Plants People Planet. 2021;3:747–760. | 747wileyonlinelibrary.com/journal/ppp3 Received: 12 February 2021  | Revised: 29 April 2021  | Accepted: 6 May 2021 DOI: 10.1002/ppp3.10211 R E S E A R C H A R T I C L E Quantifying apple diversity: A phenomic characterization of Canada’s Apple Biodiversity Collection Sophie Watts1  | Zoë Migicovsky1  | Kendra A. McClure1 | Cindy H. J. Yu1 | Beatrice Amyotte2 | Thomas Baker2 | David Bowlby2 | Karen Burgher- MacLellan2 | Laura Butler1 | Richard Donald1 | Lihua Fan2 | Sherry Fillmore2 | John Flewelling2 | Kyle Gardner1 | Mark Hodges2 | Tim Hughes2 | Vinetha Jagadeesan1 | Naomi Lewis2 | Edward MacDonell1 | Laura MacVicar2 | Michel McElroy1 | Daniel Money1 | Matthew O’Hara2 | Quang Ong1 | Leslie Campbell Palmer2 | Jason Sawler1 | Melinda Vinqvist- Tymchuk2 | HP Vasantha Rupasinghe1 | John M. DeLong2 | Charles F. Forney2 | Jun Song2 | Sean Myles1 This is an open access article under the terms of the Creative Commons Attribution- NonCommercial- NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non- commercial and no modifications or adaptations are made. 1Dalhousie Plant, Food and Environmental Sciences, Faculty of Agriculture, Dalhousie University, Truro, Canada 2Agriculture and Agri- Food Canada, Kentville Research and Development Centre, Kentville, Canada Correspondence Sean Myles, Department of Plant, Food and Environmental Sciences, Faculty of Agriculture, Dalhousie University, Truro, Nova Scotia, Canada. Email: sean.myles@dal.ca Present address Kendra A. McClure, Perennia Food and Agriculture Inc, Kentville, Canada Kyle Gardner, Agriculture and Agri- food Canada, Fredericton Research and Development Centre, Fredericton, Canada Naomi Lewis, Department of Chemistry, Dalhousie University, Halifax, Canada Funding information This research was supported in part by funding from the Canadian Horticultural Council (SM), Agriculture and Agri- Food Canada's AgriScience Program (SM), the Nova Scotia Department of Agriculture (SM), the Nova Scotia Fruit Growers Association (SM), Dalhousie University (SM), the Canada Research Chairs program (SM), the National Sciences and Engineering Research Council of Canada (SM), and A- Base funding (NOI- 1767) Societal Impact Statement A future with a secure and safe food supply requires humanity to preserve and ex- ploit the vast variation available across agricultural plant species. Apples are one of the most widely consumed fruits and provide significant nutritional value worldwide. Here, we characterize key agricultural traits in a diverse collection of apples to pro- vide a foundation for future apple improvement. We show that commercially success- ful apple varieties capture only a small fraction of apple diversity, and demonstrate that significant improvement is possible by tapping into existing genetic diversity. Summary ● Here we present a comprehensive evaluation of apple diversity through phe- notyping of Canada's Apple Biodiversity Collection (ABC) which contains over 1000 apple accessions. ● We assessed, over a 4- year period, more than 20,000 individual apples and quantified variation across 39 phenotypes, including phenology and fruit quality both at harvest and after 3 months of cold storage. ● We observe that apples in the ABC display a wide range of phenotypic varia- tion that may prove useful for future apple improvement. For example, apples can differ by nearly 61- fold in weight, 18- fold in acidity, and 100- fold in phenolic content. We quantified the dramatic changes to apple physiology that occur dur- ing 3 months of cold storage: on average, apples lost 39% of their firmness, 31% of their acidity, and 9% of their weight, but gained 7% in soluble solids. Harvest date, flowering date, and time to ripen were all positively correlated with firm- ness, which suggests that the developmental pathways that drive phenological © 2021 The Authors. Plants, People, Planet published by John Wiley & Sons Ltd on behalf of New Phytologist Foundation. www.wileyonlinelibrary.com/journal/ppp3 https://orcid.org/0000-0001-6857-3903 https://orcid.org/0000-0002-3931-1258 mailto: https://orcid.org/0000-0001-9754-5987 http://creativecommons.org/licenses/by-nc-nd/4.0/ mailto:sean.myles@dal.ca http://crossmark.crossref.org/dialog/?doi=10.1002%2Fppp3.10211&domain=pdf&date_stamp=2021-06-23 1  |  INTRODUCTION The domesticated apple (Malus domestica) is thought to have been cultivated for over 3,000 years (Zohary et al. 2012). Genetic evidence suggests that the main progenitor species of the domesticated apple is Malus sieversii from Central Asia (Velasco et al. 2010); however, sig- nificant gene flow from Malus sylvestris has also been detected (Sun et al. 2020; Duan et al., 2017; Cornille et al. 2012). Red color, reduced acidity, larger fruit, and firmness appear to have been under selec- tion during apple domestication and improvement (Ma et al., 2015; Migicovsky et al., 2021). The apple's widespread geographic distribu- tion and long- standing popularity have resulted in over 10,000 named apple cultivars with a fascinating diversity of phenotypes (Liang et al., 2015). Despite this tremendous diversity, a small number of cultivars make up a significant proportion of production. For example, in 2018, only four cultivars accounted for over 50% of apple production in the USA (“US Apple & Pear Forecast”). In addition, commercial apples have been shown to be closely related to each other genetically (Migicovsky et al., 2021; Noiton & Alspach, 1996). Reliance on a limited number of closely related cultivars means that consumers experience an ex- tremely limited fraction of the apple's genetic diversity. However, the extent of phenotypic diversity captured by the top commercial culti- vars remains to be quantified. Fruit firmness is strongly associated with consumers “likeness” of cultivars and is, therefore, a key target for breeders (Szczesniak 2002; Nybom et al., 2013). In addition, breeding apples that retain their firm- ness after long- term storage have been a key breeding target (Kouassi et al., 2009). Not only is it important for breeders to select for firmness and firmness retention, but it is also crucial to breed for the phenologi- cal traits associated with firmness and firmness retention. This is espe- cially important given that the optimal phenological breeding targets are expected to shift over time due to climate change. Characterizing how fruit texture and phenological traits are associated with each other can enable the development of new cultivars that adapt to cli- mate change and meet consumer preferences. Germplasm collections serve as important reservoirs of genetic diversity for crop improvement. They contain the high levels of di- versity that are essential for identifying valuable phenotypes that can be leveraged to develop improved cultivars (Bramel & Volk, 2019). Extensive germplasm resources are already available for major crops such as maize and rice which can be stored as seeds, but resources for perennial fruit crops are lacking as they are expensive to establish and maintain (Flint- Garcia et al., 2005; Jackson, 1997; Migicovsky & Myles, 2017). Despite the high cost, living germplasm collections are essential for woody perennials which are difficult to conserve in seed banks or tissue culture. There is increasing interest in using germplasm collections to not only preserve biodiversity and exploit valuable phenotypes but also to perform genetic mapping (Migicovsky et al., 2019). The high genetic diversity and sample sizes present in many germplasm collections can result in well- powered genetic mapping studies that identify genetic markers useful for genomics- assisted breeding (Zhu et al., 2008). In addition, germ- plasm collections often contain wild relatives with novel traits that can be introgressed into elite cultivars (Migicovsky & Myles, 2017). There are numerous apple germplasm collections that have been established worldwide (e.g., USA (Gross et al., 2013), China (Gao et al., 2015), New Zealand (Kumar et al., 2010), and Europe (Jung et al., 2020)). Within Canada, there are apple collections located across the country, for example, in Ontario and British Columbia (Hampson et al., 2009; Ward, 1978). Most recently, Canada's Apple Biodiversity Collection (ABC) was established in Nova Scotia, Canada, with over 1,000 accessions that includes trees primarily be- longing to the domesticated