University of Florida Thesis Or Dissertation Formatting

EXPLORING AVOCADO VARIABILITY FOR LAUREL WILT RESISTANCE AND EXCELLENT FRUIT QUALITY AND HORTICULTURAL TRAITS FOR PRODUCTION IN EAST-CENTRAL FLORIDA

CRISTINA PISANI

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2016

To my husband, Federico, you are my helper, best friend, and support. You are my blessing and I love you Love is of all passions the strongest, for it attacks simultaneously the head, the heart, and the senses. — Lao Tzu

ACKNOWLEDGMENTS

I would like to express my gratitude to my supervisor Dr. Mark Ritenour and co- advisor Dr. Ed Stover, who with their expertise and moral support have helped me in my personal, professional, and academic development. I thank them for making me feel welcome in their labs, homes, and for allowing me to share great moments with their families. I would also like to send my most sincere appreciation to the rest of my committee Dr. Randy Ploetz for his invaluable expertise in tropical fruit plant pathology,

Dr. Osman Gutierrez for his breeding background, and Dr. Gloria Moore for making me feel welcome in her lab while living in Gainesville. I thank them for giving me valuable feedback and making significant contributions to this study.

I would like to thank all that have helped me in so many ways with my research.

The USDA-ARS Picos farm crew Steve Mayo, Sean Reif, David Peabody, and Patrick

Zagorski for all their hard work in maintaining the trees used in this study. Special thanks to the postharvest lab, in particular Mac Hossain, Shamima Hossain and Cuifeng

Hu for their help and expertise with postharvest equipment and data processing. I thank

Malu Oliveira for her molecular expertise and the Stover lab for allowing me to be part of their crew for part of my experiments and Shatters lab for all their support. Special thanks to Dr. Dov Borovsky in Shatters lab, Dr. Jeyaprakash, and David Davison at the

Florida Department of Agriculture and Consumer Services in Gainesville for all their help, time, and expertise with the molecular portion of my research. I want to greatly thank Dr. Rocco Alessandro for allowing me to use his lab and equipment and mentoring me in chemistry and lipid extraction. I would also like to thank Dr. Anne Plotto and the sensory team Dave Wood, Skylar Sirmans, and Carly Franko for all their help and knowledge in sensory panels. Special thanks to Carly Franko who has helped me

with lipid extractions and whom without her help, the chapter would not exist. I would also like to thank Dr. Raymond Schnell for conceiving of and developing the ‘Hass’ x

‘Bacon’ and ‘Bacon’ x ‘Hass’ populations and Mike Winterstein for growing the trees in the USDA/ARS Miami nursery. I would like to extend considerations to Dr. David Kuhn and Barbara Freeman at USDA/ARS Miami, Josh Konkol at TREC in Homestead for their knowledge and help with any avocado questions.

I thank the University of Florida and the Horticultural Sciences staff for giving me the administrative conditions and support for me to smoothly conduct my research and to all my fellow graduate students for all the help and advice during my experiments. I would also like to thank the Simpson family, Mr. Bud Adams, St. Lucie County master gardeners club, Mr. Harold E. Kendall Sr., and the garden club of Ft. Pierce for believing in my research and awarding scholarships to make better contributions.

Lastly, I send my most sincere appreciation to my husband, Federico Caro for constantly supporting me through this hard life journey and always sticking by my side with all his love. I also thank my family and close friends for all their continuous support.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 9

LIST OF FIGURES ...... 11

LIST OF ABBREVIATIONS ...... 14

ABSTRACT ...... 17

CHAPTER

1 LITERATURE REVIEW ...... 19

Avocado Origin and Industry ...... 19 Fruit Development and Composition ...... 23 Laurel Wilt Disease ...... 25 The Pathogen ...... 25 The Vector(s) ...... 29 Host Trees ...... 34 Management ...... 37

2 FLOWERING OF HYBRIDS FROM RECIPROCAL CROSSES OF ‘HASS’ AND ‘BACON’ AS AN INDICATION OF ADAPTATION TO CENTRAL FLORIDA CONDITIONS ...... 42

Introduction ...... 42 Materials and Methods...... 45 Results ...... 46 Discussion ...... 49 Conclusion ...... 50

3 POSTHARVEST EVALUATION OF HYBRIDS FROM RECIPROCAL CROSSES BETWEEN AVOCADO CVS ‘HASS’ AND ‘BACON’ ...... 63

Introduction ...... 63 Materials and Methods...... 66 Tree Descriptions ...... 66 Fruit (Preharvest evaluations) ...... 67 Postharvest Quality Parameters ...... 68 Peel color ...... 68 Whole fruit firmness ...... 68 Weight loss ...... 69 Fruit pulp firmness ...... 69

Internal quality and decay ...... 69 Fruit phenotypic data ...... 70 Dry matter ...... 70 Statistical analysis ...... 71 Results ...... 71 Postharvest Quality Parameters ...... 71 Phenotypic data, dry matter, and weight loss...... 71 Peel color ...... 72 Whole fruit firmness and pulp firmness ...... 72 Internal quality and decay ...... 74 Discussion ...... 74 Conclusion ...... 77

4 CHARACTERIZATION OF FATTY ACID CONTENT FROM THE PULP OF AVOCADO FRUITS OF SELECTED ‘HASS’-‘BACON’ HYBRIDS GROWN IN EAST CENTRAL FLORIDA ...... 101

Introduction ...... 101 Materials and Methods...... 103 Raw material ...... 103 Lipid extraction and methylation ...... 104 Fatty acid analysis ...... 105 Results ...... 106 Discussion ...... 107 Conclusion ...... 111

5 SENSORY EVALUATION OF SELECTED ‘HASS’-‘BACON’ AVOCADO HYBRIDS GROWN IN AN EAST CENTRAL FLORIDA CLIMATE ...... 117

Introduction ...... 117 Materials and Methods...... 120 Fruit description ...... 120 Sensory evaluation ...... 122 Statistical analysis ...... 123 Results ...... 124 Fruit description ...... 124 Fatty acid analysis ...... 126 Sensory evaluation ...... 127 Discussion ...... 128 Conclusion ...... 133

6 SCREENING AVOCADO GERMPLASM FOR RESISTANCE TO LAUREL WILT DISEASE ...... 144

Introduction ...... 144 Materials and Methods...... 146 Results ...... 148

Discussion ...... 149 Conclusion ...... 152

7 PRELIMINARY STUDIES ON DEVELOPING A HIGH THROUGHPUT SCREENING ASSAY USING QUANTITATIVE PCR FOR DETECTING RAFFAELEA LAURICOLA, CAUSAL AGENT FOR LAUREL WILT DISEASE, IN PERSEA AMERICANA SHOOT CUTTINGS ...... 160

Introduction ...... 160 Materials and Methods...... 162 Preliminary study ...... 162 Screening using qPCR assay ...... 163 Plant material ...... 163 DNA extraction and amplification ...... 164 Results ...... 165 Preliminary study ...... 165 Screening using qPCR assay ...... 166 Discussion ...... 166 Conclusion ...... 169

8 FINAL CONCLUSIONS ...... 174

APPENDIX

SEQUENCE OF QUESTIONS ASKED TO PANELISTS DURING SENSORY ANALYSIS OF FRESH AVOCADO ...... 179

LIST OF REFERENCES ...... 180

BIOGRAPHICAL SKETCH ...... 192

LIST OF TABLES

Table page

2-1 Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/27/2013 and 4/11/2013...... 52

2-2 Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/20/2014, 4/3/2014, and 4/16/2014...... 53

2-3 Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/18/2015 and 4/1/2015...... 54

3-1 Peel color of unripe and ripe avocado fruit from different hybrids in 2013...... 80

3-2 Peel color of unripe and ripe avocado fruit from different hybrids in 2014...... 82

3-3 Peel color of unripe and ripe avocado fruit from different hybrids in 2015...... 83

3-4 Change in peel color between ripe and unripe avocado fruit in 2013...... 84

3-5 Change in peel color between ripe and unripe avocado fruit in 2014...... 86

3-6 Change in peel color between ripe and unripe avocado fruit in 2015...... 87

3-7 Pulp firmness of unripe avocado fruit from different hybrids in 2013...... 93

3-8 Pulp firmness of ripe avocado fruit in 2013...... 95

3-9 Pulp firmness of unripe avocado fruit from different hybrids in 2014...... 97

3-10 Pulp firmness of ripe avocado fruit from different hybrids in 2014...... 98

3-11 Pulp firmness of ripe avocado fruit from different hybrids in 2015...... 99

3-12 Postharvest disorders on ripe avocado fruit harvested over three seasons (2013-2015)...... 100

4-1 Fatty acid composition of ‘Hass’-‘Bacon’ hybrids grown in east-central Florida 113

4-2 Average fatty acid methyl ester percentage composition (±SE) of avocado pulp oils from trees of a mapping population in 2013...... 114

4-3 Average fatty acid methyl ester percentage composition (±SE) of avocado pulp oils from trees of a mapping population in 2014...... 116

5-1 Phenotypic fruit data on selected avocado trees in 2013...... 134

5-2 Phenotypic fruit data on selected avocado trees in 2014...... 135

5-3 Phenotypic fruit data on selected avocado trees in 2015...... 135

5-4 Peel and pulp color of ripe avocado fruit in 2014...... 136

5-5 Peel and pulp color of ripe avocado fruit in 2015...... 136

5-6 Whole fruit firmness of ripe avocado fruit in 2014...... 137

5-7 Whole fruit firmness of ripe avocado fruit in 2015...... 137

5-8 Pulp firmness of ripe avocado fruit in 2014...... 138

5-9 Pulp firmness of ripe avocado fruit in 2015...... 138

5-10 Average fatty acid (FAME) percentage composition (±SE) of avocado pulp oils by GC-FID in 2014 and 2015 taste panels...... 139

5-11 Average fatty acid (FAME) percentage composition (±SE) of avocado pulp oils of individual selections by GC-FID in 2014 and 2015 taste panels...... 139

6-1 Response of different avocado seed-source families to laurel wilt disease.a .... 155

7-1 RT-qPCR results of R. lauricola DNA extracted from avocado wood tissues. .. 173

LIST OF FIGURES

Figure page

1-1 Raffaelea lauricola extracted from a cutting of ‘Simmonds’ avocado through plating onto the selective medium CSMA+...... 40

1-2 Foliar symptoms of laurel wilt disease on avocado at the USDA Horticultural Research Laboratory Picos Farm, Ft. Pierce...... 41

1-3 Tree bolt showing Xyleborus glabratus frass tubes and galleries on redbay ...... 41

2-1 Female flowers...... 51

2-2 Male flowers...... 51

2-3 Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2013 season...... 55

2-4 Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2014 season...... 56

2-5 Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2015 season...... 57

2-6 Frequency plot showing number of trees demonstrating max bloom at given dates during the 2013 season...... 58

2-7 Frequency plot showing number of trees demonstrating max bloom at given dates during the 2014 season...... 58

2-8 Frequency plot showing number of trees demonstrating max bloom at given dates during the 2015 season...... 59

2-9 Average percentage of open flowers per tree per bi-week of the California mapping population of ‘Hass’-‘Bacon’ hybrids for the years 2013-2015...... 59

2-10 Frequency plots showing number of trees demonstrating max flower clusters during the 2013, 2014, and 2015 seasons...... 60

2-11 Average fruit length and width after fruit set measured monthly up until harvest for the years 2013, 2014, and 2015...... 61

2-12 FAWN (Florida Automated Weather Network) data for average temperature in a Fort Pierce & Saint Lucie West weather station for the years 2013-2015. ... 62

3-1 Mean percentage of daily fruit weight loss rate over 5 days after harvest in 2013...... 78

3-2 Mean percentage of daily fruit weight loss rate over 6 days after harvest in 2014...... 78

3-3 Mean percentage of daily fruit weight loss rate over 5 days after harvest in 2015...... 79

3-4 Average whole fruit firmness (N) of avocado for three harvest seasons...... 88

3-5 Whole fruit firmness average peak force at days 0 and 8 of 10/31/2013 harvest...... 89

3-6 Whole fruit firmness average peak force at days 0 and 7 of 11/14/2013 harvest...... 90

3-7 Whole fruit firmness average peak force at days 0 and 8 of 10/28/2014 harvest...... 91

3-8 Whole fruit firmness average peak force at days 0 and 7 of 10/23/2015 harvest...... 92

4-1 Relative percent composition of avocado fatty acids of fruit harvested in 2013 and 2014...... 112

5-1 Biplot of the Principal Component Analysis of fatty acid analysis of avocado fruit used in sensory taste panels in 2014 and 2015...... 140

5-2 Percentage of panelists in 2013, 2014, and 2015 characterizing the avocado flesh texture using the indicated descriptors...... 141

5-3 Percentage of panelists in 2013, 2014, and 2015 characterizing the avocado flesh flavor using the indicated descriptors...... 142

5-4 Overall liking of avocado selection in 2013, 2014, and 2015...... 143

6-1 Foliar symptoms of laurel wilt disease on avocado ‘Choquette’ and ‘Winter Mexican’ at the USDA Horticultural Research Laboratory Picos Farm, Ft. Pierce...... 154

6-2 Mean disease severity to laurel wilt in 2013 ...... 156

6-3 Mean disease severity to laurel wilt in 2014 ...... 157

6-4 Percentage of tree mortality after disease severity ratings...... 158

6-5 Relationship between stem diameter of avocado families and the severity of laurel wilt that developed after field inoculations in 2013 and 2014...... 159

7-1 Shoot cuttings set-up before inoculation with Raffaelea lauricola...... 170

7-2 Measurements of vascular staining distance from inoculation point and disease severity assessments on shoot cuttings of ‘Hass’-‘Bacon’ hybrids, ‘Lula’, and ‘Marcus Pumpkin’...... 171

7-3 Calibration curve used for quantitative real time PCR assay...... 172

LIST OF ABBREVIATIONS

AgMRC Agricultural Marketing Resource Center a.m. ante meridiem

AUDPC Area under disease pressure curve

BLAST Basic Local Alignment Search Tool

BC Before Christ cm centimeter cv. cultivar

˚C degrees Celsius

C2H4 ethylene

F Fahrenheit

FAMEs Fatty acid methyl esters

GC-MS gas chromatography-mass spectrometry

GLMM generalized linear mixed model g gram

GFP Green Fluorescent Protein

G Guatemalan

G-M Guatemalan-Mexican hybrid

HDL high density lipoprotein

HCl hydrochloric acid in inch kg kilogram

KN kilonewton

LSU Large subunit

< less than

LDL low density lipoprotein

MEA malt extract agar m meter

NaOH/MeOH methanolic sodium hydroxide

1-MCP 1-methylcyclopropene

MT metric tons

M Mexican

µl microliter

µm micrometer mg milligram ml milliliter mm millimeter min minute

MUFAs monounsaturated fatty acids

> more than ng nanogram

NASS National Agricultural Statistics Service

NCGR National Clonal Germplasm Repository

N Newton ppm parts per million

% percent

PAA peroxyacetic acid pg picogram

PCR Polymerase Chain Reaction psi pounds per square inch

PCA Principal Components Analysis rpm revolutions per minute rDNA ribosomal deoxyribonucleic acid

SSR Simple Sequence Repeat

SNP Single Nucleotide Polymorphism

SSU Small subunit

NaCl sodium chloride

SCFS South Coast Field Station sp. species

SAS Statistical Analysis Software developed by the SAS Institute

Subtropical Horticulture Research Station-Agricultural Research SHRS-ARS Service

UF-IFAS University of Florida-Institute of Food and Agricultural Sciences IRREC Indian River Research and Education Center

UF, TREC University of Florida, Tropical Research and Education Center

United States Department of Agriculture-Agricultural Research USDA-ARS Service

United States Horticultural Research Laboratory- Agricultural USHRL-ARS Research Service var. variety

WI West Indian

WI-G West Indian-Guatemalan hybrid

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

EXPLORING AVOCADO VARIABILITY FOR LAUREL WILT RESISTANCE AND EXCELLENT FRUIT QUALITY AND HORTICULTURAL TRAITS FOR PRODUCTION IN EAST-CENTRAL FLORIDA

Cristina Pisani

May 2016

Chair: Mark A. Ritenour Co-chair: Eddie W. Stover Major: Horticultural Science

Avocado (Persea americana) is newly threatened by the fungal laurel wilt pathogen (Raffaelea lauricola) vectored by an Asian ambrosia beetle (Xyleborus glabratus), which has devastated native populations of the Lauraceae. This study aimed to assess laurel wilt resistance by broadly screening half-sibling populations from the

Miami National Clonal Germplasm Repository. Additionally, this study evaluated fruit quality and horticultural traits for production in east central Florida of Mexican-

Guatemalan hybrid trees consisting of a population of Hass and Bacon reciprocal crosses for potential cultivar identification.

During two consecutive years of inoculation with the R. lauricola pathogen,

Melendez, Booth 8, and Yon families had least trees succumb to the disease, while

Ettinger had most trees die to laurel wilt without recovery. Argui 1, Booth 8, and

Melendez showed lowest disease severity, while No. 21 was most susceptible during both years.

To facilitate higher throughput screening, we also assessed resistance evaluation using cuttings. Preliminary experiments showed that R. lauricola can be quantified using qPCR, but the assay needs further development.

Promising selections with potential as cultivars were identified with excellent horticultural traits and fruit quality. During sensory evaluations R8T18 and R6T56 were rated as the creamiest two consecutive years in a row. All selections were liked in flavor and texture attributes and appeared to have fruit quality similar to Hass.

When fatty acid content was analyzed as a measure for fruit maturity, oleic acid was the main fatty acid in avocado mesocarp tissue (34.9% in 2013 and 36.6% in 2014 respectively). Palmitic acid was also abundant in fruit (up to 27.9% and 30.0%) and the third most abundant fatty acid was linoleic acid (21.2% and 18.4%). Both stearic and linolenic acids were scarce (0.4-2%), with myristic only found in trace amounts as reported elsewhere.

Overall, selections chosen to be included in sensory panels have shown the possible potential for Hass-like avocados grown in a Florida climate. Families showing more tolerance/resistance to R. lauricola were identified and will be included in future genotyping studies.

CHAPTER 1 LITERATURE REVIEW

Avocado Origin and Industry

Avocado (Persea americana Miller) is in the Laurales order and the Lauraceae family. It developed distinct subgroups due to geographical isolation that are considered either distinct races, subspecies or even different species by different taxonomists

(Scora et al. 2002). The Mexican (M) subspecies (Persea americana var. drymifolia) is native to the highlands of Mexico, has anise-scented leaves and is the most cold hardy of all races. Trees of the Mexican race can withstand temperatures as low as (-4˚C), and its fruit is small, with a smooth skin and pulp with a higher oil content than the other avocado races (up to 30%). The Guatemalan (G) subspecies (Persea americana var. guatemalensis) is native to the highlands of Central America and is somewhat cold tolerant. The fruit is round, with thick, rough, brittle skin and pulp with an intermediate oil content of 8-15% and nutty flavor. The West Indian (WI) subspecies (Persea americana var. americana) is native to the lowlands of Central America and Northern South

America. It is the most cold sensitive of the three races and cannot withstand temperatures below -1.2˚C. Its fruit have smooth, leathery, glossy skin and pulp with the lowest oil content at 3-10% (Purseglove 1968; Yahia & Woolf 2011).

Avocado probably originated in Mesoamerica (southern Mexico, Guatemala, and

Honduras) (Kopp 1966; Litz et al. 2005). Archeological evidence (seed remains) in the

Tehuacán valley, Puebla, Mexico indicated that avocado may have been one of the earliest trees domesticated in the Neotropics (Smith 1969; Galindo-Tovar et al. 2008).

Primitive avocado trees have been found in the Oriental Sierra Madre from the State of

Nuevo León in Mexico all the way south to Costa Rica (Yahia & Woolf 2011). Forest

gardens of trees were cultivated between 4,500-2,000 B.C., and subsequent

Mesoamerican cultures, such as the Mokaya, Olmecs, Mayas, and Valdivia, continued the domestication process and were responsible for the tree’s dispersal to Honduras,

Central America, and northern South America (Buckler et al. 1998; Galindo-Tovar et al.

2008). The Aztecs called the avocado fruit “Ahuacatl” from which the modern Spanish term “aguacate” originated (Yahia & Woolf 2011). For Mesoamerican cultures, the fruit was considered an aphrodisiac, was used as medicine and tribute, and had religious significance (Galindo-Tovar et al. 2008).

Avocados from Mexico and Central America where first planted in the United

States in the 1800s. The first trees were planted in Florida in 1833 and in California in

1856 (Schaffer et al. 2013). In 1911 nurseryman F.O. Popenoe brought the first high quality avocado fruit from Mexico and introduced it to Southern California. This cultivar from Atlixco, Mexico is a natural Guatemalan-Mexican hybrid called ‘Fuerte’. ‘Fuerte’ became the dominant cultivar in the California avocado industry until the 1970s and was later replaced in the 1980s by a new cultivar developed in California, the ‘Hass’, another

Guatemalan-Mexican hybrid. Today, ‘Hass’ is responsible for over 90% of California production is the main cultivar in other countries such as South Africa, Australia, New

Zealand, Spain, Chile, Israel, and Mexico (Schaffer et al. 2013). Other cultivars in

California include ‘Bacon’, ‘Gwen’, ‘Reed’, ‘Zutano’, and ‘Pinkerton’. In Florida, the main avocado cultivars are West Indian and West Indian-Guatemalan hybrids that are more suited for humid tropical climates.

Florida has three commercial avocado seasons. The early season begins in late

May and runs through August and is predominated by West Indian cultivars. West

Indian-Guatemalan hybrids predominate mid and late season, which run from

September through October and November through March, respectively (Crane et al.

2007). The most popular cultivars in Florida include ‘Bernecker’, ‘Beta’, ‘Choquette’,

‘Donnie’, ‘Dupuis’, ‘Hall’, ‘Lula’, ‘Monroe’, ‘Nadir’, ‘Nesbitt’, ‘Simmonds’, and ‘Tonnage’

(Ploetz et al. 1994; Crane et al. 2007; Crane et al. 2013).

Commercial avocado trees are grafted onto open pollinated seedling and clonal rootstocks. In California, most plants are on Mexican or Guatemalan-Mexican hybrid clonal rootstocks while in Florida West Indian seedling rootstocks are more common.

Clonal rootstocks, which often provide resistance to Phytophthora root rot caused by

Phytophthora cinnamomi, have been developed by several different programs. For example, ‘Duke 7’ was developed at the University of California-Riverside whereas

‘Dusa’, a current industry standard, was developed by the Westfalia group in South

Africa (Schaffer et al. 2013).

California leads the US in avocado production followed by Florida and Hawaii

(<0.5%) (NASS 2014). The top five avocado producing countries are Mexico (1,288,804 metric tons (MTs)), followed by the Dominican Republic (315,330 MT), Indonesia

(267,685 MT), Colombia (244,769 MT), and Peru (238,736 MT) (FAOSTAT 2014)

The bearing acreage in the U.S. decreased from 66,270 in 2009 to 61,300 in

2014 with a corresponding decline in production from 298,520 tons in 2009 to 197,450 tons in 2014. Value of production dropped from $429.6 million to $351.1 million (NASS

2014).

Demand for avocado in the U.S. continues to increase. According to the

Agricultural Marketing Resource Center (AgMRC) and the USDA Economic Research

Service (USDA ERS), U.S. consumption of fresh avocados has had a strong increasing trend since 1970. More recently, in 1989 consumption was 1.1 pounds per capita which increased to 2.1 pounds per capita in 2000, 4.8 pounds per capita in 2011, and 5.2 pounds per capita in 2013. Despite its growing demand and a growing population, prices have generally increased since 1980 although they have been highly variable. In

2013, the average price of U.S. avocados reached $1,990 per ton (net to grower), a decrease of $180 per ton from the previous two years. Average 2013 California prices were $2,400 per ton while average Florida prices were $800 per ton. However, Florida prices were still far below the 2005 price of $940 per ton. The price for California avocados is higher than that of Florida avocados largely due to consumer preference for the higher oil varieties produced in California (AgMRC 2013; NASS 2014).

Avocados are an excellent source of monounsaturated oils and a whole medium avocado (‘Hass’) contains 15 percent of the FDA’s recommended daily amount of unsaturated fat (15.41 g fat per 100 g-1 fresh weight in a California avocado and 10.06 g fat per 100 g-1 fresh weight in a Florida avocado, respectively) (AgMRC 2013; USDA

NNDSR 2014). Avocados are also a great source of potassium since a fruit can contain about 60 percent more potassium than a banana (480 vs 358 mg 100 g-1) (Yahia &

Woolf 2011; USDA NNDSR 2014). They are also rich in B vitamins, vitamin E, vitamin K and folate.

In 2002, the USDA lifted a ban that had prohibited the entry of ‘Hass’ avocado from Mexico and Central America since 1914 and established the U.S. Federal Hass

Avocado Promotion, Research and Information Order. Avocados could then be shipped to the United States from approved orchards in Mexico but with certain restrictions to

avoid saturation of the US market. Mandatory annual field surveys were required to help prevent the entry of avocado pests. Most U.S. avocado imports are now from Mexico,

Chile, and the Dominican Republic. According to the USDA’s Foreign Agricultural

Service, the value of avocado imports reached $711.2 million in 2009 but dropped to

$574.7 million in 2010 (AgMRC 2013). All other global avocado exports (mainly to

Canada, Japan, Mexico, and South Korea) were valued at $4.5 million in 2009 with a drastic increase to $31.3 million in 2010 (AgMRC 2013).

Fruit Development and Composition

The avocado fruit is botanically a berry consisting of one seed surrounded by pulp. The fruit is variable in size, shape (round, oval, pyriform), peel, flesh, and seed characteristics (Scora et al. 2002). Fruit tissue is comprised of the rind or exocarp, the edible flesh/pulp is the mesocarp, and a thin layer next to the seed coat called the endocarp (Scora et al. 2002). Vacuoles of the mesocarp tissue cells contain oil droplets as well as larger, specialized oil-containing cells known as ideoblasts (Scora et al.

2002). Unlike most other fruits, in which most cell division occurs in early development followed by further growth only through cell enlargement, avocado exhibits continuous cell division and enlargement while the fruit is on the tree (Lee & Young 1983). There are three stages of the sigmoidal growth. During the first stage, the fruit grows slowly and 90% of initially set fruit abscise. Fruit growth is rapid during the second stage in which 80% of total fruit weight (mesocarp tissue) is amassed. During this stage, carbohydrates, proteins, and fatty acids (especially oleic acid) are accumulated in the mesocarp tissue (Salazar-García et al. 2013). Physiological maturity occurs during the third stage when the seed reaches full development (when anatomically isolated from the flesh) and cell division decreases (Lee & Young 1983; Scora et al. 2002). Fruit does

not ripen while on the tree, but begins to become ethylene sensitive, and will ripen after picking (Schroeder 1953). This may be explained by the unusual accumulation of C7 sugars during fruit development (Liu et al. 1999). Studies have shown that the C7 sugar

D-manno-heptulose and its sugar alcohol perseitol might be responsible for continued sugar uptake while inhibiting respiration and protecting against damage by reactive oxygen species (Cowan 2004; Bertling & Bower 2006). It is thought that the C7 sugars, paired with other biochemical changes and hormone concentrations, play a role in preventing ripening of fruit while still on the tree (Liu et al. 1999; Liu et al. 2002; Salazar-

García et al. 2013). Avocados ripen best (fruit softening occurs more evenly with a lower incidence of rot) if fruit is stored at <15˚C for one to two weeks and then ripened at 20˚C. Avocado is a climacteric fruit and once off the tree, ethylene production

-1 -1 increases to more than 100 µL C2H4 kg h when ripened at 20˚C (Yahia & Woolf

2011). Fruit in commercial market channels are often treated with 1-MCP (1- methylcyclopropene). This chemical is an ethylene inhibitor that binds to ethylene receptors; this reduces subsequent ethylene action and delays, but does not prevent ripening. The use of 1-MCP allows for fruit to be stored for longer periods of time and reduce chilling injury symptoms postharvest (Pesis et al. 2002; Scora et al. 2002;

Hershkovitz et al. 2005).

Moisture content of the avocado mesocarp is inversely proportional to oil content

(Slater et al. 1975; Du Plessis 1979). Studies by Du Plessis (1979) have shown that oil content increases during fruit development in parallel with dry matter increase. Percent dry matter is therefore used as a maturity index in California (Yahia & Woolf 2011).

Cultivars that are grown in Florida, do not reach oil contents as high as those in

California, and maturity indices there are based on picking dates, minimum fruit weights and diameters (Hatton & Campbell 1959; Hatton et al. 1964).

Laurel Wilt Disease

The Pathogen

Widespread mortality of redbay (Persea borbonia (L.) Spreng.) was first observed along the Savannah River and Sea Islands of southeastern South Carolina and around Savannah, GA, USA in 2003. Affected stems and branches presented symptoms including streaks of black discoloration in the sapwood with beetle bore holes on stems and branches of affected trees. Species of ambrosia beetles that were found in symptomatic trees included Xyleborinus gracilis Eichhoff and Ambrosiodmus obliquus

LeConte, which are native to the southeastern USA, and Xyleborus glabratus Eichhoff

(Coleoptera: Curculionidae: Scolytinae), which is native to Southeast Asia (Fraedrich et al. 2008). Eventually, X. glabratus was shown to transmit a lethal pathogen, the fungus

Raffaelea lauricola (Figure 1-1), which caused laurel wilt and the noted mortality of redbay (Fraedrich et al. 2008; Mayfield III et al. 2008a; Hanula et al. 2008; Kendra et al.

2013a).

Laurel wilt is a vascular disease that affects many plants in the family Lauraceae in the USA (Mayfield III et al. 2008a; Hanula et al. 2008). The fungus grows in beetle- excavated galleries and adjacent sapwood of host trees, leading to the disruption of water and nutrient flows. Wilt symptoms are associated with the production of gels and tyloses in infected trees (Inch & Ploetz 2012). In the USA, it is most devastating on native hosts such as northern spicebush (Lindera benzoin L.), redbay, sassafras

(Sassafras albidum (Nutt.) Nees), silkbay (Persea humilis Nash) and swampbay

(Persea palustris (Raf.) Sarg.) (Fraedrich et al. 2008).

Laurel wilt disease has had a deleterious ecological impact in North America, threatening plant communities and fauna associated with members of the Lauraceae family. As the native Lauraceae are lost, the availability of their fruits, nectar, and faunal species that depend on them may be directly and indirectly affected by this disease. For example, larvae of the palamedes swallowtail butterfly, Papilio palamedes Drury, only feed on lauraceous hosts and the spicebush swallowtail, Papilio troilus L., use sassafras and northern spicebush as preferred host plants (Gramling 2010). Laurel wilt may cause changes in habitat by altering light availability, increased woody debris, changes in soil, and other factors. Similar cascades of events have been caused by other exotic tree diseases such as chestnut blight and Dutch elm disease (Gramling 2010). Furthermore, the distribution of non-native, but economically important species such as camphor tree and avocado may serve as “bridging” or “corridor species” between native host plant habitats, enhancing disease spread to hosts which have discontinuous distributions

(Gramling 2010).

R. lauricola was first reported in avocado (Persea americana Miller) on a 10-year old dooryard tree in Jacksonville Florida in September 2007. The tree manifested foliar wilt symptoms (Figure 1-2) and extensive vascular staining with evidence of burrowing beetle holes. The pathogen was isolated from the discolored sapwood samples and small subunit (18S) sequences from the rDNA were amplified by PCR and sequenced.

BLAST nucleotide searches revealed a 100% homology with a Raffaelea sp., which was described later as R. lauricola (Mayfield III et al. 2008a). R. lauricola is isolated with a semi-selective medium that is used for related anamorphs of the genus Ophiostoma

and contains cycloheximide, which inhibits the growth of most fungi, but generally not those in the Ophiostomatales (Harrington 1981).

The taxonomy of Raffaelea spp. is poorly defined (Dreaden et al. 2014a;

Dreaden et al. 2014b), and a better understanding of the phylogeny of the genus is needed to help diagnosticians, facilitate quarantine efforts, and understand the epidemiology of diseases caused by this phytopathogen. Most Raffaelea spp. live as saprophytes; but, R. lauricola, R. quercivora, and R. quercus-mongolicae affect economically and ecologically important members of the family Lauraceae (laurel wilt) and Quercus spp. (Japanese and Korean oak wilt) (Ploetz et al. 2013). Laurel wilt disease is now established in the southeastern United States and has the potential to reach the U.S. Pacific coast, Mexico, and Central and South America, threatening their avocado industries as well as other Lauraceae. Studies have shown that the California bay laurel (Umbellularia californica Hook. & Arn. (Nutt.)) as well as other Lauraceous hosts are susceptible to laurel wilt, hence posing a major threat to the Pacific coastal ecosystem (Kendra et al. 2013a; Gramling 2010). Public awareness is needed on the risks of transport and movement of infested wood as well as the implementation of adequate quarantine protocols.

Multigene genealogies have demonstrated that Raffaelea is polyphyletic, and that the currently described species fall in two clades (Dreaden et al. 2014a; Dreaden et al. 2014b). The Ophiostoma clade includes R. lauricola, R. brunnea, and one undescribed species of the genus from Canada, while Raffaelea spp. such as R. quercivora, R. montetyi, R. sulphurea, and R. amasae fall into the Leptographium clade

(Dreaden et al. 2014a; Dreaden et al. 2014b). There are still undescribed Raffaelea taxa that need to be properly classified.

PCR amplification of small subunit (SSU; 18S) or large subunit (LSU; 28S) sequences from the rDNA have been used to detect R. lauricola and diagnose laurel wilt (Mayfield III et al. 2008a; Harrington et al. 2011; Jeyaprakash et al. 2014). However, there are pitfalls when pathogens reside in poorly defined genera such as Raffaelea. In one case, the SSU method gave a false positive for a dead avocado tree that was examined in 2009, identifying an isolate, PL1004, as R. lauricola. PL1004 was later shown to be non-pathogenic on avocado and was recently shown to be a new species

(Dreaden et al. 2014a; Dreaden et al. 2014b). A detection method developed by

Jeyaprakash et al. (2014), which uses a section of the LSU, also fails to distinguish

PL1004 from R. lauricola. Although the SSU and LSU sequences are not R. lauricola- specific and, thus, cannot be used conclusively for diagnostic purposes, the SSU amplicon is useful in experimental situations (e.g. after artificial inoculations) since it can detect ca. 0.0001 ng of the pathogen DNA. Recently, Dreaden et al. (Dreaden et al.

2014a) developed two small sequence repeat (SSR or microsatellite) markers for R. lauricola (microsatellites in fungi have limited intertaxon transferability as individual alleles are very uncommon, vary in length and exhibit less polymorphism than in other organisms (Dutech et al. 2007; Cristancho & Escobar 2008). The SSR markers of

Dreaden et al. (Dreaden et al. 2014a) are R. lauricola-specific and can be used to identify the pathogen in culture, but have a detection limit of only 0.1 ng of pathogen

DNA. Four labs tested and confirmed the consistency of the methodology, which is now currently used to diagnose laurel wilt/identify R. lauricola (Dreaden et al. 2014a). In

summary, the SSU markers are not taxon specific, but are highly sensitive, whereas the microsatellite markers are unable to detect low titers of the pathogen, but are taxon- specific and both are used in the diagnosis of the pathogen.

Host colonization and disease development has been studied in avocado, swampbay, and camphor tree with a GFP-labeled strain of R. lauricola (Campbell

2014). Little colonization was observed, even in severely affected trees (<1.6%).

Nonetheless, resistance to laurel wilt was associated with lower levels of xylem colonization, assayed as the percentage of colonized xylem in a given cross-section of the stem (Campbell 2014).

R. lauricola is a clonal pathogen in the USA (putative founder effect of a single strain) (Hughes 2014). Isolates from the USA share strong homology with strains from

Taiwan and Japan, thus supporting the hypothesis that R. lauricola in the USA originated from Asia (Harrington et al. 2011). Likewise, R. fusca and R. subfusca were also isolated from X. glabratus from Taiwan, suggesting that these Raffaelea species were also introduced into the USA from Asia (Harrington et al. 2011).

The Vector(s)

In the order Coleoptera, there are 90 families in the infraorder Curcujiformia of which Curculionidae contains the subfamilies Platypodinae and Scolytinae that include the bark and ambrosia beetles. There are about 3,400 known species of ambrosia beetles (Ploetz et al. 2013). As bark beetles colonize phloem, ambrosia beetles have evolved to colonize the xylem tissue of woody plants, and usually have symbiotic relationships with fungi. Available evidence indicates that the redbay ambrosia beetle,

Xyleborus glabratus Eichhoff (Coleoptera: Curculionidae: Scolytinae) in the tribe

Xyleborini, was introduced with its fungal symbiont, R. lauricola, into Georgia around

2002 in infested packing material, such as wooden crates or pallets (Thomas 2007;

Rabaglia et al. 2006). It has since spread north in southeastern North Carolina in 2011 and west into a few counties of Mississippi and Louisiana in 2009, 2013, and 2014.

During the spring of 2005, the beetle was detected in Duval County, Florida and has since spread along the east coast of Florida reaching Indian River County in 2006,

Brevard County in 2007, Okeechobee and Osceola counties in 2008, and Miami Dade

County in 2011 (Ploetz et al. 2011a).

