Citation
Genetics and breeding of postharvest longevity in cut flowers of Gerbera X hybrida Hort.

Material Information

Title:
Genetics and breeding of postharvest longevity in cut flowers of Gerbera X hybrida Hort.
Creator:
Wernett, Heidi Carol, 1958-
Publication Date:
Language:
English
Physical Description:
vii, 115 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Bending ( jstor )
Breeding ( jstor )
Diameters ( jstor )
Flowers ( jstor )
Heritability ( jstor )
Inflorescences ( jstor )
Longevity ( jstor )
Statistical discrepancies ( jstor )
Stems ( jstor )
Vascular bundles ( jstor )
Cut flowers ( lcsh )
Dissertations, Academic -- Horticultural Science -- UF
Gerbera ( lcsh )
Horticultural Science thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 79-85).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Heidi Carol Wernett.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022805025 ( ALEPH )
22778850 ( OCLC )

Downloads

This item has the following downloads:


Full Text










GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF GERBERA X HYBRIDA HORT.













By

HEIDI CAROL WERNET7














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1990





























In memory of my father and mother,

Benjamin Carl Wernett and Evelyn Carolyn Wernett,

who inspired my appreciation for flowers.













ACKNOWLEDGMENTS


I wish to thank the members of my committee for their professional guidance in helping me to conduct this research. Dr. T. J. Sheehan, Dr. G.J. Wilfret, Dr. F.J. Marousky, Dr. P. M. Lyrene, and Dr. D.A. Knauft. Their patience is most appreciated.

Other debts of gratitude are acknowledged to Dr. R. C. Fluck from the Dept. of

Agricultural Engineering for his help with stem strength measurements, Dr. C.J. Wilcox from the Dept. of Dairy Science and Dr. T. L. White and his research assistant, Greg Powell, from the Dept. of Forestry for their critical help with data analysis, and Dr. F.G. Martin from IFAS Statistics whose suggestions were invaluable.

Personal thanks is also deeply extended to the Acuff family, Mark and Shari Wilson, and Terry J. Smith. Without their continued support, encouragement, sweat and toil, I could still be in the greenhouse; watering, transplanting, and taking measurements. The generous hospitality of my friends shall never be forgotten. In addition, a special "thank you" is required to show my appreciation to Terry for his kindness and assistance during the lengthy period of manuscript preparation.

There are many other individuals who advanced the completion of this research over the past seven years. I am sincerely grateful to all of them for their help.













11














TABLE OF CONTENTS


ACKNOWLEDGMENTS ............................................................................................... iii

ABSTRACT .................................................. ........ v.........

CHAPTERS

1 GENERAL INTRODUCTION................................................................... 1

2 LITERATURE REVIEW............................................................................ 3

3 PART I. VASE LIFE STUDIES ............................. ...... ............... 20

Introduction............... ........... ............ ......................................... 20
Materials and Methods............................................................. ............... 21
Plant Material..................................................................... ................. 21
Selection and Mating. ............................................................................... 23
Vase Life Evaluation....................................... ............................. 25
Quantitative Analysis..................................................... 26
Results........................... ...................... ..................................................... 28
Selection and Mating............................................................................... 28
Heritability ............................................................... 29
Senescence Patterns......................................... 31
Correlations between Vase Life and Other Traits...................... ........... 38
Discussion ...................................................................... 42
Conclusion..................................................................................................... 46

4 PART II. STEM STRENGTH STUDIES ....................................... ..... 48

Introduction.............................................................................................. 48
Materials and Methods.................................................................................. 48
Selection and Mating....................................................................... 48
Stem Strength Evaluation...................................................................... 49
Production .......................................................................................... .. 51
Quantitative Analysis............................................................................ 51
Results...................................................................................................... 53
Selection and Mating............................................................ 53
Heritability .................................................................................. 56
Correlations between Stem Strength and Other Traits......................... 58
Discussion ....................................................................................... 59
Conclusion..................................................................................................... 63

5 PART III. VASE LIFE X STEM STRENGTH STUDIES......................... 64

Introduction ................................................................................................. 64
Materials and Methods................................................. 65

iv






V

Selection and Classification.................................................................. 65
Anatomy Examination. ....................................................................... 65
Results.............................................................................. ......................... 66
Correlation Between Vase Life and Stem Strength .............................. 66
Comparison of Stem Anatomy Among Genotypes.............................. 67
Correlations Between Stem Anatomy and Postharvest Longevity.......... 72
Discussion ......................................................................................... 74
Conclusion. ....................... ......................... ............................................... 76

6 SU M M A RY ................................................................................................... 77

REFEREN CES........................... .......................... ...................................................... 79

APPENDICES

A VASE LIFE STUDIES ................................................................................. 86

B STEM STRENGTH STUDIES ..................................................................... 95

C VASE LIFE X STEM STRENGTH STUDIES ................................. 97

D HERITABILITY STUDIES FOR OTHER TRAITS IN GERBERA...........102

E CORRELATIONS AMONG OTHER TRAITS IN GERBERA......... 113 BIOGRAPHICAL SKETCH. ..................................................................................1....... 15














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF G E RA X HYBRIDA HORT.

By

Heidi Carol Wernett

May 1990

Chairman: Thomas J. Sheehan
Cochairman: Gary J. Wilfret
Major Department: Horticultural Science

Two components of postharvest longevity, vase life and stem strength, were studied in cut flowers of Gerbera X hybrida Hort.. A broad based source of germplasm was evaluated initially. Progeny means for vase life resulting from a topcross between 31 plants and 'Appleblossom' were used to select plants whose flowers had high vase life. Analysis of a 5 x 5 diallel cross was made to determine vase life heritability. No reciprocal effects were observed. Heritability ranged from 22 to 39 percent. Additive gene action possibly controls this character since broad sense and narrow sense heritability was nearly equal. Repeatability was moderately high (r = .57).

A senescence mode was recorded for each flower, i.e. bending or folding of the scape or wilting of the ligulae. "Flower" was defined as the composite inflorescence and its corresponding stem. After selection and mating, frequency of bending decreased; frequency of folding and wilting increased. Flowers that senesced due to wilting had the highest mean vase life before and after selection and mating. Significant correlations were found between vase life of flowers that senesced due to wilting and shorter stem length, thinner stems, and smaller infloresences.



vi






vii


A 7 x 7 diallel using plants whose flowers had high stem strength, measured using an Instron, was used to determine stem strength heritability. No reciprocal effects were observed. Heritability ranged from 28 to 53 percent. Additive gene action possibly controls this character since broad sense and narrow sense heritability was nearly equal. Repeatability was moderately high (r = .60). Significant correlations were found between stem strength and shorter stem length, thicker stems, and heavier infloresences.

Number of vascular bundles was compared among plants whose flowers exhibited

high or low levels of vase life and stem strength. Significant differences between high and low levels of vase life were observed for number of large bundles per 1.0 cm. scape circumference, measured 12 cm. below the peduncle; significant differences between high and low levels of stem strength were observed for number of small bundles per 1.0 cm. scape circumference.














CHAPTER 1
GENERAL INTRODUCTION


Gerbera X hybrida Hort. is a popular cut flower. Developed from hybrids of

interspecific crosses between Gerbera iamesonii Bolus and Gerbera viridifolia Sch., and possibly other species, it is a showy and handsome "flower" of the daisy family, Compositae. The distinctive inflorescence, generally measuring 8 14 cm. in diameter, has strap-shaped, ray floret ligulae that are bright shades of orange, yellow, or red, surrounding a yellow or black colored center of disc florets. This eye-catching "flower" is supported by a long, leafless, and upright stem, known as a scape.

The major postharvest problem in cut flowers of gerbera relates to the length of time until senescence occurs; when the ligulae wilt or when the stem no longer remains upright. Ideally, postharvest longevity in gerbera should be two weeks or longer. Unfortunately for the consumer, vase life is often less because the stem ceases to remain upright.

A critical attribute of cut flowers is postharvest longevity; therefore, research to

improve the lasting quality of gerbera is important. Postharvest longevity of cut flowers, including gerbera, can be extended by postharvest treatments but the extent to which these treatments can improve lasting quality may be limited by the plant genome. At present, evaluation for postharvest longevity is not routinely done in cut flower breeding programs. If gerbera cultivars with superior postharvest longevity are to be developed, then it is necessary to identify characteristics which are suitable for selection and breeding.

The main purpose of this research was to evaluate the potential of plant breeding as a method to improve postharvest longevity in gerbera, using a broad based source of germplasm. Specifically, objectives of this study were to estimate heritability for vase life and stem strength by diallel analysis, to observe changes in frequency of senescence


1






2


patterns due to selection and mating, to determine correlations between vase life, stem strength, and other traits of flower and stem morphology and growth, and to examine stem anatomy of genotypes that differ in vase life and stem strength.













CHAPTER 2
LITERATURE REVIEW


Researchers have investigated and discussed postharvest longevity in cut flowers including Gerbera X ykbrida Hort.:

The length of lasting quality in itself is not the aim of postharvest longevity
but the satisfaction of the consumer. Premature senescence due to abnormal
causes will leave the consumer with a sense of frustration. (Buys, 1978,
p. 256).

"Bent neck" in cut roses is a loss of pedicel rigidity which prematurely terminates the vase life of the flower (Burdett, 1970). Similar to bent neck, "knicking" (Wilberg, 1973; Buys, 1978), "folding" (De Jong, 1978a), "neck droop" (Zieslin et al., 1978), or "stem break" (Meeteren, 1978a) are terms used to describe the sudden bending of the stem in cut gerberas. Premature senescence in cut flowers contrasts with "natural" senescence. In gerbera, this natural phenomenon is identified as "wilting" (De Jong, 1978a). "Wilting" was described as the condition that occurs when the ligulae of an inflorescence on an upright stem have visibly lost their tugidity.

Vase life is often used as an indicator of postharvest longevity in cut flowers,

including gerbera. Vase life is determined by the number of days from harvest until flower senescence, whether or not senescence is considered premature. Sytsema (1975) specified six important factors which may affect the measurement of vase life in cut flowers:

1. Flower conditioning. Harvested flowers allowed to regain turgidity following
storage and/or transport can withstand adverse room conditions better than
unconditioned flowers.

2. Temperature. Vase life is usually shorter as temperature increases. Small
deviations from 200 C may be optimum.



3






4


3. Relative air humidity (RH) and air circulation speed. To reduce excessive
transpiration, RH should be at least 50% and air circulation speed should be
very low.

4. Light. Low light intensity is satisfactory, but total darkness may be less than
ideal.

5. Ethylene and air exchange. Good ventilation is the best method to avoid any
harmful effects due to ethylene concentration.

6. Use of bactericides. Bacterial contamination will interfere with reliable vase
life determinations, therefore a bactericide is recommended. In all cases, exact room conditions and experimental procedures should be recorded as a basis for
experimental comparison.

Reid and Kofranek (1981) also summarized some recommendations for standardizing vase life evaluations:

1. Use of vase solution control. Distilled water or deionized water treated for
removal of microorganisms and colloidal materials should be used. Flowers
should be placed individually in sterile containers.

2. Temperature. A good standard is 200 20 C.

3. Light. Light intensity of 600 mW/cm2 with a 12 hour diurnal cycle is practical
for simulating home conditions.

4. Relative humidity. A good standard is 60-70%.

5. Air circulation. One air exchange every two hours and wind velocity should
not exceed 0.5m/sec. Stages of flower maturity at time of harvest should be
well documented.


Historically, most postharvest research on cut flowers has focused on environmental, metabolic, and to.a lesser extent, anatomical factors that affect lasting quality. Environmental treatments have extended the vase life of cut flowers and/or reduced the incidence of premature senescence.

As early as 1962, it was suggested that the use of chemicals reduced wilting in plants. Experiments with strawberry plants showed applications of 8-hydroxyquinoline sulfate (8HQS) resulted in stomatal closing even when conditions were favorable for opening (Stoddard and Miller, 1962). Cut flowers of the rose 'Forever Yours' were reported to last






5


twice as long in an aqueous solution containing 200 ppm. of 8-HQS, an anti-bacterial agent, than in sterile water (Burdett, 1970).

By the late 1960s, many new postharvest methods and chemical additives were being tested as a means to extend vase life of major cut flower crops. Heide and Oydvin (1969) confirmed cytokinins used as a postharvest dip following storage were beneficial in prolonging vase life of cut carnations. As concentrations of cytokinin 6-benzylamino purine (BAP) increased, immersion times should decrease; otherwise, detrimental effects would result. Application of cytokinin to cut roses was also reported to increase vase life (Mayak and Halevy, 1970; Mayak et al., 1972).
Kofranek and Paul (1975) confirmed adding silver to preservative solutions was effective in extending postharvest longevity in chrysanthemums and carnations, but not gladiolus or gerbera. Mayak and Dilley (1976) found vase life of cut carnations was enhanced by solutions containing kinetin and sucrose compared with solutions containing only kinetin. Paulin and Muloway (1979) concluded that cytokinins are best used as a pretreatment, followed by a glucidic solution to increase the vase life of cut flowers.
Swart (1981) reported vase life of Lilium 'Enchantment' was improved by using a pretreatment of silver thiosulphate (STS). Sytsema (1981) found STS generally more effective in extending vase life of standard carnations than spray carnations. Kofranek and Halevy (1981) suggested a quaternary ammonium product, Physan, could be substituted for silver nitrate (AgNO3) as a chemical pulsing agent in chrysanthemums. Fujino et al. (1981) reported that use of aminooxyacetic acid (AOA) as a vase solution additive extended postharvest longevity of carnations. It was also shown that pulsing carnations with a higher concentration of AOA was effective. Staden and Beehuizen (1986) developed a pretreatment formula containing gibberellic acid (GA3), kinetin, daminozide, AOA, and detergent (Triton X-100) to be another chemical alternative to STS for extending the vase life of standard and spray carnations. Reddy (1988) reported solutions containing cobalt salts increased vase life of cut roses.






6


Experimenting with chemical additives to extend the vase life of minor cut flower crops was gaining momentum by the 1980s. Paull et al. (1981) reported that use of the preservative Floralife delayed flower wilting and leaf blackening in protea. Stimart and Brown (1982) demonstrated optimal vase life for cut flowers of zinnia could be obtained using a holding solution containing 200 ppm. of 8-hydroxyquinoline citrate (8-HQC) and 1% sucrose. Positive effects of floral preservative on postharvest quality of gypsophila were initially documented by Marousky and Nanney (1972). Tandler et al. (1986) found quaternary ammonium compounds were also effective in prolonging the vase life in the cut flower gypsophila. Postharvest treatments of STS or sugar plus a bactericide were reported to double the vase life of gypsophila (Barendse, 1986). Postharvest longevity was shown to increase in cut flowers of calendula, zinnia, and snapdragon if pulsing treatments of AgNO3 were utilized. Sweet pea and delphinium lasted longer if STS pulsing treatments were given (Awad et al., 1986). Sytsema (1986) reported STS prolonged vase life in cut freesia, but if STS was used in combination with cytokinins, vase life would be further enhanced. Kalkman (1986) stated cut flower preservatives would increase postharvest longevity in astilbe. Leeuwen (1986) proposed a combination of growth regulators and STS may be the most effective in prolonging vase life in Euphorbia fulgens. Downs and Reihana (1987) found preservative solutions enhance vase life of nerine cut flowers.

Research specifically dealing with postharvest treatments to increase lasting quality in cut flowers of Gerbera jamesonii has been reported. Waters (1964) reported treatments with flower preservative Everbloom increased vase life in gerberas 3-5 days. Kohl (1968) concluded a floral preservative should be used to increase vase life in gerbera, and that stems should be recut just prior to use. Marousky (1975) cautioned that vase solutions using fluoridated water can reduce postharvest longevity in gerbera. Meeteren (1978a) suggested the susceptibility to stem break in some gerbera cultivars could be prevented by using bactericides in vase solutions. Nowak (1981) reported vase life of gerbera may be






7


prolonged if a conditioning solution containing AgNO3, 8-HQC, and sucrose is applied for 20 hours before cold storage. A pretreatment application of STS was reported to extend the vase life of gerbera cut flowers later exposured to high concentrations of exogenous ethylene. Also, application of an ethoxy analog of rhizobitoxine (AVG) was reported to reduce endogenous ethylene production resulting in a slight increase in gerbera vase life, especially if used together with a flower preservative (Nowak and Plich, 1981). AbdelKader and Rogers, (1986) concluded that 8-HQS was preferable to AgNO3 as a component of flower preservatives used to enhance vase life of gerbera. Combining one of these anti-microbial agents with sucrose was shown to decrease the incidence of stem break. Amariutei et al. (1986) found pulsing treatments of sucrose, 8-HQS, and AgNO3 increased vase life of gerbera 'Symphonie' and 'Richard.'
In addition to chemical treatments, other postharvest handling measures have been investigated as a means to increase lasting quality in cut flowers. Carpenter and Rasmussen (1974) proposed fewer leaves remaining on roses could reduce the percentage of flowers developing bent neck. Dilley and Carpenter (1975) emphasized chemicals could not completely overcome premature senescence in cut flowers. Ferreira and Swardt (1981 a) reported defoliated cut flowers of the rose 'Sonia' lasted longer in deionized water than in flower preservative. Woltering (1986) concluded defoliated cut roses would be less likely to develop bent neck if placed in a vase solution containing a flower preservative or bactericide. Barendse (1986) reported optimum vase life in gypsophila could be achieved if the lower 10 cm. of the stems were defoliated. Meeteren (1978a) suggested the vase life of gerbera could be extended if the solid basal portion of the stem was removed, permitting water to enter the cavity of the stem. Later Nowak and Plich (1981) confirmed vase life of gerbera cut flowers could be enhanced if the basal portion of the stem was removed.
Metabolic causes for senescence in cut flowers have been discussed voluminously

during the past 40 years. Siegelman (1952) asserted that since a decrease in respiration rate reduces postharvest longevity in some fruits and vegetables, a similar decrease in






8


respiration rate could be the cause for senescence in cut flowers. The answer to the question, "Does respiration rate affect cut flower senescence?" remains elusive.

Siegelman (1952) demonstrated in experiments with roses and gardenias that when respiration rate was reduced, vase life increased. MacLean and Dedolph (1962) observed application of N6-benzylaminopurine (Verdan) reduced respiration rate in carnations and chrysanthemums and increased vase life. Kuc and Workman (1964) concluded a direct relationship existed between respiration rate and postharvest longevity in cut flowers because the respiration rate was three to four times greater in carnations than in chrysanthemums and postharvest longevity was much lower in carnations than chrysanthemums. Coorts, et al. (1965) reported that respiration rates increased in cut roses after being treated with a flower preservative. Treated roses lasted longer than those which were untreated. Gilbart and Dedolph (1965) treated cut roses with N6-benzyladenine (N6BA). They observed the respiration rate generally increased in petals and decreased in leaves following treatment. Ballantyne (1966) asserted variation in the respiration rate of daffodil cut flowers was more likely a result of senescence than a cause for senescence. Heide and 0ydvin (1969) suggested decreasing respiration rate in carnations will not delay senescence. Larsen and Frolich (1969) studied the relationship between respiration rate and water uptake in cut flowers. They observed that water uptake and respiration decreased simultaneously in carnation 'Red Sim.' Coorts (1973) reviewed factors contributing to metabolic changes in cut flowers which affect senescence. He concluded senescence can be delayed by using respiratory inhibitors and controlling hydrogen ion activity. It was proposed that if the pH of vase solutions was maintained between 3.0

5.0, vase life would be extended. It was also suggested mitochondrial activity may be linked to respiration rates of senescing cut flowers. Mayak and Halevy (1980) reviewed the subject of flower senescence and suggested that petal senescence would be an ideal system to study. Ferreira and Swardt (1981b) reported they found no correlation between respiration rate and vase life of cut roses. Changes in respiration rate depended on the






9


stage of senescence. Amariutei et al. (1986) observed higher respiration rates in gerbera cut flowers that were treated with a pulsing solution.

The role of plant hormones, particularly ethylene, has also been a source of

investigation for understanding cut flower senescence. Mayak and Halevy (1970) reported exogenous application of N6-BA can delay senescence in rose petals. Mayak et al. (1972) demonstrated exogenous applications of cytokinins increase postharvest longevity in roses whereas ethylene or abscisic acid (ABA) was shown to decrease postharvest longevity. They also suggested the presence of ethylene may trigger the production of ABA. Rogers (1973) reviewed the effects of ethylene synthesis and other plant hormones on cut flowers and concluded that anti-senescence factors must be applied to attain maximum vase life in cut flowers.

Nichols (1981) summarized results that indicated ethylene biosynthesis plays an important role in flower senescence, but the magnitude of its effect depends upon plant genera. Nowak (1981) stated a decrease in vase life of cut gerbera flowers may be caused by ethylene. Wulster et al. (1982) discussed the possibility of auxins increasing ethylene biosynthesis, thereby inducing senescence in carnation petals. Nichols (1982) reviewed the effects of growth regulators on cut flower senescence and concluded inhibiting endogenous production of ethylene will increase vase life in carnation, iris, daffodil, and chrysanthemum. Bufler (1986) stressed that studies should include investigating factors which control response to ethylene production, not just a measure of ethylene production. Reddy (1988) suggested cobalt salts may inhibit ethylene synthesis in cut roses.

A major factor considered to influence senescence in cut flowers, including gerbera, is the balance between water uptake and transpiration. Marousky (1969) verified reducing moisture stress by increasing water absorption will improve postharvest longevity in roses. Burdett (1970) proposed water loss as one of the causes of bent neck in cut roses. He specified water deficits resulting from transpiration are probably of small importance compared to an impairment in the water conducting system. Carpenter and Rasmussen






10


(1973) suggested transpiration rates have a close relationship with water uptake. Leaf area of cut roses was found to be the factor most closely associated with water uptake in both light and dark. Carpenter and Rasmussen (1974) concluded the number of stomates on the leaves and stems of cut flowers affects the rates of water uptake and loss. They observed when water loss from transpiration exceeds water uptake, vase life is reduced. It was proposed that by reducing transpiration, bent neck in roses would be reduced.

Mayak and Halevy (1974) found kinetin delayed wilting in cut roses. It was observed that water uptake increased, though transpiration also increased, due to stomatal opening. Bravdo et al. (1974) studied vase life in gladiolus. They proposed water uptake can be improved by increasing osmotic concentration of florets and leaves with the addition of sucrose to vase solutions. Marousky and Woltz (1975) also examined vase life of gladiolus. Though the addition of 8-hydroxyquinoline citrate plus sucrose was found to improve water uptake, the potential for fluoride toxicity was also increased depending on the concentration of fluoride in the water. They also reported that gerberas were highly sensitive to low levels of fluoride, especially with the addition of 8-hydroxyquinoline citrate plus sucrose to the vase solution which increased water uptake. Chrysanthemums and snapdragons did not exhibit this degree of sensitivity. Halevy (1976) stated water stress is the most common reason for decreased vase life in cut flowers. He asserted the basis for water imbalance is a decrease in water potential, water uptake, water loss, and water conductivity. Zieslin et al. (1978) postulated bent neck in cut roses occurs due to water stress conditions such as increased transpiration rates and decreased water uptake rates. Reddy (1988) suggested cobalt may increase vase life of cut roses by causing stomates to partially close, thereby reducing transpiration, yet maintaining water uptake.

A major reason cited for poor vase life in gerbera is the inability of the cut flowers to imbibe sufficient water (Kohl, 1968). It was suggested as gerbera flowers age, water holding capacity of the petals decreases (Meeteren, 1978b). As water content decreased in gerbera petals over time, ion leakage increased (Meeteren, 1979). It was proposed that by






11


increasing pressure potential in gerbera petal cells, ion leakage will decrease, resulting in longer vase life (Meeteren, 1980).

Researchers have also focused on the presence of vessel occlusion in the stem of cut flowers as a cause for early senescence. Two types of occlusion have been proposed. Occlusion due to microbes is considered to be a primary cause of bent neck in roses or stem break in gerberas. The use of a bactericide in vase solutions is a common practice in postharvest handling of cut flowers. The second type of occlusion is regarded as a result of physiological plugging.

Durkin and Kuc (1966) postulated a vascular block resulting from harvesting injury was the primary cause for premature senescence in cut roses. Marousky (1969) concurred vascular blockage was responsible for decreasing the vase life of cut roses. Burdett (1970) showed bent neck in roses coincided with the appearance of material which plugged xylem vessels. He postulated two possibilities for plugging: growth of microorganisms or a gum deposition which could be the result of pectin degradation products. It was also suggested when water uptake is deficient, sufficiently lignified stem tissue could prevent bent neck in cut roses. After eliminating the presence of microorganisms, Marousky (1972) demonstrated that in addition to moisture deficiency, a blocking mechanism reduces postharvest longevity in cut flowers. Parups and Molnar (1972) investigated the nature of vascular blockage in cut roses histochemically. They reported evidence of carbohydrates with sulfate, carboxyl, or phosphoryl groups, pectin-, lipid-, or other protein-like compounds, and some enzymes. They found no evidence of tannins, lignin, and callose in blocked xylem vessels. Rogers (1973) reviewed the literature concerning effects of physiological or microbial induced stem plugging but concluded further study was necessary to determine the cause of physiological plugging. He also suggested microbial plugging may have significance only with flower genera that typically last longer since it takes time for microbial populations to develop.






12


Carpenter and Rasmussen (1973) investigated the possibility of plugs occurring during light or dark periods that reduced water uptake rates. They observed no additional tissue degradation under daylight. Rasmussen and Carpenter (1974) used scanning electron microscopy to observe vascular occlusions in cut roses. They found vascular blockage affects vase life after the cut flower is physiologically incapable of maintaining adequate water balance; i.e. water loss exceeds water uptake. Mayak et al. (1974) postulated transpiration plays a more significant role in wilting of cut roses than vascular blockage.
Lineberger and Steponkus (1976) observed two types of vascular occlusion in cut roses. Microbial occlusions were found in the lower portion of the stem, and gum deposition was located in the stem above the solution level. Parups and Voisey (1976) reported the resistance to bending in cut roses is related to lignin content. It was concluded bent neck will occur if water stress occurs and stem lignification is insufficient. Zieslin et al. (1978) stated resistance to bending in cut roses depends particularly on secondary thickening of the vascular system and stem lignification. Zagory and Reid (1986) studied the role of microorganisms in reducing vase life of carnation 'Improved White Sim.' Only three of 25 microorganisms isolated from vase solutions reduced vase life. They suggested that ethylene-producing bacteria may be a possible factor in reducing vase life. De Witte and Doom (1988) postulated exogenous concentration of pseudomonas bacteria or Alcaligenes faecalis may account for vascular blockage in roses after three or more days, but endogenous bacteria in stems or air emboli may cause vascular blockage earlier. They did not find evidence of pectolytic breakdown in xylem cell walls. Dixon et al. (1988) indicated vase life of cut roses may be proportional to the loss of water conducting capacity caused by disfunctional xylem tissue. They proposed vase life may be lower in flowers having a greater proportion of large vascular bundles than small vascular bundles because larger bundles become disfunctional earlier than smaller bundles. Dixon and Peterson (1989) concluded physical vascular blockage in the stem of cut roses initially decreased water uptake, but xylem disfunction induced by water stress reduced vase life over time.






13


Two pathways for water uptake in gerbera were proposed: a direct path through xylem vessels and an indirect path through the stem cavity. It was suggested stem break occurs when water uptake is inhibited by bacterial growth (Meeteren, 1978a).

Some researchers have studied stem anatomy in an attempt to understand the causes for reduced postharvest longevity in gerbera. Reiman-Philip, as cited by Wilberg (1973), noted a larger proportion of large vascular bundles to small vascular bundles as a possible factor contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that such a relationship existed in only extreme cases. Steinitz, as cited by Marousky (1986) observed increases in phloem cell wall thickening and lignification in gerbera scapes when flowers were placed in a sucrose solution. In a comprehensive study, Marousky (1986) measured the size and number of vascular bundles in two gerberas, 'Tropic Gold' and 'Appleblossom.' He concluded 'Appleblossom' exhibited more resistance to stem bending than 'Tropic Gold,' partially because of various anatomical features such as fewer small vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference, and a greater percentage of dry weight per unit of scape length. It was emphasized these factors must be considered concurrently with variation in moisture stress to understand the cause of stem breakage.

Dubuc-Lebreux and Vieth (1985) also studied the histology of the gerbera stem. They postulated stem bending is linked to deficiencies in supportive elements in the stem. It was concluded that sensitivity to stem breakage depends on the degree of maturity of the stem approximately 10 cm. below the flower head at time of harvest. Marousky (1986) also noted lignification was greater in the lower portion of the stem compared to the upper portion of the stem where stem breakage is most likely to occur.

Despite postharvest treatments favorably affecting lasting quality in cut flowers, including gerbera, reliable expectations of vase life have not been achieved. Though metabolic and anatomical research has provided some clues to the cause of senescence in cut flowers, another approach for learning how to improve postharvest longevity exists.





14


The genetic basis for variation in postharvest longevity in cut flowers has been investigated; genotypic differences among plants examined, characteristics for selection proposed, and heritability estimated.

Mayak and Halevy (1970) observed differences in endogenous cytokinin levels between roses 'Lovita' and 'Golden Wave.' Cytokinin levels were higher in 'Lovita,' which has a longer vase life than 'Golden Wave.' Marousky (1973) suggested selecting chrysanthemums for succulent stems that readily translocate water. Mayak et al. (1974) reported transpiration rates were higher in rose cultivars with short vase life. They concluded these cultivars had less ability to close their stomates under water stress than cultivars with long vase life. Zieslin et al. (1978) studied the sensitivity of rose cultivars to neck droop. They concluded several factors contribute to variation in premature senescence in roses: transpiration rate, initial water uptake rates, and different organs of a flower competing for water at different rates. It was also suggested the effect of these factors may depend on the structural strength of the stem due to lignification or other anatomical factors. Paull and Criley (1981) observed clonal differences in vase life among protea cultivars. Ferreira and Swardt (1981b) concluded the vase life of rose cultivars varies due to genetic differences. They proposed three factors are genetically controlled: ability to store carbohydrates, ability to utilize an exogenous supply of sucrose, and susceptibility to bent neck. Stimart and Brown (1982) observed differences in vase life among zinnia cultivars when held in solutions containing 8-hydroxyquinoline citrate and sucrose.

Selection and breeding for postharvest longevity in tulips has been extensively

investigated. Twenty years ago, it was declared that by knowing the cultivar name of a tulip, its vase life could be predicted (Benschop and De Hertogh, 1969). This implies that genetics plays a significant role in determining the vase life of tulips.

Eijk and Eikelboom (1976) studied the possibility for selecting for postharvest longevity in tulips. Cut flower tulip vase life was described by three senescence characteristics, beginning with the number of days from flowering to start of tepal






15


discoloration, 50% of tepal discoloration, and perianth drop. All three characteristics were considered effective determinations of vase life, but perianth drop was easiest to measure. Also, evidence was reported that discoloration may occur earlier in some genotypes depending on the pigment composition of their flowers, but that this does not necessarily lead to earlier perianth drop. It was observed that some flowers with much carotenoid, delphinidin, or cyanidin and little pelargonidin discolor earlier than flowers with little carotenoid and much pelargonidin. It was concluded, however, that in a breeding program to improve postharvest longevity, all characteristics of vase life must be considered.

Eijk et al. (1977) confirmed earlier observations that cut flower vase life in tulips could be selected by evaluating flowers remaining attached to the plant, except in cases where response to flower preservatives is being evaluated. They also found vase life of field grown tulips was less predictable than vase life of greenhouse grown tulips. Therefore, it was suggested that field trials should be utilized only for initial screening. Several correlations were reported: (1) an increase in plant height during the growing period was not correlated to vase life; (2) an increase in growth of the last internode (stem elongation) did not appear correlated to vase life; and (3) vase life and water uptake were significantly correlated.

Eijk and Eikelboom (1980) demonstrated, using combining ability analysis, that

phenotypic observation was effective in predicting the results of crossing a set of parents in order to improve postharvest longevity in tulips. It was postulated additive gene action controls three senescence characteristics: start of discoloration, 50% discoloration, and perianth drop.

Eijk and Eikelboom (1986) investigated the influence of temperature on selecting for vase life in tulips. Variation in vase life for some genotypes was much greater depending on the temperature during the evaluation period. They recommended that selection at 170 C. provides an average response for each genotype; however, final selection should include screening at temperature extremes.






16


The use of tetraploids was suggested as another means to improve postharvest

longevity in tulips. Eijk and Eikelboom (1986) also reported that tetraploids have been created as a result of applying Initrous oxide (N20) to diploid cultivars. Progeny from tetraploid crosses showed improvement in vase life compared to their parents.

Improving postharvest longevity in cut flowers of Gerbera jamesonii by breeding and selection has been considered possible for many years. Smith and Nelson (1967) noted differences in vase life among cut flowers of gerbera. They suggested selection and breeding could minimize this variation. Kohl (1968) proposed selecting cultivars with an increased ability to uptake water and that cultivars having structural stem strength are requisite for maintaining the popularity of gerbera as a cut flower.

Wilberg (1973) observed differences in the frequency of stem bending among gerbera cultivars, and noted the need for breeding stems that remain upright in gerbera cut flowers. Three factors were identified which contributed to stem strength: the ratio of dry weight/cm. in the stem section prone to bending should be greater than the rest of the stem; stem elongation of harvested flowers should be small; and water content in the stem section prone to bending should be low. It was suggested all three factors should be included as part of a selection program. It was also noted thick stems were not necessarily more resistant to bending.

Meeteren (1978a) reported stem break in gerbera was greater during the summer than winter months. Barigozzi and Quagliotti (1978) noted tetraploids appeared to have stronger stems than diploids. These researchers were unable to find a relationship between flower color and vase life. Serini and De Leo (1978) supported this finding and reported no correlation between vase life and stem length, inflorescence diameter, and number of ligulae. They showed, however, vase life may be increased by a higher proportion of dry substance to water in the stem or a greater number of small vessels in the stem. Vase life and flower yield were reported to be negatively correlated. Tesi (1978) also reported vase life and flower yield were negatively correlated.






17


De Jong (1978a) presented methods for rapidly identifying structural strength or turgor strength of gerbera stems. These components of stem strength were postulated to affect vase life and "stem fold," a premature senescence phenomenon.

Structural strength was measured using a protractor to record the curvature of stems after freshly harvested flowers were stored dry for 24-48 hours. Turgor strength was measured by comparing the curvature of the stem after dry storage to the curvature of the stem following recovery (after the stored stems were placed in water for 24 hours and allowed to regain turgidity). This difference was related to water uptake ability of the stems. Rigidity of turgid stems was measured using a specially designed instrument which recorded the force needed to deflect the stem a predetermined distance.

Using these methods, De Jong observed a high frequency of folding was found in flowers with weak or firm stems, although the lowest frequency of folding occurred in flowers with firm stems.

De Jong (1978b) suggested breeding for structurally strong stems as a means to

improve postharvest longevity in gerbera. He reasoned that stronger stems may not be as likely to fold, thereby overcoming the deleterious effects of microbial infection. Also, structurally stronger stems could provide some added support if a water deficit occurs. He concluded, however, turgor strength is of primary importance.

Meeteren (1978b) observed increased ion leakage in petals of cut flower gerberas

depended on the cultivar. Meeteren and Gelden (1980) found no correlation between petalcytokinin activities in gerbera cultivars. Meeteren (1981) suggested pressure potential of petals from recently harvested flowers might be a good indicator for vase life and a possible selection criterion in breeding programs to improve postharvest longevity in gerberas. Nowak and Plich (1981) observed vase life of cut gerberas increases when stems were shorter.

De Jong and Garretsen (1985) noted turgor strength and structural strength of gerbera stems were greater and stronger during the summer than winter months in Holland. No






18


relation between stem stiffness and lignin content was observed among 25 cultivars. They also suggested the use of tetraploids may increase stem strength because tetraploids usually have thicker and shorter stems. These researchers observed no differences in the degree of curvature after dry storage between tetraploids and diploids.

Dubuc-Lebreux and Vieth (1985) suggested a scheme selecting gerbera cultivars

whose stems are more fully differentiated at the region of stem break when flowers are at the harvesting stage of maturity. They also identified a cultivar in their breeding program with this characteristic (K-9-9). Amariutei et al. (1986) reported consistent differences in vase life between two gerberas; 'Symphonie' lasted two days longer than 'Richard' when treated with pulsing agents or when untreated.

Genetic analysis of vase life in the cut flower Gerbera jamesonii has been described. Serini and De Leo (1978) estimated narrow sense heritability of vase life within full-sib families (h2 = .67). Heritability for vase life between plants was lower (h2 = .17). They concluded selection should be based more on families than individual plants. Tesi (1978) concluded vase life is strongly influenced by environmental factors. Variation due to environment and genotype was calculated (Ve2 = 85.2; Vg2 = 14.8). Mean vase life was reported to be 12 3.0 days.

Harding et al. (1981) estimated heritability of vase life from a non-random sample population of gerbera genotypes which consisted of half-sib families and clonal parents. Since the half-sib families had previously been mated and selected for cut flower yield and preference, it was suggested estimates for narrow sense heritability were biased due to a reduction in genetic variability. Components of variance were used to estimate heritability. Narrow sense heritability (h2) was calculated for two successive generations of the half-sib population; 24 and 38 percent. Broad sense heritability (H2) was calculated for two successive generations of clonal parents; 36 and 46 percent. Since heritability was moderately low, it was concluded either intense selection or selection over a large number






19


of generations would be required to increase mean vase life. Mean vase longevity was reported to be between 10 and 14 days.
De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in gerberas using a diallel mating scheme involving 12 parents and their progenies. Three characteristics were examined: vase life (days to wilting or folding); percent folding; and stem curvature. General combining ability was significant for each characteristic. It was pointed out that a large error variance (Ve2) will probably result when calculating variance components for vase life unless the mode of senescence is distinguished, i.e. early or late folding vs. wilting. Reviewing the inter-relationships among the three characteristics examined, several conclusions were made: (1) folding results in shorter vase life than wilting, (2) higher curvature may increase the incidence of folding, thereby resulting in shorter vase life; (3) no relationship exists between curvature and percent late folding; and
(4) late folding is more difficult to select against than early folding.
De Jong (1986) suggested parent choice is a major factor to consider when designing a breeding program to improve postharvest longevity in cut flowers of gerbera. For the characteristic "days to wilt," 78% of the variation between progeny means could be attributed to the twelve parents selected. It was concluded the main difficulty which remains in breeding to improve postharvest longevity is large intraplant variation for both days to wilt and percent folding.













CHAPTER 3
PART I. VASE LIFE STUDIES

Introduction


Gerbera X hybrid Hort. is a popular cut flower, but its postharvest performance is often less than desirable. Ideally, postharvest longevity in gerbera should be two weeks or longer. Unfortunately for the consumer, vase life is usually much less. Postharvest treatments, i.e. floral preservatives, are used to enhance the lasting ability of gerbera, but, developing cultivars with genetically superior postharvest longevity may provide the consumer with a reliable expectation for postharvest quality. Therefore, research to evaluate the potential of plant breeding as a method to improve postharvest longevity in gerbera is important.

At present, several researchers have estimated heritability for vase life, which is defined as the length of time until the flower senesces. Serini and De Leo (1978) concluded selection should be based more on families than individual plants since their estimate of narrow sense heritability was higher for among full-sib families (h2 = .67) than among plants (h2 = .17). Tesi (1978) concluded vase life is strongly influenced by environmental factors. He observed only 15 % of the phenotypic variation in vase life was due to genotype. Harding et al. (1981) based their results on a nonrandom sample population of gerbera genotypes from their Davis population which consisted of half-sib families and clonal parents. They concluded that since narrow sense heritability (h2 = .24 and .38) and broad sense heritability (H2 = .36 and .46) were moderately low for two successive generations, either intense selection or selection over a large number of generations would be required to increase mean vase life.



20






21


De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in gerberas using a diallel mating scheme involving 12 parents and their progenies. They distinguished vase life by different modes of senescence; petal wilt and stem fold. General combining ability was significant for days to wilting or folding and percent folding. De Jong (1986) concluded the main difficulty which remains in breeding to improve postharvest longevity is large intraplant variation for both days to wilt and percent folding.
Plant breeders can benefit from knowing the relationship between characteristics which they are trying to improve. Although breeding to improve postharvest longevity in gerbera has been suggested as a viable possibility for many years (Smith and Nelson, 1967), few studies of gerbera have reported correlations between vase life and other traits.

Tesi (1978) showed a significant negative correlation between vase life and cut flower yield. Serini and De Leo (1978) found no correlation between vase life and stem length, inflorescence diameter, and number of ligulae. Nowak and Plich (1981) observed vase life of cut gerberas increased when stems were shorter.

The objectives of this research on gerbera, using a broad based source of germplasm, were to determine broad sense heritability and narrow sense heritability for vase life by diallel analysis, observe changes in frequency of senescence patterns due to selection and mating, and to determine correlations between vase life and other traits of flower and stem morphology and growth.

Materials and Methods

Plant Material
Germplasm was randomly collected from several sources. Tissue cultured plantlets obtained were European cultivars or selections from a commercial breeding program at Sunshine Carnations in Hobe Sound, Florida. A list of these cultivars is given in table 3-1.






22

Table 3-1. List of tissue cultured cultivars.

Field # Cultivar Description Field # Cultivar Description
84-1 Amethyst 84-10 SI-1
84-2 Peach 84-11 P15-14-0
84-3 Seashell 84-12 SC300
84-4 Appleblossom 84-13 SC400-8
84-5 Raspberry 84-14 35C404-OX
84-6 Aztec 84-15 P18-5 84-7 Mandarin 84-16 SB-24
84-8 (Polish line) 84-17 SC205-X
84-9 Tropic Lady 84-18 SC501


Seed populations were obtained from different seed companies. In addition, Dr. J. Hardingi provided another seed source from his breeding program. Dr. Harding's seed mixture was the result of eight generations of breeding, but selection for vase life had been discontinued after the fifth generation. A list of the seed populations is given in table 3-2.

Table 3-2. List of seed populations.

Field# Population Description Seed Source
83-1 Davis Population U.C. Davis Res. Prog. (U.S.A.) 83-2* Ahms' F-1 Strain Herbst Seed Co. (U.S.A.) 83-3 Mardigras F-1 Strain Ball Seed Co. (U.S.A.) 83-4 Duplex Mixture Ball Seed Co. (U.S.A.) 83-5 Jongenelen Strain Ball Seed Co. (U.S.A.) 83-6 Florist Strain Mix Park Seed Co. (U.S.A.) 83-7 Ramona Mixture Sluis & Groot Corp. (Holland) 83-8 No. 4 F-1 Mix Clause Seed Co. (France)
* Poor germination, no plants survived.

Plant material varied in flower color and morphology. The term "flower" will be used to describe a composite inflorescence subtended by a stem. Plants grown from four seed


1Dr. J. Harding, Dept. of Environmental Horticulture, U. C. Davis, California






23


populations (Jongenelen, Florists Strain, Ramona, and No. 4 F-1) resembled those Dutch cultivars that have inflorescences with broad ligulae and thick fleshy stems. These populations represented a wide spectrum of flower color. The Duplex Mixture was comprised largely of pinkish hues with spindly stems and narrow ligulae. The Mardigras F-1 Strain included doubles and crested inflorescence types as well as the single, daisytype form. These single, daisy-type inflorescences were mostly of deep red hues with narrow ligulae. The Davis population mainly exhibited inflorescences with narrow ligulae also representing a wide spectrum of flower color. Flowers from these latter two populations had stems of medium thickness. Selection and Mating
Initially, a base population of 953 plants was grown. Plants which did not produce at least one flower during a flowering period prior to vase life evaluation (May 5--June 12, 1984) were discarded. Concurrently, plants were screened for inflorescence type and stem length. Plants which did not produce single, daisy-type inflorescences or stem length of 45 cm or greater, when 1-2 rows of disc florets were open, were discarded. The remaining plants of the base population were then referred to as the "parental generation." One to six flowers per plant were evaluated. After the evaluation period, plants that produced less than three flowers were also discarded. The "residual parental generation" included 325 plants. Plant means were determined from data collected on the first three flowers evaluated per plant. Thirty-one plants (approx. 10% of the residual parental generation) with highest mean vase life (- 11.3)) and lowest coefficient of variation (C.V. < 25.0) were selected. To maintain genetic diversity, selection included plants from each seed population.
A top-cross mating scheme was utilized as a screening method to determine which of these parents had the longest vase life. 'Appleblossom' was used as the male donor in the top cross mating scheme because of its excellent vase life rating and low intraplant






24


variation. Twenty-eight plants per cross were grown to produce the top-cross generation. Plants which did not produce at least one flower during a flowering period prior to vase life evaluation (March 12--May 15, 1985) were discarded. One to three flowers per plant were evaluated. After the evaluation period, plants that produced less than three flowers were also discarded. Plant means were determined from data collected on three flowers per plant. Progeny means were determined for each cross from individual plant means. Finally, five plants (approx. 1.5% of the residual parental generation) with highest progeny mean vase life were selected. To maintain genetic diversity, selection included plants from four seed populations.

A 5 x 5 diallel cross mating scheme was utilized to estimate heritability of vase life. Twenty-eight plants per cross were grown to produce the diallel generation. Plants which did not produce at least one flower during a flowering period prior to vase life evaluation (May 23--August 6, 1987) were discarded. Plant means were determined from data collected on one to three flowers per plant.

Vase Life Evaluation

Flowers with 1-2 rows of disc florets open were harvested each evening for six

weeks. Stems were then uniformly cut 30 cm long. Flowers were randomly placed into sterilized glass bottles with one flower per bottle. Each bottle contained 100 ml. of deionized water, buffered to pH = 3.0-3.4 with a citrate-phosphate buffer. The depth of water in each bottle was 4 cm. Every other day, the bottles containing deionized water were replaced until senescence occurred. Evaluations were conducted in a temperature controlled room (200-210 C) with 24 hrs./day lighting provided by overhead fluorescent lamps. Light intensity was .26-.52 W/cm2 at flower height. Relative humidity was approximately 70%. These conditions were patterned after experiments conducted by De Jong (1978a and 1978b) and Harding (1981).






25


Flower senescence was classified into three modes based on the visual condition of the stem:

1. Bending. The stem gradually, but irreversibly, loses turgidity resulting in an arc. If allowed to persist, the stem eventually appears folded.

2. Folding. The stem suddenly bends resulting in an irreversible sharp angle.

3. Wilting. The stem remains rigid and upright until the ligulae wilt.


Vase life was measured by the number of days to flower senescence.

Vase Life = Senescence Date Harvest Date In addition to vase life, five variables relating to flower morphology and growth were measured on each flower:

1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm two diameter measurements were taken, i.e. length
and width, on flowers from the top cross and diallel cross generations.)
3. Inflorescence diameter
4. Disc diameter
5. Stem length (at time of senescence)

A variable, "Vgrowth," was created to describe the amount of stem elongation

observed from harvest until senescence. "Vgrowth" was measured by the difference between stem length at time of senescence and 30 centimeters.

Vgrowth = Senescence Stem Length 30.0 cm


Production

Plants were grown in 12.5 cm standard plastic pots on raised benches in a clear glass greenhouse at the University of Florida in Gainesville, Florida. Minimum night temperature was maintained at approximately 18' C. Day temperature was set at 30' C. A fan and pad cooling system was used to control the temperature. Shade cloth of 25% density covered the greenhouse since light intensity generally exceeds 65 W/cm2 per day in Florida. Sowing, transplanting, and flowering dates for each generation are recorded in table 3-3.






26



Table 3-3. Record of production dates. Generation Sowing Transplanting Flowering

Parental 12-2-83 2-18-84 3-15-84/5-5-84 Top Cross 10-12-84 1-12-85 2-23-85/3-12-85 Diallel 1-18-87 3-14-87 4-22-87/5-23-87



Ouantitative Analysis
Vase life data were initially analyzed according to the random model for Griffing

Method 3 (Griffing, 1956). This method describes a diallel mating design which includes reciprocal crosses but excludes selfs. No reciprocal differences were observed. Subsequently, analysis of variance for combining ability, using a general least squares diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of plant means.
Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was estimated from ratios of the following variances:


VA = Additive genetic variance
VG = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental) VA VG
h2 =VA H2 V Vp Vp
(Falconer, 1960)


Genotypic and phenotypic variances were determined from the following equations using thevariance components for general combining ability (a2gca), specific combining ability (O2sa), and error (G2e) which were calculated by the diallel analysis program developed by Schaffer and Usanis (1969):






27



VA = 402gca
VG = 4O2gca + 402ca
Vp = 402gca + 4O2sca + 42e
(Hallauer, 1981)

Thus, heritability (h2 and H2) was estimated from the formulae:

h2 = 42ca H2 42gca + 4O2sca
4O2gca + 4O2sca + H2e 402gca + 4o2sca + 42e


Predicted estimates of narrow sense heritability and broad sense heritability for n measurements per plant were also made. This required partitioning the environmental variance (VE), a component of phenotypic variance (Vp), into general environmental variance (VEg) and special environmental variance (VEs) in order to determine the phenotypic variance for each case (Vp(n)).


Vp = VG + VE
VE = VEg + VEs

VP(n) = VG + VEg + n VEs
(Falconer, 1960)

Special environmental variance (VEs) or within-individual variance (Falconer, 1960) for a single measurement per plant may be derived by the error variance component (02e) from a one-way analysis of variance (Falconer, 1960). Henderson's Method 3 (Henderson, 1953) for obtaining variance components was performed on vase life data using individual flowers as observations rather than plant means. The SAS procedure VARCOMP (SAS, 1982) was used to obtain the o2 or error MS. General environmental variance (VEg) or between-individual variance (Falconer, 1960) was calculated by subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated






28


in the diallel cross, from the error MS or a2e obtained by the combining ability analysis of variance.

Repeatability for vase life was determined from the following ratio of variances whereby n = 1:

VG + VEg
r VP(n)
(Falconer, 1960)

Results


Selection and Mating

Vase life means for 31 parents selected for the top-cross mating with 'Appleblossom' ranged from 11.3 to 16.0 days. Coefficients of variation ranged from 0.0 to 24.7. Progeny means resulting from crosses ranged from 6.1 to 10.7 days. Five of these parents considered to have the "best" vase life were selected for diallel mating based on data given in table 3-4. Progeny means resulting from diallel mating ranged from 8.7 to 14.3 days. Listing of this data is given in table 3-5. Following selection and breeding, the population mean for vase life increased by three days. Comparison of vase life data for the parental and diallel generations is shown in table 3-6.


Table 3-4. Vase life data for five parents.
(Parental Generation) (Top-cross Generation) Parent Mean Progeny Mean

83-1-77 15.3 8.6 83-4-69 12.0 9.6 83-5- 109 13.0 8.9 83-7-4 12.3 10.2 83-7-10 16.0 10.7






29

Table 3-5. 5 x 5 Diallel. Vase life data from diallel crosses. reciprocalss pooled) (Diallel Generation)
Cross Progeny Mean
83-1-77 x 83-4-69 10.3 83-1-77 x 83-5-109 8.7 83-1-77 x 83-7-4 11.8 83-1-77 x 83-7-10 11.2 83-4-69 x 83-5-109 9.1 83-4-69 x 83-7-4 11.9 83-4-69 x 83-7-10 10.6 83-5-109 x 83-7-4 11.4 83-5-109 x 83-7-10 12.5 83-7-4 x 83-7-10 14.3


Table 3-6 Summary of vase life for parental and diallel enerations.
No. of Coefficient of Generation Plants Mean Std. Dev. Variation

Parental 325 7.82 2.64 33.77 Diallel 248 11.20 3.14 28.08



Heritability

Combining ability analysis of variance using plant means was performed on vase life data from a 5 x 5 diallel. General combining ability effects were significant. Specific combining ability effects were non-significant. Results are summarized in table 3-7. Heritability was estimated by the ratio of genetic variance (VA or VG) to phenotypic variance (Vp) (Falconer, 1960). Variances were derived using variance components for general combining ability (o2gca), specific combining ability (O2sa), and error (02e) according to formulae published by Hallauer (1981). Narrow sense heritability (h2 = .279) and broad sense heritability (H2 = .281) estimates were approximately equal. This indicates non-additive genetic variance (VG VA) is negligble. Variances and heritability estimates for this population are given in table 3-8.






30



Table 3-7. 5 x 5 Diallel. Combining ability analysis of variance for vase life. Source of Variation df M.S. F-ratio

General combining ability 4 106.12 8.03 Specific combining ability 5 13.21 1.01 n.s. Error 238 13.08
* Significant at P< 0.05.


Table 3-8. 5 x 5 Diallel. Variances and heritability estimates for vase life.

VA VG Vp h2 H2 5.08 5.12 18.20 .28 .28

Estimates of heritability were based on an average of 1.96 measurements per plant. This value (n) resulted by evaluating 487 flowers from 248 plants. The error variance component (a2e) for a single measurement per plant was obtained by an analysis of variance using vase life data from individual flowers of the diallel generation. Also, this analysis indicated differences in vase life of flowers among crosses and among plants within crosses were highly significant. Results are summarized in table 3-9. Table 3-9. 5 x 5 Diallel. Analysis of variance for vase life. Source of Variation df M.S. F-ratio

Among crosses 9 115.02 5.17 * Among plants 238 22.26 2.25 * Within plants 239 9.88
* *Significant at P 0.01.


General environmental variance (VEg) and special environmental variance (VEs) were derived from vase life data using calculations described by Falconer (1960). A summary of variances for this population are given in table 3-10. Predicted estimates of narrow sense






31


heritability and broad sense heritability for vase life were then made for 1, 2, 3, 5, and ** measurements per plant by the ratio of genetic variance (VA or VG) to phenotypic variance (Vp(n)) (Falconer, 1960). Estimates ranged from 22 to 39 percent. These results are given in table 3-11. Repeatability (r = .57) for vase life was moderately high.


Table 3-10. 5 x 5 Diallel. Summary of variances for vase life. n = 1.96
Genotypic Variance Environmental Variance Phenotypic Variance

VA VG VE VEg VEs Vp(n) 5.08 5.12 13.08 8.04 9.88 18.20


Table 3-11. 5 x 5 Diallel. Predicted estimates of heritability for vase life.
Number of Measurements

Heritability 1 2 3 5 00 h2 .22 .28 .31 .34 .39 H2 .22 .28 .31 .34 .39 Senescence Patterns

A shift in the population mean for vase life was examined more closely. Distribution of vase life data on individual flowers from the parental and diallel generations was compared. In addition to a shift in the mean from 7.8 days to 11.2 days, the frequency of vase life did not appear normally distributed in both generations. Before selection there was a much higher frequency of flowers with low vase life than high vase life. After selection and mating, this trend was clearly reversed. These results are shown in figure 31.

Distribution of vase life data by senescence mode was also compared. In both

generations, vase life frequency appeared somewhat normally distributed. Following






32


selection and mating, vase life means for days to bending, folding, and wilting increased. The largest increase in mean vase life occurred in flowers that folded; 3.5 days. Smaller increases were observed in flowers that bent and wilted; .4 and 1.2 days, respectively. These results are shown in figures 3-2, 3-3, and 3-4.




15.0

12.5- PARENTAL GENERATION mean = 7.8 10.0 7.5

5.0

2.5

0.0

12.5- DIALLEL GENERATION

10.0I I mean = 11.2

7.5 5.0

2.5 0.0
0.0 1 3 5 7 9 11 13 15 17 19 21 23

Vase Life (Days) Figure 3-1. Distribution of vase life data on flowers from the parental and diallel generations.






33











12.5
PARENTAL GENERATION mean = 6.1


7.5 5.0


2.5


0.0


DIALLEL GENERATION mean = 6.4
7.5


5.0


2.5


1 3 5 7 9 11 13 15 17 19 21 23 Vase Life (Days) Figure 3-2. Distribution of vase life data for flowers that bent from parental and diallel generations.






34










12.5

PARENTAL GENERATION 10.0 mean = 8.2

7.5 5.0

2.5 0.0.

10.0
DIALLEL GENERATION
7.5 mean = 11.7

5.0 2.5


0.0
1 3 5 7 9 11 13 15 17 19 21 23 Vase Life (Days) Figure 3-3. Distribution of vase life data for flowers that folded from parental and diallel generations.






35











12.5


10.0
PARENTAL GENERATION
7.5 mean = 12.2


5.0 2.5


0.0


10.0 DIALLEL GENERATION mean = 13.4
7.5


5.0 2.5


0.0
3 5 7 9 11 13 15 17 19 21 23 Vase Life (Days) Figure 3-4. Distribution of vase life data for flowers that wilted from parental and diallel generations.






36


Separate analyses of variance were made using vase life data from the diallel

generation for days to bending, days to folding, and days to wilting. Variation within plants or error M.S. for vase life was less when data were grouped by senscence modes as when an analysis of variance was made on vase life of all flowers. Differences in vase life of flowers among crosses were highly significant for all senescence modes. Differences among plants within crosses were also significant for days to folding and wilting, but nonsignificant for days to bending. Results are summarized in tables 3-12, 3-13, and 3-14. Table 3-12. Analysis of variance for vase life (days to bending). Source of Variation df M.S. F-ratio

Among crosses 9 21.72 2.65 Among plants 70 8.24 1.36 n.s. Within plants 29 6.05
* *Significant at P< 0.01.

Table 3-13. Analysis of variance for vase life (days to folding). Source of Variation df M.S. F-ratio

Among crosses 9 52.28 3.76 * Among plants 124 14.42 3.16 * Within plants 46 4.56
* *Significant at P5 0.01.