apple, M. domestica, and its primary wild ancestor, M. sieversii. The ABC was established as a dual- purpose orchard. The first purpose of the ABC is to preserve and maintain potentially valuable apple genetic and phenotypic diversity. Second, the ABC is specifically designed to enable accurate measurements of phenological and fruit quality traits primarily for the purposes of genetic mapping. The trees in the ABC were grafted at the same time to the same rootstock and planted in duplicate in a randomized block design to control for positional effects in the orchard. The result is an apple population that is ideally suited to accomplish a phenome- wide characterization of apples. Here we present a comprehensive evaluation of Canada's ABC through phenotyping of phenological traits and fruit quality traits both at harvest and after 3 months of cold storage. from Agriculture and Agri- Food Canada (JS). ZM was supported by the National Science Foundation Plant Genome Research Program 1546869. SW was supported by a Vanier Scholarship from the National Sciences and Engineering Research Council of Canada. events throughout the growing season may play a role in determining an apple's texture. Finally, we show that apple breeding has selected for a significant decline in phenolic content over the past 200 years: apple cultivars released after 1940 had a 30% lower median phenolic content than cultivars released before 1940. ● The data and analyses presented here not only provide a comprehensive quanti- fication of the range across, and relationships among diverse apple phenotypes, but they also enable genetic mapping studies that will provide the foundation for future apple improvement via genomics- assisted breeding. K E Y W O R D S apple, fruit quality, germplasm, Malus domestica, Malus sieversii, phenomics, plant breeding |    WATTS eT Al.748 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 2  | MATERIALS AND METHODS 2.1  |  The apple biodiversity collection The Apple Biodiversity Collection (ABC) is located at the Agriculture and Agri- Food Canada (AAFC) Kentville Research and Development Centre in Nova Scotia, Canada (45.071767, −64.480466). The ABC contains 1,119 apple accessions that were grafted to M.9 rootstock in August 2011 and allowed to grow outdoors until November 2012 when they were removed from the orchard and stored in moist sawdust at 2°C until planting. On May 31, 2013, we planted each of the 1,119 accessions in duplicate in a 5- acre orchard that was tile drained and fumigated with Telone® soil fumigant. The trees were spaced 1.5 m within rows and 5 m between rows. The trees were trained to a trellis system with wires at 1.5 and 2.4 meters above the ground. Soil amendments, training, thinning, and pruning were performed to industry standards. The ABC consists of apple accessions from the United States Department of Agriculture (USDA) Plant Genetic Resources Unit apple germplasm collection in Geneva, New York, USA; commercial cultivars from the Nova Scotia Fruit Growers’ Association Cultivar Evaluation Trial; and advanced breeding material from the AAFC Kentville breeding program. The collection contains mostly M. do- mestica accessions including cider, dessert, processing, heritage, and elite cultivars. The orchard also contains 78 accessions of the wild progenitor species M. sieversii from Central Asia. It is possible that pairs of accessions within the ABC may be clonally related in some cases. Here we treat each accession as a unique sample in down- stream analyses and future genetic investigation will reveal the de- gree of clonal relatedness in the collection. The trees in the ABC are not available for propagation as most of the material was imported from the USDA under a section 43 import permit from the Canadian Food Inspection Agency (Permit #P- 2011– 00222) which prohibits the sale or distribution of the germplasm. 2.2  |  Experimental design The ABC was planted in an incomplete randomized block design with a North and South block (Figure S1). Each block is divided into grids that include nine trees within a row across three adjacent rows. Within each grid is a “check tree” of “Ambrosia”, “SweeTango”, or “Honeycrisp”, amounting to 18 check trees of each of these culti- vars across the orchard. These “check trees” are used as standards in the restricted maximum likelihood (REML) model that accounts for phenotypic variation due to position in the orchard. We used the following model to generate phenotype values: This mixed- model accounts for fixed effects of an accession and the random effects of position depending on Block (north/south), north- to- south position within the block (rGrid), east- to- west position within the block (cGrid), and the interactions between these random effects. The measurements resulted in 39 phenotypic variables col- lected from 2014 to 2018, which are summarized in Data S1. 2.3  |  Phenology All phenology traits that were recorded as dates (e.g., harvest date and flowering date) were converted into Julian days. Harvest date was recorded for the 2016 and 2017 harvest seasons. During both seasons, 20 apples were picked randomly from each tree. Trees ready for harvest were flagged at the beginning of the week and harvested over the subsequent days of that week. Due to the diver- sity of the accessions and the variation in ripening time, a variety of methods were used to determine when to harvest. Dropped apples or changes in background skin color were indicators of harvestable trees (Watkins, 2003). In addition, a sample apple was taken from each tree and touched to assess firmness, tasted to assess starch and sweetness, cut in half to check browning of seeds, and sprayed with iodine solution to evaluate starch content (Blanpied & Silsby, 1992). The combination of these indicators was used to determine whether an accession was ready to be harvested. Flowering date was measured as the date when the youngest wood displayed >80% of flowers at the king bloom stage (McClure, 2017). Flowering date was recorded in 2016, with trees being assessed every 3 days during this period. Time to ripen was calculated as the time (in days) between flowering date and harvest date. Precocity was recorded as the year in which a tree first bloomed after establishment in the orchard and converted into a score: a score of 1 corresponded to 2014, a score of 2 corresponded to 2015, a score of 3 corresponded to 2016, and a score of 4 indicated that the tree had not yet bloomed as of 2016. 2.4  |  Fruit quality The fruit quality traits measured included weight (g), firmness (kg/ cm2), acidity (g/L malic acid), soluble solids content (SSC) (°Brix), and juiciness ((weight of juice/total fruit weight) × 100). In 2017, meas- urements were taken from 5 apples, while a sample of 10 apples was used in 2016 whenever possible and a mean measurement for each tree was calculated. Measurements were not recorded for trees with fewer than three representative apples available for as- sessment. In 2016, an automated phenotyping machine was used (the Pimprenelle, Setop Giraud Technologie, France— https://www. setop.eu/fr/produ it/pimpr enelle) to collect fruit quality data. In 2017, measurements were collected manually as described below. Since the measuring techniques differed slightly between 2016 and 2017, we assessed the correlation between years for each trait (Figure S2). Total weight (g) of all the apples in a sample were measured (Mettler- Toledo, MS3002S), and an average weight per apple was calculated by dividing the recorded total weight by the number of apples in the sample. Firmness measurements (kg/cm2) for five phenotype ∼ accession + (1 | Block) + (1 | rGrid) + (1 | cGrid) + (1 | cGrid: rGrid)    |WATTS eT Al. 749 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.setop.eu/fr/produit/pimprenelle https://www.setop.eu/fr/produit/pimprenelle individual apples were recorded using a penetrometer (Guss Fruit Texture Analyser, GS- 14). Using a vegetable peeler, a small section of skin along the side of the apple was removed to allow the pene- trometer to enter the apple flesh. Each apple was then placed on the penetrometer platform so that the piston entered the middle of the apple where the skin had been shaved off. There was no preference for the side of the apple from which the measurement was taken. The firmness measurements were averaged to get a mean value for