Breeding and Selection of Kalmia Latifolia L. for the Southeastern

BREEDING AND SELECTION OF KALMIA LATIFOLIA L. FOR THE SOUTHEASTERN

UNITED STATES

HE LI

(Under the Direction of Donglin Zhang)

ABSTRACT

Kalmia latifolia L. (mountain laurel) is an outstanding flowering shrub and a promising ornamental native to the eastern U.S. To enhance its production and landscape performance in the southeastern U.S., plants having adaptability to southeastern environmental conditions are needed. Traditional breeding of woody plants takes 10-20 years, however it could be facilitated with aid of modern technologies. To efficiently breed and select mountain laurel for the southeast, we explored and collected germplasm in the U.S.; evaluated existing cultivars for container and landscape performance; assessed genetic relationship among taxa using ISSR; and improved seed germination using in vitro method. Fifteen wild populations were explored in

Alabama, Florida, Georgia, Louisiana, Massachusetts, and South Carolina and collections were from 10 of them based on our breeding goals. A total of 277 plants was obtained from wild collection. From over 100 cultivars, 21 of them were selected for collection based on their desirable morphology and superior performance and 197 individual plants were obtained. In our evaluation, all cultivars yielded considerable growth in the first year of container trials and consistently performed well thereafter, indicating production of mountain laurel as a one-year container plant is feasible. Although most cultivars had difficulty establishing in southeastern landscapes, fall planting could improve establishment. ‘Ostbo Red’, ‘Pristine’, and ‘Tinkerbell’ consistently excelled in field trials, indicating their greater adaptability to southeastern environmental conditions. Unique ISSR profile of each cultivar and clustering of related cultivars in UPGMA indicated accurate identification of cultivars and their pedigree can be achieved using ISSR. A low level of genetic diversity among populations was found while individuals within populations tended to be genetically different, illuminating the efficient means to increase diversity. A clear separation of cultivars and wild plants was observed, which might be due to the loss of genetic information during artificial breeding. Using in vitro seed germination protocol, seeds could be collected one month prior to full maturation and germinate in one month on WPM, which shortened the period from crossing to the seedling stage from 13-

15 to 6 months and improved germination from 30% to over 90% compared with traditional method.

INDEX WORDS: mountain laurel, ericaceous, woody ornamentals, rapid breeding,

germplasm collection, growth index, morphological variation, container

production, field trial, genetic relationship, genetic diversity, ISSR, in

vitro, seed germination, fruit maturity, WPM

BREEDING AND SELECTION OF KALMIA LATIFOLIA L. FOR THE SOUTHEASTERN

UNITED STATES

HE LI

B.S., Central South University of Forestry and Technology, China, 2012

M.S., Central South University of Forestry and Technology, China, 2014

A Dissertation Submitted to the Graduate Faculty of The University of Georgia in Partial

Fulfillment of the Requirements for the Degree

DOCTOR OF PHILOSOPHY

ATHENS, GEORGIA

2018

He Li

BREEDING AND SELECTION OF KALMIA LATIFOLIA L. FOR THE SOUTHEASTERN

UNITED STATES

HE LI

Major Professor: Donglin Zhang Committee: Matthew Chappell Anish Malladi Scott Merkle Wayne Parrott

Electronic Version Approved:

Suzanne Barbour Dean of the Graduate School The University of Georgia December 2018

DEDICATION

I would like to dedicate this work to my parents and husband, who have always loved and supported me!

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Donglin Zhang, for his guidance and support throughout this project. Donglin has been a patient and supportive advisor, who always encourages me to think critically. He advised me on numerous occasions and provided me with opportunities that helped me develop my academic and teaching skills. It has been my privilege and honor to work with him. I hope to become a great advisor to my future students like Donglin has been to me.

I would like to thank my advisory committee, Dr. Chappell, Dr. Malladi, Dr. Merkle, and

Dr. Parrott for their support, advice, and feedbacks on my research. Special thanks go to Dr.

Chappell for helping me contact with local informants, gather population information, and organize collection trips.

I thank our lab technician Jinying Dong for helping me in my fieldwork and greenhouse operations. I thank all members of the UGA Woody Plant Research Lab for their assistance and friendship. I would like to acknowledge faculty and staff in Horticulture Department for their support and help. I thank all Horticulture graduate students for their encouragements. I also thank

Shuyang Zhen for setting up sensors for my study, guiding me how to use equipment, sharing of her ideas, and always supporting me.

Special thanks go to Dr. Ron Miller (Pensacola, FL), Dr. Ron Hooper (Aiken, SC), Mr.

Jack Johnston (Lakemont, GA), and Mr. Connor Ryan (Athens, GA) for exploring wild populations and collecting plant materials for this project.

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ...... v

LIST OF TABLES ...... ix

LIST OF FIGURES ...... xi

CHAPTER

1 INTROCUTION AND LITERATURE REVIEW ...... 1

Introduction ...... 1

Natural Distribution ...... 2

Cultivar Development ...... 2

Propagation ...... 3

Management Practices ...... 5

Breeding ...... 6

Research Needs ...... 7

Research Objectives ...... 7

Literature Cited ...... 9

2 EXPLORATION AND COLLECTION OF KALMIA LATIFOLIA L. GERMPLASM

IN THE U.S...... 12

Abstract ...... 12

Introduction ...... 13

Materials and Methods ...... 16

Results and Discussion ...... 18

Conclusions ...... 23

Literature Cited ...... 23

3 EVALUATION OF TWENTY-ONE MOUNTAIN LAUREL CULTIVARS FOR

CONTAINER AND LANDSCAPE PERFORMANCE IN THE SOUTHEASTERN

U.S...... 34

Summary ...... 35

Introduction ...... 36

Materials and Methods ...... 38

Results and Discussion ...... 41

Literature Cited ...... 48

4 GENETIC RELATIONSHIP AMONG KALMIA LATIFOLIA L. TAXA USING

ISSR MARKERS ...... 60

Abstract ...... 61

Introduction ...... 62

Materials and Methods ...... 64

Results and Discussion ...... 66

Conclusions ...... 72

Literature Cited ...... 73

5 IN VITRO SEED GERMINATION OF KALMIA LATIFOLIA L. HYBRIDS: A

MEANS FOR IMPROVING GERMINATION AND SPEEDING UP BREEDING

CYCLE ...... 86

Abstract ...... 87

vii

Introduction ...... 88

Materials and Methods ...... 90

Results and Discussion ...... 93

Conclusions ...... 98

Literature Cited ...... 99

6 CONCLUSIONS...... 106

viii

LIST OF TABLES

Page

Table 2.1: The geographical information, soil type, and collection information of 15 mountain

laurel populations in the U.S...... 25

Table 2.2: Major morphological characteristics of 21 mountain laurel cultivars in collection ....26

Table 2.3: Inventory of wild mountain laurel collected by the UGA Woody Plant Research

Laboratory crew and collaborators ...... 27

Table 2.4: Inventory of mountain laurel cultivars in 1-gallon and 3-gallon containers ...... 28

Table 3.1: Eleven morphologic characters describing habit, flowering, and leaf characteristics

observed on 21 mountain laurel cultivars for morphological measurements ...... 50

Table 3.2: Analysis of variance table for growth index of 21 mountain laurel cultivars in

container evaluation; performance rating of 21 cultivars in 2014 and 2015 field trial; and

performance rating, net growth, and decrease in Fv/Fm (%) of 10 cultivars in 2016 field

trial ...... 51

Table 3.3: Growth index of container-grown plants of 21 mountain laurel cultivars at the UGA

Hort Farm in Watkinsville, GA after being grown for 1 (2014), 2 (2015), and 4 years

(2017) ...... 52

Table 3.4: Eigenvectors and eigenvalues generated by PCA applied on 11 morphological

characteristics of 21 mountain laurel cultivars ...... 54

Table 3.5: Field performance rating, net growth, and percentage of decrease in maximum

quantum yield of photosystem II (Fv/Fm) of 10 mountain laurel cultivars at the UGA

Riverbend Research Center in Athens, GA in 2016 ...... 55

Table 4.1: Geographic information on sampling populations of mountain laurel ...... 77

Table 4.2: Sequence, total number of bands (# B), number of polymorphic bands (# PB), and the

percentage of polymorphic bands (PPB) of ISSR primers in 69 mountain laurel taxa .....78

Table 4.3: Nei’s unbiased genetic identity (above diagonal) and genetic distance (below

diagonal) for 21 mountain laurel cultivars ...... 79

Table 4.4: Percentage of polymorphic bands (PPB), total genetic diversity (HT), average genetic

diversity within populations (HS), and proportion of total genetic diversity attributed to

among-population variation (GST) for eight mountain laurel populations ...... 80

Table 4.5: Genetic variation among and within populations of eight mountain laurel populations

based on analysis of molecular variation (AMOVA) ...... 81

Table 4.6: Nei’s unbiased genetic identity (above diagonal) and genetic distance (below

diagonal) for eight mountain laurel populations ...... 82

Table 5.1: Effect of basal medium on in vitro germination of 4-MAP hybrid seeds of Kalmia

latifolia ...... 102

Table 5.2: Effect of pH on germination and seedling vigor of 4-MAP Kalmia latifolia hybrid

seeds cultivated in vitro ...... 103

LIST OF FIGURES

Page

Figure 2.1: Natural distribution of mountain laurel in the U.S...... 29

Figure 2.2: Map of 15 mountain laurel populations that were identified and investigated in the

U.S. by the UGA Woody Plant Laboratory breeding program crews and collaborators ..30

Figure 2.3: The variations in flower traits among collected 21 mountain laurel cultivars; cultivars

are diverse in flower size, shape, bud color, corolla color, and pigment pattern ...... 31

Figure 2.4: A) Tree-type mountain laurel plants formed in a big patch at the bottom of a dell that

partially shaded by pine trees in the Red Hill population (Monroe County, AL); (B) the

Fish River population in Baldwin County, AL; (C) mountain laurel (left) that had dark-

pink flowers was found in the Sepulga River population (Escambia County, AL) by Dr.

Ron Miller; (D) the slope at the edge of road in an opening area in the Black Rock

population (Rabun County, GA) was covered by a number of mountain laurel seedlings;

(E and F) the plant having intense blooms and bright pink flowers was found in the

Hitchcock Woods population (Aiken County, SC) ...... 32

Figure 2.5: Wild plants of mountain laurel that rooted from stem cuttings were potted up in 1-

gallon containers and maintained in the shade-house (left). Seedlings germinated from

seeds of wild collection and yielded 15-20 cm of height in 6 months after being sown on

peat moss (right) ...... 33

Figure 3.1: Diversity in size, shape, color, and pigment pattern of flowers among popular

mountain laurel cultivars ...... 57

Figure 3.2: Scatter plot of 21 mountain laurel cultivars obtained performing PCA on 11

morphological characteristics related to habit, stem, leaf, and flower. PC1 and PC2

explain 30.07% and 21.65% of morphological variation, respectively, with cumulative

variance being 51.72%. Twenty-one cultivars were classified into five phenotypic groups

based on PC1 and PC2 values ...... 58

Figure 3.3: Field performance rating of 21 mountain laurel cultivars at the UGA Hort Farm in

Watkinsville, GA after the 2014 growing season (black bar) and 2015 growing season

(gray bar). Plants were evaluated using a scale of 1 (dead) to 5 (excellent) and the rating

was based on overall plant size and vigor, leaf color, incidence of leaf spot, and abiotic

leaf and shoot damage. Data are presented as means + SE (n=4) by the descending order

of ranking in 2015. Different letters indicate that performance ratings of cultivars in 2014

are significantly different at P < 0.05 according to Fisher’s LSD ...... 59

Figure 4.1: UPGMA dendrogram of 21 mountain laurel cultivars generated based on Nei’s

unbiased genetic identity and distance matrix. The accuracy of dendrogram was

demonstrated by the clustering of cultivars with their relatives. Five groups were

observed in dendrogram as shown in the figure ...... 83

Figure 4.2: ISSR fingerprints generated from mountain laurel cultivars using primer UBC835.. 84

Figure 4.3: UPGMA dendrogram of 48 mountain laurel wild accessions (six individuals sampled

from each of eight populations) constructed based on Nei’s unbiased genetic identity and

distance matrix. Six individuals within the population were clustered together.

Meanwhile, populations located in a geographic region were grouped into a cluster and

four clusters correlated with four geographic regions were observed ...... 85

xii

Figure 5.1: In vitro seed germination of mountain laurel hybrids. (A) Capsule development

during the collecting time (from July to October) of crosses ‘Elf’ x ‘Little Linda’ (top)

and ‘Red Bandit’ x ‘Minuet’ (bottom). (B) Seed extracted from the disinfected capsule in

the laminar flow hood, which would then be sowed on medium in 6-cm petri dish. (C)

Seed germinated in 2 weeks after culture as radicle emergence was observed. (D) Eight

weeks after culture, seedlings were photographed for data collection and then

subcultivated to fresh medium. (E) Seedlings in 10-cm petri dish after two cycles of

subcultures; vigorous ones would be transplanted to presoaked peat moss for

acclimatization and further development. (F) Seedling growth after being acclimatized in

the growth chamber for 3 weeks. And (G and H) after 3-month acclimatization, seedlings

were hardened off and yielded significant shoot growth ...... 104

Figure 5.2: Effect of collecting time on in vitro seed germination, including (A) germination

percentage, (B) contamination percentage, (C) T100, indicator for germination speed, and

(D) number of true leaf per seedling, of mountain laurel hybrids ‘Elf’ x ‘Little Linda’

(solid line with closed circle) and ‘Red Bandit’ x ‘Minuet’ (dash line with open circle).

Germination %, contamination %, and number of true leaf were measured 8 weeks after

in vitro culture. T100 was calculated as the number of weeks reaching 100% of total

germination. Data are present as mean of four replications ± SE. Different letters within

each cross combination indicate that they are significantly different at P < 0.05 ...... 105

xiii

CHAPTER 1

INTRODUCTION AND LITERATURE REVIEW

Introduction

Kalmia latifolia L., also known as mountain laurel, is considered by many horticulturists, nursery growers, and gardeners to be one of the most beautiful flowering shrubs and a promising ornamental in the eastern U.S. (Dirr, 2009; Jaynes, 1988). It is a member of genus Kalmia, which belongs to Ericaceae (heath family), along with Rhododendron and Vaccinium and comprises seven native species that are primarily distributed in temperate America (Zomlefer, 1994).

Mountain laurel is usually a tall and spreading shrub that can grow to a mature height of three meters and a width of three meters. Stems range from light green to reddish. The plant has alternate, simple, light green to dark green leathery foliage, especially showy in winter. Leaf blades are elliptic. They are 5 to 12 cm long and less than 5 cm wide. It usually blooms in late spring or early summer after new shoot growth has begun. The inflorescence consists of a terminal compound corymb with numerous flowers that have glandular and sticky stalks. The calyx is five-lobed and is usually persistent in fruit. The corolla has five-lobed petals and bases are fused into a short tube. Each corolla has 10 small pouches holding the anthers and is usually light pink fading to nearly white with purple spots around each anther pocket (Jaynes, 1974). The fruit is a capsule, which turns brown and splits off as it ripens in early fall, containing numerous tiny light brown seeds (Jaynes, 1988).

Natural Distribution

Mountain laurel is restricted to the eastern U.S. and is distributed from southern Maine, west through southern New York to central Ohio, south to eastern Louisiana, southern

Mississippi, Alabama, and Georgia, and northwestern Florida (Kurmes, 1967). Mountain laurel commonly forms dense thickets in rocky and sandy forests throughout most of its range. It is also found in pastures and open fields and often forms thickets at the edges of roads. Like most members of the family Ericaceae, mountain laurel is dependent on a mycorrhizal fungus associated with its roots. This symbiotic relationship ensures adequate absorption of water and minerals by the plants, particularly in acid soils, thus reducing fertilizer and maintenance requirement (Jaynes, 1988).

Cultivar Development

Since the early 1960s, the breeding of mountain laurel, combined with micropropagation to propagate selections, has resulted in the release of many new cultivars. Approximately 140 cultivars have been introduced (The European Kalmia Society), of which about 30 gained popularity with their attractive characteristics and superior landscape performance. Efforts from professional and amateur breeders have yielded many flower colors, such as pure white, deep pink, red, cinnamon, and burgundy with maroon band inside the corolla. Dr. Richard Jaynes at

Broken Arrow Nursery (Hamden, CT) has pursued his passion for genus Kalmia for more than

40 years, and his efforts have significantly increased the availability of cultivars and market shares of mountain laurel in the northeastern U.S. Mountain laurel thus became a popular landscaping plant in the northeastern region and was elected as the state flower of Connecticut

(Jaynes, 1988). In contrast, there is still considerable uncertainty about these plants and their

performance in areas of USDA zone 7, particularly in the southeastern U.S. In the southeastern

U.S., few cultivars are available in retail settings and those cultivars are principally produced by niche Ericaceous nurseries. Ironically, although produced in several southeastern nurseries, mountain laurel is rarely used in southeastern landscapes. The majority of mountain laurel cultivars were evaluated (prior to release) and selected for release in the northeastern U.S. As a result, the performance of cultivars outside the region is relatively unknown, which has historically limited cultivar acceptance in other regions of the U.S. (Jaynes, 1982, 1988). An example of this regionality in cultivar development and use is ‘Sarah’, selected by Dr. Richard

Jaynes at the Connecticut Agricultural Experiment Station. While well adapted to northeastern landscapes, no documentation exists regarding its broader adaptability due to a lack of evaluation outside the northeast.

Propagation

Seed germination. In England, mountain laurel is raised from seed in large quantities, which is a relatively cheaper way of producing plants. Seeds of mountain laurel are very small

(0.9 mm long, 0.3 mm wide) and require careful handling for acceptable germination. A mix of peat moss and perlite in a 2:1 ratio is used as substrate that should be watered thoroughly before sowing. Seeds should be sown on the surface of the mix and not covered, since they require light to germinate (Malek et al., 1989). The moist substrate and high humidity are necessary for germination (Jaynes, 1971). Seed germination could take up to three months. Cold treatment for two months or soaking seeds overnight in 200 ppm GA3 will increase germination to 50%

(Flemer, 1949; Taylor et al., 2009). In addition to slow and poor germination, it takes two to three years to produce a plant 7.5 cm tall when propagated by seeds. Better germinating seeds

and improving seedling growth of mountain laurel are essential to breeding projects (Jaynes,

1988).

Cutting propagation. Cutting propagation is the most common way for large-scale commercial production of clonal plants (Dirr, 2006). Unfortunately, mountain laurel is generally difficult to root. Although many researchers and propagators found that the use of auxin, mist, temperature, rooting media, and cutting types helped its rooting, these techniques are not helpful enough for widespread commercial success (Hansen and Potter, 1997; Jaynes, 1982; Williams and Bilderback, 1980). Eichelser (1978) recommended that treating juvenile cuttings with 5,000 ppm 3‐indolebutyric acid (IBA) and 1‐naphthalene acetic (NAA) and rooting them in a mixture of 40% fir sawdust, 40% cedar sawdust, 10% perlite, and 10% peat moss. A mist system (10 s every 10 mins during daylight hours) and bottom heat of 75 °F need to be supplied. 60‐75% rooting can be obtained after five months.

Micropropagation. Although propagating mountain laurel from cuttings is unreliable, tissue culture has made the large-scale propagation of mountain laurel not only possible but also highly successful (Dirr, 2006). The standard procedure for micropropagation of mountain laurel is as follows: a segment of shoot with axillary bud is cultured in Woody Plant Medium (Lloyd and McCown, 1980) supplemented with 1.0-3.0 mg·L-1 6-(γ,γ-Dimethylallylamino)purine (2iP).

Healthy shoots are then transferred to multiplication medium, and subculture needs to be made every four weeks to keep shoots growing vigorously. In general, microcuttings are dipped with

IBA and rooted in trays filled with screened peat moss and perlite at the ratio of 1:1. Trays are placed in the growth chamber, and light (14 h) and moisture are provided. After microcuttings produce roots, the trays are moved to the greenhouse for acclimation. Cuttings are transplanted to small pots after they have been hardened-off (Lloyd and McCown, 1980; Nishimura et al., 2004).

Management Practices

Container production. In the northeastern U.S., shade is not needed to grow high-quality mountain laurel since shading reduced their growth. Moderate levels of shading could be provided during the first year of production if enhanced color of foliage was desired (Brand,

1997). However, in the southeastern U.S., container-grown mountain laurel required shade from high pine trees or 50% shade for good production (Bir and Bilderback, 1989). Growth of mountain laurel was also affected by day/night temperature. Malek et al. (1992) found that root growth was more sensitive to high temperature than top growth, and the optimal day and night temperatures for both type of growth were 26 °C and 22 °C, respectively. Additionally, fertilizer played an important role on container-grown mountain laurel. It was considered to have the same requirements as other ericaceous plants that need low fertility; the medium rate of nitrogen at 80 mg per pot applied as a liquid once every two weeks induced good root and shoot growth

(Hummel et al., 1990).

Establishment. Landscape establishment of mountain laurel was difficult, and customers were often disappointed when plants died after transplanting to their home landscapes.

Improving the survival of transplanted container-grown mountain laurel may encourage an increase in its commercial production and landscape utilization. The main difficulty in getting container-grown mountain laurel to quickly establish in landscapes was probably the relatively short root extension into backfill-soil in the first season after transplanting. In warmer climates, the smaller size mountain laurel, fall transplants, and east exposure had longer root extensions, which contributed to better landscape establishment. In cooler climates, initial plant size and exposure might not be critical factors (Hanson et al., 2004; Wright et al., 2005).

