JANUARY 2019 Volume 73 Number 1 AMERICAN POMOLOGICAL SOCIETY Founded in 1848 Incorporated in 1887 in Massachusetts

2018-2019 PRESIDENT FIRST SECOND M. PRITTS VICE PRESIDENT VICE PRESIDENT N. BASSIL K. GASIC SECRETARY RESIDENT AGENT MASSACHUSETTS T. EINHORN W. R. AUTIO EDITOR R. P. MARINI

EXECUTIVE BOARD M. WARMUND M. PRITTS N. BASSIL Past President President 1st Vice President K. GASIC T. EINHORN 2nd Vice President Secretary E. HOOVER G. PECK A. ATUCHA ('16 - '19) ('17 - '20) ('18 - '21)

ADVISORY COMMITTEE 2016-2019 R. MORAN E. GARCIA S. YAO M. EHLENFELDT D. BRYLA

2017-2020 B. BLACK G. FERNANDEZ D. KARP I. MINAS S. SERRA

2018-2021 G. LANG T. KON D. CHAVEZ T. VASHISTH S. NAMBEESAN

CHAIRS OF STANDING COMMITTEES Editorial Shepard Award Membership R. PERKINS-VEAZIE F. TAKEDA P. HIRST Wilder Medal Awards Nominations U. P. Hedrick Award J. CLARK P. HIRST E. FALLAHI

Registration of New Fruit and Nut Cultivars Website K. GASIC & J. PREECE M. OLMSTEAD 1

January 2019 Volume 73 Number 1 CONTENTS A Retrospective Analysis of Pawpaw (Asimina triloba [L.] Dunal) Production Data from 2005-2012 – Laine M. Greenawalt, Ron Powell, Janet Simon, and Robert G. Brannan...... 2

Regional Evaluation of Seven Newly Introduced Sweet Cherry Cultivars in North-East of Iran (Shahrood) – N. Salehabadi, M. Rezaei, A. Sarkhosh, H. Hokmabadi, and M. Abedini Esfahani...... 12 An Analysis of Strawberry (Fragaria ×ananassa) Productivity in Northern Latitudinal Aquaponic Growing Conditions – Marie Abbey, Neil O. Anderson, Chengyan Yue, Gianna Short, and Michele Schermann...... 22 Peach Bagging in the Southeastern U.S. – Jaine Allran, Guido Schnabel, and Juan Carlos Melgar...... 38 The Importance of Being ‘Boysen’: Examining Genotypic Variation with SSR Markers (U.P. Hedrick Award - Second Place) – Katie A. Carter, Jason D. Zurn, Nahla V. Basssil, Chad E. Finn, and Kim E. Hummer...... 47 The Response of ‘Montmorency’ Tart Cherry to Renewal Pruning Strategies in a High Density System (U.P. Hedrick Award – First Place) – Sheriden Hansen and Brent L. Black...... 53 Diversity of Pathogenic Fungi Associated with Apples in Cold Storage Facilities in Tunisia – Bochra A. Bahri, Yosra Belaid, Ghaya Mechichi, and Wafa Rouissi...... 62 About The Cover ...... 11 Correction ...... 46 Index for Volume 72 ...... 76 Instruction to Authors...... 80

Published by THE AMERICAN POMOLOGICAL SOCIETY Journal of the American Pomological Society (ISSN 1527-3741) is published by the American Pomological Society as an annual volume of 4 issues, in January, April, July and October. Membership in the Society includes a volume of the Journal. Most back issues are available at various rates. Paid renewals not received in the office of the Business Manager by January 1 will be temporarily suspended until payment is received. For current membership rates, please consult the Business Manager. Editorial Office: Manuscripts and correspondence concerning editorial matters should be addressed to the Editor: Richard Marini, 203 Tyson Building, Department of Plant Science, University Park, PA 16802-4200 USA; Email: [email protected]. Manuscripts submitted for publication in Journal of the American Pomological Society are accepted after recommendation of at least two editorial reviewers. Guidelines for manuscript preparation are the same as those outlined in the style manual published by the American Society for Horticultural Science for HortScience, found at http://c.ymcdn.com/sites/www.ashs.org/resource/resmgr/files/style_manual.pdf. Postmaster: Send accepted changes to the Business office. Business Office: Correspondence regarding subscriptions, advertising, back issues, and Society membership should be addressed to the Business Office, C/O Heather Hilko, ASHS, 1018 Duke St., Alexandria, VA 22314; Tel 703-836- 4606; Email: [email protected] Page Charges: A charge of $50.00 per page for members and $65.00 per page ($32.00 per half page) will be made to authors. In addition to the page charge, there will be a charge of $40.00 per page for tables, figures and photographs. Society Affairs: Matters relating to the general operation of the society, awards, committee activities, and meetings should be addressed to Michele Warmund, 1-31 Agriculture Building, Division of Plant Sciences, University of Missouri, Columbia MO 65211; Email:[email protected]. Society Web Site: http://americanpomological.org 2 3

Journal of the American Pomological Society 73(1): 2-11 2019 A Retrospective Analysis of Pawpaw (Asimina triloba [L.] Dunal) Production Data from 2005-20121 Laine M. Greenawalt1, Ron Powell2, Janet Simon1, and Robert G. Brannan1,3

Additional index words: Asimina triloba, Pawpaw, Asimina triloba, fruit production, average FW

Abstract This study examined 52 cultivars of pawpaw (Asimina triloba) grown at three locations in southwestern Ohio, of which 24 fit into previously identified genetic groups based on simple sequence repeat markers, harvested from 2005-2012. The harvest duration ranged from 31 days in 2005 to 74 days in 2010, and most of the fruit was harvested in Sept. A frequency analysis of average FW conducted for all cultivars revealed that average FW was normally distributed. The average FW ranged from 10 g (‘Cullman Late’) to 244 g (‘Davis’), but of the more prolific cultivars the average FW ranged from 72 g (‘Rappahannock’) to 172 g (‘NC-1’). Average FW and total number of fruit collected were negatively correlated (r = -0.21, p < 0.001). FW was not affected by location. Genetically similar cultivars were compared based on previously identified groupings. Cultivars assigned to the Susquehanna and Overleese genetic groups had significantly higher average FW and harvest length than other genetic groups. Results of this study expand research on pawpaw production and allow growers to select cultivars with market potential based on factors such as fruit size, yield, and harvest duration.

The pawpaw (Asimina triloba) is the larg- dition, the normal bloom period of pawpaw est fruit native to North America. It has a vast flowers can last up to 4 weeks, resulting in native growing range that spans 26 states an elongated harvest period in the fall which corresponding to U.S. Department of Agri- requires more labor and time intensive har- culture (USDA) plant hardiness zones 5-8 vesting techniques (Pomper et al., 2008a). (Galli et al., 2007). The family Annonaceae, In the early 1900s, the pawpaw was con- to which pawpaw belongs, contains over sidered for increased commercialization ef- 2,400 species, all of which are tropical or forts as a native, North American cash crop, subtropical except for pawpaw, making the but these efforts did not materialize (Pomper temperate-growing fruit unique. Moreover, and Layne, 2005). Over the next 60 years, pawpaw trees are exclusive in that they are over four dozen pawpaw cultivars were the only species of the 10 in the genus Asim- named. However, because of neglect or aban- ina to produce a commercially viable fruit. donment only a few of these remain, which Pawpaw trees are shrub-like, understory potentially could have eroded the genetic trees that can grow up to 12 meters tall. They base of current pawpaw cultivars. Since then, have long membranous leaves and maroon breeders and hobbyists have increased ef- flowers that produce clusters of up to nine forts to domesticate the pawpaw. One result fruit in late summer to early fall (Geneve et of this effort has been the establishment of al., 2003; Pomper and Layne, 2005). Howev- the officialAsimina satellite repository of the er, pawpaw flowers are likely self-unfruitful, USDA National Clonal Germplasm Reposi- making pollination an important factor that tory (Corvallis, OR) at Kentucky State Uni- could limit commercial possibilities (Layne, versity and the development of the Pawpaw 1996; Willson and Schemske, 1980). In ad- Regional Variety Trial (Pomper et al., 2008b;

1 School of Applied Health Sciences and Wellness, Ohio University, Athens, OH 2 Fox Paw Ridge Farm, Cincinnati, OH 3 Author to whom correspondence should be addressed: Robert G. Brannan, E170 Grover Center, Athens, OH 45701, (740) 593-2879 (office), (740) 593-0289, [email protected] Pawpaw 213 3

Pomper et al., 2003b). In addition, attempts The plantings at the Butler County locations have been made to evaluate the genetic di- began in 2001 and tree site and cultivar se- versity in the pawpaw. Although an early lection was primarily random. The Adams study suggested that genetic diversity in County location, Fox Paw Ridge Farm, is the pawpaw was low (Rogstad et al., 1991), a plot that contained pawpaw trees planted sampling techniques with greater discrimina- from 2003-2006. Fox Paw Ridge Farm tion showed that pawpaw has moderate to plantings were organized in three sections. high genetic diversity and is similar to other At the time of data collection, the 0.61-ha temperate woody perennials (Pomper et al., north section contained over 200 trees in 2003a; Pomper et al., 2010). east-to-west rows with 2.5 m between trees. In 2000, the Ohio Pawpaw Growers As- The 0.2-ha east section contained four rows sociation (OPGA) was formed and ulti- running north-south with trees 2.5 m apart. mately became the first state chapter of the The 0.2-ha west section contained nine rows North American Pawpaw Growers Associa- running north-south with 3 m between trees. tion (NAPGA). Currently, there are many In all three sections, rows were planted 4.5 cultivars with wide differences in fruit size, m apart and cultivar selection was primar- yield and other characteristics, but there are ily random. There is notable soil variation scant data on cultivars outside the Pawpaw between the Butler and Adams County loca- Regional Variety Trial. Research has yet to tions. Soil at the Butler County locations is determine which cultivars of pawpaw would fine and loamy with good drainage (ideal for be best for commercial purposes. pawpaw growth) while soil at the Fox Paw The objective of this research was to per- Ridge Farm location is heavy clay (less ideal form a retrospective analysis of pawpaw fruit for pawpaw growth). The pawpaw trees were production data collected from three sites fertilized three times per season using 10-10- over a period of eight growing seasons. 10 fertilizer. Rain water and cisterns were used to water the trees at the rural location, Materials and Methods and the suburban locations were watered as Plantings. Pawpaw fruit was harvested at needed using municipal water. The pawpaw three locations in southwest Ohio, namely trees were pruned and limbed as needed, a farm in rural Adams County (38.655517, however, the lower branches were retained -83.698615) that consists of three plots to prevent deer rubbing. No pesticides were that total 1.0 ha and two discontinuous but applied. The ground around the pawpaw adjacent plots in suburban Butler County trees was mowed and trimmed to control (39.294718, -84.365151 and 39.294162, grass and weeds. -84.364819) that are approximately 0.05 ha Growing degree days (GDDs) were cal- each. Trees were grafted onto seedling root- culated using a base temperature of 10°C stocks that were at least two years old. Seed- according to previous research for pawpaw lings were grown by R. Powell from the col- (Pomper et al., 2010). The Ornamental/ lection at Kentucky State University that has Horticultural Insect Degree Day Calculator/ served as the USDA National Clonal Germ- Forecaster (http://weather.uky.edu/php/ge- plasm Repository since 1994. The plants neric_dd_www.php) from the University of were grafted in late spring to early summer. Kentucky Agricultural Weather Center was The newly grafted plants were held for a year used to calculate temperature data for each at one of the suburban locations before be- harvest year from 1 Jan. to 1 Aug. ing planted at the orchard sites. The number Data collection. Fruit were hand-collected of trees per cultivar ranged from 1 to 34 and and data were recorded in notebooks taken the same cultivar was usually not planted at from spring 2005 through fall 2012. Fruit more than one location. were collected each morning and evening, 4 Journal of the American Pomological Society and only fruit that had fallen from the tree p < 0.05, Duncan’s Multiple Range Test was was collected. Individual fruits were not used to separate means. Pearson product- weighed; rather, fruits from each tree were moment correlation coefficient (r) was used pooled and weighed in a batch using a Cen- to measure the relationship between average Tech Digital Scale (Harbor Freight Tools), FW (g), total number of fruit produced per recorded in ounces, and later converted into tree, tree yield (g), and harvest days. Signifi- grams. Data collected for individual trees cance for all analyses was set at p < 0.05. included cultivar, tree location, date of col- lection, total number of fruit collected, and Results and Discussion total average FW. Two harvest variables were Harvest data for all cultivars and years. calculated. The variable “harvest days” is the The overall harvest duration, measured as the number of days that fruit from a cultivar was first collection day to the last collection day harvested, whereas the variable “harvest du- per year for all trees on all sites is shown in ration” is calculated as the number of days Figure 1. The harvest duration ranged from each year from the first day a cultivar was 31 days in 2005 to 74 days in 2010. Most of harvested until the last day a cultivar was the fruit were harvested in Sept. Almost all harvested. individual trees had a first-recorded harvest Statistical analyses. The data were tran- date in Aug. or Sept. (64% and 35%, respec- scribed from paper to an electronic format tively) with the remaining 1% with a first- by at least two investigators. Independent recorded harvest date in either July or Oct. variables were identified as cultivar, harvest The midpoint of harvest duration in terms year, and tree location. Dependent variables of number of fruit migrated from mid-Sept. were identified as harvest dates, total number from 2005-2009 to late Aug. or early Sept. of fruit collected per tree, total average FW, for 2010-2012 (Fig. 1). During the period and average FW. under study, there were three years for which Data were analyzed using statistical analy- events could have affected the data. The year sis software SPSS Statistics 22 (IBM Corpo- 2005 was the first year of data recording, and ration, Armonk, New York). Of the 60 culti- there were fewer mature trees, so it is not a vars for which there were data, 24 cultivars surprising result that 2005 had the shortest were placed into one of five genetic groups harvest duration. A drought occurred in 2007, based on findings by Pomper et al. (2010), which may account for the short, 41-day har- which used polymorphic microsatellite mark- vest duration, 9 days shorter than the average er loci to identify genetically similar pawpaw range (Fig. 1). While 2010 had the potential cultivars. This created an independent vari- to be a bumper year (large FW), Hurricane able of genetic group with five levels (Tay- Ike brought damaging winds to the area in lor & Wilson, Susquehanna, Wabash, Wells, early Sept. resulting in the loss of over 45 kg Overleese). Genetic information was not of unripe fruit in total across the three loca- available for 36 cultivars so these cultivars tions. This produced an overall smaller yield. were not placed into a genetic group. Eight of The year 2010 was also the first year that these cultivars (‘PA Golden 2’, ‘Allegheny’, fruit from the Fox Paw Ridge Farm location ‘Forest Keeling’, ‘Vicky Russel’, ‘Pickle’, were included in the analysis, resulting in an ‘Danica’, ‘Pepper’, ‘Cherokee Ridge’) were increase in the number of trees and may be excluded from data analysis because fruit a reason that 75% of the fruit were collected collection only occurred once. by early Sept. (Fig. 1). In contrast, median Multivariate ANOVAs were used to deter- harvest date, i.e., 50% of fruit collection, for mine differences in average FW (g), yield (g), the surrounding years 2009 and 2011 fell in and harvest days between genetic groups, lo- mid- to late-Sept. The year 2010 also had the cation, and year. If significance was noted at longest harvest range of 74 days. Pawpaw 5

Fig. 1. Pawpaw harvest date ranges (from first to last record) divided into quartiles for 52 pawpaw cultivars grown in Ohio in two suburban and one rural location from 2005-2012, with growing degree days (GDD) as of August 1 for eachFig year 1. Pawpaw indicated harvest in parenthesis. date ranges (from first to last record) divided into quartiles for 52 pawpaw cultivars grown in Ohio in two suburban and one rural location from 2005-2012, with growing degree days (GDD) as of August 1 for each year indicated in parenthesis. The average number of harvest days, de- with those reported in the most recent Paw- fined as the number of days that a tree pro- paw Regional Variety Trial (Pomper et al., duced harvestable fruit, was 19 days and 2008b) organized by Kentucky State Univer- ranged from 5 to 25 days depending on the sity, the USDA National Clonal Repository year. There was a significant (p = 0.047) pos- for pawpaw. Average harvest duration for the itive correlation (r = 0.714) between harvest 28 cultivars reported in the trial was 22 days, duration and harvest days. Growing degree 4 days longer than the fruit for this study (18 days (GDD) are shown in Figure 1 and are days). This small difference could be related a measure of the heat accumulation. GDD in to a number of factors including growing lo- this study were calculated as of 1 Aug. for cation (Kentucky versus Ohio) or cultivars the respective year (Fig. 1). GDD was not sampled. correlated with harvest duration (p = 0.830) Frequency data for pawpaw average FW or the midpoint of harvest, i.e. the number of for all cultivars and years. A frequency (f) days required for half of the fruit to be har- analysis of average FW was conducted for all vested that year (p = 0.693). This suggests cultivars (Fig. 2A) and average FW was nor- that factors other than temperature through- mally distributed. No previous information out the entirety of the growing season may be about the average FW distribution of pawpaw important. For example, research has shown or other Annonaceous fruits has been report- that early season temperatures and crop load ed. As evidenced by the Shapiro-Wilk’s test can affect both days from bloom to harvest (S-W test) of normality, the average FW over and fruit size at harvest for peach where the the collection period for all cultivars are nor- relationship between GDD during the first mally distributed with a slight positive skew. 30 days after bloom was positive for harvest Individually, 41 of the 52 cultivars exhibited days but negative for fruit weight (Marini, normality for average FW (data not shown). 2018). A limitation to this analysis is that average The results of this study generally agree fruit weight per cultivar was used rather than 6 Journal of the American Pomological Society individual fruit weight. Thus, the individual search (Pomper et al., 2010) and the remain- fruit weight may have a weight distribution ing 28 cultivars uncategorized because no ge- that could be obscured by the measurement netic information was available. The genetic method employed. groups were designated based on the name Across all of the cultivars, several vari- of a member, or in one case two members. ables were significantly correlated. Average The S-W test of normality indicated that the FW was correlated with total number of fruit average FW of the 24 cultivars within a ge- collected (r = -0.21, p < 0.001) and number netic group were normally distributed with a of harvest days (r = -0.11, p < 0.001). To- positive skew (Fig. 2B). Individually, three tal number of fruit also was significantly of the genetic groups (Taylor & Wilson, Wa- correlated with number of harvest days and bash, and Wells) were normally distributed, yield (r = 0.62 and r = 0.90 respectively, p < and two (Susquehanna and Overleese) did 0.001), and yield was significantly correlated not follow normal distribution for FW (data with harvest days (r = 0.61, p < 0.001). This not shown). The S-W test of normality for analysis indicates, among other things, that the remaining 28 cultivars indicates normal- as the number of total mature fruit increas- ity with a high degree of positive skew (Fig. es, average FW decreases. These findings 2C), however, these varieties may be inher- support a generally accepted phenomenon ently dissimilar from one another and have among horticulturalists and one that has been not had their genetic profiles determined. The confirmed for pawpaw (Crabtree and- Pom positive skew observed in the cultivars that per, 2007; Crabtree et al., 2010), but these re- were not placed in a genetic group (Fig. 2C) searchers caution that it is important that the likely is responsible for the positive skew increased costs of hand-thinning are offset by seen across all 52 cultivars (Fig. 2A). the increased profits brought by larger fruit Comparative analysis of pawpaw across before it should be recommended to growers all cultivars. The main effects of location (Crabtree et al., 2009). (two suburban, one rural) and year (2008- The average FW for each of the 52 cul- 2012) were analyzed for FW across all 52 tivars of pawpaw is shown in Table 1, with cultivars of pawpaw. With respect to loca- twenty-four (24) of the cultivars categorized tion, no significant differences between the into a genetic group based on previous re- three locations were observed for average

Number of Occurrences Number Number of Occurrences Number

Fig. 2: Frequency (f) distribution (number of occurrences of the average fruit weight per cultivar per year) versus average fruit weight (g) for (A) pawpaws of all cultivars grown in Ohio in two suburban and one rural location from 2005-2012 and two subsets: (B) pawpaw fruit characterized into genetic groups by simple sequence repeat markers; and (C) pawpaw fruit of unknown genetic group. Figure 2. Frequency (f) distribution (number of occurrences of the average fruit weight per cultivar per year) versus average fruit weight (g) for (A) pawpaws of all cultivars grown in Ohio in two suburban and one rural location from 2005-2012 and two subsets: (b) pawpaw fruit characterized into genetic groups by simple sequence repeat markers; and (c) pawpaw fruit of unknown genetic group.

Pawpaw 7

Table 1. Total number of fruit, average fruit weight, total yield per tree, and the number of days between the harvest of the first and last fruit (Harvest Duration) of individual pawpaw cultivars grown in Ohio from two suburban and one rural location from 2005-2012 that are 1) categorized into one of five genetic groupings based on simple sequence repeat markers, 2) of unknown genetic group, and 3) of unknown genetic group and excluded from the data analysis because harvest occurred only once. Values are means ±SD. Total # Avg. Fruit Harvest Group Cultivar of Fruit of Weight (g) Total Yield (g) Duration

1) Categorized into a Genetic Groupz Overleese Genetic Group ‘Davis’ 8 244 ± 95 1017 ± 946 3 ± 5 ‘NC-1’ 978 172 ± 43 9550 ± 12791 20 ± 13 ‘Rebecca’s Gold’ 473 167 ± 59 9953 ± 5396 21 ± 8 ʻOverleeseʼ 1036 160 ± 49 7208 ± 5151 17 ± 8 ‘Shenandoah’ 577 153 ± 47 3660 ± 3930 20 ± 10 ʻSunflowerʼ 1319 148 ± 36 10998 ± 11570 25 ± 15 ʻTaytwoʼ 261 121 ± 47 3024 ± 3086 13 ± 16 ʻIXLʼ 8 121 ± 25 4888 ± 6687 22 ± 17 ʻMitchellʼ 559 117 ± 34 5706 ± 6976 20 ± 14

Susquehanna Genetic Group ʻSusquehannaʼ 75 194 ± 76 1846 ± 2392 15 ± 8 ʻSAA Zimmermanʼ 727 170 ± 24 18182 ± 14847 24 ± 10 ʻProlificʼ 116 105 ± 33 2132 ± 1169 13 ± 8 ʻPA Golden 1ʼ 1139 93 ± 28 11161 ± 8147 23 ± 15 ʻPA Golden 3ʼ 6 92 ± 9 276 ± 26 --

Taylor and Wilson Genetic Group ʻTaylorʼ 181 119 ± 49 2079 ± 1855 9 ± 8 ʻWilsonʼ 1217 80 ± 28 4433 ± 5120 18 ± 12

Wabash Genetic Group ʻSweet Aliceʼ 12 141 ± 19 792 ± 684 4.5 ± 5 ʻPotomacʼ 16 121 ± 53 427 ± 447 3 ± 3 ʻPA Golden 4ʼ 43 107 ± 39 542 ± 473 8 ± 11 ʻGreen River Belleʼ 22 106 ± 22 10818 ± 8016 26 ± 15 ʻRappahannockʼ 251 72 ± 23 1218 ± 1247 9 ± 7

Wells Genetic Group ʻWellsʼ 312 144 ± 46 4672 ± 5779 12 ± 12 ʻMiddletownʼ 33 96 ± 15 1038 ± 119 10 ± 10 ʻSueʼ 2087 96 ± 43 10215 ± 11349 19 ± 14 2) Unknown Genetic Group Cultivars ʻMangoʼ 12 210 ± 83 1321 ± 794 5 ± 6 ‘SAB Overleese’ 17 208 ± 75 1021 ± 879 1 ± 0 ‘SAA Overleese’ 7 156 ± 69 472 ± 90 1 ± 0 ʻCawoodʼ 13 153 ± 11 667 ± 371 22 ± 4 8 Journal of the American Pomological Society

ʻLady Dʼ 12 142 ± 13 822 ± 531 1 ± 0 ʻKYSUʼ (Atwood) 1120 141 ± 67 11349 ± 10202 22 ± 12 ʻMary Foos Johnsonʼ 121 139 ± 80 1234 ± 1476 10 ± 10 ʻEstilʼ 3 136 ±17 734 ± 392 15 ± 13 ʻLynn’s Favoriteʼ 487 135 ± 55 3791 ± 4471 14 ± 10 ʻ421ʼ 11 132 ± 15 722 ± 10 17 ± 6 ʻQuaker Delightʼ 50 127 ± 69 2010 ± 2594 7 ± 9 ʻBelleʼ 18 125 ± 26 1194 ±934 8 ± 7 ʻTollgateʼ 217 120 ± 27 3853 ± 2141 17 ± 9 ʻBroadʼ 8 112 ±26 483 ± 419 7 ± 8 ʻSweet Virginiaʼ 47 110 ± 26 931 ± 675 12 ± 12 ʻGlaserʼ 4 103 ± 45 564 ± 639 6 ± 10 ʻKYSU 2-11ʼ 1682 99 ± 50 8335 ± 9206 33 ± 19 ʻKYSU 2-7ʼ 882 85 ± 27 5316 ± 4041 24 ± 15 ʻKYSU Seedlingʼ 303 81 ± 29 2901 ± 2415 22 ± 13 ʻRuby Keenanʼ 12 79 ± 36 359 ± 405 8 ± 12 ʻSunGloʼ 99 79 ± 14 1514 ± 1295 15 ± 9 ʻShawnee Trialʼ 3 77 ± 44 99 ± 12 1 ± 0 ʻKristenʼ 2 76 ± 1 337 ± 261 2 ± 1 ʻWildʼ 944 76 ± 1 264 ±156 11 ± 14 ʻConvisʼ 22 74 ± 6 791 ± 565 2.0 ± 1 ʻRanaʼ 7 73 ± 23 281 ± 237 5 ± 5 ʻLA Nativeʼ 19 68 ± 10 654 ± 146 15 ± 4 ʻCullman Lateʼ 16 10 ± 20 563 ± 212 1 ± 0

3) Unknown genetic group cultivars that were excluded from the analysis ʻPA Golden 2ʼ 1 99 692 1 ʻAlleghenyʼ 2 96 1338 1 ʻForest Keelingʼ 3 83 250 1 ʻVicky Russelʼ 3 79 238 1 ʻPickleʼ 14 78 1089 1 ʻDanicaʼ 19 68 1293 1 ʻPepperʼ 4 67 267 1 ʻCherokee Ridgeʼ 1 53 532 1 z Genetic classifications from (Pomper et al., 2010)

FW (data not shown). With respect to har- soil characteristics (pH or fertility), sunlight, vest year, significant differences for average and irrigation. Soil conditions of the subur- FW, yield, and harvest days for all cultivars ban locations (loam) were different from that for each growing year are shown in Table 2. of the rural location (clay). More research is While significant differences were noted, it required to determine what environmental is challenging to elucidate what caused these factors affect fruit production and harvest differences or discern any patterns. Yearly pre- duration of various pawpaw cultivars. harvest factors such as weather could have a Shown in Table 1, the average FW ranged significant effect on fruit production, as could from 10 g (‘Cullman Late’) to 244 g (‘Da- Pawpaw 9

Table 2. Fruit weight, yield per tree, number of days that individual fruit was harvested (harvest days), and first date and last date of harvest averaged over 52 pawpaw cultivars grown in Ohio in two suburban and one rural location from 2005-2012. Values are means ±SD. Year Average fruit weight (g) Yield (g) Harvest Days Harvest Duration 2005 121abcz ± 43 679d ± 564 5d ± 6 Aug 29 - Sep 27 2006 153a ± 48 4134bcd ± 6770 13bcd ± 11 Aug 24 – Oct 6 2007 101c ± 47 1254cd ± 1008 9cd ± 6 Aug 19 – Sep 28 2008 100c ± 46 5389abcd ± 4837 21ab ± 15 Aug 30 – Oct 29 2009 138ab ± 54 9145a ± 9087 25a ± 13 Aug 16 – Oct 20 2010 143a ± 61 3013bcd ± 5045 16bc ± 13 Jul 31 – Oct 13 2011 122abc ± 55 6027abc ± 8560 17bc ± 11 Aug 24 – Oct 3 2012 108bc ± 47 6881ab ± 9080 16bc ± 13 Jul 24 – Oct 11 z Means within columns followed by common letters do not differ, by Duncan’s Multiple Range test, p<0.05. vis’), but both of these cultivars produced fruit than the groups Taylor and Wilson, Wa- few fruit per tree (16 and 8 fruits, respective- bash, and Wells. The Susquehanna group ly). Of the more prolific cultivars, arbitrarily had larger fruit than the groups Taylor and defined as the 19 cultivars that produced Wilson and Wabash. There were significant more than 250 fruit per tree over the duration differences in harvest days , however harvest of the study, the average FW ranged from days only ranged from 13-19 days. Although 72 g (‘Rappahannock’) to 172 g (‘NC-1’). Overleese and Susquehanna share some phe- Comparative analysis of pawpaw from notypical similarities, there is little genotypic cultivars included in a genetic group. The av- commonality between the Overleese and erage FW, yield, and harvest days for the five Susquehanna genetic groups. According to genetic groups was compared (Table 3). The Pomper, et al. (2010) the Susquehanna group two most prolific genetic groups in terms of is distantly similar to the Wabash and Wells average FW and harvest days were Overleese groups, not Overleese. and Susquehanna. With respect to average The ramifications of phenotypic similarity FW, the Susquehanna and Overleese groups coupled with genotypic dissimilarity is diffi- were not significantly different, whereas cult to interpret. Pawpaw characteristics have the Overleese group had significantly larger been shown to vary by cultivar. One study re-

Table 3. Number of trees, fruit weight, yield per tree, and number of days that individual fruit was har- vested (harvest days) for the pawpaw cultivars grown in Ohio in two suburban and one rural location from 2005-2012 that were characterized into one of five genetic groups by simple sequence repeat markers. Except for number of tree, values are means±SD. Genetic group Number Average fruit (# of cultivars) of trees weight (g) Yield (g) Harvest Days

Taylor & Wilson (2) 24 92dz ± 39 3747b ± 4509 16ab ± 11 Susquehanna (5) 30 137ab ± 62 7851a ± 10144 19a ± 12 Wabash (5) 30 98cd ± 36 3570b ± 6000 13b ± 13 Wells (3) 30 112bc ± 48 7449a ± 9634 16ab ± 14 Overleese (9) 119 155a ± 53 6517ab ± 8158 19a ± 13 z Means within columns followed by common letters do not differ, by Duncan’s Multiple Range test, p<0.05. 10 Journal of the American Pomological Society ported that pawpaw size, pulp and skin color, netic groups. Two individual cultivars within sugar content, and phytochemical content the Overleese genetic group (‘Overleese’ and varied in 12 cultivars (Brannan et al., 2015). ‘IXL’) have been shown to be less prone to ox- Polyphenol oxidase, the enzyme responsible idative browning due to their low polyphenol for the quick browning of pawpaw pulp, oxidase activity (Brannan and Wang, 2017) also varied by cultivar (Brannan and Wang, and may be potential cultivars of interest. 2017). However, these studies made no at- tempt to characterize the cultivars based on Literature Cited the genetic classifications from Pomper et al. Brannan, R.G., T. Peters, and S.T. Talcott. 2015. Phy- (2010). It is worth noting that two individual tochemical analysis of ten varieties of pawpaw cultivars within the Overleese genetic group (Asimina triloba L. Dunal) fruit pulp. Food Chem. 168:656-661. (‘Overleese’ and ‘IXL’) had low polyphenol Brannan, R.G. and G. Wang. 2017. Effect of Frozen oxidase activity but what is not known con- Storage on Polyphenol Oxidase, Antioxidant Con- clusively is the relationship between poly- tent, and Color of Pawpaw (Asimina Triloba [L.] phenol oxidase activity and the onset of dete- Dunal) Fruit Pulp. J. Food. Res. 6(3). riorative browning in pawpaw. Crabtree, S. and K. Pomper. 2007. Within-cluster hand- thinning increases fruit size in pawpaw (Asimina Conclusion triloba). Hortscience 42(4):881-881. Crabtree, S., K. Pomper, and J. Lowe. 2009. Utilizing Demand and consumption of exotic goods Within-cluster Hand-thinning to Increase Pawpaw such as pawpaw have grown over the last two Fruit Weight. Hortscience 44(4):1108-1108. decades, and specialty produce is one of the Crabtree, S.B., K.W. Pomper, and J.D. Lowe. 2010. fastest growing segments of the produce in- Within-Cluster Hand-Thinning Increases Fruit dustry. Pawpaw has the potential to be a prof- Weight in North American Pawpaw Asimina triloba itable addition to new and existing orchards. (L.) Dunal. J. Amer. Pomol. Soc. 64(4):234-240. Its nutritional value and tropical flavor make Galli, F., D.D. Archbold, and K.W. Pomper. 2007. Pawpaw: An old fruit for new needs, p. 461-466, it a strong candidate for both fresh and pro- Proceedings of the 1st International Symposium on cessed markets. For pawpaw to be success- Human Health Effects of Fruits and Vegetables. fully commercialized, a stronger grower base Geneve, R.L., K.W. Pomper, S.T. Kester, J.N. Egilla, must be established. C.L.H. Finneseth, S.B. Crabtree, and D.R. Layne. Research must empower growers to select 2003. Propagation of pawpaw - A review. Horttech- and grow cultivars that show strong market nology 13(3):428-433. potential based on factors such as large fruit Layne, D.R. 1996. The pawpaw [Asimina triloba (L) Dunal]: A new fruit crop for Kentucky and the Unit- size with fewer seeds, acceptable fruit flavor ed States. Hortscience 31(5):777-784. and texture, large yields per tree, and har- Marini, R.P., Fallahi, E., Francescatto, P., Lordan, J., vest length. Results from this study provide Newell, M.J., Ouellete, D., Reighard, G., Robinson, some data for growers and unmasks oppor- T.L., Wolfe, D. 2018. The NC-140 Multi-Location tunities for researchers. Some of the results Peach Physiology Trial: Relationships Between from this study were already known to ex- Peach Fruit Weight, Crop Density and Early Season Temperature. J. Amer. Pomol. Soc. In Press. perienced pawpaw growers, such as a late- Pomper, K. and D.R. Layne. 2005. The North Ameri- summer onset of harvest ranging from one can Pawpaw: Botany and Horticulture. Hort. Rev. to more than 2 months with most of the fruit 31:349-382. harvest occurring in Sept. Other results are Pomper, K.W., S.B. Crabtree, S.P. Brown, S.C. Jones, more sophisticated such as the data from this T.M. Bonney, and D.R. Layne. 2003a. Assessment study showing that average FW is normally of genetic diversity of pawpaw (Asimina triloba) distributed. Another significant finding is that cultivars with intersimple sequence repeat markers. J. Amer. Soc. Hort. Sci. 128(4):521-525. cultivars in the genetic group Susquehanna Pomper, K.W., S.B. Crabtree, D.R. Layne, and R.N. and Overleese have significantly higher av- Peterson. 2008a. Flowering and fruiting character- erage FW and harvest length than other ge- istics of eight pawpaw [Asimina triloba (L.) Du- Pawpaw 11

nal] selections in Kentucky. J. Amer. Pomol. Soc. Dutta, K. Schneider, and J. Tidwell. 2010. Charac- 62(3):89-97. terization and Identification of Pawpaw Cultivars Pomper, K.W., S.B. Crabtree, D.R. Layne, R.N. Pe- and Advanced Selections by Simple Sequence Re- terson, J. Masabni, and D. Wolfe. 2008b. The Ken- peat Markers. J. Hort. Sci. 135(2):143-149. tucky pawpaw regional variety trial. J. Amer. Po- Rogstad, S.H., K. Wolff, and B.A. Schaal. 1991. Geo- mol. Soc. 62(2):58-69. graphical Variation in Asimina-Triloba Dunal (An- Pomper, K.W., D.R. Layne, R.N. Peterson, and D. nonaceae) Revealed by the M13 DNA Fingerprint- Wolfe. 2003b. The Pawpaw Regional Variety ing Probe. Amer. J. Bot. 78(10):1391-1396. Trial: Background and early data. Horttechnology Willson, M.F. and D.W. Schemske. 1980. Pollinator 13(3):412-417. Limitation, Fruit Production, And Floral Display Pomper, K.W., J.D. Lowe, L. Lu, S.B. Crabtree, S. In Pawpaw (Asimina-Triloba). B. Torrey Bot. Club 107(3):401-408.

