone 32], preliminary results are shown in section 2.2 (below) and these will be developed into a publication before making available via the NFC website.
  
• 
  Genetic fingerprinting by SSR analysis was carried out across the new and old pear collections to sign off the new collection and allow the removal of the old trees. Our initial attempt involved scoring SSR profiles across trees in the new collection before comparing scored data to the previously established dataset (Defra project GC0139/40); markers were used as per the previous project. Complications were found in this approach, caused by the need to adjust datasets to align with the existing scored data, as well as occasional missing alleles. A second approach was therefore established to provide a more robust analysis whereby DNA samples were extracted and analysed together for both the old and new collections (Milestone 22). This allowed a more consistent approach to be taken where alleles needed to be scored subjectively (the previously scored data were then used for further confirmation and to compare in cases of missing data). A detailed report of the process and findings is included in Appendix 3, but in summary: 495 out of 498 accessions were confirmed to be present and correct in at least one tree position in the new collections; in the vast majority (490/495) of cases the labelled tree (tree 1) was found to be correct and in most of these cases this was therefore the only tree analysed; in a small subset of cases the labelled tree was found to be incorrect (often a failed graft that had reverted to the Doyenne du Comice interstock) but the second tree was confirmed as correct; in two cases, both trees were found to be incorrect and, along with a number of the re-tested samples these were grafted for safety and replacement. One accession 1948-185 Summer Bergamot was found to have two failed grafts in the new collection and the old tree contained no further new growth for grafting. Summer Bergamot was therefore lost from the collection during the repropagation. A large number of curational queries were resolved during the process and these are presented in detail in the full report (Appendix 3) [Milestone 26];
  
• 
  Genetic fingerprinting by SSR analysis was also carried out across the Cobnut collection. The NFC 3 Cobnut collection was propagated in 2002 but due to a combination of limited resource, limited morphological descriptors and/or expertise this had not yet been verified. Leaf samples were collected and DNA was extracted from each of the new trees as well as from the labelled tree in the old collection. Markers were selected as per the EU SAFENUT project and are listed in section 2.3 (below). A comparison was made around whether the two new trees were matching and whether they both matched the original tree in NFC 2. A detailed summary is included in Appendix 4, but in summary: data were generated for all apart from one sample (where the reactions failed) in the new collection (94/95 trees [one tree is currently missing in the new collection]) and all 42 samples in the old collection (six accessions in the new collection were added at the time of propagation and were not represented in the old collection); in all but one case, where possible, at least one tree matching the accession in the old collection was found to be present and correct in the new collection (in 40/42 cases all trees analysed were found to match); in one case, the new collection was found to contain a mixed plot with one tree correct and one tree false and in the second case (accession 1978-378 Corylus colurna) all three trees were found to be different. All three Corlyus colurna trees were brought in as seedlings and this variation was therefore not surprising (but was not previously documented, and was consequently noted for the record). One further mixed plot did not have an equivalent in the old collection and will need to be resolved at a later date. The analysis was sufficient to accept the new collection as being present and correct (with a note to replace one missing tree and one false tree); a small number of potential duplicates were identified and these will be the subject of further investigation;
  
• 
  Genetic fingerprinting data were generated for a subset of the cherry collection which had previously been suspected to contain duplicates (Milestone 31). Leaves were collected for 93 samples and DNA was extracted. The microsatellite markers used were as per the ECPGR proposed set (Clarke and Tobutt, 2009 Acta Hort., 814:615-618) with modifications based on recommendations by Frei et al., (J. Hort. Sci. Biotech., 2010, 85(4):277-282) and are listed in section 2.4 (below). Data were not fully analysed but will be utilised to assess the suspect accessions in line with the planned repropagation of the collection, with a view to avoiding the repropagation of known duplicates or mislabelled accessions in the future;
  
• 
  Further, more detailed genetic analysis of the apple collection was planned to be carried out in collaboration with the EU FruitBreedomics project. As part of the EU funded project the Breedomics consortium developed a customised high density genotyping array expected to contain in the region of 480K single nucleotide polymorphism (SNP) markers. As part of a collaboration to expand the scope of this analysis we supplied further samples from the NFC and helped to support their inclusion in the study. The overall aim of this study is to characterise a wide range of germplasm, selected from a number of the key European germplasm collections, with a view to identifying genetic markers for fruit quality through an association genetics approach. This therefore presents a complimentary and additional approach to the analysis of gene linked markers and genetic diversity through DArT. Additionally, this collaborative approach allowed us access to an immensely valuable set of background knowledge and activity. Due to challenges faced by the consortium in developing the genotyping arrays these data will not be available until late 2014.
  
• 
  Collection of flowering data continued throughout the project phase; measurements included first flower (10% of blossom open), full flower (80% of blossom open), end of flowering (90% petal fall) and blossom intensity (estimated on 2-year and older wood at full flower). Flowering time was scored across the previously established set of 90 standard accessions (see section 2.5) in all years. New data were also collected for approximately 50 apple, 26 cherry, 9 plum and 7 pear accessions to generate 10-year means. Flowering time, together with meteorological data have been analysed within BSc student projects at the University. In general it has been found that there has been an advancement in time to first blossom, with flowering across the apple collection occurring 17 days earlier than in the early 1960’s, as a result of an increase in mean spring temperature of approximately 1^oC. Year to year variation in flowering time of individual accessions can be explained largely by yearly variation in accumulated spring temperature. Our analysis has shown that the amount of winter chill at Brogdale has declined by almost a third over the last fifty years. However, there is no evidence, at least in the varieties that have been analysed so far, that there has been any detectable effect of this reduction in winter chill on the pattern of flowering in apple;
  
• 
  Further phenological measurements were taken through a number of student BSc projects at the University. These included attempts to assess the time of ceasing extension growth and leaf fall and were measured over 2-3 seasons on a subset of accessions. It was generally felt that, whilst these may prove to be interesting characters for phenological study, they were labour intensive (requiring multiple repeated measures per tree) and it was beyond the capacity of the curational work to expand these further within the collections. These characters are likely to be included within the climate change trial (although extension may be taken as absolute growth rather than studying the timing in detail);
  
• 
  Further characterisation of the collections was planned to assess the potential effects of a changing climate on accessions within the collection. This relied upon the establishment of a major long term experiment and fell outside of the direct remit of the curational project but, in collaboration with the National Fruit Collections Trust, a trial was planned and the trees of 21 accessions which will form the basis of the long term experiment were planted and are now established (further details are in section 6).
  

Examples of individual allele scores for a subset of accessions (full details will be available in the NFC database):