Breeding

Flower morphology and breeding ecology. The pollen-discharge mechanism is one of the most distinctive characteristics of the Kalmia genus (Jaynes, 1988). The flower has five fused petals and 10 small pouches at the middle of the corolla. Before the flower bud opens, 10 filaments elongate and push the anthers into the 10 pouches. When the flower opens, slender filaments bend backward under tension while anthers are held in the pouches. When a pollinator lands on the flower and touches the filaments, one or more anthers are released. The strong tension of the filament allows the pollen to be thrown from the flower and stick to the pollinator.

The pollinator then deposits the pollen on the stigmas of subsequently visited flowers (Nimmo et al., 2014). This mechanism ensures cross-pollination in most cases (Jaynes, 1988). However, the anthers of mountain laurel will eventually release without a visitor at the end of floral life. This autonomous selfing assures reproduction in the absence of pollinators (Nagy et al., 1999).

Although mountain laurel can self, Rathcke and Real (1993) documented reduced fruit set on self-pollinated mountain laurel growing in Virginia and Rhode Island compared to out-crossed plants, which could be the result of self-incompatibility or inbreeding depression. Jaynes (1968) found no significant difference in pollen tube growth between crosses and selves, nor differences in germination percentage of seeds from self and cross pollinations. Therefore, mountain laurel is no longer considered as a self-incompatible species. However, Jaynes (1968) found that there was a significant reduction in the survival and vigor of seedlings from self-pollinated plants, as well as a reduction in seed set, which supports the inbreeding depression theory.

Hand pollination. Reciprocal crosses of mountain laurel are identical and usually have equal seed sets. Therefore, it does not matter which cultivar is used as the female parent in a cross (Jaynes, 1988). Flower buds that are just about to open can be selected as female parents.

Stamens and anthers need to be emasculated before buds open, and the corolla of emasculated flowers and other flower buds of the inflorescence need to be removed to decrease insect activity and save energy for the female parent. When the stigma is viscid and moist, indicating it is receptive, pollen from the selected male parent can be applied onto the stigma. The filament must be gently removed from the male plant and tapped with forceps on the surface of the stigma, and pollen will be released through the anther pore and caught by the sticky stigmatic surface. To avoid pollen contamination, pollinated flowers should be bagged (Jaynes, 1988).

Research Needs

Additional information and research are needed to continue making progress in mountain laurel breeding. Firstly, new cultivars that have combined favorable traits are needed. Favorable traits include compact habit, dark green foliage, stunning flowers, and etc. Moreover, it is desired to breed and select plants in the southernmost or northernmost limits of the native range in order to obtain widely adapted mountain laurel selections. Additionally, it is important to define existing cultivars, which requires information on heat and cold tolerance, shade preference at multiple locations, leafspot resistance, morphological traits, and landscape performance at different growing conditions (Jaynes, 1988).

Research Objectives

In the past five years, I visited a large number of nurseries and gardens in the eastern

U.S., including Maine Coastal Botanic Garden (Maine), Arnold Arboretum (Massachusetts),

Broken Arrow Nursery (Connecticut), Duke Garden (North Carolina), Bartlett Tree Lab (South

Carolina), Dick’s private garden (South Carolina), Woodlanders Nursery (South Carolina),

Transplant Nursery (Georgia), Brasstown Bald (Georgia), and Tallulah State Park (Georgia), to investigate container production and landscape utilization of mountain laurel. We found that mountain laurel is popular in the northeastern U.S., where it is widely used in yards and landscapes. By contrast, there is still considerable uncertainty about these plants in the southeastern U.S. The main reason is due to that majority of mountain laurel cultivars were developed, evaluated, and selected for release in the northeastern U.S., while their performance in other regions was unknown. Southeastern climate varies significantly from that of the northeastern U.S. Milder winter temperatures lead to less cold damage in nurseries and landscapes. Yet hotter and more humid summer conditions, as well as a lack of winter chilling hours, could limit adaptability of cultivars selected in the northeastern U.S. and transitioned into the southeastern nurseries and landscapes.

As a promising ornamental plant for the southeastern U.S., mountain laurel plants having adaptability to southeastern environmental conditions are desired. Breeding of woody plants has been challenging for breeders, since the breeding cycle is generally long, taking up to 10-20 years. For instance, the first two mildew resistant Lagerstroemia hybrids of L. indica x L. fauriei,

‘Natchez’ and ‘Muskogee’, were introduced in 1978, which was 19 year later after the disease resistant L. fauriei was found in Japan (Egolf, 1981; Pooler, 2017). As reviewed literatures and communications with breeders indicate, the main difficulties in mountain laurel breeding are obtaining the resources that provide traits of interest, good seed germination, and fast growth of seedlings. Additionally, limited information is available on the genetic relationships of mountain laurel and performance of existing cultivars in the southeastern U.S. Therefore, it is of significance to 1) collect mountain laurel germplasm; 2) evaluate existing cultivars for container

and landscape performance in the southeastern U.S.; 3) investigate genetic relationships among mountain laurel taxa; 4) enhance seed germination for mountain laurel breeding program.

Literature Cited

Bir, R.E. and T.E. Bilderback. 1989. Growing better mountain laurel in containers. Intl. Plant

Prop. Soc. Proc. 39:442-447.

Brand, M.H. 1997. Shade influences plant growth, leaf color, and chlorophyll content of Kalmia

latifolia L. cultivars. HortScience 32:206-208.

Dirr, M.A. 2006. The reference manual of woody plant propagation. Timber Press, Portland,

London.

Dirr, M.A. 2009. Manual of woody landscape plants. 6th ed. Stipes Publishing, Champaign, IL.

Egolf, D.R. 1981. ‘Muskogee’ and ‘Natchez’ Lagerstroemia. HortScience 16:576.

Eichelser, J.E. 1978. Propagation of Kalmia. Intl. Plant Prop. Soc. Proc. 28:133-134.

Flemer, W. 1949. The propagation of Kalmia latifolia from seed. Bul. Torrey Bot. Club 76:12-

16.

Hansen, O.B. and J.R. Potter. 1997. Rooting of apple, rhododendron, and mountain laurel

cuttings from stock plants etiolated under two temperatures. HortScience 32:304-306.

Hanson, A.M., J.R. Harris, and R. Wright. 2004. Effects of transplant season and container size

on landscape establishment of Kalmia latifolia L. J. Environ. Hort. 22:133-138.

Hummel, R.L., C.R. Johnson, and O.M. Lindstrom. 1990. Root and shoot growth response to

three container-grown Kalmia latifolia L. cultivars at two locations to growing medium

and nitrogen form. J. Environ. Hort. 8:10-13.

Jaynes, R.A. 1968. Self-incompatibility and inbreeding depression in three laurel species. Amer.

Soc. Hort. Sci. Proc. 93:618-622.

Jaynes, R.A. 1971. Seed germination of six Kalmia species. J. Amer. Soc. Hort. Sci. 96:668-672.

Jaynes, R.A. 1974. Propagation of chestnut and other difficult-to-root plants from softwood

cuttings. HortScience 9:301-304.

Jaynes, R.A. 1982. New mountain laurel selections and their propagation. Intl. Plant Prop. Soc.

Proc. 32:431-434.

Jaynes, R.A. 1988. Kalmia: The laurel book. Timber Press, Portland, OR.

Kurmes, E.A. 1967. The distribution of Kalmia latifolia L. Amer. Midland Naturalist 77:525-

526.

Lloyd, G. and B. McCown. 1980. Commercially-feasible micropropagation of mountain laurel,

Kalmia latifolia, by use of shoot-up culture. Intl. Plant Prop. Soc. Proc. 30:421-427.

Malek, A.A., F.A. Blazich, S.L. Warren, and J.E. Shelton. 1989. Influence of light and

temperature on seed germination of mountain laurel. J. Environ. Hort. 7:161-162.

Malek, A.A., F.A. Blazich, S.L. Warren, and J.E. Shelton. 1992. Initial growth of seedlings of

mountain laurel as influenced by day/night temperature. J. Amer. Soc. Hort. Sci.

117:736-739.

Nagy, E.S., L. Strong, and L.F. Galloway. 1999. Contribution of delayed autonomous selfing to

reproductive success in mountain laurel, Kalmia latifolia (Ericaceae). Amer. Midland

Naturalist 142:39-46.

Nimmo, J.R., P.M. Hermann, M.B. Kirkham, and E.R. Landa. 2014. Pollen dispersal by catapult:

Experiments of Lyman J. Briggs on the flower of mountain laurel. Phys. in Perspective

16:371-389.

Nishimura, T., S. Hasegawa, A. Meguro, R. Yoshida, and H. Kunoh. 2004. Micropropagation

technique for mountain laurel (Kalmia latifolia L.) using cytokinin and growth retardant.

Jpn. J. Crop Sci. 73:107-113.

Pooler, M. 2017. Good genes - Using germplasm and breeding to create new plants at the U.S.

National Arboretum. Acta Hort. 1185:1-5.

Rathcke, B. and L. Real. 1993. Autogamy and inbreeding depression in mountain laurel, Kalmia

latifolia (Ericaceae). Amer. J. Bot. 80:143-146.

Taylor, L.L., R. Conev, and J.R. Harris. 2009. Germination response to seed pretreatments in two

genotypes of mountain laurel (Kalmia latifolia L.). HortScience 44:1102-1103.

The European Kalmia Society. Laurel cultivars - The international Kalmia register and checklist.

27 Oct. 2018. < http://www.kalmia-society.org/cultivars.php?&lang=gb>.

Williams, R.F. and T.E. Bilderback. 1980. Factors affecting rooting of Rhododendron maximum

and Kalmia latifolia stem cuttings. HortScience 15:827-828.

Wright, A.N., S.L. Warren, F.A. Blazich, J.R. Harris, and R.D. Wright. 2005. Initial plant size

and landscape exposure affect establishment of transplanted Kalmia latifolia ‘Olympic

Wedding’. J. Environ. Hort. 23:91-96.

Zomlefer, W.B. 1994. Guide to flowering plant families. University of North Carolina Press,

Chapel Hill, NC.

CHAPTER 2

EXLORATION AND COLLECTION OF KALMIA LATIFOLIA L. GERMPLASM IN THE

U.S.

Abstract. Kalmia latifolia L. (mountain laurel) is a promising ornamental shrub by its evergreen foliage and attractive inflorescences. A breeding program was initiated by the Woody Plant

Laboratory at the University of Georgia in 2014, the aim of which was to breed and select mountain laurel cultivars as ornamentals for the southeastern U.S. Breeding elite cultivars relies on plant genetic resources including wild plants and existing cultivars. This study was conducted to collect mountain laurel germplasm that might provide desirable traits for the current breeding program. An exploration and collection mission for wild populations was carried out from 2015 to 2018 in the eastern U.S. A total of 15 populations were identified in Alabama, Florida,

Georgia, Louisiana, Massachusetts, and South Carolina; and collections were made from 10 of them. A total of 277 plants was obtained from wild collections, of which 69, 186, and 22 were from cuttings, seeds, and individual plants, respectively. From more than 100 existing cultivars, twenty-one were selected for collection because of their reported superior performance, popularity, and diverse traits. We currently collected 197 plants for these 21 cultivars, with 93 and 104 in 1-gallon and 3-gallon containers, respectively. The collection of mountain laurel wild plants and cultivars benefits the current breeding program by providing promising genetic resources and the exploration and documentation of wild populations will provide detailed information in subsequent plant collection and population genetic studies.

Additional index words. mountain laurel, ornamental breeding, wild populations, cultivars, southernmost range, documentation

Introduction

Kalmia latifolia L. (mountain laurel), an evergreen flowering shrub, is a member of

Ericaceae or heath family, along with rhododendrons and blueberries, that occurs throughout temperate areas of the world (Jaynes, 1988). Mountain laurel has been considered by many horticulturists, breeders, and gardeners to be a fascinating native species to the United States

(Dirr, 2009; Jaynes, 1988). Its lush foliage, outstanding inflorescences, and hardiness made mountain laurel a valuable ornamental shrub in the nursery and landscape industries (Jaynes,

1988). More than 100 mountain laurel selections have been developed since the early 1960s, yet there is still considerable ignorance about these plants, particularly in the southeastern U.S.

Fewer than 30 mountain laurel cultivars gained popularity due to their attractive ornamental traits and superior landscape performance whereas others are no longer in commercial production. The majority of commercial cultivars were developed and selected for release in the northeastern U.S. As a result, the performance of cultivars outside the northeast is relatively unknown and this uncertainty has therefore limited cultivar acceptance in other regions of the

U.S., particularly in the southeast. In the southeastern U.S., few cultivars are available in retail settings, and they are principally produced in niche Ericaceous nurseries. Ironically, although produced in several southeastern nurseries, mountain laurel is rarely used in southeastern landscapes (Jaynes, 1988). In order to improve the production and use of mountain laurel in the southeast, adaptability to southeastern environmental conditions is considered to be the major improvement by breeders. In terms of ornamental traits, as both container and landscape plants,

mountain laurel that has compact or dense habits, lush foliage, and stunning flowers is always desired (Jaynes, 1982). Moreover, since consumer needs and environmental challenges are constantly changing, breeders must be prepared to obtain new traits and meet future needs

(Engels et al., 1995).

Being responsible to develop new elite cultivars, plant breeders always need to rely on plant genetic resources, including wild sources and cultivated varieties, to introduce the desirable traits (Fehr, 1987). Without this germplasm, breeders would be struggling to find sources providing favorable traits. For instance, crape myrtle is one of the most widely planted and recognized woody ornamental plants in the southern U.S. However, it was not such a popular plant until the mid-1970s since all Lagerstroemia indica cultivars were susceptible to powdery mildew. In 1959, a single new germplasm introduction, PI237884 L. fauriei, collected from

Japan by John Creech was identified to be resistant to powdery mildew. Later, the hybrids of L. indica x fauriei, ‘Natchez’ and ‘Muskogee’, having excellent mildew resistance were introduced, which completely changed the fate of crape myrtle (Egolf, 1981; Pooler, 2017). In order to develop mountain laurel cultivars for the southeastern U.S., exploring and collecting germplasm that would enhance adaptability to southeastern environmental conditions and introduce desirable ornamental traits and other novel traits is important to breeding programs.

In the United States, mountain laurel is naturally distributed in the east, specifically from southern Maine (USDA Zone 4a) west through southern New York (4a) to central Ohio (6a), south to eastern Louisiana (8b), southern Mississippi (8b), Alabama (8b), Georgia (8b), and northwestern Florida (8b) (Kurmes, 1967, Fig. 2.1). The southernmost mountain laurel populations have evolved in southeastern environmental conditions. As a result, wild plants from this geographic region are well adapted to hot and humid summer conditions and mild winter

temperatures and they are therefore promising germplasm that should improve adaptability of mountain laurel cultivars to the southeastern environmental conditions. Wild plants from other portions of its natural range, particularly northeastern populations, were reported to have greater variation in flower traits, which were therefore regarded as an essential starting point for introduction of novel ornamental traits. In addition to traits of interest, greater genetic diversity is needed to develop new mountain laurel cultivars. Historically, collecting wild species from regions with various habitats has been a means to increase genetic diversity for breeding programs worldwide (Engels et al., 1995). The wild populations in different geographic areas are valuable resources beneficial to breeders. On the other hand, wild mountain laurel plants generally carry some unfavorable morphological traits, like sparse habit and diminished leaf growth, due to the lack of artificial selection (Jaynes, 1988). In this situation, breeders could not solely rely on wild resources while also need to turn to existing cultivars. Mountain laurel cultivars having desirable traits like dense habits, lustrous dark green foliage, and showy flowers should be targeted and collected. Followed by evaluation, cultivars that could be adapted to the southeast might be directly selected for landscape use or as valuable materials for breeding mountain laurel plants for the southeastern U.S.

A breeding program was initiated by the Woody Plant Laboratory at the University of

Georgia in 2014 for breeding and selecting mountain laurel cultivars as ornamentals for the southeastern U.S. The objective of this study was to document the exploration and collection of mountain laurel wild plants and cultivars in the U.S., which was the foundation for the current breeding program by collecting genetic resources and providing detailed wild population information for further research.

Materials and Methods

Gathering Collection Information. Wild populations were identified through the USDA distribution map (Fig. 2.1), which provides county-level records of mountain laurel natural distribution, description from “Kalmia-Mountain Laurel and Related Species” (Jaynes, 1988), and information obtained from local informants. The main target, southernmost distribution, was reported to be located in southern Alabama and northwestern Florida. Other populations in the southern range, including Georgia and South Carolina, were targeted because of their diverse habitats. The northeast range, wild populations in Massachusetts specifically, was explored due to the potential of having unique traits and increasing genetic diversity.

Collecting trips and documentation. In November 2015, a trip was made to explore and collect wild mountain laurel plants in the southernmost range. The itinerary included six populations in Monroe County (AL), Baldwin County (AL), and Santa Rosa County (FL) (Fig.

2.2). Using a state vehicle, we drove to Monroeville, AL and met up with Dr. Ron Miller who aided in the localization of these southernmost populations. A motorboat was operated to get access to the Fish River population in Baldwin County, AL and other populations were easily reached by the state vehicle. A seed-lot collected from Pushepetappa Creek in Louisiana by Dr.

Ron Miller was shared with us during the trip. In May 2016, another trip was made in this area for the blooming season and recording flower traits. On 12 May 2017, populations in Aiken

County and Lexington County, SC were identified under the guidance of Dr. Ron Hooper (Fig.

2.2). In the population located in Aiken County, one plant was targeted due to its vigorous growth, intense blooms, and bright pink flowers, a collection was thus conducted from this plant on 21 November 2017. On 13 February 2018, with the help of Mr. Jack Johnston, localization of mountain laurel natural distribution in Rabun County, GA (north mountain population) was

achieved and plants from the Black Rock population were collected. One Georgia piedmont population located in Clarke County was identified through communication with Dr. Dayton

Wilde and was explored on 22 February 2018 (Fig. 2.2). GPS coordinates and altitude of each population were recorded using Compass on an iPhone. All sites with collected or un-collected samples were recorded. Soil information was obtained from USDA Web Soil Survey based on

GPS coordinates (https://websoilsurvey.sc.egov.usda.gov/App/HomePage.htm) (Table 2.1). One mountain laurel plant that had dark pink flowers from an additional southernmost population was exploited by Dr. Ron Miller, though only GPS coordinate information was provided. Mature capsules were collected from this plant by him and shipped to us in December 2016. Two wild sites in Massachusetts were explored by Mr. Connor Ryan on 30 July and 2 August 2016, respectively, and only GPS coordinates were available (Table 2.1, Fig. 2.2).

Wild plant collection. For some wild populations, the healthy and good-looking plants were selected for collection and GPS coordinate of each collected plant was recorded along with altitude. Whole plants were dug out and collected from the Black Rock population, while stem cuttings and/or mature capsules were collected from other populations. Semi-hardwood cuttings were placed in black plastic bags and sprayed with water after being collected from plants. They were trimmed to 10-15 cm long and leaves at the bottom 5-7 cm were stripped. Cuttings were dipped into 8,000 ppm K-IBA for 10 seconds, then air-dried for at least 10 minutes. Treated cuttings were inserted into the rooting medium that contained Fafard ProMix (Sun Gro

Horticulture, Agawam, MA) and perlite at 1:1 (v:v). Cuttings were rooted in 32-cell flat trays and thoroughly watered. Trays were then placed on a mist bench and the mist system was set for

20 seconds every 20 minutes for the first week, then 10 seconds every 20 minutes thereafter.

Collected mature capsules were stored in envelopes at room temperature to air dry for a few

days. Capsules were crushed on a piece of wax paper and seeds were then cleaned using a 0.5- mm round-hole sieve. Seeds were carefully sown on the surface of pre-soaked peat moss since germination requires light. Germination took place in a growth chamber at 25 ± 2 °C under a 16- h photoperiod provided by cool-white fluorescent lamps at PPF of 66 mol·m-2·s-1. Seeds were watered as needed. Surviving plants were maintained in an unheated cold frame covered with

60% shade cloth at University of Georgia Durham Horticulture Farm (Watkinsville, GA). Plants were irrigated twice a day (15 min per irrigation event) using 10 ft. full-circle nozzles monitored by a timer.

Cultivar collection. Twenty-one mountain laurel cultivars were selected based on their popularity, reported superior landscape performance in the northeast, diversity in morphological traits, and desired ornamental traits including dense habits, lustrous dark green foliage, and showy flowers (Table 2.2, Fig. 2.3). Liners of 21 cultivars were shipped from Briggs Nursery,

Inc. (Elma, WA) to the University of Georgia Durham Horticulture Farm on 11 March 2013 in

49-cell flats. Liners were micropropagated and ranged 8-10 cm tall and 5-6 cm wide. Plants were immediately transplanted into 1-gallon containers filled with pine bark-based soilless substrate

(85% Bark Pine and 15% Compost; Sun Gro) after being received. Potted plants were top- dressed with 5 g of 20N-3P-9K controlled-release fertilizer (Everris, Marysville, OH) and held in the same shade-house as described above.

Results and Discussion

A total of 15 mountain laurel wild populations was explored in the U.S (Table 1.1, Fig.