About the cover ʻBoysenʼ blackberry was an industry standard for nearly 50 years, but until re- cently its origin was unclear. DNA fingerprinting suggests ʻBoysenʼ is a hybrid of ʻLoganʼ × ʻAustin Mayesʼ and ʻLucretiaʼ is not a parent. Photo courtesy of Kim Hummer. 12 Journal of the American Pomological Society

Journal of the American Pomological Society 73(1): 12-21 2019 Regional Evaluation of Seven Newly Introduced Sweet Cherry Cultivars in North-East of Iran (Shahrood) N. Salehabadi1, M. Rezaei1*, A. Sarkhosh2, H. Hokmabadi3 and M. Abedini Esfahani3

Additional index words: Tree Growth, Prunus avium L., Fruit quality, Test panel, Flower phenology

Abstract We evaluated five new introduced cultivars of sweet cherry (‘Stella’, ‘Sunburst’, ‘Summit’, ‘Subima’, and ‘Germesdorfi Coln 3’) with two control cultivars (‘Siah Mashhad’ and ‘Sileg Belamarka’) in one of the main temperate fruit production regions in North-East of Iran (Shahrood) during 2014 to 2016. Tree growth, flowering phenology, pollination, fruit set characteristics, and fruit quality were assessed in ten-year-old trees. ‘Summit’ and ‘Sunburst’ had the maximum and minimum tree height and tree canopy extension, respectively. ‘Stella’ had the highest yield and received the highest score in the organoleptic test. The lowest yield was observed in ‘Sunburst’ with the highest fruit weight (7.40 g). ‘Sileg Belamarka’ had the highest percentage of fruit set when open-pollinated. Complete compatibility in a field self-pollination experiment was confirmed for ‘Sunburst’ and ‘Stella’ and relative compatibility was observed in ‘Germesdorfi Coln 3’. All the introduced cultivars especially ‘Stella’ appeared well adapted to the Shahrood climate conditions.

Introduction al., 2014); ‘Stella’, ‘Summit’ and ‘Sileg Introducing and evaluating the adaptability Belamarka’ to Meshkinshahr climate (Fathi of new cultivars in different production et al., 2013); and ‘Stella’, ‘Sunburst’, regions is one aspect of breeding. Successful ‘Summit’, ‘Subima’ and ‘Germesdorfi Coln production of new fruit tree cultivars in a 3’ to Karaj climate (Akbari et al., 2014). region requires, not only a complete survey These researchers indicated a diverse of the regional climate condition, but also adaptability of these new cultivars to the information on pollination, fruit set, and different environmental conditions. ‘Stella’ consumer preferences (Badenes and Byrne, is the first self-compatible cultivar of sweet 2012). Shahrood is one of the most important cherry that was introduced in 1968 (Lapins, regions for stone fruit production in Iran. 1971). Self-compatible cultivars which have Sweet cherry (Prunus avium L.) is one of been released to date include: ‘Symphony’, the important stone fruits which have a ‘Sweet heart’, ‘Lapins’, ‘Sunburst’, unique economic value in Iran. In 1999, five ‘Tehrani vee’, ‘White Gold’, and ‘Samba’ new sweet cherry cultivars were imported (Arzani, 2005). Self-incompatible cultivars to Iran (Arzani, 2005). These cultivars usually have less fruit set (3-5%) than self- were evaluated in three main temperate compatible cultivars (Choi et al., 2002). Self- fruit production regions in Iran including incompatibility in the most popular Iranian Mashhad, Karaj, and Meshkinshahr. Former sweet cherry cultivar was reported for ‘Siah studies confirmed adaptability of ‘Stella’, Mashhad’ (Arzani, 1988). ‘Summit’ with 11% ‘Germesdorfi Coln 3’, and ‘Sunburst’ to fruit set in a self-pollinated test was reported Mashhad climate (Ganji Moghadam et as a self-compatible cultivar (Wlodzimierz et

1 Horticulture Science Department, Agriculture Faculty, Shahrood University of Technology, Shahrood, Iran 2 Horticultural Science Department, University of Florida, Gainesville, FL 32611 USA 3 Agricultural and Natural Resources Research and Education Center of Semnan Province (Shahrood), Agricul- tural Research, Education and Extension Organization (AREEO), Shahrood, Iran 4 Corresponding author; email: [email protected] Cherry 13 al., 2008). However, in a study on pollination (longitude: 54°57′ E, latitude 36°25′ N, 1367 requirement with 13 sweet cherry cultivars m elevation). Shahrood has a cold and dry in Serbia, ‘Summit’ and ‘Germesdorfi Coln climate with an average annual temperature 3’ were introduced as self-incompatible of 14.4 °C, relative humidity of 63%, and cultivars (Radicevic et al., 2015), showing average rainfall is about 160 mm per year. that environmental conditions can influence Five newly introduced sweet cherry self-incompatibility of some cultivars. Self- cultivars from Hungary including; ‘Stella’, compatible sweet cherry cultivars with cross- ‘Sunburst’, ‘Germesdorfi Coln3’, ‘Summit’, pollination have greater final fruit set than and ‘Subima’ (Arzani, 2005) along with a self-pollinated trees (Blazková, 1996). Fruit native compatible cultivar, ‘Siah Mashhad’, set of ‘Summit’ increased from 20.7% to and a non-native compatible cultivar, ‘Sileg 38.2% in self-pollination to cross-pollination Belamarka’were grafted on ‘Mahaleb’ (Blazková, 1996). After open pollination, rootstock (Prunus mahaleb L.), and self-fertile sweet cherry seedlings produced transferred to the field in April 2006. Trees 5–10% higher yield than the self-sterile trees were planted at 5 × 4 meters in a completely (Blazková, 1996). randomized block design (CRBD) with three Genotype and environmental parameters replications, and four trees of each cultivar influence the time of anthesis and the length in each plot. All horticulture practices such of the flowering period. Effective pollination as drip irrigation, fertigation and foliar period in sweet cherry is about 4 to 5 days, applications, pruning, pest’s management, indicating that knowledge of flowering and harvesting were performed similarly on phenology in fruit trees is an essential factor all of them and trees were trained in a spindle in effective pollination and fertilization form. The cultivars were evaluated in two (Stosser and Anvari, 1983). Hosseini years, 2015 to 2016. (2009) evaluated the genetic diversity of Growth characters such as annual growth, 25 sweet cherry cultivars and based on the tree height, trunk cross-sectional area, and flowering phenology, the cultivars were canopy volume of each tree were measured divided into three groups: early, middle, and at the end of the growing season in Sept. late flowering. Environmental conditions 2015 and 2016. Trunk cross-sectional area initiate double fruiting in sweet cherry calculated from trunk diameter at 20 cm cultivars (Engin and Unal, 2004); e.g., water above the graft union. Tree canopy volume stress and high temperature at the time of was calculated as V=1.33πa2b (if height was flower initiation accelerate the formation of more than width) or V=1.33πab2 (if the width doubling fruits (Beppu and Kataoka, 1999; was more than height) formula (a =half of big Engin and Unal, 2004). diameter, and b = half of small diameter. This study aimed to evaluate growth and Flowering phenological stages of the yield parameters of five newly introduced whole tree such as flower bud break, balloon sweet cherry cultivars in comparison with stage, full bloom, end of flowering, petal two standard cultivars in the Shahrood region. fall, and fruit maturity was recorded in each Also, the flowering phenology, pollination year for all cultivars. The full bloom stage compatibility, and fruit set of these cultivars was considered when 75% of the flowers were studied, as well as an organoleptic test were open, end of flowering was recorded panel judged the fruit quality. when over 95% of the flowers opened, and when 5% of the flowers still had petals was Materials and Methods considered as a petal fall stage (Tzoner The research was conducted at the and Yamaguchi, 1999), measured on four Shahrood Agricultural and Natural branches on different sides of each tree. Resources Research and Education Centre Self-incompatibility was evaluated by 14 Journal of the American Pomological Society the level of fruit set in self-pollination (SP) traits assessment. and open-pollination (OP) tests. Four shoots Yield (kg per tree) and fruit weight (average with 100 to 150 flowers approximately of 20 fruit per tree) were recorded annually. around the canopy for each tree were used Fruit length and diameter were measured for each pollination test. SP was measured with a digital caliper. Fruit firmness was by determining the rate of self-fertility in measured with a Penetrometer (Wagner FT isolated emasculated and non-emasculated 30, Greenwich, CT, USA) and total soluble (intact) flowers. To prevent unwanted solids concentration (TSS) of fruit juice pollination in both tests, shoots were isolated obtained from 100 g commercially ripened by covering shoots with cotton bags before fruits was measured with a refractometer the balloon stage. For the SP test, flowers (ATAGO master 5EM, Japan). Total acidity were emasculated by removing petals and (TA) of fruit juice was also assessed by anthers at the balloon stage, and then shoots titration NaOH (0.1 N). were bagged. Flowers were pollinated Data for tree characteristics were analyzed with a pen using pollen collected from the as a randomized complete block design with same tree. This operation was repeated 24 SAS 9.4 software (SAS/STAT Software , hours later to ensure adequate pollination. 2014). The experimental unit was the four- For the SP test, flowering shoots with non- tree plot for each cultivar randomized within emasculated (intact flower) flowers were each of the three blocks. Means for the four- bagged before flowers opened. After bloom, tree plots were analyzed with SAS’s Proc to ensure pollination, a fine brush was used Glimmix, where block was specified as a to transfer pollen from the stamens to the random effect, and LSmeans were comparing stigmas of the same flowers. Four non-treated with Tukey’s test. Data from evaluations shoots per tree were used to evaluate fruit set for fruit characteristics were analyzed as of open-pollinated shoots. The number of a split-plot, where year was the whole-plot harvested fruit per shoot was used calculate and cultivar was the split-plot. When the the percentage of flowers that set fruit. year×cultivar interaction was significant, Ten trained panelists, five males, and five cultivar LSmeans were compared within females, between 25 to 40 years, were asked each year with the Slicediff adjustment. to evaluated fruit quality. All the panelists assessed three fruits per cultivar labeled Results and Discussion with a code with three replications. Panelists The results of data analysis (P-value) evaluated the size, color, appearance, texture, were presented in Table 1. The cultivars had aroma, flavor, taste, and overall acceptability significantly different canopy characteristics by scoring each index on a scale from 1 to (Table 1). ‘Subima’ trees were the largest, 9 (1= lowest quality and 9= best quality). with a volume of 17.07 m3, but did not Coffee powder was used as blind for quality differ from ‘Summit’ and ‘Stella’ (Table 2).

Table 1. P-values from analysis of variances for various characteristics of trees and fruit for seven sweet cherry cultivars in two years. Pr > F Effect DF CVz TCSA TH FSOP FSIF FSEF FL FW FEW FWR FF DF TSS TA TSS/TA Year 1 0.0081 0.6332 0.0232 0.0400 0.0030 0.0023 0.1995 0.0879 0.0935 <.0001 0.0002 0.2165 0.1414 <.0001 <.0001 Cultivar 6 0.0378 0.0003 0.0106 <.0001 <.0001 <.0001 0.2007 0.0154 0.0030 0.0080 0.0005 <.0001 0.0105 0.0607 0.0719 year*cultivar 6 0.8321 0.8613 0.6782 0.0002 <.0001 <.0001 0.0021 0.0002 <.0001 0.0017 0.0052 0.0019 0.5937 0.1464 0.1150

Z Canopy volume (CV), trunk Cross-section area (TCSA), Tree height (TH), Final fruit set after open pollination (FSOP), fruit set after Self pollination of intact flower (FSIF), fruit set after Self pollination of emasculated flower (FSEF), Fruit length (FL), Fruit width (FW), Fruit weight (FEW), Fruit to stone weight ratio (FWR), Fruit firmnes (FF), Double fruiting (DF), Total soluble solids concentration (TSS), titratable acidity (TA) and TSS/TA ratio. Cherry 15

Table 2. Canopy volume, trunk cross sectional area (TCSA), and tree height of 11 year-old trees of seven sweet cherry cultivars in Shahrood, Iran. Cultivar name Canopy volume (m3) TCSA (cm2) Tree height (cm) Germesdorfi Coln3 12.87 abz 93.11 cd 275.9 b Sileg Belamarka 12.49 b 86.65 d 287.4 b Subima 17.07a 133.51 b 318.3 a Siah Mashhad 11.15 b 94.07 cd 291.5 ab Summit 16.5 a 109.80 c 340.4 a Sunburst 9.38 b 105.06 c 267.3 b Stella 16.05 a 159.82 a 313.5 ab Z Means within columns followed by common letters do not differ at the 5% level of significance, by Tukey’s test.

‘Sunburst’ trees were the smallest, with an cultivars. For most cultivars, trees started average canopy volume of 9.38 m3 and were blooming earlier in 2016 than in 2015 bloom similar to ‘Siah Mashhad’ (Table 2). The ranged from 3 days in ‘Sileg Belamarka’ to 9 starting date of the fruit-bearing in ‘Sunburst’ days in ‘Sunburst’. Seasonal environmental was earlier than other cultivars (Data not changes and weather conditions influence shown) and this may be one of the reasons date of flowering (Tooke and Nicholas, for low canopy volume of this cultivar. 2010). Based on the number of days from ‘Stella’ had the largest trunks, whereas anthesis to end of flowering the cultivars cultivars with the smallest trunks included were divided into three groups that can be ‘Sileg Belamarka’, ‘Germesdorfi Coln3’ and used as pollinizer for each other: 1) early ‘Siah Mashhad’ (Table 2). The tallest and included ‘Germesdorfi Coln3’ and ‘Siah shortest trees were ‘Summit’ and ‘Sunburst’, Mashhad’, 2) middle included ‘Subima’, respectively (Table 2). ‘Summit’, ‘Subima’ ‘Sileg Belamarka’, and 3) late included and ‘Stella’ trees were the tallest (Table 2). ‘Summit’ and ‘Stella’. Brozik (1971) showed A cursory investigation of flower that the optimum blooming duration in morphological characteristics showed cherry is 10–14 days, and for incompatible that all cultivars had similar numbers of cultivars, at least 4–6 days overlapping was stamen, pistils, sepals, and petals. Flower necessary for good pollination and fruit set. phenological stages including swelling Also, early cultivars cannot be recommended flower buds, tip green stage, balloon stage, as pollinators for late cultivars (Kiris, 1992), full bloom, end of flowering, petal fall stages e.g. ‘Stella’ and ‘Summit’ cannot be used as and leaf initiation in each cultivar were pollinizers for the early flowering cultivars, presented as a schematic (Fig. 1). For most ‘Germesdorfi Coln3’ and ‘Siah Mashhad’. cultivars, the bud break stage lasted 2 to 3 Fathi et al. (2013) studied the adaptation of days, and balloon stage lasted 3 to 4 days, but these cultivars in the Meshkinshar region in cultivar differences were not significant in Iran and recommended ‘Stella’ and ‘Summit’ the two years of study. The time of anthesis as late-flowering cultivars, which agree with varied with cultivar (Fig. 1). The difference our results. However, they indicated that between early- and late-blooming cultivars ‘Sileg Belamarka’ and ‘Subima’ bloomed was about six days in 2015 and eight days in early which conflicts with our results in 2016. The earliest flowering cultivars in both Shahrood. ‘Sileg Belamarka’ and ‘Summit’ years were ‘Siah Mashhad’ and ‘Germesdorfi was reported as early and late flowering Coln3’, and the latest flowering cultivars cultivars in Mashhad climate (Iran), were ‘Stella’ and ‘Summit’ cultivars (Fig. respectively (Ganji Moghadam et al., 2014). 1). Petal abscission lasted 4 to 6 days for all In the Shahrood region, ‘Germesdorfi

16 Journal of the American Pomological Society

March April

24 25 26 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Year 2015 Siah Mashhad Year 2016

Year 2015 Germesdorfi Coln3 Year 2016

Year 2015 Sunburst Year 2016

Year 2015 Sileg Belamarka Year 2016

Year 2015 Subima Year 2016

Year 2015 Summit Year 2016

Year 2015 Stella Year 2016

¢ Swelling flower buds ¢ Tip green stage ¢Balloon stage ¢ Full bloom ¢End flowering ¢ Petal abscission ¢ Leaf initiation Figure 1. FigureFlowering 1. Flowering phenological phenological stages stages(Webster (Webster and Looney,and Looney, 1996) 1996 for) for seven seven sweet sweet cherry cherry cultivars in Shahrood, Iran. cultivars in Shahrood, Iran

Coln3’ and ‘Siah Mashhad’ fruit ripened was no significant difference in the early on 20 June 2015 and 9 June 2016, whereas stages of flower bud development, swelling other cultivars ripened on 6 June 2015 and bud and green tip stages. However, there 3 June 2016. However, in Mashhad ‘Sileg was a significant difference between the Belamarka’ required 43 days for fruit beginning of flowering among genotypes maturity and was introduced as an early from 23 March in SH4 genotype to 2 April cultivar and ‘Siah Mashhad’ required 73 in SH21 genotype and petal fall stage for all days for fruit maturity was introduced as a genotypes was six days. late ripening cultivar (Ganji Moghadam et Cultivars significantly influenced fruit 16 al., 2014). Evaluation of phenological stages set for SP (artificial and natural SP) and OP indicated obvious differences among studied (Table 2). For all cultivars, OP limbs had cultivars, which confirmed other research on significantly higher fruit set than SP limbs. phenology in sweet cherry (Hossieni 2009; For the OP treatment, ‘Stela’ had the lowest Ahmadi Moghdam et al., 2012). Ahmadi fruit set in 2015, but not in 2016 (Table 3). Moghdam et al., (2012) studied phenological ‘Sileg Belamarka’, ‘Siah Mashhad’, and flowering stage of 13 sweet cherry genotypes ‘Summit’ had the highest fruit set in both obtained from ‘Siah Mashhad’ and there years. The same results were reported by Cherry 17

Table 3. Final fruit set (%) for open pollinated (OP) and self-pollinated (SP) flowers of seven sweet cherry cultivars in Shahrood in 2015 and 2016. Fruit set at Harvest (%) SP of SP of intact Year Cultivars OP emasculated flower flower Germesdorfi Coln3 50.44 bz 3.74 c 4.67 c Sileg Belamarka 60.93 a 0.00 d 0.00 d Subima 54.12 ab 0.17 d 0.00 d 2015 Siah Mashhad 45.28 b 0.63 d 0.95 d Summit 50.68 b 0.00 d 0.00 d Sunburst 55.37 ab 16.08 a 14.61 a Stella 17.14 b 12.12 b 11.72 b Germesdorfi Coln3 56.65 ab 8.67 c 8.89 c Sileg Belmarka 65.17 a 0.16 d 0.00 d Subima 56.99 ab 1.48 d 1.88 d 2016 Siah Mashhad 64.71 a 0.00 d 0.95 d Summit 61.56 ab 0.34 d 0.64 d Sunburst 55.46 b 32.06 a 23.57 b Stella 58.94 ab 17.62 b 29.90 a z LSmeans within years and columns followed by common letters are not significant at the 5% level, by the SLICEDIFF adjustment.

(Akbari et al., 2012; Fathi et al., 2013), kg/tree for ‘Siah Mashhad’ (Akbari, 2012), where ‘Summit’ had the highest percentage which was not confirmed by our results. In of fruit set. For both years the SP treatments addition to the genotype, yield is influenced had less than 3% fruit set for ‘Subima’, ‘Sileg by other factors such as management system, Belamarka’, ‘Summit’, and ‘Siah Mashhad’, nutritional conditions, climate conditions, which are self-incompatible. ‘Sunburst’ and tree age and biotic and abiotic stress. (Iezzoni ‘Stella’ had more than 5% self-pollinated et al., 1990). However, yield performance for fruit set and are complete self-compatible tree fruit cultivars requires several years of cultivars, and ‘Germesdorfi Coln 3’ is a data. relative self-compatible cultivar (Table 3). In the judgment of panelists, ‘Stella’ had Similar results were reported by (Akbari et the highest overall acceptance and ‘Siah al., 2012; Ganji Moghadam et al., 2014). ‘Sunburst’ with 39.4% fruit set in an isolated SP test was considered as self-compatible Table 4. Yield of seven newly introduced sweet in Turkey climate (Sutyemez, 2011) which cherry cultivars years 2015. coincides with our results. These results were Cultivar Yield (kg/tree) also confirmed by (Fathi, 2000; Choi et al., Germesdorfi Coln3 21.66 dz 2002), however in their study most sweet Sileg Belamarka 38.33 bc cherry cultivars were self-incompatible. Subima 36.66 c ‘Stella’ had the highest yield, with 55 kg/ Siah Mashhad 25.00 d tree, and the lowest yield of 15 kg/tree was Summit 46.66 b recorded for ‘Sunburst’ (Table 4). Over two Sunburst 15.00 e years in a study in Karaj, Iran the highest Stella 55.00 a yield was recorded for ‘Sileg Belamarka’ z Means followed by common letters do not differ a the 5% with 21.8 kg/tree, and the lowest yield was 6.4 level of signficance, by Tukey’s test. 18 Journal of the American Pomological Society

Table 5. Organoleptic ratings for seven sweet cherry cultivars based on test panels in 2015. Cultivar Juicinessz Sizey Colory Texturex Aroma Flavorv Overall and Odorx Acceptanceu Germesdorfi Coln3 5.37 at 5.04 b 6.70 a 5.06 b 4.66 ab 5.45 bc 6.08 b Sileg Belamarka 5.72 a 6.18 ab 6.82 a 3.96 c 4.60 ab 5.70 b 5.88 bc Subima 5.53 a 5.83 ab 6.93 a 4.48b c 4.76 ab 5.65 b 6.05 b Siah Mashhad 4.45 b 3.15 c 5.15 b 4.50b c 3.25 b 3.85 c 4.20 c Sunburst 5.36 a 5.13 b 5.76 b 5.03 b 4.40 ab 4.56 bc 5.20 b Stella 5.93 a 7.23 a 7.10 a 6.33 a 5.36 a 6.30 a 7.13 a Summit 5.85 a 5.65 ab 4.40 c 3.90 c 5.1 a 6.87 a 5.75 b t Means within columns followed by common letters do not differ at the 5% level of significance, by Tukey’s test. z 1-5 low water - juicy x 1-9 very soft - crisp v 1-9 bad taste - delicious y 1-9 weak - excellent w 1-9 weak - excellent u 1-9 inedible - excellent

Mashhad’ had the lowest (Table 5). In a weight to fruit stone weight, and double test panel with the same cultivars, ‘Stella’ fruiting was revealed significant differences and ‘Sunburst’ had the highest acceptance among cultivars (Table 6). ‘Sileg Belamarka’ (Akbari, 2012). All of the introduced and ‘Subima’ had the most double fruit cultivars received higher scores for juiciness, both years and ‘Germezdorfi Coln3’, ‘Siah size, aroma, and aroma and flavor than the Mashhad’ and ‘Stella’ the lowest percentage control cultivar ‘Siah Mashhad’. ‘Summit’ of double fruits (Table 6).‘ Low and high received the lowest score for fruit texture and fruit doubling in sweet cherry cultivars was colour (Table 5). reported by other researchers (Bouzari and Fruit parameters such as fruit length, Arzani, 2006). Fruit doubling in ‘Stella’ was diameter, weight, the ratio of fruit pulp 2% in 1977 and 29% in 1980 in southern

Table 6. The length, width, weight, firmness and the ratio of fruit weight to stone weight of seven sweet cherry cultivars in 2015 and 2016. Cultivar Year Fruit length Fruith width Fruit wt. Fruit wt. Firmness Double (mm) (mm) (g) /Stone wt. (kg/cm2) fruiting % Germesdorfi Coln3 2015 18.75 bz 16.97 c 4.83 c 13.06 c 1.18 b 1.39 d Sileg Belamarka 20.35 ab 18.29 bc 5.58 bc 20.40 a 1.24 ab 8.65 a Subima 20.53 a 18.15 bc 5.63 bc 16.34 bc 1.16 b 7.35 b Siah Mashhad 19.11 ab 17.03 c 4.25 c 14.41 c 1.12 b 1.18 d Sunburst 19.01 ab 18.06 bc 4.68 c 14.60 c 1.32 ab 2.74 c Stella 20.66 a 20.14 a 6.75 a 19.06 ab 1.38 a 0.99 d Summit 20.64 a 18.78 b 5.94 b 17.56 b 1.24 ab 2.62 c Germesdorfi Coln3 2016 21.26 ab 20.26 ab 6.53 b 12.06 ab 2.20 a 1.18 e Sileg Belamarka 20.30 ab 19.21 b 5.33 c 11.61 b 1.55 d 5.33 b Subima 21.83 ab 19.53 b 6.80 ab 14.04 a 1.88 bc 6.56 a Siah Mashhad 20.46 ab 19.89 ab 5.60 c 9.80 b 1.79 c 1.37 e Sunburst 22.08 a 21.26 a 7.40 a 14.12 a 2.00 b 4.48 d Stella 19.65 b 19.51 b 5.69 c 11.68 b 2.11 ab 1.11 e Summit 18.59 b 17.34 c 4.95 c 10.37 b 1.90 bc 1.27 e z Means within years and columns followed by common letters do not differ at the 5% level of significance, by the SLICE adjustment. Cherry 19

San Joaquin valley of California (Mike et Table 7. Total soluble solids concentration (TSS), al., 1983) Temperatures above 28 °C and titratable acidity (TA) and TSS/TA ratio of fruits of drought stress during the time of flower bud seven sweet cherry cultivars grown in Shahrood, initiation increased double fruiting in peach Iran. (Johnson et al., 1992). Cultivar TSS TA TSS/TA In 2015 ‘Germesdorfi’, ‘Sunburst’, Germesdorfi Coln3 19.48 abz 0.59 a 34.13 b and ‘Siah Mashhad’ had the lowest fruit Sileg Belamarka 20.08 ab 0.59 a 36.70 b length and diameter, but results in 2016 Subima 20.47 a 0.52 b 44.72 a were quite different. Akbari et al. (2014) Siah Mashhad 18.88 b 0.58 a 35.73 b studied the same cultivars in Karaj, Iran Sunburst 19.12 ab 0.60 a 38.81 b and reported the highest and lowest fruit Stella 19.33 ab 0.56 ab 40.12 b length for ‘Sunburst’ and ‘Sileg Belamarka’, Summit 18.93 b 0.45 c 43.13 ab respectively, whereas ‘Stella’ had the highest z Means within columns followed by common letters do not fruit diameter and ‘Subima’ had the lowest differ at the 5% level of signficance, by Tukey’s test. in one year. Fruit weight was not consistent both years. Fruit weight was highest in 2015 for ‘Stella’ and for ‘Sunburst’ in 2016. The Conclusions big fruit of ‘Sunburst’ was also reported by The evaluation of seven sweet cherry others researchers (Radicevic et al., 2008; cultivars in Shahrood, Iran indicated that Dzhurinov, 2009). The ratio of fruit pulp to ‘Stella’, ‘Summit’, and ‘Subima’ produce stone weight ranged between 14.18% for larger trees than the other cultivars. There ‘Siah Mashhad’ to 24.67% for ‘Sunburst’ were obvious differences in flowering (Table 6). The maximum and minimum of phenological stage, and based on bloom fruit firmness were recorded for ‘Germesdorfi dates, the cultivars were divided into three Coln3’ in 2016 and ‘Siah Mashhad’ in 2015, groups, which can be used to pollinate for respectively (Table 6). TSS, TA and TSS/ each other within a group. Pollination tests TA were affected significantly by cultivars indicated that ‘Subima’, ‘Sileg Belamarka’, (Table 1). ‘Sublime’ had highest TSS ‘Summit’, and ‘Siah Mashhad’ were self- which was significantly higher than ‘Siah incompatible.‘Sunburst’ and ‘Stella’, with Mashhad’ and ‘Summit’ (Table 7). The TA more than 5% self-pollinated fruit set, of ‘Summit’ fruits was significantly lower were complete self-compatible cultivars, than other cultivars. ‘Subima’ had higher and ‘Germesdorfi Coln3’ is a relative fruit flavor index, TSS/TA ratio, than the self-compatible cultivar. In two years other cultivars except ‘Summit’ (Table 7). of evaluation, high double fruiting was TSS in sweet cherries is mostly dependent observed in ‘Sileg Belamarka’; ‘Subima’ on environmental conditions (Akbari et had more double fruit than other’s cultivars al., 2014). TSS in sweet cherry fruit ranges and ‘Germezdorfi Coln3’, ‘Siah Mashhad’ between 11 and 25%, mainly due to glucose and ‘Stella’ had low double fruiting. ‘Stella’ and fructose, and less to the presence of had the highest overall acceptance rating. sucrose and sorbitol and TSS is a cultivar- ‘Sunburst’, ‘Subima’, and ‘Stella’ had the dependent parameter in sweet cherry largest fruits with high fruit flavor rating. (Martinez-Romero et al., 2006). Acidity In general, all the introduced cultivars changes are low during fruit maturation, and were well adapted to Shahrood climate. ripening of sweet cherry (Akbari et al., 2014) However, based on tree growth, pollination and the malic acid in sweet cherry is much requirement, yield and fruit quality, ‘Stella’ less than other of temperate fruits (Crisosto is recommended for cultivation in Shahrood, et al., 2002). Iran. 20 Journal of the American Pomological Society