Females of X. glabratus and of other ambrosia beetles carry fungal spores in specialized mandibular structures called mycangia. Males are haploid, smaller than females and cannot fly, whereas females are diploid, strong fliers, and are responsible for dispersion of R. lauricola (Fraedrich et al. 2008; Hanula et al. 2008). Females bore into the xylem of host trees to create galleries that they inoculate with R. lauricola; adults and larvae then feed on R. lauricola that proliferates in galleries (Figure 1-3). X. glabratus has haplo-diploid sex determination and females are able to reproduce without mating. This allows isolated females to locate a new host and lay unfertilized eggs that will give rise to males. The parental female can then mate and produce diploid eggs that will hatch into females able to disperse to colonize new hosts (Kendra et al.

2013a). The fungal diet of X. glabratus enables it to colonize the xylem of host trees, which is nutrient-poor but provides a protected habitat for brood development. Studies have shown that flight patterns are species specific for the Scolytinae (Brar et al. 2012;

Kendra et al. 2012). Xyleborus spp. generally initiates flight at about one hour before sunset. However, X. glabratus was observed to start flight a few hours earlier, which suggests the use of visual cues for host location (Kendra et al. 2012). This can help

improve its detection and develop attractant traps for pest control. Studies show that

Scolytinae beetles fly close to the ground with the highest number of X. glabratus captured at 35-100 cm above the ground (Brar et al. 2012). Trap captures in Florida revealed peak activities for X. glabratus in September-October and February-April.

Small peaks were observed in May-August and December-January which may be influenced by degradation of the manuka oil lure used that is known to last only two weeks, as well as rainfall, and lower temperatures in the winter (Brar et al. 2012).

The ecological niche of ambrosia beetles is typically stressed, dying or dead trees in which the insects can propagate. Thus, interactions of ambrosia beetles with

“healthy” trees have been viewed as atypical and a probable indication that the host trees were stressed by drought, flooding, freeze damage, wind damage, and poor cultural practices (Ranger et al. 2010), or biotic stresses such as Phytophthora root rot

(cause by P. cinnamomi) (Ploetz et al. 2012; Dann et al. 2013).

As laurel wilt spread in the USA, it was noted that X. glabratus attacked both healthy trees and those that were already affected by laurel wilt disease. To explain the supposed atypical interaction of X. glabratus with healthy hosts, several hypotheses were proposed. Hulcr and Dunn (2011) suggested that an “olfactory mismatch” occurred in which plant volatiles that are specific to stressed trees in the native habitat of X. glabratus were produced by healthy trees in the new regions. A “permissive choice hypothesis” proposed that selection in the beetle’s native environment ensured that they were attracted to stressed or dying trees rather than healthy trees with greater natural plant defenses (Hulcr & Dunn 2011). However, in new ranges or habitats natural selection resulted in beetles that were attracted to healthy trees, enabling them to avoid

competition with native beetles. An alternative explanation for these relationships may be that selection pressure for susceptibility in the home range of R. lauricola culled all hosts except those that had useful resistance. Ploetz et al. (2013) suggested that ambrosia beetles as a group may actually interact with healthy trees more often than the above hypotheses suggest. They noted that the only reason X. glabratus was known to interact with healthy trees was due to its dissemination of a lethal pathogen.

Experimental results for other ambrosia beetle species transmitting R. lauricola were published recently (Carrillo et al. 2014). Clearly, better understandings are needed for how and when ambrosia beetles interact with healthy trees.

Studies on plant and fungal volatiles have sought to identify attractants to help deter or capture the redbay ambrosia beetle (Kendra et al. 2011; Kendra et al. 2013b;

Kendra et al. 2014a; Hulcr et al. 2011). Phoebe oil lures were very efficient in attracting

X. glabratus, but are no longer available due to the vulnerability of overharvesting and the scarcity of phoebe oil trees in their natural habitat in Brazil (Hanula et al. 2013).

Manuka oil lures are used instead, but research shows that these lures only last up to two weeks in the field (Kendra et al. 2014b). Evaluation of seven essential oils has shown that the greatest number of redbay ambrosia beetles was captured using cubeb, manuka, and phoebe oils and that the addition of ethanol as a potential synergist had no effect on the number of beetles captured (Kendra et al. 2014b). Studies have shown that plant volatiles such as α-copaene, β-caryophyllene, and α-humulene have been positively correlated with field captures of X. glabratus. Most recently, four sesquiterpenes were confirmed to attract beetles: α-copaene, α-humulene, α-cubebene, and calamenene. Alpha-cubebene and α-copaene are the two major components in

cubeb oil and were found to be the major attractants in susceptible hosts with α- cubebene being the stronger attractant of the two. Further studies confirm that cubeb lures are currently the best attractants for X. glabratus detection and can last for at least eight weeks in the field (Kendra et al. 2011; Kendra et al. 2013b; Kendra et al. 2014a). It is likely that females are attracted to multiple volatile compounds emitted by the

Lauraceae, not a single kairomone. However, this research did not examine the female beetles’ specific preferences among the three known horticultural races of avocado

(races discussed below), despite the different chemical profiles produced by each avocado race as determined by gas chromatography-mass spectrometry (GC-MS) analysis (Kendra et al. 2011). Beetles have shown a preferential attraction to lychee,

Litchi chinensis Sonnerat, which is not susceptible to laurel wilt, but more females bore into avocado wood over a longer period of time (Kendra et al. 2013b). This suggests that beetles are initially more strongly attracted to lychee wood volatiles, but find it to be an unsatisfactory substrate to grow their fungal symbiont and hence relocate to look for a more suitable long-term host (Kendra et al. 2011). Ambrosia beetles are also attracted to volatiles produced by their corresponding fungal symbionts, which may enable their orientation within a gallery or location of established fungal gardens of conspecific beetles (Hulcr et al. 2011). These findings may ultimately help engineer species-specific lures for beetle and disease control.

There is usually a high level of specificity between ambrosia beetles and their fungal symbionts. Recently, the lateral transfer of R. lauricola to, and its dissemination by, six other ambrosia beetle species was reported (Carrillo et al. 2014). Xyleborus affinis, X. ferrugineus, X. volvulus and Xyleborinus gracilis are endemic to tropical

America and the southeastern USA, whereas Xylosandrus crassiusculus Motschulsky and Xyleborinus saxeseni Ratzeburg are non-native beetles that have established in the

USA. Capacity of several beetle species to carry this pathogen could conceivably enable an expanded host range for laurel wilt, as the other species have wider host ranges than X. glabratus. More importantly, the ability of these beetles to transmit R. lauricola to avocado and/or redbay (Carrillo et al. 2014), and the current absence of X. glabratus in laurel wilt-affected avocado groves in Miami-Dade County (Carrillo et al. unpublished data) indicate that the other species may play a role in the epidemiology of this disease on this crop.

Host Trees

To date, laurel wilt disease has not been reported in the native range (Taiwan,

Japan, India, and Myanmar) of X. glabratus (Harrington et al. 2011). It has been hypothesized that Asian members of the family coevolved with the pathogen, resulting in varying levels of resistance, while North American hosts have not (Ploetz et al. 2013).

For example, the camphor tree (Cinnamomum camphora L.) is a Lauraceae species of

Asian origin that typically recovers after infection (Fraedrich et al. 2015; Smith et al.

2009). Although trees may have coevolved an accommodating response to these fungi in Asia, the host features that are responsible for susceptibility in some trees in the

Western Hemisphere are not clear. Nonetheless, avocado does respond rapidly to artificial inoculation with R. lauricola. Tyloses and gels were induced in the xylem shortly after inoculation (Inch et al. 2012), which was rapidly colonized by the pathogen (Ploetz et al. 2012). Yet, there was surprisingly little histological evidence of the pathogen in these trees. Xylem function and xylem conductivity were significantly correlated and reduced dramatically after infection (Inch & Ploetz 2012; Inch et al. 2012). Trees with a

larger stem diameter develop more severe and rapid wilting symptoms compared with smaller diameter trees (Ploetz et al. 2012). Other studies indicate that as few as 100 conidia can kill an entire tree (Hughes et al. 2015). More information is needed on the features of resistant host species.

Redbay is found along the Atlantic coast from southern Virginia to southern

Florida and west along the Gulf of Mexico to eastern Texas. Its concentrations are found in southern Georgia and the Albemarle Peninsula of eastern North Carolina (Koch

& Smith 2008). Other Lauraceous hosts such as sassafras have much wider ranges, with low densities in the southeastern United States, Michigan, and north into New

Hampshire. Sassafras’s highest concentrations are found in eastern Oklahoma to Ohio and West Virginia (Koch & Smith 2008). The broad range of redbay and other native host trees throughout Florida facilitated the spread of R. lauricola and X. glabratus to the commercial avocado production area in Miami-Dade County, where it was detected in early 2011 (Ploetz et al. 2011a). Now that it is established in South Florida, laurel wilt is a major threat to commercial avocado production and the valuable collections in the

USDA-ARS national avocado germplasm repository in Miami. Cuttings of the collection are currently in quarantine in Beltsville, Maryland so that is can be moved to a safe location in Hawaii.

Avocado is a high-value specialty crop in the US, grown commercially primarily in

California and Florida. The Florida avocado industry contributes nearly $30 million to the local economy, with production of 31,100 tons and with more than 6,773 production acres in Miami-Dade County alone. A production and marketing report illustrated how direct cost estimates are broken down into potential sale losses, decreased property

values, and increased management costs as a consequence of laurel wilt disease. It suggested that the entire industry could be destroyed and agricultural property values could lose one-half of their market value if the disease continues to spread (Evans et al.

2010). As a consequence, avocado imports and/or domestic production by other avocado-producing states would need to increase to meet current demand in the United

States. California is the major producer of avocado in the United States with a production of 195,000 tons in 2012 (NASS). California laurel Umbellularia californica

(Hook. & Arn.) Nutt. is susceptible and could help spread the disease within California if introduced into that state (Gramling 2010). Other avocado-producing countries such as

Mexico and Chile could have devastating losses from laurel wilt disease. The immediate threat in South Florida and the possibility of spread to other states has made identification and implementation of control measures a high priority.

Avocado is an evergreen subtropical fruit tree of neotropical origin in the family

Lauraceae. This species is characterized by three botanical races that originated in

Guatemala, Mexico and Central America (Schaffer & Wolstenholme 2013; Litz et al.

2005). Scion cultivars used in California include ‘Bacon’, ‘Gwen’, ‘Reed’, ‘Zutano’,

‘Pinkerton’, ‘Hass’, and ‘Fuerte’, all of which are derived from the Mexican and/or

Guatemalan races. In Florida, commercial cultivars are all Antillean (i.e., West Indian) or

Antillean x Guatemalan hybrids such as ‘Lula’, ‘Booth 8’, ‘Waldin’, ‘Simmonds’, ‘Donnie’ and ‘Choquette’ (Litz et al. 2005; Ploetz et al. 1994). The three avocado races are easily distinguished from each other in that Mexican types are semi-tropical, usually more tolerant to colder environments and are smaller trees with anise-scented leaves. Fruits of Mexican cultivars have thin, smooth, and dark skin and take up to 6 months to reach

maturity. Guatemalan varieties are subtropical, intermediate in cold tolerance and are able to grow at high attitudes (900 m to 2,400 m). In contrast, West Indian cultivars prefer tropical environments and generally are more cold-sensitive. Ploetz et al. (2012) reported differences in the severity of laurel wilt disease development across different avocado cultivars, with West Indian cultivars being more susceptible than those with

Mexican or Guatemalan backgrounds.

On avocado, external laurel wilt symptoms appear as wilting of the terminal leaves that rapidly change color from dark green to brown right after wilting occurs

(Figure 1-2). Unlike redbay, in which leaves do not detach from the tree for a year or longer, avocado can defoliate in as little as 2 to 3 months from the first symptom development (Ploetz et al. 2011b). This may be due to the higher susceptibility of redbay that leads to a faster disease development and inadequate time for leaf abscission zones to develop. Internal symptoms develop faster than external symptoms and when wilting of leaves is observed internal symptoms are advanced. Sapwood turns to a reddish brown to blue-grey with streaks (Ploetz et al. 2012), resembling symptoms of Dutch elm disease.

Management

Currently, laurel wilt challenges the Florida industry and threatens those in

Texas, California, Mexico and the Caribbean. The following integrated pest management program has been recommended wherein early detection is based on growers’ visual scouting: suspect wood samples are collected and taken to county and university diagnostic labs for confirmation of R. lauricola; positive trees are then promptly removed, chipped and sprayed with insecticide containing permethrin

(Permethrin 3.2 AG; Arysta LifeScience North America, Cary NC or Permethrin 3.2 EC;

Helena Chemical, Collierville, TN) (Evans et al. 2010); and adjacent, surrounding

“healthy” trees are treated with Tilt, propiconazole (Syngenta Crop Protection LLC,

Greensboro, NC, USA) to impede root graft transmission of the pathogen (primary means of disease spread in affected orchards) (Ploetz personal communication). Wood that is too large to chip should be burned. Studies show that disease development was prevented when Tilt fungicide treatment occurred before inoculation compared to treatment after inoculation (Ploetz et al. 2011b). This may be due to the fact that tyloses may prevent systemic movement of the fungicide up the xylem vessels. The fungicide is delivered via macro-infusion, which is a most effective application measure but is slow and expensive and not commercially viable for treating entire groves. Current research is being conducted on the effectiveness and residual lifespan of additional fungicides within the tree. Although thiabendazole has been shown to have a longer lifespan than propiconazole and is currently being used to protect against Dutch elm disease, it is ineffective against laurel wilt (Ploetz et al. 2011b). Studies to develop effective delivery methods for Propiconazole Pro, tebuconazole, and Tilt are underway (Ploetz unpublished). Unfortunately, it is impractical to use quarantine practices to limit spread of the beetle as it is already well established along the Southeastern United States. It is however strongly advised to not move or sell redbay and other host trees as firewood to minimize the spread of the beetle and pathogen to unaffected areas.

Extensive laurel wilt infection of native redbay and swampbay in Merritt Island,

FL has been a concern for the small avocado groves in the area, which is far north of the primary avocado production area. It was observed that avocado groves in this area were only randomly attacked compared to redbay stands over a 3-4 year period.

However, this could change once the redbay population declines and beetles scout for alternative hosts. Another explanation could be that X. glabratus may play a much more limited role in the spread of laurel wilt to avocado than originally thought and alternative vectors needed for transmission to avocado are not present in the area (Crane et al.

2011).

Recently, entomopathogenic fungi were tested against the redbay ambrosia beetle. Two commercial strains of Isaria fumosorosea and one strain of Beauveria bassiana were found to effectively kill female beetles in galleries, thus preventing beetle reproduction and suppressing the establishment of their fungal symbiont in galleries

(Carrillo et al. 2015). It was shown that median survivorship times of female beetles ranged from as little as 3 days for B. bassiana to 5 days for I. fumosorosea strains.

These biocontrol strains were not tested for effects on R. lauricola establishment and the potential for disease transmission via root graft needs to also be considered.

Vector and inoculum management strategies as well as fungicide applications are critical to protect established plantings, but use of resistant avocado cultivars would provide the most sustainable long-term solution. Screening of the germplasm by artificial inoculation of the R. lauricola pathogen will aid in identifying tolerance or resistance.

Seedlings from diverse avocado parents are being subjected to field assessments (Pisani et al. unpublished). Protocols to facilitate higher throughput screening are also in development. Ideally, once truly resistant materials are available,

Simple Sequence Repeat (SSR) and Single Nucleotide Polymorphism (SNP) markers could help identify tolerance or resistance genes to the disease more efficiently and

accurately. It is hoped that promising laurel wilt resistant selections with desirable horticultural traits and resistance to priority diseases for commercial production will be identified.

Since avocado consumption is increasing and a large proportion of avocados eaten in the United States are currently imported, expanded U.S. avocado production has significant potential. However, laurel wilt resistance will be a key factor in the sustainability of this crop. Without prompt identification and disposal of affected trees and fungicide treatment of vulnerable adjacent trees, it is currently impossible to impede the disease’s spread in affected orchards (Kendra et al. 2013a). Substantial research is needed on the disease’s epidemiology and improved measures for its management.

Figure 1-1. Raffaelea lauricola extracted from a cutting of ‘Simmonds’ avocado through plating onto the selective medium CSMA+. (photo courtesy of Cristina Pisani)

Figure 1-2. Foliar symptoms of laurel wilt disease on avocado at the USDA Horticultural Research Laboratory Picos Farm, Ft. Pierce. (photo courtesy of Cristina Pisani)

A B Figure 1-3. Tree bolt showing Xyleborus glabratus frass tubes (A) and galleries (B) on redbay, Persea borbonia. (photo courtesy of Jiri Hulcr)

CHAPTER 2 FLOWERING OF HYBRIDS FROM RECIPROCAL CROSSES OF ‘HASS’ AND ‘BACON’ AS AN INDICATION OF ADAPTATION TO CENTRAL FLORIDA CONDITIONS

Introduction

The avocado tree is heterozygous, with a long juvenile period that can exceed 15 years (Litz et al. 2005; Salazar-García et al. 2013). Once in the adult phase, trees can produce 100,000 to more than 1,000,000 flowers but also experience a high rate of subsequent flower abscission and fruit drop (Bergh 1986; Litz et al. 2005; Salazar-

García et al. 2013). Flowers are protogynous and dichogamous; pistils mature before stamens to promote out-crossing and limit self-pollination. Each flower opens twice during two consecutive days and cultivars are classified based on whether the flowers open as functionally female in the morning (Type A) or in the afternoon (Type B).

Therefore, Type A cultivars (‘Hass’) have flowers that open as functionally female (stage

I) (Figure 2-1) in the morning, close midday, and then open a second time as functionally male (stage II) (Figure 2-2) in the afternoon of the following day to ensure that self-pollination cannot occur (Davenport 1986; Litz et al. 2005). The opposite is true for type B cultivars (‘Bacon’). During the female stage, the stigma becomes receptive and the stamens are bent outward with anthers not dehisced. During the male stage, the stigma is no longer receptive and the anthers become dehisced on upright stamens

(Davenport 1986). Studies have shown that at temperatures between 12 (53.6F) to

17˚C (62.6F) most flowers open in the male stage and most successful fruit set occurs between 20 (68F) to 25˚C (77F) when female and male stages overlap (Sedgley 1977;

Lovatt 1990). Flowering and fruit set are affected by low temperatures. Cool temperatures reduce the viability of the ovule and the time for the pollen tube to grow

from the stigma to the ovule is increased. This makes for shorter periods of pollination and effective fruit set (Lovatt 1990). Type B cultivars are less productive than type A cultivars under cool conditions, with the exception of ‘Bacon’ that is a type B cultivar with good yield in lower temperatures (Sedgley & Grant 1982/83). Floral cycling, pollen tube growth, and embryo development in ‘Hass’, a type A cultivar, was most favorable at 25˚C during the day and 20˚C during the night with increased fruitlet drop at higher temperatures and slow fruit development at lower temperatures (Sedgley & Annells

1981).

The chromosome number in avocado is 2n = 2x = 24 and major cultivars are derived from open pollinated seedlings (Litz et al. 2005). The species Persea americana has evolved into three subspecies/races that comprise the common avocado: P. americana var. drymifolia (Mexican race), P. americana var. guatemalensis

(Guatemalan race), and P. americana var. americana (West Indian race) (Bergh 1986).

There are no genetic barriers among the three subspecies and many cultivars today are hybrids of two or more subspecies. California cultivars ‘Bacon’ and ‘Hass’ are cultivars considered to be Guatemalan-Mexican hybrids while the Florida avocado industry is based on pure West Indian and Guatemalan-West Indian hybrids (Litz et al. 2005). All three races exhibit similar growth and flower characteristics.

‘Bacon’ is a Type B flowering tree with medium to large ovate fruit, thin green peel and large seed. This cultivar is rather cold tolerant (can withstand frost down to

4.4˚C (39.9F) and production is suited for colder regions such as California, Australia, and Italy (Crane et al. 2013). This cultivar is extremely susceptible to anthracnose and is hence not suitable for humid subtropical areas like Florida. ‘Hass’ is the most

predominant cultivar on today’s market and accounts for more than 90% of the export market (Crane et al. 2013). This cultivar grows best in cool subtropical areas of New

Zealand, California, Chile, Peru, Mexico, Israel, Spain, Australia, and Colombia. ‘Hass’ is a type A flowering tree with small to medium size ovate fruit. The fruit peel is thick, leathery with a corky texture that helps this cultivar be more tolerant to insect pests and diseases (Crane et al. 2013). Unlike ‘Bacon’, a change in peel color from green to black aides in identifying ripeness and masks minor exterior imperfections (Crane et al. 2013).

Unfortunately, if grown is warmer climates, the fruit tends to be undersized (<200g)

(Crane et al. 2013).

Much work has been done on finding an acceptable maturity index for avocado.

Based on studies conducted by Hatton and Campbell in 1958-1961 on different Florida cultivars, it was found that the weight and diameter of individual fruit were correlated and also correlated with oil content. Additionally, palatability, determined using taste panels, increased as fruit matured and was highly correlated with oil content and fruit weight. Other indices that were tested, such as color, were found not to be commercially adequate. As a result, two sets of maturity standards have been developed and are used in the United States today. The California avocado industry bases their maturity standard on oil content, with a minimum of 8% oil based on fruit fresh weight. Actual picking dates vary with cultivar and climatic zone (Barmore 1976). Since most Florida cultivars have very low oil content that may never reach 8%, the Florida Avocado

Administrative Committee established the Florida avocado maturity index based on specific picking dates (days after full bloom) as well as minimum fruit weight and diameter, depending on variety (Barmore 1976).

The current study evaluated flowering and cropping of California avocado reciprocal crosses of ‘Bacon’ and ‘Hass’ as an indication of adaptation to conditions in east-central Florida (especially chilling requirements). Flowering and fruit set was monitored for three consecutive seasons (2013-2015) and was part of a larger study to identify California-type avocado hybrids that demonstrated good horticultural traits while growing under conditions in east-central Florida.

Materials and Methods

Trees originated from seeds collected from a commercial orchard in California by

Dr. Raymond Schnell (SHRS-ARS, Miami) and planted at USHL-ARS, Fort Pierce,

Florida. Molecular marker analysis confirmed which seedlings were hybrids before planting at the USDA-ARS Picos farm in Fort Pierce, Florida on August 29, 2008. Trees were on double row beds in Riviera fine sand soil type, and sprayed with horticultural oil and copper (CS-2005, Magna-Bon II, LLC, Okeechobee, FL), and received foliar fertilization with a 20:10:20 N:P:K soluble fertilizer every two weeks as part of a regular maintenance regimen. Soil applied Ridomil (Syngenta) and foliar applied Lexx-A-Phos

(Foliar Nutrients, Inc., Cairo, GA) for Phytophthora control was applied twice a year together with a granular dry fertilizer (12-2-14) at approximately 500-600 pounds/acre.

Flower assessments began on 4 March, 20 February, and 4 February in 2013,

2014, and 2015, respectively, and were conducted biweekly between 8:00 and 11:00 a.m. until fruit set. Due to the great number of flowers produced by trees, flowers were counted in cluster groups of 20 per terminal with estimation of open flowers around the whole tree canopy or on one side of tree canopy. Length of the flower shoot was also measured and rated on a 1 to 5 scale where 1 = <1 inches long, 2 = 1-2 inches, …, 5 =

>8 inches long. Flower clusters were rated on a 1 to 5 scale where 1 = <5 clusters on a

tree, 2 = 5-10 clusters, 3 = 10-20 clusters, 4 = >20 clusters (but not enough to cover the entire tree), and 5 = cluster on almost every shoot apex. General notes such as number of flowers per shoot, presence of fruit set, flower distribution around the tree, and presence of new flush were taken. Overall percentage of open flowers was also estimated to calculate peak bloom for each tree. Flowers were screened as either male or female in the morning hours to assess whether trees were of either A or B flowering types. Once flowering period ended and fruit began to develop, fruit growth was measured on a monthly basis by taking the average length and width measurements of three random fruits per tree with a caliper. Since avocado has a huge rate of fruit drop in

June, fruit yield was assessed before harvest on 29 October and 14 November in 2013,

28 October and 23 October in 2014 and 2015, respectively. As Florida uses different avocado fruit maturity indices than California, fruit were considered ready for harvest when they stopped growing, which is often when they began naturally dropping from the trees.

Results

The most important and widely distributed disease of avocado is Phytophthora root rot, Phytophthora cinnamomi. On avocado, the oomycete (fungal-like organism) causes a root rot and destroys feeder roots causing the tree to decline and eventually die if left untreated. Phosphonate fungicides paired with good horticultural practices and integrated management are used against Phytophthora root rot (Ploetz et al. 1994).

Unfortunately, the study block had a Phytophthora outbreak in 2014 and many trees were lost and others that declined did not fruit. Thus, fruit data were reduced in 2014 and 2015 compared to 2013.

During the study, the date of peak bloom varied greatly throughout the California mapping population. Frequency plots show how many trees had most of their flowers open at each given time (Figure 2-3; 2-4; 2-5; 2-6; 2-7; 2-8). In 2013, 143 trees that flowered were assessed and the greatest number of trees had most flowers open (max bloom) between March 27 and April 11 (Table 2-1; Figure 2-3; 2-6). Only one tree in

2013, R7T12, had a majority of its flowers open (80%) on a single date (Table 2-1). In

2014, 181 trees flowered and for most trees, peak bloom occurred between March 20 and April 16 (Table 2-2; Figure 2-4; 2-7). Again, only one tree, R7T46, had a high percentage (80%) of open flowers on a single date, April 3, 2014 (Table 2-2). In 2015,

153 trees flowered and peak bloom for most trees occurred on March 18 and April 1,

2015 (Table 2-3; Figure 2-5; 2-8). In 2015, the highest percentage of open flowers,

65%, was measured on R3T57 and R8T14 April 1, 2015 (Table 2-3). Frequency plots of major peak bloom dates show that most trees had their highest bloom percentages on

11 April 2013 (Figure 2-3; 2-6), on 16 April 2014 (Figure 2-4; 2-7), and on 1 April 2015

(Figure 2-5; 2-8). When looking at the mapping population as a whole, peak bloom for most trees occurred 7 weeks after bud break in 2013 and 9 weeks after budbreak in

2014 and 2015 (Figure 2-9). Flower numbers were assessed based on cluster presence on each tree. Most trees had flowers on almost every shoot during all three years

(Figure 2-10). In 2013, 33 trees had 10-20 flower clusters, 32 trees had more than 20 clusters, and 44 trees had clusters on almost every shoot (Figure 2-10A). In 2014 and

2015, 110 and 89 trees had flower clusters on almost every shoot, respectively (Figure

2-10A; 2-10B).

Beginning soon after fruit set, fruit size was measured on a monthly. Fruit stopped growing at the beginning of October in all three years (Figure 2-11). In

California, fruit are harvested at predetermined dates based on dry matter and size for each cultivar (Yahia & Woolf 2011). In Florida, avocados have assigned picking dates based on fruit size (Barmore 1976; Lee 1981). As our fruit consisted of California cultivar hybrids grown in a Florida climate, fruit growth was considered as an acceptable maturity index prior to harvest. As a result, harvests occurred on 29 October and 14

November in 2013, 28 October in 2014, and 23 October in 2015. Two harvests occurred in 2013 as there was too much fruit to handle at one give time.

Maximum and minimum daily average temperatures (January-December) were very similar throughout all three years (Figure 2-12). In 2013 the average daily temperatures ranged between 29˚C (84.1F) and 8.4˚C (47.1F), 29.2˚C (84.5F) and

7.8˚C (46.1F) in 2014, and 28.5˚C (83.3F) and 7˚C (44.6F) in 2015 (Figure 2-12). During the flowering period (February-May), temperatures ranged from 27.1˚C (80.7F) to

11.3˚C (52.3F) in 2013, 26.3˚C (79.4F) to 12.6˚C (54.7F) in 2014, and 24.7˚C (76.5F) to

7˚C (44.6F) in 2015 (Figure 2-12). The spring of 2015 was slightly colder than in the previous two years and the only observation was that flowering began slightly earlier in

February and ended in early April instead of early May. However, the lowest temperature during flower bud initiation (October-November) (Davenport 1986) was in

2014 at 11.2˚C (52.3F) (Figure 2-12). Most trees in 2015 reached a peak bloom in April

(Table 2-1), which was also true in 2013 (Table 2-2) and 2014 (Table 2-3). Most trees had reached peak bloom by April 11 in 2013 (Figure 2-3), by April 16 in 2014 (Figure 2-

4), and by April 1 in 2015 (Figure 2-5). However, in 2014, many trees had reached peak

bloom by March 20 (Table 2-2). Out of 187 total trees that flowered during 2014 and

2015, 96 were of type A and 91 were of type B (data not shown). Trees ranged from producing no fruits each year to as many as 225, 284, 212 fruits in 2013, 2014, and

2015, respectively (data not shown).

Discussion

In avocado, outcrossing is controlled by the protogynous and dichogamous nature of its flowers, with different cultivars being of type A or B. Having A and B cultivars in close proximity enables cross pollination. It has been observed that flower behavior is very sensitive to temperature and varies greatly among cultivars (Davenport

1986). Avocado flower bud induction occurs sometime after the last vegetative flush in the autumn and inflorescence buds are apparent (October-November) (Davenport

1986). Daylength and/or temperature play critical roles in flower bud induction

(Davenport 1986). Flower induction occurs when day temperatures do not exceed 25˚C

(77F), regardless of daylength. Low temperatures of 20˚C (68F) during the day and night temperatures between 5 and 15˚C have been shown to induce flowering, while daytime temperatures of 25-30˚C completely inhibited flower initiation (Buttrose &

Alexander 1978). Nonetheless, most selections flowered well under east-central Florida conditions. From 2013-2015, temperatures during flower bud induction (October 1-

November 30) never exceeded 27˚C, and the lowest temperatures were 13, 11, and

17˚C in 2013, 2014, and 2015, respectively. Overall, temperature ranges during the flower bud induction period (October-November) were 13-27˚C in 2013, 11-27˚C in

2014, and 17-27˚C in 2015. During peak bloom times (March-April), temperatures ranged between 11 and 25˚C, 12 and 26˚C, and 14 and 24˚C in 2013, 2014, and 2015, respectively.

During optimum conditions (25˚C day/20˚C night) (Sedgley & Grant 1982/83), the asynchronous opening and closing of male and female flowers on the same tree ensures little self-pollination. However, if cool temperatures arise during the flowering period, flower opening in type A cultivars may be delayed and reversed to mimic type B flowering. Type B cultivars are even more sensitive to low temperatures (Sedgley &

Grant 1982/83) and some varieties fail to open stage I flowers (Davenport 1986).

However, Sedgley and Grant (1982/83) observed ‘Bacon’ yielded well at lower temperatures (17˚C day/12˚C night). It was also observed that type A cultivars such as

‘Hass’, were less affected by higher temperatures and that for all cultivars, the midrange temperatures of 25˚C during the day and 20˚C during the night was optimal (Sedgley &

Grant 1982/83).

Flowering, pollination, and eventual fruit set for the ‘Hass’-‘Bacon’ hybrids in the present study occurred over a wide range of temperatures. From 2013 to 2015, temperatures during flower bud initiation (October-November) and peak bloom times

(March-April) were within the range for successful flowering and fruit set. Approximately equal numbers of A and B types occurred among the hybrids, and their bloom and yield traits were fairly consistent throughout the study.

Conclusion

Minimum and maximum temperatures did not vary much during the study, and they were similar to those recorded in the past ten years in the Fort Pierce area. Since

2005, average temperatures were always between 21-22˚C (70-73F) with average lows between 9-12˚C (49-54F) and average highs between 31-33˚C (88-91F) (data not shown). Years of this study can therefore be considered normal and not due to warm or cold years, and can be considered typical as temperatures averages seem to only vary

by a few degrees throughout the past ten years. Clearly, fertilization and fruit set were successful in the east-central Florida climate during the three year study.

Figure 2-1. Female flowers. (photo courtesy of Cristina Pisani)

Figure 2-2. Male flowers. (photo courtesy of Cristina Pisani)

Table 2-1. Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/27/2013 and 4/11/2013. 3/27/2013 4/11/2013

Flower shoot length, Flower shoot length,

Peak bloom % Peak bloom % Tree ID Tree ID R7T12 (5, 80) R3T51 (2, 75) R6T36 (4, 70) R8T51 R7T15 R4T46 R4T21 (5, 70) R4T20 R6T40 (3, 70) R3T23 (2, 70) R7T8 R6T38 (2, 70) R8T49 R7T6 R6T19 (5, 60) R8T8 R6T44 (4, 60) R6T50 R6T46 (4, 60) R3T57 (5, 50) R4T23 (1, 60) R3T52 R3T31 R3T30 (3, 50) R8T41 R6T34 R5T43 (5, 50) R3T18 (2, 50) R8T53 R8T38 R7T11 R6T52 (4, 50) R5T19 R7T19 R3T28 (2, 40) R6T22 R4T43 (3, 50) R8T23 R3T20 (5, 30) R6T24 R5T46 (2, 50) R8T36 (3, 30) R8T11 R6T42 R6T6 R5T13 (5, 40) R3T39 R8T15 (2, 30) R6T28 R6T23 R5T47 R5T40 (4, 40) R4T17 (3, 25) R8T24 R8T21 R7T21 R6T21 (3, 40) R3T22 R7T51 (2, 25) R7T45 (1, 40) R3T49 (1, 25) R5T41 R4T50 (5, 30) R6T9 (4, 20) R8T18 R7T34 R6T18 R5T51 (4, 30) R8T3 (3, 20) R8T1 R7T30 (3, 30) R8T48 R4T4 (2, 20) R5T35 (2, 30) R8T32 R7T39 (5, 15) R7T36 (3, 25) R7T13 (4, 15) R4T52 (5, 20) R7T35 (2, 15) R8T43 R6T29 R5T26 R5T16 (4, 20) R7T48 (5, 10) R8T16 R7T44 R6T49 R6T30 (3, 20) R5T38 R4T40 R5T25 R4T18 (3, 10) R7T27 (2, 20) R6T33 (1, 10) R5T21 R4T6 (1, 20) R7T54 (2, 5) R6T7 (4, 10) R8T19 (1, 5) R4T27 (3, 10) R8T26 R8T25 R7T5 (2, 10) R7T17 R6T27 R5T53 R5T29 (1, 10) R4T35 R4T3 R6T41 (1, 5)

Table 2-2. Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/20/2014, 4/3/2014, and 4/16/2014. 3/20/2014 4/3/2014 4/16/2014 Flower Flower Flower shoot shoot shoot length, length, length, Peak Peak Peak bloom Tree ID bloom % Tree ID bloom % Tree ID % R3T3 (4, 60) R7T46 (1, 80) R3T39 (5, 75) R8T49 R8T38 R6T42 (5, 50) R4T46 (5, 50) R4T18 (5, 70) R3T57 R8T51 R7T47 R6T52 (5, 40) R7T37 (5, 45) R3T49 (5, 65) R6T19 R5T51 R6T40 R5T14 (4, 40) R7T10 (1, 40) R5T31 (5, 60) R8T56 R7T48 R6T18 (5, 35) R7T6 (5, 35) R3T56 (1, 60) R6T30 (4, 35) R7T32 (4, 30) R7T57 R6T28 R5T25 (5, 50) R4T50 R4T21 R3T23 R7T52 R7T45 R7T34 (5, 30) R8T26 R8T16 (5, 25) R5T32 (4, 50) R4T36 R4T43 R7T25 R7T13 R3T46 (4, 30) R8T20 (4, 25) R5T2 (3, 50) R3T26 (3, 30) R6T6 (5, 20) R8T35 (2, 50) R7T19 (4, 25) R8T18 R8T5 (4, 20) R4T8 (1, 50) R7T44 R7T15 R4T34 (5, 20) R8T45 (1, 20) R4T52 R3T16 (5, 45) R3T52 R8T12 R7T27 R7T7 (4, 20) R8T48 (4, 15) R3T51 (4, 45) R7T16 R4T40 (5, 15) R7T5 (3, 15) R5T13 R4T6 R3T30 (5, 40) R3T20 R8T30 R7T28 (4, 15) R8T9 R6T22 (5, 10) R6T29 (4, 40) R3T10 (3, 15) R6T16 (4, 10) R6T34 R5T40 R4T31 (5, 35) R7T35 (1, 15) R6T51 (3, 10) R5T38 R4T44 R4T9 (5, 30) R3T36 R8T25 R8T17 R7T17 (4, 10) R7T20 (4, 5) R4T3 (4, 30) R6T25 R7T41 (3, 10) R4T57 R4T45 (3, 30) R8T33 (2, 10) R8T21 R3T22 (5, 25) R8T24 R8T23 R7T18 (4, 5) R4T47 (4, 25) R7T40 (3, 5) R3T5 (2, 25) R7T39 (1, 5) R4T35 R3T54 (5, 20) R8T7 (2, 1) R6T27 R6T15 (4, 20) R8T37 (1, 1) R5T19 (5, 15) R6T44 R5T22 (4, 15) R5T10 R4T25 (2, 15) R5T16 R5T6 (5, 10) R5T4 R3T17 (4, 10) R7T21 R3T35 (3, 10) R7T36 R6T43 (1, 5) R8T22 R4T28 (3, 1)

Table 2-3. Panicle lengths and percentage of open flowers at peak bloom dates for individual hybrids assessed on 3/18/2015 and 4/1/2015. 3/18/2015 4/1/2015 Flower shoot Flower shoot

length, Peak length, Peak Tree ID Tree ID bloom % bloom % R8T37 (2, 30) R3T57 (5, 65) R7T43 R7T29 R5T19 (5, 25) R8T14 (1, 65) R7T31 (4, 25) R6T34 R5T56 R4T40 (5, 50) R8T13 (1, 25) R8T20 R6T10 (4, 45) R8T56 R8T38 R8T9 R7T37 (5, 20) R8T49 R8T26 R7T21 R4T21 (5, 40) R7T34 R7T13 R7T6 R6T46 R5T53 R5T14 R8T36 R8T33 R7T51 R7T45 (4, 20) R7T40 R3T36 (4, 40) R6T16 R8T30 R6T13 (2, 20) R4T44 R4T35 R3T39 (5, 35) R7T36 R7T15 R6T18 R5T31 (5, 15) R5T54 (3, 35) R5T2 R4T9 R3T49 R8T48 R8T45 R7T54 R7T10 (4, 15) R6T56 R6T29 R5T47 R5T38 (5, 30) R4T45 R3T10 R5T25 R5T29 R4T18 (5, 10) R7T17 R7T7 R5T40 R3T52 (4, 30) R3T26 R7T28 R4T8 (4, 10) R8T27 (3, 30) R3T34 (3, 10) R8T12 R8T7 R6T42 R5T43 (5, 25) R5T44 (2, 10) R8T34 R8T25 R4T47 R4T33 (4, 25) R4T46 (5, 5) R8T54 R4T4 (3, 25) R4T50 (3, 5) R8T35 R8T21 R8T18 R7T57 (5, 20) R7T48 R7T39 R7T19 R6T38 R6T30 R5T49 R4T2 R3T18 R7T46 (4, 3) R7T30 R6T43 R6T40 R5T13 (4, 20) R4T25 R4T3 R8T51 R8T22 R8T1 R7T35 (5, 15) R7T12 R6T36 R6T15 R5T48 R5T16 R3T20 R8T44 R4T57 R4T52 R4T31 (4, 15) R4T28 R3T22 R8T29 (3, 15) R7T1 R6T51 (2, 15) R4T17 R3T47 R3T16 (5, 10) R7T20 R5T6 (4, 10) R5T32 (3, 10) R8T11 R6T49 (2, 10) R5T8 (3, 5)

Figure 2-3. Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2013 season.