1.2). Eight southernmost populations, including five in southern Alabama, one in northeastern

Louisiana, and two in northwestern Florida were targeted. Southeastern populations, comprising

three Georgia and two South Carolina populations, were identified. Two northeastern populations were located in Massachusetts. Similar as what Dr. Richard Jaynes described

(Jaynes, 1988), mountain laurel commonly forms dense thickets in rocky and sandy forests throughout most of its range, particularly where there are soils with good drainage and openings in the canopy. We will review each of the regions in closer detail.

Southernmost ranges. Mountain laurel is naturally distributed in the eastern U.S., specifically from Maine south to Florida. With the help of Dr. Ron Miller, the southernmost ranges were successfully identified in southern Alabama and northwestern Florida during the trip in November 2015. As the main target of our mountain laurel breeding program, six populations located in this area were visited and collected.

The Big Flat Creek population in Monroe County, AL was found in the canopy at the edge of Road 17. A few mountain laurel plants growing along with evergreen rhododendrons could be accessed. These plants exhibited somewhat poor growth, which was probably due to the slow runoff from not well-drained soil. Since most capsules were dried out and dehisced, only semi-hardwood cuttings were collected. Twenty-four stem cuttings were prepared and six of them were rooted and survived (GHW15003, Table 2.3). In contrast, a big population was present in Haines Island Park in Monroe, AL. The majority of mountain laurel plants were distributed on the top of the hill under a pine forest. Two plants growing along the cliff were selected for collection due to the foliage colors of new shoots and attractive morphology.

Twenty-two and 24 cuttings were collected, with six (GHW15004) and 16 plants (GHW15005) being obtained from them, respectively. The Red Hill population in Monroe, AL was the most impressive one that we visited in Monroe County. This population was located on the top of a hill, being separated into two different habitats. Some tree-type plants formed a big patch at the

bottom of a dell that was partially shaded by pine trees (Fig. 2.4A), while others, including some newly grown plants, were growing in open areas along with Ilex opaca. Capsules and cuttings were collected from plants growing in both habitats. Collections GHW15011-13 (semi-hardwood cuttings) and GHW15017-19 (seeds) were made from three tree-type plants. Cuttings from these plants yielded almost no rooting, which did not surprise us too much, because mountain laurel is a difficult-to-root species and cuttings from mature plants are less likely to survive. Seed germination was also challenging for us, yet we obtained considerable number of seedlings from each collection. Cuttings of juvenile mountain laurel (GHW15014&15) in the open area produced relatively more plants while seeds (GHW15020) had similar germination as those of tree-type plants (Table 2.3).

The population along the Fish River in Baldwin County, AL has a very distinct habitat compared with Alabama populations mentioned above (Fig. 2.4B). Fish River is part of the tidal system associated with Mobile Bay and contains both freshwater and saltwater. During summer, this population can be flooded by the Fish River and mountain laurel plants are submerged for a few days without any problem. One plant growing at the edge of the land (GHW15026&28) and another growing at inner land (GHW15030&31) were selected for collection. Both stem cuttings and fruit capsules were accessed. Similarly, a number of seedlings was obtained, while only a few cuttings were survived due to maturity (Table 2.3).

Two spots, Indian Ford Bridge and Blackwater Forest, in Santa Rosa County, FL were investigated. The habitats of these two populations were similar, where wild mountain laurel grew under the tree canopy along a creek. Cuttings from both populations were rooted. A total of

23 seedlings was produced by the seeds collected from Indian Ford Bridge (GHW15040). The

majority of capsules had dehisced in Blackwater Forest, so only three seedlings were available

(GHW15039).

A seed-lot collected from Pushepetappa Creek area in Washington Parish County, LA by

Dr. Ron Miller was shared with us during the trip (GHW15043), and 13 seedlings were obtained from this collection. An additional seed collection made along the Sepulga River in Escambia

County, AL (GHW16001) was shipped to us by Dr. Ron Miller in December 2016. Wild mountain laurel plants in southernmost populations generally have white or faded-pink flowers.

However, Dr. Ron Miller found one plant in the Sepulga River population that had dark-pink flowers (Fig. 2.4C). He therefore went back to this population in the fall season in order to collect capsules from this plant and send these valuable resources to us. From this seed-lot, 24 seedlings have been obtained so far.

Georgia and South Carolina. Georgia is a geographically diverse state and the main geographical features include mountains in the northeast, the piedmont in the central portion of the state, and coastal plain in the south. Mountain laurel populations in the southwest coastal plain are more like the southernmost ones we have exploited, while plants naturally distributed in the northern mountains and central piedmont were expected to differ due to diverse habitats.

Two populations, Piedmont College and Black Rock, in Rabun County were identified by Mr.

Jack Johnston. Both of them were at an elevation of more than 2,000 feet and grew along with evergreen rhododendrons under pines and oaks. In the Black Rock population, one slope at the edge of road in an open area was covered by a number of mountain laurel seedlings (Fig. 2.4D).

These seedlings germinated from seeds that dispersed in fall season 2017, successfully overwintered, and went through the slow-growing phase and continued to grow in spring 2018.

Twenty-two seedlings, instead of cuttings or capsules, were thus directly collected from the

cleared roadside and potted in 1-gallon containers, and all of them reached 25 cm of height in seven months after being collected (GHW18001). A small patch located in Sandy Creek Park at

Clarke County, GA was investigated and about 20 plants were found along the East Oconee

River in February 2018. Leaf samples for DNA extraction were collected.

With the help of Dr. Ron Hooper, the localization of the Hitchcock Woods population in

Aiken County, SC and Peachtree Rock population in Lexington County, SC was achieved. A point should be highlighted here was that the only one cultivar, ‘Pristine’, developed in the southeast, was selected from a seed-lot collected around Aiken. Therefore, the Hitchcock Woods population in Aiken was collected. In this population, mountain laurel only grew along one side of Kalmia Trail and extended to the south-facing slope, which was somewhat open. One plant growing at the slope bottom had impressive plant size and shape, vigorous growth, intense blooms, and bright pink flowers caught our eyes (Fig. 2.4E&F). In order to collect both capsules and cuttings from it, another trip was made in November 2017. Twenty-one seedlings were obtained (GHW17001); unfortunately none of the cuttings rooted (GHW17002).

Massachusetts. Mr. Connor Ryan explored two sites, Middlesex Fells and Blue Hills, in

Massachusetts in summer 2016. Both places were in deep and dark woodland and mountain laurel plants were scattered rather than in clusters. Leaf samples were collected for DNA extraction.

Cultivar collection. Twenty-one cultivars displayed diverse growth and morphological traits, with foliage and flower traits being the major variations (Table 2.2, Fig. 2.3). All mountain laurel cultivars reached more than 95 cm of plant size (height + width at the widest direction + width at the perpendicular direction) after being grown in 1-gallon containers for two years

(Table 2.4). On 21 May 2015, four plants of each cultivar were transplanted from 1-gallon to 3-

gallon containers filled with bark-based substrate as described previously for subsequent growth evaluation. Plants were fertilized with 15 g of 20N-3P-9K controlled-release fertilizer annually and maintained in the same shade-house as described above. Cultivars in 3-gallon containers yielded plant sizes ranging from 156 to 301 cm after being grown for another two year. The number of plants in both 1-gallon and 3-gallon pots were recorded (Table 2.4).

Conclusions

With the objective of breeding mountain laurel plants for the southeastern U.S., a total of

15 populations in the eastern U.S. that may provide desirable resources was identified and collections were made from 10 of them. From our wild collection, 277 plants were obtained, of which 69, 186, and 22 were from cuttings, seeds, and plants, respectively (Fig. 2.5). Twenty-one commercial cultivars were selected for collection because of their reported superior performance and diverse and favorable traits. A total of 197 cultivar plants are currently on our hands, with 93 and 104 in 1-gallon and 3-gallon containers, respectively. The exploration and documentation of mountain laurel wild populations will assist in plant collection and population genetics study in the future by providing detailed geographic location and other information. The collection of a variety of mountain laurel wild plants and commercial cultivars will benefit breeding programs in the southeastern U.S. by providing promising genetic resources.

Literature Cited

Dirr, M.A. 2009. Manual of woody landscape plants. 6th ed. Stipes Publishing, Champaign, IL.

Egolf, D.R. 1981. ‘Muskogee’ and ‘Natchez’ Lagerstroemia. HortScience 16:576.

Engels, J.M.M, R.K. Arora, and L. Guarino. 1995. An introduction to plant germplasm

exploration and collecting: Planning, methods, and procedures, follow-up, p. 21-63. In:

Guarino, l., V. R. Rao, and R. Reid (eds). Collecting plant genetic diversity: Technical

guidelines. CAB Intl., Wallingford, U.K.

Fehr, W.R. 1987. Principles of cultivar development. Volume 1. Theory and technique.

Macmillan Publishing, New York.

Jaynes, R.A. 1982. New mountain laurel selections and their propagation. Intl. Plant Prop. Soc.

Proc. 32:431-434.

Jaynes, R.A. 1988. Kalmia: The laurel book. Timber Press, Portland, OR.

Kurmes, E.A. 1967. The distribution of Kalmia latifolia L. Amer. Midland Naturalist 77:525-

526.

Pooler, M. 2017. Good genes - Using germplasm and breeding to create new plants at the U.S.

National Arboretum. Acta Hort. 1185:1-5.

Table 2.1. The geographical information, soil type, and collection information of 15 mountain laurel populations in the U.S.

Population County State Coordinate Altitude Soil type Collected or not? 31°36'31.47"N Big Flat Creek Monroe AL 140 ft Iuka and Mantachie soil Yes 87°24'51.29"W 31°43'16.70"N Haines Island Monroe AL 180 ft Arundel loam Yes 87°27'48.14"W 31°44'17.53"N Red Hill Monroe AL 450 ft Gravelly sandy loam Yes 87°21'30.42"W 30°28'16.89"N Fish River Baldwin AL 0 ft Wahee silt loam Yes 87°48'1.92"W 30°57'6"N Pushepetappa Creek Washington Parish LA - Ouachita, Bibb, and Jena soils Yes 89°55'0"W 31°12'6.77"N Chewacla-Lenoir-Riverview Sepulga River Escambia AL - Yes 86°46'20.32"W association 30°43'35.42"N Indian Ford Bridge Santa Rosa FL 30 ft Bibb-Kinston association Yes 86°53'56.38"W 30°47'23.63"N Blackwater Forest Santa Rosa FL 130 ft Troup loamy sand Yes 86°47'33.46"W 33°33'9"N The Hitchcock Woods Aiken SC 390 ft Vaucluse-Ailey complex Yes 81°44'33"W 33°49'47"N Peachtree Rock Lexington SC 490 ft Vaucluse loamy sand No 81°11'57"W 34°51'50"N Piedmont College Rabun GA 2120 ft Hayesville fine sandy loam No 83°23'16"W 34°55'3"N Black Rock Rabun GA 2650 ft Edneyville-Ashe association Yes 83°24'38"W 34°0'57.97"N Congaree soils and alluvial Sandy Creek Clarke GA - No 83°22'32.59"W land Hollis-Rock outcrop-Charlton 42°26'51"N Middlesex Fells Middlesex MA - complex No 71°5'26"W

42°12'55"N Blue Hills Norfolk MA - Freetown muck No 71°3'3"W

Table 2.2. Major morphological characteristics of 21 mountain laurel cultivars in collection.

Cultivar Habit Leaf sizez Leaf color Flower sizey Bud color Corolla color Bullseye Sparse and typical Large Dark green Medium Pink White Carol Dense and typical Large Dark green Large Red Light pink Elf Sparse and dwarf Small Medium to dark green Small Light pink White Firecracker Dense and typical Medium Dark green Medium Red Light pink Forever Red Sparse and typical Large Dark green Medium Red White Freckles Sparse and typical Large Dark green Medium Pink Pink Heart of Fire Sparse and typical Large Dark green Large Red Light pink Kaleidoscope Dense and typical Medium Medium to dark green Medium White White Little Linda Dense and dwarf Small Dark green Medium Pink Light pink Minuet Sparse and dwarf Small Dark green Small Light pink White Nathan Hale Dense and typical Large Dark green Medium Red Pink Olympic Fire Sparse and typical Medium Medium to dark green Medium Red Light pink Ostbo Red Sparse and typical Medium Dark green Medium Red Light pink Peppermint Sparse and typical Large Dark green Medium Pink White Pink Charm Sparse and typical Large Medium to dark green Medium Deep pink Deep pink Pristine Dense and typical Large Medium green Medium White White Red Bandit Sparse and typical Large Medium to dark green Medium Pink White Sarah Dense and typical Medium Dark green Medium Red Deep pink Snowdrift Dense and typical Large Medium green Medium White White Starburst Sparse and dwarf Small Dark green Small Light pink White Tinkerbell Sparse and dwarf Small Dark green Medium Pink Deep pink z Small: length + width < 5 cm; medium: 5 cm < length + width < 8 cm; large: length + width > 8 cm. y Small: diameter < 1.8 cm; medium: 1.8 cm < diameter < 2.5 cm; large: diameter > 2.5 cm.

Table 2.3. Inventory of wild mountain laurel collected by the UGA Woody Plant Research

Laboratory crew and collaborators.

Collection Material Number of Date Population Quantity series type plants obtained GHW15003 11/21/15 Big Flat Creek, AL Cutting 24 6 GHW15004 11/21/15 Haines Island, AL Cutting 22 6 GHW15005 11/21/15 Haines Island, AL Cutting 24 16 GHW15011 11/21/15 Red Hill, AL Cutting 21 1 GHW15012 11/21/15 Red Hill, AL Cutting 20 1 GHW15013 11/21/15 Red Hill, AL Cutting 44 0 GHW15014 11/21/15 Red Hill, AL Cutting 32 15 GHW15015 11/21/15 Red Hill, AL Cutting 22 3 GHW15017 11/21/15 Red Hill, AL Seed 0.143 g 7 GHW15018 11/21/15 Red Hill, AL Seed 0.170 g 30 GHW15019 11/21/15 Red Hill, AL Seed 0.369 g 22 GHW15020 11/21/15 Red Hill, AL Seed 0.200 g 15 GHW15026 11/22/15 Fish River, AL Seed 0.068 g 18 GHW15028 11/22/15 Fish River, AL Cutting 40 2 GHW15030 11/22/15 Fish River, AL Seed 0.335 g 10 GHW15031 11/22/15 Fish River, AL Cutting 64 1 GHW15037 11/22/15 Blackwater Forest, FL Cutting 20 2 GHW15038 11/22/15 Blackwater Forest, FL Cutting 20 6 GHW15039 11/22/15 Blackwater Forest, FL Seed 0.084 g 3 GHW15040 11/22/15 Indian Ford Bridge, FL Seed 0.290 g 23 GHW15041 11/22/15 Indian Ford Bridge, FL Cutting 12 9 GHW15042 11/22/15 Indian Ford Bridge, FL Cutting 32 1 GHW15043 09/05/15 Pushepetappa Creek, LA Seed 0.080 g 13 GHW16001 12/07/16 Sepulga River, AL Seed 0.312 g 24 GHW17001 11/21/17 The Hitchcock Woods, SC Seed 0.474 g 21 GHW17002 11/21/17 The Hitchcock Woods, SC Cutting 60 0 GHW18001 02/13/18 Black Rock, GA Plant 22 22

Table 2.4. Inventory of mountain laurel cultivars in 1-gallon and 3-gallon containers.

1-gallon 3-gallon Cultivar Quantity Plant sizez (cm) Quantity Plant sizey (cm) Bullseye 5 159 6 301 Carol 5 154 6 252 Elf 13 141 5 227 Firecracker 10 115 4 165 Forever Red 6 146 6 244 Freckles n/a n/a 8 232 Heart of Fire n/a n/a 4 275 Kaleidoscope n/a n/a 5 175 Little Linda 7 136 4 215 Minuet 8 144 6 156 Nathan Hale n/a n/a 4 214 Olympic Fire 12 171 7 265 Ostbo Red n/a n/a 3 254 Peppermint n/a n/a 3 271 Pristine 10 105 6 256 Red Bandit n/a n/a 6 219 Sarah 1 129 7 201 Snowdrift 5 99 3 221 Starburst 10 104 7 217 Tinkerbell 1 134 4 170 z Height, width 1 at the widest point, and width 2 at perpendicular direction of 1-gallon plants were measured in 2015 (two year after being grown in 1-gallon pots). Plant size was then calculated as height + width 1 + width 2. y Height and widths of 3-gallon plants were measured in 2017 in the same manner as described above (two year after being transplanted to 3-gallon pots). Plant size was then calculated as height + width 1 + width 2.

Figure 2.1. Natural distribution of mountain laurel in the U.S.

(http://plants.usda.gov/core/profile?symbol=KALA).

Figure 2.2. Map of 15 mountain laurel populations that were identified and investigated in the

U.S. by the UGA Woody Plant Laboratory breeding program crews and collaborators.

Figure 2.3. The variations in flower traits among collected 21 mountain laurel cultivars; cultivars were diverse in flower size, shape, bud color, corolla color, and pigment pattern.

Figure 2.4. (A) Tree-type mountain laurel plants formed in a big patch at the bottom of a dell that partially shaded by pine trees in the Red Hill population (Monroe County, AL); (B) the Fish

River population in Baldwin County, AL; (C) mountain laurel (left) that had dark-pink flowers was found in the Sepulga River population (Escambia County, AL) by Dr. Ron Miller; (D) the slope at the edge of road in an opening area in the Black Rock population (Rabun County, GA) was covered by a number of mountain laurel seedlings; (E and F) the plant having intense blooms and bright pink flowers was found in the Hitchcock Woods population (Aiken County,

SC).

Figure 2.5. Wild plants of mountain laurel that rooted from stem cuttings were potted up in 1- gallon containers and maintained in the shade-house (left). Seedlings germinated from seeds of wild collection and yielded 15-20 cm of height in 6 months after being sown on peat moss

(right).

CHAPTER 3

EVALUATION OF TWENTY-ONE MOUNTAIN LAUREL CULTIVARS FOR CONTAINER

AND LANDSCAPE PERFORMANCE IN THE SOUTHEASTERN U.S.1

1Li, H., M. Chappell, and D. Zhang. Accepted by HortTechnology.

Reprinted here with permission of the publisher.

Summary. Mountain laurel (Kalmia latifolia) is an outstanding ornamental shrub due to its attractive foliage and showy inflorescences. Breeding efforts have led to improved selections that have predominantly been developed and evaluated in the northeastern U.S. Consequently, most cultivars have largely been dismissed as incompatible for the southeastern U.S. environmental conditions by nursery growers and consumers. This study was conducted over a 4-year period to evaluate 21 popular mountain laurel cultivars, primarily developed in the northeastern U.S., for container and field performance in Georgia. All cultivars yielded considerable growth in the first year of container trials, indicating production of mountain laurel as a 1-year container crop is feasible. Cultivars displayed significantly different total growth index throughout the container trial. Fast-growing cultivars like Bullseye and Ostbo Red yielded over 100, 150, and 250 cm of growth index in 1, 2, and 4 years, respectively. Conversely, cultivars that grew slower, like

Firecracker and Tinkerbell, had less than 80, 115, and 180 cm in 1, 2, and 4 years, respectively.

Cultivars were classified into five groups, using principal component analysis (PCA), that included dwarf habit with pink flower, dwarf habit with non-pink flower, non-dwarf habit with green stem and white flower, non-dwarf habit with pigment-patterned flower, and non-dwarf habit with pink flower. In a field study, performance rating of 21 cultivars ranged from 2.0 to 4.8

(out of 5.0) in 2014 and from 2.0 to 5.0 in 2015. Ten cultivars that received the highest ratings over the 2 years were selected for a subsequent field trial in 2016. Cultivars showed overall decreased ratings (1.0-3.3) from the previous 2 years because of late spring planting. ‘Ostbo

Red’, ‘Pristine’, and ‘Tinkerbell’ had higher performance ratings, more net growth, and less decrease in maximum quantum yield, which indicated suitable adaptation to southeastern environmental conditions. Nursery growers and consumers should benefit from regional cultivar trial information derived from this study. ‘Ostbo Red’, ‘Pristine’, and ‘Tinkerbell’ performed

well across trials and therefore are recommended for southeastern landscapes based on superior container and field performance, leaf spot tolerance, and morphologic distinctions.

Additional index words. Kalmia latifolia, heath family, growth index, morphology, performance rating, maximum quantum yield, stress tolerance

Introduction

Mountain laurel (Kalmia latifolia) is an evergreen flowering shrub in the heath family

(Ericaceae) that occurs throughout temperate areas of the world (Jaynes, 1988). Mountain laurel is native to the eastern U.S., specifically from southern Maine west through southern New York to central Ohio, south to eastern Louisiana, southern Mississippi, Alabama, Georgia, and northwestern Florida (Kurmes, 1967). The plant has been considered by many horticulturists, nursery growers, and gardeners to be an outstanding flowering native species (Dirr, 2009;

Jaynes, 1988). Its attractive lustrous green foliage, showy inflorescences, and variations in morphological traits have made mountain laurel a valuable ornamental shrub in the nursery and landscape industries (Jaynes, 1988).