Literature Cited the formation of double fruit in sweet cherry. J. Ege Ahmadi, M. H., M. E. Ganji and S. Akhavan. Faculty Agri. 41: 19-28. 2012. Evaluation of quantitative and qualitative Fathi, H. 2000. Germination of sweet cherry hybrids characteristics of some selected sweet cherry in outdoor and laboratory. Isla. Azad Univ., Sci. and genotypes. Seed Plant Impro. J. 28 (1):187– 200. Res. Branch, Tehran, Irn. MSc. Thesis. Akbari, A. 2012. Morphological and pomological Fathi, H., U. Jahani and N. Bouzari. 2013. Evaluation studies of some sweet cherry cultivars based on adaptability and comparison quantitative and testing distinctness. Uniformity and sustainable. qualitative traits of new sweet cherry cultivars Univ. of Zanjan, Irn. MSc. Thesis. Pp. 168. under Meshkinshahr environmental condition. Plant Akbari, A., M.E. Amiri, and K. Arzani. 2012. Production Tech. 14: 29-42. Comparison open pollination and fruit set isolated Ganji Moghadam, E., N. Bouzari, M. Momeni and A. on some sweet cherry cultivars. The first national Asghar Zadeh. 2014. Effect of Pollination on Fruit conferences on strategies to achieve sustainable Set and Evaluation of Phenological, Pomological development in agriculture, natural resource and and Morphological Characteristics of some environment. Pp. 1-7. Introduced Sweet Cherry Cultivars Under Khorasan Akbari, A., N. Bouzari, M. E. Amiri, and K. Arzani. Razavi Province. J. Hort. Sci. 4: 785-806. 2014. Evaluation of pomological some new sweet Hosseini, P. 2009. Evaluation of Genetic Diversity cherry cultivars in Karaj weather conditions. J. in some Iranian sweet cherry cultivars. Isla. Azad Agri. 46: 17-25. Univ., Sci. and Res. Branch, Tehran, Irn. MSc. Arzani, K. 1988. Selection of the best pollinizer for Thesis. Pp. 168. ‘Siah Mashhad’ sweet cherry. University of Tehran, Iezzoni, A., H. Schmidt and A. Albertini. 1990. Irn. MSc.Thesis. Pp. 218. Cherries.In: Genetic resources of temperate fruit Arzani, K. 2005. Introduction, propagation, quarantine and nut crops. Acta Hort. 209: 109-173. inspection and evaluation of compatibility studies Johnson, R. S., D. F. Handley and T. M. DeJong. 1992. of sweet cherry cv. Stella in Iran. Proc. of the 4th Long-term response of early maturing peach trees Iranian Cong. Hort. Sci. 8-10 Nov., Mashhad, Irn. to postharvest water deficit. J. Amer. Soc. Hort. Sci. Badenes, M. L., and D. H Byrne. 2012. Fruit Breeding 117: 881-886. Series: Handbook of Plant Breeding, Springer, Vol. Kiris, N. 1992. Dalbasti kirazinin (Prunus avium cv. 8. Pp. 875. Dalbasti) pomolojik o zellikleri ve do¨ lleyicilerin Beppu K. and I. Kataoka. 1999. High Temperature tespiti u zerinde bir arastirma (Yu ksek Lisans Tezi) Rather Than Drought Stress is Responsible for The Ege U niversitesi Fen Bilimleri Enstituu su Bahce Occurrence of Double Pistil in ‘Satohishiki’ Sweet Bitkileri Ana Bilim Dali, Bornova, I´ zmir. 51. Cherry. Sci. Hort. 81: 125-134. Lapins, K. O. 1971. Stella, a self-fruitful sweet cherry. Blazková, J. 1996. Inheritance of self-fertility in sweet Can. J. Plant Sci. 51: 252-253. cherry (from Stella cultivar) and productivity of Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. seedlings. Acta Hort. 423: 125-134. Wolfinger, and O. Schabenberger. 2006. SAS for Bouzari, N., and K. Arzani. 2006. Fruit doubling of mixed models. 2nd ed. SAS Inst. Inc. Cary, NC sweet cherry; genotype and temperature effects. Martinez-Romero, D., Alburquerque, J. M., Valverde, 27th Intentional Horticulture Congrgrasess (IHC). F., Guillén, S., Castillo, D. and M. Valero Serrano. 13-19. 2006. Postharvest sweet cherry quality and safety Brozik S. 1971. Fertility Conditions of Major Cherry maintenance by Aloe vera treatment: a new edible Varieties. Evifobb Kutatasi Jelentesei, Budapest. coating. Postharvest Biol. Tech, 39: 93-100. Choi, C. H., R. Tao and R. L. Andersen. 2002. Micke, W., J. Doyle and J. Yeager. 1983. Doubling Identification of self-incompatibility alleles and potential of sweet cherry cultivars. California Agr. pollen incompatibility groups in sweet cherry 37(3): 24-25. by PCR based S-allele typing and controlled Radicevic, S., O. Mitrovic and I. Glisic. 2008. pollination. Euphytica 123: 9-20. Pomological characteristics and biochemical Crisosto, C. H., G. M.Crisosto and M. A. Ritenour. fruit composition of some Canadian sweet cherry 2002. Testing the reliability of skin color as an cultivars. Acta Hort. 795: 283-286. indicator of quality for early season ‘Brooks’ Radicevic, S., S. Maric and R. Cerovic. 2015. S-allele (Prunus avium L.) cherry. Postharvest Biol. Tech. constitution and flowering time synchronization 24: 147-154. preconditions for effective fertilization in sweet Dzhurinov, V. K. 2009. Fruit bearing habit of nine cherry (Prunus avium L.) Orchards. Roma Biotech sweet cherry cultivars. Acta Hort. 814: 245-250. Lett. 20: 10997-11006. Eng, H. and A. Unal.2004. Effects of water deficits on SAS/STAT Software, Version 9.4 (2014). Cary, NC: Cherry 21

SAS Institute, USA. www.sas.com/en_us/software/ Tzoner, R. and M. Yamaguchi. 1999. Investigations sas9.html on some far-east prunus species, phenology. Acta Stosser, R. and S. F. Anvari. 1983. Pollen tube growth Hort. 488: 239-242. and fruit set as influenced by senescence of stigma, Webster, A. D. and N. E. Looney. 1996. Cherries, Crop style and ovules. Acta Hort. 139: 13-22. Physiology, Prod. and Uses Cambridge Univ. Press, Sutyemez, M. 2011. Pollen quality, quantity and fruit Cam.e, UK..356 pp. set of some self-compatible and self-incompatible Wlodzimierz, L., M. Malodobry, E. Dziedzic, M. cherry cultivars with artificial pollination. Afr. J. Bieniasz and S. Doniec. 2008. Biology of sweet Biotechnol. 10: 3380-3386. cherry flowering. J. Fruit. Ornam.Plant Res. 16: Tooke, F. and H. B. Nicholas. 2010. Temperate 189-199. flowering phenology. J. Exp. Bot. 61: 2853-2862. 22 Journal of the American Pomological Society

Journal of the American Pomological Society 73(1): 22-37 2019 An Analysis of Strawberry (Fragaria ×ananassa) Productivity in Northern Latitudinal Aquaponic Growing Conditions Marie Abbey1, Neil O. Anderson2, Chengyan Yue3, Gianna Short1, Michele Schermann4, Nicholas Phelps5, Paul Venturelli5 and Zata Vickers6 Additional index words: Perca flavescens(yellow perch), Carassius auratus (goldfish), Oreochromis spp. (tila- pia), Cyprinus carpio (koi)

Abstract Aquaponics, the combination of hydroponics and aquaculture into one growing system, is a controlled envi- ronment production system that potentially has increased environmental and consumer benefits over traditional production methods. Typical horticulture aquaponic production focuses on leafy greens and herbs with no known studies on the production potential of strawberries (Fragaria ×ananassa Duchesne) or any other perennial fruit crop. This study compares day-neutral strawberry yield of ‘Albion’, ‘Evie 2’ and ‘Portola’ in aquaponic produc- tions with different variables of strawberry yield in greenhouse production using soilless medium. There was no addition of supplemental nutrients or pollinators to the systems in order to evaluate the differences between treatments. We found a significant difference among cultivars in number of fruit, fresh fruit weight, and dry fruit weight with ‘Evie 2’ having the highest yield in all. There was no significant difference in the number of fruit produced by strawberries grown in soilless medium and those grown aquaponically. We did, however, find that aquaponic strawberries had a significantly higher fresh fruit weight while strawberries grown in soilless medium had a significantly higher dry fruit weight. This indicates that strawberries grown in soilless medium had a higher mass to water ratio, although aquaponic-grown strawberries sometimes had higher fresh weight yield.

Aquaponics is the integration of hy- waste which is converted into plant available droponics (the soilless growing of plants) nutrients (nitrates) by nitrifying bacteria in and aquaculture (the raising of fish) into a a biofilter and taken up by the plants (Diver closed-loop, recirculating system (Rakocy and Rinehart, 2006). et al., 2006). The fish waste provides a nu- Aquaponics developed as a way to control trient source for the plants after processing waste water from recirculating aquaculture through a biofilter while the plants provide a systems (RAS; Costa-Pierce et al., 1997). natural “filter” utilizing the resulting nitrate Though RAS has many benefits, water con- N and other nutrients (Rakocy et al., 2006). servation is not one of them and there was The only nutritional input (food or nutrients) a demand to develop a cost effective filtra- into the system is fish food. Fish then excrete tion system. Plants can be used as filters for

1 Graduate Research Assistant; Department of Horticultural Science, University of Minnesota, 286 Alderman Hall, 29170 Folwell Avenue, St. Paul, MN 55108 2 Professor; Corresponding author email: [email protected]; Department of Horticultural Science; University of Minnesota, 286 Alderman Hall, 2970 Folwell Avenue, St. Paul, MN 55108 3 Associate Professor; Department of Applied Economics and Department of Horticultural Science, University of Minnesota, 458 Alderman Hall, 2970 Folwell Avenue, St. Paul, MN 55108 1 Graduate Research Assistant; Department of Applied Economics, University of Minnesota, 213 Ruttan Hall, 2994 Buford Avenue, St. Paul, MN 55108 4 Research Scientist, Research 5; Department of Horticultural Science, University of Minnesota, 286 Alderman Hall, 2970 Folwell Avenue, St. Paul, MN 55108 5 Assistant Professor; Department of Fisheries, Wildlife and Conservation Biology, University of Minnesota, 135 Skok Hall, 2003 Upper Buford Circle, Suite 135, St. Paul, MN 55108 6 Professor; Dept. of Food Science and Nutrition, Univ. of Minnesota, 1334 Eckles Ave., St. Paul, MN, USA Strawberry 23

RAS wastewater and also as a secondary mum 30cm tall) and spread both sexually by crop which led to the development of closed seed and vegetatively via stolons (Vincent et loop aquaponic systems (Rakocy et al., 2006; al., 1990). Flowers grow in clusters on in- Lewis, 1978). There was also a motivation dividual stalks to an even height or slightly from hydroponic growers to develop more above the foliage and bloom successively cost effective and environmentally friendly (Cold Climate Strawberry Farming, 2014). nutrient solution sources (Rakocy et al., The first bud to flower is called the ‘king 2007). flower’ and is significantly larger than sub- Aquaponic production is the fastest grow- sequent flowers on the cluster; resultant fruit ing sector of agriculture (Kloas et al., 2015), from this are termed ‘king berries’. Subtend- in part, due to pressure from population ing lateral fruits are smaller in size. Flowers growth, drought, and increased water de- have both stamens and pistils and are able to mand (Hundley and Navarro, 2013). Total self-pollinate although complete pollination aquaponic industry growth worldwide is ex- and fruit fill requires additional stimulation pected to exceed 10% by 2020 (Aquaponic besides wind (Vincent et al., 1990). The fruit Farming: Global Market Intelligence, 2016). is an aggregate accessory fruit and is formed Urban agriculture, including aquaponics, is from the receptacle, which holds the ovary; continuing to grow with over 100 million the sum of these receptacles forms the straw- growers estimated (Eigenbrod and Gruda, berry fruit. 2015). Strawberries are classified by their photo- Most aquaponic producers are considered periodic response into three categories. June- small farms both in size and revenue (Love bearing cultivars need short day lengths (<12 et al., 2014). Aquaponic systems can vary hr) to initiate flower buds, ever-bearing (re- dramatically in scale, system design, plant montant) types need long day lengths (>12 crops, fish species and management -proce hr) to initiate flower buds, and day-neutral dures. The type of system ultimately chosen cultivars are not affected by day length. depends on the location, production goals, There is also a temperature effect on flow- market demand, and many other factors al- ering response. Bradford et al. (2010) found though the vast majority are in controlled en- that non-remontant flowering ‘Honeoye’ had vironments such as greenhouses (Love et al., a photoperiod-insensitive temperature range 2015); warehouses have also been used. The of 14-20 °C whereas remontant types had a most common commercial aquaponic system higher range of 23-26 °C. Whenever tem- is called deep water culture (DWC), where peratures surpassed this range, genotypes the plants and fish are physically separated required either short or long days to flower and the plants grow on floating rafts with regardless of remontancy. June-bearing their roots suspended in the nutrient-rich wa- strawberries produce only one large crop of ter (Taiz, 2010). fruit in the spring and are currently the most Commercial aquaponic growers also use popular choice for northern latitude farmers ebb and flow systems in which the plant because of their ability to overwinter (Cold roots are intermittently submerged in water, Climate Strawberry Farming, 2014). How- though it is less common than DWC. The ever, day-neutral cultivar popularity has been other types of systems, e.g. aeroponic and growing in northern latitudes with the in- nutrient film techniques, are rarely used in crease in season extension (high tunnels and aquaponic production because of issues with low tunnels) and climate controlled produc- solids clogging the system (Søberg, 2016). tion (Petran et al., 2017). The day-neutral, re- Strawberries (Fragaria ×ananassa, Rosa- montant ‘Portola’ and ‘Albion’ strawberries ceae) are an herbaceous, perennial crop that are recommended for outdoor production due grow relatively close to the ground (maxi- to their high yields, large fruit size, and sweet 24 Journal of the American Pomological Society fruit (Petran et al., 2017) while ‘Evie 2’ has as used in hydroponic systems and a bio- performed well in northern hydroponic trials based liquid nutrient source (similar to aqua- (Wortman et al., 2016). ponic systems) were compared, strawberries The United States is the largest producer of grown with the synthetic nutrient source had strawberries in the world with the largest pro- 15% higher yield (Wortman et al., 2016). duction in California and Florida (Morgan, Year-round aquaponic production studies 2015). As consumer demand continues to have focused on the most popular aquaponic grow so does the opportunity for controlled crops, leafy greens and herbs (Love et al., climate strawberry production (USDA Eco- 2014). There has been limited research on nomic Researh Service, 2016). It is estimated year-round production of perennial crops, that increased offseason production could ex- such as strawberries, in aquaponic produc- pand the US strawberry industry by $520 M tion particularly in northern latitudes. The annually (Arnade and Kuchler, 2015). objective of this study was to produce day- The majority of strawberries are field- neutral strawberries in year-round aquaponic grown, although recently high tunnels are production systems in northern latitudes to being used for production, particularly in create a baseline for potential yield. The hy- coastal areas of California and to extend potheses tested in this experiment include: the season in cooler climates (Poppe et al., Ho: There is no difference in yield between 2016; Pritts and Mcdermott, 2017). Hydro- strawberries produced aquaponically and ponic strawberry production had comparable those grown in soilless medium. Further- yields to field production (Wortman et al., more, Ho: There is no difference in yield 2016) and hydroponic strawberry runner between strawberries grown with different production (production of bare-root straw- aquaponic treatments. berry plants) was as effective as field pro- duction (Takeda and Hokanson, 2003). Us- Materials and Methods ing aquaponic systems to grow strawberries Genotypes Tested. has not been investigated to the best of our This experiment was conducted for a knowledge. There has been speculation that 13-month period (Jan. 2016-Feb. 2017). aquaponic growing methods could be used to Three cultivars of day neutral strawberry produce strawberry runners (Mattner et al., plants were used for this experiment: ‘Por- 2017) but no proof of concept has been at- tola’, ‘Albion’, and ‘Evie 2’, based on previ- tempted. ous recommendations (Wortman et al., 2016; Adequate nutrient levels are critical for Petran et al., 2017), previous winter cultiva- strawberry fruit maturation and achieving tion studies (Paparozzi et al., 2010; Petran market fruit sizes. The recommended liquid et al., 2017) and availability. Plants were fertilizer for strawberry irrigation in hydro- acquired as pre-chilled, bare root transplants ponic nutrient solutions has a low nitrate from Nourse Farms (South Deerfield, MA) concentration, less than half that of recom- and were received in week 55 (2015). The mended tomato nutrient solution. Increased plants were held in a cooler at 3-5°C (dark- nitrate levels leads to tip-burn in strawber- ness) until planting in each system tested. ries and decreased production (Cold Climate Bare-root plants were planted into each sys- Strawberry Farming, 2014). In a short-term tem in week 4 (2016). The experiment was experiment examining nutrient levels in conducted at the St. Paul Campus of the Uni- aquaponic strawberry production nutrient versity of Minnesota Plant Growth Facilities levels were adequate for growth but varied (44°59’17.8” N, -93°10’51.6” W). greatly depending on fish species and density Fish species grown in the various aqua- (Villarroel et al., 2011). When production be- ponic production facilities were: Perca flave- tween a fully synthetic nutrient source, such scens (yellow perch), Oreochromis spp. (ti- Strawberry 25 lapia), Cyprinus carpio (koi) and Carassius systems except (c) used a randomized block auratus (goldfish). The goldfish were- pur design. System (c) was randomized by PVC chased at PetSmart (Roseville, MN) in 2014. pipeline of plants (see Figure 1). The number Tilapia were obtained as fingerlings from of experimental units varied by treatment, the Arrowhead Fisheries, LLC (Canon City, CO) exact number depending on the space avail- in Jan. 2015. Yellow perch were obtained as able in each system. In the soilless medium, fingerlings from Will Allen Farms,- Grow 32 plants/cultivar were grown; the A-frame ing Power (Milwaukee, WI) in March 2015. ebb and flow had 8 plants/tube for a total of Koi were obtained from Tangletown Gardens 48 plants/A-frame; the tray ebb and flow had (Plato, MN) in Feb. 2016. All fish were ac- 9 plants/tub; the floating raft DWC systems climated to each system’s environment and had 4 plants/tank whereas the warehouse had then placed into each system as soon as the 30 plants/cultivar equally divided between biofilters were functioning. the two growing tubs.

Experimental Setup. Environmental Conditions. Environmental systems tested included: Soilless Medium (Control). The green- (a) soilless medium (control), (b) floating raft house environmental conditions for soil- deep-water culture (DWC), (c) A-frame ebb less medium strawberry production were and flow, (d) tray ebb and flow, and (e) ware- 24.4±3.0/18.3±1.5°C day/night daily integral house. Systems (a) through (d) were located and a 16 hr long photoperiod (0600–2200 in greenhouses while (e) was in a warehouse. HR) lighting (400 w high pressure sodium Systems (c) and (e) used koi, treatment (b) high intensity discharge lamps, HPS-HID) used yellow perch, and treatment (d) used at a minimum of 150 μmol m-2 s-1 PAR. The both goldfish and tilapia. Fish types were greenhouse, located in the St. Paul campus the treatments while strawberry cultivars Plant Growth Facilities (University of Min- were the variable; all treatments had an equal nesota, St. Paul, MN), was an A-frame even- number of plants for each strawberry culti- span construction, sharing one inner wall var randomized throughout each system. All with each adjacent house. The roof, shared

Figure 1. Diagram of the A-frame ebb and flow system showing the location of supplemental lighting, biofilter, fish tank, and plant lines with arrows indicating direction of water flow. 26 Journal of the American Pomological Society inner and interior walls adjoining the service period (long day) and bio control methods, walkway were glazed with double-strength as instituted in the soilless medium treatment float glass whereas the exterior walls had were also used herein. Electric generators chambered acrylic (Exolite®; Cyro Indus- served as the electrical power backup system tries, Mt. Arlington, NJ) glazing. Heating for this and all other aquaponic setups. was delivered from the University of Min- This system consisted of eight aluminum nesota heating plant via hot water into the tanks (193x77.5x75 cm, l x w x h with 6.5 perimeter pipes of the greenhouse with gal- cm thick walls) for fish/plant production. vanized fins for enhanced heat exchange. Each tank had a floating raft system (2/tank; All environmental settings were controlled 60x60x5.5 cm, Owens Corning FOAMU- via an Argus Control Systems Ltd. computer LAR 150, R-10 insulation sheathing; Owens (Surrey, British Columbia, Canada). Corning Co., Toledo, OH); the water volume Strawberries were transplanted into square in each tank was ~550 L or 0.55 m3. Two plas- 754 cm3 plastic pots (Landmark Plastic, Ak- tic, hemispherical tanks (68x47x26 cm) were ron, Ohio) filled with Sunshine LC8 soilless connected to each fish tank and served as the potting medium (Sun Gro Horticulture, Aga- biofilters. Each biofilter was filled with 8-10 wam, MA). Plants were fertilized twice daily, cm dia. gravel (D-Rock Center, New Brigh- between the hours of 0700-0800 and 1600- ton, MN). In greenhouse 369-C2, ammonium 1700, using a constant liquid feed (CLF) of chloride (1 g/biofilter; Hawkins Chemical 125 ppm N from water-soluble 20N–4.4P– Co., Roseville, MN) was used to start the 16.6K (Scotts, Marysville, OH). Fungicide biological filter or biofilter in 8-10 cm dia. drenches of Banrot (Scotts, Marysville, OH), lava rock (D-Rock Center, New Brighton, Subdue (Syngenta, Basel, Switzerland), Me- MN) to produce ~1 mg/L ammonia with an dallion (Syngenta, Basel, Switzerland), and initial start of Carassius auratus (goldfish) Clearys 3336 (Nufarm, Melbourne, Austra- whereas ammonium carbonate was used in lia) were applied in monthly rotations. 369-C4. Two plastic, hemi cylindrical tanks Insect control consisted of bio con- (68x47x26 cm) were mounted above one end trol methods of using yellow sticky cards of each fish tank and served as the biofilters. (12.7 x 7.6 cm; Evergreen Growers Supply, Each biofilter was filled with 2 cm dia. gran- Clackamas, OR) to catch flying insects. Ad- ite gravel (Hedberg Aggregates, Stillwater, ditionally, a variety of mites, Amblyseius MN). A low density (approx. 25-30 fish / andersoni, A. cucumeris, A. swirskii, Neosei- tank) of Carassius auratus was used to start ulus fallacis, Galendromus occidentalis, N. the biological filter in the gravel; these were californicus, Phytoseiulus persimilis (Ben- later removed before the experiment com- eficial Insectary, Redding, CA and Rincon menced and replaced with Perca flavescens Vitova, Ventura, CA), were released rota- (yellow perch). Water was lifted to the bio- tionally for bio control in this and all other filter tanks by a Danner Supreme 700 GPH greenhouses and warehouse over the course mag drive pump. The outflow was had valves of the experiment in order to control for spi- and was split between the two biofilter tanks der-mites (Tetranychus urticae), white flies and a third outlet which discharged directly (Trialeuroides vaporariorum), and thrips to the fish tank for added aeration and circu- (Thysanoptera spp.). Cease fungicide (Bio- lation. Each biofilter received approximately works Inc., Victor, NY) was applied weekly 4 l/min. An automatic bell siphon in each of during Nov. 2016 to control powdery mildew the biofilter tanks allowed the water level to (Podosphaera xanthii). rise in the gravel from a low point of approx- Floating Rafts. This aquaponic greenhouse imately 2 cm depth to a high of around 15 had a 23.6+0.8°C daily integral; the tempera- cm. At the high point the siphon would start ture set point was 23.5°C. The same photo- and the water would draw down (returning to Strawberry 27 the fish tank), creating an ebb and flow in the One fish tank (aluminum; identical speci- gravel. Potential plant spacing on each raft fications as used in the floating raft and A- could be a max. of 16 plants in a 4x4 grid, frame ebb and flow systems) was used for each plant could be grown in a 12cm dia. Net each separate galvanized steel framed, ad- Cup (Hydrofarm Central, Grand Prairie, TX) justable shelving rack system (Ebb and Flow filled with T-rock rockwool (medium grade, systems). One fish tank contained Oreo- 4CF, 30/PL; Therm-O-Rock East, Inc., New chromis spp. (tilapia), which fed one shelv- Eagle, PA). ing rack system, while Carassius auratus Water quality was monitored daily (5/wk (goldfish) were grown in the other tank. All excluding weekends). Temperature measure- fish were at varying densities, depending on ments averaged 22.3+0.9°C and closely ap- fish age. proximated the air temperature set point. The Each system had two shelves/rack (Figure fish species grown in this house and used for 2). Two tubs/shelf (123x94x18 cm; Polytank the duration of the experiment was Perca Co., Litchfield, MN), each of which could flavescens (yellow perch) at varying densi- hold six 50.8x25.4 cm (10”x20”) trays into ties (from 20-30 fish), depending on age which separate plug trays (50s or 72s) were (Sorensen et al., 2015). The same bio control inserted to hold the plants. The top shelf of methods for insect and arachnid pest control each rack system was exposed to natural were instituted in this and all other green- and supplemental lighting (high pressure so- houses, as delineated earlier. dium HID lights) whereas the second shelf A-frame ebb and flow. In this greenhouse, has supplemental light emitting diode (LED) 21.7+0.4°C was the daily integral and the lighting supplied by either Sunshine Systems temperature set point was 21.5°C. Tempera- Grow Pan (450-470, 630 nm; 300 Watt; Sun- ture measurements averaged 23.5+0.9°C and shine Systems, LLC, Wheeling, IL) or Green approximated the air temperature set point. Power LED (450-470, 660 nm; 300 Watt; The same photoperiod (long days) and bio 152x12 cm; 110v strips; Royal Philips N.V., control methods, as instituted in the soilless Andover, MA). One plastic, rectangular tub medium treatment, were also used herein. (123x186x18 cm; Polytank Co., Litchfield, Two tanks in this greenhouse each feed MN) serves as a biofilter for each tank and separate A-frame ebb and flow systems is filled with 3-4 cm dia. lava rock (D-Rock (Figure 1). Fish species grown in this house Center, New Brighton, MN). Each tub is lo- were Cyprinus carpio (koi) at varying den- cated on the concrete floor. sities, depending on fish age. Airlift pumps moved the water from the fish tank to the Warehouse. biofilter; a Danner Supreme 700 GPH mag The warehouse system was a retrofitted drive pump lifts the water from the biofilter walk-in cooler (7.19m x 4.87m x 2.74m), in to the A-frame lines, draining back to the the basement of the plant growth facilities fish tank. Four plastic, hemi cylindrical tanks head house, with galvanized interior walls (68x47x26 cm) were mounted below each A- where a F5 (Fantastically Fun Fresh Food frame were filled with 3-4 cm dia. lava rock Factory) commercial type systems from Nel- (D-Rock Center, New Brighton, MN) and son and Pade Company (http://aquaponics. served as the biofilters. com/; Montello, WI) was installed. The F5 Tray ebb and flow. In this greenhouse, system consisted of one 110-gallon fish tank 21.7+0.4°C was the daily integral and the with separate bio filters and 2—3’ x 5’ plastic temperature set point was 21.5°C. The same tubs that hold floating rafts. There were 15 - photoperiod (long days) and bio control 2” net pots / raft and 2 rafts/tub for a total of methods, as instituted in the soilless medium 90 plants possible to be grown in this system. treatment, were used in this greenhouse. The LED lighting system from Agrivolution 28 Journal of the American Pomological Society

Figure 2. Tray ebb and flow aquaponic system setup showing the location of the supplemental light sourc- es, grow beds, biofilter, and fish tank. Arrows indicate the direction of water flow. Figure modified from Gebhardt, et al. (2015).

LLC (http://www.agrivolution.co/; South and dry weight in grams of fruit per plant per Windsor, CT) was a triple-band LED light harvest. The analysis was conducted using a (bar) above the plants, which could telescope split-plot design with the type of fish as the vertically, depending on the plant height. The treatment and the cultivar as the variable. LEDs had single-chip diodes emitting blue, The systems and fish are interchangeable as green, and red light with full photosyntheti- treatments to the split-plot design in this case cally active radiation (PAR=400 nm to 700 because the cultivars were evenly and ran- nm). Supplemental cooling was supplied to domly distributed no matter whether ‘fish’ or maintain average growing temperatures of ‘system’ is used as a treatment. Using ‘fish’ 20-21°C day/night. lowers the sample size, specifically in the Data Collection. Fruit was harvested when case of the tray system of C4 going from ripened (fully red); dates of harvest were re- both tilapia and goldfish included in one to corded. Fruit were grouped by plant, counted them being separated but there were still (fruit count), weighed (fresh fruit weights, equal numbers of cultivars randomly placed g), and then placed in a high temperature into each tray system. Thus, it still meets the oven (76.67°C) (Hotpack, Philadelphia, PA) requirements of a split plot design. for seven days. Fruit were then removed The koi treatments in system C and E were and weighed after drying was complete (dry analyzed separately. Dependent variables in fruit weights, g). Average fresh or dry fruit this analysis are the number of fruit, fresh weights (g) per plant were calculated as: av- fruit weight, and dry fruit weight. erage total fresh or dry weight / average fruit The GLIMMIX determines the correlation number. between data that has been fit to a specified Statistical Analyses. A Generalized Linear statistical model. Common non-normal dis- Mixed Model test using non-normal distribu- tribution models were tested and the lognor- tion (GLIMMIX) was used to analyze num- mal distribution was found to have the low- ber of fruit per plant per harvest, weight in est AIC and BIC and, therefore, the best fit. grams of fresh fruit per plant per harvest, GLIMMIX assumes a normal distribution of Strawberry 29

Table 1. Significance levels of fruit count, fresh fruit weight (g) by plant, and dry fruit weight (g) by cul- tivar in treatment comparisons. Treatment comparisons Fruit count Fresh fruit weight Dry fruit weight (g) by plant (g) by plant Systems A-E .449 z ns .923 z .085* Fish with soilless medium .517 z .883 z .072* Fish without soilless medium included .445 z .861 z .468 z z ns, *indicates not significant andP <0.10, respectively. random effects, which this experiment has. Tables 2-4 show the average number of All tests were conducted using the software fruit per plant, fresh fruit weight, and dry SAS v.9.4 (Cary, North Carolina). fruit weight along with the standard devia- tion for each cultivar per system treatment Results and fish type, including the soilless medium. System (systems A-E) and fish with soil- Variation was high within all treatments as less medium significantly (P<0.10) affected evidenced by the large standard deviations dry fruit weight per plant, but not number of (Tables 2-4). ‘Evie 2’ in the A-frame ebb and fruit or the fresh fruit weight per plant (Table flow (Koi fish type) had the highest mean 1). When data for soilless medium treatment fruit count (5.30), followed by the floating were excluded from the analysis, dry fruit raft DWC (perch, 4.66) (Table 2). Of the weight was affected by fish treatment, - in three cultivars, ‘Evie 2’ had the highest yield dicating that differences in dry fruit weight across all treatments and fish types, with the were likely due to differences between aqua- exception of the warehouse (Koi) where ‘Por- ponically-grown and soilless-medium-grown tola’ was the highest (Table 2). ‘Evie 2’ in the treatments. floating raft DWC, which had perch as the

Table 2. Means and standard deviations for fruit count per plant for three day neutral strawberry cultivars grown in soilless medium broken down by treatment, fish type, and cultivar. Treatment Fish types Cultivar Mean Standard deviation Soilless medium - Albion 2.98 2.22 Portola 2.72 1.79 Evie 2 3.90 2.63 Floating raft DWC Perch Albion 2.87 1.98 Portola 3.20 2.45 Evie 2 4.66 3.47 Tray ebb and flow Goldfish Albion 2.91 1.91 Portola 2.69 2.42 Evie 2 3.46 2.76 A-frame ebb and flow Koi Albion 2.59 2.69 Portola 3.42 2.73 Evie 2 5.30 4.17 Tray ebb and flow Tilapia Albion 2.64 2.08 Portola 3.54 3.67 Evie 2 4.09 3.20 Warehouse Koi Albion 1.00 0.00 Portola 2.67 1.94 Evie 2 2.1 2.07 30 Journal of the American Pomological Society

Table 3. Means and standard deviations for fresh individual fruit weight (g) for three day neutral straw- berry cultivars grown in soilless medium broken down by treatment, fish type, and cultivar. Treatment Fish types Cultivar Mean Standard deviation Soilless medium - Albion 16.15 11.11 Portola 17.19 12.84 Evie 2 21.02 13.97 Floating raft DWC Perch Albion 15.11 10.63 Portola 21.96 14.93 Evie 2 31.59 24.08 Tray ebb and flow Goldfish Albion 19.51 16.82 Portola 16.83 19.98 Evie 2 21.08 22.89 A-frame ebb and flow Koi Albion 13.66 11.32 Portola 20.77 15.46 Evie 2 26.49 19.13 Tray ebb and flow Tilapia Albion 11.99 10.25 Portola 21.14 21.06 Evie 2 27.76 28.32 Warehouse Koi Albion 11.55 4.35 Portola 12.37 13.55 Evie 2 14.89 10.89