Figure 2-4. Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2014 season.

Figure 2-5. Frequency plots showing number of trees demonstrating a specific percentage full bloom at given dates during the 2015 season.

Figure 2-6. Frequency plot showing number of trees demonstrating max bloom at given dates during the 2013 season.

Figure 2-7. Frequency plot showing number of trees demonstrating max bloom at given dates during the 2014 season.

Figure 2-8. Frequency plot showing number of trees demonstrating max bloom at given dates during the 2015 season.

Figure 2-9. Average percentage of open flowers per tree per bi-week of the California mapping population of ‘Hass’-‘Bacon’ hybrids for the years 2013-2015.

C Figure 2-10. Frequency plots showing number of trees demonstrating max flower clusters during the 2013 (A), 2014 (B), and 2015 (C) seasons. Clusters were counted using a 1-5 rating scale where 1 = < 5 flower clusters per tree, 2 = 5- 10, 3 = 10-20, 4 = >20, 5 = flower cluster on almost every shoot.

C Figure 2-11. Average fruit length and width after fruit set measured monthly up until harvest for the years 2013 (A), 2014 (B), and 2015 (C).

Figure 2-12. FAWN (Florida Automated Weather Network) data for average temperature in a Fort Pierce & Saint Lucie West weather station for the years 2013-2015.

CHAPTER 3 POSTHARVEST EVALUATION OF HYBRIDS FROM RECIPROCAL CROSSES BETWEEN AVOCADO CVS ‘HASS’ AND ‘BACON’

Introduction

Avocado is a subtropical climacteric fruit from round to oblong in shape. As fruit ripen, skin may remain green or change from green to black or purple depending on cultivar. The ‘Hass’ cultivar is a black-skinned avocado when ripe and is the predominant commercial variety in most avocado producing countries such as United

States, Chile, and Mexico (Yahia & Woolf 2011; Crane et al. 2013). Its popularity is due to its precocity and regular, heavy crops as well as postharvest qualities such as its excellent storage and shipping ability and its change in peel color (from green to black) that can be used as an index for ripeness and masks minor peel imperfections (Crane et al. 2013). Avocado fruit possess a physiological mechanism to inhibit ethylene while on the tree and begin the ripening process only after detaching from the tree, hence, fruit can be ‘stored’ on the tree for up to 12 months depending on variety and environmental conditions (Schroeder 1953; Yahia & Woolf 2011). The mechanism that inhibits ethylene while on the tree is still unknown; however, studies have observed the unusual accumulation of C7 sugars during fruit development (Liu et al. 1999). In particular, the C7 sugar D-manno-heptulose and its sugar alcohol perseitol might be responsible for continued sugar uptake while inhibiting respiration and protecting against damage by reactive oxygen species (Cowan 2004; Bertling & Bower 2006). It is thought that the C7 sugar may help prevent ripening of fruit while still on the tree together with other biochemical changes and hormone concentrations (Liu et al. 1999;

Liu et al. 2002; Salazar-García et al. 2013). After harvest, the ethylene inhibition is lost within one or two days and fruit will start to ripen fast, however, the longer fruit stays on

the tree, the less time needed for fruit to ripen off the tree (Yahia & Woolf 2011). Once off the tree, there is a surge in ethylene production associated with ripening (Jeong et al. 2003). Avocado has a relatively high respiration rate during the climacteric phase of ripening compared to other fruit and can complete ripening in as little as five days after harvest (Seymour & Tucker 1993; Jeong et al. 2003).

Determining a good maturity index for avocado is problematic compared to other crops because maturity indicators such as softening and color change do not occur while on the tree. If fruit is harvested too early, it may lead to poor quality after ripening, such as rubbery texture, ‘stringy’ vascular tissue, bland watery flavors, and increased incidence of postharvest rots with off flavors (Yahia & Woolf 2011). Fruit may reach physiological maturity (i.e. will ripen after harvest), but may still be commercially unacceptable due to low dry matter and oil content and may exhibit uneven ripening, lack of flavor, with shorter storage potential (Yahia & Woolf 2011). Dry matter and oil content are highly correlated in avocados, regardless of genetic background (all three horticultural races) and are used as maturity indices in most avocado producing countries.

In the United States, California and Florida are the top avocado producing states, but grow avocados with different genetic backgrounds. In California, cultivars are of

Guatemalan and Mexican backgrounds and mature harvest requires dry matter content between 17 to 25% with an 8% minimum oil content (depending on cultivar) (Barmore

1976; Lee et al. 1983; Yahia & Woolf 2011). The California avocado industry has used dry matter content as its maturity index since 1983 and fruit is released into the market on specific dates, depending on the cultivar (Yahia & Woolf, 2011). The cultivar ‘Bacon’

requires a dry matter content of at least 17.7% and ‘Hass’ 20.8% (Yahia & Woolf 2011).

Florida avocados (West Indian or West Indian-Guatemalan hybrids) have much lower oil content and rarely reach the 8% absolute minimum required in California. The Florida

Avocado Administrative Committee therefore established maturity indices for Florida avocados based on days after full bloom as well as specific weights and diameters of fruit, depending on variety (Barmore 1976). Since the trees in the present study were hybrids of typical California cultivars (reciprocal crosses of ‘Hass’ and ‘Bacon’) in a

Florida climate, fruit was considered ready to be harvested when it stopped growing and began to drop from the tree, assuming that these indicated physiological maturity.

Postharvest rots are a key issue in avocado growing regions with hot moist environments, such as Florida. It is therefore, extremely important to follow good management practices before and after harvest. Avocado is very susceptible to stem end rots and body rots caused by a variety of fungal organisms. The most serious postharvest disease of avocado is anthracnose, a body rot, caused by the fungus

Colletotrichum gloeosporioides (its teleomorph Glomerella cingulata is thought to have a minor role in this disease) (Dann et al. 2013; White et al. 2009). Stem end rots are caused by several fungal species. In the United States, stem end rot is most commonly caused by Botryosphaeria dothidea, but Colletotrichum gloeosporioides alone or together with other fungi can also cause this rot. It is very hard to pinpoint which fungus or combination of fungi is causing the rot as the predominant pathogen varies with environmental conditions and local inoculum. The pathogens can colonize dying tissue associated with other diseases, such as scab, caused by Sphaceloma perseae, which is very common in humid growing regions, such as Florida (Menge & Ploetz 2003). A

combination of preharvest and postharvest fungicide treatments, good cultural practices, and resistant cultivars are some of the management strategies used for these problems.

Evaluation is underway for quality avocado selections well adapted to east- central Florida which experiences more frequent freeze events than the established south Florida avocado production zone. Extensive phenotypic data on fruit quality (fruit weight, seed weight, fruit size and diameter, oil content, number of fruits per tree, dry matter) was collected over the 2013-2015 growing seasons. Promising selections with potential as cultivars were identified with good horticultural traits and fruit quality.

Materials and Methods

Tree Descriptions

About 350 Mexican x Guatemalan hybrid trees have been planted in the

USDA-ARS station in Fort Pierce. These trees are a population of ‘Hass’ x ‘Bacon’ and

‘Bacon’ x ‘Hass’ for potential cultivar identification, and a detailed genetic mapping study using a newly developed SNP chip led by scientists at the Subtropical Horticulture

Research Station (SHRS) in Miami, Florida. Trees originated from seeds collected from a commercial orchard in California by Dr. Raymond Schnell working at SHRS-ARS,

Miami. Molecular marker analysis was used to determine which seedlings were truly hybrids before planting. Trees were planted at the USDA-ARS Picos farm in Fort Pierce,

Florida on August 29, 2008 on double row beds in Riviera fine sand soil type. A regular biweekly maintenance regimen includes a spray with horticultural oil and copper (CS-

2005, Magna-Bon II, LLC, Okeechobee, FL), and fertilization with a 20:10:20 N:P:K, together with a granular dry fertilizer (12 2 14) at approximately 500-600 pounds/acre.

Soil applied Ridomil (Syngenta) and foliar applied Lexx-A-Phos (Foliar Nutrients, Inc.,

Cairo, GA) were used twice per year for Phytophthora control.

Fruit (Preharvest evaluations)

Length and width measurements of three initially randomly selected fruit were followed monthly on each tree. Measurements were taken post bloom up until the harvest date to determine an appropriate harvest index for picking these California-type avocados. Fruit on the tree was counted right before harvest as a measure of yield for each individual tree. If fruit was observed to have recently dropped off the tree, it was counted and included in yield measurements. Sixteen fruits or less were harvested from each tree for non-destructive tests to evaluate ripening, peel color, and the development of postharvest diseases and disorders. An additional three fruits were picked for destructive tests to evaluate initial quality attributes such as pulp firmness, dry weight, and oil content. Harvest occurred on 31 October and 14 November 2013, 28 October

2014, and 23 October 2015. Two harvests occurred in 2013 due to the large amount of fruit to process at one given time. Fruit was stored at room temperature (approximately

24-25˚C) in plastic containers with lids. Approximately twelve 1mm holes were manually drilled in a line in the trough on the bottom of the container. Holes would help drain free moisture that ran to the bottom of the container and allow for gas exchange. Containers were vented and dried daily to avoid moisture buildup. Data on phenotypic characteristics (length, width, weight, flesh to seed ratio), peel color, whole fruit/pulp firmness, weight loss, dry weight, oil content, and postharvest rots/disorders was collected for the 2013, 2014, and 2015 seasons.

Postharvest Quality Parameters

Peel color

A Minolta Colorimeter (Konica Minolta Sensing, Inc., Japan) was used to assess color, just after harvest (unripe), and a second time after the fruit were commercially ripe

(reached a firmness of 20-30 N), but before internal assessments. Peel color was measured at the equatorial region (three readings per fruit). The Chroma Meter was calibrated with a white standard tile and the CIELAB values L* (lightness, where 0 = black, 100 = white), a* (bluish-green/red-purple hue component) and b* (yellow/blue hue component) were measured. The results are presented as the Lightness (L*), hue angle

(hº), and chroma value (C*, color saturation; degree of departure from gray toward pure chromatic color) with hue values of 90º representing a yellow color and 180º a green color. The chroma and hue were calculated from the measured a* and b* values using the formulas C*=(a*2+ b*2)1/2 and h=arc tangent (b*/a*) (McGuire 1992).

Whole fruit firmness

Firmness was determined by a non-destructive compression test on whole un- peeled fruit using a Texture Analysis system from Texture Technologies Corp. fitted with a flat-plate (5 cm diameter) and 50 kg load cell (0.5 kN). After establishing zero force contact between the probe and the horizontal region of the fruit, two measurements were taken per fruit and the fruit were rotated 90˚ between measurements. The probe compressed the fruit 2.5 mm with a crosshead speed of 20 mm/min-1.

The same fruits were measured every other day until fruits reached the fully ripe stage. Fruits were considered commercially ripe upon reaching 20-30 N firmness. Flesh bruising at the areas of compression was not evaluated during internal rot assessments.

Weight loss

Fruit were weighed every other day on the same day as whole fruit firmness measurements. The following equation was used to calculate weight loss on a fresh weight basis:

WL (%) = W0 - Wf x 100 W0

Where: WL is the weight loss given in percentage, Wo is the initial weight, and Wf is the final weight.

Results were also presented as percentage of fruit weight loss per day.

Fruit pulp firmness

Pulp firmness was determined by a destructive compression test on three unripe and ripe fruits. Cross-section slices of 1.5 cm were used from both stem end and blossom end for this analysis. The extremities (stem and blossom ends) of the fruit were cut off and the slices were taken avoiding the seed cavity. Pulp firmness was determined by a puncture test using the Texture Analysis system from Texture

Technologies Corp. fitted with an 8 mm diameter convex probe and 50 kg load cell. The slice was placed on a flat surface for measurement in the center of the exposed tissue.

After establishing zero force contact between the probe and the mesocarp tissue, the probe was driven with a crosshead speed of 50 mm/min-1 for 5 mm depth, puncturing the mesocarp tissue in an inside-out (seed cavity to extremity) direction. The maximum force was recorded.

Internal quality and decay

When fruit was commercially ripe according to firmness, the sixteen fruit per tree were cut in half and internal quality assessed using the rating scale from the

International Avocado Quality Manual where 0=healthy, 0.5=5, 1=10, 1.5=15, 2=25,

2.5=33, 3=50% affected (White et al. 2009). This rating included common disorders such as vascular browning, stem end rot, and body rot, as well as less common disorders including uneven ripening, pink staining, tissue breakdown, seed cavity browning, vascular leaching, and stones in the flesh. This rating did not include flesh bruising that may have been caused by the firmness test.

Fruit phenotypic data

Three unripe fruits were used to take several measurements. Fruit length, width, weight, and seed weight were taken. Seed weight and fruit weight were used to calculate flesh weight of each fruit to obtain seed to flesh ratios. The same measurements were taken for three fruits that reached commercial ripeness.

Dry matter

After taking fruit weight, length, width, and seed weight, three fruits per tree were chosen for sampling dry weight. Two samples per fruit were taken for both ripe and unripe fruit. A core of the mesocarp or pulp (no exocarp or skin and no seed coat) was sampled for dry weight. Fruit were first cut in half (through the seed), laid the flat side down and cut a wedge out of the middle (approximately 1/8th of the fruit). These wedges were used for sampling. A potato peeler or knife was used to remove the skin.

The seed and all traces of the seed coat were removed. When unripe, the mesocarp tissue was cut into smaller pieces and placed into the food processor with a chopping blade. The food processor was run until the avocado was chopped into fine pieces

(starts to stick to the side of the food processor container) and had the size and consistency of grated Parmesan cheese. Five grams were weighed out onto a scale and placed in petri dishes. Samples were dried in an oven at 146 F for at least five days

to obtain dry weight of tissue. The same procedure was done with three ripe fruits except that mesocarp tissue was mashed with a knife rather than chopped in a food processor.

Statistical analysis

The experiments were conducted in a Completely Randomized Design. For peel color, whole fruit firmness, weight loss, and decay, sixteen fruits were used as replicates for each individual tree in the population. For pulp firmness, fruit length, width, seed weight, and dry matter, three unripe fruits were used from each tree and three ripe fruits from the sixteen used in the other evaluations. Statistical procedures were performed using Statistical Analysis Software (SAS) version 9.4 (SAS Institute Inc., Cary, NC).

Differences between means were determined using Tukey’s studentized range test

(honest significant difference-HSD).

Results

Postharvest Quality Parameters

Phenotypic data, dry matter, and weight loss

Unripe fruit ranged between 72.1 mm and 166.8 mm mean length by 51.9 mm and 106.0 mm mean width in 2013, 80.5 mm and 138.0 mm by 53.6 mm and 99.0 mm in 2014, 71.1 mm and 129.4 mm by 58.0 mm and 77.7 mm in 2015 (individual hybrid data not shown). The average of the mean flesh weight was 186.4 g (unripe) and 182.6 g (ripe) in 2013, 195.1 g (unripe) and 206.5 g (ripe) in 2014, 171.6 g (unripe) and 170.8 g (ripe) in 2015 (individual hybrid data not shown). Mean dry matter ranged from 15.5 to

34.4% (unripe) and from 14.5 to 36.0% (ripe) in 2013, from 17.4 to 35.3% (unripe) and from 13.2 to 29.6% (ripe) in 2014, from 12.7 to 29.9% (unripe) and 16.2 to 25.4% in

2015 (individual hybrid data not shown). Phenotypic data on selected avocado trees in

2014 and 2015 are reported in Chapter 5 (Table 5-1; 5-2).

The rate of weight loss was calculated on fruit stored at room temperature

(approximately 24-25˚C) in plastic containers with lids and was highly variable among fruit from different hybrids. Mean rates of fruit weight loss ranged between 0.14% and

0.63% per day over five days in 2013 (Figure 3-1), 0.23% and 0.71% in 2014 over six days (Figure 3-2), and 0.28% and 0.51% over five days in 2015 (Figure 3-3).

Peel color

Lightness (L*), chroma (C*), and hue angle (hº) were evaluated on unripe and ripe fruit for each of the three seasons between 2013 and 2015. Fruit that turned darker upon ripening was considered desirable as they resemble commercial ‘Hass’ in that the color change tends to mask minor imperfections. Upon ripening, all avocados exhibited a loss in green color expressed by lower h˚ values throughout all three years and some appeared to become yellow-green when hue angle values were close to 90 (Table 3-1;

Table 3-2; Table 3-3). Lightness and chroma values of ripe fruit show that most fruit became darker (negative lightness difference values) and dull (negative chroma difference values) upon ripening except fruit from trees R3T39, R5T21, R6T22, R7T12,

R7T47, R7T30, R7T20, R7T57, R8T36, and R8T21 in 2013 (Table 3-4). In 2014 and

2015, all fruit became darker and duller when ripe (Table 3-5; Table 3-6).

Whole fruit firmness and pulp firmness

In 2013, fruit had an average initial whole fruit firmness of 126 N on the first harvest (31 October 2013) and 144 N on the second harvest (14 November 2013) and reached full ripe stage (20-30 N) between 12 and 16 days after harvest (Figure 3-4).

Initial whole fruit firmness (unripe, day 0) was in the ranges of 82-169 N for the October

harvest (Figure 3-5) and 92-181 N for the November harvest (Figure 3-6). After ripening for eight days in October, or seven days in November, some fruit reached full ripeness stage and fruit firmness ranged between 10 N and 127 N, or 12 N and 149 N, respectively (Figure 3-5; Figure 3-6). Fruit with greater than 30 N firmness was still not considered ripe and measurements were continued every other day until reaching the fully ripe stage (20-30 N). Upon reaching ripeness, fruit was subjected to destructive tests such as pulp firmness, internal disorder assessments, and dry weight measurements. In 2014 and 2015, initial average whole fruit firmness was 135 N and

136 N, respectively, and reached full ripe stage within 10 to 14 days after harvest

(Figure 3-4). In 2014, initial whole fruit firmness at harvest (day 0) ranged between 74 N and 197 N and between 14 N and 140 N after eight days ripening (Figure 3-7). In 2015, initial whole fruit firmness at harvest (day 0) was between 91 N and 162 N and between

17 N and 116 N after seven days ripening (Figure 3-8). Pulp firmness peak force values were generally higher on the stem end than the blossom end of unripe fruit throughout

2013 (Table 3-7) and 2014 (Table 3-9). Some differences were observed among fruits from different trees. Pulp firmness values of unripe fruit for the 2015 season were not collected. There were no significant differences among pulp firmness of the blossom and stem ends of ripe fruit in 2013 with the exception of R3T31 and R8T17, possibly due to uneven ripening (Table 3-8). Overall, ripe fruit from the different hybrids did not differ significantly in pulp firmness (with exception of R8T9 in 2014) at the stem or blossom ends throughout 2013 (Table 3-8) and 2014 (Table 3-10) seasons (P<0.001).

However, peak force results for pulp firmness of ripe fruit were not significant in 2015

(P=0.0619) (Table 3-11).

Internal quality and decay

Upon ripening, fruits were evaluated for internal disorders and decay. Significant

(P<0.0001) differences in physiological and pathological disorders were detected each of the three seasons. The highest mean severities in 2013 were observed in disorders such as uneven ripening, vascular browning, pink staining, and tissue breakdown (Table

3-12). Uneven ripening and tissue breakdown had the highest mean severities in 2014.

Seed cavity browning was the most common disorder in 2015, but there were no significant differences among other disorders (Table 3-12). Uneven ripening was absent in the fruit evaluated in 2015 and pink staining and stones in the flesh were not found in any fruit evaluated in 2014 and 2015. Tissue breakdown may be associated with stem end and body rot fungi. Other possible causes for tissue breakdown may have been bacteria or high temperature during ripening (above 25˚C) (Dann et al. 2013).

According to all data collected and analyzed, trees with the best fruit were chosen as “selections” to be used in sensory panels as potential ‘Hass’-like avocados that can do well in a Florida climate (see Chapter 5).

Discussion

The Florida avocado industry is comprised of avocados with West Indian (WI) and West Indian-Guatemalan (WI-G) hybrid backgrounds. Guatemalan (G), Mexican

(M), and their hybrids that are commonly grown in California, reportedly do not do well in the warm, humid Florida climate (Bergh & Ellstrand 1986; Crane et al. 2007).

Cultivars ‘Hass’ and ‘Bacon’, both Guatemalan-Mexican hybrids (G-M), are common in hot dry avocado growing areas such as California, Mexico, Chile, and Israel (Yahia &

Woolf 2011; Crane et al. 2013). Reciprocal crosses of these two cultivars were planted in east-central Florida in the hopes to identify selections well-adapted to this region.

Furthermore, the Florida avocado industry is now affected by laurel wilt, and previous studies have shown that cultivars with Guatemalan and Mexican backgrounds may have higher tolerance to the disease (Mayfield III et al. 2008c; Ploetz et al. 2012).

Compared to other crops, determining when to harvest avocados can be problematic as their fruit do not change color or soften until they are ready to harvest (Yahia & Woolf

2011). Avocados are normally harvested when minimum flesh dry matter content ranges from 17% to 25% with a minimum of 8% oil content, depending on cultivar and location. In California, minimum maturity standards require flesh dry matter contents of

20.8% for ‘Hass’ and 1707% for ‘Bacon’ (Yahia & Woolf 2011). Florida avocados are generally harvested based on days after full bloom and on fruit weight and size, depending on cultivar. Since our study involved California-type avocado cultivar hybrids grown in a Florida climate with no identified maturity index, fruit were harvested when they stopped growing and started to drop from the tree.

In this study, flesh dry matter contents at harvest ranged from 15.5% to 34.4% in

2013, 17.4% to 35.3% in 2014, and 12.7% to 29.9% in 2015. Since dry matter and oil content are highly correlated, they can both be used as maturity indices for avocado.

Fatty acid composition was analyzed for fruit in the present study (see Chapter 4).

Although dry matter has been a good indicator of fruit quality at harvest, dry matter and oil content are not reliable indicators of fruit maturity and quality for ‘Hass’ harvested late in the season (Hofman et al. 2000). Therefore, it was necessary to ensure that fruit was harvested upon reaching physiological maturity and not picked too late in the season.

Another important challenge is managing diseases in the field and postharvest.

Tissue breakdown, uneven ripening, and vascular browning were among the most severe fruit problems during the study (Table 3-12). Fungi associated with stem-end rots and body rots, such as Colletotrichum gloeosporioides, may be to blame.

Anthracnose is a very common postharvest disease on avocado and it thrives in hot humid tropical and subtropical environments like Florida. In addition, temperature management postharvest may also influence anthracnose severity. Maintaining appropriate temperatures during the handling and ripening is important for fruit quality and reducing postharvest rots (Yahia & Woolf 2011). Fruit did not develop anthracnose when stored for up to 21 days at 5˚C (Kotze’ 1978). The fruit in this study was stored at room temperature (24-25˚C), temperatures at which anthracnose developed within 7 days of harvest (Kotze’ 1978). Fruit in this study were stored in containers with holes to allow for gas exchange, and containers were vented and dried every day to avoid excess moisture buildup. This was done to simulate the challenges that would normally exist throughout the entire handling chain from harvest to consumption. Further postharvest studies on appropriate storage temperature should be considered in the future to further enhance this study.

‘Hass’ avocado stored at 15˚C, 20˚C, and 25˚C took 8, 12, and 7 days respectively to reach eating ripeness (Cox et al. 2004). Skin color changed from green to purple then black over 8-12 days during ripening at 20˚C and 25˚C (h˚, C*, L* decreased) (Cox et al. 2004). Obenland et al. (2012) reported that optimum eating pulp firmness of ‘Hass’ should be from 4.4 N to 6.7 N. Florida-type avocados were considered commercially ripe upon reaching whole fruit firmness of 10-20 N (Pereira

2010) and 15-20 N in ‘Booth 7’ avocados when treated with 1-MCP (Zhang et al. 2011).

In this study, optimum eating whole fruit firmness was set at 20-30 N as fruit began to show higher incidence of postharvest rots at lower firmness values. Some fruit reached ripeness stage within 7 and 8 days in 2013, and within 10 to 14 days in 2014 and 2015.

Upon ripening, all avocados exhibited a loss in green color expressed by lower h˚ values throughout all three years with most fruit becoming darker and duller with some exceptions. Due to similarities with commercial ‘Hass’, some hybrids were chosen as selections to be included in sensory panels (see Chapter 5) suggesting that some might make good hybrids to market.

Conclusion

Postharvest evaluations of reciprocal crosses of California-type avocado cultivars

‘Hass’ and ‘Bacon’ grown in east-central Florida, show some similarities with commercial ‘Hass’. Fruit reached good maturity and was ready for harvest towards the end of October through mid-November when California ‘Hass’ fruit is off the market.

Flesh mean percentage dry weight ranges were within the accepted values accepted in

California (17.7% for ‘Bacon’ and 20.8% for ‘Hass’). Some trees have fruit with darker skin color upon ripening to mask some imperfections and resemble ‘Hass’, the most common cultivar on the market. Optimal eating ripeness stage was reached within 8-12 days after harvest in 2013 and within 10-14 days in 2014 and 2015, respectively.

Different trees with better horticultural traits were chosen as preferred selections and included in trained taste panels (see Chapter 5).

Figure 3-1. Mean percentage of daily fruit weight loss rate over 5 days after harvest in 2013.

Figure 3-2. Mean percentage of daily fruit weight loss rate over 6 days after harvest in 2014.

Figure 3-3. Mean percentage of daily fruit weight loss rate over 5 days after harvest in 2015.

Table 3-1. Peel color of unripe and ripe avocado fruit from different hybrids in 2013. Hybrids are ranked in descending order based on hue of unripe fruit. Unripe Ripe Lightness Hue Chroma Lightness Hue Chroma Tree ID (L) (h˚) (C) (L) (h˚) (C) R8T36 35.7 e-q 125.9 a 18.2 f-u 40.2 bc 116.8 a-c 24.2 b-d R5T19 34.1 k-t 125.2 a 16.5 j-v 29.0 o-z 84.6 e-m 6.6 q-x R8T21 33.7 l-v 124.6 a 15.2 n-x 35.0 e-j 123.2 a 16.5 e-k R7T30 35.3 f-r 124.3 a 18.6 e-t 37.6 b-e 121.3 a 22.2 b-e R7T43 35.0 g-r 123.9 ab 19.4 d-r 27.1 a, v-z 60.8 m-t 5.1 s-x R8T5 33.9 k-t 123.8 a-c 15.5 l-w 30.4 k-x 107.9 a-f 8.3 n-x R6T34 38.8 a-e 123.8 a-c 23.1 b-h 29.9 l-y 46.1 q-u 7.7 o-x R8T48 33.9 k-t 123.8 a-c 17.5 i-u 32.1 g-s 112.1 a-d 12.5 j-q R7T20 38.5 a-f 123.6 a-c 23.6 b-g 39.7 bc 122.0 a 24.9 bc R8T43 34.0 k-t 123.6 a-c 15.9 l-w 30.5 k-x 91.2 c-k 8.3 n-x R7T47 36.0 d-p 123.5 a-c 19.2 e-r 37.3 b-f 121.5 a 21.2 b-f R6T22 36.9 b-m 123.4 a-c 20.1 d-p 40.5 b 117.4 a-c 25.0 bc R5T47 35.6 e-q 123.3 a-c 20.6 c-n 35.4 d-h 117.7 a-c 18.8 d-i R8T50 33.2 n-v 123.0 a-c 15.8 l-w 32.6 g-q 118.6 ab 13.9 h-n R8T23 38.0 a-h 122.9 a-c 23.7 b-g 35.1 e-i 100.6 a-h 15.6 f-m R7T48 34.7 h-r 122.7 a-c 17.7 h-u 31.6 h-u 109.3 a-e 11.4 j-r R5T40 33.8 k-u 122.7 a-d 16.5 j-v 33.4 f-m 120.2 a 16.0 f-l R5T35 39.0 a-e 122.6 a-d 25.8 a-c 31.7 h-t 81.7 f-n 10.4 l-u R6T27 34.9 g-r 122.5 a-d 15.9 l-w 33.2 f-n 108.6 a-f 10.4 l-v R6T29 35.7 e-q 122.3 a-d 18.2 f-u 28.0 a, t-z 33.4 u 5.2 s-x R7T19 38.3 a-g 122.3 a-d 22.7 b-i 27.7 a, t-z 48.3 p-u 5.0 t-x R6T38 34.0 k-t 122.3 a-d 17.5 i-u 33.7 e-l 119.7 ab 17.3 e-j R6T44 32.6 q-w 122.2 a-d 15.2 n-x 28.0 a, t-z 89.7 d-l 6.2 r-x R5T21 36.3 d-o 122.2 a-d 20.9 c-m 36.1 c-g 114.1 a-d 20.1 c-g R7T35 37.8 b-i 122.1 a-d 20.4 c-o 34.2 e-k 98.8 a-i 12.8 i-p R5T25 37.6 b-i 122.1 a-d 22.0 b-j 30.4 k-y 80.4 g-o 7.4 p-x R6T36 37.1 b-k 122.0 a-d 22.7 b-i 31.3 i-u 92.7 b-j 9.8 m-w R8T18 41.2 a 121.9 a-d 29.6 a 30.0 l-y 62.7 m-s 8.1 n-x R7T12 39.1 a-d 121.8 a-d 24.9 a-d 41.4 b 117.7 a-c 26.8 b R8T3 33.2 n-v 121.7 a-d 14.9 o-x 31.1 i-v 108.6 a-f 10.0 l-w R7T21 31.0 t-w 121.6 a-d 10.5 w-y 30.6 k-x 103.3 a-g 10.7 k-t R5T46 34.9 h-r 121.6 a-d 17.4 i-u 32.9 g-p 113.8 a-d 14.1 g-n R7T6 37.8 a-i 121.3 a-d 21.7 b-k 34.2 e-k 108.0 a-f 13.9 h-n R7T13 36.5 c-n 121.2 a-d 20.1 d-o 29.3 m-z 73.4 i-p 8.1 n-x R7T39 35.0 g-r 121.2 a-d 18.1 g-u 33.0 g-o 101.9 a-g 13.6 h-o R6T21 35.0 g-r 121.0 a-d 18.5 e-t 34.4 e-k 111.6 a-d 15.8 f-m R7T15 33.1 o-v 121.0 a-d 12.8 u-y 27.9 a, t-z 59.3 m-u 3.5 x R7T8 34.9 h-r 120.8 a-d 17.5 i-u 34.4 e-k 116.7 a-c 15.9 f-l

Table 3-1. Continued Unripe Ripe Lightness Hue Chroma Lightness Hue Chroma Tree ID (L) (h˚) (C) (L) (h˚) (C) R7T57 38.1 a-h 120.7 a-d 23.8 b-f 39.2 b-d 117.8 a-c 25.0 bc R3T39 39.8 a-c 120.5 a-d 26.7 ab 47.8 a 111.5 a-e 36.5 a R4T46 33.0 o-v 120.3 a-d 13.9 r-y 32.4 g-r 118.1 a-c 13.1 i-p R5T13 36.8 b-m 120.1 a-e 21.1 b-l 32.5 g-r 97.8 a-j 13.2 i-p R5T16 33.3 n-v 120.0 a-e 16.6 j-v 24.8 a 34.4 tu 4.0 wx R8T9 39.9 ab 120.0 a-e 23.9 b-e 32.6 g-q 73.1 i-p 9.5 n-x R7T36 32.6 q-w 119.9 a-e 15.2 n-x 26.2 a, yz 41.8 r-u 4.8 t-x R8T24 37.0 b-k 119.9 a-e 22.4 b-i 27.9 a, t-z 59.8 m-u 6.6 q-x R7T5 36.2 d-o 119.5 a-e 19.6 d-p 30.5 k-x 74.8 h-p 8.3 n-x R8T17 32.8 p-w 118.9 a-e 14.9 o-x 28.4 a, r-z 71.8 j-q 6.2 r-x R4T13 33.5 n-v 118.9 a-e 16.2 k-v 31.3 i-u 114.5 a-d 11.1 k-s R8T44 34.1 k-t 118.8 a-e 15.1 n-x 31.0 i-v 91.4 c-k 8.4 n-x R6T25 35.2 f-r 118.7 a-f 19.0 e-s 35.5 d-h 104.4 a-g 17.3 e-j R6T18 37.5 b-j 118.5 a-f 22.6 b-i 26.5 a, x-z 55.9 n-u 5.3 r-x R3T20 30.5 u-w 118.3 a-f 11.1 v-y 25.7 a-z 50.8 p-u 4.2 wx R5T51 36.5 c-n 117.8 a-f 18.3 f-u 31.7 h-t 82.1 f-n 8.2 n-x R6T49 37.0 b-l 117.7 a-f 21.1 b-l 36.3 c-g 112.1 a-d 19.3 c-h R6T6 36.3 d-o 117.6 a-f 19.5 d-q 28.2 a, s-z 64.1 l-s 5.7 r-x R3T52 33.6 m-v 117.4 a-f 15.4 m-w 28.8 a, p-z 65.5 k-r 5.5 r-x R6T50 34.5 i-s 117.2 a-f 16.4 j-v 29.4 m-z 108.4 a-f 11.1 k-s R7T7 33.0 o-v 116.9 a-f 12.7 u-y 28.0 a, t-z 71.1 j-q 4.1 wx R5T56 31.3 s-w 116.7 a-f 12.8 u-y 26.6 a, x-z 62.4 m-s 4.5 u-x R6T28 35.3 f-r 116.4 a-h 17.9 h-u 28.9 o-z 71.9 i-q 7.5 p-x R4T52 36.2 d-o 116.3 a-g 17.4 i-u 30.8 k-w 79.2 g-o 7.3 p-x R3T31 33.3 n-v 115.4 a-g 15.1 n-x 27.6 a, u-z 57.8 m-u 5.4 r-x R4T39 38.0 a-h 114.5 a-g 22.4 b-i 28.1 a, s-z 53.6 o-u 5.7 r-x R3T30 32.0 r-w 111.4 b-h 13.1 t-y 30.9 j-v 106.0 a-g 11.3 j-r R6T40 32.7 p-w 111.1 c-h 13.8 r-y 29.2 n-z 71.6 j-q 7.3 p-x R7T16 34.4 j-s 109.9 d-h 14.0 q-x 29.4 m-z 60.2 m-u 5.1 s-x R6T19 30.4 vw 107.4 e-h 9.7 xy 27.1 a, v-z 65.7 k-r 4.3 v-x R6T54 33.9 k-t 106.0 f-h 14.5 p-x 29.3 m-z 60.2 m-u 7.4 p-x R8T51 35.1 f-r 103.8 gh 17.9 h-u 26.7 a, w-z 39.3 r-u 5.2 s-x R8T11 32.5 q-w 100.3 hi 13.5 s-y 28.6 a, q-z 38.3 s-u 4.7 t-x R4T44 29.5 w 88.1 i 8.3 y 27.0 a, v-z 49.6 p-u 4.9 t-x