the tree. To measure acidity in a sample of apples, a quarter of each apple was juiced using a handheld garlic press. The juice from the apples belonging to a single tree was combined to make a composite juice sample. Titratable acidity was measured using the 865 Dosimat Plus (Metrohm). This involved titrating 1 ml of the composite juice sample with 0.1 M of NaOH. Soluble solids content (SSC) was measured on the composite juice sample using the Pocket Refractometer (Atago, PAL- 1). The SSC/acid ratio was calculated by dividing SSC measure- ments by acidity measurements. After harvest, apples that were not phenotyped were placed in cold storage (4°C). After 3 months in storage, these apples were re- moved and placed at room temperature until fruit quality could be measured on the following day using the same methods described above. Thus, apples from each accession were assessed for fruit quality both at harvest and after 3 months of storage. Post- storage measurements were taken from a minimum of 3 and a maximum of 10 apples per accession. During the 2016 harvest, whole fruit samples were collected and ground into a fine powder using liquid nitrogen and stored at −80°C. Frozen ground samples (approximately 0.5 g each) were ex- tracted in duplicate for the quantification of phenols using sonica- tion in an 80% methanol solution in two 2 ml microcentrifuge tubes. A ferric- reducing antioxidant power (FRAP) assay for total antioxi- dant capacity (TAC) was performed to detect the reduction poten- tial of compounds by measuring the intensity of the violet- blue color produced. FRAP reagents were prepared as described previously (Benzie & Strain, 1996; Rupasinghe et al., 2008). The ferric reducing antioxidant powers of apple extracts were reported in micromolar of Trolox equivalents (TE) per gram fresh weight of an apple. The Folin– Ciocalteu assay estimates the total phenolic content (TPC) present in plant foods and was carried out as previously described (Huber & Rupasinghe, 2009; Singleton et al., 1999). TPC of apple extracts was reported in micromolar of gallic acid equivalents (GAE) per gram fresh weight of an apple. TPC and TAC were strongly correlated (ρ = 0.948; p < 1 × 10−15; Figure S3), so we only present TPC in the main body of the manuscript. 2.5  | Apple classification Apple accessions were classified into binary categories based on species (M. domestica vs. M. sieversii), geographic origin (new world vs. old world), and end use (cider vs. dessert) using information re- trieved primarily from the USDA Germplasm Resource Information Network (GRIN; https://www.ars- grin.gov/), as well as online sources. Accessions classified as M. pumila (N = 5) in the GRIN da- tabase were considered M. domestica in our analyses since these terms are synonymous. Accessions classified by GRIN as M. sylvestris (N = 3) or hybrids (N = 5) were excluded from analysis of comparisons between species. All accessions that were listed as dessert, cooking, or eating were classified as “dessert” (N = 804) while only those la- beled as cider were classified as “cider” (N = 104). Accessions labeled as wild (N = 78), crab (N = 5), and rootstock (N = 1) cultivars were ex- cluded from the analysis of comparisons between cider and dessert. For geographic origin, accessions that originated in Europe or Asia were classified as “old world” (N = 450), while those that originated in North America, South America, South Africa, Australia, and New Zealand were classified as “new world” (N = 529). The year of release was also retrieved from the GRIN database or from online sources for 343 accessions that were introduced as named cultivars. 2.6  |  Statistical analysis All data curation, handling, and analysis were performed in R (R Core Team 2018). Each accession was planted in duplicate, but if one of an accession's two trees died or did not produce fruit, then the data came from only a single tree. The phenology and fruit trait measure- ments before storage were adjusted for their location in the orchard by running the REML model (see above) using the “lme4” package (Bates et al., 2015) in R, which resulted in one adjusted mean meas- urement per accession. Phenolic measures (TPC and TAC) were not adjusted since, most often, only a single tree was sampled per ac- cession. Fruit quality measurements taken after storage were also not adjusted for location in the orchard because, for each accession, the fruit from the two duplicate trees was combined before being placed in cold storage. Thus, post- storage measurements are sim- ply the mean value across apples within an accession. To evaluate how fruit quality changed during storage, the difference between the measurements before storage (adjusted) and after storage (un- adjusted) was calculated for each accession. Linear correlations between phenotypes were assessed using Pearson correlations via the “cor.test” function in R. To assess how the most commercially successful cultivars differ from the remain- der of the collection, we compared phenotypes from 9 of the top 10 cultivars produced in the USA (“US Apple & Pear Forecast”) that are found in the ABC to the phenotypes of the remaining ac- cessions in the ABC. The 9 of the top 10 cultivars that are present in the ABC are as follows: “Red Delicious,” “Gala,” “Fuji,” “Granny Smith,” “Honeycrisp,” “Golden Delicious,” “McIntosh,” “Cripps Pink,” and “Empire.” The median, standard error, minimum, and maximum trait measurements were calculated for the “Top 9” and for the other ABC accessions. Mann– Whitney U tests were performed to determine whether measurements of the “top 9” differed significantly from the measurements of the other ABC accessions. The proportion of the ABC’s range in trait measure- ments encompassed by the “top 9” cultivars was calculated by |    WATTS eT Al.750 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.ars-grin.gov/ dividing the range in measurements of the “Top 9” by the range of the entire ABC. Mann- Whitney U tests were also performed to determine whether phenotypic traits differed significantly among Malus species, geographic origin, and end use classifications. Standard error was calculated using “std.error” function from the plotrix package in R (Lemon 2006). The Bonferroni correction was applied to account for multiple testing where appropriate. 3  |  RESULTS Canada's Apple Biodiversity Collection (ABC) aims to capture much of the world's diversity of apples and contains 1,119 accessions from 47 countries, with a strong emphasis on apples from Canada (N = 278) and the USA (N = 234) (Figure 1). The measurements from over 20,000 individual fruit from more than 1,000 unique apple accessions across 2 years are presented in Data S1. Numerous phenotypes were measured in both 2016 and 2017. With the ex- ception of flowering date, juiciness, and precocity, all other phe- notypes presented in the main body of the manuscript are from 2017 due to the larger sample size in that year. The correlations between years for phenotypes ranged from 0 to 0.69, and 13 of 17 of the phenotypes had significant (p < 0.05) between- year correla- tions (Figure S2). The weakest correlations were for the phenotypic changes observed during storage. For example, the weakest cor- relation was for change in weight during storage (R2 = 1.2 × 10−4, p = 1). The strongest correlation between years was for harvest date (R2 = 0.691, p < 1 × 10−15). 