5. _ ... if. ,Izrb -
r: . V22 r n .
. ....r...

_ to :‘Y...

,.

. 3R..." I" .. xx. .

..uarfiu. ....mu‘. ‘ ‘ , .. ,

.5 why. “UTA. 2 . ‘ .. .m‘wmfifl.‘
, _ um... yrs. ‘ , arguawwg 2;

$3 ““5... x ...... . a

«.51. . tin.” $39. , ‘

.:r: a
V. .r... .. _. ..
.

L
.

m

a

<
c l
.. .. ...

. 4.x ”I :2?
F». .32. r .. .
.Su. "Mum.

.

.1 x . ‘
1 , .. ‘ . .

L: . 3 1

35mm...“ pummvva . . .

«wwfiwflfimxu . . x, fir . ‘ . ‘ . .

73) HQ». t V . . . . t J- . 9

...... ...,a. ..

‘ . ....PfiJJL :

. ‘ V m1}; , ,

.

iii-3“---

... .
V “as... ......w.~...k.
.,. 8.3. .1

.

 

r I
is. ..t‘

u.

q i ; ..

:mm fiqauuufswah
fin;
1d;

. . .
.fivwww.
. . V . . . . . . a...
. _ earn.
an? n
. Ian;
t.

v... I
.. ...}...13. V

413.3...

J 7.9.: x

.\

z . . ...
v ......“ h. 3.3 w .
« 1 f . A fit“!

”1%.“? :

3:32.)“
5

.§.

cm ‘-

' 'Jn
.v::

’6

u C. .V ...
wtfikuf;

..Lx... ,
. 5...... . , ‘ J
(.... 1 . . T any? .
, ‘ T ....» P. ...x v
a

‘9‘ J m M
I" 3“. "Y

1 .11.
‘ .Yuifi:

I").
\

A . . . . ‘ , u....W.....L:. human?!.szkggutimmfiiégn:

..B...1F. ‘
In .3).

 

THESIS

’7
BLOC:

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

Environmental Constraints on Marketing, Production, and
Postharvest Shelf Life of Edible Flowers

presented by

Kathleen M. Kelley

has been accepted towards fulfillment
of the requirements for

 

Ph.D. degree“, Horticulture and Botany
and Plant Pathology

 

 

Major professor

Wake/W?

Date [I *2? ‘60

MS U it an Affirmative Action/Equal Opportunity Institution 0. 12771

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m cJCIRCJDatoDquS-ma

ENVIRONMENTAL CONSTRAINTS ON MARKETING,
PRODUCTION AND POSTHARVEST SHELF LIFE
OF EDIBLE FLOWERS

By

Kathleen M. Kelley

A DISSERTATION

Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Departments of Horticulture and
Botany and Plant Pathology

2000

E.“

E‘fz

TCC .' m it

 

on that at

ABSTRACT

ENVIRONMENTAL CONSTRAINTS ON MARKETING, PRODUCTION,
AND POSTHARVEST SI-IELF LIFE OF EDIBLE FLOWERS

By

Kathleen M. Kelley

Experiments were conducted to develop marketing, production, and postharvest storage
recommendations for edible flowers. Chefs and consumer participants rated edible flowers based
on characteristics such as fragrance, taste, and visual characteristics. Chefs were more likely to
rate the attributes and uses of the flowers lower than consumers, with the exception of Tropaeolum
maj us L. ‘Jewel Mix’. Consumer participants also evaluated edible-flower color and color
combinations, container size, and price. A second group rated all characteristics except price.
Flower color was allocated the most points in the purchasing decision (63% for the first group and
95% for the second), with the mixtures of all three colors (blue, yellow, and orange) being the most
desirable. To determine the level of flower quality consumers would accept, two groups of
participants were shown photographic slides of the flowers with visual quality ratings on a scale of
( 1 -5, 5 being flawless). Both groups awarded identical visual quality ratings for all species except
Borago oflicinalis which varied with ratings of 5 to 3 or 5 to 4. Eight species were grown for 1 2
to 18 weeks in a certifiable organic (30% mineral soil) potting medium and fertilized with two
organic fertilizers at 300 or 600 ppm N every two weeks. Growth and nutrient content were

compared to plants grown with the same mineral soil or a soilless medium fertilized with a synthetic

 

 

med 3:

 

W C! \V CC»..-

water soluble fertilizer at 300 ppm Shoot dryweight and tissue P concentrations were generally
higher with a 6 N-2 . 6 P -5 K organic fertilizer (Omega 6-6-6) than with the S N-O.4 P-O. 8 K fish
emulsion fertilizer or the l 9 N- 1 .8 P- 1 9 K complete water soluble fertilizer. Shoot tissue N, K,
Ca, Mg and micronutrients were in the sufliciency range. Root media pH and EC were in the
acceptable range except for the 600 ppm rate of the Omega 6-6-6. The organic fertilizers were
not more acidic than the water soluble. Three of five edible flowers received ratings of 5 when
stored at O to 2.5 °C after two weeks, with B. oflicinalis flowers still marketable, 3 or higher, after
two weeks at -2. 5 °C. Phaseolus coccineus L. flowers were marketable at O to 1 0 °C, after one

week, but unmarketable after 10 d at 0 and 2.5 °C, 9 d at 5 °C, and 7 d at 10 °C.

 

eix er. n1 '

CXCEIIC:
V

the 5:16:

 

 

1
to 5451:.
CSAS I,

“A!
leaching

ACKNOWLEDGMENTS

I would like to thank my committee members for the help and encouragement they have
given me for the last three years. Each has made a contribution to my program to make me a
better student, researcher, and thinker. I would like to thank my major professors, Drs. John A.
Biernbaum and Kenneth L. Poff for the opportunity to be involved in a project that has been
exciting every step of the way and one that I feel has a real purpose to add knowledge not only to
the scientific community, but to help small farmers practice sustainability with a niche crop.

I would like to thank Drs. Bridget K Behe and Richard R. Harwood for introducing me
to subjects that have interested me and that I have developed a deep passion for: Marketing,
CSAs/ local food production, and food security. I would like to thank Dr. Arthur C. Cameron for
teaching how to write a paper for a scientific audience.

To Jerry, my best friend and husband, thank you for all your help, love, and assurance.
My parents Martin and Ellen Boisvert, in-laws Charles and Carol Kelley, and brother and sister-in-
law Prescott and Seon Hee Boisvert, have always given me their support and love and have taken
an interest in my research. My good friends Bridget Behe and Beth Fausey have made me laugh
and helped me enjoy my time as a Ph. D. student. I would also like to thank my friends and fellow
grad students at MSU: Jill Hardy, Elizabeth Moore, Emily Clough, Marcus Duck, Mary-Slade
Morrison, Hyeon-Hye Kim, Meredith Phares, Susan Baldyga, Lindsey Henige, Erin Nausieda, J irn
Heilig, Cathy Whitman, Eric Runkle and others who I have failed to mention.

iv

TABLE OF CONTENTS

Page
List of Tables ............................................................................................................... vii
List of Figures ............................................................................................................... ix
DISSERTATION PREFACE ....................................................................................... 1
SECTION 1. Consumer and professional chef perceptions of three edible-flower
species .................................................................................................. 10
Abstract .................................................................................................................... 11
Introduction ............................................................................................................. 1 2
Materials and Methods ............................................................................................. 13
Results ..................................................................................................................... 16
Conclusion ............................................................................................................... 22
Literature Cited ........................................................................................................ 25

SECTION 11. Consumer preference for edible-flower color, container size and

price ................................................................................................... 34

Abstract ................................................................................................................... 35
Introduction ............................................................................................................. 35
Materials and Methods ............................................................................................. 37
Results ..................................................................................................................... 40
Marketing Implications ...... -. ...................................................................................... 44
Literature Cited ........................................................................................................ 47
SECTION 111. Consumer ratings of edible-flower quality, mix, and color ................... 54
Abstract .................................................................................................................... 55
Introduction .............................................................................................................. 56
Materials and Methods .............................................................................................. 58
Results and Discussion .............................................................................................. 61
Literature Cited ......................................................................................................... 64

SECTION IV. Organic nutrient management of greenhouse production of edible

flowers in containers ........................................................................... 74
Abstract .................................................................................................................... 75
Introduction .............................................................................................................. 76
Materials and Methods .............................................................................................. 80
Results and Discussion .............................................................................................. 83

v9 ,1'!"
14.6me

SEC TIC
Abs-tr:
Intrcél
Maren

Result

 

TABLE OF CONTENTS (CONT’D.)

Literature Cited ......................................................................................................... 90
SECTION V. Postharvest shelf life of five edible flowers ......................................... 103
Abstract ................................................................................................................ 104
Introduction .......................................................................................................... 1 04
Materials and Methods .......................................................................................... 106
Results and Discussion .......................................................................................... 108
Conclusion ............................................................................................................ 110
Literature Cited ..................................................................................................... 112
DISSERTATION CONCLUSION AND RESEARCH SUMMARY ........................ 120
APPENDICES ....................................................................................................... 124

Appendix A. Survey instrument used for an edible flower tasting with Garden
Days participants at Michigan State University, 7 August 1998. Title of
Survey: Consumer perceptions of three edible-flower species. ............................... 125

Appendix B. Survey instrument used for an edible flower tasting with members of

the Michigan Chef de Cuisine Inc. Association at the Detroit Athletic Club in

Detroit, Mich, 8 March 1999. Title of Survey: Professional chef perceptions

of three edible-flower species. ............................................................................... 128

Appendix C. Survey instrument used at Bloomfest at Cobo Hall, Detroit, Mich,
9 and 10 April 1999. Title of Survey: Consumer preference of edible-flower
color, container size, and price. ............................................................................. 131

Appendix D. Survey instrument used at a Master Gardener Conference 29 June
1999 and at Garden Days at Michigan State University, 5 and 6 August 1999.
Title of Survey: Consumer ratings of edible-flower quality, mix, and color. ............ 136

Appendix E. The University Committee on Research Involving Human Subjects
approval letter. Title of research: Consumers Evaluate Edible Flowers and Culinary
Herbs. .................................................................................................................. 139

 

Table

edul

'H
H-
. .r
(I

 

 

IJ

LIST OF TABLES

Table Page

SECTION I

. Consumer and professional chef ratings of attributes of three species of edible
flowers. Percentage of responses with a composite acceptable rating (5, 6, or 7) ..... 26

. Effects of income, marital status and age of respondents, gender, level of

education, and number of people in the household on mean ratings of consumers’
acceptability of viola, borage, and nasturtium based on visual, fragrance, taste,

and usage attributes ................................................................................................. 27

. Effect of gender, number of meals served each week, price of least expensive

entree, years as a chef, and certification of respondents on mean ratings of chefs’
acceptability of viola, borage, and nasturtium based on visual, fragrance, taste,

and usage attributes ................................................................................................. 30

. Mean ratings of chef vs. consumer acceptability of three species of edible flowers
based on visual, fragrance, taste, and usage attributes .............................................. 32

SECTION II

. Conjoint analysis of two groups of participants who viewed and rated either:
1.) flower color, color combination, container size, and price in 27 photographs
or 2.) flower color, color combination, and container size in 14 photographs ............ 49

. Description of three consumer segments derived from cluster analysis based on

participants’ responses to variables, including attitudes about edible flowers and

salad consumption ................................................................................................... 50
SECTION III

. Comparison of demographic characteristics and edible flower preferences for
Master Gardeners and Garde Days participants ....................................................... 66

. Percent of Garden Days (GD) or Master Gardener (MG) participants who rated
edible-flower quality acceptable for five species. Scale of 1-5 (5 = perfect) ............ 67

vii

|»

J

I“

A

tor .
left;

5,“.
4.-

Sign“.

 

 

Spec

PDS‘

Edi:

LIST OF TABLES (CONT’D.)

. Comparison and percentage of GD and MG participants who would be very
likely to eat or purchase selected edible flowers ....................................................... 68

SECTION IV

. Shoot dry weight (dry wt) in grams and percent dry weight (% dw) for eight
edible flower species grown with one of six media/fertilizer combinations. Each
value is the mean of four samples, each composed of three or four plants ............... 92

. Shoot-tissue analysis for Viola tricolor with one of six media/fertilizer
combinations. Each value is the mean of four samples, each composed of four
plants ...................................................................................................................... 94

. Total flower number, average flower size in cm2 and plant dry weight in grams,

for Viola tricolor and V. xwittrockiana grown in a 30% soil medium and
fertilized with Omega 6-6-6 300 or 600 ppm N ....................................................... 96

. Summary of analysis of variance for media pH and EC at the sixth, 12‘”, and 18‘h
sampling week after the first fertilizer application, for eight edible flower
species ................................................................................................................... 97

SECTION V

. Postharvest visual quality characteristics, as influenced by temperature, for five
edible flowers rated 1-5 (5 being the greatest) ....................................................... 114

viii

 

 

 

' Fig.1:

\ \ I13;

 

‘ 2 A 11;:

 

LIST OF FIGURES

Figure Page
S ECTION III
1 . A nasturtium flower that was flawless and was rated a 5 .................................................. 70
2. A nasturtium flower that had the greatest amount of damage and was rated a 1 ................ 72
SECTION IV

1. Root media pH of eight species in soilless (O) or 30% soil medium with five
water soluble fertilizer treatments (MSU Special (19-1.8-19) a water soluble
fertilizer at 300 (0,0) ppm N; Omega 6-6-6 (6-2.6-5) a commercially available
blend of organic nutrient sources at 300 (V) and 600 (v) ppm N; and Fish
Emulsion (5-04-08) at 300 (I) and 600 (D) ppm N). Data points are means of
two samples of three pots each ............................................................................... 99

2. Root media EC of eight species in soilless (O) or 30% soil medium with five
water soluble fertilizer treatments (MSU Special (l9-1.8-l9) a water soluble
fertilizer at 300 (0,0) ppm N', Omega 6-6-6 (6-2.6-5) a corrnnercially available
blend of organic nutrient sources at 300 (v) and 600 (v) ppm N', and Fish
Emulsion (5-04-08) at 300 (I) and 600 (D) ppm N). Data points are means of
two samples of three pots each .............................................................................. 101

SECTION V
1. The postharvest visual quality assessment for nasturtium, viola, and pansy

flowers stored at -2.5 to 20 °C for one and two weeks. Visual quality ratings for
one and two weeks were significantly different at P = 0.001 ............................................ 1 16

2. The postharvest visual quality assessment for borage and scarlet runner bean

flowers stored at -2.5 to 20 °C for one and two weeks. Visual quality ratings for
one and two weeks were significantly different at P = 0.00 l ............................................ 1 1 8

ix

DISSERTATION PREFACE

 

 

State

“1111-;

Lin

IA‘r'er

DISSERTATION PREFACE

This research project, conducted during the years of 1997-2000 began with a goal to serve
the cornrrmnity of the Moapa Band of Paiute Indians who reside in Southeast Nevada The tribe
owns three greenhouses, covering approximately 1 5,000 square feet, which were used during the
1970's for tomato production. Tomato varieties were grown in bag culture in the glass, Dutch
Venlo style greenhouses. During the mid 1 970's, a series ofhail storms destroyed the glazing and
halted production. Over the next 20-3 0 years, vandals further damaged the structure and removed
materials that could be sold off the reservation.

For 10 years, the greenhouses had created a source of income for the reservation. With
the damage to the greenhouse and surrounding properties the tribal community lost a valuable
source of income. As a result, many of the younger generation left the reservation for opportunities
of higher wages and the potential for a better quality of life.

Dr. Ken Poff, a professor in the Department ofBotany and Plant Pathology at Michigan
State University, is interested in minority student affairs and is actively recruiting minority students
with an interest in science for Michigan State. Ken began to talk with tribal elders and the tribal
leader about potential opportunities for their members. After touring the greenhouses, surrounding
outbuildings, and landscape, Ken talked with Dr. John Biembaum, a Professor in Horticulture with
greenhouse crop production and interest in Native American culture.

Ken and John decided to pursue the possibility of finding funding to repair the damage to
the greenhouse, for labor, and for a graduate student to conduct research at Michigan State

University in the areas of marketing, production, and postharvest shelf life of a viable crop that

could lead to a sustainable operation. The venture would need to be both economically and
environmentally sustainable. Cultural sustainability was also considered since a major importance
to the Paiute Indians is their relationship to plants, however, this was not a component of the
research. It would need to employ members for the oomrrrunity, provide income for the community
and if practical would produce crops that were important to their culture. The most apparent
market identified was hotels and resorts in the Las Vegas casino area. Crops that were considered
were culinary herbs, edible flowers, medicinal herbs, and leafy greens.

Kathleen Kelley interviewed with both professors and was chosen as the graduate student
for this research and to pursue her Ph. D. with Drs. Biernbaum and Poff as co-advisors. Degree
certification is joint with Horticulture and Botany and Plant Pathology. Work began during the fall
of 1 997 with a literature review for the project. No published articles were found in the refereed
journals on either the marketing, production, or postharvest shelf life of edible flowers. A few
relevant articles were found in farm or greenhouse trade magazines. Faculty from the Departments
of Horticulture and Crop and Soil Science were recruited for the Ph. D. committee. Members
included: Dr. Bridget Behe, Associate professor in Horticulture with an expertise in marketing; Dr.
Richard Harwood, Professor and C. S. Mott Chair of Sustainable Agriculture; and Dr. Art
Cameron, Professor of Horticulture who conducts postharvest research. All committee members
felt that the research was viable and preliminary greenhouse studies began in late fall of 1997. A
list of objectives was developed which included:

1 . Identify appropriate crops

a. Identify and contact experts in the area and solicit information regarding best

crops for first stages of the project.

b. Grow a variety of crops and evaluate feasrbility for inclusion in future studies.