P values <.0001*** <.0001*** <.0001*** <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Table 3-2. Peel color of unripe and ripe avocado fruit from different hybrids in 2014. Hybrids are ranked in descending order based on hue of unripe fruit. Unripe Ripe Lightness Hue Chroma Lightness Hue Chroma Tree ID (L) (h˚) (C) (L) (h˚) (C) R8T9 37.6 a-c 122.4 a 21.3 b 31.5 e-i 85.6 c-h 7.4 e-h R7T54 36.8 b-d 122.3 a 18.0 b-f 34.4 c-f 122.9 a 16.4 cd R8T17 35.0 c-g 121.9 a 20.8 bc 29.9 g-k 75.7 f-h 7.0 e-h R7T7 32.9 d-h 121.9 a 12.9 f-j 28.9 h-k 84.5 d-h 4.6 gh R8T5 33.5 d-h 121.1 a 14.6 d-i 29.2 g-k 94.4 b-e 5.9 f-h R8T38 39.4 ab 121.0 a 15.3 d-h 34.2 c-f 121.9 a 16.5 cd R8T3 32.7 e-h 121.0 a 11.7 h-j 31.7 e-h 119.2 ab 10.1 ef R8T11 33.1 d-h 120.3 a 17.1 b-h 27.8 jk 46.0 hi 5.1 f-h R8T21 35.1 c-g 120.2 a 16.3 b-h 36.8 bc 121.6 a 18.4 c R7T34 37.6 a-c 119.9 a 19.9 b-d 36.2 c 114.7 a-c 17.1 cd R8T54 35.4 c-g 119.7 a 19.2 b-d 32.5 d-g 99.6 a-f 9.2 e-g R3T39 41.1 a 119.2 a 28.4 a 44.5 a 117.0 ab 32.7 a R8T18 36.4 b-e 119.0 a 21.2 b 30.3 g-j 66.6 g-i 9.6 e-g R7T19 36.4 b-e 118.8 a 18.5 b-e 31.3 f-j 80.9 e-h 7.2 e-h R3T16 35.4 c-g 117.1 ab 18.3 b-f 35.3 cd 113.5 a-d 16.8 cd R4T35 36.0 b-e 117.1 ab 19.0 b-d 34.9 c-e 121.2 a 17.7 c R8T26 31.7 f-h 115.8 ab 11.8 g-j 28.4 h-k 79.4 e-h 5.4 f-h R7T52 33.8 c-h 115.7 a 15.7 b-h 31.1 f-j 91.9 c-f 8.0 e-h R7T48 35.5 c-g 113.0 ab 17.4 b-g 32.6 d-g 105.3 a-e 11.9 de R7T15 31.9 f-h 112.9 ab 13.0 e-j 28.2 h-k 63.9 g-i 4.8 f-h R7T57 35.6 b-f 112.3 ab 21.3 b 40.1 b 115.5 ab 25.6 b R5T25 33.9 c-h 112.2 ab 14.6 d-i 28.2 i-k 67.7 g-i 4.8 f-h R6T18 33.7 c-h 111.4 ab 15.3 c-h 28.0 i-k 66.9 g-i 5.6 f-h R8T4 31.6 gh 110.6 ab 14.8 d-h 28.6 h-k 75.7 f-h 6.6 e-h R7T16 31.6 gh 101.0 b 9.1 ij 28.5 h-k 76.2 e-h 5.4 f-h R8T51 30.3 h 67.9 c 7.6 j 26.6 k 39.3 i 3.0 h

Table 3-3. Peel color of unripe and ripe avocado fruit from different hybrids in 2015. Hybrids are ranked in descending order based on hue of unripe fruit. Unripe Ripe Lightness Hue Chroma Lightness Hue Chroma Tree ID (L) (h˚) (C) (L) (h˚) (C) R3T16 32.8 e 125.2 a 15.4 a 33.7 b-e 122.8 a 14.8 a R7T54 36.1 b-d 123.2 a 18.9 c-e 34.9 bc 119.6 a 15.5 b-e R8T21 35.3 b-e 123.1 a 16.9 c-e 35.9 b 123.2 a 17.5 c-e R7T35 35.7 b-d 122.1 a 17.5 b-e 33.4 b-e 110.9 ab 11.9 b-d R5T47 35.8 b-d 122.1 a 20.5 a-c 36.1 b 119.3 a 19.4 bc R6T56 34.6 c-e 121.9 a 18.2 a-d 34.5 b-d 118.3 a 17.4 bc R7T29 33.5 de 121.8 a 16.0 b-d 31.3 d-g 110.2 ab 10.4 bc R8T18 37.4 a-c 121.7 a 23.9 c-e 30.9 e-g 87.4 c 10.0 b-e R8T38 32.7 e 121.1 a 15.1 c-e 30.7 e-g 117.1 a 11.8 c-e R8T4 32.5 ef 121.0 a 15.6 c-e 32.6 c-e 115.7 a 13.8 d-f R5T25 34.5 de 120.9 a 17.2 a-c 29.3 gh 85.3 c 5.5 bc R7T20 37.9 ab 120.8 a 20.1 a-d 36.2 b 119.7 a 17.8 bc R7T48 34.9 c-e 120.1 a 17.3 c-e 32.5 c-g 112.4 ab 11.2 b-d R6T36 36.3 b-d 120.1 a 20.1 a-d 31.3 d-g 94.1 bc 9.5 bc R3T52 34.5 de 119.4 a 17.2 ab 34.6 bc 113.8 ab 14.7 ab R8T17 34.4 de 118.4 a 17.6 c-e 29.7 fg 78.2 c 7.0 b-e R5T2 39.4 a 115.8 ab 23.1 ab 40.7 a 115.0 a 24.3 b R3T20 29.8 f 104.6 bc 9.3 a 25.9 i 54.8 d 3.8 a R8T51 32.7 ef 101.5 c 14.3 c-e 26.2 hi 40.2 d 4.4 d-g

Table 3-4. Change in peel color between ripe and unripe avocado fruit in 2013. Hybrids are ranked in descending order based on changes in hue. Change in Change in Change in Chroma Tree ID Lightness (ΔL) Hue (Δh˚) (ΔC) R6T29 -7.7 p-w -88.9 z -13.0 a, r-z R5T16 -8.5 s-x -85.5 yz -12.7 a, q-z R7T36 -6.1 m-u -78.0 x-z -10.0 a, n-z R6T34 -9.0 t-x -77.6 w-z -15.7 a-c, x-z R7T19 -12.8 x -75.9 v-z -20.9 bc R8T11 -5.1 i-u -70.9 u-z -10.7 a, o-z R3T20 -5.0 h-u -68.1 t-z -7.0 g-t R8T51 -8.8 s-x -67.4 s-z -13.3 a, s-z R7T15 -5.4 k-u -62.6 r-z -9.8 m-z R6T18 -11.0 v-x -62.5 q-z -17.3 a-c R7T43 -8.0 q-x -62.3 q-z -15.2 a-c, w-z R8T24 -8.5 s-x -60.1 q-z -14.9 a-c, v-z R8T18 -11.6 w-x -58.9 p-z -22.3 c R3T31 -6.0 m-u -57.7 o-y -10.0 a, n-z R3T52 -5.7 l-u -54.7 o-x -11.9 a, p-z R4T39 -9.7 u-x -54.5 o-x -16.6 a-c, z R5T56 -4.6 h-t -54.1 o-x -7.9 i-w R6T6 -8.1 r-x -53.5 o-x -13.8 ab, t-z R7T16 -5.3 j-u -52.4 n-x -9.5 m-z R8T9 -7.8 p-w -52.1 m-x -16.2 a-c, yz R8T17 -4.5 g-t -47.6 l-w -9.1 l-y R7T13 -6.1 m-u -47.2 l-v -10.9 a, o-z R6T54 -4.6 h-t -45.8 k-v -7.1 g-t R7T7 -4.8 h-t -44.8 k-u -8.4 j-x R7T5 -5.8 l-u -44.4 k-u -11.4 a, o-z R6T28 -6.7 n-v -44.4 j-u -11.0 a, o-z R6T19 -3.5 f-r -43.0 i-u -5.8 f-r R5T25 -7.2 o-w -41.7 h-u -14.6 ab, u-z R5T35 -7.4 p-w -40.8 h-u -16.2 a-c, yz R6T40 -3.5 f-r -39.5 g-t -6.5 g-t R4T44 -2.5 d-o -37.4 f-s -3.3 d-n R4T52 -5.4 k-u -37.1 f-r -10.2 a, n-z R5T51 -4.9 h-u -35.4 e-r -10.2 a, n-z R6T44 -4.8 h-u -33.5 d-r -9.6 m-z R5T19 -4.1 g-s -33.4 c-r -9.0 k-y R8T43 -3.5 f-r -32.4 b-q -7.7 h-v R6T36 -5.9 m-u -29.2 a-p -12.9 a, r-z R8T44 -3.2 e-q -28.3 a-o -6.8 g-t R7T35 -3.4 f-r -23.3 a-n -7.3 g-u

Table 3-4. Continued Change in Change in Change in Tree ID Lightness (ΔL) Hue (Δh˚) Chroma (ΔC) R8T23 -2.9 e-p -22.3 a-n -8.1 j-w R5T13 -4.8 h-t -22.0 a-m -8.2 j-w R7T39 -1.6 d-m -19.9 a-l -4.2 e-o R7T21 -0.5 b-j -18.6 a-l 0.0 b-g R8T5 -3.5 f-r -16.0 a-k -7.3 g-u R6T25 -0.5 b-j -14.3 a-j -3.2 d-n R6T27 -1.8 d-m -14.0 a-i -5.3 e-p R7T48 -3.1 e-p -13.3 a-i -6.3 g-s R7T6 -3.4 f-r -13.3 a-i -7.3 g-u R8T3 -2.0 d-n -13.1 a-i -4.9 e-p R8T48 -1.7 d-m -11.7 a-h -4.9 e-p R6T50 -4.8 h-t -9.5 a-g -4.7 e-p R3T39 7.0 a -9.3 a-f 8.4 a R6T21 -0.8 c-k -9.2 a-f -2.9 d-n R5T46 -1.5 d-m -9.0 a-f -2.5 d-m R8T36 4.4 ab -8.8 a-f 5.9 ab R5T21 0.3 b-g -8.7 a-f -0.2 b-g R4T13 -2.0 d-n -6.9 a-e -5.4 e-q R6T22 3.6 a-c -6.5 a-e 5.1 a-c R6T49 -0.4 b-i -5.2 a-d -1.7 c-k R5T47 -0.1 b-h -5.2 a-d -1.5 b-j R3T30 -0.9 c-l -4.9 a-d -1.4 b-j R8T50 -0.6 c-k -4.8 a-d -2.1 c-l R7T12 1.6 b-e -4.5 a-d 1.5 a-f R7T8 -0.8 c-k -4.2 a-d -2.0 c-l R5T40 -0.5 c-j -3.3 a-c -0.7 b-i R7T47 1.2 b-f -3.0 ab 1.8 a-e R7T30 2.3 a-d -2.9 ab 3.6 a-d R6T38 -0.6 c-k -2.3 ab -0.5 b-h R4T46 -0.7 c-k -1.9 a -0.7 b-i R7T20 1.3 b-f -1.7 a 1.3 a-f R8T21 1.3 b-f -1.4 a 1.3 a-f R7T57 1.3 b-f -1.4 a 1.3 a-f

P values <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Table 3-5. Change in peel color between ripe and unripe avocado fruit in 2014. Hybrids are ranked in descending order based on changes in hue. Change in Change in Hue Change in Tree ID Lightness (ΔL) (Δh˚) Chroma (ΔC) R8T9 -37.2 f-i -123.3 c -20.9 h-j R8T17 -35.0 b-g -121.9 c -20.8 h-j R7T7 -32.9 a-e -121.9 c -12.9 b-d R8T26 -32.6 a-e -121.6 c -13.9 b-e R8T5 -33.5 a-f -121.5 c -14.6 c-f R8T38 -39.4 hi -121.0 c -15.3 c-g R8T3 -32.7 a-e -121.0 c -11.7 a-c R8T11 -32.5 a-e -120.9 c -16.6 d-h R7T54 -35.4 c-g -120.6 c -17.7 d-i R8T21 -35.1 b-g -120.2 c -16.3 c-h R4T35 -36.2 e-h -120.1 c -20.3 h-j

R8T18 -37.0 f-h -119.9 c -23.0 j R7T34 -37.6 g-i -119.7 c -19.9 g-j R8T54 -35.4 c-g -119.7 c -19.2 f-j R3T39 -41.1 i -119.2 c -28.4 k R7T19 -36.1 e-h -118.5 bc -18.1 e-i R3T16 -35.4 c-g -117.1 bc -18.2 e-j R7T15 -31.8 a-d -116.7 bc -12.9 b-d R7T52 -33.6 a-f -115.3 bc -14.1 c-e R5T25 -34.1 a-g -113.9 bc -15.2 c-g R7T57 -35.8 d-h -113.3 bc -22.1 ij R7T48 -35.6 d-h -113.2 bc -17.7 d-i R6T18 -33.7 a-g -111.4 bc -15.3 c-g R8T4 -31.4 ab -110.5 bc -14.9 c-f R7T16 -31.6 a-c -101.0 b -9.0 ab R8T51 -30.3 a -67.9 a -7.6 a P values <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Table 3-6. Change in peel color between ripe and unripe avocado fruit in 2015. Hybrids are ranked in descending order based on changes in hue. Change in Change in Change in Tree ID Lightness (ΔL) Hue (Δh˚) Chroma (ΔC) R3T16 -32.8 b -125.2 c -15.4 b R7T54 -36.1 c-e -123.2 c -18.9 b-e R8T21 -35.3 b-e -123.1 c -16.9 b-d R7T35 -35.7 c-e -122.1 c -17.5 b-d R5T47 -35.8 c-e -122.1 c -20.5 d-f R6T56 -34.6 b-d -121.9 c -18.2 b-d R7T29 -33.5 bc -121.8 c -16.0 b-d R8T18 -37.4 d-f -121.7 c -23.9 f R8T38 -32.7 b -121.1 c -15.1 b R8T4 -32.5 ab -121.0 c -15.6 bc R5T25 -34.5 bc -120.9 c -17.2 b-d R7T20 -37.9 ef -120.8 c -20.1 c-f R7T48 -34.9 b-d -120.1 c -17.3 b-d R6T36 -36.3 c-e -120.1 c -20.1 c-f R3T52 -34.5 bc -119.4 c -17.2 b-d R8T17 -34.4 bc -118.4 c -17.6 b-d R5T2 -39.4 f -115.8 bc -23.1 ef R3T20 -29.8 a -104.6 ab -9.3 a R8T51 -32.7 ab -101.5 a -14.3 b P values <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Figure 3-4. Average whole fruit firmness (N) of avocado for three harvest seasons. Dashed line at 30N indicates fruit that were considered “fully ripe”.

Figure 3-5. Whole fruit firmness average peak force at days 0 and 8 of 10/31/2013 harvest.

Figure 3-6. Whole fruit firmness average peak force at days 0 and 7 of 11/14/2013 harvest.

Figure 3-7. Whole fruit firmness average peak force at days 0 and 8 of 10/28/2014 harvest.

Figure 3-8. Whole fruit firmness average peak force at days 0 and 7 of 10/23/2015 harvest.

Table 3-7. Pulp firmness of unripe avocado fruit from different hybrids in 2013. Peak force (N) Mean force (N)

Tree ID blossom stem blossom stem end R3T20 120.1 e-q 147.1 c-q 55.9 b 75.5 a-r R3T30 207.0 a-i 218.6 a-g 112.0 a-h 113.2 a-h R3T31 110.7 f-q 115.3 f-q 47.3 i-r 45.0 j-r R3T39 35.6 q 56.8 pq 23.3 r 36.1 n-r R3T52 146.4 c-q 219.1 a-g 62.5 b-r 114.3 a-g R4T13 128.2 d-q 151.8 b-p 61.3 c-r 79.3 a-r R4T39 157.8 b-p 183.3 a-m 66.1 b-r 79.2 a-r R4T44 134.0 c-q 141.8 c-q 64.5 b-r 69.3 a-r R4T46 174.9 a-o 179.2 a-n 95.7 a-p 106.1 a-j R4T52 265.5 ab 280.0 a 116.5 a-e 130.2 a R5T13 116.7 f-q 158.0 b-p 49.8 h-r 69.3 a-r R5T16 163.9 b-p 173.1 a-o 87.0 a-q 82.7 a-r R5T19 123.9 e-q 210.9 a-h 60.5 c-r 118.4 a-d R5T21 95.5 i-q 134.3 c-q 45.8 i-r 71.6 a-r R5T25 162.2 b-p 159.7 b-p 85.2 a-r 78.4 a-r R5T35 96.4 i-q 137.0 c-q 37.8 m-r 51.9 g-r R5T40 145.4 c-q 189.5 a-k 69.5 a-r 113.8 a-g R5T46 192.5 a-k 187.9 a-k 89.8 a-q 92.2 a-q R5T47 139.8 c-q 165.9 a-p 64.5 b-r 91.5 a-q R5T51 161.3 b-p 207.3 a-i 66.8 a-r 96.9 a-o R5T56 122.6 e-q 173.8 a-o 60.8 c-r 105.9 a-k R6T18 162.3 b-p 198.7 a-k 81.0 a-r 103.5 a-k R6T19 170.9 a-p 205.0 a-i 80.3 a-r 106.5 a-j R6T21 135.0 c-q 151.9 b-p 62.8 b-r 77.0 a-r R6T22 233.4 a-e 199.5 a-j 122.0 a-c 112.5 a-h R6T25 140.5 c-q 176.0 a-o 67.0 a-r 97.0 a-o R6T27 170.7 a-p 244.6 a-c 81.2 a-r 123.0 a-c R6T28 140.1 c-q 144.8 c-q 75.6 a-r 79.5 a-r R6T29 127.6 d-q 166.7 a-p 62.6 b-r 84.4 a-r R6T34 123.5 e-q 102.0 h-q 63.2 b-r 54.0 e-r R6T36 164.7 b-p 192.2 a-k 87.5 a-q 95.6 a-p R6T38 144.1 c-q 195.1 a-k 79.1 a-r 107.2 a-j R6T40 130.7 c-q 168.6 a-p 52.4 f-r 96.4 a-o R6T44 154.8 b-p 169.5 a-p 80.7 a-r 94.4 a-q R6T49 164.1 b-p 179.1 a-n 83.5 a-r 97.6 a-n R6T50 143.1 c-q 152.3 b-p 69.9 a-r 77.2 a-r R6T6 145.5 c-q 119.0 f-q 64.0 b-r 51.4 g-r

Table 3-7. Continued Peak force (N) Mean force (N) Tree ID blossom stem blossom stem end R7T12 109.4 f-q 115.7 f-q 55.8 d-r 58.3 d-r R7T13 105.1 g-q 113.2 f-q 58.3 d-r 60.0 c-r R7T15 214.3 a-h 219.5 a-f 101.1 a-m 115.9 a-f R7T16 170.0 a-p 175.7 a-o 81.6 a-r 85.1 a-r R7T19 103.2 h-q 157.9 b-p 50.8 g-r 91.3 a-q R7T20 142.7 c-q 168.8 a-p 74.4 a-r 86.6 a-r R7T21 175.1 a-o 171.6 a-o 100.5 a-m 97.5 a-n R7T30 143.2 c-q 168.2 a-p 78.1 a-r 82.9 a-r R7T35 161.7 b-p 181.3 a-m 79.8 a-r 89.9 a-q R7T36 163.1 b-p 190.2 a-k 73.3 a-r 103.1 a-l R7T39 69.1 m-q 64.2 o-q 32.0 p-r 32.6 p-r R7T43 163.9 b-p 174.0 a-o 78.9 a-r 89.7 a-q R7T47 106.6 f-q 65.6 n-q 47.5 i-r 33.7 o-r R7T48 191.7 a-k 194.9 a-k 102.1 a-l 100.4 a-m R7T57 128.5 d-q 152.6 b-p 75.0 a-r 81.7 a-r R7T5 165.2 b-p 156.1 b-p 84.5 a-r 75.3 a-r R7T6 197.3 a-k 241.6 a-d 92.2 a-q 104.4 a-k R7T7 138.3 c-q 134.6 c-q 63.6 b-r 67.9 a-r R7T8 161.0 b-p 159.2 b-p 78.5 a-r 67.1 a-r R8T11 159.0 b-p 129.3 d-q 83.5 a-r 70.5 a-r R8T17 122.5 e-q 177.8 a-o 60.8 c-r 97.1 a-o R8T18 196.2 a-k 185.1 a-l 100.6 a-m 95.4 a-p R8T21 198.5 a-k 173.1 a-o 108.9 a-i 92.7 a-q R8T23 152.6 c-p 183.2 a-m 76.6 a-r 103.1 a-l R8T24 72.8 l-q 126.8 e-q 31.3 qr 62.3 c-r R8T36 90.7 j-q 107.6 f-q 39.7 l-r 42.3 k-r R8T3 186.2 a-l 198.8 a-k 89.8 a-q 101.1 a-m R8T43 113.3 f-q 205.2 a-i 54.3 e-r 117.7 a-e R8T44 84.9 k-q 145.2 c-q 47.2 i-r 83.6 a-r R8T48 164.4 b-p 189.4 a-k 91.4 a-q 102.6 a-l R8T50 149.8 c-q 139.7 c-q 78.8 a-r 72.7 a-r R8T51 147.6 c-q 134.2 c-q 67.6 a-r 69.7 a-r R8T5 177.4 a-o 212.8 a-h 90.9 a-q 126.1 ab R8T9 141.3 c-q 167.6 a-p 76.9 a-r 90.5 a-q P values <.0001*** <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slice.

Table 3-8. Pulp firmness of ripe avocado fruit in 2013. Peak force (N) Mean force (N) Tree ID blossom stem end blossom stem end R3T20 4.4 c 4.0 c 3.5 cd 3.0 cd R3T30 2.6 c 2.9 c 2.3 cd 2.0 cd R3T31 2.2 c 34.6 b 1.8 cd 17.7 b R3T39 3.4 c 2.7 c 2.6 cd 1.9 cd R3T52 1.9 c 0.5 c 1.6 cd 0.3 d R4T13 5.9 c 2.9 c 3.5 cd 2.2 cd R4T39 2.5 c 7.3 c 1.9 cd 4.6 cd R4T44 3.1 c 3.7 c 2.4 cd 2.8 cd R4T46 3.5 c 4.9 c 2.7 cd 3.8 cd R4T52 4.3 c 4.0 c 3.5 cd 2.8 cd R5T13 2.8 c 3.4 c 2.2 cd 2.5 cd R5T16 5.2 c 6.8 c 3.9 cd 5.3 cd R5T19 3.3 c 6.1 c 2.7 cd 4.7 cd R5T21 1.1 c 2.3 c 0.9 cd 1.6 cd R5T25 3.4 c 2.7 c 2.7 cd 2.2 cd R5T35 1.8 c 2.4 c 1.5 cd 1.7 cd R5T40 3.9 c 7.0 c 2.8 cd 5.3 cd R5T46 4.1 c 4.0 c 3.0 cd 2.9 cd R5T47 2.5 c 3.0 c 2.1 cd 2.2 cd R5T51 2.9 c 4.6 c 2.3 cd 3.2 cd R5T56 2.3 c 9.1 c 1.9 cd 6.6 cd R6T18 2.4 c 2.1 c 1.9 cd 1.6 cd R6T19 2.3 c 2.2 c 1.8 cd 1.6 cd R6T21 2.2 c 2.9 c 1.8 cd 2.3 cd R6T22 7.9 c 4.0 c 5.6 cd 3.1 cd R6T25 1.6 c 2.2 c 1.4 cd 1.6 cd R6T27 3.7 c 3.1 c 2.8 cd 2.5 cd R6T28 3.1 c 8.3 c 2.3 cd 5.4 cd R6T29 2.5 c 3.0 c 1.9 cd 2.1 cd R6T34 2.2 c 2.4 c 1.7 cd 1.8 cd R6T36 6.0 c 11.3 c 4.5 cd 8.1 c R6T38 2.6 c 5.2 c 2.1 cd 3.9 cd R6T40 1.7 c 5.9 c 1.4 cd 4.2 cd R6T44 2.5 c 3.7 c 1.9 cd 2.8 cd R6T49 3.6 c 5.4 c 2.4 cd 3.8 cd R6T50 3.2 c 5.7 c 2.7 cd 4.5 cd R6T6 6.8 c 10.5 c 5.5 cd 6.8 cd R7T12 1.5 c 1.5 c 1.2 cd 1.1 cd

Table 3-8. Continued Peak force (N) Mean force (N) Tree ID blossom stem end blossom stem end R7T13 1.6 c 2.1 c 1.3 cd 1.6 cd R7T15 5.4 c 6.1 c 4.3 cd 4.9 cd R7T16 5.7 c 5.0 c 4.5 cd 4.0 cd R7T19 1.8 c 1.9 c 1.5 cd 1.5 cd R7T20 4.0 c 4.8 c 2.7 cd 3.7 cd R7T21 2.0 c 2.9 c 1.4 cd 2.3 cd R7T30 1.8 c 2.1 c 1.5 cd 1.5 cd R7T35 2.8 c 2.2 c 2.2 cd 1.8 cd R7T36 3.8 c 5.9 c 3.0 cd 4.2 cd R7T39 3.2 c 3.6 c 2.5 cd 2.9 cd R7T43 1.2 c 6.9 c 1.0 cd 4.0 cd R7T47 2.3 c 3.2 c 1.8 cd 2.3 cd R7T48 7.0 c 7.1 c 5.7 cd 6.0 cd R7T57 1.8 c 2.5 c 1.5 cd 1.8 cd R7T5 4.9 c 4.0 c 3.8 cd 3.1 cd R7T6 4.5 c 5.4 c 3.5 cd 3.9 cd R7T7 5.3 c 3.2 c 4.2 cd 2.5 cd R7T8 3.1 c 2.7 c 2.2 cd 2.1 cd R8T11 4.0 c 3.0 c 2.6 cd 2.4 cd R8T17 3.7 c 1.3 a 2.8 cd 57.1 a R8T18 3.2 c 1.9 c 2.7 cd 1.5 cd R8T21 2.3 c 4.6 c 1.7 cd 3.1 cd R8T23 1.8 c 3.4 c 1.6 cd 2.5 cd R8T24 2.0 c 1.4 c 1.7 cd 1.2 cd R8T36 1.7 c 2.0 c 1.3 cd 1.5 cd R8T3 3.4 c 2.8 c 2.5 cd 2.2 cd R8T43 3.0 c 8.8 c 2.5 cd 6.8 cd R8T44 1.9 c 2.6 c 1.6 cd 2.0 cd R8T48 3.3 c 5.9 c 2.7 cd 4.7 cd R8T50 3.5 c 5.7 c 2.7 cd 4.3 cd R8T51 4.0 c 4.2 c 3.4 cd 3.4 cd R8T5 6.7 c 6.5 c 5.1 cd 5.4 cd R8T9 4.1 c 9.7 c 2.9 cd 6.4 cd P values <.0001*** <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slices.

Table 3-9. Pulp firmness of unripe avocado fruit from different hybrids in 2014. Peak force (N) Mean force (N)

Plant ID blossom stem blossom stem R3T16 212.2 a-d 210.0 a-d 107.5 a-e 100.6 a-g R4T35 156.2 a-g 174.1 a-g 92.0 a-g 111.6 a-c R5T25 119.3 e-h 129.9 d-g 61.2 d-h 71.4 b-g R7T15 221.5 a-c 171.9 a-g 106.7 a-e 81.1 b-g R7T19 33.0 h 118.5 e-h 20.8 h 53.3 f-h R7T34 86.9 gh 135.3 c-g 53.2 f-h 77.0 b-g R7T48 177.5 a-f 224.3 a-c 91.8 a-g 110.2 a-d R7T52 132.1 d-g 148.6 b-g 68.0 c-h 86.6 b-g R7T54 120.0 e-h 144.7 c-g 60.3 e-h 91.1 a-g R7T57 94.3 f-h 138.8 c-g 52.0 gh 70.7 b-g R8T11 162.0 a-g 186.7 a-e 77.4 b-g 98.2 a-g R8T17 192.9 a-e 242.4 a 92.4 a-g 138.6 a R8T18 164.1 a-g 163.3 a-g 78.3 b-g 89.1 b-g R8T21 161.6 a-g 177.1 a-f 86.0 b-g 102.1 a-f R8T3 223.1 a-c 236.9 ab 106.5 a-e 118.9 ab R8T4 155.1 a-g 188.8 a-e 78.3 b-g 106.2 a-e R8T54 114.2 e-h 146.1 c-g 53.6 f-h 70.8 b-g R8T5 185.2 a-e 179.7 a-f 93.8 a-g 109.8 a-d R8T9 118.9 e-h 155.1 a-g 53.7 f-h 78.0 b-g P values <.0001*** <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slices.

Table 3-10. Pulp firmness of ripe avocado fruit from different hybrids in 2014. Peak force (N) Mean force (N) Plant ID blossom stem blossom stem R3T16 5.3 ab 3.5 b 4.0 ab 2.3 ab R3T39 5.3 ab 7.8 ab 4.5 ab 6.2 ab R4T35 3.6 b 8.3 ab 2.8 ab 6.3 ab R5T25 3.4 b 3.6 b 2.7 ab 2.8 ab R6T18 4.4 ab 4.1 ab 3.1 ab 3.3 ab R7T15 5.1 ab 6.9 ab 3.9 ab 5.6 ab R7T19 2.6 b 6.6 ab 2.1 ab 5.4 ab R7T34 3.2 b 5.0 ab 2.4 ab 3.7 ab R7T48 5.7 ab 4.7 ab 4.4 ab 3.3 ab R7T52 3.3 b 2.1 b 2.6 ab 1.6 b R7T54 3.6 b 4.7 ab 2.9 ab 3.9 ab R7T57 2.2 b 2.7 b 1.7 b 2.0 b R7T7 5.7 ab 6.7 ab 3.5 ab 5.1 ab R8T11 3.0 b 2.6 b 2.4 ab 2.0 b R8T17 3.0 b 4.4 ab 2.2 ab 3.5 ab R8T18 6.8 ab 3.9 ab 5.3 ab 3.0 ab R8T21 2.3 b 3.1 b 1.7 b 2.1 ab R8T26 3.6 b 3.9 ab 2.7 ab 3.0 ab R8T38 3.2 b 7.0 ab 2.6 ab 4.7 ab R8T3 5.0 ab 3.0 b 3.9 ab 2.5 ab R8T4 2.3 b 4.8 ab 1.8 b 3.7 ab R8T51 4.1 ab 6.2 ab 3.0 ab 4.7 ab R8T54 5.2 ab 3.3 b 3.8 ab 2.2 ab R8T5 5.3 ab 5.4 ab 4.0 ab 4.2 ab R8T8 3.4 b 5.0 ab 2.7 ab 3.5 ab R8T9 3.6 b 10.8 a 3.0 ab 7.3 a P values <.0003*** <.0003*** <.0002*** <.0002*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slices.

Table 3-11. Pulp firmness of ripe avocado fruit from different hybrids in 2015. Peak force (N) Mean force (N) Plant ID blossom stem blossom stem R3T16 7.4 ab 2.0 b 5.6 ab 1.4 b R3T20 4.7 ab 9.3 ab 3.5 b 7.3 ab R3T52 6.4 ab 3.5 ab 4.9 ab 2.7 b R5T2 4.8 ab 4.2 ab 3.6 b 3.4 b R5T25 17.2 ab 4.7 ab 9.7 ab 3.4 b R5T47 6.4 ab 5.5 ab 4.8 ab 3.8 ab R6T36 3.6 ab 2.8 ab 2.7 b 2.0 b R6T56 2.7 ab 3.2 ab 2.1 b 2.6 b R7T20 5.7 ab 4.4 ab 3.6 b 3.3 b R7T29 4.3 ab 5.6 ab 3.3 b 2.8 b R7T35 3.1 ab 2.6 ab 2.3 b 2.0 b R7T48 6.5 ab 4.7 ab 4.9 ab 2.9 b R7T54 2.6 ab 1.9 b 2.0 b 1.2 b R8T4 5.4 ab 7.1 ab 3.8 ab 5.1 b R8T17 4.3 ab 17.8 a 3.1 b 12.9 a R8T18 5.4 ab 3.7 ab 4.3 ab 2.6 b R8T21 4.4 ab 5.2 ab 2.8 b 3.7 b R8T38 3.9 ab 10.6 ab 3.1 b 7.7 ab R8T51 5.2 ab 7.3 ab 4.0 ab 5.6 ab P values 0.0619 NS 0.0619 NS 0.0045 ** 0.0045 ** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slices.

Table 3-12. Postharvest disorders on ripe avocado fruit harvested over three seasons (2013-2015). Mean severity Postharvest disorders 0 to 3 index 2013 N = 737 2014 N = 249 2015 N = 249

Vascular leaching 1.58ab 1.95a 2.00a Uneven ripening 2.40a 2.33a NA Vascular browning 2.15ab 1.72a 2.07a Pink staining 2.00ab NA NA Tissue breakdown 1.97ab 2.20a 2.03a Seed cavity browning 1.26a-c 1.55a 2.10a Stem end rot 1.04a-c 1.11ab 1.36a Body rot 0.92bc 1.64a 1.48a Stones in flesh 1.00a-c NA NA P values <.0001*** <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). The ratings were as follows: 0=healthy, 0.5=5%, 1=10%, 1.5=15%, 2=25%, 2.5=33%, 3=50%. NA indicates absence of postharvest disorder.

100

CHAPTER 4 CHARACTERIZATION OF FATTY ACID CONTENT FROM THE PULP OF AVOCADO FRUITS OF SELECTED ‘HASS’-‘BACON’ HYBRIDS GROWN IN EAST CENTRAL FLORIDA

Introduction

The avocado fruit has gained much interest and popularity due to the beneficial nutritional and cosmetic properties of its oil. The avocado (Persea americana Mill.) belongs to the Lauraceae family which includes other species such as pondspice

(Litsea aestivalis L.), camphor laurel (Cinnamomum camphora L.), redbay (Persea borbonia (L.) Spreng.), and over 150 other species mostly originating in tropical America

(Bora et. al 2001). The oil content of avocado fruit is greatly dependent on the cultivar and its ecological origin. Avocado originates from three main races; Mexican (M),

Guatemalan (G), and West Indian (WI). Pulp of the Mexican-type fruit is known to have the highest oil content with up to 30 percent. Fruit of Guatemalan origin have medium oil content of 8-15 percent and the West Indian type have the lowest oil content of 3-10 percent (Purseglove 1968). Regardless of the many positive attributes similar to olive oil, avocado is not considered to be a primary source of oil compared to olive, and as a result, only a few studies have been conducted on oil extraction from the pulp (Moreno et al. 2003). Lipid content of the three races of avocado fruit is reportedly comprised of

15-20% saturated fats, 60-80% monounsaturated fatty acids, and about 10% polyunsaturated fatty acids (Yahia & Woolf 2011). A diet recommended as healthy by

Wang et al. (2015) is one high in monounsaturated fatty acids (MUFAs). Even though avocado is rich in linoleic acid (omega-6), which has been considered unhealthy, the omega-3/omega-6 ratio may not matter after all since omega-6 linoleic acid from plants does not convert at a high rate to arachidonic acid, a precursor to inflammatory

101

cytokines. Avocado oil has a very similar profile to olive oil and studies that compared the two showed that oil stability during heating was very similar from these two sources

(Berasategi et al. 2012). Like olive oil, avocado oil can be cold pressed without altering amounts of bioactive phytochemicals in the fruit (Berasategi et al. 2012). In addition to its high concentration of oleic acid, MUFAs, antioxidant vitamins, and phytosterols

(known to decrease cholesterol absorption) (Wang et al. 2015), avocado oil has significant amounts of other beneficial compounds such as tocopherols (vitamin E), phosphorus, calcium, magnesium, folate, and also has more potassium per gram of flesh than bananas (Requejo et al. 2003; Yahia & Woolf 2011; Berasategi et al. 2012).

Avocado is therefore considered a healthy addition to the diet (Yahia & Woolf 2011).

Studies have shown that fatty acid content increases during fruit development.

According to a study by Du Plessis (1979), composition of avocado oil changes seasonally as fruits develop. Oleic acid content generally increases, while palmitic

(omega-6) and linolenic acid contents decrease. These changes vary with cultivar and climate. Other minor fatty acid contents also change or remain constant during growth and development of the fruit (Du Plessis 1979). Fruit mesocarp and seed, also known as the almond, contain many fatty acids, but the main ones in mesocarp tissue are oleic

(C18:1), palmitoleic (C16:1), palmitic (C16:0), linoleic (C18:2), and linolenic (C18:3) (Moreno et al. 2003; Pacetti et al. 2007) and are the ones we analyzed during our study.