Many cultivars of mountain laurel have been introduced into the marketplace, beginning in the early 1960s (The European Kalmia Society, 2018). Yet there is still considerable uncertainty about these plants and their performance in areas south of U.S. Department of

Agriculture (USDA) Zone 7, particularly in the southeastern U.S. In the southeastern U.S., few cultivars are available in retail settings and those cultivars that are produced are principally done so by niche Ericaceous nurseries. Ironically, although produced in several southeastern nurseries, mountain laurel is rarely used in southeastern landscapes (Jaynes, 1988). The majority of

mountain laurel cultivars were evaluated (prior to release) and selected for release in the northeastern U.S. As a result, the performance of cultivars outside the northeastern U.S. is relatively unknown, which has historically limited cultivar acceptance in other regions of the

U.S. (Jaynes, 1982, 1988). An example of this regionality in cultivar development and use is

‘Sarah’, selected by Richard Jaynes at the Connecticut Agricultural Experiment Station. While well-adapted to northeastern landscapes, no documentation exists regarding its broader adaptability due to a lack of evaluation outside the northeastern U.S.

Southeastern climate varies significantly from that of the northeastern U.S. Milder winter temperatures lead to less cold damage in nurseries and landscapes of the southeastern U.S. Yet hotter and more humid summer conditions, as well as a lack of winter chilling hours, could limit adaptability of cultivars selected in the northeastern U.S. and transitioned into southeastern nurseries and landscapes. Limited information on physiologic and adaptive responses; including heat tolerance, photosynthetically active radiation (PAR) tolerance, growth rates, disease and insect tolerance, required chilling hours to flower, soil moisture tolerance, etc. for most of mountain laurel cultivars is available (Jaynes, 1982, 1988). Additionally, while phenotypic traits of many mountain laurel cultivars have been well described at the time of release, these traits may differ across drastically different physiographic regions (Jaynes, 1988) due to genotype by environment interactions (Schlichting, 1986). This phenotypic variability was documented to occur in inkberry (Ilex glabra), whereby leaf traits varied when grown in varying environments

(Givnish, 1984). This study evaluated 21 popular mountain laurel cultivars for their potential as both nursery and landscape plants. Our objectives were to provide comprehensive regional cultivar information for growers, landscape contractors, and breeders that could enhance

mountain laurel production and landscape performance by identifying those cultivars with greatest adaptability to the southeastern U.S.

Materials and Methods

Plant materials. Liners of 21 cultivars were commercially micropropagated, rooted, and shipped from Briggs Nursery, Inc. (Elma, WA) to the University of Georgia Durham

Horticulture Farm (UGA Hort Farm) [(Watkinsville, GA) lat. 33.88689ºN, long. 83.41941ºW,

236 ft elevation, USDA Zone 8a (USDA, 2012)] on 11 Mar. 2013 in 49-cell flats. These cultivars were selected based on their popularity, reported superior landscape performance, and their diversity in habit, foliage, and inflorescence in the northeastern U.S. (Fig. 3.1). Liners (8-10 cm tall, 5-6 cm wide) were immediately transplanted into 1-gal containers (PF310; Nursery

Supplies, Inc., Chambersburg, PA) filled with pine bark-based soilless substrate (85% bark pine and 15% compost; Sun Gro Horticulture, Agawam, MA). Potted plants were top-dressed with 5 g of 20N-1.3P-7.5K controlled-release fertilizer (Everris, Marysville, OH). Containers were held in an unheated cold frame covered with 60% shade cloth and irrigated twice per day (15 min/irrigation event) using 10-ft full-circle nozzles (Rain Bird, Azusa, CA) monitored by a timer

(Node-100; Hunter, San Marcos, CA) until container and field trials were initiated.

Container evaluation. Container performance was assessed at the UGA Hort Farm beginning on 4 May 2013. On 21 May 2015, the plants were transplanted from 1- to 3-gal containers (PF1200; Nursery Supplies, Inc.) filled with bark-based substrate as described previously and fertilized with 15 g of 20N-1.3P-7.5K controlled-release fertilizer (Everris).

Plants were maintained in the same shade-house as described above for continued evaluation as

3-gal containers through the end of the trial in 2017 and top-dressed with 20N-1.3P-7.5K controlled-release fertilizer (Everris) annually.

Total growth index measurements were taken in the first week of May 2014, 2015, and

2017. Measurements included height (ground level to the tallest point), width at the widest point, and width perpendicular to the widest point. Growth index was then calculated as height + width

1 + width 2. Morphological traits were observed on container-grown plants throughout the container trial. Morphological measurements were taken on 11 traits that described habit, flowering, and leaf characteristics. Traits described by Jaynes (1988) that are responsible for major variations among cultivars were documented. These are described in Table 3.1 and included habit, stem and petiole color, leaf size, leaf blade, leaf surface, leaf color, flower size, bud color, corolla color, pigment pattern, and bloom time. Size measurements, including flower and leaf size, were taken on three flowers in full bloom and three mature leaves on each plant.

Additionally, leaf spot incidence was recorded, but severity was not assessed.

Field trial. The first field trial was conducted at the UGA Hort Farm from 2013 to 2015.

Four 1-gal containerized plants of the same 21 cultivars evaluated in containers were transplanted to the field on 1 Oct. 2013. The planting bed was tilled to a 10-inch depth prior to planting and existing Cecil clay soil (unamended) was used as backfill into planting holes. The trial was divided into four replications (rows) and one plant of each cultivar was randomly assigned to each replication. Plants were transplanted on 6-ft centers. The root-balls were slightly split and then covered with backfill-soil. Plants were top-dressed with 15 g of 20N-1.3P-7.5K controlled-release fertilizer (Everris) and then beds were mulched with a 2-inch-thick layer of pine straw. Plants were watered using drip irrigation with emitters (2 gallons per hour)

(XB20PC; Rain Bird) immediately after planting and then watered as needed by manually

turning on and off an irrigation valve. Plant performance rating was evaluated on 8 Oct. 2014 and 7 Oct. 2015. The performance rating was made using a scale of 1 (dead) to 5 (excellent: vigorous growth, lush and green foliage, no leaf spot, and no above-ground damage); based on overall plant size and vigor, leaf color, incidence of leaf spot and leaf and shoot damage due to abiotic conditions.

Based on the results of 2014 and 2015 field trial, 2-year-old 1-gal container plants of the highest rated 10 cultivars were planted in a field plot in the same manner as described above at the UGA Riverbend Research Center, Athens, GA [lat. 33.93014ºN, long. 83.36364ºW, 670 ft elevation, USDA Zone 8a (USDA, 2012)] on 14 Apr. 2016. Four replications were installed, each with one plant of each cultivar randomized within replication. Plants were installed in unamended Cecil clay field soil on 6-ft centers within two cultivated rows tilled to a depth of 10 inches. Light energy absorbed by chlorophyll in a leaf undergoes three fates: being used to drive photosynthesis, being emitted as heat, and being re-emitted as chlorophyll fluorescence. The spectrum of chlorophyll fluorescence is different to that of absorbed light. As a result, the yield of chlorophyll fluorescence can be measured and this information served as an indication of photosynthetic efficiency. Maximum quantum yield of photosystem II (Fv/Fm) is calculated as

(Fm-F0)/Fm, where Fm and F0 indicate maximum and minimal level of fluorescence, respectively.

Fv/Fm has been utilized as a determination of damage to photosystem II due to abiotic (e.g. heat and light) stress (Baker and Rosenquist, 2004; Maxwell and Johnson, 2000). This study quantified decreases in Fv/Fm over the course of a growing season to determine abiotic stress tolerance of cultivars in the 2016 field study. Fv/Fm was measured on three leaves of each plant beginning at dawn and ending at sunrise using a portable fluorometer (FluorPen FP 100; Photon

Systems Instruments, Albuquerque, NM). Baseline Fv/Fm and growth index were recorded while

plants were in containers, the day before cultivars were planted (13 Apr. 2016). Final growth index and Fv/Fm were measured on plants in field plots on 9 Nov. 2016. Net growth was then calculated as final growth index - initial growth index. The percentage of decrease in Fv/Fm was calculated as [(initial Fv/Fm - final Fv/Fm)/initial Fv/Fm x 100%]. Plant performance ratings were collected using the same 1 to 5 scale as described in the first field trial.

Experimental design and data analysis. Experimental design for the container evaluation study was a completely randomized design with four replications. Each replication included one plant of 21 cultivars. Likewise, randomized complete block design was employed for the field trial, which included four replications with one plant of each cultivar per replication. Growth index, net growth, performance rating, and the percentage change in Fv/Fm were subjected to analysis of variance using GLM procedures in SAS (University Edition; SAS Institute, Cary,

NC). Fisher’s LSD at P < 0.05 was applied for means separation. PCA was performed on morphological traits using SPSS (Version 21.0; IBM Crop, Armonk, NY). The first two axes were plotted according to the extracted Eigen vectors in Excel (Microsoft, Redmond, WA).

Results and Discussion

Container evaluation. Twenty-one container-grown mountain laurel cultivars reached 60 cm or greater total growth in the first year (2013-14). At this size, plants would be marketable after a 1-year production cycle, indicating it is feasible to commercially produce mountain laurel as container plants in the southeastern U.S. Yet, cultivars displayed differences in plant size during container evaluation (Table 3.2). In 2014, ‘Bullseye’, ‘Forever Red’, ‘Freckles’, ‘Heart of

Fire’, ‘Olympic Fire’, ‘Ostbo Red’, and ‘Pink Charm’ had growth indices of ≥100 cm (Table

3.3). These large cultivars continued vigorous growth rates in containers after 2 years of

production in 2015 (≥145 cm) and 4 years of production in 2017 (≥230 cm). ‘Carol’ and

‘Peppermint’ showed similar growth indices in 2015 (≥145 cm) and 2017 (≥230 cm), despite having smaller size in 2014 compared with ‘Bullseye’, ‘Forever Red’, ‘Freckles’, ‘Heart of Fire’,

‘Olympic Fire’, ‘Ostbo Red’, and ‘Pink Charm’. Based on these results, nine mountain laurel cultivars including Bullseye, Carol, Forever Red, Freckles, Heart of Fire, Olympic Fire, Ostbo

Red, Peppermint, and Pink Charm would be good candidates for production in 1-gal containers and could be sold after a single year in production. Transplanting these cultivars in the second year from 1- to 3-gal containers and selling them as 3-gal containers is also an option for growers. Although this increases input(s), labor, facility, and shipping resources, the fast growth rate of these cultivars and higher prices of 3-gal containers may compensate for additional inputs.

In contrast, ‘Firecracker’, ‘Kaleidoscope’, ‘Minuet’, ‘Nathan Hale’, and ‘Tinkerbell’ yielded less than 80 cm of total growth in 2014 (Table 3.3). ‘Firecracker’ and ‘Kaleidoscope’ were considered to be dense/compact growers, and accordingly had smaller growth indices of

<115 and 180 cm in 2015 and 2017, respectively. Interestingly, two cultivars frequently marketed as dwarf, Minuet and Tinkerbell, had greater growth indices in 2015 when compared to

Firecracker and Kaleidoscope. Three cultivars marketed as dwarf spreading cultivars including

Elf, Little Linda, and Starburst and four non-dwarf cultivars including Pristine, Red Bandit,

Sarah, and Snowdrift displayed moderate growth throughout the 4-year container evaluation.

These slow and moderate growers could also reach a salable size in a 1-gal container in 1 year.

An additional 2 years would be required to reach a salable size in a 3-gal container as defined by a growth index of 150 cm that is common among plants at the time of sale.

‘Elf’, ‘Freckles’, and ‘Goodrich’ were trialed in container production at Griffin, GA and

Puyallup, Washington by Hummel et al. (1990). Rooted liners of ‘Elf’ and ‘Freckles’ yielded more than 25 cm of growth index (height + width/2) in 6 months of production at both locations, with ‘Freckles’ having overall greater size than ‘Elf’. This corroborates our results, in that

‘Freckles’ was a fast grower while ‘Elf’ had moderate size throughout the evaluation period.

Since limited studies have been conducted to quantify the growth of container-grown mountain laurel, container trial results from this study provide novel detailed growth information for the 21 cultivars tested.

Morphology. The 21 mountain laurel cultivars trialed in this study were subjected to morphologic analysis of 11 phenotypic traits (Table 3.1). PCA indicated that the first four components accounted for 30.07%, 21.65%, 13.08%, and 11.08% of the morphological variance respectively, their cumulative variance being 75.88% (Table 3.4). On the basis of eigenvector values for traits along the first two components, the major traits responsible for morphological diversity were flower size, bud color, and habit along the first component (PC1). Leaf color, leaf size, and leaf blade fell along the second component (PC2). The first two components were used to visualize a scatter plot whereby cultivars were divided into five groups (Fig. 3.2). The first group (group 1) consisted of 10 cultivars including Bullseye, Carol, Firecracker, Forever Red,

Heart of Fire, Nathan Hale, Olympic Fire, Ostbo Red, Pink Charm, and Sarah that had higher

PC1 values. These cultivars had typical (non-dwarf) habit and deep pink to red flower buds. Six cultivars including Freckles, Kaleidoscope, Little Linda, Peppermint, Red Bandit, and Tinkerbell having PC1 value near 0, were classified into two groups based on their different PC2 values.

The first group (group 2) contained two dwarf cultivars, Little Linda and Tinkerbell, that had V- shaped foliage and pink flowers, while the second group (group 3) comprised four non-dwarf

cultivars including Freckles, Kaleidoscope, Peppermint, and Red Bandit that had pigmented (in spot or band pattern) pink flowers. Cultivars Pristine and Snowdrift comprised group 4, separated from other cultivars by low PC2 values. These two cultivars displayed medium green stem and foliage and pure white inflorescence that were very distinctive. Three dwarf cultivars,

Elf, Minuet, and Starburst, comprised group 5, based on low PC1 values. These three cultivars had more dense habit and lighter pink buds compared to Little Linda and Tinkerbell. The results confirmed visual observations, in that mountain laurel cultivars have a wide range of morphological combinations. Yet, PCA also indicated sharing of traits among five distinct groups. This classification information should provide plant breeders with useful information that affords insight into breeding for optimal growth index combined with favorable phenotypic characteristics. Additionally, growers in the southeastern U.S. could use this information to select cultivars having optimal growth indices from each of five groups. This would afford growers the ability to select fast-growing cultivars while maximizing phenotypic diversity in a minimum number of cultivars.

Leaf spot incidence. The majority of cultivars selected for this study were reported as leaf spot tolerant in the northeastern U.S. No leaf spot was observed on cultivars Ostbo Red, Pristine, and Tinkerbell, whereas all other cultivars in this study had observed leaf spot incidence.

Similarly, Jaynes (1988) observed that ‘Ostbo Red’ and ‘Tinkerbell’ showed higher tolerance to leaf spot compared with other cultivars. ‘Pristine’ was selected from a seed-lot collected in a wild population growing within 30 km of Woodlander Nursery [Aiken, SC, USDA Zone 8a

(USDA, 2012)]. This is one of few cultivars on the market that was introduced from a selection program in the southeastern U.S. As a result of continual disease pressure observed in the southeastern U.S. and the fact that ‘Pristine’ was selected from a native population in the region

that evolved in these conditions, ‘Pristine’ is more likely to have higher level of tolerance to leaf spot. Similar results were obtained in leaf spot tolerance studies in rabbiteye blueberry selections that were derived from genotypes developed and/or selected from wild-types in the southeastern

U.S. In the case of rabbiteye blueberries, leaf spot tolerance was tied to greater adaptability of genotypes derived from southeastern germplasm that were subjected to hot and humid conditions of the southeastern U.S., compared to northern highbush blueberries (Scherm, 2008).

Field trial. Plants were rated using a 1-5 scale at the conclusion of the 2014 and 2015 growing seasons. Performance rating of 21 cultivars ranged from 2.0 to 4.8 in 2014 and from 2.0 to 5.0 in 2015, with differences observed among cultivars in 2014 (Table 3.2, Fig. 3.3). Four cultivars (rating ≥ 4) performed well during the 2014 growing season; Elf, Ostbo Red, Red

Bandit, and Tinkerbell. Seven of 21 cultivars received similar ratings (rating ≥ 4) in 2015, including Carol, Little Linda, Minuet, Peppermint, Pristine, Sarah, and Tinkerbell. The majority of cultivars had superior or equivalent field performance ratings in 2015 when compared to

2014, while Red Bandit, Starburst, and Tinkerbell had lower performance ratings in 2015 (Fig.

3.3). Improved field performance of most cultivars in year two indicated mountain laurel could have enhanced performance in landscapes after establishment (observed in year two) and adapt to local southeastern conditions (high heat and humidity). Based on the rating results from 2014 and 2015, the 10 highest rated cultivars (Carol, Elf, Firecracker, Little Linda, Minuet, Ostbo Red,

Peppermint, Pristine, Sarah, and Tinkerbell) were subsequently evaluated for field performance at the UGA Riverbend Research Center in 2016.

There was no significant difference in performance rating among the 10 cultivars field- trialed in 2016 (Table 3.2). Performance ratings in 2016 ranged from 1.0 to 3.3, lower than the initial field trial in 2014-15 (Table 3.5). ‘Elf’ received consistently high ratings after the initial

field trial in 2014 (4.0) and 2015 (3.8), yet its performance rating dropped to 1.8 in 2016 (Table

3.5, Fig. 3.3). This reduction in performance rating could be due to transplant date. The initial field trial, conducted in 2014-15, was planted in Fall 2013 (1 Oct. 2013). Hanson et al. (2004) reported the main difficulty in establishing container-grown mountain laurel plants in landscapes was the lack of root extension into backfill-soil after transplanting in spring. Fall planting, as observed in the first field study conducted in 2014-15, was proceeded by fall and winter conditions typified by extended cool temperatures and higher soil moisture when compared to spring plantings. This resulted in lower transpiration in the Fall 2013-planted field trials and likely led to enhanced establishment that improved later performance when compared to plants established in Spring 2016. Although spring-summer establishment conditions following transplant in 2016 were overall unfavorable (higher physiologic stress) when compared to the

Fall 2013 establishment period; ‘Ostbo Red’, ‘Pristine’, and ‘Tinkerbell’ continued to exhibit higher performance ratings. ‘Ostbo Red’ and ‘Tinkerbell’ had more net growth (>50 cm) than other cultivars, while ‘Pristine’ exhibited slightly less net growth (35 cm) due to its comparably dense habit (Table 3.5).

Maximum quantum yield of photosystem II (Fv/Fm) is a measure of plant stress, particularly heat stress (Baker and Rosenquist, 2004). Fv/Fm for most plants grown in minimal environmental stress conditions is approximately 0.79-0.83, while values lower than this range suggest that plants have been exposed to stress and PSII reaction centers are damaged (Kalaji et al., 2012; Maxwell and Johnson, 2000). The Fv/Fm measured on mountain laurel plants before

2016 transplant ranged from 0.78 to 0.84, whereas all cultivars showed decreased Fv/Fm (0.00 to

0.80) at the conclusion of field trials. Decrease in Fv/Fm was therefore used to compare the ability of mountain laurel cultivars to adapt to southeastern environmental conditions. In this scenario, a

lower percentage decrease in Fv/Fm indicated less damage of photosystem II and therefore enhanced stress tolerance. Negative correlation between percentage decrease in Fv/Fm and performance rating was observed among all cultivars. ‘Ostbo Red’, ‘Pristine’, and ‘Tinkerbell’ had less decrease in Fv/Fm (<51%), while other cultivars had greater decrease in Fv/Fm (>70%)

(Table 3.5). When combined with performance ratings, Fv/Fm appeared to be a parameter that could reflect stress tolerance of mountain laurel cultivars.

Conclusion. Twenty-one mountain laurel cultivars displayed various responses to container and field trials when evaluated in the southeastern U.S. (USDA Zone 8a). Although the

21 cultivars displayed differing growth indices in container production, all cultivars could yield sufficient growth the first year after being transplanted into 1-gal containers and consistently perform well thereafter in a container production environment. This indicated that growing mountain laurel for 1 year as 1-gal container plant should be feasible for southeastern producers.

Fast-growing cultivars including Bullseye, Carol, Forever Red, Freckles, Heart of Fire, Olympic

Fire, Ostbo Red, Peppermint, and Pink Charm could be shifted up to 3-gal containers.

Additionally, mountain laurel cultivars had a variety of morphological characteristics and fell into five phenotypic (PCA) groups based on shared morphologic traits when grown in a container production environment; principally on habit, foliage, and inflorescence traits. Results of this study should benefit plant breeders and commercial producers by providing regional cultivar information for the southeastern U.S. (and other USDA Zone 8-9 environments). ‘Ostbo

Red’, Pristine’, and ‘Tinkerbell’ would be recommended to the southeastern producers in order to have both phenotypic variability in the cultivars produced and leaf spot tolerance. Although most cultivars in this study had difficulty establishing in southeastern landscapes; fall planting could facilitate improved establishment. Three cultivars, Ostbo Red, Pristine, and Tinkerbell,

consistently excelled in field trials, indicating they are more adapted to landscape environmental conditions observed in the southeastern U.S.

Literature Cited

Baker, N.R. and E. Rosenquist. 2004. Applications of chlorophyll fluorescence can improve crop

production strategies: An examination of future possibilities. J. Expt. Bot. 55:1607-1621.

Dirr, M.A. 2009. Manual of woody landscape plants. 6th ed. Stipes Publishing, Champaign, IL.

Givnish, T.J. 1984. Leaf and canopy adaptations in tropical forests. Physiol. Ecol. Plants Wet

Trop.: 51-84.