3.1  |  Phenology and fruit quality traits The distributions of phenology and fruit quality traits are shown in Figure 2. The distributions of these traits for the year 2016 are provided in Figure S4. The summary statistics for each of the phe- notypes are provided in Data S2. Flowering date spanned 21 days with “Kaz 95 18- 02P- 33,” a M. sieversii accession, flowering first on May 22nd and “Frostproof,” a M. domestica cultivar, flowering last on June 12th. Harvest spanned 65 days; more than three times as long as flowering. The earliest harvested accession was “C.P. Close” on August 12th and the latest harvested accession was “Red Spy” on October 17th. Time to ripen spanned 68 days. “Lunost” had the shortest time to ripen (71 days) and “KAZ 7” had the longest time to ripen (139 days). We observed a 61- fold difference in apple weight among ac- cessions, ranging from 7.57 g to 460 g. The lightest accession was “Uralskoje Nalivnoje” and the heaviest was “Bietigheimer.” Acidity varied by 18- fold: the least acidic (1.41 g/L) accession was “Dunning” and the most acidic (26.35 g/L) was “Barenhecke 3 Klipphausen” a M. sylvestris accession. There was a sevenfold difference in firm- ness among accessions at harvest: the softest accession was “Miron Sacharanij” with a measurement of 1.99 kg/cm2, whereas the firmest accession was “Oekonomierat Echter- meyer” at 14.21 kg/cm2. The least sweet accession was “Julyred” at 7.04 °Brix, and the sweet- est accession, “Golden Delicious,” was more than twice as sweet at 16.4 °Brix. Phenolic content varied by nearly two orders of magni- tude from 0.293 to 27.9 μmol/g. The two accessions with the low- est phenolic content were “Zestar” and the advanced breeding line F IGURE 1 World map showing geographic origins of apple accessions from Canada's Apple Biodiversity Collection Afghanistan (1)Algeria (1) 27 of the 1119 samples did not have any geographic origin information. Zimbabwe (1) India (2) Pakistan (2) Turkmenistan (2) Israel (4) Brazil (4) Tajikistan (1) New Zealand (7) South Africa (6) Russia (50) Uzbekistan (2) Japan (21) Australia (16) China (4) Canada (278) Kazakhstan (76) United States (234) Austria (1) Belgium (12) Bulgaria (2) Czech Republic (7) Denmark (10) Estonia (2) Europe - unknown (7) Finland (3) France (78) Germany (54) Hungary (1) Ireland (5) Italy (3) Latvia (8) Lithuania (1) Macedonia (1) Netherlands (20) Poland (8) Portugal (2) Romania (9) Serbia (1) Spain (9) Sweden (9) Switzerland (13) Ukraine (4) United Kingdom (108) Yugoslavia (2) EUROPE    |WATTS eT Al. 751 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense “S19– 23– 52,” while the cider apple cultivar “Marachal” had the high- est phenolic content. None of the phenotypes of the “top 9” cultivars differed signifi- cantly from the rest of the accessions in the ABC after accounting for multiple tests. The phenotype measurements of the “top 9” cultivars capture between 10.6% and 70.6% of the total variation in the ABC across phenotypes. For phenolic content, the “top 9” cultivars are found close to the median of the ABC and capture only 14.2% of the ABC’s total variation in phenolic content. In contrast, 70.6% of the total variation in soluble solids is captured by the “top 9”, and values for the “top 9” tend to be above the median of the ABC. Precocity ranged from 1 year to over 4 years across all accessions, but all of the "top 9" culti- vars fruited in the first 2 years after establishment in the orchard. 3.2  |  Fruit quality after storage Significant correlations were observed between measurements taken before and after storage for all fruit quality traits (Figure 3, Figure S5). On average, apples lost 31% of their acidity, 39.9% of F IGURE 2 Distribution of apple phenotypes from Canada's Apple Biodiversity Collection. Shaded grey areas represent the range of values observed across the "top 9" cultivars grown in the USA Soluble Solids Content (ºBrix) N=857 Acidity (g/L) N=834 Phenolic Content (μmol/g) N=472 Harvest Date (Julian date) N=866 Weight (g) N=862 Firmness (kg/cm2) N=863 Precocity N=1109 Flowering Date (Julian date) N=1069 Time to Ripen (days) N=834 7.5 10.0 12.5 15.0 5 10 15 20 25 0 10 20 240 260 280 0 100 200 300 400 5 10 1 2 3 4 145 150 155 160 80 100 120 0 10 20 30 40 50 0 25 50 75 100 0 30 60 90 0 100 200 300 0 25 50 75 0 25 50 75 0 100 200 300 0 25 50 75 0 25 50 75 C ou nt |    WATTS eT Al.752 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense F IGURE 3 Fruit quality measurements before and after 3 months of cold storage. Correlations of fruit quality measures taken before and after storage are found in the left column (a, c, e, g). Distributions of the percent change of each trait during storage are shown in the right column (b, d, f, h). The shaded grey areas represent the range occupied by the “top 9” cultivars 0 10 20 0 10 20 Acidity at harvest (g/L) A ci di ty a fte r s to ra ge (g /L ) R2 = 0.75 p < 1x10-15 (a) 0 10 20 30 40 −80 −40 0 40 Change in acidity during storage (%) N=568 C ou nt (b) 5 10 15 20 5 10 15 20 SSC at harvest (ºBrix) SS C a fte r st or ag e ( ºB ri x) R2 = 0.261 p < 1x10-15 (c) 0 10 20 30 40 −50 −25 0 25 50 Change in SSC during storage (%) N=601 C ou nt (d) 5 10 15 5 10 15 Firmness at harvest (kg/cm2) Fi rm ne ss a fte r st or ag e (k g/ cm 2 ) R2 = 0.536 p < 1x10-15 (e) 0 10 20 30 −50 −25 0 Change in firmness during storage (%) N=538 C ou nt (f) 0 100 200 300 400 0 100 200 300 400 Weight at harvest (g) W ei gh t a fte r s to ra ge (g ) R2 = 0.826 p < 1x10-15 (g) 0 10 20 30 40 50 −40 0 40 Change in weight during storage (%) N=536 C ou nt (h)    |WATTS eT Al. 753 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense their firmness, and 9.1% of their weight, but increased in soluble sol- ids by 7.1% during 3 months of cold storage (Data S2). The change in acidity, firmness, soluble solids, and weight for the "top 9" cultivars did not differ significantly from the rest of the accessions. However, the “top 9” cultivars tended to experience a less severe loss of firm- ness (29%) and acidity (25.9%) compared to the rest of the ABC. 3.3  |  Relationships among traits We calculated correlations among all the fruit quality, phenology, and storage traits (Figure 4, Data S3). Correlations for the 2016 phenotype data are provided in Figure S6 and Data S4. Of the 105 pairwise comparisons, 32 were significant after correcting for mul- tiple comparisons (p < .01). Of those significant correlations, 28 were positive and 4 were negative. Harvest date was significantly correlated with weight (R2 = 0.11, p < 1 × 10−15), juiciness (R2 = 0.04, p = 1.85 × 10−4), percent change in SSC (R2 = 0.33, p < 1 × 10−15), and percent change in acidity (R2 = 0.18, p = 1.14 × 10−3). Juiciness and weight at harvest were significantly correlated (R2 = 0.06, p = 2.05 × 10−6). Phenolic content and weight at harvest were negatively correlated (R2 = 0.06, p = 1.1 × 10−4). Acidity and flow- ering date were negatively correlated (R2 = 0.04, p = 4.4 × 10−7). Firmness at harvest was correlated with phenology traits, including flowering date (R2 = 0.0292, p = 5 × 10−7), harvest date (R2 = 0.246, p < 1 × 10−15), and time to ripen (R2 = 0.259, p < 1 × 10−15) (Figure 5). Change in firmness during storage was positively asso- ciated with flowering date (R2 = 0.0263, p = 1.7 × 10−4), harvest date (R2 = 0.0847, p = 5.8 × 10−12), and time to ripen (R2 = 0.0939, p = 7.8 × 10−13) (Figure 5). F IGURE 4 Heat map of pairwise correlations among phenology, fruit quality, and storage traits. Significant correlations (p < .01) after correcting for multiple comparisons are indicated with an asterisk | WATTS eT Al.754 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 3.4  |  Phenotypes according to apple classifications For 9 of the 15 phenotypes, we observed significant differences between apples classified according to species, geographic ori- gin, and/or use (Figure 6). When comparing median values be- tween groups, M. domestica accessions flowered 3 days later (p < 1 × 10−15), ripened 14 days later (p = 1.78 × 10−9), were harvested 15 days later (p = 5.35 × 10−13), were 250% heavier (p < 1 × 10−15), and had 67% less phenolic content (p = 5.63 × 10−3) than M. siev- ersii accessions (Data S5). New world apples flowered 0.76 days earlier (p = 1.39 × 10−4), were 20% heavier (p = 7.45 × 10−3), and had 22% less phenolic content (p = 3.20 × 10−4) than old world apples. Finally, dessert apples flowered 0.75 days earlier (p = 6.36 × 10−6), were 43% heavier (p = 3.59 × 10−9), were 11% softer (p = 1.74 × 10−4), and had 34% less phenolic content than cider apples (p = 1.12 × 10−7). 3.5  |  Phenotypic changes over time Phenolic content was the only phenotype of the 39 tested that showed a significant change over historical time after correct- ing for multiple comparisons: the year of release of named culti- vars was negatively correlated with phenolic content (R2 = 0.12; p = 1.0 × 10−4) (Data S6). Apple cultivars released after 1940 had a median phenolic content of 3.38 μmol/g, which is 30% lower than cultivars released before 1940 (4.86 μmol/g) (Figure 7). 