The Florida avocado industry grows cultivars that are mainly West Indian and

West Indian-Guatemalan hybrids that are lower in oil content and are considered best suited to warmer climates. ‘Hass’, the most popular cultivar worldwide, and ‘Bacon’ are

Mexican-Guatemalan hybrids with higher oil content and are more tolerant of freezing

102

winter conditions that sometimes occur in the subtropics. Percentage dry matter and oil content have been used in California-type avocados as indicators of fruit maturity and quality. In our study, we focus on the fatty acid composition of hybrids from reciprocal crosses of cultivars ‘Hass’ and ‘Bacon’ in relation to their suitability to growing in a warmer east central Florida climate.

Materials and Methods

Raw material

Avocado (Persea americana Mill.) fruit was obtained from bearing seedling trees of ‘Hass’ and ‘Bacon’ hybrids located at USDA-ARS Picos Farm, Fort Pierce, Florida.

Trees originated from seeds collected from a commercial orchard in California by Dr.

Raymond Schnell. Molecular marker analysis was used to determine which seedlings were truly hybrids before planting. Trees were planted at the USDA-ARS Picos farm in

Fort Pierce, Florida on August 29, 2008 on double row beds in Riviera fine sand soil type. A regular biweekly maintenance regimen includes a spray with horticultural oil and copper (CS-2005, Magna-Bon II, LLC, Okeechobee, FL), and fertilization with a N:P:K

20:10:20 soluble fertilizer. Soil applied Ridomil (Syngenta) and foliar applied Lexx-A-

Phos (Foliar Nutrients, Inc., Cairo, GA) was applied twice per year for Phytophthora control together with a granular dry fertilizer (12 2 14) at approximately 500-600 pounds/acre. Fruit was harvested on 29 October 2013, 14 November 2013, and 28

October 2014 and stored at room temperature until ripe. Samples were taken from 72 trees in 2013 and 28 trees in 2014. Trees are normally topped to 15 feet towards the end of November; however, trees were topped in February (during bloom) in 2014, which resulted in considerably less fruit in 2014. Two samples per fruit and three fruits per tree were analyzed for total percent oil content and main fatty acid composition. The

103

fatty acids analyzed were ones that were previously reported in the literature on avocado (Slater et al. 1975; Du Plessis 1979; Bora et al. 2001; Ozdemir & Topuz 2004).

Due to time constraint, some of the duplicate samples were not analyzed. For each sample, 500mg of ripe mesocarp tissue at the equatorial region of the fruit was chopped with a scalpel, weighed and put in a 15mL glass tube with PTFE

(Polytetrafluoroethylene) lined caps and 3mL isopropanol. Tubes were stored at -40˚C until analysis.

Lipid extraction and methylation

Samples were brought to room temperature and heated at 80˚C for 10 minutes.

Samples were then homogenized using a glass rod and 100µL of a 50mg/mL 17:0 glyceryl triheptadecanoate surrogate (TAG) was added. One mL of isopropanol and

6mL hexane were added for a final ratio of 4:6 isopropanol:hexane. Samples were shaken and let stand for 3 minutes. Five mL of a saturated sodium chloride (NaCl) solution was added to the samples, shaken vigorously and allowed to separate.

Samples were then spun at 2000 rpm for 5 min at room temperature in a Sorvall RC-5C

Plus centrifuge (Sorvall, Kendro Laboratory Products, Newtown, CT, USA). The upper phase was removed with a glass Pasteur pipette and transferred to a pre-labeled 15mL glass vial. Solvents were then evaporated under a continuous stream of nitrogen in a

40˚C water bath (Nitrogen evaporator, N-EVAP III OA-SYS Heating System,

Organomation Associates, Inc. Berlin, MA, USA). The lipid residue remaining after evaporation was redissolved in 1mL chloroform and either stored in the freezer or processed immediately. When ready for processing, 100µL of the sample in chloroform was aliquoted into a new 7mL glass vial with PTFE lined caps and dried down in the nitrogen evaporator. Saponification reaction was carried out by adding 1mL of a 0.5N

104

NaOH/MeOH solution, and heating for approximately 1-2 minutes at 80˚C until saponification was achieved as noted by dissolution of oil. Vial was cooled to room temperature. Then, 1mL of a 1N methanolic hydrochloric acid (HCl) was added to the vial, which was incubated at room temperature for one hour allowing methylation to take place. The reaction was then quenched with 1mL milliQ water. Lipids were then extracted using hexane with an internal lipid standard for the fatty acid methyl ester

(FAME 15:0 50ng/µL methyl pentadecanoate). This sample was vortexed for 30 seconds and the vials were kept at room temperature (24-25˚C) 1-2 minutes until phase separation was complete. The supernatant was transferred to a new 7mL glass vial containing approximately 500 mg sodium sulfate to absorb any moisture. Hexane (with the internal standard) extraction of the aqueous phase was repeated two more times and upper phases combined to get a total final volume of 6mL. 200 µL of the extract and

800 µL of the hexane (with the internal standard) for a final ratio of 1:5 was pipetted into a labeled glass GC 2 mL autosampler vial ready for GC analysis.

Fatty acid analysis

Fatty acid methyl esters (FAMEs) were identified using the Supelco 37 component standard. The Supelco 37 component standard has 37 different FAMEs that run from C6 out to C24. These FAMEs represent considerable diversity (ex. cis and trans isomers). The chromatographic resolution was maximized and used in setting up the

GC method. There were no other FAMEs detected other than the ones that were analyzed, therefore, the most common fatty acids previously reported on avocado were the ones included in the analysis of this study.

Fatty acid methyl ester (FAME) determination was performed using a gas chromatograph HP 6890 series (Hewlett Packard) equipped with CTC Analytics Combi

105

PAL autosampler split/splitless injector, flame ionization detector and a 100m x 0.25mm i.d. x 0.2µm film SP-2560 fused silica open tubular (FSOT) capillary column using helium carrier gas. The initial head pressure of the carrier gas was 38.6 psi and was electronically increased throughout the run to maintain a constant column flow rate of the carrier gas at 1.5ml/min. The oven temperature was held initially at 60˚C for 2 min, later increased at 16˚C/min to 170˚C and held for 6 min, and then increased to 250˚C at

4˚C/min for 10 min. The temperatures of the injection port and the detector were 250˚C and 280˚C, respectively. Fatty acid column eluants were positively identified by matching their retention times with those of reference standards. Calibration curve verification (CCV) and calibration curve blank (CCB) vials were loaded onto the GC after every 10 samples. Calibration blank runs verified that we never had carry over after running the calibration verification vials. Calibration curve verifications were always within 15% of a known concentration and the coefficient of determination showed that the regression line perfectly fit the data (R2>0.98).

Results

The analysis of avocado pulp from reciprocal crosses of cultivars ‘Hass’ and

‘Bacon’ is presented in Table 4-1. Fatty acid composition was similar both years. Oleic acid (C18:1) was the main fatty acid in avocado mesocarp tissue (34.9% in 2013 and

36.6% in 2014 respectively). Palmitic (C16:0) acid was also abundant in fruit (up to 27.9% and 30.0%) and the third most abundant fatty acid was linoleic acid (C18:2) (21.2% and

18.4%). Both stearic (C18:0), and linolenic (C18:3) acids were scarce (0.4-2%) with myristic only found in trace amounts as previously reported for other avocado cultivars

(Bora et al. 2001; Moreno et al. 2003; Pacetti et al. 2007). There was no correlation found between oleic acid and total oil contents in 2013 (P > 0.099) and 2014 (P >

106

0.324), respectively. Fatty acid composition for each tree in the population that set fruit in 2013 and 2014 is presented in Tables 4-2 and 4-3, respectively. A diet high in monounsaturated fatty acids, such as oleic and palmitoleic has been considered as healthy and recommended to lower bad cholesterol (Hall et al. 1980; Wang et al. 2015).

Oleic acid is the most abundant fatty acid found in avocado mesocarp tissue. R8T3 had the lowest percentage of oleic acid throughout both years (16.6 and 16.1%). R7T7 had the highest oleic acid content in 2014 (47.7%) and was among the trees with the highest oleic content in 2013 (47.4%). The relative percent composition of avocado fatty acids of avocado pulp is also represented in Figure 4-1.

Discussion

The results in this study are similar to past reports on ‘Hass’ and ‘Fuerte’. Oleic acid was the principal monounsaturated fatty acid. Palmitic, linoleic, and palmitoleic acids were the second, third, and fourth most abundant fatty acids found in mesocarp tissue. Stearic, linoleic, and myristic were found in small numbers and trace amounts.

However, the concentrations of these fatty acids were different than the cited reports.

‘Fuerte’ avocados grown in the Northeast region of Brazil were reported to contain 64% oleic acid (Bora et al. 2001). Oleic acid was observed to be the only fatty acid to increase continuously in ‘Hass’ and ‘Fuerte’ from November to January with percentages ranging from 59.3% to 73.0% in ‘Fuerte’ and from 47.2% to 59.5% in

‘Hass’ when grown in Turkey (Ozdemir & Topuz 2004). Results observed on the same varieties grown in California averaged 75.1% in ‘Fuerte’ and 78.4% in ‘Hass’ (Slater et al. 1975). In this study, average oleic acid was 34.9% and 36.6% in 2013 and 2014, respectively when harvested at the end of October and mid-November. Some of the hybrids in this study had higher oleic acid content and were comparable to previous

107

reports such as R8T9, R6T44, R5T56, and R4T39 in 2013 (53.8%, 51.0%, 54.7%, and

56.8%). Overall, this quantitative difference observed on most of the hybrids in this report can be explained by the varietal difference, geographical origin and maturity index of the fruit.

Vegetable oils, mainly palm and olive, are the major source of edible lipids consumed in the world. Avocado is another very important oil fruit that has gained popularity due to its high nutritional content and health benefits. Avocado is a high caloric fruit with a high content of monounsaturated fatty acids (MUFAs) such as oleic and palmitoleic acids (Yahia & Woolf 2011). A diet high in monounsaturated fatty acids has been shown to improve the ratio of good HDL to bad LDL cholesterol and benefit people with diabetes and high blood pressure (Hall et al. 1980; Wang et al. 2015).

Studies have shown that regions with a diet high in monounsaturated fatty oils, such as in olive oil, have had a lower incidence in atherosclerotical cardiovascular disease associated with high blood cholesterol levels (Moreno et al. 2003). Similar studies conducted in Mexico with diets rich in avocado pulp, have had similar lower cholesterol results (López et al. 1996). Health benefits of avocado have been questioned due to the relatively high concentrations of the saturated fatty acid palmitic acid, and omega-6 fatty acid known to have inflammatory properties. However, omega-6 linoleic acid from plants does not convert at high rate to arachidonic acid, the precursor to inflammatory cytokines and palmitic acid content in avocado is not considered to be deleterious to a healthy diet. Avocado is also rich in ascorbic acid, folic acid, vitamin E, vitamin B6, β- carotene, and minerals such as potassium, phosphorus, calcium, and magnesium

(Yahia & Woolf 2011).

108

Determination of a good maturity index is difficult in avocado, as fruit matures only when removed from the tree and the ripening process is not accompanied by visible external changes. Oil and moisture contents of avocado mesocarp tissue are inversely proportional; as oil content increases steadily beginning a few weeks after fruit set, and water content decreases by the same amount (Du Plessis 1979; Ozdemir &

Topuz 2004). Dry matter and oil content are therefore used as maturity indices in avocado growing regions. Studies done by Slater (1975) and Du Plessis (1979) investigate the seasonal changes on the chemical composition of avocado oil of different cultivars. It was observed that oleic acid increases whereas palmitic acid decreases throughout the season (Du Plessis 1979). Similar results were observed when oleic acid was reported to be the only fatty acid to increase continuously while the fruit was on the tree across three harvesting times from November to January (Ozdemir

& Topuz 2004). In this study, it was assumed that fruit reached physiological maturity when fruit stopped growing (end of October), and higher oleic acid values were expected when compared to other fatty acids. These results agree with previous findings (Ratovohery et al. 1988; Bora et al. 2001; Moreno et al. 2003; Ozdemir & Topuz

2004; Pacetti et al. 2007). Oleic acid was the most abundant fatty acid found in mesocarp tissue on both years even if in lower amounts (35 and 37%). However, some trees had as much as 54% (R8T9) and 55% (R5T56) in 2013 (Table 4-2). Other reports have found oleic acid to be in the 60-70% range in ‘Fuerte’ and ‘Hass’, depending on the harvest season (Slater 1975; Bora et al. 2001; Ozdemir & Topuz 2004). Similar content (63.5%) was found in the store-bought ‘Hass’ from Chile used as a control in the

2015 sensory panel (see Chapter 5 Table 5-10). The difference could be due to different

109

variables such as cultivar difference, geographic area, and physiological maturity of the fruit. Bora et al. (2001) have compared fatty acid composition in pulp and seed of cultivar ‘Fuerte’ in the Northeast region of Brazil. Oleic acid in pulp tissue was the main monounsaturated fatty acid (64.3%) and palmitic acid was the principal saturated fatty acid (21.3%). Ozdemir and Topuz (2004) found oleic acid content to be 53% in ‘Hass’ and 65% in ‘Fuerte’. As fruit in this study was considered ready for harvest at the end of

October, similar oleic acid content was expected. Studies conducted by Slater (1975) on

‘Hass’ and ‘Fuerte’ in California have shown an average of 75% oleic acid in ‘Fuerte’ and 78% on ‘Hass’. Again, palmitic acid content decreased throughout the season.

Regardless, these findings matched previous reports in which oleic acid was found to be the principal fatty acid component of mesocarp tissue, followed by palmitic, linoleic, and palmitoleic as the second, third, and fourth most abundant fatty acids, and linolenic, stearic, and myristic only found in very small or trace amounts. Previous studies by

Ozdemir and Topuz (2004) show that content of fatty acids differ at different harvest dates as well as during postharvest ripening stages. They also show a cultivar difference observed between ‘Hass’ and ‘Fuerte’ (Ozdemir & Topuz 2004). Lipid content in this study was analyzed on postharvest ripe fruit on cultivar crosses ‘Hass’ and

‘Bacon’, which are normally not grown in a hot and humid environment like Florida.

Nonetheless, our results agree with previous reports but can also be considered novel findings as this is the first report on avocado pulp oil content analysis of California-type avocados grown in an east-central Florida environment. Growing location may have little effect on oil content as some hybrids did show similar oil composition values to previous studies conducted on ‘Hass’. Average oleic acid values may increase and

110

resemble previous studies if fruit is harvested later in the season. Therefore, future studies on seasonal changes on the chemical composition of avocado oil should be considered.

Conclusion

Fatty acid composition was analyzed on fruit from hybrids produced from reciprocal crosses of California cultivars ‘Hass’ and ‘Bacon’ grown in an east-central

Florida climate. Findings of this study are similar to previous reports of fatty acid composition of mesocarp tissue of ripe avocado. Oleic acid was found to be the most abundant fatty acid followed by palmitic, linoleic, and palmitoleic acids. Linolenic, stearic, and myristic were found in very small and/or trace amounts. The value variability of this study is most likely due to geographic origin and time of harvest.

However, values found in this study confirm that the east-central Florida climate supports good quality fruit based on lipid content as fruit had a minimum of 8% oil content at harvest (except R5T35 in 2014) as set by the California avocado industry. Oil content is highly correlated with percent dry weight and both are generally used as maturity indices in avocado producing areas (except Florida). In previous findings by

Lee et al. (1983), acceptable taste was noted at a minimum of 20.0% and 22.8% dry weight for ‘Bacon’ and ‘Hass’, respectively (see Chapter 3 for dry weight data). Percent oil for acceptable taste was 8.7% for ‘Bacon’ and 11.2% for ‘Hass’ (Lee et al. 1983).

Hybrids in this study greatly varied in oil content (from 8% to 25.5% in 2013 and from

9.0% to 20.5% in 2014) and fruit from trees that showed better horticultural traits, such as fruit quality and yield were selected and included in trained sensory panels. Lipid content from those fruit was also analyzed to see if our chosen hybrids were comparable to commercial ‘Hass’ (see Chapter 5).

111

Figure 4-1. Relative percent composition of avocado fatty acids of fruit harvested in 2013 and 2014.

112

Table 4-1. Fatty acid composition of ‘Hass’-‘Bacon’ hybrids grown in east-central Florida % total fatty acid*a Fatty Acid 2013b N=403 2014b N=91 Saturated fatty acids: Tetradecanoic acid myristic C14:0 Tr Tr Hexadecanoic acid palmitic C16:0 27.88 ± 0.23 30.01 ± 0.56 Octadecanoic acid stearic C18:0 0.67 ± 0.02 2.41 ± 0.31

Monounsaturated fatty acids:

9-hexadecenoic acid palmitoleic C16:1 15.05 ± 0.33 14.05 ± 0.68 9-octadecenoic acid oleic C18:1 34.89 ± 0.51 36.60 ± 0.88

Polyunsaturated fatty acids:

9,12-octadecadienoic acid linoleic C18:2 21.19 ± 0.21 18.40 ± 0.53 9,12,15-octadecatrienoic acid linolenic C18:3 0.53 ± 0.03 0.47 ± 0.06 *Tr - Traces (Concentration less than 0.02% of the total fatty acids). aAverage fatty acid (FAME) percentage composition (±SE) of avocado pulp oils by GC- FID. bAverage of oil determinations on the ripe fruit from a composite sample of 3 fruit per tree of 72 trees in 2013 and 28 trees in 2014. Two samples per fruit and three fruits per tree were analyzed for total percent oil content and main fatty acid composition. Due to time constraint, some of the duplicate samples were not analyzed. For every ten samples, a sample was extracted in triplicate for extraction efficiency verification.

113

Table 4-2. Average fatty acid methyl ester percentage composition (±SE) of avocado pulp oils from trees of a mapping population in 2013. % Total Fatty Acida oil Tree ID palmitic palmitoleic stearic oleic linoleic linolenic content R8T51 24.4±0.5 11.1±0.1 0.7±0.0 37.2±0.4 25.4±0.5 1.1±0.1 13.0±0.6 R8T50 33.4±1.5 14.6±1.8 0.5±0.0 34.3±2.2 17.1±1.8 0.1±0.1 17.7±5.0 R8T48 30.7±0.4 17.1±0.6 0.5±0.0 25.6±1.8 25.8±1.2 0.2±0.2 11.9±0.8 R8T44 26.0±0.3 10.7±0.2 0.6±0.0 43.8±0.7 18.6±0.4 0.2±0.1 21.6±1.5 R8T43 26.9±1.0 7.7±0.4 0.9±0.1 33.5±2.3 30.2±1.2 0.9±0.1 14.0±0.8 R8T36 24.0±0.3 9.6±0.4 0.7±0.0 44.7±0.5 20.3±0.6 0.7±0.1 15.4±0.6 R8T24 21.4±0.4 12.8±0.6 0.5±0.1 37.3±0.8 25.8±0.3 2.2±0.2 12.1±0.5 R8T23 26.2±0.2 8.5±0.2 0.6±0.0 38.4±1.0 25.9±0.9 0.3±0.2 14.6±0.5 R8T21 34.9±1.0 20.7±1.3 0.6±0.0 26.7±2.0 17.0±0.2 0.1±0.1 19.8±1.9 R8T18 26.9±1.4 15.5±0.4 0.5±0.1 36.2±1.6 20.8±0.4 0.1±0.1 14.8±0.9 R8T17 32.0±0.8 18.2±0.2 0.5±0.0 29.1±0.4 20.1±0.5 0.0±0.0 16.7±1.1 R8T11 28.9±0.3 10.8±0.4 0.5±0.1 30.5±1.0 27.9±0.9 1.2±0.1 10.6±0.9 R8T9 21.3±0.5 6.0±0.8 0.5±0.0 53.8±0.8 18.3±0.8 0.0±0.0 17.0±0.8 R8T5 26.0±0.5 10.0±0.2 0.5±0.0 44.7±0.8 18.7±0.4 0.2±0.2 19.5±0.6 R8T3 32.7±0.9 33.3±0.1 0.2±0.1 16.6±0.5 17.2±0.3 0.0±0.0 11.5±1.1 R7T57 32.8±0.9 9.7±0.4 0.5±0.2 38.4±1.9 18.4±0.6 0.3±0.3 12.7±1.2 R7T48 24.1±0.9 10.4±0.7 0.6±0.0 43.2±2.2 21.1±0.6 0.6±0.1 17.2±0.7 R7T47 34.4±0.2 19.7±0.2 0.6±0.0 26.8±0.5 17.6±0.2 0.9±0.1 15.6±0.5 R7T43 25.0±0.3 11.0±0.2 0.8±0.0 40.1±0.8 22.8±0.7 0.2±0.1 13.5±0.4 R7T39 31.6±0.4 22.6±0.3 0.3±0.1 18.9±0.3 25.6±0.4 1.1±0.2 8.4±0.5 R7T36 25.4±0.4 4.7±0.2 0.8±0.0 44.6±0.9 24.2±1.1 0.4±0.1 12.4±0.6 R7T35 26.4±0.4 8.4±0.2 0.5±0.0 42.2±0.6 22.5±0.8 0.0±0.0 14.0±0.4 R7T30 26.4±0.7 12.6±0.2 0.6±0.0 42.0±0.8 18.4±1.2 0.0±0.0 16.5±1.8 R7T21 31.4±0.3 19.7±0.1 0.6±0.0 30.5±0.5 17.7±0.3 0.1±0.1 17.5±0.8 R7T20 29.6±0.3 29.4±0.7 0.1±0.1 21.2±1.5 19.8±2.2 0.0±0.0 11.3±0.9 R7T19 23.7±0.4 13.7±0.1 0.6±0.0 37.2±0.7 23.8±0.8 1.0±0.2 8.0±0.7 R7T16 32.0±0.3 22.7±0.6 0.2±0.1 26.6±1.2 18.6±0.8 0.0±0.0 11.1±0.5 R7T15 22.0±0.4 6.9±0.2 0.6±0.0 42.5±3.3 27.8±3.1 0.3±0.3 18.9±1.9 R7T13 30.3±0.2 20.0±0.4 0.6±0.0 22.1±0.5 25.9±0.6 1.0±0.1 9.8±0.7 R7T12 30.2±0.2 16.6±0.4 0.7±0.0 33.2±0.6 18.6±0.9 0.7±0.2 12.5±0.6 R7T8 31.5±0.7 24.9±1.1 0.5±0.0 27.4±2.3 14.9±0.7 0.7±0.1 16.6±0.4 R7T7 21.1±0.2 14.2±0.2 0.6±0.0 47.4±0.6 16.0±0.2 0.7±0.0 14.1±0.9 R7T6 19.7±0.5 6.1±0.2 0.8±0.0 45.1±2.7 27.4±2.6 0.9±0.5 16.1±1.2 R7T5 22.6±0.3 15.2±0.2 0.5±0.0 41.4±0.7 20.2±0.4 0.1±0.1 18.2±0.7 R6T54 25.5±0.1 12.1±0.3 0.7±0.0 44.1±0.1 17.4±0.3 0.2±0.0 25.2±0.7 R6T50 31.1±0.6 20.9±1.0 0.5±0.0 33.4±1.0 14.2±0.9 0.0±0.0 23.0±1.6 R6T49 26.9±0.4 18.7±0.5 0.5±0.0 30.3±0.6 22.8±0.3 0.8±0.0 16.5±0.5 R6T44 16.8±0.3 8.4±0.3 0.5±0.0 51.0±0.6 22.9±0.4 0.4±0.0 23.4±1.4 R6T40 27.7±0.8 17.0±0.8 0.7±0.1 30.9±1.4 22.8±0.5 0.8±0.2 20.2±0.7 R6T38 29.9±0.7 21.8±0.6 1.1±0.1 23.4±1.3 22.5±0.4 1.3±0.1 16.7±0.5

114

Table 4-2. Continued % Total Fatty Acida oil Tree ID palmitic palmitoleic stearic oleic linoleic linolenic content R6T36 27.8±0.3 18.5±0.7 1.3±0.0 30.8±0.7 20.6±0.4 0.9±0.1 18.6±1.1 R6T34 30.8±1.7 16.4±1.2 1.8±0.2 33.2±0.7 19.5±1.2 0.1±0.1 17.3±1.4 R6T29 27.6±1.5 6.1±0.6 0.3±0.2 44.1±4.1 21.9±2.5 0.0±0.0 11.6±1.9 R6T28 31.4±0.4 18.4±0.8 0.5±0.0 29.5±0.2 20.1±0.9 0.0±0.0 15.8±0.8 R6T27 31.4±0.8 17.3±0.8 0.5±0.0 33.5±1.5 17.4±0.7 0.0±0.0 16.1±0.4 R6T25 34.4±0.6 28.4±0.6 1.4±0.4 18.4±2.0 26.0±3.6 0.9±0.2 17.7±3.9 R6T22 33.3±1.8 17.8±1.0 0.7±0.3 20.4±1.4 25.1±2.5 0.6±0.4 18.6±0.7 R6T21 30.2±0.4 14.3±0.4 0.6±0.0 39.7±0.4 15.2±0.7 0.0±0.0 21.0±0.8 R6T19 22.9±0.5 10.0±0.2 0.5±0.0 48.6±1.5 17.9±1.4 0.0±0.0 16.9±0.4 R6T18 27.9±0.2 17.5±0.2 0.6±0.0 31.7±0.6 21.5±0.3 0.9±0.1 19.5±0.8 R6T6 26.9±0.8 9.7±1.0 0.6±0.0 39.4±0.5 22.4±1.3 1.1±0.1 15.9±0.1 R5T56 17.5±0.2 5.3±0.2 0.6±0.0 54.7±0.6 21.8±0.6 0.2±0.1 14.7±0.8 R5T51 36.0±0.7 16.6±1.2 0.5±0.0 26.7±0.3 20.0±1.6 0.1±0.1 17.0±2.8 R5T47 25.7±0.4 11.4±0.2 0.6±0.0 41.3±0.6 21.0±0.5 0.0±0.0 14.8±0.7 R5T46 33.4±0.6 21.9±1.4 0.6±0.0 27.7±1.6 16.4±0.4 0.0±0.0 19.7±0.7 R5T40 25.8±0.4 8.8±0.2 0.8±0.0 43.5±0.8 21.1±0.8 0.0±0.0 15.8±1.2 R5T35 33.6±0.4 6.3±0.2 0.8±0.0 31.2±1.2 28.2±1.1 0.0±0.0 6.8±0.5 R5T25 30.3±1.0 18.5±1.1 0.5±0.0 26.3±2.6 23.1±0.9 1.5±0.3 12.5±1.0 R5T21 23.0±0.4 8.2±0.2 0.7±0.0 44.8±1.1 23.2±0.8 0.1±0.1 16.6±0.7 R5T19 23.8±0.1 11.0±0.1 0.6±0.0 40.6±0.3 23.1±0.5 0.9±0.1 17.8±0.4 R5T16 30.0±0.6 28.0±2.2 0.2±0.1 21.0±0.2 20.1±1.9 0.6±0.3 9.8±2.6 R5T13 25.8±0.7 13.4±0.5 0.6±0.0 42.1±0.8 17.9±0.6 0.2±0.1 25.5±1.2 R4T52 25.6±2.1 13.3±0.5 0.4±0.0 43.2±2.2 16.9±1.8 0.5±0.0 17.8±1.9 R4T46 25.8±0.4 8.3±0.4 0.4±0.0 48.6±1.3 17.0±0.9 0.0±0.0 11.2±0.4 R4T44 24.6±0.6 10.3±0.4 0.5±0.0 48.9±0.4 15.7±0.5 0.0±0.0 17.2±0.8 R4T39 18.7±0.9 7.3±0.2 0.7±0.0 56.8±1.3 16.2±2.0 0.3±0.2 22.0±2.0 R4T13 29.3±0.5 22.2±0.9 0.5±0.0 31.2±0.9 16.7±0.7 0.1±0.1 21.5±0.5 R3T52 26.5±3.4 7.9±2.2 0.7±0.0 44.7±6.5 20.1±1.2 0.3±0.1 18.8±2.7 R3T39 30.5±2.2 8.9±1.7 0.4±0.2 46.4±1.5 13.8±1.8 0.0±0.0 11.5±2.6 R3T31 23.6±0.2 8.8±0.2 0.7±0.0 48.3±0.5 18.5±0.8 0.1±0.1 19.5±0.9 R3T30 33.1±0.1 17.8±1.5 0.5±0.0 29.6±3.3 19.0±1.6 0.0±0.0 16.1±1.7 R3T20 34.2±0.4 19.6±1.0 0.5±0.0 25.1±1.0 20.7±1.3 0.0±0.0 14.5±2.3 aAverage of oil determinations on ripe fruit from 3 fruit per tree by GC-FID.

115

Table 4-3. Average fatty acid methyl ester percentage composition (±SE) of avocado pulp oils from trees of a mapping population in 2014. % Total Fatty Acida Tree oil palmitic palmitoleic stearic oleic linoleic linolenic ID content R8T54 23.4±0.6 7.3±0.3 0.5±0.0 46.0±0.5 22.2±0.4 0.5±0.1 15.3±0.7 R8T51 28.7±0.4 14.1±0.2 0.5±0.0 37.5±0.7 18.8±0.7 0.4±0.2 14.4±0.5 R8T38 23.5±0.6 9.4±0.4 0.6±0.0 47.7±1.0 18.8±1.0 0.0±0.1 17.0±0.7 R8T26 27.7±0.6 17.6±0.3 0.4±0.0 39.3±0.2 15.0±0.4 0.0±0.0 13.4±0.8 R8T21 36.8±0.5 21.3±0.4 0.5±0.0 26.6±1.4 14.7±0.6 0.1±0.1 19.0±0.4 R8T18 27.9±0.5 16.4±0.7 0.5±0.0 34.3±0.9 20.7±0.4 0.3±0.1 13.5±0.8 R8T17 34.0±0.3 14.0±0.5 0.4±0.0 32.0±0.7 19.3±0.7 0.3±0.1 14.8±0.5 R8T16 27.9±0.6 14.2±0.7 0.4±0.1 38.4±0.8 18.7±0.6 0.4±0.0 15.4±3.0 R8T11 35.5±3.4 12.0±1.0 0.6±0.1 36.8±2.0 14.7±6.8 0.4±0.4 9.0±1.0 R8T9 21.0±1.2 5.6±0.5 0.5±0.0 54.2±2.1 18.7±0.6 0.0±0.0 10.9±0.9 R8T5 32.6±4.2 8.4±2.2 0.6±0.1 39.0±4.5 19.2±1.6 0.2±0.1 13.1±0.2 R8T4 29.0±0.9 16.3±0.2 0.5±0.0 37.5±0.7 16.8±0.5 0.0±0.0 11.8±1.0 R8T3 37.3±0.2 31.9±2.8 0.0±0.0 16.1±2.1 14.7±1.1 0.0±0.0 9.6±1.0 R7T57 36.1±0.3 11.6±0.8 0.5±0.0 41.8±0.7 10.0±0.2 0.0±0.0 10.7±0.9 R7T54 28.0±0.4 12.8±0.2 0.3±0.2 42.4±1.0 16.5±0.6 0.0±0.0 11.4±1.1 R7T52 31.2±0.4 18.5±0.9 0.5±0.0 32.5±0.9 16.8±0.4 0.5±0.1 17.3±0.6 R7T48 29.0±0.8 11.8±0.5 0.6±0.0 40.2±1.6 18.0±0.5 0.4±0.0 15.1±0.5 R7T34 38.5±2.3 7.1±6.5 0.7±0.1 34.2±2.8 18.9±1.6 0.6±0.1 20.5±1.6 R7T19 28.0±1.1 16.2±1.6 0.5±0.1 33.5±0.6 20.7±2.8 1.0±0.4 9.4±0.7 R7T16 35.3±3.3 12.3±11.6 0.6±0.1 27.0±5.4 23.2±2.7 1.7±0.2 15.7±1.7 R7T15 23.4±0.2 6.6±0.1 0.5±0.0 37.0±0.9 30.9±1.0 1.5±0.2 13.3±2.0 R7T7 22.1±0.3 14.3±0.6 0.4±0.0 47.7±1.1 14.9±0.2 0.5±0.0 13.8±0.8 R6T18 30.8±0.1 20.3±0.1 0.5±0.0 28.0±0.7 19.3±0.7 1.1±0.0 14.4±0.2 R5T35 31.1±0.4 27.0±0.9 0.4±0.0 24.2±1.2 16.6±0.5 0.7±0.1 13.8±1.4 R5T25 28.7±0.4 16.9±0.4 0.5±0.0 28.2±0.3 24.4±0.3 1.4±0.2 9.6±1.3 R3T39 33.8±2.0 9.2±0.5 0.5±0.1 44.9±3.2 11.4±1.0 0.2±0.2 7.1±0.6 aAverage of oil determinations on ripe fruit from 3 fruit per tree by GC-FID.

116

CHAPTER 5 SENSORY EVALUATION OF SELECTED ‘HASS’-‘BACON’ AVOCADO HYBRIDS GROWN IN AN EAST CENTRAL FLORIDA CLIMATE

Introduction

The center of origin of wild avocado (Persea americana) trees is considered to be the humid tropical highlands of Central America (southern Mexico, Guatemala,

Honduras) and the three subspecies/races appear to have evolved in different climatic environments isolated from each other geographically (Kopp 1966; Scora et al. 2002;

Litz et al. 2005). Persea americana var. drymifolia race Mexican evolved in the highlands of south-central Mexico and is adapted to tropical highlands (semi-tropical climate). P. americana var. guatemalensis race Guatemalan is adapted to the medium elevations in the tropics and therefore prefers subtropical climate. Lastly, P. americana var. americana race West Indian (or Antillean) is adapted to the lowlands and humid subtropics and therefore grows best in tropical areas (Unknown 1935; Litz et al. 2005).

California and other countries with similar Mediterranean and subtropical climates grow mostly avocados that have adaptations to cooler temperatures and the physiological attributes suited for these areas, such as Guatemalan and Guatemalan-Mexican hybrids. The Florida avocado industry grows cultivars that are more conducive to tropical climates with West Indian and West Indian-Guatemalan hybrid backgrounds

(Litz et al. 2005; Crane et al. 2013).

To compete in the commercial market, avocado fruit has to meet several quality standards. For Florida avocados, maturity standards were set in 1954 based on studies conducted by Stahl (1933) and Harding (1954). Florida maturity standards were set with the help of taste panels, after Harding found that there was a close relationship between

117

fruit maturity, specific picking dates (days from full bloom to harvest), and size and weight within a variety (Lee 1981). The Florida industry is divided into three seasons: early (late May-August), mid (September-October), and late (November-March) (Crane et al. 2007). Cultivars with West Indian background predominate the market during the early season. Such popular cultivars, ‘Bernecker’ must weigh 397-623 g and size 82-86 mm in diameter, ‘Donnie’ 397-680 g and 82-84 mm in diameter, ‘Simmonds’ 453-963 g and 78-98 mm in diameter. Cultivars with Guatemalan-West-Indian hybrid background predominate the mid and late seasons such as ‘Beta’ 453-680 g and 84-89 mm in diameter, ‘Choquette’ 510-1133 g and 95-111 mm in diameter, ‘Hall’ 510-850 g and 89-

98 mm in diameter, ‘Lula’ 397-680 g and 81-105 mm in diameter, and ‘Monroe’ 453-

1133 g and 92-111 mm in diameter (Crane et al. 2007; Crane et al. 2013).

California avocado maturity standards are quite different from those used in

Florida. Avocados with Mexican and Guatemalan backgrounds have much higher oil content (up to 30% on some varieties) when compared to those with West Indian backgrounds (Lee et al. 1983; Gibson 1984). In 1925, California set its first avocado quality standards when growers established a minimum requirement of 8% oil content needed before avocados could be harvested (Anonymous 1925; Lee 1981; Obenland et al. 2012). A 16-year study led by Hodgkin (1939) found that each individual variety was rated higher by taste panels when oil content was higher. However, there is a huge variability among cultivars, and varietal differences in accumulation rates make it difficult to use oil content alone as a maturity index. Morris and O’Brien (1980) observed that there was a correlation between oil content and dry weight and saw that during maturation, the increase in percent dry weight was primarily due to the increase in

118

percent oil content. Additionally, determining dry weight was much easier and practical for growers, and new maturity regulations based on dry weight were established for

New South Wales, Australia. Percent dry matter (or water content in South Africa) has subsequently been the main maturity index in most avocado producing areas, except

Florida. In 1983, California adopted a maturity index standard based on dry matter content and specific release dates depending on variety. These were further modified in

1990, with maturity standards dictating that fruit must have a minimum of 18.4 to 21.9% dry weight of flesh, and have specific size, depending on cultivar (Kader 2002; http://ucavo.ucr.edu/General/Maturity.html). In addition to physiological maturity, other quality factors such as shape, texture, skin and flesh color, lack of decay, and other defects need to be considered in order to meet U.S. grade standards for both Florida and California varieties (https://www.ams.usda.gov/grades-standards/florida-avocado- grades-and-standards).

Having accurate maturity standards is important because when picked too early, avocado eating quality has been described as having a grassy aftertaste, bland flavor, and rubbery and/or watery texture (Yahia & Woolf 2011; Obenland et al. 2012). In

California, major cultivars such as ‘Bacon’ and ‘Hass’, both Guatemalan-Mexican hybrids (crosses used in this study), must have minimum dry matter contents of 17.7% and 20.8% respectively (Yahia & Woolf 2011). More recently, fruit flavor components have been studied by analyzing flesh volatiles. Terpenes and aldehydes have been found to be some of the more predominant compounds in avocado extracts (Pino et al.