2. Crop scheduling and flowering
a. Identifying appropriate production stages and times such as propagation,
growth, length of harvest time.
b. Flowering requirement and methods to induce flowering.
c. Regrowth potential and time for crops suitable for sequential harvesting.
3. Production Methods- Production Facility Arrangements for Maximizing Production
Efficiency
a. Use of ground beds and predominately mineral soils versus containers of
predominately soilless media (peat, bark, and coir).
b. Combination of container production and ground beds to maximize space
utilization and scheduling.
c. Types of containers, pots, bags, baskets, and columns.
4. Nutrient Management Strategies
a. Comparison of plant grth and product quality using water-soluble fertilizer
versus compost, organic matter, and slowly soluble minerals accepted in organic
production.
b. Identification and comparison of various sources of organic matter, compost,
and slowly soluble minerals as nutrient sources.
5. Pest and Disease Control *
Facilities most likely will not allow replicated trials. Plant production for research
if production methods and nutrient management will provide opportunities were
pest and disease control will be essential. Predators and biological control agents
used in greenhouses are readily available and will be used on an as needed basis.
6. Packaging
a. Defining the effect of storage parameters (temperature and humidity) on
longevity and quality of edible flowers and culinary herbs.
b. Investigate the use of specialized films for maintaining quality.
7. Marketing
a. Consider consumer perceptions and possible constraints.
b. Develop consumer education information.
c. Investigate potential marketing or brokering channels.

The objectives were refined as the research progressed. Objectives two through five
would firlfil the requirement for environmental sustainability, while objective seven would be
required for economically sustainability. The objectives for crop scheduling and flowering were
completed, however, they will be presented in nonscientific papers after the dissertation is
completed. The use of ground beds and comparison of types of containers, outline in objective

three, were not evaluated because it becarrre obvious that many strategies would be feasible once

 

 

 

 

 

cultural methods and growth characteristics for each species were identified. Several varieties of
lettuce were planted and grown with organic fertilizers in ground beds in a minimally-heated
polyhouse at the Michigan State Horticulture Teaching Research Center. This lettuce research was
not conducted for the dissertation. David Cappert, Research Associate for the Department of
Entomology, assisted with pest control and has written a final report on methods and beneficials
used and the outcomes. The final written report may be used as a basis for future funding
opportunities. Packaging evolved to postharvest experiments which concentrated on the visual
quality of five flowers exposed to -2. 5 to 20°C for two weeks.

The marketing efforts beganwith focusing on both professional chefs and other consumers.
After conducting survey experiments with professional chefs, the researchers found that the rate
of surveys returned was quite low. The researchers then decided to focus efforts on consumers
who were involved with Master Gardeners, Garden Days (an annual event held by the Department
ofHorticulture), and Bloomfest (an annual flower show held at the Cobo Center in Detroit, Mich),
since the rate of survey return and completion were near 100%.

Efforts for the focus of the research also shifted fi‘om completely serving the Moapa Band
of Piaute Indians to serving all small farmers who wanted to grow edible flowers, herbs or other
food crops. Comrmmication with the Tribal Council is still occurring. However, they have not fully
supported the idea of growing edible flowers in their greenhouse to be sold to hotels and resorts
in Las Vegas.

During the fall of 1997, a list of 100 species of edible flowers, compiled fi'om edible-flower
cookbooks and web sites, was created for consideration for the researchers to consider for the

marketing, production, and postharvest shelflife experiments. From that list, 28 species of culinary

herbs and annuals and perennials, with flowers that were edible, were grown in the Plant Science
Research Greenhouses on campus during fall 1997 and spring and surrrmer 1998. Species were
chosen based on recommendations found in cookbooks and by individuals who grew edible
flowers or had tasted them previously.

During the first year (97-98) flowers were evaluated by the researchers for taste, visual
appearance, and marketing potential. Many species were started from seeds and were evaluated
based on ease of propagation, time to germination, visrble bud, first color, beginning of flowering,
firll flower, and senescence. Media consisting of coconut coir and soilless mediumwere used as
growing material with water-soluble fertilizers and fish emulsion supplying nutrients. The
researchers also sampled flowers that were grown in gardens during the summer of 1998 for these
same characteristics. A final list was developed for 1998/1999 greenhouse research which

included 12 species, including the following annuals and perennials:

 

 

Species i Annual (A) or Perennial (P)
Agastachefoeniculum Pursh (Anise hyssop) P
Althea oflicinalis L. (Marsh Mallow) P
Begonia X tuberhybrida ‘Ornament Pink’ Voss A
(Tuber begonia)

Borago oflicinalis L. (Borage) A
Coriandrum sativum L. (Coriander) A
Dianthus superbus L. ‘Super Fantasy Mixed’ P
(Dianthus)

0cimum basilicum L. ‘Siam Queen’ (Basil) A
Origanum vulgare L. (Oregano) P
Phaseolus coccineus L. ‘Dwarf Bees’ (Scarlet A

Runner Bean)

 

 

 

 

 

Tropaeolum majus ‘Jewel Mix’ L. (Nasturtium) A
Viola tricolor L.‘Helen Mount’ (Viola) A

V. X wittrockiana L. ‘Accord Banner Clear A
Mixture’ (Pansy)

Selected plants were used in various experiments presented in this dissertation. Several additional
species were evaluated in 1 999-2000 including: A Ilium schoenoprasum L. (chives), A.
tuberosum L. (garlic chives), Hem erocallis spp. L. (daylily), Rosmarinus officinalis L.
(rosemary), Salvia elegans Vahl. (pineapple sage), and S. officinalis L. (garden sage).

The species were grown using conventional or certifiable organic fertilizers and media.
Organic production was chosen since food grown using this method can sometimes command
higher prices. We also found evidence that no pesticides have been registered for use on edible
flowers. Finally, organic production was considered to be more sustainable and desirable since
amendments were renewable. To be compliant with the certification committees in Michigan, pest
control methods were employed that were established by the California Certified Organic Farmers
Association (Michigan Certifying Committees follow these rules). Beneficial insects and biorational
sprays were used to control whitefly, thrip, aphid, spiderrnite, and mealy bud populations. Data
taken included the root medium pH and EC for each species fertilized with the various organic and
conventional fertilizers. At the conclusion of the experiment, fresh and dryweights were taken,

percent dry weight data were calculated, and ground plant samples were sent to an independent

lab for nutrient analysis.

mmel

  
 

ccndu.

COQSLT‘

1A.;

. 11*“.
‘ L

B.

..Ci‘n

During the experiment, flowers were harvested for tastings held during Garden Days, and
at the Detroit Athletic Club with members of the Michigan Chef de Cuisine. Surveys were also
conducted at Bloomfest, and at Garden Days and Master Gardener Conferences to determine
consumer preferences of edible-flower color, container size, price, and quality.

Flowers were also harvested for postharvest shelflife studies. Flowers were packaged and
placed in -2. 5 to 20° control chambers and evaluated daily for visual quality. Rather than
errrphasizing research comparing packaging films, efforts focused on traditional identification of
potential storage temperature and product longevity.

Five manuscripts, prepared for publication in referred journals, were selected for inclusion
in this dissertation. The results fiom these experiments are presented in the following chapters of
this Dissertation. Additional research completed is summarized in the conclusion of the dissertation.
Chapter one includes three manuscripts containing the marketing experiments conducted with Drs.
Bridget Behe, John Biembaum, and Ken Poff. The first paper included in Chapter One is entitled
“Consumer and Chef Perception of Three Edible-Flower Species.” This paper was formatted and
accepted for the journal HortScience. The second paper entitled “Consumer Preference ofEdrble-
Flower Color, Container Size, and Price” was also accepted for publication in HortScience. A
third paper “Consumer ratings of edible-flower quality, color, and mix” will be submitted to the
Journal HortTech.

The second chapter contains a paper entitled “Organic Nutrient Management of
Greenhouse Production of Edible Flowers in Containers” coauthored with Dr. John Biernbaum.
This paper presents the results from one of the production experiments and will be submitted to the

Journal of American Society of Horticultural Science.

hDrs

litre:
intent;

 

The third chapter describes the postharvest experiment and contains that paper coauthored
by Drs. Art Cameron, John Biernbaum, and Ken Poff entitled “P ostharvest shelflife of five edible
flowers”. This paper was formatted for the journal HortScience and will be submitted after the

internal review is completed.

SECTION I

CONSUMER AND PROFESSIONAL CHEFS PERCEPTIONS
OF THREE EDIBLE-FLOWER SPECIES

HORTSCIENCE
Accepted for publication: May 16, 2000

fonsu

 

Consumer and Professional Chef Perceptions of Three Edible-Flower Species
Kathleen M. Kelley‘,

Michigan Agricultural Experiment Station and Departments ofHorticulture and Botany and Plant
Pathology, Michigan State University, East Lansing, MI 48824-1325

Bridget K. Behe2 and John A. Biernbaum3

Michigan Agricultural Experiment Station and Department of Horticulture, Michigan State
University, East Lansing, MI 48824-1325

Kenneth L. Pofls

Michigan Agricultural Experiment Station and Department of Botany and Plant Pathology,

Michigan State University, East Lansing, MI 48824-1325 -

 

Received for publication January 25, 2000. Accepted for publication May 16, 2000. Research
conducted at Michigan State University, East Lansing, Mich. Use of trade names does not imply
endorsement of the products named nor criticism of similar ones not named. The authors thank
Mike Rutter, Statistical Counseling Center, Michigan State University and Judy L. Pfaff,
Programmer Analyst, Department of Agricultural Economics, Michigan State University, for their
statistical support. We also thank the Garden Days coordinators at Michigan State University for
permission to conduct the survey and the Michigan Chef de Cuisine Association.

1Graduate Student

2Associate Professor. To whom reprint requests should be sent. E-mail
address:behe@msu.edu

3Professor

10

 

Qfi {lair

fwd-lib.

Fit;
‘4wa
\ l

Marketing and Economics

Additional index words. Survey, econorrrics, visual, taste, fragrance, Viola tricolor, Borago
officinalis, T ropaeolum majus

Abstract. Two surveys were conducted to assess consumer and professional chefs’ perceptions
of three edible-flower species. Our objectives were to determine opinions, preferences, and uses
of Viola tricolor L. ‘Helen Mount’ (viola), Borago oflicinalis L. (borage), and Tropaeolum
maj us L. ‘Jewel Mix’ (nasturtium). Flowers were grown using certifiable organic methods and
chosen to reflect a variety of flower tastes, textures, and appearances. We quantified three
attributes (taste, fragrance, and visual appeal) with a total of seven semantic, differential scales
adapted from a scaling authority. The attributes were rated as: visual -- “appealing”, “desirable,”
and “very interested in tasting”; fragrance -- “appealing” and “pleasant”; and taste -- “tasty” and
“desirable”. Garden Day participants were self-selected to evaluate and taste flowers from a
consumer perspective. When asked to rate the three species on visual appeal and desire, no less
than 76% of consumers awarded all flowers an acceptable rating. We found similar results when
consumers answered questions regarding the taste of two of the three species. Results from this
study support our hypothesis that customers would rate edible flower attributes highly and would
be likely to purchase and serve the three species tested. Members of the Michigan Chefs de
Cuisine Inc. Association participated in a similar survey. At least 66% of these chefs rated the
three visual attributes and two fragrance attributes of viola and nasturtium acceptable. Chefs ’
ratings of the fragrance of borage as “appealing” and “pleasant” were higher than those of
consumers, but the ratings were still low, 21 % and 25%, respectively. Unlike consumers, chefs’

ratings of the taste of viola as “appealing” and “desirable” were low (29% and 3 6% , respectively).

11

We found some minor differences in ratings when groups were compared, using demographic
variables as a basis for segmentation, indicating a homo genous marketing strategy may be
employed.

Introduction

The popularity of edible flowers has increased since the late 19805, as evidenced by
upscale restaurants where edible flowers garnish meals. Such flowers are featured in popular
articles, and are also the subjects of cookbooks (Barash, 1993', Belsinger, 1991 ; McVicar, 1997).
The list of edible flowers is extensive, with over 55 known genera represented (Badertscher and
Newman, 1996; Barash, 1993; Belsinger, 1991, McVicar, 1997). Flowers canbeusedirrsalads,
to garnish soups and entrees, as ingredients in main dishes, sprinkled on desserts, frozen in ice
cubes and floated in drinks (Barash, 1998a, b).

Not only do edible flowers add excitement to food presentations, they also have nutritional
benefits. The vitamin content of viola on a weight basis is higher than that of oranges [Citrus
sin ensis (L.) Osbeck], and viola also has a higher concentration ofbeta-carotene (Kosztolnyik,
1996). Because most flowers are more than 95% water, the nutrient value does not appear to be
significant. Flowers, however, are nearly calorie-free (Evans , 1 993 ) , a real advantage to health-
and weight-conscious consumers. Visual, taste, and nutritional appeal enhance the potential
marketability of edible flowers.

Consumers, including professional chefs, may find sorrre edible flowers more appealing than
others, and research can facilitate efforts to determine which edible flowers deserve further
investigation. With better consumer infomration, growers, wholesalers, and retailers can selectively

target consumer groups with more effective marketing strategies. Consumer information is lacking,

12

although a few articles include suggestions as to which flowers are edible and indicate that growers
are producing flowers for both restaurants and consumers (Evans, 1 993, Greenhouse Business,
1998, Whitman, 1991).

As edible flowers become more popular, defining the preferences of consumers and chefs
will become more important to producers and marketers. Our objectives were to: a.) determine
which characteristics appeal to consumers and chefs, and b) how similar the preferences of these
two groups are, which may dictate separate or similar marketing strategies. Edible flowers can be
grown using certified organic methods. Since production method is a consideration for the grower
and retailer, we also wanted to determine if production method had an effect on edible flower
desirability by chefs and/ or consumers.

Material and Methods

General. Flowers of viola, borage, and nasturtium were grown in a greenhouse at a
constant 20 °C, in Strong-Lite Universal Mix (Strong-Lite, Pinebluff, Ark), and fertilized every
other week with fish emulsion (400 mg-L'l , 4N-0.4P-0 . 8 K; Northeast Organics, Manchester-by-
the Sea, Mass), MSU special (500 mg-L“, 19N-1.8P-19K; Greencare, Chicago, 111.), or dried
blood (12N-0P-0K', Dragon Corporation, Roanoke, Va.) Plants were not treated with
pesticides. Corrrrnercially raised predators and parasites, including A ph idius colemani Viereck
(aphid parasite), Hypoaspis miles (Berlese) (fungus gnat predatory mite) and Amblyseius
cucumeris (Oudernans) (thrips predatory mite), were released in the greenhouse to help control
aphid and mite populations. Flowers were harvested the mornings of the surveys and stored in a
5 °C cooler several hours until needed. They were rinsed with distilled water 2 h before the tasting

to remove any visible debris. Rinsed flowers were placed in labeled, 50-mL paper cups. Order

13

 

...1
L.‘

of tasting was randomized among participants. One nasturtium (9 10 mg), three borage (290 mg
each), or three viola flowers (80 mg each) were used per cup.

Two surveys were administered, one of consumers who were attending a state-wide annual
event, the second of chefs at a monthly professional chefs association meeting in a large
metropolitan area. A 43-item consumer-directed questionnaire and a 38-item chef-directed
questionnaire were developed to assess perceptions of visual, taste, and fragrance attributes of
three edible-flower species. The survey instruments consisted of either twelve preference items
for each of the three species plus either seven (consumer-directed) or five (professional chef-
directed) demographic questions . We selected three questions for visual evaluation and two
questions for fragrance evaluation and adapted two questions for taste from an authority on
measurement scales for various product attributes (Bruner and Hensel, 1996). We quantified the
attributes with a total of seven, seven-point, semantic differential scales in the reference. We asked
both groups of participants if, from a visual perspective, the flowers were “appealing,” and
“desirable” and if the participants were “very interested in tasting” the flowers after seeing them;
if, after smelling the flowers, participants found them “appealing” and “pleasant”;and if the
participants, after tasting the flowers considered them to be “tasty” and “desirable”. We also asked
how likely consumers were to grow or to purchase a particular species and how likely they would
be to use them in entrees. Professional chefs were asked if they were more likely to purchase the
flow er if it were grown using certified organic methods, whether they would purchase the flower
if it had 10% insect damage, and if they would identify and describe the flavor and fragrance of the
flower using their own words. Demographic questions for consumers included age, gender,

education, family status, household size and income. Professional chefs were asked questions

14

 

Self-e

‘-

P-s _

LIV)...»
....

the r.