Table 3-14. Analysis of variance for vase life (days to wilting). Source of Variation df M.S. F-ratio

Among crosses 9 28.18 3.21 * Among plants 120 8.77 1.49 Within plants 68 5.88
*Significant at P 0.05. *Significant at P 0.01.






37


Changes in the proportion of flowers that senesced due to bending, folding, and

wilting were also observed. Differences between senscence mode frequency in the parental and diallel generations are shown in figure 3-5. After selection and mating, the incidence of bending was reduced by an average of 41 percent. Folding and wilting increased by an average of 23 and 18 percent, respectively. In both generations, the frequency of wilting exceeded that of folding.


120

%bending
100- 0 % folding O % wilting
80
Frequency
(%)
60 40

20

0 1
parental diallel Generation

Figure 3-5. Distribution of senescence modes for parental and diallel generations.


The frequency of bending, folding, and wilting in progeny of the diallel generation was compared to vase life means. Among ten progenies, the proportion of bending generally decreased as vase life increased. With the exception of one cross, approximately 50 percent of the flowers wilted in progenies with vase life of 11 days or greater. Approximately 25 percent of the flowers wilted in progenies with vase life of less than 11 days. The proportion of folding did not appear related to progeny means for vase life. Distribution of this data is shown in figure 3-6.






38





120 % bending E % folding
100 0[ % wilting

80
Frequency


40 20

0
8.7 9.1 10.3 10.6 11.2 11.4 11.8 11.9 12.5 14.3
vase life progeny means

Figure 3-6. Distribution of senesence mode frequency in progeny of diallel generation. Correlations between Vase Life and Other Traits

Phenotypic correlations between vase life and five stem and inflorescence traits: stem length, stem diameter, inflorescence diameter, disc diameter, and post-harvest stem elongation (vgrowth), were determined for both parental and diallel generations. Correlation coefficients based on data from individual flowers were obtained using a SAS program for Pearson's product-moment correlation procedure (SAS, 1986).

Correlations determined from the parental generation demonstrated the linear

relationship between vase life and these traits in a random population before selection for improving vase life. Correlations determined from the diallel generation demonstrated the linear relationship between vase life and these traits after selection. Additionally, correlations were made between the vase life of flowers that bent, folded, or wilted and the five stem and inflorescence traits. These results are summarized in tables 3-15, 3-16, 3-17, 3-18, and 3-19.






39



Table 3-15. Phenotypic correlation coefficients between vase life and stem length before and after selection for vase life.
(Senescence Modes)
Generation N Total Bending Folding Wilting

Parental 1508 -.14** -.15 ** -.24* -.05

Diallel 487 -.14** -.01 -.04 -.27
* Significant at P< 0.01.


Table 3-16. Phenotypic correlation coefficients between vase life and stem diameter before and after selection for vase life.
(Senescence Modes)
Generation N Total Bending Folding Wilting

Parental 1508 .02 -.11 * -.08 .08

Diallel 487 -.11 .08 -.15 -.24 *
* Significant at P< 0.05. * Significant at P< 0.01.


Table 3-17. Phenotypic correlation coefficients between vase life and inflorescence diameter before and after selection for vase life.
(Senescence Modes)
Generation N Total Bending Folding Wilting

Parental 1508 -.11* -.10** -.33** -.29**

Diallel 487 -.05 .14 .00 -.23 *
* Significant at P< 0.01.


Table 3-18. Phenotypic correlation coefficients between vase life and disc diameter before and after selection for vase life.
(Senescence Modes)
Generation N Total Bending Folding Wilting

Parental 1508 -.05 -.12 -.26** .02

Diallel 487 .03 .02 .09 -.07
* Significant at P< 0.01.






40

Table 3-19. Phenotypic correlation coefficients between vase life and post-harvest stem elongation before and after selection for vase life.
(Senescence Modes)
Generation N Total Bending Folding Wilting Parental 1508 .22 * .14 * .20** .37 *

Diallel 487 .24** .23* .17* .13
* Significant at P5 0.05. * Significant at P< 0.01.


Vase life and stem length. Significant negative correlations were observed, based on data from the total number of flowers evaluated, before and after selection. Among flowers that bent, folded, or wilted in the parental generation, the negative correlations observed between vase life and stem length were significant only for bending and folding. In contrast, among flowers in the diallel generation, only the negative correlation observed between vase life and stem length for flowers that wilted was significant. Overall, it appeared from these correlations that vase life was highest when stem length was shorter.

Vase life and stem diameter. Based on data from the total number of flowers

evaluated, no correlation was observed before selection, however, a significant correlation was observed after selection. Among flowers that bent, folded, or wilted in the parental generation, negative correlations were observed between vase life and stem diameter for flowers that bent or folded; but, the correlation was significant only for flowers that bent. The positive correlation observed for flowers that wilted was non-significant. In contrast, among flowers in the diallel generation, negative correlations were observed between vase life and stem diameter for flowers that folded or wilted, but the correlation was significant only for flowers that wilted. The positive correlation observed for flowers that bent was non-significant. Overall, it appeared from these correlations that vase life was highest when stem diameter was smaller.

Vase life and inflorescence diameter. Negative correlations were observed, based on data from the total number of flowers evaluated, before and after selection; however, only the correlation before selection was significant. Among flowers that bent, folded, or wilted






41


in the parental generation, significant negative correlations were observed between vase life and inflorescence diameter. Among flowers that bent, folded, or wilted in the diallel generation, a significant negative correlation was observed between vase life and inflorescence diameter only for flowers that wilted. The positive correlation for flowers that bent was non-significant and no correlation was observed for flowers that folded. In both cases, before and after selection, the negative correlation observed between vase life and inflorescence diameter was highly significant for flowers that wilted. Thus, it appeared that vase life increased as inflorescence diameter decreased.

Vase life and disc diameter. No significant correlations were observed, based on data from the total number of flowers evaluated, before or after selection. The only significant correlations observed between vase life and disc diameter were among flowers from the parental generation that bent or folded. These highly significant correlations were negative. A negative correlation was also observed for flowers that wilted in the diallel generation, though this correlation was non-significant. In general, these correlations demonstrated a very weak negative relationship, if any, between vase life and disc diameter. The significant correlations that were observed in the parental generation for vase life of flowers that bent and folded seemed of little interest since selection would be made against these senescence modes in a breeding program to improve vase life.

Vase life and vgrowth. Significant positive correlations were observed, based on data from the toal number of flowers evaluated, before and after selection. Among flowers that bent, folded, or wilted in the parental generation, highly significant positive correlations were observed between vase life and vgrowth. Among flowers that bent, folded, or wilted in the diallel generation, positive correlations were also observed between vase life and vgrowth, however, correlations were significant for flowers that bent or folded. It appeared from these correlations that the positive relationship which was observed between vase life and post-harvest stem elongation prior to selection and mating weakened as vase life increased.






42



Discussion


Heritability estimates for a given character can vary based on the population of plants evaluated, selection intensity, mating design and environment. (Simmonds, 1979). Despite this, estimates of heritability for vase life determined from this experiment (h2 = .28 and H2 = .28) were within proximity of those determined from investigations by other researchers (Serini and De Leo, 1978; Tesi, 1978; and Harding et al., 1981). This reflects some consistency in the proportion of genetic variance to phenotypic variance for vase life in gerbera regardless of genetic diversity in populations sampled, breeding procedures, and environment.
The estimate of 28 percent for broad sense heritability (H2), based on 1.96

measurements per plant, is moderately low. Similarly, the estimate of 28 percent for narrow sense heritability (h2), also based on 1.96 measurements per plant, is moderately low. This estimate is between the range of prior estimates: 15 percent (Tesi, 1978); 17 percent (Serini and De Leo, 1978); and 0, 24, and 38 percent (Harding et al., 1981). It appears that genetic variation may be largely controlled by additive gene action since broad sense and narrow sense heritability were approximately equal. Therefore, in a fixed model experiment, progeny means obtained from a top-cross mating would be effective in determining parents with good combining ability for increasing vase life.

Falconer (1960) demonstrated a method to predict estimates of heritability for a specified number of measurements per experimental unit. This involved partitioning environmental variance (VE) into general environmental variance (VEg) and special environmental variance (VEs). Special environmental variance (VE), or within-plant variation, is the environmental variation for a single observation per experimental unit. The magnitude for special environmental variance (VE) is then divided by a specified number of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic variance for each special case. Ideally, if n = oo, then VEs will be reduced to zero, thereby






43


deleting a significant source of environmental variation. In that case, the highest possible estimate of heritability could be obtained for a given population.

Using this method, predicted estimates of heritability for vase life in gerbera ranged from 22 to 39 percent with n specified as 1, 2, 3, 5, and -c flowers per plant. Since these calculations assumed no change in genotypic effects, broad sense and narrow sense heritability estimates for each case were approximately equal. It is interesting to note that in spite of varying the magnitude of environmental variance, the range of estimates remained within proximity to those determined by other researchers (Serini and De Leo, 1978; Tesi, 1978; and Harding et al., 1981).

The repeatability estimate for vase life (r = .57) is moderately high, indicating that two to three flowers per plant is adequate for determining the average vase life per plant. Falconer (1960) recommends that further gain in accuracy by more than two measurements does not justify additional expense or time required to collect more data when repeatability is high. This was proven by comparing the relative increase in heritability from predicted estimates based on 1, 2, 3, 5, and measurements per plant. Between one and three measurements, heritability increased by nine percent, while beyond three measurements through infinity, the gain in heritability was only eight percent.

Thus far, this research has confirmed that improvement of postharvest longevity in gerbera can be obtained by selecting and mating plants with good combining ability for high vase life. The overall population mean for vase life resulting from a diallel cross among five plants selected, based on their combining ability with 'Appleblossom,' a cultivar with high vase life, yielded an increase of more than three days. Additional information has been gained, however, by classifying vase life determinations by three distinct modes of senescence: bending, folding, and wilting. The frequency of vase life days based on data from all flowers was not normally distributed before or after selection and mating. Yet, when frequency of vase life days was classified by senescence mode, distribution was normal. Moreover, after selection and mating, increases in mean vase life






44


for flowers that folded, bent, or wilted differed. For example, the increase in mean vase life for flowers that folded was much greater than for flowers that bent or wilted. Combining this evidence, it is suggested that vase life may be a composite character of at least three components, represented by each senescence mode.

Further evidence to support this suggestion is found by comparing the magnitude of the error variance component (a2e) or error MS from an analysis of variance based on the total number of vase life observations versus the error MS from individual analyses based on the number of vase life observations for each senescence mode. De Jong and Garretsen (1985) previously reported that if termination of vase life is not distinguished by stem collapse or petal wilt in an analysis of variance, a relativlely large error variance would result. In fact, the error MS based on the total number of vase life observations from this data was nearly double the arithmetic mean of the error MS from individual analyses based on flowers that bent, folded, and wilted (9.88 ya 5.50).

This study was designed to evaluate the potential of improving postharvest longevity in gerbera by selecting plants with high vase life. Plants were not selected based on the specific number of days to bending, folding, or wilting of their flowers. However, it does appear, that selecting plants based on the number of days to bending, folding, or wilting, rather than the composite character of vase life, may prove to be a useful approach to improving postharvest longevity in gerbera, since general combining ability for vase life of flowers was determined to be highly significant for all three senescence modes.

The incidence of bending, folding, or wilting, not only the vase life of these

senescence modes, is proposed to be another important aspect of postharvest longevity in gerbera. Before selection and mating to improve vase life, a greater proportion of flowers bent than wilted. Flowers that bent generally exhibited lower vase life than flowers that wilted. After selection and mating, a greater proportion of flowers wilted than bent. Additionally, the shift in proportion of these two modes that occurred was rather dramatic






45


after only one generation of selection. Therefore, it is postulated that the incidence of bending vs wilting may be a qualitatively based trait controlled by relatively few genes.

In studies on breeding for keeping quality in gerbera, De Jong (1986) distinguished senescence by three classes; early stem fold, late stem fold, and petal wilt. His definition for early stem fold was slightly different from that of bending in this study, but he reported that if selection were made against the phenomenon, its incidence could be reduced or eliminated fairly easily. He showed data for 59 progenies resulting from a diallel mating; four progenies did not exhibit early fold. After studying these results, it seems possible that the trait "early fold ys no early fold" could fit a genetic model involving only two genes with epistatic effects (15:1). Hence, it seems plausible to hypothesize that a similar model may apply to the trait "bending vs wilting." Among ten progenies, bending was not completely eliminated, but, in one case, its incidence was reduced to only five percent.
The incidence of stem folding may be distinct from the incidence of bending or
wilting. Before selection and mating to improve vase life, the frequency of folding was relatively low. After selection and mating, the frequency of folding nearly tripled. Accompanying this increase, vase life also increased. This created some speculation as to whether a higher proportion of folding is the result of longer lasting flowers. Yet when the proportion of folding for each progeny was compared to progeny means for vase life, this speculation could not be confirmed because some progeny with higher vase life showed a smaller proportion of folding than some progeny with lower vase life. Instead, the most striking observation of this comparison, among ten progenies, was that the proportion of folding could be grouped into four classes of approximately 20, 30, 40, and 65 percent. This gives cause to wonder whether data from more progenies might yield a sufficient number of classes to identify a genetic model that would indicate folding is also qualitatively inherited. If so, it seems likely that the incidence of folding is controlled by at least several genes.






46


Correlations between vase life and other morphological traits were particularly interesting because a broad based germplasm was examined. The wide variation in morphological phenotypes permitted an extensive comparison. Correlation coefficients were generally low. Despite this, they were often significant due to the large number of flowers evaluated.

Since it has been discussed that eliminating senescence due to bending and folding among flowers may be possible, correlations between the vase life of flowers that wilted and other traits are probably the most useful to a breeder. Significant relationships between longer vase life of flowers that wilted and shorter stem length, smaller stem diameter, and smaller inflorescence diameter were determined.

Unfortunately, the current floriculture market evidently prefers gerberas with long, thick stems and large inflorescences, as most commercial cultivars tend to be of this type. Therefore, it is encouraged that postharvest longevity be deliberately included as an objective into cut flower gerbera breeding programs or the consumer will have to be satisfied with only the expected vase life possible from postharvest treatments such as floral preservatives. If attention to this situation is not given, a decline in the popularity of gerbera might result as other flower varieties with better, more predictable vase life become increasingly available.

Conclusion

Breeding to improve vase life in gerbera has potential, despite the fact that heritability was confirmed to be moderately low. It is concluded that the best approach for establishing lines with superior lasting quality requires recognizing vase life as a composite character. Distinguishing quantitatively inherited traits, i.e. days to bending, days to folding, and days to wilting and qualitatively inherited traits, i.e. bending ya wilting and folding vs non-folding, may be the key to a successful breeding program.






47


Evidence suggests that obtaining plants homozygous for wilting may be possible.

Recurrent selection then would be useful to increase the number of days to wilting. Further studies to determine the heritability of this vase life component could indicate the intensity of selection necessary to increase mean vase life for a gerbera population. Consistent selection against plants whose flowers exhibit bending or folding is imperative.

Finally, after several lines with superior vase life are established, it is recommended that breeders incorporate this trait into other lines with desireable plant morphology using a backcross mating scheme, especially if wilting can be confirmed to be a qualitative trait.













CHAPTER 4
PART II. STEM STRENGTH STUDIES

Introduction


Gerbera X hybrid Hort. is a popular cut flower that is recognized for its long, leafless, and upright stem, known as a scape. Postharvest longevity of gerbera often abruptly ends when the stem ceases to remain upright.. This phenomenon has been termed "knicking" (Wilberg, 1973; Buys, 1978), "folding" (De Jong, 1978a), "neck droop" (Zieslin et al., 1978), or "stem break" (Meeteren, 1978a). De Jong (1978b) suggested breeding for structurally strong stems may be a means to improve postharvest longevity in gerbera.

Descriptions of the relationship between stem strength and morphological and growth traits in gerbera are few. Wilberg (1973) identified that higher ratios of dry weight/cm for stem sections prone to bending contribute to stem strength in gerbera. Barigozzi and Quagliotti (1978) noted tetraploids appeared to have stronger stems than diploids. De Jong and Garretsen (1985) supported this observation by suggesting thicker and shorter stems of tetraploids may result in increased stem strength.

The objectives of this research on gerbera were to determine broad sense heritability and narrow sense heritiability estimates for stem strength and to determine correlations between stem strength and other traits of flower and stem morphology and growth.

Materials and Methods

Selection and Mating

Initially, a population of 278 plants was randomly selected for stem strength evaluation (May 25--June 14, 1984). They were selected from the "residual parental generation" of 48






49


325 plants that had been previously established for vase life studies (Chapter 3). Plants selected had already produced at least three flowers that had been evaluated for vase life. One to six flowers per plant were evaluated. After the evaluation period plants that produced less than three flowers were discarded. This "residual parental generation" included 73 plants. Plant means were determined from data collected on the first three flowers evaluated per plant. Seven plants (approx. 10% of the total number of plants evaluated for stem strength) with highest mean stem strength (-x >14.00) and lowest coefficient of variation (C.V. 27.00), were selected. To maintain genetic diversity, selection included plants from five seed populations.

A 7 x 7 diallel cross mating scheme was utilized to estimate heritability of stem

strength. Thirty plants per cross were grown to produce the diallel generation. Plants which did not produce at least one flower during a flowering period prior to stem strength evaluation (June 12-July 30, 1985) were discarded. Plant means were determined from data collected on one to three flowers per plant. Stem Strength Evaluation

Flowers with 1-2 rows of open disc florets were harvested each evening for six weeks. After 24 hours dry storage at room temperature (200-210 C), a 15 cm stem segment from each flower was evaluated for stem strength. The portion of the stem 4.5 19.5 cm below the base of the peduncle was the area from which the segment was taken. This is shown in figure 4-1.

Stem strength was determined by the amount of force required to deflect the midpoint of each segment a specified distance (F=g/cm). The segments were supported at two points, 10 cm apart; equidistant to their midpoint. Measurements were made using an Instron.1 Instron specifications are listed in table 4-1. Initially, a maximum deformation of

1.0 cm was specified with measurements taken at five equal intervals: .2 cm, .4 cm, .6


1Provided by the Dept. of Agricultural Engineering, Univ. of Florida, Gainesville, Florida, 33610.






50


cm, .8 cm, and 1.0 cm Later, flowers from the diallel cross generation were measured only for deformation at .2 cm.






15 cm. Stem Segment



i- 12 cm.


4 30 cm.


Figure 4-1. Portion of stem used for stem strength evaluation.




Table 4-1. Instron specifications Instron: Model TM Load Cell Compression 100 grams

Crosshead Speed .2 inches/minute Chart Speed 2 inches/minute Deformation Chart Distance .2 cm .79 inches .4 cm 1.57 inches .6 cm 2.36 inches .8 cm 3.15 inches 1.0 cm 3.93 inches






51


In addition to stem strength evaluation, five variables relating to flower morphology and growth were measured for each flower from plants belonging to the parental generation:

1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm)
3. Inflorescence diameter.
4. Disc diameter
5. Inflorescence weight (after storage)

Three variables were measured on each flower from plants belonging to the diallel cross generation:

1. Stem length (at time of harvest)
2. Stem diameter (at 30.0 cm)
3. Inflorescence weight (before storage)


Production

Plants were grown according to the production regime described for vase life studies (Chapter 3). Sowing, transplanting, and flowering dates for both generations are recorded in table 4-2.


Table 4-2. Record of production dates Generation Sowing Transplanting Flowering

Parental 12-2-83 2-18-84 3-15-84/5-25-84 Diallel 1-30-85 4-7-85 5-15-85/6-12-85


Ouantitative Analysis

Stem strength data were initially analyzed according to the random model for Griffing Method 3 (Griffing, 1956). This method describes a diallel mating design which includes reciprocal crosses but excludes selfs. No reciprocal differences were observed. Subsequently, analysis of variance for combining ability, using a general least squares






52


diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of plant means.

Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was estimated from ratios of the following variances:


VA = Additive genetic variance
VG = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental)


VG
h2 VA H2 = VG Vp Vp (Falconer, 1960)


Genotypic and phenotypic variances were determined from the following equations using the variance components for general combining ability (a2gca), specific combining ability (G2sca), and error (02e) which were calculated by the diallel analysis program developed by Schaffer and Usanis (1969):

VA = 4O2gca

VG = 4O2gca + 4O2sca
Vp = 4O2gca + 4O2sca + (2e (Hallauer, 1981)

Thus, heritability (h2 and H2) was estimated from the formulae:

h2- 42gca H2 = 42 gca + 4O2sca
4O2gca + 4O2sca + y2e 4O2gca + 4G2sca + y2e


Predicted estimates of narrow sense heritability and broad sense heritability for n measurements per plant were also made. This required partitioning the environmental variance (VE), a component of phenotypic variance (Vp), into general environmental






53


variance (VEg) and special environmental variance (VEs) in order to determine the phenotypic variance for each case (Vp(n)).


Vp= VG + VE
VE = VEg + VEs

VP(n) = VG + VEg + I VEs
(Falconer, 1960)

Special environmental variance (VEs) or within-individual variance (Falconer, 1960) for a single measurement per plant may be derived by the error variance component (o2e) from a one-way analysis of variance (Falconer, 1960). Henderson's Method 3 (Henderson, 1953) for obtaining variance components was performed on vase life data using individual flowers as observations rather than plant means. The SAS procedure VARCOMP (SAS, 1982) was used to obtain the a2e or error MS. General environmental variance (VEg) or between-individual variance (Falconer, 1960) was calculated by subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated in the diallel cross, from the error MS or o2e obtained by the combining ability analysis of variance.

Repeatability for vase life was determined from the following ratio of variances whereby n = 1:

VG + VEg
Vp(n)
(Falconer, 1960)

Results

Selection and Mating

Stem strength was initially measured at five intervals of deformation between zero and one centimeter. Correlations between stem strength measured at .2 cm and other intervals were






54


approximately equal to 1.0. The relationship among these measurements for each flower was generally non-linear. Stem strength means for five intervals are shown in figure 4-2.


30



20

grams

10
(Force = g/cm)


0.0 0.2 0.4 0.6 0.8 1.0 1.2
centimeters


Figure 4-2. Stem strength means at five deformation intervals.

Measurements made beyond .2 cm were often observed to be a function of the damage incurred to the stem segment as a result of force being applied. For example, some segments folded during the measurement process, hence impairing the contact between the segment and apparatus. Measurements taken at .2 cm were considered, however, to be a reasonable estimate of stem strength.

Seven parents considered to have the "best" stem strength were selected for diallel

mating based on data given in table 4-3. Stem strength means for these plants ranged from 14.2 to 31.5 g/.2 cm and coefficients of variation ranged from 5.6 to 26.8. Progeny means resulting from diallel mating ranged from 10.3 to 33.7 g/.2 cm. Listing of this data is given in table 4-4. Following selection and breeding, the population mean for stem strength more than doubled. Comparison of stem strength data for the parental and diallel generations is shown in table 4-5.






55

Table 4-3. Stem strength data of seven parents. (F = g/.2 cm) Coefficient of
Parent Mean Variation
83- 1 10 22.0 26.8 83-1-31 31.5 15.6 83-1-96 18.0 5.6
83-4-8 14.2 25.5 83 5 76 14.3 23.2 83 -7 48 24.0 24.2 83-8-7 16.0 23.6



Table 4-4. 7 x 7 Diallel. Stem strength data from diallel crosses. (F = g/.2cm; reciprocals pooled)
(Diallel Generation)
Cross Progeny Mean
83-1-10 x 83-1-31 31.3 83-1-10 x 83-1-96 26.6 83-1-10 x 83-4-8 17.1 83-1-10 x 83-5-76 32.3 83-1-10 x 83-7-48 32.4 83-1-10 x 83-8-7 19.7 83-1-31 x 83-1-96 23.2 83-1-31 x 83-4-8 14.2 83-1-31 x 83-5-76 27.5 83-1-31 x 83-7-48 26.7 83-1-31 x 83-8-7 23.9 83-1-96 x 83-4-8 14.2 83-1-96 x 83-5-76 33.7 83-1-96 x 83-7-48 29.7 83-1-96 x 83-8-7 21.2 83-4-8 x 83-5-76 15.9 83-4-8 x 83-7-48 14.4 83-4-8 x 83-8-7 10.3 83-5-76 x 83-7-48 24.2 83-5-76 x 83-7-48 29.9 83-7-48 x 83-8-7 25.1






56


Table 4-5. S of stem stength for parental and diallel generations. (F = g/.2cm) No. of Coefficient of Generation Plants Mean Std. Dev. Variation

Parental 73 10.17 7.27 71.48 Diallel 642 22.52 11.32 50.28


Heritabilit

Combining ability analysis of variance using plant means was performed on stem strength data from a 7 x 7 diallel. General combining ability effects were highly significant. Specific combining ability effects were non-significant. Results are summarized in table 4-6. Heritability was estimated by the ratio of genetic variance (VA or VG) to phenotypic variance (Vp) (Falconer, 1960). Variances were derived using variance components for general combining ability (o2gca), specific combining ability (a2sca), and error (o2e) according to formulae published by Hallauer (1981). The difference between narrow sense heritability (h2 = .38) and broad sense heritability (H2 = .42) estimates was small. This indicates the effect attributable to non-additive genetic variance (VG VA) is minimal. Variances and heritability estimates for this population are given in table 4-7.


Table 4-6. 7 x 7 Diallel. Combining ability analysis of variance for stem strength. (F = g/.2cm)

Source of Variation df M.S. F-ratio General combining ability 6 4218.62 16.32 * Specific combining ability 14 258.47 1.63 n.s. Error 621 158.22
* Significant at P< 0.01.


Estimates of heritability were based on an average of 2.66 measurements per plant. This value (n) resulted by evaluating 1710 flowers from 642 plants. The error variance






57




Table 4-7. 7 x 7 Diallel. Variances and heritability estimates for stem strength. (F = .2 cm)

VA VG Vp h2 H2

103. 64 116.76 274.98 .38 .42


component (2,) for a single measurement per plant was obtained by an analysis of variance using stem strength data from individual flowers of the diallel generation. Also, this analysis indicated differences in stem strength of flowers among crosses and among plants within crosses were highly significant. Results are summarized in table 4-8.


Table 4-8. 7 x 7 Diallel. Analysis of variance for stem strength. (F=g/.2 cm) Source of Variation df M.S. F-ratio

Among crosses 20 3995.98 10.68 * Among plants 621 374.08 2.12 * Within plants 1068 146.06
* *Significant at P< 0.01.


General environmental variance (VEg) and special environmental variance (VEs) were derived from stem strength data using calculations described by Falconer (1960). A summary of variances for this population is given in table 4-9. Predicted estimates of narrow sense heritability and broad sense heritability for stem strength were then made for 1, 2, 3, 5, and ** measurements per plant by the ratio of genetic variance (VA or VG) to phenotypic variance (Vp(n)) (Falconer, 1960). Estimates ranged from 28 to 53 percent. These results are given in table 4-10. Repeatability (r = .60) for stem strength was moderately high.






58




Table 4-9. 7 x 7 Diallel. Summary of variances for stem strength. (F = g/.2 cm) n = 2.66
Genotypic Variance Environmental Variance Phenotypic Variance

VA VG VE VEgVEs Vp(n)

103.64 116.76 158.22 103.31 146.06 274.98


Table 4-10. 7 x 7 Diallel. Predicted estimates of heritability for stem strength. (F = g/.2 cm)
Number of Measurements

Heritability 1 2 3 5 0 h2 .28 .35 .39 .42 .47 H2 .32 .40 .43 .47 .53 Correlations between Stem Strength and Other Traits
Phenotypic correlations between stem strength and three stem and inflorescence traits, stem length, stem diameter, and inflorescence weight, were determined for both parental and diallel generations. Correlation coefficients based on data from individual flowers were obtained using a SAS program for Pearson's product-moment correlation procedure (SAS, 1986).

Correlations determined from the parental generation demonstrated the linear

relationship between stem strength and these traits in a random population before selection for improving stem strength. Correlations determined from the diallel generation demonstrated the linear relationship between stem strength and these traits after selection. These results are summarized in table 4-11.

Stem strength and stem length. Significant negative correlations were observed

between stem strength and stem length, before and after selection. It appeared, however,






59

Table 4-11. Phenotypic correlation coefficients between stem strength (F = g/.2 cm) and three traits before and after selection.
Generation N StemLengttemStem Diameter Inflorescence Weight
(after storage)
Parental 548 -.28 * .10* .16 *
(before storage)
Diallel 1710 -.05 .22 * .38 *
* Significant at P< 0.05. * Significant at P5 0.01.

that following selection and mating, the relationship weakened.

Stem strength and stem diameter. Significant positive correlations were observed between stem strength and stem diameter, before and after selection. Unlike the relationship between stem strength and stem length, it appeared that following selection and mating, the relationship strengthened.

Stem strength and inflorescence weight. Significant positive correlations were observed between stem strength and inflorescence weight, before and after selection. Similar to the relationship between stem strength and stem diameter, in this case, following selection and mating, the correlation coefficient was higher.

Correlations between stem strength and two other traits, inflorescence diameter and disc diameter, were also determined prior to selection. The correlation (r = -.08) between stem strength and inflorescence diameter was not significant. A positive correlation (r = .11) observed between stem strength and disc diameter was highly significant; as stem strength increased, disc diameter increased.

Discussion

Strength of gerbera stem segments was determined using an Instron. Results were recorded in terms of the amount of force required to deflect the segment a specified distance. This method, similar to flexture tests made on engineering materials (Mohsenin, 1970) was developed to test the segment as a simple beam. The intention of using an Instron to determine stem strength was to increase the accuracy of determinations.






60


Although this apparatus provided a relatively sophisticated method to quantitatively measure mechanical strength, a major problem remained due to the variability in gerbera stems. Gerbera stems are not structurally homogeneous. In addition to differences in stem diameter, stems can be hollow or solid and round or oval. Measurements for determining the force required to deflect a stem segment a specified distance can vary depending on these factors.

Mohsenin (1970) suggested accounting for variation, often found in agricultural materials, by using a formula to calculate apparent stiffness, known as modulus of elasticity (E). Parups and Voisey (1976), who studied the resistance to bending of the pedicel in greenhouse-grown roses, which are also not structurally homogeneous, explained their calculations for determining modulus of elasticity of the pedicel part of rose stems. There was concern about the variability in gerbera stems, therefore stem strength results obtained from raw data were initially compared to modulus of elasticity values (E). The following formula was used to calculate modulus of elasticity: FL3
E=
Xd4
D48 i )
(Mohsenin, 1970)

where F is force required to deflect the segment a specified distance, L is length of the segment, D is deflection distance, and d is diameter of the stem segment at mid-span.

Since no differences were found between ranking plants using raw data directly from Instron measurements and ranking plants by modulus of elasticity, stem strength values obtained from raw data were used to estimate heritability and to correlate stem strength to morphological traits.
Another problem with accurately evaluating stem strength in gerbera relates to the moisture content of the stem segment measured. Segments were stored at room temperature for 24 hours; however, this did not specify their exact level of moisture content at the time of measurement. Evidently the proportion of evaporation differed between stem






61


segments. Frequently, but not always, stem segments with small diameters dehydrated more than segments with larger diameters. These segments became brittle and resulted in stem strength greater than 50g/.2 cm, much higher than most stem strength values recorded.

Despite efforts to obtain an accurate evaluation of stem strength for gerbera flowers,

the coefficient of variation for stem strength in our experiment exceeded fifty percent before and after selection and mating to increase stem strength. This demonstrates that environmental conditions greatly influence this character.

Broad sense heritability for stem strength (H2 = .42), based on 2.66 measurements per plant, appears moderate; therefore improvment for this character can be expected with even a moderate rate of selection intensity. Narrow sense heritability was also fairly moderate (h2 = .38), indicating parental phenotypes could be expected to correspond to some degree to their genotypes.

Genetic variation may be largely controlled by additive gene action since the difference between broad sense and narrow sense heritability was small. Therefore, in a fixed model experiment, progeny means obtained from a top-cross mating would be effective in determining parents with good combining ability for increasing vase life.

As discussed in Chapter 3, environment is a critical variance component of heritability. Different estimates of heritability would be expected if the magnitude of environmental conditions varied, assuming effects due to genotype remained constant for a given population.

Falconer (1960) demonstrated a method to predict estimates of heritability for a specified number of measurements per experimental unit. This involved partitioning environmental variance (VE) into general environmental variance (VEg) and special environmental variance (VEs). Special environmental variance (VEs) or within-plant variation is the environmental variation for a single observation per experimental unit. The magnitude for special environmental variance (VEs) is then divided by a specified number






62


of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic variance for each special case. Ideally, if n = -, then VEs will be reduced to zero, thereby deleting a significant source of environmental variation. In that case, the highest possible estimate of heritability could be obtained for a given population.

Using this method, predicted estimates of heritability for vase life in gerbera ranged from 28 to 53 percent with n specified as 1, 2, 3, 5, and flowers per plant. Since these calculations assumed no change in genotypic effects, the differences between broad sense and narrow sense heritability estimates for each case were small. This range of heritability reinforces the initial description that heritability is moderate, and improvement of stem strength in gerbera can be obtained even with moderate selection.

The repeatability estimate for stem strength (r = .60) is moderately high, indicating that two to three flowers per plant is adequate for determining the average stem strength per plant. Falconer (1960) recommends that further gain in accuracy by more than two measurements does not justify additional expense or time required to collect more data when repeatability is high. This was proven by comparing the relative increase in heritability from predicted estimates based on 1, 2, 3, 5, and measurements per plant. Between one and three measurements, heritability increased by eleven percent, while beyond three measurements through infinity, the gain in heritability was only eight percent.

Correlations between stem strength and other morphological traits in this population of gerbera are of interest, given the wide variation in morphological phenotypes studied. Despite correlation coefficients were often significant due to the large number of flowers evaluated, they were generally low.

Significant relationships between stronger stem strength of flowers and shorter stem length, larger stem diameter, and heavier inflorescence weight were determined before and after selection and mating. Already, Barigozzi and Quagliotti (1978) and De Jong and Garretsen (1985) observed that tetraploids appear to have stronger stems than diploids. These findings provide strong evidence why the use of tetraploids, whose flowers






63


generally have shorter, thicker stems and heavier inflorescences, has been suggested as a possible means for increasing stem strength in gerbera.

Conclusion


Improving structural stem strength in gerbera can be realized through breeding efforts. Despite this encouraging conclusion, the breeder faces difficulty with selecting plants whose flower stems are structurally strong. Large intraplant variation associated with this character is a major problem.

This situation highlights several unanswered questions, "What determines structural strength in flower stems?" Is it a composite character? If so, what are the individual components of structural stem strength? How do they contribute to postharvest longevity in gerbera? It is recommended that further research to identify factors that contribute to structural stem strength in gerbera be conducted, particularly if accuracy in evaluating stem strength can be attained.

Another important concern for the breeder involves defining the conditions under which stem strength should be measured. An investigation of the effects of pre-harvest environmental conditions versus morphological or anatomical variability of the stem could yield useful information for developing a method to evaluate the magnitude of stem strength more accurately.

Clarifying breeding objectives and developing appropriate evaluation methods could be the key to a successful breeding program to improve structural stem strength in gerbera.













CHAPTER 5
PART III. VASE LIFE X STEM STRENGTH STUDIES Introduction


Postharvest longevity is a critical attribute of cut flowers. De Jong (1978a) proposed two main components, vase life and stem strength, contribute to postharvest longevity in gerbera. In a second paper, De Jong (1978b) suggested structurally strong stems could extend postharvest longevity by providing added support to the flower should a water deficit occur.

The basis for vase life and stem strength in gerbera remains under investigation. Stem anatomy studies have been conducted in an effort to understand the causes for variation in postharvest longevity. Reiman-Philip, as cited by Wilberg (1973), noted a greater proportion of large vascular bundles to small vascular bundles as a possible factor contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that such a relationship existed in only extreme cases. In a comprehensive study, Marousky (1986) measured the size and number of vascular bundles in two gerberas, 'Tropic Gold' and 'Appleblossom.' He concluded 'Appleblossom' exhibited more resistance to stem bending than 'Tropic Gold' partially because of various anatomical features such as fewer small vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference, and a greater percentage of dry weight per unit of scape length.

The objectives of this study were to determine the relationship between vase life and stem strength from a population of Gerbera X hybrida Hort. which varied greatly in flower and stem morphology and to compare stem anatomy of flowers from plants that were classified by differences in vase life and stem strength.



64






65

Materials and Methods


Selection and Classification

Seventy-three plants were evaluated for both stem strength and vase life in 1984 (Chapter 4). Three flowers per plant were evaluated to obtain mean values for each variable. Based on these means, 22 plants were selected and classified into four categories. The categories were defined by vase life and stem strength ratings described in table 5-1. To represent genetic diversity, selection included plants from five seed populations.

Table 5-1. Vase life x stem strength classification
Plant
Category Vase Life Stem Strength

I High > 11.0 days High > 11.5g/.2cm II High 2 12.0 days Low 5 5.5g/.2cm III Low <6.0 days High > 11.3g/.2cm

IV Low 5 4.0 days Low -2.5g/.2cm



Stem anatomy of flowers from plants assigned to each category was compared (February 20-July 30, 1985). Approximately 14 flowers per plant were examined. 'Appleblossom,' assigned to category I, was also examined because of its low intraplant variation for vase life and stem strength (Chapters 3 and 4).

Anatomy Examination

Flowers at various stages of maturity with at least one row of disc florets open were randomly sampled. Using a light microscope with 40x magnification, fresh stem sections, approximately 50-60 microns thick, were cut, using a razor blade, 12 cm below each peduncle and examined. No stains were used. The number of large and small vascular bundles was recorded for each flower. Determination of large and small bundles was relative between plants. Therefore, before counting, the smallest "large" bundle was






66


measured. All bundles greater or equal to that size were described "large," all others; "small." This method was described by F. J. Marousky, personal communication, 1985. Circumference and area of each cross section were calculated using the diameter of each stem, measured using a vernier caliper, at 12 cm below the peduncle.

A set of variables was created which described a series of mathematical relationships determined by direct and indirect measurements. A list of these variables is given in table 5-2.

Table 5-2. List of variables from anatomy evaluation of gerbera.
LBUN = # of large bundles SBUN = # of small bundles
TBUN = # of large bundles + # of small bundles
LBUNARAT = # of large bundles per 1.0 cm2 stem area @ 12 cm SBUNARAT = # of small bundles per 1.0 cm2 stem area @ 12 cm TBUNARAT = Total # of bundles per 1.0 cm2 stem area @ 12 cm
LBUNCRAT = # of large bundles per 1.0 cm stem circumference @ 12 cm SBUNCRAT = # of small bundles per 1.0 cm stem circumference @ 12 cm TBUNCRAT = Total # of bundles per 1.0 cm stem circumference @ 12 cm



Results


Correlation Between Vase Life and Stem Strength

A significant positive correlation (r = .28) was observed when vase life means were plotted against stem strength means from 73 plants belonging to the parental generation described in Chapters 3 and 4. The relationship between these two components of postharvest longevity appeared moderately weak, despite significance. Except for one plant, plants with the weakest mean stem strength had a mean vase life of less than five days. Distribution of these means is shown in figure 5-1.






67

40 y= 5.15 +.52x r = .28
30 N


Stem strength 20.
(F g/.2cm) N

10



0 10 20
Vase life (days)

Figure 5-1. Distribution of vase life and stem strength means for seventy-three plants Comparison of Stem Anatomy Among Genotypes

Twenty two plants were selected and classified by high and low levels of vase life and stem strength according to the categories described. Listing of these plants and their corresponding mean vase life and stem strength is given in table 5-3.

Analyses of variance were made using plant means for nine stem anatomy variables.

Vase life and stem strength were treated as main effects. No significant interaction between vase life and stem strength was determined for any of the nine variables. Analyses of variance for the number of vascular bundles directly measured per stem section (TBUN, LBUN, and SBUN) only yielded a significant difference between high and low levels of vase life for number of large bundles (LBUN). The mean number of large vascular bundles was significantly different for plants with high vase life compared to those with low vase life. Plants with high vase life exhibited a smaller number of large vascular bundles than plants with low vase life. A summary of these results is given in tables 5-4 and 5-5.

Analyses of variance for the number of vascular bundles calculated per unit stem area (1.0 cm2) at 12 cm below the peduncle (TBUNARAT, LBUNARAT, and SBUNARAT), only yielded significant differences between high and low levels of vase life for total






68

Table 5-3. Vase life and stem strength data used for classification of twenty-two plants including 'Appleblossom.'
Category I Category III

Plant Vase life Stem Plant Vase life Stem strength strength
(days) (g/.2cm) (days) (g/.2cm) 83-1-19 12.0 13.3 83- 1-2 5.7 13.3 83-1-26 16.0 20.5 83-4-8 4.3 14.2 83-1-31 11.0 31.5 83-5-1 5.7 14.2 83-1-54 11.0 12.7 83-5-76 5.7 14.3 83-7-47 12.7 11.5 83-7-22 6.0 16.8 83-8-7 11.3 16.0 83-7-48 4.7 24.0 'Appleblossom' 11.0 25.7 83-7-35 5.7 11.3
Category II Category IV

Plant Vase life Stem Plant Vase life Stem strength strength
(days) (g/.2cm) (days) (g/.2cm)

83- 1-6 13.0 4.3 83-1-24 3.7 1.2 83-1-32 12.3 4.7 83-1-79 4.0 2.0 83-1-64 15.0 5.5 83-4-29 3.3 1.5 83-1-77 15.3 4.2 83-4-16 3.0 2.3


Table 5-4. Analysis of variance for stem anatomy. (# of vascular bundles) TBUN LBUN SBUN Source df M.S. M.S. M.S.

Vase Life 1 1.03 n.s. 382.68 423.50 n.s. (high ys low)

Stem Strength 1 380.60 n.s. 8.55 n.s. 275.07 n.s. (high yS low)

Vase Life x Stem Strength 1 274.46 n.s. 15.14 n.s. 418.53 n.s. Error 18 535.68 71.62 447.56
* Significant at P< 0.05.






69


Table 5-5. Meansz for number of vascular bundles between high and low levels of vase life and stem strength in gerbera.
TBUN LBUN SBUN

Vase High 44.92a 16.37a 33.55a
Life
Low 49.81a 18.59b 31.22a

Stem High 48.76a 17.31a 31.45a Strength
Low 50.97a 17.64a 33.33a
z Means followed by different letters within each bundle group for vase life or stem strength are significant at the 5 % level.


number of bundles (TBUNARAT) and number of large bundles (LBUNARAT). The means for total number of vascular bundles and number of large vascular bundles for this unit area in gerbera stem sections were significantly different for plants with high vase life compared to those with low vase life. Plants with high vase life exhibited a smaller total number of vascular bundles per unit area than plants with low vase life. Also, plants with high vase life exhibited a smaller number of large vascular bundles per unit area than plants with low vase life. A summary of these results is given in tables 5-6 and 5-7.



Table 5-6. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm 2 scape area.)
TBUNARAT LBUNARAT SBUNARAT
Source df M.S. M.S. M.S.

Vase Life 1 444840.20 130549.39 93420.23 n.s. (high xa low)

Stem Strength 1 196864.30 n.s. 12087.29 n.s. 111390.20 n.s. (high ys low)

Vase Life x Stem Strength 1 10271.08 n.s. 64.98 n.s. 8702.12 n.s. Error 18 79568.77 20121.40 26395.89
* Significant at P5 0.05.






70

Table 5-7. Meansz for number of vascular bundles per 1.0 cm 2 scape area between high and low levels of vase life and stem strength in gerbera.
TBUNARAT LBUNARAT SBUNARAT Vase High 239.06a 78.85a 160.22a
Life
Low 314.71b 119.83b 194.88a

Stem High 251.72a 93.10a 158.62a Strength
Low 302.05a 105.57a 196.48a z Means followed by different letters within each bundle group for vase life or stem strength are significant at the 5 % level.


Significant differences between high and low levels of vase life and high and low levels of stem strength for total number of vascular bundles calculated per unit stem circumference (1.0 cm) at 12 cm below the peduncle (TBUNCRAT) were determined. The means for total number of vascular bundles per unit circumference in gerbera stem sections were significantly different for plants with high vase life compared to those with low vase life. Plants with high vase life exhibited a smaller total number of vascular bundles per unit circumference than plants with low vase life. Also, the means for total number of vascular bundles per unit circumference in gerbera stem sections were significantly different for plants with high stem stength compared to those with low stem strength. Plants with high stem strength exhibited a smaller total number of vascular bundles per unit circumference than plants with low stem strength.

A highly significant difference between high and low levels of vase life was

determined for the number of large vascular bundles calculated per unit stem circumference (1.0 cm) at 12 cm below the peduncle (LBUNCRAT). The means for number of large vascular bundles per unit circumference in gerbera stem sections were significantly different for plants with high vase life compared to those with low vase life. Plants with high vase life exhibited a smaller number of large vascular bundles per unit circumference than plants with low vase life. A significant difference between high and low levels of stem strength was determined for the number of small vascular bundles calculated per unit






71


stem circumference (1.0 cm) at 12 cm below the peduncle (SBUNCRAT). The means for number of small vascular bundles per unit circumference in gerbera stem sections were significantly different for plants with high stem strength compared to those with low stem strength. Plants with high stem strength exhibited a smaller number of small vascular bundles per unit circumference than plants with low stem strength. A summary of these results is given in tables 5-8 and 5-9.


Table 5-8. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm scape circumference.)
TBUNCRAT LBUNCRAT SBUNCRAT

Source df M.S. M.S. M.S.

Vase Life 1 1467.31 742.67 122.18 n.s. (high s low)

Stem Strength 1 1108.91 61.55 n.s. 647.96 (high v low)

Vase Life x Stem Strength 1 143.85 n.s. .69 n.s. 164.47 n.s. Error 18 237.65 80.43 124.06
* Significant at P< 0.05. * Significant at P 0.01.


Table 5-9. Meansz for number of vascular bundles per 1.0 cm scape circumference between high and low levels of vase life and stem strength in gerbera.
TBUNCRAT LBUNCRAT SBUNCRAT Vase High 30.46a 10.02a 20.44a
Life
Low 34.81b 13.1 lb 21.70a

Stem High 30.75a 11.12a 19.63a Strength
Low 34.52b 12.01a 22.51b
z Means followed by different letters within TBUNCRAT and SBUNCRAT for vase life or stem strength are significant at the 5 % level and significant at the 1% level within LBUNCRAT.






72


The only variable analyzed for which both significant differences were found between high and low levels of vase life and stem strength was the total number of vascular bundles per unit of circumference (TBUNCRAT). The means for this variable with respect to both levels of vase life and stem strength are presented in table 5-10. The mean for total number of vascular bundles per unit stem circumference (1.0 cm) at 12 cm below the peduncle was lowest for plants with high vase life and high stem strength. Conversely, the mean was highest for plants with low vase life and low stem strength. The difference between the means of plants with high and low vase life for total number of vascular bundles per unit of circumference among plants with high stem strength is not equal to the difference between the means of plants with high and low vase life for total number of vascular bundles per unit of circumference among plants with low stem strength (28 34 # 33 36). Similarly, the difference between the means of plants with high and low stem strength for total number of vascular bundles per unit of circumference among plants with high vase life is not equal to the difference between the means of plants with high and low stem strength for total number of vascular bundles per unit of circumference among plants with low vase life (28 33 # 34 36). These inequalities possibly indicate a source of experimental error since interaction effects were non-significant.


Table 5-10. Means for total number of vascular bundles per 1.0 cm scape circumference between high and low levels of vase life and stem strength of twenty two gerbera plants.
Vase Life
High Low x Stem High 28 34 31.0
Strength
Low 33 36 34.5 R 30.5 35.0
Correlations Between Stem Anatomy and Postharvest Longevity

Correlation coefficients between nine stem anatomy variables and two components of postharvest longevity, vase life and stem strength, were determined using means from






73


twenty two plants. Significant negative relationships were determined between vase life

and two stem anatomy variables which were based on total number of vascular bundles

(TBUNARAT, and TBUNCRAT). Significant relationships between vase life and three

stem anatomy variables which were based on number of large vascular bundles (LBUN,

LBUNARAT, and LBUNCRAT) were also negative. Highly significant at the 1 % level,

the strongest correlation was between vase life and number of large bundles per unit of

circumference. Correlations relating to stem anatomy variables based on small vascular

bundles and vase life or stem strength were not significant. Summary of these results are

given in tables 5-11, 5-12, and 5-13.

Table 5-11. Correlation coefficients between stem anatomy (number of vascular bundles) and postharvest longevity.
TBUN TBUNARAT TBUNCRAT Vase Life
(days) .06 -.46 -.44 *
Stem Strength
(F=g/.2 cm) -.27 -.33 -.39
* Significant at P 0.05.

Table 5-12. Correlation coefficients between stem anatomy (number of large vascular bundles) and postharvest longevit.
LBUN LBUNARAT LBUNCRAT Vase Life
(days) -.51* -.53* -.67 *
Stem Strength
(F=g/.2 cm) -.21 -.21 -.27
* Significant at P< 0.05. * Significant at P5 0.01.


Table 5-13. Correlation coefficients between stem anatomy (number of small vascular bundles) and postharvest longevity.
SBUN SBUNARAT SBUNCRAT
Vase Lie
(days) .37 -.34 -.23
Stem Strength
(F=g/.2 cm) -.08 -.40 -.40






74




The relationship between vase life and structural stem strength described in this study was weak ( r = .28) but, significant at the level of P< 0.05. Therefore, it would be expected that flowers with the highest vase life will only have a slight tendency for high stem strength. Breeding for both characters may improve keeping quality in gerbera cut flowers only marginally.

Whether or not the correlation provides a good description of the relationship deserves comment. First of all, plant means for each variable were plotted against each other. Since evaluating vase life and stem strength requires destructive tests, plant means for each variable were determined by different sets of flowers. Also, vase life means were determined from flowers evaluated during the period May 5 to June 12, 1985, and stem strength means were determined from flowers evaluated during the period June 12 to July 30, 1985. Environmental differences during these two periods may not have affected all of the plants similarly. In addition, problems associated with the accuracy of stem strength determinations (Chapter 4) may have skewed the comparison.

Variation in the number of vascular bundles per cross-section of gerbera flower stems cut 12 cm below the peduncle was compared among plants whose flowers exhibited high or low vase life and stem strength. Large and small bundles were counted. Their sum was calculated to obtain a value for total number of bundles. The diameter of each cross-section was also measured. Using the diameter measurement, the number of bundles per unit area (1.0 cm2) or unit circumference (1.0 cm) for each cross-section was also calculated.

Overall, differences in the number of large bundles appears related to vase life. Fewer number of large bundles, fewer number of large bundles per unit area, and fewer number of large bundles per unit circumference were observed in plants with high vase life. In contrast, only the difference in the number of small bundles per unit circumference appears to affect stem strength. Fewer numbers of small bundles were observed in plants with high






75


stem strength. Therefore, it follows that the fewer total number of vascular bundles per unit circumference was observed in plants with high vase life and high stem strength.

Results from correlations determined between these stem anatomy variables and vase life or stem strength enhance the link suggested between size and number of vascular bundles to postharvest longevity. Fewer bundles, small or large, were found in plants with high vase life or high stem strength. Negative correlations between stem anatomy variables involving large bundles were significant for vase life but not stem strength at the level of P 0.05. Correlations between stem anatomy variables involving small bundles were not significant for vase life or stem strength at the level of P< 0.05. However, negative correlations between numbers of small bundles per unit area or unit circumference and stem strength were significant at the level of P< 0.10.

Strong turgor strength in gerbera stems has been considered an important factor for

maintaining postharvest longevity (De Jong, 1978a). Assuming a constant supply of water available to flowers with equal stem diameters, fewer vascular bundles would be expected to increase the flow rate of water in the stem. It seems likely that this situation would serve to maintain an upright stem, if stem rigidity depended on the maintenance of turgor. Additionally, an increase in flow rate of water in the stem may reduce the opportunity for microbial growth to occur, a factor cited to decrease vase life in gerbera (Meeteren, 1978a). Since differences in large vascular bundles and not small vascular bundles were found in plants with high and low levels of vase life, it is theorized that large vascular bundles are mainly responsible for differences in water uptake rates in gerbera flowers.

As for the role of vascular bundles in stem strength, only small bundles appeared to be a factor. De Jong (1986) proposed that stronger structural stem strength would help to decrease the incidence of stem folding. Marousky (1986) examined stem anatomy, including number and size of vascular bundles, of two cultivars whose flowers differed in resistance to scape breakage. The cultivar that showed less resistance to breakage had more small bundles than the cultivar that showed more resistance to breakage. One reason for






76


stronger stems exhibiting less incidence of folding could be because stronger stems have fewer small vascular bundles.

Dubuc-Lebreux and Vieth (1985) observed that a greater proportion of support tissues, such as sclerenchyma, was found in lower regions of the stem where breakage did not normally occur. They suggested that plants whose stems showed less incidence of folding at the typical zone of folding (10-15 cm below the peduncle) might have more support tissues in this area compared to plants whose stems showed more incidence of folding. Perhaps in stonger stems, small vascular bundles are displaced by more support tissue; therefore, the fewer small vascular bundles in a stem cross-section, the greater proportion of support tissue will be present.

Conclusion

It is concluded that ifvase life and stem strength are two distinct components of

postharvest longevity in Gerbera X hybrid Hort., then the breeder wishing to improve postharvest quality in this cut flower might consider selecting plants with good combining ability for both high vase life and high stem strength.

Stem anatomy studies yielded interesting results. It appears that vase life is affected by the number of large vascular bundles and stem strength is affected by the number of small vascular bundles. Fewer bundles of each type per unit of circumference were found in plants with high vase life and high stem strength.

Additional genetic investigations to determine the heritability of stem anatomy variables involving number of large vascular bundles and stem anatomy variables involving number of small vascular bundles are recommended.













CHAPTER 6
SUMMARY


Breeding cultivars with superior postharvest longevity may be the best approach for satisfying the floral consumer's expectation of cut flower postharvest quality. Combining this effort with postharvest treatments, such as the use of floral preservatives, would help maintain the popularity of gerbera for years to come.

The heritability of two characteristics, vase life and stem strength, considered to be components of postharvest longevity, was determined. Although results indicate that intense selection of plants for these traits will yield increases in postharvest longevity, it is of utmost importance to determine appropriate breeding objectives which will best improve postharvest performance.

In the case of gerbera, a simple analysis of vase life appears inadequate. Breeding programs which incorporate selection against frequency of bending and folding and selection for increased days to wilting may be more effective. Further research is needed to prove that undesireable modes of senescence, i.e. bending and folding, are qualitatively inherited by only several genes.

An attempt to document the role of stem strength in postharvest longevity of gerbera was made; however, the function of this component cannot be specified at this time. Difficulties associated with measuring stem strength were discussed. As factors are identified which affect stem strength, breeding programs to improve postharvest performance can be refined.

While vase life and stem strength have been postulated as components of postharvest longevity, their effects appear independent of each other. Correlations demonstrated that




77






78


vase life increased as stem diameter decreased. In contrast, stem strength increased as stem diameter also increased.

The number of vascular bundles observed in cross sections of scapes at the zone of bending or folding suggested that the ratio of small bundles per unit scape circumference may affect stem strength and the ratio of large bundles per unit circumference may affect vase life. Heritability studies on these anatomical ratios in gerbera could provide pertinent information that would be useful to plant breeders.













REFERENCES


Abdel-Kader, H. and Rogers, M.N., 1986. Postharvest treatment of Gerbera iamesonii. Acta Hort., 181: 169-176.

Amariutei, A., Burzo, I. and Alexe, C., 1986. Researches concerning some metabolism aspects of cut gerbera flowers. Acta Hort., 181: 331-337.

Awad, A.R.E., Meawad, A., Dawh, A.K. and Elsakaa, M., 1986. Cut flower longevity as affected by chemical pretreatment. Acta Hort., 181: 177-172.

Ballantyne, D.J., 1963. Respiration of floral tissue of the daffodil (Narcissus pseudonarcissus Linn.) treated with benzyladenine and auxin. Can. J. Bot., 44: 117-119.

Barigozzi, C. and Quagliotti, L., 1978. Current research on breeding of gerbera. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 57-68.

Benschop, M. and De Hertogh, A.A., 1969. An analysis of the post-harvest characteristics of cut tulips. Flor. Rev., 145: 24-25, 62-65.

Borghi, B. and Baldi, V., 1970. Variabilita tra cloni di gerbera (Gerber jamesonii) allevati in diverse condizioni ambientali. Sementi elette, 6: 25-32.

Bravdo, B., Mayak, S. and Gravrieli, Y., 1974. Sucrose and water uptake from concentrated sucrose solutions by gladiolus shoots and the effect of these treatments on floret life. Can. J. Bot., 52: 1271-1281.

Bufler, G., 1986. Ethylene biosynthesis and action. Acta Hort., 181: 93-98.

Burdett, A.N., 1970. The cause of bent neck in cut roses. J. Amer. Soc. Hort. Sci., 95: 427-431.

Buys, C., 1978. Quality problems in gerbera. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 2428 April, at Alassio, Italy, pp. 255-262.

Carpenter, W.J. and Rasmussen, H.P., 1973. Water uptake rates by cut roses (Rosa hybrida) in light and dark. J. Amer. Soc. Hort. Sci., 98: 309-313.

Carpenter, W.J. and Rasmussen, H.P., 1974. The role of flower and leaves in cut flower water uptake. Scientia Hortic., 2: 293-298.

Cocozza Talia, M., Porceddu, E., and Stellacci, P., 1978. Genetic variability in gerbera. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 163-177.