2000). Obenland et al. (2012) showed that avocado eating quality increases as the fruit matures. As avocado fruit maturation progresses, panelists rated fruit to be creamier

119

with a less watery texture, and a less grassy flavor. Volatile analyses of these samples showed that aldehydes are produced as a result of lipid degradation and are associated with fruit that is less mature with higher grassy sensory rating (Obenland et al. 2012). In this study, after extensive postharvest evaluations from reciprocal hybrids of ‘Hass’-

‘Bacon’ grown in Fort Pierce, Florida (see Chapters 2 and 3), several selections demonstrating good horticultural and postharvest qualities were identified. These selections were evaluated in sensory taste panels. Attributes assessed included dry matter and lipid content to determine whether these hybrids grown in an east-central

Florida correspond to the California maturity index.

Materials and Methods

Fruit description

Avocado fruit from a population of ‘Hass’ x ‘Bacon’ and ‘Bacon’ x ‘Hass’ hybrids were obtained in December 2013 as a preliminary sensory study (see Chapter 2 for source of trees and cultural practices) (Pisani et al. 2015). Six selections were chosen based on good horticultural and postharvest traits and were included in the taste panel with store-bought ‘Hass’ (from Mexico) used as a control. Ten fruits for each selection and 10 store-bought unripe ‘Hass’ fruit were ripened at room temperature (~23°C) without added ethylene in the lab until they were 10-20 N or less in firmness as determined by a Stable Micro Systems Texture Analyser (Model TA-XT2i, Scarsdale,

NY, USA) using a flat-plate (5 cm diameter) and a 50 kg load cell. The fruit were measured every other day, at 8 and 12 days postharvest until reaching the fully ripe stage. Only fruits that reached 20-30 N on the Texture Analyser (TA-XT2i, Texture

Technologies Corp., New York), on the day prior to the scheduled date for tasting were used for the taste panel. This panel was used as a background study for the more

120

extensive panels in 2014 and 2015. Fruit was harvested on 28 October 2014, 7

November 2014, and 23 October 2015 from the USDA-ARS US Horticultural Research

Laboratory (USHRL) farm in Fort Pierce, FL and transferred to the UF/IFAS, Indian

River Research and Education Center postharvest lab in Fort Pierce for ripening. Nine selections were chosen in 2014 and four selections in 2015 based on good horticultural and postharvest traits (see Chapter 3). These were included in two taste panels conducted in 2014 with store-bought ‘Hass’ (from Mexico and the Dominican Republic, respectively) used as controls. Two additional taste panels were conducted in 2015 in which ‘Hass’ from the same Florida plot (R2T50) was included as a control, together with store-bought ‘Hass’ (from Chile). The second taste panel in 2015 was a sensory difference testing comparing the Florida grown ‘Hass’ to the store-bought ‘Hass’. Each year, origin of the store-bought ‘Hass’ was based on fruit availability at the time of purchase. Fruit for each selection and store-bought unripe ‘Hass’ were ripened in a controlled room at 22ºC with added ethylene (for synchronized ripening) at 87-95 ppm until fruit started to ripen (get softer) and then placed in a cold room (10ºC) until all reached ideal ripeness for the panels (20-30N). When ripe, fruit were transferred to the nearby USDA-ARS USHRL sensory lab for washing, sanitizing, and sensory evaluation.

Fruits were washed with 200 mL of commercial fruit detergent (Fruit Cleaner 395, JBT

Food Tech, Lakeland, Florida) per ~10 L lukewarm water, followed by a 3 min. dip for sanitization with 100 ppm peroxyacetic acid (PAA) (Peraclean® 15, Degussa, Ont.,

Canada). Fruit were air-dried for at least two hours before placing at 13.5ºC until sensory evaluation the next day.

121

General fruit data was collected to supplement postharvest data (Table 5-1; 5-2;

5-3). Postharvest data included fruit and pulp firmness, exterior and pulp color, and lipid content. (See Chapter 3 for general postharvest color and firmness methods; Chapter 4 for fatty acid extraction methods).

Sensory evaluation

Panelists for the preliminary panel in 2013 consisted of personnel from the

IRREC and USDA ARS HRL as well as California and Florida commercial avocado and citrus industry representatives. The panel consisted of ten panelists and was not a consumer panel, due to its small size and level of panel experience. This panel served as a background study to identify which descriptors would best characterize avocados from this study.

In 2014 and 2015, panelists consisted of personnel from the IRREC and USDA

ARS USHRL, as well as Florida commercial avocado and citrus industry representatives and consisted of 55 panelists each year. Fruit was prepared just prior to tasting by cutting each avocado vertically from the stem to blossom end, separating the halves, and removing the seed. Flesh at the stem and blossom ends, above and below the seed was removed and the remaining portions were peeled and cubed. Three pieces (~

2 cm3 each) were placed into 30 mL plastic cups that were labeled with three-digit random numbers for each selection, and served at room temperature. The tasting was conducted in individual booths and under red lighting. Panelists rated overall preference using a 1 to 9 hedonic scale with 1 being dislike extremely and 9 being like extremely.

Then, they completed a multiple choice questionnaire to best describe each sample.

Textural descriptors included: firm, mushy, stringy, gritty, creamy, smooth, dry, watery, and oily. Flavor and aromatics descriptors included: bland, grassy, woody, piney-

122

terpiney, sweet, fruity, nutty, buttery, savory, oily-fatty, and rancid. Those descriptors were selected based on previous research in California and the preliminary panel in

2013 (Obenland et al. 2012; Pisani et al. 2015). Panelists were also instructed to take a bite of carrot or cracker and a drink some water to rinse the palate between each sample (Obenland et al. 2012).

Sensory difference test: Tetrad. As fruit from Florida-grown ‘Hass’ were not available to serve panelists in 2014, and harvest date differed from the selections tested, store-bought ‘Hass’ fruit were presented to panelists. However, in 2015, both

Florida-grown and store-bought ‘Hass’ fruit were available. Therefore a separate taste panel was performed to compare Florida-grown and store-bought ‘Hass’. For this, a

Tetrad testing was used. In that test, samples of each Florida-grown and store-bought

‘Hass’ were presented in duplicate with different code numbers. Panelists were instructed that they had to select the two samples that were identical among the four samples (two pairs) presented. Results were analyzed using published tables (Bi &

O’Mahony 2013).

Statistical analysis

Sample serving was arranged in a William’s design (balanced block) with each selection representing a treatment and the panelists as the replicates. Data collection and analysis were performed using Compusense five® sensory software (Guelph, Ont.,

Canada). Differences between treatments (selections) were calculated using the

Tukey’s HSD tests (P = 0.05). In addition, the fatty acid composition for each selection was evaluated by multivariate analysis using Principal Components Analysis (PCA)

(Chabanet 2000). PCA data analyses were performed using XLSTAT 2014 software

(New York, NY, USA). PCA reduces the number of original variables into fewer principal

123

components (unobservable variables) that are linear combinations of the original ones

(Borgognone et al. 2001). The PCA was performed on the variance-covariance matrix in order to give more weight to descriptors with a wider range of scores. The objective of the PCA was to explain as much of the variability of the original data with as few of the original principle components as possible.

Results

Fruit description

In 2013, fruit varied from 99.8 mm length and 65.9 mm width (R8T23) to 142.0 mm length and 79.7 mm width (R5T56) (Table 5-1). Mean fruit weight of each selection ranged from 213.3 (R8T23) to 400.9 g (R5T56) with mean flesh weight ranging from

147.7 g (R8T23) to 356.7 g (R5T56). Trees produced as little as 55 fruits (R5T56) to as many as 225 fruits (R8T36).

Fruit varied from 80.8 mm length and 62.1 mm width (R8T18) to 148.1 mm length and 79.4 mm width (R5T56) in 2014 (Table 5-2). In 2015, fruit varied from 82.0 mm length and 65.4 mm width (R8T18) to 138.0 mm length and 78.4 mm width (R5T56)

(Table 5-3). In 2014, mean fruit weight of each selection ranged from 152.9 g (R8T5) to

449.8 g (R5T56) with mean flesh weight ranging from 118.8 g (R8T18) to 412.4 g

(R5T56) (Table 5-2). In 2015, mean fruit weight ranged from 168.4 g (‘Hass’ Chile) to

387.4 g (R5T56) with mean flesh weight ranging from 129.8 g (R8T18) to 346.4 g

(R5T56) (Table 5-3). Field and postharvest data on Florida grown ‘Hass’ (R2T50) were not collected as it was only used as a control in the tetrad taste panel. Trees produced as little as 12 and 20 fruits (R5T56) during both years, to as many as 284 fruits (R8T5) in 2014 and 212 fruits (R8T18) in 2015 respectively. The six selections in 2013, nine selections in 2014, and four selections in 2015 in this study were chosen for sensory

124

evaluation because earlier evaluations demonstrated their fruit developed acceptably low incidence of postharvest disorders and rots and had acceptable seed to flesh ratios compared to the other selections (see Chapter 3).

Peel and pulp color measurements of ripe fruit were taken in 2014 and 2015 to compare with commercial ‘Hass’. Selections in 2014 had peel that was green-yellow in color when ripe (h˚ 97.7-123.0) similar to the ‘Hass’ samples from the Dominican

Republic (h˚ = 96.7), except for R5T56 that was more yellow than green (h˚ = 60.5) and was more comparable to ‘Hass’ samples from Mexico (h˚ = 67.3) (Table 5-4). There was no significant difference in pulp color among the selections. Lightness (L*) and chroma

(C*) values indicate that all fruit selections had very dark peel color and were comparable to ‘Hass’ except R7T54, R8T18, and R8T9 were significantly different in L* and R818 in C* (Table 5-4). Selections in 2015, including store-bought ‘Hass’ from Chile and field ‘Hass’ (R2T50), showed no significant differences in peel or pulp color, except for pulp of R5T56 that had the most yellow pulp when compared to other selections (h˚ value closer to 90) (Table 5-5).

There was a gradient in firmness among the selections in 2014 (Table 5-6).

‘Hass’ D.R. (21.3 N) was among the firmest cultivars, and ‘Hass’ from Mexico among the soft selections (12.9 N) together with R5T56 and R8T54 (Table 5-6). There were no significant differences in pulp firmness at both the stem and blossom ends between the selections and ‘Hass’ (Table 5-8). In 2015, ‘Hass’ samples from Chile were the most firm (35.3 N) when included in the sensory panel which may explain why it was liked the least (Table 5-7; 5-9).

125

Fatty acid analysis

Fatty acid composition is described in Tables 5-10 and 5-11. Results are similar to the ones obtained on the whole mapping population described in Chapter 4.

Avocados are rich in the monounsaturated fatty acid oleate (C18:1), which is the most abundant fatty acid found in this study’s fruit during both years (45 and 40%) (Table 5-

10). Palmitate (C16:0), linoleate (C18:2), and palmitoleate (C16:1) are the second, third, and fourth most abundant fatty acids. Stearate (C18:0), linolenate (C18:3), and myristate (C14:0) were only found in small or trace amounts (Table 5-10).

During the 2014 taste panel, store-bought ‘Hass’ from Mexico had higher oleate values compared to ‘Hass’ from the Dominican Republic (51 and 36% respectively)

(Table 5-11). ‘Hass’ from Chile varied significantly in oleate content from the ‘Hass’ grown in Florida (R2T50) in 2015 (64 and 33% respectively) (Table 5-11). Most selections had similar fatty acid composition to the Florida grown ‘Hass’, except R5T56 which was more similar to the store-bought ‘Hass’ from Chile in 2015. PCA was used to reduce our fatty acids dataset into fewer dimensions in order to explain observed similarities and differences. 99.8% of the variance was explained by the first two PCs

(F1 and F2) (Figure 5-1). PCA shows that palmitate and palmitoleate are highly correlated as they are together in the first dimension; R8T21, R8T18, R6T56, and ‘Hass’

D.R. have the highest percentage composition of these two fatty acids. Furthermore, palmitate and palmitoleate are negatively correlated to oleate as they are in the opposite quadrant of the first dimension. ‘Hass’ from Chile, ‘Hass’ from Mexico, R8T9,

R8T5, R5T56, R8T54, and R7T54 are the hybrids showing higher percentage composition of oleate. Linolenate and stearate are in the third dimension and are

126

uncorrelated to the other fatty acids (Figure 5-1). PCA allows us to characterize all our selections using all fatty acids rather than observing them individually.

Sensory evaluation

In 2013, the R8T36 selection and commercial ‘Hass’ cultivar had the highest preference rating (6.5 = like somewhat) of all selections, but there were no significant differences among any of the selections (P = 0.752) for overall preference, the lowest,

R6T29, was rated 5.40 (Figure 5-4A). This similarity in rating might be expected, since the selected hybrids were chosen to be close to ‘Hass’ as a fruit, and the panel size was small, and not representative of any consumer population. However, the high level of acceptance overall supports the potential that high quality new selections may be identified. Each panelist was also asked to rank their four favorite selections: ‘Hass’ was ranked first, followed by R5T56 and R8T36 (data not shown). Over 50% of panelists characterized each of the evaluated selections as creamy, smooth, and firm

(Figure 5-2A). ‘Hass’ was rated as the most creamy and the R8T36 selection was characterized to be equally smooth as ‘Hass’. Only a low percentage of panelists used the terms stringy, gritty, watery, or oily to characterize any of the selections. Two flavor attributes, nutty and buttery, were identified by sensory panelists to be characteristic of the R8T36, R6T29, and R5T56 selections (Figure 5-3A). None of the panelists thought that ‘Hass’ had negative attributes like stringy or gritty, and only 10% characterized it as rancid (Figure 5-2A; 5-3A). However, only three selections were described as stringy or gritty and only by less than 20% of panelists. Therefore, some descriptors were combined on the following years with the large scale panels.

In 2014 and 2015, over 50% of panelists characterized each of the evaluated selections as creamy during both years with the exception of R8T11 and R7T54 in 2014

127

(Figure 5-2B) and ‘Hass’ from Chile in 2015 (Figure 5-2C). Selection R8T18 in 2014 and

R6T56 in 2015 were rated as the most creamy. ‘Hass’ was among the most creamy in

2014. However, ‘Hass’ from Chile was rated as firm in 2015 and was the least liked overall (Figure 5-2C; 5-4C). Only a low percentage of panelists used the terms stringy, gritty, dry, or watery, to characterize any of the selections.

Two flavor attributes, nutty and buttery, were identified by sensory panelists to be characteristic of the R8T18 and R7T48 selections in 2014 (Figure 5-3B). Less than 5% of the panelists thought that ‘Hass’ had negative attributes in 2014 like stringy or gritty, or characterized it as rancid in both years (Figure 5-3A; 5-3B). ‘Hass’ flesh flavor was rated as sweet and buttery in 2014 (Figure 5-3B) and bland and grassy in 2015 (Figure

5-3C). Selections R8T5, R6T56, and R5T56 were characterized to be equally buttery as

‘Hass’ from Mexico.

In 2014, ‘Hass’ from D.R. and Mexico were preferred, followed by R8T18 and

R8T9 and R6T56 and R5T56 in each panel, respectively (Figure 5-4B). The R7T54 selection was liked significantly less than other selections in 2014 (Figure 5-4B). In

2015, the R6T56 selection had the highest preference rating 6.18, but ‘Hass’ from Chile had the lowest ranking rating (4.84 = dislike slightly) (Figure 5-4C). The difference test

(tetrad) showed significant differences when choosing between Florida grown ‘Hass’ and store-bought ‘Hass’ (P = 0.000). Out of 55 panelists, 44 chose the Florida grown

‘Hass’ to the store-bought ‘Hass’ (data not shown).

Discussion

Avocado fruit do not ripen while on the tree and only begin to ripen after harvest.

The lack of change in color and firmness while still on the tree makes identification of

128

fruit physiological maturity difficult. Dry matter, picking date, fruit size, and oil content are some characteristics used as maturity indices depending on cultivar and geographic location. The sensory study conducted in 2013 was considered preliminary and indicated good consumer acceptance of California-type avocados (reciprocal crosses of cultivars ‘Hass’ and ‘Bacon’) grown in an east-central Florida climate. In the current study, dry matter of the avocado selections ranged between 19.9% (R7T36) and 27.9%

(R8T36) in 2013 (Table 5-1), 18.4% (R7T54) and 25.7% (R6T56) in 2014 (Table 5-2), and 20.0% (R8T18) and 26.0% (Hass Chile) in 2015 (Table 5-3). Dry matter percentage of store-bought ‘Hass’ was 23.6% (Dominican Republic) and 24.0% (Mexico), respectively in 2014 and 26.0% (Chile) in 2015. All of these are above the minimum maturity standards listed for California ‘Bacon’ and ‘Hass’ (17.7% and 20.8% dry weight, respectively) (Yahia & Woolf 2011). Dry weights of the current selections corresponded to those of California ‘Hass’ avocados harvested in April-May (Obenland et al. 2012). In

California, avocados can be stored on the tree and ‘Hass’ harvest can occur from April through October, while ‘Bacon’ is harvested November through March. Based on the growth data of the current hybrids, fruit stopped growing at the beginning of October in all three years (see Chapter 2 Figure 2-7) and we can estimate the range of maturity dates of our trees to be between October and December, before California ‘Hass’ fruit is mature. Dry matter content increases with later picking dates (longer tree stored fruit) and once harvested, fruit maturation was observed to be associated with creamier, less watery texture, and less grassy flavor (Obenland et al. 2012). The current results were similar to other published reports (Obenland et al. 2012) where the most watery

129

selections (e.g., R7T54) have the lowest dry matter content (18.4%) and are the firmest

(28.3N) (Table 5-6).

Avocado fruit is one of the most important natural sources of monounsaturated fatty acids such as oleate and its low content of saturated fatty acids makes it an excellent source of healthy fat (Ozdemir & Topuz 2004). Lipid concentrations range from 15-30% in cultivars such as ‘Hass’, ‘Bacon’, and ‘Fuerte’ (Bora et al. 2001). In the current study, total oil content ranged from 10-17% (Table 5-11). Oleate is reported as the most abundant monounsaturated fatty acid, which was also found true in the current study. In ‘Hass’, oleate was reported to range from 47.2-59.5% depending on harvest date (Ozdemir & Topuz 2004). In the current study, oleate ranged from 19.4-63.5% and was very similar to other studies that found oleate, palmitate, palmitoleate, and linoleate to be the major fatty acids found in avocado pulp (Ratovohery et al. 1988; Bora et al.

2001; Moreno et al. 2003; Ozdemir & Topuz 2004; Pacetti et al. 2007). Oleate is a monounsaturated fatty acid known to lower “bad” cholesterol (low density lipoprotein) and is the most abundant fatty acid in avocado mesocarp. Linolenate, stearate, and myristate were found in very low numbers or trace amounts (Table 5-11). Similar results were also found in the fatty acid analysis for the whole mapping population (see

Chapter 4). Previous reports reported 6-8 common fatty acids in pulp of avocado, and are the ones analyzed in the current study, but another study by Bora et al. (2001) has identified as many as 22 fatty acids in avocado mesocarp.

In 2014, about 30% of panelists rated store-bought ‘Hass’ (Mexico and D.R.) as having the most grassy flavor (Figure 5-3B). In 2015, the same was true for store- bought ‘Hass’ (Chile) and R5T56 (Figure 5-3C). Selection R7T54 in 2014 and store-

130

bought ‘Hass’ in 2015 were least liked by panelists (Figure 5-5) which can be explained by their higher firmness values (Table 5-6; 5-7). Additionally, R7T54 was rated as the most firm (70%) and watery (40%) by panelists (Figure 5-2B). Such a difference in firmness values may be attributed to ripening duration of fruit. Fruit that is less mature tends to have been associated with grassy aftertaste, bland flavor, rubbery and/or watery texture (Yahia & Woolf 2011; Obenland et al. 2012). With those exceptions, there were no significant differences in overall liking among the selections in 2014 and

2015 (Figure 5-5). The positive feedback from panelists is an overall indication that new selections with good qualities may be identified.

Oil content of fruit varies with cultivar and ecological origin. Cultivars with

Mexican backgrounds are reported to have the highest oil content of all races with up to

30% (Purseglove 1968). Guatemalan cultivars range from 10% to 13% and West Indian types have the lowest oil content of 2.5% to 5% (Bora et al. 2001). The selections in this study ranged from 9.5% oil content (Hass D.R) to 17.3% (R8T54) in 2014 and 2015

(Table 5-11) and met the minimum 8% oil standard used in California-type cultivars. In a

1981 study, ‘Bacon’ cultivar was shown to have between 18.8% and 20.0% dry weight at 8% oil in different southern California locations (Lee et al. 1983). Similarly, ‘Hass’ was shown to have between 18.8% and 20.6% dry weight. Acceptable taste was noted for

‘Bacon’ at minimum of 20.0% dry weight and at 22.8% for ‘Hass’ (Lee et al. 1983). All of the selections evaluated in the current study met the minimum acceptable dry weight percentages throughout all three years (Table 5-1; 5-2; 5-3). It was shown that dry weight and oil content were highly correlated and that percent oil for acceptable taste was 8.7% for ‘Bacon’ and 11.2% for ‘Hass’ (Lee et al. 1983).This same study also

131

showed that acceptable taste did not vary from year to year but it did vary among locations in Southern California. In the current study, overall liking of the selections that were sampled for more than one year (R5T56, R8T18, and R6T56) did not vary from year to year (Figure 5-4). All our selections (‘Hass’-‘Bacon’ hybrids) had dry weight that ranged between 19.9% and 27.9% in 2013, 18.4% and 25.7% in 2014, and 20.0% and

26.0% in 2015. These values were well within the range of accepted values for these individual cultivars when grown in a California climate and were shown to be acceptable when grown in an east-central Florida environment. In 2013, selections that were rated as firm, were also rated as dry, and watery (R7T36, R7T21, R8T23, R8T36, and R5T56)

(Figure 5-2A). R7T54 was rated as firm, watery, and bland in 2014 (Figure 5-2B; 5-3B) and ‘Hass’ (Chile) was rated as firm, dry, bland, and grassy in 2015 (Figure 5-2C; 5-3C) and were the selections that were liked the least (Figure 5-4B; 5-4C). Oil content and dry matter for R7T54 (11.8% and 18.4%, respectively) (Table 5-2; 5-11) was among the lowest and ‘Hass’ (Chile) (16.0% and 26.0%, respectively) was among the highest in oil content and the highest in dry matter (Table 5-3; 5-11). Therefore, palatability cannot be solely related to oil and dry matter from results of this study.

Based upon results of this study, the selections evaluated appear to have fruit quality similar to commercial ‘Hass’. ‘Hass’ from Chile was least liked, rating as bland and grassy. Firmness tables show that fruit had not yet reached optimum ripeness stage (Table 5-7; 5-9). We hope to identify suitable selections for production in east- central Florida and address further concerns regarding growing “Hass”-like avocados in the region.

132

Conclusion

In summary, our results agree with previous studies in that taste scores generally decrease with less mature fruit (higher firmness values) (Lee et al. 1983; Obenland et al. 2012). Avocados that were most liked were described as creamy in texture, and buttery and nutty in flavor. None of our selections were disliked except R7T54 in 2014 and store-bought ‘Hass’ (Chile) in 2015, which may be due to the lack of ripeness.

R7T54 had the lowest dry weight (18.4%) and among the lowest oil content (11.8%).

However, the opposite is true for store-bought ‘Hass’ (Chile) that had the highest dry matter (26.0%) and among the highest oil content (16.0%). Fruit that is less mature tends to have been associated with grassy aftertaste, bland flavor, rubbery and/or watery texture. Even though our selections meet minimum requirements set for cultivars

‘Hass’ and ‘Bacon’ in California, whether oil content and dry matter can be used as good maturity indices for California-type avocado hybrids growing in a Florida climate needs further evaluation. Since almost all the selections evaluated were of similar or better acceptability compared to store-bought and Florida grown ‘Hass’, further studies are needed to evaluate the effect of harvest date on oil content and consumer acceptance of these selections. Most noteworthy was the fact that favorability of almost all of our hybrid selections was similar to ‘Hass’ in sensory panels.

133

Table 5-1. Phenotypic fruit data on selected avocado trees in 2013. flesh dry fruit per weighta lengtha widtha Tree weighta mattera tree (g) (mm) (mm) (g) (%) 10/21/13 R8T36 248.4 189.5 100.2 69.3 27.9 225 R8T23 213.3 147.7 99.8 65.9 23.4 66 R7T36 306.8 268.6 132.7 73.2 19.9 117 R7T21 264.1 209.0 101.1 72.9 24.8 133 R6T29 268.3 203.7 106.4 71.1 20.1 102 R5T56 400.9 356.7 142.0 79.7 21.3 55 SD 65.0 73.6 18.7 4.6 3.1 61.0 SE 26.5 30.0 7.6 1.9 1.3 24.9 a Values are means of fruit used in this study. Data on store-bought ‘Hass’ is missing.

134

Table 5-2. Phenotypic fruit data on selected avocado trees in 2014. flesh fruit per weighta lengtha widtha dry mattera Tree weighta tree (g) (mm) (mm) (%) (g) 10/20/14 Hass Mexicob 210.3 182.7 94.8 65.7 24.0 NA R8T9 198.1 164.1 88.6 64.6 21.9 59 R8T5 152.9 121.5 94.3 57.4 22.5 284 R6T56 291.2 249.3 134.1 67.8 25.7 17 R5T56 449.8 412.4 148.1 79.4 24.0 12 Hass D.R.b 203.0 171.5 86.0 67.2 23.6 NA R8T54 180.7 135.8 85.0 62.4 25.0 45 R8T18 161.8 118.8 80.8 62.1 21.4 20 R8T11 163.4 123.4 84.2 62.0 19.4 160 R7T54 208.5 172.8 104.2 63.5 18.4 110 R7T48 217.8 185.8 95.5 66.2 22.1 56 SD 84.6 84.4 21.8 5.5 2.2 88.8 SE 25.5 25.4 6.6 1.7 0.7 26.8 a Values are means of fruit used in this study. bStore-bought ‘Hass’ indicates selections included in two separate sensory panels.

Table 5-3. Phenotypic fruit data on selected avocado trees in 2015. flesh dry fruit per weighta lengtha widtha Tree weighta mattera tree (g) (mm) (mm) (g) (%) 10/20/15 Hass Chile 168.4 145.8 89.5 61.1 26.0 NA R8T21 171.4 144.0 86.0 63.6 23.5 40 R8T18 185.4 129.8 82.0 65.4 20.0 212 R6T56 276.1 226.1 115.1 71.7 22.5 91 R5T56 387.4 346.4 138.0 78.4 22.0 20 Florida Hass 174.4 151.2 92.1 61.8 23.8 NA SD 88.5 83.6 21.8 6.7 2.0 86.2 SE 26.7 25.2 6.6 2.0 0.6 26.0 a Values are means of fruit used in this study.

135

Table 5-4. Peel and pulp color of ripe avocado fruit in 2014. Lightness (L*) Hue (h˚) Chroma (C*) Tree peel pulp peel pulp peel pulp Hass 28.7 bc 79.0 a 67.3 bc 102.6 ab 6.4 cd 35.8 de Mexico Hass 29.8 bc 77.6 ab 96.7 ab 103.3 a 9.3 b-d 38.3 c-e D.R. R5T56 26.2 c 74.0 a-c 60.5 c 94.3 bc 3.0 d 41.1 bc R8T5 29.3 bc 72.1 bc 97.7 a 95.8 a-c 5.5 cd 38.3 c-e R8T54 32.7 ab 73.6 a-c 103.8 a 93.6 a 9.8 a-d 40.2 b-d R6T56 31.3 a-c 75.8 ab 106.3 a 95.0 a-c 9.2 b-d 41.8 bc R8T11 32.8 ab 69.2 c 106.7 a 97.1 a-c 12.9 a-c 35.8 de R8T9 35.4 a 76.7 ab 106.8 a 95.7 a-c 14.0 ac 38.9 c-e R8T18 35.9 a 74.6 a-c 110.0 a 98.0 a-c 19.5 a 35.5 e R7T48 32.4 ab 75.1 a-c 112.6 a 95.5 a-c 11.6 a-d 43.7 ab R7T54 35.9 a 75.6 ab 123.0 a 100.5 a-c 17.4 ab 47.0 ab P values 0.0003*** 0.0036** 0.0001*** 0.007** 0.0009*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Table 5-5. Peel and pulp color of ripe avocado fruit in 2015. Lightness (L*) Hue (h˚) Chroma (C*)

Tree peel pulp peel pulp peel pulp Florida Hass 32.4 a 78.7 a 92.3 a 95.0 bc 14.3 a 45.0 ab Hass Chile 28.2 a 79.9 a 53.2 a 103.8 ab 5.5 a 39.5 b R5T56 26.1 a 77.1 a 64.7 a 93.6 c 3.8 a 44.2 ab R8T18 32.2 a 78.2 a 84.4 a 96.9 bc 12.7 a 41.7 ab R6T56 34.2 a 78.6 a 118.7 a 95.9 bc 17.7 a 46.8 a R8T21 35.0 a 76.3 a 123.8 a 100.3 ab 16.9 a 46.6 a 0.2577 P values 0.2906 NS 0.3864 NS 0.1328 NS 0.0001*** 0.0113** NS Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

136

Table 5-6. Whole fruit firmness of ripe avocado fruit in 2014. Tree Peak Force (N) Mean Force (N) HASS D.R. 21.3 ab 10.1 ab HASS Mexico 12.9 c 6.6 bc R5T56 11.6 c 5.5 c R6T56 15.6 bc 8.1 bc R7T48 17.8 bc 8.5 bc R7T54 28.3 ab 12.9 a R8T11 21.9 ab 9.9 ab R8T18 15.2 bc 7.8 bc R8T5 17.9 bc 9.1 a-c R8T54 13.4 c 6.4 bc R8T9 15.5 bc 7.8 bc P values <.0001*** <.0001*** Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

Table 5-7. Whole fruit firmness of ripe avocado fruit in 2015. Tree Peak Force (N) Mean Force (N) Florida Hass 30.4 ab 15.4 a Hass Chile 35.3 a 15.3 ab R5T56 17.9 c 9.5 b R6T56 23.4 bc 13.3 ab R8T18 21.3 bc 10.7 ab R8T21 20.3 bc 10.5 ab P values 0.0006*** 0.0148* Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

137

Table 5-8. Pulp firmness of ripe avocado fruit in 2014. Peak Force (N) Mean Force (N) Tree blossom stem blossom stem Hass D.R. 4.4 ab 4.5 ab 3.5 ab 3.2 b HASS Mexico 2.9 b 2.5 b 2.6 b 1.9 b R5T56 3.7 b 3.3 b 2.5 b 2.6 b R6T56 2.6 b 4.5 ab 2.1 b 3.5 ab R7T48 2.9 b 3.9 b 1.9 b 2.9 b R7T54 4.0 b 9.0 a 3.3 b 6.8 a R8T11 5.0 ab 4.4 ab 3.7 ab 3.5 ab R8T18 2.5 b 2.7 b 2.0 b 1.8 b R8T54 3.7 b 4.2 ab 2.7 b 3.0 b R8T5 4.0 b 3.9 b 3.2 b 3.0 b R8T9 3.0 b 3.4 b 2.3 b 2.3 b P values 0.121 NS 0.060 NS Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05). Firmness was taken at the stem end and blossom end of cross section slices.

Table 5-9. Pulp firmness of ripe avocado fruit in 2015. Peak Force (N) Mean Force (N) Tree blossom stem blossom stem Florida Hass 8.7 4.8 6.9 3.9 Hass Chile 33.4 26.9 19.7 15.3 R5T56 6.4 21.8 4.9 14.1 R6T56 3.8 7.8 2.6 6.3 R8T18 3.4 3.2 2.9 2.4 P values 0.1299 NS 0.0663 NS Means followed by the same letter in the column do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

138

Table 5-10. Average fatty acid (FAME) percentage composition (±SE) of avocado pulp oils by GC-FID in 2014 and 2015 taste panels. % total fatty acid* Fatty Acid 2014 N=57 2015 N=32 Saturated fatty acids: Tetradecanoic acid myristate C14:0 Tr Tr Hexadecanoic acid palmitate C16:0 24.89 ± 0.50 25.91 ± 1.37 Octadecanoic acid stearate C18:0 0.49 ± 0.02 0.45 ± 0.02 Monounsaturated fatty acids:

9-hexadecenoic acid palmitoleate C16:1 10.71 ± 0.49 11.63 ± 1.25 9-octadecenoic acid oleate C18:1 44.93 ± 0.94 40.01 ± 2.83 Polyunsaturated fatty acids:

9,12-octadecadienoic acid linoleate C18:2 19.23 ± 0.59 21.08 ± 0.75 9,12,15-octadecatrienoic acid linolenate C18:3 0.15 ± 0.04 0.91 ± 0.05 *Tr - Traces (Concentration less than 0.02% of the total fatty acids).

Table 5-11. Average fatty acid (FAME) percentage composition (±SE) of avocado pulp oils of individual selections by GC-FID in 2014 and 2015 taste panels. 2014 palmitate palmitoleate stearate oleate linoleate linolenate oil content HASS Mexico 25.0±1.86 11.9±2.08 0.4±0.09 51.3±4.51 11.4±0.73 0.1±0.07 14.2±2.09 R8T9 23.5±0.07 6.4±0.52 0.6±0.03 51.4±0.59 18.1±0.03 0.0±0.00 13.6±0.29 R8T5 23.2±0.52 8.2±0.40 0.6±0.01 49.3±1.93 18.4±0.97 0.3±0.06 11.8±0.58 R6T56 29.2±0.07 14.5±1.89 0.6±0.00 39.7±0.31 15.9±1.46 0.2±0.04 16.6±0.83 R5T56 18.1±0.84 4.9±0.53 0.6±0.02 50.8±1.68 25.5±1.30 0.3±0.05 12.8±1.07 HASS D.R. 31.3±1.08 14.8±2.35 0.2±0.17 36.6±0.47 17.2±1.92 0.0±0.00 9.5±1.78 R8T54 22.6±0.20 7.0±0.02 0.6±0.00 49.2±2.52 20.3±2.05 0.3±0.29 17.3±0.60 R8T18 27.2±1.33 15.5±0.02 0.5±0.03 35.1±0.84 21.5±0.38 0.2±0.16 13.2±0.47 R8T11 29.3±1.33 10.6±0.50 0.5±0.05 34.4±1.14 25.0±0.25 0.0±0.00 10.2±0.41 R7T54 22.7±1.12 11.3±1.51 0.5±0.01 45.1±1.40 20.1±0.90 0.3±0.33 11.8±0.25 R7T48 26.2±0.77 12.9±0.52 0.6±0.03 37.9±1.24 22.3±0.14 0.1±0.12 13.2±0.78 2015

HASS Chile 17.0±0.99 4.3±0.75 0.3±0.00 63.5±1.41 14.0±0.25 1.0±0.08 16.0±1.83 Florida HASS 32.9±0.27 18.2±1.32 0.5±0.07 22.9±0.92 24.4±0.16 1.1±0.09 9.9±1.24 (R2T50) R8T21 37.2±0.58 20.8±1.14 0.5±0.16 19.4±0.12 21.4±1.58 0.7±0.10 11.7±0.10 R8T18 29.7±0.53 16.0±1.21 0.4±0.03 31.8±0.49 21.4±1.26 0.7±0.02 12.9±0.82 R6T56 29.8±2.16 15.3±0.94 0.5±0.04 32.9±0.15 20.7±2.76 0.8±0.15 13.3±1.20 R5T56 20.1±1.81 3.6±0.28 0.5±0.03 48.6±1.02 26.1±0.30 1.0±0.20 10.4±0.59

139

Figure 5-1. Biplot of the Principal Component Analysis of fatty acid analysis of avocado fruit used in sensory taste panels in 2014 and 2015. The percent contribution to the variance of each factor is given in parenthesis.

140

C Figure 5-2. Percentage of panelists in 2013 (A), 2014 (B), and 2015 (C) characterizing the avocado flesh texture using the indicated descriptors. Store-bought ‘Hass’ served as a commercial standard.

141

C Figure 5-3. Percentage of panelists in 2013 (A), 2014 (B), and 2015 (C) characterizing the avocado flesh flavor using the indicated descriptors. Store-bought ‘Hass’ served as a commercial standard.

142

C Figure 5-4. Overall liking of avocado selection in 2013 (A), 2014 (B), and 2015 (C). Rating based on a 1-9 scale (1 = dislike extremely; 5 = neither like nor dislike; 9 = like extremely). Means followed by the same letter do not differ significantly according to Tukey’s studentized range (HSD) test (P<0.05).