Table 4. Means and standard deviations for dry individual fruit weight (g) for three day neutral strawberry cultivars grown in soilless medium broken down by treatment, fish type, and cultivar. Treatment Fish types Cultivar Mean Standard deviation Soilless medium - Albion 2.32 1.54 Portola 1.54 1.10 Evie 2 2.13 1.36 Floating raft DWC Perch Albion 1.16 10.63 Portola 1.46 1.14 Evie 2 1.99 1.72 Tray ebb and flow Goldfish Albion 1.47 1.13 Portola 1.03 0.95 Evie 2 1.25 1.16 A-frame ebb and flow Koi Albion 1.11 1.25 Portola 1.04 0.87 Evie 2 1.67 1.44 Tray ebb and flow Tilapia Albion 1.30 1.15 Portola 1.41 1.24 Evie 2 1.81 1.27 Warehouse Koi Albion 1.23 0.32 Portola 1.90 1.46 Evie 2 1.10 0.57 Strawberry 31 fish treatment, had the highest average fresh strawberries than those from the soilless me- fruit weight (31.59 g, Table 3). For all treat- dium treatment. ‘Evie 2’ had the highest av- ments and fish types, ‘Evie 2’ had the highest erage dry weight in only three treatments and average fresh fruit weights. ‘Albion’ in the fish types: floating raft DWC/perch, A-frame warehouse system, with koi, had the lowest ebb and flow/Koi, and tray ebb and flow/Ti- fresh fruit weight (11.55 g, Table 3). Average lapia (Table 4). dry individual fruit weight was highest for Since ‘Evie 2’ produced the most fruit on ‘Albion’ in soilless medium treatment (2.32 average (Tables 2 and 3), it follows that it g) whereas ‘Portola’ in the tray ebb and flow also had the highest total fresh fruit weight goldfish treatment and the ebb and flow A- per harvest (Table 5). It is interesting to note, frame treatment had the lowest average dry however, that though ‘Albion’ had the lowest weight (1.03 and 1.04 g, respectively, Table average fruit count and lowest average fresh 4). This indicates there was significantly fruit weight, ‘Portola’ had the lowest dry more water content in aquaponically grown fruit weight. This indicates that ‘Albion’ had

Table 5. Fruit count by month (from April 2016 to February 2017), total number of fruit, fresh fruit weight (g), and individual fresh fruit weight (g) for three day neutral strawberry cultivars averaged over all aqua- ponic treatments and soilless medium treatment. Measurement Month ‘Albion’ ‘Evie 2’ ‘Portola’ Totals Fruit count April 2016 92.0 361.0 305.0 758.0 June 2016 70.0 452.0 139.0 661.0 July 2016 223.0 368.0 168.0 759.0 August 2016 128.0 341.0 105.0 574.0 October 2016 64.0 72.0 29.0 165.0 November 2016 320.0 631.0 229.0 1180.0 December 2016 107.0 109.0 61.0 277.0 January 2017 63.0 119.0 106.0 288.0 February 2017 5.0 4.0 8.0 17.0 Total 1072.0 2457.0 1150.0 4679.0 Fresh fruit weight (g) April 2016 473.48 2567.63 1894.31 4935.42 June 2016 471.73 3067.68 903.29 4442.7 July 2016 1369.28 2234.99 1114.57 4718.84 August 2016 645.08 1323.59 620.47 2589.14 October 2016 378.76 438.56 252.93 1070.25 November 2016 1126.36 1768.94 910.9 3806.2 December 2016 550.7 610.8 489.2 1650.7 January 2017 37.53 168.85 288.43 494.81 February 2017 47.11 26.77 81.23 155.11 Total 5100.03 12207.81 6555.33 23863.17 Average individual April 5.15 7.11 6.21 fresh fruit weight (g) June 6.74 6.79 6.50 per month July 6.14 6.07 6.63 August 5.04 3.88 5.91 October 5.92 6.09 8.72 November 3.52 2.80 3.98 December 5.15 5.60 8.02 January 0.60 1.42 2.72 32 Journal of the American Pomological Society a higher mass to water ratio. 5.43g and 5.82g, respectively (Table 2). Though plants were put into the system in The soilless medium treatment had the Jan. 2016, the first harvest did not occur until lowest average fresh fruit weight (18.1g), April 2016 – the time required for leaf un- significantly less than the DWC floating raft folding, flowering and fruit set. Harvests var- treatment (25.5g). However, it had the high- ied by month and by cultivar (Table 5), as is est average dry fruit weight (2.04g), which typical of day neutral strawberries (Rowley was significantly more than the tray ebb and et al., 2011). The highest number of fruit for flow goldfish and the A-frame ebb and flow ‘Albion’ and ‘Portola’ was harvested in Nov. koi treatments (1.30g). This suggests that 2016 whereas for ‘Evie 2’ the highest fruit berries grown in soilless medium have a count occurred in June 2016 (Table 5). How- higher mass to water ratio, which may impact ever, the highest total fresh fruit weight was taste and nutrition, though this study did not harvested in June for ‘Evie 2’ and July for address those factors. ‘Albion’, although for ‘Portola’ this occurred While ‘Albion’ had the lowest fresh fruit even earlier in April 2016 (Table 5). This weight and the lowest average number indicates that fruit produced in later months of fruit, ‘Portola’ had the lowest dry fruit were smaller than those produced in the first weight, suggesting ‘Albion’ has a higher large harvest period. In general, average in- mass to water ratio. This is consistent with dividual fresh fruit weight (g) per month de- previous research confirming ‘Albion’ had clined over time for all cultivars (Table 5). higher total soluble solid concentration than other cultivars (Petran et al., 2017). Discussion In a study comparing differences between The average fruit weight in this study is row covers in plasticulture field strawberry half as much for ‘Albion’ and less than half production, average fresh fruit weight was for ‘Evie 2’ than any other study examining 14.32g for ‘Albion’, 12.96g for ‘Evie 2’, and strawberry production using alternative pro- 14.78g for ‘Portola’ (Jordan, 2013). For low duction methods (Table 6). The hydroponic tunnels in the same experiment, average fresh study of Paparozzi, et al. (2010), reported fruit weight of ‘Albion’, ‘Evie 2’, ‘Portola’ average fruit sizes of 11.68g for ‘Albion’ and was 15.16g, 14.86g, and 18.61g, respectively 16.31g for ‘Evie 2’ (Table 6), while the aver- (Jordan, 2013). A follow up study, conducted age fresh weight per berry in this study was in 2017 using the same plasticulture row cov-

Table 6. Reported average fruit weight (g) for the day neutral strawberries ‘Albion’, ‘Evie 2’, and ‘Portola’ grown in aquaponic (all treatments), hydroponic, low tunnel, high tunnel, and field production systems. Low High Cultivar Aquaponic Hydroponic Tunnel Tunnel Field Citations ‘Albion’ 5.43 10.81-11.68 15.16 13.7 13.9-16.1 Moore, et al. 2013; Jordan, 2017, 2013; Rowley, et al. 2011; Paparozzi, et al. 2010; Miranda, et al. 2014 ‘Evie 2’ 6.3g 16.31 14.86 12.1 12.92 Jordan, 2013; Rowley, et al. 2011; Paparozzi, et al. 2010

‘Portola’ 5.82g - 18.61 - 14.78-18.1 Jordan 2013, 2017

Strawberry 33 ers, had average fresh fruit weight of 16.1g leading to stunting of the plant and eventual for ‘Albion’ and 18.1g for ‘Portola’ (Jordan, death. Strawberries grown hydroponically 2017). In a high tunnel experiment, average are recommended to have 2.5g/L iron che- fruit size for ‘Albion’ was 13.7g while ‘Evie late included in the nutrient solution for ideal 2’ had 12.1g (Rowley et al., 2011). These re- growth (de Villiers, 2008) although the ef- sults are consistent with other studies of field fectiveness of plant uptake depends on the and low tunnel production using the chosen type of chelating agent used and the pH of cultivars (cf. Table 5, Petran et al., 2017). the system (Lucena et al., 1990). All studies reported over double the average To correct for iron chlorosis in hydroponic fresh berry weight found in this study. or field settings either a solution of chelated A marketable berry is considered to be over iron would be added directly to the roots or 10g, completely and evenly red, and evenly alkalinity is lowered (pH =6.5). Keeping the filled without major deformities (Rowley et pH <6.5 makes iron more soluble and able al., 2011). In terms of fresh weight we did to be taken up by the plants. The other op- not reach an average berry weight above tion, adding iron directly to the water of the 10g in any cultivar during any month (Table aquaponic system, may have negative conse- 4). Though we did not distinguish between quences on the fish and/or nitrifying bacteria marketable and unmarketable berries in this in the biofilter. Fish tolerance of high iron study, it is important to keep marketability in levels is not known but probably varies by mind when considering the issues of straw- species (Kwong and Niyogi, 2008). Iron ab- berry production in aquaponic systems. sorption in fish has been shown to be very Small fruit in this study was not completely inefficient (Bury and Grosell, 2003) which unexpected. Aquaponics has several known could indicate that addition of chelated iron intrinsic issues like iron deficiency (Graber to water in aquaponic systems is possible. A and Junge, 2009) which reduces photosyn- study done subsequent to this study by Ru, thesis and a lack of pollinators which directly et al. (2017) added 2mg/L Fe-EDTA weekly affects fruit fill. All strawberry cultivars in all with positive effects to plant growth and no aquaponic systems had severe iron deficien- negative effects to the fish. Modifying the cy symptoms, due to insufficient levels of Fe fish feed to include more iron may also in- in the aquaponic tanks. There are methods in crease availability in the system. Fish feeds development to overcome these issues but range from 30 to 170mg iron per Kg (Wata- they were excluded in order to understand nabe et al., 1997). the potential yield of strawberry in aquaponic While strawberry roots can absorb iron systems without supplemental inputs. more efficiently than any other part of the Iron deficiency has a significant - nega plant, the difficulties in getting chelated tive impact on strawberry fruit production iron to them has led to the development of (Roosta, 2014). Strawberry iron uptake fluc- iron foliar sprays (Dordas, 2008). Iron foliar tuates over the life cycle of the plant increas- sprays can be effective in reversing chlorosis ing when vegetative growth occurs and de- on strawberry (Pestana et al., 2011). Iron can creasing during fruit formation and ripening also be absorbed through stomatal openings (Chow et al., 1992). Iron is a micronutrient as it has been found that foliar applications used in the production of chlorophyll and in to the underside of leaves are more effective reproductive processes. It is the micronutri- than those applied to the top side (Schlegel ent needed in strawberries at highest quanti- et al., 2006). Aquaponic growers operating ties (Kobayashi and Nishizawa, 2012). Lack in greenhouses are suggested to apply iron of iron in plants causes interveinal chlorosis foliar sprays in the morning in order to stim- or yellowing of leaves. Symptoms intensi- ulate photosynthesis and maximize uptake fy until the leaf becomes completely white, (Brüggemann et al., 1993). Growers with 34 Journal of the American Pomological Society fully artificial light are recommended to add dicates there was substantial yield loss due to a strong blue light to contribute to iron re- powdery mildew infection. duction and stomatal opening (Brüggemann Pollinators were not added to any of the et al., 1993). tested production systems, which may have There is also the threat of pests and dis- had an impact on fruit size. Fruit fill in straw- eases, which did not spare this study. Spider berries is determined by successful polli- mites (Tetranychus urticae) and powdery nation and fertilization of the ovule which mildew (Podosphaera ×anthii) were issues stimulates fruit fill (Nitsch, 1950). We relied for all cultivars, despite the bio control meth- on gravity and air movement to pollinate ods, especially in the greenhouse housing the strawberry flowers which has been shown A-frame ebb and flow and the tray ebb and to have a pollination success rate of up to flow systems. No strawberry cultivar showed 60% (Vincent et al., 1990). The addition of any resistance to spider mites. Thus, there are pollinators to other greenhouse experiments multiple opportunities for day neutral straw- significantly increased fruit size and weight berry breeders for cultivar improvement (Abrol et al., 2017). Poor pollination most before greenhouse or warehouse aquaponic likely contributed to small fruit size through- strawberry production can be viable. out this experiment. Though we did not dis- Spider mite infestation was an ongoing tinguish between marketable and non-mar- problem in the greenhouse housing the A- ketable fruit, the lack of insect pollinators frame ebb and flow and the tray ebb and flow led to uneven pollination and misshapen fruit systems and affected all treatments therein. (Nye and Anderson, 1974). Due to toxicity to the fish it was not possible If there had been sufficient plant acces- to treat these with pesticides, as would be sible iron, adequate pollination, and had standard practice in regular greenhouse cul- there not been a spider mite infestation, it is tivation. Bio controls in the form of various possible average fruit size could have been species of mites, which are parasitic to spi- over 10g. Iron foliar sprays to iron deficient der mites, were regularly released. However, plants increased yield by up to 56% (Zaiter this bio control method takes 2-3 weeks to et al., 1993). Spider mite infestation reduces reduce populations of spider mites and has strawberry yield an average of 23% (Walsh et to be continuously renewed (Pundt, 2014). al., 2002). Finally, the lack of pollinators re- Spider mite infestations reduce day neutral duces fruit set and total yield by up to 4 fold strawberry yield by an average 23% (Walsh (Abrol et al., 2017). It is reasonable that av- et al., 2002). erage fruit size could have been comparable Powdery mildew was observed on straw- to field production methods if these essential berry plants in the greenhouse housing the A- strawberry production needs had been met. frame ebb and flow and the tray ebb and flow There was a significant difference in cul- systems beginning in Nov. 2016. It was con- tivars in this study, specifically ‘Evie 2’ pro- trolled within a month with weekly sprays of duced more significantly heavier fruit (fresh a biological fungicide. Extensive infection weight) than the other cultivars (Table 3). can cause yield losses of up to 60% in straw- ‘Evie 2’ consistently performed better in berry (Asalf et al., 2012). Even when infec- controlled environments (Petran et al., 2017) tion is not severe, powdery mildew can still and was among the highest yielding cultivars reduce yield by 5% when observed infection in hydroponic trials in the Midwest (Wort- is less than 20% (Carisse et al., 2013). Dur- man et al., 2016). ‘Evie 2’ was bred to im- ing the month of Nov. 2016, fruit yield in the prove upon day neutral cultivars in northern total soilless media was 2324g and only 845g climates with hot summers (Edward Vinson fruit for the greenhouse housing the A-frame Limited, 2018). Its improved tolerance to ebb and flow and tray ebb and flow. This -in temperatures over 33oC and moderate resis- Strawberry 35 tance to powdery mildew make it ideal for Phytoseiulus persimilis on two-spotted spider mite production in controlled environments. To in strawberry. Biol. Control 61:121–127. https://doi. further improve day neutral strawberries for org/10.1016/j.biocontrol.2012.01.015 Bradford, A., J.F. Hancock, and R.M. Warner. 2010. aquaponic production in both greenhouses Interactions of temperature and photoperiod deter- and warehouses, plant breeders should focus mine expression of repeat flowering in strawberry. on increasing iron uptake or chlorosis resis- J. Amer. Soc. Hort. Sci. 135:102-107. tance and resistance to spider mites while Brüggemann, W., K. Maas-Kantel, and P.R. Moog. maintaining or increasing high fresh fruit 1993. Iron uptake by leaf mesophyll cells: The yield and quality. role of the plasma membrane-bound ferric-che- late reductase. Planta 190:151–155. https://doi. org/10.1007/BF00196606 Conclusion Bury, N.R. and M. Grosell. 2003. Waterborne iron ac- Strawberry production using aquaponic quisition by a freshwater teleost fish, zebrafish, Da- systems has many challenges that cannot be nio rerio. J. Exptl. Biol. 206:3529–3535. ignored if yield is to be equivalent to other Carisse, O., V. Morissette-Thomas, and H. Van Der systems. Though the number of fruit per Heyden. 2013. Lagged association between pow- plant across systems was equivalent without dery mildew leaf severity, airborne inoculum, supplemental inputs to the system, aquaponic weather, and crop losses in strawberry. Phytopathol. 103:811–821. https://doi.org/10.1094/PHYTO-11- strawberry fruit size was significantly lower 12-0300-R than both field and hydroponic cultivation Chow, K.K., T. V Price, and B.C. Hanger. 1992. Nutri- methods. Adding iron, including pollinators, tional requirements for growth and yield of straw- as well as controlling arthropods pests and berry in deep flow hydroponic systems. Scientia disease should increase aquaponic produc- Hort. 52:95–104. tion. Choosing a cultivar such as ‘Evie 2’ that Cold Climate Strawberry Farming. 2014. 1st Ed. Univ. is suited to indoor production will make a of Minnesota Dept. of Hort. Sci., Saint Paul, MN. Costa-Pierce, B.A., J. Rakocy, and W.A. Society. 1997. significant difference in yield, unlike choos- Tilapia aquaculture in the Americas. World Aqua- ing between the different species of fish. culture Soc.; American Tilapia Association, Baton Further research should focus on breeding Rouge, La. of new strawberry cultivars better suited to de Villiers, J.J. 2008. The influence of different pro- aquaponic growing conditions and standard- duction systems, planting densities and levels of ization of recommendations regarding inputs shading on the yield, quality and growth potential of into aquaponic systems for supplemental ‘Chandler’ strawberry plants (Fragaria ×ananassa) grown in coir. M.S. Thesis. Stellenbosch University, iron treatments. Stellenbosch, S. Africa. Diver, S. and L. Rinehart. 2006. Aquaponics — Inte- Literature Cited gration of hydroponics with aquaculture. ATTRA Abrol, D.P., A.K. Gorka, M.J. Ansari, A. Al-Ghamdi, Natl. Sustain. Agric. Inf. Serv.,Washington, DC. and S. Al-Kahtani. 2017. Impact of insect pollina- http://www.attra.ncat.org Accessed Feb.12.2018 tors on yield and fruit quality of strawberry. Saudi Dordas, C. 2008. Role of nutrients in controlling plant J. Biol. Sci. https://doi.org/https://doi.org/10.1016/j. diseases in sustainable agriculture. A review. Agron. sjbs.2017.08.003 Sustain. Dev. 28: 33. https://doi-org.ezp3.lib.umn. Aquaponic Farming: Global Market Intelligence. edu/10.1051/agro:2007051 2016. M2 Press. URL https://www.prnewswire. Edward Vinson Limited. 2018. Evie 2 strawberry. com/news-releases/aquaponic-farming-global-mar- Plant Breeders Rights. Can. Food Insp. Agency. ket-intelligence-2011-2020-300395192.html Ac- http://www.inspection.gc.ca/english/plaveg/pbr- cessed Feb.12.2018. pov/cropreport/str/app00006055e.shtml Accessed Arnade, C. and F. Kuchler. 2015. Measuring the im- 15 Aug. 2017. pacts of off-season berry imports. USDA Econ. Res. Eigenbrod, C. and N. Gruda. 2015. Urban vegetable Serv. No. 197. for food security in cities. A review. Agron. Sustain. Asalf, B., N. Trandem, A. Stensvand, V.W. Wekesa, Dev. 35:483–498. https://doi.org/10.1007/s13593- G.J. de Moraes, and I. Klingen. 2012. Influence 014-0273-y of sulfur, powdery mildew, and the predatory mite Gebhardt, B., M. White, C. Montouri, and K. Vinzenz. 36 Journal of the American Pomological Society

2015. Comparing multiple forms of light on dif- soilless systems for strawberry transplant produc- ferent cultivars of Swiss chard in an abb and flow tion in Australia. Acta Hort. 1176:53–64. https:// bench aquaponics system. www.aquaponics.umn. doi.org/10.17660/ActaHortic.2017.1176.8 edu Univ. of Minnesota. Accessed 17 Oct. 2017. Miranda, F.R. de, V.B. Da Silva, F.S.R. Dos Santos, Graber, A. and R. Junge. 2009. Aquaponic systems: A.G. Rossetti, and C. de F.B. Da Silva. 2014. Pro- Nutrient recycling from fish wastewater by vegeta- duction of strawberry cultivars in closed hydropon- ble production. Desalination 246:147–156. https:// ic systems and coconut fibre substrate. Rev. Ciência doi.org/https://doi.org/10.1016/j.desal.2008.03.048 Agronômica 45:833–841. https://doi.org/10.1590/ Hundley, G.C. and R.D. Navarro. 2013. Aquaponis: S1806-66902014000400022 Integration between aquaculture and hydroponics. Moore, P., D. Collins, A. Bary, and C. Cogger. 2013. Rev. Bras. Agropecuária Sustentável 3. https://doi. Evaluation of day-neutral production in Washing- org/10.21206/rbas.v3i2.218 Accessed 1 Nov. 2017. ton cultivars for organic strawberry. Acta Hort. Jordan, E. 2017. 2017 Results. Univ. of Minnesota 1001:167–174. Fruit Res. http://fruit.cfans.umn.edu/2017-results/ Morgan, K. 2015. Commodity strawberry profile. Ag- Accessed 1 Nov. 2017. ric. Mark. Resour. Cent. URL https://www.agmrc. Jordan, E. 2013. Individual berry weight for day- org/commodities-products/fruits/strawberries/ Ac- neutral strawberries. Univ. of Minnesota Fruit Res. cessed 10 Dec. 2017. URL http://fruit.cfans.umn.edu/individual-berry- Nitsch, J.P. 1950. Growth and morphogenesis of the weight-day-neutral-strawberries/ strawberry as related to auxin. Amer. J. Bot. 37:211– Kloas, W., R. Groß, J. Graupner, H. Monsees, M. Tsch- 215. https://doi.org/10.1002/j.1537-2197.1950. irner, B. Wittstock, S. Wuertz, A. Zikova, B. Renne- tb12183.x rt, D. Baganz, G. Staaks, J. Suhl, and U. Schmidt. Nye, W.P. and J.L. Anderson. 1974. Insect pollinators 2015. A new concept for aquaponic systems to frequenting strawberry blossoms and the effect of improve sustainability, increase productivity, and honey bees on yield and fruit quality. J. Amer. Soc. reduce environmental impacts. Aquac. Environ. In- Hort. Sci. 99:40–44. teract. 7:179–192. https://doi.org/10.3354/aei00146 Ostertagová, E., O. Ostertag, and J. Kováč. 2014. Kobayashi, T. and N.K. Nishizawa. 2012. Iron up- Methodology and application of the Kruskal- Wal- take, translocation, and regulation in higher plants. lis test. Appl. Mech. Mater. https://doi.org/10.4028/ Annu. Rev. Plant Biol. 63:131–152. https://doi. www.scientific.net/AMM.611.115 Accessed 1 Jan. org/10.1146/annurev-arplant-042811-105522 2018. Kwong, R.W.M. and S. Niyogi. 2008. An in vitro ex- Paparozzi, E., S. Adams, G. Meyer, M.E. Conley, V. amination of intestinal iron absorption in a freshwa- Schlegel, E. Blankenship, and P. Read. 2010. Se- ter teleost, rainbow trout (Oncorhynchus mykiss). lecting strawberry cultivars for winter greenhouse J. Comp. Physiol. B 178:963–975. https://doi. production. HortSci. 45:S230. org/10.1007/s00360-008-0279-3 Pestana, M., I. Domingos, F. Gama, S. Dandlen, M. Lewis, W.M. 1978. Use of hydroponics to maintain Miguel, J. Castro Pinto, A. de Varennes, and P. Cor- quality of recirculated water in a fish culture system. reia. 2011. Strawberry recovers from iron chlorosis Trans. Amer. Fish Soc. 107:92–99. after foliar application of a grass-clipping extract. Love, D.C., J.P. Fry, L. Genello, E.S. Hill, J.A. Fred- J. Plant Nutr. Soil Sci. 174:473–479. https://doi. erick, X. Li, and K. Semmens. 2014. An interna- org/10.1002/jpln.201000215 tional survey of aquaponics practitioners. PLoS Petran, A., E. Hoover, L. Hayes, and S. Poppe. 2017. ONE 9(7):e102662. https://doi.org/10.1371/journal. Yield and quality characteristics of day-neutral pone.0102662 strawberry in the United States Upper Midwest Love, D.C., J.P. Fry, X. Li, E.S. Hill, L. Genello, K. using organic practices. Biol. Agric. & Hortic. Semmens, and R.E. Thompson. 2015. Commercial 33:73–88. https://doi.org/10.1080/01448765.2016. aquaponics production and profitability: Findings 1188152 from an international survey. Aquaculture 435:67– Poppe, S., E. Hoover, A. Petran, and J. Tillman. 2016. 74. https://doi.org/https://doi.org/10.1016/j.aqua- University of Minnesota day-neutral low tunnel culture.2014.09.023 strawberry project. Univ. of Minnesota. http://fruit. Lucena, J.J., A. Garate, A.M. Ramon, and M. Man- cfans.umn.edu/files/2016/01/Low_Tunnel_MFV- zanares. 1990. Iron nutrition of a hydroponic straw- GA_2016_FINAL.pdf Accessed 20 Dec. 2017. berry culture (Fragaria vesca L.) supplied with dif- Pritts, M. and L. Mcdermott. 2017. Protected culture ferent Fe chelates. Plant and Soil 15:9–15. for strawberries using low tunnels. Horticulture Mattner, S.W., C.B. Horstra, M. Milinkovic, P.R. Mer- Section, School of Integrative Plant Sci., Cornell riman, and F.C. Greenhalgh. 2017. Evaluation of University, Ithaca, NY. http://www.hort.cornell.edu/ Strawberry 37

fruit/pdfs/low-tunnel-strawberries.pdf Accessed 1 Sorensen, M., L. Robinson, and T. Roth-Krosnoski. Jan. 2018. 2015. A comparison of rockwool, lava rock, ex- Pundt, L. 2014. Biological control of two-spotted spi- panded clay aggregate, and coconut coir as growing der mites. Univeristy of Connecticut, Department of substrate in a floating raft aquaponic system exam- Plant Sci. and Landscape Architecture. Fact Sheet. ining growth rates of ‘Improved Amethyst’ basil http://ipm.uconn.edu/documents/raw2/html/664. and ‘Nancy’ butterhead lettuce. University of Min- php?aid=664 Accessed 1 Feb. 2018. nesota. http://www.aquaponics.umn.edu/spring- Rakocy, J.E., M.P. Masser, and T.M. Losordo. 2006. 2015-final-projects Accessed 21 Dec. 2017. Recirculating aquaculture tank production systems: Taiz, L. 2010. Plant physiology, 5th ed. Sinauer As- Aquaponics- integrating fish and plant culture. sociates, Sunderland, MA. SRAC Publ. - South. Reg. Aquac. Cent. 16. https:// Takeda, F. and S.C. Hokanson. 2003. Strawberry doi.org/454 Accessed 2 Feb. 2018. fruit and plug plant production in the greenhouse. Rakocy, J.E., D.S. Bailey, R.C. Shultz, and J.J. Da- Acta Hort. 626:283-285 DOI: 10.17660/ActaHor- naher. 2007. Preliminary evaluation of organic tic.2003.626.39 waste from two aquaculture systems as a source USDA Economic Research Service. 2016. U.S. of inorganic nutrients for hydroponics. Acta Hort. strawberry consumption continues to grow. USDA 17660:201–207. https://doi.org/10.17660/ActaHor- Econ. Res. Serv. URL https://www.ers.usda. tic.2007.742.27 gov/data-products/chart-gallery/gallery/chart- Roosta, H.R. 2014. Effect of ammonium:nitrate ratios detail/?chartId=77884 Accessed 20 Dec. 2017. in the response of strawberry to alkalinity in hydro- Villarroel, M., J.M.R. Alvariño, and J.M. Duran. 2011. ponics. J. Plant Nutr. 37:10. https://doi.org/10.1080 Aquaponics: integrating fish feeding rates and ion /01904167.2014.888749 waste production. Spanish J. Agric. Res. 9:537–545. Rowley, D., B.L. Black, D. Drost, and D. Feuz. 2011. Vincent, C., D.D. de Oliveira, and A. Bélanger. 1990. Late-season strawberry production using day-neu- The management of insect pollinators and pests in tral cultivars in high-elevation high tunnels. Hort- Quebec strawberry plantations. Monit. Integr. Man- Science 46:1480–1485. ag. Arthropod Pests Small fruit Crops, p.177-192. Ru, D., J. Liu, Z. Hu, Y. Zou, L. Jiang, X. Cheng, and Walsh, D.B., F.G. Zalom, D. V Shaw, and K.D. Lar- Z. Lv. 2017. Improvement of aquaponic perfor- son. 2002. Yield reduction caused by twospotted mance through micro- and macro-nutrient addition. spider mite feeding in an advanced-cycle straw- Environ. Sci. Pollut. Res. 24:16328–16335. https:// berry breeding population. J. Amer. Soc. Hort. Sci. doi.org/10.1007/s11356-017-9273-1 127:230–237. Schlegel, T.K., J. Schönherr, and L. Schreiber. 2006. Watanabe, T., V. Kiron, and S. Satoh. 1997. Trace min- Rates of foliar penetration of chelated Fe(III): erals in fish nutrition. Aquaculture 151:185–207. Role of light, stomata, species, and leaf age. J. https://doi.org/10.1016/S0044-8486(96)01503-7 Agric. Food Chem. 54:6809–6813. https://doi. Wortman, S., M.S. Douglass, and J.D. Kindhart. 2016. org/10.1021/jf061149i Cultivar, growing media, and nutrient source influ- Søberg, E.-E. 2016. The growth and development of ence strawberry yield in a vertical, hydroponic, high lettuce, coriander and swiss chard in a cold water tunnel system. HortTech. 26:466–473. aquaponic system optimized for lettuce production. Zaiter, H.Z., I. Saad, and M. Nimah. 1993. Yield of iron M.S. Thesis. Norwegian Unviersity of Life Sci- sprayed and non-sprayed strawberry cultivars grown ences, Ås, Norway. on high pH calcareous soil. J. Plant Nutr. 16:281– 296. https://doi.org/10.1080/01904169309364531 38 Journal of the American Pomological Society

Journal of the American Pomological Society 73(1): 38-46 2019 Peach bagging in the Southeastern U.S. Jaine Allran, Guido Schnabel, and Juan Carlos Melgar1 Additional index words: IPM, organic, Prunus persica, pesticide, brown rot

Abstract The control of pests and diseases is pivotal in producing high quality peaches in the humid climate of the Southeastern United States. Fruit bagging is proposed as a strategy to physically exclude pests, prevent disease development, and consequently help reduce the reliance on pesticides. Six peach trees from two cultivars (early and mid-season cultivars) were chosen from one organic farm and one conventional farm, and 150 fruit/tree were bagged with 75 being marked as control and left unbagged. In order to understand the effect of bagging on coloring, 50 bags were removed per tree 10 days before harvest. Fruit quality analysis (size, weight, SSC, acidity, color) and disease incidence were assessed at harvest and post-harvest. Bagging increased marketable yield at the organic farm but not at the conventional farm. Fruit quality (size, brix, acidity) of bagged peaches was similar when compared to non-bagged fruit. The intensity of the red blush of the peach was reduced for the bagged fruit compared to control fruit but peaches that were unbagged 10 days before harvest had blush comparable to the control. The effect of bags on postharvest disease incidence was not conclusive. Public surveys showed that po- tential consumers preferred bagged peaches and were willing to pay more for them when informed that the fruit had been less exposed to pesticides. Conventional control practices for pests orchards under these conditions could benefit and diseases for peach production can reduce from new techniques to help combat pests the sustainability and profitability of -com and diseases. mercial orchards and can be challenging in Fruit bagging is an extensively used prac- regions with warm and humid climates such tice in countries such as China and Japan as the southeastern U.S. As a consequence, (Sharma et al., 2014; Shen et al., 2014;) and organic production of peaches under these on a smaller scale in other countries such as climatic conditions is almost nonexistent Spain (Faci et al., 2014). In the United States, because of high pest and disease pressure fruit bagging has been used on a very small and the lack of OMRI-approved, effica- scale in commercial orchards on the West cious chemicals (Ames, 2012). Nevertheless, Coast to improve late-season red color devel- peach had the highest demand among organic opment in ‘Fuji’ apples for export to Asian fruits in recent years; between 2008 and 2011 countries (Bessin and Hartman, 2003), and the value of the organic peach industry rose also for pollination of Calimyrna figs (Gerdts by 41%, and it expanded in both harvested and Clark, 1979), but bagging has been con- acreage and harvested quantity, mostly in sidered impractical for most commercial California, where dry summer conditions growers. Fruit bags have been used for a are ideal for growing fruit organically (Perez wide variety of fruit crops including bananas, and Plattner, 2013). On the other hand, the mangos, pears, apples, and grapes (Hofman frequent application of pesticides in conven- et al., 1997; Fan and Mattheis, 1998; Ama- tionally-produced peaches in the southeast- rante et al., 2002; Signes et al., 2007; Muchui ern U.S. is of environmental concern, may et al., 2010). They are mostly used for pest pose increased risks to workers and consum- and disease protection but also to manipu- ers, and promote pathogen resistance devel- late aesthetic parts of the fruit such as color opment. Thus, both organic and conventional (Sharma et al., 2013) and produce unblem-