79






80


Coorts, G.D., 1973. Internal metabolic changes in cut flowers. HortScience, 8: 195-198.

Coorts, G.D., McCollum, J.P. and Gartner, J.B., 1965. Effect of senescence and preservative on mitochondrial activity in flower petals of Roa hybrida 'Velvet Times.' Proc. Amer. Soc. Hort. Sci., 86: 791-797.

De Jong, J., 1978a. Dry storage and subsequent recovery of cut gerbera flowers as an aid in selection for longevity. Scientia Hort., 9: 389-397.

De Jong, J., 1978b. Selection for keeping quality in gerbera. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 263-268.

De Jong, J., 1986. Breeding for keeping quality in gerbera. Acta Hort., 181: 353-357.

De Jong, J. and Garretsen, F., 1985. Genetic analysis of cut flower longevity in gerbera. Euphytica, 34: 779-784.

De Leo, V. and Ottaviano, E., 1978. Genetical analysis of morphological traits in Gerbera jamesonii. Clonal and diallel families. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 179-191.

De Witte, Y. and Doom, W.G. van, 1988. Identification of bacteria in the vase water of roses, and the effect of the isolated strains on water uptake. Scientia Hortic., 35: 285-291.

Dilley, D.R. and Carpenter, W.J., 1975. The role of chemical adjuvants and ethylene synthesis on cut flower longevity. Acta Hort., 41: 117-132.

Dixon, M.A. and Peterson, C.A., 1989. A re-examination of stem blockage in cut roses. Scientia Hortic., 38: 277-288.

Dixon, M.A., Butt, J.A., Murr, D.P. and Tsujita, M.J., 1988. Water relations of cut greenhouse roses:the relationships between stem water potential, hydraulic conductance and cavitation. Scientia Hortic., 36: 109-118.

Downs, C. and Reihana, M., 1987. Extending vaselife and improving quality of nerine cut flowers with preservatives. HortScience, 22: 670-671.

Drennan, D., Harding, J. and Byrne, T.G., 1986. Heritability of inflorescence and floret traits in gerbera. Euphytica, 35: 319-330.

Dubuc-Lebreux, M.A. and Vieth, J., 1985. Histologie du pedoncule inflorescentiel de Gerbera iamesonii. Acta Bot. Neerl., 34(2): 171-182.

Durkin, D. and Kuc, R., 1966. Vascular blockage and senescence of the cut rose flower. Proc. Amer. Soc. Hort. Sci., 89: 683-688.

Eijk, J.P. van and Eikelboom, W., 1976. Possibilities of selection for keeping quality in tulip breeding. Euphytica, 25: 353-359.

Eijk, J.P. van and Eikelboom, W., 1980. Methods of selection in tulip breeding. Acta Hort., 109: 217-225.






81


Eijk, J.P. van and Eikelboom, W., 1986. Aspects of breeding for keeping quality in TuliDa Acta Hort., 181: 237-244.
Eijk, J.P. van, Eikelboom, W. and Sparnaaij, L.D., 1977. Possibilities of selection for keeping quality in tulip breeding. Euphytica, 26: 825-828.

Falconer, D.S., 1960. Introduction to Quantitative Genetics. Oliver and Boyd, Edinburgh, 365 pp.

Ferreira, D.I. and Swardt, G.H., 1981a. The influence of the number of foliage leaves on the vase life of cut rose flowers in two media. Agroplante, 13: 73-76.

Ferreira, D.I. and Swardt, G.H., 1981b. A comparison of the vase life and respiration rate of ten cut rose cultivars and the influence of a flower preservative thereupon. Agroplante, 13: 77-81.

Fujino, D.W., Reid, M.S. and Yang, S.F., 1981. Effects of aminooxyacetic acid on postharvest characteristics of carnation. Acta Hort., 113: 59-64.

Gilbart, D.A. and Dedolph, R.R., 1965. Phytokinin effects on respiration and photosynthesis in roses and broccoli. Proc. Amer. Soc. Hort. Sci., 86: 774-778.

Griffing, B., 1956. Concept of general and specific combining ability in relation to diallel crossing systems. Australian J. of Biol. Sci., 9: 463-493.

Halevy, A.H., 1976. Treatments to improve water balance of cut flowers. Acta Hort., 64: 223-230.

Hallauer, A.R. and Miranda, J.B., 1981. Quantitative Genetics in Maize Breeding. Iowa State University Press, Ames, Iowa, 480 pp.

Harding, J., Byrne, T.G. and Nelson, R.L., 1981. Heritability of cut-flower vase longevity in gerbera. Euphytica, 30: 653-657.

Heide, O.M. and 0ydvin, J., 1969. Effects of 6-benzylaminopurine on the keeping quality and respiration of glasshouse carnations. Hort. Res., 9: 26-36.

Henderson, C.R., 1953. Estimation of variance and covariance components. Biometrics, 9: 226-252.

Kalkman, E.Ch., 1985. Postharvest treatment of Astilbe hybrida. Acta Hort., 181: 389392.

Kofranek, A.M. and Halevy, A. H., 1981. Chemical pretreatment of chrysanthemums before shipment. Acta Hort., 113: 89-96.

Kofranek, A.M. and Paul, J.L., 1975. The value of impregnating cut stems with high concentration of silver nitrate. Acta Hort., 41: 199-206.

Kohl, H.C., 1968. Gerberas: Their culture and commercial possibilities. S. Flor. and Nur., 28: 18,24-26.






82


Kuc, R. and Workman, M., 1964. The relation of maturity to the respiration and keeping quality of cut carnations and chrysanthemums. Proc. Amer. Soc. Hort. Sci., 84: 575-581.

La Malfa, G. and Noto, G., 1978. Variations in the qualitative characteristics of Gerbera jamesonii hbrida flowers in relation to environmental conditions. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.233-244.

Larsen, F.E. and Frolich, M., 1969. The influence of 8-hydroxyquinoline citrate, ndimethylamino succinamic acid, and sucrose on respiration and water flow in 'Red Sim' cut carnations in relation to flower senescence. J. Amer. Soc. Hort. Sci., 94: 289-292.

Leeuwen, P.J. van, 1986. Post-harvest treatment of Euphorbia fulgens. Acta Hort., 181: 467-472.

Lineberger, R.D. and Steponkus, P.L., 1976. Identification and localization of vascular occlusions in cut roses. J. Amer. Soc. Hort. Sci., 101: 246-250.

MacLean, D.C. and Dedolph, P.L., 1962. Effects of N6-benzylaminopurine on postharvest respiration of Chrysanthemum morifolium and Dianthus cIarvoyllus. Bot. Gaz., 124: 20-21.

Marousky, F.J., 1969. Vascular blockage, water absorption, stomatal opening, and respiration of cut 'Better Times' roses treated with 8-hydroxyquinoline citrate and sucrose. J. Amer. Soc. Hort. Sci., 94: 223-226.

Marousky, F.J., 1972. Water relations, effects of floral preservatives on bud opening, and keeping quality of cut flowers. HortScience, 7: 114-116.

Marousky, F.J., 1973. Recent advances in opening bud-cut chrysanthemum flowers. HortScience, 8: 199-201.

Marousky, F.J., 1986. Vascular structure of the gerbera scape. Acta Hort., 181: 399406.

Marousky, F.J. and Nanney, J., 1972. Influence of storage temperatures, handling and floral preservatives on post harvest quality of gypsophila. Proc. Fla. State Hort. Soc., 85:419-422.

Marousky, F.J. and Woltz, S.S., 1975. Relationship of floral preservatives to water movement, fluoride distribution, and injury in gladiolus and other cut flowers. Acta Hort., 41: 171-182.

Maurer, J., 1968. Genetisch-ziichterische untersuchungen bei Gerbera iamesonii H. Bolus. Z. Pflanzenziichtung, 60: 113-143.

Mayak, S. and Dilley, D.R., 1976. Effect of sucrose on response of cut carnation to kinetin, ethylene, and abscisic acid. J. Amer. Soc. Hort. Sci., 101: 583-585.

Mayak, S. and Halevy, A.H., 1970. Cytokinin activity in rose petals and its relation to senescence. Plant Physiol., 46: 497-499.






83


Mayak, S. and Halevy, A.H., 1974. The action of kinetin in improving the water balance and delaying senescence processes of cut rose flowers. Physiol. Plant, 32: 330-336.

Mayak, S. and Halevy, A.H., 1980. Flower senescence, In: Senescence in Plants, K.V. Thimann (Editor), CRC Press, Boca Raton, Florida, 131-156.

Mayak, S., Halevy, A.H. and Katz, M., 1972. Correlative changes in phtyohormones in relation to senescence processes in rose petals. Physiol. Plant, 27: 1-4.

Mayak, S., Halevy, A.H., Sagie, S., Bar-Yoseph, A. and B. Bravdo, 1974. The water balance of cut rose flowers. Physiol. Plant, 31: 15-22.

Meeteren, U. van, 1978a. Water relations and keeping quality of cut gerbera flowers. I. The cause of stem break. Scientia Hort., 8: 65-74.

Meeteren, U. van, 1978b. Water relations and keeping quality of cut gerbera flowers. II. Water balance of ageing flowers. Scientia Hort., 9: 189-197.

Meeteren, U. van, 1979. Water relations and keeping quality of cut gerbera flowers. III. Water content, permeability, and dry weight of ageing petals. Scientia Hort., 11: 83-93.

Meeteren, U. van, 1980. Water relations and keeping quality of cut gerbera flowers. VI. Role of pressure potential. Scientia Hort., 12: 283-292.

Meeteren, U. van, 1981. Role of pressure potential in keeping quality of cut gerbera inflorescences. Acta Hort., 113: 143-150.

Meeteren, U. van and Gelder, H. van, 1980. Water relations and keeping quality of cut gerbera flowers. V. Role of endogenous cytokinins. Scientia Hort., 12: 273-281.

Mohsenin, N.N., 1970. Physical Properties of Plant and Animal Materials. Gordon and Breach Science Publication, New York, 123 pp.

Mutsenietse, G. Ya., Rashal', I. D. and Dishler, V. Ya., 1978. Inheritance of quantitative characters in gerbera in diallel crosses. II. Inflorescence characters. Genetika, 14: 779783.

Nichols, R., 1981. Ethylene, present and future. Acta Hort., 113: 11-18.

Nichols, R., 1982. Growth regulators and flower senescence. British Plant Growth Regulator Group, Monograph, 8: 113-120.

Nowak, J., 1981. Regulation of bud opening, storage period, vase-life and senescence of cut flowers. I. The effect of silver complexes and sucrose on longevity of cut gerbera inflorescences stored for different periods of time. Res. Inst. of Pom. and Floric., Ann. Report, 2: 3-10.

Nowak, J. and Plich, H., 1981. Regulation of bud opening, storage period, vase-life, and senescence of cut flowers. II. The effect of silver ions and other anti-ethylene agents on ethylene synthesis and senescence of gerbera inflorescences. Res. Inst. of Pom. and Floric., Ann. Report, 2: 10-21.






84


Ottaviano, E., Schiva, T., and Sari-Gorla, M., 1974. Analisi multivariata per lo studio delle differenze genetiche in popolazioni di Gerbera iamesonii. Genet. Agr. 28: 292-306.

Parups, E.V. and Molnar, J. M., 1972. Histochemical study of xylem blockage in cut roses. J. Amer. Soc. Hort. Sci., 97: 532-534.

Parups, E.V. and Voisey, P.W., 1976. Lignin content and resistance to bending of the pedicel in greenhouse-grown roses. J. Hort. Sci., 51: 253-259.

Paulin, A. and Muloway, K., 1979. Perspectives in the use of growth regulators to increase the cut flower vase life. Acta Hort., 91: 135-141.

Paull, R., Goo, T., Criley, R.A. and Parvin, P.E., 1981. Leaf blackening in cut Protea eimia: Importance of water relations. Acta Hort., 113: 159-168.
Rasmussen, H.P. and Carpenter, W.J., 1974. Changes in the vascular morphology of cut rose stems: A scanning electron microscope study. J. Amer. Soc. Hort. Sci., 99: 454-459.

Reddy, T.V., 1988. Mode of action of cobalt extending vase life of cut roses. Scientia Hortic., 36: 303-313.

Reid, M.S. and Kofranek, A.M., 1981. Recommendations for standardized vase-life evaluations. Acta Hort., 113: 171-174.

Rogers, M.N., 1973. An historical and critical review of postharvest physiology research on cut flowers. HortScience, 8: 189-194.

SAS Institute Inc., 1982. SAS User's Guide: Statistics, 1982 Edition. SAS Institute Inc., Cary, N.C., 584 pp.

Schaffer, H.E. and Usanis, R.A., 1969. General least squares analysis of diallel experiments. A computer program-DIALL. Genetics Dept., North Carolina State Univ., Raleigh, N.C. 27607.

Schiva, T., 1973. Analisi dei caratteri metrici di importanza commerciale su Gerbera iamesonii e conseguenze dell'inincrocio. Estrato dagli Annali della facolth di Scienze Agrarie dell'Universith degli Studi di Torino, 8: 355-366.

Schiva, T., 1975. Miglioramento genetico dell gerbera. II. Analisi genetica di caratteri metrici di importanza commerciale e conseguenze dell'inincrocio. Genet. Agr., 29: 233240.

Serini, G. and De Leo, V., 1978. Phenotypic characters and preservability of gerbera flowers. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 269-277.

Siegelman, H.W., 1952. The respiration of rose and gardenia flowers. Proc. Amer. Soc. Hort. Sci., 59: 496-500.

Simmonds, N.W., 1979. Principles of Crop Improvement. Longman Group Limited, London, 408 pp.

Smith, D.E. and Nelson, R.L., 1967. Gerbera propagation. Calif. Agric., 21(12): 7.






85


Staden, O.L. and Beekhuizen, J.G., 1986. A new formula for the anti-ethylene pretreatment of cut carnations. Acta Hort., 181: 425-428.

Stimart, D.P. and Brown, D.J., 1982. Regulation of postharvest flower senescence in Zinnia elegans Jacq. Scientia Hortic., 17: 391-396.

Stoddard, E.M. and Miller, P.M., 1962. Chemical control of water loss in growing plants. Science, 137: 224-225.

Swart, A., 1981. Quality of Lilium 'Enchantment' flowers as influenced by season and silver thiosulphate. Acta Hort., 113: 45-50.

Sytsema, W., 1975. Conditions for measuring vase life of cut flowers. Acta Hort., 41: 217-226.

Sytsema, W., 1981. Vase life and development of carnations as influenced by silver thiosulphate. Acta Hort., 113: 33-38.

Sytsema, W., 1986. Post-harvest treatment of freesia with silver thiosulphate and cytokinins. Acta Hort., 181: 439-442.

Tandler, J., Mor, Y., Spiegelstein, H. and Mayak, S., 1986. Chemical treatments to improve the quality of cut gypsophila flowers. Acta Hort., 181: 443-450.

Tesi, R., 1978. Variation of some characters of flowers in clones of Gerbera iamesonii lybrida. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.227-232.

Waters, W.E., 1964. Influence of chemical preservatives on keeping quality of asters, carnations, chrysanthemums, and gerbera daisies. Proc. Fla. St. Hort. Soc., 77: 466-470.

Wilberg, B., 1973. Physiological investigations on the problem of freedom from stem break as a prerequisite in breeding gerbera forms which remain upright when used as cut flowers. Z. Pflanzenzuchtung, 69: 107-114.

Woltering, E.J., 1986. Postharvest treatment of GvDsohila paniculata. Acta Hort., 181: 338.

Wricke, G., Weber, W. E. and Ottleben, R., 1982. Die analyse der genetischen varianz bei quantitativ vererbten Merkmalen von Gerbera iamesonii. Z. Pflanzenziichtung, 89: 329-336.

Wulster, G., Sacalis, J. and Janes, H.W., 1982. Senescence in isolated carnation petals. Plant Physiol., 70: 1039-1043.

Zagory, D. and Reid, M.S., 1986. Evaluation of the role of vase micro organisms in the postharvest life of cut flowers. Acta Hort., 181: 207-218.

Zieslin, N., Kohl, H.C., Kofranek, A.M. and Halevy, A.H., 1978. Changes in the water status of cut roses and its relationship to bent-neck phenomenom. J. Amer. Soc. Hort. Soc., 103: 176-179.






























APPENDIX A
VASE LIFE STUDIES











A-1. Vase life data for thirty-one parents from a population of gerbera.
(Parental Generation) (Top-cross Generation) Parent Mean C.V. Progeny Mean
83-1-6 13.0 20.4 6.8 83- 1-13 11.3 5.1 6.8 83-1-26 16.0 16.6 7.7 83-1 -32 12.3 4.7 8.2 83- 1-61 12.3 9.4 6.8 83-1-64 15.0 17.6 7.9 83-1-74 13.7 23.5 6.9 83-1-77 15.3 24.7 8.6 83-3-4 11.7 9.9 7.5 83-3-34 13.3 11.5 7.2 83-3-38 15.7 7.4 8.7 83-3-47 13.0 20.4 7.4 83-3-77 13.0 7.7 6.4 83-3-96 13.3 11.5 7.4 83-4-4 11.7 5.0 6.1 83-4-17 12.3 20.4 6.3 83-4-38 11.3 5.1 7.8 83-4-69 12.0 8.3 9.6 83-5-33 14.0 7.1 7.6 83-5-39 16.0 22.5 10.4 83-5- 109 13.0 21.7 8.9 83-6-42 14.3 17.6 7.9 83-7-4 12.3 18.7 10.2 83-7-6 13.7 16.9 7.8 83-7-10 16.0 0.0 10.7 83-7-14 12.7 18.2 6.6 83-7- 18 14.3 10.7 6.9 83-7-26 12.3 18.7 9.4 83-7-29 16.0 0.0 9.4 83-7-47 12.7 4.6 8.6 83-8-7 11.3 5.1 7.6


87






88















A-2. 5 x 5 Diallel. Vase life means based on progeny means of gerbera crosses.

83-1-77 83-4-69 83-5-109 83-7-4 83-7-10

83-1-77 R = 9.2 x = 9.3 i = 12.5 i = 9.8 n=13 n=18 n=19 n=13 83-4-69 i = 11.0 x = 8.5 = 12.2 i = 11.7
n = 19 n = 16 n = 18 n =21

83-5-109 R = 8.6
n = 12

83-7-4 i = 11.8 R= 11.1 i = 11.0 i = 13.4
n = 21 n=9 n = 22 n = 25 83 7 10 R = 15.5 i = 14.0 x = 12.5
n=3 n=1 n=18





89




A-3. Parental Generation. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera.
Stem Inflorescence Disc Traits Diameter Diameter Diameter Vgrowth

Stem Length .07 .16* .20 * .06 *

Stem Diameter .10 * .64 * .26 * Inflorescence Diameter .23 *-.07 *

Disc Diameter .17 *
* Significant at P< 0.05. * Significant at P< 0.01.



A-4. Parental Generation. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera by senescence mode.
Stem Inflorescence Disc Traits Diameter Diameter Diameter Vgrowth

Stem Length
Bending .07 .11 * .18** .07 Folding .12 .31* .35** -.02 Wilting .16* .21 ** .29** .09

Stem Diameter
Bending .15 * .63 * .28 *
Folding .15 .58 * .07
Wilting -.01 .76 * .14* Inflorescence Diameter
Bending .27 * -.02
Folding .34* -.28 ** Wilting .06 -.20 **

Disc Diameter
Bending .20 *
Folding -.08 Wilting .09
Significant at P< 0.05. * Significant at P< 0.01.





90







A-5. Gerbera X hvbrida Hort.. Vase life data for backcrosses.
Backcross Number of Progeny Mean Vase Life
84-27-2 x 83-1-77 13 10.4 84- 27 -2 x 'Appleblossom' 2 8.0 84-27-4 x 83-1-77 9 10.1 84- 27 -4 x 'Appleblossom' 6 8.3 84-27-7 x 83-1-77 1 3.0 84-27-12 x 83-1-77 1 13.0 84-37-4 x 83-4-69 5 9.4 84- 37 -4 x 'Appleblossom' 19 8.4 84-37-10 x 83-4-69 12 9.3 84- 37 10 x 'Appleblossom' 16 9.9 84-37-12 x 83-4-69 21 9.6 84- 37 12 x 'Appleblossom' 11 9.6 84-41-3 x 83-7-4 19 10.7 84-41 3 x 'Appleblossom' 1 10.0 84-41-5 x 83-7-4 1 13.0 84-41-8 x 83-7-4 1 10.0 84- 41 13 x 'Appleblossom' 16 10.0 84-41-15 x 83-7-4 22 13.7 84-41-15 x 'Appleblossom' 1 13.5 84-41-19 x 83-7-4 1 12.0 84-41 20 x 'Appleblossom' 3 8.3 84-43-2 x 83-7-10 27 12.4 84-43 2 x 'Appleblossom' 11 8.9 84- 43 -5 x 'Appleblossom' 7 9.5 84-43-18 x 83-7-10 19 11.2 84 43 18 x 'Appleblossom' 2 9.3






91






A-6. Gerbera X hbrida Hort.. Frequency of senescence modes for backcrosses.
Number of % % % Backcross Flowers Bending Folding Wilting
84-27-2 x 83-1-77 30 26.7 40.0 33.3 84- 27 2 x 'Appleblossom' 6 16.7 33.3 50.0 84-27-4 x 83-1-77 13 30.8 23.1 46.2 84 27 4 x 'Appleblossom' 12 8.3 58.3 33.3 84-27-7 x 83-1-77 1 100 84-27-12 x 83-1-77 1 100 84-37-4 x 83-4-69 8 12.5 12.5 75.0 84- 37 -4 x 'Appleblossom' 33 33.3 24.2 42.4 84-37-10 x 83-4-69 22 40.9 22.7 36.4 84- 37 10 x 'Appleblossom' 34 26.5 47.1 26.5 84-37-12 x 83-4-69 35 11.4 51.4 37.1 84- 37 12 x 'Appleblossom' 20 15.0 50.0 35.0 84-41-3 x 83-7-4 34 41.2 58.8 84 -41 3 x 'Appleblossom' 1 - 100 84-41-5 x 83-7-4 1 - 100 84-41-8 x 83-7-4 1 100 84-41 13 x 'Appleblossom' 26 23.1 30.8 46.2 84-41-15 x 83-7-4 38 2.6 26.3 71.1 84-41 15 x 'Appleblossom' 2 - 100 84-41-19 x 83-7-4 1 - 100 84-41-20 x 'Appleblossom' 3 33.3 33.3 33.3 84-43-2 x 83-7-10 62 14.5 24.2 61.3 84- 43 2 x 'Appleblossom' 32 28.1 37.5 34.4 84- 43 5 x 'Appleblossom' 19 31.6 26.3 42.1 84-43-18 x 83-7-10 30 20.0 40.0 40.0 84-43 18 x 'Appleblossom' 3 66.7 33.3






92












A-7. Gerbera X hybrid Hort.. Vase life data for full-sib crosses.
Cross Number of Progeny Mean Vase Life
84-27-2 x 84-27-12 10 9.7 84-27-4 x 84-27-12 2 9.5 84-37-4 x 84-37-10 3 9.7 84-37-12 x 84-37-10 8 10.6 84-41-3 x 84-41-5 13 9.5 84-41-5 x 84-41-8 7 8.1 84-41-8 x 84-41-13 4 6.8 84-41-13 x 84-41-8 9 9.2 84-41-15 x 84-41-5 23 9.6 84-41-15 x 84-41-8 1 14.7 84-41-20 x 84-41-13 3 12.7 84-43-2 x 84-43-3 20 11.0 84-43-5 x 84-43-3 1 15.0 84-43-7 x 84-43-3 22 9.6 84-43-18 x 84-43-3 1 14.0






93












A-8. Qerbera X hvbrida Hort.. Frequency of senescence modes for full-sib crosses.
Number of % % Cross Flowers Bending Folding Wilting
84-27-2 x 84-27- 12 25 36.0 32.0 32.0 84-27-4 x 84-27-12 2 100 84-37-4 x 84-37-10 6 16.7 83.3 84-37- 12 x 84-37-10 16 12.5 43.7 43.7 84-41-3 x 84-41-5 23 21.7 78.3 84-41-5 x 84-41-8 13 15.4 30.8 53.8 84-41-8 x 84-41-13 5 100 84-41-13 x 84-41-8 21 9.5 57.1 33.3 84-41-15 x 84-41-5 45 11.1 17.8 71.1 84-41-15 x 84-41-8 3 66.7 33.3 84-41-20 x 84-41-13 6 16.7 33.3 50.0 84-43-2 x 84-43-3 46 17.4 28.3 54.3 84-43-5 x 84-43-3 1 100 84-43-7 x 84-43-3 42 19.0 35.7 45.2 84-43-18 x 84-43-3 2 - 100




Full Text
10
(1973) suggested transpiration rates have a close relationship with water uptake. Leaf area
of cut roses was found to be the factor most closely associated with water uptake in both
light and dark. Carpenter and Rasmussen (1974) concluded the number of stomates on the
leaves and stems of cut flowers affects the rates of water uptake and loss. They observed
when water loss from transpiration exceeds water uptake, vase life is reduced. It was
proposed that by reducing transpiration, bent neck in roses would be reduced.
Mayak and Halevy (1974) found kinetin delayed wilting in cut roses. It was observed
that water uptake increased, though transpiration also increased, due to stomatal opening.
Bravdo et al. (1974) studied vase life in gladiolus. They proposed water uptake can be
improved by increasing osmotic concentration of florets and leaves with the addition of
sucrose to vase solutions. Marousky and Woltz (1975) also examined vase life of
gladiolus. Though the addition of 8-hydroxyquinoline citrate plus sucrose was found to
improve water uptake, the potential for fluoride toxicity was also increased depending on
the concentration of fluoride in the water. They also reported that gerberas were highly
sensitive to low levels of fluoride, especially with the addition of 8-hydroxyquinoline
citrate plus sucrose to the vase solution which increased water uptake. Chrysanthemums
and snapdragons did not exhibit this degree of sensitivity. Halevy (1976) stated water
stress is the most common reason for decreased vase life in cut flowers. He asserted the
basis for water imbalance is a decrease in water potential, water uptake, water loss, and
water conductivity. Zieslin et al. (1978) postulated bent neck in cut roses occurs due to
water stress conditions such as increased transpiration rates and decreased water uptake
rates. Reddy (1988) suggested cobalt may increase vase life of cut roses by causing
stomates to partially close, thereby reducing transpiration, yet maintaining water uptake.
A major reason cited for poor vase life in gerbera is the inability of the cut flowers to
imbibe sufficient water (Kohl, 1968). It was suggested as gerbera flowers age, water
holding capacity of the petals decreases (Meeteren, 1978b). As water content decreased in
gerbera petals over time, ion leakage increased (Meeteren, 1979). It was proposed that by


TABLE OF CONTENTS
ACKNOWLEDGMENTS iii
ABSTRACT vi
CHAPTERS
1 GENERAL INTRODUCTION 1
2 LITERATURE REVIEW 3
3 PART I. VASE LIFE STUDIES 20
Introduction 20
Materials and Methods 21
Plant Material 21
Selection and Mating. 23
Vase Life Evaluation 25
Quantitative Analysis 26
Results 28
Selection and Mating. 28
Heritability 29
Senescence Patterns 31
Correlations between Vase Life and Other Traits 38
Discussion 42
Conclusioa 46
4 PART II. STEM STRENGTH STUDIES 48
Introduction 48
Materials and Methods 48
Selection and Mating. 48
Stem Strength Evaluation 49
Production 51
Quantitative Analysis 51
Results 53
Selection and Mating. 53
Heritability 56
Correlations between Stem Strength and Other Traits 58
Discussion 59
Conclusioa 63
5 PART III. VASE LIFE X STEM STRENGTH STUDIES 64
Introduction 64
Materials and Methods 65
iv


D-L5x5D^kLO^er^ait_me^Sibas^_on_£ro|en^means_ofjer^acro^si
83-1-77
83-4-69
83-5-109
83-7-4
83-7-10
X! = 56.95
X! = 54.77
X! = 52.67
X! = 52.95
x2 = .45
x2 = .58
x2 = .46
x2 = -49
83-1-77
x3 = 10.18
x3 = 10.59
x3 = 10.16
x3 = 9.81
X4 = 1.42
X4 = 1.85
x4= 1.74
X4 = 1.61
x5= 1.17
x5= 1.94
x5= 1.81
x5= 1.59
n= 13
n = 18
n = 19
n = 13
xj = 54.37
X! = 60.46
X! = 49.48
xj = 55.18
x2 = .43
x2 = .55
x2 = .46
x2 = .45
83-4-69
x3 = 9.86
x3 = 10.04
x3= 9.11
x3 = 8.50
X4 = 1.40
X4 = 1.83
X4 = 1.54
X4 = 1.51
x5 = .84
X5 = 1.30
x5= 1.59
X5 = 1.06
n= 19
n = 16
n = 18
n = 21
x¡ = 50.63
x2 = -62
Q'l S 10Q
1 n 8s
oj j i\jy
1 v OJ
X4 = 1.80
X5 = 1.63
n= 12
xi = 47.60
X! = 50.69
X! = 48.73
X! = 49.32
x2 = .46
x2= .51
x2 = .55
x2 = .45
83-7-4
x3 = 9.90
x3= 9.05
x3= 9.79
x3 = 8.80
x4 = 1.66
x4= 1.73
x4 = 2.00
x4= 1.77
x5= 1.69
x5= 1.71
X5 = 1.65
x5= 1.57
n = 21
n = 9
n = 22
n = 25
x¡ = 45.83
X! = 43.00
xj = 54.96
x2 = .46
x2 = .36
x2 = .54
83-7-10
x3= 9.00
x3= 7.50
x3= 9.33
x4 = 1.53
x4= 1.10
X4 = 1.83
x5 = 2.00
x5 = .50
x5= 1.77
n = 3
n = 1
n = 18
xj = stem length
x2 = stem diameter
x3 = inflorescence diameter
X4 = disc diameter
X5 = stem elongation
n = number of progeny
103


8
respiration rate could be the cause for senescence in cut flowers. The answer to the
question, Does respiration rate affect cut flower senescence? remains elusive.
Siegelman (1952) demonstrated in experiments with roses and gardenias that when
respiration rate was reduced, vase life increased. MacLean and Dedolph (1962) observed
application of N6-benzylaminopurine (Verdan) reduced respiration rate in carnations and
chrysanthemums and increased vase life. Kuc and Workman (1964) concluded a direct
relationship existed between respiration rate and postharvest longevity in cut flowers
because the respiration rate was three to four times greater in carnations than in
chrysanthemums and postharvest longevity was much lower in carnations than
chrysanthemums. Coorts, et al. (1965) reported that respiration rates increased in cut roses
after being treated with a flower preservative. Treated roses lasted longer than those which
were untreated. Gilbart and Dedolph (1965) treated cut roses with N6-benzyladenine (N6-
BA). They observed the respiration rate generally increased in petals and decreased in
leaves following treatment. Ballantyne (1966) asserted variation in the respiration rate of
daffodil cut flowers was more likely a result of senescence than a cause for senescence.
Heide and 0ydvin (1969) suggested decreasing respiration rate in carnations will not delay
senescence. Larsen and Frolich (1969) studied the relationship between respiration rate
and water uptake in cut flowers. They observed that water uptake and respiration
decreased simultaneously in carnation Red Sim. Coorts (1973) reviewed factors
contributing to metabolic changes in cut flowers which affect senescence. He concluded
senescence can be delayed by using respiratory inhibitors and controlling hydrogen ion
activity. It was proposed that if the pH of vase solutions was maintained between 3.0 -
5.0, vase life would be extended. It was also suggested mitochondrial activity may be
linked to respiration rates of senescing cut flowers. Mayak and Halevy (1980) reviewed
the subject of flower senescence and suggested that petal senescence would be an ideal
system to study. Ferreira and Swardt (1981b) reported they found no correlation between
respiration rate and vase life of cut roses. Changes in respiration rate depended on the


APPENDIX A
VASE LIFE STUDIES


76
stronger stems exhibiting less incidence of folding could be because stronger stems have
fewer small vascular bundles.
Dubuc-Lebreux and Vieth (1985) observed that a greater proportion of support tissues,
such as sclerenchyma, was found in lower regions of the stem where breakage did not
normally occur. They suggested that plants whose stems showed less incidence of folding
at the typical zone of folding (10-15 cm below the peduncle) might have more support
tissues in this area compared to plants whose stems showed more incidence of folding.
Perhaps in stonger stems, small vascular bundles are displaced by more support tissue;
therefore, the fewer small vascular bundles in a stem cross-section, the greater proportion
of support tissue will be present
Conclusion
It is concluded that ifvase life and stem strength are two distinct components of
postharvest longevity in Gerbera X hvbrida Hort., then the breeder wishing to improve
postharvest quality in this cut flower might consider selecting plants with good combining
ability for both high vase life and high stem strength.
Stem anatomy studies yielded interesting results. It appears that vase life is affected by
the number of large vascular bundles and stem strength is affected by the number of small
vascular bundles. Fewer bundles of each type per unit of circumference were found in
plants with high vase life and high stem strength.
Additional genetic investigations to determine the heritability of stem anatomy variables
involving number of large vascular bundles and stem anatomy variables involving number
of small vascular bundles are recommended.


9
stage of senescence. Amariutei et al. (1986) observed higher respiration rates in gerbera
cut flowers that were treated with a pulsing solution.
The role of plant hormones, particularly ethylene, has also been a source of
investigation for understanding cut flower senescence. Mayak and Halevy (1970) reported
exogenous application of N^-BA can delay senescence in rose petals. Mayak et al. (1972)
demonstrated exogenous applications of cytokinins increase postharvest longevity in roses
whereas ethylene or abscisic acid (ABA) was shown to decrease postharvest longevity.
They also suggested the presence of ethylene may trigger the production of ABA. Rogers
(1973) reviewed the effects of ethylene synthesis and other plant hormones on cut flowers
and concluded that anti-senescence factors must be applied to attain maximum vase life in
cut flowers.
Nichols (1981) summarized results that indicated ethylene biosynthesis plays an
important role in flower senescence, but the magnitude of its effect depends upon plant
genera. Nowak (1981) stated a decrease in vase life of cut gerbera flowers may be caused
by ethylene. Wulster et al. (1982) discussed the possibility of auxins increasing ethylene
biosynthesis, thereby inducing senescence in carnation petals. Nichols (1982) reviewed
the effects of growth regulators on cut flower senescence and concluded inhibiting
endogenous production of ethylene will increase vase life in carnation, iris, daffodil, and
chrysanthemum. Bufler (1986) stressed that studies should include investigating factors
which control response to ethylene production, not just a measure of ethylene production.
Reddy (1988) suggested cobalt salts may inhibit ethylene synthesis in cut roses.
A major factor considered to influence senescence in cut flowers, including gerbera, is
the balance between water uptake and transpiration. Marousky (1969) verified reducing
moisture stress by increasing water absorption will improve postharvest longevity in roses.
Burdett (1970) proposed water loss as one of the causes of bent neck in cut roses. He
specified water deficits resulting from transpiration are probably of small importance
compared to an impairment in the water conducting system. Carpenter and Rasmussen


C-LPheTOt^£c_COTelation_c^fficients_tetween_stem^nd_innorescenceffaitsJn_£ertera.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
.05
.06
.08
Stem Diameter @ 12 cm
.76**
.87 **
Disc Diameter
.80 **
* Significant at P< 0.01.
C-2. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera
by vase life (VL) ratings.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
VL = high
-.09
-.05
-.04
VL = low
Stem Diameter @ 12 cm
.22*
.19
.22* *
VL = high
.78 **
.88 **
VL = low
Disc Diameter
.78 **
.89 **
VL = high
.12**
£
_o
n
>
.86**
* Significant at P< 0.05. *
* Significant at P< 0.01.
C-3. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera
by stem strength (SS) ratings.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
SS = high
.10
.12
.16*
SS = low
Stem Diameter @ 12 cm
.11
.20 *
.28 **
SS = high
.70**
.87 **
SS = low
Disc Diameter
.84 *
.89 **
SS = high
.80 **
SS = low
.78 **
* Significant at P< 0.05. *
* Significant at P< 0.01.
98


21
De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in
gerberas using a diallel mating scheme involving 12 parents and their progenies. They
distinguished vase life by different modes of senescence; petal wilt and stem fold. General
combining ability was significant for days to wilting or folding and percent folding. De
Jong (1986) concluded the main difficulty which remains in breeding to improve
postharvest longevity is large intraplant variation for both days to wilt and percent folding.
Plant breeders can benefit from knowing the relationship between characteristics which
they are trying to improve. Although breeding to improve postharvest longevity in gerbera
has been suggested as a viable possibility for many years (Smith and Nelson, 1967), few
studies of gerbera have reported correlations between vase life and other traits.
Tesi (1978) showed a significant negative correlation between vase life and cut flower
yield. Serini and De Leo (1978) found no correlation between vase life and stem length,
inflorescence diameter, and number of ligulae. Nowak and Plich (1981) observed vase life
of cut gerberas increased when stems were shorter.
The objectives of this research on gerbera, using a broad based source of germplasm,
were to determine broad sense heritability and narrow sense heritability for vase life by
diallel analysis, observe changes in frequency of senescence patterns due to selection and
mating, and to determine correlations between vase life and other traits of flower and stem
morphology and growth.
Materials and Methods
Plant Material
Germplasm was randomly collected from several sources. Tissue cultured plantlets
obtained were European cultivars or selections from a commercial breeding program at
Sunshine Carnations in Hobe Sound, Florida. A list of these cultivars is given in table 3-1.


71
stem circumference (1.0 cm) at 12 cm below the peduncle (SBUNCRAT). The means for
number of small vascular bundles per unit circumference in gerbera stem sections were
significantly different for plants with high stem strength compared to those with low stem
strength. Plants with high stem strength exhibited a smaller number of small vascular
bundles per unit circumference than plants with low stem strength. A summary of these
results is given in tables 5-8 and 5-9.
Table 5-8. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm
scape circumference.)
TBUNCRAT
LBUNCRAT
SBUNCRAT
Source
df
M.S.
M.S.
M.S.
Vase Life
(high vs low)
1
1467.31 *
742.67 *
122.18 ns-
Stem Strength
(high vs low)
1
1108.91 *
61.55 n s-
647.96 *
Vase Life x Stem Strength
1
143.85 n-s-
69 n.s.
164.47
Error
18
237.65
80.43
124.06
* Significant at P< 0.05. Significant at P< 0.01.
Table 5-9. Means2 for number of vascular bundles per 1.0 cm scape circumference
between hi
gh and low levels of vase life and stem strength in gerbera.
TBUNCRAT LBUNCRAT SBUNCRAT
Vase
Life
High 30.46a 10.02a 20.44a
Low 34.81b 13.11b 21.70a
Stem
Strength
High 30.75a 11.12a 19.63a
Low 34.52b 12.01a 22.51b
2 Means fo
lowed by different letters within TBUNCRAT and SBUNCRAT for vase life
or stem strength are significant at the 5 % level and significant at the 1% level within
LBUNCRAT.


13
Two pathways for water uptake in gerbera were proposed: a direct path through xylem
vessels and an indirect path through the stem cavity. It was suggested stem break occurs
when water uptake is inhibited by bacterial growth (Meeteren, 1978a).
Some researchers have studied stem anatomy in an attempt to understand the causes
for reduced postharvest longevity in gerbera. Reiman-Philip, as cited by Wilberg (1973),
noted a larger proportion of large vascular bundles to small vascular bundles as a possible
factor contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that
such a relationship existed in only extreme cases. Steinitz, as cited by Marousky (1986)
observed increases in phloem cell wall thickening and lignification in gerbera scapes when
flowers were placed in a sucrose solution. In a comprehensive study, Marousky (1986)
measured the size and number of vascular bundles in two gerberas, Tropic Gold and
Appleblossom. He concluded Appleblossom exhibited more resistance to stem bending
than Tropic Gold, partially because of various anatomical features such as fewer small
vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference,
and a greater percentage of dry weight per unit of scape length. It was emphasized these
factors must be considered concurrently with variation in moisture stress to understand the
cause of stem breakage.
Dubuc-Lebreux and Vieth (1985) also studied the histology of the gerbera stem. They
postulated stem bending is linked to deficiencies in supportive elements in the stem. It was
concluded that sensitivity to stem breakage depends on the degree of maturity of the stem
approximately 10 cm. below the flower head at time of harvest. Marousky (1986) also
noted lignification was greater in the lower portion of the stem compared to the upper
portion of the stem where stem breakage is most likely to occur.
Despite postharvest treatments favorably affecting lasting quality in cut flowers,
including gerbera, reliable expectations of vase life have not been achieved. Though
metabolic and anatomical research has provided some clues to the cause of senescence in
cut flowers, another approach for learning how to improve postharvest longevity exists.


43
deleting a significant source of environmental variation. In that case, the highest possible
estimate of heritability could be obtained for a given population.
Using this method, predicted estimates of heritability for vase life in gerbera ranged
from 22 to 39 percent with n specified as 1, 2, 3, 5, and flowers per plant. Since these
calculations assumed no change in genotypic effects, broad sense and narrow sense
heritability estimates for each case were approximately equal. It is interesting to note that in
spite of varying the magnitude of environmental variance, the range of estimates remained
within proximity to those determined by other researchers (Serini and De Leo, 1978; Tesi,
1978; and Harding et al., 1981).
The repeatability estimate for vase life (r = .57) is moderately high, indicating that two
to three flowers per plant is adequate for determining the average vase life per plant.
Falconer (1960) recommends that further gain in accuracy by more than two measurements
does not justify additional expense or time required to collect more data when repeatability
is high. This was proven by comparing the relative increase in heritability from predicted
estimates based on 1,2, 3, 5, and < measurements per plant. Between one and three
measurements, heritability increased by nine percent, while beyond three measurements
through infinity, the gain in heritability was only eight percent.
Thus far, this research has confirmed that improvement of postharvest longevity in
gerbera can be obtained by selecting and mating plants with good combining ability for
high vase life. The overall population mean for vase life resulting from a diallel cross
among five plants selected, based on their combining ability with 'Appleblossom,' a
cultivar with high vase life, yielded an increase of more than three days. Additional
information has been gained, however, by classifying vase life determinations by three
distinct modes of senescence: bending, folding, and wilting. The frequency of vase life
days based on data from all flowers was not normally distributed before or after selection
and mating. Yet, when frequency of vase life days was classified by senescence mode,
distribution was normal. Moreover, after selection and mating, increases in mean vase life


11
increasing pressure potential in gerbera petal cells, ion leakage will decrease, resulting in
longer vase life (Meeteren, 1980).
Researchers have also focused on the presence of vessel occlusion in the stem of cut
flowers as a cause for early senescence. Two types of occlusion have been proposed.
Occlusion due to microbes is considered to be a primary cause of bent neck in roses or stem
break in gerberas. The use of a bactericide in vase solutions is a common practice in
postharvest handling of cut flowers. The second type of occlusion is regarded as a result
of physiological plugging.
Durkin and Kuc (1966) postulated a vascular block resulting from harvesting injury
was the primary cause for premature senescence in cut roses. Marousky (1969) concurred
vascular blockage was responsible for decreasing the vase life of cut roses. Burdett (1970)
showed bent neck in roses coincided with the appearance of material which plugged xylem
vessels. He postulated two possibilities for plugging: growth of microorganisms or a gum
deposition which could be the result of pectin degradation products. It was also suggested
when water uptake is deficient, sufficiently lignified stem tissue could prevent bent neck in
cut roses. After eliminating the presence of microorganisms, Marousky (1972)
demonstrated that in addition to moisture deficiency, a blocking mechanism reduces
postharvest longevity in cut flowers. Parups and Molnar (1972) investigated the nature of
vascular blockage in cut roses histochemically. They reported evidence of carbohydrates
with sulfate, carboxyl, or phosphoryl groups, pectin-, lipid-, or other protein-like
compounds, and some enzymes. They found no evidence of tannins, lignin, and callse in
blocked xylem vessels. Rogers (1973) reviewed the literature concerning effects of
physiological or microbial induced stem plugging but concluded further study was
necessary to determine the cause of physiological plugging. He also suggested microbial
plugging may have significance only with flower genera that typically last longer since it
takes time for microbial populations to develop.


40-,
67
y = 5.15 + .52x
r= .28
30-
M
Stem strength 20-
(F = g/.2cm)
10
0
0 10 20
Vase life (days)
Figure 5-1. Distribution of vase life and stem strength means for seventy-three plants
Comparison of Stem Anatomy Among Genotypes
Twenty two plants were selected and classified by high and low levels of vase life and
stem strength according to the categories described. Listing of these plants and their
corresponding mean vase life and stem strength is given in table 5-3.
Analyses of variance were made using plant means for nine stem anatomy variables.
Vase life and stem strength were treated as main effects. No significant interaction between
vase life and stem strength was determined for any of the nine variables. Analyses of
variance for the number of vascular bundles directly measured per stem section (TBUN,
LBUN, and SBUN) only yielded a significant difference between high and low levels of
vase life for number of large bundles (LBUN). The mean number of large vascular
bundles was significantly different for plants with high vase life compared to those with
low vase life. Plants with high vase life exhibited a smaller number of large vascular
bundles than plants with low vase life. A summary of these results is given in tables 5-4
and 5-5.
Analyses of variance for the number of vascular bundles calculated per unit stem area
(1.0 cm2) at 12 cm below the peduncle (TBUNARAT, LBUNARAT, and SBUNARAT),
only yielded significant differences between high and low levels of vase life for total


82
Kuc, R. and Workman, M., 1964. The relation of maturity to the respiration and keeping
quality of cut carnations and chrysanthemums. Proc. Amer. Soc. Hort. Sci., 84: 575-581.
La Malfa, G. and Noto, G., 1978. Variations in the qualitative characteristics of Gerbera
iamesonii hvbrida flowers in relation to environmental conditions. In: L. Quagliotti and A.
Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a
Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.233-244.
Larsen, F.E. and Frolich, M., 1969. The influence of 8-hydroxyquinoline citrate, n-
dimethylamino succinamic acid, and sucrose on respiration and water flow in Red Sim
cut carnations in relation to flower senescence. J. Amer. Soc. Hort. Sci., 94: 289-292.
Leeuwen, P.J. van, 1986. Post-harvest treatment of Euphorbia fulgens. Acta Hort., 181:
467-472.
Lineberger, R.D. and Steponkus, P.L., 1976. Identification and localization of vascular
occlusions in cut roses. J. Amer. Soc. Hort. Sci., 101: 246-250.
MacLean, D.C. and Dedolph, P.L., 1962. Effects of N6-benzylaminopurine on post-
harvest respiration of Chrysanthemum morifolium and Dianthus carvophvllus. Bot. Gaz.,
124: 20-21.
Marousky, F.J., 1969. Vascular blockage, water absorption, stomatal opening, and
respiration of cut Better Times roses treated with 8-hydroxyquinoline citrate and sucrose.
J. Amer. Soc. Hort. Sci., 94: 223-226.
Marousky, F.J., 1972. Water relations, effects of floral preservatives on bud opening, and
keeping quality of cut flowers. HortScience, 7: 114-116.
Marousky, F.J., 1973. Recent advances in opening bud-cut chrysanthemum flowers.
HortScience, 8: 199-201.
Marousky, F.J., 1986. Vascular structure of the gerbera scape. Acta Hort., 181: 399-
406.
Marousky, F.J. and Nanney, J., 1972. Influence of storage temperatures, handling and
floral preservatives on post harvest quality of gypsophila. Proc. Fla. State Hort. Soc.,
85:419-422.
Marousky, F.J. and Woltz, S.S., 1975. Relationship of floral preservatives to water
movement, fluoride distribution, and injury in gladiolus and other cut flowers. Acta Hort.,
41: 171-182.
Maurer, J., 1968. Genetisch-zlichterische untersuchungen bei Gerbera iamesonii H.
Bolus. Z. Pflanzenzlichtung, 60: 113-143.
Mayak, S. and Dilley, D.R., 1976. Effect of sucrose on response of cut carnation to
kinetin, ethylene, and abscisic acid. J. Amer. Soc. Hort. Sci., 101: 583-585.
Mayak, S. and Halevy, A.H., 1970. Cytokinin activity in rose petals and its relation to
senescence. Plant Physiol., 46: 497-499.


112
D-13. Summary of published heritability estimates for stem and inflorescence traits in
Source
Stem Length
h2 H2
Stem Diameter
h2 H2
Inflorescence Diameter
h2 H2
Disc Diameter
h2 H2
Maurer
1968
.30



.58

Borghi & Baldi
1970
.35



.30


Schiva
1973


.34
.57
.13
.68
Ottaviano et al.
1974

.57

.62

.24
.58
Schiva
1975
.06
.47
.50
.66



De Leo & Ottaviano
1978
.25

.61

.13

.70
Mutsenietse
1978


.35
.45
.45
.48
.52
.51
Tesi
1978

.34



.74

Wricke
1982

.77



.79
.69
Drennan et al.
1986


.36

.08

.23
D-14. Summary of published repeatability estimates for stem and inflorescence traits in
Source
Stem Length
r
Stem Diameter
r
Inflorescence Diameter
r
Disc Diameter
r
De Leo & Ottaviano
1978
.69
.79
.88
.84
Drennan et al.
1986

.61
.69
.54


38
H

Frequency
7%)
8.7 9.1 10.3 10.6 11.2 11.4 11.8 11.9 12.5 14.3
vase life progeny means
120
100
80
60
40
20
0
% bending
% folding
% wilting
Figure 3-6. Distribution of senesence mode frequency in progeny of diallel generation.
Correlations between Vase Life and Other Traits
Phenotypic correlations between vase life and five stem and inflorescence traits: stem
length, stem diameter, inflorescence diameter, disc diameter, and post-harvest stem
elongation (vgrowth), were determined for both parental and diallel generations.
Correlation coefficients based on data from individual flowers were obtained using a SAS
program for Pearsons product-moment correlation procedure (SAS, 1986).
Correlations determined from the parental generation demonstrated the linear
relationship between vase life and these traits in a random population before selection for
improving vase life. Correlations determined from the diallel generation demonstrated the
linear relationship between vase life and these traits after selection. Additionally,
correlations were made between the vase life of flowers that bent, folded, or wilted and the
five stem and inflorescence traits. These results are summarized in tables 3-15, 3-16, 3-17,
3-18, and 3-19.


100
C-5. Phenotypic correlation coefficients between stem anatomy and three traits in gerbera
by vase life (VL) ratings.
Stem Anatomy Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN VL = high
-.04
.16*
.16*
VL = low
.04
.09
-.01
SBUN VL = high
-.07
.09
.18*
VL = low
.12
.33 *
.35**
TBUN VL = high
-.08
.14
.23*
VL = low
.13
.36
.34**
LBUNCRAT VL = high
.02
-.39 *
-.45 *
VL = low
-.12
-.53 *
-.67 *
SBUNCRAT VL = high
-.01
-.44**
-.41 *
VL = low
.00
-.20 *
-.24 *
TBUNCRAT VL = high
.01
-.54 *
-.53 *
VL = low
-.06
-.42 *
-.53**
LBUNARAT VL = high
.07
-.63 *
-.70 *
VL = low
-.15*
-.65 *
-.77 *
SBUNARAT VL = high
.05
o
*
#
-.71 *
VL = low
-.08
-.57 *
-.63 *
TBUNARAT VL = high
.06
-.75 *
-.78 *
VL = low
0.13
-.67 *
-.77 *
* Significant at P< 0.05. Significant at P< 0.01.


I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Thomas J. Sheehan^ Chair
Professor of Horticultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
'Cochair
forticultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
0
"rancis J. Mari
Professor of Hi
¡cultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
David A. Knauft 1/
Associate Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Paul M. Lyrene (j
Professor of Horticultural Science


GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF GERBERA X HYBRIDA HORT.
By
HEIDI CAROL WERNETT
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1990


58
Table 4-9. 7x7 Diallel. Summary of variances for stem strength. (F = g/.2 cm) n
= 2.66
Genotypic Variance
VA vG
vE
Environmental Variance
VEg
Ves
Phenotypic Variance
VP(n)
103.64
116.76
158.22
103.31
146.06
274.98
Table 4-10. 7x7 Diallel. Predicted estimates of heritability for stem strength.
(F = g/.2 cm)
Heritability
1
2
Number of Measurements
3 5
oo
h2
.28
.35
.39 .42
.47
H2
.32
.40
.43 .47
.53
Correlations between Stem Strength and Other Traits
Phenotypic correlations between stem strength and three stem and inflorescence traits,
stem length, stem diameter, and inflorescence weight, were determined for both parental
and diallel generations. Correlation coefficients based on data from individual flowers
were obtained using a SAS program for Pearsons product-moment correlation procedure
(SAS, 1986).
Correlations determined from the parental generation demonstrated the linear
relationship between stem strength and these traits in a random population before selection
for improving stem strength. Correlations determined from the diallel generation
demonstrated the linear relationship between stem strength and these traits after selection.
These results are summarized in table 4-11.
Stem strength and stem length. Significant negative correlations were observed
between stem strength and stem length, before and after selection. It appeared, however,


53
variance (VEg) and special environmental variance (V^) in order to determine the
phenotypic variance for each case (Vp(n)).
Vp = VG + VE
VE = vEg + vEs
VP(n) = VG + VEg + ~ VEs
(Falconer, 1960)
Special environmental variance (VEs) or within-individual variance (Falconer, 1960)
for a single measurement per plant may be derived by the error variance component (c2^)
from a one-way analysis of variance (Falconer, 1960). Hendersons Method 3
(Henderson, 1953) for obtaining variance components was performed on vase life data
using individual flowers as observations rather than plant means. The SAS procedure
VARCOMP (SAS, 1982) was used to obtain the a2e or error MS. General environmental
variance (VEg) or between-individual variance (Falconer, 1960) was calculated by
subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated
in the diallel cross, from the error MS or c2e obtained by the combining ability analysis of
variance.
Repeatability for vase life was determined from the following ratio of variances
whereby n = 1:
VG + VEg
VP(n)
(Falconer, 1960)
Results
Selection and Mating
Stem strength was initially measured at five intervals of deformation between zero and one
centimeter. Correlations between stem strength measured at .2 cm and other intervals were


41
in the parental generation, significant negative correlations were observed between vase life
and inflorescence diameter. Among flowers that bent, folded, or wilted in the diallel
generation, a significant negative correlation was observed between vase life and
inflorescence diameter only for flowers that wilted. The positive correlation for flowers
that bent was non-significant and no correlation was observed for flowers that folded. In
both cases, before and after selection, the negative correlation observed between vase life
and inflorescence diameter was highly significant for flowers that wilted. Thus, it appeared
that vase life increased as inflorescence diameter decreased.
Vase life and disc diameter. No significant correlations were observed, based on data
from the total number of flowers evaluated, before or after selection. The only significant
correlations observed between vase life and disc diameter were among flowers from the
parental generation that bent or folded. These highly significant correlations were negative.
A negative correlation was also observed for flowers that wilted in the diallel generation,
though this correlation was non-significant. In general, these correlations demonstrated a
very weak negative relationship, if any, between vase life and disc diameter. The
significant correlations that were observed in the parental generation for vase life of flowers
that bent and folded seemed of little interest since selection would be made against these
senescence modes in a breeding program to improve vase life.
Vase life and vgrowth. Significant positive correlations were observed, based on data
from the toal number of flowers evaluated, before and after selection. Among flowers that
bent, folded, or wilted in the parental generation, highly significant positive correlations
were observed between vase life and vgrowth. Among flowers that bent, folded, or wilted
in the diallel generation, positive correlations were also observed between vase life and
vgrowth, however, correlations were significant for flowers that bent or folded. It
appeared from these correlations that the positive relationship which was observed
between vase life and post-harvest stem elongation prior to selection and mating weakened
as vase life increased.