143

CHAPTER 6 SCREENING AVOCADO GERMPLASM FOR RESISTANCE TO LAUREL WILT DISEASE

Introduction

Avocado, Persea americana Mill., is an evergreen subtropical fruit whose archeological evidence shows that it was first domesticated in Mesoamerica where its human consumption was documented as far back as 8000-7000 BC (Smith 1966; Litz et al. 2005). Archeological studies suggest that human selection may have taken place as early as 4000 BC (Smith 1966). Avocado originates from three ecological races or subspecies known by their centers of origin, the Mexican (M) (Persea americana var. drymifolia Blake), Guatemalan (G) (Persea americana var. guatemalensis Williams), and West Indian (WI) (Persea americana var. americana Miller). Genetic marker studies suggest that these three races went through separate domestication and only came into contact with each other in the 16th century following European settlement (Chen et al.

2009).

In the United States, avocado (Persea americana Mill.) is grown commercially in

California and Florida. It is now threatened by laurel wilt, caused by Raffaelea lauricola, a nutritional symbiont of an introduced Asian ambrosia beetle, Xyleborus glabratus

(Fraedrich et al. 2008; Ploetz et al. 2011a). Female beetles carry fungal spores in specialized structures known as mycangia and usually attack stressed or dying trees.

Unfortunately, in their new environment, X. glabratus are also attracted to healthy trees

(Hulcr & Dunn 2011). Laurel wilt affects all Lauraceous hosts such as redbay (Persea borbonia (L.) Spreng.), northern spicebush (Lindera benzoin L.), sassafras (Sassafras albidum (Nutt.) Nees), silkbay (Persea humilis Nash), swampbay (Persea palustris

(Raf.) Sarg.), and avocado (Fraedrich et al. 2008). On avocado, external symptoms

144

include wilting of the leaves that stay attached to the branches up to 2-3 months after initial infection (Ploetz et al. 2012). Foliar symptoms usually develop rapidly within two weeks of infection. Internal symptoms include a red-purple stain on the outer sapwood, which is typical of the Ophiostomaceae fungi such as Ophiostoma novo-ulmi, causal agent of Dutch elm disease (Fraedrich et al. 2008; Harrington et al. 2008).

First seen in Savannah, GA, in 2002 on redbay, laurel wilt was later positively identified on trees in commercial avocado groves in Miami-Dade County (Ploetz et al.

2011a) and in 2015, has also been detected as far west as two counties in eastern

Texas on the related species swamp bay and silk bay (U.S. Department of Agriculture

Forestry Service 2015). Laurel wilt causes the rapid wilting and death of mature trees within a period of 6-8 weeks and poses a significant risk to the avocado industry in

Florida (Inch et al. 2012; Inch & Ploetz 2012). Because of laurel wilt, the USDA-ARS

Subtropical Horticultural Research Station (SHRS) avocado germplasm collection is being treated yearly with a systemic fungicide (Ploetz et al. 2011b; Ayala-Silva et al.

2012) and germplasm trees are being grafted and protected in greenhouses in the

USDA Ft. Detrick, MD facility until budwood can be sent to the USDA-ARS genebank in

Hilo, HI.

Cultivars of the West Indian race of avocado appear to be most susceptible to laurel wilt (Inch et al. 2012) with Guatemalan and Mexican accessions showing greater resistance. To identify parents with heritable resistance to the pathogen, a project was initiated to plant and challenge in Ft. Pierce, Florida, open-pollinated half-sib families from a broad range of accessions in the SHRS germplasm collection. In addition, open pollinated seeds from Mexican and Guatemalan germplasm accessions that are not

145

present in the SHRS collection are being procured from University of California

Riverside for planting at Ft. Pierce and SHRS. Assaying seedlings for tolerance/resistance to the laurel wilt fungus (Raffaelea lauricola) has not yet been optimized. Since the seedlings are open pollinated progeny, we are screening for the heritability of the resistance trait from the maternal clonal cultivar.

Laurel wilt-resistant germplasm is needed for currently threatened production areas, as well as areas where avocado has potential as an alternative crop (e.g. where citrus is impacted by huanglongbing). Half-sib families were generated from diverse open pollinated source trees in the USDA Miami avocado germplasm collection and planted at the U.S. Horticultural Research Laboratory in Fort Pierce, FL. Trees were inoculated with an isolate of R. lauricola during two consecutive years (2013-2014) in the hope of identifying laurel wilt tolerant genotypes of avocado.

Materials and Methods

Half sib populations (open pollinated) from Miami National Clonal Germplasm

Repository (NCGR) was planted at the ARS-USHRL Picos Farm in Fort Pierce, Florida in 2012. Plants were treated with standard fertilizer and irrigation practices (see Chapter

2). The RL4 (CBS 127349) isolate of Raffaelea lauricola (Ploetz et al. 2011b) was used for field studies in 2013 and 2014 to challenge trees and observe disease development.

Repeated inoculations were conducted on surviving trees, and trees were monitored for external symptom development. This was done in the hope of identifying laurel wilt tolerant genotypes of avocado. Two inoculation techniques were used to induce laurel wilt symptoms. On 16 June 2013, 5 mm2 mycelial plugs were obtained from a one week old culture grown on malt extract agar (MEA). Plugs were inserted in a slit cut downward at a 45º angle on the main trunk right under a main branch to infect

146

associated vascular traces and the cut wrapped in Parafilm to protect the wound from other pathogens and environmental stresses. The following two inoculations on 12

September 2013 and 9 June 2014 were conducted by drilling four 3 mm diameter holes at 45º angle down with a portable drill around the trunk under a main branch to infect associated vascular traces and mimic beetle entry holes. 20 µL conidial suspension containing 1 x 106 conidia was pipetted into each hole for a total of 80 µL per tree. The holes were then wrapped in Parafilm for protection.

A total of 517 trees (18 families) (Table 6-1) were screened for tolerance to laurel wilt disease. Plants were rated for external symptoms biweekly after inoculation until temperature was no longer optimal for fungal growth. Canopy affected was rated as a percentage of the total canopy. Mean symptom percentage was calculated for each family and standard error of the difference of the means was calculated to identify families that were either more resistant or more susceptible than the mean for all families. Dead and affected branches were removed prior to subsequent inoculations to avoid rating of the same material for each inoculation.

The generalized linear mixed model (GLMM) GLIMMIX procedure in SAS 9.4 was used to separate mean disease severities for all half-sib families in 2013 and 2014 field experiments.

DNA from all trees was extracted by the USDA Miami laboratory where data will be analyzed by running a bulked segregant analysis to identify alleles associated with resistance and susceptibility. Selected trees showing promise in the ongoing evaluation for laurel wilt resistance will be genotyped using 384 SNP markers developed by Dr.

David Kuhn at Subtropical Horticulture Research Station-Agricultural Research Service

147

(SHRS-ARS), Miami Florida. These markers are evenly spaced across the 12 linkage groups (25 per linkage group, sufficient for quantitative trait loci mapping) of the avocado genome and will include markers specific for flower type prediction. A version of bulk segregant analysis will genotype the three best and three worst individuals from each family in future studies.

Additionally, previous studies by Ploetz et al. (2012) on potted trees of the cv.

‘Simmonds’ show that there is a correlation among tree size and susceptibility to laurel wilt; larger trees are more susceptible to the disease. To see if there was a correlation among cultivars and tree size in the field, tree diameter was taken with a caliper at the inoculation point during both years on the inoculation dates. A regression analysis was performed to see if any correlation existed.

Results

External foliar symptoms began to develop within two weeks of inoculation, regardless of whether mycelial plug or liquid inoculum was used to infect trees.

Symptoms observed were rapidly wilting leaves that turned olive green to brown and remained attached to the branches for up to a month after initial symptoms (Figure 6-1).

Often, only a portion of tree canopy was affected and occasionally, the entire tree succumbed to the disease. Some trees recovered and new shoots grew out near or below the inoculation point. After two inoculations in 2013, the data were combined and analyzed. There were significant differences in subsequent disease severity among families in both 2013 and 2014 (Table 6-1). Throughout this chapter, the name of the seed parent for each family will be used to indicate each half-sib family. During both years, ‘Argui 1’, ‘Booth 8’, and Melendez showed least disease severity and ‘No. 21’ the highest disease severity (Figure 6-2A; Figure 6-3A). This can also be observed with the

148

area under disease pressure curve (AUDPC). Lower AUDPCs represent slower disease progression and greater apparent resistance to the disease, while higher AUDPCs represent faster disease progression and greater apparent susceptibility to laurel wilt disease (Figure 6-2B; Figure 6-3B). The working hypothesis was that West Indian families (e.g. ‘Bernecker’ and ‘Dade’) would be more susceptible based on previous findings using similar inoculation techniques (Ploetz et al. 2012). Mexican (‘La Piscina’),

Guatemalan (‘R14T06’), or G-M hybrids (‘Bacon’, ‘Ettinger’) were expected to be more resistant than pure West Indian or G-WI hybrids. Overall, ‘Melendez’, ‘Booth 8’, and

‘Yon’ (all G-WI hybrids) families had the lowest proportion of trees succumb to the disease, while ‘Ettinger’, a G-M hybrid, had the highest proportion of trees die to laurel wilt without recovery (Figure 6-4). ‘Argui 1’, ‘Booth 8’, and ‘Melendez’ showed lowest disease severity in 2013 and 2014. ‘No. 21’, a G-M hybrid, was most susceptible during both years.

To check for the influence of plant size to disease severity, trunks of each tree were measured with a caliper at the inoculation point both years. Regression plots show that symptom severity was significantly (P < 0.05) and positively correlated with plant size during both years (Figure 6-5), regardless of genotype. However, the R2 were very low, under 0.10 in both years.

Discussion

Expanding avocado production has significant potential, but laurel wilt resistance will be a key factor in future sustainability. A genetic mapping study may determine which genes are linked to key traits for future use in marker assisted selection. The inoculation challenge was conducted in Fort Pierce primarily because there is no local avocado industry that may be affected by laurel wilt inoculations.

149

Avocado (Persea americana Mill.) is a high-value specialty crop commercially grown in California and Florida. The Florida avocado industry is among the state’s largest fruit industry after citrus and contributes nearly $14 million to the local economy and $30 million to the wholesale market with more than 6,773 production acres in

Miami-Dade County alone (Evans et al. 2010). However, the avocado industry is newly threatened by the fungal laurel wilt pathogen (Raffaelea lauricola) vectored by an Asian ambrosia beetle (Xyleborus glabratus), which has devastated native populations of avocado relatives (Fraedrich et al. 2008). Vector management strategies and fungicide applications have so far been effective on a small scale, but may be difficult to apply in commercial settings (Ploetz et al. 2011b). Currently, growers are being advised to closely monitor orchards and scout for beetle activity. Affected trees in orchards should be removed and burned (Crane et al. 2011). Surrounding “healthy” trees are treated with Tilt, propiconazole (Syngenta Crop Protection LLC, Greensboro, NC, USA) to inhibit root graft transmission of the pathogen (Ploetz personal communication). The delivery of the fungicide via macro-infusion is the most effective application measure but is very slow, expensive, and not commercially viable. Laurel wilt-resistant cultivars are needed for currently threatened production areas, as well as areas where avocado has potential as an alternative crop (e.g. where citrus is impacted by huanglongbing). The collection at Miami National Clonal Germplasm Repository (NCGR) in Miami, Florida contains 224 accessions with all three avocado races and their hybrids. Half-sib families were generated from diverse open pollinated source trees in the USDA Miami avocado germplasm collection and planted at the U.S. Horticultural Research Laboratory in Fort

Pierce, Florida. Trees were inoculated with an isolate of R. lauricola during two

150

consecutive years. Repeated inoculations were conducted on surviving trees, and trees were monitored for external symptom development. There are preliminary indications that some avocado genotypes from the more freeze-tolerant Mexican and Guatemalan races are more resistant than the West Indian material which is the foundation of South

Florida production (Ploetz et al. 2012). Screening a broad range of avocado genotypes may help find tolerance in these genomes. Field and greenhouse inoculation studies conducted by Ploetz et al. (2012) show that disease symptoms began to develop within

10-14 days of inoculation, regardless of whether the plants were inoculated with mycelia or conidial suspension. Furthermore, WI cultivars developed more severe symptoms than the pure G, G-M and G-WI hybrids (Ploetz et al. 2012). Additional studies by Ploetz et al. (2015) show that ‘Russell’ (WI) was more susceptible to laurel wilt than ‘Brogdon’

(M-G-WI) and ‘Marcus Pumpkin’ (G) and that there was a relationship between the development of external symptoms and net CO2 assimilation, stomatal conductance, transpiration, water use efficiency, and mean daily sap flow in ‘Russell’. Symptoms were associated with a decrease of these parameters, however, these studies were done using clonal scions grafted onto ‘Waldin’ (WI) seedling rootstocks, and the possibility that rootstocks might influence the susceptibility of the scion to laurel wilt needs further attention. In the current study, foliar symptoms began to develop within 14 days, regardless of whether trees were inoculated with mycelia or conidia. However, some aspects of the results in this study do not agree with previous reports (Ploetz et al.

2012). The most susceptible families were seedlings of ‘No. 21’, a G-M hybrid, with trees showing laurel wilt symptoms in 69% of their canopies in 2013 and 61% in 2014.

‘Ettinger’ and ‘Winter Mexican’ seedlings, also G-M hybrids, had the greatest mortality

151

from laurel wilt disease after the first year’s inoculations. Trees of ‘Argui 1’ (WI-(G-M) and ‘Booth 8’ (G-WI) showed the lowest disease severity with more than 37% of trees displaying laurel wilt symptoms in less than 20% of their canopies, over both years.

Considerable variability was observed in most half-sibling families which most likely reflects multigenic contributions to resistance/tolerance within these very genetically heterogeneous plants. Laurel wilt symptoms are usually limited to mature trees rather than seedlings due to host choice by the beetle vector (Fraedrich et al. 2008; Ploetz et al. 2012). However, trunk diameters of trees in this study were highly variable and diameter measurements were taken at the time of inoculation to observe if a correlation with disease severity existed. To avoid this issue, we directly inoculated the trees with the fungus, but a general recovery of many trees from all families after initial laurel wilt symptoms were observed.

Additional avocado genotypes will be screened for disease tolerance/resistance in the near future. Seedlings from SCFS are currently being generated and will be planted in Fort Pierce. This will allow for comparisons among California and Florida seedlings in finding parent genotypes and associated molecular markers that confer a higher degree of laurel wilt tolerance.

Conclusion

The current study demonstrated that the family ‘No. 21’ was most susceptible to the disease and ‘Ettinger’ had the highest proportion of trees die to laurel wilt without recovery. Both cultivars are G-M hybrids. Lowest disease severity was observed for

‘Argui 1’ (WI-(G-M)), ‘Booth 8’ (G-WI), and ‘Melendez (G-WI). In contrast, Ploetz et al.

(2012) found that cultivars with WI background developed most severe symptoms.

Laurel wilt symptoms are usually limited to mature trees rather than seedlings due to

152

host choice by the beetle vector. Furthermore, regression analysis shows that there is a correlation between disease severity and trunk size, which corroborates with previous reports (Ploetz et al. 2012). Additional inoculations on older trees need to occur to see if disease severity changes among genotypes.

153

A B Figure 6-1. Foliar symptoms of laurel wilt disease on avocado ‘Choquette’ (A) and ‘Winter Mexican’ (B) at the USDA Horticultural Research Laboratory Picos Farm, Ft. Pierce. (photos courtesy of Cristina Pisani)

154

Table 6-1. Response of different avocado seed-source families to laurel wilt disease.a Disease severity (% canopy affected)b Families No of trees Genomec 2013d 2014 Argui 1 34 WI x (G x M) 29.8 c 17.3 d Booth 8 27 G x WI 33.1 bc 22.2 d Melendez 37 G x WI 35.2 a-c 18.1 b-d Semil 38 16 G x WI 38.1 a-c 35.6 b-d Bernecker 44 WI 38.9 a-c 25.7 b-d Bacon 6 G x M 38.2 a-c 17.0 d R14T06 26 G 42.2 a-c 34.7 b-d Yon 32 G x WI 46.2 ab 32.5 b-d Dade 20 WI 51.2 a 41.9 a-c Choquette 32 G x WI 53.8 a 31.0 b-d Suardia 37 G x WI 47.6 ab 46.0 a-c No. 21 7 G x M 61.6 a 69.4 a Winter Mexican 40 G x M 49.8 a 41.3 a-d Gripina 24 WI x (G x WI) 52.0 a 23.8 cd La Piscina 16 M 52.5 a 48.8 ab Booth 7 46 ? 54.3 a 43.7 a-c Ettinger 40 G x M 55.7 a 40.6 b-d Borrego 33 G x WI 56.5 a 40.7 a-c aTrees were planted in 2012 and inoculated with Raffaelea lauricola in 2013 and 2014. Disease was rated biweekly after inoculation. Results are from field experiments conducted at ARS-USHRL, Fort Pierce, Florida. bMean external disease severity rated as a percentage of canopy affected. Means were compared using generalized linear mixed model; means followed by the same letter are not significantly different (P < 0.05). cGenome indicates the background of the cultivar: Guatemalan (G), Mexican (M), West Indian (WI). dIncludes mean disease severity of two inoculations.

155

B Figure 6-2. Mean disease severity to laurel wilt in 2013 rated as a percentage of canopy affected (A), area under disease pressure curve (B). Lower AUDPCs represent slower disease progression and greater resistance to the disease; higher AUDPCs represent faster disease progression and greater susceptibility to laurel wilt disease.

156

B Figure 6-3. Mean disease severity to laurel wilt in 2014 rated as a percentage of canopy affected (A), area under disease pressure curve (B). Lower AUDPCs represent slower disease progression and greater resistance to the disease; higher AUDPCs represent faster disease progression and greater susceptibility to laurel wilt disease.

157

30 a 25 ab ab ab 20 ab ab ab ab 15 b b b 10 b b b b b b 5 b Mortality(%) b 0

Figure 6-4. Percentage of tree mortality after disease severity ratings.

158

B Figure 6-5. Relationship between stem diameter of avocado families and the severity of laurel wilt that developed after field inoculations in 2013 (A) and 2014 (B).

159

CHAPTER 7 PRELIMINARY STUDIES ON DEVELOPING A HIGH THROUGHPUT SCREENING ASSAY USING QUANTITATIVE PCR FOR DETECTING RAFFAELEA LAURICOLA, CAUSAL AGENT FOR LAUREL WILT DISEASE, IN PERSEA AMERICANA SHOOT CUTTINGS

Introduction

The laurel wilt pathogen, Raffaelea lauricola, has had a huge ecological effect on plant communities in the United States. Since its vector was first discovered in Port

Wentworth, GA in 2002, the disease has had a devastating impact on native redbay,

Persea borbonia and other native plants that belong to the family Lauraceae (Gramling

2010). Such hosts affected by laurel wilt include swampbay (P. palustris), silk bay (P. borbonia var. humilis), pondberry (Lindera melissifolia), camphor tree (Cinnamomum camphora), pondspice (Litsea aestivalis), and avocado (P. americana) (Fraedrich et al.

2008; Crane et al. 2008). Raffaelea lauricola is a fungal symbiont of the redbay ambrosia beetle, Xyleborus glabratus. Similar to other ambrosia beetles, X. glabratus carries fungal spores in its mycangia and cultivates the fungus in gallery systems of wood to use as food for their larvae (Fraedrich et al. 2008). In less than two years, more than 92-100% of the redbays (depending on size) were killed (Dreaden et al. 2014a).

Since its first discovery in 2002, laurel wilt has been identified in other counties in North and South Carolina, Georgia, Alabama, Mississippi, Louisiana, and as far south as

Miami-Dade County, Florida (Ploetz et al. 2011a; Ploetz et al 2013; U.S. Dep. Agric.

For. Serv. 2015) Avocado, Persea americana, and bay laurel, Laurus nobilis, are the few members of the Lauraceae family that are of commercial importance. Florida is the second most important producer of avocado in the United States (ca. $55 million per year) after California (ca. $350 per year) (Ploetz et al. 2013). For global commerce, the

160

biggest concern is that the disease will spread to California and eventually Mexico, the world’s lead producer (Ploetz et al. 2013).

Currently, R. lauricola-infected wood is identified by attempted isolation on a variant of the semi-selective medium containing cycloheximide and streptomycin malt extract agar (CSMA) developed by Harrington (1981), CSMA+. CSMA+ contained 0.6 g cycloheximide, 0.3 g streptomycin, 0.25 g ampicillin, and 0.005 L-1 rifamycin (Ploetz et al. 2012). Alternatively genomic DNA is amplified by PCR procedure and sequencing of the 18S or 28S ribosomal DNA is used to identify the presence of the pathogen

(Dreaden et al. 2014; Jeyaprakash et al. 2014). Drawbacks such as false positives can occur. Unfortunately, documented small subunit (SSU; 18S) markers are not taxon- specific at the species level for R. lauricola but are highly sensitive in indicating presence other fungal families and can detect very small amounts of DNA (up to 0.0001 ng per PCR) (Dreaden et al. 2014). Small sequence repeat markers (SSRs) developed by Dreaden et al. (2014) are R. lauricola-specific and can be used to identify the pathogen, but can only detect up to 0.1 ng of pathogen DNA (Jeyaprakash et al. 2014;

Dreaden et al. 2014). The discrepancy in detection limits between the SSU and the microsatellite markers may be due to the difference in copy number between the PCR targets as rDNA loci have multiple copies while microsatellite loci are low in copy number and require more genomic DNA template for the PCR to reach detection limit

(Dreaden et al. 2014). In this study we wanted to use the approach previously used by

Jeyaprakash et al. (2014) to develop a high throughput screening assay using quantitative real time PCR (qPCR) to be able to quantify pathogen DNA in shoot cuttings of avocado. If successful, this assay could help screen for proliferation of R.

161

lauricola as a possible test of resistance as well as screen for the presence or absence of the laurel wilt pathogen very quickly and inexpensively helping in field detection and preventing further spread.

Materials and Methods

Preliminary study

Preliminary tests were conducted to see it would be possible to inoculate avocado shoot cuttings with the laurel wilt pathogen, Raffaelea lauricola, and recover the pathogen at a range of distances from the inoculation site. The RL4 (CBS 127349) isolate of Raffaelea lauricola (Ploetz et al. 2011b) was used for this study. Inoculations were done on shoot cuttings from ‘Hass’-‘Bacon’ hybrids (G-M), ‘Lula’ (WI), and ‘Marcus

Pumpkin’ (G) to include a range of genetic backgrounds. Shoot source trees were located at USDA-ARS Picos farm and UF-IFAS IRREC, Fort Pierce, Florida, respectively. Thirty-six ‘Hass’-‘Bacon’, Thirty-two Lula, and fourteen Marcus Pumpkin shoot cuttings were collected, approximately 20-25 cm in length. Cuttings were washed under water to remove sooty mold. Leaves were removed except for 2-3 apical leaves that were cut in half to permit transpiration without excessive water stress (Figure 7-1).

Shoots were placed in 25 x150 mm glass tubes containing approximately 12 mL water and 0.75 mL/L Kathon Plant Preservative Mixture (PPM) (Sigma-Aldrich, USA) as a biocide. Cuttings were inoculated with a 5 mm2 mycelial plug obtained from a one week old culture grown on malt extract agar (MEA). Plugs were inserted in a slit cut at 45º angle and wrapped in Parafilm to protect the wound. A clean malt agar plug and uncut cuttings were used as controls. Cuttings were then placed in a growth chamber set at

28˚C for 16 hour days. After 2 weeks, shoots were cut longitudinally and vascular staining, a typical laurel wilt symptom, was measured from the inoculation point with a

162

ruler. Disease severity was assessed based on the following scale: 0 = absent (no vascular staining), 1 = slight (up to 1cm above plug), 2 = present (from 1-5 cm above plug), 3 = severe (> 5 cm above plug), 4 = dead (dead cutting). Pieces of wood tissue were removed above (8 and 16 cm) and at the inoculation point, surface disinfected, and plated on a semi-selective malt agar medium containing 0.6 g cycloheximide, 0.3 g streptomycin, 0.25 g ampicillin, 0.01 g/mL rifampicin (CSMA+) to isolate the pathogen

(Ploetz et al. 2012). Plates were observed after 2 weeks for growth of the fungus which was phenotypically identified as having hyaline conidiophores (mostly aseptate and unbranched) sometimes septate at branches, with hyaline blastospores forming as clusters at tips of conidiophores, cottony white colony with a slimy yeast phase at the center as described by Harrington et al. (2008). Results from these sets of experiments would allow us to evaluate whether a similar experiment could be conducted by assaying samples with a real-time qPCR assay to determine the presence of Raffaelea lauricola.

Screening using qPCR assay

Plant material

In this experiment shoot cuttings of ‘Hass’ (G-M) and ‘Waldin’ (WI) cultivars were used. Shoots were obtained from trees located at the USDA-ARS Picos farm in Fort

Pierce, FL. One hundred and twenty cuttings of each cultivar approximately 20-25 cm in length were collected and were brought back to the lab for processing. Leaves were removed except for two at the apex that were cut in half to facilitate transpiration. Shoot cuttings were placed in 25 x150 mm glass tubes containing approximately 12 mL water and 0.75 mL/L Kathon Plant Preservative Mixture (PPM) (Sigma-Aldrich, USA) as a biocide (Figure 7-1). Inoculation was done by drilling a 3 mm diameter hole with a

163

portable drill, 26 cm from the shoot apex. Conidial suspension was obtained from a 2 week old mycelial colony growing on MEA medium. 20 µL of conidial suspension containing 1 x 106 conidia/mL was pipetted into the hole. Controls were either inoculated with water or not inoculated at all. Cuttings were kept in a growth chamber set at 28˚C for 16 hour days until DNA extraction.

DNA extraction and amplification

DNA was extracted from wood tissue (~100 mg) as described by Jeyaprakash et al. (2014) at three different points: at the inoculation point (0 cm), 8 cm above the inoculation point, and at the shoot apex at 16 cm above the inoculation point. DNA was extracted from 24 different ‘Hass’ and ‘Waldin’ cuttings at time of the inoculation (week

0) and every week thereafter for four weeks. Primers and probe designed by

Jeyaprakash et al. (2014) were not used as the sequence of the amplicon they amplified was unknown for the setup of the standard curve. Primers were designed with

DNASTAR primer design software (Madison, Wisconsin, USA) from the RL4 18S ribosomal RNA gene (CBS 127349) and obtained from a commercial source (Integrated

DNA Technologies, Inc.). Primer pairs used were forward 5’-

CGGCTGGGTCTTGGCCAGC-3’ and reverse 5’-AAGTCTGGTGCCAGCAGCCG-3’ to amplify a 251 bp amplicon. The gene fragment (5’-

CGGCTGGGTCTTGGCCAGCCATGGTGACAACGGGTAACGGAGGGTTAGGGCTCG

ACCCCGGAGAAGGAGCCTGAGAAACGGCTACTACATCTAAGGAAGGCAGCAGGC

GCGCAAATTACCCAATCCCGACGCGGGGAGGTAGTGACAATAAATACTGATACAG

GGCTCTTTAGGGTCGTGTAATTGGAATGAGTACAGTTTAATTCCCTTAACGAGGAA

CAATTGGAGGGCAAGTCTGGTGCCAGCAGCCG-3’) was synthesized and diluted to

10 ng/µL (without avocado DNA) and used to create dilutions for a standard curve. 5 µL

164

of extracted DNA (20 ng/µL) was amplified using 10 pmol each of the forward and reverse primers. Quantitative real time PCR (RT-qPCR) on DNA from Persea americana tissues was performed using the GoTaq qPCR Master Mix (12.5 µL)

(Promega) protocol in a 25 µL volume. The amplification profile included (i) 95˚C for 2 min (1 cycle) and (ii) 95˚C for 15s, 65˚C for 30s, and 95˚C for 15s (40 cycles). The melt curve profile included 60˚C for 1 min, 95˚C for 30s, and 60˚C for 15s. Amplification was performed using an ABI 7500 Detection System (96-well) (Applied Biosystems) and cycle threshold (Ct) value recorded for each sample. The number of molecules for each sample was calculated using the following formula:

Number of copies = (pg * 6.022x1023)/(length * 1x10-12 * 650)

23 Where: pg is the amount of DNA in picograms, 6.022x10 is Avogadro’s number, length is the length of the DNA fragment in base pairs, 1x10-12 is used to convert the answer to picograms, and 650 is the average weight of a single DNA base pair.

Results

Preliminary study

Preliminary experiments showed that avocado shoot cuttings could be kept alive for almost two months when the cut basal end was placed in a water solution containing

0.75 mL/L Kathon Plant Preservative Mixture (PPM) biocide. The goal was to observe whether it would be possible to inoculate shoot cuttings with the laurel wilt pathogen,

Raffaelea lauricola and see if the pathogen would translocate up the shoot and be recovered on semi-selective medium CSMA+. The experiment was carried out using avocado cuttings with different genetic backgrounds. All three experiment sets had very similar results. All controls, whether inoculated with a clean agar plug or just water without any inoculation, showed no recovery of the pathogen on CSMA+. Raffaelea

165

lauricola was recovered as low as 0.6 to as high as15 cm above the inoculation point in

‘Hass’-‘Bacon’ hybrid cuttings (Figure 7-2A). The pathogen was also recovered in the range of 0.2-5.5 cm in Lula (Figure 7-2B), and 1.4-4.1 cm in Marcus Pumpkin (Figure 7-

2C) shoot cuttings. These results proved that it should be viable to inoculate shoot cuttings with the laurel wilt pathogen and use quantitative real time PCR to quantify small amounts of DNA as part of a high throughput screening assay.

Screening using qPCR assay

Shoot cuttings from ‘Hass’, a Guatemalan-Mexican hybrid (G-M) and ‘Waldin’, a pure West Indian (WI) were used in this experiment. A calibration curve was run during each qPCR with 6 different concentrations (2000-0.02 pg) (Figure 7-3). All calibration curves had R2 values close to 1 and were considered acceptable (data not shown).

Large Ct values indicate less DNA titer. Ct values from 17-32 were considered optimal, values 32-37 were accepted but were questionable, and values 37-40 were considered undetectable for pathogen DNA. For most of our samples we did not amplify R. lauricola

DNA (data not shown). However, all our positive controls from extracted DNA from a fungal colony grown on malt agar had Ct values below 37 (Table 7-1). All Ct values and their corresponding number of molecules are reported in Table 7-1. Fungal DNA was recovered at the apex (16 cm) of a ‘Hass’ cutting at week 0 (Ct = 10.1). DNA was also recovered four weeks after the inoculation date at 0, 8, and 16 cm on ‘Hass’ cuttings showing that the pathogen moves in the vessels of the cutting and can be recovered at different time periods (Table 7-1).

Discussion

Amplification of rDNA markers (18S, 28S, and ITS regions) have been the norm when identifying fungal organisms. However, amplification of such markers for

166

Raffaelea lauricola has been very difficult and problematic. It is necessary to have a practical diagnostic method for identifying Raffaelea lauricola as laurel wilt is a major threat to avocado producing areas such as Florida, California, and Mexico. The 28S region has unique species-specific sequences that allowed for the design of a unique probe used in qPCR assays (Jeyaprakash et al. 2014). Major advantages of qPCR include faster processing times in which many samples from wood extracted DNA could be processed without the need to initially culture putative R. lauricola (Jeyaprakash et al. 2014). Unfortunately, a false positive was obtained with the SSU region of isolate PL

1004, proving that the use of a single marker to distinguish closely related taxa can be difficult and designing taxon-specific assays for new pathogens can be a real challenge

(Dreaden et al. 2014). Therefore, Dreaden et al. (2014) designed a set of microsatellite markers to be able to distinguish closely related taxa of R. lauricola but with the minor drawback of having a detection limit of only up to 0.1 ng DNA compared to 0.0001 ng per PCR for the SSU. Ribosomal DNA loci have multiple copies and vary greatly among fungi, while microsatellite markers are presumed to be low in copy number. The more copy number, the less genomic DNA template is needed for a PCR to reach detection limit (Dreaden et al. 2014). Even though the microsatellite markers are taxon-specific, they are reported to not be sensitive enough to routinely detect pathogen DNA that is normally found in low titers of diseased wood tissue, but may be used to quickly screen for R. lauricola DNA from cultures, and are therefore, used today in the University of

Florida diagnostic clinics (Dreaden et al. 2014). In the current experiment, the amount of

R. lauricola DNA was not able to be quantified using the LW28S forward and reverse primers used by Jeyaprakash et al. (2014) because the origin of the primer sequence

167

could not be identified in the genomic data on NCBI and hence the sequence of the amplicon needed for development of the standard curve was unknown. Unfortunately,

LWD3 and LWD4 primers used in previous screening studies (Ploetz et al. 2012) are not taxon-specific and therefore could not exclude the possibility of false positives.

Additionally, microsatellite markers used by Dreaden et al. (2014) are not sensitive enough to detect R. lauricola in wood samples since titers of this pathogen are very low in diseased tissue (Dreaden et al. 2014). In the current study, primers were designed from 18S rRNA gene of the RL4 isolate (CBS127349) (Ploetz et al. 2011b), but gave us false positives as well as unreliable results. Jeyaprakash et al. (2014) attributes such complications to premature stopping of the Taq polymerase enzyme during the extension process when it comes across GC- or AT-rich tandem repeats, or interference from other complex secondary structures such as stem and loop structures

(Jeyaprakash et al. 2014). The experiment should be repeated and perfected with different inoculation approaches. Furthermore, extracted DNA should be amplified using primers that amplify avocado DNA to confirm that amplification is occurring and that there are no plant inhibitors present. It was also observed that DNA extraction using sorbitol yielded more fungal DNA compared to extraction using the Qiagen kit reported by Jeyaprakash et al. (2014). If the DNA extraction method from woody avocado tissue could be improved, Dr. Dreaden’s microsatellite primers could be a viable option in fiture experiments. More research needs be done in order to confirm the efficacy of the screening assay.

Having a quick and reliable screening assay could help identify laurel wilt- infected plant tissue at the nursery level as strategies to reduce the spread of this

168

pathogen, which is in dire need for the avocado industry as well as our forest ecosystem.

Conclusion

Preliminary experiments show that avocado shoot cuttings could be kept alive for almost two months when the cut basal end was placed in a water solution containing

Kathon Plant Preservative Mixture (PPM) biocide. It was shown that the laurel wilt pathogen, Raffaelea lauricola, can be inoculated in avocado shoot cuttings, that the pathogen would translocate up the shoot, and be recovered and isolated on semi- selective media CSMA+. However, a reliable qPCR assay needs more work. Primers that we designed were inconsistent, maybe due to the premature stopping of the polymerase during the extension phase. Microsatellite markers may be a better choice for the fine tuning of this assay even though DNA titers may be too small for such markers. It was also observed that sorbitol extraction protocols yielded more pathogen

DNA compared to the extraction method used by Jeyaprakash et al. (2014). If more pathogen DNA could be extracted with a more efficient protocol, microsatellite markers could be a good alternative to perfect this assay.

169

Figure 7-1. Shoot cuttings set-up before inoculation with Raffaelea lauricola. (photo courtesy of Cristina Pisani)

170

Figure 7-2. Measurements of vascular staining distance from inoculation point and disease severity assessments on shoot cuttings of ‘Hass’-‘Bacon’ hybrids (A), ‘Lula’ (B), and ‘Marcus Pumpkin’ (C). Disease severity assessed based on the following scale: 0 = absent (no vascular staining), 1 = slight (up to 1cm above plug), 2 = present (from 1-5 cm above plug), 3 = severe (> 5 cm above plug), 4 = dead (dead cutting).

171

C Figure 7-2. Continued.

Figure 7-3. Calibration curve used for quantitative real time PCR assay.

172

Table 7-1. RT-qPCR results of R. lauricola DNA extracted from avocado wood tissues. Weeka Sampleb Ct No. of molecules/µl H2O (-) Undetermined NA

Raf (+) 35.0 2.5E+10

Raf (+) 32.3 9.8E+10

Raf (+) 26.7 7.0E+11

Raf (+) 33.4 3.1E+10

0 Hs-A 10.1 1.1E+16 1 Wd-IP 5.3 1.5E+16 1 Hs-A 10.8 1.3E+15 2 Hs-IP 38.7 4.7E+09 2 Wd-A 35.0 2.7E+10 3 Wd-IP 12.9 8.5E+14 4 Hs-IP 12.8 4.5E+14 4 Hs-A 4.6 2.1E+16 4 Hs-IP 5.1 1.6E+16 4 Hs-M 5.3 1.5E+16 aDNA was extracted from shoot cuttings immediately after inoculation (week 0) and every week for 4 weeks thereafter. bDNA extracted from a fungal colony grown on MEA was used as a (+) control. DNA was extracted from ‘Hass’ (Hs) and ‘Waldin’ (Wd) shoot cuttings at three different points: inoculation point (IP) at 0 cm, at 8 cm (M), and at the shoot apex 16 cm (A).