(Jr

".1
(la
(1

including gender, meals served each week, price of least expensive entree, years as a chef, and if
they were certified. The order in which participants evaluated the three species was randomized.
Results were tested for significance using logistic regression and chi-square with Fisher’s exact test
(SAS Institute Inc. , 1998). The experiment was approved by the Institutional Review Board for
Protection of Human Subjects at Michigan State University prior to implementation.
Consumer-directed survey. Forty-one participants were self—selected from 287
registrants who attended Garden Days at Michigan State University on 7 August 1998. During this
popular annual event, participants who are highly involved in gardening pay a fee to attend a variety
of activities, such as garden walks and seminars on flower paper making, and herb use and
production. Minors (under age 1 8), pregnant women, and those with severe food allergies were
self-excluded from our study. Volunteers reflected a diverse sample, including, but not limited to,
college students, employed persons, retirees, athletes, and gardeners. When participants entered
the room, and after they signed a waiver, they were given three cups, each containing a flower
species and a survey form The survey took approximately 20 min to complete.
Participants ranged in age from 1 9 to 72 years, 66% were ferrule, and 55% had graduated
from college; 72% of the participants had a 1 997 household income that ranged fiom $20,000 to
$79,999. Fifty-three percent of the participants had no dependents, and 4 1 % had more than two
people in their household. Participants were asked to identify themselves by zip code. Of the 37
who responded, 46% lived within 32 km, the remainder from 3 3 to 2 58 km from the survey site.
Participant responses using the seven-point Likert scale were divided into three categories:
not acceptable (composite of responses rated either 1 , 2, or 3); neutral (response of 4); and

acceptable (composite response of either 5, 6, or 7). We assumed that if at least 30% of the

15

 

S r u t .... t
.3 ... r a d
....r. n. r e W P
'“Ln 9th F H“ A .5

 

E‘s E

Ch:

a::n'~_

\

‘isua:

Ilr

responses for any itemwere acceptable, asubstantial market segment was pleased with that flower
attribute.

Professional ch ef-directed survey. Twenty-six member-chefs of the Michigan Chef de
Cuisine Inc. Association participated in a survey at the Detroit Athletic Club in Detroit, Mich, on
8 March 1999. Pregnant or potentially allergic members were self-excluded fi'om the study. The
chefs signed a waiver, and were given three cups , each containing a flower species and a survey
form. These participants took less than 20 rrrin per person to complete the survey. Chefs who
participated in the survey answered demographic questions. Six of the chefs were women, 20
were men. About half of them served 500 or fewer meals each week. The highest number of
meals that a chef reported serving in 1 week was 35,000. Half the chefs that participated also
served entrees that cost $8 or less, were certified chefs, and had been chefs for 1 1 years or more;
five had been chefs for 20 years. Of those who were certified, six were Certified Executive Chefs.
Chefs rated the flower attributes on the same 1-7 scale that consumers used. Their responses were
also divided into the same three categories: Not acceptable, neutral, and acceptable.

Results and Discussion

Consum er-directed survey. For all three species, at least 74% of participants rated
visual characteristics (appeal, desire, and interest in tasting) acceptable (Table 1). Participants
rated visual appeal and desirability of viola higher than those of borage or nasturtium.

The majority of consumer-participants rated all three species acceptable for both taste
attributes. For viola and borage, ratings for taste appeal and desire were lower than ratings for

visual characteristics. Of all three species, nasturtium was rated highest for appealing taste (80%),

16

but lowest for interest in tasting after visual examination (74%) , visual desire (76%), and visual
appeal (74%).

Few participants (8%) rated the fragrance of borage as pleasing. A much higher
percentage rated viola (68%) and nasturtium (54%) as having acceptable fragrance. Verbal
comments indicated that participants either did not detect or did not approve of the odor of the
borage flower. More responses for borage were neutral (80% and 8 1% for an appealing fragrance
and pleasing fragrance, respectively), than acceptable. Thus, borage fragrance may not be a
germane issue.

Each species was also evaluated for potential use. At least 68% of participants rated viola
acceptable for use as a garnish or in a salad, and 85% would purchase it if available. Similar
responses were observed for nasturtium, but only 25% of participants would use it in a salad.
Borage received ratings of at least 48% for potential uses, with 65% of the participants finding it
acceptable for serving to guests.

We investigated differences in mean ratings by gender, education (college graduate or not),
income (household income s $40,000), marital status, and household size (Table 2). The mean
ratings of the higher income (2 $40,000) group were consistently higher on 14 items across the
three species tested They had a higher mean perception of the visual desirability of viola, and were
more likely to purchase this flower. They rated borage higher than the other two species in visual
appeal, interest in tasting, pleasant fragrance, desirable to taste, use as a garnish and in a salad, and
purchasing and serving. They were also more interested in tasting the nasturtium flower, found that

it had a pleasant fi‘agrance and desirable taste, and were more likely to use it in a salad.

17

 

 

 

Di]

Married participants rated 1 5 items consistently higher than did single participants; these
included visual desire of viola, visual appeal ofborage, interest in tasting borage, use ofborage in
a salad, and purchase of borage, as well as nearly all attributes of nasturtium.

We divided participants into two age groups at the median age: 43 years. Participants who
were 43 or older rated both fragrance and both taste characteristics of viola higher than did the
younger age group. They also rated all of the usage attributes for all species, except for the use
of viola as a garnish and in a salad, and purchase of the flower, higher than did their younger
counterparts. Aside from the appealing and pleasing fragrance ofborage and its appealing taste,
the older participants rated all of the other visual, fragrance, and taste attributes of borage and
nasturtium lower than did younger participants.

Across all three species tested, we saw only three differences were apparent in mean
preference by gender. Males rated the visual appeal of nasturtium lower than did females and
indicated that they were less likely to use it as a garnish and grow it.

We found only one difference in mean preference rating when comparing college graduates
with non- graduates. College graduates rated use of nasturtium as a garnish higher (5.2) than did
non-college graduates (4.0).

Mean preferences of small households (one person) vs. larger households (two persons)
differed in only two cases. Respondents from one-person households rated both pleasant and
appealing fragrance of viola higher.

Income and marital status were the two variables for which we found greatest differences,

and these would be usefirl in a segmentation scenario. Age, gender, education level, and number

18

 

 

12151111

”‘1

'5;

n
(I

pun

of people in the household showed few differences and these variables would not be a good basis
for segmenting the market for edible flowers.

Professional chef-directed survey. At least 72% of the chefs rated the Visual
characteristics of viola and nasturtium (appeal, desire, and interest in tasting) as acceptable (Table
1 ). Nasturtiumhad the highest rating (96%) for visual appeal, but the lowest (72%) for interest in
tasting. Borage did not receive a rating higher than 54% for any of the three characteristics.

Nasturtium received the highest ratings for appealing (87%) and pleasing fragrance
(79%) . In contrast with the results from consurrrer participants, viola received the second highest
ratings in both categories, 66% and 76%. Although chefs rated borage higher in both categories,
21% and 25%, than did consumer participants, in both categories borage received the lowest
ratings among the three species.

For the chef s fragrance ratings, nasturtiurnreceived a 72% rating for appealing taste and
a 7 4% rating for desirable taste, and the highest ratings in both taste categories. Unlike consumer
participants, the chefs rated viola’s appealing taste (29%) and desirable taste (36%) lower than
borage’s (63 % and 48%, respectively).

When chefs were asked whether they were likely to use the three species in a salad, 65%
reported that they would use nasturtium, 42% that they would use viola, and 40% that they would
use borage. In contrast, consumers rated viola highest, followed by borage and nasturtium While
chefs were more likely to purchase nasturtium (64%), consumer participants were more likely to

purchase viola (85%). Only 39% of the chefs gave viola an acceptable rating.

19

ll

£7311”

for b

 

When chefs were asked if they were more likely to purchase the flower because it was
grown organically, the acceptable ratings increased from 3 9 to 54% for viola and from 38 to 46%
for borage. The acceptable rating for nasturtium decreased by 2%.

When chefs were asked if they would purchase the three flowers if they had 10% insect
damage, only 8% gave viola an acceptable rating. When asked whether they would purchase a
borage flower or anasturtiurn flower with 1 0% damage, only 4% of the chefs, in both instances,
indicated that they would.

In order to determine differences in mean ratings for viola, borage, and nasturtiurrr, we
divided chefs into groups based on gender, average number of meals served weekly, meal cost,
professional certification, and years of experience. Ratings varied little with gender, with the
exception of appealing fragrance (3. 5 for females vs 5 .4 for males) (Table 3). Mean ratings for
borage characteristics were similar except for desirable taste, which females rated higher (5.5 vs
3.8). Males were more likely to purchase nasturtium (5.3 vs 3 .0).

We found no differences in ratings between professional chefs who served more vs. less
than 900 meals per week.

Chefs employed by restaurants where the least expensive entree cost $8 or less were more
inclined to purchase viola if grown organically than were those working at restaurants where
entrees were more expensive (5.4 vs 3 .4). Mean ratings were similar for chefs in both categories
for all attributes. We found three differences in mean ratings for nasturtium. Chefs at less
expensive restaurants gave higher ratings to visual desirability (6 .2 vs 4. 6) and interest in tasting
(6.3 vs 3 .3), and were more likely to purchase if plants were grown organically (6. 1 vs 4.7) (Table

3).

20

Years of employment did not affect ratings for attributes of viola, but several attributes of
borage were rated lower by chefs who had worked for 1 l or more years (Table 3).

Certification had little effect on ratings for viola and borage, but non-certified chefs rated
fragrance ofnasturtiumboth more appealing (5.6 vs 4.9) and pleasing (5.6 vs 5.0) than did certified
chefs (Table 3). Across all five of these comparisons, chefs were more similar in their evaluations
than were consumers.

Chefs were also asked to describe the fragrance and taste of the flowers, and how they
used them If they did not use them, they were asked if they intended to use them in the future.
If the answer was positive, they were also asked how many days they would use them, and how
much they would pay for a dozen (data not shown). Most of those who responded described viola
as having a floral to perfume scent. Chefs were not able to describe the fi‘agrance ofborage, which
corresponds with the low acceptance ratings. Chefs described the fragrance of nasturtium as
having a floral, perfumed, sweet, spicy, or fi'uit-like scent. The tastes of viola and borage ranged
from undetectable to bland, “vegetable,” or, in the case ofborage, sweet or clam-like. Most chefs
described nasturtiurn’s taste as peppery, and/ or cabbage-radish-or citrus-like. Nearly all chefs said
they would use all three species in a salad, but less than half would use the flowers as a garnish.
Several chefs reported that they would use borage in seafood presentations, while one reported
that he/ she would use nasturtium in Asian presentations.

Fifteen percent of the chefs used viola in their presentations , eight used borage, and five
used nasturtium. Forty-six percent of the chefs would consider using the flowers at least once a
week, and the rmjority (58%) would use them two to three times aweek. Prices that chefs were

willing to pay for the flowers varied greatly. They were willing to pay $1 .00 to $1 0.00 per dozen

21

 

 

PASTE

Chfil

Fen

Ex:

IX:

ha:

r:

for viola, $1.00 to $5.00 for borage, and $2.00 to $1 5.00 for nasturtium. At least one chef
reported that the restaurant grew edible flowers in the garden. Although actual use is low, the
potential for increasing use of edible flowers among these chefs appears to be great.

Sample comparisons, consum er and professional chefs. We analyzed the mean ratings
ofboth chefs and consumer participants to determine if the two groups had similar opinions about
the edible flower species. For the visual and fragrance attributes of viola, consumers and chefs
expressed similar perceptions (Table 4). However, consumers liked the taste better and were
more likely to purchase and use viola than were chefs. Consumers liked borage better than did
chefs, except for fragrance. In contrast, three nasturtium attributes were rated higher by chefs.
Few mean ratings for nasturtium were significantly different between the two groups, with the
exception of visual appeal, appealing fragrance, and purchase of the flower. Similarities in ratings
may indicate that no separate marketing strategy is warranted. Consumers and chefs appear to
have similar preferences about attributes for nasturtium, but this is less true for viola and borage.

There was a significant correlation between the decision to purchase and flower taste for
both participant groups. The relationship between taste and purchase for viola was stronger for
chefs (r=0.51) than for consumers (r=0. 1 3). For borage, these correlations were higher (chefs
r=0. 7 1 , consumers r=0 . 64), but, for nasturtium they were similar (chefs r=0 . 78 and consumers
r=0.60). This indicates that consumer ratings for “tasty” flowers are strongly related to their
willingness to purchase the flower.

Conclusions
Most of the participants found nasturtium and viola visually acceptable, while more than

half found borage visually acceptable. Viola and nasturtium had an acceptable fragrance rating

22

 

 

pszer

! the;

\V r.‘

x.

I"
(r

..fi—._h .,.74 A

pv<

from more than halfof all participants. The acceptability of flower taste varied for consumers vs.
professional chefs. At least a portion of both groups liked the taste and would appear to be
potential target markets for edible flowers. The species tested appear to have moderate to high
market potential, suggesting that a new niche market exists for growers and retailers. We saw
many similarities between consumers and chefs in their perceptions of edible flowers.

Given these ratings, we believe that there is adequate reason to further investigate market
potential of these flowers. In future studies to determine consumer preferences, determining the
effect fragrance has on taste, or on consumers’ choice to buy an edible flower product, may be
possible. However, packaging could eliminate this as a factor in the decision to buy.

When analyzing the use characteristic data, one important factor is whether the participant
would consider purchasing the flower if it were available. Whether a person will buy a product has
a pro found effect on whether the product will be available in the long term The fact that at least
58% of the participants reported that they would buy the flowers, if available, supports the
hypothesis that more consumers would purchase them if they were readily available. Another
consumer segment may choose to grow their own edible flowers. In this study, at least 48%
indicated that they would be likely to grow them These results indicate that grocery stores or
specialty food stores may be ready for product and price trials. This may be a more effective
variable for market segmentation.

Chefs indicated they would be more likely to purchase viola and borage if grown
organically, suggesting that chefs are not only concerned about serving attractive, fragrant, and
palatable flowers, but also prefer flowers that have been grown organically. Nevertheless, few

would accept flowers that had 10% insect damage. These two responses pose a challenge to

23

producers. Chefs would prefer to purchase organically grown flowers, but would not buy them
if they were flawed. Research can help marketers close the gap between an expectation of no
tolerance for blemishes and efficacy of pest control methods acceptable in organic production.
Flowers grown using certified organic methods may incur more insect damage since pesticides
cannot be used for rapid control of outbreaks. Researchers may need to provide information
indicating how growers can produce edible flowers with minimal insect damage.

Other areas should be investigated as well. These questions include: the effect of the visual
and fragrance characteristics on the decision to purchase, how much will customers pay for edible
flowers, how many flowers customers will purchase, and what effect color intensity has on the

decision to purchase.

24

Literature Cited

Badertscher, KB. and SE. Newman. 1996. Flowers. Colorado State Univ. Coop. Ext. Bul.
7.237.

Barash, CW. 1993. Edrble flowers: From garden to palate. Fulcrum Publishing, Golden, Colo.

Barash, CW. 1998a. The flavors of flowers. The Herb Companion 10(4):32-37.

Barash, C.W. 1998b. Please eat the flowers. Horticulture 95(5):36-40.

Belsinger, S. 1991. Flowers in the kitchen: A bouquet of tasty recipes. Interweave Press,
Loveland, Colo.

Bruner II, G.C. and P.J. Hensel.1996. Marketing scales handbook. Amer. Mktg. Assoc,
Chicago, Ill.

Evans, RD. 1993. Flowers as food. Small Farm Today 10(2): 18-21.

Kosztolnyik, L. 1996. Selling edible flowers. Natural Food Merchandiser 17(7):74.

McVicar, J. 1992. Good enough to eat: Growing and cooking edible flowers. Kyle Cathie Ltd,
London.

SAS Institute, Inc. 1998. ver. 6.12 for Win95, SAS Institute, Cary, NC.

Whitman, AT. 1991. Edible flowers and culinary herbs: New uses for traditional crops, new

crops for traditional growers. GrowerTalks 54(13):22-23, 25, 27, 31, 33.

25

 

—r.. ..v.u:—h.‘—‘ .4.\....—....:- .1

 

93:51—53.) 3 2:3 ....or...1_ru._._:91.22.4971... 2.5.5:: aft—re .3 7.2:?!

.36 w m 60.58%“: osflwo— 3 Coco 3.66835 98 685303 @0333 320% can 958 £55 53838 $36

 

 

26

a v a v a m own-Ewe 885 a mm a no p cm 2650

go 2:3 9 atom 2:03

03:83 2:03 a we a mm a on 28.9:

a me a we m cm oanco 8233 38m 2:03

usage 28% a mm a S a 2 sac: as

630: $5 3283 2:03

a 8 a mm a on 03:23 283 a mm as we a we 38

m we a ov a me 92mm an mm a on a me 5:55
35 35

n we as we m cm oBfiRoQ a we a em a we Eastman”

a Nb 8 8 a on 338%.. a ow a S a me 3383
8.85 mag

n E a mm a as wanna a am a w a K msaua

a S a a n 8 manages, 9 mm a w a we manage.
oozEMEK mozexMEk

n S. a we p cm wccmfi E 33285 a we a E m cm megs E eoumocBE
a No a on 9 mm 058605 a on a E n no ofigmom
a 8 a 4m .9 S 31.8%. a we a S L mm 338%
35.3 3me

Ear—3mm Z owmcom 32> 23.52 Eatfimmz owmcom 22> 8352

 

 

£05 35.5835 3 86on do mwcnmm 805338 3 86on do mweumm

 

 

 

 

 

C. .5 .0 .m :38 053 coon
Bach—Eco a 5:5 noncommoCo omfieoouom €250: 05:6 do 86on 685.8 83953 do mwecfi deco Eeoimod 05 9a 8:25:00 ._ 2an

 

 

.... ......z: ....

.1: 13......

.....22; ...: ...

.._._...:_ ...