Hanson, A.M., J.R. Harris, and R. Wright. 2004. Effects of transplant season and container size

on landscape establishment of Kalmia latifolia L. J. Environ. Hort. 22:133-138.

Hummel, R.L., C.R. Johnson, and O.M. Lindstrom. 1990. Root and shoot growth response of

three container-grown Kalmia latifolia L. cultivars at two locations to growing medium

and nitrogen form. J. Environ. Hort. 8:10-13.

Jaynes, R.A. 1982. New mountain laurel selections and their propagation. Intl. Plant Prop. Soc.

Proc. 32:431-434.

Jaynes, R.A. 1988. Kalmia: The laurel book. Timber Press, Portland, OR.

Kalaji, H.M., R. Carpentier, S.I. Allakhverdiev, and K. Bosa. 2012. Fluorescence parameters as

early indicators of light stress in barley. J. Photochem. and Photobiol. B. 112:1-6.

Kurmes, E.A. 1967. The distribution of Kalmia latifolia L. Amer. Midland Naturalist 77:525-

526.

Maxwell, K. and G.N. Johnson. 2000. Chlorophyll fluorescence - A practical guide. J. Expt. Bot.

51:659-668.

Scherm, H. 2008. Disease management in organic rabbiteye blueberries. Intl. J. Fruit Sci. 8:69-

80.

Schlichting, C.D. 1986. The evolution of phenotypic plasticity in plants. Ann. Rev. Ecol. Syst.

17:667-693.

The European Kalmia Society. The European Kalmia Society. Laurel cultivars - The

international Kalmia register and checklist. 2 Oct. 2018. < http://www.kalmia

society.org/cultivars.php?&lang=gb>.

U.S. Department of Agriculture. 2012. USDA plant hardiness zone map. 1 Aug. 2018.

Table 3.1. Eleven morphologic characters describing habit, flowering, and leaf characteristics observed on 21 mountain laurel cultivars for morphological measurements.

Numerical description

Character 0 1 2 3 4

Habit Sparse and dwarf Dense and dwarf Sparse and typical Dense and typical

Stem and petiole color Green Slightly purplish red Moderately purplish red Purplish red

Leaf sizez Length + width < 5 cm 5 cm < length + width < 8 cm Length + width > 8 cm

Leaf blade Elliptic Elliptic to ovate Elliptic to lanceolate Lanceolate Lanceolate to ovate

Leaf surface Flat V-shaped Cupped Wavy

Leaf color Medium green Medium to dark green Dark green

Flower size Diameter < 1.8 cm 1.8 cm < diameter < 2.5 cm Diameter > 2.5 cm

Bud color White Light pink Pink Deep pink Red

Corolla color White Light pink Pink Deep pink

Pigment pattern No pigment Normal band Full band Radiation Spot

Bloom time Before 1 May Between 1 and 15 May After 15 May z Size measurements, including leaf and flower size, were taken on three mature leaves and three flowers in full bloom on each plant;

1 cm = 0.3937 inch.

Table 3.2. Analysis of variance table for growth index of 21 mountain laurel cultivars in container evaluation; performance rating of 21 cultivars in 2014 and 2015 field trial; and performance rating, net growth, and decrease in Fv/Fm (%) of 10 cultivars in 2016 field trial.

Source df Mean square F P

Container evaluation

Cultivar

2014 growth index 20 622.8 7.1 <0.0001

2015 growth index 20 2277.7 7.1 <0.0001

2017 growth index 20 5929.3 9.3 <0.0001

2014 and 2015 field trial

Cultivar

2014 performance rating 20 2.5 1.83 0.0380

2015 performance rating 20 3.1 1.2 0.2987

2016 field trial

Cultivar

Performance rating 9 2.2 1.1 0.4052

Net growth 9 425.1 1.4 0.2363

Decrease in Fv/Fm (%) 9 1482.5 0.6 0.7914

Table 3.3. Growth index of container-grown plants of 21 mountain laurel cultivars at the UGA

Hort Farm in Watkinsville, GA after being grown for 1 (2014), 2 (2015), and 4 years (2017).

Growth index [means ± SE (n=4) (cm)]z

Cultivar 2014 2015y 2017

Bullseye 100.1±9.3 abx 158.5±12.4 abcde 301.2±12.2 a

Carol 86.4±3.4 cdef 154.0±10.7 abcdef 251.6±8.0 bcdef

Elf 86.4±5.8 cdef 140.5±15.5 cdef 227.2±16.7 defgh

Firecracker 68.3±1.2 hi 114.9±6.6 ghij 164.9±11.9 j

Forever Red 99.6±1.4 abc 146.4±16.6 abcdef 244.2±18.7 bcdefg

Freckles 105.3±3.8 a 160.2±7.0 abcd 232.1±19.9 cdefgh

Heart of Fire 100.7±5.4 ab 168.5±10.1 ab 275.3±7.5 ab

Kaleidoscope 78.9±2.6 efgh 110.2±4.7 hij 175.1±13.0 ij

Little Linda 88.4±4.0 bcde 135.5±4.7 defg 215.3±9.1 gh

Minuet 74.5±4.3 fghi 143.8±2.8 bcdef 156.1±5.3 j

Nathan Hale 62.3±2.3 i 96.8±7.0 j 213.6±4.6 gh

Olympic Fire 99.4±6.9 abc 171.1±5.2 a 264.8±14.2 bc

Ostbo Red 101.6±3.4 ab 149.5±6.9 abcdef 253.5±4.0 bcde

Peppermint 97.0±4.2 abcd 157.4±7.2 abcde 270.5±15.7 ab

Pink Charm 105.9±7.4 a 164.2±8.6 abc 229.9±2.3 cdefgh

Pristine 89.5±5.8 bcde 105.3±3.5 ij 255.8±15.2 bcd

Red Bandit 89.9±5.0 bcde 140.0±8.3 cdefg 218.5±6.2 efgh

Sarah 91.9±2.9 bcde 129.2±12.3 fghi 201.1±14.2 hi

Snowdrift 85.7±3.5 def 98.7±4.9 j 221.2±17.8 defgh

Starburst 81.7±3.5 efg 103.7±8.5 j 216.6±22.0 fgh

Tinkerbell 71.2±2.4 ghi 134.1±5.9 efgh 169.9±10.7 ij z Growth index was calculated as height + width 1 + width 2. Width 1 and width 2 measured perpendicular to each other with width 1 at the widest point; 1 cm = 0.3937 inch. y Measurements were taken before plants were transplanted to 3-gal (11.4 L) containers in 2015. x Data followed by different letters within the column mean they were significantly different at P

< 0.05 according to Fisher’s LSD.

Table 3.4. Eigenvectors and eigenvalues generated by PCA applied on 11 morphological characteristics of 21 mountain laurel cultivars.

Eigenvectors

Variable PC1 PC2 PC3 PC4

Habit 0.626 -0.650 0.044 0.032

Stem and petiole color 0.599 0.324 0.517 -0.055

Leaf size 0.573 -0.617 0.026 0.416

Leaf blade 0.262 -0.604 0.251 -0.470

Leaf surface 0.580 0.148 -0.654 0.156

Leaf color 0.288 0.783 0.284 0.284

Flower size 0.823 -0.264 -0.007 0.133

Bud color 0.807 0.431 -0.034 0.056

Corolla color 0.485 0.286 0.374 -0.541

Pigment pattern -0.277 -0.124 0.585 0.633

Blooming time 0.315 0.348 -0.337 0.028

Eigenvalues

Eigenvalue 3.308 2.381 1.439 1.219

Percent of total variance explained 30.070 21.650 13.081 11.078

Cumulative percent of total variance explained 30.070 51.720 64.800 75.879

Table 3.5. Field performance rating, net growth, and percentage of decrease in maximum quantum yield of photosystem II (Fv/Fm) of 10 mountain laurel cultivars at the UGA Riverbend

Research Center in Athens, GA in 2016.

z y x Rating Net growth (cm) Decrease in Fv/Fm (%)

Cultivar means ± SE (n=4)

Ostbo Red 2.5±0.7 58.7±11.7 aw 43.8±33.0

Peppermint 1.5±0.5 55.6±10.6 ab 73.8±26.2

Tinkerbell 3.3±1.0 54.8±13.9 ab 49.8±29.0

Little Linda 1.8±0.8 49.0±4.6 abc 75.3±24.7

Firecracker 1.0±0.0 43.6±1.5 abc 100.0±0.0

Minuet 1.5±0.5 40.5±17.0 abc 75.9±24.1

Sarah 1.0±0.0 39.4±7.5 abc 100.0±0.0

Pristine 2.8±1.0 35.3±4.1 abc 50.9±29.0

Elf 1.8±0.8 32.7±5.9 bc 72.8±27.2

Carol 2.0±1.0 28.9±16.0 c 74.4±25.6 z Field performance rating was evaluated using a scale of 1 (dead) to 5 (excellent) and the rating was based on overall plant size and vigor, leaf color, incidence of leaf spot, and leaf and shoot abiotic damage. y Growth index was calculated as height + width 1 + width 2. Width 1 and width 2 measured perpendicular to each other with width 1 at the widest point. Net growth was calculated as final growth index-initial growth index; 1 cm = 0.3937 inch.

x Initial and final Fv/Fm were measured on plants before being planted to the field and after growing season, respectively. The percentage of decrease in Fv/Fm was then calculated as [(initial

Fv/Fm - final Fv/Fm)/initial Fv/Fm x 100%]. w Data are sorted by descending numerical values of net growth. Data followed by different letters within the column indicate differences in mean values at P < 0.05 according to Fisher’s

LSD.

Figure 3.1. Diversity in size, shape, color, and pigment pattern of flowers among popular mountain laurel cultivars.

Figure 3.2. Scatter plot of 21 mountain laurel cultivars obtained performing PCA on 11 morphological characteristics related to habit, stem, leaf, and flower. PC1 and PC2 explain

30.07% and 21.65% of morphological variation, respectively, with cumulative variance being

51.72%. Twenty-one cultivars were classified into five phenotypic groups based on PC1 and

PC2 values.

Figure 3.3. Field performance rating of 21 mountain laurel cultivars at the UGA Hort Farm in

Watkinsville, GA after the 2014 growing season (black bar) and 2015 growing season (gray bar).

Plants were evaluated using a scale of 1 (dead) to 5 (excellent) and the rating was based on overall plant size and vigor, leaf color, incidence of leaf spot, and abiotic leaf and shoot damage.

Data are presented as means + SE (n=4) by the descending order of ranking in 2015. Different letters indicate that performance ratings of cultivars in 2014 are significantly different at P < 0.05 according to Fisher’s LSD.

CHAPTER 4

GENETIC RELATIONSHIP AMONG KALMIA LATIFOLIA L. TAXA USING ISSR

MARKERS2

2Li, H., M. Chappell, and D. Zhang. To be submitted to Journal of American Society for

Horticultural Science (ASHS).

Abstract. Kalmia latifolia L. (mountain laurel), an attractive flowering shrub, is considered to be a potential ornamental plant for the eastern U.S. Limited information on genetic diversity and relationship of mountain laurel is available, which would obstruct the breeding of elite cultivars.

Genetic relationship among 48 wild accessions sampled from eight populations and 21 cultivars and genetic diversity among and within eight populations were assessed. Sixty-nine taxa were analyzed using eight Inter Simple Sequence Repeat (ISSR) markers. A total of 116 bands were amplified, 90.52% of which (105) were polymorphic. The genetic similarity among 21 cultivars ranged from 0.5636 to 0.7909 and cultivars were classified into five groups in UPGMA dendrogram based on similarity coefficient. The robustness of ISSR was demonstrated by high similarity among related cultivars and clustering of them. All cultivars were uniquely identified by their ISSR profiles, therefore accurate cultivar identification would be achieved using ISSR.

The relatively low GST (0.38) corresponding with the low percentage of variation among populations from AMOVA (30%) indicated that individual plants within populations were likely to be genetically different, but each population contained a similar complement of alleles in similar frequencies. A relatively large proportion of diversity was attributed to within-population variation, yet the low actual diversity within populations (HS=0.19) was observed due to the small geographic size of populations. UPGMA dendrogram exhibited the clustering of nearby populations because of relatively high genetic similarity between them (0.8994-0.9169) and four clusters correlated with geographic regions. A clear separation of cultivars and wild accessions was revealed, which might be due to the loss of genetic information during breeding and selection of cultivars. The overall results would benefit breeding of mountain laurel on identifying cultivars, understanding pedigree, collecting germplasm, selecting parents, and increasing genetic diversity.

Additional index words. mountain laurel, ornamental breeding, cultivar identification, pedigree, among-population diversity, within-population diversity, AMOVA, UPGMA dendrogram

Introduction

The genus Kalmia is a member of Ericaceae (heath family) along with genus

Rhododendron and Vaccinium. It comprises seven species that are primarily distributed in temperate zones. Most of them are shrubs and subshrubs, some are herbs, and a few are trailing vines (Zomlefer, 1994). Kalmia latifolia L. (mountain laurel) is the most famous species in genus

Kalmia. It is an outstanding spring-summer blooming woody ornamental and has gained popularity in northeastern U.S. landscapes.

More than 100 mountain laurel cultivars have been developed since the early 1960s (The

European Kalmia Society, 2018). Characteristics, including habit, petiole and stem color, leaf size and color, and bud and flower color, have been used to identify them (Jaynes, 1988).

However, morphological variation among cultivars have been found to be limited, particularly among those in the same phenotypic group (results from a separate study). Furthermore, morphological characters might vary depending on growing environmental conditions, such as climate, light, soil, and altitude (Jaynes, 1988). For instance, ‘Kaleidoscope’ is recognized by the purplish red petiole and stem of its new growth, whereas the color would be faded when it is heavily shaded. As a consequence, some mountain laurel cultivars are being sold under the wrong name (Jaynes, 1988), indicating that it is not adequate to identify them only by morphology. Additionally, as we documented our mountain laurel collection, the pedigree information of more than half of the cultivars was found to be unavailable, which is an obstacle to efficiently use them for future breeding.

Historically, wild plant resources have been involved in breeding programs worldwide to introduce desirable traits and increase genetic diversity (Engels et al., 1995). It is widely accepted that reproductive strategies, gene flow, and geographic ranges of plants have essential effects on the partitioning of genetic diversity within and among wild populations (Liu et al.,

2012; Nybom, 2004). Due to the consideration of lowering facility and labor cost, it is always desirable to minimize the quantity of plants collected meanwhile maximizing the genetic diversity that collection may provide. Studying the genetic diversity of populations is therefore necessary for efficient germplasm collection and plant improvement. Many studies have been carried out to determine the diversity of wild populations of Rhododendrons spp. that closely related to mountain laurel in recent years. The genetic diversity of Rhododendrons was principally attributed to within-population variation (62%-83%), while a low level of genetic diversity was observed among populations (Chappell et al., 2008; De Keyser et al., 2010; Liu et al., 2012). These results illuminated the importance of collecting individual plants within a population, which would benefit breeding programs to efficiently maximize genetic diversity.

Similar studies on mountain laurel populations are needed as an essential step toward better gaining germplasm.

Yet as a cross-pollinating species, mountain laurel was found to be able to self. While, the significant reduction in survival and vigor of seedlings from self-pollinated plants indicated inbreeding depression in mountain laurel (Jaynes, 1968). From a breeding standpoint, in order to prevent inbreeding depression as well as increase genetic variations among progenies, individuals that are genetically distinct should be used as parents in cross hybridization.

Therefore, genetic relativeness information will be tremendously beneficial to breeders.

DNA fingerprinting is a reliable and efficient tool to detect genetic relationships and diversity within species. The Inter Simple Sequence Repeat (ISSR) marker system detects polymorphisms in inter-microsatellite loci and can produce much larger numbers of fragments per primer, with the advantages of high reproducibility and relatively low cost (Nagaoka and

Ogihara, 1997). Moreover, ISSR is regardless of the availability of genome sequence information (Shi et al., 2010). In recent years, the relationships among accessions within genus

Rhododendron have been successfully determined using ISSR markers (Liu et al., 2010; Zheng et al., 2011). In the study, ISSR markers were employed to determine the genetic relationship among 69 mountain laurel taxa, including 21 cultivars and 48 wild accessions, and assess the genetic diversity among and within wild populations. The results would assist in identifying cultivars and their pedigree, collecting wild genetic resources, selecting parents, and increasing genetic diversity for mountain laurel breeding.

Materials and Methods

Plant materials and collection. Micropropagated liners of 21 mountain laurel cultivars were obtained from Briggs Nursery Inc. (Elma, WA) and grown at the Horticulture Farm at

University of Georgia (Watkinsville, GA). The youngest leaves were collected and then genomic

DNA was extracted from 100 mg leaf tissue using Plant DNAzol® Reagent (Invitrogen,

Carlsbad, CA) following manufacturer’s protocols. The DNA samples were diluted to 10 ng·L-1 and then stored at -20 °C until further analysis.

Mountain laurel is naturally distributed from southern Maine west through southern New

York to central Ohio, south to east Louisiana, southern Mississippi, Alabama, and Georgia, and northwest Florida (Kurmes, 1967). Wild populations were identified in the native range and eight

populations with a minimum of 15 individuals were selected for sampling (Table 4.1). In each population, a random sample of six individual plants was obtained. Individuals were evenly distributed throughout the population and were at least 10 m away from each other. The youngest leaves were individually collected from six sampled plants and placed in six envelopes. Envelops were carefully labeled and placed in a heavy-duty zip-lock bag filled with 100 g silica gel. Leaf tissues were maintained in the bag until being transported back to the lab and stored at -80 °C.

Genomic DNA samples were prepared in the same manner as described above.

ISSR procedure. Cultivars ‘Olympic Fire’, ‘Ostbo Red’ (maternal parent of ‘Olympic

Fire’), and ‘Starburst’ were used to screen 96 ISSR primers designed by the University of British

Columbia (UBC 801-896). Eight primers were selected based on the strong, clear, reproducible, and polymorphic banding patterns they produced (Table 4.2). ISSR-PCR amplifications were performed in a total volume of 20 L consisting of 2 L (20 ng) template DNA, 2 L primer, 10

L master mix (Applied Biosystems, Foster city, CA), and 6 L double-distilled water. The amplifications were programmed in a Mastercycler nexus gradient (Eppendorf, Hamburg,

Germany) under following conditions: an initial denaturation step at 94 °C for 5 min; 40 cycles of 94 °C for 30 s, 52 °C for 50 s, 72 °C for 120 s; and followed by an extension for 7 min at 72

°C. Samples were then cooled to 4 °C and placed in a refrigerator until being loaded into gels.

The PCR products were electrophoresed on 1.2% (w/v) agarose gels that in 0.5 x TBE buffer solution (Fisher Scientific, Fair Lawn, NJ) (80 V for 3.5 h) and stained with 0.5 mg·L-1 ethidium bromide solution (Sigma-Aldrich, St. Louis, MO). 100-bp DNA ladder (Invitrogen, Carlsbad,

CA) was loaded onto the outside two lanes of each gel to estimate the size of amplified fragments. Gels were visualized and photographed using a BioDoc-ItTM Imaging System (UVP,

Upland, CA).

Data analysis. ISSR amplified fragments were scored based on the presence (1) or absence (0) for each sample, and values recorded in Excel. POPGENE version 1.32 (Yeh et al.,

1997) was used to calculate the percentage of polymorphic bands (PPB) and Nei’s unbiased genetic identity and distance (Nei, 1978). An unweighted pair group method with arithmetic mean (UPGMA) dendrogram, based on Nei’s genetic distance (Nei, 1978), was generated to determine the relationships among taxa using NTSYS-pc version 2.10e (Rohlf, 2008.).

Additionally, the total genetic diversity (HT), mean genetic diversity within populations (HS), and proportion of total genetic diversity among populations [GST=(HT-HS)/HT] were calculated for eight populations using POPGENE version 1.32. Analysis of molecular variance (AMOVA) was employed to determine the partitioning of genetic variance within and among populations using

Arlequin version 3.5.2.2 (Excoffier and Lischer, 2010). The significance of this F-statistic was tested with 1,000 random permutations.

Results and Discussion

ISSR polymorphism. A total of 116 bands was generated in 69 mountain laurel taxa by eight ISSR markers, of which 105 (90.52%) were polymorphic. The number of bands amplified per primer ranged from 10 to 19, with an average of 14.5 (Table 4.2). The molecular weight of bands was found to be between 150 and 2000 base pairs (bps). Although no DNA fingerprinting study has been carried out on mountain laurel, numerous similar studies were conducted on two other genera, Rhododendron and Vaccinium, in heath family. Zheng et al. (2011) found that 13

ISSR primers amplified 106 fragments in Rhododendron hybridum in total, with 85.48% being polymorphic, and each primer produced 9.5 bands on average. In nine cold-hardy Rhododendron spp., thirteen ISSR primers amplified fragments between 300 to 2000 bps and 90.63% of them

were polymorphic (Liu et al., 2010). In blueberries, six ISSR primers generated 87 bands, of which 80.4% were polymorphic (Garriga et al., 2013). Similarly, the reproducible and highly polymorphic fragments produced by ISSR in the present study indicated that ISSR was a reliable and effective technique to amplify the loci between microsatellites in mountain laurel.