1 Department of Plant and Environmental Sciences, Clemson University, Clemson, SC 29634 Corresponding author: [email protected] Peach 39 ished fruits that meet consumer’s high qual- season cultivars were ‘Scarletprince’ and ity requirements. Depending on the purpose ‘Julyprince’ (for the conventional farm the of the bag and fruit involved, single or dou- original ‘Scarletprince’ block was removed ble layer bags, yellow bags, and polyethylene after the first year and the second year we plastic have been used (Kim et al., 2008; Xu used a block of ‘Julyprince’ trees; for the or- et al., 2010; Muchui et al., 2010). ganic farm, we used ‘Julyprince’ both years). We hypothesized that fruit bags could be a All trees were mature trees, 7-10-year-old, valuable tool to reduce the incidence of pests grafted onto GuardianTM peach seedling root- and diseases, and increase marketable yields stock, at a planting density of 346 trees per in organic orchards located in warm and hu- ha. The conventional orchards strictly fol- mid climates such as the southeastern U.S. lowed a chemical spray program based on Fruit bags may also be used to produce high- the current Southeastern peach, nectarine quality peaches in conventional orchards and plum management guidelines, including for a premium market and reduce reliance dormant sprays for scale and bacterial spot on pesticides. The objectives of this project control, pre-bloom, bloom, and post-bloom were to evaluate fruit bagging as a strategy to for peach scab and blossom blight control, increase production of high-quality peaches summer cover sprays for summer disease in both organic and conventional farms, and control, and preharvest sprays for brown rot to obtain consumer feedback on preferences control. One to three days before bagging, and willingness to pay a premium for this conventional trees were treated with captan product. and phosmet, while the organic orchards were treated with sulfur-based products, py- Materials and Methods rethroids, and kaolin clay. All orchards were Orchard selection. This research was car- within 15 km from each other (Table 1) and ried out in two commercial farms in South the soil was well-drained sandy loam at both Carolina: one organic farm and one conven- locations. Also, the environmental conditions tional farm. The organic farm was located were practically the same for all orchards. in Monetta, SC. The conventional farm was Precipitation from 1 March thru 30 May located in Ridge Spring, SC. Two peach (bloom-harvest period for early-season culti- [Prunus persica (L.) Batsch] orchards were vars) in 2015 and 2016 was 216 mm and 185 selected at each farm: one with an early-sea- mm, respectively, with an average tempera- son cultivar and another one with a mid-sea- ture of 17.4 ˚C for both years. Precipitation son cultivar. The early-season cultivar was and average temperature between 1 March ‘Crimson Lady’ for both farms, and the mid- and 10 July (bloom-harvest period for mid-

Table 1: Location (latitude and longitude) and harvest date for each orchard and each year. Orchard Latitude / Longitude Harvest date 2015 2016 Organic ‘Crimson Lady’ 33° 50’ 2.4” N / 81° 37’ 21.7” W May 26 June 3 ‘Julyprince’ 33° 47’ 39.8” N / 81° 35’ 8.5” W July 14 July 1 Conventional ‘Crimson Lady’ 33° 49’ 4.8” N / 81° 44’ 7.1” W May 26 June 3 ‘Scarletprince’ 33° 49’ 31.1” N / 81° 42’ 7.9” W July 6 -z ‘Julyprince’ 33° 49’ 28.9” N / 81° 42’ 31.3” W - July 1 z ‘Scarletprince’ trees at the conventional farm were pulled out after the first year of data and we used ‘Julyprince’ trees the second year. 40 Journal of the American Pomological Society season cultivars) was 473 mm and 19.5 ˚C in with a temperature-compensated refractome- 2015, and 253 mm and 20.2 ˚C in 2016. ter (model ATC-1, Atago Co., Tokyo, Japan), Experimental setup. The experiment was and data were given as °Brix. The pH and performed in a completely randomized de- TA were determined by autotitration with 0.1 sign with six trees (reps) for each cultivar N NaOH to pH 8.2 (Titrosampler, Metrohm and orchard. In the third week of April 2015, Riverview, FL, USA) and data were given as when fruitlets were about 2.5 cm in diam- % malic acid per 100 g of fresh weight (FW). eter, right after thinning, seventy-five fruit In addition, SSC/TA ratio was calculated. per tree were tagged as control (non-bagged) Postharvest. At harvest, 30 fruit per tree, fruit, and 150 fruit per tree were bagged. Ten cultivar, and location were set aside for post- days before harvest the bags of 50 bagged harvest disease assessment. The fruit were fruit were removed to evaluate the effect of stored in an air-conditioned room at 20°C sunlight on color (bagged-10d). Thus, three and were evaluated three and seven days treatments were established: control, bagged, later to determine the incidence of brown rot. and bagged-10d fruit. At harvest, fruit were Consumer surveys. Consumer surveys collected and separated by treatments: 75 were conducted in two separate locations af- control, 100 bagged, and 50 bagged-10d ter fruit was harvested the first year. The first fruit per tree. The number of unblemished was at the local farmer’s market in Clemson, (marketable) and rotten/damaged fruit was SC where 57 participants agreed to be sur- recorded. Since there was no difference in veyed. The second location was on Clemson color between the bagged-10d and control University’s campus within walking distance fruit for any of the locations and cultivars in of the downtown Clemson area where 29 2015, the removal of bags ten days before participates agreed to be surveyed. A table harvest was not repeated in 2016. In 2016, was set up with two baskets of unlabeled fruit were tagged and bags were placed dur- peaches: control (never bagged) peaches, ing the first week of May. The number of and bagged peaches. Participants were asked bagged fruit was reduced to 100 bags per tree which batch of peaches they found more ap- with 75 tagged as control but everything else pealing. They were allowed to hold or smell was the same as in 2015. the peaches but they did not taste them. The Fruit quality. Fruit was harvested when concept of bagging was then explained to control fruit was at commercial ripening the participant and then asked again which stage. Since the early-season cultivar (‘Crim- batch of peaches was more attractive. The fi- son Lady’) was the same for both farms, fruit nal question was if the consumer would pay a was harvested on the same date; for the mid- premium for bagged peaches and, if so, how season cultivars, there was a difference of a much that premium would be. Demographic few days between farms due to the slightly data (sex, age and race) were recorded. different ripening dates of ‘Scarletprince’ Data analyses. Data from each cultivar, and ‘Julyprince’. Size (diameter, mm) and commercial farm, and year were analyzed weight (g) were measured using a Fruit Tex- separately (i.e. each year, we compared ture Analyzer (FTA; GÜSS, Strand, South bagged and non-bagged fruit from each cul- Africa) from a subsample of five fruit (Frett tivar separately, for both the organic and the et al., 2012) per treatment, per tree, per culti- commercial orchard). Data were subjected to var and location. Afterwards, fruit juice was analysis of variance (ANOVA) using JMP® squeezed from a composite sample com- 12.2.0 software (SAS Institute, Cary, NC). prised of two slices from each of the five fruit Percentage data was transformed using arc- and subsequently used to measure soluble sine transformation and then subjected to solids concentration (SSC), pH, and titrat- ANOVA. able acidity (TA). The SSC was measured Peach 41

Table 2: Quality parameters for fruit at the organic farm including diameter (mm), mass (g), SSC (%) and TA (%). Data shown for “Crimson Lady’ and ‘Julyprince’ for two years. 2015 2016 Diam. Mass SSC TA Diam. Mass SSC TA ‘Crimson Lady’ Control 61.3bz 113.5b 9.3 0.8a 61.8 127.6 9.3 0.6 Bagged 66.1a 139.0a 9.1 0.7b 61.2 125.7 8.6 0.6 ‘Julyprince’ Control 69.2 183.9 10.7 0.9 59.8 111.3 11.1 0.8 Bagged 72.1 202.9 10.2 0.9 60.2 112.3 10.7 0.8 z Values within columns and season followed by different letters indicate significant differences at P<0.05.

Results 1). For ‘Crimson Lady’ in 2016 there were Organic peaches. For ‘Crimson Lady’ in 29% marketable fruit for the control treat- 2015, but not 2016, the bagged treatment ment vs. 30% for the bagged treatment. For had significantly higher values for size and ‘Julyprince’, the control treatment had 88% mass, and lower TA than control (Table 2). marketable fruit and the bagged had 99% ‘Julyprince’ showed no significant difference marketable fruit (P ≤ 0.05; Fig. 1). in fruit quality between treatments in both Regarding postharvest disease assessment, years. The color of the control and bagged- the control ‘Crimson Lady’ fruit in 2015 had 10d peach fruit was comparable; however, 82% and 87% rotten fruit, and the bagged bagged peaches had less intense red blush fruit showed 38% and 97% rotten fruit three (data not shown). and seven days after harvest, respectively In 2015 for ‘Crimson Lady’, 69% of con- (Fig. 2). Disease incidence of ‘Julyprince’ trol fruit were marketable vs. 82% of bagged was lower in both experimental years fruit. For ‘Julyprince’, 92% and 95% of the likely due to the hotter and warmer summer fruit were marketable for the control and months suppressing disease development. bagged treatments, respectively, and the dif- ‘Julyprince’ had 16% and 38% rotten fruit for ference was not significant (P ≤ 0.05; Fig. the control treatment, and 7% and 28% for

Control 2015 2016 a 100 a a Bagged b a 80 b

60

40 a a

Marketable Fruit (%) 20

0 'Crimson Lady' 'Julyprince' 'Crimson Lady' 'Julyprince' Figure 1. Percentage of marketable (unblemished) fruit at harvest for control (non-striped bar) and bagged (striped bar) treatments for ʻCrimson Ladyʼ and ʻJulyprinceʼ for two years. These data were collected at Figurethe organic 1: Percentage farm. Values of marketablewithin year (unblemished) and season of fruitmaturity at harves followedt for controlby common (non- stripedletters dobar) not and differ baggedsignificantly (striped at thebar) 5% treatments level. for ‘Crimson Lady’ and ‘Julyprince’ for two years. These data were collected at the organic farm. Values within year and season of maturity followed by common letters do not differ significantly at the 5% level.

42 Journal of the American Pomological Society

2015 'Crimson Lady' 2016 'Crimson Lady' 100 a a a a 80 b a 60 a b 40

20

0 2015 'Julyprince' 2016 'Julyprince' 100 Control 80 Bagged

Disease Incidence (%) Incidence Disease 60 a 40 a a 20 a a a a a 0 3 7 3 7 Days in Storage Figure 2. Incidence of brown rot after 3 and 7 days of storage at 20oC. Bagged and control treatments are shown forFigure ʻCrimson 2: Incidence Ladyʼ of and brown ʻJulyprinceʼ rot after 3 and for 7 two days years. of storage These at 20 data˚C. Baggedwere collected and contr olat treatmentsthe organic are farm. Values withinshown foryear ‘Crimson and season Lady’ ofand maturity ‘Julyprince’ followed for two years.by common These data letters were do collected not differ at the significantly organic farm at the 5% level.Values within year and season of maturity followed by common letters do not differ significantly at the 5% level. the bagged treatment, respectively. In 2016 in 2015 had significantly higher SSC for the ‘Crimson Lady’ had 47% and 72% rotten control vs. the bagged fruit (Table 3). Bagging fruit for the control treatment, and 61% and significantly increased fruit size and mass for 88% rotten fruit for the bagged treatment, ‘Crimson Lady’ and ‘Julyprince’ in 2016 respectively. Virtually no storage rot was (Table 3). All fruit was marketable at harvest observed in any treatment for ‘Julyprince’. for both years and both treatments (data not None of the fruit showed symptoms of insect shown). Likewise, no post-harvest rot was damage. observed, except for ‘Crimson Lady’ in 2016 Conventional Peaches. ʻCrimson Lady’ when the fruit were improperly bagged after

Table 3: Quality parameters for fruit at the conventional farm including diameter (mm), mass (g), SSC (%) and TA (%). Data shown for ʻCrimson Lady’ and ‘Scarletprince’ (2015) and ʻJulyprinceʼ (2016) for two years. 2015 2016 Diam. Mass SSC TA Diam. Mass SSC TA ‘Crimson Lady’ Control 60.1 113.8 8.1az 0.8 55.0b 93.9b 9.8 0.6 Bagged 59.3 108.6 7.5b 0.8 66.4a 161.6a 9.9 0.6 ‘Julyprince’ Control 59.7 121.4 12.2 0.9 70.9b 183.6b 12.6 0.7 Bagged 61.8 135.7 12.3 0.9 74.5a 212.8a 12.5 0.8 z Values within columns and season followed by different letters indicate significant differences at P<0.05. Peach 43 a rain event (data not shown). Color develop- lower SSC and TA than control fruit but dif- ment in conventional peaches was the same as ferences were not significant for all years and described for organic peaches. locations. Consumer Surveys. Surveys at the farmers’ For the conventional farm the only dif- market showed that when individuals were ference in fruit quality occurred in 2015 for asked about their preference of the unmarked ‘Crimson Lady’ for SSC and for size and control and bagged peaches, 62% preferred mass in 2016. For peach there have been the control fruit, 23% indicated no prefer- some instances where the fruit SSC was ence, and 15% preferred the bagged fruit, lower for bagged fruit vs. control fruit (Li et which had slightly less red color. When the al., 2001) but our experiments did not show concept of bagging was explained to the con- that trend. Lower SSC was reported for other sumer and they were subsequently asked the bagged fruits such as pear, apple, mandarin, same questions, 92% preferred the bagged and plum (Chen et al., 2012; Hiratsuka et al., fruit, 7% preferred the controls, and 1% had 2012; Lin et al., 2008; Murray et al., 2005). no preference. At the survey carried out near Treatments did not affect marketable yield in Clemson campus, 93% of participants ini- the conventionally grown fruit because fruit tially preferred the control peaches, but most from both treatments did not rot preharvest changed their minds and 86% then preferred or postharvest likely due to successful IPM bagged fruit. The average increase in price implementation at this farm. individuals at the farmers market were will- Color development of fruit while bagged ing to pay was $0.38 extra per pound for depends on factors such as time of bagging, bagged fruit, with 35% people answering type of bag used, and timing of bag removal. $0.50 or more (85% of these were women) Bagged peaches in this study did show a dif- whereas at the Campus survey the average ference in red color intensity but not as dras- premium was $0.25 extra per pound. tic as reported in other fruits. For instance, apples had drastic color reduction, especially Discussion red-finish apples, if the bag was kept on the Bagging did not significantly affect fruit fruit until harvest. If the bag was removed size, mass, SSC, and TA consistently between three to four days before harvest, the expo- years or location. Bagged fruit for ‘Crim- sure to UV light gave the fruit a deeper red son Lady’ at the organic farm in 2015 and color than before (Bai et al., 2016; Fan and both ‘Crimson Lady’ and ‘Julyprince’ from Mattheis, 1998; Shen et al., 2014). For this the conventional farm in 2016 had a larger reason, double layer bags can be used to help mass and size than non-bagged fruit. Since improve color. The first layer is opaque and commercial peach fruit sizing is categorical has 0% transmittance of light but can be re- (based on diameter) these differences did moved before harvest to allow light to the not change the fruit size category and would fruit. The second bag still gives the fruit the have no impact on price for the grower. Fruits same protection while allowing light trans- respond differently to bagging in term of size mittance. This practice cannot be done in the and mass, for instance some fruits such as United States due to high cost of labor of par- date palms, mangos, longan, and carambola tial and eventually complete unbagging the had increased size when bagged (Harshash fruit (Huang et al., 2009; Xu et al., 2010). and Al-Obeed, 2010; Watanawan et al., 2007; Regarding postharvest quality, the bags Xu et al., 2008; Yang et al., 2009), but other helped reduce the amount of postharvest fruits such as loquat and some ‘Conference’ brown rot for three days at the organic farm in pears had reduced size (Hudina and Stampar, 2015 but results were not consistent for both 2012; Xu et al., 2010). On the other hand, years. Numerous factors influence posthar- bagged ‘Crimson Lady’ fruit tended to have vest brown rot infections such as amount of 44 Journal of the American Pomological Society rain prior to or at harvest (which was the main To understand the consumer’s view on this cause of yield reduction for ʻCrimson Ladyʼ new agriculture practice, surveys were con- in 2016), number and type of fungicide sprays ducted. When asked, surveyed individuals prior to harvest, amount of inoculum pres- initially tended to prefer the deep red color ent, and postharvest treatments. The data for of the control peaches but after the concept the conventional farm are not shown because of bagging was explained, color was no lon- there was no postharvest rot in any treatment ger the decisive factor. The overwhelming even after 7 days of incubation. consensus was that fruit exposed to less pes- Marketable yield (i.e. amount of har- ticides were preferred. Consumers indicated vested fruit without blemishes caused by willingness to pay a 0.38 cents per pound av- pests or diseases) of bagged fruit was higher erage premium for bagged peaches over the than that of control fruit in the organic set- current price. These data indicate growers ting. The overall decrease in yield observed could benefit of marketing strategies directed for ‘Crimson Lady’ in 2016 was a conse- to niche markets. If the labor force was avail- quence of bagging the fruit immediately af- able, the increase in price could help offset ter a rainfall. The rain probably washed off the cost of bags and extra labor involved. For the previous application of OMRI-approved instance, if a farmer harvests 113 kg of fruit fungicide spray and fruit were bagged while per tree and sells them at a price increase of still wet. We did not see symptoms of insect $0.84/kg, the potential profit per tree would damage, and fruit rot diseases were the main be $80. On the other hand, based on our field factor that caused reduction in marketable trials, it takes about one hour for a worker to yield. Also, bagging did not affect harvest bag an entire mature tree. At the current price time (data not shown). Wang et al. (2010) re- of $12.00 per hour for labor, and bags them- ported that peaches of the cultivar ‘Wanmi’ selves running about $3.00 for one mature had a harvest date around one month ahead tree (400 fruit), the farm has the potential to of schedule when bags were used but we did profit for up to $65.00 per tree from bagging, not observe any change in ripening between depending on marketing strategies and num- bagged and non-bagged fruit. ber of intermediaries. This amount can also Bagging is an agricultural practice that increase if the farm’s packout increased by can be costly but can also produce some cost using the bags. Positive results were reported savings. The decision to use bags to help ex- for bagged lychee fruit in Hawaii; when ly- clude pests and diseases is based on factors chee fruit was bagged there was an increase such as packout yield, organic certification, in marketable yield from 57% to 84% and scale of farm, labor availability, increase in had a potential to earn $173 per tree if the profitability, and costs. The addition to add- fruits were bagged (Kawabata and Nakamo- ing fruit bags to the farm’s IPM program to, 2013). could be costly due to increased labor and On the other hand, conventional farms have material cost, but could also reduce the already low losses due to pests and diseases amount and frequency of sprays of expen- because of strict spray schedules and use of sive pesticides needed throughout the season efficacious chemicals. Thus, the addition of and boost prices. However, the trees would bags is not expected to increase marketable still have to be protected throughout the sea- yield for these farms. Nevertheless, they may son from damaging insects such as scale or find opportunities when marketing these fruit damaging pathogens such as Pseudomonas as prime fruit, for instance, “grown-in-bag” sp., Botryosphaeria spp., Phomopsis sp., and peaches. The idea of fruit bagging appeals to Leucostoma sp. Those pathogens may be consumers since peaches are known as one controlled, however, with cheaper products of the most highly sprayed fruit on the mar- and in greater application intervals. ket; thus, bagged peaches could be sold at a Peach 45 premium at some niche markets. phenolic compounds in the peel and flesh of ‘Gold- One of the downsides to using bags is the en Delicious’, ‘Red Delicious’ and ‘Royal Gala’ loss of crop if they are not utilized correct- apples. Sci. Hort. 142:68-73. Faci, J.M., E.T. Medina, A. Martínez-Cob, and J.M. ly. The bags need to be placed on the fruit Alonso. 2014. Fruit yield and quality response of when they are dry and shortly after they were a late season peach orchard to different irrigation sprayed. If there is any moisture left on the regimes in a semi-arid environment. Agric. Water fruit and tree from dew or a storm event the Manage. 143:102-112. day before and the fruit has not been sani- Fan, X. and J.P. Mattheis. 1998. Bagging ‘Fuji’ apples tized, a small amount of disease inoculum during fruit development affects color development can cause rot and other detrimental problems and storage quality. HortScience 33:1235-1238. Frett, T.J., K. Gasic, J.R. Clark, D. Byrne, T. Gradziel, for the fruit. Training workers to properly and C. Crisosto. 2012. Standardized phenotyp- secure bags onto branches is also important ing for fruit quality in peach [Prunus persica (L.). for the success of the bags. If tied properly Batsch]. J. Amer. Pomol. Soc. 66:214-219. the bags are able to withstand high winds and Gerdts, M. and J.K. Clark. 1979. Caprification: a heavy thunderstorms. unique relationship between plant and insect. Calif. Overall adding fruit bags to the IPM regi- Agric. 33:12-14. ment is a double-edged sword but may be Harshash, M.M. and R.S. Al-Obeed. 2010. Effect of bunch bagging color on yield and fruit quality of beneficial for some farms. The bags do not af- date palm. Am. Eurasian J. Agric. Environ. Sci. fect fruit quality other than slightly reducing 7:312-319. red color, but could help improve yield for Hiratsuka, S., Y. Yokoyama, H. Nishimura, T. Miyaza- the farm if utilized properly. The integration ki, and K. Nada. 2012. Fruit photosynthesis and of bags could also help reduce the amount of phosphoenolpyruvate carboxylase activity as affect- pesticides needed, which in turn may reduce ed by lightproof fruit bagging in Satsuma mandarin. costs. The “grown-in-bag” fruits are viewed J. Amer. Soc. Hort. Sci. 137:215-220. Hofman, P.J., L.G. Smith, D.C. Joyce, G.I. Johnson, positively by consumers and could potential- and G.F. Meiburg. 1997. Bagging of mango (Man- ly be sold at a higher price to help offset the gifera indica cv. ‘Keitt’) fruit influences fruit qual- price of the bags. ity and mineral composition. Postharvest Biol. Technol. 12:83-91. Acknowledgments Huang, C., B. Yu, Y. Teng, J. Su, Q. Shu, Z. Cheng, and The authors thank the Southern IPM Cen- L. Zeng. 2009. Effects of fruit bagging on coloring ter and Southern SARE (project #OS16-100) and related physiology, and qualities of red Chi- nese sand pears during fruit maturation. Sci. Hort. for funding this research. 121:49-158. Hudina, M. and F. Stampar. 2012. Effect of fruit bag- Literature Cited ging on quality of ‘Conference’ pear (Pyrus com- Amarante, C., N.H. Banks, and S. Max. 2002. Effect munis L.). Eur. J. Hort. Sci. 76:176-181. of preharvest bagging on fruit quality and posthar- Kawabata, A. and S.T. Nakamoto. 2013. Lychee fruit vest physiology of pears (Pyrus communis). N. Z. J. bagging for commercial and home growers. Univ. Crop Hort. Sci. 30:99-107. Hawai’i at Manoa. F&N-14. Ames, G. 2012. Peaches: organic and low-spray pro- Kim, Y.H., H.Y. Kim, C.K. Youn, S.J. Kweon, H.J. duction. ATTRA, NCAT. Publication IP047. Jung, and C.H. Lee. 2008. Effects of bagging mate- Bai, S., P.A. Tuan, T. Saito, C. Honda, Y. Hatsuyama, rial on fruit coloration and quality of ‘Janghowon A. Ito, and T. Moriguchi. 2016. Epigenetic regu- Hwangdo’ peach. Acta Hort. 772:81-86. lation of MdMYB1 is associated with paper bag- Li, S.H., M. Génard, C. Bussi, J.G. Huguet, R. Habib, ging-induced red pigmentation of apples. Planta J. Besset, and J. Laurent. 2001. Fruit quality and 244:573-586. leaf photosynthesis in response to microenviron- Bessin, R. and J. Hartman. 2003. Bagging apples: al- ment modification around individual fruit by cov- ternative pest management for hobbyists. Univ. of ering the fruit with plastic in nectarine and peach Kentucky Coop. Ext. Serv. ENTFACT-218. trees. J. Hort. Sci. Biotechnol. 76:61-69. Chen, C.S., D. Zhang, Y.Q. Wang, P.M. Li, and F.W. Lin, J., Y. Chang, Z. Yan, and X. Li. 2008. Effect of Ma. 2012. Effects of fruit bagging on the contents of bagging on the quality of pear fruit and pesticide 46 Journal of the American Pomological Society

residues. Acta Hort. 772:315-318. Signes, A.J., F. Burlo, F. Martínez-Sánchez, and A.A. Muchui, M.N., F.M. Mathooko, C.K. Njoroge, E.M. Carbonell-Barrachina. 2007. Effects of preharvest Kahangi, C.A. Onyango, and E.M. Kimani. 2010. bagging on quality of black table grapes. World J. Effect of perforated blue polyethylene bunch cov- Agric. Sci. 3:32-38. ers on selected postharvest quality parameters of Wang, Y., C. Yang, C. Liu, M. Xu, S. Li, L. Yang, and tissue-cultured bananas (Musa spp.) cv. Williams Y. Wang. 2010. Effects of bagging on volatiles and in Central Kenya. J. Stored Prod. Postharvest Res. polyphenols in “Wanmi” peaches during endocarp 1:29-41. hardening and final fruit rapid growth stages. J. Murray, X.J., D.M. Holcroft, N.C. Cook, and S.J.E. Food Sci. 75:S455-S460. Wand. 2005. Postharvest quality of ‘Laetitia’ and Watanawan, A., C. Watanawan, and J. Jarunate. 2007. ‘Songold’ (Prunus salicina Lindell) plums as af- Bagging ‘Nam Dok Mai #4’ mango during devel- fected by preharvest shading treatments. Posthar- opment affects color and fruit quality. Acta Hort. vest Biol. Technol. 37:81-92. 787:325-328. Perez, A. and K. Plattner. 2013. Fruit and tree nuts out- Xu, C.X., H.B. Chen, R.Y. Huang, and Y.J. He. 2008. look: commodity highlight. Organic fruit and ber- Effects of bagging on fruit growth and quality of ries. U.S.D.A., Econ. Res. Serv. FTS-355. carambola. Acta Hort. 773:195-200. Sharma, R.R., R.K. Pal, R. Asrey, V.R. Sagar, M.R. Xu, H.-X., J.-W. Chen, and M. Xie. 2010. Effect of dif- Dhiman, and M.R. Rana. 2013. Pre-harvest fruit ferent light transmittance paper bags on fruit qual- bagging influences fruit color and quality of apple ity and antioxidant capacity in loquat. J. Sci. Food cv. Delicious. Agric. Sci. 4:443-448. Agric. 90:1783-1788. Shen, J.Y., L. Wu, H.R. Liu, B. Zhang, X.R. Yin, Y.Q. Yang, W.H., X.C. Zhu, J.H. Bu, G.B. Hu, H.C. Wang, Ge, and K.S. Chen. 2014. Bagging treatment influ- and X.M. Huang. 2009. Effects of bagging on fruit ences production of C6 aldehydes and biosynthesis- development and quality in cross-winter off-season related gene expression in peach fruit skin. Mol- longan. Sci. Hort. 120:194-200. ecules 19:13461-13472.

Correction: On the web page (https://www.pubhort.org/aps/72/72.htm#2018_7_3) for volume 72(3), for the article by Mehmet Sutyemez, Akide Özcan, and Ş. Burak Bükücü “Walnut Cul- tivars Through Cross-Breeding: ‘Dirilis’ and ’15 temmuz’”, the last author’s name was misspelled as Özkan and should be Özcan. 47

Journal of the American Pomological Society 73(1): 47-52 2019 The Importance of Being ʽBoysenʼ: Examining Genotypic Variation With SSR Markers Katie A. Carter1, Jason D. Zurn2, Nahla V. Bassil2, Chad E. Finn3, and Kim E. Hummer2,4

Additional index words: blackberry, hybrid berry, simple sequence repeat markers, SSR, germplasm

Abstract Multiple variants of ‘Boysen’ have been horticulturally recognized since the cultivar was introduced in 1938. An improved 8-SSR DNA fingerprinting set was applied to ‘Boysen’ and ‘Logan’ variants and their putative par- ents obtained from private growers and commercial nurseries. A ‘Boysen’ genotype likely to be from the original clone was identified. ‘Boysen’ appears to have resulted from ‘Logan’ x ‘Austin Mayes’. ‘Lucretia’ is not a parent. Continuing work will establish pedigree or relational links for wild and cultivated blackberries. ‘Boysen’ blackberry was introduced in vice) senior pomologist and berry breeder in 1935 in California and became a standard Beltsville, MD, of the existence of this ex- in the industry through the 1900s (Jennings, citing cultivar. Darrow visited California to 1988). This berry has strong public recog- meet Boysen and identify the original plants. nition, although growers have shifted away Walter Knott trialed these plants and found from large-scale production of this cultivar. the fruit to be large and flavorful. He began It is an important founding cultivar for the selling plants in 1935 prior to Darrow’s offi- Pacific Northwest blackberry industry (Hall cial release of the cultivar in June 1938 (Stel- and Funt, 2017). lar, 1937; Wood et al. 1999). The origin of ‘Boysen’ remains obscure. It Over the years, variants of ‘Boysen’ were was developed by Rudolph Boysen, a hob- selected internationally. These variants in- cluded the New Zealand selection ‘Riwaka byist plant breeder who reportedly crossed a Choice’ and an Oregon selection named raspberry, a blackberry, and ‘Logan’ to cre- ‘Boysen 43’ that became an industry stan- ate ‘Boysen’ (Stellar, 1937). Jennings (1988) dard in the 1990s (J. DeFrancesco, personal later hypothesized that ‘Boysen’ resulted communication 2018). from a cross between ‘Logan’ and a trail- The USDA-ARS National Clonal Germ- ing blackberry such as ‘Lucretia’ or ‘Aus- plasm Repository (NCGR) Rubus germ- tin Mayes’ (Jennings, 1981; 1988). He gave plasm includes heritage cultivars, such as the plants of his new hybrid to D.W. Coolidge ‘Boysen’ variants, along with crop wild rela- of Coolidge Rare Plant Gardens (Pasadena, tives. DNA-based fingerprinting of this valu- CA). Coolidge then informed Dr. George able germplasm complements morphological Darrow, USDA Bureau of Plant Industry characterization. Simple sequence repeat (now the USDA-Agricultural Research Ser- (SSR) markers are used to establish genetic

1 Department of Horticulture, Oregon State University, Corvallis, Oregon 97331 2 United States Department of Agriculture Agricultural Research Service, National Clonal Germplasm Reposi- tory, 33447 Peoria Road, Corvallis, Oregon 97333-2521 3 U.S. Department of Agriculture, Agricultural Research Service, Horticultural Crops Research Unit, 3420 NW Orchard Avenue, Corvallis, OR 97330 4 The authors thank USDA ARS CRIS 2072-21000-049-00D and Jeanette Boysen Fitzgerald for support of this research. In addition, technical assistance by April Nyberg, Jill Bushakra and Mary Peterson is appreciated. 4 Corresponding author [email protected] 48 Journal of the American Pomological Society identify and familial relationships and assess package ‘Poppr’ v 2.6.1 (Bruvo et al., 2004; germplasm diversity (Gökirmak et al., 2009; Paradis et al., 2004; Kamvar et al., 2014; R Bassil et al., 2015). Recently, an 8-SSR fin- Core Team, 2015) as described in Zurn et al. gerprinting set was developed for blackberry (2018). Allele compositions of samples were (Zurn et al., 2018) that could clarify the par- compared to determine if historical pedigree entage of ‘Boysen’ and its relationship with reports could be supported. other heritage hybrid berries. The objectives of this study were to assess Results and Discussion variation in ‘Boysen’ clones from different Misidentified hybrid berries abound sources and examine putative pedigree rela- The eight SSRs amplified in all 29 -indi tionships between other early hybrid berries viduals and generated 16 unique genotypes. using SSR markers. Many hybrid berries with different names had matching fingerprints (Fig. 1). ‘Young Materials and Methods Thornless’ from PF3, which originated in DNA extraction a private garden in Washington State, had Young leaf tissue was collected from fingerprints that matched that of ‘Thornless ‘Boysen’, ‘Logan’, and associated cultivars Logan’ and is likely misidentified. Speci- from the NCGR Rubus collection, three inde- mens for ‘Nectar Boysenberry’ and ‘Austin pendent private grower sites (PF1, PF2, and Mayes’ from N2 in 2012 had matching gen- PF3), the USDA-ARS National Clean Plant otypes (Fig. 1), and were hexaploid (Hum- Network (NCPN), the USDA-ARS, Horti- mer et al., 2015) when expected to be octo- cultural Crops Research Unit (Corvallis, OR) ploid and septaploid, respectively (Jennings, (ORUS) breeding program, and four com- 1988). Their identity is unknown at this time. mercial nurseries (N1, N2, N3, and N4; Ta- ‘Austin Dewberry’ from N2 obtained in 2018 ble 1). Specimens for ‘Nectar Boysenberry’ matched a ‘Boysen’ genotype and is prob- and ‘Austin Mayes’ were acquired from N2 ably mislabeled. in 2012. New ‘Austin Dewberry’ and ‘Nec- tar Boysenberry’ plants were acquired from Five ‘Boysen’ genotypes N2 in 2018 due to inconsistencies. DNA was Five genotypes were observed for “‘Boy- extracted using a modified Puregene proto- sen’-like” accessions (Fig. 1). One ‘Boysen’ col (Gentra Systems Inc. Minneapolis, MN, genotype was observed for the ‘Boysen’ ac- U.S.A.; Gilmore, 2011). cessions from PF1 and PF2, the ‘Thornless Boysen’ from N1, and the ‘Austin Dewberry’ Genotyping and data analysis from N2 obtained in 2018. The second geno- All samples were genotyped with an 8-SSR type was exhibited by ‘Riwaka Choice’, the multiplexed assay developed for blackberries three plants from the NCPN, and ‘Boysen (Zurn et al., 2018). DNA was PCR amplified ORUS’. ‘Boysen 43’, ‘Boysenberry’ from using the Type-it® Microsatellite PCR Kit the NCGR, and ‘Boysen’ from N3 each had (Qiagen, Inc., Valencia, CA, U.S.A.) as unique genotypes (Fig. 1). described by Zurn et al., 2018. PCR products The identical allele compositions, Califor- were separated by capillary electrophoresis nian origin, and history of the ‘Boysen’ plants with a Beckman CEQ 8000 (Beckman from PF1 and PF2 suggest that these likely Coulter, Inc.). Electropherogram visualization represent the original heritage ‘Boysen’ gen- and allele calling was performed with the CEQ otype. The PF1 grower is the granddaughter 8000 software fragment analysis module. of Rudolph Boysen, and the plant originated An unrooted neighbor-joining phylogeny from the Napa Valley Creek stand originally was constructed using the allele sharing model established by Rudolph Boysen (Jeanette of R v 4.3.3 Bruvo’s infinite allele, using the Fitzgerald, Personal Communication; Court- Boysen 49