4  | DISCUSSION Humanity's food security, and the long- term sustainability of agricul- ture, relies on the preservation and characterization of germplasm collections that house the genetic diversity required for future crop improvement. Detailed quantification of the phenotypes within these germplasm collections is vital to enabling breeders to more rapidly and efficiently generate improved plant varieties. To this end, our comprehensive apple phenome evaluation involved phenotyp- ing of Canada's Apple Biodiversity Collection (ABC), which contains over 1000 apple accessions. The degree of phenotypic variation in the ABC was substantial for some traits (Figure 2). For example, ap- ples can differ by 61 - fold in weight, 18- fold in acidity, and nearly 100 - fold in phenolic content. Phenological traits, such as harvest date, also varied considerably across the orchard, spanning 66 days. In contrast, flowering took place within a 21- day window. Our findings are in line with observations from diverse apples planted throughout Europe where harvest season was also approximately three times longer than the flowering period (Jung et al., 2020). Cultivars that are prone to early floral development under warm temperatures are at increased risk of damage from spring frosts, and late flowering F IGURE 5 Correlation of firmness at harvest during storage with (a) flowering date, (b) harvest date, and (c) time to ripen 5 10 145 150 155 160 165 F ir m ne ss ( kg /c m 2 ) R2 = 0.0292 p = 5x10-7 (a) 5 10 240 260 280 (b) 5 10 80 100 120 140 (c) −50 −25 0 145 150 155 160 165 Flowering date (Julian date) F ir m ne ss lo ss d ur in g st or ag e (% ) (d) −50 −25 0 240 260 280 Harvest date (Julian date) (e) −50 −25 0 80 100 120 140 Time to ripen (days) (f) R2 = 0.246 p < 1x10-15 R2 = 0.259 p < 1x10-15 R2 = 0.0263 p = 1.7x10-4 R2 = 0.0847 p = 5.8x10-12 R2 = 0.0939 p = 7.8x10-13    |WATTS eT Al. 755 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense may, therefore, be a desirable breeding target (Gottschalk & van Nocker, 2013). Time to ripen is also a critical phenotype for apple producers and we observed a nearly twofold difference in the time to ripen between the cultivar with the shortest (71 days) and long- est (139 days) times to ripen. There is a strong preference among producers for fruit that ripens quickly: the longer fruit hangs on the tree exposed to potentially adverse weather events or pests and dis- ease, the higher the risk of fruit damage and yield loss. Therefore, there is interest in breeding late flowering and fast ripening culti- vars to reduce the risks imposed by adverse environmental events (Gottschalk & van Nocker, 2013; Seeley & Anderson, 2003). Our ob- servations provide not only an overall view of the levels of variation across apple phenotypes but also quantifies phenotypic variation that is informative for breeding improved apple cultivars. Widening the apple breeding pool to include lesser- known cul- tivars, as well as wild relatives such as M. sieversii, can lead to the development of elite cultivars with improved health benefits, tastes, and fruit quality. The “top 9” cultivars sold in the USA did not differ significantly from the entire collection as a whole for any phenotype, suggesting that the most popular apples in the USA are not outliers in terms of their phenotypes compared to the variation found within the ABC. However, there is considerable room for improvement through conventional breeding to move toward more desirable val- ues for numerous phenotypes. For instance, there are apples within F IGURE 7 Apple phenolic content decreased over the past 200 years. Each dot represents a named cultivar. A cultivar's “year of release” refers to its year of commercialization, release to the public, or initial mention in historical records 0 5 10 15 1800 1850 1900 1950 2000 Year of Release Ph en ol ic s ( µm ol /g ) R2 = 0.12 p = 1.0x10-4 F IGURE 6 Comparisons of fruit quality traits among species, geographic origins, and uses. The arrows indicate phenotypes where the difference between groups was significant. The arrow points to the group with the higher value Phenolics ∆ Acidity ∆ Soluble solids ∆ Firmness ∆ Weight Soluble solids/acid Acidity Soluble solids Firmness Weight Juiciness Harvest date Time to ripen Flowering time Precocity Phenotype M. domestica (N=1030) vs M. sieversii (N=78) New world (N=529) vs Old world (N=450) Dessert (N=804) vs Cider (N=104) > 0.05 < 0.05Bonferroni corrected p-value: direction of increase: Phenology Storage Fruit quality domestica sieversii New Old Dessert Cid |    WATTS eT Al.756 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense the ABC with acidity levels, phenolic content, ripening times, and flowering dates that lie beyond the values observed within the “top 9” cultivars. With the exception of soluble solids, where “Golden Delicious” was the sweetest accession in the ABC, none of the “top 9” cultivars had extreme values for fruit quality and phenology measurements. In addition, the range of trait measurements for the “top 9” spanned a median of 37% of the variation displayed by the broader collection. “Marachal” had the highest phenolic content, a value five times higher than “Granny Smith,” which had the high- est phenolic content among the “top 9” cultivars. In addition, the “top 9” only captured 10.6% of the range in flowering date observed across the entire collection, and the “top 9” cultivars took a median of 13 days longer to ripen than the average apple in the orchard. This indicates that the most popular apples in the USA capture only a fraction of the phenotypic diversity in the ABC, and that most phenotypic variation remains largely unavailable to consumers and producers. In addition, recent genetic analysis of the USDA apple collection showed that over half the collection is related through a series of first- degree relationships due to the prolific use of the most popular cultivars in breeding programs (Migicovsky et al., 2021). A review of the US apple genetic resources indicates that, despite the relatively robust germplasm resources available, the apple industry remains vulnerable due to the limited number of cultivars employed in commercial production (Volk et al., 2015). Although apples did not experience a diversity- reducing domestication bottleneck like many crops, if breeding programs continue to make use of only a small fraction of apple diversity, it could lead to reductions in genetic and phenotypic diversity in the future (Gross et al., 2014; Noiton & Alspach, 1996). Exploiting the diversity of M. domestica and M. sieversii is not only useful to avoid narrowing the genetic base, but it is also critical for introducing novel traits into new cultivars that can meet apple growers’ challenges and the demands of consumers. Storability is a critically important target for breeders because consumers generally eat apples that have been stored for months. Accessions within the ABC that performed well during storage with minimal losses of firmness, acidity, and weight may serve as use- ful breeding material to develop cultivars with superior storability. When examining how acidity, soluble solids, firmness, and weight changed over 3 months of storage, we found that the direction of change in each trait was not universal across cultivars. For example, while firmness loss is expected during storage, a small number of cultivars appeared to gain firmness during storage (Figure 3). These rare observations can be explained by sampling noise due to the de- structive nature of our assays: the apples from a particular accession measured at harvest were destroyed, so the apples measured after storage from that accession were not the same apples as those mea- sured at harvest. Despite this sampling noise, the overall direction of phenotypic change during storage was consistent with previous studies: firmness, acidity, and weight decrease during storage, while SSC increases (Guerra et al., 2010). For example, apples generally lost about a third (31%) of their acidity during storage, which is con- sistent with previous work (Kouassi et al., 2009; Verma et al., 2019). Apple flavor is strongly influenced by acidity, most predominantly malic acid content, and the loss of acidity during storage leads to de- creased fruit quality (Harker et al., 2003; Zhang et al., 2010). One of the other key determinants of fruit quality both at harvest and after months of storage is fruit firmness (Guerra et al., 2010; Johnston et al., 2010). Despite the apple's superior ability to retain its firmness during storage compared to many other horticultural crops, there is still tremendous opportunity to improve apples by breeding novel cultivars that