90
A-5. Gerbera X hvbrida Hort.. Vase life data for backcrosses.
Backcross
Number of Progeny
Mean Vase Life
84-27 -2
X
83-1-77
13
10.4
84-27-2
X
Appleblossom
2
8.0
84-27-4
X
83-1-77
9
10.1
84-27-4
X
Appleblossom
6
8.3
84-27-7
X
83-1-77
1
3.0
84-27 12
X
83-1-77
1
13.0
84- 37 -4
X
83-4-69
5
9.4
84- 37 -4
X
Appleblossom
19
8.4
84 37 10
X
83-4-69
12
9.3
84 37 10
X
Appleblossom
16
9.9
84 37 12
X
83-4-69
21
9.6
84 37 12
X
Appleblossom
11
9.6
84-41-3
X
83-7-4
19
10.7
84-41-3
X
Appleblossom
1
10.0
84-41-5
X
83-7-4
1
13.0
84-41-8
X
83-7-4
1
10.0
84-41 -13
X
Appleblossom
16
10.0
84-41 15
X
83-7-4
22
13.7
84-41 -15
X
Appleblossom
1
13.5
84-41 19
X
83-7-4
1
12.0
84 41 20
X
Appleblossom
3
8.3
84-43-2
X
83-7-10
27
12.4
84-43-2
X
Appleblossom
11
8.9
84-43-5
X
Appleblossom
7
9.5
84-43- 18
X
83-7-10
19
11.2
84-43- 18
X
Appleblossom
2
9.3


81
Eijk, J.P. van and Eikelboom, W., 1986. Aspects of breeding for keeping quality in
Tulipa. Acta Hort, 181: 237-244.
Eijk, J.P. van, Eikelboom, W. and Spamaaij, L.D., 1977. Possibilities of selection for
keeping quality in tulip breeding. Euphytica, 26: 825-828.
Falconer, D.S., 1960. Introduction to Quantitative Genetics. Oliver and Boyd,
Edinburgh, 365 pp.
Ferreira, D.I. and Swaidt, G.H., 1981a. The influence of the number of foliage leaves on
the vase life of cut rose flowers in two media. Agroplante, 13: 73-76.
Ferreira, D.I. and Swardt, G.H., 1981b. A comparison of the vase life and respiration rate
of ten cut rose cultivars and the influence of a flower preservative thereupon. Agroplante,
13: 77-81.
Fujino, D.W., Reid, M.S. and Yang, S.F., 1981. Effects of aminooxyacetic acid on
postharvest characteristics of carnation. Acta Hort., 113: 59-64.
Gilbart, D.A. and Dedolph, R.R., 1965. Phytokinin effects on respiration and
photosynthesis in roses and broccoli. Proc. Amer. Soc. Hort. Sci., 86: 774-778.
Griffing, B., 1956. Concept of general and specific combining ability in relation to diallel
crossing systems. Australian J. of Biol. Sci., 9: 463-493.
Halevy, A.H., 1976. Treatments to improve water balance of cut flowers. Acta Hort., 64:
223-230.
Hallauer, A.R. and Miranda, J.B., 1981. Quantitative Genetics in Maize Breeding. Iowa
State University Press, Ames, Iowa, 480 pp.
Harding, J., Byrne, T.G. and Nelson, R.L., 1981. Heritability of cut-flower vase
longevity in gerbera. Euphytica, 30: 653-657.
Heide, O.M. and 0ydvin, J., 1969. Effects of 6-benzylaminopurine on the keeping
quality and respiration of glasshouse carnations. Hort. Res., 9: 26-36.
Henderson, C.R., 1953. Estimation of variance and covariance components. Biometrics,
9: 226-252.
Kalkman, E.Ch., 1985. Postharvest treatment of Astilbe hvbrida. Acta Hort., 181: 389-
392.
Kofranek, A.M. and Halevy, A. H., 1981. Chemical pretreatment of chrysanthemums
before shipment. Acta Hort., 113: 89-96.
Kofranek, A.M. and Paul, J.L., 1975. The value of impregnating cut stems with high
concentration of silver nitrate. Acta Hort., 41: 199-206.
Kohl, H.C., 1968. Gerberas: Their culture and commercial possibilities. S. Flor, and
Nut., 28: 18,24-26.


68
Table 5-3. Vase life and stem strength data used for classification of twenty-two plants
indu^n|^A£gleblossom/_________==_==_______^^_^_^__
Category I
Category III
Plant
Vase life
Stem
strength
Plant
Vase life
Stem strength
(days)
(g/.2cm)
(days)
(g/.2cm)
83-1-19
12.0
13.3
83-1-2
5.7
13.3
83-1-26
16.0
20.5
83-4-8
4.3
14.2
83-1-31
11.0
31.5
83-5-1
5.7
14.2
83-1-54
11.0
12.7
83-5-76
5.7
14.3
83-7-47
12.7
11.5
83-7-22
6.0
16.8
83-8-7
11.3
16.0
83-7-48
4.7
24.0
Appleblossom
11.0
25.7
83-7-35
5.7
11.3
Category II
Category IV
Plant
Vase life
Stem
strength
Plant
Vase life
Stem strength
(days)
(g/.2cm)
(days)
(g/.2cm)
83-1-6
13.0
4.3
83-1-24
3.7
1.2
83-1-32
12.3
4.7
83-1-79
4.0
2.0
83-1-64
15.0
5.5
83-4-29
3.3
1.5
83-1-77
15.3
4.2
83-4-16
3.0
2.3
Table 5-4. Analysis of variance for stem anatomy. (# of vascular bundles)
TBUN
LBUN
SBUN
Source
df
M.S.
M.S.
M.S.
Vase Life
(high ys low)
1
1.03 n s-
382.68 *
423.50 n s-
Stem Strength
(high ys low)
1
380.60 n s-
8.55 n-s-
275.07
Vase Life x Stem Strength
1
274.46 "-S.
15.14 n.s.
418.53 n-s-
Error
18
535.68
71.62
447.56
* Significant at P< 0.05.


A-l. Vase life data for thirty-one parents from a population of gerbera.
Parent
(Parental Generation)
Mean C.V.
(Top-cross Generation)
Progeny Mean
83-1-6
13.0
20.4
6.8
83-1-13
11.3
5.1
6.8
83-1-26
16.0
16.6
7.7
83-1-32
12.3
4.7
8.2
83-1-61
12.3
9.4
6.8
83-1-64
15.0
17.6
7.9
83-1-74
13.7
23.5
6.9
83-1-77
15.3
24.7
8.6
83-3-4
11.7
9.9
7.5
83-3-34
13.3
11.5
7.2
83-3-38
15.7
7.4
8.7
83-3-47
13.0
20.4
7.4
83-3-77
13.0
7.7
6.4
83-3-96
13.3
11.5
7.4
83-4-4
11.7
5.0
6.1
83-4-17
12.3
20.4
6.3
83-4-38
11.3
5.1
7.8
83-4-69
12.0
8.3
9.6
83-5-33
14.0
7.1
7.6
83-5-39
16.0
22.5
10.4
83-5-109
13.0
21.7
8.9
83-6-42
14.3
17.6
7.9
83-7-4
12.3
18.7
10.2
83-7-6
13.7
16.9
7.8
83-7-10
16.0
0.0
10.7
83-7-14
12.7
18.2
6.6
83-7-18
14.3
10.7
6.9
83-7-26
12.3
18.7
9.4
83-7-29
16.0
0.0
9.4
83-7-47
12.7
4.6
8.6
83-8-7
11.3
5.1
7.6
87


78
vase life increased as stem diameter decreased. In contrast, stem strength increased as stem
diameter also increased.
The number of vascular bundles observed in cross sections of scapes at the zone of
bending or folding suggested that the ratio of small bundles per unit scape circumference
may affect stem strength and the ratio of large bundles per unit circumference may affect
vase life. Heritability studies on these anatomical ratios in gerbera could provide pertinent
information that would be useful to plant breeders.


CHAPTER 2
LITERATURE REVIEW
Researchers have investigated and discussed postharvest longevity in cut flowers
including Gerbera X hybrida Hort.:
The length of lasting quality in itself is not the aim of postharvest longevity
but the satisfaction of the consumer. Premature senescence due to abnormal
causes will leave the consumer with a sense of frustration. (Buys, 1978,
p. 256).
Bent neck in cut roses is a loss of pedicel rigidity which prematurely terminates the
vase life of the flower (Burdett, 1970). Similar to bent neck, knicking (Wilberg, 1973;
Buys, 1978), folding (De Jong, 1978a), neck droop (Zieslin et al., 1978), or stem
break (Meeteren, 1978a) are terms used to describe the sudden bending of the stem in cut
gerberas. Premature senescence in cut flowers contrasts with natural senescence. In
gerbera, this natural phenomenon is identified as wilting (De Jong, 1978a). Wilting
was described as the condition that occurs when the ligulae of an inflorescence on an
upright stem have visibly lost their tugidity.
Vase life is often used as an indicator of postharvest longevity in cut flowers,
including gerbera. Vase life is determined by the number of days from harvest until flower
senescence, whether or not senescence is considered premature. Sytsema (1975) specified
six important factors which may affect the measurement of vase life in cut flowers:
1. Flower conditioning. Harvested flowers allowed to regain turgidity following
storage and/or transpon can withstand adverse room conditions better than
unconditioned flowers.
2. Temperature. Vase life is usually shoner as temperature increases. Small
deviations from 20 C may be optimum.
3


4
3. Relative air humidity (RH) and air circulation speed. To reduce excessive
transpiration, RH should be at least 50% and air circulation speed should be
very low.
4. Light Low light intensity is satisfactory, but total darkness may be less than
ideal.
5. Ethylene and air exchange. Good ventilation is the best method to avoid any
harmful effects due to ethylene concentration.
6. Use of bactericides. Bacterial contamination will interfere with reliable vase
life determinations, therefore a bactericide is recommended. In all cases, exact
room conditions and experimental procedures should be recorded as a basis for
experimental comparison.
Reid and Kofranek (1981) also summarized some recommendations for standardizing vase
life evaluations:
1. Use of vase solution control. Distilled water or deionized water treated for
removal of microorganisms and colloidal materials should be used. Flowers
should be placed individually in sterile containers.
2. Temperature. A good standard is 20 2 C.
3. Light Light intensity of 600 mW/cm2 with a 12 hour diurnal cycle is practical
for simulating home conditions.
4. Relative humidity. A good standard is 60-70%.
5. Air circulation. One air exchange every two hours and wind velocity should
not exceed 0.5m/sec. Stages of flower maturity at time of harvest should be
well documented.
Historically, most postharvest research on cut flowers has focused on environmental,
metabolic, and to a lesser extent, anatomical factors that affect lasting quality.
Environmental treatments have extended the vase life of cut flowers and/or reduced the
incidence of premature senescence.
As early as 1962, it was suggested that the use of chemicals reduced wilting in plants.
Experiments with strawberry plants showed applications of 8-hydroxyquinoline sulfate (8-
HQS) resulted in stomatal closing even when conditions were favorable for opening
(Stoddard and Miller, 1962). Cut flowers of the rose Forever Yours were reported to last


26
Table 3-3.
Record of production dates.
Generation
Sowing
Transplanting
Flowering
Parental
12-2-83
2-18-84
3-15-84/5-5-84
Top Cross
10-12-84
1-12-85
2-23-85/3-12-85
Diallel
1-18-87
3-14-87
4-22-87/5-23-87
Quantitative Analysis
Vase life data were initially analyzed according to the random model for Griffing
Method 3 (Griffing, 1956). This method describes a diallel mating design which includes
reciprocal crosses but excludes seifs. No reciprocal differences were observed.
Subsequently, analysis of variance for combining ability, using a general least squares
diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of
plant means.
Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was
estimated from ratios of the following variances:
VA = Additive genetic variance
Vq = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental)
(Falconer, 1960)
Genotypic and phenotypic variances were determined from the following equations using
thevariance components for general combining ability (a2gca), specific combining ability
(o2sca), and error (a2e) which were calculated by the diallel analysis program developed by
Schaffer and Usanis (1969):


12
Carpenter and Rasmussen (1973) investigated the possibility of plugs occurring during
light or dark periods that reduced water uptake rates. They observed no additional tissue
degradation under daylight. Rasmussen and Carpenter (1974) used scanning electron
microscopy to observe vascular occlusions in cut roses. They found vascular blockage
affects vase life after the cut flower is physiologically incapable of maintaining adequate
water balance; i.e. water loss exceeds water uptake. Mayak et al. (1974) postulated
transpiration plays a more significant role in wilting of cut roses than vascular blockage.
Lineberger and Steponkus (1976) observed two types of vascular occlusion in cut
roses. Microbial occlusions were found in the lower portion of the stem, and gum
deposition was located in the stem above the solution level. Parups and Voisey (1976)
reported the resistance to bending in cut roses is related to lignin content. It was concluded
bent neck will occur if water stress occurs and stem lignification is insufficient. Zieslin et
al. (1978) stated resistance to bending in cut roses depends particularly on secondary
thickening of the vascular system and stem lignification. Zagory and Reid (1986) studied
the role of microorganisms in reducing vase Ufe of carnation Improved White Sim. Only
three of 25 microorganisms isolated from vase solutions reduced vase life. They suggested
that ethylene-producing bacteria may be a possible factor in reducing vase life. De Witte
and Doom (1988) postulated exogenous concentration of pseudomonas bacteria or
Alcaligenes faecalis may account for vascular blockage in roses after three or more days,
but endogenous bacteria in stems or air emboti may cause vascular blockage earher. They
did not find evidence of pectolytic breakdown in xylem cell walls. Dixon et al. (1988)
indicated vase Ufe of cut roses may be proportional to the loss of water conducting capacity
caused by disfunctional xylem tissue. They proposed vase life may be lower in flowers
having a greater proportion of large vascular bundles than small vascular bundles because
larger bundles become disfunctional earlier than smaller bundles. Dixon and Peterson
(1989) concluded physical vascular blockage in the stem of cut roses initially decreased
water uptake, but xylem disfunction induced by water stress reduced vase life over time.


62
of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic
variance for each special case. Ideally, if n = then VEs will be reduced to zero, thereby
deleting a significant source of environmental variation. In that case, the highest possible
estimate of heritability could be obtained for a given population.
Using this method, predicted estimates of heritability for vase life in gerbera ranged
from 28 to 53 percent with n specified as 1,2,3, 5, and < flowers per plant. Since these
calculations assumed no change in genotypic effects, the differences between broad sense
and narrow sense heritability estimates for each case were small. This range of heritability
reinforces the initial description that heritability is moderate, and improvement of stem
strength in gerbera can be obtained even with moderate selection.
The repeatability estimate for stem strength (r = .60) is moderately high, indicating that
two to three flowers per plant is adequate for determining the average stem strength per
plant. Falconer (1960) recommends that further gain in accuracy by more than two
measurements does not justify additional expense or time required to collect more data
when repeatability is high. This was proven by comparing the relative increase in
heritability from predicted estimates based on 1,2, 3, 5, and < measurements per plant.
Between one and three measurements, heritability increased by eleven percent, while
beyond three measurements through infinity, the gain in heritability was only eight percent.
Correlations between stem strength and other morphological traits in this population of
gerbera are of interest, given the wide variation in morphological phenotypes studied.
Despite correlation coefficients were often significant due to the large number of flowers
evaluated, they were generally low.
Significant relationships between stronger stem strength of flowers and shorter stem
length, larger stem diameter, and heavier inflorescence weight were determined before and
after selection and mating. Already, Barigozzi and Quagliotti (1978) and De Jong and
Garretsen (1985) observed that tetraploids appear to have stronger stems than diploids.
These findings provide strong evidence why the use of tetraploids, whose flowers


V
Selection and Classification 65
Anatomy Examinatioa 65
Results 66
Correlation Between Vase Life and Stem Strength 66
Comparison of Stem Anatomy Among Genotypes 67
Correlations Between Stem Anatomy and Postharvest Longevity 72
Discussion 74
Conclusion. 76
6 SUMMARY 77
REFERENCES 79
APPENDICES
A VASE LIFE STUDIES 86
B STEM STRENGTH STUDIES 95
C VASE LIFE X STEM STRENGTH STUDIES 97
D HERITABILITY STUDIES FOR OTHER TRAITS IN GERBERA 102
E CORRELATIONS AMONG OTHER TRAITS IN GERBERA 113
BIOGRAPHICAL SKETCH 115


5
twice as long in an aqueous solution containing 200 ppm. of 8-HQS, an anti-bacterial
agent, than in sterile water (Burdett, 1970).
By the late 1960s, many new postharvest methods and chemical additives were being
tested as a means to extend vase life of major cut flower crops. Heide and 0ydvin (1969)
confirmed cytokinins used as a postharvest dip following storage were beneficial in
prolonging vase life of cut carnations. As concentrations of cytokinin 6-benzylamino
purine (BAP) increased, immersion times should decrease; otherwise, detrimental effects
would result. Application of cytokinin to cut roses was also reported to increase vase life
(Mayak and Halevy, 1970; Mayak et al., 1972).
Kofranek and Paul (1975) confirmed adding silver to preservative solutions was
effective in extending postharvest longevity in chrysanthemums and carnations, but not
gladiolus or gerbera. Mayak and Dilley (1976) found vase life of cut carnations was
enhanced by solutions containing kinetin and sucrose compared with solutions containing
only kinetin. Paulin and Muloway (1979) concluded that cytokinins are best used as a
pretreatment, followed by a glucidic solution to increase the vase life of cut flowers.
Swart (1981) reported vase life of Lilium Enchantment was improved by using a
pretreatment of silver thiosulphate (STS). Sytsema (1981) found STS generally more
effective in extending vase life of standard carnations than spray carnations. Kofranek and
Halevy (1981) suggested a quaternary ammonium product, Physan, could be substituted
for silver nitrate (AgNC^) as a chemical pulsing agent in chrysanthemums. Fujino et al.
(1981) reported that use of aminooxyacetic acid (AOA) as a vase solution additive extended
postharvest longevity of carnations. It was also shown that pulsing carnations with a
higher concentration of AOA was effective. Staden and Beehuizen (1986) developed a
pretreatment formula containing gibberellic acid (GA3), kinetin, daminozide, AOA, and
detergent (Triton X-100) to be another chemical alternative to STS for extending the vase
life of standard and spray carnations. Reddy (1988) reported solutions containing cobalt
salts increased vase life of cut roses.


2
patterns due to selection and mating, to determine correlations between vase life, stem
strength, and other traits of flower and stem morphology and growth, and to examine stem
anatomy of genotypes that differ in vase life and stem strength.


25
Flower senescence was classified into three modes based on the visual condition of the
stem:
1. Bending.
2. Folding.
3. Wilting.
The stem gradually, but irreversibly, loses turgidity resulting in an
arc. If allowed to persist, the stem eventually appears folded.
The stem suddenly bends resulting in an irreversible sharp angle.
The stem remains rigid and upright until the ligulae wilt
Vase life was measured by the number of days to flower senescence.
Vase Life = Senescence Date Harvest Date
In addition to vase life, five variables relating to flower morphology and growth were
measured on each flower:
1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm two diameter measurements were taken, i.e. length
and width, on flowers from the top cross and diallel cross generations.)
3. Inflorescence diameter
4. Disc diameter
5. Stem length (at time of senescence)
A variable, Vgrowth, was created to describe the amount of stem elongation
observed from harvest until senescence. Vgrowth was measured by the difference
between stem length at time of senescence and 30 centimeters.
Vgrowth = Senescence Stem Length 30.0 cm
Production
Plants were grown in 12.5 cm standard plastic pots on raised benches in a clear glass
greenhouse at the University of Florida in Gainesville, Florida. Minimum night
temperature was maintained at approximately 18 C. Day temperature was set at 30 C. A
fan and pad cooling system was used to control the temperature. Shade cloth of 25%
density covered the greenhouse since light intensity generally exceeds 65 W/cm2 per day in
Florida. Sowing, transplanting, and flowering dates for each generation are recorded in
table 3-3.


27
Va 4o2gca
Vq = 4 o2ar + 4o2
gca
sea
Vp = 4 G2ar* + 4a2sca + O2,
gca
(Hallauer, 1981)
Thus, heritability (h2 and H2) was estimated from the formulae:
h2 =
4a2
gca
4a2gca+ 4a2 sea +
H2 =
4o2gca + 4a2
sea
4o2gca + 4a2sca + o2e
Predicted estimates of narrow sense heritability and broad sense heritability for n
measurements per plant were also made. This required partitioning the environmental
variance (VE), a component of phenotypic variance (Vp), into general environmental
variance (VEg) and special environmental variance (V^) in order to determine the
phenotypic variance for each case (Vp(n)).
Vp = VG + VE
VE = VEg + VEs
VP(n) = VG + VEg + VEs
(Falconer, 1960)
Special environmental variance (VEs) or within-individual variance (Falconer, 1960)
for a single measurement per plant may be derived by the error variance component (o2e)
from a one-way analysis of variance (Falconer, 1960). Hendersons Method 3
(Henderson, 1953) for obtaining variance components was performed on vase life data
using individual flowers as observations rather than plant means. The SAS procedure
VARCOMP (SAS, 1982) was used to obtain the o2e or error MS. General environmental
variance (VEg) or between-individual variance (Falconer, 1960) was calculated by
subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF GERBERA X HYBRIDA HORT.
By
Heidi Carol Wemett
May 1990
Chairman: Thomas J. Sheehan
Cochairman: Gary J. Wilfret
Major Department: Horticultural Science
Two components of postharvest longevity, vase life and stem strength, were studied in
cut flowers of Gerbera X hvbrida Hort.. A broad based source of germplasm was
evaluated initially. Progeny means for vase life resulting from a topcross between 31
plants and Appleblossom were used to select plants whose flowers had high vase life.
Analysis of a 5 x 5 diallel cross was made to determine vase life heritability. No reciprocal
effects were observed. Heritability ranged from 22 to 39 percent. Additive gene action
possibly controls this character since broad sense and narrow sense heritability was nearly
equal. Repeatability was moderately high (r = .57).
A senescence mode was recorded for each flower, i.e. bending or folding of the scape
or wilting of the ligulae. Flower was defined as the composite inflorescence and its
corresponding stem. After selection and mating, frequency of bending decreased;
frequency of folding and wilting increased. Flowers that senesced due to wilting had the
highest mean vase life before and after selection and mating. Significant correlations were
found between vase life of flowers that senesced due to wilting and shorter stem length,
thinner stems, and smaller infloresences.
vi


109
D-8. 7x7 Diallel. Parents and progeny means for three traits in gerbera.
N
Stem Length
Stem Diameter
Inflorescence Weight
(cm)
(cm)
(s)
Parents
7
46.8
.57
3.6
Progeny
642
52.8
.50
4.3


59
Table 4-11. Phenotypic correlation coefficients between stem strength (F = g/.2 cm) and
three traits before and after selection.
Generation
N
Stem Length
Stem Diameter
Inflorescence
Weight
Parental
548
-.28 *
.10*
(after storage)
.16**
Diallel
1710
-.05*
.22**
(before storage)
.38**
* Significant at P< 0.05. Significant at P< 0.01.
that following selection and mating, the relationship weakened.
Stem strength and stem diameter. Significant positive correlations were observed
between stem strength and stem diameter, before and after selection. Unlike the
relationship between stem strength and stem length, it appeared that following selection and
mating, the relationship strengthened.
Stem strength and inflorescence weight. Significant positive correlations were
observed between stem strength and inflorescence weight, before and after selection.
Similar to the relationship between stem strength and stem diameter, in this case, following
selection and mating, the correlation coefficient was higher.
Correlations between stem strength and two other traits, inflorescence diameter and
disc diameter, were also determined prior to selection. The correlation (r = -.08) between
stem strength and inflorescence diameter was not significant. A positive correlation (r =
.11) observed between stem strength and disc diameter was highly significant; as stem
strength increased, disc diameter increased.
Discussion
Strength of gerbera stem segments was determined using an Instron. Results were
recorded in terms of the amount of force required to deflect the segment a specified
distance. This method, similar to flexture tests made on engineering materials (Mohsenin,
1970) was developed to test the segment as a simple beam. The intention of using an
Instron to determine stem strength was to increase the accuracy of determinations.


70
Table 5-7. Means2 for number of vascular bundles per 1.0 cm 2 scape area between high
andlowlevelsofva^J^^djtemjffenghm^ierbCTa^^^^^^^^^^^^^^
TBUNARAT
LBUNARAT
SBUNARAT
Vase
Life
High
Low
239.06a
314.71b
78.85a
119.83b
160.22a
194.88a
Stem
Strength
High
Low
251.72a
302.05a
93.10a
105.57a
158.62a
196.48a
2 Means fo
strength are
lowed by different letters within each bundle group for vase life or stem
significant at the 5 % level.
Significant differences between high and low levels of vase life and high and low
levels of stem strength for total number of vascular bundles calculated per unit stem
circumference (1.0 cm) at 12 cm below the peduncle (TBUNCRAT) were determined. The
means for total number of vascular bundles per unit circumference in gerbera stem sections
were significantly different for plants with high vase life compared to those with low vase
life. Plants with high vase life exhibited a smaller total number of vascular bundles per unit
circumference than plants with low vase life. Also, the means for total number of vascular
bundles per unit circumference in gerbera stem sections were significantly different for
plants with high stem stength compared to those with low stem strength. Plants with high
stem strength exhibited a smaller total number of vascular bundles per unit circumference
than plants with low stem strength.
A highly significant difference between high and low levels of vase life was
determined for the number of large vascular bundles calculated per unit stem circumference
(1.0 cm) at 12 cm below the peduncle (LBUNCRAT). The means for number of large
vascular bundles per unit circumference in gerbera stem sections were significantly
different for plants with high vase life compared to those with low vase life. Plants with
high vase life exhibited a smaller number of large vascular bundles per unit circumference
than plants with low vase life. A significant difference between high and low levels of
stem strength was determined for the number of small vascular bundles calculated per unit


45
after only one generation of selection. Therefore, it is postulated that the incidence of
bending ys wilting may be a qualitatively based trait controlled by relatively few genes.
In studies on breeding for keeping quality in gerbera, De Jong (1986) distinguished
senescence by three classes; early stem fold, late stem fold, and petal wilt His definition
for early stem fold was slightly different from that of bending in this study, but he reported
that if selection were made against the phenomenon, its incidence could be reduced or
eliminated fairly easily. He showed data for 59 progenies resulting from a diallel mating;
four progenies did not exhibit early fold. After studying these results, it seems possible
that the trait early fold vs no early fold could fit a genetic model involving only two
genes with epistatic effects (15:1). Hence, it seems plausible to hypothesize that a similar
model may apply to the trait bending vs wilting. Among ten progenies, bending was not
completely eliminated, but, in one case, its incidence was reduced to only five percent.
The incidence of stem folding may be distinct from the incidence of bending or
wilting. Before selection and mating to improve vase life, the frequency of folding was
relatively low. After selection and mating, the frequency of folding nearly tripled.
Accompanying this increase, vase life also increased. This created some speculation as to
whether a higher proportion of folding is the result of longer lasting flowers. Yet when the
proportion of folding for each progeny was compared to progeny means for vase life, this
speculation could not be confirmed because some progeny with higher vase life showed a
smaller proportion of folding than some progeny with lower vase life. Instead, the most
striking observation of this comparison, among ten progenies, was that the proportion of
folding could be grouped into four classes of approximately 20, 30,40, and 65 percent.
This gives cause to wonder whether data from more progenies might yield a sufficient
number of classes to identify a genetic model that would indicate folding is also
qualitatively inherited. If so, it seems likely that the incidence of folding is controlled by at
least several genes.


17
De Jong (1978a) presented methods for rapidly identifying structural strength or turgor
strength of gerbera stems. These components of stem strength were postulated to affect
vase life and stem fold, a premature senescence phenomenon.
Structural strength was measured using a protractor to record the curvature of stems
after freshly harvested flowers were stored dry for 24-48 hours. Turgor strength was
measured by comparing the curvature of the stem after dry storage to the curvature of the
stem following recovery (after the stored stems were placed in water for 24 hours and
allowed to regain turgidity). This difference was related to water uptake ability of the
stems. Rigidity of turgid stems was measured using a specially designed instrument which
recorded the force needed to deflect the stem a predetermined distance.
Using these methods, De Jong observed a high frequency of folding was found in
flowers with weak or firm stems, although the lowest frequency of folding occurred in
flowers with firm stems.
De Jong (1978b) suggested breeding for structurally strong stems as a means to
improve postharvest longevity in gerbera. He reasoned that stronger stems may not be as
likely to fold, thereby overcoming the deleterious effects of microbial infection. Also,
structurally stronger stems could provide some added support if a water deficit occurs. He
concluded, however, turgor strength is of primary importance.
Meeteren (1978b) observed increased ion leakage in petals of cut flower gerberas
depended on the cultivar. Meeteren and Gelden (1980) found no correlation between petal-
cytokinin activities in gerbera cultivan. Meeteren (1981) suggested pressure potential of
petals from recently harvested flowers might be a good indicator for vase life and a possible
selection criterion in breeding programs to improve postharvest longevity in gerberas.
Nowak and Plich (1981) observed vase life of cut gerberas increases when stems were
shorter.
De Jong and Garretsen (1985) noted turgor strength and structural strength of gerbera
stems were greater and stronger during the summer than winter months in Holland. No


91
A-6. Gerbera X hvbrida Hort.. Frequency of senescence modes for backcrosses.
Backcross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-27-2
X
83-1-77
30
26.7
40.0
33.3
84-27-2
X
Appleblossom
6
16.7
33.3
50.0
84-27-4
X
83-1-77
13
30.8
23.1
46.2
84-27-4
X
Appleblossom
12
8.3
58.3
33.3
84-27-7
X
83-1-77
1

100

84 27 12
X
83-1-77
1

100

84- 37 -4
X
83-4-69
8
12.5
12.5
75.0
84-37-4
X
Appleblossom
33
33.3
24.2
42.4
84 37 10
X
83-4-69
22
40.9
22.7
36.4
84 37 10
X
Appleblossom
34
26.5
47.1
26.5
84-37- 12
X
83-4-69
35
11.4
51.4
37.1
84- 37 12
X
Appleblossom
20
15.0
50.0
35.0
84-41-3
X
83-7-4
34

41.2
58.8
84-41-3
X
Appleblossom
1


100
84-41-5
X
83-7-4
1


100
84-41-8
X
83-7-4
1

100

84-41 13
X
Appleblossom
26
23.1
30.8
46.2
84-41 15
X
83-7-4
38
2.6
26.3
71.1
84-41 15
X
Appleblossom
2


100
84-41 19
X
83-7-4
1


100
84 41 20
X
Appleblossom
3
33.3
33.3
33.3
84-43-2
X
83-7-10
62
14.5
24.2
61.3
84-43-2
X
Appleblossom
32
28.1
37.5
34.4
84-43-5
X
Appleblossom
19
31.6
26.3
42.1
84-43- 18
X
83-7-10
30
20.0
40.0
40.0
84-43- 18
X
Appleblossom
3

66.7
33.3


APPENDIX B
STEM STRENGTH STUDIES


40
Table 3-19. Phenotypic correlation coefficients between vase life and post-harvest stem
elongation before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
.22**
.14**
.20**
.37**
Diallel
487
.24**
.23 *
.17*
.13
* Significant at P< 0.05. Significant at P< 0.01.
Vase life and stem length. Significant negative correlations were observed, based on
data from the total number of flowers evaluated, before and after selection. Among flowers
that bent, folded, or wilted in the parental generation, the negative correlations observed
between vase life and stem length were significant only for bending and folding. In
contrast, among flowers in the diallel generation, only the negative correlation observed
between vase life and stem length for flowers that wilted was significant. Overall, it
appeared from these correlations that vase life was highest when stem length was shorter.
Vase life and stem diameter. Based on data from the total number of flowers
evaluated, no correlation was observed before selection, however, a significant correlation
was observed after selection. Among flowers that bent, folded, or wilted in the parental
generation, negative correlations were observed between vase life and stem diameter for
flowers that bent or folded; but, the correlation was significant only for flowers that bent.
The positive correlation observed for flowers that wilted was non-significant. In contrast,
among flowers in the diallel generation, negative correlations were observed between vase
life and stem diameter for flowers that folded or wilted, but the correlation was significant
only for flowers that wilted. The positive correlation observed for flowers that bent was
non-significant. Overall, it appeared from these correlations that vase life was highest
when stem diameter was smaller.
Vase life and inflorescence diameter. Negative correlations were observed, based on
data from the total number of flowers evaluated, before and after selection; however, only
the correlation before selection was significant Among flowers that bent, folded, or wilted


CHAPTER 5
PART HI. VASE LIFE X STEM STRENGTH STUDIES
Introduction
Postharvest longevity is a critical attribute of cut flowers. De Jong (1978a) proposed
two main components, vase life and stem strength, contribute to postharvest longevity in
gerbera. In a second paper, De Jong (1978b) suggested structurally strong stems could
extend postharvest longevity by providing added support to the flower should a water
deficit occur.
The basis for vase life and stem strength in gerbera remains under investigation. Stem
anatomy studies have been conducted in an effort to understand the causes for variation in
postharvest longevity. Reiman-Philip, as cited by Wilberg (1973), noted a greater
proportion of large vascular bundles to small vascular bundles as a possible factor
contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that such
a relationship existed in only extreme cases. In a comprehensive study, Marousky (1986)
measured the size and number of vascular bundles in two gerberas, Tropic Gold and
Appleblossom. He concluded Appleblossom exhibited more resistance to stem bending
than Tropic Gold partially because of various anatomical features such as fewer small
vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference,
and a greater percentage of dry weight per unit of scape length.
The objectives of this study were to determine the relationship between vase life and
stem strength from a population of Gerbera X hvbrida Hort. which varied greatly in flower
and stem morphology and to compare stem anatomy of flowers from plants that were
classified by differences in vase life and stem strength.
64


BIOGRAPHICAL SKETCH
Heidi Carol Wemett, was bom August 17, 1958, in Los Angeles, California. Her
enthusiasm for ornamental horticulture began during childhood. Since her father was an
avid gardener, family outings usually meant a trip to Lotusland in Santa Barbara or
Descanso Gardens in La Caffada, where friends and relatives resided.
She attended St. Johns College, known for its Great Books Program of western
civilization from 1976-1979. There, she received recognition for a prize essay in
mathematics. In May, 1979, she was granted her B.A. degree in liberal arts from the Santa
Fe, New Mexico, campus.
Two weeks later, she enrolled in the horticulture department at Pennsylvania State
University to begin graduate studies in plant breeding and genetics. After conducting
research for three years on the inheritance of orange flower color in Pelargonium x
hortorum L.H. Bailey she obtained her M.S. degree in November, 1982.
In January, 1983, Miss Wemett moved south to attend the University of Florida at
Gainesville where she conducted research on breeding and genetics of postharvest
longevity in cutflowers of Gerbera X hvbrida Hort. to fulfill part of the requirements for
the Doctor of Philosophy degree. Following the successful completion of her candidacy
exam and formal admission to the doctoral program in 1985, she interrupted her studies to
spend a year in Japan. She wishes to continue travelling after graduation.
115


92
Cross
Number of Progeny
Mean Vase Life
84-27 -2
X
84 27 12
10
9.7
84-27 -4
X
84-27- 12
2
9.5
84- 37 -4
X
84 37 10
3
9.7
84- 37 12
X
84 37 10
8
10.6
84-41-3
X
84-41-5
13
9.5
84-41-5
X
84-41-8
7
8.1
84-41-8
X
84-41 -13
4
6.8
84-41 13
X
84-41-8
9
9.2
84-41 15
X
84-41-5
23
9.6
84-41 15
X
84-41-8
1
14.7
84-41-20
X
84-41 13
3
12.7
84-43 -2
X
84-43-3
20
11.0
84-43-5
X
84-43-3
1
15.0
84-43 -7
X
84-43-3
22
9.6
84-43- 18
X
84-43-3
1
14.0


APPENDIX D
HERITABELITY STUDIES FOR OTHER TRAITS IN GERBERA


19
of generations would be required to increase mean vase life. Mean vase longevity was
reported to be between 10 and 14 days.
De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in
gerberas using a diallel mating scheme involving 12 parents and their progenies. Three
characteristics were examined: vase life (days to wilting or folding); percent folding; and
stem curvature. General combining ability was significant for each characteristic. It was
pointed out that a large error variance (Ve2) will probably result when calculating variance
components for vase life unless the mode of senescence is distinguished, i.e. early or late
folding vs. wilting. Reviewing the inter-relationships among the three characteristics
examined, several conclusions were made: (1) folding results in shorter vase life than
wilting; (2) higher curvature may increase the incidence of folding, thereby resulting in
shorter vase life; (3) no relationship exists between curvature and percent late folding; and
(4) late folding is more difficult to select against than early folding.
De Jong (1986) suggested parent choice is a major factor to consider when designing a
breeding program to improve postharvest longevity in cut flowers of gerbera. For the
characteristic days to wilt, 78% of the variation between progeny means could be
attributed to the twelve parents selected. It was concluded the main difficulty which
remains in breeding to improve postharvest longevity is large intraplant variation for both
days to wilt and percent folding.


37
Changes in the proportion of flowers that senesced due to bending, folding, and
wilting were also observed. Differences between senscence mode frequency in the parental
and diallel generations are shown in figure 3-5. After selection and mating, the incidence
of bending was reduced by an average of 41 percent. Folding and wilting increased by an
average of 23 and 18 percent, respectively. In both generations, the frequency of wilting
exceeded that of folding.
120
100
Frequency
(%)
80
60
40
20
0
% bending
% folding
% wilting
parental diallel
Generation
Figure 3-5. Distribution of senescence modes for parental and diallel generations.
The frequency of bending, folding, and wilting in progeny of the diallel generation
was compared to vase life means. Among ten progenies, the proportion of bending
generally decreased as vase life increased. With the exception of one cross, approximately
50 percent of the flowers wilted in progenies with vase life of 11 days or greater.
Approximately 25 percent of the flowers wilted in progenies with vase life of less than 11
days. The proportion of folding did not appear related to progeny means for vase life.
Distribution of this data is shown in figure 3-6.


14
The genetic basis for variation in postharvest longevity in cut flowers has been investigated;
genotypic differences among plants examined, characteristics for selection proposed, and
heritability estimated.
Mayak and Halevy (1970) observed differences in endogenous cytokinin levels
between roses Lovita and Golden Wave. Cytokinin levels were higher in Lovita,
which has a longer vase life than Golden Wave. Marousky (1973) suggested selecting
chrysanthemums for succulent stems that readily translocate water. Mayak et al. (1974)
reported transpiration rates were higher in rose cultivis with short vase life. They
concluded these cultivars had less ability to close their stomates under water stress than
cultivars with long vase life. Zieslin et al. (1978) studied the sensitivity of rose cultivars to
neck droop. They concluded several factors contribute to variation in premature senescence
in roses: transpiration rate, initial water uptake rates, and different organs of a flower
competing for water at different rates. It was also suggested the effect of these factors may
depend on the structural strength of the stem due to lignification or other anatomical factors.
Pauli and Criley (1981) observed clonal differences in vase life among protea cultivars.
Ferreira and Swardt (1981b) concluded the vase life of rose cultivars varies due to genetic
differences. They proposed three factors are genetically controlled: ability to store
carbohydrates, ability to utilize an exogenous supply of sucrose, and susceptibility to bent
neck. Stimart and Brown (1982) observed differences in vase life among zinnia cultivars
when held in solutions containing 8-hydroxyquinoline citrate and sucrose.
Selection and breeding for postharvest longevity in tulips has been extensively
investigated. Twenty years ago, it was declared that by knowing the cultivar name of a
tulip, its vase life could be predicted (Benschop and De Hertogh, 1969). This implies that
genetics plays a significant role in determining the vase life of tulips.
Eijk and Eikelboom (1976) studied the possibility for selecting for postharvest
longevity in tulips. Cut flower tulip vase life was described by three senescence
characteristics, beginning with the number of days from flowering to start of tepal


18
relation between stem stiffness and lignin content was observed among 25 cultivars. They
also suggested the use of tetraploids may increase stem strength because tetraploids usually
have thicker and shorter stems. These researchers observed no differences in the degree of
curvature after dry storage between tetraploids and diploids.
Dubuc-Lebreux and Vieth (1985) suggested a scheme selecting gerbera cultivars
whose stems are more fully differentiated at the region of stem break when flowers are at
the harvesting stage of maturity. They also identified a cultivar in their breeding program
with this characteristic (K-9-9). Amariutei et al. (1986) reported consistent differences in
vase life between two gerberas; Symphonic lasted two days longer than Richard when
treated with pulsing agents or when untreated.
Genetic analysis of vase life in the cut flower Gerbera jamesonii has been described.
Serini and De Leo (1978) estimated narrow sense heritability of vase life within full-sib
families (h2 = .67). Heritability for vase life between plants was lower (h2 = .17). They
concluded selection should be based more on families than individual plants. Tesi (1978)
concluded vase life is strongly influenced by environmental factors. Variation due to
environment and genotype was calculated (Ve2 = 85.2; Vg2 = 14.8). Mean vase life was
reported to be 12 3.0 days.
Harding et al. (1981) estimated heritability of vase life from a non-random sample
population of gerbera genotypes which consisted of half-sib families and clonal parents.
Since the half-sib families had previously been mated and selected for cut flower yield and
preference, it was suggested estimates for narrow sense heritability were biased due to a
reduction in genetic variability. Components of variance were used to estimate heritability.
Narrow sense heritability (h2) was calculated for two successive generations of the half-sib
population; 24 and 38 percent. Broad sense heritability (H2) was calculated for two
successive generations of clonal parents; 36 and 46 percent. Since heritability was
moderately low, it was concluded either intense selection or selection over a large number


E-l. Summary of published phenotypic correlation coefficients between stem and
inflorescence traits in gerbera.
Traits
Stem Diameter
Inflorescence Diameter
Disc Diameter
Stem Length
Borghi & Baldi
1970

.13

Ottaviano et al.
1974
-.25 .50 *
n.s.
-.25 .50 *
Schiva
1975
n.s.
.25 .50 *
n.s.
Cocozza et al.t
1978
n.s.
.34
n.s.
De Leo & Ottaviano
1978
n.s.
-.20 .40
n.s.
La Malfa & Noto
1978
.89**


Stem Diameter
Ottaviano et al.
1974
n.s.
.75 1.00**
Schiva
1975
.50 .75 *
.50 .75 *
Cocozza et al.'
1978
.29
.37
De Leo & Ottaviano
1978
n.s.
>.80
Inflor. Diameter
Ottaviano et al.
1974
n.s.
Schiva
1975
.50 .75 *
Cocozza et alJ
1978
.47
De Leo & Ottaviano
1978
n.s.
* Significant at P< 0.05. Significant at P< 0.01.
f Level of significance was not reported.
114


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EKOS7K1FN_ASDJ0G INGEST_TIME 2015-04-01T19:17:27Z PACKAGE AA00029861_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


22
Table 3-1.
List of tissue cultured cultivars.
Field #
Cultivar Description
Field #
Cultivar Description
84-1
Amethyst
84-10
SI-1
84-2
Peach
84-11
P15-14-0
84-3
Seashell
84-12
SC300
84-4
Appleblossom
84-13
SC400-8
84-5
Raspberry
84-14
35C404-0X
84-6
Aztec
84-15
PI 8-5
84-7
Mandarin
84-16
SB-24
84-8
(Polish line)
84-17
SC205-X
84-9
Tropic Lady
84-18
SC501
Seed populations were obtained from different seed companies. In addition, Dr. J.
Harding1 provided another seed source from his breeding program. Dr. Hardings seed
mixture was the result of eight generations of breeding, but selection for vase life had been
discontinued after the fifth generation. A list of the seed populations is given in table 3-2.
Table 3-2. List of seed populations.
Field#
Population Description
Seed Source
83-1
Davis Population
U.C. Davis Res. Prog. (U.S.A.)
83-2*
Ahms F-l Strain
Herbst Seed Co. (U.S.A.)
83-3
MardigrasF-1 Strain
Ball Seed Co. (U.S.A.)
83-4
Duplex Mixture
Ball Seed Co. (U.S.A.)
83-5
Jongenelen Strain
Ball Seed Co. (U.S.A.)
83-6
Florist Strain Mix
Park Seed Co. (U.S.A.)
83-7
Ramona Mixture
Sluis & Groot Corp. (Holland)
83-8
No. 4 F-l Mix
Clause Seed Co. (France)
* Poor germination, no plants survived.
Plant material varied in flower color and morphology. The term flower will be used
to describe a composite inflorescence subtended by a stem. Plants grown from four seed
'Dr. J. Harding, Dept, of Environmental Horticulture, U. C. Davis, California


60
Although this apparatus provided a relatively sophisticated method to quantitatively
measure mechanical strength, a major problem remained due to the variability in gerbera
stems. Gerbera stems are not structurally homogeneous. In addition to differences in stem
diameter, stems can be hollow or solid and round or oval. Measurements for determining
the force required to deflect a stem segment a specified distance can vary depending on
these factors.
Mohsenin (1970) suggested accounting for variation, often found in agricultural
materials, by using a formula to calculate apparent stiffness, known as modulus of
elasticity (E). Parups and Voisey (1976), who studied the resistance to bending of the
pedicel in greenhouse-grown roses, which are also not structurally homogeneous,
explained their calculations for determining modulus of elasticity of the pedicel part of rose
stems. There was concern about the variability in gerbera stems, therefore stem strength
results obtained from raw data were initially compared to modulus of elasticity values (E).
The following formula was used to calculate modulus of elasticity:
FL3
E =
D48^fr>
(Mohsenin, 1970)
where F is force required to deflect the segment a specified distance, L is length of the
segment, D is deflection distance, and d is diameter of the stem segment at mid-span.
Since no differences were found between ranking plants using raw data directly from
Instron measurements and ranking plants by modulus of elasticity, stem strength values
obtained from raw data were used to estimate heritability and to correlate stem strength to
morphological traits.
Another problem with accurately evaluating stem strength in gerbera relates to the
moisture content of the stem segment measured. Segments were stored at room
temperature for 24 hours; however, this did not specify their exact level of moisture content
at the time of measurement. Evidently the proportion of evaporation differed between stem


94
A-9. Qgikgia X bykda Hort.. Vase life data for half-sib crosses.
Cross
Number of Progeny
Mean Vase Life
84-37- 12 x 84-43-3
12
11.9
84-41 19 x 84-27-2
1
13.0
84-43 3 x 84- 41 -8
16
10.0
A-10. Gerbera X hvbrida Hort.. Frequency of senescence modes for half-sib crosses.
Cross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-37 12 x 84-43 -3
23
13.0
34.8
52.2
84-41 19 x 84-27-2
1


100
84-43- 3 x 84-41 -8
29
24.1
58.6
17.2


89
A-3. Parental Generation. Phenotypic correlation coefficients between stem and
inflorescence traits in gerbera.
i 1 g", mmm m == ==
Traits
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Stem Length
b
-j
*
.16 *
.20* *
.06*
Stem Diameter
.10* *
.64**
.26**
Inflorescence Diameter
.23* *
-.07 *
Disc Diameter
.17* *
* Significant at P< 0.05. Significant at P< 0.01.
A-4. Parental Generation. Phenotypic correlation coefficients between stem and
inflorescence traits in gerbera by senescence mode.
Traits
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Stem Length
Bending
.07
.11**
.18 *
.07
Folding
.12
.31 **
.35 **
-.02
Wilting
.16*
.21 **
.29**
.09
Stem Diameter
Bending
.15**
.63**
.28**
Folding
.15
.58 **
.07
Wilting
-.01
.76**
.14*
Inflorescence Diameter
Bending
.21**
-.02
Folding
.34**
-.28 *
Wilting
.06
-.20 *
Disc Diameter
Bending
.20* *
Folding
-.08
Wilting
.09
* Significant at P< 0.05. Significant at P< 0.01.


84
Ottaviano, E., Schiva, T., and Sari-Gorla, M., 1974. Analisi multivariata per lo studio
delle differenze genetiche in popolazioni di Gerbera jamesonii. Genet. Agr. 28: 292-306.
Parups, E.V. and Molnar, J. M., 1972. Histochemical study of xylem blockage in cut
roses. J. Amer. Soc. Hort. Sci., 97: 532-534.
Parups, E.V. and Voisey, P.W., 1976. Lignin content and resistance to bending of the
pedicel in greenhouse-grown roses. J. Hort. Sci., 51: 253-259.
Paulin, A. and Muloway, K., 1979. Perspectives in the use of growth regulators to
increase the cut flower vase life. Acta Hort., 91: 135-141.
Pauli, R., Goo, T., Criley, R.A. and Parvin, P.E., 1981. Leaf blackening in cut Protea
eximia: Importance of water relations. Acta Hort, 113: 159-168.
Rasmussen, H.P. and Carpenter, W.J., 1974. Changes in the vascular morphology of cut
rose stems: A scanning electron microscope study. J. Amer. Soc. Hort. Sci., 99: 454-459.
Reddy, T.V., 1988. Mode of action of cobalt extending vase life of cut roses. Scientia
Hortic., 36: 303-313.
Reid, M.S. and Kofiranek, A.M., 1981. Recommendations for standardized vase-life
evaluations. Acta Hort., 113: 171-174.
Rogers, M.N., 1973. An historical and critical review of postharvest physiology research
on cut flowers. HortScience, 8: 189-194.
SAS Institute Inc., 1982. SAS Users Guide: Statistics, 1982 Edition. SAS Institute Inc.,
Cary, N.C., 584 pp.
Schaffer, H.E. and Usanis, R.A., 1969. General least squares analysis of diallel
experiments. A computer program-DIALL. Genetics Dept., North Carolina State Univ.,
Raleigh, N.C. 27607.
Schiva, T., 1973. Analisi dei caratteri metrici di importanza commerciale su Gerbera
iamesonii e conseguenze dellinincrocio. Estrato dagli Annali della facolt di Scienze
Agrarie dellUniversita degli Studi di Torino, 8: 355-366.
Schiva, T., 1975. Miglioramento gentico dell gerbera. II. Analisi gentica di caratteri
metrici di importanza commerciale e conseguenze dellinincrocio. Genet Agr., 29: 233-
240.
Serini, G. and De Leo, V., 1978. Phenotypic characters and preservability of gerbera
flowers. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and
Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 269-277.
Siegelman, H.W., 1952. The respiration of rose and gardenia flowers. Proc. Amer. Soc.
Hort. Sci., 59: 496-500.
Simmonds, N.W., 1979. Principles of Crop Improvement. Longman Group Limited,
London, 408 pp.
Smith, D.E. and Nelson, R.L., 1967. Gerbera propagation. Calif. Agrie., 21(12): 7.


80
Coorts, G.D., 1973. Internal metabolic changes in cut flowers. HortScience, 8: 195-198.
Coorts, G.D., McCollum, J.P. and Gartner, J.B., 1965. Effect of senescence and
preservative on mitochondrial activity in flower petals of Rosa hybrida. Velvet Times.
Proc. Amer. Soc. Hort. Sci., 86: 791-797.
De Jong, J., 1978a. Dry storage and subsequent recovery of cut gerbera flowers as an aid
in selection for longevity. Scientia Hort., 9: 389-397.
De Jong, J., 1978b. Selection for keeping quality in gerbera. In: L. Quagliotti and A.
Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a
Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 263-268.
De Jong, J., 1986. Breeding for keeping quality in gerbera. Acta Hort., 181: 353-357.
De Jong, J. and Garretsen, F., 1985. Genetic analysis of cut flower longevity in gerbera.
Euphytica, 34: 779-784.
De Leo, V. and Ottaviano, E., 1978. Genetical analysis of morphological traits in Gerbera
iamesonii. Clonal and diallel families. In: L. Quagliotti and A. Baldi (Editors), Genetics
and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April,
at Alassio, Italy, pp. 179-191.
De Witte, Y. and Doom, W.G. van, 1988. Identification of bacteria in the vase water of
roses, and the effect of the isolated strains on water uptake. Scientia Hortic., 35: 285-291.
Dilley, D.R. and Carpenter, W.J., 1975. The role of chemical adjuvants and ethylene
synthesis on cut flower longevity. Acta Hort., 41: 117-132.
Dixon, M.A. and Peterson, C.A., 1989. A re-examination of stem blockage in cut roses.
Scientia Hortic., 38: 277-288.
Dixon, M.A., Butt, J.A., Muir, D.P. and Tsujita, M.J., 1988. Water relations of cut
greenhouse roses:the relationships between stem water potential, hydraulic conductance
and cavitation. Scientia Hortic., 36: 109-118.
Downs, C. and Reihana, M., 1987. Extending vaselife and improving quality of nerine cut
flowers with preservatives. HortScience, 22: 670-671.
Drennan, D., Harding, J. and Byrne, T.G., 1986. Heritability of inflorescence and floret
traits in gerbera. Euphytica, 35: 319-330.
Dubuc-Lebreux, M.A. and Vieth, J., 1985. Histologie du pedoncule inflorescentiel de
Gerbera iamesonii. ActaBot. Neerl., 34(2): 171-182.
Durkin, D. and Kuc, R., 1966. Vascular blockage and senescence of the cut rose flower.
Proc. Amer. Soc. Hort. Sci., 89: 683-688.
Eijk, J.P. van and Eikelboom, W., 1976. Possibilities of selection for keeping quality in
tulip breeding. Euphytica, 25: 353-359.
Eijk, J.P. van and Eikelboom, W., 1980. Methods of selection in tulip breeding. Acta
Hort., 109: 217-225.


57
Table 4-7. 7x7 Diallel. Variances and heritability estimates for stem strength.
(F = g/,2 cm)
VA
vG
vP
h2
H2
103. 64
116.76
274.98
OO
m
.42
component (c2^ for a single measurement per plant was obtained by an analysis of
variance using stem strength data from individual flowers of the diallel generation. Also,
this analysis indicated differences in stem strength of flowers among crosses and among
plants within crosses were highly significant. Results are summarized in table 4-8.
Table 4-8. 7x7 Diallel. Analysis of variance for stem strength. (F=g/.2 cm)
Source of Variation
df
M.S.
F-ratio
Among crosses
20
3995.98
10.68 *
Among plants
621
374.08
2.12**
Within plants
1068
146.06
* ^Significant at P< 0.01.
General environmental variance (VEg) and special environmental variance (Vg^ were
derived from stem strength data using calculations described by Falconer (1960). A
summary of variances for this population is given in table 4-9. Predicted estimates of
narrow sense heritability and broad sense heritability for stem strength were then made for
1, 2, 3, 5, and measurements per plant by the ratio of genetic variance (VA or Vq) to
phenotypic variance (VP(n)) (Falconer, 1960). Estimates ranged from 28 to 53 percent.
These results are given in table 4-10. Repeatability (r = .60) for stem strength was
moderately high.


In memory of my father and mother,
Benjamin Carl Wemett and Evelyn Carolyn Wemett,
who inspired my appreciation for flowers.


83
Mayak, S. and Halevy, A.H., 1974. The action of kinetin in improving the water balance
and delaying senescence processes of cut rose flowers. Physiol. Plant, 32: 330-336.
Mayak, S. and Halevy, A.H., 1980. Rower senescence, In: Senescence in Plants, K.V.
Thimann (Editor), CRC Press, Boca Raton, Florida, 131-156.
Mayak, S., Halevy, A.H. and Katz, M., 1972. Correlative changes in phtyohormones in
relation to senescence processes in rose petals. Physiol. Plant, 27: 1-4.
Mayak, S., Halevy, A.H., Sagie, S., Bar-Yoseph, A. and B. Bravdo, 1974. The water
balance of cut rose flowers. Physiol. Plant, 31: 15-22.
Meeteren, U. van, 1978a. Water relations and keeping quality of cut gerbera flowers. I.
The cause of stem break. Scientia Hort., 8: 65-74.
Meeteren, U. van, 1978b. Water relations and keeping quality of cut gerbera flowers. II.
Water balance of ageing flowers. Scientia Hort, 9: 189-197.
Meeteren, U. van, 1979. Water relations and keeping quality of cut gerbera flowers. III.
Water content, permeability, and dry weight of ageing petals. Scientia Hort., 11: 83-93.
Meeteren, U. van, 1980. Water relations and keeping quality of cut gerbera flowers. VI.
Role of pressure potential. Scientia Hort., 12: 283-292.
Meeteren, U. van, 1981. Role of pressure potential in keeping quality of cut gerbera
inflorescences. Acta Hon., 113: 143-150.
Meeteren, U. van and Gelder, H. van, 1980. Water relations and keeping quality of cut
gerbera flowers. V. Role of endogenous cytokinins. Scientia Hort., 12: 273-281.
Mohsenin, N.N., 1970. Physical Properties of Plant and Animal Materials. Gordon and
Breach Science Publication, New York, 123 pp.
Mutsenietse, G. Ya., RashaT, I. D. and Dishler, V. Ya., 1978. Inheritance of quantitative
characters in gerbera in diallel crosses. II. Inflorescence characters. Genetika, 14: 779-
783.
Nichols, R., 1981. Ethylene, present and future. Acta Hort., 113: 11-18.
Nichols, R 1982. Growth regulators and flower senescence. British Plant Growth
Regulator Group, Monograph, 8: 113-120.
Nowak, J., 1981. Regulation of bud opening, storage period, vase-life and senescence of
cut flowers. I. The effect of silver complexes and sucrose on longevity of cut gerbera
inflorescences stored for different periods of time. Res. Inst, of Pom. and Floric., Ann.
Report, 2: 3-10.
Nowak, J. and Plich, H., 1981. Regulation of bud opening, storage period, vase-life, and
senescence of cut flowers. II. The effect of silver ions and other anti-ethylene agents on
ethylene synthesis and senescence of gerbera inflorescences. Res. Inst, of Pom. and
Floric., Ann. Report, 2: 10-21.


Frequency (%)
34
12.5-1
10.0-
7.5-
PARENTAL GENERATION
mean = 8.2
Vase Life (Days)
Figure 3-3. Distribution of vase life data for flowers that folded from parental and diallel
generations.


69
Table 5-5. Means2 for number of vascular bundles between high and low levels of vase
lfeandstemstrengunjerbera;___=>^_B^_aig!aB_____==¡=a=_a=^_^___^
TBUN
LBUN
SBUN
Vase
High
44.92a
16.37a
33.55a
Life
Low
49.81a
18.59b
31.22a
Stem
High
48.76a
17.31a
31.45a
Strength
Low
50.97a
17.64a
33.33a
2 Means fo
strength are
lowed by different letters within each bundle group for vase life or stem
significant at the 5 % level.
number of bundles (TBUNARAT) and number of large bundles (LBUNARAT). The
means for total number of vascular bundles and number of large vascular bundles for this
unit area in gerbera stem sections were significantly different for plants with high vase life
compared to those with low vase life. Plants with high vase life exhibited a smaller total
number of vascular bundles per unit area than plants with low vase life. Also, plants with
high vase life exhibited a smaller number of large vascular bundles per unit area than plants
with low vase life. A summary of these results is given in tables 5-6 and 5-7.
Table 5-6. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm 2
scape area.)
TBUNARAT
LBUNARAT
SBUNARAT
Source
df
M.S.
M.S.
M.S.
Vase Life
(high vs low)
1
444840.20 *
130549.39 *
93420.23 n s-
Stem Strength
(high is low)
1
196864.30
12087.29 n s-
111390.20 n-s-
Vase Life x Stem Strength
1
10271.08 n s-
64.98 n-s-
8702.12 n s-
Error
18
79568.77
20121.40
26395.89
* Significant at P< 0.05.