Table 1. ʻBoysenʼ and associated heritage cultivars examined with an 8-SSR fingerprinting set. PI and local numbers are listed for accessions acquired from the USDA NCGR. Thorniness is indicated with ʻYʼ and thornlessness with a ʻTʼ. Thorniness Pedigree/origin based Name as received (T/Y) Ploidy Source Local number PI number on name as received Austin_Dewberry_N2_18 Y 2n=8x=56 Nursery 2 n/a n/a R. baileyanus x R. argutus

Austin_Mayes_N2_12 Y 2n=8x=56 Nursery 2 CRUB 2653.001 PI 672692 R. baileyanus x R. argutus

Austin_Thornless T 2n=8x=56 NCGR CRUB 357.001 PI 553292 Thornless seedling/sport of 'Austin Mayes'

Bauers_Thornless Logan T 2n=6x=42 NCGR CRUB 81.001 PI 553258 Thornless sport of 'Logan'

Boysen_43 Y 2n=7x=49 NCGR CRUB 1108.001 PI 553341 Variant of 'Boysen' Boysen_ORUS Y 2n=7x=49 USDA ARS Lab 1 n/a n/a ‘Logan' x unknown trailing blackberry Boysen_NCPN1 Y 2n=7x=49 USDA ARS Lab 2 n/a n/a ‘Logan' x unknown trailing blackberry Boysen_NCPN2 Y 2n=7x=49 USDA ARS Lab 2 n/a n/a ‘Logan' x unknown trailing blackberry Boysen_NCPN3 Y 2n=7x=49 USDA ARS Lab 2 n/a n/a ‘Logan' x unknown trailing blackberry Boysen_PF1 Y 2n=7x=49 CA Private Farm 1 n/a n/a ‘Logan' x unknown trailing blackberry

Boysenberry Y 2n=7x=49 NCGR CRUB 1025.001 PI 553336 ‘Logan' x unknown trailing blackberry Boysen_N3 Y 2n=7x=49 Nursery 3 CRUB 2894.001 PI 687271 ‘Logan' x unknown trailing blackberry

Lincoln_Logan T 2n=6x=42 NCGR CRUB 737.001 PI 553314 TC derived, thornless 'Logan'

Logan_BE1132 T 2n=6x=42 NCGR CRUB 2535.001 PI 687013 ‘Aughinbaugh' x 'Red Antwerp'

Loganberry_seedling1 Y 2n=6x=42 NCGR CRUB 2657.002 PI 672696 ‘Logan' x unknown

Loganberry_seedling2 Y 2n=6x=42 NCGR CRUB 2657.001 PI 672696 ‘Logan' x unknown

Lucretia Y 2n=7x=49 NCGR CRUB 74.001 PI 553253 R. baileyanus x R. argutus

Nectar_Boysenberry_N2_12 Y 2n=7x=49 Nursery 2 CRUB 2654.001 PI 672693 ‘Young' x unknown; 'Boysen' variant Nectar_Boysenberry_N2_18 Y 2n=7x=49 Nursery 2 n/a n/a ‘Young' x unknown; 'Boysen' variant

Nectarberry Y 2n=7x=49 NCGR CRUB 1864.001 PI 618436 ‘Young' x unknown; 'Boysen' variant

Phenomenal Y 2n=6x=42 NCGR CRUB 1817.001 PI 618478 ‘Aughinbaugh' x 'Cuthbert'

Riwaka_Choice Y 2n=7x=49 NCGR CRUB 2207.001 PI 638239 ‘Boysen' selection in New Zealand Thornless_Boysen_N1_18 T 2n=7x=49 Nursery 1 CRUB 2892.001 PI 687270 Thornless sport of 'Boysen' Thornless_Logan_N1_18 T 2n=6x=42 Nursery 1 n/a n/a Thornless sport of 'Logan' Thornless_Logan_N3 T 2n=6x=42 Nursery 3 n/a n/a Thornless sport of 'Logan' Logan_N4 T 2n=6x=42 Nursery 4 n/a n/a Thornless sport of 'Logan'

Young Y 2n=7x=49 NCGR CRUB 131.001 PI 553266 ‘Phenomenal' x 'Austin Mayes'

Young_Thornless_PF3 T 2n=6x=42 NCGR CRUB 2003.001 PI 618391 Thornless sport of 'Young' Boysen_PF2 Y 2n=7x=49 CA Private Farm 2 CRUB 2809.001 PI 684800 ‘Logan' x unknown trailing blackberry

ney, 2016). The ‘Boysen’ plant from the relationship between ‘Boysen’, ‘Young’ and PF2 grower was originally purchased from ‘Austin Thornless’ (Nybom and Hall, 1991). Knott’s Berry Farm about 70 years ago and Because we were unable to acquire a reliable was grown in Buena Vista, California. ‘Austin Mayes’ genotype, ‘Austin Thorn- ‘Thornless Boysen’ from N1 is likely a less’, a thornless sport of ‘Austin Mayes’ clonal thornless variant of the heritage geno- (Butterfield 1928), was used as its proxy in type because these samples are indistinguish- allele comparisons. Two ‘Logan’ genotypes able by the 8-SSR assay. Clonal variants are were observed; one is exhibited by accession rarely distinguished by SSRs (Bassil and Logan-BE1132 (thornless) from Bologna, Postman, 2010; Cretazzo et al., 2010). Other Italy, and the other is seen in six other thorn- ‘Boysen’ variants differed from the heritage less variants of ‘Logan’. Each ‘Logan’ geno- genotype by multiple alleles for at least five type was assessed for parentage of ‘Boysen’ of the eight SSR loci. The level of variation using either ‘Lucretia’ or ‘Austin Thornless’ observed is typical of a parent-offspring or as the pollen parent. With Logan-BE1132 as half-sibling relationship. the maternal parent and ‘Lucretia’ as the pol- len parent, the hypothesized original ‘Boy- Parentage of ‘Boysen’ sen’ genotype had unique alleles not found in Historically, ‘Boysen’ was hypothesized either parent at six SSR loci. When ‘Austin by Jennings to be a cross between ‘Logan’ Thornless’ was the pollen parent, unique al- and an unknown trailing blackberry similar leles were only seen at two loci. The original to ‘Lucretia’ (1988) or ‘Austin Mayes’ (1981, ‘Boysen’ genotype also had unique alleles at 1988). DNA fingerprinting uncovered a close six loci when assessing the ‘Thornless Logan’ Journal50 of the American PomologicalJournal Society of the xx:xx American-xx. 2018. Pomological Society

Figure 1. Neighbor joining cluster analysis of 29 hybrid berry accessions based on Bruvoʼs distance from an 8-SSR blackberry fingerprinting set. Multiple ʻBoysenʼ genotypes, ʻLoganʼ genotypes, and hybrid berry cultivars from private farms (PF) and nurseries (N), some of which are likely mislabeled, are illustrated. ORUS refers to the Oregon, USDA-ARS Horticultural Crop Research program in Corvallis, OR. NCPN is the National Clean Plant Network in Corvallis, OR. Hybrid berry plants were obtained from four nurser- ies (N1, N2, N3 and N4), three private farms (PF1, PF2, and PF3) and the USDA-ARS National Clonal Germplasm Repository collection in Corvallis, OR (NCGR). Thorniness is indicated by black font while lack of thorns is indicated by grey font. The scale bar measures Bruvo's genetic distance along each branch. 9

x ‘Lucretia’ cross. However, ‘Thornless Lo- the maternal parent of ‘Boysen’. ‘Lincoln gan’ x ‘Austin Thornless’ could have resulted Logan’ is a genetically thornless cultivar de- in the hypothesized original ‘Boysen’ geno- rived using tissue culture from ‘Thornless type. We can eliminate ‘Lucretia’ as a pos- Logan’ (Hall et al., 1986). ‘Thornless Logan’ sible parent and infer that the pollen parent may be an abbreviated name adopted by the was ‘Austin Mayes’ or a close relative. This industry for ‘Bauer Thornless Loganberry’, is supported by attempts in New Zealand to explaining the identical fingerprints of these re-create ‘Boysen’ from crossing ‘Austin three accessions. The original ‘Logan’ culti- Thornless’ with ‘Lincoln Logan’ (Hall et al., var was thorny, but we were unable to acquire 2002) and previous DNA fingerprinting indi- any ‘Logan’ types with prickles after search- cating a close relationship between ‘Austin ing local nurseries and national breeding pro- Thornless’ and ‘Boysen’ (Nybom and Hall, grams. This genotype may have been lost in 1991). The six accessions with the ‘Thorn- favor of the thornless type. Logan-BE1132, less Logan’ genotype are likely true thornless as well as the two Loganberry seedlings 1 variants of the original ‘Logan’ cultivar due and 2 are possibly ‘Logan’ offspring, as they to their consistent genotype and the plausi- share alleles with ‘Logan’ at all loci but have bility that a cultivar with this genotype was unique alleles, putatively from the other par- ent, at multiple loci. Boysen 51

Alternatively, the ‘Lucretia’, ‘Austin enburg,. 2004. A simple method for the calculation Mayes’, and ‘Logan’ plants evaluated may of microsatellite genotype distances irrespective of not represent historical specimens of these ploidy level. Mole. Ecol. 13: 2101-2106. Butterfield, H.M. 1928. The origin of certain thornless cultivars. Mutations and propagation er- blackberries and dewberries. J. Hered. 19:135-138 rors could have led to different genotypes. Courtney, K. 2016. A story of Napa, the boysenberry As seen in the five ‘Boysen’ genotypes, in- and me. 18 Sept. Napa Valley Register. dividual plants with widely differing geno- Cretazzo, E., S. Meneghetti, M. De Andrés, L. Ga- types can be acquired under the same name. forio, E. Frare, and J. Cifre. 2010. Clone differen- A search of US berry breeding programs and tiation and varietal identification by means of SSR, private nurseries yielded no other reliable AFLP, SAMPL and M-AFLP in order to assess the clonal selection of grapevine: the case study of sources of ‘Logan’, ‘Lucretia’, or ‘Austin Manto Negro, Callet and Moll, autochthonous culti- Mayes’. Other global sources of these heri- vars of Majorca. Ann. App. Biol. 157: 213-227. tage cultivars will be investigated. Gilmore, B.S., N.V. Bassil, K.E. Hummer. 2011. DNA extraction protocols from dormant buds of twelve Relationships of ‘Boysen’ “clones” woody plant genera. J. Amer. Pomol. Soc. 65:201- ‘Nectarberry’, ‘Nectar Boysenberry’, ‘Ri- 207. waka’s Choice’, and ‘Boysen 43’ differed Gökirmak, T., S.A. Mehlenbacher, and N.V. Bassil. 2009. Characterization of European hazelnut (Cory- from the putative ‘Boysen’ heritage genotype lus avellana) cultivars using SSR markers. Genet. at five to seven SSRs and are more likely Resour. Crop Ev. 56: 147-172. seedlings of the heritage genotype or relat- Hall, H. K. and R.C. Funt. 2017. Blackberries and their ed by pedigree than clonal variants. At this Hybrids. Crop Production Science in Horticulture, time parentage inference is not possible in CABI. Boston, MA. polyploids and we rely on comparing allele Hall, H.K., M.J. Stephens, C.J. Stanley, C. Finn, and composition between putative parents and B. Yorgey. 2001. Breeding new ‘Boysen’ and ‘Mar- ion’ cultivars. Acta Hortic. 585: 91-95. offspring to confirm such relationships. Hall, H.K., R.M. Skirvin, and W.F. Braam. 1986. In conclusion, genotyping heritage hybrid Germplasm release of ‘Lincoln Logan’, a tissue cul- blackberry cultivars at eight SSR loci ex- ture-derived genetic thornless ‘Loganberry’. Fruit emplifies the variability in accessions with Var. J. 40:134-135. similar names. Genetic characterization of Hummer, K., N. Bassil, and L. Alice. 2015. Rubus germplasm is necessary to curate true-to-type ploidy assessment. Acta Hortic. 1133:81-88. specimens. Breeders or buyers may need to Jennings, D.L. 1981. A hundred years of loganber- ries. Fruit Var. J. 35: 34-37. confirm the authenticity of desired cultivars Jennings, D. L. 1988. Raspberries and blackberries: because of misidentified plants in the trade. their breeding, diseases and growth, Academic The hypothesized original ‘Boysen’ genotype Press.San Diego, CA. was likely derived from a ‘Logan’ x ‘Austin Kamvar, Z. N., J.F. Tabima, and N.J. Grünwald. 2014. Mayes’ or similar cross. Further research is Poppr: an R package for genetic analysis of popu- planned to contrast morphological and fruit lations with clonal, partially clonal, and/or sexual quality differences for ‘Boysen’ variants. reproduction. PeerJ, 2: e281. Nybom, H. and H.K. Hall. 1991. Minisatellite DNA ‘fingerprints’ can distinguish Rubus cultivars and Literature Cited estimate their degree of relatedness. Euphyti- Bassil, N., A. Nyberg, C. Finn, J. Clark, C. Peace, and ca 53(2):107-114. A. Iezzoni. 2015. Development of a multiplexed Paradis, E., J. Claude, and K. Strimmer. 2004. APE: fingerprinting set in blackberry. Acta Hortic. 1133: analyses of phylogenetics and evolution in R lan- 89-96. guage. Bioinformatics 20: 289-290. Bassil, N. and J.D. Postman. 2010. Identification of R Core Team. 2017. R: a language and environment European and Asian pears using EST-SSRs from for statistical computing. R Foundation for Statis- Pyrus. Genet. Resour. Crop Ev. 57: 357-370. tical Computing, Vienna, Austria. https://www.R- Bruvo, R., N.K. Michiels, T.G. D’souza, and H. Schul- project.org/.

52 Journal of the American Pomological Society

Stellar, O.A. 1937. The Giant Boysenberry Goes Na- with this disease. New Zealand J. Crop Hort. Sci. tional. Better Fruit. February, p. 20. 27: 281-295. Wood, G., M. Andersen, R. Forster, M. Braithwaite, Zurn, J.D., K.A. Carter, M.H. Yin, M. Worthing- and H. Hall. 1999. History of Boysenberry and ton, J.R. Clark, C.E. Finn, and N.V. Bassil. 2018. Youngberry in New Zealand in relation to their Validating blackberry seedling pedigrees and de- problems with Boysenberry decline, the association veloping an improved multiplexed microsatellite of a fungal pathogen, and possibly a phytoplasma, fingerprinting set. J. Amer. Soc. Hort. Sci. 143(5): 381-390. https://doi:10.21273/JASHS04474-18.

Begin well.

End well.

Adams County Nursery recognizes the importance of starting with quality nursery stock. We know it is your goal to produce high quality fruit. We strive to produce quality trees for the commercial industry. Let us help you get started. Begin with us. Begin well.

Adams County Nursery, Inc. • Aspers, PA (800) 377-3106 • (717) 677-4124 fax • email: [email protected] • www.acnursery.com 53

Journal of the American Pomological Society 73(1): 53-61 2019 The Response of ʽMontmorencyʼ Tart Cherry to Renewal Pruning Strategies in a High Density System Sheriden Hansen1, and Brent L. Black2

Additional index words: sour cherry, Prunus cerasus, mechanization

Abstract Tart cherry (Prunus cerasus) production in the U.S. is based on low-density plantings designed to accommo- date large trunk-shake mechanical harvesters. Availability of canopy-shake harvesters adapted to smaller trees could facilitate transitioning to high-density (HD) precocious systems, based on continuous fruiting walls that are highly efficient at light capture. HD systems would require specialized pruning techniques to maintain long-term productivity while maximizing the efficiency of a limited labor supply. Experimental HD orchards with multiple rootstocks, training systems and tree densities were used for investigating renewal pruning strategies. Pruning cuts of predetermined lengths ranging from 0 cm to 25 cm were made on branches of differing size (0.6 cm to 4.7 cm diameter), and renewal growth was monitored for shoot number and length. The minimum stub length for generating at least one renewal shoot was approximately 10 cm. However, this differed somewhat with rootstock and diameter of the cut branch, where the critical length was shorter for larger diameter branches and on the more vigorous rootstock. Results provide guidelines on pruning of HD plantings to renew fruiting wood and maintain productivity. In many temperate fruit tree species, re- Lakso, 2000). As light filters through the newal pruning is used to replace large, rigid canopy, a gradient of light interception is es- branches with smaller, flexible branches with tablished that contributes to both whole tree healthy, young spurs and fruiting shoots. In light interception and light microenviron- apples and peaches, renewal pruning often ment around the fruiting spur. By selectively involves leaving short, 2 cm long bevel cut renewing the largest branches in the canopy, stubs, referred to as Dutch or stub cuts. From light penetration may be increased. This in- this short stub, a bud will break, typically on creased light in the canopy promotes tree the underside of the stub. The resulting shoot health, increased flower bud formation, and will grow with a wide crotch angle and at a the development of quality fruit. flatter angle, effectively replacing or renew- Renewal pruning for improved light dis- ing the cut branch (Robinson, 2003). Re- tribution can increase air circulation and im- newal pruning is an important practice aiding prove spray penetration through the canopy in improvement of light penetration and air- (Sutton and Unrath, 1984; Ferree and Hall, flow in the canopy and helps to maintain the 1980). Tart cherry is susceptible to a host of health of the tree (Cain, 1972). fungal infections, such as powdery mildew It has been well documented in multiple (Podosphaera clandestine), as well as arthro- fruit crops, that dry matter production, fruit pod pests that thrive in dense canopies where size, color, soluble solids concentration air circulation is poor. Selective renewal and total fruit yield are directly related to pruning can improve air circulation, decrease the amount of sunlight intercepted in the canopy humidity, decrease the prevalence of tree (Campbell and Marini,1992; Flore and disease, and at the same time may make crop Layne, 1999; Palmer, 1997; Wünsche and protectant applications more effective by in-

1 Present address: Utah State University Extension - Davis County, Kaysville, UT 84037 2 Corresponding author: [email protected] Plants, Soils and Climate Department, Utah State University, 4820 Old Main Hill, Logan, UT 84322 54 Journal of the American Pomological Society creasing distribution uniformity through the lengths to regrow fruiting wood. Long stub canopy. Hedging is being used to efficiently lengths (>15 cm from the base) in sweet cher- prune and reduce labor costs in the orchard, ries have been found to increase the number but this practice can create a denser outer of flower buds and lateral shoots (Guimond canopy in the trees (Nugent, 2002). Renewal et al., 1998). Nugent (2002) recommended pruning may be needed in conjunction with 10 to 15 cm long stubs to promote renewal hedging to maintain appropriate canopy den- growth in ‘Montmorency’ tart cherry, but did sity. It has been suggested that one-fifth of not cite data from which this recommenda- the largest branches in tart cherry trees be re- tion was based. newed annually to keep wood small, flexible, The objective of this research was to find and fruitful (Crandall, 1979). Maintaining the minimum stub length to generate at least flexible fruiting wood may also be important one new shoot on ‘Montmorency’ tart cherry for over-the-row harvesters that remove fruit in a high density orchard, and to determine from the plant by shaking the canopy instead whether this critical length was affected by of the tree trunk. These canopy shake har- rootstock vigor. Our hypothesis was that tart vesters are commonly used in grapes, rasp- cherry requires 10 cm long stubs to generate berries, blueberries, olives, and some nut a renewal shoot. crops. Fruiting in tart cherries occurs primarily Materials and Methods on spurs on two-year-old and older wood, but Renewal pruning strategies were applied in certain circumstances, may also occur at to a high density (HD) tart cherry orchard the base of one-year-old wood, which can re- planted in 2010 at the Utah Agricultural Ex- sult in blind wood after the first fruiting year. periment Station research farm in Kaysville, In contrast, vegetative buds formed on one- Utah (41°01’16°N latitude, 1328 m eleva- year-old wood result in spurs that will fruit tion, 165 freeze-free days). The orchard con- for several years (Perry et al., 1998). Any sists of combinations of 3 rootstocks and 3 management practice that promotes more training systems with ‘Montmorency’ tart precocious fruiting, potentially risks exces- cherry as the scion cultivar. Rootstocks in- sive flower bud formation on one-year-old cluded the dwarfing Gisela® 3 (Gi.3) and wood, further contributing to the amount of Gisela® 5 (Gi.5), and the commercial stan- blind wood in the canopy. This can be man- dard ‘Mahaleb’. Tree training included a sin- aged with the application of gibberellic acid gle leader, a double leader, and a quad leader, (Anderson et al., 1996) which suppresses with leaders oriented in line with the row to flower bud formation in favor of vegetative facilitate machine harvest. Annual dormant bud formation, resulting in less blind wood pruning was based on a columnarized system formation in the canopy. However, renew- with renewal cuts made back to the leaders al pruning is also required to replace blind in a 3-4 year cycle, patterned after protocols wood with spur-bearing branches by remov- for tall spindle apples (Robinson et al.,2006). ing unfruitful wood and encouraging new po- Briefly, annual dormant pruning involved re- tentially fruitful growth in its place. newal cuts to 2 to 3 of the largest branches Although apples and peaches respond best of each tree. The targeted result was 1, 2 or to cutting branches to at least 2 cm stubs 4 permanent leaders with weaker fruiting (Robinson, 2003), the stub length for tart lateral shoots that are frequently replaced. cherry branch renewal is a topic that has not Row orientation was approximately north- been well researched. Preliminary results in- east to southwest corresponding to the slope dicate that cherries respond to short renewal of the field to improve air drainage. Each cuts differently than apples and peaches, rootstock-training system combination was where cherries appear to need longer stub replicated in ~9 m long plots. Orchard soil Cherry 55 was a Kidman fine sandy loam with 0 to 1 GLM procedure in SAS statistical analysis percent slope. Fertilizer application rates dif- software, version 9.4 (Cary, NC). Data for fered between years, with nitrogen applica- each year were analyzed separately. Means tion rates of 55 kg·ha-1 in 2015 and 25 kg·ha-1 separations were determined using the pdiff in 2016 banded within the tree row. option in the LS-Means statement of SAS. For the purposes of this study, branches Quadratic regression were calculated using identified for renewal during the 2015 and the estimate option in the GLM procedure. 2016 dormant pruning seasons were cut to ~25 cm long. Selected branches represented Results a range of locations within the tree, height Shoot number in the canopy and orientation. The diameter The amount of renewal growth, as deter- of each cut branch was measured at the base mined by the number of new shoots originat- of the branch using hand held calipers, cat- ing from a stub cut, was affected by both the egorized by diameter class (small <1.5 cm, length and the diameter of the cut branch in medium 1.5 to 2.5 cm, and large >2.5 cm), both 2015 and 2016. In 2016, the tree train- flagged, numbered and the diameter record- ing system and rootstock also affected new ed. Approximately equal numbers of flagged shoot number. Except for a marginally sig- branches representing each diameter class, nificant interaction (P < 0.075) between root- rootstock and training system combination stock and stub diameter in 2016, there were were then randomly assigned to a stub length no significant interactions (P < 0.10) among treatment. Stub length treatments varied any factors, and the data are presented as slightly between 2015 and 2016. In 2015, main effects for each factor. target stub lengths were 0, 10, 18, and 25 cm. In both study years, the number of new In order to better determine optimum stub shoots per renewal cut was linearly related length, the number of treatments in 2016 was to the length of the remaining stub (Fig. 1), increased to include target lengths of 0, 5, where the 25 cm stub lengths resulted in 10, 15, 20, and 25 cm. Branches were cut more than 2.5 times more new shoots per cut to the assigned stub length on 23-27 March stub than the 0 cm length. Stub cuts approxi- 2015 and 11-15 April 2016 and assessed for mately 10 cm in length resulted in an average growth in early September in both years. of one new shoot per renewal cut. The num- Growth was evaluated by number and length ber of new shoots was also related to branch of new shoots. diameter in both study years. However, the Overall pruning severity (the number of magnitude of this diameter effect depended branch cuts per tree) was relatively consis- somewhat on the rootstock, where the larg- tent across training system and rootstock est diameter cuts on ‘Mahaleb’ rootstock treatments for both years, with the exception had disproportionately more new shoots of the single-leader trees in 2016. In order to when compared to large diameter cuts on the accommodate the interior space in the over- dwarfing rootstocks (Fig. 2). Another way to the-row harvester, tree height had to be re- visualize this effect is in comparing the stub duced in the single leader system between length that is required for regrowth of an av- the 2015 and 2016 seasons, resulting in more erage of one branch per renewal cut. Linear severe pruning compared to the multi-leader regression was used to calculate this critical systems. stub length for each rootstock and branch di- Data for the number and length of re- ameter combination (Table 1). To regrow a growth were analyzed as a completely ran- single renewal shoot, smaller diameter stubs domized design with 3 branch diameter × 5 on Gi.3 rootstocks required 14 cm stub length branch length × 3 rootstock × 3 training sys- as compared to Gi.3 large diameter stubs that tem factorial treatment structure, using the required 8 cm. In contrast, large diameter 56 Journal of the American Pomological Society

Table 1. Critical stub length required to regrow per stub in the quad leader system (data not one shoot for each rootstock and branch diameter shown). category. Linear regressions were calculated for each combination. Asterisks indicate the R2 of the Shoot length regression. The average length of new shoots is anoth- Stub Diameter er way to quantify renewal growth response. Small Medium Large Average shoot length was affected by root- Rootstock (<1.5 cm) (1.5-2.5 cm) (>2.5 cm) stock, diameter, and stub length in both 2015 Length of stub cut (cm) and 2016. There were significant interactions Gi.3 14.1** 10.6*** 7.8** between training system and stub length (P Gi.5 11.5*** 10.2*** 8.4*** = 0.001) and between training system and Mahaleb 10.2** 8.6* 5.1*** branch diameter (P = 0.038) in 2016, but no *R2 = 0.80-0.89, **R2 = 0.90-0.94, ***R2 = 0.95-1.00 significant interaction in 2015 (P > 0.10). P value for all regressions were <0.0001 The effects of training system and stub length on shoot number is shown in Fig. 3. In general, average new shoot length increased stubs on ‘Mahaleb’ could be as short as 5 cm with stub length up to 10 cm, with minimal and regrow an average of one shoot per cut. increase with stubs longer than 10 cm (Fig. A significant effect of training system on 3). In 2016 for example, average new shoot number of new shoots was found in 2016, but length in the 4-leader system increased from not 2015. Mean number of shoots per stub 5.9 cm to 38.6 cm as stub length increased in the single leader system in 2016 was 1.45 from 0 to 10 cm, but was only 41.9 cm for 290 shoots, compared to 1.02 shoots per stub Figuresin 25 cm stub lengths. The interaction between the double leader system, and 1.26 shoots training system and stub length observed in 291

292 Figure 1. The effect of stub length on regrowth as determined by the number of new shoots formed per branch cut. Values are averaged across rootstock, training system and stub diameter to show general trends 293 Figsoure that 1 .each The value effect represents of stub lengththe mean on of regrowth at least 200 as anddetermined 100 observed by the renewal number stubs of innew 2015 shoots and 2016, formed respectively. R2 values are for a linear regression model. Within the various treatment combinations, the 294 perrelationship branch cut. between Values stub are length averaged and new across shoot rootstock, number remained training linear. system P values and stubare <0.0001 diameter for toboth show 2015 and 2016. 295 general trends so that each value represents the mean of at least 200 and 100 observed renewal

296 stubs in 2015 and 2016, respectively. R2 values are for a linear regression model. Within the

297 various treatment combinations, the relationship between stub length and new shoot number

298 remained linear. P values are <0.0001 for both 2015 and 2016.

299 300 Cherry 57 301

302Figure 2. The effect of rootstock and branch diameter on regrowth as determined by the number of new shoots formed per cut branch in 2015 (A) and 2016 (B). Values are averaged across training system and 303stub Figlength,ure 2and. The represent effect of the rootstock means of and at least branch 80 anddiameter 50 stubs on perregrowth branch as diameter determined and byrootstock the number combina - tion in 2015 and 2016, respectively. 304 of new shoots formed per cut branch in 2015 (A) and 2016 (B). Values are averaged across training

3052016 system appears and tostub be length, the resultand represent of dispropor the means- ofthis at leasteffect 80 wasand 50less stubs pronounced per branch diameterin 2016 for tionately longer new shoots originating in the larger diameter branches. 306the single-leaderand rootstock combination system, particularly in 2015 and 2016,at in respe- ctively. termediate (5 cm and 15 cm) stub lengths. Discussion 307The interaction between training system and The objective of this study was to deter- 308branch diameter is illustrated in Fig. 4. In mine the minimum length of renewal stub this interaction, the single leader training cuts needed to regenerate branches in high 309system results in disproportionately longer density ‘Montmorency’ tart cherry. It is well new shoots, but only for the large diameter known that short stub cuts lead to renewal branches. Both of these interactions may be growth in apple and peach. Critical length due to the more intense pruning severity re- of renewal pruning cuts for tart cherry is not quired in the single-leader system. well documented. Nugent (2002) recom- The effect of rootstock on new shoot length mended that renewal stub lengths be left be- differed slightly between years, but was gen- tween 10 and 14 cm for tart cherry, but this erally related to overall rootstock vigor. The recommendation did not reference any pub- effect of rootstock and branch diameter are lished data. A study with sweet cherry found shown in Fig. 5. In general, the more vigor- that long renewal stub lengths (>15 cm) re- ous ‘Mahaleb’ rootstock had longer shoots sulted in greater shoot and flower formation than the less vigorous rootstock, although (Guimond et al., 1998). Our results confirm 58 Journal of the American Pomological Society 310

Figure 3. The effect of branch stub length and training system on renewal growth as measured by the aver- 311 age length of new shoots. Values are averaged across rootstock and stub diameter and represent the means of at least 70 and 30 renewal stubs observed per treatment in 2015 and 2016, respectively. Values of R2 are 312 Figforure a quadratic 3. The effectregression of branch model, stub where length all regressions and training were system significant on renewal atP values growth <0.0004. as measured by

313 the average length of new shoots. Values are averaged across rootstock and stub diameter and

314 represent the means60 of at least 70 and 30 renewal stubs observed per treatment in 2015 and 2016, 2016 a 315 respectively. Values50 of Single R2 are forDouble a quadraticQuad regression model, where all regressions were b 40 316 significant at P < 0.004. c cd cd c 30 317 e f 20 f 318

New Shoot Length (cm) Length Shoot New 10

0 Small Medium Large Diameter Class 319 Figure 4. The effect of branch diameter and training system on the average length of new shoots in 2016. 320 LettersFigure denote 4. The significant effect of differences branch diameter (P < 0.05) and using training the pdiff system options on the from average the LSMeans length ofstatement new shoots in SAS. 321 in 2016. Letters denote significant differences (P < 0.05) using the pdiff option from the

322 LSMeans statement in SAS.

323

324

325

326

327 Cherry 59

Figure 5. The effect of rootstock and diameter on regrowth, as measured by the average length of shoots 328formed per renewal cut. Branch diameter was classified into small (<1.5 cm), medium (1.5-2.5 cm) and large (>2.5 cm) categories. Values are the means of at least 105 (2015) and 61 (2016) renewal stubs ob- 329served per rootstocks. Letters denote significant differences (P values <0.05) using the pdiff option from 330the LS-Means statement in SAS.