......ZZZ ...:

 

0.080

 

 

 

 

am am 3 0a 3 6m 3 mm 3 we 0w 3 335
9.280%

3 3 am ..0 am mm 3 4.0 30 em :0 ow 35$ 888
£050
3 mm on an on am an em 3 cm 3 ..m 2 atom
mm 5. mm 3 on we 3 mm 3 3 10.0 3 0300 as
8285
0030:
0.0 3. 3 mm mm 0.0 ma mm 0.0 mm ow 0.4 35 380
020m
3 3 3 ..m 3. cm 3. 3 we 3. 3 S. a E a:
5.5%
on 3. ea am 3 mm 0% 3. om cm 3 mm a mm .8:
8mg
3 ow 3 we. 3. 5. S. E. 3. 3. 3 3 858000
2mg
3. 3. 3 3. 3 3. mm 3. we 3. we 3 méeaea.
.2 mm E 2 4m 2 on gm mam 5. 3 0m 8503
means—m
.3 9m 4w 4.... 2 cm am am mm 3 mm am 85020
wan—309$
mans
G E 0.0 no 3 cc mm 4.0 me am 3 mm 5. e225
0500
4.0 me no No 4.0 4.0 S 0.0 .3 m0 tee on 335
mEanam

no he 60 0.0 4.0 0.0 4.0 we we m0 m0 G 335 22>

NN 3 saw :80 eases 252 amend an an BEE ”new 054% 05va 2382. 868m
3260 30:8
a 002
2023.00: 05 5 0000005 80:00 me 0 < £52m 2:005

303030 .02

023.53 owlmm: 0:0 .208 .8emcwwa .335 :0 083 8005080 05 Noumea .203 do 3593385 Eon—0300
00 maze :88 :0 202820: 05 E 03000 00 0005:: 000 00:00:00 d0 6%: 500% .300000030 ..0 0mm 0:0 mama RES: .0885 (.0 macaw .N 2an

27

m.m
wé

wd

9m

5m

ow

wé

ed

wd

vé

mé

vé

wé
5m
Wm

aim

Nm
ow

Nam

Wm

9m

<6

ed

3.

ad

06

N6

m6
0%
_.v

in

Wm
me

oh

Wm

Wm

«in.

N
V

0%

5m

w.m

04V
QM
w.m

dim

fim
_.m

fin

oh

oh

oh

5%

mfi

06

we

we

«xv

9v
w.m
QM

Wm

N6
0.?

wd

.8040

mi

5.?

ed

0.?

oh

ofi

5v
m.m
9m

fim

Ym
me

5m

6.0

50

mé

m0

Q?

N.
‘1’

mé

mé

ed
5v
0%

Om

...mav

vé

we

..md

_.m

N.
In

md

:0

we

oé

_.v

me

0%
ad
ad

wé

N0

in

Wm

_.o

mo

50

md

0%

mm

me

me.

md

mé
5m
5m

_.©

...w.m
:h

_.m

{00

...N.c

«mé

cd

we

wé
5m
Qm

50

NV
0.?

me.

vd

wfi

wd

Qm

mi

md

Wm

Wm

5m

me
0d
5m

0%

...—.0

:6

...—h

...m.m

NW

*0ng

.06.».

...a.v

m6
00M
0%

{0.0

v.0

vd

wd

_.m

N.
v

mé

N.m

QM

Wm

06.
w.m
Wm

5v

Ems
waging.
808mg
@5003

0030a
w0=800<
magma

E 6825

0:30
:50;

30000
.035
$050

2 030m
$300 $5
8280a
0030:
0.5 305
030m

0 E 8:
55am

0 mm 35
23
205080
ems
méaané
00:33.0
wEmmuE
8&3
w§800<
warms

5 $835

825022

30:80 m 205

28

Abzfimooa 33-8 “mob 83m mLonmi so 3me E .o m K E EmocEmB 68m. .oEoUEV mutowemo £53, 8:553 .«o moocouobfi :82...

 

v5.50
mé m6 vd wé m6. Nd *wd Fm twm mam on Wm 0. 2Com
wd Nam QM md :u 3‘ 3M vw ...m.m Wm we. :U bkoa ME.
3.285
.526:
we 0% me _.m 1% iv an Wm ..m.m w.m ow w.m mi: 305
62mm

. . . i . . «5 3
v6 06 gm fixv 5m 5? 1m wv ...wv hr *0? mm . D
532%
ch NV ...o.v Wm 35 0..“ ...w.m Um ...m.m Wm wd Wm m mm 353
8mg

..- . . . . . 0 Saw
we 5% wd we. we we ion #0 Law mm ...mm VM 3 . G

3:53 N 29$

9
2

 

 

.7. 1.12:... 2.5:. 2.. 127.433.:7...

... __..__-.:__—_..,a t...“ ._.:_.a ._ I_,

7......)

....:_:.. .3.;:.:?2 9...... .1 ..

66 ch in N? QM fin 5m Nd cm 5m 633 m E 3.3

 

 

 

 

 

 

. . . . . . . . . . 530:
3 3 3 3 3 3m 3 3 3 3 méommfiam
3 3 .3 3 3 3 3 3 3 3 23 25230
3 3 .3 3 3 3 3 3 3 3 322 3:89?
3 3 L 3 3 3 3 3 3 3 .3 8:38. 9:355
. . . . . . . . . 8:38
a m w m ...3 3 3 m m m a m m m c m m 3 35.895
3 3 3 3 3 3 3 3 3 3 953 a .325
3 3 .3 3 3 3 3 3 3 3 83% 32>
3 3 .3 3 3 3 3 3 3 3 35.8%“ 335 930m
. own—3c
3.. 3 m: 3 f E E Z M: 3 535:.2
5:5 93:83
>=8an5 850%
3 3 3 3 .3 3 3 3 3 3 a 3223
2 382 8oz
3 3 3 3 3 3 3 3 3 3 33. a a 8:
530a $5
3 3 3 3 3 3 3 3 3 3 $205
3 3 3 3 3 3 3 3 3 . 3 23 03839
3 3 3 3 3 3 3 .3 3 3 222 3:23
3 3 3 3 3 3 3 3 3 3 8:33 33a
. . . . . . . . . . 8:38
3 3 3 3 3 3 3 3 3 3 33%?
3 3 3 3 3 3 3 3 3 3 mafia 5. 5225
3 3 3 3 3 3 3 3 3 3 23% 32>
3 3 3 3 3 3 3 3 3 3 3:83 :33 22>
oz 8». _ _ N o. m 3N mm“ 8% 8% 222 225“. 2392 862%
cosmocfou :35 8:5 03.3598 x83 :03 $950
a 3 an“; ~32 .6 ~35 338 £32:

u ho 3%
.8255; mag: 28 “Ems “eczema. .333 so 323 52:33: 23 6 88 £65 “o :53 88 £23
‘6 mucus SEE no 2528?: go :88E:8 23 £5 a ,3 3% .855 «>553 332 go 85 283 58 Bag 238 go 59:5: 63% “o «Sam .m 22%

 

3O

3:5805

32-8 and. 53m mLonmE :0 3.25 2,: w R E EmocEw; 33 233 5% 33% ESE .3 52:5: .55qu maromuza 25:3 7,8333 ho mouceéfi :32;

m.—

—.m

0.?
ad
Vm.
ch
ion
tow
mé
cw
v.0
ON

9v

0N

mé

in
0%
9%
Wm
oh
ad
06
3.0
No
b.—

0.—

...ch

c.m

o.—

..m

..m

N
V

«.6.
Wm
WV
06.
New
cc

4.

f

N.
V?

N..—

“Wm
.96.?
am
0.—

_.m

4N

ac

cm
0%
«Em
0.?
be.
We
m6
co

m;

_.m

N..—

fm

5v
mé
oh
oh
ah
ah
wé
0m
w.m
w._

ed

_.M

mm

0.0
m.m
Om
0.0
5v
0.?
ch
mc
vc
N..—

m.m

_.m

m.m
vim
Wm
9m
m.m
m4»
Tm
We
0;

oh

w.—

may

we
ck
5m
md
md
5%
WV
mm
m...»
w.—

CM

338%

63.5 $2

53» 328:;
onEmmS :3on
ba 8223

2 39.: 832
228 a E 3.:
530:

mp: amazoam
oucfimg wEmmuE
~33 3:39?
uugag 9585
8:33
@5895

mczmfi E $825
0580 333
wczwoaam 15$ >
ommEmc

635 $2

2:3 amazoam
3.8.8me
ESE : 3223

2 >3: 802

EsEammZ

.5209 m 23

31

Table 4. Mean rating of chef vs. consumer acceptability of three species of edible flowers based on
visual, fragrance, taste, and usage attributes.

 

 

Viol_a Borage Ngsuutimn
Attribute Chef Consrmier Chef Consumer Chef Consumer
Visual appealing 6.1 6.7 4.7 5.8* 6.2 5.7*
Visual desire 6.0 6.4 4.5 56* 5.9 5.6
Interest in tasting 5.5 6.1 4.2 5.5* 5.6 5.5
Appealing fragrancc 5.3 5.4 3.8 3.9* 5.8 50*
Pleasing fragrance 5.6 5.5 4.0 3.9* 5.8 4.9
Appcaling taste 3.8 4.8* 4.3 4.6 5.4 5.4
Desirable taste 4.2 4.7* 4.7 4.5 5.5 4.7
Use in a salad 4.] 4.8* 3.7 4.0* 5.2 4.0
Purchase this flower 3.7 50* 3.8 4.3 5.0 4.2*

 

*Mean differences of attributes within categories (chef vs. consumer) significant at P s 0.10
based on Fisher’s Exact Test (2-tail probability).

32

SECTION II

CONSUMER PREFERENCE FOR EDIBLE-FLOWER COLOR,
CONTAINER SIZE, AND PRICE

HORTSCIENCE
Accepted for publication: August 1, 2000

33

Consumer Preference for Edible-Flower Color, Container Size, and Price

Kathleen M. Kelleyl

Departments of Horticulture and Botany and Plant Pathology, Michigan State University, East
Lansing, MI 48824-1325

Bridget K. Behe2 and John A. Biernbaum3

Department of Horticulture, Michigan State University, East Lansing, MI 48824-1325
Kenneth L. Poi‘l3

Department of Botany & Plant Pathology, Michigan State University, East Lansing, MI 48824-

1312

 

Received for publication 5 Apn'12000. Accepted for publication 1 August 2000. Use of trade
names does not imply endorsement of the pro ducts named or criticism of similar ones not named
The authors thank the Bloomfest coordinators at Cobo Hall, Detroit, Michigan.

1Graduate Student

2Associate Professor. To whom reprint requests should be sent. E-mail address:

behe@msu.edu

3Professor

34

Marketing and Economics
A dditional index words. Survey, economics, conjoint analysis, cluster analysis, visual, package
size, price, color, pansy, Viola >< wittrockiana Gams. ‘Accord Banner Clear Mixture’
Abstract. Two surveys were conducted to determine characteristics important in containerized
edible flowers that could be sold in retail outlets. Self-selected participants at Bloomfest at Cobo
Hall, Detroit, Mich. , were assigned to one group that rated the importance of attributes such as
color of pansy ( Viola >< wittrockiana Gams. ‘Accord Banner Clear Mixture’), color combinations,
container size, and price. Participants assigned to a second group rated color, color combinations,
and container size. Flower color was allocated the most points in the purchasing decision (63%
for the first group and 95% for the second), with a mixture of all three colors (blue, yellow, and
orange) being the most desirable. Responses were subjected to Cluster Analysis (SPSS Inc. ,
Chicago, 111.), which resulted in the formation of three distinct groups. The groups were labeled
“Likely Buyer” (those who had eaten and purchased edible flowers before and rated characteristics
of edible flowers favorablY); “Unlikely Consumer” (those who had eaten edible flowers before and
had rated characteristics of edible flowers unfavorabIY); and “Persuadable Garnisheis” (those who
had not eaten edible flowers before, but were very likely to purchase edible flowers for a meal’s
garnish).
Introduction

Edible flowers can be used to add color, fragrance, and flavor to food such as salads,
soups, entrees, desserts, and drinks (Barash, 1998a, b). Since the late 19805, there has been a
resurgence in the popularity of edible flowers used by chefs and people entertaining at home.

Consumers have been exposed to an increasing number of edible-flower cookbooks, culinary

35

magazine articles, and television segments (Rusnak, 1999). Potential consumers who do not have
the space, patience, or time to grow edible flowers can purchase them.

Small, rigid, plastic packages of edible flowers are now available in some retail stores in
the United States. Discussions with produce managers have revealed that many of these packages
are not sold before the expiration date. Reasons for this may include flower color, species
assortment, package size, price, package design, or some other attributes. If unfavorable opinions
predominate, marketing efforts can improve an item’s appearance, taste, or portion size, or replace
it with amore desirable one. Although edible flowers are not one of the top sellers in the produce
department, a greater number of baby boomers are using them in entertaining (Rusnak, l 999).

To promote sales, retailers must educate customers about the uses of edible flowers and
their ability to enhance food at holiday dinners and other special occasions (Rusnak, 1999).
Successful marketing and advertising techniques used to promote sales of culinary herbs can also
be used for edible flowers. Rusnak( 1999) reported that consumers who purchase fresh herbs are
likely to purchase edible flowers. Signage, point-of-purchase material, and storage-condition or
recipe information included in the package can draw consumers’ attention and attempt to inform
them (Moore, 1998).

Marketers also need to understand potential consumers ’ needs and concentrate on
characteristics that should be included in the final product (Food Product Development, 1 979).
Conjoint analysis is a tool that helps researchers determine the importance of various factors that
affect the consumers’ inthe purchase decision (Behe et a1. , 1999; Gineo, 1990; Hardy et a1. , 2000;
Price et a1, 1980; Robertson and Chatfield, 1982; Shafer and Kelly, 1 986; Townsley-Brascamp

et a1, 1995). Marketers are interested in understanding the components of the edible—flower

36

package and determining changes that should be made to encourage other marketing segments to
purchase the product. To date, no data are available regarding consumers ’ preferences for edible
flowers’ color, package size, and price. The objective of this analysis was to address this
deficiency.

Materials and Methods

Two surveys were conducted at Bloomfest at the Cobo Hall Center in Detroit on 9 and
10 April 1999. Detroit was chosen as a survey cite based on articles that defined the Detroit
metropolitan area as a suitable test rmrket (WaldrOp, 1 992). Bloomfest, a highly advertized event,
allows garden enthusiasts to view the offerings of garden center exhibits, nonprofit booths, and
various vendors. Participants pay a fee to enter the 4-day event. Participants who agreed to
complete a survey were self-selected. Thus, this sample may be more reflective of a population
that is more interested in gardening and flowers than is the general public.

A nonorthogonal design was developed by using OrthoPlan, a computer software program
component of the SPSS software package (SPSS Inc. , Version 8.0, Chicago, Ill). The survey was
developedby using three single-color and four multicolor combinations of pansy (blue, yellow,
orange, blue and yellow, blue and orange, and all three colors); two sizes (nine pansies per
package; 8-oz. (227-g) container and 18 pansies per package, l6-oz. (454-g) container) of a
clear, rigid, plastic container, and three prices, basing the first price on packages seen in the retail
market by the researchers ($2.99). Doubling the $2.99 price would have created an increment
that was between what retailers consider apsychological price barrier (Mason and Mayer, 1984),

so it was reduced to the nearest perceived increment ($4.98) and doubled ($9.95). The three

colors of pansies were chosen to best reflect contrasting, natural colors. The total number of

37

possible container combinations (color x container size x price) was 42. An orthogonal
arrangement was developedwith Orthoplan (SPSS Inc.) by using 27 of the possible combinations
to reduce participant fatigue while providing data for the complete orthogonal design. Equal weight
was given to color, size, and price, and each color and color combination was featured.

Photographs of a single edible-flower package that measured 1 0.2 by l 5.2 cm accurately
and reliably portrayed the real objects (Daniel and Boster, 1976; Shafer and Richards, 1974).
These were glued onto foam core board and placed in a random layout on four adjacent boards
covered with black cloth. Each board measured 0.6 1 by 0.92 m and was placed on a metal easel
on one of the 0.61 msides. The four boards were placed on a 0.6 l by 1.83 mtable along the 1.83
m side. On the two outer boards, six photographs were arranged at equal intervals. On one of
the two inner boards, seven photographs were arranged. Eight photographs were arranged on the
other inner board.

A description of the number of flowers in the package and the price was affixed to apiece
of foam core board 5. 1 cm tall and 1 5.2 cm wide and placed under each of the photographs. On
the upper right corner next to the photograph, black, vinyl, self-sticking letters fi'omA to AA were
placed on 3 . 8-by-3 . 8-cm pieces of foam core board. Each letter corresponded to an item on the
survey. Small pieces of velcro on the back of each picture, description, and letter board secured
the foam core board to the black cloth board.

Expt. 1. The arrangement of the pictures on the board corresponded with a schematic that was
placed on a two-page survey sheet. In each box of the schematic, a letter from A to AA
corresponded to the letter next to the pictures on the board. The experiment was a lOO-point

exhaustion designed to determine which color or color combinations, price, and size the

38

 

 

Pam
Pant
fog-i
111th

m

r1
(8

(If)
(.1

Ti

participants would choose to purchase if shopping for a package of edible flowers to serve to
family and friends. Participants were told to assign the 100 points to the containers of edible
flowers, giving their favorite the most points, and to continue to assign points until all had been used
Participants did not need to assign points to all the containers, but just the ones they liked.
Participants were asked 2 1 questions, including which edible flowers they had eaten, what their
food shopping and gardening habits were, and their demographic status. Demographic questions
included year ofbirth, gender, education level, income (using nine discrete $20,000 categories),
family status, number of persons in the household, and zip code. Participants answered questions
regarding purchasing edible flowers a.) to be used for garnishes and salads, b.) that were insect—
damaged, and c. that were grown without pesticides. A semantic differential on a seven-point
Likert scale (7 being the highest rating) was used.