CHAPTER 4
PART II. STEM STRENGTH STUDIES
Introduction
Gerbera X hvbrida Hort. is a popular cut flower that is recognized for its long,
leafless, and upright stem, known as a scape. Postharvest longevity of gerbera often
abruptly ends when the stem ceases to remain upright.. This phenomenon has been termed
knicking (Wilberg, 1973; Buys, 1978), folding (De Jong, 1978a), neck droop
(Zieslin et al., 1978), or stem break (Meeteren, 1978a). De Jong (1978b) suggested
breeding for structurally strong stems may be a means to improve postharvest longevity in
gerbera.
Descriptions of the relationship between stem strength and morphological and growth
traits in gerbera are few. Wilberg (1973) identified that higher ratios of dry weight/cm for
stem sections prone to bending contribute to stem strength in gerbera. Barigozzi and
Quagliotti (1978) noted tetraploids appeared to have stronger stems than diploids. De Jong
and Garretsen (1985) supported this observation by suggesting thicker and shorter stems of
tetraploids may result in increased stem strength.
The objectives of this research on gerbera were to determine broad sense heritability
and narrow sense heritiability estimates for stem strength and to determine correlations
between stem strength and other traits of flower and stem morphology and growth.
Materials and Methods
Selection and Mating
Initially, a population of 278 plants was randomly selected for stem strength evaluation
(May 25June 14, 1984). They were selected from the residual parental generation of
48


93
A-S^GerberaXhvbrida^Hort^Fre^uencv^ofsenescence^m^es^for^full-sibcrosses^
Cross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-27-2
X
84 27 12
25
36.0
32.0
32.0
84-27-4
X
84 27 12
2

100

84- 37 -4
X
84 37 10
6

16.7
83.3
84 37 12
X
84 37 10
16
12.5
43.7
43.7
84-41-3
X
84-41-5
23
21.7

78.3
84-41-5
X
84-41-8
13
15.4
30.8
53.8
84-41-8
X
84-41 13
5

100

84-41 13
X
84-41-8
21
9.5
57.1
33.3
84-41 15
X
84-41-5
45
11.1
17.8
71.1
84-41 15
X
84-41-8
3

66.7
33.3
84 41 20
X
84-41 13
6
16.7
33.3
50.0
84-43-2
X
84-43-3
46
17.4
28.3
54.3
84-43-5
X
84-43 -3
1

100

84-43-7
X
84-43 -3
42
19.0
35.7
45.2
84-43- 18
X
84-43-3
2


100


D^T^T^x^T^DialleL^Other^aitmeans^asedon^ro^en^meansofjerberacrosses.
83-1-10
83-1 -31
83-1-96
xj = 57.33
X! = 55.54
83-1-10
x2 = .48
x2 = .47
x3= 4.23
x3= 4.94
n = 16
n= 16
xj = 56.84
xj = 48.48
83-1 -31
x2 = .47
x2 = .46
x3= 4.28
x3= 3.63
n = 16
n = 11
X! =52.10
xj = 46.34
83-1-96
x2 = .46
x2 = .47
x3= 4.21
x3= 4.01
n = 16
n = 14
X] = 53.44
xj = 52.81
xt = 46.78
83-4-8
x2 = .47
x2 = .46
x2 = .45
x3= 3.47
x3= 3.05
x3= 3.18
n= 16
n = 15
n= 15
X! = 55.03
X! = 49.05
xj = 40.97
83-5-76
x2 = .52
x2 = .53
x2 = .49
x3= 5.13
x3= 5.01
x3= 4.32
n = 16
n = 16
n = 13
8.45
X! = 52.95
xj =53.14
xj = 50.28
83-7-48
x2 = .50
x2 = .50
x2 = .50
x3= 4.90
x3= 4.24
x3= 4.66
n = 16
n = 16
n = 13
83-4-8
83-5-76
83-7-48
83-8-7
xj = 52.62
x2 = .43
x3= 3.24
n= 16
X! = 58.23
x2 = .54
x3= 5.53
n = 16
x, = 59.03
x2 = .51
x3= 5.13
n = 16
xj = 67.61
x2 = .50
x3= 4.47
n= 16
X! = 48.63
x2 = .44
x3= 2.83
n = 16
X! = 47.39
x2 = .52
x3= 4.96
n = 16
X! = 54.90
x2 = .50
x3= 4.36
n = 16
X! = 61.38
x2 = .52
i3= 4.32
n = 16
xj = 45.93
x2 = .46
x3= 3.40
n = 15
X! = 42.54
x2 = .49
x3= 4.63
n= 14
X! = 46.44
x2 = .49
x3= 4.59
n= 16
X! = 57.93
x2 = .53
x3= 4.99
n= 15
X] = 45.41
x2 = .49
x3= 3.70
n = 16
X] = 45.66
x2 = -45
x3= 3.23
n = 16
X! = 57.29
x2 = .52
x3= 3.78
n = 16
X! = 43.20
x2 = .48
x3= 3.61
n = 15
xj = 45.40
x2 = .55
x3= 4.76
n= 15
x, = 49.49
x2 = .57
x3= 5.22
n = 13
xj = 48.89
x2 = .48
x3= 3.56
n = 16
X! =51.64
x2 = .55
x3= 4.70
n = 15
X! = 57.83
x2 = .56
x3= 5.31
n = 16


CHAPTER 3
PARTI. VASE LIFE STUDIES
1-ngQd.UCLiQa
Gerbera X hvbrida Hort. is a popular cut flower, but its postharvest performance is
often less than desirable. Ideally, postharvest longevity in gerbera should be two weeks or
longer. Unfortunately for the consumer, vase life is usually much less. Postharvest
treatments, i.e. floral preservatives, are used to enhance the lasting ability of gerbera, but,
developing cultivars with genetically superior postharvest longevity may provide the
consumer with a reliable expectation for postharvest quality. Therefore, research to
evaluate the potential of plant breeding as a method to improve postharvest longevity in
gerbera is important.
At present, several researchers have estimated heritability for vase life, which is
defined as the length of time until the flower senesces. Serini and De Leo (1978)
concluded selection should be based more on families than individual plants since their
estimate of narrow sense heritability was higher for among full-sib families (h2 = .67) than
among plants (h2 = .17). Tesi (1978) concluded vase life is strongly influenced by
environmental factors. He observed only 15 % of the phenotypic variation in vase life was
due to genotype. Harding et al. (1981) based their results on a nonrandom sample
population of gerbera genotypes from their Davis population which consisted of half-sib
families and clonal parents. They concluded that since narrow sense heritability (h2 = .24
and .38) and broad sense heritability (H2 = .36 and .46) were moderately low for two
successive generations, either intense selection or selection over a large number of
generations would be required to increase mean vase life.
20


23
populations (Jongenelen, Florists Strain, Ramona, and No. 4 F-l) resembled those Dutch
cultivare that have inflorescences with broad ligulae and thick fleshy stems. These
populations represented a wide spectrum of flower color. The Duplex Mixture was
comprised largely of pinkish hues with spindly stems and narrow ligulae. The Mardigras
F-l Strain included doubles and crested inflorescence types as well as the single, daisy-
type form. These single, daisy-type inflorescences were mostly of deep red hues with
narrow ligulae. The Davis population mainly exhibited inflorescences with narrow ligulae
also representing a wide spectrum of flower color. Flowers from these latter two
populations had stems of medium thickness.
Selection and Mating
Initially, a base population of 953 plants was grown. Plants which did not produce at
least one flower during a flowering period prior to vase life evaluation (May 5June 12,
1984) were discarded. Concurrently, plants were screened for inflorescence type and stem
length. Plants which did not produce single, daisy-type inflorescences or stem length of 45
cm or greater, when 1-2 rows of disc florets were open, were discarded. The remaining
plants of the base population were then referred to as the parental generation. One to six
flowers per plant were evaluated. After the evaluation period, plants that produced less
than three flowers were also discarded. The residual parental generation included 325
plants. Plant means were determined from data collected on the first three flowers
evaluated per plant. Thirty-one plants (approx. 10% of the residual parental generation)
with highest mean vase life (x ^ 11.3)) and lowest coefficient of variation (C.V. £ 25.0)
were selected. To maintain genetic diversity, selection included plants from each seed
population.
A top-cross mating scheme was utilized as a screening method to determine which of
these parents had the longest vase life. Appleblossom was used as the male donor in the
top cross mating scheme because of its excellent vase life rating and low intraplant


This dissertation was submitted to the Graduate Faculty of the College of Agriculture
and to the Graduate School and was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May, 1990
Dean,
:ufmre
Allege of Agricu
Dean, Graduate School


Ill
^D-H^T^xTKdle^Summ^ofvm^cesfor^ree^^^mjer^^n^^^
Variances
Stem Length
Stem Diameter
Inflorescence Weight
vA
90.36
.0034
1.27
vG
103.96
.0035
1.42
VE
45.09
.0020
0.90
VHg
35.43
.0016
0.73
Vrs
. 25.71
.0010
0.46
VP(n)
149.05
.0055
2.32
D-12. 7x7 Diallel. Estimates of heritability and repeatability for three traits in gerbera.
Stem Length
Stem Diameter
Inflorescence Weight
heritability
h2
.61
.62
.55
H2
.70
.64
.61
repeatability
r
.84
.84
.82


72
The only variable analyzed for which both significant differences were found between
high and low levels of vase life and stem strength was the total number of vascular bundles
per unit of circumference (TBUNCRAT). The means for this variable with respect to both
levels of vase life and stem strength are presented in table 5-10. The mean for total number
of vascular bundles per unit stem circumference (1.0 cm) at 12 cm below the peduncle was
lowest for plants with high vase life and high stem strength. Conversely, the mean was
highest for plants with low vase life and low stem strength. The difference between the
means of plants with high and low vase life for total number of vascular bundles per unit of
circumference among plants with high stem strength is not equal to the difference between
the means of plants with high and low vase life for total number of vascular bundles per
unit of circumference among plants with low stem strength (28 34 ^ 33 36). Similarly,
the difference between the means of plants with high and low stem strength for total
number of vascular bundles per unit of circumference among plants with high vase life is
not equal to the difference between the means of plants with high and low stem strength for
total number of vascular bundles per unit of circumference among plants with low vase life
(28 33 ^ 34 36). These inequalities possibly indicate a source of experimental error
since interaction effects were non-significant.
Table 5-10. Means for total number of vascular bundles per 1.0 cm scape circumference
between high and low levels of vase life and stem strength of twenty two gerbera plants.
Vase Life
High
Low
X
Stem
Strength
High
28
34
31.0
Low
33
36
34.5
X
30.5
35.0
Correlations Between Stem Anatomy and Postharvest Longevity
Correlation coefficients between nine stem anatomy variables and two components of
postharvest longevity, vase life and stem strength, were determined using means from


47
Evidence suggests that obtaining plants homozygous for wilting may be possible.
Recurrent selection then would be useful to increase the number of days to wilting. Further
studies to determine the heritability of this vase life component could indicate the intensity
of selection necessary to increase mean vase life for a gerbera population. Consistent
selection against plants whose flowers exhibit bending or folding is imperative.
Finally, after several lines with superior vase life are established, it is recommended
that breeders incorporate this trait into other lines with desireable plant morphology using a
backcross mating scheme, especially if wilting can be confirmed to be a qualitative trait.


106
D-5. 5x5 Diallel. Summa
ry of variances for five traits in gerbera. n = 1.96
Variances
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
vA
22.76
.0084
1.38
.089
0.15
vG
27.20
.0100
1.30
.089
0.21
VE
67.40
.0035
0.92
.040
0.84
VHg
48.54
.0017
0.69
.025
0.47
VEs
36.96
.0035
0.46
.029
0.72
VP(n)
94.60
.0135
2.20
.129
1.05
D-6. 5x5 Diallel. Estimates of heritability
and repeatability for five traits in
gerbera.
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
heritability
h2
.24
.62
.62
.69
.14
H2
.29
.74
.59
.69
.20
repeatability
r
.74
.77
.81
.80
.49


APPENDIX E
CORRELATIONS AMONG OTHER TRAITS IN GERBERA


Vil
A 7 x 7 diallel using plants whose flowers had high stem strength, measured using an
Instron, was used to determine stem strength heritability. No reciprocal effects were
observed. Heritability ranged from 28 to 53 percent Additive gene action possibly
controls this character since broad sense and narrow sense heritability was nearly equal.
Repeatability was moderately high (r = .60). Significant correlations were found between
stem strength and shorter stem length, thicker stems, and heavier infloresences.
Number of vascular bundles was compared among plants whose flowers exhibited
high or low levels of vase life and stem strength. Significant differences between high and
low levels of vase life were observed for number of large bundles per 1.0 cm. scape
circumference, measured 12 cm. below the peduncle; significant differences between high
and low levels of stem strength were observed for number of small bundles per 1.0 cm.
scape circumference.


6
Experimenting with chemical additives to extend the vase life of minor cut flower
crops was gaining momentum by the 1980s. Pauli et al. (1981) reported that use of the
preservative Floralife delayed flower wilting and leaf blackening in protea. Stimart and
Brown (1982) demonstrated optimal vase life for cut flowers of zinnia could be obtained
using a holding solution containing 200 ppm. of 8-hydroxyquinoline citrate (8-HQC) and
1% sucrose. Positive effects of floral preservative on postharvest quality of gypsophila
were initially documented by Marousky and Nanney (1972). Tandler et al. (1986) found
quaternary ammonium compounds were also effective in prolonging the vase life in the cut
flower gypsophila. Postharvest treatments of STS or sugar plus a bactericide were
reported to double the vase life of gypsophila (Barendse, 1986). Postharvest longevity
was shown to increase in cut flowers of calendula, zinnia, and snapdragon if pulsing
treatments of AgNC>3 were utilized. Sweet pea and delphinium lasted longer if STS pulsing
treatments were given (Awad et al., 1986). Sytsema (1986) reported STS prolonged vase
life in cut ffeesia, but if STS was used in combination with cytokinins, vase life would be
further enhanced. Kalkman (1986) stated cut flower preservatives would increase
postharvest longevity in astilbe. Leeuwen (1986) proposed a combination of growth
regulators and STS may be the most effective in prolonging vase life in Euphorbia fulgens.
Downs and Reihana (1987) found preservative solutions enhance vase life of nerine cut
flowers.
Research specifically dealing with postharvest treatments to increase lasting quality in
cut flowers of Gerbera iamesonii has been reported. Waters (1964) reported treatments
with flower preservative Everbloom increased vase life in gerberas 3-5 days. Kohl (1968)
concluded a floral preservative should be used to increase vase life in gerbera, and that
stems should be recut just prior to use. Marousky (1975) cautioned that vase solutions
using fluoridated water can reduce postharvest longevity in gerbera. Meeteren (1978a)
suggested the susceptibility to stem break in some gerbera cultivars could be prevented by
using bactericides in vase solutions. Nowak (1981) reported vase life of gerbera may be


D-7.-continued.
83- 1 10 83 1 -31 83 1 -96
83-8-7
x, = 68.69
X2 = .51
x3= 4.45
n = 15
xj = 60.98
x2 = .51
x3= 4.24
n = 16
x, = 52.57
x2= -51
x3= 4.52
n = 15
X] = stem length
x2 = stem diameter
x3 = inflorescence weight
n = number of progeny
83- 4- 8 83 5 -76 83-7 -48
xj = 55.64
x2 = .47
x3= 3.37
n = 14
xj =49.13
x2 = .59
x3= 5.68
n= 13
xj = 56.63
x2 = .57
x3= 5.07
n = 15
83-8-7
o
OO


Frequency (%)
35
12.5-1
10.0-
7.5-
PARENTAL GENERATION
mean = 12.2
Vase Life (Days)
Figure 3-4. Distribution of vase life data for flowers that wilted from parental and diallel
generations.


110
D-9. 7x7 Diallel. General combining ability analysis of variance for three traits in
Source of Variation
df
Stem Length
M.S.
Stem Diameter
M.S.
Inflorescence Weight
M.S.
General combining
ability
6
3600.57 *
.1316 *
50.60 *
Specific combining
ability
14
148.82 *
.0032 n s-
2.08 *
Error
621
45.09
.0020
0.90
* Significant at P< 0.01.
D-10. 7x7 Diallel. Analysis of variance for three traits in gerbera.
Source of Variation
df
Stem Length
M.S.
Stem Diameter
M.S.
Inflorescence Weight
M.S.
Among crosses
20
3169.60 *
.0998 *
43.49 *
Among plants
621
109.69 *
.0053 *
2.31 *
Within plants
1069
25.71
.0010
0.46
* Significant at P< 0.01.


51
In addition to stem strength evaluation, five variables relating to flower morphology
and growth were measured for each flower from plants belonging to the parental
generation:
1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm)
3. Inflorescence diameter.
4. Disc diameter
5. Inflorescence weight (after storage)
Three variables were measured on each flower from plants belonging to the diallel
cross generation:
1. Stem length (at time of harvest)
2. Stem diameter (at 30.0 cm)
3. Inflorescence weight (before storage)
Production
Plants were grown according to the production regime described for vase life studies
(Chapter 3). Sowing, transplanting, and flowering dates for both generations are recorded
in table 4-2.
Table 4-2. Record of production dates
Generation
Sowing
Transplanting
Flowering
Parental
12-2-83
2-18-84
3-15-84/5-25-84
Diallel
1-30-85
4-7-85
5-15-85/6-12-85
Quantitative Analysis
Stem strength data were initially analyzed according to the random model for Griffing
Method 3 (Griffing, 1956). This method describes a diallel mating design which includes
reciprocal crosses but excludes seifs. No reciprocal differences were observed.
Subsequently, analysis of variance for combining ability, using a general least squares


39
Table 3-15. Phenotypic correlation coefficients between vase life and stem length before
and after selection for vase life.
-- i
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.14 *
-.15**
-.24 *
1
b
U\
Diallel
487
-.14**
-.01
-.04
-.27 *
* Significant at P< 0.01.
Table 3-16. Phenotypic correlation coefficients between vase life and stem diameter
before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
.02
-.11**
-.08
.08
Diallel
487
-.11*
.08
-.15
-.24 *
* Significant at P< 0.05. *
* Significant at P< 0.01.
Table 3-17. Phenotypic correlation coefficients between vase Ufe and inflorescence
diameter before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.11**
-.10**
-.33 *
-.29**
Diallel
487
-.05
.14
.00
-.23 *
* Significant at P< 0.01.
Table 3-18. Phenotypic correlation coefficients between vase life and disc diameter before
and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.05
-.12**
-.26 *
.02
Diallel
487
.03
.02
.09
-.07
Significant at P< 0.01.


75
stem strength. Therefore, it follows that the fewer total number of vascular bundles per
unit circumference was observed in plants with high vase life and high stem strength.
Results from correlations determined between these stem anatomy variables and vase
life or stem strength enhance the link suggested between size and number of vascular
bundles to postharvest longevity. Fewer bundles, small or large, were found in plants with
high vase life or high stem strength. Negative correlations between stem anatomy variables
involving large bundles were significant for vase life but not stem strength at the level of
P< 0.05. Correlations between stem anatomy variables involving small bundles were not
significant for vase life or stem strength at the level of P< 0.05. However, negative
correlations between numbers of small bundles per unit area or unit circumference and stem
strength were significant at the level of P< 0.10.
Strong turgor strength in gerbera stems has been considered an important factor for
maintaining postharvest longevity (De Jong, 1978a). Assuming a constant supply of water
available to flowers with equal stem diameters, fewer vascular bundles would be expected
to increase the flow rate of water in the stem. It seems likely that this situation would serve
to maintain an upright stem, if stem rigidity depended on the maintenance of turgor.
Additionally, an increase in flow rate of water in the stem may reduce the opportunity for
microbial growth to occur, a factor cited to decrease vase life in gerbera (Meeteren, 1978a).
Since differences in large vascular bundles and not small vascular bundles were found in
plants with high and low levels of vase life, it is theorized that large vascular bundles are
mainly responsible for differences in water uptake rates in gerbera flowers.
As for the role of vascular bundles in stem strength, only small bundles appeared to be
a factor. De Jong (1986) proposed that stronger structural stem strength would help to
decrease the incidence of stem folding. Marousky (1986) examined stem anatomy,
including number and size of vascular bundles, of two cultivars whose flowers differed in
resistance to scape breakage. The cultivar that showed less resistance to breakage had more
small bundles than the cultivar that showed more resistance to breakage. One reason for


29
^Tabl^^S^x^S^^kL^VaseUfe^tafromdidlel^ros^Jrcdg^ds^TOl^)
Cross
(Diallel Generation)
Progeny Mean
83-1-77
X
83-4-69
10.3
83-1-77
X
83-5-109
8.7
83-1-77
X
83-7-4
11.8
83-1-77
X
83-7-10
11.2
83 -i
1-69
X
83-5-109
9.1
83
1-69
X
83-7-4
11.9
83 -i
1-69
X
83-7-10
10.6
83-!
5-109
X
83-7-4
11.4
83-:
5-109
X
83-7-10
12.5
83-'
1 4
X
83-7-10
14.3
Table 3-6 Summary of vase life for parental and diallel generation^
Generation
No. of
Plants
Mean
Std. Dev.
Coefficient of
Variation
Parental
325
7.82
2.64
33.77
Diallel
248
11.20
3.14
28.08
Heritabilitv
Combining ability analysis of variance using plant means was performed on vase life
data from a 5 x 5 diallel. General combining ability effects were significant. Specific
combining ability effects were non-significant. Results are summarized in table 3-7.
Heritability was estimated by the ratio of genetic variance (VA or Vq) to phenotypic
variance (Vp) (Falconer, 1960). Variances were derived using variance components for
general combining ability (a2gca), specific combining ability (a2^), and error (a2e)
according to formulae published by Hallauer (1981). Narrow sense heritability (h2 = .279)
and broad sense heritability (H2 = .281) estimates were approximately equal. This
indicates non-additive genetic variance (Vq VA) is negligble. Variances and heritability
estimates for this population are given in table 3-8.


B-l, 7x7 Diallel. Stem strength means based on progeny means of gerbera crosses.
83-1-10
83- 1 -31
83-1-96
83-4-8
83-5-76
83-7-48
83-8-7
83-1-10
x = 28.9
n = 16
x = 29.4
n = 16
x = 19.5
n = 16
x = 32.5
n = 16
x = 29.6
n = 16
x = 20.0
n = 16
83-1-31
x = 31.2
n = 16
x = 22.0
n = 11
x = 13.3
n = 16
x = 30.3
n = 16
x = 21.3
n = 16
x = 25.3
n = 16
83-1-96
x = 25.6
n= 16
x = 24.9
n = 14
x= 12.7
n= 15
x = 31.5
n = 14
x = 27.5
n = 16
x = 19.7
n= 15
83-4-8
x = 16.1
n = 16
x = 15.1
n = 15
x = 17.3
n = 15
x = 16.6
n = 16
x= 14.3
n= 16
x = 8.7
n = 16
83-5-76
x = 33.9
n = 16
x = 20.4
n = 16
x = 29.5
n= 13
x = 16.4
n= 15
x = 24.1
n= 15
x = 25.1
n= 13
83-7-48
x = 31.3
n = 16
x = 31.5
n = 16
x = 33.4
n= 13
x = 14.4
n = 16
x = 25.8
n= 15
x = 23.5=
n = 16
83-8-7
x = 20.3
n= 15
x = 24.4
n = 16
x = 24.8
n = 15
x = 12.9
n = 14
x = 23.1
n = 13
x = 24.6
n= 15


101
C-6. Phenotypic correlation coefficients between stem anatomy and three traits in gerbera
by stem strength (SS) ratings.
Stem Anatomy Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN
SS = high
-.11
.11
-.01
SS =low
.15
.07
.12
SBUN
SS = high
.02
.34* *
.42* *
SS = lw
-.09
.13
.17
TBUN
SS = high
-.02
.38 *
.40* *
SS = lw
.03
.15
.20*
LBUNCRAT
SS = high
.02
-.39 *
-.45 *
SS =low
-.12
-.53 *
-.67 *
SBUNCRAT
SS = high
-.01
-.44 *
-.41 *
SS =low
.00
-.20 *
-.24 *
TBUNCRAT
SS = high
.01
-.54 *
-.53 *
SS = low
-.06
-.42 *
-.53 *
LBUNARAT
SS = high
.07
-.63 *
-.70 *
SS =low
-.15 *
-.65 *
* 77 *
SBUNARAT
SS = high
.05
-.70 *
-.71 *
SS =low
-.08
-.57 *
-.63 *
TBUNARAT
VL = high
.06
-.75 *
-.78 *
VL = low
0.13
-.67 *
-.77 *
* Significant at P< 0.05. *
* Significant at P< 0.01.


105
D-3. 5x5 Diallel. General combining ability analysis of variance for five traits in
Source of Variation
df
Stem
Length
M.S.
Stem
Diameter
M.S.
Inflorescence
Diameter
M.S.
Disc
Diameter
M.S.
Vgrowth
M.S.
General combining ability
4
512.84 *
.1630 *
25.81 *
1.670**
3.99
n.s.
Specific combining ability
5
93.74 n-s-
.0121 *
0.46 n s-
0.042 ns-
1.18
n.s.
Error
238
67.40
.0035
0.92
0.040
0.84
* Significant at P< 0.05. Significant at P< 0.01.
D-4. 5x5 Diallel. Analysis of variance for five traits in gerbera.
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Source of Variation
df
M.S.
M.S.
M.S.
M.S.
M.S.
Among crosses
9
495.53 *
.1529 **
21.50**
1.640 *
6.70 *
Among plants
238
115.31 **
.0061 *
1.67 *
0.073 *
1.35 **
Within plants
239
36.96
.0035
0.46
0.029
0.72
* Significant at P< 0.01.


74
Discussion
The relationship between vase life and structural stem strength described in this study
was weak (r = .28) but, significant at the level of P< 0.05. Therefore, it would be
expected that flowers with the highest vase life will only have a slight tendency for high
stem strength. Breeding for both characters may improve keeping quality in gerbera cut
flowers only marginally.
Whether or not the correlation provides a good description of the relationship deserves
comment First of all, plant means for each variable were plotted against each other. Since
evaluating vase life and stem strength requires destructive tests, plant means for each
variable were determined by different sets of flowers. Also, vase life means were
determined from flowers evaluated during the period May 5 to June 12,1985, and stem
strength means were determined from flowers evaluated during the period June 12 to July
30, 1985. Environmental differences during these two periods may not have affected all of
the plants similarly. In addition, problems associated with the accuracy of stem strength
determinations (Chapter 4) may have skewed the comparison.
Variation in the number of vascular bundles per cross-section of gerbera flower stems
cut 12 cm below the peduncle was compared among plants whose flowers exhibited high
or low vase life and stem strength. Large and small bundles were counted. Their sum was
calculated to obtain a value for total number of bundles. The diameter of each cross-section
was also measured. Using the diameter measurement, the number of bundles per unit area
(1.0 cm2) or unit circumference (1.0 cm) for each cross-section was also calculated.
Overall, differences in the number of large bundles appears related to vase life. Fewer
number of large bundles, fewer number of large bundles per unit area, and fewer number
of large bundles per unit circumference were observed in plants with high vase life. In
contrast, only the difference in the number of small bundles per unit circumference appears
to affect stem strength. Fewer numbers of small bundles were observed in plants with high


85
Staden, O.L. and Beekhuizen, J.G., 1986. A new formula for the anti-ethylene
pretreatment of cut carnations. Acta Hort., 181: 425-428.
Stimart, D.P. and Brown, D J., 1982. Regulation of postharvest flower senescence in
Zinnia elegans Jacq. Scientia Hortic., 17: 391-396.
Stoddard, E.M. and Miller, P.M., 1962. Chemical control of water loss in growing
plants. Science, 137: 224-225.
Swan, A., 1981. Quality of Lilium Enchantment flowers as influenced by season and
silver thiosulphate. Acta Hort., 113: 45-50.
Sytsema, W., 1975. Conditions for measuring vase life of cut flowers. Acta Hort., 41:
217-226.
Sytsema, W., 1981. Vase life and development of carnations as influenced by silver
thiosulphate. Acta Hort., 113: 33-38.
Sytsema, W., 1986. Post-harvest treatment of freesia with silver thiosulphate and
cytokinins. Acta Hort., 181: 439-442.
Tandler, J., Mor, Y., Spiegelstein, H. and Mayak, S., 1986. Chemical treatments to
improve the quality of cut gypsophila flowers. Acta Hort., 181: 443-450.
Tesi, R., 1978. Variation of some characters of flowers in clones of Gerbera iamesonii
hvbrida. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and
Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.227-232.
Waters, W.E., 1964. Influence of chemical preservatives on keeping quality of asters,
carnations, chrysanthemums, and gerbera daisies. Proc. Fla. St. Hort. Soc., 77: 466-470.
Wilberg, B 1973. Physiological investigations on the problem of freedom from stem
break as a prerequisite in breeding gerbera forms which remain upright when used as cut
flowers. Z. Pflanzenzuchtung, 69: 107-114.
Woltering, E.J., 1986. Postharvest treatment of Gvpsophila paniculata. Acta Hort., 181:
338.
Wricke, G., Weber, W. E. and Ottleben, R., 1982. Die analyse der genetischen varianz
bei quantitativ vererbten Merkmalen von Gerbera iamesonii. Z. Pflanzenziichtung, 89:
329-336.
Wulster, G., Sacalis, J. and Janes, H.W., 1982. Senescence in isolated carnation petals.
Plant Physiol., 70: 1039-1043.
Zagory, D. and Reid, M.S., 1986. Evaluation of the role of vase micro organisms in the
postharvest life of cut flowers. Acta Hort., 181: 207-218.
Zieslin, N., Kohl, H.C., Kofranek, A.M. and Halevy, A.H., 1978. Changes in the water
status of cut roses and its relationship to bent-neck phenomenom. J. Amer. Soc. Hort.
Soc., 103: 176-179.


65
Materials and Methods
Selection and Classification
Seventy-three plants were evaluated for both stem strength and vase life in 1984
(Chapter 4). Three flowers per plant were evaluated to obtain mean values for each
variable. Based on these means, 22 plants were selected and classified into four categories.
The categories were defined by vase life and stem strength ratings described in table 5-1.
To represent genetic diversity, selection included plants from five seed populations.
Table 5-1. Vase life x stem strength classification
Plant
Category
Vase Life
Stem Strength
I
High ^ 11.0 days
High ^ 11.5g/.2cm
II
High > 12.0 days
Low < 5.5g/.2cm
III
Low <6.0 days
High ^ 11.3g/.2cm
IV
Low ^ 4.0 days
Low ^2.5g/.2cm
Stem anatomy of flowers from plants assigned to each category was compared
(February 20-July 30, 1985). Approximately 14 flowers per plant were examined.
Appleblossom, assigned to category I, was also examined because of its low intraplant
variation for vase life and stem strength (Chapters 3 and 4).
Anatomy Examination
Flowers at various stages of maturity with at least one row of disc florets open were
randomly sampled. Using a light microscope with 40x magnification, fresh stem sections,
approximately 50-60 microns thick, were cut, using a razor blade, 12 cm below each
peduncle and examined. No stains were used. The number of large and small vascular
bundles was recorded for each flower. Determination of large and small bundles was
relative between plants. Therefore, before counting, the smallest large bundle was


16
The use of tetraploids was suggested as another means to improve postharvest
longevity in tulips. Eijk and Eikelboom (1986) also reported that tetraploids have been
created as a result of applying lnitrous oxide (N20) to diploid cultivare. Progeny from
tetraploid crosses showed improvement in vase life compared to their parents.
Improving postharvest longevity in cut flowers of Gerbera jamesonii by breeding and
selection has been considered possible for many years. Smith and Nelson (1967) noted
differences in vase life among cut flowers of gerbera. They suggested selection and
breeding could minimize this variation. Kohl (1968) proposed selecting cultivare with an
increased ability to uptake water and that cultivare having structural stem strength are
requisite for maintaining the popularity of gerbera as a cut flower.
Wilberg (1973) observed differences in the frequency of stem bending among gerbera
cultivare, and noted the need for breeding stems that remain upright in gerbera cut flowers.
Three factors were identified which contributed to stem strength: the ratio of dry
weight/cm. in the stem section prone to bending should be greater than the rest of the stem;
stem elongation of harvested flowers should be small; and water content in the stem section
prone to bending should be low. It was suggested all three factors should be included as
part of a selection program. It was also noted thick stems were not necessarily more
resistant to bending.
Meeteren (1978a) reported stem break in gerbera was greater during the summer than
winter months. Barigozzi and Quagliotti (1978) noted tetraploids appeared to have stronger
stems than diploids. These researchers were unable to find a relationship between flower
color and vase life. Serini and De Leo (1978) supported this finding and reported no
correlation between vase life and stem length, inflorescence diameter, and number of
ligulae. They showed, however, vase life may be increased by a higher proportion of dry
substance to water in the stem or a greater number of small vessels in the stem. Vase life
and flower yield were reported to be negatively correlated. Tesi (1978) also reported vase
life and flower yield were negatively correlated.


GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF GERBERA X HYBRIDA HORT.
By
HEIDI CAROL WERNETT
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1990

In memory of my father and mother,
Benjamin Carl Wemett and Evelyn Carolyn Wemett,
who inspired my appreciation for flowers.

ACKNOWLEDGMENTS
I wish to thank the members of my committee for their professional guidance in
helping me to conduct this research. Dr. T. J. Sheehan, Dr. G.J. Wilfret, Dr. F.J.
Marousky, Dr. P. M. Lyrene, and Dr. D.A. Knauft. Their patience is most appreciated.
Other debts of gratitude are acknowledged to Dr. R. C. Fluck from the Dept, of
Agricultural Engineering for his help with stem strength measurements, Dr. C.J. Wilcox
from the Dept, of Dairy Science and Dr. T. L. White and his research assistant, Greg
Powell, from the Dept, of Forestry for their critical help with data analysis, and Dr. F.G.
Martin from IFAS Statistics whose suggestions were invaluable.
Personal thanks is also deeply extended to the Acuff family, Mark and Shari Wilson,
and Terry J. Smith. Without their continued support, encouragement, sweat and toil, I
could still be in the greenhouse; watering, transplanting, and taking measurements. The
generous hospitality of my friends shall never be forgotten. In addition, a special "thank
you" is required to show my appreciation to Terry for his kindness and assistance during
the lengthy period of manuscript preparation.
There are many other individuals who advanced the completion of this research over
the past seven years. I am sincerely grateful to all of them for their help.

TABLE OF CONTENTS
ACKNOWLEDGMENTS iii
ABSTRACT vi
CHAPTERS
1 GENERAL INTRODUCTION 1
2 LITERATURE REVIEW 3
3 PART I. VASE LIFE STUDIES 20
Introduction 20
Materials and Methods 21
Plant Material 21
Selection and Mating. 23
Vase Life Evaluation 25
Quantitative Analysis 26
Results 28
Selection and Mating. 28
Heritability 29
Senescence Patterns 31
Correlations between Vase Life and Other Traits 38
Discussion 42
Conclusioa 46
4 PART II. STEM STRENGTH STUDIES 48
Introduction 48
Materials and Methods 48
Selection and Mating. 48
Stem Strength Evaluation 49
Production 51
Quantitative Analysis 51
Results 53
Selection and Mating. 53
Heritability 56
Correlations between Stem Strength and Other Traits 58
Discussion 59
Conclusioa 63
5 PART III. VASE LIFE X STEM STRENGTH STUDIES 64
Introduction 64
Materials and Methods 65
iv

V
Selection and Classification 65
Anatomy Examinatioa 65
Results 66
Correlation Between Vase Life and Stem Strength 66
Comparison of Stem Anatomy Among Genotypes 67
Correlations Between Stem Anatomy and Postharvest Longevity 72
Discussion 74
Conclusion. 76
6 SUMMARY 77
REFERENCES 79
APPENDICES
A VASE LIFE STUDIES 86
B STEM STRENGTH STUDIES 95
C VASE LIFE X STEM STRENGTH STUDIES 97
D HERITABILITY STUDIES FOR OTHER TRAITS IN GERBERA 102
E CORRELATIONS AMONG OTHER TRAITS IN GERBERA 113
BIOGRAPHICAL SKETCH 115

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
GENETICS AND BREEDING OF POSTHARVEST LONGEVITY
IN CUT FLOWERS OF GERBERA X HYBRIDA HORT.
By
Heidi Carol Wemett
May 1990
Chairman: Thomas J. Sheehan
Cochairman: Gary J. Wilfret
Major Department: Horticultural Science
Two components of postharvest longevity, vase life and stem strength, were studied in
cut flowers of Gerbera X hvbrida Hort.. A broad based source of germplasm was
evaluated initially. Progeny means for vase life resulting from a topcross between 31
plants and Appleblossom were used to select plants whose flowers had high vase life.
Analysis of a 5 x 5 diallel cross was made to determine vase life heritability. No reciprocal
effects were observed. Heritability ranged from 22 to 39 percent. Additive gene action
possibly controls this character since broad sense and narrow sense heritability was nearly
equal. Repeatability was moderately high (r = .57).
A senescence mode was recorded for each flower, i.e. bending or folding of the scape
or wilting of the ligulae. Flower was defined as the composite inflorescence and its
corresponding stem. After selection and mating, frequency of bending decreased;
frequency of folding and wilting increased. Flowers that senesced due to wilting had the
highest mean vase life before and after selection and mating. Significant correlations were
found between vase life of flowers that senesced due to wilting and shorter stem length,
thinner stems, and smaller infloresences.
vi

Vil
A 7 x 7 diallel using plants whose flowers had high stem strength, measured using an
Instron, was used to determine stem strength heritability. No reciprocal effects were
observed. Heritability ranged from 28 to 53 percent Additive gene action possibly
controls this character since broad sense and narrow sense heritability was nearly equal.
Repeatability was moderately high (r = .60). Significant correlations were found between
stem strength and shorter stem length, thicker stems, and heavier infloresences.
Number of vascular bundles was compared among plants whose flowers exhibited
high or low levels of vase life and stem strength. Significant differences between high and
low levels of vase life were observed for number of large bundles per 1.0 cm. scape
circumference, measured 12 cm. below the peduncle; significant differences between high
and low levels of stem strength were observed for number of small bundles per 1.0 cm.
scape circumference.

CHAPTER 1
GENERAL INTRODUCTION
Gerbera X hvbrida Hort. is a popular cut flower. Developed from hybrids of
interspecific crosses between Gerbera iamesonii Bolus and Gerbera viridifolia Sch., and
possibly other species, it is a showy and handsome flower of the daisy family,
Compositae. The distinctive inflorescence, generally measuring 8-14 cm. in diameter,
has strap-shaped, ray floret ligulae that are bright shades of orange, yellow, or red,
surrounding a yellow or black colored center of disc florets. This eye-catching flower is
supported by a long, leafless, and upright stem, known as a scape.
The major postharvest problem in cut flowers of gerbera relates to the length of time
until senescence occurs; when the ligulae wilt or when the stem no longer remains upright.
Ideally, postharvest longevity in gerbera should be two weeks or longer. Unfortunately for
the consumer, vase life is often less because the stem ceases to remain upright.
A critical attribute of cut flowers is postharvest longevity; therefore, research to
improve the lasting quality of gerbera is important. Postharvest longevity of cut flowers,
including gerbera, can be extended by postharvest treatments but the extent to which these
treatments can improve lasting quality may be limited by the plant genome. At present,
evaluation for postharvest longevity is not routinely done in cut flower breeding programs.
If gerbera cultivars with superior postharvest longevity are to be developed, then it is
necessary to identify characteristics which are suitable for selection and breeding.
The main purpose of this research was to evaluate the potential of plant breeding as a
method to improve postharvest longevity in gerbera, using a broad based source of
germplasm. Specifically, objectives of this study were to estimate heritability for vase life
and stem strength by diallel analysis, to observe changes in frequency of senescence
1

2
patterns due to selection and mating, to determine correlations between vase life, stem
strength, and other traits of flower and stem morphology and growth, and to examine stem
anatomy of genotypes that differ in vase life and stem strength.

CHAPTER 2
LITERATURE REVIEW
Researchers have investigated and discussed postharvest longevity in cut flowers
including Gerbera X hybrida Hort.:
The length of lasting quality in itself is not the aim of postharvest longevity
but the satisfaction of the consumer. Premature senescence due to abnormal
causes will leave the consumer with a sense of frustration. (Buys, 1978,
p. 256).
Bent neck in cut roses is a loss of pedicel rigidity which prematurely terminates the
vase life of the flower (Burdett, 1970). Similar to bent neck, knicking (Wilberg, 1973;
Buys, 1978), folding (De Jong, 1978a), neck droop (Zieslin et al., 1978), or stem
break (Meeteren, 1978a) are terms used to describe the sudden bending of the stem in cut
gerberas. Premature senescence in cut flowers contrasts with natural senescence. In
gerbera, this natural phenomenon is identified as wilting (De Jong, 1978a). Wilting
was described as the condition that occurs when the ligulae of an inflorescence on an
upright stem have visibly lost their tugidity.
Vase life is often used as an indicator of postharvest longevity in cut flowers,
including gerbera. Vase life is determined by the number of days from harvest until flower
senescence, whether or not senescence is considered premature. Sytsema (1975) specified
six important factors which may affect the measurement of vase life in cut flowers:
1. Flower conditioning. Harvested flowers allowed to regain turgidity following
storage and/or transpon can withstand adverse room conditions better than
unconditioned flowers.
2. Temperature. Vase life is usually shoner as temperature increases. Small
deviations from 20 C may be optimum.
3

4
3. Relative air humidity (RH) and air circulation speed. To reduce excessive
transpiration, RH should be at least 50% and air circulation speed should be
very low.
4. Light Low light intensity is satisfactory, but total darkness may be less than
ideal.
5. Ethylene and air exchange. Good ventilation is the best method to avoid any
harmful effects due to ethylene concentration.
6. Use of bactericides. Bacterial contamination will interfere with reliable vase
life determinations, therefore a bactericide is recommended. In all cases, exact
room conditions and experimental procedures should be recorded as a basis for
experimental comparison.
Reid and Kofranek (1981) also summarized some recommendations for standardizing vase
life evaluations:
1. Use of vase solution control. Distilled water or deionized water treated for
removal of microorganisms and colloidal materials should be used. Flowers
should be placed individually in sterile containers.
2. Temperature. A good standard is 20 2 C.
3. Light Light intensity of 600 mW/cm2 with a 12 hour diurnal cycle is practical
for simulating home conditions.
4. Relative humidity. A good standard is 60-70%.
5. Air circulation. One air exchange every two hours and wind velocity should
not exceed 0.5m/sec. Stages of flower maturity at time of harvest should be
well documented.
Historically, most postharvest research on cut flowers has focused on environmental,
metabolic, and to a lesser extent, anatomical factors that affect lasting quality.
Environmental treatments have extended the vase life of cut flowers and/or reduced the
incidence of premature senescence.
As early as 1962, it was suggested that the use of chemicals reduced wilting in plants.
Experiments with strawberry plants showed applications of 8-hydroxyquinoline sulfate (8-
HQS) resulted in stomatal closing even when conditions were favorable for opening
(Stoddard and Miller, 1962). Cut flowers of the rose Forever Yours were reported to last

5
twice as long in an aqueous solution containing 200 ppm. of 8-HQS, an anti-bacterial
agent, than in sterile water (Burdett, 1970).
By the late 1960s, many new postharvest methods and chemical additives were being
tested as a means to extend vase life of major cut flower crops. Heide and 0ydvin (1969)
confirmed cytokinins used as a postharvest dip following storage were beneficial in
prolonging vase life of cut carnations. As concentrations of cytokinin 6-benzylamino
purine (BAP) increased, immersion times should decrease; otherwise, detrimental effects
would result. Application of cytokinin to cut roses was also reported to increase vase life
(Mayak and Halevy, 1970; Mayak et al., 1972).
Kofranek and Paul (1975) confirmed adding silver to preservative solutions was
effective in extending postharvest longevity in chrysanthemums and carnations, but not
gladiolus or gerbera. Mayak and Dilley (1976) found vase life of cut carnations was
enhanced by solutions containing kinetin and sucrose compared with solutions containing
only kinetin. Paulin and Muloway (1979) concluded that cytokinins are best used as a
pretreatment, followed by a glucidic solution to increase the vase life of cut flowers.
Swart (1981) reported vase life of Lilium Enchantment was improved by using a
pretreatment of silver thiosulphate (STS). Sytsema (1981) found STS generally more
effective in extending vase life of standard carnations than spray carnations. Kofranek and
Halevy (1981) suggested a quaternary ammonium product, Physan, could be substituted
for silver nitrate (AgNC^) as a chemical pulsing agent in chrysanthemums. Fujino et al.
(1981) reported that use of aminooxyacetic acid (AOA) as a vase solution additive extended
postharvest longevity of carnations. It was also shown that pulsing carnations with a
higher concentration of AOA was effective. Staden and Beehuizen (1986) developed a
pretreatment formula containing gibberellic acid (GA3), kinetin, daminozide, AOA, and
detergent (Triton X-100) to be another chemical alternative to STS for extending the vase
life of standard and spray carnations. Reddy (1988) reported solutions containing cobalt
salts increased vase life of cut roses.

6
Experimenting with chemical additives to extend the vase life of minor cut flower
crops was gaining momentum by the 1980s. Pauli et al. (1981) reported that use of the
preservative Floralife delayed flower wilting and leaf blackening in protea. Stimart and
Brown (1982) demonstrated optimal vase life for cut flowers of zinnia could be obtained
using a holding solution containing 200 ppm. of 8-hydroxyquinoline citrate (8-HQC) and
1% sucrose. Positive effects of floral preservative on postharvest quality of gypsophila
were initially documented by Marousky and Nanney (1972). Tandler et al. (1986) found
quaternary ammonium compounds were also effective in prolonging the vase life in the cut
flower gypsophila. Postharvest treatments of STS or sugar plus a bactericide were
reported to double the vase life of gypsophila (Barendse, 1986). Postharvest longevity
was shown to increase in cut flowers of calendula, zinnia, and snapdragon if pulsing
treatments of AgNC>3 were utilized. Sweet pea and delphinium lasted longer if STS pulsing
treatments were given (Awad et al., 1986). Sytsema (1986) reported STS prolonged vase
life in cut ffeesia, but if STS was used in combination with cytokinins, vase life would be
further enhanced. Kalkman (1986) stated cut flower preservatives would increase
postharvest longevity in astilbe. Leeuwen (1986) proposed a combination of growth
regulators and STS may be the most effective in prolonging vase life in Euphorbia fulgens.
Downs and Reihana (1987) found preservative solutions enhance vase life of nerine cut
flowers.
Research specifically dealing with postharvest treatments to increase lasting quality in
cut flowers of Gerbera iamesonii has been reported. Waters (1964) reported treatments
with flower preservative Everbloom increased vase life in gerberas 3-5 days. Kohl (1968)
concluded a floral preservative should be used to increase vase life in gerbera, and that
stems should be recut just prior to use. Marousky (1975) cautioned that vase solutions
using fluoridated water can reduce postharvest longevity in gerbera. Meeteren (1978a)
suggested the susceptibility to stem break in some gerbera cultivars could be prevented by
using bactericides in vase solutions. Nowak (1981) reported vase life of gerbera may be

7
prolonged if a conditioning solution containing AgNOj, 8-HQC, and sucrose is applied for
20 hours before cold storage. A pretreatment application of STS was reported to extend the
vase life of gerbera cut flowers later exposured to high concentrations of exogenous
ethylene. Also, application of an ethoxy analog of rhizobitoxine (AVG) was reponed to
reduce endogenous ethylene production resulting in a slight increase in gerbera vase life,
especially if used together with a flower preservative (Nowak and Plich, 1981). Abdel-
Kader and Rogers, (1986) concluded that 8-HQS was preferable to AgNC>3 as a
component of flower preservatives used to enhance vase life of gerbera. Combining one of
these anti-microbial agents with sucrose was shown to decrease the incidence of stem
break. Amariutei et al. (1986) found pulsing treatments of sucrose, 8-HQS, and AgN3
increased vase life of gerbera Symphonie and Richard.
In addition to chemical treatments, other postharvest handling measures have been
investigated as a means to increase lasting quality in cut flowers. Carpenter and
Rasmussen (1974) proposed fewer leaves remaining on roses could reduce the percentage
of flowers developing bent neck. Dilley and Carpenter (1975) emphasized chemicals could
not completely overcome premature senescence in cut flowers. Ferreira and Swardt
(1981a) reported defoliated cut flowers of the rose Sonia lasted longer in deionized water
than in flower preservative. Woltering (1986) concluded defoliated cut roses would be less
likely to develop bent neck if placed in a vase solution containing a flower preservative or
bactericide. Barendse (1986) reponed optimum vase life in gypsophila could be achieved if
the lower 10 cm. of the stems were defoliated. Meeteren (1978a) suggested the vase life of
gerbera could be extended if the solid basal portion of the stem was removed, permitting
water to enter the cavity of the stem. Later Nowak and Plich (1981) confirmed vase life of
gerbera cut flowers could be enhanced if the basal portion of the stem was removed.
Metabolic causes for senescence in cut flowers have been discussed voluminously
during the past 40 years. Siegelman (1952) asserted that since a decrease in respiration rate
reduces postharvest longevity in some fruits and vegetables, a similar decrease in

8
respiration rate could be the cause for senescence in cut flowers. The answer to the
question, Does respiration rate affect cut flower senescence? remains elusive.
Siegelman (1952) demonstrated in experiments with roses and gardenias that when
respiration rate was reduced, vase life increased. MacLean and Dedolph (1962) observed
application of N6-benzylaminopurine (Verdan) reduced respiration rate in carnations and
chrysanthemums and increased vase life. Kuc and Workman (1964) concluded a direct
relationship existed between respiration rate and postharvest longevity in cut flowers
because the respiration rate was three to four times greater in carnations than in
chrysanthemums and postharvest longevity was much lower in carnations than
chrysanthemums. Coorts, et al. (1965) reported that respiration rates increased in cut roses
after being treated with a flower preservative. Treated roses lasted longer than those which
were untreated. Gilbart and Dedolph (1965) treated cut roses with N6-benzyladenine (N6-
BA). They observed the respiration rate generally increased in petals and decreased in
leaves following treatment. Ballantyne (1966) asserted variation in the respiration rate of
daffodil cut flowers was more likely a result of senescence than a cause for senescence.
Heide and 0ydvin (1969) suggested decreasing respiration rate in carnations will not delay
senescence. Larsen and Frolich (1969) studied the relationship between respiration rate
and water uptake in cut flowers. They observed that water uptake and respiration
decreased simultaneously in carnation Red Sim. Coorts (1973) reviewed factors
contributing to metabolic changes in cut flowers which affect senescence. He concluded
senescence can be delayed by using respiratory inhibitors and controlling hydrogen ion
activity. It was proposed that if the pH of vase solutions was maintained between 3.0 -
5.0, vase life would be extended. It was also suggested mitochondrial activity may be
linked to respiration rates of senescing cut flowers. Mayak and Halevy (1980) reviewed
the subject of flower senescence and suggested that petal senescence would be an ideal
system to study. Ferreira and Swardt (1981b) reported they found no correlation between
respiration rate and vase life of cut roses. Changes in respiration rate depended on the

9
stage of senescence. Amariutei et al. (1986) observed higher respiration rates in gerbera
cut flowers that were treated with a pulsing solution.
The role of plant hormones, particularly ethylene, has also been a source of
investigation for understanding cut flower senescence. Mayak and Halevy (1970) reported
exogenous application of N^-BA can delay senescence in rose petals. Mayak et al. (1972)
demonstrated exogenous applications of cytokinins increase postharvest longevity in roses
whereas ethylene or abscisic acid (ABA) was shown to decrease postharvest longevity.
They also suggested the presence of ethylene may trigger the production of ABA. Rogers
(1973) reviewed the effects of ethylene synthesis and other plant hormones on cut flowers
and concluded that anti-senescence factors must be applied to attain maximum vase life in
cut flowers.
Nichols (1981) summarized results that indicated ethylene biosynthesis plays an
important role in flower senescence, but the magnitude of its effect depends upon plant
genera. Nowak (1981) stated a decrease in vase life of cut gerbera flowers may be caused
by ethylene. Wulster et al. (1982) discussed the possibility of auxins increasing ethylene
biosynthesis, thereby inducing senescence in carnation petals. Nichols (1982) reviewed
the effects of growth regulators on cut flower senescence and concluded inhibiting
endogenous production of ethylene will increase vase life in carnation, iris, daffodil, and
chrysanthemum. Bufler (1986) stressed that studies should include investigating factors
which control response to ethylene production, not just a measure of ethylene production.
Reddy (1988) suggested cobalt salts may inhibit ethylene synthesis in cut roses.
A major factor considered to influence senescence in cut flowers, including gerbera, is
the balance between water uptake and transpiration. Marousky (1969) verified reducing
moisture stress by increasing water absorption will improve postharvest longevity in roses.
Burdett (1970) proposed water loss as one of the causes of bent neck in cut roses. He
specified water deficits resulting from transpiration are probably of small importance
compared to an impairment in the water conducting system. Carpenter and Rasmussen

10
(1973) suggested transpiration rates have a close relationship with water uptake. Leaf area
of cut roses was found to be the factor most closely associated with water uptake in both
light and dark. Carpenter and Rasmussen (1974) concluded the number of stomates on the
leaves and stems of cut flowers affects the rates of water uptake and loss. They observed
when water loss from transpiration exceeds water uptake, vase life is reduced. It was
proposed that by reducing transpiration, bent neck in roses would be reduced.
Mayak and Halevy (1974) found kinetin delayed wilting in cut roses. It was observed
that water uptake increased, though transpiration also increased, due to stomatal opening.
Bravdo et al. (1974) studied vase life in gladiolus. They proposed water uptake can be
improved by increasing osmotic concentration of florets and leaves with the addition of
sucrose to vase solutions. Marousky and Woltz (1975) also examined vase life of
gladiolus. Though the addition of 8-hydroxyquinoline citrate plus sucrose was found to
improve water uptake, the potential for fluoride toxicity was also increased depending on
the concentration of fluoride in the water. They also reported that gerberas were highly
sensitive to low levels of fluoride, especially with the addition of 8-hydroxyquinoline
citrate plus sucrose to the vase solution which increased water uptake. Chrysanthemums
and snapdragons did not exhibit this degree of sensitivity. Halevy (1976) stated water
stress is the most common reason for decreased vase life in cut flowers. He asserted the
basis for water imbalance is a decrease in water potential, water uptake, water loss, and
water conductivity. Zieslin et al. (1978) postulated bent neck in cut roses occurs due to
water stress conditions such as increased transpiration rates and decreased water uptake
rates. Reddy (1988) suggested cobalt may increase vase life of cut roses by causing
stomates to partially close, thereby reducing transpiration, yet maintaining water uptake.
A major reason cited for poor vase life in gerbera is the inability of the cut flowers to
imbibe sufficient water (Kohl, 1968). It was suggested as gerbera flowers age, water
holding capacity of the petals decreases (Meeteren, 1978b). As water content decreased in
gerbera petals over time, ion leakage increased (Meeteren, 1979). It was proposed that by

11
increasing pressure potential in gerbera petal cells, ion leakage will decrease, resulting in
longer vase life (Meeteren, 1980).
Researchers have also focused on the presence of vessel occlusion in the stem of cut
flowers as a cause for early senescence. Two types of occlusion have been proposed.
Occlusion due to microbes is considered to be a primary cause of bent neck in roses or stem
break in gerberas. The use of a bactericide in vase solutions is a common practice in
postharvest handling of cut flowers. The second type of occlusion is regarded as a result
of physiological plugging.
Durkin and Kuc (1966) postulated a vascular block resulting from harvesting injury
was the primary cause for premature senescence in cut roses. Marousky (1969) concurred
vascular blockage was responsible for decreasing the vase life of cut roses. Burdett (1970)
showed bent neck in roses coincided with the appearance of material which plugged xylem
vessels. He postulated two possibilities for plugging: growth of microorganisms or a gum
deposition which could be the result of pectin degradation products. It was also suggested
when water uptake is deficient, sufficiently lignified stem tissue could prevent bent neck in
cut roses. After eliminating the presence of microorganisms, Marousky (1972)
demonstrated that in addition to moisture deficiency, a blocking mechanism reduces
postharvest longevity in cut flowers. Parups and Molnar (1972) investigated the nature of
vascular blockage in cut roses histochemically. They reported evidence of carbohydrates
with sulfate, carboxyl, or phosphoryl groups, pectin-, lipid-, or other protein-like
compounds, and some enzymes. They found no evidence of tannins, lignin, and callse in
blocked xylem vessels. Rogers (1973) reviewed the literature concerning effects of
physiological or microbial induced stem plugging but concluded further study was
necessary to determine the cause of physiological plugging. He also suggested microbial
plugging may have significance only with flower genera that typically last longer since it
takes time for microbial populations to develop.