soften very little during extended periods of storage. We observed significant softening during storage: on average, ap- ples lost 39% of their firmness over 3 months of storage at 4°C. However, there were 14 accessions that experienced ≤10% firmness loss (Figure 3). Our results suggest that apple breeders’ use of culti- vars with superior acid and firmness retention during storage could result in novel cultivars with superior eating quality after storage. Post- storage fruit quality is key to an apple cultivar's commercial success and its biggest threat is loss of firmness due to softening. Here, we focused on the relationship between firmness and phe- nological traits, such as flowering date, time to ripen, and harvest date. Previous work has found that late flowering and late harvested apples tend to be firmer (Migicovsky et al., 2016; Nybom et al., 2013; Oraguzie et al., 2004), and we confirmed this relationship here. We found that later harvested apples are firmer at harvest and soften less over 3 months of storage than early harvested apples (Figures 4, 5). We determined that, on average, for every week of on- tree ripening, an apple experiences a 25% increase in firmness at har- vest. It has been observed that early harvested apples have larger cells with greater intercellular space and, therefore, weaker tissue, which may account for their softer texture both at harvest and after storage (Khan and Vincent 1990; Johnston et al. 2002). In addition, early harvested cultivars produce higher levels of ethylene than late- harvested cultivars, which likely contributes to their softer texture after storage (Watkins, 2003). Firmness at harvest is also a strong predictor of firmness after storage (R2 = 0.54, p < 1 × 10−15), indicat- ing that the texture of an apple after storage is largely dictated by how firm it was when picked. The time it takes an apple to ripen is the period of time be- tween its flowering date and its harvest date. While long ripen- ing periods may appear beneficial because they result in firm fruit that soften less, they also result in increased exposure to disease pressures and potential adverse weather events. In addition, late flowering cultivars are desirable because they are less prone to frost damage (Mehlenbacher & Voordeckers, 1991). While growers may desire cultivars with short ripening periods, selection for rapid ripening will likely result in apples with poor texture attributes that make them undesirable to consumers. Apple cultivars with short ripening periods (i.e., late- flowering and early- harvested cultivars) would certainly be selected for by breeders if they had desirable texture. It is possible that a few of the outlier cultivars we analyzed here that ripen relatively quickly but also possess desirable firm- ness attributes (e.g., “Tayeshnoe” and “Ivan”) may hold the key to decoupling ripening time from firmness. Genomic techniques that enable the relationship between ripening time and firmness to be broken could deliver novel apple cultivars in the future that ripen    |WATTS eT Al. 757 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense quickly, are harvested early, and are firm both at harvest and after storage. Comparisons between domesticated crops and their wild ances- tors can reveal which traits were targeted during domestication and improvement (Meyer et al., 2012). We found that the domesticated accessions (M. domestica) flower later, ripen slower, are harvested later, are heavier, less acidic, higher in soluble solids, and lower in phenolics than accessions belonging to the apple's primary wild progenitor, M. sieversii. These findings are consistent with studies of apple domestication showing that larger fruit, low acidity, and high SSC content appear to have been selected for during apple domes- tication and improvement (Duan et al., 2017; Ma et al., 2015; Miller & Gross, 2011). Examples of fruit enlargement over the course of domestication can be seen in other crops as well, such as tomato, peach, and pear (Frary et al., 2000; Li, Liu, et al., 2019; Li, Cao, et al., 2019). Our results suggest that a decrease in phenolics and later flowering and harvest times may have also been targets of selection during apple improvement. Consistent with previous work (Migicovsky et al., 2016), we determined that New World cultivars are larger and have lower phenolic content than Old World cultivars. These differences may be due to a difference in breeding targets between regions. It is well known that consumer preferences for fruit quality can be in- fluenced by cultural factors like geography, as seen in peach, for instance (Li, Cao, et al., 2019), and this may explain the divergent apple phenotypes between different geographic regions observed here. While most of the apples we evaluated are intended for fresh eating and are, thus, classified as “dessert apples”, about 10% of the ABC consists of “cider apples” that have attributes that make them desirable for fermenting into cider. We found that dessert apples flower earlier, are heavier, are less firm, and have lower phenolic content than cider apples. Higher phenolic content in cider apples compared to dessert apples is consistent with previ- ous work (Sanoner et al., 1999; Valois et al., 2006) and reflects the desire of cider makers to use apples with high astringency, a juice quality imparted by phenolics (Thompson- Witrick et al., 2014). Furthermore, genomic regions associated with flavonoid biosyn- thesis in dessert apples showed signatures of recent positive se- lection, suggesting that there was perhaps intentional breeding for reduced phenolic content in dessert apples (Migicovsky et al., 2021). Previous work is also consistent with our observation that dessert apples are generally larger than cider apples (Migicovsky et al., 2016). We investigated whether the traits we measured changed over historical time and determined that, over the past 200 years, commercial apples have experienced a reduction in phenolic content. Apple cultivars released after 1940 have 30% lower phenolic content than those released prior to 1940. Phenolic compounds impart many health benefits and apples are one of the major sources of phenolics in the human diet. It is estimated that apples provide 22% of the phenolics that North Americans consume, therefore, a reduction in these compounds during modern cultivar development may be a health concern (Vinson et al., 2001). Breeding apples with increased phenolic content is predicted to be relatively simple since phenolic content is highly heritable and has a simple genetic architecture (McClure et al., 2019). Similar to a previous study (Kschonsek et al., 2018), we hypothesize that the observed decline in phenolic content of newly released apple cultivars is due to active selection against astringent taste and enzymatic browning. Enzymatic browning is driven by high phenolic content (Amiot et al., 1992; Holderbaum et al., 2010; Murata et al., 1995) and is associated with undesir- able color and flavor, and reduced browning has been a target for selection by apple breeders (Toivonen, 2006). In fact, the only genetically modified apple on the market today has been engi- neered to be non- browning (Armstrong & Lane, 2014). It is, thus, possible that the decrease in phenolics over the past 200 years is due to selection for reduced browning, a phenotype that we did not measure but that is correlated with phenolic content. However, breeding can be both the problem as well as the solu- tion, and the present study provides the foundation for the de- velopment of novel apple cultivars that forfeit neither nutrition nor quality. ACKNOWLEDG MENTS General: The authors gratefully acknowledge the Nova Scotia Fruit Growers’ Association and the Farm Services team at AAFC- Kentville for establishing and maintaining the apple trees studied here. We also thank the following people for their help in establishing and/ or maintaining the collection: Charles Embree, Casi Sutherland, Emily Anderson, Dylan Troop, Scott Cann, Laura Porter- Muntz, Andrew Jamieson, Christiane DesLauriers, Thomas Gregoire, Benoit Girard, Leah Beveridge, Michael Jordan, Gavin Douglas, Samantha Lawrence, Jason Sawler, Melanie Eelman, Emma Gillis, Kerstin Surgenor, Conny Bishop, David Baldwin, Nancy Raymond, Cherie Collins, Tamara Dondi, and Leticia Reis. CONFLIC T OF INTERE S T The authors declare no competing