331 Figure 5. The effect of rootstock and diameter on regrowth, as measured by the average length the recommendation by Nugent (2002) that (Long et al., 2014). Under Utah conditions, 332lengths of shootsgreater formed than per10 renewalcm generally cut. Branch pro diameter- ‘Montmorency’ was classified tartinto cherrysmall (<1.5 on Gi.3cm), mediumand Gi.5 duced at least one new renewal shoot per cut are 32% and 33% the size of ‘Mahaleb’ trees, 333in the(1.5 season-2.5 cm) following and large pruning.(>2.5 cm) However,categories. Valuesrespectively are the means (Roper of at leastet al.,105 (2015)unpublished). and 61 334this critical(2016) renewal length stubs is influenced observed per rootstock.somewhat LettersThese denote dwarfing, significant more differences precocious (P

flower initiation and shoot regrowth in sweet cherry. In: D. Ferree and I. Warrington, eds., Apple-orchard HortScience 33:647-649. planting systems, 1st ed. Cambridge: CABI Pub- Long, L.E. and C. Kaiser. 2010. Sweet cherry root- lishing. 345-407. stocks for the pacific northwest. PNW. 619. Corval- Robinson, T.L., S. A. Hoying, and G.H. Reginato. lis, OR: Oregon State University. 2006. The tall spindle apple production system. Long, L.E., L.J. Brewer, and C. Kaiser. 2014. Sweet New York Fruit Qrtly. 14(2):21-28. cherry rootstocks for the modern orchard. Proc. 57th Roper, T., B. Black, M. Stasiak, R. Marini, J. Cline, Annu. Intl. Fruit Tree Assn. Conf. 10 Nov. 2017. T. Robinson, G. Lang, L. Anderson, R. Anderson, . ‘Montmorency’ tart cherry (Prunus cerasus L) on Nugent, J.E. 2002. Training tart cherry trees. North- Potential Size Controlling Rootstocks at Six Loca- west Michigan Hort. Ctr. Res. Rept. Michigan State tions in North America: Results of the 1998 NC-140 University. Trial. J. Amer. Pomol. Soc. In review. Palmer, J.W. 1997. Apples light and orchard design for Schupp, J.R., H.E. Winzeler, T.M. Kon, R.P. Marini, enhancement of yield and fruit quality. Proc. 1997: T.A. Baugher, L.F. Kime, and M.A. Schupp. 2017. Searching for Quality. Joint Mtg. Australian Avo- A method for quantifying whole-tree pruning se- cado Grower’s Federation, Inc. and NZ Avocado verity in mature tall spindle apple plantings. Hort- Growers Assn., Inc. 156-172. Science 52:1233-1240. Perry, R., G. Lang, R. Andersen, L. Anderson, A. Aza- Sutton, T.B. and C.R. Unrath. 1984. Evaluation of the renko, T. Facteau, D. Ferree, A. Gaus, F. Kappel, tree-row-volume concept with density adjustments F. Morrison, C. Rom, T. Roper, S. Southwick, G. in relation to spray deposits in apple orchards. Plant Tehrani, and C. Walsh. 1998. Performance of the Dis. 68:480-484. NC-140 cherry rootstock trials in North America. Wünsche, J.N. and A.N. Lakso. 2000. Apple tree phys- Acta Hort. 468:291-296. iology: Implications for orchard and tree manage- Robinson, T. 2003. Apple-orchard planting systems. ment. Compact Fruit Tree 33:(3)82-88. 62 Journal of the American Pomological Society

Journal of the American Pomological Society 73(1): 62-75 2019 Diversity of Pathogenic Fungi Associated with Apples in Cold Storage Facilities in Tunisia Bochra A. Bahri1*, Yosra Belaid1, Ghaya Mechichi1, and Wafa Rouissi1ǂ

Additional index words: Aggressiveness, pathogenicity, Penicillium expansum, postharvest

Abstract Postharvest storage fungi are major limiting factors for the apple industry. In a study conducted in Tunisia in 2014, six refrigerated fruit storage facilities were surveyed in order to determine disease incidence, identify the pathogenic fungal species on apples and study mycelial growth in vitro and lesion diameters in vivo of sampled fungal isolates. Results showed that Penicillium expansum (42.3%) was the predominant fungus in apples, fol- lowed by Alternaria spp. (23%), Botrytis spp. (19.2%), Aspergillus spp. (13.5%) and Fusarium spp. (2%). Isolates collected from storage facilities with highly diverse fungal species had significantly greater mycelial growth at 24°C in vitro and larger lesion diameters on apples and oranges than isolates collected from storage facilities with less diversity. Moreover, a trade-off between the diversity in fungal species and the disease incidence in the stor- age facilities was suggested; facilities with a low or high disease incidence showed the lowest diversity in fungal species. Controlling disease incidence in cold storage facilities is recommended to limit the diversity in fungal species and the development of virulent isolates. Further studies are needed to determine the fruit storage condi- tions that influence fungal species diversity and their variability in pathogenicity, in order to implement efficient postharvest disease management. Due to their organoleptic and nutritional la’, ‘Anna’, ‘Aziza’, ‘Golden Delicious’ and qualities, including antioxidants, Apple (Ma- ‘Gala’ (GIFruits, 2016; Mlika et al., 2002). lus x domestica) is among the fruits experi- The preservation of fruits and vegetables encing increasing international interest, with in Tunisia depends heavily on the cold stor- a world annual production of 84.6 million age industry. In fact, about 70% of the total tons in 2014 (FAOSTAT, 2014). Countries refrigerated storage capacity in Tunisia is in the northern hemisphere produce the high- used for fruits and vegetables and is located est quantities of apples. Asia accounts for at Beni Khaled area (API, 2006). These stor- more than 62% of global apple production, age technologies keep fruit at < 4°C and 70% while the European Union countries produce relative humidity using a sandwich panel 20.7% (FAOSTAT, 2014). In Tunisia, apples structure as thermal insulation. Unfortunate- are a critically important local fresh fruit. ly they are precarious and poorly managed Apple orchards cover 26,000 ha, 3 % of the (Jammes, 2012; Jraidi, 2010). This contrib- total cultivated land. Approximately 90,000 utes to development of postharvest fungal tons of apples were produced in Tunisia in diseases, which are the major factor limit- 2015. The crop is produced widely across the ing the storage life of fruits and vegetables country, but the main production area is the (Ogawa and English, 1991). In fact, while governorate of Kasserine with 27.7% of the cold storage delays fruit maturation and se- national apple production. The main cultivat- nescence, it leads to numerous phytosanitary ed local and introduced cultivars are ‘Chah- problems. Postharvest diseases can cause

1 The National Agronomic Institute of Tunisia, Department of Plant Protection & Post-harvest Diseases; Labora- tory "Bio-agresseurs et protection intégrée en agriculture", 43 Avenue Charles Nicolle, Tunis 1082, Tunisia. ǂ Current address; The National Institute of Agronomic Research of Tunisia Laboratory of Plant Protection, Rue Hedi Karray, 2049 Ariana, Tunisia. * Corresponding author: [email protected] Apple 63 losses up to 50 % to the apple industry during age diversities in fungal species on the po- storage, transport and marketing (ElGhaouth, tential pathogenicity of the phytopathogens. 1997; Jurick et al., 2011). These diseases can originate from infection in the orchard, or Methods and Materials are more commonly associated with post- Identification of post-harvest pathogenic fun- harvest fruit handling practices. Leibinger gi of apples in cold storage et al. (1997) reported more than 90 fungal Storage sites and sampling protocol. In species infecting apples during storage. The 2014 apples were sampled from six cold most prevalent post-harvest fungal disease fruit storage facilities distributed throughout of apple is blue mold caused by Penicillium the main post-harvest storage areas in Tu- expansum. However, Botrytis cinerea (gray nisia (Beni Khaled, Mornag, Khledia, Ras mold) and Gloeoporus spp. also are of eco- Jebel and Korba), to identify common phy- nomic concern in Europe (Dean et al., 2012; topathogenic fungal species (Table 1). Apple Droby and Lichter, 2004; Elad et al., 2015; samples (n = 30 fruits per storage unit) were Rupp et al., 2016). Traditionally, the appli- collected in storage units from Feb. to April, cation of synthetic fungicides is considered and each storage unit was sampled once. At the most essential practice in a profitable each site, 10 storage boxes and three apples post-harvest fruit storage (Conway et al., per box were randomly selected based on box 1999; Eckert and Ogawa, 1988; Eckert and and fruit position in the storage unit and box, Sommer, 1967). However, the emergence respectively. Selected apples did not have of fungal strains that are resistant to fungi- disease symptom and were kept separately in cides, especially resistance of P. expansum to a bag to avoid cross-contaminations. For all benzimidazoles (Holmes and Eckert, 1999; storage units, fruits came from more than one Rupp et al., 2016; Spotts and Cervantes, orchard and were stored at 2-4°C; the num- 1986), and regulations that restrict chemical ber of fruit suppliers varied with the storage application (Eckert et al., 1994; Gullino and capacity. However, because traceability sys- Kuijpers, 1994; Ragsdale and Sisler, 1994) tem is not implemented in the storage facili- have stimulated a renewed interest in biocon- ties, origin of fruits was unknown. trol strategies. Biocontrols have the potential Isolation of fungi. A piece (~1 cm2; 0.5 to achieve sustainable post-harvest disease cm deep) of apple epicarp was taken from management while ensuring human health, each apple collected from the different cold environmental safety and consumer expecta- storage facilities and plated on potato dex- tions of fruit quality (Janisiewicz and Kor- trose agar (PDA), after surface wash in three sten, 2002; Romanazzi et al., 2016; Wilson different changes of distilled water. Each et al., 1993). sample was processed separately to avoid In Tunisia, the phytopathogenic fungi that cross-contamination. After incubation for occur in stored fruits have not been surveyed. seven days at 24°C in the dark, each fungal Prior to exploring the feasibility of biocontrol culture was purified by sub-culturing to new and other non-chemical management strate- PDA plates. Subsequently, single-spore cul- gies, it is critical to survey the types of fungi tures were prepared from pure colonies; all associated with cold apple storage in Tuni- isolates were stored as 7-day old single-spore sia, and the genetic diversity of these species. cultures on PDA plates at 4°C and cultured in Hence, the goals of this research were 1) to fresh PDA medium in darkness at 24°C for 7 identify the main phytopathogenic fungi in- days when needed. fecting apples in the cold storage facilities Pathogenicity tests. Pathogenicity tests in Tunisia, 2) to evaluate the relative aggres- were performed with all isolated fungi in vivo siveness of these fungal species in vitro and by inoculating apple and pear fruits when in vivo and 3) to study the effect of the stor- Penicillium expansum was identified. Before 64 Journal of the American Pomological Society inoculation, fruits were surface-sterilized represents the potential post-harvest loss in with 90% alcohol, washed with 1% sodium each storage. Furthermore, the frequency of hypochlorite, rinsed with distilled water and each fungal species identified was assessed air-dried at room temperature for 5 minutes. in each storage unit. The diversity in phy- Surface sterilized fruits were then wounded topathogenic fungal species of each storage twice at equal distance with a sterile tip by unit was calculated according to the Shannon making an injury of ~0.5 cm diameter. Then, diversity index calculated as follows: 20 μl of conidial suspension (prepared from s 7 days old PDA cultures) of each isolate ad- H’ = ∑ - (Pi * ln Pi) justed to 103 spores/ml were applied into i=1 each wound; two fruits were inoculated per where S is the number of species observed isolate. Four fruits inoculated with sterilized and Pi is the fraction of the entire population distilled water were used as controls. Inocu- made up of species i. The fruit storage facili- lated fruit were covered with plastic film to ties were grouped as having diverse and non- avoid contamination and to ensure 100% diverse phytopathogenic fungal populations RH and incubated at 24°C in the dark. Fungi when H’>1 and H’≤ 1, respectively. were re-isolated from infected apples and compared with the original isolates in order Characterization of pathogenic fungi isolat- to confirm Koch’s postulates. Isolates that ed from apples in cold storage induced rot symptoms on inoculated apples Mycelial growth. In vitro mycelial growth were considered pathogenic. was measured for 14 isolates of the five phy- Identification of pathogenic fungi and topathogenic species isolated from apples in diversity in fungal species associated with cold storage facilities. These isolates were apple fruits in cold storage facilities. Fungal selected arbitrarily and included six Penicil- species isolated from stored apple fruit were lium expansum, three Aspergillus sp., two identified based on macroscopic observa- Botrytis sp., two Alternaria sp. and one Fu- tions of the monosporic colonies on PDA, sarium sp. isolates. Agar plugs (~5 mm di- followed by microscopic observations of ameter) were taken from each monosporic spores and pathogenicity tests on inoculated culture on PDA and plated onto new PDA apple fruit. Only monosporic colonies that plates. Each isolate was incubated at 24°C were pathogenic by inducing symptoms on and 4°C and colony diameter measurements inoculated apples were considered for iden- were recorded at 7 days post-incubation tification. The macro and micro-morpholog- (DPI). There were three replicates per isolate ical characteristics included colony growth x temperature treatment. and colour, presence or absence of wrinkles, Lesion diameter. The diameter of the le- presence or absence of septa, morphology sions 7 DPI of five isolates of four -phyto and size of conidia, observation of the struc- pathogenic species (two Penicillium, one tures bearing spores, and rot symptoms on Aspergillus, one Botrytis, one Alternaria) inoculated fruits. Fungal organisms were collected from stored apples was evaluated identified in accordance with (Domsch et al., in vivo by inoculating mature ‘Golden Deli- 2007; Klich, 2002; Leslie and Summerell, cious’ fruits. Inoculation method was similar 2006; Pitt and Hocking, 2009; Vico et al., to that described above in the pathogenicity 2014; Xiao and Kim, 2008). The incidence of tests; however in this test five fruits were in- disease was calculated for each cold storage oculated once at the equator with each iso- unit as the number of apple tissue samples late. Controls included fruits treated with that yielded pathogenic fungal cultures di- sterilized distilled water. The pathogenicity vided by the total number of apple fruit. This of the five isolates was also assessed with a disease incidence evaluated on 30 apples, similar approach on mature orange fruits of Apple 65 the cultivar ‘Maltaise’, the main orange cul- ues H’>1 (high diversity in phytopathogenic tivar grown and stored in Tunisia. species) and storage units with index H’≤ 1 (low diversity in phytopathogenic species). Data analysis In addition, Pearson’s correlation between le- Data were analysed as a completely ran- sion diameters on apple and orange, and be- domized design with analysis of variance tween lesion diameters on apple, and mycelia (ANOVA) using R 3.0.2 (R Core Team, growth on PDA media at 24°C were tested 2013). Data for in vitro mycelial growth of under R 3.0.2 (R Core Team, 2013) using the fungal species at 24°C and at 4°C were ana- rcorr function in the Hmisc package. lysed independently with 1-way ANOVAs. Similar analyses were performed for in vitro Results mycelial growth of fungal isolates within Identification of pathogenic fungi isolated species at 24°C and 4°C. Following ANO- from apples and diversity in fungal species VA, contrasts were used to compare mycelial in cold storage. growth means between isolates within spe- From the pathogenicity tests, 52 isolates cies at 5% level of significance. In addition, from 180 apples tissue cultures were patho- the in vitro growth of fungal species and fun- genic on inoculated apples. Morphological gal isolates within species at 4°C and 24°C analysis revealed the presence of five genera were analysed as 2-way factorials. Contrasts in these 52 pathogenic fungi. These includ- were performed to compare mycelial growth ed Penicillium expansum, Aspergillus sp., I means between species and between isolates sp., Fusarium sp. and Botrytis sp. (Table 1). within species. The lesion diameter of fun- These five pathogens had characteristic colo- gal isolates on apple and orange fruits were ny morphology and spores on PDA medium, analysed as 2-way factorial. Contrasts were and typical disease symptoms on inoculated also performed to compare mycelial growth apple fruits (Fig. 1). and lesion diameter means between isolates Penicillium expansum was the most com- sampled from storage units with index val- monly observed fungus in stored apples with TABLES

Table 1. NumberTable 1. of Number fungal of species fungal speciesand isolates and isolates recovered recovered from from apple apple fruits fruits from from sixsix cold storage storage facilities in Tunisia betweenfacilities inFebruary Tunisia between and April February 2014. and April 2014. Storage # # fungal Pen ID Region isolates species z Alt Bot Asp Fus H'y Isolate ID x Beni 1 Khaled 10 4 4 3 2 0 1 1.84 Pen4, Bot1, Fus* 2 Korba 2 1 2 0 0 0 0 0.00 Pen1 Beni Pen2, Alt3*, 3 Khaled 14 3 6 7 0 1 0 1.29 Asp3 Pen6, Bot2, Alt2, 4 Mornag 8 4 2 2 1 3 0 1.90 Asp2* 5 Khelidia 5 2 2 0 0 3 0 0.97 Pen3*, Asp1 6 Ras Jebel 13 2 6 0 7 0 0 0.99 Pen5* Total 52 5 22 12 10 7 1 1.75 z Pen, Alt, Bot,Z Pen, Asp, Alt, Fus Bot, correspond Asp, Fus to correspond Penicillium to Peniciliumexpansum, expansum Alternaria, Alternaria spp., Botrytis spp., spp., Botrytis Aspergilus spp., Aspergilus spp., and Fusarium spp. isolates observedspp., and Fusarium spp. isolates observed y Hʼ: diversity index calculated according to the Shannon index y x Isolates ID corresponds H’: diversity to index the isolates calculated used according for the in to vitro the Shannontests on PDA index media; asterisks indicate isolates used for pathogenicity assays on orangex Isolates and ID apple corresponds fruits to the isolates used for the in vitro tests on PDA media; asterisks indicate isolates used for pathogenicity assays on orange and apple fruits

66 Journal of the American Pomological Society

Figure 1. Mycelia cultures and spores characteristics on PDA media and typical disease symptoms on inoculated fruits for the 5 genera isolated Penicillium expansum (A), Alternaria spp. (B), Fusarium spp. (C), Botrytis spp. (D), and Aspergilus spp. (E). a frequency of 42.3%, followed by Alternar- not differ significantly (Penicillium, Asper- ia sp., Botrytis sp., Aspergillus sp. and Fu- gillus, Alternaria and Botrytis) (P = 0.858). sarium sp. with frequencies of 23% 19.2%, However, temperature did significantly -af 13.5% and 2%, respectively. The six storage fect in vitro mycelial growth of 14 isolates facilities surveyed had different levels of on PDA medium. At 24°C, isolates Asp2, disease incidence and the diversity index H’ Asp3 grew to a mean diameter of 4.48 cm in isolated fungal species. Storage facilities whereas the Pen5, Pen3 and Asp1 isolates 3 and 6 had the highest overall disease inci- had mean diameters of 1.96 cm. Fifty-seven dences (46.6% and 43.3%, respectively), fol- percent of the fungal isolates showed inter- lowed by storage facilities 1, 4, 5 and 2 with mediate mycelial growth ranging from 2.00 disease incidences of 33.3%, 26.6%, 16.6 % to 3.93 cm (Tables 2 and 3). Analysis by spe- and 6.6%, respectively. Storage facilities 1, 3 cies showed that mycelial growth was sig- and 4 had the highest mean diversity indices nificantly different among Aspergillus sp. (P (H’ = 1.68) with three to four species of fun- = 0.0002) and among Penicillium expansum gi. Storage facilities 2, 5 and 6 had the least isolates (P = 0.020) at 24°C. Pen2 (Storage diverse fungal populations (from one to two facility 3), Pen6 (Storage 4) and Pen4 (Stor- species) with mean diversity indices ranging age 1) isolates had mycelial growth that was from 0 to 0.97 (Table 1). significantly greater than Pen5 (Storage 6) Characterization of pathogenic fungi iso- and Pen3 (Storage 5). The mycelial growth lated from stored apples. In vitro mycelial of isolates Asp2 (Storage 4) and Aps3 (Stor- growth of 14 fungal isolates was measured age 3) was significantly greater than Asp1 on PDA medium. Mycelial growth between (Storage 5). However, mycelial growth of the four fungal species tested at 24°C did the Botrytis sp. isolates, Bot1 (Storage 1) and Apple 67

Bot2 (Storage 4) (P = 0.08) and the Alter- isolates, Pen3 (Storage 1) and Pen5 (Storage naria sp. isolates Alt2 (Storage 4) and Alt3 6) were the least virulent with barely visible (Storage 3) (P = 0.39) did not differ at 24°C lesions on apples after 7 days of incubation on PDA. In addition, significant interaction at 24°C. Mean lesion diameters were grew effects of ‘temperature x species’ (P < 0.001) to 0.5 cm after 14 days of incubation. The and ‘temperature x isolate within species’ (P most virulent isolates on apples were also the < 0.001) were observed on mycelial growth. most virulent on oranges and lesion diameter Mycelial growth at 4°C was greatest for on apples and oranges were correlated (Fig. Bot1, Pen2 and Pen1 isolates (average of 1.2 2; R² =0.982; P= 0.0011). cm), which were intermediate at 24°C. In ad- Correlations between mycelial growth, dition, isolates Fus, Asp3 and Asp2, the most lesion diameter, disease incidence and H’ fit at 24°C did not grow after 7 days of incu- index. Lesion diameter and in vitro mycelia bation at 4°C (Tables 2 and 3). growth for the tested isolates at 24°C were Lesion diameter of fungi isolated from strongly correlated (Fig. 3; R² = 0.864 %; P= stored apples. After 7 days of incubation at 0.0221). Asp2 isolate, with greatest mycelial 24°C, lesion diameter of the five isolates test- growth in vitro at 24°C, had the highest mean ed were significantly different on both apple lesion diameter. Isolates from storage facili- and orange fruit (P < 0.0001). The most viru- ties with a H’ index > 1 (Storage facilities 1, lent isolate on apple was Aps2 (Storage 4) 3 and 4) had significantly higher mycelial (mean lesion diameter of ~ 1.2 cm), followed growth at 24°C than those from storage fa- by Alt3 (Storage 3) (diameter = 0.72 cm). The cilities with a H’ index ≤ 1 (Storage facilities 2, 5 and 6) for all fungal species (P < 0.0001) and for Aspergillus sp. (P < 0.0001) and for Table 2. Mycelia growth (cm) of 14 isolates recov- Penicillium expansum (P = 0.0038) isolates eredTable from 2. appleMycelia fruits growth in cold (cm) storage of facilities14 isolates in recovered from apple fruits in cold storage facilities tested separately. At 4°C, storage facilities Tunisiain Tunisia after after7 days 7 daysof incubation of incubation on potato on dexpotato- dextrose agar media at 24°C and at 4°C. trose agar media at 25oC and at 4oC. with a H’ index > 1 (Storage facilities 1, 3 and 4) and H’ index ≤ 1(Storage facilities 2, 5 Isolate 4°Cz 24°Cz Meanz and 6) did not differ (P= 0.6383) for mycelial Alt3 0.45A 3.17A 1.81A growth (Tables 2 and 3, Fig. 4). The lesion B A A diameters were significantly higher for the Alt2 0.74 3.53 2.14 isolates from storage facilities with a H’ in- A A A Asp1 0.00 2.03 1.02 dex > 1 (Storage facilities 1, 3 and 4) than for Asp3 0.00A 4.30B 2.15B those from storage facilities with a H’ index Asp2 0.00A 4.67B 2.33B ≤ 1 on both apples and oranges (P < 0.0001) after 7 days of incubation at 24°C (Table 4, 0.13A 3.30A 1.71A Bot2 Fig. 5). The plot of the diversity H’ index Bot1 1.25B 2.30A 1.78A against the disease incidence of the six cold F 0.00 3.93 1.97 storage facilities showed that the H’ reaches a maximum value for an intermediate disease Pen5 0.60A 1.87A 1.23A incidence of 30 %. Storage facilities with low A AB AB Pen3 0.73 2.00 1.36 and high disease incidences had a lower di- Pen4 0.70A 3.13C 1.92BC versity in fungal species (Fig. 6). Pen6 0.68A 3.20C 1.94BC Discussion Pen1 1.11B 2.93BC 2.02C Phytopathogens can cause significant eco- B C C Pen2 1.19 3.27 2.23 nomic losses in post-harvest fruit storage. z z Contrasts Contrasts were were performed performed between between isolates within isolates species. within This species. study Isolates focused with on common identifying letters and do charnot -differ at theIsolate 5% with level common of significance letters do not. differ at the 5% level of acterizing phytopathogenic fungi common in significance.

Table 3. Analysis of variance of the mycelia growth (cm) of five fungal species and 14 isolates within species recovered from apple fruits in cold storage facilities in Tunisia after 7 days of incubation on potato dextrose agar media at 24°C and at 4°C. Contrasts were shown between H'≤1 and H'> 1 overall and, at 24°C and at 4°C independently.

Source of variation Df Sum Sq Mean Sq F value Pr>F Temperature 1 139.4 139.4 351.399 < 2e-16 *** Species 4 0.52 0.13 0.328 0.858 Temperature x Species 4 12.97 3.24 8.174 1.63e-05 *** Residuals 74 29.36 0.4

Source of variation Df Sum Sq Mean Sq F value Pr>F Temperature 1 139.4 139.4 1198.184 < 2e-16 *** Isolate within species 9 11.09 1.23 10.59 2.28e-09 *** Temperature x Isolate within species 13 24.72 1.9 16.346 1.45e-14 *** Residuals 56 6.52 0.12 Contrasts (Pr>F): H'≤1 vs. H'> 1 <.0001 H'≤1 vs. H'> 1 at 4°C 0.6383

Table 2. Mycelia growth (cm) of 14 isolates recovered from apple fruits in cold storage facilities 68in Tunisia after 7 Jdaysournal of incubation of the A mericanon potato P dextroseomological agar Smocietyedia at 24°C and at 4°C.

Isolate 4°Cz 24°Cz Meanz Alt3 0.45A 3.17A 1.81A Alt2 0.74B 3.53A 2.14A Asp1 0.00A 2.03A 1.02A Asp3 0.00A 4.30B 2.15B Asp2 0.00A 4.67B 2.33B Bot2 0.13A 3.30A 1.71A Bot1 1.25B 2.30A 1.78A F 0.00 3.93 1.97 Pen5 0.60A 1.87A 1.23A Pen3 0.73A 2.00AB 1.36AB Pen4 0.70A 3.13C 1.92BC Pen6 0.68A 3.20C 1.94BC Pen1 1.11B 2.93BC 2.02C Figure 2. Correlation between mean lesion diameter on orange and apple fruits after 7 days of incubation at B C C 24oCPen2 for five fungal isolates1.19 recovered3.27 from2.23 apple fruits stored in cold storage facilities in Tunisia in 2014. z Each Contrasts value represents were performed the mean of between four observations. isolates within species. Isolates with common letters do not differ at the 5% level of significance. Fig. 2. Correlation between mean lesion diameter on orange and apple fruits after 7 days of

incubation at 24°C for five fungal isolates recovered from apple fruits stored in cold storage

Tablefacilities 3. Analy in sisTunisia of variance in 2014. of Eachthe mycelia value represents growth (cm) the ofmean five offungal four observationsspecies and 14 isolates Tablewithin 3. Analysis species of recoveredvariance of fromthe mycelia apple growth fruits (cm) in cold of five storage fungal facilities species and in Tunisia14 isolates after within 7 days of speciesincubation recovered on from potato apple dextrose fruits in agarcold storage media facilitiesat 24°C in and Tunisia at 4°C after. Contrasts 7 days of incubationwere shown on potato between H'≤1 dextrose agar media at 24oC and at 4oC. Contrasts were shown between H'<1 and H'>1 overall and, at 24oC and H'> 1 overall and, at 24°C and at 4°C independently. independently.

Source of variation Df Sum Sq Mean Sq F value Pr>F Temperature 1 139.4 139.4 351.399 < 2e-16 *** Species 4 0.52 0.13 0.328 0.858 Temperature x Species 4 12.97 3.24 8.174 1.63e-05 *** Residuals 74 29.36 0.4

Source of variation Df Sum Sq Mean Sq F value Pr>F Temperature 1 139.4 139.4 1198.184 < 2e-16 *** Isolate within species 9 11.09 1.23 10.59 2.28e-09 *** Temperature x Isolate within species 13 24.72 1.9 16.346 1.45e-14 *** Residuals 56 6.52 0.12 Contrasts (Pr>F): H'≤1 vs. H'> 1 <.0001 H'≤1 vs. H'> 1 at 4°C 0.6383

H'≤1 vs. H'> 1 at 24°C <.0001

Df: Degrees of freedom; Sum Sq: Sum of squares; Mean Sq: Mean squares.

Table 4. Analysis of variance of the lesion diameter of five isolates recovered from apple fruits in cold storage facilities in Tunisia after 7 days of incubation at 24°C on apple and orange fruits (Host). Contrasts were shown between H'≤1 and H'> 1 overall and, on apple and orange fruits independently.

Source of variation Df Sum Sq Mean Sq F value Pr>F Host 1 0.01 0.006 0.074 0.788 Isolate 4 42.14 10.535 136.319 < 2e-16 *** Host x Isolate 4 0.01 0.002 0.021 0.999 Residuals 40 3.09 0.077 Contrasts (Pr>F): H'≤1 vs. H'> 1 <.0001 H'≤1 vs. H'> 1 on Apple <.0001 H'≤1 vs. H'> 1 on Orange <.0001 Df: Degrees of freedom; Sum Sq: Sum of squares; Mean Sq: Mean squares.

H'≤1 vs. H'> 1 at 24°C <.0001 Df: Degrees of freedom; Sum Sq: Sum of squares; Mean Sq: Mean squares.

Apple 69 Table 4. Analysis of variance of the lesion diameter of five isolates recovered from apple fruits in cold storage facilities in Tunisia after 7 days of incubation at 24°C on apple and orange fruits (Host). TableContrasts 4. Analysis were of variance shown ofbetween the lesion H'≤1 diameter and of H'>five 1isolates overall recovered and, fromon appleapple fruits and in orange cold fruits storage facilities in Tunisia after 7 days of incubation at 24oC on apple and orange fruits (Host). Contrasts wereindependently. shown between H'<1 and H'>1 overall and, at 24oC independently.

Source of variation Df Sum Sq Mean Sq F value Pr>F Host 1 0.01 0.006 0.074 0.788 Isolate 4 42.14 10.535 136.319 < 2e-16 *** Host x Isolate 4 0.01 0.002 0.021 0.999 Residuals 40 3.09 0.077 Contrasts (Pr>F): H'≤1 vs. H'> 1 <.0001 H'≤1 vs. H'> 1 on Apple <.0001 H'≤1 vs. H'> 1 on Orange <.0001 Df: Degrees of freedom; Sum Sq: Sum of squares; Mean Sq: Mean squares.

stored apples in Tunisia. In 2014, five fungal by Botrytis cinerea is the main cause of post- species associated with apples in cold stor- harvest losses on pears, and the second most age conditions were identified (Penicillium common disease on apples after blue mold expansum, Alternaria spp., Botrytis spp., (Rosenberger, 1990). The lower economic Aspergillus spp. and Fusarium spp.). These importance of Fusarium infections in storage fungi are also among the most widely re- facilities has also been reported (Snowdon, ported pathogens in cold storage facilities 1990). The low prevalence of Fusarium spp. around the world (Leibinger et al., 1997; in storage facilities suggests it is opportunis- Louw and Korsten, 2014; Pepejnjak et al., tic on apples. A second survey performed in 2002; Romanazzi et al., 2016; Rosenberger, 2015 across eight cold storage facilities in 1990). Because of their ubiquitous nature, Tunisia similarly revealed the prevalence of broad host range in post-harvest conditions Penicillium expansum and Alternaria spp. and necrotrophic lifestyle, they represent a on over 137 isolates sampled, with respec- significant economic threat (Rosenberger, tive frequencies of 53% and 46% (Bahri et 1990; Rosenberger et al., 2006; Sommer et al., Unpublished results). In addition, P. ex- al., 2002). According to our survey, Peni- pansum is a necrotrophic pathogen that has a cillium expansum was the most prevalent saprotrophic phase on the fruit surface before post-harvest fungus on apple in cold storage initiating infections (Blakeman and Brodie, facilities in Tunisia, followed by Alternaria 1977). This saprotrophic phase gives P. ex- spp., Botrytis spp., Aspergillus spp. and Fu- pansum a postharvest survival advantage and sarium spp. The detection frequencies that explains its high prevalence under storage we observed for these fungi reflect previous conditions. It is likely that primary infec- reports worldwide. Blue mold caused by P. tion by these fungi begins in the orchard, and expansum is the most common disease of presence or absence of a fungal species in a post-harvest apples (Janisiewicz and Kor- storage unit could be due to differences in sten, 2002; Leibinger et al., 1997; Louw orchard origin of fruits. Because the origin and Korsten, 2014; Pepejnjak et al., 2002; of fruits is unknown, we don’t know the ef- Rosenberger, 1990). This decay impacts also fect of apple growing conditions and protec- fruit processing by the production of the car- tive stray practices in the different orchards cinogenic mycotoxin patulin (Barkai-Golan, on pathogen population variabilities in the 2008). In the United States, gray mold caused storage facilities. However, out of 308 iso- 70 Journal of the American Pomological Society

FigureFig. 4. 4 .Box Box plots plots for for mycelial mycelial growth growth by by the the diversity diversity index index H' H’ for for fungal fungal isolates isolates recovered recovered from from apple fruits in cold storage facilities in Tunisia in 2014. Isolates were grown at 24oC (left) and 4oC (right) for 7 daysapple on potato fruits dextrosein cold storage agar media. facilities The in bottom Tunisia and in 2014.top of Isolates each box were are grownthe first at 24°Cand third (left) quartiles and 4°C of the data(right) set, respectively. for 7 days onThe potato band dextrose inside the agar box media. is the Themedian. bottom The and ends top of of the each dashed box are lines the (whiskers) first and are the minimum and maximum observations. third quartiles of the data set, respectively. The band inside the box is the median. The ends of the dashed lines (whiskers) are the minimum and maximum observations.