Expt. [1. Another arrangement of the pictures was used to ask participants who did not participate
in Expt. 1. about color, color combinations, and container size. A total of 14 pictures were
randomly arranged on the black boards. A description of the package size was placed under the
picture and a letter from A to N was placed to the upper right of the photograph. Eighty-six
participants filled out the survey and a schematic that corresponded to the arrangement of pictures
on the board was on the survey form. Participants who rated these photographs also used the
lOO-point exhaustion process and answered the 21 questions asked in Expt. 1. Participants’
answers to both 1 00-point exhaustion surveys were analyzed with Conjoint Analyzer software

(SPSS Inc); other questions were analyzed with a two-tailed t-test and the Kruskal-Wallis test.

39

Results

Expt. 1. Participants who viewed and rated the 27 pictures of containers indicated that color was
the rrrost influential factor (63% of the decision) when deciding which package they would purchase
(Table 1). Price was the next most important factor (24%), with container size having the least
importance ( 13%). The utility for all factors (Table 1 ), indicates a preference that is more or less
over an average or ideal. A negative utility for price indicates that participants valued the container
less than an idea] or average because of its higher price. The container combinations of color,
price, and size were ranked by points allocated by the participants. The nine-count container with
all three colors and a price of $2.99 received the most points, with amean of 13. 1 . The 18-count
container with all three colors and a price of $4. 98 received the next highest number, with anrean
of 1 l .7, followed by the 1 8-count container with blue flowers and a price of $4.98 , with a mean
of 7.0.
Expt. [1. Ratings trade by participants who rated containers by color and package size alone were
similar to those of the group that rated the containers on all three attributes. Again, color was the
most important characteristic in the decision (95%) in choice of container (Table 1), whereas
container size had little (5%). Again, the nine-count container with all three colors received the
most points, with a mean of 18.2. The ranking of the second and third containers, however,
differed from those chosen by the group that viewed all 27 photographs. The second highest
number of points was assigned to the 1 8-count container with blue flowers, with a mean of 13.7,
followed by the nine-count container with blue flowers.

To increase the sample size for segmentation analysis, the similarity ofboth groups was

tested. Considering six demographic questions we decided to combine samples if there were no

40

differences on four of the six demographic questions (Behe et al., 1999). Kruskal-Wallis tests
revealed that only gender distribution was significantly different, so samples were combined for a
cluster analysis on common variables. Of the 224 participants, 146 had eaten them before, and
33 of them had purchased edible flowers before. The participants had eaten a total of 2 1 different
species, including nasturtium (Tropaeolum majus L.) (eaten by 45% of participants), pansy
(28%), and violets (Viola tricolor L.) (21%). A few of the participants reported that they had
eaten flowers that are not recognized as being safe, which supports the current belief that
consumers should be told what is edible and whit is not (Barash 1998a, b). Participants were
asked how much they would pay for a nine-or 1 8-count container if they purchased it in a well-
known grocery store. For a nine-count container, 3% would pay up to $1 0, with 27%, the highest
percentage of participants, willing to pay $2. 99. For the l8-count container, 2.9% of the
participants would pay $15, with the greatest percentage of participants, 20%, willing to pay
$5.00. These results are consistent with those of Expt. 1.

Participants were asked about their purchasing habits. We hypothesized that if they were
likely to purchase prepared salad mixes, they might purchase other packaged iterrs or salad mixes
with edible flowers as an ingredient. Edible flowers can add value to the product and increase the
price that can be charged for the mix. Participants purchased an average of two salad mixes a
month and were more likely to eat salad in the summer (9 1 %), but were more likely to purchase
separate salad ingredients (70%) than salad mixes (30%).

We asked the participants to answer questions about their gardening habits in a broad
sense, including lawn care. They reported that they either did not garden (4.3%), or gardened

between 1 and 9 (45%), 10 and 19 (30%), 20 and 45 (20.2%), or 54 h each week (0.5%). Of

41

those who gardened, their gardening area averaged 16. 3% vegetables, 5 1 .6% lawn, and 40. 7%
flowers. Participants who are active gardeners and grow vegetables maybe arrrarketing segment
that could be targeted. If edible-flower consumers grow vegetables, they rmy grow edible flowers
for consumption.

Pest management strategies may impact consumer preferences for edible flowers. The
percentages of participants who were very likely to purchase edible flowers was 62% if they were
grown pesticide free, 52% as a garnish for a meal, and 47% to eat in a salad. Only 14% were
likely to purchase edible flowers if they had 1 0% insect damage, and 75% were very unlikely to
purchase such flowers.

Cluster analysis (SPSS Inc.) was used to determine whether meaningful customer
segmentations, based on participants’ answers to several questions, could be created. These
groups could then be targeted by producers or retailers. Variables were used for clustering based
on attitudes about edible flowers. By using K-Means, clusters of size 2, 3, and 4 were examined
by using six cluster algorithms. After examination of each cluster size, the three-cluster solution was
selected to develop the customer market segments. Because of nonresponse of several
participants to several questions, only answers from 175 of the 224 participants could be used to
create the segments.

Of the three segments that were created, the one labeled “Likely Buyer” comprised 6 l %
of the sample (Table 2). The “Likely Buyer” had eaten edible flowers before, but had not
purchased them before taking part in the survey. “Likely Buyers” were likely to purchase edible
flowers for use as a garnish and were very likely to use edible flowers in a salad; 80% were very

likely to purchase edible flowers because they were grown pesticide free. They showed a

42

significantly greater preference for pesticide- free products and were more willing to buy insect-
darrraged products than the other two segments.

“Likely Buyers” spent amean number of 12 h in the garden each week, with 40% of their
garden planted with flowers, 1 7% with vegetables, and 53% with lawn. This segment awarded the
nine-count container of flowers with all three colors the highest mean number of points ( 1 8) which
was significantly different from zero. When the demographics of the segments were analyzed,
groups were similar with regard to education level, family size and status, income, and number of
dependents. The group of “Likely Buyers” contained a greater percentage of males than did
“Persuadable Garnishers.”

The second segment, “Unlikely Consumer,” contained 23% of the sample. They had eaten
edible flowers before, yet had never purchased them. However, they were unlikely to purchase
edible flowers for a garnish (93%) or a salad ingredient (73%). When corrrpared with the
combined responses of the other two segments, their response was significantly different for the
question of purchasing edible flowers as a garnish.

“Unlikely Consumers” were people who were the least “very likely” to purchase edible
flowers because they were grown pesticide-free (26%). This response was also significantly
different from the combined responses of the other two segments. Only 3% of the respondents
were very likely to purchase the flowers with 10% insect damage. This segment included
individuals who would also pay the least amount of money for an 1 8-count container of edible
flowers at a major grocery store chain. Other differences between the “Unlikely Consumer”

segment and the other two were they spent the least time gardening each week (8 h), allocated the

43

least garden space to flowers (37%), and assigned the least amount of points to the nine-count
container of edible flowers with all three colors (10).

The third segment, “Persuadable Garnishers,” comprised 17% of the sample. The
participants in this segment had never tasted edible flowers but would be very likely to use them
as a garnish. Forty-six percent of the “Persuadable Garnishers” segment would be very likely to
purchase edible flowers because they were grown pesticide-free and, as with the “Unlikely
Consumers,”only 3% would be very likely to purchase edible flowers with 10% insect damage.
Participants in this segment were likely to pay the highest mean price for a nine—or 18-count
container of edrble flowers at a grocery store, ($4.20 and $7 .42, respectively). This group hadthe
highest percentage of females. “Persuadable Garnishers” purchased a mean number of three
packages of salad mixes in a month vs. two for the other two segments.

Marketing implications

Results of this study suggest there are several approaches that edible-flower producers
must use to effectively rmrket the product. Eating food involves more than simply oral senses,
flavor, and texture; it also involves color, appearance, and stimulation of the visual senses (Little,
1980). We eat first with our eyes and then our mouths. Food must be visually appealing if
consumers are expected to eat it. Consumers are likely to reject food based on appearance, even
if the food is considered appealing and pleasing (Little, 1 980). Food is selected on variables such
as color and defects or the lack thereof (Little, 1980). This selection may be done consciously;
however, color and similar characteristics are associated with every object that consumers see in

their daily lives (Food Product Development, 1 979). Choosing which color of edible pansies to

44

sell is not atrivial matter. If the wrong color is used, consurrrers rmy be less likely to purchase the
product, and the package may remain on the shelf until the product is no longer marketable.

The same attention must be paid to color combinations in the package. The potential
consumer’s preference for a package that contains one vs. three colors of pansy must be
determined. This preference may be based on the situation. S everal participants stated that they
would purchase a certain color or color combination of edible flowers to use as decoration on a
cake, while they would use a completely different color or color combination in a salad. Results
here suggest that a mix of flower colors is preferred over single colors, and some specific color
contrasts preferred over others. The mix of blue, yellow, and orange pansies was much more
highly valued than any other combination or single color. Research should enrphasize what situation
is more likely to occur or whether equal numbers of packages would be purchased in either
situation. The greater the understanding of what potential edrble-flower consumers want in their
proposed packages, the greater the probability that packages will be purchased.

Defining the consumer segments with cluster analysis makes it possible to target those
segments that are more likely to purchase edible flowers, and is an effective use of marketing
dollars. When targeting potential consumers, limited resources, such as time and money, may be
used on the consumers who fit this profile only. Coupons or special advertisements might be
rmiled or otherwise distributed to “Likely Buyers” who have not purchased edible flowers before,
but are very likely to purchase the edible flower for use as a garnish or in a salad. The coupons
and other inducements may persuade these potential consumers to purchase the item.

Similar marketing efforts can be used to increase the number of packages of edible flowers

purchased by segments such as “Persuadable Garnishers Though the participants in this segment

45

have never eaten edible flowers, they are very likely to purchase edible flowers for use as a garnish
and could be as valuable as customers as those in the “Likely Buyer” category. By purchasing the
flowers for a garnish, those in this segment might experiment with the edible flowers and begin to
use them in food items, possrbly becoming “Likely Buyers” more willing to purchase the flowers
more often. Edible flowers can be packaged and promoted as a garnish, much as baking soda is
used as a refiigerator deodorizer rather than in baking, its traditional use.

In order to convince the “Unlikely Consumer”to buy edible flowers, marketers may have
to spend a great deal of time and energy, which may not be an effective strategy. Conversations
with several gardening groups indicate that personal production of edible flowers may be more
desirable than purchase of them This segment may purchase edible flowers for applications other
than use in salads and garnishes. Future marketing studies will be implemented to address such
issues.

Growers should be aware of consumers’ attitudes toward the use of pesticides on food
items and the consumers’ greater inclination to purchase edible flowers grown pesticide-free.
Although this sample was drawn from what appeared to be a population interested in gardening
and flowers, gardening is the hobby in which the largest percentage of Americans engage (Gallup
Organization, Inc. , 1999). Consumer’s opinion and feelings about the use of pesticides coincide
with current production trends and edible-flower growers are aware that consumers prefer
pesticide-free products (Whitman, 1991). Each year, more food products are grown using organic

methods (DiMartino, l 999), and the demand for such items will continue to grow in other areas
of production (DiMartino, 1999). Edible flowers must not be sprayed with chemicals

( Kosztolnyik, 1996), since no pesticides are labeled for their production.

46

Literature Cited

Barash, C. W. 1998a. The flavors of flowers. Herb Companion 10(4):32-37.

Barash, C. W. 1998b. Please eat the flowers. Horticulture 95(5):36-40.

Behe, B. , R. Nelson, S. Barton, C. Hall, C.D. Safley, and S. Turner. 1999. Consumer preferences
for geranium flower, color, leaf variegation, and price. HortScience 34:740-742.

Bunn, D., G. W. Feenstra, L. Lynch, and R. Sorrrrner. 1 990. Consumer acceptance of cosmetically
imperfect produce. J. Consumer Affairs 24(2):268—279.

Daniel, TC. and RS. Boster. 1976. Measuring landscape esthetics: The scenic beauty
estimation method. US. Dept. of Agr. For. Serv. Res. Paper RM-167.

DiMartino, C. 1999. Organics growing in foodservice. Produce Business 15(3):41-42.

Food Product Development. 1 979. Market research system keys on product “personality Food
Prod. Dev. l3(4):54-55.

Gallup Organization, Inc. 1999. National gardening survey 1998-1999. The Natl. Gardening
Assoc, Inc. Burlington, Vt.

Gineo, W. M. 1990. A conjoint/logit analysis of nursery stock purchases. Northeastern J. Agr.
Resource Econ. 19:49-58.

Hardy, J. , BK. Behe, T]. Page, Jr., and R. Schutzki. 2000. Consumer segmentation based on

perceived plant knowledge and gardening involvement. Southern Nurserymen’s Assoc. Res. Conf.
Proc. (In press.)

Kosztolnyik, L. 1996. Selling edible flowers. Natural Food Merchandiser 17(7):74.

Little, AC. 1980. The eyes have it. J. Amer. Dietetic Assoc. 77(6):688-691.

47

Mason, J. B. and M. L. Mayer. 1 984. Modern retailing: Theory and practice. Business Publ. Plano,
Texas.

Moore, C. 1998. Sales can rise with education. Packer lO6(25):6D-7D.

Price, T.L., J .L. Robertson, and L.H. Chatfield. 1980. Factors affecting the marketability of roses.
J Amer. Soc. Hort. Sci. 105:388-393.

Robertson, J.L. and L.H. Chatfield. 1982. Fresh flower merchandising in loose bunches.
HortScience 17:593-595.

Rusnak, J. 1999. Overcoming organic supply issues. Produce Business 15(6):60, 63-64.

Shafer, B. S. and J. W. Kelley. 1 986. The influence of cultivar, price, and longevity on consumer
preferences for potted chrysanthemurns. HortScience 21: 1412-1413.

Shafer, EL. Jr., and TA. Richards. 1974. A comparison of viewer reactions to outdoor scenes
and photographs of those scenes. US. Dept. of Agr. For. Serv. Res. Paper NE-302.

SPSS Inc. 1998. SPSS Base 8.0 for Windows User’s Guide. Chicago, Ill.

Townsley-Brascamp, W. , N. E. Marr, E. Matsuo, and PD. Relf. 1995 . Evaluation and analysis
of consumer preferences for outdoor ornamental plants. Acta Hort. 391 : 199-206.

Whitman, A. T. 1991 . Edible flowers and culinary herbs: New uses for traditional crops, new
crops for traditional growers. GrowerTalks 54(13):22-23, 25, 27, 31, 33.

Waldrop, J. 1992. All-American markets. Amer. Demographics l4(l):24-30.

48

Table 1. Conjoint analysis of two groups of participants who viewed and rated either: 1.) flower
color, color conbination, container size, and price in 27 photographs or 2.) flower color, color
combination, and container size in 14 photographs.

 

Factor Relative importancez Utilityy
Group 1
Price (8 US.) 23.89
2.99 1.36 *
4.98 0.52 NS
9.95 -].87 *
Size 12.77
9-count container -0.86 *
18-count container 0.86 *
Color 63.34
Blue -0.87 *
Yellow -3.12 *
Orange -2.75 *
Blue and yellow 0.81 NS
Blue and orange 1.21 *
Yellow and orange -0.73 NS
All three colors 5.45 *
Pearson’s R = 0.919 Significance < 0.00001
Group 2
Size 5.04
9-count container 0.28 NS
18-count container -O.28 NS
Color 94.96
Blue 4.67 * '
Yellow -4.17 *
Orange -4.70 *
Blue and yellow -0.33 NS
Blue and orange 1.51 NS
Yellow and orange -2.95 *
All three colors 5.98 *
Pearson’s R = 0.870 Significance < 0.00001

 

2A higher value indicates a greater importance.
yA more positive value is more desirable.
”5' ' Nonsignifrcant or significantly different from 0 at P s 0.05, as based on a two—tailed

t-test.

49

m.m ... we om ow A85 8de283 550%
883 8% 8:83 2830: 83:8 88:83 8 box: bo>

 

o o a 385 as,
R a 2 bag 388882
a ... Q 8 o 285::
28.83%: 355. e .Bsx resobxo‘efim 3.535% 2 E35
an o 5
z a E be": es>
an .. o 8 m be: 8858:
38%:
32: e .6.\ 525% e aekfeioKmEfim 3.33:3 2 28::
me we we
m.m ... N m N Axe Ao: n m .8» n 3 283 2830c 85:8 88:83;
on me ow
m: ... N ~ ~ .x. Ao: n m .8» H C 8&3 8830: 83:8 :88 88$
2 mm ow .x.
om ov me— e
Noofiommemmm cor—EEO 85:28.0 893m 335 288$
83.8888 385.5

 

“85mg 8:25:00

 

coca-cameo“. mew—8 wen 2830c 83:8 Sons 895:8
$56305 £38.52, 9 880%»: £89698 :0 882 $588 8326 EB.“ 83.8w 388mg 863:8 8:: Mo 2059589 .N 055.

50

m2

m2

m2
m2
m2

m2
m2

m2

Wm ...

2

mm
@—
mm

S
.235

om."V

O—

on
5
NM

NVN

ohm

E

mm
2
cc

~—

how

ooh

on

3:88 88.55
2: wctsv 37% 208 “mm Aoc n N 8% n 3
H8 38:4

8830: 30:8.»