Genetic relationship of cultivars and cultivar identification. The genetic identity among

21 mountain laurel cultivars ranged from 0.5636 to 0.7909 (Table 4.3). The robustness of ISSR for determining the genetic relationships of cultivars was demonstrated by the observed high genetic identities among related cultivars. For instance, the highest identity (0.7909) was observed between ‘Olympic Fire’ and its maternal parent ‘Ostbo Red’, and between ‘Olympic

Fire’ and its sibling ‘Heart of Fire’ that shared the same maternal parent. Another high genetic similarity value (0.7727) was found between ‘Tinkerbell’ and its female parent ‘Olympic Fire’.

Also, the similarity between ‘Kaleidoscope’ and its maternal parent ‘Sarah’ and between ‘Sarah’ and its paternal parent ‘Pink Charm’ was found to be relatively high at 0.7455.

Five groups were observed in UPGMA tree (Fig. 4.1). The first group consisted of three related cultivars (‘Ostbo Red’ and its progenies ‘Heart of Fire’ and ‘Olympic Fire’) and five other cultivars ‘Bullseye’, ‘Forever Red’, ‘Carol’, ‘Firecracker’ and ‘Snowdrift’. Besides

‘Bullseye’ and ‘Snowdrift’, all others are red-bud selections. Red bud is a recessive trait and breeding for this trait requires intercrosses between red-bud selections (Jaynes, 1988), which probably accounted for the grouping of these red-budded cultivars. The second group contained pink to red bud cultivars, including three dwarf ones ‘Tinkerbell’, ‘Starburst’, and ‘Little Linda’ and a non-dwarf one ‘Nathan Hale’. The dwarf form was found to be under the control of a single recessive gene and all dwarf cultivars were derived from a common ancestor (Jaynes,

1988). ‘Sarah’, its male parent ‘Pink Cham’, its progeny ‘Kaleidoscope’, and an additional

cultivar ‘Red Bandit’ fell into the third group. This clustering allowed us to estimate the pedigree of ‘Red Bandit’, which might be a relative of other cultivars in the same group. The fourth cluster contained three white-bud cultivars. In this cluster, ‘Freckles’ and ‘Minuet’ shared the same maternal parent that is ‘Star Cluster’, which indicated that the other cultivar ‘Elf’ might also developed from ‘Star Cluster’. Two cultivars, ‘Peppermint’ and ‘Pristine’, formed the last group. Interestingly, one of the parents of ‘Peppermint’ was developed in South Carolina and

‘Pristine’ was selected from a seed-lot collected in South Carolina, which probably resulted in the grouping of these two cultivars. Again, the accuracy of ISSR was proved by the observed clustering of cultivars with their relatives in UPGMA dendrogram (Fig. 4.1). The genetic relativeness of mountain laurel cultivars revealed by ISSR technique would be of essential significance for pedigree certification, parental selection, and cultivar development.

All cultivars were uniquely identified by their ISSR profiles. Although primer UBC835 alone was able to discriminate 21 cultivars (Fig. 4.2), a combination of a few primers would achieve greater reliability in cultivar identification. As suggested by Garriga, et al. (2013), accurate identification of 13 blueberry cultivars could be reached using three ISSR primers.

Hence, the morphologically similar cultivars could be distinguished, and morphological differences of cultivars induced by environmental conditions could be identified by ISSR.

Genetic diversity among and within wild populations. For dominant markers, genetic differentiation among populations is often estimated with the proportion of genetic diversity attributed to among-population variation (GST=(HT-HS)/HT), a critical indicator of genetic diversity at the population level (Nybom, 2004). The GST of eight mountain laurel wild populations was 0.38, which indicated that a low-to-moderate proportion of total genetic diversity was observed among populations (38%) whereas a relatively high level of diversity was

found within populations (62%) (Table 4.4). The low GST value indicated a high level of gene flow among populations (Falk et al., 2001), particularly between those distributed in geographically closed area as indicated by reduced GST values within geographic groups (0.18-

0.20) (Table 4.4). Mountain laurel is a cross-pollinated species that relies on insects, bumble bees in particular, for pollination (Jaynes, 1988). Similar low levels of genetic diversity among populations have been observed on numerous species that rely on insects for cross pollination

(Nybom, 2004; Liu et al., 2012). Because of outcrossing, gene flow among populations led to a similar complement of alleles in similar frequencies and thus a low level of genetic diversity observed at the population level. Nearby populations, in contrast to the geographically isolated populations where gene flow could be obstructed by distance or barriers, insects could transport pollen easily between them, resulting to more frequent gene flow and hence lower GST (Falk et al., 2001). AMOVA results (Table 4.5) indicated a small proportion of variation was observed among populations (30%; P<0.001) while a high level of variation was found within populations

(70%; P<0.001), which was consistent with observed GST value (0.38). The relatively low GST value corresponding with the low percentage of variation among populations from AMOVA indicated that individual plants within populations were likely to be genetically different, but each population contained a similar complement of alleles in similar frequencies (Falk et al.,

2001).

One of the most commonly employed methods to estimate within-population diversity is

HS, Nei’s unbiased gene diversity. In contrast to the proportion of genetic diversity partitioned to among and within populations estimated by GST and AMOVA, HS is a direct measure of average diversity within populations (Nybom, 2004). HS of eight mountain laurel wild populations was

0.19 (Table 4.4), which was lower compared with that of other outcrossing species (mean=0.27)

(Nybom, 2004) yet reasonable. One likely cause for lower average diversity within populations was small population size. A trend toward lower genetic diversity for smaller populations was illustrated by studies on Rhododendron spp. (Chappell et al., 2008; Liu.et al., 2012). The populations surveyed in this study inhabited a relatively small geographic area. Differing from larger populations, insects could cover the entire population and easily transport pollen among all individuals in the population, leading to gene flow among individuals and thus a lower level of within-population genetic diversity. In addition to population size, the ability to self of mountain laurel might contribute to the lower diversity within population compared with other out- crossers, supported by the HS approaching to that of species having mixed breeding system

(mean=0.18) (Nybom, 2004). Selfing could increase the expression of recessive deleterious alleles and result in inbreeding depression in individuals that were likely get eliminated, which reduced the genetic diversity within populations (Liu et al., 2012).

The information on genetic diversity of populations is important for both conservation and breeding programs. Since variation among mountain laurel populations accounted for a low proportion of total genetic diversity (indicated by GST and AMOVA results), collecting individuals from a large number of populations will not necessarily increase diversity in a breeding program. The low average diversity within mountain laurel populations as indicated by low HS exhibited the potential risk of losing genetic variation due to the small geographic size of most populations.

Genetic relationship of wild populations. Nei’s unbiased genetic identity indicated the higher genetic similarity between geographically closed populations (0.8994-0.9169) compared with that between geographically isolated populations (0.8024-0.8913). For instance, two populations in Massachusetts, Blue Hills and Middlesex Fells, had 91.69% of genetic similarity

while Blue Hills only showed 79.75% similarity with the Sandy Creek population in Georgia

(Table 6). Correspondingly, the UPGMA dendrogram revealed four clusters, which correlated with geographic regions (Fig. 4.3). Several factors related to breeding system, population size, and environmental conditions were most likely to lead to this population structure. Cross pollination of mountain laurel relies on insects. The small size of surveyed populations in the study would allow pollinators to cover all individuals throughout the entire population, relatively high genetic similarities were thus observed among individuals within the population and they were clustered together (Fig. 4.3). Also due to insect pollination, pollen could be easily transported between populations that are geographically closed, leading to frequent gene flow between nearby populations (Falk et al., 2001). By contrast, gene flow between geographically isolated populations was reduced, which could be due to different environmental conditions and large distances (Bao et al., 2017). Environmental factors, including temperature and light, could greatly affect phenological phases. According to our investigation, mountain laurel populations at the southernmost range usually have a flowering time one month earlier than populations in

Massachusetts. This gap in blooming period, coupled with the large geographic distance, could prevent insect pollination and reduce gene flow among geographically distant populations.

Hence, nearby populations had relatively high genetic similarity and fall into a cluster; and four clusters correlated with four geographic regions were observed in UPGMA dendrogram (Table

4.6; Fig. 4.3).

Genetic relationship between cultivars and wild accessions. The genetic identity between all possible pairs of 69 mountain laurel taxa widely ranged from 0.5259 to 0.9397 (data not shown). Such wide range was mainly due to the presence of wild accessions, since the lowest

(0.5259) and highest similarity (0.9397) was observed between individual SC6 (The Hitchcock

Woods, Aiken, SC) and GA12 (Sandy Creek, Clarke, GA) and between MA1 and MA4

(Middlesex Fells, Middlesex, MA), respectively. When omitting the similarity among wild accessions, the highest similarity between cultivars and wild plants was reduced to 0.7759, which indicated genetic differentiation between them. Cluster analysis separated mountain laurel cultivars and wild plants into two major groups (data not shown), also indicating a clear differentiation between them. A similar trend has been widely observed in other studies. A clear separation of cultivars, landraces, and wildtypes of hazelnut was observed by Martins et al.

(2014) using both ISSR and AFLP analysis. Sun (2010) observed wild and cultivated accessions of Ilex glabra were classified into two distinct groups using AFLP. Numerous studies working on grapevine also found a clear differentiation between cultivated and wild accessions (Snoussi et al., 2004). This genetic divergence could be due to the fact that these cultivars were principally developed through controlled crosses of cultivated genotypes and followed by artificial selection for certain desirable traits. As a result of this breeding scenario, cultivars lost some alleles and diverged from wild plants. The clear differentiation and relatively low genetic similarity between cultivars and wild accessions informed breeders that inbreeding depression may not be an issue when making crosses between them. Cross hybridization between wild plants and cultivars would be therefore an approach to develop elite cultivars by introducing specific traits from wild plants into a cultivar-adapted background and increasing genetic variations in offspring populations.

Conclusions

ISSR markers exhibited a high level of efficiency in determining genetic relationships among mountain laurel taxa and diversity of wild populations. Five points have been illustrated

by the present study for the breeding of mountain laurel. First, the robustness of ISSR for determining the genetic relationship of cultivars was demonstrated by the observed high genetic identities among related cultivars. Hence, our results can help breeding programs to understand the pedigree of other cultivars and select cultivars for use of parents. Second, all cultivars were uniquely identified by their ISSR profiles; therefore morphologically similar cultivars could be distinguished, and morphological differences of cultivars induced by environmental conditions could be identified using ISSR. Third, a low level of genetic diversity observed among populations indicated that collecting individuals from a large number of populations might not increase diversity in a breeding program. Fourth, although a relatively large proportion of diversity was attributed to within-population variation, the low actual diversity within populations illuminated the potential risk of losing genetic variation due to the small geographic size of most populations. Fifth, the genetic divergence between cultivars and wild accessions due to the loss of genetic information of cultivars during breeding indicated that incorporating wild plants to increase genetic variation and introduce desirable traits would be a potentially useful approach to develop elite cultivars.

Literature Cited

Bao, W., T. Wuyun, T. Li, H. Liu, Z. Jiang, X. Zhu, H. Du, and Y. Bai. 2017. Genetic diversity

and population structure of Prunus mira (Koehne) from the Tibet plateau in China and

recommended conservation strategies. PLos ONE 12:e0188685.

Chappell, M., C. Robacker, and T.M. Jenkins. 2008. Genetic diversity of seven deciduous azalea

species (Rhododendron spp. section Pentanthera) native to the Eastern United States. J.

Amer. Soc. Hort. Sci. 133:374-382.

De Keyser, E., V. Scariot, N. Kobayashi, T. Handa, and J. De Riek. 2010. Azalea phylogeny

reconstructed by means of molecular techniques. Methods Mol.Biol. 589:349-364.

Engels, J.M.M., R.K. Arora, and L. Guarino. 1995. An introduction to plant germplasm

exploration and collecting: planning, methods, and procedures, follow-up, p. 21-63. In:

Guarino, l., V. R. Rao, and R. Reid (eds). Collecting plant genetic diversity: Technical

guidelines. CAB Intl., Wallingford, U.K.

Excoffier, L. and H.E.L. Lischer. 2010. Arlequin suite ver 3.5: A new series of programs to

perform population genetics analyses under Linux and Windows. Mol. Ecol. Resources

10:564-567.

Falk, D.A., E.E. Knapp, and E.O. Guerrant. 2001. An introduction to restoration genetics. Soc.

Ecol. Restoration, Sci. Policy Paper No. 1.

Garriga, M., P.A. Parra, P.D.S. Caligari, J.B. Retamales, B.A. Carrasco, G.A. Lobos, and R.

Garcı´a-Gonza´les. 2013. Application of inter-simple sequence repeats relative to simple

sequence repeats as a molecular marker system for indexing blueberry cultivars. Can. J.

Plant Sci. 93:913-921.

Jaynes, R.A. 1968. Self-incompatibility and inbreeding depression in three laurel species. Amer.

Soc. Hort. Sci. Proc. 93:618-622.

Jaynes, R.A. 1988. Kalmia: The laurel book. Timber Press, Portland, OR.

Kurmes, E.A. 1967. The distribution of Kalmia latifolia L. Amer. Midland Naturalist 77:525-

526.

Liu, X., X. Shen, and Y. Zhang. 2010. Genetic relationship of nine hardy Rhododendrons by

ISSR makers. Acta Agr. Boreali-occidetalis Sinica 19:89-92.

Liu, Y., M. Xing, W. Zhao, R. Fan, S. Luo, and X. Chen. 2012. Genetic diversity analysis of

Rhododendron aureum Georgi (Ericaceae) located on Changbai Mountain using ISSR

and RAPD markers. Plant Systematics Evolution 298:921-930.

Martins, S., F. Simoes, J. Matos, A.P. Silva, and V. Carnide. 2014. Genetic relationship among

wild, landraces and cultivars of hazelnut (Corylus avellana) from Portugal revealed

through ISSR and AFLP markers. Plant Systematics Evolution 300:1035-1046.

Nagaoka, T. and Y. Ogihara. 1997. Applicability of inter-simple sequence repeat polymorphisms

in wheat for use as DNA markers in comparison to RFLP and RAPD markers.

Theoretical Appl. Genet. 94:597-602.

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of

individuals. Genet. 89:583-590.

Nybom, H. 2004. Comparison of different nuclear DNA markers for estimating intraspecific

genetic diversity in plants. Mol. Ecol. 13:1143-1155.

Rohlf, F.J. 2008. NTSYSpc: Numerical Taxonomy System, ver. 2.20. Exeter Publishing, Ltd.:

Setauket, NY.

Shi, A., S. Kantartzi, M. Mmbaga, and P. Chen. 2010. Development of ISSR PCR markers for

diversity study in dogwood (Cornus spp.). Agric. Biol. J. N. Amer. 1:189-194.

Snoussi, H., M.H.B. Slimane, L. Ruiz-García, J.M. Martínez-Zapater, and R. Arroyo-García.

2004. Genetic relationship among cultivated and wild grapevine accessions from Tunisia.

Genome 47:1211-1219.

Sun, Y. 2010. Genetic diversity, propagation, and cold hardiness of Ilex glabra (L.) A. Gray.

Univ. Maine, B, Electronic Theses and Dissertations. 261.

The European Kalmia Society. Laurel cultivars-the international Kalmia register and checklist.

27 Oct. 2018. < http://www.kalmia-society.org/cultivars.php?&lang=gb>.

Yeh, F.C., R.C. Yang, T.B.J. Boyle, Z.H. Ye, and J.X. Mao. 1997. POPGENE, the user-friendly

shareware for population genetic analysis. Molecular Biology and Biotechnology Centre,

University of Alberta, Canada.

Zheng, Y., T. He, L. Chen, L. Chen, J. Rong, and Y. Zheng. 2011. ISSR analysis on the cultivars

of Rhododendron hybridum. J. Fujian Agr. For. Univ. 40:271-275.

Zomlefer, W.B. 1994. Guide to flowering plant families. University of North Carolina Press,

Chapel Hill, NC.

Table 4.1. Geographic information on sampling populations of mountain laurel.

Population Location Latitude Longitude Sample nos.

Middlesex Fells Middlesex, MA 42°26'51"N 71°5'26"W MA1-6

Blue Hills Norfolk, MA 42°12'55"N 71°3'3"W MA7-12

Peachtree Rock Lexington, SC 33°49'47"N 81°11'57"W SC1-6

Hitchcock Woods Aiken, SC 33°33'9"N 81°44'33"W SC7-12

Piedmont College Rabun, GA 34°51'50"N 83°23'16"W GA1-6

Sandy Creek Clarke, GA 34°0'58"N 83°22'32"W GA7-12

Red Hill Monroe, AL 31°44'17"N 87°21'30"W AL1-6

Blackwater Forest Santa Rosa, FL 30°47'23"N 86°47'33"W FL1-6

Table 4.2. Sequence, total number of bands (# B), number of polymorphic bands (# PB), and the percentage of polymorphic bands (PPB) of ISSR primers in 69 mountain laurel taxa.

ISSR primers Sequence (5’ – 3’) # B # PB PPB

UBC808 (AG)8C 14 13 92.86%

UBC814 (CT)8A 10 8 80.00%

UBC835 (AG)8YC 13 12 92.31%

UBC836 (AG)8YA 16 13 81.25%

UBC841 (GA)8YC 13 11 84.62%

UBC856 (AC)8YA 15 14 93.33%

UBC864 (ATG)6 19 19 100.00%

UBC873 (GACA)4 16 15 93.75%

Total 116 105 90.52%

Table 4.3. Nei’s unbiased genetic identity (above diagonal) and genetic distance (below diagonal) for 21 mountain laurel cultivars.

Forever Heart of Kaleido- Nathan Olympic Pink Red Snow- Fire- Ostbo Pepper- Little Tinker- Star- Bullseye Freckles Carol Pristine Sarah Elf Minuet Red Fire scope Hale Fire Charm Bandit drift cracker Red mint Linda bell burst Bullseye **** 0.7636 0.6636 0.7091 0.7000 0.7637 0.7364 0.7000 0.6636 0.7000 0.7455 0.6727 0.7455 0.7182 0.6636 0.6455 0.6545 0.6727 0.6636 0.6909 0.6909 Forever 0.2697 **** 0.7000 0.7091 0.7000 0.6909 0.7182 0.6818 0.6091 0.6818 0.7636 0.7818 0.6909 0.7364 0.6273 0.7364 0.6364 0.6364 0.6273 0.6909 0.7273 Red Freckles 0.4100 0.3567 **** 0.7182 0.5636 0.6636 0.6545 0.6909 0.6000 0.6545 0.7182 0.7000 0.7000 0.7091 0.6727 0.6545 0.6818 0.6636 0.7091 0.7000 0.6818 Heart of 0.3438 0.3438 0.3310 **** 0.6455 0.7273 0.7909 0.7182 0.6818 0.7182 0.7455 0.7818 0.6909 0.6455 0.6273 0.6455 0.6182 0.6727 0.6636 0.6909 0.7273 Fire Kaleido- 0.3567 0.3567 0.5733 0.4378 **** 0.7182 0.6727 0.7818 0.7636 0.6000 0.6818 0.6455 0.7000 0.6000 0.6545 0.7455 0.6818 0.6455 0.6727 0.7364 0.6455 scope Nathan 0.2697 0.3697 0.4100 0.3185 0.3310 **** 0.7727 0.7364 0.6818 0.6818 0.7273 0.6364 0.7091 0.6636 0.6818 0.6636 0.6727 0.7273 0.7000 0.7818 0.7273 Hale Olympic 0.3060 0.3310 0.4238 0.2346 0.3964 0.2578 **** 0.6727 0.6909 0.7091 0.7364 0.6818 0.7909 0.6545 0.6000 0.6545 0.6636 0.7182 0.6364 0.7000 0.7000 Fire Pink 0.3567 0.3830 0.3697 0.3310 0.2461 0.3060 0.3964 **** 0.7091 0.6545 0.6818 0.7000 0.7182 0.6909 0.6909 0.7455 0.6636 0.5727 0.7455 0.7727 0.6636 Charm Red 0.4100 0.4958 0.5108 0.3830 0.2697 0.3830 0.3697 0.3438 **** 0.5818 0.7364 0.7000 0.7364 0.6364 0.6909 0.6909 0.5909 0.7182 0.7273 0.7545 0.5727 Bandit Snow- 0.3567 0.3830 0.4238 0.3310 0.5108 0.3830 0.3438 0.4238 0.5416 **** 0.7182 0.6455 0.7364 0.6364 0.6000 0.6545 0.6636 0.6818 0.6727 0.6455 0.7000 drift Carol 0.2938 0.2697 0.3310 0.2938 0.3830 0.3185 0.3060 0.3830 0.3060 0.3310 **** 0.7818 0.7818 0.7182 0.6455 0.6818 0.6727 0.7455 0.6818 0.7273 0.6727 Fire- 0.3964 0.2461 0.3567 0.2461 0.4378 0.4520 0.3830 0.3567 0.3567 0.4378 0.2461 **** 0.6364 0.6455 0.6273 0.7000 0.6000 0.6545 0.6818 0.6909 0.6909 cracker Ostbo 0.2938 0.3697 0.3567 0.3697 0.3567 0.3438 0.2346 0.3310 0.3060 0.3060 0.2461 0.4520 **** 0.6455 0.6455 0.6455 0.7091 0.6727 0.7182 0.7273 0.6909 Red Pepper- 0.3310 0.3060 0.3438 0.4378 0.5108 0.4100 0.4238 0.3697 0.4520 0.4520 0.3310 0.4378 0.4378 **** 0.7091 0.6364 0.6455 0.6091 0.6727 0.7364 0.6636 mint Pristine 0.4100 0.4664 0.3964 0.4664 0.4238 0.3830 0.5108 0.3697 0.3697 0.5108 0.4378 0.4664 0.4378 0.3438 **** 0.6545 0.6455 0.6818 0.7455 0.7364 0.6273

Sarah 0.4378 0.3060 0.4238 0.4378 0.2938 0.4100 0.4238 0.2938 0.3697 0.4238 0.3830 0.3567 0.4378 0.4520 0.4238 **** 0.6818 0.7182 0.6909 0.6818 0.6273

Elf 0.4238 0.4520 0.3830 0.4810 0.3830 0.3964 0.4100 0.4100 0.5261 0.4100 0.3964 0.5108 0.3438 0.4378 0.4378 0.3830 **** 0.6909 0.7727 0.7273 0.6182 Little 0.3964 0.4520 0.4100 0.3964 0.4378 0.3185 0.3310 0.5573 0.3310 0.3830 0.2938 0.4238 0.3964 0.4958 0.3830 0.3310 0.3697 **** 0.6455 0.7273 0.6727 Linda Minuet 0.4100 0.4664 0.3438 0.4100 0.3964 0.3567 0.4520 0.2938 0.3185 0.3964 0.3830 0.3830 0.3310 0.3964 0.2938 0.3697 0.2578 0.4378 **** 0.7545 0.6636 Tinker- 0.3697 0.3697 0.3567 0.3697 0.3060 0.2461 0.3567 0.2578 0.2816 0.4378 0.3185 0.3697 0.3185 0.3060 0.3060 0.3830 0.3185 0.3185 0.2816 **** 0.7636 bell Star- 0.3697 0.3185 0.3830 0.3185 0.4378 0.3185 0.3567 0.4100 0.5573 0.3567 0.3964 0.3697 0.3697 0.4100 0.4664 0.4664 0.4810 0.3964 0.4100 0.2697 **** burst

Table 4.4. Percentage of polymorphic bands (PPB), total genetic diversity (HT), average genetic diversity within populations (HS), and proportion of total genetic diversity attributed to among- population variation (GST) for eight mountain laurel populations.