FigureFig. 5 5.. Box plotsplots for for lesion lesion diameter diameter influenced influenced by the by fungal the fungalspecies speciesdiversity indexdiversity H' for index apple H’(left) for apple and orange (right) fruits after 7 days of incubation at 24oC with fungal isolates recovered from apples in cold(left) storage and facilitiesorange in(right) Tunisia. fruits The bottomafter 7 and days top of eachincubation box are the at first24°C and with third fungal quartiles isolates of the data recovered set, respectively. The band inside the box is the median. The ends of the dashed lines (whiskers) are the minimumfrom apples and maximum in cold storageobservations. facilities in Tunisia. The bottom and top of each box are the first and third quartiles of the data set, respectively. The band inside the box is the median. The ends of the dashed lines (whiskers) are the minimum and maximum observations.

Apple 71

Figure 6. Relationship between the fungal species diversity index H' and disease incidence on apple fruits from six coldFig. storages 6. Relationship facilities in between Tunisia in the 2014. fungal species diversity index H’ and disease incidence on apple fruits from six cold storages facilities in Tunisia in 2014. lates sampled in 29 Tunisian apple orchards served among isolates collected from storage distributed over nine governorates, no vari- facilities with H’> 1 compared to those from ability in fungal species frequencies were ob- storage facilities with H’≤ 1. P. expansum served across regions, and Alternaria sp. was isolates, Pen2, Pen6 and Pen4 and Aspergil- the most prevalent fungus (96%) on symp- lus spp. isolates, Asp2 and Aps3, all derived tomatic leaves although orchards differed in from H’> 1 storage facilities, had mycelial disease severity and fungicide application growth that on average was 2 times greater practices. Furthermore, fungicide treated or- than Pen5 and Aps1, respectively. Pen5 and chards had similar levels of disease severity Aps1 recovered from storage units with H’≤ to untreated orchards (Bahri et al., Unpub- 1. In addition, Botrytis spp. and Alternaria lished results). This study, based on 180 ap- spp. isolates that were all recovered from ples sampled from six storage units, provides H’> 1 storage units did not significant differ the first pathogenic fungal assessment in cold in mycelial growth at 24°C. Similarly, iso- storage facilities in Tunisia. However, a more lates Aps2, Aps3 and Fus, derived from H’> exhaustive sampling could allow us to draw 1 storage units had a mean lesion diameter clearer conclusions on the relative propor- that was 28 times greater than that produced tions of these fungi in storage facilities. by Pen3 and Pen5. Pen3 and Pen5 both origi- In the current study, significant differences nated from storage units with H’≤ 1. Addi- were observed between the isolates for my- tionally, growth characteristics of the isolates celial growth on PDA and lesion diameter on were positively correlated. Specifically, there apple and orange fruits incubated at 24°C. was a strong positive correlation between le- This variability could be explained by differ- sion diameter on apples and oranges for the ences in storage conditions as well as differ- tested isolates. Because lesion diameters on ences in orchard origins. The growth traits of oranges were similar on apples, we think the isolates correlated significantly with the that the fungal species isolated from stored H’ Shannon diversity index for fungi isolated apple fruits are not opportunistic pathogens from fruits in the different storage facilities. of orange in Tunisian storage conditions. Cit- The only significant differences in mycelial rus, mainly orange, is usually processed with growth at 24°C on PDA and in lesion diam- apples in the storage facilities. Hence, cross- eters on apple and orange fruits were ob- contaminations are likely to occur exposing 72 Journal of the American Pomological Society citrus to typical post-harvest fungi such as disease incidence in storage. The level of P. expansum and Alternaria spp. (Louw and competition between isolates may be an im- Korsten, 2014, 2015). In addition, in vitro portant factor in phytopathogenic population mycelial growth of the isolates were cor- structure (species diversity and virulence) in related with their lesion diameters in vivo. storage conditions. The classical theory of Isolates collected from storage facilities with pathogen evolution predicts that within-host H’> 1 grew abundantly on PDA medium and competition increases pathogen virulence had the highest pathogenicity in vivo on ap- (Ebert and Mangin, 1997). In fact, resource ples and oranges. Further analyses are need- competition among strains was suggested to ed to test the current hypothesis where the favor more virulent strains that transmit at diversity in fungal species could determine faster rates (May and Anderson, 1983; May potential pathogen pathogenicity in storages. and Nowak, 1994). However, as in the cur- In addition, sporulation rate is also an impor- rent study, empirical studies do not always tant life-trait component of pathogen fitness support the classical theory of pathogen evo- (Bruns et al., 2012; Pariaud et al., 2009). In lution. Recent observations suggested that storage units, infected fruits can produce in- pathogen evolution in multiple infections oculum and spores can be disseminated by differ from what the theory predicts and does cooling fans and contaminate neighbour- not lead to an increase in virulence (Gower ing fruits (Janisiewicz and Korsten, 2002; and Webster, 2005; Staves and Knell, 2010). Nair et al., 1995). Thus, the contribution of Although we observed significant differ- fungal sporulation to variation in frequency ences among fungal isolates based on the pa- of disease infection in storage units should rameters measured, most isolates from stored also be considered in future works. Because apple fruits showed intermediate levels of trade-off between virulence and transmission mycelial growth and pathogenicity. Virulence fungal traits was previously shown in several during the infectious phase is influenced pathosystems (Ebert and Bull, 2003; Ewald by several evolutionary processes, and may and De Leo, 2002; Frank, 1996; Koella and be counter-balanced at other stages in the Agnew, 1999), investigation on sporulation pathogen’s life cycle to avoid an unlimited capacity could lead to contrasting findings increase and stabilize around an intermedi- compared to mycelial growth results shown ate value. Some studies showed that the most in this study. virulent phytopathogenic fungal isolates are Our results suggest a trade-off between not necessarily the fittest over one or several the fungal species diversity and disease in- seasons (Flier et al., 1998). Thus, highly vir- cidence in the storage facilities for post- ulent isolates favored during infection con- harvest apple pathogens. Storage units with ditions at 24°C in our study could be under a low and high disease incidence had low H’ negative selection pressure during fruit stor- indices. The diversity in fungi in the storage age conditions. In fact, our results showed facilities was highest at an average disease significant ‘temperature x isolate’ interaction incidence of 30%. A high disease incidence and isolates with the highest mycelia growth in storage facilities can lead to strong com- at 24°C had the lowest mycelia growth at petition between isolates that could select for 4°C. In addition, the most fit isolates at 4°C a low diversity in phytopathogenic species. had intermediate mycelial growth at 24°C. Our results may suggest that a competitive This can be explained by a trade-off between equilibrium between isolates in storage con- virulence in the field and survival in cold ditions must be achieved for the selection of storage that would maintain the diversity in isolates with high pathogenicity. This study fungi isolates on stored apples in Tunisia. In highlights the importance of limiting the de- natural systems, trade-offs explain the di- velopment of virulent isolates by reducing versity in pathogen populations, the mainte- Apple 73 nance of pathogenicity polymorphisms, and facilities to limit the diversity in fungal spe- the counter-selection of highly virulent iso- cies and the development of virulent isolates. lates (Ebert and Bull, 2003; Ewald and De Leo, 2002; Frank, 1996; Koella and Agnew, Acknowledgements 1999). Pathogen adaptation to their host and This project was funded by the Research to their environment also depends on genetic Laboratory LR14AGR02 “Bio-agresseurs et factors affecting their fitness components protection intégrée en agriculture”. The au- (Pariaud et al., 2009); these genetic factors thors are grateful to Ron Walcott, professor could also explain the mycelial growth varia- at the University of Georgia (Dept. of Plant tions observed in this study. The diversity Pathology) for reviewing the manuscript. in pathogenicity of phytopathogenic fungi could also be driven by conditions in storage. Literature Cited In our study, storage facilities 1, 3 and 4 had API. 2006. Etude de positionnement de la branche H’> 1 and were characterized by the high- entreposage frigorifique. Ministère de l’Industrie, est fruit storage capacity, the highest number de l’Energie et des Petites et Moyennes Entreprises, Agence de Promotion de l’Industrie. Tunisie, 2 p. of fruit suppliers and were probably the most Baert, K., F. Devlieghere, H. Flyps, M. Oosterlinck, diversified in fruit species stored. Some stud- M. M. Ahmed, A. Rajkovic, B. Verlinden, B. Nico- ies have also shown the influence of different lai, J. Debevere, and B. De Meulenaer. 2007. Influ- storage conditions, e.i. pH and temperature ence of storage conditions of apples on growth and of controlled atmosphere storages, on the de- patulin production by Penicillium expansum. Intl. J. cay caused by P. expansum in apples and the Food Microbiol. 119:170-181. ability of the fungus to growth and produce Barkai-Golan, R. 2008. Penicillium mycotoxins. In: Barkai-Golan R. and N. Paster (eds.). Mycotoxins patulin (Baert et al., 2007; Morales et al., in fruits and vegetables, Elsevier, San Diego, USA, 2007; Sydenham et al., 1995). Further analy- pp.153-185. sis is needed to understand the diversity in Blakeman, J. P. and I. D. S. Brodie. 1977. Competition fungal species and to determine the main fac- for Nutrients between Epiphytic Microorganisms tors explaining the maintenance of the vari- and Germination of Spores of Plant Pathogens on ability in pathogen pathogenicity in Tunisian Beetroot Leaves. Physiol. Plant Pathol. 10: 29-42. storages. Bruns, E., M.Carson and G. May. 2012. Pathogen and host genotype differently affect pathogen fitness through their effects on different life-history stages. Conclusions Bmc Evol. Biol. 12:135. Results indicate that Penicillium expan- Conway, W. S., W. J. Janisiewicz, J. D. Klein, and C. sum was the predominant pathogenic fungal E. Sams. 1999. Strategy for combining heat treat- species isolated in apples from cold storage ment, calcium infiltration, and biological control facilities in Tunisia, followed by Alternaria to reduce postharvest decay of ‘Gala’ apples. Hort- spp., Botrytis spp., Aspergilus spp. and Fu- science 34: 700-704. Dean, R., J. A. L.Van Kan, Z. A. Pretorius, K. E. Ham- sarium spp. In addition, fungal diversity in mond-Kosack, A. Di Pietro, P. D. Spanu, J. J. Rudd, a storage may impact the evolution of viru- M. Dickman, R. Kahmann, J. Ellis, and G. D. Fos- lence in fungal pathogens. Isolates collected ter. 2012. The Top 10 fungal pathogens in molecular from storage facilities with highly diverse plant pathology. Mol. Plant Pathol. 13(7):804. fungal species had significantly more - my Domsch, K. H., W. Gams and T. H. Anderson. 2007. celial growth at 24°C in vitro and larger le- Compendium of Soil Fungi. IHW-Verlag, Eching, sion diameters on apples and oranges than Germany, pp. 672. Droby, S. and A. Lichter. 2004. Post-harvest Botrytis isolates collected from storage facilities with infection: etiology, development and management. less diversity. Storage facilities with a low or In: Elad, Y., B. Williamson, P. Tudzynski and N. high disease incidence had the lowest diver- Delen (eds.). Botrytis: Biology, Pathology and Con- sity in fungal species. Thus, we recommend trol. Kluwer Academic Publishers, Dordrecht, The controlling disease incidence in cold storage Netherlands, pp. 349–367. 74 Journal of the American Pomological Society

Ebert, D. and J. J. Bull. 2003. Challenging the trade-off Jammes, D. 2012. Les Chambres Froides “Pièce model for the evolution of virulence: is virulence Équipée D’installations Frigorifiques Utilisée Pour management feasible? Trends in Microbiol. 11: 15- Stocker Des Denrées Périssables.” Fiche technique. 20. Inter-Reseau Regional Agriculture - Energie - Envi- Ebert, D. and K. L. Mangin. 1997. The influence of ronnement. PACA: Bio de Provence. 2012. http:// host demography on the evolution of virulence of www.bio-provence.org/IMG/pdf/Fiche_technique_ a microsporidian gut parasite. Evolution 51: 1828- chambre_froide_IRRAEE_action_4_.pdf. 1837. Janisiewicz, W. J. and L. Korsten. 2002. Biological Eckert, J. W. and J. M. Ogawa. 1988. The Chemical control of postharvest diseases of fruits. Ann. Rev. Control of Postharvest Diseases - Deciduous Fruits, Phytopathol. 40: 411-441. Berries, Vegetables and Root Tuber Crops. Ann. Jraidi, B. 2010. Foire méditerranéenne de l’agriculture Rev. Phytopathol. 26: 433-469. et des industries agroalimentaires de Sfax, 8ème Eckert, J. W., J. R. Sievert, and M. Ratnayake. 1994. édition du 26 au 29 mai 2010, Tunisie. Reduction of Imazalil Effectiveness against Citrus Jurick, W. M., W. J. Janisiewicz, R. A. Saftner, I. Vico, Green Mold in California Packinghouses by Resis- V. L. Gaskins, E. Park, P. L. Forsline, G. Fazio, and tant Biotypes of Penicillium digitatum. Plant Dis. W. S. Conway. 2011. Identification of wild apple 78: 971-974. germplasm (Malus spp.) accessions with resistance Eckert, J. W. and N. F. Sommer. 1967. Control of Dis- to the postharvest decay pathogens Penicillium ex- eases of Fruits and Vegetables by Postharvest Treat- pansum and Colletotrichum acutatum. Plant Breed- ment. Ann. Rev. Phytopathol. 5: 391-432. ing 130: 481-486. Elad, Y., M. Vivier and S. Fillinger. 2015. Botrytis: Klich, M. A. 2002. Identification of Common Asper- the good, the bad and the ugly. In: Fillinger, S., Y. gillus species, Centraalbureau voor Schimmelcul- Elad and M. Vivier (eds.). Botrytis—the Fungus, tures, Utrecht, The Netherlands. pp 116. the Pathogen and Its Management in Agricultural Koella, J. C. and P. Agnew. 1999. A correlated re- Systems. Springer, Heidelberg, Germany, pp. 1–15. sponse of a parasite’s virulence and life cycle to ElGhaouth, A. 1997. Biologically-based alternatives selection on its host’s life history. J. Evolutionary to synthetic fungicides for the control of postharvest Biol. 12: 70-79. diseases. J. Industrial Microbiol.Biotech. 19: 160-162. Leibinger, W., B. Breuker, M. Hahn, and K. Mendgen. Ewald, P. W. and G. De Leo. 2002. Alternative trans- 1997. Control of postharvest pathogens and coloni- mission modes and the evolution of virulence. In: zation of the apple surface by antagonistic micro- Dieckmann, U. , J. A. J. Metz, M. W. Sabelis and organisms in the field. Phytopathol. 87: 1103-1110. K. Sigmund (eds.). Adaptive dynamics of infec- Leslie, J. F. and B. A. Summerell. 2006. The Fusarium tious diseases: in pursuit of virulence management.. Laboratory Manual, Wiley-Blackwell, Ames, Iowa Cambridge University Press, Cambridge, UK. USA. pp 388. FAOSTAT. 2014. http://www.fao.org/faostat/en/#data/ Louw, J. P. and L. Korsten. 2014. Pathogenic Penicil- QC/visualize. Visited on September of 2017. lium spp. on Apple and Pear. Plant Dis. 98: 590-598. Flier, W. G., L.J.Turkensteen, and A. Mulder. 1998. Louw, J. P. and L. Korsten. 2015. Pathogenicity and Variation in tuber pathogenicity of Phytophthora in- Host Susceptibility of Penicillium spp. on Citrus. festans in the Netherlands. Potato Res. 41: 345-354. Plant Dis. 99: 21-30. Frank, S. A. 1996. Models of parasite virulence. Quar- May, R. M. and R. M. Anderson. 1983. Epidemiology terly Review of Biology 71: 37-78. and Genetics in the Coevolution of Parasites and GIFruits. 2016. Production fruits 2015. Groupement Hosts. Proceedings of the Royal Society of London Interprofessionnel des Fruits. http://gifruits.com/ Series a-Mathematical Physical and Engineering wp-content/uploads/prod.pdf. Sciences 390: 219-219. Gower, C. M. and J. P. Webster. 2005. Intraspecific May, R. M. and M. A. Nowak. 1994. Superinfection, competition and the evolution of virulence in a par- Metapopulation Dynamics, and the Evolution of asitic trematode. Evolution 59: 544-553. Diversity. J. Theoret. Biol. 170: 95-114. Gullino, M. L. and L. A. M. Kuijpers. 1994. Social and Mlika, M., B. Jraidi, S. Aouini, and N. Ben AbdelAali. Political Implications of Managing Plant-Diseases 2002. Document technique Arbres fruitiers Variétés with Restricted Fungicides in Europe. Ann. Rev. fruitières recommandées en Tunisie, n°14. Phytopathol. 32: 559-579. Morales, H., V. Sanchis, A. Rovira, A. J. Ramos, and Holmes, G. J. and J. W. Eckert. 1999. Sensitivity of S. Marin. 2007. Patulin accumulation in apples dur- Penicillium digitatum and P. italicum to postharvest ing postharvest: Effect of controlled atmosphere citrus fungicides in California. Phytopathol. 89: storage and fungicide treatments. Food Control 18: 716-721. 1443-1448. Apple 75

Nair, N. G., S. GuilbaudOulton, I. Barchia, and R. Rupp, S., R.W.S. Weber, D. Rieger, P. Detzel, and M. Emmett. 1995. Significance of carry over inocu- Hahn. 2016. Spread of Botrytis cinerea Strains with lum, flower infection and latency on the incidence Multiple Fungicide Resistance in German Horticul- of Botrytis cinerea in berries of grapevines at har- ture. Front Microbiol. 7, 2075. vest in New South Wales. Austral. J. Expt. Agr. 35: Snowdon, A. L. 1990. Pome fruits. In: Snowdon, A.L. 1177-1180. (eds). A colour atlas of post-harvest diseases and Ogawa, J. M. and H. English. 1991. Diseases of tem- disorders of fruits and vegetables. Vol. 1: General perate zone tree fruit and nut crops. Davis, CA: introduction and fruits. Wolfe Scientifi c Ltd., Lon- Univ. Calif. Div. Agr. and Nat. Resources. don, pp 170-218. Pariaud, B., V. Ravigne, F. Halkett, H. Goyeau, J. Sommer, N., R. J. Fortlage, and D. C. Edwards. 2002. Carlier, and C. Lannou. 2009. Aggressiveness and Postharvest diseases of selected commodities. In: its role in the adaptation of plant pathogens. Plant A. Kader (eds.). Postharvest technology of horticul- Pathol. 58: 409-424. tural crops (3rd ed.). Univ. Calif. Agr. Natural Res., Pepejnjak, S., M. Segvic, and L. Ozegovic. 2002. Ci- Davis, CA. trininotoxinogenicity of Penicillium spp. isolated Spotts, R. A. and L. A. Cervantes. 1986. Populations, from decaying apples. Brazilian J. Microbiol. 33: Pathogenicity, and Benomyl Resistance of Botrytis 134-137. spp, Penicillium spp, and Mucor piriformis in Pack- Pitt, J. I. and A. D. Hocking. 2009. Fungi and Food inghouses. Plant Dis. 70: 106-108. Spoilage, Third Edition.Springer, Berlin, Germany, Staves, P. A. and R. J. Knell. 2010. Virulence and pp 519. Competitiveness: Testing the Relationship during R Core Team. 2013. R: A language and environment Inter- and Intraspecific Mixed Infections. Evolution for statistical computing. R Foundation for Statis- 64: 2643-2652. tical Computing, Vienna, Austria. ISBN 3-900051- Sydenham, E. W., H. F. Vismer, W. F. O. Marasas, N. 07-0. http://www.R-project.org/. Brown, M. Schlechter, L. Vanderwesthuizen, and Ragsdale, N. N. and H. D. Sisler. 1994. Social and J. P. Rheeder. 1995. Reduction of Patulin in Apple Political Implications of Managing Plant-Diseases Juice Samples - Influence of Initial Processing. with Decreased Availability of Fungicides in the Food Control 6: 195-200. United-States. Ann. Rev. Phytopathol. 32: 545-557. Vico, I., N. Duduk, M. Vasić, and M. Nikolić. 2014. Romanazzi, G., J. L. Smilanick, E. Feliziani, and S. Identification of Penicillium expansum causing Droby. 2016. Integrated management of postharvest postharvest blue mold decay of apple fruit. Pesticidi gray mold on fruit crops. Postharvest Biol. Technol. i Fitomedicina 29: 257-266. 113: 69-76. Wilson, C. L., M. E. Wisniewski, S. Droby, and E. Rosenberger, D. A. 1990. Blue mold. In: A. L. Jones Chalutz. 1993. A Selection Strategy for Microbial and H. S. Aldwinkle (eds). Compendium of apple Antagonists to Control Postharvest Diseases of and pear diseases. APS Press, St. Paul, MN. Fruits and Vegetables. Sci. Hort. 53: 183-189. Rosenberger, D. A., C. A. Engle, F. W. Meyer and, C. Xiao, C. L. and Y. K. Kim. 2008. Postharvest fruit rots B. Watkins. 2006. Penicillium expansum invades in apples caused by Botrytis cinerea, Phacidiopyc- apples through stems during controlled atmosphere nis washingtonensis, and Sphaeropsis pyriputres- storage. Online. Plant Health Progress. cens. Online. Plant Health Progress.

76 Journal of the American Pomological Society

Index for Volume 72 A Cold-hardy ...... 80, 166 Abdelghafar, A...... 40 Container-grown blueberry...... 222 Agricultural byproduct ...... 101 Crain, B...... 122 AL Zein, M...... 260 Crane, J...... 242 Anderson, N.O...... 146 Crop density...... 231 Anthracnose canker...... 113 Cross-breeding ...... 173 Antioxidant capacity ...... 40 Cultivars...... 157 Apple ...... 113 Cultural control ...... 113 Aradhya, M...... 242 As-Sadi, F...... 260 D Atucha, A...... 80 Days from bloom to harvest ...... 231 Devetter, L.W...... 113 B Diospyros kaki...... 74 Battle, I...... 21 Diriliş ...... 173 Berries ...... 157 Disease ...... 94 Berry size ...... 128 Distribution ...... 260 Bioactive compounds...... 40 Dormancy ...... 166 Black berry cultivars ...... 181 Black walnut ...... 135 E Blueberry ...... 2, 101 Early cropping ...... 222 Borda, M.P...... 222 Early season temperature...... 231 Botritis cinerae ...... 122 Ehlenfedlt, M.K...... 2 Brianna ...... 80 EL Riachy, M...... 260 Brower, T.M...... 242 Brown rot ...... 94 F Brown-Rytlewski ...... 122 Fakih, L...... 260 Bükücü, Ş.B...... 173 Fallahi, E...... 231 Burrell, R...... 40, 94 Farga cultivar ...... 21 Fertilization ...... 212 C Floricane...... 122, 181, 251 Carob gerplasm...... 260 Flower initiation ...... 212 Carya illinoinensis ...... 101 Flower bud removal ...... 22 Castanea mollissima ...... 12 Francescatto, P...... 231 Ceratonia siliqua ...... 260 Freezingtolerance...... 166 Chalak, L...... 260 Frontenac ...... 80 Chami, M...... 260 Fruit ...... 74 Chater, J.M...... 157 Fruit characteristics...... 74 Chavez, D.J...... 166 Fruit quality...... 212 Chemical control ...... 113 Fruit rot ...... 122 Chineese chestnut ...... 12 Fruit set ...... 173 Cho K.S...... 74 Fruit size...... 212 Citrus huanglongbing...... 242 Fruit weight ...... 231 Cline, J...... 221 Fu, W...... 94 Index 77

G K GA4+7 ...... 135 Kahwaji, J...... 260 Gariglio, N.F...... 222 Kang, J.H...... 74 Garton, W.J...... 113 Kapps, M.L...... 128 Gasic, K...... 40, 94 Kessler, J.R., Jr...... 101 Genetic resources ...... 21 Khalil, R. J...... 212 Germplasm...... 40, 157 Kim, Y.K...... 74 Ghaith, S...... 260 Kostick, S.A...... 146 Gottwald, T...... 242 Hwana, H.S...... 74 Gozlekci, S...... 242 Grape ...... 80 L Growing degree days...... 231 La Crescent...... 80 Growth ...... 12 La Crosse ...... 80 Guava SSR analysis ...... 242 Laterally fruitful ...... 173 Guldan, S...... 202 Lebanon ...... 260 Liberibacter ...... 242 H Linge, C.D.S...... 94 Hajj, A...... 260 Lordan, J...... 231 Hall, D...... 242 LT50 ...... 166 Han, J.H...... 74 Hanson, E...... 122 M Haploblock ...... 94 Ma, K.B...... 74 Harvest dates...... 128 Mallardi ...... 166 Hedicke, J...... 74 Malus ...... 29 Hermoso, J.F...... 21 Malus xdomestica ...... 113,212 Heyduck, R...... 203 Marini, R.P...... 231 Highbush ...... 2 Marketable yield ...... 128 High elevation ...... 202 Marquette ...... 80 High pH soil ...... 202 Mazzola, M...... 113 High tunnels ...... 122, 202, 251 Merhaut, D.J...... 157 Hirst, P.M...... 212 Metaxenia ...... 212 History ...... 29 Microsatellite ...... 242 Hoover, E.E...... 146, 195 Miles, C.A...... 113 Hoppers, A.A...... 101 Mineral nutrient concerntration ...... 181 Hormones ...... 135 Molecular marker...... 242 Huanglongbing ...... 242 Monilinia spp...... 94

I N Interspecific hybrid...... 166 Nabbout, R...... 260 Invasive...... 146 Navarro-Cerillo, R...... 260 NC-140 peach physiology trial...... 231 J Neofabraea malicorticis...... 113 Jia, Z...... 157 Newell, M.J...... 231 Juglans nigra ...... 135 Ninot, A...... 21 Juglans regia...... 173 Nitrogen ...... 12 78 Journal of the American Pomological Society

Nut yield...... 12 Prunus salicina ...... 146 Nutrients ...... 12 Psidium guajava...... 242

O Q Olive ...... 21 QTL ...... 94 Ouellette, D...... 231 R Organic Blackberry...... 202 Rabbiteye ...... 2, 101 Organic system ...... 181 Ramong ...... 74 Özcan, A...... 173 Redpath, L.E...... 166 Red raspberry ...... 122, 195, 251 P Reighard, G...... 40, 231 Palacios-Rodriguez, G...... 260 Return bloom...... 212 Paul Howe Shepard...... 29 Riley, C.M...... 242 Parthenocarpy ...... 2 Robinson, T.L...... 231 Peach breeding...... 40, 94 Romero, A...... 21 Pecan shell mulch ...... 101 Root distribution ...... 101 Persimmon ...... 74 Root growth ...... 101 Pescie, M.A...... 222 RosBREED ...... 94 Phenolic compounds...... 40 Rubus idaeus ...... 122, 195, 251 Phenological ...... 173 Rubus L. subgenus Rubus Watson ...... 202 Phenotypic variability ...... 260 Ruiz Gomez, F.J...... 260 Photosynthesis ...... 157 Physiology ...... 157 S Physiological disorder ...... 74 Santiago, L.S...... 157 Plant breeding ...... 29, 74 Scarification ...... 146 Plant growth regulators...... 135 Schnabel, G...... 94 Pod and seed morphological Season extension...... 202 characteristics...... 260 Seed germination ...... 146 Pollination ...... 212 Seed number ...... 212 Pollen source ...... 212 Sensorial profile ...... 21 Pomegranate ...... 157 Shepard award ...... 221 Pomological...... 173 Shin, L.S...... 74 Powell, A.A...... 101 Smaha, M...... 260 Preece, J.E...... 157 Smith, E.D...... 166 Primocane...... 122, 181 Soluble solids concentration...... 212 Productivity ...... 202 Southern highbush blueberry...... 166, 222 ProGibb ...... 135 Spiers, J.D...... 101 Provide ...... 135 Stover, E...... 242 Prunica granatum ...... 157 Stratification ...... 135 Prunus ...... 29, 40 Strawberry ...... 128 Prunus americana ...... 146 Strik, B.C...... 181 Prunus armeniaca ...... 146 Stringer, S.J...... 2 Prunus cerasus ...... 146 Sufficiency range ...... 12 Prunus domestica ...... 146 Survey range ...... 12 Prunus persica ...... 231 Sütyemez, M...... 173 Index 79

T W 15 Temmuz ...... 173 Walnut ...... 173 Tepe, E...... 146 Warmund, M.R...... 12, 29, 135 Tillman, J...... 146 Water relations ...... 157 Tous, J...... 260 Weber, C.A...... 195, 251 Total yield ...... 128 Winter damage ...... 202 Tree canker ...... 113 Winter hardy...... 146 Wisconsin...... 80 U Wolfe, D...... 231 USDA ...... 157 Workmaster, B.A...... 80 Wright, A...... 101 V Vaccinium ashei ...... 2 Y Vaccinium corymbosum ...... 2, 166 Yield ...... 195 Vaccinium corymbosum interspecific Yao, S...... 202 hybrids ...... 222 Vaccinium virgatum ...... 2, 101 Z Van Sambeek, J.W...... 135 Zee, F...... 242 Vance, A.J...... 181 Vietnam ...... 242 Vitus ...... 29 80 Journal of the American Pomological Society

Instructions to Authors Journal of the American Pomological Society

The prime purpose of the Journal of the detailed instructions for manuscript prepa- American Pomological Society is to provide ration can be found at: http://www.ameri- a repository for information on all aspects of canpomological.org/journal/journal.instruc- fruit and nut crops. The long-term emphasis tions.html of the journal on cultivars and rootstocks Before submission, manuscripts should continues, but manuscripts reporting origi- be reviewed by at least two colleagues and nal research on a wide range of fruit and nut revised accordingly. At the time of submis- crops are welcomed. Acceptable areas of re- sion, the corresponding author must attest in search including pruning, nutrition, growth the covering letter to the Editor that all coau- regulators, cultural practices, economics, and thors on the paper have had the opportunity pest control. Studies involving the interaction to review it before to submission, that it has of one or more of these aspects with either not been published previously, and that it is cultivars and/or rootstocks are particularly not presently under consideration for publi- appropriate. If in doubt about the suitability cation elsewhere. In addition, the names and of a particular manuscript, please contact the full contact information (mailing address, Editor. e-mail and telephone numbers) for three po- Reports on field studies are expected to tential reviewers should be provided. Submit contain data from multiple years. Reports manuscripts electronically to the Editor: Dr. are to be the result of adequately replicated Richard Marini, 203 Tyson Building, Depart- trials and the data should be subjected to ap- ment of Plant Science, University Park, PA propriate statistical analysis. Manuscripts 16802-4200 USA; E-mail: richmarini1@ submitted for publication in the Journal must gmail.com. Acceptable format is MSWord. not have been previously published, and sub- Manuscripts are sent to two reviewers mission implies no concurrent submission competent to evaluate scientific content. Ac- elsewhere. ceptance for publication depends upon the Scientific names and authorities for plants, combined judgement of the two reviewers disease organisms, and insects should be in- and the Editor. In unusual circumstances the cluded parenthetically when the organism is Editor, without further review, may return a first mentioned. American spelling conven- manuscript, which obviously does not meet tions and SI units should be used. Manu- Journal standards, to the author. scripts should be double spaced throughout. A charge of $50.00 per page for APS Typical organization is as follows: Title, members (at least one author is a member) Authors, Abstract, Introduction, Materials and $65.00 per page ($32.50 per half page) and Methods, Results, Discussion, Literature for nonmembers will be made to authors for Cited, Tables, Figures. The Results and Dis- those articles constituting publication of re- cussion sections are often combined. Author search. In addition to the page charge, there addresses, email adresses and acknowledge- will be a charge of $40.00 per page for tables, ments are in footnotes on the first page. More figures and photographs. American Pomological Society 2019 Annual Meeting to be held in conjunction with The Annual Conference of the American Society for Horticultural Science 22 – 25 July • Las Vegas, Nevada The Tropicana Las Vegas

Call for Wilder Silver Medal Nominations The Wilder Committee of the American Pomological Society (APS) invites nominations for the 2019 Wilder Silver Medal Award. All active members of APS are eligible to submit nominations. The award was established in 1873 in honor of Marshall P. Wilder, the founder and first president of APS. The award consists of a beautifully engraved medal which is presented to the recipient at the annual meeting of APS, held during the ASHS Annual Meeting. The Wilder medal is presented to individuals or organizations that have rendered outstanding service to horticulture in the area of pomology. Special consideration is given to work relating to the origination and introduction of meritorious fruit cultivars. Individuals associated with either commercial concerns or professional organizations will be considered if their introductions are truly superior and have been widely planted. Significant contributions to the science and practice of pomology other than through fruit breeding will also be considered. Such contributions may relate to any important area of fruit production such as rootstock development and evaluation, anatomical and morphological studies, or noteworthy publications in any of the above subjects. Information about the award, past recipients, etc. can be found on the APS website at: http://americanpomological.org/wilder1.html To obtain nomination guidelines, please contact committee chairperson, Dr. John R. Clark Dept. of Horticulture, University of Arkansas phone: 479-575-2810 fax 479-575-8619 e-mail: [email protected] Nominations must be submitted by 1 May 2019. JOIN THE American Pomological SOCIETY!

Membership Benefits: • Online and/or printed access to the Journal of APS • Networking opportunities with world-renowned pomologists • Award opportunities for superior pomology-related research and student papers • Leadership opportunities

Individual, Student, and Join Online Library Rates TODAY! Available

http://americanpomological.org