Aowcfio 98 56:8» “2:3 8200 830a ooh: =<
“:23 858:8
€28-85: m 8 @2823 353 38 go 898:: :82

:33
85m8mo>

@8263
HE 52mm mo “8ch AR; 8.8 :82

$5 x83 :08 mam—82% E27. 2:: :82

258$ _

“550$
E20 89m buoohw SEE m E 8250c 85.68
mo 853:8 a 5m >3 8 9:53 6WD 3 8E :82

AR; 0mg
835 $3 53» Boaoc 83:8 88:83 8 box: bo>

.8283 N 2an

51

::8£:m_m 808388

8820 =m H w J umfimwm :28: :8 8:568 m :8 N H m ,N 85%: :38: :8 8:588 m :8 _ u N .m 85mm: :38: :8 8:588
N :8 _ 8820 n _ .63 £__m>>-_mxm3¥ :8 H83 8:863: m :o 885 8 mod U Q a: “88.5%: AL bugumEmG :o AmZV EnemiwficoZN

 

m2
m2
m2
m2
m2
_ ...
m2
m2

m2

mm
mm
mN
mm
3
no

NV

mm
on
NN
m:
vw
mm.

_m

m:
we
NN
mm
mo
w:
w:

$5888: 53» Ax; 3:88:89
38:88 XL maflm

88.8-08de £3 5 2:85
A858: 025 5:? £3 86 538m
Amman—88% 58:88:08 Ax; 8588‘”.—
EE: 8 520:

A525 0?»

5:9:
m :_ 8858:: 888 :38 .5 888m: .5 85:5: :82

8x5: 8.3 888—885 9m

.5253 N 2%:

52

SECTION III

CONSUMER RATINGS OF EDIBLE-F LOWER
QUALITY, MIX, AND COLOR

53

Consumer Ratings of Edible-Flower Quality, Mix, and Color

Kathleen M. Kelleyl

Michigan Agricultural Experiment Station and Departments ofHorticulture and Botany & Plant
Pathology, Michigan State University, Plant and Soil Sciences Bldg. , East Lansing, MI 48824-
1325

Bridget K. Behe2 and John A. Biernbaum3

Michigan Agricultural Experiment Station and Department of Horticulture, Michigan State
University, Plant and Soil Sciences Bldg, East Lansing, MI 48824-1325

Kenneth L. Pof‘lfl

Michigan Agricultural Experiment Station and Department ofBotany & Phnt Pathology, Michigan

State University, Plant Biology Bldg, East Lansing, MI 48824-1325

 

Received for publication . Research conducted at Michigan State University, East
Lansing, Michigan. Use of trade names does not imply endorsement of the products named nor
criticism of similar ones not named. The authors thank the Garden Days and Master Gardener
coordinators at Michigan State University for permission to conduct these surveys.

1Graduate Student

2Associate Professor. To whom reprint requests should be sent. E—mail address:
behe@msu.edu

3Professor

54

Marketing and Economics

Additional index words: survey, economics, visual, perception, damage, Viola X wittrockiana,
Viola tricolor, Borago officinalis, Tropaeolum majus, and Begonia >< tuberhybrida.
Summary. Two identical surveys were conducted with separate samples to determine consumer
perceptions of quality of five edible flower species. Participants were either members of a class
which reviewed the history and uses of edible flowers at Garden Days (n = 23) or were Michigan
Master Gardeners (n = 5 l) attending a similar class. Participants were shown a randomized series
of projected photographic slides of five edible flower species and asked to indicate whether or not
they found the flower quality acceptable. The slides depicted a range of ratings of mechanical
damage, insect damage or flower senescence of the taxa on a scale (l-S) developed by the
researchers. A flower that was rated 5 was flawless while a flower that was rated 1 had substantial
damage. Nearly half of all participants had eaten edible flowers prior to the study and 5 7 to 59%
had grown them personally for their own consumption, indicating many had previous direct
experience. Both samples rated flower quality equally. Analysis showed that Viola
><wittrockiana Gams. ‘Accord Banner Clear Mixture’ (pansy), Begonia >< tuberhybrida Voss
‘Ornament Pink’ (begonia), and Viola tricolor L. ‘Helen Mount’ (viola) were acceptable from
stage 5 to 3. Both groups rated the Tropaeolum majus L. ‘Jewel Mix’ (Nasturtium) flowers
acceptable at only stage 5 . Garden Days participants rated Borago ojficinalis L. (borage) and
results showed it was acceptable from stages 5 to 3, while the Master Gardeners results showed
acceptability from stages 5 to 4. Participants also rated flower color (yellow, orange, and blue),

with results showing all colors equally acceptable.

55

Introduction

Rigid plastic containers of edible flowers can now be found in some grocery store produce-
departrnents. Though the number of containers purchased is small, upscale clientele and caterers
will likely purchase edible flowers as an ingredient or garnish for meals, especially at holiday times
( Rusnak, 1999a). For the hotel and restaurant market, Quail Mountain Herbs, in Watsonville,
CA, seasonally produces 50 varieties of edible flowers including nasturtium, pansies, and
chrysanthernums (Hunn, 1999). While larger quantities and varieties are offered to chefs,
consumers may purchase smaller quantities of a mix of edible flowers.

Consumer perception of product quality is ofien the factor that creates customer loyalty
and the continued purchase of that product (F oodservice Equipment and Supplies, 1 999). Food
service venues such as restaurants strive to improve or keep food quality high. When choosing a
restaurant, food-quality is the most important factor among consumers that were surveyed (Dulen,
1999). For instance, a restaurant’s reputation can be tarnished if the coffee served is of poor
quality (Bendall, 1999). With a product, such as edible flowers, which maybe placed on a plate
as a garnish and/ or eaten, flower quality may be as important as the coffee that is served. If the
flower isn’t appetizing, it may reduce visual appeal of the meal and may determine whether a
person will eat it or not (Little, 1980).

In the floriculture industry, quality is also essential for customer satisfaction. Consumers
want more varieties of top quality plants with a longer shelf-life (Shaw, 1998). As with most
products, if customers do not feel that they are purchasing ahigh quality item, they may choose a
competitor’s product. Customers are ofien willing to pay more for ahigher—quality product (Shaw,

1998). It is reasonable to assume that packages of edible

56

flowers should contain a desirable mix of species that are palatable, have a reasonably long shelf-
life, are of high quality, and contain information such as how to store and use the product.

For edible-flowers, the product qualitymust begin with planing during the production stage.
Since no pesticides are registered for use on edible flowers (Kosztolnyik, 1 996), other alternatives
must be used to control pests. With an increase in the demand for organic products, edible flower
producers may consider growing them using certified organic methods. The supply of organic
products as well as product-quality has improved over the past few years (Rusnak, 1999b), but
product-quality must continue to meet or exceed consumer’s expectations to be profitably
competitive with conventional products.

Delicate edible flowers should be protected from damage with proper packaging. They
can be marketed in rigid plastic containers similar to containers used to store and protect
strawberries and other highly perishable items. Not only do packages protect the easily darmged
items, but they also reduce condensation when used to package strawberries (Fite, 1 998). Even
if all precautions are taken to protect the product, store managers should monitor product quality
and remove any containers of edible flowers from the shelves once noticeable deterioration is
apparent. Prior research has shown that edible-flower color is the most important product
attribute, followed by container price and size (Kelley et al., 2000b). Further, amix of species and
cultivars was preferred to a single color or species (Kelley et al. , unpublished data), but it is not
know the extent of deterioration that consumers will tolerate in a mix of edible flower species? The
researchers hypothesized that various species of edible flowers could be sold through grocery

stores with some minimal damage, and still be perceived as acceptable.

57

Our objective was to provide information to help answer questions regarding consumer

preferences and quality perceptions of edible flowers.
Material and Methods

Two major Michigan cities, Detroit and Grand Rapids, are considered variable test
markets (Waldrop, 1992). This means that their populations closely resemble an “average” United
States city. Results collected in test markets are often extrapolated to other typical cities as one
indication ofhow well the product tested might be perceived. We were more interested in how
a sample of cons umers who may be more experienced with flowers , and perhaps more critical of
flower quality, might rate the edible flowers. Less critical consumers should have similar or lower
expectations. Master Gardeners (MG) are more interested and/ or experienced in gardening with,
perhaps, a heightened sensitivity and awareness of flower quality. MG participants learn about
various aspects of gardening and receive certification sponsored by Michigan State University
Extension Office. Garden Day (GD) participants pay to attend an annual event and learn about
gardening related subjects. Participants are self-selected to attend gardening programs because
of their interest in gardening, with a similar level of flower awareness. Both groups were Michigan
residents, some of whom resided in test market areas. Twenty-nine percent ofMG were residents
of Metro Detroit with an additional 1 5% residing in the Grand Rapids metro area. Twenty-two
percent of GD participants were residents ofMetro Detroit, none were from Grand Rapids. The
remainder ofMG and GD participants were fromthe Greater Lansing Area, cities surrounding the
two metro areas, and states bordering Michigan. While drawn from cities identified as

representative we found these samples to be an indication, perhaps more critical

58

than what we might expect of groups of “average” Americans who are less involved in gardening.

The Michigan MG participants attended atwo-day annual prograrnthat allowed members
to enroll in classes on various gardening-related topics. Fifty-one self-selected members
preregistered and attended one of two l-h 30-min seminars on edible flowers on 29 June 1999.
GD participants also had the opportunity to attend other seminar topics during Garden Days, a
two-day annual meeting. Twenty-three self-selected GD participants enrolled in an edible-flower
seminar on either 5 or 6 August 1999. All participants were given a survey form that included a
scale for visual quality assessment and several other questions pertaining to edible flower
preferences, uses, and demographic questions.

Participants were shown 25 projected 3 5 mrnphoto graphic slides (with aprojected size
of 1.22m x 1.73m) of five edible flower species: Viola xwittrockiana ‘Accord Banner Clear
Mixture’ (Pansy), Begonia ><tuberhybrida ‘Omament Pink’ (Tuberous Begonia), Viola tricolor
‘Helen Mount’ (Viola), Borago officinalis (Borage), and Tropaeolum majus ‘Jewel Mix’
(Nasturtium). Each slide showed a flower with a quality rating on a scale from 1 to 5. Each stage
of the scale corresponded with a postharvest visual assessment scale that was developed by the
researchers because no damage assessment scales for edible flowers were located in the literature.
A flower rated a 5 was flawless (Figure 1). As the numerical rating decreased, the amount of
damage or flaws increased. Flowers rated a 1 had the greatest amount of damage from either
mechanical damage, insect darmge, or senescence (Figure 2). Flower color was also affected and

flowers that had ratings lower than 5 had increasing degrees of petal discoloration.

59

1‘

As each slide was shown, participants were asked to indicate on their survey form whether
or not the flower quality was acceptable. Each slide was shown a second time, in the same order,
to insure that participants had an adequate opportunity to examine each flower carefully. Our
hypothesis was that consumers would consider flower quality acceptable from stages 5 through 3.
Damage was rated on a 5 to 1 scale, 5 = no visible blemishes. Based on the amount and types of
damage, we hypothesized that flowers that received a rating of either 1 or 2 would have a
substantial amount of damage and significant browning to be unappealing to consumers. After
viewing slides, participants were asked to examine and rate slides of three different colored pansies
(yellow, orange, and blue), using a seven point Likert scale (7 == highest rating). They were asked,
based on flower color, how likely they would be to eat the pansy shown. Participants then rated
how likely they would be to purchase edible flowers if grown organically, purchase flowers with
10% insect damage, if they had ever eaten edible flowers before, how they obtained the edible
flowers, and if they preferred mixtures of edible flower species or one single species. Finally,
participants responded to aseries of demo graphic questions. Results were tested for significance
using the SPSS software for Windows 95 program (SPSS Inc. , Chicago, IL).

We hypothesized that Garden Days participants and the Master Gardeners would
tolerate a similar threshold of edible flower damage before indicating an unwillingness to purchase
a container of edible flowers because they had similar experiences with edible flowers and similar
demographic characteristics. Further, the researchers hypothesized that, if at least 50% of the
participants found a flower acceptable at a particular rating, that the flower would likely be

saleable. We would recommend that packages of edible flowers

remain in the store until the threshold rating was reached. By leaving the packages of edible

6O

flowers in the store longer and not discarding them at the first sign of a minor change in appearance,
store managers may be able to sell more packages if they remain in their display area longer. This
would reduce product shrinkage and enhance profitability. We would recommend that store
managers not leave the packages in the display area past the acceptable stage. Consumers should
only see packages of edible flowers that are at or above the threshold level.

Results and Discussion

The demographic characteristics ofboth groups were similar except that 23% more MG
participants were married (Table 1). Nearly half of both groups indicated that they had eaten
edible flowers in the past 3 months; namely pansy, dandelion, nasturtium, lavender, chives, violet,
borage, mint, tulip, calendula, arugula, and Signet marigold. This result indicated that participants
had some recent experience with the product, perhaps sensitizing them to the product attributes.
Only 1 2% of the participants in both groups purchased edible flowers in the past year, while slightly
more than half GD participants and MG participants, respectively, were growing edible flowers for
personal consumption. More than half of both groups preferred mixtures of edible flowers. A
good percentage ofboth of these groups had prior experience with edible flowes and could assess
damage with a critical eye.

More than half ofboth MG and GD participants rated flower quality of pansy, tuberous
begonia, and viola acceptable at stages 5 through 3 (Table 2). Less than half ofboth groups rated
nasturtium acceptable at stages lower than 5 . Flower quality ofborage was acceptable at stage
3, for half of the GD participants but for only 47% of MG participants. W e a s k e d
participants to answer a variety of questions regarding their preferences for edible flower color

(Table 3). For all three single pansy flower-colors (yellow, orange, and

61

blue) participants in both groups were very likely (composite rating of 5, 6, or 7), to eat them all.
Behe et al. , ( 1 999) found there was a group of consumers who may purchase blue geraniums. It
may be possrble that consumers would prefer blue flower-colors for edible flowers while others
may see this color as less appetizing. At least 86% of participants responded that they would
purchase edible flowers if grown organically, however, only 30% stated they would purchase
edible flowers with 10% insect damage. Information such as this will aid growers who will be
making decisions as to what colors to include in the package and how the flowers should be grown.
Future studies will help determine if consumers can detect small amount of damage on edrble
flowers in packages.

Based on these data from these perhaps more highly-sensitized customer groups, we
see that potential consumers may purchase edible flowers at quality levels that are less than perfect.
However, there is a likely lower-threshold that below which flowers would not be purchased. We
can conclude that flowers do not have to be perfect but may be sold with minor flaws, even to
consumers with some experience growing and/ or eating flowers.

Information that is generatedwith regards to edible flower quality will help producers and
marketers create mixes, package sizes, and pricing techniques that may increase market share and
possibly boost profits. Participants in several marketing surveys have indicated that there is an
interest in edible flowers and that potential consumers care about flower quality as well as how the
flower is grown and treated (Kelley et al., 2000a; Kelley et al., 2000b; Kelley et al., In
Preparation). Few companies are producing edible flowers to be sold through retail outlets. As

the interest in edible flowers as a garnish and as an ingredient increases, new mixes, new color, and

62

larger quantities may be sold Edible flowers may also have the potential as being included in ready
to go meals or prepackaged salad mixes.

Other factors should be considered when creating edible flower products. The mix of
edible flowers may be a factor in the purchasing decision as well as the flower- quality. Robertson
and Chatfield( 1982) found that flower bouquet composition affected a consumers ’ preference.
Little is know regarding which species or mix of edible flowers consumers would prefer and how
quality affects the purchasing decision (Kelley et a1. , 2000a and unpublished data). Currently
marketed containers of edible flowers contain several species. It is possible that consumers may
prefer a more homo genous or heterogeneous rrrix. The desiredmix of edible flowers may also be
situational, whether the flowers will be used as a garnish or in a food item. Future studies should

concentrate on gaining information in these areas.

63

Literature Cited
Anonymous, 1999. Keys to maintaining customer loyalty. Foodservice Equipment & Supplies
52(6): 1 10.
Bendall, D. 1999. Coffee, tea and espresso. Restaurant Hospitality 83(5): 129-136.
Behe, B., R. Nelson, S. Barton, C. Hall, C.D. Safley, and S. Turner. 1999. Consumer preferences
for geranium flower color, leaf variegation, and price. HortScience 34(4):740-742.
Dulen, J. 1999. Quality control. Restaurants & Institutions lO9(5):38-52.
Fite, S. 1998. Companies put new twists. The Packer September 7 page 18.

Forkush, RH. 1999. Exotics - What to do with them. Produce Business lS(8):29-30.
Kelley, K.M., B.K. Behe, J.A. Biernbaum, and KL. Pofi. 2000a. HortScience (accepted for
publication). Consumer and chef perceptions of three edible flower species.

Kelley, K.M., B.K. Behe, J.A. Biernbaum, and KL. Poff. 2000b. HortScience (accepted for
publication). Consumer preference of edible-flower color, container size, and price.

Kelley, K.M., B.K. Behe, J.A. Biernbaum, and KL. Poff. (In Preparation) Combinations of
edible-flower species and colors in container effect consumer preference.

Kosztolnyik, L. 1996. Selling edible flowers. Natural Food Merchandiser 17(7):74.

Little, A.C. , 1980. The eyes have it. Journal ofThe American Dietetic Association 77(6): 688-691.

Robertson, J.L.,and L.H. Chatfield. 1982. Fresh flower merchandising in loose bunches.
HortScience 179(4): 593-595.

Rusnak, J. 1 999a. Edible flowers, fresh herbs, baby vegetables: Still room for growth. Produce
Business 15(3):33-34, 36-37.

Rusnak, J. 1999b. Overcoming organic supply issues. Produce Business. 15(6):60, 63-64.