Population PPB HT HS GST

Middlesex Fells, MA 45.69

Blue Hills, MA 40.52

MA group 59.48 0.2059 0.1643 0.2023

Peachtree Rock, SC 55.17

Hitchcock Woods, SC 50.86

SC group 69.83 0.2523 0.2062 0.1827

Piedmont College, GA 56.03

Sandy Creek, GA 47.41

GA group 75 0.2474 0.1988 0.1965

Red Hills, AL 50.86

Blackwater Forest, FL 54.31

Southernmost group 69.83 0.2391 0.1956 0.1818

Total 90.52 0.3089 0.1912 0.3809

Table 4.5. Genetic variation among and within populations of eight mountain laurel populations based on analysis of molecular variation (AMOVA).

Source of d.f. Sum of Variance Percentage of P variation squares component variance

Among 7 327.542 5.58 29.54 <0.001 populations

Within 40 532.333 13.31 70.46 <0.001 populations

Total 47 859.875 18.89

Table 4.6. Nei’s unbiased genetic identity (above diagonal) and genetic distance (below diagonal) for eight mountain laurel populations.

Piedmont Middlesex Blue Hills, Peachtree Hitchcock Red Hill, Blackwater Sandy College, Fells, MA MA Rock, SC Woods, SC AL Forest, FL Creek, GA GA Middlesex **** 0.9169 0.8888 0.8472 0.8290 0.8481 0.8913 0.8162 Fells, MA Blue Hills, 0.0868 **** 0.8650 0.8160 0.8024 0.8193 0.8734 0.7975 MA Peachtree 0.1178 0.1451 **** 0.9053 0.8564 0.8544 0.8745 0.8524 Rock, SC Hitchcock 0.1659 0.2033 0.0994 **** 0.8636 0.8306 0.8595 0.8462 Woods, SC Red Hill, 0.1875 0.2201 0.1550 0.1466 **** 0.9121 0.8304 0.8107 AL Blackwater 0.1647 0.1993 0.1574 0.1857 0.0920 **** 0.8561 0.8092 Forest, FL Piedmont 0.1151 0.1354 0.1341 0.1514 0.1859 0.1554 **** 0.8994 College, GA Sandy 0.2031 0.2263 0.1597 0.1670 0.2098 0.2117 0.1061 **** Creek, GA

Figure 4.1. UPGMA dendrogram of 21 mountain laurel cultivars generated based on Nei’s unbiased genetic identity and distance matrix. The accuracy of dendrogram was demonstrated by the clustering of cultivars with their relatives. Five groups were observed in dendrogram as shown in the figure.

Figure 4.2. ISSR fingerprints generated from mountain laurel cultivars using primer UBC835.

Figure 4.3. UPGMA dendrogram of 48 mountain laurel wild accessions (six individuals sampled from each of eight populations) constructed based on Nei’s unbiased genetic identity and distance matrix. Six individuals within the population were clustered together. Meanwhile, populations located in a geographic region were grouped into a cluster and four clusters correlated with four geographic regions were observed.

CHAPTER 5

IN VITRO SEED GERMINATION OF KALMIA LATIFOLIA L. HYBRIDS: A MEANS FOR

IMPROVING GERMINATION AND SPEEDING UP BREEDING CYCLE3

3Li, H. and D. Zhang. 2018. HortScience 53:535-540.

Reprinted here with permission of the publisher.

Abstract. Kalmia latifolia L. (mountain laurel), a member of Ericaceae, is a beautiful ornamental shrub native to the eastern United States. The plant is not common in the southeastern United

States landscapes because of the limited heat tolerance of most commercial cultivars. Breeding of heat-tolerant cultivars can be achieved by cross hybridization, but is often challenged by low germination percentage, long germination time, and potential abortion of cross-hybridized seeds.

We used in vitro seed germination to enhance germination and shorten germination time and investigated the appropriate collecting time, optimal basal medium, and pH for this approach.

Collecting time affected in vitro seed germination, with more mature hybrid seeds [collected 4–5 months after pollination (MAP)] having higher germination rate (90% in 4 weeks) than the less mature seeds collected in 2 MAP (20% in 7 weeks). Seedlings from the mature seeds also produced two true leaves on average after 8 weeks of culture, whereas seedlings from the less mature seeds had no true leaves. Woody Plant Medium (WPM) better enhanced in vitro seed germination compared with Murashige and Skoog (MS) or Gamborg’s B5 (B5) medium. WPM yielded higher germination (98%) than MS (90%) and significantly greater total leaf area per seedling (67 mm2) than MS (50 mm2) and B5 (52 mm2) for seeds of ‘Firecracker’ x ‘Snowdrift’.

Similar effects had been observed on seeds from ‘Little Linda’ x ‘Starburst’ and ‘Pristine’ x

‘Peppermint’. The pH ranging from 4.2 to 5.4 did not affect seed germination and seedling development of mountain laurel hybrids. Our protocol enabled early collection of mountain laurel hybrid seeds 1 month before their full maturation and permitted seeds to germinate in 4 weeks on WPM, which shortened the period from crossing to the seedling stage from up to 15 to

6 months and enhanced germination percentage from 30% to more than 90% compared with traditional seed germination. This protocol should be applied to promote the breeding and selection of new mountain laurel cultivars for the southeastern United States landscapes.

Additional index words. ericaceous, mountain laurel, ornamental plants, rapid breeding, tissue culture

Introduction

Kalmia latifolia L. (mountain laurel), a member of Ericaceae (heath family), is a large evergreen flowering shrub native to the eastern United States. It has been considered by many horticulturists and gardeners to be one of the most beautiful flowering species because of its attractive lustrous green foliage and showy inflorescence (Dirr, 2011; Jaynes, 1988). It usually blooms in late spring or early summer on new shoots. The inflorescence consists of a terminal compound corymb with numerous showy flowers. The variations in flower shape and color, leaf shape and size, and habit have made mountain laurel a valuable ornamental shrub in gardens and landscapes (Jaynes, 1988). More than 100 mountain laurel cultivars have been released through the effort of ornamental breeders since the early 1960s. However, most cultivars were selected and only evaluated in the northeastern United States and often were not widely distributed until the plants were named and introduced. Their responses over a broad range of growing sites were thus unknown (Jaynes, 1982, 1988). As a result, most of mountain laurel cultivars failed to have reliable performance and gain market share in the southeastern United States. To improve the popularity of mountain laurel in the southeastern landscapes, plants that have capacity of coping with heat are desired.

Hybridization crosses between the southernmost mountain laurel populations and existing cultivars provide the main source of heat tolerant cultivars. Unfortunately, seed germination of mountain laurel has been challenging for breeders. Mountain laurel seeds generally become fully mature in 5 months after fertilization (Jaynes, 1988). Mature seeds only

yield approximately 30% to 40% of germination and often require up to 3 months to germinate even under favorable environmental conditions (Jaynes, 1971). The low and uneven germination and long germination period thus slow down the breeding cycle. Furthermore, seeds from cross hybridization may abort because of the incompatibility. A better and faster method to germinate seeds of mountain laurel hybrids is needed to speed up breeding of new cultivars.

In vitro culture is the cultivation of plant tissues or organs under aseptic conditions in an artificial medium of known chemical composition. For the last 50 years, this culture technique has been widely used on many plants to obtain seedlings from interspecific and intergeneric breeding, of which embryos lack of endosperm and seeds fail to germinate using conventional methods (Bridgen, 1994; De Jeu, 2000; Sharma, 1995). By culturing immature embryos or seeds in the nutrient medium, some can overcome the incompatibility and avoid abortion. Successful case has been documented for another member of heath family - rhododendron hybrids

(Eeckhaut et al., 2007). Some types of seeds only germinate after a period of storage to overcome dormancy, and in vitro culture can break seed dormancy and shorten seed germination time (Sharma et al., 1996). Therefore, in vitro seed germination methods would be a potential means to enhance seed germination of mountain laurel.

Successful in vitro culture depends on many factors and the protocol varies depending on genotypes. Medium selection is considered as one of the most important factors for tissue culture.

Several formulations such as MS (Murashige and Skoog, 1962), B5 medium (Gamborg et al.,

1968), and WPM (Lloyd and McCown, 1980) with certain degrees of modification, are the most widely used basal media in embryo and seed culture (Askari-Khorasgani et al., 2013; Hao et al.,

2014; Pavlovic et al., 2012). The pH of media can also affect plant growth (Reeves et al., 1983).

Generally, the pH of tissue culture media is maintained around 5.7-5.9, whereas adjusted pH can

improve the embryo development and plant growth for some species (Ribeiro et al., 1999; Zhang et al., 2004). In addition to these cultivation conditions, determining the appropriate time to collect seeds is a key factor for the success of in vitro culture. Seeds may fail to germinate when they are harvested in their early stage whereas later collection could lead to increasing contamination rates (Yang et al., 2015).

In this study, we investigated the effects of seed maturity, basal medium, and pH on the germination percentage and speed, as well as seedling vigor for several mountain laurel hybrids.

Our objective was to establish an efficient in vitro seed germination protocol for mountain laurel hybrid seeds to enable seed harvest before full maturity, shorten the time required for germination by breaking down dormancy, and improve seed germination by providing favorable conditions. This protocol would be an approach to speed up the breeding cycle and promote the breeding and selection of new mountain laurel cultivars for the southeastern landscapes.

Materials and Methods

Plant materials, sterilization, and culture conditions. All fruit capsules used in the study were obtained from artificial crosses between mountain laurel cultivars. Female parents were emasculated and hand pollinated in early May 2016 under greenhouse conditions at the

University of Georgia Horticulture Farm (Watkinsville, GA). Capsules were harvested from mother plants and washed with running tap water for 30 min. Capsules were then surface disinfected with 50% Clorox solution (4.15% sodium hypochlorite; The Clorox Company,

Oakland, CA) added with one drop of Tween 20 (Hoefer, Holliston, MA) for 6 min, and subsequently rinsed four times with autoclaved distilled water in a laminar flow hood. The surface-sterilized capsules were kept in sterile water until dissection. Seeds were extracted under

a dissecting microscope (Fisher Scientific Education, Nazareth, PA) and isolated seeds were transferred onto the surface of medium in 6-cm petri dishes. Each petri dish contained 10 mL of basal media supplemented with 3% sucrose (Sigma-Aldrich, St. Louis, MO) and 0.6% agar

(Fisher Science Education) and was sealed with parafilm (Bemis Company, Oshkosh, WI) after sowing. Eight weeks after culture, measurements were taken and seedlings were subcultured to

10-cm petri dishes filled with 25 mL of same basal medium. Subculture was then made every 4 weeks. All petri dishes were incubated in a culture room at 25 ± 2 °C under cool-white fluorescent lamps [photosynthetic photon flux (PPF) of 66 mol·m-2·s-1] for 14-h photoperiod because light is required for germination of mountain laurel seeds (Malek et al., 1989).

Effect of seed-collecting time on germination. Fruit capsules from the crosses ‘Elf’ x

‘Little Linda’ and ‘Red Bandit’ x ‘Minuet’ were harvested every month from July to Oct. 2016

(2-5 MAP). Capsule growth was measured by a ruler and the appearance was photographed.

Extracted seeds were observed and photographed under dissecting microscope. Culture medium contained B5 basal medium supplemented with 3% sucrose and 0.6% agar, and pH was adjusted to 5.0. All cultures were maintained under the aforementioned environmental conditions.

Effect of basal medium on seed germination. Hybrid capsules from ‘Little Linda’ x

‘Starburst’, ‘Firecracker’ x ‘Snowdrift’, and ‘Pristine’ x ‘Peppermint’ were collected on 15 Sept.

2016. The surface-disinfected capsules were dissected and seeds were isolated from capsules under a microscope. The isolated seeds were cultured in 6-cm petri dishes containing either 1)

MS, 2) WPM, or 3) B5 basal medium plus 3% sucrose and 0.6% agar. The pH of media was adjusted to 5.0 before adding agar.

Effect of pH on seed germination. Hybrid capsules from ‘Little Linda’ x ‘Starburst’,

‘Firecracker’ x ‘Snowdrift’, and ‘Pristine’ x ‘Peppermint’ were harvested on 16 Sept. 2016.

Capsules were disinfected following the aforementioned procedure. The extracted seeds were sowed into B5 medium with following pH: 1) 4.2, 2) 4.6, 3) 5.0, or 4) 5.4. All petri dishes were incubated in the culture room as previously mentioned.

Acclimatization. After two cycles of subcultures, more than 300 vigorous seedlings (with at least four true leaves) of all crosses were taken out from petri dishes and washed with distilled water to remove remains of agar. Seedlings were then transplanted to presoaked peat moss in 32- cell flat trays for acclimatization and subsequent growth. Trays were covered with plastic dome to maintain high humidity. The dome was gradually vented after seedlings started to harden off and was eventually removed during acclimatization. All plants were kept in a growth chamber at

25 ± 2 °C under a 16-h photoperiod provided by cool-white fluorescent lamps at PPF of 66

mol·m-2·s-1. Seedlings were fertilized with 50 ppm 15N-4P-15K water-soluble fertilizer (J.R.

Peters, Inc., Allentown, PA) weekly and watered as needed.

Experimental design and data analysis. All studies were performed in a completely randomized design with four replications (petri dishes) per treatment and 10 seeds per treatment per replication. Seed germination was observed weekly under microscope and visible radicle emergence was counted as germination (Fig. 5.1C). Germination percentage was calculated as

(number of seeds showing radicle emergence 8 weeks after culture/total number of seeds cultured) x 100%. Germination speed (T100) was calculated as the number of weeks reaching

100% of total germination. Individual petri dishes were photographed 8 weeks after culture for seedling growth analysis. Number of true leaf and total leaf area were measured using ImageJ

1.x (Fig. 5.1D; Schneider et al., 2012). Data were analyzed by analysis of variance and GLM program in SAS University Edition (SAS Institute, Cary, NC) and the significance level was set at P < 0.05.

Results and Discussion

The effect of collecting time on capsule development and seed germination. Capsules were successfully harvested from both crosses (‘Elf’ x ‘Little Linda’ and ‘Red Bandit’ x

‘Minuet’) 2 to 5 MAP. Capsule expansion completed in the first 2 months with a significant increase in capsule diameter and no further growth thereafter. Capsule color changed from green to yellow and then to brown during collecting period, with the capsule of ‘Red Bandit’ x

‘Minuet’ ripened earlier than that of ‘Elf’ x ‘Little Linda’. The capsule from the cross between

‘Elf’ and ‘Little Linda’ was smaller compared with that from ‘Red Bandit’ x ‘Minuet’ at the same collecting time (Fig. 5.1A). This is because that the female parent ‘Elf’ is a miniature cultivar with a dwarf growth habit and reduced leaf and flower size and therefore produces smaller fruit than nondwarf cultivar ‘Red Bandit’. Seeds from both crosses were small, ranging from 0.7 to 0.9 mm in length and 0.3 to 0.4 mm in width (Fig. 5.1B). This is the normal seed size of mountain laurel (Jaynes, 1988). Both crosses did not show significant changes in seed size or seed color from 2 to 5 MAP.

The germination capability of hybrid seeds of mountain laurel varied with the collecting time. Less than 20% of seeds germinated when the fruit capsule was collected in July (2 MAP), whereas germination percentage of August collection for cross ‘Red Bandit’ x ‘Minuet’ and cross ‘Elf’ x ‘Little Linda’ increased significantly to 65.0% and 92.5%, respectively (Fig. 5.2A).

In September and October collections, both crosses yielded more than 90.0% of germination.

However, contamination percentage increased significantly in October collection, particularly for the hybrids between ‘Red Bandit’ x ‘Minuet’ (12.5%) that ripened earlier than the other cross

(Fig. 5.2B). Hybrid seeds from both crosses tended to have faster and more even germination when they were harvested later, indicated by a negative correlation between T100 and collecting

time (Fig. 5.2C). T100 was 6.5 weeks in early collection for the cross ‘Elf’ x ‘Little Linda’ whereas T100 in September and October collections was reduced to approximate 4 weeks, although the difference was not significant. A similar trend was observed with the hybrid seeds from cross ‘Red Bandit’ x ‘Minuet’ (Fig. 5.2C). Germinated seedlings became increasingly vigorous as harvesting time went later. Seedlings from seeds that were harvested at the early stage (2 MAP) produced no true leaf after 8 weeks of culture, only radicles or cotyledons were observed. In contrast, seedlings were able to yield significantly more true leaves when seeds were collected in later collections, particularly October collection having two true leaves on average per seedling (Fig. 5.2D). Results showed that collecting time affected in vitro seed germination, with more mature hybrid seeds having higher capacity to germinate compared with immature seeds. Similar results were reported on closely related ericaceous species rhododendron by Michishita et al. (2001) that less mature hybrid seeds had lower ability to germinate in vitro compared with mature seeds. Geng et al. (2014) also found that younger rhododendron hybrid embryos were more likely to only form callus while relatively mature embryos tended to germinate successfully into seedlings. Mountain laurel seeds generally start to ripen in 4 MAP and become fully mature in 5 MAP, although maturation time varies a bit from year to year and among plants. Two-month-old capsules did not yield any seedlings because of the underdeveloped embryos at this stage, whereas most of the seeds harvested in September (4

MAP) were able to germinate evenly because seeds became mature and embryos enabled the radicle to penetrate the endosperm and seedcoat (Jaynes, 1988; Taiz and Zeiger, 2010). Later collection thus yielded higher germination percentages, more leaf growth, and faster germination than early collection. However, capsules might start to dehisce in 5 MAP and the longer the capsules were exposed to field conditions, the higher the infection occurred, which led to

increased contamination rates in October collection. Four MAP therefore would be recommended to be the appropriate time for collecting seeds for in vitro culture. The seeds of

‘Red Bandit’ x ‘Minuet’ had overall lower germination ability, including germination percentage and seedling vigor, compared with that of ‘Elf’ x ‘Little Linda’ (Fig. 5.2). This is probably because mountain laurel is a cross-pollinated species and cultivars ‘Red Bandit’ and ‘Minuet’ had relatively higher genetic similarity (72.73%) than that of the other cross (69.09%) (results from a separate study) and their hybrid seeds are less capable to germinate because of higher inbreeding depression. The overall results showed that collecting and cultivating 4-month-old hybrid seeds of both crosses in vitro could shorten the period from crossing to the seedling stage from 13-15 months to 6 months and improve germination percentage from 30% to more than

90%, which is of significance for plant breeding programs.