173

CHAPTER 8 FINAL CONCLUSIONS

A California mapping population (reciprocal crosses of ‘Hass’ and ‘Bacon’) was established in Fort Pierce, Florida to identify new hybrids that appear to be well adapted to east-central Florida. Flowering occurred from mid-February to mid-May throughout the course of this study (2013-2015). Peak bloom for most trees occurred between the dates of March 27-April 11 (Table 2-1; Figure 2-3; 2-6), March 20-April 16 (Table 2-2;

Figure 2-4; 2-7), and March 18-April 1 (Table 2-3; Figure 2-5; 2-8) in 2013, 2014, and

2015 respectively. Temperature during the flowering period (February-May) ranged from

11.3˚C-27.1˚C, 12.6˚C -26.3˚C, and 7˚C-24.7˚C in 2013-2015 (Figure 2-12). Flower induction occurred during October-November when day temperatures did not exceed

25˚C, regardless of daylength. Low temperatures of 20˚C during the day and night temperatures between 5 and 15˚C have been shown to induce flowering, and daytime temperatures of 25-30˚C completely inhibit flower initiation (Sedgley & Annells 1981). In our study, temperatures during flower bud initiation (October-November) ranged from

11˚C and 27˚C (Figure 2-12) and lowest temperatures were caused by passing cold fronts. Nonetheless, temperatures were well within optimal range for successful flowering.

Fruit stopped growing and was ready for harvest during the last week of October across all seedlings for all three years (Figure 2-11). At harvest, the fruit had dry matter content ranging between 13% and 35% (individual hybrid data not shown). According to

California maturity standards, minimum dry matter for ‘Hass’ is 20.8%, while for ‘Bacon’ it is 17.7% at harvest. Dry matter is correlated with oil content and both have been used in California-type avocados as indicators of fruit maturity and quality. In addition, the

174

types and abundance of different fatty acids found in the current hybrids was similar to previous reports showing oleic acid (C18:1) as the most abundant, followed by palmitic

(C16:0), linoleic (C18:2), and palmitoleic (C16:1) acids. Linolenic (C18:3), stearic (C18:0), and myristic (C14:0) fatty acids were found in very small and/or trace amounts (Table 4-1).

Hybrids varied markedly in the presence of the omega-6 linoleic acid, which is considered unhealthy by some medical experts, when taken over 12-17 g/day (5%-10% adequate intake) (Academy of Nutrition and Dietetics 2014). Optimal dietary intake of omega-6: omega-3 ratio should be around 1-4:1 and has now increased to be within the range of 10:1 to 20:1 (Patterson et al. 2011). However, omega-6 linoleic acid from plants does not convert at a high rate to arachidonic acid, a precursor for inflammatory cytokines. Dry matter percentages in the current study were above the minimum maturity standards listed for California ‘Bacon’ and ‘Hass’ and apparently can be used with oil content as maturity indices for any selections derived from these hybrids. Upon ripening, fruit of most of the hybrids became darker and dull (like ‘Hass’), with the exception of some selections in which fruit remained green but darkened slightly (like

‘Bacon’) (Table 3-1; 3-2; 3-3). Tissue breakdown of fruits was the postharvest disorder/disease of greatest severity throughout all three seasons (Table 3-12). Fungi causing stem end rots and body rots, such as Colletotrichum gloeosporioides (causal agent of anthracnose) may be the primary postharvest disease agent. Anthracnose is a very common disease on avocado and it thrives in hot humid tropical and subtropical environments like Florida. Temperature management postharvest may be implicated as warm and humid conditions allow the pathogens to proliferate. Maintenance minimal spray program was used to reveal weaknesses in the hybrids under evaluation, for

175

commercial production a better IPM program needs to be implemented both in the field and postharvest to reduce rot incidences.

Selections chosen to be included in sensory panels have shown the potential for producing California-type avocados grown in a Florida climate. Selections evaluated appear to have fruit similar in quality to commercial ‘Hass’. The R8T18 selection had the highest preference rating in 2014 and again in 2015 together with R6T56 (Figure 5-4B;

5-4C). Selection R8T18 in 2014 and R6T56 in 2015 were rated as the most creamy with nutty and buttery flavor (Figure 5-2B; 5-2C). However, there were no significant differences among any of the selections (except R7T54 and commercial ‘Hass’ from

Chile) (Figure 5-4B; 5-4C) which supports the possibility that high quality new selections were identified.

When half sib families were screened for laurel wilt tolerance/resistance, ‘Argui 1’ and ‘Booth 8’ showed least disease severity and ‘No. 21’ highest disease severity

(Table 6-1). The working hypothesis was that West Indian families (e.g. ‘Bernecker’,

‘Dade’) would be more susceptible based on previous findings (Ploetz et al. 2012).

Mexican (‘La Piscina’) or Guatemalan (‘R14T06’) or G-M hybrids (‘Bacon’, ‘Ettinger’) were expected to be more resistant than pure West Indian or WI-G hybrids. Overall, the opposite was true in our study, although pure M or G material likely was not evaluated.

‘Melendez’ (G-WI), ‘Booth 8’ (G-WI), and ‘Yon’ (G-WI) families had the lowest proportion of trees succumb to the disease, while ‘Ettinger’ (G-M) had the highest proportion of trees die due to laurel wilt. ‘Argui 1’ WI-(G-M), ‘Booth 8’ (G-WI), and ‘Melendez’ (G-WI) showed lowest disease severity and ‘No. 21’, a G-M hybrid, was most susceptible

176

(Table 6-1). Additional avocado genotypes will be screened for disease tolerance/resistance in the near future.

When developing a qPCR assay to screen for laurel wilt, preliminary experiments showed that avocado shoot cuttings could be kept alive in test tubes for almost two months. It was possible to inoculate these cuttings with the laurel wilt pathogen,

Raffaelea lauricola and re-isolate it on semi-selective medium CSMA+. However, the screening using a qPCR assay needs more work. Primers that we designed from the

18S rRNA gene gave us unreliable results with false positives. Furthermore, DNA extraction using sorbitol yielded more DNA (ng/µL) and the experiment should be repeated with DNA sorbitol extraction and using microsatellite markers identified in previous research (Dreaden et al. 2014) for the fine tuning of this assay. The downfall of the microsatellite markers is that they can only detect a minimum of 0.1 ng of pathogen

DNA. If more pathogen DNA could be extracted with a more efficient protocol, microsatellite markers could be a good alternative to perfect this assay.

Future studies consist of a continuation of the above studies. Selected trees already planted from the ongoing evaluation for laurel wilt resistance will be genotyped by Dr. David Kuhn using 384 SNP markers that are evenly spaced across the 12 linkage groups (25 per linkage group, sufficient for quantitative trait loci mapping) and will include markers specific for flower type prediction. In addition, a version of bulk segregant analysis that will consist on genotyping the three best and three worst from each family will be used.

Three hundred additional reciprocal crosses of ‘Hass’ and ‘Bacon hybrids were planted at Ft. Pierce in 2011 and some were flowering in 2014. Sister hybrids were also

177

planted in the Miami ARS station in 2011, and may provide valuable selections for

South Florida. Tree growth (height), general health, flowering, and cropping (number of fruit and weight) will be evaluated on the above trees and samples of promising selections harvested for postharvest quality and shelf life evaluations.

The three best hybrid selections from the project described above are being field- tested alongside 10 cultivars with ‘Hass’-like fruit quality, and the two best selections from the University of California at Riverside avocado breeding project. Trees were propagated on three rootstocks, ‘Waldin’ seedling, which is widely used in Florida,

‘Dusa’ and ‘Duke 7’ clonal rootstocks which are Phytophthora-resistant and widely used in California. Trees were planted on 4 sites (3 in the Indian River area and 1 in Miami-

Dade County) in the spring of 2015.

178

APPENDIX SEQUENCE OF QUESTIONS ASKED TO PANELISTS DURING SENSORY ANALYSIS OF FRESH AVOCADO

Please take a bite of sample _____ and tell us how much you like it:

Sample ______

Dislike Dislike Dislike Dislike Neither Like Like Like Like extremely very moderately slightly like or slightly moderately very extremely much dislike much

Please choose which describes best sample ____ (circle as many descriptors as necessary):

Texture Flavor/aromatics Firm Bland Mushy Grassy Stringy/Gritty Woody/Piney Creamy Sweet Dry Nutty Watery Buttery Savory Rancid

Comments______

Take a bite of carrot or cracker and a drink of water to rinse your mouth before evaluating the next sample.

179

LIST OF REFERENCES

Academy of Nutrition and Dietetics, 2014. Position of the Academy of Nutrition and Dietetics: dietary fatty acids for healthy adults. Journal of the Academy of Nutrition and Dietetics 114: 136-153.

AgMRC – Agricultural Marketing Resource Center, 2013. Available from http://www.agmrc.org/commodities-products/fruits/avocados/ Last access: January 6, 2016.

Anon., 1925. Standardization committee report California Avocado Association Annual Report. California Avocado Association Pasadena (CA) pp. 46-47.

Ayala-Silva T, Schnell R, Gordon G, Winterstein MC, 2012. Application of propiconazole in management of laurel wilt disease in avocado (Persea Americana Mill.) trees. In: Proceedings of the First International Symposium on Wild Realtives of Subtropical and Temperate Fruit and Nut Crops. Acta Horticulturae pp. 71-78.

Barmore CR, 1976. Avocado fruit maturity. In: Proceedings of the First International Tropical Fruit Short Course: The Avocado. Florida Cooperative Extension Service. Institute of Food and Agricultural Sciences, University of Florida, Gainesville pp. 103-109.

Berasategi I, Barriuso B, Ansorena D, Astiasarán I, 2012. Stability of avocado oil during heating: comparative study to olive oil. Food Chemistry 132: 439-446.

Bergh BO, 1986. Persea americana. In: Halevy HA (eds), Handbook of Flowering, Volume 5, 1st edn. CRC Press, Boca Raton, Florida, pp. 253-268.

Bergh B, Ellstrand N, 1986. Taxonomy of the avocado. California Avocado Society 1986 Yearbook 70: 135-146.

Bertling I, Bower JP, 2006. Avocado sugars during early fruit development. South African Avocado Growers’ Association Yearbook 29: 38-39.

Bi J, O’Mahony M, 2013. Variance of dꞌ for the tetrad test and comparisons with other forced-choice methods. Journal of Sensory Studies 28: 91-101.

Bora PS, Narain N, Rocha RVM, Paulo MQ, 2001. Characterization of the oils from the pulp and seeds of avocado (cultivar: Fuerte) fruits. Grasas y Aceites 52(3-4): 171-174.

Borgognone MG, Bussi J, Hough G, 2001. Principal component analysis in sensory analysis: covariance or correlation matrix? Food Quality and Preference 12: 323- 326.

180

Brar GS, Capinera JL, McLean S, Kendra PE, Ploetz RC, Peña JE, 2012. Effect of trap size, trap height and age of lure on sampling Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae), and its flight periodicity and seasonality. Florida Entomologist 95(4): 1003-1011.

Buckler ES IV, Pearsall DM, Holtsford TP, 1998. Climate, plant ecology, and Central American archaic subsistence. Current Anthropology 39: 152-164.

Buttrose MS, Alexander DM, 1978. Promotion of floral initiation in ‘Fuerte’ avocado by low temperature and short daylength. Scientia Horticulturae 8: 213-219.

Campbell AS, 2014. Characterizing Pathogen: vector:host interactions in laurel wilt, a disease of avocado. Ph.D dissertation University of Florida, Gainesville (FL).

Carrillo D, Duncan RE, Ploetz JN, Campbell AF, Ploetz RC, Peña JE, 2014. Lateral transfer of a phytopathogenic symbiont among native and exotic ambrosia beetles. Plant Pathology 63: 54-62.

Carrillo D, Dunlap CA, Avery PB, Navarrete J, Duncan RE, Jackson MA, Behle RW, Cave RD, Crane J, Rooney AP, Peña JE, 2015. Entomopathogenic fungi as biological control agents for the vector of laurel wilt disease, the redbay ambrosia beetle, Xyleborus glabratus (Coleoptera: Curculionidae). Biological Control 81: 44-55.

Chabanet C, 2000. Statistical analysis of sensory profiling data. Graphs for presenting results (PCA and ANOVA). Food Quality and Preference 11: 159-162.

Chen H, Morrell PL, Ashworth VETM, De La Cruz M, Clegg MT, 2009. Tracing the geographic origins of major avocado cultivars. Journal of Heredity 100(1): 56-65.

Cowan AK, 2004. Metabolic control of avocado fruit growth: 3-hydroxy-methylglutaryl coenzyme a reductase, active oxygen species and the role of C7 sugars. South African Journal of Botany 70(1): 75-82.

Cox KA, McGhie TK, White A, Woolf AB, 2004. Skin colour and pigment changes during ripening of ‘Hass’ avocado fruit. Postharvest Biology and Technology 31: 287- 294.

Crane J, Evans E, Balerdi C, 2007. A review of the Florida avocado industry. Proceedings VI World Avocado Congress, Viña del Mar, Chile, Nov. 12-16. Available from http://www.avocadosource.com/WAC6/en/Extenso/5a-209.pdf Last access: January 7, 2016.

Crane JH, Peña J, Osborne JL, 2008. Redbay ambrosia beetle-laurel wilt pathogen: a potential major problem for the Florida avocado industry. UF-IFAS EDIS Doc. # HS1136. Available from http://edis.ifs.ufl.edu Last access: January 24,2016.

181

Crane J, Peña JE, Ploetz RC, Smith J, Evans EA, 2011. Proposed strategies for decreasing the threat of laurel wilt (LW) and its vector, the redbay ambrosia beetle (RAB) to commercial avocado groves in Miami-Dade County. IFAS doi: c://ext/handouts/2011/Proposed grove strategy RAB-LW 4-1-11.docx

Crane JH, Douhan G, Faber BA, Arpaia ML, Bender GS, Balerdi CF, Barrientos-Priego AF, 2013. Cultivars and Rootstocks. In: Schaffer B, Wolstenholme BN, Whiley AW, (eds). The avocado: botany, production and uses, 2nd edn. CABI, Massachusetts, pp. 200-233.

Cristancho M, Escobar C, 2008. Transferability of SSR markers from related Uredinales species to the coffee rust Hemileia vastatrix. Genetics and Molecular Research 7: 1186-1192.

Dann EK, Ploetz RC, Coates LM, Pegg KG, 2013. Foliar, fruit, and soilborne diseases. In: Schaffer B, Wolstenholme BN, Whiley AW (eds), The avocado: botany, production and uses, 2nd edn. CABI, Boston, pp. 380-415.

Davenport TL, 1986. Avocado flowering. In: Janick J (eds), Horticultural Reviews, AVI Publishing Co., Inc., Westport, CN, 8: pp. 257-289.

Davis DR, Wagner DL, 2011. Biology and systematics of the New World Phyllocnistis Zeller leafminers of the avocado genus Persea (Lepidoptera, Gracillariidae). ZooKeys 97: 39-73.

Dreaden TJ, Davis JM, Harmon CL, Ploetz RC, Palmateer AJ, Soltis PS, Smith JA, 2014a. Development of multilocus PCR assays for Raffaelea lauricola, causal agent of laurel wilt disease. Plant Disease 98(3): 379-383.

Dreaden TJ, Davis JM, De Beer ZW, Ploetz RC, Soltis PS, Wingfield MJ, Smith JA, 2014b. Phylogeny of ambrosia beetle symbionts in the genus Raffaelea. Fungal Biology 118: 970-978.

Du Plessis LM, 1979. Seasonal changes in the composition of avocado oil recovered by centrifugation. South African Avocado Growers’ Association Research Report 3: 74-79.

Dutech C, Enjalbert J, Fournier E, Delmotte F, 2007. Challenges of microsatellite isolation in fungi. Fungal Genetics and Biology 44: 933-949.

Evans EA, Crane J, Hodges A, Osborne JL, 2010. Potential economic impact of laurel wilt disease on the Florida avocado industry. HortTechnology 20(1): 234-238.

Evans JP, Scheffers BR, Hess M, 2014. Effect of laurel wilt invasion of redbay populations in a maritime forest community. Biological Invasions 16: 1581-1588.

182

FAOSTAT – Food and Agriculture Organization of the United Nations, 2014. FAO Statistics Division, Rome, Italy. Available from http://faostat.fao.org/. Last access: January 6, 2016.

Fraedrich SW, Harrington TC, Rabaglia RJ, Ulyshen MD, Mayfield III AE, Hanula JL, Eickwort JM, Miller DR, 2008. A fungal symbiont of the redbay ambrosia beetle causes a lethal wilt in redbay and other Lauraceae in the southeastern United States. Plant Disease 92(2): 215-224.

Fraedrich SW, Harrington TC, Best GS, 2015. Xyleborus glabratus attacks and systemic colonization by Raffaelea lauricola associated with dieback of Cinnamomum camphora in the southeastern United States. Forest Pathology doi: 10.1111/efp.12124.

Galindo-Tovar ME, Ogata-Aguilar N, Arzate-Fernández AM, 2008. Some aspects of avocado (Persea americana Mill.) diversity and domestication in Mesoamerica. Genetic Resources and Crop Evolution 55: 441-450.

Gibson AC, 1984. The guacamole tree. In: Writeups and illustrations of economically important plants. < http://www.botgard.ucla.edu/html/botanytextbooks/economicbotany/Persea/index .html> Last access: February 17, 2016.

Gramling LM, 2010. Potential effects of laurel wilt on the flora of North America. Southeast Naturalist 9(4): 827-836.

Hall NT, Smoot JM, Knight Jr. RJ, Nagy S, 1980. Protein and amino acid compositions of ten tropical fruits by gas-liquid chromatography. Journal of Agricultural and Food Chemistry 28: 1217-1221.

Hanula JL, Mayfield AE, Fraedrich SW, Rabaglia RJ, 2008. Biology and host associations of redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae), exotic vector of laurel wilt killing redbay trees in the southeastern United States. Journal of Economic Entomology 101(4): 1276-1286.

Hanula JL, Sullivan BT, Wakarchuk D, 2013. Variation in manuka oil lure efficacy for capturing Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae), and cubeb oil as an alternative attractant. Chemical Ecology 42(2): 333-340.

Harding PL, 1954. The relation of maturity to quality in Florida avocados. Florida State Horticultural Society 67: 276-280.

Harrington TC, 1981. Cycloheximide sensitivity as a taxonomic character in Ceratocystis. Mycologia 73: 1123-1129.

Harrington TC, Fraedrich SW, Aghayeva DN, 2008. Raffaelea lauricola, a new ambrosia beetle symbiont and pathogen on the Lauraceae. Mycotaxon 104: 399-404.

183

Harrington TC, Yun HY, Lu S-S, Goto H, Aghayeva DN, Fraedrich SW, 2011. Isolations from the redbay ambrosia beetles, Xyleborus glabratus, confirm that the laurel wilt pathogen, Raffaelea lauricola, originated in Asia. Mycologia 103(5): 1028- 1036.

Hatton Jr. TT, Campbell CW, 1959. Evaluation of indices for Florida avocado maturity. Proceedings of the Florida State Horticultural Society 72: 349-353.

Hatton TT, Harding PL, Reeder WF, 1964. Seasonal changes in Florida avocados, USDA, Technical Bulletin # 1310. Washington D.C.

Hershkovitz V, Saguy SI, Pesis E, 2005. Postharvest application of 1-MCP to improve the quality of various avocado cultivars. Postharvest Biology and Technology 37: 252-264.

Hodgkin GB, 1939. Avocado standardization. California Avocado Society 1939 Yearbook 24: 141-148.

Hofman PJ, Jobin-Décor M, Giles J, 2000. Percentage of dry matter and oil content are not reliable indicators of fruit maturity or quality in late-harvested ‘Hass’ avocado. HortScience 35(4): 694-695.

Hughes MA, 2013. The evaluation of natural resistance to laurel wilt disease in redbay (Persea borbonia). Ph.D dissertation University of Florida, Gainesville (FL).

Hughes MA, Inch SA, Ploetz RC, Er HL, van Bruggen AHC, Smith JA, 2015. Responses of swamp bay, Persea palustris, and avocado, Persea americana, to various concentrations of the laurel wilt pathogen, Raffaelea lauricola. Forest Pathology doi: 10.1111/efp.12134.

Hulcr J, Dunn RR, 2011a. The sudden emergence of pathogenicity in insect-fungus symbioses threatens naïve forest ecosystems. Proceedings of the Royal Society of Biological Sciences 278: 2866-2873.

Hulcr J, Mann R, Stelinski LL, 2011b. The scent of a partner: Ambrosia beetles are attracted to volatiles from their fungal symbionts. Journal of Chemical Ecology 37: 1374-1377.

Inch S, Ploetz R, Held B, Blanchette R, 2012. Histological and anatomical responses in avocado, Persea americana, induced by the vascular wilt pathogen, Raffaelea lauricola. Botany-Botanique 90(7): 627-635.

Inch SA, Ploetz RC, 2012. Impact of laurel wilt, caused by Raffaelea lauricola, on xylem function in avocado, Persea americana. Forest Pathology 42(3): 239-245.

Jeong J, Huber DJ, Sargent SA, 2003. Delay of avocado (Persea americana) fruit ripening by 1-methylcyclopropene and wax treatments. Postharvest Biology and Technology 28: 247-257.

184

Jeyaprakash A, Davison DA, Schubert TS, 2014. Molecular detection of the laurel wilt fungus, Raffaelea lauricola. Plant Disease 98(4): 559-564.

Kader AA, 2002.Postharvest Technology of Horticultural Crops, 3rd edn. University of California Agriculture and Natural Resources, California.

Kendra PE, Montgomery WS, Niogret J, Peña JE, Capinera JL, Brar G, Epsky ND, Heath RR, 2011. Attraction of the redbay ambrosia beetle, Xyleborus glabratus, to avocado, lychee, and essential oil lures. Journal of Chemical Ecology 37: 932- 942.

Kendra PE, Montgomery WS, Niogret J, Deyrup MA, Guillén L, Epsky ND, 2012. Xyleborus glabratus, X. affinis, and X. ferrugineus (Coleoptera: Curculionidae: Scolytinae): Electroantennogram responses to host-based attractants and temporal patterns in host-seeking flight. Chemistry and Ecology 41(6): 1597- 1605.

Kendra PE, Montgomery WS, Niogret J, Epsky ND, 2013a. An uncertain future for American Lauraceae: a lethal threat from redbay ambrosia beetle and laurel wilt disease (a review). American Journal of Plant Sciences 4: 727-738.

Kendra PE, Ploetz RC, Montgomery WS, Niogret J, Peña JE, Brar GS, Epsky ND, 2013b. Evaluation of Litchi chinensis for host status to Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae) and susceptibility to laurel wilt disease. Florida Entomologist 96(4): 1442-1453.

Kendra PE, Montgomery WS, Niogret J, Pruett GE, Mayfield AE III, MacKenzie M, Deyrup MA, Bauchan GR, Ploetz RC, Epsky ND, 2014a. North American Lauraceae: Terpenoid emissions, relative attraction and boring preferences of redbay ambrosia beetle, Xyleborus glabratus (Coleoptera: Curculionidae: Scolytinae). PLoS ONE 9(7): e102086.

Kendra PE, Montgomery WS, Niogret J, Schnell EQ, Deyrup MA, Epsky ND, 2014b. Evaluation of seven essential oils identifies cubeb oil as most effective attractant for detection of Xyleborus glabratus. Journal of Pest Science 87: 681-689.

Koch FH, Smith WD, 2008. Spatio-temporal analysis of Xyleborus glabratus (Coleoptera: Circulionidae: Scolytinae) invasion in Eastern U.S. forests. Environmental Entomology 37(2): 442-452.

Kopp L, 1966. A taxonomic revision of the genus Persea in the western hemisphere (Perseae-Lauraceae). Memoirs of the New York Botanical Garden 14: 1-120.

Kotze’ JM, 1978. Anthracnose of avocados. South African Avocado Growers’ Association Research Report for 1978 2: 45-47.

185

Lederhouse RC, Ayres MP, Nitao JK, Scriber JM, 1992. Differential use of lauraceous hosts by swallowtail butterflies, Papilio troilus and P. palamedes (Papilionidae). Oikos 63: 244-252.

Lee SK, 1981. A review and background of the avocado maturity standard. California Avocado Society 1981 Yearbook 65: 101-109.

Lee SK, Young RE, Schiffman PM, Coggins Jr. CW, 1983. Maturity studies of avocado fruit based on picking dates and dry weight. Journal of the American Society for Horticultural Science 108(3): 390-394.

Liu X, Robinson PW, Madore MA, Witney GW, Arpaia ML, 1999. ‘Hass’ avocado carbohydrate fluctuations. II. Fruit growth and ripening. Journal of the American Society for Horticultural Science 124(6): 676-681.

Liu X, Sievert J, Arpaia ML, Madore MA, 2002. Postulated physiological roles of the seven-carbon sugars, mannoheptulose, and perseitol in avocado. Journal of the American Society for Horticultural Science 127(1): 108-114.

Litz RE, Witjaksono, Raharjo S, Efendi D, Pliego-Alfaro F, Barceló-Muñoz A, 2005. Persea americana Avocado. In: Litz RE (eds), Biotechnology of fruit and nut crops, 1st edn. CABI, Cambridge, pp.326-347.

López LR, Frati MAC, Hernández DCB, Cervantes MS, Hernández LMH, Juárez C, Morán LS, 1996. Monounsaturated fatty acid (avocado) rich diet for mild hypercholesterolemia. Archives of Medical Research 27(4): 519-523.

Lovatt CJ, 1990. Factors affecting fruit set/early fruit drop in avocado. California Avocado Society 1990 Yearbook 74: 193-199.

Mayfield III AE, Smith JA, Hughes M, Dreaden TJ, 2008a. First report of laurel wilt disease caused by Raffaelea sp. on avocado in Florida. Plant Disease 92(6): 976.

Mayfield III AE, Barnard EL, Smith JA, Bernick SC, Eickwort JM, Dreaden TJ, 2008b. Effect of propiconazole on laurel wilt disease development in redbay trees and on the pathogen in vitro. Arboriculture & Urban Forestry 34(5): 317-324.

Mayfield III AE, Peña JE, Crane JH, Smith JA, Branch CL, Ottoson ED, Hughes M, 2008c. Ability of the redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae) to bore into young avocado (Lauraceae) plants and transmit the laurel wilt pathogen (Raffaelea sp.). Florida Entomologist 9(3): 485-487.

McGuire RG, 1992. Reporting of objective color measurements. HortScience 27(12): 1254-1255.

Menge JA, Ploetz, RC, 2003. Diseases of avocado. In: Ploetz RC (eds), Diseases of Tropical Fruit Crops, 1st edn. CABI, Cambridge, pp. 35-72.

186

Milgoom MG, 2004. Biological control of chestnut blight with hypovirulence: a critical analysis. Annual Review of Pathology 42: 311-338.

Moreno AO, Dorantes L, Galíndez J, Guzmán RI, 2003. Effect of different extraction methods on fatty acids, volatile compounds, and physical and chemical properties of avocado (Persea americana Mill.) oil. Journal of Agricultural and Food Chemistry 51: 2216-2221.

Morris R, O’Brien K, 1980. Testing avocados for maturity. Agricultural Gazette of New South Wales 42: 44.

NASS – National Agricultural Statistics Service, 2014. Noncitrus fruits and nuts – 2014 summary. Available from http://www.nass.usda.gov. Last access: January 6, 2016.

Newhouse AE, Schrodt F, Liang H, Maynard CA, Powell WA, 2007. Transgenic American elm shows reduced Dutch elm disease symptoms and normal mycorrhizal colonization. Plant Cell Reports 26: 977-987.

Obenland D, Collin S, Sievert J, Negm F, Arpaia ML, 2012. Influence of maturity and ripening on aroma volatiles and flavor in ‘Hass’ avocado. Postharvest Biology and Technology 71: 41-50.

Ozdemir F, Topuz A, 2004. Changes in dry matter, oil content and fatty acids composition of avocado during harvesting time and post-harvesting ripening period. Food Chemistry 86: 79-83.

Pacetti D, Boselli E, Lucci P, Frega NG, 2007. Simultaneous analysis of glycolipids and phospholipids molecular species in avocado (Persea americana Mill) fruit. Journal of Chromatography A 1150: 241-251.

Patterson E, Wall R, Fitzgerald GF, Ross RP, Stanton C, 2011. Health implications of high dietary omega-6 polyunsaturated fatty acids. Journal of Nutrition and Metabolism 2012: 1-16.

Pereira MEC, 2010. Ripening, volatiles and sensory attributes of West Indian and Guatemalan-West Indian hybrid avocados as affected by 1-methylcyclopropene and ethylene. Ph.D dissertation University of Florida, Gainesville (FL).

Pesis E, Ackerman M, Ben-Arie R, Feygenberg O, Feng X, Apelbaum A, Goren R, Prusky D, 2002. Ethylene involvement in chilling injury symptoms of avocado during cold storage. Postharvest Biology and Technology 24: 171-181.

Pino J, Rosado A, Aguero J, 2000. Volatile components of avocado (Persea americana Mill.) fruits. Journal of Essential Oil Research 12: 377-378.

187

Pisani C, Ritenour MA, Stover E, Plotto A, Gutierrez OA, Kuhn D, 2015. California avocados in Florida? Finding the perfect avocado for production in East-Central Florida. Proceedings of Florida State Horticultural Society 127: 131-134.

Ploetz RC, Zentmyer GA, Nishijima WT, Rohrbach KG, Ohr HD, 1994. Compendium of Tropical Fruit Diseases, 1st edn. The American Phytopathological Society, Minnesota.

Ploetz RC, Peña JE, Smith JA, Dreaden TJ, Crane JH, Schubert T, Dixon W, 2011a. Laurel wilt, caused by Raffaelea lauricola, is confirmed in Miami-Dade county, center of Florida's commercial avocado production. Plant Disease 95(12):1589- 1589.

Ploetz RC, Pérez-Martínez JM, Evans EA, Inch SA, 2011b. Toward fungicidal management of laurel wilt of avocado. Plant Disease 95: 977-982.

Ploetz RC, Pérez-Martínez JM, Smith JA, Hughes M, Dreaden TJ, Inch SA, Fu Y, 2012. Responses of avocado to laurel wilt, caused by Raffaelea lauricola. Plant Pathology 61: 801-808.

Ploetz RC, Hulcr J, Wingfield MJ, de Beer ZW, 2013. Destructive tree diseases associated with ambrosia and bark beetles: Black swan events in tree pathology? Plant Disease 97(7): 856-872.

Ploetz RC, Schaffer B, Vargas AI, Konkol JL, Salvatierra J, Wideman R, 2015. Impact of laurel wilt, caused by Raffaelea lauricola, on leaf gas exchange and xylem sap flow in avocado, Persea americana. Phytopathology 105(4): 433-440.

Purseglove JW, 1968. Tropical Crops. Dicotyledons 1, 1st edn. Longmans, London.

Rabaglia RJ, Dole SA, Cognato AI, 2006. Review of American Xyleborina (Coleoptera: Curculionidae: Scolytinae) occurring north of Mexico, with an illustrated key. Annals of the Entomological Society of America 99: 1034-1056.

Ranger CM, Reding ME, Persad AB, Herms DA, 2010. Ability of stress-related volatiles to attract and induce attacks by Xylosandrus germanus and other ambrosia beetles. Agricultural and Forest Entomology 12: 177-185.

Ratovohery JV, Lozano YF, Gaydou EM, 1988. Fruit development effect on fatty acid composition of Persea americana fruit mesocarp. Journal of Agricultural and Food Chemistry 36(2): 287-293.

Requejo AM, Ortega RM, Robles F, Navia B, Faci M, Aparicio A, 2003. Influence of nutrition on cognitive function in a group of elderly, independently living people. European Journal of Clinical Nutrition 57: S54-S57.

188

Salazar-García S, Garner LC, Lovatt CJ, 2013. Reproductive Biology. In: Schaffer B, Wolstenholme BN, Whiley AW, (eds). The avocado: botany, production and uses, 2nd edn. CABI, Massachusetts, pp. 118-167.

Schaffer B, Wolstenholme BN, Whiley AW, 2013. The avocado: botany, production and uses, 2nd edn. CABI, Massachusetts.

Schnell RJ, Brown JS, Olano CT, Power EJ, Krol CA, Kuhn DN, Motamayor JC, 2003. Evaluation of avocado germplasm using microsatellite markers. Journal of the American Society of Horticultural Science 128(6): 881-889.

Schroeder CA, 1953. Growth and development of the ‘Fuerte’ avocado fruit. Proceedings of the American Society for Horticultural Science 61: 103-109.

Scora RW, Wolstenholme BN, Lavi U, 2002. Taxonomy and Botany. In: Schaffer B, Wolstenholme BN, Whiley AW, (eds). The avocado: botany, production and uses, 1st edn. CABI, Massachusetts, pp. 15-37.

Sedgley M, 1977. Physiology of pollination and fruit set and possibilities of manipulation. Proceedings of the Avocado Research Workshop, South Queensland, pp. 59-65.

Sedgley M, Annells CM, 1981. Flowering and fruit-set response to temperature in the avocado cultivar ‘Hass’. Scientia Horticulturae 14: 27-33.

Sedgley M, Grant, WJR, 1982/83. Effect of low temperatures during flowering on floral cycle and pollen tube growth in nine avocado cultivars. Scientia Horticulturae 18: 207-213.

Seymour GB, Tucker GA, 1993. Avocado. In: Seymour GB, Tayler J, Tucker GA (eds), Biochemistry of Fruit Ripening, 1st edn. Chapman & Hall, London, pp. 53-81.

Slater GG, Shankman S, Sheperd JS, Alfin-Slater RB, 1975. Seasonal variation in the composition of California avocados. Journal of Agricultural and Food Chemistry 23(3): 468-474.

Smith CE, 1966. Archeological evidence for selection in avocado. Economic Botany 23: 135-140.

Smith JA, Mount L, Mayfield AE III, Bates CA, Lamborn WA, Fraedrich SW, 2009. First report of laurel wilt disease caused by Raffaelea lauricola on camphor in Florida and Georgia. Plant Disease 93(2): 198.

Spence DJ, Smith JA, Ploetz R, Hulcr J, Stelinski LL, 2013. Effect of chipping on emergence of the redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae) and recovery of the laurel wilt pathogen from infested wood chips. Forest Entomology 106(5): 2093-2100.

189

Stahl AL, 1933. Changes in composition of Florida avocados, in relation to maturity. Agricultural Experiment Station, Gainesville, Florida, May Bulletin 259.

Thomas MC, 2007. Two Asian ambrosia beetles recently established in Florida (Curculionidae: Scolytinae). Dept. Plant Industry, Florida Department of Agriculture and Consumer Services.

UC-ANR. Available from http://ucavo.ucr.edu/General/Maturity.html Last access: February 17, 2016.

Unknown, 1935. Origin of the cultivated races of avocados. California Avocado Association 1935 Yearbook 20: 184-194.

United States Department of Agriculture, 2014. Ranking of potassium content in common foods, per 100 g. National Nutrient Database for Standard Reference, Release 28. Available from http://ndb.nal.usda.gov/ndb/nutrients/index Last access: February 10, 2016.

United States Department of Agriculture, Agricultural Marketing Service. Florida Avocado Grades and Standards. Available from https://www.ams.usda.gov/grades-standards/florida-avocado-grades-and- standards Last access: February 28, 2016.

United States Department of Agriculture Economic Research Service, 2015. Per capita consumption of fresh avocados in the United States from 2000 to 2013 (in pounds)*. In Statista - The Statistics Portal. Available from http://www.statista.com/statistics/257192/per-capita-consumption-of-fresh- avocados-in-the-us/ Last access: February 10, 2016.

U.S. Dep. Agric. For. Serv. 2015. Distribution of counties with laurel wilt disease by year of initial detection. Available from http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprd3832188.pdf Last access: January 24, 2016.

Wang L, Bordi PL, Fleming JA, Hill AM, Kris-Etherton PM, 2015. Effect of a moderate fat diet with and without avocados on lipoprotein particle number, size and subclasses in overweight and obese adults: a randomized, controlled trial. Journal of the American Heart Association 4: 1-14.

White A, Woolf A, Hofman P, Arpaia ML, 2009. The International Avocado Quality Manual, 1st edn. Plant and Food Research, New Zealand.

Yahia EM, Woolf AB, 2011. Avocado (Persea americana Mill.). In: Yahia EM (eds), Postharvest biology and technology of tropical and subtropical fruits Volume 2: Açai to citrus, 1st edn. Woodhead Publishing, Cambridge, pp.125-185.

190

Zhang Z, Huber DJ, Rao J, 2011. Ripening delay of mid-climacteric avocado fruit in response to elevated doses of 1-methylcyclopropene and hypoxia-mediated reduction in internal ethylene concentration. Postharvest Biology and Technology 60: 83-91.

191

BIOGRAPHICAL SKETCH

Cristina Pisani was born in 1980 in Florence, Italy and grew up on a winery in the

Chianti region. She moved to the United States at the age of eight but returned to Italy to start high school. There she studied languages in a private high school where she got basic knowledge on literature and science. Growing up on a wine farm gave her the opportunity to observe the process of wine making and to also help out with the bottling and shipping processes. After high school she returned to the United States where universities offer a wider range of degrees and where she got the opportunity to get hands on training. She graduated with a B.S. in biological sciences in 2006 at Florida

International University, Miami, Florida and started working in a laboratory conducting environmental research. She then received a master's degree in plant pathology focusing on grapes at University of California Davis in 2011 where she got to work closely with farm advisors on a variety of perennial crops. In 2016, Cristina received her

Ph.D. from the Horticultural Science Department at the University of Florida under the supervision of Mark A. Ritenour and Ed Stover. Cristina plans to pursue a career in academia or USDA with job opportunities dealing with horticulture and pathology of perennial crops and working closely with growers.

192