interests. AUTHOR CONTRIBUTIONS SM conceived and designed the study. SF aided in statistical de- sign and analysis. SW and ZM performed statistical analyses. SW and SM wrote the manuscript. All other authors collected data in the field or the laboratory or helped coordinate fieldwork or labo- ratory work. DATA AVAIL ABILIT Y S TATEMENT All data and code used for analyses are available through GitHub at github.com/MylesLab/abc- phenomics. ORCID Sophie Watts https://orcid.org/0000-0001-6857-3903 Zoë Migicovsky https://orcid.org/0000-0002-3931-1258 Sean Myles https://orcid.org/0000-0001-9754-5987 |    WATTS eT Al.758 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://orcid.org/0000-0001-6857-3903 https://orcid.org/0000-0001-6857-3903 https://orcid.org/0000-0002-3931-1258 https://orcid.org/0000-0002-3931-1258 https://orcid.org/0000-0001-9754-5987 https://orcid.org/0000-0001-9754-5987 R E FE R E N C E S Amiot, M. J., Tacchini, M., Aubert, S., & Nicolas, J. (1992). Phenolic com- position and browning susceptibility of various apple cultivars at maturity. Journal of Food Science, 57, 958– 962. Armstrong, J., & Lane, W. (2014). Genetically modified reduced- browning fruit- producing plant and produced fruit thereof, and method of ob- taining such. US Patent. Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed- effects models using lme4. Journal of Statistical Software, 67, 1– 48. Benzie, I. F., & Strain, J. J. (1996). The ferric reducing ability of plasma (FRAP) as a measure of ‘antioxidant power’: The FRAP assay. Analytical Biochemistry, 239, 70– 76. Blanpied, G. D., & Silsby, K. J. (1992). Predicting harvest date windows for apples. Cornell University Library. Bramel, P., & Volk, G. M. (2019). A global strategy for the conservation and use of apple genetic resources. Cornille, A., Gladieux, P., Smulders, M. J. M., Roldán- Ruiz, I., Laurens, F., Le Cam, B., Nersesyan, A., Clavel, J., Olonova, M., Feugey, L., Gabrielyan, I., Zhang, X.- G., Tenaillon, M. I., & Giraud, T. (2012). New insight into the history of domesticated apple: Secondary con- tribution of the european wild apple to the genome of cultivated varieties. PLoS Genetics, 8, e1002703. https://doi.org/10.1371/ journ al.pgen.1002703 Duan, N., Bai, Y., Sun, H., Wang, N., Ma, Y., Li, M., Wang, X., Jiao, C., Legall, N., Mao, L., Wan, S., Wang, K., He, T., Feng, S., Zhang, Z., Mao, Z., Shen, X., Chen, X., Jiang, Y., … Chen, X. (2017). Genome re- sequencing reveals the history of apple and supports a two- stage model for fruit enlargement. Nature Communications, 8, 249. Flint- Garcia, S. A., Thuillet, A., Yu, J., Pressoir, G., Romero, S. M., Mitchell, S. E., Doebley, J., Kresovich, S., Goodman, M. M., & Buckler, E. S. (2005). Maize association population: a high- resolution platform for quantitative trait locus dissection. The Plant Journal: for Cell and Molecular Biology, 44, 1054– 1064. Frary, A., Nesbitt, T. C., Grandillo, S., Knaap, E., Cong, B., Liu, J., Meller, J., Elber, R., Alpert, K. B., & Tanksley, S. D. (2000). fw2.2: a quanti- tative trait locus key to the evolution of tomato fruit size. Science, 289, 85– 88. Gao, Y., Liu, F., Wang, K., Wang, D., Gong, X., Liu, L., Richards, C. M., Henk, A. D., & Volk, G. M. (2015). Genetic diversity of Malus culti- vars and wild relatives in the Chinese national repository of apple germplasm resources. Tree Genetics & Genomes, 11, 106. Gottschalk, C., & van Nocker, S. (2013). Diversity in seasonal bloom time and floral development among apple species and hybrids. Journal of the American Society for Horticultural Science. American Society for Horticultural Science, 138, 367– 374. Gross, B. L., Henk, A. D., Richards, C. M., Fazio, G., & Volk, G. M. (2014). Genetic diversity in Malus ×domestica (Rosaceae) through time in response to domestication. American Journal of Botany, 101, 1770– 1779. Gross, B. L., Volk, G. M., Richards, C. M., Reeves, P. A., Henk, A. D., Forsline, P. L., Szewc- McFadden, A., Fazio, G., & Chao, C. T. (2013). Diversity captured in the USDA- ARS national plant germplasm system apple core collection. Journal of the American Society for Horticultural Science. American Society for Horticultural Science, 138, 375– 381. Guerra, M., Sanz, M. A., & Casquero, P. A. (2010). Influence of storage conditions on the sensory quality of a high acid apple. International Journal of Food Science & Technology, 45, 2352– 2357. Hampson, C. R., McNew, R., Cline, J., Embree, C., Embree, C., Zandstra, J., & Wilson, K. (2009). Regional differences in performance of Canadian- bred apple cultivars and implications for breeding. Canadian journal of plant science. Revue Canadienne De Phytotechnie, 89, 81– 91. Harker, F. R., Gunson, F. A., & Jaeger, S. R. (2003). The case for fruit qual- ity: An interpretive review of consumer attitudes, and preferences for apples. Postharvest Biology and Technology, 28, 333– 347. Holderbaum, D. F., Kon, T., Kudo, T., & Guerra, M. P. (2010). Enzymatic browning, polyphenol oxidase activity, and polyphenols in four apple cultivars: Dynamics during fruit development. HortScience: a Publication of the American Society for Horticultural Science, 45: 1150– 1154. Huber, G. M., & Rupasinghe, H. P. V. (2009). Phenolic profiles and anti- oxidant properties of apple skin extracts. Journal of Food Science, 74, C693– 700. Jackson, M. T. (1997). Conservation of rice genetic resources: The role of the International Rice Genebank at IRRI. Plant Molecular Biology, 35, 61– 67. Johnston, J. W., Hewett, E. W., & Hertog, M. L. A. T. M. (2010). Postharvest softening of apple (Malus domestica) fruit: A review. New Zealand Journal of Crop and Horticultural Science, 30, 145– 160. Johnston, J. W., Hewett, E. W., Hertog, M. L. A. T. M., Harker, F. R. (2002). Harvest date and fruit size affect postharvest softening of apple fruit. The Journal of Horticultural Science and Biotechnology, 77, (3), 355– 360. https://doi.org/10.1080/14620 316.2002.11511505 Jung, M., Roth, M., Aranzana, M. J., Auwerkerken, A., Bink, M., Denancé, C., Dujak, C., Durel, C.- E., Font i Forcada, C., Cantin, C. M., Guerra, W., Howard, N. P., Keller, B., Lewandowski, M., Ordidge, M., Rymenants, M., Sanin, N., Studer, B., Zurawicz, E., … Muranty, H. (2020). The apple REFPOP— a reference population for genomics- assisted breeding in apple. Horticulture Research, 7, 189. Khan, A. A., & Vincent, J. F. V. (1990). Anisotropy of apple parenchyma. Journal of the Science of Food and Agriculture, 52, 455– 466. https:// doi.org/10.1002/jsfa.27405 20404 Kouassi, A. B., Durel, C.- E., Costa, F., Tartarini, S., van de Weg, E., Evans, K., Fernandez- Fernandez, F., Govan, C., Boudichevskaja, A., Dunemann, F., Antofie, A., Lateur, M., Stankiewicz- Kosyl, M., Soska, A., Tomala, K., Lewandowski, M., Rutkovski, K., Zurawicz, E., Guerra, W., & Laurens, F. (2009). Estimation of genetic parame- ters and prediction of breeding values for apple fruit- quality traits using pedigreed plant material in Europe. Tree Genetics & Genomes, 5, 659– 672. Kschonsek, J., Wolfram, T., Stöckl, A., & Böhm, V. (2018). Polyphenolic compounds analysis of old and new apple cultivars and contribu- tion of polyphenolic profile to the in vitro antioxidant capacity. Antioxidants (Basel, Switzerland) 7. Kumar, S., Volz, R. K., Alspach, P. A., & Bus, V. G. M. (2010). Development of a recurrent apple breeding programme in New Zealand: A synthe- sis of results, and a proposed revised breeding strategy. Euphytica/ Netherlands Journal of Plant Breeding, 173, 207– 222. Lemon, J. (2006). Plotrix: A package in the red light district of R. R- News, 6, 8- 12. Li, X., Liu, L., Ming, M., Hu, H., Zhang, M., Fan, J., Song, B., Zhang, S., & Wu, J. (2019). comparative transcriptomic analysis provides insight into the domestication and improvement of pear (P. pyrifolia) fruit. Plant Physiology, 180, 435– 452. Li, Y., Cao, K. E., Zhu, G., Fang, W., Chen, C., Wang, X., Zhao, P., Guo, J., Ding, T., Guan, L., Zhang, Q., Guo, W., Fei, Z., & Wang, L. (2019). Genomic analyses of an extensive collection of wild and culti- vated accessions provide new insights into peach breeding history. Genome Biology, 20, 36. Liang, W., Dondini, L., Franceschi, P. D., Paris, R., Sansavini, S., & Tartarini, S. (2015). Genetic diversity, population structure and construction of a core collection of apple cultivars from Italian Germplasm. Plant Molecular Biology Reporter / ISPMB, 33, 458– 473. Ma, B., Chen, J., Zheng, H., Fang, T., Ogutu, C., Li, S., Han, Y., & Wu, B. (2015). Comparative assessment of sugar and malic acid composition in cultivated and wild apples. Food Chemistry, 172, 86– 91.    |WATTS eT Al. 759 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.1371/journal.pgen.1002703 https://doi.org/10.1371/journal.pgen.1002703 https://doi.org/10.1080/14620316.2002.11511505 https://doi.org/10.1002/jsfa.2740520404 https://doi.org/10.1002/jsfa.2740520404 McClure, K. (2017). Exploring horticulturally important traits in apple using genetic mapping in populations of differing size and composition. McClure, K. A., Gong, Y. H., Song, J., Vinqvist- Tymchuk, M., Campbell Palmer, L., Fan, L., Burgher- MacLellan, K., Zhang, Z. Q., Celton, J.- M., Forney, C. F., Migicovsky, Z., & Myles, S. (2019). Genome- wide association studies in apple reveal loci of large effect controlling apple polyphenols. Horticulture Research, 6, 107. Mehlenbacher, S. A., & Voordeckers, A. M. (1991). Relationship of flow- ering time, rate of seed germination, and time of leaf budbreak and usefulness in selecting for late- flowering apples. Journal of the American Society for Horticultural Science. American Society for Horticultural Science, 116, 565– 568. Meyer, R. S., DuVal, A. E., & Jensen, H. R. (2012). Patterns and processes in crop domestication: an historical review and quantitative analy- sis of 203 global food crops. The New Phytologist, 196, 29– 48. Migicovsky, Z., Gardner, K. M., Money, D., Sawler, J., Bloom, J. S., Moffett, P., Chao, C. T., Schwaninger, H., Fazio, G., Zhong, G.- Y., & Myles, S. (2016). Genome to phenome mapping in apple using historical data. The Plant Genome, 9. Migicovsky, Z., Gardner, K. M., Richards, C., Thomas Chao, C., Schwaninger, H. R., Fazio, G., Zhong, G.- Y., & Myles, S. (2021). Genomic consequences of apple improvement. Horticulture Research, 8, 9. Migicovsky, Z., & Myles, S. (2017). Exploiting wild relatives for genomics- assisted breeding of perennial crops. Frontiers in Plant Science, 8, 1– 16. Migicovsky, Z., Warschefsky, E., Klein, L. L., & Miller, A. J. (2019). Using living germplasm collections to characterize, improve, and conserve woody perennials. Crop Science, 59, 2365– 2380. Miller, A. J., & Gross, B. L. (2011). From forest to field: Perennial fruit crop domestication. American Journal of Botany, 98, 1389– 1414. Murata, M., Noda, I., & Homma, S. (1995). Enzymatic browning of ap- ples on the market: Relationship between browning, polyphenol content, and polyphenol oxidase. Nippon Shokuhin Kagaku Kogaku Kaishi, 42, 820– 826. Noiton, D. A. M., & Alspach, P. A. (1996). Founding clones, inbreeding, coancestry, and status number of modern apple cultivars. Journal of the American Society for Horticultural Science. American Society for Horticultural Science, 121, 773– 782. Nybom, H., Ahmadi- Afzadi, M., Sehic, J., & Hertog, M. (2013). DNA marker- assisted evaluation of fruit firmness at harvest and post- harvest fruit softening in a diverse apple germplasm. Tree Genetics & Genomes, 9, 279– 290. Oraguzie, N. C., Iwanami, H., Soejima, J., Harada, T., & Hall, A. (2004). Inheritance of the Md- ACS1 gene and its relationship to fruit softening in apple (Malus x domestica Borkh.). TAG. Theoretical and applied genetics. Theoretische Und Angewandte Genetik, 108, 1526– 1533. R Core Team (2018). R: A language and environment for statistical comput- ing. R Foundation for Statistical Computing. https://www.R- proje ct.org/ Rupasinghe, H. P. V., Wang, L., Huber, G. M., & Pitts, N. L. (2008). Effect of baking on dietary fibre and phenolics of muffins incorporated with apple skin powder. Food Chemistry, 107, 1217– 1224. Sanoner, P., Guyot, S., Marnet, N., Molle, D., & Drilleau, J.- F. (1999). Polyphenol profiles of french cider apple varieties (Malus domes- tica sp.). Journal of Agricultural and Food Chemistry, 47, 4847– 4853. Seeley, S. D., & Anderson, J. L. (2003). Apple- orchard freeze protection. In D. C. Ferree, & I. J. Warrington (Eds.), Apples: botany, production and uses (pp. 521– 538). CABI Publishing. Singleton, V. L., Orthofer, R., & Lamuela- Raventós, R. M. (1999). Analysis of total phenols and other oxidation substrates and antioxidants by means of folin- ciocalteu reagent. Methods in Enzymology (pp. 152– 178). Academic Press. Sun, X., Jiao, C., Schwaninger, H., Chao, C. T., Ma, Y., Duan, N., Khan, A., Ban, S., Xu, K., Cheng, L., Zhong, G., & Fei, Z. (2020). Phased dip- loid genome assemblies and pan- genomes provide insights into the genetic history of apple domestication. Nature Genetics, 52, 1423– 1432. https://doi.org/10.1038/s4158 8- 020- 00723 - 9 Szczesniak, A. S. (2002). Texture is a sensory property. Food Quality and Preference, 13, 215– 225. https://doi.org/10.1016/s0950 - 3293(01)00039 - 8 Thompson- Witrick, K. A., Goodrich, K. M., Neilson, A. P., Hurley, E. K., Peck, G. M., & Stewart, A. C. (2014). Characterization of the Polyphenol Composition of 20 Cultivars of Cider, Processing, and Dessert Apples (Malus × domestica Borkh.) Grown in Virginia. Journal of Agricultural and Food Chemistry, 62, 10181– 10191. Toivonen, P. M. A. (2006). Fresh- cut apples: Challenges and opportuni- ties for multi- disciplinary research. Canadian journal of plant science. Revue Canadienne De Phytotechnie, 86, 1361– 1368. US Apple and Pear Forecast. The world apple and pear association (WAPA). Valois, S., Merwin, I. A., & Padilla- Zakour, O. I. (2006). Characterization of fermented cider apple cultivars grown in Upstate New York. Journal- American Pomological Society, 60, 113– 128. Velasco, R., Zharkikh, A., Affourtit, J., Dhingra, A., Cestaro, A., Kalyanaraman, A., Fontana, P., Bhatnagar, S. K, Troggio, M., Pruss, D., Salvi, S., Pindo, M., Baldi, P., Castelletti, S., Cavaiuolo, M., Coppola, G., Costa, F., Cova, V., Dal Ri, A., … Viola, R. (2010). The genome of the domesticated apple (Malus × domestica Borkh.). Nature Genetics, 42, 833– 839. https://doi.org/10.1038/ng.654 Verma, S., Evans, K., Guan, Y., Luby, J. J., Rosyara, U. R., Howard, N. P., Bassil, N., Bink, M. C. A. M., van de Weg, W. E., & Peace, C. P. (2019). Two large- effect QTLs, Ma and Ma3, determine genetic potential for acidity in apple fruit: breeding insights from a multi- family study. Tree Genetics & Genomes, 15, 18. Vinson, J. A., Su, X., Zubik, L., & Bose, P. (2001). Phenol antioxidant quan- tity and quality in foods: Fruits. Journal of Agricultural and Food Chemistry, 49, 5315– 5321. Volk, G. M., Chao, C. T., Norelli, J., Brown, S. K., Fazio, G., Peace, C., McFerson, J., Zhong, G.- Y., & Bretting, P. (2015). The vulnerability of US apple (Malus) genetic resources. Genetic Resources and Crop Evolution, 62, 765– 794. Ward, G. M. (1978). A History of the Research Station in Harrow, Ontario. Research Branch, Canada Department of Agriculture. Watkins, C. B. (2003). Principles and Practices of Postharvest Handling and Stress. In D. C. Ferree, & I. J. Warrington (Eds.), Apples: Botany, Production and Uses (pp. 585– 614). CABI Publishing. Zhang, Y., Li, P., & Cheng, L. (2010). Developmental changes of carbo- hydrates, organic acids, amino acids, and phenolic compounds in ‘Honeycrisp’ apple flesh. Food Chemistry, 123, 1013– 1018. Zhu, C., Gore, M., Buckler, E. S., & Yu, J. (2008). Status and Prospects of Association Mapping in Plants. The Plant Genome Journal, 1, 5– 20. Zohary, D., Hopf, M., & Weiss, E. (2012). Domestication of Plants in the Old World: The origin and spread of domesticated plants in Southwest Asia, Europe, and the Mediterranean Basin, Oxford University Press. SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section. How to cite this article: Watts S, Migicovsky Z, McClure KA, et al. Quantifying apple diversity: A phenomic characterization of Canada’s Apple Biodiversity Collection. Plants, People, Planet. 2021;3:747–760. https://doi. org/10.1002/ppp3.10211 |    WATTS eT Al.760 25722611, 2021, 6, D ow nloaded from https://nph.onlinelibrary.w iley.com /doi/10.1002/ppp3.10211 by C anadian A griculture L ibrary A griculture & A gri-Food C anada, W iley O nline L ibrary on [09/01/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.R-project.org/ https://www.R-project.org/ https://doi.org/10.1038/s41588-020-00723-9 https://doi.org/10.1016/s0950-3293(01)00039-8 https://doi.org/10.1016/s0950-3293(01)00039-8 https://doi.org/10.1038/ng.654 https://doi.org/10.1002/ppp3.10211 https://doi.org/10.1002/ppp3.10211