12
Carpenter and Rasmussen (1973) investigated the possibility of plugs occurring during
light or dark periods that reduced water uptake rates. They observed no additional tissue
degradation under daylight. Rasmussen and Carpenter (1974) used scanning electron
microscopy to observe vascular occlusions in cut roses. They found vascular blockage
affects vase life after the cut flower is physiologically incapable of maintaining adequate
water balance; i.e. water loss exceeds water uptake. Mayak et al. (1974) postulated
transpiration plays a more significant role in wilting of cut roses than vascular blockage.
Lineberger and Steponkus (1976) observed two types of vascular occlusion in cut
roses. Microbial occlusions were found in the lower portion of the stem, and gum
deposition was located in the stem above the solution level. Parups and Voisey (1976)
reported the resistance to bending in cut roses is related to lignin content. It was concluded
bent neck will occur if water stress occurs and stem lignification is insufficient. Zieslin et
al. (1978) stated resistance to bending in cut roses depends particularly on secondary
thickening of the vascular system and stem lignification. Zagory and Reid (1986) studied
the role of microorganisms in reducing vase Ufe of carnation Improved White Sim. Only
three of 25 microorganisms isolated from vase solutions reduced vase life. They suggested
that ethylene-producing bacteria may be a possible factor in reducing vase life. De Witte
and Doom (1988) postulated exogenous concentration of pseudomonas bacteria or
Alcaligenes faecalis may account for vascular blockage in roses after three or more days,
but endogenous bacteria in stems or air emboti may cause vascular blockage earher. They
did not find evidence of pectolytic breakdown in xylem cell walls. Dixon et al. (1988)
indicated vase Ufe of cut roses may be proportional to the loss of water conducting capacity
caused by disfunctional xylem tissue. They proposed vase life may be lower in flowers
having a greater proportion of large vascular bundles than small vascular bundles because
larger bundles become disfunctional earlier than smaller bundles. Dixon and Peterson
(1989) concluded physical vascular blockage in the stem of cut roses initially decreased
water uptake, but xylem disfunction induced by water stress reduced vase life over time.

13
Two pathways for water uptake in gerbera were proposed: a direct path through xylem
vessels and an indirect path through the stem cavity. It was suggested stem break occurs
when water uptake is inhibited by bacterial growth (Meeteren, 1978a).
Some researchers have studied stem anatomy in an attempt to understand the causes
for reduced postharvest longevity in gerbera. Reiman-Philip, as cited by Wilberg (1973),
noted a larger proportion of large vascular bundles to small vascular bundles as a possible
factor contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that
such a relationship existed in only extreme cases. Steinitz, as cited by Marousky (1986)
observed increases in phloem cell wall thickening and lignification in gerbera scapes when
flowers were placed in a sucrose solution. In a comprehensive study, Marousky (1986)
measured the size and number of vascular bundles in two gerberas, Tropic Gold and
Appleblossom. He concluded Appleblossom exhibited more resistance to stem bending
than Tropic Gold, partially because of various anatomical features such as fewer small
vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference,
and a greater percentage of dry weight per unit of scape length. It was emphasized these
factors must be considered concurrently with variation in moisture stress to understand the
cause of stem breakage.
Dubuc-Lebreux and Vieth (1985) also studied the histology of the gerbera stem. They
postulated stem bending is linked to deficiencies in supportive elements in the stem. It was
concluded that sensitivity to stem breakage depends on the degree of maturity of the stem
approximately 10 cm. below the flower head at time of harvest. Marousky (1986) also
noted lignification was greater in the lower portion of the stem compared to the upper
portion of the stem where stem breakage is most likely to occur.
Despite postharvest treatments favorably affecting lasting quality in cut flowers,
including gerbera, reliable expectations of vase life have not been achieved. Though
metabolic and anatomical research has provided some clues to the cause of senescence in
cut flowers, another approach for learning how to improve postharvest longevity exists.

14
The genetic basis for variation in postharvest longevity in cut flowers has been investigated;
genotypic differences among plants examined, characteristics for selection proposed, and
heritability estimated.
Mayak and Halevy (1970) observed differences in endogenous cytokinin levels
between roses Lovita and Golden Wave. Cytokinin levels were higher in Lovita,
which has a longer vase life than Golden Wave. Marousky (1973) suggested selecting
chrysanthemums for succulent stems that readily translocate water. Mayak et al. (1974)
reported transpiration rates were higher in rose cultivis with short vase life. They
concluded these cultivars had less ability to close their stomates under water stress than
cultivars with long vase life. Zieslin et al. (1978) studied the sensitivity of rose cultivars to
neck droop. They concluded several factors contribute to variation in premature senescence
in roses: transpiration rate, initial water uptake rates, and different organs of a flower
competing for water at different rates. It was also suggested the effect of these factors may
depend on the structural strength of the stem due to lignification or other anatomical factors.
Pauli and Criley (1981) observed clonal differences in vase life among protea cultivars.
Ferreira and Swardt (1981b) concluded the vase life of rose cultivars varies due to genetic
differences. They proposed three factors are genetically controlled: ability to store
carbohydrates, ability to utilize an exogenous supply of sucrose, and susceptibility to bent
neck. Stimart and Brown (1982) observed differences in vase life among zinnia cultivars
when held in solutions containing 8-hydroxyquinoline citrate and sucrose.
Selection and breeding for postharvest longevity in tulips has been extensively
investigated. Twenty years ago, it was declared that by knowing the cultivar name of a
tulip, its vase life could be predicted (Benschop and De Hertogh, 1969). This implies that
genetics plays a significant role in determining the vase life of tulips.
Eijk and Eikelboom (1976) studied the possibility for selecting for postharvest
longevity in tulips. Cut flower tulip vase life was described by three senescence
characteristics, beginning with the number of days from flowering to start of tepal

15
discoloration, 50% of tepal discoloration, and perianth drop. All three characteristics were
considered effective determinations of vase life, but perianth drop was easiest to measure.
Also, evidence was reported that discoloration may occur earlier in some genotypes
depending on the pigment composition of their flowers, but that this does not necessarily
lead to earlier perianth drop. It was observed that some flowers with much carotenoid,
delphinidin, or cyanidin and little pelargonidin discolor earlier than flowers with little
carotenoid and much pelargonidin. It was concluded, however, that in a breeding program
to improve postharvest longevity, all characteristics of vase life must be considered.
Eijk et al. (1977) confirmed earlier observations that cut flower vase life in tulips could
be selected by evaluating flowers remaining attached to the plant, except in cases where
response to flower preservatives is being evaluated. They also found vase life of field
grown tulips was less predictable than vase life of greenhouse grown tulips. Therefore, it
was suggested that field trials should be utilized only for initial screening. Several
correlations were reported: (1) an increase in plant height during the growing period was
not correlated to vase life; (2) an increase in growth of the last intemode (stem elongation)
did not appear correlated to vase life; and (3) vase life and water uptake were significantly
correlated.
Eijk and Eikelboom (1980) demonstrated, using combining ability analysis, that
phenotypic observation was effective in predicting the results of crossing a set of parents in
order to improve postharvest longevity in tulips. It was postulated additive gene action
controls three senescence characteristics: start of discoloration, 50% discoloration, and
perianth drop.
Eijk and Eikelboom (1986) investigated the influence of temperature on selecting for
vase life in tulips. Variation in vase life for some genotypes was much greater depending
on the temperature during the evaluation period. They recommended that selection at
17 C. provides an average response for each genotype; however, final selection should
include screening at temperature extremes.

16
The use of tetraploids was suggested as another means to improve postharvest
longevity in tulips. Eijk and Eikelboom (1986) also reported that tetraploids have been
created as a result of applying lnitrous oxide (N20) to diploid cultivare. Progeny from
tetraploid crosses showed improvement in vase life compared to their parents.
Improving postharvest longevity in cut flowers of Gerbera jamesonii by breeding and
selection has been considered possible for many years. Smith and Nelson (1967) noted
differences in vase life among cut flowers of gerbera. They suggested selection and
breeding could minimize this variation. Kohl (1968) proposed selecting cultivare with an
increased ability to uptake water and that cultivare having structural stem strength are
requisite for maintaining the popularity of gerbera as a cut flower.
Wilberg (1973) observed differences in the frequency of stem bending among gerbera
cultivare, and noted the need for breeding stems that remain upright in gerbera cut flowers.
Three factors were identified which contributed to stem strength: the ratio of dry
weight/cm. in the stem section prone to bending should be greater than the rest of the stem;
stem elongation of harvested flowers should be small; and water content in the stem section
prone to bending should be low. It was suggested all three factors should be included as
part of a selection program. It was also noted thick stems were not necessarily more
resistant to bending.
Meeteren (1978a) reported stem break in gerbera was greater during the summer than
winter months. Barigozzi and Quagliotti (1978) noted tetraploids appeared to have stronger
stems than diploids. These researchers were unable to find a relationship between flower
color and vase life. Serini and De Leo (1978) supported this finding and reported no
correlation between vase life and stem length, inflorescence diameter, and number of
ligulae. They showed, however, vase life may be increased by a higher proportion of dry
substance to water in the stem or a greater number of small vessels in the stem. Vase life
and flower yield were reported to be negatively correlated. Tesi (1978) also reported vase
life and flower yield were negatively correlated.

17
De Jong (1978a) presented methods for rapidly identifying structural strength or turgor
strength of gerbera stems. These components of stem strength were postulated to affect
vase life and stem fold, a premature senescence phenomenon.
Structural strength was measured using a protractor to record the curvature of stems
after freshly harvested flowers were stored dry for 24-48 hours. Turgor strength was
measured by comparing the curvature of the stem after dry storage to the curvature of the
stem following recovery (after the stored stems were placed in water for 24 hours and
allowed to regain turgidity). This difference was related to water uptake ability of the
stems. Rigidity of turgid stems was measured using a specially designed instrument which
recorded the force needed to deflect the stem a predetermined distance.
Using these methods, De Jong observed a high frequency of folding was found in
flowers with weak or firm stems, although the lowest frequency of folding occurred in
flowers with firm stems.
De Jong (1978b) suggested breeding for structurally strong stems as a means to
improve postharvest longevity in gerbera. He reasoned that stronger stems may not be as
likely to fold, thereby overcoming the deleterious effects of microbial infection. Also,
structurally stronger stems could provide some added support if a water deficit occurs. He
concluded, however, turgor strength is of primary importance.
Meeteren (1978b) observed increased ion leakage in petals of cut flower gerberas
depended on the cultivar. Meeteren and Gelden (1980) found no correlation between petal-
cytokinin activities in gerbera cultivan. Meeteren (1981) suggested pressure potential of
petals from recently harvested flowers might be a good indicator for vase life and a possible
selection criterion in breeding programs to improve postharvest longevity in gerberas.
Nowak and Plich (1981) observed vase life of cut gerberas increases when stems were
shorter.
De Jong and Garretsen (1985) noted turgor strength and structural strength of gerbera
stems were greater and stronger during the summer than winter months in Holland. No

18
relation between stem stiffness and lignin content was observed among 25 cultivars. They
also suggested the use of tetraploids may increase stem strength because tetraploids usually
have thicker and shorter stems. These researchers observed no differences in the degree of
curvature after dry storage between tetraploids and diploids.
Dubuc-Lebreux and Vieth (1985) suggested a scheme selecting gerbera cultivars
whose stems are more fully differentiated at the region of stem break when flowers are at
the harvesting stage of maturity. They also identified a cultivar in their breeding program
with this characteristic (K-9-9). Amariutei et al. (1986) reported consistent differences in
vase life between two gerberas; Symphonic lasted two days longer than Richard when
treated with pulsing agents or when untreated.
Genetic analysis of vase life in the cut flower Gerbera jamesonii has been described.
Serini and De Leo (1978) estimated narrow sense heritability of vase life within full-sib
families (h2 = .67). Heritability for vase life between plants was lower (h2 = .17). They
concluded selection should be based more on families than individual plants. Tesi (1978)
concluded vase life is strongly influenced by environmental factors. Variation due to
environment and genotype was calculated (Ve2 = 85.2; Vg2 = 14.8). Mean vase life was
reported to be 12 3.0 days.
Harding et al. (1981) estimated heritability of vase life from a non-random sample
population of gerbera genotypes which consisted of half-sib families and clonal parents.
Since the half-sib families had previously been mated and selected for cut flower yield and
preference, it was suggested estimates for narrow sense heritability were biased due to a
reduction in genetic variability. Components of variance were used to estimate heritability.
Narrow sense heritability (h2) was calculated for two successive generations of the half-sib
population; 24 and 38 percent. Broad sense heritability (H2) was calculated for two
successive generations of clonal parents; 36 and 46 percent. Since heritability was
moderately low, it was concluded either intense selection or selection over a large number

19
of generations would be required to increase mean vase life. Mean vase longevity was
reported to be between 10 and 14 days.
De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in
gerberas using a diallel mating scheme involving 12 parents and their progenies. Three
characteristics were examined: vase life (days to wilting or folding); percent folding; and
stem curvature. General combining ability was significant for each characteristic. It was
pointed out that a large error variance (Ve2) will probably result when calculating variance
components for vase life unless the mode of senescence is distinguished, i.e. early or late
folding vs. wilting. Reviewing the inter-relationships among the three characteristics
examined, several conclusions were made: (1) folding results in shorter vase life than
wilting; (2) higher curvature may increase the incidence of folding, thereby resulting in
shorter vase life; (3) no relationship exists between curvature and percent late folding; and
(4) late folding is more difficult to select against than early folding.
De Jong (1986) suggested parent choice is a major factor to consider when designing a
breeding program to improve postharvest longevity in cut flowers of gerbera. For the
characteristic days to wilt, 78% of the variation between progeny means could be
attributed to the twelve parents selected. It was concluded the main difficulty which
remains in breeding to improve postharvest longevity is large intraplant variation for both
days to wilt and percent folding.

CHAPTER 3
PARTI. VASE LIFE STUDIES
1-ngQd.UCLiQa
Gerbera X hvbrida Hort. is a popular cut flower, but its postharvest performance is
often less than desirable. Ideally, postharvest longevity in gerbera should be two weeks or
longer. Unfortunately for the consumer, vase life is usually much less. Postharvest
treatments, i.e. floral preservatives, are used to enhance the lasting ability of gerbera, but,
developing cultivars with genetically superior postharvest longevity may provide the
consumer with a reliable expectation for postharvest quality. Therefore, research to
evaluate the potential of plant breeding as a method to improve postharvest longevity in
gerbera is important.
At present, several researchers have estimated heritability for vase life, which is
defined as the length of time until the flower senesces. Serini and De Leo (1978)
concluded selection should be based more on families than individual plants since their
estimate of narrow sense heritability was higher for among full-sib families (h2 = .67) than
among plants (h2 = .17). Tesi (1978) concluded vase life is strongly influenced by
environmental factors. He observed only 15 % of the phenotypic variation in vase life was
due to genotype. Harding et al. (1981) based their results on a nonrandom sample
population of gerbera genotypes from their Davis population which consisted of half-sib
families and clonal parents. They concluded that since narrow sense heritability (h2 = .24
and .38) and broad sense heritability (H2 = .36 and .46) were moderately low for two
successive generations, either intense selection or selection over a large number of
generations would be required to increase mean vase life.
20

21
De Jong and Garretsen (1985) analyzed combining ability for postharvest longevity in
gerberas using a diallel mating scheme involving 12 parents and their progenies. They
distinguished vase life by different modes of senescence; petal wilt and stem fold. General
combining ability was significant for days to wilting or folding and percent folding. De
Jong (1986) concluded the main difficulty which remains in breeding to improve
postharvest longevity is large intraplant variation for both days to wilt and percent folding.
Plant breeders can benefit from knowing the relationship between characteristics which
they are trying to improve. Although breeding to improve postharvest longevity in gerbera
has been suggested as a viable possibility for many years (Smith and Nelson, 1967), few
studies of gerbera have reported correlations between vase life and other traits.
Tesi (1978) showed a significant negative correlation between vase life and cut flower
yield. Serini and De Leo (1978) found no correlation between vase life and stem length,
inflorescence diameter, and number of ligulae. Nowak and Plich (1981) observed vase life
of cut gerberas increased when stems were shorter.
The objectives of this research on gerbera, using a broad based source of germplasm,
were to determine broad sense heritability and narrow sense heritability for vase life by
diallel analysis, observe changes in frequency of senescence patterns due to selection and
mating, and to determine correlations between vase life and other traits of flower and stem
morphology and growth.
Materials and Methods
Plant Material
Germplasm was randomly collected from several sources. Tissue cultured plantlets
obtained were European cultivars or selections from a commercial breeding program at
Sunshine Carnations in Hobe Sound, Florida. A list of these cultivars is given in table 3-1.

22
Table 3-1.
List of tissue cultured cultivars.
Field #
Cultivar Description
Field #
Cultivar Description
84-1
Amethyst
84-10
SI-1
84-2
Peach
84-11
P15-14-0
84-3
Seashell
84-12
SC300
84-4
Appleblossom
84-13
SC400-8
84-5
Raspberry
84-14
35C404-0X
84-6
Aztec
84-15
PI 8-5
84-7
Mandarin
84-16
SB-24
84-8
(Polish line)
84-17
SC205-X
84-9
Tropic Lady
84-18
SC501
Seed populations were obtained from different seed companies. In addition, Dr. J.
Harding1 provided another seed source from his breeding program. Dr. Hardings seed
mixture was the result of eight generations of breeding, but selection for vase life had been
discontinued after the fifth generation. A list of the seed populations is given in table 3-2.
Table 3-2. List of seed populations.
Field#
Population Description
Seed Source
83-1
Davis Population
U.C. Davis Res. Prog. (U.S.A.)
83-2*
Ahms F-l Strain
Herbst Seed Co. (U.S.A.)
83-3
MardigrasF-1 Strain
Ball Seed Co. (U.S.A.)
83-4
Duplex Mixture
Ball Seed Co. (U.S.A.)
83-5
Jongenelen Strain
Ball Seed Co. (U.S.A.)
83-6
Florist Strain Mix
Park Seed Co. (U.S.A.)
83-7
Ramona Mixture
Sluis & Groot Corp. (Holland)
83-8
No. 4 F-l Mix
Clause Seed Co. (France)
* Poor germination, no plants survived.
Plant material varied in flower color and morphology. The term flower will be used
to describe a composite inflorescence subtended by a stem. Plants grown from four seed
'Dr. J. Harding, Dept, of Environmental Horticulture, U. C. Davis, California

23
populations (Jongenelen, Florists Strain, Ramona, and No. 4 F-l) resembled those Dutch
cultivare that have inflorescences with broad ligulae and thick fleshy stems. These
populations represented a wide spectrum of flower color. The Duplex Mixture was
comprised largely of pinkish hues with spindly stems and narrow ligulae. The Mardigras
F-l Strain included doubles and crested inflorescence types as well as the single, daisy-
type form. These single, daisy-type inflorescences were mostly of deep red hues with
narrow ligulae. The Davis population mainly exhibited inflorescences with narrow ligulae
also representing a wide spectrum of flower color. Flowers from these latter two
populations had stems of medium thickness.
Selection and Mating
Initially, a base population of 953 plants was grown. Plants which did not produce at
least one flower during a flowering period prior to vase life evaluation (May 5June 12,
1984) were discarded. Concurrently, plants were screened for inflorescence type and stem
length. Plants which did not produce single, daisy-type inflorescences or stem length of 45
cm or greater, when 1-2 rows of disc florets were open, were discarded. The remaining
plants of the base population were then referred to as the parental generation. One to six
flowers per plant were evaluated. After the evaluation period, plants that produced less
than three flowers were also discarded. The residual parental generation included 325
plants. Plant means were determined from data collected on the first three flowers
evaluated per plant. Thirty-one plants (approx. 10% of the residual parental generation)
with highest mean vase life (x ^ 11.3)) and lowest coefficient of variation (C.V. £ 25.0)
were selected. To maintain genetic diversity, selection included plants from each seed
population.
A top-cross mating scheme was utilized as a screening method to determine which of
these parents had the longest vase life. Appleblossom was used as the male donor in the
top cross mating scheme because of its excellent vase life rating and low intraplant

24
variation. Twenty-eight plants per cross were grown to produce the top-cross generation.
Plants which did not produce at least one flower during a flowering period prior to vase life
evaluation (March 12May 15,1985) were discarded. One to three flowers per plant
were evaluated. After the evaluation period, plants that produced less than three flowers
were also discarded. Plant means were determined from data collected on three flowers per
plant Progeny means were determined for each cross from individual plant means.
Finally, five plants (approx. 1.5% of the residual parental generation) with highest progeny
mean vase life were selected. To maintain genetic diversity, selection included plants from
four seed populations.
A 5 x 5 diallel cross mating scheme was utilized to estimate heritability of vase life.
Twenty-eight plants per cross were grown to produce the diallel generation. Plants which
did not produce at least one flower during a flowering period prior to vase life evaluation
(May 23August 6, 1987) were discarded. Plant means were determined from data
collected on one to three flowers per plant.
Vase Life Evaluation
Flowers with 1-2 rows of disc florets open were harvested each evening for six
weeks. Stems were then uniformly cut 30 cm long. Flowers were randomly placed into
sterilized glass bottles with one flower per bottle. Each bottle contained 100 ml. of
deionized water, buffered to pH = 3.0-3.4 with a citrate-phosphate buffer. The depth of
water in each bottle was 4 cm. Every other day, the bottles containing deionized water
were replaced until senescence occurred. Evaluations were conducted in a temperature
controlled room (20-21 C) with 24 hrs./day lighting provided by overhead fluorescent
lamps. Light intensity was ,26-.52 W/cm2 at flower height. Relative humidity was
approximately 70%. These conditions were patterned after experiments conducted by De
Jong (1978a and 1978b) and Harding (1981).

25
Flower senescence was classified into three modes based on the visual condition of the
stem:
1. Bending.
2. Folding.
3. Wilting.
The stem gradually, but irreversibly, loses turgidity resulting in an
arc. If allowed to persist, the stem eventually appears folded.
The stem suddenly bends resulting in an irreversible sharp angle.
The stem remains rigid and upright until the ligulae wilt
Vase life was measured by the number of days to flower senescence.
Vase Life = Senescence Date Harvest Date
In addition to vase life, five variables relating to flower morphology and growth were
measured on each flower:
1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm two diameter measurements were taken, i.e. length
and width, on flowers from the top cross and diallel cross generations.)
3. Inflorescence diameter
4. Disc diameter
5. Stem length (at time of senescence)
A variable, Vgrowth, was created to describe the amount of stem elongation
observed from harvest until senescence. Vgrowth was measured by the difference
between stem length at time of senescence and 30 centimeters.
Vgrowth = Senescence Stem Length 30.0 cm
Production
Plants were grown in 12.5 cm standard plastic pots on raised benches in a clear glass
greenhouse at the University of Florida in Gainesville, Florida. Minimum night
temperature was maintained at approximately 18 C. Day temperature was set at 30 C. A
fan and pad cooling system was used to control the temperature. Shade cloth of 25%
density covered the greenhouse since light intensity generally exceeds 65 W/cm2 per day in
Florida. Sowing, transplanting, and flowering dates for each generation are recorded in
table 3-3.

26
Table 3-3.
Record of production dates.
Generation
Sowing
Transplanting
Flowering
Parental
12-2-83
2-18-84
3-15-84/5-5-84
Top Cross
10-12-84
1-12-85
2-23-85/3-12-85
Diallel
1-18-87
3-14-87
4-22-87/5-23-87
Quantitative Analysis
Vase life data were initially analyzed according to the random model for Griffing
Method 3 (Griffing, 1956). This method describes a diallel mating design which includes
reciprocal crosses but excludes seifs. No reciprocal differences were observed.
Subsequently, analysis of variance for combining ability, using a general least squares
diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of
plant means.
Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was
estimated from ratios of the following variances:
VA = Additive genetic variance
Vq = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental)
(Falconer, 1960)
Genotypic and phenotypic variances were determined from the following equations using
thevariance components for general combining ability (a2gca), specific combining ability
(o2sca), and error (a2e) which were calculated by the diallel analysis program developed by
Schaffer and Usanis (1969):

27
Va 4o2gca
Vq = 4 o2ar + 4o2
gca
sea
Vp = 4 G2ar* + 4a2sca + O2,
gca
(Hallauer, 1981)
Thus, heritability (h2 and H2) was estimated from the formulae:
h2 =
4a2
gca
4a2gca+ 4a2 sea +
H2 =
4o2gca + 4a2
sea
4o2gca + 4a2sca + o2e
Predicted estimates of narrow sense heritability and broad sense heritability for n
measurements per plant were also made. This required partitioning the environmental
variance (VE), a component of phenotypic variance (Vp), into general environmental
variance (VEg) and special environmental variance (V^) in order to determine the
phenotypic variance for each case (Vp(n)).
Vp = VG + VE
VE = VEg + VEs
VP(n) = VG + VEg + VEs
(Falconer, 1960)
Special environmental variance (VEs) or within-individual variance (Falconer, 1960)
for a single measurement per plant may be derived by the error variance component (o2e)
from a one-way analysis of variance (Falconer, 1960). Hendersons Method 3
(Henderson, 1953) for obtaining variance components was performed on vase life data
using individual flowers as observations rather than plant means. The SAS procedure
VARCOMP (SAS, 1982) was used to obtain the o2e or error MS. General environmental
variance (VEg) or between-individual variance (Falconer, 1960) was calculated by
subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated

28
in the diallel cross, from the error MS or o2e obtained by the combining ability analysis of
variance.
Repeatability for vase life was determined from the following ratio of variances
whereby n = 1:
, VG + vEg
Vp(n)
(Falconer, 1960)
Results
Selection and Marine
Vase life means for 31 parents selected for the top-cross mating with Appleblossom
ranged from 11.3 to 16.0 days. Coefficients of variation ranged from 0.0 to 24.7.
Progeny means resulting from crosses ranged from 6.1 to 10.7 days. Five of these parents
considered to have the best vase life were selected for diallel mating based on data given
in table 3-4. Progeny means resulting from diallel mating ranged from 8.7 to 14.3 days.
Listing of this data is given in table 3-5. Following selection and breeding, the population
mean for vase life increased by three days. Comparison of vase life data for the parental
and diallel generations is shown in table 3-6.
Table 3-4. Vase life data for five parents.
Parent
(Parental Generation)
Mean
(Top-cross Generation)
Progeny Mean
83-1-77
15.3
8.6
83-4-69
12.0
9.6
83-5-109
13.0
8.9
83-7-4
12.3
10.2
83-7-10
16.0
10.7

29
^Tabl^^S^x^S^^kL^VaseUfe^tafromdidlel^ros^Jrcdg^ds^TOl^)
Cross
(Diallel Generation)
Progeny Mean
83-1-77
X
83-4-69
10.3
83-1-77
X
83-5-109
8.7
83-1-77
X
83-7-4
11.8
83-1-77
X
83-7-10
11.2
83 -i
1-69
X
83-5-109
9.1
83
1-69
X
83-7-4
11.9
83 -i
1-69
X
83-7-10
10.6
83-!
5-109
X
83-7-4
11.4
83-:
5-109
X
83-7-10
12.5
83-'
1 4
X
83-7-10
14.3
Table 3-6 Summary of vase life for parental and diallel generation^
Generation
No. of
Plants
Mean
Std. Dev.
Coefficient of
Variation
Parental
325
7.82
2.64
33.77
Diallel
248
11.20
3.14
28.08
Heritabilitv
Combining ability analysis of variance using plant means was performed on vase life
data from a 5 x 5 diallel. General combining ability effects were significant. Specific
combining ability effects were non-significant. Results are summarized in table 3-7.
Heritability was estimated by the ratio of genetic variance (VA or Vq) to phenotypic
variance (Vp) (Falconer, 1960). Variances were derived using variance components for
general combining ability (a2gca), specific combining ability (a2^), and error (a2e)
according to formulae published by Hallauer (1981). Narrow sense heritability (h2 = .279)
and broad sense heritability (H2 = .281) estimates were approximately equal. This
indicates non-additive genetic variance (Vq VA) is negligble. Variances and heritability
estimates for this population are given in table 3-8.

30
Tabl^^^^^DiaJleL^^^mini^abili^and^sisofv^anceforva^Jifc^
Source of Variation
df
M.S.
F-ratio
General combining ability
4
106.12
8.03 *
Specific combining ability
5
13.21
1.01 ns-
Error
238
13.08
* Significant at P< 0.05.
Table 3-8. 5x5 Diallel. Variances and heritability estimates for vase life.
vA vG
vP
h2
H2
5.08 5.12
18.20
.28
.28
Estimates of heritability were based on an average of 1.96 measurements per plant.
This value (n) resulted by evaluating 487 flowers from 248 plants. The error variance
component (o2^ for a single measurement per plant was obtained by an analysis of
variance using vase life data from individual flowers of the diallel generation. Also, this
analysis indicated differences in vase life of flowers among crosses and among plants
within crosses were highly significant. Results are summarized in table 3-9.
Table 3-9. 5x5 Diallel. Analysis of variance for vase life.
Source of Variation
df
M.S.
F-ratio
Among crosses
9
115.02
5.17 *
Among plants
238
22.26
2.25 *
Within plants
239
9.88
* Significant at P< 0.01.
General environmental variance (VEg) and special environmental variance (VEs) were
derived from vase life data using calculations described by Falconer (1960). A summary of
variances for this population are given in table 3-10. Predicted estimates of narrow sense

31
heritability and broad sense heritability for vase life were then made for 1, 2, 3,5, and
measurements per plant by the ratio of genetic variance (VA or Vq) to phenotypic variance
(Vp(n>) (Falconer, I960). Estimates ranged from 22 to 39 percent. These results are given
in table 3-11. Repeatability (r = .57) for vase life was moderately high.
Genotypic Variance
Environmental Variance
Phenotypic Variance
vA
vG
vE
VHg
Ves
VP(n)
5.08
5.12
13.08
8.04
9.88
18.20
Table^U^SxJ^dle^^^ctedestima^of^ritbilit^tovasehfc^
Number of Measurements
Heritability
1
2
3
5
oo
h2
.22
.28
.31
34
.39
H2
.22
.28
.31
34
.39
Senescence Patterns
A shift in the population mean for vase life was examined more closely. Distribution
of vase life data on individual flowers from the parental and diallel generations was
compared. In addition to a shift in the mean from 7.8 days to 11.2 days, the frequency of
vase life did not appear normally distributed in both generations. Before selection there
was a much higher frequency of flowers with low vase life than high vase life. After
selection and mating, this trend was clearly reversed. These results are shown in figure 3-
1.
Distribution of vase life data by senescence mode was also compared. In both
generations, vase life frequency appeared somewhat normally distributed. Following

32
selection and mating, vase life means for days to bending, folding, and wilting increased.
The largest increase in mean vase life occurred in flowers that folded; 3.5 days. Smaller
increases were observed in flowers that bent and wilted; .4 and 1.2 days, respectively.
These results are shown in figures 3-2, 3-3, and 3-4.
Vase Life (Days)
Figure 3-1. Distribution of vase life data on flowers from the parental and diallel
generations.

Frequency (%)
33
10.0-
7.5-
DIALLEL GENERATION
mean = 6.4
Figure 3-2. Distribution of vase life data for flowers that bent from parental and diallel
generations.

Frequency (%)
34
12.5-1
10.0-
7.5-
PARENTAL GENERATION
mean = 8.2
Vase Life (Days)
Figure 3-3. Distribution of vase life data for flowers that folded from parental and diallel
generations.

Frequency (%)
35
12.5-1
10.0-
7.5-
PARENTAL GENERATION
mean = 12.2
Vase Life (Days)
Figure 3-4. Distribution of vase life data for flowers that wilted from parental and diallel
generations.

36
Separate analyses of variance were made using vase life data from the diallel
generation for days to bending, days to folding, and days to wilting. Variation within
plants or error M.S. for vase life was less when data were grouped by senscence modes as
when an analysis of variance was made on vase life of all flowers. Differences in vase life
of flowers among crosses were highly significant for all senescence modes. Differences
among plants within crosses were also significant for days to folding and wilting, but non
significant for days to bending. Results are summarized in tables 3-12, 3-13, and 3-14.
Table 3-12. Analysis of variance for vase life^daysjojrending^
Source of Variation
df
M.S.
F-ratio
Among crosses
9
21.72
2.65 *
Among plants
70
8.24
1.36 n s-
Within plants
29
6.05
* Significant at P< 0.01.
Table 3-13. Analysis of variance for vase life (days
to folding).
Source of Variation
df
M.S.
F-ratio
Among crosses
9
52.28
3.76
Among plants
124
14.42
3.16
Within plants
46
4.56
* Significant at P< 0.01.
Table 3-14. Analysis of variance for vase life (days
to wilting).
Source of Variation
df
M.S.
F-ratio
Among crosses
9
28.18
3.21 *
Among plants
120
8.77
1.49 *
Within plants
68
5.88
Significant at P< 0.05. Significant at P< 0.01.

37
Changes in the proportion of flowers that senesced due to bending, folding, and
wilting were also observed. Differences between senscence mode frequency in the parental
and diallel generations are shown in figure 3-5. After selection and mating, the incidence
of bending was reduced by an average of 41 percent. Folding and wilting increased by an
average of 23 and 18 percent, respectively. In both generations, the frequency of wilting
exceeded that of folding.
120
100
Frequency
(%)
80
60
40
20
0
% bending
% folding
% wilting
parental diallel
Generation
Figure 3-5. Distribution of senescence modes for parental and diallel generations.
The frequency of bending, folding, and wilting in progeny of the diallel generation
was compared to vase life means. Among ten progenies, the proportion of bending
generally decreased as vase life increased. With the exception of one cross, approximately
50 percent of the flowers wilted in progenies with vase life of 11 days or greater.
Approximately 25 percent of the flowers wilted in progenies with vase life of less than 11
days. The proportion of folding did not appear related to progeny means for vase life.
Distribution of this data is shown in figure 3-6.

38
H

Frequency
7%)
8.7 9.1 10.3 10.6 11.2 11.4 11.8 11.9 12.5 14.3
vase life progeny means
120
100
80
60
40
20
0
% bending
% folding
% wilting
Figure 3-6. Distribution of senesence mode frequency in progeny of diallel generation.
Correlations between Vase Life and Other Traits
Phenotypic correlations between vase life and five stem and inflorescence traits: stem
length, stem diameter, inflorescence diameter, disc diameter, and post-harvest stem
elongation (vgrowth), were determined for both parental and diallel generations.
Correlation coefficients based on data from individual flowers were obtained using a SAS
program for Pearsons product-moment correlation procedure (SAS, 1986).
Correlations determined from the parental generation demonstrated the linear
relationship between vase life and these traits in a random population before selection for
improving vase life. Correlations determined from the diallel generation demonstrated the
linear relationship between vase life and these traits after selection. Additionally,
correlations were made between the vase life of flowers that bent, folded, or wilted and the
five stem and inflorescence traits. These results are summarized in tables 3-15, 3-16, 3-17,
3-18, and 3-19.

39
Table 3-15. Phenotypic correlation coefficients between vase life and stem length before
and after selection for vase life.
-- i
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.14 *
-.15**
-.24 *
1
b
U\
Diallel
487
-.14**
-.01
-.04
-.27 *
* Significant at P< 0.01.
Table 3-16. Phenotypic correlation coefficients between vase life and stem diameter
before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
.02
-.11**
-.08
.08
Diallel
487
-.11*
.08
-.15
-.24 *
* Significant at P< 0.05. *
* Significant at P< 0.01.
Table 3-17. Phenotypic correlation coefficients between vase Ufe and inflorescence
diameter before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.11**
-.10**
-.33 *
-.29**
Diallel
487
-.05
.14
.00
-.23 *
* Significant at P< 0.01.
Table 3-18. Phenotypic correlation coefficients between vase life and disc diameter before
and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
-.05
-.12**
-.26 *
.02
Diallel
487
.03
.02
.09
-.07
Significant at P< 0.01.

40
Table 3-19. Phenotypic correlation coefficients between vase life and post-harvest stem
elongation before and after selection for vase life.
Generation
N
Total
Bending
(Senescence Modes)
Folding
Wilting
Parental
1508
.22**
.14**
.20**
.37**
Diallel
487
.24**
.23 *
.17*
.13
* Significant at P< 0.05. Significant at P< 0.01.
Vase life and stem length. Significant negative correlations were observed, based on
data from the total number of flowers evaluated, before and after selection. Among flowers
that bent, folded, or wilted in the parental generation, the negative correlations observed
between vase life and stem length were significant only for bending and folding. In
contrast, among flowers in the diallel generation, only the negative correlation observed
between vase life and stem length for flowers that wilted was significant. Overall, it
appeared from these correlations that vase life was highest when stem length was shorter.
Vase life and stem diameter. Based on data from the total number of flowers
evaluated, no correlation was observed before selection, however, a significant correlation
was observed after selection. Among flowers that bent, folded, or wilted in the parental
generation, negative correlations were observed between vase life and stem diameter for
flowers that bent or folded; but, the correlation was significant only for flowers that bent.
The positive correlation observed for flowers that wilted was non-significant. In contrast,
among flowers in the diallel generation, negative correlations were observed between vase
life and stem diameter for flowers that folded or wilted, but the correlation was significant
only for flowers that wilted. The positive correlation observed for flowers that bent was
non-significant. Overall, it appeared from these correlations that vase life was highest
when stem diameter was smaller.
Vase life and inflorescence diameter. Negative correlations were observed, based on
data from the total number of flowers evaluated, before and after selection; however, only
the correlation before selection was significant Among flowers that bent, folded, or wilted

41
in the parental generation, significant negative correlations were observed between vase life
and inflorescence diameter. Among flowers that bent, folded, or wilted in the diallel
generation, a significant negative correlation was observed between vase life and
inflorescence diameter only for flowers that wilted. The positive correlation for flowers
that bent was non-significant and no correlation was observed for flowers that folded. In
both cases, before and after selection, the negative correlation observed between vase life
and inflorescence diameter was highly significant for flowers that wilted. Thus, it appeared
that vase life increased as inflorescence diameter decreased.
Vase life and disc diameter. No significant correlations were observed, based on data
from the total number of flowers evaluated, before or after selection. The only significant
correlations observed between vase life and disc diameter were among flowers from the
parental generation that bent or folded. These highly significant correlations were negative.
A negative correlation was also observed for flowers that wilted in the diallel generation,
though this correlation was non-significant. In general, these correlations demonstrated a
very weak negative relationship, if any, between vase life and disc diameter. The
significant correlations that were observed in the parental generation for vase life of flowers
that bent and folded seemed of little interest since selection would be made against these
senescence modes in a breeding program to improve vase life.
Vase life and vgrowth. Significant positive correlations were observed, based on data
from the toal number of flowers evaluated, before and after selection. Among flowers that
bent, folded, or wilted in the parental generation, highly significant positive correlations
were observed between vase life and vgrowth. Among flowers that bent, folded, or wilted
in the diallel generation, positive correlations were also observed between vase life and
vgrowth, however, correlations were significant for flowers that bent or folded. It
appeared from these correlations that the positive relationship which was observed
between vase life and post-harvest stem elongation prior to selection and mating weakened
as vase life increased.

42
Discussion
Heritability estimates for a given character can vary based on the population of plants
evaluated, selection intensity, mating design and environment. (Simmonds, 1979). Despite
this, estimates of heritability for vase life determined from this experiment (h2 = .28 and H2
= .28) were within proximity of those determined from investigations by other researchers
(Serini and De Leo, 1978; Tesi, 1978; and Harding et al., 1981). This reflects some
consistency in the proportion of genetic variance to phenotypic variance for vase life in
gerbera regardless of genetic diversity in populations sampled, breeding procedures, and
environment.
The estimate of 28 percent for broad sense heritability (H2), based on 1.96
measurements per plant, is moderately low. Similarly, the estimate of 28 percent for
narrow sense heritability (h2), also based on 1.96 measurements per plant, is moderately
low. This estimate is between the range of prior estimates: 15 percent (Tesi, 1978); 17
percent (Serini and De Leo, 1978); and 0, 24, and 38 percent (Harding et al., 1981). It
appears that genetic variation may be largely controlled by additive gene action since broad
sense and narrow sense heritability were approximately equal. Therefore, in a fixed model
experiment, progeny means obtained from a top-cross mating would be effective in
determining parents with good combining ability for increasing vase life.
Falconer (1960) demonstrated a method to predict estimates of heritability for a
specified number of measurements per experimental unit. This involved partitioning
environmental variance (V£) into general environmental variance (VEg) and special
environmental variance (V^. Special environmental variance (VEs), or within-plant
variation, is the environmental variation for a single observation per experimental unit. The
magnitude for special environmental variance (VEs) is then divided by a specified number
of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic
variance for each special case. Ideally, if n = then VEs will be reduced to zero, thereby

43
deleting a significant source of environmental variation. In that case, the highest possible
estimate of heritability could be obtained for a given population.
Using this method, predicted estimates of heritability for vase life in gerbera ranged
from 22 to 39 percent with n specified as 1, 2, 3, 5, and flowers per plant. Since these
calculations assumed no change in genotypic effects, broad sense and narrow sense
heritability estimates for each case were approximately equal. It is interesting to note that in
spite of varying the magnitude of environmental variance, the range of estimates remained
within proximity to those determined by other researchers (Serini and De Leo, 1978; Tesi,
1978; and Harding et al., 1981).
The repeatability estimate for vase life (r = .57) is moderately high, indicating that two
to three flowers per plant is adequate for determining the average vase life per plant.
Falconer (1960) recommends that further gain in accuracy by more than two measurements
does not justify additional expense or time required to collect more data when repeatability
is high. This was proven by comparing the relative increase in heritability from predicted
estimates based on 1,2, 3, 5, and < measurements per plant. Between one and three
measurements, heritability increased by nine percent, while beyond three measurements
through infinity, the gain in heritability was only eight percent.
Thus far, this research has confirmed that improvement of postharvest longevity in
gerbera can be obtained by selecting and mating plants with good combining ability for
high vase life. The overall population mean for vase life resulting from a diallel cross
among five plants selected, based on their combining ability with 'Appleblossom,' a
cultivar with high vase life, yielded an increase of more than three days. Additional
information has been gained, however, by classifying vase life determinations by three
distinct modes of senescence: bending, folding, and wilting. The frequency of vase life
days based on data from all flowers was not normally distributed before or after selection
and mating. Yet, when frequency of vase life days was classified by senescence mode,
distribution was normal. Moreover, after selection and mating, increases in mean vase life

44
for flowers that folded, bent, or wilted differed. For example, the increase in mean vase
life for flowers that folded was much greater than for flowers that bent or wilted.
Combining this evidence, it is suggested that vase life may be a composite character of at
least three components, represented by each senescence mode.
Further evidence to support this suggestion is found by comparing the magnitude of
the error variance component (o^g) or error MS from an analysis of variance based on the
total number of vase life observations versus the error MS from individual analyses based
on the number of vase life observations for each senescence mode. De Jong and Garretsen
(1985) previously reported that if termination of vase life is not distinguished by stem
collapse or petal wilt in an analysis of variance, a relatively large error variance would
result. In fact, the error MS based on the total number of vase life observations from this
data was nearly double the arithmetic mean of the error MS from individual analyses based
on flowers that bent, folded, and wilted (9.88 vs. 5.50).
This study was designed to evaluate the potential of improving postharvest longevity
in gerbera by selecting plants with high vase life. Plants were not selected based on the
specific number of days to bending, folding, or wilting of their flowers. However; it does
appear, that selecting plants based on the number of days to bending, folding, or wilting,
rather than the composite character of vase life, may prove to be a useful approach to
improving postharvest longevity in gerbera, since general combining ability for vase life of
flowers was determined to be highly significant for all three senescence modes.
The incidence of bending, folding, or wilting, not only the vase life of these
senescence modes, is proposed to be another important aspect of postharvest longevity in
gerbera. Before selection and mating to improve vase life, a greater proportion of flowers
bent than wilted. Flowers that bent generally exhibited lower vase life than flowers that
wilted. After selection and mating, a greater proportion of flowers wilted than bent.
Additionally, the shift in proportion of these two modes that occurred was rather dramatic

45
after only one generation of selection. Therefore, it is postulated that the incidence of
bending ys wilting may be a qualitatively based trait controlled by relatively few genes.
In studies on breeding for keeping quality in gerbera, De Jong (1986) distinguished
senescence by three classes; early stem fold, late stem fold, and petal wilt His definition
for early stem fold was slightly different from that of bending in this study, but he reported
that if selection were made against the phenomenon, its incidence could be reduced or
eliminated fairly easily. He showed data for 59 progenies resulting from a diallel mating;
four progenies did not exhibit early fold. After studying these results, it seems possible
that the trait early fold vs no early fold could fit a genetic model involving only two
genes with epistatic effects (15:1). Hence, it seems plausible to hypothesize that a similar
model may apply to the trait bending vs wilting. Among ten progenies, bending was not
completely eliminated, but, in one case, its incidence was reduced to only five percent.
The incidence of stem folding may be distinct from the incidence of bending or
wilting. Before selection and mating to improve vase life, the frequency of folding was
relatively low. After selection and mating, the frequency of folding nearly tripled.
Accompanying this increase, vase life also increased. This created some speculation as to
whether a higher proportion of folding is the result of longer lasting flowers. Yet when the
proportion of folding for each progeny was compared to progeny means for vase life, this
speculation could not be confirmed because some progeny with higher vase life showed a
smaller proportion of folding than some progeny with lower vase life. Instead, the most
striking observation of this comparison, among ten progenies, was that the proportion of
folding could be grouped into four classes of approximately 20, 30,40, and 65 percent.
This gives cause to wonder whether data from more progenies might yield a sufficient
number of classes to identify a genetic model that would indicate folding is also
qualitatively inherited. If so, it seems likely that the incidence of folding is controlled by at
least several genes.

46
Correlations between vase life and other morphological traits were particularly
interesting because a broad based germplasm was examined. The wide variation in
morphological phenotypes permitted an extensive comparison. Correlation coefficients
were generally low. Despite this, they were often significant due to the large number of
flowers evaluated.
Since it has been discussed that eliminating senescence due to bending and folding
among flowers may be possible, correlations between the vase life of flowers that wilted
and other traits are probably the most useful to a breeder. Significant relationships between
longer vase life of flowers that wilted and shorter stem length, smaller stem diameter, and
smaller inflorescence diameter were determined.
Unfortunately, the current floriculture market evidently prefers gerberas with long,
thick stems and large inflorescences, as most commercial cultivars tend to be of this type.
Therefore, it is encouraged that postharvest longevity be deliberately included as an
objective into cut flower gerbera breeding programs or the consumer will have to be
satisfied with only the expected vase life possible from postharvest treatments such as floral
preservatives. If attention to this situation is not given, a decline in the popularity of
gerbera might result as other flower varieties with better, more predictable vase life become
increasingly available.
Conclusion
Breeding to improve vase life in gerbera has potential, despite the fact that heritability
was confirmed to be moderately low. It is concluded that the best approach for establishing
lines with superior lasting quality requires recognizing vase life as a composite character.
Distinguishing quantitatively inherited traits, i.e. days to bending, days to folding, and
days to wilting and qualitatively inherited traits, i.e. bending vs wilting and folding ys
non-folding, may be the key to a successful breeding program.

47
Evidence suggests that obtaining plants homozygous for wilting may be possible.
Recurrent selection then would be useful to increase the number of days to wilting. Further
studies to determine the heritability of this vase life component could indicate the intensity
of selection necessary to increase mean vase life for a gerbera population. Consistent
selection against plants whose flowers exhibit bending or folding is imperative.
Finally, after several lines with superior vase life are established, it is recommended
that breeders incorporate this trait into other lines with desireable plant morphology using a
backcross mating scheme, especially if wilting can be confirmed to be a qualitative trait.

CHAPTER 4
PART II. STEM STRENGTH STUDIES
Introduction
Gerbera X hvbrida Hort. is a popular cut flower that is recognized for its long,
leafless, and upright stem, known as a scape. Postharvest longevity of gerbera often
abruptly ends when the stem ceases to remain upright.. This phenomenon has been termed
knicking (Wilberg, 1973; Buys, 1978), folding (De Jong, 1978a), neck droop
(Zieslin et al., 1978), or stem break (Meeteren, 1978a). De Jong (1978b) suggested
breeding for structurally strong stems may be a means to improve postharvest longevity in
gerbera.
Descriptions of the relationship between stem strength and morphological and growth
traits in gerbera are few. Wilberg (1973) identified that higher ratios of dry weight/cm for
stem sections prone to bending contribute to stem strength in gerbera. Barigozzi and
Quagliotti (1978) noted tetraploids appeared to have stronger stems than diploids. De Jong
and Garretsen (1985) supported this observation by suggesting thicker and shorter stems of
tetraploids may result in increased stem strength.
The objectives of this research on gerbera were to determine broad sense heritability
and narrow sense heritiability estimates for stem strength and to determine correlations
between stem strength and other traits of flower and stem morphology and growth.
Materials and Methods
Selection and Mating
Initially, a population of 278 plants was randomly selected for stem strength evaluation
(May 25June 14, 1984). They were selected from the residual parental generation of
48

49
325 plants that had been previously established for vase life studies (Chapter 3). Plants
selected had already produced at least three flowers that had been evaluated for vase life.
One to six flowers per plant were evaluated. After the evaluation period plants that
produced less than three flowers were discarded. This residual parental generation
included 73 plants. Plant means were determined from data collected on the first three
flowers evaluated per plant. Seven plants (approx. 10% of the total number of plants
evaluated for stem strength) with highest mean stem strength (x £ 14.00) and lowest
coefficient of variation (C.V. ^ 27.00), were selected. To maintain genetic diversity,
selection included plants from five seed populations.
A 7 x 7 diallel cross mating scheme was utilized to estimate heritability of stem
strength. Thirty plants per cross were grown to produce the diallel generation. Plants
which did not produce at least one flower during a flowering period prior to stem strength
evaluation (June 12July 30, 1985) were discarded. Plant means were determined from
data collected on one to three flowers per plant.
Stem Strength Evaluation
Flowers with 1-2 rows of open disc florets were harvested each evening for six
weeks. After 24 hours dry storage at room temperature (20-21 C), a 15 cm stem
segment from each flower was evaluated for stem strength. The portion of the stem 4.5 -
19.5 cm below the base of the peduncle was the area from which the segment was taken.
This is shown in figure 4-1.
Stem strength was determined by the amount of force required to deflect the midpoint
of each segment a specified distance (F=g/cm). The segments were supported at two
points, 10 cm apart; equidistant to their midpoint. Measurements were made using an
Instron.1 Instron specifications are listed in table 4-1. Initially, a maximum deformation of
1.0 cm was specified with measurements taken at five equal intervals: .2 cm, .4 cm, .6
'Provided by the Dept of Agricultural Engineering, Univ. of Florida, Gainesville, Florida, 33610.

50
cm, .8 cm, and 1.0 cm Later, flowers from the diallel cross generation were measured
only for deformation at .2 cm.
Figure 4-1. Portion of stem used for stem strength evaluation.
Table 4-1. Instron specifications
Instron:
Model TM
Load Cell Compression
100 grams
Crosshead Speed
.2 inches/minute
Chart Speed
2 inches/minute
Deformation
Chart Distance
.2 cm
.79 inches
.4 cm
1.57 inches
.6 cm
2.36 inches
.8 cm
3.15 inches
1.0 cm
3.93 inches

51
In addition to stem strength evaluation, five variables relating to flower morphology
and growth were measured for each flower from plants belonging to the parental
generation:
1. Stem length (at time of harvest)
2. Stem diameter (at 30 cm)
3. Inflorescence diameter.
4. Disc diameter
5. Inflorescence weight (after storage)
Three variables were measured on each flower from plants belonging to the diallel
cross generation:
1. Stem length (at time of harvest)
2. Stem diameter (at 30.0 cm)
3. Inflorescence weight (before storage)
Production
Plants were grown according to the production regime described for vase life studies
(Chapter 3). Sowing, transplanting, and flowering dates for both generations are recorded
in table 4-2.
Table 4-2. Record of production dates
Generation
Sowing
Transplanting
Flowering
Parental
12-2-83
2-18-84
3-15-84/5-25-84
Diallel
1-30-85
4-7-85
5-15-85/6-12-85
Quantitative Analysis
Stem strength data were initially analyzed according to the random model for Griffing
Method 3 (Griffing, 1956). This method describes a diallel mating design which includes
reciprocal crosses but excludes seifs. No reciprocal differences were observed.
Subsequently, analysis of variance for combining ability, using a general least squares

52
diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of
plant means.
Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was
estimated from ratios of the following variances:
VA = Additive genetic variance
VG = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental)
(Falconer, 1960)
Genotypic and phenotypic variances were determined from the following equations using
the variance components for general combining ability (a2gca), specific combining ability
(o2sca). and error (o2e) which were calculated by the diallel analysis program developed by
Schaffer and Usanis (1969):
Va = 42gca
VG = 4a2gca + 4g2
sea
VP = 42gca + 4o2sca + CJ2e
(Hallauer, 1981)
Thus, heritability (h2 and H2) was estimated from the formulae:
4o2
h2 = T^Lgca
4a2gca + 4a2
sea +
4a2gca + 4o2sca
4a2gca + 4o2
sea + 2e
Predicted estimates of narrow sense heritability and broad sense heritability for n
measurements per plant were also made. This required partitioning the environmental
variance (VE), a component of phenotypic variance (Vp), into general environmental

53
variance (VEg) and special environmental variance (V^) in order to determine the
phenotypic variance for each case (Vp(n)).
Vp = VG + VE
VE = vEg + vEs
VP(n) = VG + VEg + ~ VEs
(Falconer, 1960)
Special environmental variance (VEs) or within-individual variance (Falconer, 1960)
for a single measurement per plant may be derived by the error variance component (c2^)
from a one-way analysis of variance (Falconer, 1960). Hendersons Method 3
(Henderson, 1953) for obtaining variance components was performed on vase life data
using individual flowers as observations rather than plant means. The SAS procedure
VARCOMP (SAS, 1982) was used to obtain the a2e or error MS. General environmental
variance (VEg) or between-individual variance (Falconer, 1960) was calculated by
subtracting the quotient of this error MS/n, whereby n = # of flowers /# of plants evaluated
in the diallel cross, from the error MS or c2e obtained by the combining ability analysis of
variance.
Repeatability for vase life was determined from the following ratio of variances
whereby n = 1:
VG + VEg
VP(n)
(Falconer, 1960)
Results
Selection and Mating
Stem strength was initially measured at five intervals of deformation between zero and one
centimeter. Correlations between stem strength measured at .2 cm and other intervals were

54
approximately equal to 1.0. The relationship among these measurements for each flower
was generally non-linear. Stem strength means for five intervals are shown in figure 4-2.
grams
Figure 4-2. Stem strength means at five deformation intervals.
Measurements made beyond .2 cm were often observed to be a function of the damage
incurred to the stem segment as a result of force being applied. For example, some
segments folded during the measurement process, hence impairing the contact between the
segment and apparatus. Measurements taken at .2 cm were considered, however, to be a
reasonable estimate of stem strength.
Seven parents considered to have the best stem strength were selected for diallel
mating based on data given in table 4-3. Stem strength means for these plants ranged from
14.2 to 31.5 g/.2 cm and coefficients of variation ranged from 5.6 to 26.8. Progeny means
resulting from diallel mating ranged from 10.3 to 33.7 g/.2 cm. Listing of this data is
given in table 4-4. Following selection and breeding, the population mean for stem
strength more than doubled. Comparison of stem strength data for the parental and diallel
generations is shown in table 4-5.

55
Parent
Mean
Coefficient of
Variation
83-1-10
22.0
26.8
83-1-31
31.5
15.6
83-1-96
18.0
5.6
83-4-8
14.2
25.5
83-5-76
14.3
23.2
83-7-48
24.0
24.2
83-8-7
16.0
23.6
Table 4-4. 7x7 Diallel. Stem strength data from diallel crosses. (F = g/.2cm; reciprocals
pooled)
Cross
83-1-10
X
83-1-31
31.3
83-1-10
X
83-1-96
26.6
83-1-10
X
83-4-8
17.1
83-1-10
X
83-5-76
32.3
83-1-10
X
83-7-48
32.4
83-1-10
X
83-8-7
19.7
83-1-31
X
83-1-96
23.2
83-1-31
X
83-4-8
14.2
83-1-31
X
83-5-76
27.5
83-1-31
X
83-7-48
26.7
83-1-31
X
83-8-7
23.9
83-1-96
X
83-4-8
14.2
83-1-96
X
83-5-76
33.7
83-1-96
X
83-7-48
29.7
83-1-96
X
83-8-7
21.2
83-4-8
X
83-5-76
15.9
83-4-8
X
83-7-48
14.4
83-4-8
X
83-8-7
10.3
83-5-76
X
83-7-48
24.2
83-5-76
X
83-7-48
29.9
83-7-48
X
83-8-7
25.1
(Diallel Generation)

56
Table^S^Sum^^ofstemstnghfor^enta^anddialle^ienerations^F^^^m)
Generation
No. of
Plants
Mean
Std. Dev.
Coefficient of
Variation
Parental
73
10.17
7.27
71.48
Diallel
642
22.52
11.32
50.28
Heritabilitv
Combining ability analysis of variance using plant means was performed on stem
strength data from a 7 x 7 diallel. General combining ability effects were highly
significant. Specific combining ability effects were non-significant. Results are
summarized in table 4-6. Heritability was estimated by the ratio of genetic variance (VAor
Vq) to phenotypic variance (Vp) (Falconer, 1960). Variances were derived using variance
components for general combining ability (o2gca), specific combining ability (o2^), and
error (a2e) according to formulae published by Hallauer (1981). The difference between
narrow sense heritability (h2 = .38) and broad sense heritability (H2 = .42) estimates was
small. This indicates the effect attributable to non-additive genetic variance (Vq VA) is
minimal. Variances and heritability estimates for this population are given in table 4-7.
Table 4-6. 7x7 Diallel. Combining ability analysis of variance for stem strength.
(F = g/-2cm)
Source of Variation
df
M.S.
F-ratio
General combining ability
6
4218.62
16.32**
Specific combining ability
14
258.47
1.63 n s-
Error
621
158.22
* Significant at P< 0.01.
Estimates of heritability were based on an average of 2.66 measurements per plant.
This value (n) resulted by evaluating 1710 flowers from 642 plants. The error variance

57
Table 4-7. 7x7 Diallel. Variances and heritability estimates for stem strength.
(F = g/,2 cm)
VA
vG
vP
h2
H2
103. 64
116.76
274.98
OO
m
.42
component (c2^ for a single measurement per plant was obtained by an analysis of
variance using stem strength data from individual flowers of the diallel generation. Also,
this analysis indicated differences in stem strength of flowers among crosses and among
plants within crosses were highly significant. Results are summarized in table 4-8.
Table 4-8. 7x7 Diallel. Analysis of variance for stem strength. (F=g/.2 cm)
Source of Variation
df
M.S.
F-ratio
Among crosses
20
3995.98
10.68 *
Among plants
621
374.08
2.12**
Within plants
1068
146.06
* ^Significant at P< 0.01.
General environmental variance (VEg) and special environmental variance (Vg^ were
derived from stem strength data using calculations described by Falconer (1960). A
summary of variances for this population is given in table 4-9. Predicted estimates of
narrow sense heritability and broad sense heritability for stem strength were then made for
1, 2, 3, 5, and measurements per plant by the ratio of genetic variance (VA or Vq) to
phenotypic variance (VP(n)) (Falconer, 1960). Estimates ranged from 28 to 53 percent.
These results are given in table 4-10. Repeatability (r = .60) for stem strength was
moderately high.