64

Shaw, J. 1998. In my opinion: Consumers want value. Green Profit. January/February254

Waldrop, J. 1992. All-American markets. American Demographics l4(1):24-30.

65

“-1-

Table 1 . Comparison of demo graphic characteristics and edible flower preferences for Master
Gardeners and Garden Days partiCJiJants.

 

Master Gardener Garden Day
Attribute participant participant Significance
Mean Age (years) 45 46 NS
College or tech. graduate
(percent) 53 59 NS
Two person families 36 42 NS
Income $40,000 to
$60,000 (percent) 2l 37 NS
Married (percent) 82 59 *
Female (percent) 86 91 NS
Eaten edible flowers before
(percent) 44 _ 45 NS
Purchased edible flowers
before (percent) 12 12 NS
Grew edible flowers before
(percent) 57 59 NS

Preferred a mixture of
edible flower species
(percent) 65 68 NS

NS , * Nonsignificant or significantly different at P s 0.05 in rows, based on a two-tailed t-test and
chi-square.

 

66

wugéoa mocmufim .mod w m 3 A30 .595 .anwon 3832 .353 830: 036.... some .8 Eobbfi 3:8

amok £33-3me 2: was:

owtcmi co EmoEcwfico Z ... .mZ

 

 

...w on mZo o mZo m mZS mm mZ N v _

*N 5 m2 0 0 m2 3 mm mZ v o mZE om N
m2 3 mm m2 2 m mZ cw om mZ mm ow ... K mm m
m2 I. am m2 vm mm ... E E. mZ mm 00 m2 cm 2: v
m2 No oo mZ mm 2. m2 2: 00— m2 co co mZ cm 00— m

02 no 02 no 02 no 02 no 02 no madam
omeaQ
mwflom Bataan Z «Eowom £5833. amen.—

860mm

 

Assam“ a 3 to 28m

.86on 2E Sm 058368 5:55 630F056“. BE on? BEWQBEE GS: cocuwéw dogma—2 no 50v mama 56.80 mo :8ch .m 2an

67

Table 3. Comparison and percentage of GD and MG participants who would be very likely to eat

or purchase selected edible flowers.
Variable GD MG Significance

 

n 23 51

Eat a yellow pansy 91 90 NS
Eat an orange pansy 74 67 NS
Eat a blue pansy 91 94 NS
Purchase edible flowers

if grown organically 90 86 NS
Purchase edible flowers

with 10% insect 24 35 *
damage

 

NS, * Nonsignificant or significantly different at P s 0.05 in rows, based on a two-tailed t-test.

68

Figure 1. A nasturtium flower that was flawless and was rated a 5.

69

 

7O

Figure 2. A nasturtium flower that had the greatest amount of damage and was rated a l.

71

 

72

 

SECTION IV

Organic Nutrient Management of Greenhouse
Production of Edible Flowers in Containers

73

Organic Nutrient Management of Greenhouse Production of Edible Flowers in Containers
Kathleen M. Kelleyl

Michigan Agriculture Experiment Station and Departments of Horticulture and Botany and Plant
Pathology, Michigan State University, East Lansing, MI 48824-1325

John A. Biernbaum2

Michigan Agricultural Experiment Station and Department of Horticulture, Michigan State

University, East Lansing, MI 48824-1325

 

Received for publication . Accepted for publication . Research
conducted at Michigan State University, East Lansing, Mich. Use of trade names does not imply
endorsement of the products named nor criticism of similar ones not named. The authors thank
Susan Baldygas, Lindsey Henige, and Meredith Phares for their assistance.

 

1Graduate Student

2Professor. To whom reprint requests should be sent. E-mail address:
biembauflimsuedu

74

Additional index words: Collodial rock phosphate, dried blood, fish emulsion, greensand, nutrient
analysis, Omega 6-6-6, pH and EC, water-soluble fertilizer.

Abstract. Eight species with edible flowers: Begonia Xtuberhybrida ‘Ornarnent Pink’ Voss,
Borago oflicinalis L. , Coriandrum sativum L. , Dianthus superbus L. ‘Super Fantasy Mixed’,
Phaseolus coccin eus L. ‘Dwarf Bees ’ , T ropaeolum majus ‘Jewel Mix’ L. , Viola tricolor L.
‘Helen Mount’, and V. X wittrockiana L. ‘Accord Banner Clear Mixture’, were grown in 12. 5 cm
(1.5 L) square containers in a 30% by volume mineral soil, root medium suitable for organic
certification. Plants were fertilized biweekly with 200 ml of either fish emulsion (5-0.4-0. 8), or
Omega 6-6-6 (a certified organic, commercially available soluble fertilizer (6-2. 6-5)), each at 3 00
or 600 ppmN. Plants in either the 30% soil medium or a standard peat and perlite medium were
fertilized with a synthetic water-soluble fertilizer( 19-1 .8-19) at 300 ppmN (six treatments total).
Balanced plant growth and flowering occurred with both organic fertilizers, but shoot dryweight
was largerwith Omega 6-6-6. There was an increase in shoot dry weight for only three of the 16
comparisons of the 300 vs. 600 fertilizer rate. Shoot dry weight with the synthetic, water-soluble
fertilizer in both the soil based and soilless media was either the same or smaller than the organic
fertilizer treatments, but the rate of application was increased to weekly, midway through the
experiment, to correct and prevent leaf chlorosis. There were no differences in the total number
of flowers and flower size for V. tricolor and V. X wittrockiana species when fertilized with 300
and 600 ppmN Omega 6-6-6. In general, the root media pH increased over the range of 6.0 to
7.0, from the sixth to the 12th sampling week after the first fertilizer application, but only increased
or decreased slightly fiom l 2 to 1 8 weeks and there was little effect of fertilizer type. Root media

EC decreased initially with minimal change later, generally in the range of 0.25 to 0.75 dS-m‘l in

75

,_V_

 

a l :2 extract, with little effect of fertilizer type. For three species with Omega 6-6-6 at 600 ppm
N, root media EC increased above, and pH decreased below acceptable levels. Differences in
whole shoot nutrient concentration were greater between species than between fertilizer types.
Nitrogen levels ranged from 1.18 to 2.18% for all treatments, but were lowest in the soilless
medium control. The fish emulsion and synthetic fertilizer treatments maintained low tissue level of
P (0. 19 to 0.24%) and adequate K (2.69 to 3.22%). Phosphorus (0.39 to 0.49 %) and K (3.26
to 3.63%) tissue levels were higher with Omega 6-6—6 treatments and increased P most likely was
the greatest factor influencing the increased shoot dry weight. Shoot tissue concentrations of
micronutrients were in the sufficiency range for all treatments. There was little or no effect of
fertilizer rate on shoot tissue nutrient concentration, accept where 600 ppmN of Omega 6-6-6
resulted in increased medium EC and lower medium pH resulting in increased Fe and Mn tissue
levels.
Introduction

Greenhouse production of food crops could supply additional income for small farms.
Edible flowers are one example of a value-added crop that could be locally produced and
marketed. The popularity of edible flowers has increased since the late 1 980's. They are featured
in popular articles, and are also the subject of several cookbooks (Belsinger, 1991; McVicar,
1997; Barash, 1993).

Edible flowers can be grown using organic methods (Barash, 1993). Certifiable organic
media and amendments, such as fish emulsion, dried blood, and other plant and animal by-products
are used instead of chemically derived products. Suppliers list several factors as to why consumers

prefer organic produce. Some reasons include: “Knowledge ofhealth issues, concern for dietary

76

var

 

welfare ofloved ones, pathogen fears, and awareness of ecological issues.” Predictions have been
made that consumer demand for organic products will continue to increase (Di Martino, 1999).
Consumers’ responses to questions regarding the use of pesticides on edible flowers indicated that
they would prefer the product to be grown without such chemicals (Kelley et a]. , 20003, b).

Various sources have reported the benefits of using organic amendments. Since organic
fertilizers release their nutrients slowly (due to the micro andrmcro organisrm in the soil), the plants
are able to capture a greater portion (Mattem, 1996), whereas with synthetic fertilizers, the
nutrients are readily available and more vulnerable to leaching. Organic amendments not only
supply available nutrients, but they also “influence plant grth indirectly by improvements in the
soil’s physical condition (e. g. , soil tilth and water infiltration)” (Chaney et al. , 1992) and the level
of biological activity in the soil or medium.

Little has been written about greenhouse crop production using certified organic
methods. Biological activity could be important in determining the availability of organic fertilizers
and in preventing development of plant pathogens that are common in soilless potting media
(Chaney et al. , 1992; Hoitink et a1. , 1991, 1997). Short term greenhouse experiments with media
in containers would not focus on physical properties, but instead would concentrate on identifying
organic nutrient sources that are readily available and if the ratio between elements is acceptable.

Certified organic potting mixes used in the greenhouse should provide optimum aeration
and water holding capacities, retain nutrients, and can include ingredients such as peat, vermiculite,
and perlite (ATTRA, 1 998). However, for a potting mix to be certified organic it must not contain
synthetically derived starter fertilizers or wetting agents. The organic mix may contain corrrponents

such as soil, compost, kenaf, coconut coir, and organic nutrient amendments approved by the

77

certifying committee (ATTRA, 1998). Prepared organic potting mixes are available for both home
gardeners and commercial greenhouse establishments and recipes are available (ATTRA, 1998;
Coleman, 1 992). Coleman ( 1 992) has developed a compost based potting mix that includes dried
blood, greensand, and colloidal rock phosphate to supply nitrogen, potassium, phosphate and
micronutrients.

According to Chaney et al. (1992) animal byproducts such as dried blood may have
nutrient release rates similar to synthetic fertilizers due to the small amount of organic carbon. Bunt
( 1988), however, states that biological activity is needed to transform organic sources such as dried
blood into a readily available form that plants can use. Colloidal rock phosphate reportedly has
acceptable release rates since only little biological activity is needed to convert it to an available
form (Matten, 1 996). Greensand, however, is released slowly for l 0 years or more, but contains
high quantities of micronutrients (Mattem, 1996). When added as a preplant nutrient charge,
arnendrrrents such as greensand are added in larger quantities of 2.3 to 7 kg-m‘3(5 to 15 Ib6°yd‘3)
as opposed to 0.03 to 0. 1 1 k g-m'3(1 to 3 oz. -yd'3) for corrrrnercially manufactured fiitted trace
elements (Nelson, 1998).

Fish emulsion and Omega 6-6-6 are two certified, water-soluble organic fertilizers
identified in conversations with organic transplant growers. Growers also apparently prefer
biweekly or interval fertilization over constant liquid fertilization, which is more commonly used in
commercial greenhouses. Unlike corrrrnonly used synthetic, water-soluble fertilizer, organic
fertilizer cannot be maintained for long periods as diluted stock solutions suitable for fertilizer
injection equipment. Concentrated stock of organic fertilizer is usually at pH <40 to decrease

biological degradation. Bacterial and! or fungal growth is common once the stock is diluted with

78

water. With interval fertilization, diluted stock solution would be made and used in the same day.
Some organic fertilizers, hke fish emulsion, have a strong odor which also favors limited rather than
regular use. Interval-or-as-needed fertilization also favors more controlled plant growth than a
constant fertilization pro gram (Nelson, 1 998). Application concentrations or rates used by growers
for fish emulsion and Omega 6-6-6 could not be determined along with the effect of these fertilizers
on root media pH and electrical conductivity (EC).

Normally inorganic acid rrrust be injected to control media pH of soilless media when using
irrigation water with >250 ppmbicarbonate alkalinity (Argo and Biembaum, 1996; Nelson, 1998).
Acid injection is not acceptable for organic certification, so alternative strategies must be
considered in areas of the country with high alkalinity water. High alkalinity irrigation water is a
common problem in the Midwest. Root media containing rrrineral soil can provide greater pH and
nutrient buffering than soilless media (Nelson, 1 998).

This research was conducted to develop recommendations for organic fertilization methods
and to compare the growth, development, flowering, and whole shoot nutrient concentration of
edible flower species using organic media amendments with organic water-soluble fertilizers or
synthetic water—soluble fertilizer. To our knowledge, few published studies have focused on the
use of certified organic media or fertilizers in greenhouse container production. Other objectives
were to determine whether a common organic fertilizer strategy and rate would be acceptable for
a wide range of edrble flower species; to compare the organic fertilizer effect (100% arrrrnonium
(NI-I4-N) nitrogen plus carbon sources) on the root medium pH and EC as opposed to the
synthetic fertilizer (25%NH4-N); and to identify possrble differences between a higher-N fertilizer

(fish emulsion 5-0.4-0.8) and a more balanced NPK fertilizer (Omega 6-6-6 [6-2.6-5]).

79

Materials and Methods

Seeds of Begonia ><tuberhybrida ‘Ornament Pink’ Voss, Borago officinalis L. ,
Coriandrum sativum L. , Dianthus superbus L. ‘Super Fantasy Mixed’, Phaseolus coccineus
L. ‘Dwar'f Bees’, T ropaeolum majus ‘Jewel Mix’ L., Viola tricolor L. ‘Helen Mount’, and V.
><wittrockiana L. ‘Accord Banner Clear Mixture’ were sown between September 25 and
October 23, 1 998, into 30 cell (80 ml) transplant half- flats with a medium consisting of 40% (by
volume) of peat (Sun Gro Horticulture, Bellevue, WA), 40% coconut coir (Chrystal Company
Gro-Brick, St. Louis, MO), and 20% coarse vermiculite (Strong-lite, Pine Bluff, AR). Pulverized
dolomitic lime (9 kg-m'3 ( 1 5 lbs -yd'3))', National Lime and Stone Co. , Corey, OH) was added to
the mix to raise the pH to 6.3. The flats were placed into a 25 °C (77 °F) day and night
terrrperature propagation greenhouse with a 16—hour daylength using high-pressure sodium lamps
(HPS). After germination the plug trays were placed into a 21 °C (70 °F) day and night
temperature greenhouse, also with a 16-hour daylength. The seedlings were watered as needed
and were fertilized with 500 ppm N, from fish emulsion (FE) (5 -0.4-0. 8; Alaska Fish Fertilizer
Co., Renton, WA) to supply nutrients for initial growth and development. The water applied to
the crops between fertilizer treatments had an EC of 0 . 6 dS m"; 105 Ca“, 3 5 Mg“, 12 Na“, and
23 804-8 (ppm); and a titratable alkalinity to pH 4. 5 (Chau, 1984) of 3 20 mg CaCO3/liter (well
water) (Argo and Biernbaurrr, 1996).

As a species reached transplant size, over the period from] 9 November to 9 December,
1 12 plants per species were transplanted into a medium composed of 30% (by volume)
composted sandy-loam soil and equal parts (15% by volume) of coarse texture peat (Sun Gro

Horticulture), finer texture peat (Select Peat Moss Inc. , New Brunswick, Canada), coconut coir,

80

 

perlite, and vermiculite. To supply rrracronutrients and nricronutrients, 4.25 kg-m'3 (7 lbs -yd'3) of
equal parts (by volume) of greensand (0-0-7; North Country Organics, Bradford, VT), colloidal
rock phosphate (0-2-0; North Country Organics), and dried blood ( 1 2-0-0; Dragon, Roanoke,
VA), were added. This rate is approximately one half the rate used by Coleman ( l 992). Based
on measured fertilizer density, approximately 1.9, 1.4, and 1.0 kg-m‘3 (3 .2, 2.3, and 1.6 lbs-yd'3)
of greensand, rock phosphate, and dried blood were added. The pH of the media was 5 . 6 at the
time of planting. No lime was necessary due to the high pH of the composted mineral soil, which
offset the low pH of the peat. Growing containers were 14. 5-cm—tall x 15.5-cm-wide (1 .S-L)
plastic, square containers.

An additional 1 6 plants per species were transplanted into a soilless medium consisting of
70% (by volume) peat and 30% perlite (Strong-lite) mixture with a standard starter nutrient charge
(0.6 kg (llboyd’3) 10103, 0.3 kg (0.5 lb-yd‘3) triple superphosphate (0498-0), and 0.9 kg
(1 .5 lb-yd'3) gypsum; 0.07 kg (0. 1 lb-yd°3) fritted trace elements; 0.3 kg (0.5 lb°yd'3) wetting agent
(Aquagro “G”, Aquatrols, Pennsaulken, NJ) per m3 of medium) and 1 .5 kg-m'3 (2.5 lbs-yd'3) of
dolomitic hydrated lime. This medium was used extensively in previous research (Argo and
Biernbaum, 1996, 1997). The pH of this medium was 5.7 at the time of planting.

On 21 December, when plants began to look chlorotic, plants in the 30% soil medium
were divided into five fertilizer treatments. The number of plants per treatment was between 12
to 16 depending on species. Treatments included: FE at 300 or 600 ppm N; and Omega 6-6-6
(OM) (6-26—5; Petrik Laboratories, Inc. , Woodland, CA) a commercially available blend (3 liquid
starter fertilizer from microbial fermentation and digestion ofblood meal, bonemeal and potassium

sulphate (Harmony Farm Supply and Nursery, 2000)), at 300 or 600 ppm N. The fifth

81

 

treatment in the 30% soil medium and a sixth treatment in the sorlless medium were fertilized with
MSU Special ( 19-1.8-19, 25% NH4-N) a water-soluble fertilizer at a rate of 300 ppm N,
(Greencare, Chicago, Ill.). The plants were fertilized every two weeks with 200 ml of the
respective fertilizer treatment to insure that the plants were well fertilized without any leaching. On
29 January it was observed that all species fertilized

with the MSU Special 300 ppm N in the soilless and 30% soil medium looked chlorotic.
The These treatments were fertilized weekly at 200 ml ofMSU Special until February 26. The B.
officinalis and T. majus plants in these treatments still looked nutrient deficient after this period
of time and continued to be fertilized with 200 ml of MSU Special weekly until April 2.