Effect of basal medium on seed germination. The highest germination percentages were obtained with the treatment of WPM in 8 weeks for all three crosses (Table 5.1). The cross

‘Little Linda’ x ‘Starburst’ yielded significantly more germination on WPM (90.0%) than that of

MS (72.5%) and B5 (72.5%), whereas ‘Firecracker’ x ‘Snowdrift’ and ‘Pristine’ x ‘Peppermint’ had 97.5% and 92.5% of germination, respectively, with both WPM and B5. Although WPM and

B5 induced similar germination percentage, the smaller T100 of WPM (2.0-3.5 weeks) indicated that WPM tended to have higher capacity of shortening germination time compared with B5

(T100 of 2.8-4.5 weeks). In terms of seedling vigor, the results showed that the effect of basal medium on true leaf number per seedling was genotype-specific. MS promoted more leaf production of the seedlings of ‘Little Linda’ x ‘Starburst’ and ‘Firecracker’ x ‘Snowdrift’ whereas WPM enhanced that of ‘Pristine’ x ‘Peppermint’, but the difference was not statistically significant. In contrast, there was a trend between total leaf area per seedling and medium types

that WPM yielded larger leaf area than B5 and MS and the difference was significant with the cross ‘Firecracker’ x ‘Snowdrift’. Foliage chlorosis was observed on seedlings in B5. The data showed that WPM better enhanced in vitro seed germination of mountain laurel hybrids than MS and B5, which was consistent with the findings of Geng et al. (2014). In their study, WPM was found to be the optimal medium for seed germination of rhododendron hybrids that indicated by higher germination percentage and better seedling growth compared with other media including

MS, 1/4MS, and Anderson (Anderson, 1980). WPM is originally designed for micropropagation of mountain laurel by Lloyd and McCown (1980) for the purpose of propagating mountain laurel clones in large scales to fulfill commercial demand. WPM has then been widely used on many other dicot ornamental species, such as rose, birch, and oak, for their tissue cultures (McCown,

2000). The major difference between WPM and MS is that MS has overall higher nutrient contents, particularly nitrogen (>4-fold), than WPM (Lloyd and McCown, 1980; Murashige and

Skoog, 1962). This high nutrient content of MS might be excessive to tiny young seeds and thus suppressed their germination and seedling development. In contrast, B5 has low inorganic nutrient levels, such as calcium and magnesium, which might lead to nutrient deficiency and inhibit seedling growth. In addition to the low nutrient levels, the primary nitrogen source of B5

- is nitrate (NO3 ), which might account for the stunted growth as well. Many ericaceous members,

+ including rhododendrons and blueberries, have a preference for the ammonium form (NH4 ) over

- the nitrate form (NO3 ) of inorganic nitrogen. In the absence of this preferred condition, these ericaceous plants may display foliar chlorosis and stunted shoot growth (Alt et al., 2017; Clark et

+ - al., 2003). In B5 medium, the NH4 / NO3 ratio is 0.08, much lower than that of WPM and MS

- (0.52) (Gamborg et al., 1968). Therefore, the low nutrient level and as NO3 the dominant nitrogen form of B5 medium probably limited seed germination and resulted in foliage chlorosis

and smaller seedlings of mountain laurel hybrids. No significant difference in seed germination between cross combinations was observed; however, ‘Little Linda’ x ‘Starburst’ had a smaller total leaf area but similar leaf number than other two crosses. Because ‘Little Linda’ and

‘Starburst’ are miniature cultivars that produce smaller leaves than other normal cultivars and this trait is controlled by a single recessive gene (Jaynes, 1988), their progenies were all miniatures producing reduced leaf and had relatively smaller total leaf area.

Effect of pH on seed germination. pH values did not show significantly different effect on in vitro seed germination of three mountain laurel hybrids (Table 5.2). Seeds collected from all crosses (‘Little Linda’ x ‘Starburst’, ‘Firecracker’ x ‘Snowdrift’, and ‘Pristine’ x ‘Peppermint’) yielded more than 80% of germination in 4 weeks under all pH treatments. Neither number of true leaf nor total leaf area displayed significant difference between treatments within each cross combination. Like most ericaceous members such as rhododendrons and blueberries, mountain laurel is considered to prefer acidic soil condition ranging between 4.0 and 5.5 (Jaynes, 1988).

However, our results showed that none of these hybrid seeds exhibited a pH preference by their in vitro germination. This could be due to that the pH range (4.2-5.4) used in our study fell into the desired pH conditions of mountain laurel, and thus, all treatments favored seed germination.

Alternatively, the preference on acidic soil condition may not be attributed by pH per se but by the nutrient availability under low pH values. Under acidic soil conditions, NH4+ is present as

- the primary form of inorganic nitrogen rather than NO3 (Alt et al., 2017). As previously

+ - mentioned, many of ericaceous plants prefer NH4 over NO3 of inorganic nitrogen (Alt et al.,

2017; Clark et al., 2003). This probably accounts for adaptation of heath family plants to the soil with low pH. Availability of other nutrients, such as iron, may also cause this preference for low pH. Under our tissue culture condition, the hybrid seeds were exposed to the same nutrient

components when testing between different pH treatments. Therefore, pH did not significantly influence in vitro seed germination and seedlings vigor in our study. When comparing germination of three hybrid seeds, the similar trend was found as in basal medium study that seedlings from ‘Little Linda’ x ‘Starburst’ had smaller total leaf area than that of other two crosses because two parents involved in this cross are miniatures.

Acclimatization. Seedlings having at least four true leaves were transplanted into presoaked peat moss in 32-cell flat trays for acclimatization and subsequent growth. Seedlings displayed significant aboveground growth in 3 weeks following transplanting (Fig. 5.1F). After

3-month acclimatization, 83.9% of seedlings survived and yielded well-developed shoot and root system (Fig. 5.1G and H). Seedlings were subsequently transplanted to 1-gallon pots filled with pine bark as substrate for future evaluation.

Conclusions

Mountain laurel seeds are generally collected in 5-6 MAP when they are fully mature.

Seeds will be then stored at 4 °C and sowed in the following spring using traditional seed germination. A mix of peat moss and perlite at a ratio of 2:1 is used as substrate, and seeds should be sown on the surface of the mix and not covered. The moist substrate and high humidity are necessary for germination (Jaynes, 1971). This protocol requires careful handling but only yields low germination percentage and the germination could take up to 3 months. The subsequent seedling growth is generally slow, which takes approximately 1 year to reach a plant height of 10 cm. While using our in vitro seed germination protocol, hybrid seeds could be collected 1 month before their full maturation and seeds were able to germinate in 1 month on

WPM, which shortened the period from crossing to the seedling stage from 13-15 months to 6

months and improved germination percentage from 30% to more than 90% compared with the traditional method. After two cycles of subcultures, seedlings were vigorous enough to be transplanted from petri dish to peat moss for acclimatization. Seedlings with both great shoot and root system were obtained in 3 months after acclimatization. The protocol developed in this study therefore enables to increase the number of hybrid seedlings and to speed up the breeding cycle by improving germination percentage, shortening germination time, and enhancing seedling growth. This protocol thus can be an approach to promote the breeding and selection of new mountain laurel cultivars for the southeastern United States landscapes.

Literature Cited

Alt, D.S., J.W. Doyle, and A. Malladi. 2017. Nitrogen-source preference in blueberry

(Vaccinium sp.): Enhanced shoot nitrogen assimilation in response to direct supply of

nitrate. J. Plant Physiol. 216:79-87.

Anderson, W.C. 1980. Tissue culture propagation of red and black raspberries, Rubus idaeus and

R. occidentalis. Acta Hort. 112:13-20.

Askari-Khorasgani, O., F. Mortazaeinezhad, M. Otroshy, and A.R. Golparvar. 2013. Breaking

seed dormancy of endangered medicinal plant Kelussia odoratissima using zygotic

embryo culture technique. Tech. J. Eng. Appl. Sci. 3:1712-1718.

Bridgen, M.P. 1994. A review of plant embryo culture. HortScience 29:1243-1246.

Clark, M.B., H.A. Mills, C.D. Robacker, and J.G. Latimer. 2003. Influence of nitrate:

Ammonium ratios on growth and elemental concentration in two azalea cultivars. J. Plant

Nutr. 26:2503-2520.

De Jeu, M. 2000. In vitro techniques for ornamental breeding. Acta Hort. 508:55-60.

Dirr, M.A. 2011. Manual of woody landscape plants. 6th ed. Stipes Publishing, Champaign, IL.

Eeckhaut, T., E. De Keyser, J. Van Huylenbroeck, J. De Riek, and E. Van Bockstaele. 2007.

Application of embryo rescue after interspecific crosses in the genus Rhododendron.

Plant Cell Tissue Organ Cult. 89:29-35.

Gamborg, O.L., R.A. Miller, and K. Ojima. 1968. Nutrient requirements of suspension cultures

of soybean root cells. Expt. Cell Res. 50:151-158.

Geng, X., Y. Wu, and H. Zhao. 2014. Preliminary study on hybrid embryo rescue of

Rhododendrons. J. Yunnan Agr. Univ. 29:533-539.

Hao, Y., H. Wang, R. Yang, F.L. Wang, and J. Lu. 2014. Effectiveness of culture media on the

success of embryo rescue among different crosses of seedless grapes. Acta Hort.

1046:259-263.

Jaynes, R.A. 1971. Seed germination of six Kalmia species. J. Amer. Soc. Hort. Sci. 96:668-672.

Jaynes, R.A. 1982. New mountain laurel selections and their propagation. Intl. Plant Prop. Soc.

Proc. 32:431-434.

Jaynes, R.A. 1988. Kalmia: The laurel book. Timber Press, Portland, OR.

Lloyd, G. and B. McCown. 1980. Commercially-feasible micropropagation of mountain laurel,

Kalmia latifolia, by use of shoot-up culture. Intl. Plant Prop. Soc. Proc. 30:421-427.

Malek, A.A., F.A. Blazich, S.L. Warren, and J.E. Shelton. 1989. Influence of light and

temperature on seed germination of mountain laurel. J. Environ. Hort. 7:161-162.

McCown, B.H. 2000. Recalcitrance of woody and herbaceous perennials plants: Dealing with

genetic predeterminism. In Vitro Cell. Dev. Biol. Plant 36:149-154.

100

Michishita, A., K. Ureshino, and I. Miyajima. 2001. Shortening the period from crossing to the

seedling stage through ovule culture of interspecific crosses of azalea (Rhododendron

spp.). J. Jpn. Soc. Hort. Sci. 70:54-59.

Murashige, T. and F. Skoog. 1962. A revised medium for rapid growth and bioassays with

tobacco tissue cultures. Physiol. Plant. 15:473-497.

Pavlovic, S., B. Vinterhalter, S. Zdravkovic-Korac, D. Vinterhalter, J. Zdravkovic, D. Cvikic,

and N. Mitic. 2012. Recurrent somatic embryogenesis and plant regeneration from

immature zygotic embryos of cabbage (Brassica oleracea var. capitata) and cauliflower

(Brassica oleracea var. botrytis). Plant Cell Tissue Organ Cult. 113:397-406.

Reeves, D.W., B.D. Horton, and C.A. Couvillon. 1983. Effect of media and media pH on in vitro

propagation of ‘Nemaguard’ peach rootstock. Scientia Hort. 21:353-357.

Ribeiro, V.G., M. Psaqual, J.D. Ramos, G.R. Carvalho, and A.F. Oliveira Jr. 1999. Effect of agar

and pH on in vitro orange ‘Natal’ cultivar embryo. Rev. Ceres 46:587-595.

Schneider, C. A., W.S. Rasband, and K.W. Eliceiri. 2012. NIH Image to ImageJ: 25 years of

image analysis. Nat. Methods 9:671-675.

Sharma, D.R., R. Kaur, and K. Kumar. 1996. Embryo rescue in plants - A review. Euphytica

89:325-337.

Sharma, H.C. 1995. How wide can a wide cross be? Euphytica 82:43-64.

Taiz, L. and E. Zeiger. 2010. Plant physiology. 5th ed. Sinauer Assoc., Sunderland, MA.

Yang, Y., D. Zhang, Z. Li, X. Jin, and J. Dong. 2015. Immature embryo germination and its

micropropagation of Ilex crenata Thunb. HortScience 50:733-737.

Zhang, M., M. Han, X. Wang, Y. Tian, S. Wang, A. Wang, and X. Zhu. 2004. Factors affecting

embryo rescue of early ripening nectarines. J. Fruit Sci. 21:382-384.

101

Table 5.1. Effect of basal medium on in vitro germination of 4-MAP hybrid seeds of Kalmia latifolia.

Germination T100 No. of true Total leaf area Cross combination Medium (%)z (weeks)y leaves (mm2)

MS 90.0 ±5.8 2.25±0.25 2.75±0.23 50.41±3.51 b ‘Firecracker’ x B5 97.5±2.5 3.50±0.65 2.24±0.32 52.26±4.44 b ‘Snowdrift’ WPM 97.5±2.5 2.75±0.48 2.48±0.24 67.03±6.27 a

MS 72.5±4.8 b 3.50±0.50 2.70±0.44 38.02±6.68 ‘Little Linda’ x B5 72.5±2.5 b 4.50±1.19 2.47±0.28 39.55±7.77 ‘Starburst’ WPM 90.0±4.1 a 3.50±0.29 1.61±0.25 40.98±3.27

MS 85.0±6.5 4.50±1.19 1.50±0.23 44.25±8.01 ‘Pristine’ x B5 92.5±4.8 2.75±0.75 1.41±0.15 53.70±5.67 ‘Peppermint’ WPM 92.5±4.8 2.00±0.00 1.75±0.35 66.83±14.89 z Data are present as mean of four replications ± SE. Means followed by different letters within each cross combination in a column indicate that they are significantly different at p < 0.05. y Seed germination was recorded weekly and T100 (indicating germination speed) was calculated as the number of weeks reaching 100% of total germination.

MS = Murashige and Skoog; B5 = Gamborg’s B5; WPM = Woody Plant Medium; MAP = months after pollination.

102

Table 5.2. Effect of pH on germination and seedling vigor of 4-MAP Kalmia latifolia hybrid seeds cultivated in vitro.

Cross Germination T100 No. of true Total leaf area pH combination (%)z (weeks)y leaves (mm2)

4.2 90.0±7.1 2.75±0.25 2.15±0.39 48.28±5.75

‘Firecracker’ x 4.6 92.5±4.8 2.50±0.29 2.20±0.12 52.78±3.98

‘Snowdrift’ 5.0 95.0±2.9 2.75±0.25 2.32±0.28 44.38±5.31

5.4 95.0±2.9 2.50±0.29 2.28±0.21 49.72±2.52

4.2 80.0±5.8 3.25±0.48 2.33±0.41 37.60±5.12

‘Little Linda’ x 4.6 77.5±10.3 3.75±0.85 1.98±0.15 38.43±3.16

‘Starburst’ 5.0 82.5±4.8 4.00±0.41 1.89±0.13 40.26±0.98

5.4 85.0±2.9 2.75±0.25 1.74±0.06 35.77±2.87

4.2 95.0±2.9 3.00±0.41 1.35±0.21 45.95±6.29

‘Pristine’ x 4.6 100.0±0 2.50±0.50 1.53±0.08 49.25±1.90

‘Peppermint’ 5.0 90.0±4.1 2.75±0.48 1.37±0.13 44.25±4.83

5.4 87.5±4.8 3.25±0.75 1.25±0.06 42.42±3.28 z Data are present as mean of four replications ± SE. Data were statistically analyzed and were not significantly different at p < 0.05. y Seed germination was recorded weekly and T100 (indicating germination speed) was calculated as the number of weeks reaching 100% of total germination.

MAP = months after pollination.

103

Figure 5.1. In vitro seed germination of mountain laurel hybrids. (A) Capsule development during the collecting time (from July to October) of crosses ‘Elf’ x ‘Little Linda’ (top) and ‘Red

Bandit’ x ‘Minuet’ (bottom). (B) Seed extracted from the disinfected capsule in the laminar flow hood, which would then be sowed on medium in 6-cm petri dish. (C) Seed germinated in 2 weeks after culture as radicle emergence was observed. (D) Eight weeks after culture, seedlings were photographed for data collection and then subcultivated to fresh medium. (E) Seedlings in

10-cm petri dish after two cycles of subcultures; vigorous ones would be transplanted to presoaked peat moss for acclimatization and further development. (F) Seedling growth after being acclimatized in the growth chamber for 3 weeks. And (G and H) after 3-month acclimatization, seedlings were hardened off and yielded significant shoot growth.

104

Figure 5.2. Effect of collecting time on in vitro seed germination, including (A) germination percentage, (B) contamination percentage, (C) T100, indicator for germination speed, and (D) number of true leaf per seedling, of mountain laurel hybrids ‘Elf’ x ‘Little Linda’ (solid line with closed circle) and ‘Red Bandit’ x ‘Minuet’ (dash line with open circle). Germination %, contamination %, and number of true leaf were measured 8 weeks after in vitro culture. T100 was calculated as the number of weeks reaching 100% of total germination. Data are present as mean of four replications ± SE. Different letters within each cross combination indicate that they are significantly different at P < 0.05.

105

CHAPTER 6

CONCLUSIONS

Mountain laurel is an outstanding flowering shrub and a promising ornamental that is native to the eastern U.S. It has gained popularity in the northeastern U.S., yet there is still considerable uncertainty about this plant in the southeastern U.S. To enhance the production and landscape performance of mountain laurel in the southeastern U.S., plants having adaptability to southeastern environmental conditions are needed. Traditional breeding of woody plants has been challenging for breeders due to the long breeding cycle; however it could be facilitated with the aid of modern technologies. In this project, we 1) explored and collected mountain laurel germplasm in the U.S.; 2) evaluated existing cultivars for container and landscape performance in the southeastern U.S.; 3) assessed genetic relationships among mountain laurel taxa using

ISSR markers; and 4) improved seed germination of mountain laurel hybrids using in vitro culture, which would effectively speed up the breeding and selection of mountain laurel for the southeastern U.S.

In our first study, there were 15 populations in the eastern U.S. that may provide desirable resources for a mountain laurel breeding program were identified, and collections were made from 10 of them. From this collection, 277 plants were obtained, of which 69, 186, and 22 were from cuttings, seeds, and plants, respectively. Twenty-one commercial cultivars were selected for collection because of their reported superior performance and diverse and favorable traits. A total of 197 cultivar plants is currently available, with 93 and 104 in 1- and 3-gal

106

containers, respectively. The exploration and documentation of mountain laurel wild populations will assist in plant collection and population genetics studies in the future by providing detailed geographic location and other information. The collection of a variety of mountain laurel wild plants and commercial cultivars will benefit a breeding program in the southeastern U.S. by providing promising genetic resources.

A 4-year trial was conducted on 21 mountain laurel cultivars in the southeastern U.S in our second study and cultivars displayed various responses to container and field trials. Although the 21 cultivars displayed differing growth indices in container production, all cultivars could yield sufficient growth the first year after being transplanted into 1-gal containers and consistently perform well thereafter in a container production environment. This performance indicated that growing mountain laurel for 1 year as 1-gal container plant should be feasible for southeastern producers. Fast-growing cultivars including ‘Bullseye’, ‘Carol’, ‘Forever Red’,

‘Freckles’, ‘Heart of Fire’, ‘Olympic Fire’, ‘Ostbo Red’, ‘Peppermint’, and ‘Pink Charm’ could be shifted up to 3-gal containers. Additionally, cultivars had a variety of morphological characteristics and fell into five phenotypic (PCA) groups based on shared morphologic traits when grown in a container production environment; principally on habit, foliage, and inflorescence traits. Results of this study should benefit plant breeders and commercial producers by providing regional cultivar information for the southeastern U.S. (and other USDA Zone 8-9 environments). ‘Ostbo Red’, Pristine’, and ‘Tinkerbell’ would be recommended to the southeastern producers in order to have both phenotypic variations and leaf spot tolerance.

Although most cultivars in this study had difficulty establishing in southeastern landscapes, fall planting could facilitate improved establishment. ‘Ostbo Red’, ‘Pristine’, and ‘Tinkerbell’

107

consistently excelled in field trials, indicating they are more adapted to environmental conditions observed in the southeastern U.S.

In the third study, ISSR markers efficiently determined genetic relationship among mountain laurel taxa and diversity of wild populations. Five points were illustrated by this study in relation to the breeding of mountain laurel. First, the robustness of ISSR for determining the genetic relationship of cultivars was demonstrated by the observed high genetic identities among related cultivars. Hence, our results would help breeding program to understand the pedigree of other cultivars and select parents from genetically diverse cultivars. Second, all cultivars were uniquely identified by their ISSR profiles, making it possible to distinguish morphologically similar cultivars and identify morphological differences of cultivars induced by environmental conditions using ISSR. Third, a low level of genetic diversity was observed among populations, indicated that collecting individuals from a large number of populations might not increase diversity for a breeding program. Fourth, although a relatively large proportion of diversity was attributed to within-population variation, the low actual diversity within populations illuminated the potential risk of losing genetic variation due to the small geographic size of most populations.

Fifth, the genetic divergence between cultivars and wild accessions due to the loss of genetic information of cultivars during breeding indicated that incorporating germplasm from wild plants to increase genetic variation and introduce desirable traits could be an approach to develop elite cultivars.

In vitro seed germination of mountain laurel was investigated in our fourth study.

Germinating mountain laurel seeds using traditional sowing method only yields low germination percentages and takes a long time. The subsequent seedling growth is generally slow, which takes approximately 1 year to reach a plant height of 10 cm. While using our in vitro seed

108

germination protocol, hybrid seeds collected 1 month before their full maturation were able to germinate in 1 month on WPM, which shortened the period from crossing to the seedling stage from 13-15 months to 6 months and improved the germination percentage from 30% to more than 90% compared with the traditional method. After two cycles of subcultures, seedlings were vigorous enough to be transplanted from petri dish to peat moss for acclimatization. Seedlings with both great shoot and root system were obtained in 3 months after acclimatization. The protocol developed in this study therefore enables a substantial increase in the number of hybrid seedlings and an acceleration of the breeding cycle by improving germination percentage, shortening germination time, and enhancing seedling growth. This protocol thus can be an approach to promote the breeding and selection of new mountain laurel cultivars for southeastern

U.S. landscapes.

109