58
Table 4-9. 7x7 Diallel. Summary of variances for stem strength. (F = g/.2 cm) n
= 2.66
Genotypic Variance
VA vG
vE
Environmental Variance
VEg
Ves
Phenotypic Variance
VP(n)
103.64
116.76
158.22
103.31
146.06
274.98
Table 4-10. 7x7 Diallel. Predicted estimates of heritability for stem strength.
(F = g/.2 cm)
Heritability
1
2
Number of Measurements
3 5
oo
h2
.28
.35
.39 .42
.47
H2
.32
.40
.43 .47
.53
Correlations between Stem Strength and Other Traits
Phenotypic correlations between stem strength and three stem and inflorescence traits,
stem length, stem diameter, and inflorescence weight, were determined for both parental
and diallel generations. Correlation coefficients based on data from individual flowers
were obtained using a SAS program for Pearsons product-moment correlation procedure
(SAS, 1986).
Correlations determined from the parental generation demonstrated the linear
relationship between stem strength and these traits in a random population before selection
for improving stem strength. Correlations determined from the diallel generation
demonstrated the linear relationship between stem strength and these traits after selection.
These results are summarized in table 4-11.
Stem strength and stem length. Significant negative correlations were observed
between stem strength and stem length, before and after selection. It appeared, however,

59
Table 4-11. Phenotypic correlation coefficients between stem strength (F = g/.2 cm) and
three traits before and after selection.
Generation
N
Stem Length
Stem Diameter
Inflorescence
Weight
Parental
548
-.28 *
.10*
(after storage)
.16**
Diallel
1710
-.05*
.22**
(before storage)
.38**
* Significant at P< 0.05. Significant at P< 0.01.
that following selection and mating, the relationship weakened.
Stem strength and stem diameter. Significant positive correlations were observed
between stem strength and stem diameter, before and after selection. Unlike the
relationship between stem strength and stem length, it appeared that following selection and
mating, the relationship strengthened.
Stem strength and inflorescence weight. Significant positive correlations were
observed between stem strength and inflorescence weight, before and after selection.
Similar to the relationship between stem strength and stem diameter, in this case, following
selection and mating, the correlation coefficient was higher.
Correlations between stem strength and two other traits, inflorescence diameter and
disc diameter, were also determined prior to selection. The correlation (r = -.08) between
stem strength and inflorescence diameter was not significant. A positive correlation (r =
.11) observed between stem strength and disc diameter was highly significant; as stem
strength increased, disc diameter increased.
Discussion
Strength of gerbera stem segments was determined using an Instron. Results were
recorded in terms of the amount of force required to deflect the segment a specified
distance. This method, similar to flexture tests made on engineering materials (Mohsenin,
1970) was developed to test the segment as a simple beam. The intention of using an
Instron to determine stem strength was to increase the accuracy of determinations.

60
Although this apparatus provided a relatively sophisticated method to quantitatively
measure mechanical strength, a major problem remained due to the variability in gerbera
stems. Gerbera stems are not structurally homogeneous. In addition to differences in stem
diameter, stems can be hollow or solid and round or oval. Measurements for determining
the force required to deflect a stem segment a specified distance can vary depending on
these factors.
Mohsenin (1970) suggested accounting for variation, often found in agricultural
materials, by using a formula to calculate apparent stiffness, known as modulus of
elasticity (E). Parups and Voisey (1976), who studied the resistance to bending of the
pedicel in greenhouse-grown roses, which are also not structurally homogeneous,
explained their calculations for determining modulus of elasticity of the pedicel part of rose
stems. There was concern about the variability in gerbera stems, therefore stem strength
results obtained from raw data were initially compared to modulus of elasticity values (E).
The following formula was used to calculate modulus of elasticity:
FL3
E =
D48^fr>
(Mohsenin, 1970)
where F is force required to deflect the segment a specified distance, L is length of the
segment, D is deflection distance, and d is diameter of the stem segment at mid-span.
Since no differences were found between ranking plants using raw data directly from
Instron measurements and ranking plants by modulus of elasticity, stem strength values
obtained from raw data were used to estimate heritability and to correlate stem strength to
morphological traits.
Another problem with accurately evaluating stem strength in gerbera relates to the
moisture content of the stem segment measured. Segments were stored at room
temperature for 24 hours; however, this did not specify their exact level of moisture content
at the time of measurement. Evidently the proportion of evaporation differed between stem

61
segments. Frequently, but not always, stem segments with small diameters dehydrated
more than segments with larger diameters. These segments became brittle and resulted in
stem strength greater than 50g/.2 cm, much higher than most stem strength values
recorded.
Despite efforts to obtain an accurate evaluation of stem strength for gerbera flowers,
the coefficient of variation for stem strength in our experiment exceeded fifty percent before
and after selection and mating to increase stem strength. This demonstrates that
environmental conditions greatly influence this character.
Broad sense heritability for stem strength (H2 = .42), based on 2.66 measurements per
plant, appears moderate; therefore improvment for this character can be expected with even
a moderate rate of selection intensity. Narrow sense heritability was also fairly moderate
(h2 = .38), indicating parental phenotypes could be expected to correspond to some degree
to their genotypes.
Genetic variation may be largely controlled by additive gene action since the difference
between broad sense and narrow sense heritability was small. Therefore, in a fixed model
experiment, progeny means obtained from a top-cross mating would be effective in
determining parents with good combining ability for increasing vase life.
As discussed in Chapter 3, environment is a critical variance component of heritability.
Different estimates of heritability would be expected if the magnitude of environmental
conditions varied, assuming effects due to genotype remained constant for a given
population.
Falconer (1960) demonstrated a method to predict estimates of heritability for a
specified number of measurements per experimental unit. This involved partitioning
environmental variance (V^) into general environmental variance (VEg) and special
environmental variance (VEs). Special environmental variance (VEs) or within-plant
variation is the environmental variation for a single observation per experimental unit. The
magnitude for special environmental variance (VEs) is then divided by a specified number

62
of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic
variance for each special case. Ideally, if n = then VEs will be reduced to zero, thereby
deleting a significant source of environmental variation. In that case, the highest possible
estimate of heritability could be obtained for a given population.
Using this method, predicted estimates of heritability for vase life in gerbera ranged
from 28 to 53 percent with n specified as 1,2,3, 5, and < flowers per plant. Since these
calculations assumed no change in genotypic effects, the differences between broad sense
and narrow sense heritability estimates for each case were small. This range of heritability
reinforces the initial description that heritability is moderate, and improvement of stem
strength in gerbera can be obtained even with moderate selection.
The repeatability estimate for stem strength (r = .60) is moderately high, indicating that
two to three flowers per plant is adequate for determining the average stem strength per
plant. Falconer (1960) recommends that further gain in accuracy by more than two
measurements does not justify additional expense or time required to collect more data
when repeatability is high. This was proven by comparing the relative increase in
heritability from predicted estimates based on 1,2, 3, 5, and < measurements per plant.
Between one and three measurements, heritability increased by eleven percent, while
beyond three measurements through infinity, the gain in heritability was only eight percent.
Correlations between stem strength and other morphological traits in this population of
gerbera are of interest, given the wide variation in morphological phenotypes studied.
Despite correlation coefficients were often significant due to the large number of flowers
evaluated, they were generally low.
Significant relationships between stronger stem strength of flowers and shorter stem
length, larger stem diameter, and heavier inflorescence weight were determined before and
after selection and mating. Already, Barigozzi and Quagliotti (1978) and De Jong and
Garretsen (1985) observed that tetraploids appear to have stronger stems than diploids.
These findings provide strong evidence why the use of tetraploids, whose flowers

63
generally have shorter, thicker stems and heavier inflorescences, has been suggested as a
possible means for increasing stem strength in gerbera.
Conclusion
Improving structural stem strength in gerbera can be realized through breeding efforts.
Despite this encouraging conclusion, the breeder faces difficulty with selecting plants
whose flower stems are structurally strong. Large intraplant variation associated with this
character is a major problem.
This situation highlights several unanswered questions, What determines structural
strength in flower stems? Is it a composite character? If so, what are the individual
components of structural stem strength? How do they contribute to postharvest longevity
in gerbera? It is recommended that further research to identify factors that contribute to
structural stem strength in gerbera be conducted, particularly if accuracy in evaluating stem
strength can be attained.
Another important concern for the breeder involves defining the conditions under
which stem strength should be measured. An investigation of the effects of pre-harvest
environmental conditions versus morphological or anatomical variability of the stem could
yield useful information for developing a method to evaluate the magnitude of stem strength
more accurately.
Garifying breeding objectives and developing appropriate evaluation methods could be
the key to a successful breeding program to improve structural stem strength in gerbera.

CHAPTER 5
PART HI. VASE LIFE X STEM STRENGTH STUDIES
Introduction
Postharvest longevity is a critical attribute of cut flowers. De Jong (1978a) proposed
two main components, vase life and stem strength, contribute to postharvest longevity in
gerbera. In a second paper, De Jong (1978b) suggested structurally strong stems could
extend postharvest longevity by providing added support to the flower should a water
deficit occur.
The basis for vase life and stem strength in gerbera remains under investigation. Stem
anatomy studies have been conducted in an effort to understand the causes for variation in
postharvest longevity. Reiman-Philip, as cited by Wilberg (1973), noted a greater
proportion of large vascular bundles to small vascular bundles as a possible factor
contributing to stem strength. Siewert, also cited by Wilberg (1973), determined that such
a relationship existed in only extreme cases. In a comprehensive study, Marousky (1986)
measured the size and number of vascular bundles in two gerberas, Tropic Gold and
Appleblossom. He concluded Appleblossom exhibited more resistance to stem bending
than Tropic Gold partially because of various anatomical features such as fewer small
vascular bundles, smaller stem diameter, more vascular bundles per unit of circumference,
and a greater percentage of dry weight per unit of scape length.
The objectives of this study were to determine the relationship between vase life and
stem strength from a population of Gerbera X hvbrida Hort. which varied greatly in flower
and stem morphology and to compare stem anatomy of flowers from plants that were
classified by differences in vase life and stem strength.
64

65
Materials and Methods
Selection and Classification
Seventy-three plants were evaluated for both stem strength and vase life in 1984
(Chapter 4). Three flowers per plant were evaluated to obtain mean values for each
variable. Based on these means, 22 plants were selected and classified into four categories.
The categories were defined by vase life and stem strength ratings described in table 5-1.
To represent genetic diversity, selection included plants from five seed populations.
Table 5-1. Vase life x stem strength classification
Plant
Category
Vase Life
Stem Strength
I
High ^ 11.0 days
High ^ 11.5g/.2cm
II
High > 12.0 days
Low < 5.5g/.2cm
III
Low <6.0 days
High ^ 11.3g/.2cm
IV
Low ^ 4.0 days
Low ^2.5g/.2cm
Stem anatomy of flowers from plants assigned to each category was compared
(February 20-July 30, 1985). Approximately 14 flowers per plant were examined.
Appleblossom, assigned to category I, was also examined because of its low intraplant
variation for vase life and stem strength (Chapters 3 and 4).
Anatomy Examination
Flowers at various stages of maturity with at least one row of disc florets open were
randomly sampled. Using a light microscope with 40x magnification, fresh stem sections,
approximately 50-60 microns thick, were cut, using a razor blade, 12 cm below each
peduncle and examined. No stains were used. The number of large and small vascular
bundles was recorded for each flower. Determination of large and small bundles was
relative between plants. Therefore, before counting, the smallest large bundle was

66
measured. All bundles greater or equal to that size were described large, all others;
small. This method was described by F. J. Marousky, personal communication, 1985.
Circumference and area of each cross section were calculated using the diameter of each
stem, measured using a vernier caliper, at 12 cm below the peduncle.
A set of variables was created which described a series of mathematical relationships
determined by direct and indirect measurements. A list of these variables is given in table
5-2.
Table 5-2. List of variables from anatomy evaluation of gerbera.
LBUN = # of large bundles
SB UN = # of small bundles
TBUN = # of large bundles + # of small bundles
LBUNARAT = # of large bundles per 1.0 cm2 stem area @ 12 cm
SBUNARAT = # of small bundles per 1.0 cm2 stem area @ 12 cm
TBUN A RAT = Total # of bundles per 1.0 cm2 stem area @ 12 cm
LBUNCRAT = # of large bundles per 1.0 cm stem circumference @ 12 cm
SBUNCRAT = # of small bundles per 1.0 cm stem circumference @ 12 cm
TBUNCRAT = Total # of bundles per 1.0 cm stem circumference @ 12 cm
Results
Correlation Between Vase Life and Stem Strength
A significant positive correlation (r = .28) was observed when vase life means were
plotted against stem strength means from 73 plants belonging to the parental generation
described in Chapters 3 and 4. The relationship between these two components of
postharvest longevity appeared moderately weak, despite significance. Except for one
plant, plants with the weakest mean stem strength had a mean vase life of less than five
days. Distribution of these means is shown in figure 5-1.

40-,
67
y = 5.15 + .52x
r= .28
30-
M
Stem strength 20-
(F = g/.2cm)
10
0
0 10 20
Vase life (days)
Figure 5-1. Distribution of vase life and stem strength means for seventy-three plants
Comparison of Stem Anatomy Among Genotypes
Twenty two plants were selected and classified by high and low levels of vase life and
stem strength according to the categories described. Listing of these plants and their
corresponding mean vase life and stem strength is given in table 5-3.
Analyses of variance were made using plant means for nine stem anatomy variables.
Vase life and stem strength were treated as main effects. No significant interaction between
vase life and stem strength was determined for any of the nine variables. Analyses of
variance for the number of vascular bundles directly measured per stem section (TBUN,
LBUN, and SBUN) only yielded a significant difference between high and low levels of
vase life for number of large bundles (LBUN). The mean number of large vascular
bundles was significantly different for plants with high vase life compared to those with
low vase life. Plants with high vase life exhibited a smaller number of large vascular
bundles than plants with low vase life. A summary of these results is given in tables 5-4
and 5-5.
Analyses of variance for the number of vascular bundles calculated per unit stem area
(1.0 cm2) at 12 cm below the peduncle (TBUNARAT, LBUNARAT, and SBUNARAT),
only yielded significant differences between high and low levels of vase life for total

68
Table 5-3. Vase life and stem strength data used for classification of twenty-two plants
indu^n|^A£gleblossom/_________==_==_______^^_^_^__
Category I
Category III
Plant
Vase life
Stem
strength
Plant
Vase life
Stem strength
(days)
(g/.2cm)
(days)
(g/.2cm)
83-1-19
12.0
13.3
83-1-2
5.7
13.3
83-1-26
16.0
20.5
83-4-8
4.3
14.2
83-1-31
11.0
31.5
83-5-1
5.7
14.2
83-1-54
11.0
12.7
83-5-76
5.7
14.3
83-7-47
12.7
11.5
83-7-22
6.0
16.8
83-8-7
11.3
16.0
83-7-48
4.7
24.0
Appleblossom
11.0
25.7
83-7-35
5.7
11.3
Category II
Category IV
Plant
Vase life
Stem
strength
Plant
Vase life
Stem strength
(days)
(g/.2cm)
(days)
(g/.2cm)
83-1-6
13.0
4.3
83-1-24
3.7
1.2
83-1-32
12.3
4.7
83-1-79
4.0
2.0
83-1-64
15.0
5.5
83-4-29
3.3
1.5
83-1-77
15.3
4.2
83-4-16
3.0
2.3
Table 5-4. Analysis of variance for stem anatomy. (# of vascular bundles)
TBUN
LBUN
SBUN
Source
df
M.S.
M.S.
M.S.
Vase Life
(high ys low)
1
1.03 n s-
382.68 *
423.50 n s-
Stem Strength
(high ys low)
1
380.60 n s-
8.55 n-s-
275.07
Vase Life x Stem Strength
1
274.46 "-S.
15.14 n.s.
418.53 n-s-
Error
18
535.68
71.62
447.56
* Significant at P< 0.05.

69
Table 5-5. Means2 for number of vascular bundles between high and low levels of vase
lfeandstemstrengunjerbera;___=>^_B^_aig!aB_____==¡=a=_a=^_^___^
TBUN
LBUN
SBUN
Vase
High
44.92a
16.37a
33.55a
Life
Low
49.81a
18.59b
31.22a
Stem
High
48.76a
17.31a
31.45a
Strength
Low
50.97a
17.64a
33.33a
2 Means fo
strength are
lowed by different letters within each bundle group for vase life or stem
significant at the 5 % level.
number of bundles (TBUNARAT) and number of large bundles (LBUNARAT). The
means for total number of vascular bundles and number of large vascular bundles for this
unit area in gerbera stem sections were significantly different for plants with high vase life
compared to those with low vase life. Plants with high vase life exhibited a smaller total
number of vascular bundles per unit area than plants with low vase life. Also, plants with
high vase life exhibited a smaller number of large vascular bundles per unit area than plants
with low vase life. A summary of these results is given in tables 5-6 and 5-7.
Table 5-6. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm 2
scape area.)
TBUNARAT
LBUNARAT
SBUNARAT
Source
df
M.S.
M.S.
M.S.
Vase Life
(high vs low)
1
444840.20 *
130549.39 *
93420.23 n s-
Stem Strength
(high is low)
1
196864.30
12087.29 n s-
111390.20 n-s-
Vase Life x Stem Strength
1
10271.08 n s-
64.98 n-s-
8702.12 n s-
Error
18
79568.77
20121.40
26395.89
* Significant at P< 0.05.

70
Table 5-7. Means2 for number of vascular bundles per 1.0 cm 2 scape area between high
andlowlevelsofva^J^^djtemjffenghm^ierbCTa^^^^^^^^^^^^^^
TBUNARAT
LBUNARAT
SBUNARAT
Vase
Life
High
Low
239.06a
314.71b
78.85a
119.83b
160.22a
194.88a
Stem
Strength
High
Low
251.72a
302.05a
93.10a
105.57a
158.62a
196.48a
2 Means fo
strength are
lowed by different letters within each bundle group for vase life or stem
significant at the 5 % level.
Significant differences between high and low levels of vase life and high and low
levels of stem strength for total number of vascular bundles calculated per unit stem
circumference (1.0 cm) at 12 cm below the peduncle (TBUNCRAT) were determined. The
means for total number of vascular bundles per unit circumference in gerbera stem sections
were significantly different for plants with high vase life compared to those with low vase
life. Plants with high vase life exhibited a smaller total number of vascular bundles per unit
circumference than plants with low vase life. Also, the means for total number of vascular
bundles per unit circumference in gerbera stem sections were significantly different for
plants with high stem stength compared to those with low stem strength. Plants with high
stem strength exhibited a smaller total number of vascular bundles per unit circumference
than plants with low stem strength.
A highly significant difference between high and low levels of vase life was
determined for the number of large vascular bundles calculated per unit stem circumference
(1.0 cm) at 12 cm below the peduncle (LBUNCRAT). The means for number of large
vascular bundles per unit circumference in gerbera stem sections were significantly
different for plants with high vase life compared to those with low vase life. Plants with
high vase life exhibited a smaller number of large vascular bundles per unit circumference
than plants with low vase life. A significant difference between high and low levels of
stem strength was determined for the number of small vascular bundles calculated per unit

71
stem circumference (1.0 cm) at 12 cm below the peduncle (SBUNCRAT). The means for
number of small vascular bundles per unit circumference in gerbera stem sections were
significantly different for plants with high stem strength compared to those with low stem
strength. Plants with high stem strength exhibited a smaller number of small vascular
bundles per unit circumference than plants with low stem strength. A summary of these
results is given in tables 5-8 and 5-9.
Table 5-8. Analysis of variance for stem anatomy. (# of vascular bundles per 1.0 cm
scape circumference.)
TBUNCRAT
LBUNCRAT
SBUNCRAT
Source
df
M.S.
M.S.
M.S.
Vase Life
(high vs low)
1
1467.31 *
742.67 *
122.18 ns-
Stem Strength
(high vs low)
1
1108.91 *
61.55 n s-
647.96 *
Vase Life x Stem Strength
1
143.85 n-s-
69 n.s.
164.47
Error
18
237.65
80.43
124.06
* Significant at P< 0.05. Significant at P< 0.01.
Table 5-9. Means2 for number of vascular bundles per 1.0 cm scape circumference
between hi
gh and low levels of vase life and stem strength in gerbera.
TBUNCRAT LBUNCRAT SBUNCRAT
Vase
Life
High 30.46a 10.02a 20.44a
Low 34.81b 13.11b 21.70a
Stem
Strength
High 30.75a 11.12a 19.63a
Low 34.52b 12.01a 22.51b
2 Means fo
lowed by different letters within TBUNCRAT and SBUNCRAT for vase life
or stem strength are significant at the 5 % level and significant at the 1% level within
LBUNCRAT.

72
The only variable analyzed for which both significant differences were found between
high and low levels of vase life and stem strength was the total number of vascular bundles
per unit of circumference (TBUNCRAT). The means for this variable with respect to both
levels of vase life and stem strength are presented in table 5-10. The mean for total number
of vascular bundles per unit stem circumference (1.0 cm) at 12 cm below the peduncle was
lowest for plants with high vase life and high stem strength. Conversely, the mean was
highest for plants with low vase life and low stem strength. The difference between the
means of plants with high and low vase life for total number of vascular bundles per unit of
circumference among plants with high stem strength is not equal to the difference between
the means of plants with high and low vase life for total number of vascular bundles per
unit of circumference among plants with low stem strength (28 34 ^ 33 36). Similarly,
the difference between the means of plants with high and low stem strength for total
number of vascular bundles per unit of circumference among plants with high vase life is
not equal to the difference between the means of plants with high and low stem strength for
total number of vascular bundles per unit of circumference among plants with low vase life
(28 33 ^ 34 36). These inequalities possibly indicate a source of experimental error
since interaction effects were non-significant.
Table 5-10. Means for total number of vascular bundles per 1.0 cm scape circumference
between high and low levels of vase life and stem strength of twenty two gerbera plants.
Vase Life
High
Low
X
Stem
Strength
High
28
34
31.0
Low
33
36
34.5
X
30.5
35.0
Correlations Between Stem Anatomy and Postharvest Longevity
Correlation coefficients between nine stem anatomy variables and two components of
postharvest longevity, vase life and stem strength, were determined using means from

73
twenty two plants. Significant negative relationships were determined between vase life
and two stem anatomy variables which were based on total number of vascular bundles
(TBUNARAT, and TBUNCRAT). Significant relationships between vase life and three
stem anatomy variables which were based on number of large vascular bundles (LBUN,
LBUNARAT, and LBUNCRAT) were also negative. Highly significant at the 1 % level,
the strongest correlation was between vase life and number of large bundles per unit of
circumference. Correlations relating to stem anatomy variables based on small vascular
bundles and vase life or stem strength were not significant Summary of these results are
given in tables 5-11, 5-12, and 5-13.
Table 5-11. Correlation coefficients between stem anatomy (number of vascular bundles)
and^os^arvesdongevit^^^^___________
TBUN
TBUNARAT
TBUNCRAT
Vase Life
(days)
.06
-.46*
-.44*
Stem Strength
(F=g/.2 cm)
-.27
-.33
-.39
* Significant at P< 0.05.
Table 5-12. Correlation coefficients between stem anatomy (number of large vascular
bundles) and postharvest longevity.
LBUN
LBUNARAT
LBUNCRAT
Vase Life
(days)
-.51*
-.53*
-.67 *
Stem Strength
(F=g/.2 cm)
-.21
-.21
-.27
* Significant at P< 0.05. Significant at P< 0.01.
Table 5-13. Correlation coefficients between stem anatomy (number of small vascular
bundles) and postharvest longevity.
SBUN
SBUNARAT
SBUNCRAT
Vase Life
(days)
.37
-.34
-.23
Stem Strength
(F=g/.2 cm)
-.08
-.40
-.40

74
Discussion
The relationship between vase life and structural stem strength described in this study
was weak (r = .28) but, significant at the level of P< 0.05. Therefore, it would be
expected that flowers with the highest vase life will only have a slight tendency for high
stem strength. Breeding for both characters may improve keeping quality in gerbera cut
flowers only marginally.
Whether or not the correlation provides a good description of the relationship deserves
comment First of all, plant means for each variable were plotted against each other. Since
evaluating vase life and stem strength requires destructive tests, plant means for each
variable were determined by different sets of flowers. Also, vase life means were
determined from flowers evaluated during the period May 5 to June 12,1985, and stem
strength means were determined from flowers evaluated during the period June 12 to July
30, 1985. Environmental differences during these two periods may not have affected all of
the plants similarly. In addition, problems associated with the accuracy of stem strength
determinations (Chapter 4) may have skewed the comparison.
Variation in the number of vascular bundles per cross-section of gerbera flower stems
cut 12 cm below the peduncle was compared among plants whose flowers exhibited high
or low vase life and stem strength. Large and small bundles were counted. Their sum was
calculated to obtain a value for total number of bundles. The diameter of each cross-section
was also measured. Using the diameter measurement, the number of bundles per unit area
(1.0 cm2) or unit circumference (1.0 cm) for each cross-section was also calculated.
Overall, differences in the number of large bundles appears related to vase life. Fewer
number of large bundles, fewer number of large bundles per unit area, and fewer number
of large bundles per unit circumference were observed in plants with high vase life. In
contrast, only the difference in the number of small bundles per unit circumference appears
to affect stem strength. Fewer numbers of small bundles were observed in plants with high

75
stem strength. Therefore, it follows that the fewer total number of vascular bundles per
unit circumference was observed in plants with high vase life and high stem strength.
Results from correlations determined between these stem anatomy variables and vase
life or stem strength enhance the link suggested between size and number of vascular
bundles to postharvest longevity. Fewer bundles, small or large, were found in plants with
high vase life or high stem strength. Negative correlations between stem anatomy variables
involving large bundles were significant for vase life but not stem strength at the level of
P< 0.05. Correlations between stem anatomy variables involving small bundles were not
significant for vase life or stem strength at the level of P< 0.05. However, negative
correlations between numbers of small bundles per unit area or unit circumference and stem
strength were significant at the level of P< 0.10.
Strong turgor strength in gerbera stems has been considered an important factor for
maintaining postharvest longevity (De Jong, 1978a). Assuming a constant supply of water
available to flowers with equal stem diameters, fewer vascular bundles would be expected
to increase the flow rate of water in the stem. It seems likely that this situation would serve
to maintain an upright stem, if stem rigidity depended on the maintenance of turgor.
Additionally, an increase in flow rate of water in the stem may reduce the opportunity for
microbial growth to occur, a factor cited to decrease vase life in gerbera (Meeteren, 1978a).
Since differences in large vascular bundles and not small vascular bundles were found in
plants with high and low levels of vase life, it is theorized that large vascular bundles are
mainly responsible for differences in water uptake rates in gerbera flowers.
As for the role of vascular bundles in stem strength, only small bundles appeared to be
a factor. De Jong (1986) proposed that stronger structural stem strength would help to
decrease the incidence of stem folding. Marousky (1986) examined stem anatomy,
including number and size of vascular bundles, of two cultivars whose flowers differed in
resistance to scape breakage. The cultivar that showed less resistance to breakage had more
small bundles than the cultivar that showed more resistance to breakage. One reason for

76
stronger stems exhibiting less incidence of folding could be because stronger stems have
fewer small vascular bundles.
Dubuc-Lebreux and Vieth (1985) observed that a greater proportion of support tissues,
such as sclerenchyma, was found in lower regions of the stem where breakage did not
normally occur. They suggested that plants whose stems showed less incidence of folding
at the typical zone of folding (10-15 cm below the peduncle) might have more support
tissues in this area compared to plants whose stems showed more incidence of folding.
Perhaps in stonger stems, small vascular bundles are displaced by more support tissue;
therefore, the fewer small vascular bundles in a stem cross-section, the greater proportion
of support tissue will be present
Conclusion
It is concluded that ifvase life and stem strength are two distinct components of
postharvest longevity in Gerbera X hvbrida Hort., then the breeder wishing to improve
postharvest quality in this cut flower might consider selecting plants with good combining
ability for both high vase life and high stem strength.
Stem anatomy studies yielded interesting results. It appears that vase life is affected by
the number of large vascular bundles and stem strength is affected by the number of small
vascular bundles. Fewer bundles of each type per unit of circumference were found in
plants with high vase life and high stem strength.
Additional genetic investigations to determine the heritability of stem anatomy variables
involving number of large vascular bundles and stem anatomy variables involving number
of small vascular bundles are recommended.

CHAPTER 6
SUMMARY
Breeding cultivare with superior postharvest longevity may be the best approach for
satisfying the floral consumers expectation of cut flower postharvest quality. Combining
this effort with postharvest treatments, such as the use of floral preservatives, would help
maintain the popularity of gerbera for years to come.
The heritability of two characteristics, vase life and stem strength, considered to be
components of postharvest longevity, was determined. Although results indicate that
intense selection of plants for these traits will yield increases in postharvest longevity, it is
of utmost importance to determine appropriate breeding objectives which will best improve
postharvest performance.
In the case of gerbera, a simple analysis of vase life appears inadequate. Breeding
programs which incorporate selection against frequency of bending and folding and
selection for increased days to wilting may be more effective. Further research is needed to
prove that undesireable modes of senescence, i.e. bending and folding, are qualitatively
inherited by only several genes.
An attempt to document the role of stem strength in postharvest longevity of gerbera
was made; however, the function of this component cannot be specified at this time.
Difficulties associated with measuring stem strength were discussed. As factors are
identified which affect stem strength, breeding programs to improve postharvest
performance can be refined.
While vase life and stem strength have been postulated as components of postharvest
longevity, their effects appear independent of each other. Correlations demonstrated that
77

78
vase life increased as stem diameter decreased. In contrast, stem strength increased as stem
diameter also increased.
The number of vascular bundles observed in cross sections of scapes at the zone of
bending or folding suggested that the ratio of small bundles per unit scape circumference
may affect stem strength and the ratio of large bundles per unit circumference may affect
vase life. Heritability studies on these anatomical ratios in gerbera could provide pertinent
information that would be useful to plant breeders.

REFERENCES
Abdel-Kader, H. and Rogers, M.N., 1986. Postharvest treatment of Gerbera iamesonii.
ActaHort., 181: 169-176.
Amariutei, A., Burzo, I. and Alexe, C, 1986. Researches concerning some metabolism
aspects of cut gerbera flowers. Acta Hort., 181: 331-337.
Awad, A.R.E., Meawad, A., Dawh, A.K. and Elsakaa, M., 1986. Cut flower longevity
as affected by chemical pretreatment. ActaHort., 181:177-172.
Ballantyne, D.J., 1963. Respiration of floral tissue of the daffodil (Narcissus
pseudonarcissus Linn.) treated with benzyladenine and auxin. Can. J. Bot., 44: 117-119.
Barigozzi, C. and Quagliotti, L., 1978. Current research on breeding of gerbera. In: L.
Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera.
Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 57-68.
Benschop, M. and De Hertogh, A.A., 1969. An analysis of the post-harvest
characteristics of cut tulips. Flor. Rev., 145: 24-25, 62-65.
Borghi, B. and Baldi, V., 1970. Variabilita tra cloni di gerbera (Gerbera jamesonii) allevati
in diverse condizioni ambientali. Sementi elette, 6: 25-32.
Bravdo, B., Mayak, S. and Gravrieli, Y., 1974. Sucrose and water uptake from
concentrated sucrose solutions by gladiolus shoots and the effect of these treatments on
floret life. Can. J. Bot., 52: 1271-1281.
Bufler, G., 1986. Ethylene biosynthesis and action. ActaHort., 181:93-98.
Burdett, A.N., 1970. The cause of bent neck in cut roses. J. Amer. Soc. Hort. Sci., 95:
427-431.
Buys, C, 1978. Quality problems in gerbera. In: L. Quagliotti and A. Baldi (Editors),
Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-
28 April, at Alassio, Italy, pp. 255-262.
Carpenter, WJ. and Rasmussen, H.P., 1973. Water uptake rates by cut roses (Rosa
hvbrida) in light and dark. J. Amer. Soc. Hort. Sci., 98: 309-313.
Carpenter, W.J. and Rasmussen, H.P., 1974. The role of flower and leaves in cut flower
water uptake. Scientia Hortic., 2: 293-298.
Cocozza Talia, M., Porceddu, E., and Stellacci, P., 1978. Genetic variability in gerbera.
In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera.
Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 163-177.
79

80
Coorts, G.D., 1973. Internal metabolic changes in cut flowers. HortScience, 8: 195-198.
Coorts, G.D., McCollum, J.P. and Gartner, J.B., 1965. Effect of senescence and
preservative on mitochondrial activity in flower petals of Rosa hybrida. Velvet Times.
Proc. Amer. Soc. Hort. Sci., 86: 791-797.
De Jong, J., 1978a. Dry storage and subsequent recovery of cut gerbera flowers as an aid
in selection for longevity. Scientia Hort., 9: 389-397.
De Jong, J., 1978b. Selection for keeping quality in gerbera. In: L. Quagliotti and A.
Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a
Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 263-268.
De Jong, J., 1986. Breeding for keeping quality in gerbera. Acta Hort., 181: 353-357.
De Jong, J. and Garretsen, F., 1985. Genetic analysis of cut flower longevity in gerbera.
Euphytica, 34: 779-784.
De Leo, V. and Ottaviano, E., 1978. Genetical analysis of morphological traits in Gerbera
iamesonii. Clonal and diallel families. In: L. Quagliotti and A. Baldi (Editors), Genetics
and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April,
at Alassio, Italy, pp. 179-191.
De Witte, Y. and Doom, W.G. van, 1988. Identification of bacteria in the vase water of
roses, and the effect of the isolated strains on water uptake. Scientia Hortic., 35: 285-291.
Dilley, D.R. and Carpenter, W.J., 1975. The role of chemical adjuvants and ethylene
synthesis on cut flower longevity. Acta Hort., 41: 117-132.
Dixon, M.A. and Peterson, C.A., 1989. A re-examination of stem blockage in cut roses.
Scientia Hortic., 38: 277-288.
Dixon, M.A., Butt, J.A., Muir, D.P. and Tsujita, M.J., 1988. Water relations of cut
greenhouse roses:the relationships between stem water potential, hydraulic conductance
and cavitation. Scientia Hortic., 36: 109-118.
Downs, C. and Reihana, M., 1987. Extending vaselife and improving quality of nerine cut
flowers with preservatives. HortScience, 22: 670-671.
Drennan, D., Harding, J. and Byrne, T.G., 1986. Heritability of inflorescence and floret
traits in gerbera. Euphytica, 35: 319-330.
Dubuc-Lebreux, M.A. and Vieth, J., 1985. Histologie du pedoncule inflorescentiel de
Gerbera iamesonii. ActaBot. Neerl., 34(2): 171-182.
Durkin, D. and Kuc, R., 1966. Vascular blockage and senescence of the cut rose flower.
Proc. Amer. Soc. Hort. Sci., 89: 683-688.
Eijk, J.P. van and Eikelboom, W., 1976. Possibilities of selection for keeping quality in
tulip breeding. Euphytica, 25: 353-359.
Eijk, J.P. van and Eikelboom, W., 1980. Methods of selection in tulip breeding. Acta
Hort., 109: 217-225.

81
Eijk, J.P. van and Eikelboom, W., 1986. Aspects of breeding for keeping quality in
Tulipa. Acta Hort, 181: 237-244.
Eijk, J.P. van, Eikelboom, W. and Spamaaij, L.D., 1977. Possibilities of selection for
keeping quality in tulip breeding. Euphytica, 26: 825-828.
Falconer, D.S., 1960. Introduction to Quantitative Genetics. Oliver and Boyd,
Edinburgh, 365 pp.
Ferreira, D.I. and Swaidt, G.H., 1981a. The influence of the number of foliage leaves on
the vase life of cut rose flowers in two media. Agroplante, 13: 73-76.
Ferreira, D.I. and Swardt, G.H., 1981b. A comparison of the vase life and respiration rate
of ten cut rose cultivars and the influence of a flower preservative thereupon. Agroplante,
13: 77-81.
Fujino, D.W., Reid, M.S. and Yang, S.F., 1981. Effects of aminooxyacetic acid on
postharvest characteristics of carnation. Acta Hort., 113: 59-64.
Gilbart, D.A. and Dedolph, R.R., 1965. Phytokinin effects on respiration and
photosynthesis in roses and broccoli. Proc. Amer. Soc. Hort. Sci., 86: 774-778.
Griffing, B., 1956. Concept of general and specific combining ability in relation to diallel
crossing systems. Australian J. of Biol. Sci., 9: 463-493.
Halevy, A.H., 1976. Treatments to improve water balance of cut flowers. Acta Hort., 64:
223-230.
Hallauer, A.R. and Miranda, J.B., 1981. Quantitative Genetics in Maize Breeding. Iowa
State University Press, Ames, Iowa, 480 pp.
Harding, J., Byrne, T.G. and Nelson, R.L., 1981. Heritability of cut-flower vase
longevity in gerbera. Euphytica, 30: 653-657.
Heide, O.M. and 0ydvin, J., 1969. Effects of 6-benzylaminopurine on the keeping
quality and respiration of glasshouse carnations. Hort. Res., 9: 26-36.
Henderson, C.R., 1953. Estimation of variance and covariance components. Biometrics,
9: 226-252.
Kalkman, E.Ch., 1985. Postharvest treatment of Astilbe hvbrida. Acta Hort., 181: 389-
392.
Kofranek, A.M. and Halevy, A. H., 1981. Chemical pretreatment of chrysanthemums
before shipment. Acta Hort., 113: 89-96.
Kofranek, A.M. and Paul, J.L., 1975. The value of impregnating cut stems with high
concentration of silver nitrate. Acta Hort., 41: 199-206.
Kohl, H.C., 1968. Gerberas: Their culture and commercial possibilities. S. Flor, and
Nut., 28: 18,24-26.

82
Kuc, R. and Workman, M., 1964. The relation of maturity to the respiration and keeping
quality of cut carnations and chrysanthemums. Proc. Amer. Soc. Hort. Sci., 84: 575-581.
La Malfa, G. and Noto, G., 1978. Variations in the qualitative characteristics of Gerbera
iamesonii hvbrida flowers in relation to environmental conditions. In: L. Quagliotti and A.
Baldi (Editors), Genetics and Breeding of Carnation and Gerbera. Proceedings of a
Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.233-244.
Larsen, F.E. and Frolich, M., 1969. The influence of 8-hydroxyquinoline citrate, n-
dimethylamino succinamic acid, and sucrose on respiration and water flow in Red Sim
cut carnations in relation to flower senescence. J. Amer. Soc. Hort. Sci., 94: 289-292.
Leeuwen, P.J. van, 1986. Post-harvest treatment of Euphorbia fulgens. Acta Hort., 181:
467-472.
Lineberger, R.D. and Steponkus, P.L., 1976. Identification and localization of vascular
occlusions in cut roses. J. Amer. Soc. Hort. Sci., 101: 246-250.
MacLean, D.C. and Dedolph, P.L., 1962. Effects of N6-benzylaminopurine on post-
harvest respiration of Chrysanthemum morifolium and Dianthus carvophvllus. Bot. Gaz.,
124: 20-21.
Marousky, F.J., 1969. Vascular blockage, water absorption, stomatal opening, and
respiration of cut Better Times roses treated with 8-hydroxyquinoline citrate and sucrose.
J. Amer. Soc. Hort. Sci., 94: 223-226.
Marousky, F.J., 1972. Water relations, effects of floral preservatives on bud opening, and
keeping quality of cut flowers. HortScience, 7: 114-116.
Marousky, F.J., 1973. Recent advances in opening bud-cut chrysanthemum flowers.
HortScience, 8: 199-201.
Marousky, F.J., 1986. Vascular structure of the gerbera scape. Acta Hort., 181: 399-
406.
Marousky, F.J. and Nanney, J., 1972. Influence of storage temperatures, handling and
floral preservatives on post harvest quality of gypsophila. Proc. Fla. State Hort. Soc.,
85:419-422.
Marousky, F.J. and Woltz, S.S., 1975. Relationship of floral preservatives to water
movement, fluoride distribution, and injury in gladiolus and other cut flowers. Acta Hort.,
41: 171-182.
Maurer, J., 1968. Genetisch-zlichterische untersuchungen bei Gerbera iamesonii H.
Bolus. Z. Pflanzenzlichtung, 60: 113-143.
Mayak, S. and Dilley, D.R., 1976. Effect of sucrose on response of cut carnation to
kinetin, ethylene, and abscisic acid. J. Amer. Soc. Hort. Sci., 101: 583-585.
Mayak, S. and Halevy, A.H., 1970. Cytokinin activity in rose petals and its relation to
senescence. Plant Physiol., 46: 497-499.

83
Mayak, S. and Halevy, A.H., 1974. The action of kinetin in improving the water balance
and delaying senescence processes of cut rose flowers. Physiol. Plant, 32: 330-336.
Mayak, S. and Halevy, A.H., 1980. Rower senescence, In: Senescence in Plants, K.V.
Thimann (Editor), CRC Press, Boca Raton, Florida, 131-156.
Mayak, S., Halevy, A.H. and Katz, M., 1972. Correlative changes in phtyohormones in
relation to senescence processes in rose petals. Physiol. Plant, 27: 1-4.
Mayak, S., Halevy, A.H., Sagie, S., Bar-Yoseph, A. and B. Bravdo, 1974. The water
balance of cut rose flowers. Physiol. Plant, 31: 15-22.
Meeteren, U. van, 1978a. Water relations and keeping quality of cut gerbera flowers. I.
The cause of stem break. Scientia Hort., 8: 65-74.
Meeteren, U. van, 1978b. Water relations and keeping quality of cut gerbera flowers. II.
Water balance of ageing flowers. Scientia Hort, 9: 189-197.
Meeteren, U. van, 1979. Water relations and keeping quality of cut gerbera flowers. III.
Water content, permeability, and dry weight of ageing petals. Scientia Hort., 11: 83-93.
Meeteren, U. van, 1980. Water relations and keeping quality of cut gerbera flowers. VI.
Role of pressure potential. Scientia Hort., 12: 283-292.
Meeteren, U. van, 1981. Role of pressure potential in keeping quality of cut gerbera
inflorescences. Acta Hon., 113: 143-150.
Meeteren, U. van and Gelder, H. van, 1980. Water relations and keeping quality of cut
gerbera flowers. V. Role of endogenous cytokinins. Scientia Hort., 12: 273-281.
Mohsenin, N.N., 1970. Physical Properties of Plant and Animal Materials. Gordon and
Breach Science Publication, New York, 123 pp.
Mutsenietse, G. Ya., RashaT, I. D. and Dishler, V. Ya., 1978. Inheritance of quantitative
characters in gerbera in diallel crosses. II. Inflorescence characters. Genetika, 14: 779-
783.
Nichols, R., 1981. Ethylene, present and future. Acta Hort., 113: 11-18.
Nichols, R 1982. Growth regulators and flower senescence. British Plant Growth
Regulator Group, Monograph, 8: 113-120.
Nowak, J., 1981. Regulation of bud opening, storage period, vase-life and senescence of
cut flowers. I. The effect of silver complexes and sucrose on longevity of cut gerbera
inflorescences stored for different periods of time. Res. Inst, of Pom. and Floric., Ann.
Report, 2: 3-10.
Nowak, J. and Plich, H., 1981. Regulation of bud opening, storage period, vase-life, and
senescence of cut flowers. II. The effect of silver ions and other anti-ethylene agents on
ethylene synthesis and senescence of gerbera inflorescences. Res. Inst, of Pom. and
Floric., Ann. Report, 2: 10-21.

84
Ottaviano, E., Schiva, T., and Sari-Gorla, M., 1974. Analisi multivariata per lo studio
delle differenze genetiche in popolazioni di Gerbera jamesonii. Genet. Agr. 28: 292-306.
Parups, E.V. and Molnar, J. M., 1972. Histochemical study of xylem blockage in cut
roses. J. Amer. Soc. Hort. Sci., 97: 532-534.
Parups, E.V. and Voisey, P.W., 1976. Lignin content and resistance to bending of the
pedicel in greenhouse-grown roses. J. Hort. Sci., 51: 253-259.
Paulin, A. and Muloway, K., 1979. Perspectives in the use of growth regulators to
increase the cut flower vase life. Acta Hort., 91: 135-141.
Pauli, R., Goo, T., Criley, R.A. and Parvin, P.E., 1981. Leaf blackening in cut Protea
eximia: Importance of water relations. Acta Hort, 113: 159-168.
Rasmussen, H.P. and Carpenter, W.J., 1974. Changes in the vascular morphology of cut
rose stems: A scanning electron microscope study. J. Amer. Soc. Hort. Sci., 99: 454-459.
Reddy, T.V., 1988. Mode of action of cobalt extending vase life of cut roses. Scientia
Hortic., 36: 303-313.
Reid, M.S. and Kofiranek, A.M., 1981. Recommendations for standardized vase-life
evaluations. Acta Hort., 113: 171-174.
Rogers, M.N., 1973. An historical and critical review of postharvest physiology research
on cut flowers. HortScience, 8: 189-194.
SAS Institute Inc., 1982. SAS Users Guide: Statistics, 1982 Edition. SAS Institute Inc.,
Cary, N.C., 584 pp.
Schaffer, H.E. and Usanis, R.A., 1969. General least squares analysis of diallel
experiments. A computer program-DIALL. Genetics Dept., North Carolina State Univ.,
Raleigh, N.C. 27607.
Schiva, T., 1973. Analisi dei caratteri metrici di importanza commerciale su Gerbera
iamesonii e conseguenze dellinincrocio. Estrato dagli Annali della facolt di Scienze
Agrarie dellUniversita degli Studi di Torino, 8: 355-366.
Schiva, T., 1975. Miglioramento gentico dell gerbera. II. Analisi gentica di caratteri
metrici di importanza commerciale e conseguenze dellinincrocio. Genet Agr., 29: 233-
240.
Serini, G. and De Leo, V., 1978. Phenotypic characters and preservability of gerbera
flowers. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and
Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 269-277.
Siegelman, H.W., 1952. The respiration of rose and gardenia flowers. Proc. Amer. Soc.
Hort. Sci., 59: 496-500.
Simmonds, N.W., 1979. Principles of Crop Improvement. Longman Group Limited,
London, 408 pp.
Smith, D.E. and Nelson, R.L., 1967. Gerbera propagation. Calif. Agrie., 21(12): 7.

85
Staden, O.L. and Beekhuizen, J.G., 1986. A new formula for the anti-ethylene
pretreatment of cut carnations. Acta Hort., 181: 425-428.
Stimart, D.P. and Brown, D J., 1982. Regulation of postharvest flower senescence in
Zinnia elegans Jacq. Scientia Hortic., 17: 391-396.
Stoddard, E.M. and Miller, P.M., 1962. Chemical control of water loss in growing
plants. Science, 137: 224-225.
Swan, A., 1981. Quality of Lilium Enchantment flowers as influenced by season and
silver thiosulphate. Acta Hort., 113: 45-50.
Sytsema, W., 1975. Conditions for measuring vase life of cut flowers. Acta Hort., 41:
217-226.
Sytsema, W., 1981. Vase life and development of carnations as influenced by silver
thiosulphate. Acta Hort., 113: 33-38.
Sytsema, W., 1986. Post-harvest treatment of freesia with silver thiosulphate and
cytokinins. Acta Hort., 181: 439-442.
Tandler, J., Mor, Y., Spiegelstein, H. and Mayak, S., 1986. Chemical treatments to
improve the quality of cut gypsophila flowers. Acta Hort., 181: 443-450.
Tesi, R., 1978. Variation of some characters of flowers in clones of Gerbera iamesonii
hvbrida. In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and
Gerbera. Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp.227-232.
Waters, W.E., 1964. Influence of chemical preservatives on keeping quality of asters,
carnations, chrysanthemums, and gerbera daisies. Proc. Fla. St. Hort. Soc., 77: 466-470.
Wilberg, B 1973. Physiological investigations on the problem of freedom from stem
break as a prerequisite in breeding gerbera forms which remain upright when used as cut
flowers. Z. Pflanzenzuchtung, 69: 107-114.
Woltering, E.J., 1986. Postharvest treatment of Gvpsophila paniculata. Acta Hort., 181:
338.
Wricke, G., Weber, W. E. and Ottleben, R., 1982. Die analyse der genetischen varianz
bei quantitativ vererbten Merkmalen von Gerbera iamesonii. Z. Pflanzenziichtung, 89:
329-336.
Wulster, G., Sacalis, J. and Janes, H.W., 1982. Senescence in isolated carnation petals.
Plant Physiol., 70: 1039-1043.
Zagory, D. and Reid, M.S., 1986. Evaluation of the role of vase micro organisms in the
postharvest life of cut flowers. Acta Hort., 181: 207-218.
Zieslin, N., Kohl, H.C., Kofranek, A.M. and Halevy, A.H., 1978. Changes in the water
status of cut roses and its relationship to bent-neck phenomenom. J. Amer. Soc. Hort.
Soc., 103: 176-179.

APPENDIX A
VASE LIFE STUDIES

A-l. Vase life data for thirty-one parents from a population of gerbera.
Parent
(Parental Generation)
Mean C.V.
(Top-cross Generation)
Progeny Mean
83-1-6
13.0
20.4
6.8
83-1-13
11.3
5.1
6.8
83-1-26
16.0
16.6
7.7
83-1-32
12.3
4.7
8.2
83-1-61
12.3
9.4
6.8
83-1-64
15.0
17.6
7.9
83-1-74
13.7
23.5
6.9
83-1-77
15.3
24.7
8.6
83-3-4
11.7
9.9
7.5
83-3-34
13.3
11.5
7.2
83-3-38
15.7
7.4
8.7
83-3-47
13.0
20.4
7.4
83-3-77
13.0
7.7
6.4
83-3-96
13.3
11.5
7.4
83-4-4
11.7
5.0
6.1
83-4-17
12.3
20.4
6.3
83-4-38
11.3
5.1
7.8
83-4-69
12.0
8.3
9.6
83-5-33
14.0
7.1
7.6
83-5-39
16.0
22.5
10.4
83-5-109
13.0
21.7
8.9
83-6-42
14.3
17.6
7.9
83-7-4
12.3
18.7
10.2
83-7-6
13.7
16.9
7.8
83-7-10
16.0
0.0
10.7
83-7-14
12.7
18.2
6.6
83-7-18
14.3
10.7
6.9
83-7-26
12.3
18.7
9.4
83-7-29
16.0
0.0
9.4
83-7-47
12.7
4.6
8.6
83-8-7
11.3
5.1
7.6
87

88
A-2. 5x5 Diallel. Vase life means based on progeny means of gerbera crosses.
83-1-77
83-4-69
83-5-109
83-7-4
83-7-10
83-1-77
x = 9.2
x = 9.3
x = 12.5
x = 9.8
n = 13
n= 18
n = 19
n = 13
83-4-69
83-5-109
x= 11.0
n = 19
x = 8.6
n = 12
x = 8.5
n = 16
x= 12.2
n = 18
x = 11.7
n = 21
83-7-4
x = 11.8
n = 21
x = 11.1
n = 9
x = 11.0
n = 22
x= 13.4
n = 25
83-7-10
x= 15.5
n = 3
x = 14.0
a = 1
x = 12.5
n = 18

89
A-3. Parental Generation. Phenotypic correlation coefficients between stem and
inflorescence traits in gerbera.
i 1 g", mmm m == ==
Traits
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Stem Length
b
-j
*
.16 *
.20* *
.06*
Stem Diameter
.10* *
.64**
.26**
Inflorescence Diameter
.23* *
-.07 *
Disc Diameter
.17* *
* Significant at P< 0.05. Significant at P< 0.01.
A-4. Parental Generation. Phenotypic correlation coefficients between stem and
inflorescence traits in gerbera by senescence mode.
Traits
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Stem Length
Bending
.07
.11**
.18 *
.07
Folding
.12
.31 **
.35 **
-.02
Wilting
.16*
.21 **
.29**
.09
Stem Diameter
Bending
.15**
.63**
.28**
Folding
.15
.58 **
.07
Wilting
-.01
.76**
.14*
Inflorescence Diameter
Bending
.21**
-.02
Folding
.34**
-.28 *
Wilting
.06
-.20 *
Disc Diameter
Bending
.20* *
Folding
-.08
Wilting
.09
* Significant at P< 0.05. Significant at P< 0.01.

90
A-5. Gerbera X hvbrida Hort.. Vase life data for backcrosses.
Backcross
Number of Progeny
Mean Vase Life
84-27 -2
X
83-1-77
13
10.4
84-27-2
X
Appleblossom
2
8.0
84-27-4
X
83-1-77
9
10.1
84-27-4
X
Appleblossom
6
8.3
84-27-7
X
83-1-77
1
3.0
84-27 12
X
83-1-77
1
13.0
84- 37 -4
X
83-4-69
5
9.4
84- 37 -4
X
Appleblossom
19
8.4
84 37 10
X
83-4-69
12
9.3
84 37 10
X
Appleblossom
16
9.9
84 37 12
X
83-4-69
21
9.6
84 37 12
X
Appleblossom
11
9.6
84-41-3
X
83-7-4
19
10.7
84-41-3
X
Appleblossom
1
10.0
84-41-5
X
83-7-4
1
13.0
84-41-8
X
83-7-4
1
10.0
84-41 -13
X
Appleblossom
16
10.0
84-41 15
X
83-7-4
22
13.7
84-41 -15
X
Appleblossom
1
13.5
84-41 19
X
83-7-4
1
12.0
84 41 20
X
Appleblossom
3
8.3
84-43-2
X
83-7-10
27
12.4
84-43-2
X
Appleblossom
11
8.9
84-43-5
X
Appleblossom
7
9.5
84-43- 18
X
83-7-10
19
11.2
84-43- 18
X
Appleblossom
2
9.3

91
A-6. Gerbera X hvbrida Hort.. Frequency of senescence modes for backcrosses.
Backcross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-27-2
X
83-1-77
30
26.7
40.0
33.3
84-27-2
X
Appleblossom
6
16.7
33.3
50.0
84-27-4
X
83-1-77
13
30.8
23.1
46.2
84-27-4
X
Appleblossom
12
8.3
58.3
33.3
84-27-7
X
83-1-77
1

100

84 27 12
X
83-1-77
1

100

84- 37 -4
X
83-4-69
8
12.5
12.5
75.0
84-37-4
X
Appleblossom
33
33.3
24.2
42.4
84 37 10
X
83-4-69
22
40.9
22.7
36.4
84 37 10
X
Appleblossom
34
26.5
47.1
26.5
84-37- 12
X
83-4-69
35
11.4
51.4
37.1
84- 37 12
X
Appleblossom
20
15.0
50.0
35.0
84-41-3
X
83-7-4
34

41.2
58.8
84-41-3
X
Appleblossom
1


100
84-41-5
X
83-7-4
1


100
84-41-8
X
83-7-4
1

100

84-41 13
X
Appleblossom
26
23.1
30.8
46.2
84-41 15
X
83-7-4
38
2.6
26.3
71.1
84-41 15
X
Appleblossom
2


100
84-41 19
X
83-7-4
1


100
84 41 20
X
Appleblossom
3
33.3
33.3
33.3
84-43-2
X
83-7-10
62
14.5
24.2
61.3
84-43-2
X
Appleblossom
32
28.1
37.5
34.4
84-43-5
X
Appleblossom
19
31.6
26.3
42.1
84-43- 18
X
83-7-10
30
20.0
40.0
40.0
84-43- 18
X
Appleblossom
3

66.7
33.3

92
Cross
Number of Progeny
Mean Vase Life
84-27 -2
X
84 27 12
10
9.7
84-27 -4
X
84-27- 12
2
9.5
84- 37 -4
X
84 37 10
3
9.7
84- 37 12
X
84 37 10
8
10.6
84-41-3
X
84-41-5
13
9.5
84-41-5
X
84-41-8
7
8.1
84-41-8
X
84-41 -13
4
6.8
84-41 13
X
84-41-8
9
9.2
84-41 15
X
84-41-5
23
9.6
84-41 15
X
84-41-8
1
14.7
84-41-20
X
84-41 13
3
12.7
84-43 -2
X
84-43-3
20
11.0
84-43-5
X
84-43-3
1
15.0
84-43 -7
X
84-43-3
22
9.6
84-43- 18
X
84-43-3
1
14.0

93
A-S^GerberaXhvbrida^Hort^Fre^uencv^ofsenescence^m^es^for^full-sibcrosses^
Cross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-27-2
X
84 27 12
25
36.0
32.0
32.0
84-27-4
X
84 27 12
2

100

84- 37 -4
X
84 37 10
6

16.7
83.3
84 37 12
X
84 37 10
16
12.5
43.7
43.7
84-41-3
X
84-41-5
23
21.7

78.3
84-41-5
X
84-41-8
13
15.4
30.8
53.8
84-41-8
X
84-41 13
5

100

84-41 13
X
84-41-8
21
9.5
57.1
33.3
84-41 15
X
84-41-5
45
11.1
17.8
71.1
84-41 15
X
84-41-8
3

66.7
33.3
84 41 20
X
84-41 13
6
16.7
33.3
50.0
84-43-2
X
84-43-3
46
17.4
28.3
54.3
84-43-5
X
84-43 -3
1

100

84-43-7
X
84-43 -3
42
19.0
35.7
45.2
84-43- 18
X
84-43-3
2


100

94
A-9. Qgikgia X bykda Hort.. Vase life data for half-sib crosses.
Cross
Number of Progeny
Mean Vase Life
84-37- 12 x 84-43-3
12
11.9
84-41 19 x 84-27-2
1
13.0
84-43 3 x 84- 41 -8
16
10.0
A-10. Gerbera X hvbrida Hort.. Frequency of senescence modes for half-sib crosses.
Cross
Number of
Flowers
%
Bending
%
Folding
%
Wilting
84-37 12 x 84-43 -3
23
13.0
34.8
52.2
84-41 19 x 84-27-2
1


100
84-43- 3 x 84-41 -8
29
24.1
58.6
17.2

APPENDIX B
STEM STRENGTH STUDIES

B-l, 7x7 Diallel. Stem strength means based on progeny means of gerbera crosses.
83-1-10
83- 1 -31
83-1-96
83-4-8
83-5-76
83-7-48
83-8-7
83-1-10
x = 28.9
n = 16
x = 29.4
n = 16
x = 19.5
n = 16
x = 32.5
n = 16
x = 29.6
n = 16
x = 20.0
n = 16
83-1-31
x = 31.2
n = 16
x = 22.0
n = 11
x = 13.3
n = 16
x = 30.3
n = 16
x = 21.3
n = 16
x = 25.3
n = 16
83-1-96
x = 25.6
n= 16
x = 24.9
n = 14
x= 12.7
n= 15
x = 31.5
n = 14
x = 27.5
n = 16
x = 19.7
n= 15
83-4-8
x = 16.1
n = 16
x = 15.1
n = 15
x = 17.3
n = 15
x = 16.6
n = 16
x= 14.3
n= 16
x = 8.7
n = 16
83-5-76
x = 33.9
n = 16
x = 20.4
n = 16
x = 29.5
n= 13
x = 16.4
n= 15
x = 24.1
n= 15
x = 25.1
n= 13
83-7-48
x = 31.3
n = 16
x = 31.5
n = 16
x = 33.4
n= 13
x = 14.4
n = 16
x = 25.8
n= 15
x = 23.5=
n = 16
83-8-7
x = 20.3
n= 15
x = 24.4
n = 16
x = 24.8
n = 15
x = 12.9
n = 14
x = 23.1
n = 13
x = 24.6
n= 15

APPENDIX C
VASE LIFE X STEM STRENGTH STUDIES

C-LPheTOt^£c_COTelation_c^fficients_tetween_stem^nd_innorescenceffaitsJn_£ertera.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
.05
.06
.08
Stem Diameter @ 12 cm
.76**
.87 **
Disc Diameter
.80 **
* Significant at P< 0.01.
C-2. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera
by vase life (VL) ratings.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
VL = high
-.09
-.05
-.04
VL = low
Stem Diameter @ 12 cm
.22*
.19
.22* *
VL = high
.78 **
.88 **
VL = low
Disc Diameter
.78 **
.89 **
VL = high
.12**
£
_o
n
>
.86**
* Significant at P< 0.05. *
* Significant at P< 0.01.
C-3. Phenotypic correlation coefficients between stem and inflorescence traits in gerbera
by stem strength (SS) ratings.
Traits
Stem Diameter
@ 12 cm
Disc
Diameter
Inflorescence
Weight
Stem Length
SS = high
.10
.12
.16*
SS = low
Stem Diameter @ 12 cm
.11
.20 *
.28 **
SS = high
.70**
.87 **
SS = low
Disc Diameter
.84 *
.89 **
SS = high
.80 **
SS = low
.78 **
* Significant at P< 0.05. *
* Significant at P< 0.01.
98

99
Stem Anatomy
Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN
-.01
.07
-.003
SBUN
.02
.22**
.28**
TBUN
.01
.24**
.27 *
LBUNCRAT
-.05
-.47 *
-.60**
SBUNCRAT
.002
-.33 *
-.34 *
TBUNCRAT
-.02
-.48 *
-.55 *
LBUNARAT
-.05
-.60**
-.72 *
SBUNARAT
-.01
-.62 *
-.68 *
TBUNARAT
-.03
-.67 *
-.76 *
Significant at P< 0.01.

100
C-5. Phenotypic correlation coefficients between stem anatomy and three traits in gerbera
by vase life (VL) ratings.
Stem Anatomy Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN VL = high
-.04
.16*
.16*
VL = low
.04
.09
-.01
SBUN VL = high
-.07
.09
.18*
VL = low
.12
.33 *
.35**
TBUN VL = high
-.08
.14
.23*
VL = low
.13
.36
.34**
LBUNCRAT VL = high
.02
-.39 *
-.45 *
VL = low
-.12
-.53 *
-.67 *
SBUNCRAT VL = high
-.01
-.44**
-.41 *
VL = low
.00
-.20 *
-.24 *
TBUNCRAT VL = high
.01
-.54 *
-.53 *
VL = low
-.06
-.42 *
-.53**
LBUNARAT VL = high
.07
-.63 *
-.70 *
VL = low
-.15*
-.65 *
-.77 *
SBUNARAT VL = high
.05
o
*
#
-.71 *
VL = low
-.08
-.57 *
-.63 *
TBUNARAT VL = high
.06
-.75 *
-.78 *
VL = low
0.13
-.67 *
-.77 *
* Significant at P< 0.05. Significant at P< 0.01.

101
C-6. Phenotypic correlation coefficients between stem anatomy and three traits in gerbera
by stem strength (SS) ratings.
Stem Anatomy Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN
SS = high
-.11
.11
-.01
SS =low
.15
.07
.12
SBUN
SS = high
.02
.34* *
.42* *
SS = lw
-.09
.13
.17
TBUN
SS = high
-.02
.38 *
.40* *
SS = lw
.03
.15
.20*
LBUNCRAT
SS = high
.02
-.39 *
-.45 *
SS =low
-.12
-.53 *
-.67 *
SBUNCRAT
SS = high
-.01
-.44 *
-.41 *
SS =low
.00
-.20 *
-.24 *
TBUNCRAT
SS = high
.01
-.54 *
-.53 *
SS = low
-.06
-.42 *
-.53 *
LBUNARAT
SS = high
.07
-.63 *
-.70 *
SS =low
-.15 *
-.65 *
* 77 *
SBUNARAT
SS = high
.05
-.70 *
-.71 *
SS =low
-.08
-.57 *
-.63 *
TBUNARAT
VL = high
.06
-.75 *
-.78 *
VL = low
0.13
-.67 *
-.77 *
* Significant at P< 0.05. *
* Significant at P< 0.01.

APPENDIX D
HERITABELITY STUDIES FOR OTHER TRAITS IN GERBERA

D-L5x5D^kLO^er^ait_me^Sibas^_on_£ro|en^means_ofjer^acro^si
83-1-77
83-4-69
83-5-109
83-7-4
83-7-10
X! = 56.95
X! = 54.77
X! = 52.67
X! = 52.95
x2 = .45
x2 = .58
x2 = .46
x2 = -49
83-1-77
x3 = 10.18
x3 = 10.59
x3 = 10.16
x3 = 9.81
X4 = 1.42
X4 = 1.85
x4= 1.74
X4 = 1.61
x5= 1.17
x5= 1.94
x5= 1.81
x5= 1.59
n= 13
n = 18
n = 19
n = 13
xj = 54.37
X! = 60.46
X! = 49.48
xj = 55.18
x2 = .43
x2 = .55
x2 = .46
x2 = .45
83-4-69
x3 = 9.86
x3 = 10.04
x3= 9.11
x3 = 8.50
X4 = 1.40
X4 = 1.83
X4 = 1.54
X4 = 1.51
x5 = .84
X5 = 1.30
x5= 1.59
X5 = 1.06
n= 19
n = 16
n = 18
n = 21
x¡ = 50.63
x2 = -62
Q'l S 10Q
1 n 8s
oj j i\jy
1 v OJ
X4 = 1.80
X5 = 1.63
n= 12
xi = 47.60
X! = 50.69
X! = 48.73
X! = 49.32
x2 = .46
x2= .51
x2 = .55
x2 = .45
83-7-4
x3 = 9.90
x3= 9.05
x3= 9.79
x3 = 8.80
x4 = 1.66
x4= 1.73
x4 = 2.00
x4= 1.77
x5= 1.69
x5= 1.71
X5 = 1.65
x5= 1.57
n = 21
n = 9
n = 22
n = 25
x¡ = 45.83
X! = 43.00
xj = 54.96
x2 = .46
x2 = .36
x2 = .54
83-7-10
x3= 9.00
x3= 7.50
x3= 9.33
x4 = 1.53
x4= 1.10
X4 = 1.83
x5 = 2.00
x5 = .50
x5= 1.77
n = 3
n = 1
n = 18
xj = stem length
x2 = stem diameter
x3 = inflorescence diameter
X4 = disc diameter
X5 = stem elongation
n = number of progeny
103

104
D-^J^SDi^eLP^ents^d^roien^^eansto^veffmts^injertera^
N
Stem Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
(cm)
(cm)
(cm)
(cm)
(cm)
Parents
5
' 49.5
.61
9.1
1.9
2.8
Progeny
248
53.0
.50
9.7
1.7
1.5

105
D-3. 5x5 Diallel. General combining ability analysis of variance for five traits in
Source of Variation
df
Stem
Length
M.S.
Stem
Diameter
M.S.
Inflorescence
Diameter
M.S.
Disc
Diameter
M.S.
Vgrowth
M.S.
General combining ability
4
512.84 *
.1630 *
25.81 *
1.670**
3.99
n.s.
Specific combining ability
5
93.74 n-s-
.0121 *
0.46 n s-
0.042 ns-
1.18
n.s.
Error
238
67.40
.0035
0.92
0.040
0.84
* Significant at P< 0.05. Significant at P< 0.01.
D-4. 5x5 Diallel. Analysis of variance for five traits in gerbera.
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
Source of Variation
df
M.S.
M.S.
M.S.
M.S.
M.S.
Among crosses
9
495.53 *
.1529 **
21.50**
1.640 *
6.70 *
Among plants
238
115.31 **
.0061 *
1.67 *
0.073 *
1.35 **
Within plants
239
36.96
.0035
0.46
0.029
0.72
* Significant at P< 0.01.

106
D-5. 5x5 Diallel. Summa
ry of variances for five traits in gerbera. n = 1.96
Variances
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
vA
22.76
.0084
1.38
.089
0.15
vG
27.20
.0100
1.30
.089
0.21
VE
67.40
.0035
0.92
.040
0.84
VHg
48.54
.0017
0.69
.025
0.47
VEs
36.96
.0035
0.46
.029
0.72
VP(n)
94.60
.0135
2.20
.129
1.05
D-6. 5x5 Diallel. Estimates of heritability
and repeatability for five traits in
gerbera.
Stem
Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
heritability
h2
.24
.62
.62
.69
.14
H2
.29
.74
.59
.69
.20
repeatability
r
.74
.77
.81
.80
.49

D^T^T^x^T^DialleL^Other^aitmeans^asedon^ro^en^meansofjerberacrosses.
83-1-10
83-1 -31
83-1-96
xj = 57.33
X! = 55.54
83-1-10
x2 = .48
x2 = .47
x3= 4.23
x3= 4.94
n = 16
n= 16
xj = 56.84
xj = 48.48
83-1 -31
x2 = .47
x2 = .46
x3= 4.28
x3= 3.63
n = 16
n = 11
X! =52.10
xj = 46.34
83-1-96
x2 = .46
x2 = .47
x3= 4.21
x3= 4.01
n = 16
n = 14
X] = 53.44
xj = 52.81
xt = 46.78
83-4-8
x2 = .47
x2 = .46
x2 = .45
x3= 3.47
x3= 3.05
x3= 3.18
n= 16
n = 15
n= 15
X! = 55.03
X! = 49.05
xj = 40.97
83-5-76
x2 = .52
x2 = .53
x2 = .49
x3= 5.13
x3= 5.01
x3= 4.32
n = 16
n = 16
n = 13
8.45
X! = 52.95
xj =53.14
xj = 50.28
83-7-48
x2 = .50
x2 = .50
x2 = .50
x3= 4.90
x3= 4.24
x3= 4.66
n = 16
n = 16
n = 13
83-4-8
83-5-76
83-7-48
83-8-7
xj = 52.62
x2 = .43
x3= 3.24
n= 16
X! = 58.23
x2 = .54
x3= 5.53
n = 16
x, = 59.03
x2 = .51
x3= 5.13
n = 16
xj = 67.61
x2 = .50
x3= 4.47
n= 16
X! = 48.63
x2 = .44
x3= 2.83
n = 16
X! = 47.39
x2 = .52
x3= 4.96
n = 16
X! = 54.90
x2 = .50
x3= 4.36
n = 16
X! = 61.38
x2 = .52
i3= 4.32
n = 16
xj = 45.93
x2 = .46
x3= 3.40
n = 15
X! = 42.54
x2 = .49
x3= 4.63
n= 14
X! = 46.44
x2 = .49
x3= 4.59
n= 16
X! = 57.93
x2 = .53
x3= 4.99
n= 15
X] = 45.41
x2 = .49
x3= 3.70
n = 16
X] = 45.66
x2 = -45
x3= 3.23
n = 16
X! = 57.29
x2 = .52
x3= 3.78
n = 16
X! = 43.20
x2 = .48
x3= 3.61
n = 15
xj = 45.40
x2 = .55
x3= 4.76
n= 15
x, = 49.49
x2 = .57
x3= 5.22
n = 13
xj = 48.89
x2 = .48
x3= 3.56
n = 16
X! =51.64
x2 = .55
x3= 4.70
n = 15
X! = 57.83
x2 = .56
x3= 5.31
n = 16

D-7.-continued.
83- 1 10 83 1 -31 83 1 -96
83-8-7
x, = 68.69
X2 = .51
x3= 4.45
n = 15
xj = 60.98
x2 = .51
x3= 4.24
n = 16
x, = 52.57
x2= -51
x3= 4.52
n = 15
X] = stem length
x2 = stem diameter
x3 = inflorescence weight
n = number of progeny
83- 4- 8 83 5 -76 83-7 -48
xj = 55.64
x2 = .47
x3= 3.37
n = 14
xj =49.13
x2 = .59
x3= 5.68
n= 13
xj = 56.63
x2 = .57
x3= 5.07
n = 15
83-8-7
o
OO

109
D-8. 7x7 Diallel. Parents and progeny means for three traits in gerbera.
N
Stem Length
Stem Diameter
Inflorescence Weight
(cm)
(cm)
(s)
Parents
7
46.8
.57
3.6
Progeny
642
52.8
.50
4.3

110
D-9. 7x7 Diallel. General combining ability analysis of variance for three traits in
Source of Variation
df
Stem Length
M.S.
Stem Diameter
M.S.
Inflorescence Weight
M.S.
General combining
ability
6
3600.57 *
.1316 *
50.60 *
Specific combining
ability
14
148.82 *
.0032 n s-
2.08 *
Error
621
45.09
.0020
0.90
* Significant at P< 0.01.
D-10. 7x7 Diallel. Analysis of variance for three traits in gerbera.
Source of Variation
df
Stem Length
M.S.
Stem Diameter
M.S.
Inflorescence Weight
M.S.
Among crosses
20
3169.60 *
.0998 *
43.49 *
Among plants
621
109.69 *
.0053 *
2.31 *
Within plants
1069
25.71
.0010
0.46
* Significant at P< 0.01.

Ill
^D-H^T^xTKdle^Summ^ofvm^cesfor^ree^^^mjer^^n^^^
Variances
Stem Length
Stem Diameter
Inflorescence Weight
vA
90.36
.0034
1.27
vG
103.96
.0035
1.42
VE
45.09
.0020
0.90
VHg
35.43
.0016
0.73
Vrs
. 25.71
.0010
0.46
VP(n)
149.05
.0055
2.32
D-12. 7x7 Diallel. Estimates of heritability and repeatability for three traits in gerbera.
Stem Length
Stem Diameter
Inflorescence Weight
heritability
h2
.61
.62
.55
H2
.70
.64
.61
repeatability
r
.84
.84
.82

112
D-13. Summary of published heritability estimates for stem and inflorescence traits in
Source
Stem Length
h2 H2
Stem Diameter
h2 H2
Inflorescence Diameter
h2 H2
Disc Diameter
h2 H2
Maurer
1968
.30



.58

Borghi & Baldi
1970
.35



.30


Schiva
1973


.34
.57
.13
.68
Ottaviano et al.
1974

.57

.62

.24
.58
Schiva
1975
.06
.47
.50
.66



De Leo & Ottaviano
1978
.25

.61

.13

.70
Mutsenietse
1978


.35
.45
.45
.48
.52
.51
Tesi
1978

.34



.74

Wricke
1982

.77



.79
.69
Drennan et al.
1986


.36

.08

.23
D-14. Summary of published repeatability estimates for stem and inflorescence traits in
Source
Stem Length
r
Stem Diameter
r
Inflorescence Diameter
r
Disc Diameter
r
De Leo & Ottaviano
1978
.69
.79
.88
.84
Drennan et al.
1986

.61
.69
.54

APPENDIX E
CORRELATIONS AMONG OTHER TRAITS IN GERBERA

E-l. Summary of published phenotypic correlation coefficients between stem and
inflorescence traits in gerbera.
Traits
Stem Diameter
Inflorescence Diameter
Disc Diameter
Stem Length
Borghi & Baldi
1970

.13

Ottaviano et al.
1974
-.25 .50 *
n.s.
-.25 .50 *
Schiva
1975
n.s.
.25 .50 *
n.s.
Cocozza et al.t
1978
n.s.
.34
n.s.
De Leo & Ottaviano
1978
n.s.
-.20 .40
n.s.
La Malfa & Noto
1978
.89**


Stem Diameter
Ottaviano et al.
1974
n.s.
.75 1.00**
Schiva
1975
.50 .75 *
.50 .75 *
Cocozza et al.'
1978
.29
.37
De Leo & Ottaviano
1978
n.s.
>.80
Inflor. Diameter
Ottaviano et al.
1974
n.s.
Schiva
1975
.50 .75 *
Cocozza et alJ
1978
.47
De Leo & Ottaviano
1978
n.s.
* Significant at P< 0.05. Significant at P< 0.01.
f Level of significance was not reported.
114

BIOGRAPHICAL SKETCH
Heidi Carol Wemett, was bom August 17, 1958, in Los Angeles, California. Her
enthusiasm for ornamental horticulture began during childhood. Since her father was an
avid gardener, family outings usually meant a trip to Lotusland in Santa Barbara or
Descanso Gardens in La Caffada, where friends and relatives resided.
She attended St. Johns College, known for its Great Books Program of western
civilization from 1976-1979. There, she received recognition for a prize essay in
mathematics. In May, 1979, she was granted her B.A. degree in liberal arts from the Santa
Fe, New Mexico, campus.
Two weeks later, she enrolled in the horticulture department at Pennsylvania State
University to begin graduate studies in plant breeding and genetics. After conducting
research for three years on the inheritance of orange flower color in Pelargonium x
hortorum L.H. Bailey she obtained her M.S. degree in November, 1982.
In January, 1983, Miss Wemett moved south to attend the University of Florida at
Gainesville where she conducted research on breeding and genetics of postharvest
longevity in cutflowers of Gerbera X hvbrida Hort. to fulfill part of the requirements for
the Doctor of Philosophy degree. Following the successful completion of her candidacy
exam and formal admission to the doctoral program in 1985, she interrupted her studies to
spend a year in Japan. She wishes to continue travelling after graduation.
115

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Thomas J. Sheehan^ Chair
Professor of Horticultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
'Cochair
forticultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
0
"rancis J. Mari
Professor of Hi
¡cultural Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
David A. Knauft 1/
Associate Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.
Paul M. Lyrene (j
Professor of Horticultural Science

This dissertation was submitted to the Graduate Faculty of the College of Agriculture
and to the Graduate School and was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May, 1990
Dean,
:ufmre
Allege of Agricu
Dean, Graduate School



CHAPTER 6
SUMMARY
Breeding cultivare with superior postharvest longevity may be the best approach for
satisfying the floral consumers expectation of cut flower postharvest quality. Combining
this effort with postharvest treatments, such as the use of floral preservatives, would help
maintain the popularity of gerbera for years to come.
The heritability of two characteristics, vase life and stem strength, considered to be
components of postharvest longevity, was determined. Although results indicate that
intense selection of plants for these traits will yield increases in postharvest longevity, it is
of utmost importance to determine appropriate breeding objectives which will best improve
postharvest performance.
In the case of gerbera, a simple analysis of vase life appears inadequate. Breeding
programs which incorporate selection against frequency of bending and folding and
selection for increased days to wilting may be more effective. Further research is needed to
prove that undesireable modes of senescence, i.e. bending and folding, are qualitatively
inherited by only several genes.
An attempt to document the role of stem strength in postharvest longevity of gerbera
was made; however, the function of this component cannot be specified at this time.
Difficulties associated with measuring stem strength were discussed. As factors are
identified which affect stem strength, breeding programs to improve postharvest
performance can be refined.
While vase life and stem strength have been postulated as components of postharvest
longevity, their effects appear independent of each other. Correlations demonstrated that
77


7
prolonged if a conditioning solution containing AgNOj, 8-HQC, and sucrose is applied for
20 hours before cold storage. A pretreatment application of STS was reported to extend the
vase life of gerbera cut flowers later exposured to high concentrations of exogenous
ethylene. Also, application of an ethoxy analog of rhizobitoxine (AVG) was reponed to
reduce endogenous ethylene production resulting in a slight increase in gerbera vase life,
especially if used together with a flower preservative (Nowak and Plich, 1981). Abdel-
Kader and Rogers, (1986) concluded that 8-HQS was preferable to AgNC>3 as a
component of flower preservatives used to enhance vase life of gerbera. Combining one of
these anti-microbial agents with sucrose was shown to decrease the incidence of stem
break. Amariutei et al. (1986) found pulsing treatments of sucrose, 8-HQS, and AgN3
increased vase life of gerbera Symphonie and Richard.
In addition to chemical treatments, other postharvest handling measures have been
investigated as a means to increase lasting quality in cut flowers. Carpenter and
Rasmussen (1974) proposed fewer leaves remaining on roses could reduce the percentage
of flowers developing bent neck. Dilley and Carpenter (1975) emphasized chemicals could
not completely overcome premature senescence in cut flowers. Ferreira and Swardt
(1981a) reported defoliated cut flowers of the rose Sonia lasted longer in deionized water
than in flower preservative. Woltering (1986) concluded defoliated cut roses would be less
likely to develop bent neck if placed in a vase solution containing a flower preservative or
bactericide. Barendse (1986) reponed optimum vase life in gypsophila could be achieved if
the lower 10 cm. of the stems were defoliated. Meeteren (1978a) suggested the vase life of
gerbera could be extended if the solid basal portion of the stem was removed, permitting
water to enter the cavity of the stem. Later Nowak and Plich (1981) confirmed vase life of
gerbera cut flowers could be enhanced if the basal portion of the stem was removed.
Metabolic causes for senescence in cut flowers have been discussed voluminously
during the past 40 years. Siegelman (1952) asserted that since a decrease in respiration rate
reduces postharvest longevity in some fruits and vegetables, a similar decrease in


28
in the diallel cross, from the error MS or o2e obtained by the combining ability analysis of
variance.
Repeatability for vase life was determined from the following ratio of variances
whereby n = 1:
, VG + vEg
Vp(n)
(Falconer, 1960)
Results
Selection and Marine
Vase life means for 31 parents selected for the top-cross mating with Appleblossom
ranged from 11.3 to 16.0 days. Coefficients of variation ranged from 0.0 to 24.7.
Progeny means resulting from crosses ranged from 6.1 to 10.7 days. Five of these parents
considered to have the best vase life were selected for diallel mating based on data given
in table 3-4. Progeny means resulting from diallel mating ranged from 8.7 to 14.3 days.
Listing of this data is given in table 3-5. Following selection and breeding, the population
mean for vase life increased by three days. Comparison of vase life data for the parental
and diallel generations is shown in table 3-6.
Table 3-4. Vase life data for five parents.
Parent
(Parental Generation)
Mean
(Top-cross Generation)
Progeny Mean
83-1-77
15.3
8.6
83-4-69
12.0
9.6
83-5-109
13.0
8.9
83-7-4
12.3
10.2
83-7-10
16.0
10.7


88
A-2. 5x5 Diallel. Vase life means based on progeny means of gerbera crosses.
83-1-77
83-4-69
83-5-109
83-7-4
83-7-10
83-1-77
x = 9.2
x = 9.3
x = 12.5
x = 9.8
n = 13
n= 18
n = 19
n = 13
83-4-69
83-5-109
x= 11.0
n = 19
x = 8.6
n = 12
x = 8.5
n = 16
x= 12.2
n = 18
x = 11.7
n = 21
83-7-4
x = 11.8
n = 21
x = 11.1
n = 9
x = 11.0
n = 22
x= 13.4
n = 25
83-7-10
x= 15.5
n = 3
x = 14.0
a = 1
x = 12.5
n = 18


32
selection and mating, vase life means for days to bending, folding, and wilting increased.
The largest increase in mean vase life occurred in flowers that folded; 3.5 days. Smaller
increases were observed in flowers that bent and wilted; .4 and 1.2 days, respectively.
These results are shown in figures 3-2, 3-3, and 3-4.
Vase Life (Days)
Figure 3-1. Distribution of vase life data on flowers from the parental and diallel
generations.


66
measured. All bundles greater or equal to that size were described large, all others;
small. This method was described by F. J. Marousky, personal communication, 1985.
Circumference and area of each cross section were calculated using the diameter of each
stem, measured using a vernier caliper, at 12 cm below the peduncle.
A set of variables was created which described a series of mathematical relationships
determined by direct and indirect measurements. A list of these variables is given in table
5-2.
Table 5-2. List of variables from anatomy evaluation of gerbera.
LBUN = # of large bundles
SB UN = # of small bundles
TBUN = # of large bundles + # of small bundles
LBUNARAT = # of large bundles per 1.0 cm2 stem area @ 12 cm
SBUNARAT = # of small bundles per 1.0 cm2 stem area @ 12 cm
TBUN A RAT = Total # of bundles per 1.0 cm2 stem area @ 12 cm
LBUNCRAT = # of large bundles per 1.0 cm stem circumference @ 12 cm
SBUNCRAT = # of small bundles per 1.0 cm stem circumference @ 12 cm
TBUNCRAT = Total # of bundles per 1.0 cm stem circumference @ 12 cm
Results
Correlation Between Vase Life and Stem Strength
A significant positive correlation (r = .28) was observed when vase life means were
plotted against stem strength means from 73 plants belonging to the parental generation
described in Chapters 3 and 4. The relationship between these two components of
postharvest longevity appeared moderately weak, despite significance. Except for one
plant, plants with the weakest mean stem strength had a mean vase life of less than five
days. Distribution of these means is shown in figure 5-1.


73
twenty two plants. Significant negative relationships were determined between vase life
and two stem anatomy variables which were based on total number of vascular bundles
(TBUNARAT, and TBUNCRAT). Significant relationships between vase life and three
stem anatomy variables which were based on number of large vascular bundles (LBUN,
LBUNARAT, and LBUNCRAT) were also negative. Highly significant at the 1 % level,
the strongest correlation was between vase life and number of large bundles per unit of
circumference. Correlations relating to stem anatomy variables based on small vascular
bundles and vase life or stem strength were not significant Summary of these results are
given in tables 5-11, 5-12, and 5-13.
Table 5-11. Correlation coefficients between stem anatomy (number of vascular bundles)
and^os^arvesdongevit^^^^___________
TBUN
TBUNARAT
TBUNCRAT
Vase Life
(days)
.06
-.46*
-.44*
Stem Strength
(F=g/.2 cm)
-.27
-.33
-.39
* Significant at P< 0.05.
Table 5-12. Correlation coefficients between stem anatomy (number of large vascular
bundles) and postharvest longevity.
LBUN
LBUNARAT
LBUNCRAT
Vase Life
(days)
-.51*
-.53*
-.67 *
Stem Strength
(F=g/.2 cm)
-.21
-.21
-.27
* Significant at P< 0.05. Significant at P< 0.01.
Table 5-13. Correlation coefficients between stem anatomy (number of small vascular
bundles) and postharvest longevity.
SBUN
SBUNARAT
SBUNCRAT
Vase Life
(days)
.37
-.34
-.23
Stem Strength
(F=g/.2 cm)
-.08
-.40
-.40


54
approximately equal to 1.0. The relationship among these measurements for each flower
was generally non-linear. Stem strength means for five intervals are shown in figure 4-2.
grams
Figure 4-2. Stem strength means at five deformation intervals.
Measurements made beyond .2 cm were often observed to be a function of the damage
incurred to the stem segment as a result of force being applied. For example, some
segments folded during the measurement process, hence impairing the contact between the
segment and apparatus. Measurements taken at .2 cm were considered, however, to be a
reasonable estimate of stem strength.
Seven parents considered to have the best stem strength were selected for diallel
mating based on data given in table 4-3. Stem strength means for these plants ranged from
14.2 to 31.5 g/.2 cm and coefficients of variation ranged from 5.6 to 26.8. Progeny means
resulting from diallel mating ranged from 10.3 to 33.7 g/.2 cm. Listing of this data is
given in table 4-4. Following selection and breeding, the population mean for stem
strength more than doubled. Comparison of stem strength data for the parental and diallel
generations is shown in table 4-5.


24
variation. Twenty-eight plants per cross were grown to produce the top-cross generation.
Plants which did not produce at least one flower during a flowering period prior to vase life
evaluation (March 12May 15,1985) were discarded. One to three flowers per plant
were evaluated. After the evaluation period, plants that produced less than three flowers
were also discarded. Plant means were determined from data collected on three flowers per
plant Progeny means were determined for each cross from individual plant means.
Finally, five plants (approx. 1.5% of the residual parental generation) with highest progeny
mean vase life were selected. To maintain genetic diversity, selection included plants from
four seed populations.
A 5 x 5 diallel cross mating scheme was utilized to estimate heritability of vase life.
Twenty-eight plants per cross were grown to produce the diallel generation. Plants which
did not produce at least one flower during a flowering period prior to vase life evaluation
(May 23August 6, 1987) were discarded. Plant means were determined from data
collected on one to three flowers per plant.
Vase Life Evaluation
Flowers with 1-2 rows of disc florets open were harvested each evening for six
weeks. Stems were then uniformly cut 30 cm long. Flowers were randomly placed into
sterilized glass bottles with one flower per bottle. Each bottle contained 100 ml. of
deionized water, buffered to pH = 3.0-3.4 with a citrate-phosphate buffer. The depth of
water in each bottle was 4 cm. Every other day, the bottles containing deionized water
were replaced until senescence occurred. Evaluations were conducted in a temperature
controlled room (20-21 C) with 24 hrs./day lighting provided by overhead fluorescent
lamps. Light intensity was ,26-.52 W/cm2 at flower height. Relative humidity was
approximately 70%. These conditions were patterned after experiments conducted by De
Jong (1978a and 1978b) and Harding (1981).


104
D-^J^SDi^eLP^ents^d^roien^^eansto^veffmts^injertera^
N
Stem Length
Stem
Diameter
Inflorescence
Diameter
Disc
Diameter
Vgrowth
(cm)
(cm)
(cm)
(cm)
(cm)
Parents
5
' 49.5
.61
9.1
1.9
2.8
Progeny
248
53.0
.50
9.7
1.7
1.5


50
cm, .8 cm, and 1.0 cm Later, flowers from the diallel cross generation were measured
only for deformation at .2 cm.
Figure 4-1. Portion of stem used for stem strength evaluation.
Table 4-1. Instron specifications
Instron:
Model TM
Load Cell Compression
100 grams
Crosshead Speed
.2 inches/minute
Chart Speed
2 inches/minute
Deformation
Chart Distance
.2 cm
.79 inches
.4 cm
1.57 inches
.6 cm
2.36 inches
.8 cm
3.15 inches
1.0 cm
3.93 inches


ACKNOWLEDGMENTS
I wish to thank the members of my committee for their professional guidance in
helping me to conduct this research. Dr. T. J. Sheehan, Dr. G.J. Wilfret, Dr. F.J.
Marousky, Dr. P. M. Lyrene, and Dr. D.A. Knauft. Their patience is most appreciated.
Other debts of gratitude are acknowledged to Dr. R. C. Fluck from the Dept, of
Agricultural Engineering for his help with stem strength measurements, Dr. C.J. Wilcox
from the Dept, of Dairy Science and Dr. T. L. White and his research assistant, Greg
Powell, from the Dept, of Forestry for their critical help with data analysis, and Dr. F.G.
Martin from IFAS Statistics whose suggestions were invaluable.
Personal thanks is also deeply extended to the Acuff family, Mark and Shari Wilson,
and Terry J. Smith. Without their continued support, encouragement, sweat and toil, I
could still be in the greenhouse; watering, transplanting, and taking measurements. The
generous hospitality of my friends shall never be forgotten. In addition, a special "thank
you" is required to show my appreciation to Terry for his kindness and assistance during
the lengthy period of manuscript preparation.
There are many other individuals who advanced the completion of this research over
the past seven years. I am sincerely grateful to all of them for their help.


APPENDIX C
VASE LIFE X STEM STRENGTH STUDIES


99
Stem Anatomy
Traits
Stem Length
Disc Diameter
Inflorescence Weight
LBUN
-.01
.07
-.003
SBUN
.02
.22**
.28**
TBUN
.01
.24**
.27 *
LBUNCRAT
-.05
-.47 *
-.60**
SBUNCRAT
.002
-.33 *
-.34 *
TBUNCRAT
-.02
-.48 *
-.55 *
LBUNARAT
-.05
-.60**
-.72 *
SBUNARAT
-.01
-.62 *
-.68 *
TBUNARAT
-.03
-.67 *
-.76 *
Significant at P< 0.01.


52
diallel analysis program (Schaffer and Usanis, 1969), was performed on pooled data of
plant means.
Narrow sense heritability (h2) and broad sense heritability (H2) for vase life was
estimated from ratios of the following variances:
VA = Additive genetic variance
VG = Total genotypic variance (additive + non-additive)
Vp = Total phenotypic variance (genotypic + environmental)
(Falconer, 1960)
Genotypic and phenotypic variances were determined from the following equations using
the variance components for general combining ability (a2gca), specific combining ability
(o2sca). and error (o2e) which were calculated by the diallel analysis program developed by
Schaffer and Usanis (1969):
Va = 42gca
VG = 4a2gca + 4g2
sea
VP = 42gca + 4o2sca + CJ2e
(Hallauer, 1981)
Thus, heritability (h2 and H2) was estimated from the formulae:
4o2
h2 = T^Lgca
4a2gca + 4a2
sea +
4a2gca + 4o2sca
4a2gca + 4o2
sea + 2e
Predicted estimates of narrow sense heritability and broad sense heritability for n
measurements per plant were also made. This required partitioning the environmental
variance (VE), a component of phenotypic variance (Vp), into general environmental


49
325 plants that had been previously established for vase life studies (Chapter 3). Plants
selected had already produced at least three flowers that had been evaluated for vase life.
One to six flowers per plant were evaluated. After the evaluation period plants that
produced less than three flowers were discarded. This residual parental generation
included 73 plants. Plant means were determined from data collected on the first three
flowers evaluated per plant. Seven plants (approx. 10% of the total number of plants
evaluated for stem strength) with highest mean stem strength (x £ 14.00) and lowest
coefficient of variation (C.V. ^ 27.00), were selected. To maintain genetic diversity,
selection included plants from five seed populations.
A 7 x 7 diallel cross mating scheme was utilized to estimate heritability of stem
strength. Thirty plants per cross were grown to produce the diallel generation. Plants
which did not produce at least one flower during a flowering period prior to stem strength
evaluation (June 12July 30, 1985) were discarded. Plant means were determined from
data collected on one to three flowers per plant.
Stem Strength Evaluation
Flowers with 1-2 rows of open disc florets were harvested each evening for six
weeks. After 24 hours dry storage at room temperature (20-21 C), a 15 cm stem
segment from each flower was evaluated for stem strength. The portion of the stem 4.5 -
19.5 cm below the base of the peduncle was the area from which the segment was taken.
This is shown in figure 4-1.
Stem strength was determined by the amount of force required to deflect the midpoint
of each segment a specified distance (F=g/cm). The segments were supported at two
points, 10 cm apart; equidistant to their midpoint. Measurements were made using an
Instron.1 Instron specifications are listed in table 4-1. Initially, a maximum deformation of
1.0 cm was specified with measurements taken at five equal intervals: .2 cm, .4 cm, .6
'Provided by the Dept of Agricultural Engineering, Univ. of Florida, Gainesville, Florida, 33610.


61
segments. Frequently, but not always, stem segments with small diameters dehydrated
more than segments with larger diameters. These segments became brittle and resulted in
stem strength greater than 50g/.2 cm, much higher than most stem strength values
recorded.
Despite efforts to obtain an accurate evaluation of stem strength for gerbera flowers,
the coefficient of variation for stem strength in our experiment exceeded fifty percent before
and after selection and mating to increase stem strength. This demonstrates that
environmental conditions greatly influence this character.
Broad sense heritability for stem strength (H2 = .42), based on 2.66 measurements per
plant, appears moderate; therefore improvment for this character can be expected with even
a moderate rate of selection intensity. Narrow sense heritability was also fairly moderate
(h2 = .38), indicating parental phenotypes could be expected to correspond to some degree
to their genotypes.
Genetic variation may be largely controlled by additive gene action since the difference
between broad sense and narrow sense heritability was small. Therefore, in a fixed model
experiment, progeny means obtained from a top-cross mating would be effective in
determining parents with good combining ability for increasing vase life.
As discussed in Chapter 3, environment is a critical variance component of heritability.
Different estimates of heritability would be expected if the magnitude of environmental
conditions varied, assuming effects due to genotype remained constant for a given
population.
Falconer (1960) demonstrated a method to predict estimates of heritability for a
specified number of measurements per experimental unit. This involved partitioning
environmental variance (V^) into general environmental variance (VEg) and special
environmental variance (VEs). Special environmental variance (VEs) or within-plant
variation is the environmental variation for a single observation per experimental unit. The
magnitude for special environmental variance (VEs) is then divided by a specified number


Frequency (%)
33
10.0-
7.5-
DIALLEL GENERATION
mean = 6.4
Figure 3-2. Distribution of vase life data for flowers that bent from parental and diallel
generations.


55
Parent
Mean
Coefficient of
Variation
83-1-10
22.0
26.8
83-1-31
31.5
15.6
83-1-96
18.0
5.6
83-4-8
14.2
25.5
83-5-76
14.3
23.2
83-7-48
24.0
24.2
83-8-7
16.0
23.6
Table 4-4. 7x7 Diallel. Stem strength data from diallel crosses. (F = g/.2cm; reciprocals
pooled)
Cross
83-1-10
X
83-1-31
31.3
83-1-10
X
83-1-96
26.6
83-1-10
X
83-4-8
17.1
83-1-10
X
83-5-76
32.3
83-1-10
X
83-7-48
32.4
83-1-10
X
83-8-7
19.7
83-1-31
X
83-1-96
23.2
83-1-31
X
83-4-8
14.2
83-1-31
X
83-5-76
27.5
83-1-31
X
83-7-48
26.7
83-1-31
X
83-8-7
23.9
83-1-96
X
83-4-8
14.2
83-1-96
X
83-5-76
33.7
83-1-96
X
83-7-48
29.7
83-1-96
X
83-8-7
21.2
83-4-8
X
83-5-76
15.9
83-4-8
X
83-7-48
14.4
83-4-8
X
83-8-7
10.3
83-5-76
X
83-7-48
24.2
83-5-76
X
83-7-48
29.9
83-7-48
X
83-8-7
25.1
(Diallel Generation)


56
Table^S^Sum^^ofstemstnghfor^enta^anddialle^ienerations^F^^^m)
Generation
No. of
Plants
Mean
Std. Dev.
Coefficient of
Variation
Parental
73
10.17
7.27
71.48
Diallel
642
22.52
11.32
50.28
Heritabilitv
Combining ability analysis of variance using plant means was performed on stem
strength data from a 7 x 7 diallel. General combining ability effects were highly
significant. Specific combining ability effects were non-significant. Results are
summarized in table 4-6. Heritability was estimated by the ratio of genetic variance (VAor
Vq) to phenotypic variance (Vp) (Falconer, 1960). Variances were derived using variance
components for general combining ability (o2gca), specific combining ability (o2^), and
error (a2e) according to formulae published by Hallauer (1981). The difference between
narrow sense heritability (h2 = .38) and broad sense heritability (H2 = .42) estimates was
small. This indicates the effect attributable to non-additive genetic variance (Vq VA) is
minimal. Variances and heritability estimates for this population are given in table 4-7.
Table 4-6. 7x7 Diallel. Combining ability analysis of variance for stem strength.
(F = g/-2cm)
Source of Variation
df
M.S.
F-ratio
General combining ability
6
4218.62
16.32**
Specific combining ability
14
258.47
1.63 n s-
Error
621
158.22
* Significant at P< 0.01.
Estimates of heritability were based on an average of 2.66 measurements per plant.
This value (n) resulted by evaluating 1710 flowers from 642 plants. The error variance


63
generally have shorter, thicker stems and heavier inflorescences, has been suggested as a
possible means for increasing stem strength in gerbera.
Conclusion
Improving structural stem strength in gerbera can be realized through breeding efforts.
Despite this encouraging conclusion, the breeder faces difficulty with selecting plants
whose flower stems are structurally strong. Large intraplant variation associated with this
character is a major problem.
This situation highlights several unanswered questions, What determines structural
strength in flower stems? Is it a composite character? If so, what are the individual
components of structural stem strength? How do they contribute to postharvest longevity
in gerbera? It is recommended that further research to identify factors that contribute to
structural stem strength in gerbera be conducted, particularly if accuracy in evaluating stem
strength can be attained.
Another important concern for the breeder involves defining the conditions under
which stem strength should be measured. An investigation of the effects of pre-harvest
environmental conditions versus morphological or anatomical variability of the stem could
yield useful information for developing a method to evaluate the magnitude of stem strength
more accurately.
Garifying breeding objectives and developing appropriate evaluation methods could be
the key to a successful breeding program to improve structural stem strength in gerbera.


CHAPTER 1
GENERAL INTRODUCTION
Gerbera X hvbrida Hort. is a popular cut flower. Developed from hybrids of
interspecific crosses between Gerbera iamesonii Bolus and Gerbera viridifolia Sch., and
possibly other species, it is a showy and handsome flower of the daisy family,
Compositae. The distinctive inflorescence, generally measuring 8-14 cm. in diameter,
has strap-shaped, ray floret ligulae that are bright shades of orange, yellow, or red,
surrounding a yellow or black colored center of disc florets. This eye-catching flower is
supported by a long, leafless, and upright stem, known as a scape.
The major postharvest problem in cut flowers of gerbera relates to the length of time
until senescence occurs; when the ligulae wilt or when the stem no longer remains upright.
Ideally, postharvest longevity in gerbera should be two weeks or longer. Unfortunately for
the consumer, vase life is often less because the stem ceases to remain upright.
A critical attribute of cut flowers is postharvest longevity; therefore, research to
improve the lasting quality of gerbera is important. Postharvest longevity of cut flowers,
including gerbera, can be extended by postharvest treatments but the extent to which these
treatments can improve lasting quality may be limited by the plant genome. At present,
evaluation for postharvest longevity is not routinely done in cut flower breeding programs.
If gerbera cultivars with superior postharvest longevity are to be developed, then it is
necessary to identify characteristics which are suitable for selection and breeding.
The main purpose of this research was to evaluate the potential of plant breeding as a
method to improve postharvest longevity in gerbera, using a broad based source of
germplasm. Specifically, objectives of this study were to estimate heritability for vase life
and stem strength by diallel analysis, to observe changes in frequency of senescence
1


42
Discussion
Heritability estimates for a given character can vary based on the population of plants
evaluated, selection intensity, mating design and environment. (Simmonds, 1979). Despite
this, estimates of heritability for vase life determined from this experiment (h2 = .28 and H2
= .28) were within proximity of those determined from investigations by other researchers
(Serini and De Leo, 1978; Tesi, 1978; and Harding et al., 1981). This reflects some
consistency in the proportion of genetic variance to phenotypic variance for vase life in
gerbera regardless of genetic diversity in populations sampled, breeding procedures, and
environment.
The estimate of 28 percent for broad sense heritability (H2), based on 1.96
measurements per plant, is moderately low. Similarly, the estimate of 28 percent for
narrow sense heritability (h2), also based on 1.96 measurements per plant, is moderately
low. This estimate is between the range of prior estimates: 15 percent (Tesi, 1978); 17
percent (Serini and De Leo, 1978); and 0, 24, and 38 percent (Harding et al., 1981). It
appears that genetic variation may be largely controlled by additive gene action since broad
sense and narrow sense heritability were approximately equal. Therefore, in a fixed model
experiment, progeny means obtained from a top-cross mating would be effective in
determining parents with good combining ability for increasing vase life.
Falconer (1960) demonstrated a method to predict estimates of heritability for a
specified number of measurements per experimental unit. This involved partitioning
environmental variance (V£) into general environmental variance (VEg) and special
environmental variance (V^. Special environmental variance (VEs), or within-plant
variation, is the environmental variation for a single observation per experimental unit. The
magnitude for special environmental variance (VEs) is then divided by a specified number
of measurements per experimental unit (n) as part of the calculation to obtain the phenotypic
variance for each special case. Ideally, if n = then VEs will be reduced to zero, thereby


44
for flowers that folded, bent, or wilted differed. For example, the increase in mean vase
life for flowers that folded was much greater than for flowers that bent or wilted.
Combining this evidence, it is suggested that vase life may be a composite character of at
least three components, represented by each senescence mode.
Further evidence to support this suggestion is found by comparing the magnitude of
the error variance component (o^g) or error MS from an analysis of variance based on the
total number of vase life observations versus the error MS from individual analyses based
on the number of vase life observations for each senescence mode. De Jong and Garretsen
(1985) previously reported that if termination of vase life is not distinguished by stem
collapse or petal wilt in an analysis of variance, a relatively large error variance would
result. In fact, the error MS based on the total number of vase life observations from this
data was nearly double the arithmetic mean of the error MS from individual analyses based
on flowers that bent, folded, and wilted (9.88 vs. 5.50).
This study was designed to evaluate the potential of improving postharvest longevity
in gerbera by selecting plants with high vase life. Plants were not selected based on the
specific number of days to bending, folding, or wilting of their flowers. However; it does
appear, that selecting plants based on the number of days to bending, folding, or wilting,
rather than the composite character of vase life, may prove to be a useful approach to
improving postharvest longevity in gerbera, since general combining ability for vase life of
flowers was determined to be highly significant for all three senescence modes.
The incidence of bending, folding, or wilting, not only the vase life of these
senescence modes, is proposed to be another important aspect of postharvest longevity in
gerbera. Before selection and mating to improve vase life, a greater proportion of flowers
bent than wilted. Flowers that bent generally exhibited lower vase life than flowers that
wilted. After selection and mating, a greater proportion of flowers wilted than bent.
Additionally, the shift in proportion of these two modes that occurred was rather dramatic


31
heritability and broad sense heritability for vase life were then made for 1, 2, 3,5, and
measurements per plant by the ratio of genetic variance (VA or Vq) to phenotypic variance
(Vp(n>) (Falconer, I960). Estimates ranged from 22 to 39 percent. These results are given
in table 3-11. Repeatability (r = .57) for vase life was moderately high.
Genotypic Variance
Environmental Variance
Phenotypic Variance
vA
vG
vE
VHg
Ves
VP(n)
5.08
5.12
13.08
8.04
9.88
18.20
Table^U^SxJ^dle^^^ctedestima^of^ritbilit^tovasehfc^
Number of Measurements
Heritability
1
2
3
5
oo
h2
.22
.28
.31
34
.39
H2
.22
.28
.31
34
.39
Senescence Patterns
A shift in the population mean for vase life was examined more closely. Distribution
of vase life data on individual flowers from the parental and diallel generations was
compared. In addition to a shift in the mean from 7.8 days to 11.2 days, the frequency of
vase life did not appear normally distributed in both generations. Before selection there
was a much higher frequency of flowers with low vase life than high vase life. After
selection and mating, this trend was clearly reversed. These results are shown in figure 3-
1.
Distribution of vase life data by senescence mode was also compared. In both
generations, vase life frequency appeared somewhat normally distributed. Following


30
Tabl^^^^^DiaJleL^^^mini^abili^and^sisofv^anceforva^Jifc^
Source of Variation
df
M.S.
F-ratio
General combining ability
4
106.12
8.03 *
Specific combining ability
5
13.21
1.01 ns-
Error
238
13.08
* Significant at P< 0.05.
Table 3-8. 5x5 Diallel. Variances and heritability estimates for vase life.
vA vG
vP
h2
H2
5.08 5.12
18.20
.28
.28
Estimates of heritability were based on an average of 1.96 measurements per plant.
This value (n) resulted by evaluating 487 flowers from 248 plants. The error variance
component (o2^ for a single measurement per plant was obtained by an analysis of
variance using vase life data from individual flowers of the diallel generation. Also, this
analysis indicated differences in vase life of flowers among crosses and among plants
within crosses were highly significant. Results are summarized in table 3-9.
Table 3-9. 5x5 Diallel. Analysis of variance for vase life.
Source of Variation
df
M.S.
F-ratio
Among crosses
9
115.02
5.17 *
Among plants
238
22.26
2.25 *
Within plants
239
9.88
* Significant at P< 0.01.
General environmental variance (VEg) and special environmental variance (VEs) were
derived from vase life data using calculations described by Falconer (1960). A summary of
variances for this population are given in table 3-10. Predicted estimates of narrow sense


15
discoloration, 50% of tepal discoloration, and perianth drop. All three characteristics were
considered effective determinations of vase life, but perianth drop was easiest to measure.
Also, evidence was reported that discoloration may occur earlier in some genotypes
depending on the pigment composition of their flowers, but that this does not necessarily
lead to earlier perianth drop. It was observed that some flowers with much carotenoid,
delphinidin, or cyanidin and little pelargonidin discolor earlier than flowers with little
carotenoid and much pelargonidin. It was concluded, however, that in a breeding program
to improve postharvest longevity, all characteristics of vase life must be considered.
Eijk et al. (1977) confirmed earlier observations that cut flower vase life in tulips could
be selected by evaluating flowers remaining attached to the plant, except in cases where
response to flower preservatives is being evaluated. They also found vase life of field
grown tulips was less predictable than vase life of greenhouse grown tulips. Therefore, it
was suggested that field trials should be utilized only for initial screening. Several
correlations were reported: (1) an increase in plant height during the growing period was
not correlated to vase life; (2) an increase in growth of the last intemode (stem elongation)
did not appear correlated to vase life; and (3) vase life and water uptake were significantly
correlated.
Eijk and Eikelboom (1980) demonstrated, using combining ability analysis, that
phenotypic observation was effective in predicting the results of crossing a set of parents in
order to improve postharvest longevity in tulips. It was postulated additive gene action
controls three senescence characteristics: start of discoloration, 50% discoloration, and
perianth drop.
Eijk and Eikelboom (1986) investigated the influence of temperature on selecting for
vase life in tulips. Variation in vase life for some genotypes was much greater depending
on the temperature during the evaluation period. They recommended that selection at
17 C. provides an average response for each genotype; however, final selection should
include screening at temperature extremes.


46
Correlations between vase life and other morphological traits were particularly
interesting because a broad based germplasm was examined. The wide variation in
morphological phenotypes permitted an extensive comparison. Correlation coefficients
were generally low. Despite this, they were often significant due to the large number of
flowers evaluated.
Since it has been discussed that eliminating senescence due to bending and folding
among flowers may be possible, correlations between the vase life of flowers that wilted
and other traits are probably the most useful to a breeder. Significant relationships between
longer vase life of flowers that wilted and shorter stem length, smaller stem diameter, and
smaller inflorescence diameter were determined.
Unfortunately, the current floriculture market evidently prefers gerberas with long,
thick stems and large inflorescences, as most commercial cultivars tend to be of this type.
Therefore, it is encouraged that postharvest longevity be deliberately included as an
objective into cut flower gerbera breeding programs or the consumer will have to be
satisfied with only the expected vase life possible from postharvest treatments such as floral
preservatives. If attention to this situation is not given, a decline in the popularity of
gerbera might result as other flower varieties with better, more predictable vase life become
increasingly available.
Conclusion
Breeding to improve vase life in gerbera has potential, despite the fact that heritability
was confirmed to be moderately low. It is concluded that the best approach for establishing
lines with superior lasting quality requires recognizing vase life as a composite character.
Distinguishing quantitatively inherited traits, i.e. days to bending, days to folding, and
days to wilting and qualitatively inherited traits, i.e. bending vs wilting and folding ys
non-folding, may be the key to a successful breeding program.


REFERENCES
Abdel-Kader, H. and Rogers, M.N., 1986. Postharvest treatment of Gerbera iamesonii.
ActaHort., 181: 169-176.
Amariutei, A., Burzo, I. and Alexe, C, 1986. Researches concerning some metabolism
aspects of cut gerbera flowers. Acta Hort., 181: 331-337.
Awad, A.R.E., Meawad, A., Dawh, A.K. and Elsakaa, M., 1986. Cut flower longevity
as affected by chemical pretreatment. ActaHort., 181:177-172.
Ballantyne, D.J., 1963. Respiration of floral tissue of the daffodil (Narcissus
pseudonarcissus Linn.) treated with benzyladenine and auxin. Can. J. Bot., 44: 117-119.
Barigozzi, C. and Quagliotti, L., 1978. Current research on breeding of gerbera. In: L.
Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera.
Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 57-68.
Benschop, M. and De Hertogh, A.A., 1969. An analysis of the post-harvest
characteristics of cut tulips. Flor. Rev., 145: 24-25, 62-65.
Borghi, B. and Baldi, V., 1970. Variabilita tra cloni di gerbera (Gerbera jamesonii) allevati
in diverse condizioni ambientali. Sementi elette, 6: 25-32.
Bravdo, B., Mayak, S. and Gravrieli, Y., 1974. Sucrose and water uptake from
concentrated sucrose solutions by gladiolus shoots and the effect of these treatments on
floret life. Can. J. Bot., 52: 1271-1281.
Bufler, G., 1986. Ethylene biosynthesis and action. ActaHort., 181:93-98.
Burdett, A.N., 1970. The cause of bent neck in cut roses. J. Amer. Soc. Hort. Sci., 95:
427-431.
Buys, C, 1978. Quality problems in gerbera. In: L. Quagliotti and A. Baldi (Editors),
Genetics and Breeding of Carnation and Gerbera. Proceedings of a Eucarpia Meeting, 24-
28 April, at Alassio, Italy, pp. 255-262.
Carpenter, WJ. and Rasmussen, H.P., 1973. Water uptake rates by cut roses (Rosa
hvbrida) in light and dark. J. Amer. Soc. Hort. Sci., 98: 309-313.
Carpenter, W.J. and Rasmussen, H.P., 1974. The role of flower and leaves in cut flower
water uptake. Scientia Hortic., 2: 293-298.
Cocozza Talia, M., Porceddu, E., and Stellacci, P., 1978. Genetic variability in gerbera.
In: L. Quagliotti and A. Baldi (Editors), Genetics and Breeding of Carnation and Gerbera.
Proceedings of a Eucarpia Meeting, 24-28 April, at Alassio, Italy, pp. 163-177.
79


36
Separate analyses of variance were made using vase life data from the diallel
generation for days to bending, days to folding, and days to wilting. Variation within
plants or error M.S. for vase life was less when data were grouped by senscence modes as
when an analysis of variance was made on vase life of all flowers. Differences in vase life
of flowers among crosses were highly significant for all senescence modes. Differences
among plants within crosses were also significant for days to folding and wilting, but non
significant for days to bending. Results are summarized in tables 3-12, 3-13, and 3-14.
Table 3-12. Analysis of variance for vase life^daysjojrending^
Source of Variation
df
M.S.
F-ratio
Among crosses
9
21.72
2.65 *
Among plants
70
8.24
1.36 n s-
Within plants
29
6.05
* Significant at P< 0.01.
Table 3-13. Analysis of variance for vase life (days
to folding).
Source of Variation
df
M.S.
F-ratio
Among crosses
9
52.28
3.76
Among plants
124
14.42
3.16
Within plants
46
4.56
* Significant at P< 0.01.
Table 3-14. Analysis of variance for vase life (days
to wilting).
Source of Variation
df
M.S.
F-ratio
Among crosses
9
28.18
3.21 *
Among plants
120
8.77
1.49 *
Within plants
68
5.88
Significant at P< 0.05. Significant at P< 0.01.