The pH and EC ( 1 : l and 1 :2 soil: water dilutions respectively) of approximately 50 ml root
media samples were tested at six week intervals from the first fertilizer application and continued
throughout the experiment. Two samples per treatment were collected from the lower half of three
pots each. Only two pH and electrical conductivity (EC) readings were taken for C. sativum and
P. coccineus, at six and 12 weeks after December 23, since the rate of flowering declined and
whole shoot samples were harvested for nutrient analysis prior to the third pH and EC soil
sampling.

The first species began to flower on October 3 1 . On December 22, to encourage larger,
high quality flowers the temperature in one of the greenhouses was lowered to 16° C (61° F) day
and night temperatures (Niu et al. , 1 999). The other greenhouse remained at 21 °C (70 °F). Plant
species such as B. officinalis, T. majus, V. Xwittrockz'ana, V. tricolor, and P. coccineus
remained in the 16 °C (61 °F) greenhouse, while species that grew slower at the lower temperature

including: B X tuberhybrz’da, C. sativum, and D. superbus were moved to the wamrer greenhouse

82

to encourage more rapid growth. Beginning on February 5 and continuing weekly until April 1 9,
the flower size (height >< width) of 1 0 V. tricolor (viola) and 10 V. ><wittrockiana (pansy) flowers
were measured in each treatment.

All shoot growth and flowers were harvested for a species was harvested on one day
during the period of March 22 until May 5. Four samples per species each consisted of three or
four plants. The fiesh weight of the shoot was recorded after the plant material was cut at the soil
line. The plant material was then dried in a 60 °C (140 °F) oven for 48 hours, weighed again to
determine the dry weight and the percent dry weight was calculated. Shoot-tissue samples,
ground to 40 mesh, were sent to an independent lab for nutrient analysis (MicroMacro
International, Athens, Ga). Data were evaluated using SAS ’5 analysis of variance (AN OVA)
procedures and mean separation was preformed using least significant differences (LSD) (SAS
Institute, Cary, NC).

Results and Discussion

For the three fertilizer solutions applied at 300 ppm N, in the 30% soil medium, three of
8 species (B. oflicinalis, C. sativum, and P. coccineus), had the highest dry weight when
fertilized with OM (Table 1). Plant appearance and leaf color were also better with OM. Based
on V. tricolor tissue analysis, which was characteristic for most species, the increased growth of
these species is likely due to increased phosphorus availability (Table 2). Nutrient analysis means
across species were calculated, but were not presented due to significant species by fertilizer
treatment interaction for 9 of 12 elements. At the 300 ppmN rate, the concentration of P applied

was at least 4.5 times higher (130 vs 28 or 24 ppm) for the OM.

83

There was an increase in shoot dry weight for only three of the l 6 comparisons of the 300
vs. 600 fertilizer rate. There were no differences in the total number of flowers and flower size for
V. tricolor and V. ><wittrockiana species when fertilized with 300 and 600 ppm N OM (Table
3). When the two fish emulsion treatments were compared there were few differences in the dry
weight and the percent dry weight.

The soilless medium and MSU Special water-soluble fertilizer provided a tested control
comparison to the organic medium and fertilizer treatments. Shoot dry weight with the synthetic,
water-soluble fertilizer in both the soil based and soilless media, was either the same or smaller than
the organic fertilizer treatments, but the rate of application was increased to weekly, midway
through the experiment, to correct and prevent leaf chlorosis. Based on plant growth and tissue
analysis the soil and incorporated amendments did not increase growth or nutrient uptake with the
synthetic water-soluble fertilizer applied. With the addition of soil the dryweight was larger for
P. coccineus and V. Xm’ttrockiana, smaller for C. sativum and V. tricolor, and not different for
the other four species.

Media pH increased fi'om the sixth to 1 2‘h sampling week, but then generally decreased
or stayed the same (Figure l , Table 4). Differences between the medium pH for the two organic
fertilizers compared to 300 ppm N, MSU Special in the 3 0% soil medium existed only for the B.
X tuberhybrida and P. coccin eus (Figure 1 ). For both of these species the pH for those fertilized
with MSU special were higher than those fertilized with OM at 300 ppmN. Differences in medium
pH were not consistent for species over the sampling weeks when comparisons were made
between 300 and 600 OM or FE. Where differences did occur, generally medium pH was lower

for the 600 ppm N rate.

84

For four of the eight species, B. oflicinalis, B. xtuberhybrida, T. majus, and V.
xwittrockiana, the greatest change in media pH over time was with the soilless rrrediurntreatrnent.
On the sixth week of sampling, for all but C. sativum, the pH for species grown in the 30% soil
medium and fertilized with 300 ppmN, MSU Special was at least 0.5 higher than those grown in
the soilless medium. Medium pH for D. superbus and P. coccin cus exhibited this trend for the
12th and B. oflicinalis and V. ><wittrockiana on the 18th sampling week.

Certain species such as T. majus had higher media pH for all treatments tested. This trend
may be a species effect and/ or the alkaline water source may have affected the media pH since T.
majus was watered more frequently due to the large plant size.

High alkalinity (3 20 ppm) in the irrigation water would be expected to increase the root
medium pH over time with a 25% NH4-N fertilizer (Argo and Biernbaurrr, 1996). Under the
conditions of this study the change in medium pH was less than we anticipated. One possible
explanation is that little or no excess water was applied at each irrigation. The plants were also
grown during the winter when irrigation is reduced.

Bunt ( 1988) states that organic fertilizers such as hoof and horn (13% nitrogen) and dried
blood ( l 0- 1 3% nitrogen), are broken down into a useable form by firngi and bacteria for plants
to absorb. Bunt further explains that organic nitrogen is transformed to ammonium faster
(mineralization) than armnonium to nitrite and then to nitrate (nitrification). It is possible that
mineralization would cause the pH to initially rise over time. When nitrification occurs hydrogen ions
(H’) are released which can make the growing medium acidic. The net effect is little or no change

in media pH.

85

 

Argo and Biembaum( 1997) demonstrated that pH of root medium was dependent on the
water alkalinity, nitrogen form, lime, and peat type. With the combination of a high alkalinity water
source, and a 100% NH4-N fertilizer such as FE or OM, in a 30% mineral soil medium the
medium pH there was no sign of declining pH except at the 600 ppmN rate of OM. It is not clear
what affect the organic fertilizers would have when applied with a “pure” water source, common
in the Northeast U. S . , and to a soilless root medium. Adding mineral soil to apotting mediumcan
reduce fluctuations in nutrient content or availability due to cation exchange capacity and buffering
related to the higher bulk density.

Media EC values for all species were less uniform and had more variability between
fertilizer treatments than for pH. The EC values were generally in the low to acceptable range
(0.25 to l .0 dS°m'1) for all species. Differences between the EC for these treatments were not
considered culturally significant. For B. X tuberhybrida and D. superbus, the EC was higher for
the 600 vs. 300 ppmN rate. Shoot dryweight of B. x tuberhybrida was reduced in the OM 600
ppm treatment. It is likely that B. X tuberhybrida is a salt sensitive plant and that the soluble salts
inhibited growth. D. superbus had the highest flesh and dry weight when grown in the 30% soil
medium and fertilized with the OM 600 ppmN. The percent dry weight was similar to the plants
in the other treatments. Therefore, the higher levels of salts were not detrimental to D. superbus
growth. Changes in EC were similar for the two media fertilized with 300 ppmN, MSU Special,
except for C. Sativum were the EC increased in the soilless medium on the 12‘h week from the
sixth week (Figure 2).

Greenhouse container plant fertilization rates for standard water-soluble fertilizers are often

adjusted based on media EC. Under the conditions of this study, similar EC readings were

86

obtained with all three fertilizers and both media, which is an indication that EC monitoring would
also be a usefirl tool with organic fertilizers.

Though soil samples were harvested for pH and EC at six, 1 2, and l 8 weeks after the first
fertilizer treatment was applied, sampling occurred at different times in relation to date of
fertilization, which was on a two week cycle. Species had been divided into groups based on
growth characteristics to decrease the number of soil samples that were handled in aweeks time.
For each sampling week, soil samples for B. oflicinalis, C. sativum, and T. majus were harvested
and analyzed two weeks after species were fertilized. Soil samples for B. Xtuberhybrida, D
superbus, P. coccineus, V. tricolor, and V. Xwittrockiana were taken five days after species
were fertilized. This may explain the EC variation between fertilizer treatments and higher EC
values for B. ><tuberhybrida, D. superbus, and V. ><wittrockiana at each of the sampling times.

Differences in whole shoot nutrient concentration were greater between species than
between fertilizer types. Nitrogen levels ranged from 1 . 1 8 to 2. 1 8% for all treatments, but were
lowest in the soilless medium control. The FE and MSU special fertilizer treatments maintained low
tissue level ofP (0. 19 to 0.24%) and adequate K (2.69 to 3 22%). Phosphorus (0.39 to 0.49 %)
and K (3 .26 to 3.63%) tissue levels were higher with OM treatments. The increased P was most
likely the greatest factor influencing the increased shoot dry weight with OM.

Nitrogen and P levels in the whole-shoot sarrrples for most species fertilized with 300 ppm
N, fish emulsion, were lower than the sufficiency range levels recorrrrnended for leaf samples (Mills
and Jones, 1 996). This would be expected, however, since material that was analyzed contained
woody stems and not just succulent leaf material. C. sativum, P. coccineus, and T. majus had

the lowestN levels of 1 .29, 1.23, and 1.35%, respectively. Nitrogen levels for V. tricolorranged

87

from 1.25 to 1.96% for the six treatments and was representative of the mean values across
species (Table 2). Potassium levels were adequate for growth and development for all species
fertilized with FE even though the N: K ratio applied was 5: 0. 8. For most water-soluble fertilizer
solutions used in most corrrrnercial greenhouses the N:K ratio is 1: 1.

There were no culturally significant differences in Ca, Mg or S concentrations in all species
although for V. tricolor there were some statistical differences. As described previously, the water
source used for irrigation has high-alkalinity and supplies ample amount of Ca, Mg, and S. Overall
Ca to Mg ratios for each species were within the recommended 2: 1 to 3: 1 range for all species
except B. officinalis and B. X tuberhybrida which had ratios of 6: 1 and 1 : l , respectively.
Suflicient amount of sulfur were found in the whole shoot analysis for all species with all treatments.

Micronutrients Fe, Mn, B, Cu, and Zn were near the sufficiency range used for leaf tissue
analysis for all treatments (Mills and Jones, 1996). Sodium levels for B. officinalis, B.
><tuberhybrida, and T. majus were in the range of 2723 to 5055 ppm, which may indicate Na
accumulation, while for all other species the concentration was less than 1 000 ppm. There was
no indication oflrigher sodium levels in the tissue due to FE fertilizer. Species grown in the soilless
mix tended to have the highest level of rrricronutrients except when Fe, Mn, and Cu availability
were influenced by low pH with the 600 pm OM fertilizer rate.

Balanced plant grth was achieved with OM and FE applied at 300 ppm N at two week
intervals, though growth of plants fertilized with OM was better primarily due to increased P
availability. Plant growth was similar to that of plants fertilized with twice the rate, hence, the lower
rate is recommended for both supplying nutrients and lower costs for growers. Additional tests

are necessary to identify rates of organic fertilizers for soilless medium. A higher P analysis in

88

 

water—soluble fertilizer, like OM, or other media sources ofP, like bone meal or compost, or higher
rate of rock phosphate, are recommended. Rock phosphate in the 30% soil medium did not
supply adequate P, although more controlled studies with varying rates of rock phosphate are
needed. The rate used in this study was one half the rate recommended by Coleman ( 1 992). He
also recorrrrnended incorporating the rock phosphate into a compost-based nrediurn several weeks

or months prior to planting, which likely would influence solubility.

89

“A ‘v-lssat .

 

Literature Cited

Argo, W.R. and J .A. Biernbaum, 1997. Lime, water source, and fertilizer nitrogen form affect
medium pH and nitrogen accumulation and uptake. HortScience 32: 71-74.

Agro, W. R. and J .A. Biernbaum. 1996. The effect of lime, irrigation-water source, and water—
soluble fertilizer on root-zone pH, electrical conductivity, and rnacronutrient management
of container root media with impatients. J. Amer. Soc. Hort. Sci. 121(3):442-452.

Appropriate Technology Transfer for Rural Areas. 1998' Organic potting Mixes. ATTRA,
F ayetteville, AR 72702.

Barash, C. W. 1993. Edible flowers: From garden to palate. Fulcrum Publishing, Golden, CO.

Belsinger, S. 1991 . Flowers in the kitchen: A bouquet of tasty recipes. Interweave Press,

Loveland, Colo.

Bunt, AC. 1998. Media and mixed for container-grown plants. Unwirr Hyman, London.

Chaney, D.E., LE. Drinkwater, and GS. Pettygrove. 1992. Organic soil amendments and

fertilizers. Oakland, Calif: UC Sustainable Agriculture Research and Education Program,

University of California, Division of Agriculture and Natural Resource.

Chau, A.S.Y. 1984. Water and salt. p. 617-638. In: S. Williams (ed). Official methods of
analysis of the association of official analytical chemists. Assoc. of Official Anal. Chemists,
Arlington, Va.

Coleman, E. 1992. The new organic grower’s four-season harvest: How to harvest fresh organic
vegetables fiom your home garden all year long. Chelsea Green Publishing, Post Mills,
BT.

Di Martino, C. 1999. Organics growing in foodservice. Produce Business 15(3); 41-42

90

Harmony Farm Supply and Nursery. 2000. Fall 2000 cover crop seeds and gardener’s gift
catalog. Sebastopol, CA.
Hoitink, H.A.J., Y. Inbar, and M.J. Boehm 1991. Status of compost-amended potting mixes

naturally suppressive to soilbome diseases of floriculture crops. Plant Disease 75(9): 869-

873.

Hoitink, H.A.J. , A.G. Stone, and D.Y. Han. 1997. Suppression of plant diseases by compost.
HortScience 32:184-187.

Mattem, V. 1996. The 0G guide to organic fertilizers. Organic Gardening 43(5):55-59

McVicar. J. 1 997. Good enough to eat: Growing and cooking edible flowers. Kyle Cathie Ltd. ,
London.

Mills, HA. and J .B. Jones, Jr. 1996. Plant analysis handbook II. MicroMacro Publishing, Inc. ,
Athens, Ga.

Nelson, P.V. 1998. Greenhouse operation and management. Prentice Hall, Englewood Cliffs, NJ.
Niu, G., A. Franc, R.D. Heins, A. Cameron, and W. Carlson. 1999. Forcing Perennials:

Greenhouse temperature affect the timing and quality of herbaceous perennials.

Greenhouse Grower l7(9):130, 132, 134, 136, 138-140

91

 

L552 533.

 

0 2m 0 u on an on on on m m 33c»:
.... .... £2: 3A 92 EN 4.2 gm 3: 0.2 at mom 5m 52 so:
.52 333.
0 an o u a o a m on an o 3er
... .... 83.2 men we. 3m 32 36 I; n? 3: e9” 8.: 3m 53888:
.moomrtmao.
D a on D m on m w £ m u @3mebe
.. ...... £52 ..8 0.8 0.2 v8 32 van 0.8 :m 9.8 3A m3 383qu
.852
35:3 Seam.
on o a pm on pm 89:35.
m2 _. 3m 02 SN 3: no“ 4.? b _ m mm 3“ 3m EN 3m “.3585
on u on m u an SE23.
mz .. emu 32 3,“ was won n. a 3m mg 4.8. NS 2.“ one §§§§e
0 pm no u a EN 2 n so a so £38K?
...... .... 2.3 w? 3; 2m .....2 56 N: 3% 92 3m mi in 883m
we:
EoEaEO.
enteafmaé
m2 m2 3 NR 3 mi 3 we 5 EN be 2: Z: mam 3.58%
En .x. E be >6 .x. E 5 >6 .x. E be 20 .x. E be E. .x. E be an .x. E 5 as .x. E be 86on
:8 son Em soon :5 scam zom $8 Em econ $2.6m
822cm 55 seesaw 5.: 0.0.0 82:0 3-8 «8:5 seam 3m: 58% am:
8583a 2 can ace 2 can 8m 2 can Se 2 Egg 0% 2 EB 8m z 5% 8m

 

.333 58 co 925 go vomomfioo :omo £2953. 58 mo :35 05 md “53> seem
.mcoszpEou SNEUQEBE x7. do 25 £3, 55% .36on 53cc 9369 Ems “8 $6 .5 Ems; be :5on 93 385 E $3 E3 EH?» 5 Beam ._ 293.

92

as; 2 EmanEm >23: can 58% o. EmouEEm .oocmoa_:m_m-:oz ...... .... .m Z

 

Emu—.256 8: 2m 5:2 2:9. 2: E 330:8 $.82 .EwBB be :8on new 3995 be .3 @2888 cam 88on some cod mace E B oucmogcmaa
.8252 820
55mm 283}
.3: 2w m o 0 pm 0 a an a m n enefieEEE
.. mi mind m5 w: mgr; c.mm v.2 men we— msm Z: of x EDS
.5253 _ 2,5

93

55.602 =om Axsom

 

 

 

00 0 0 0 0 0 00 0 0 0 020380 in

00 00.0 00 00 0.0 00 0_ .0 00.0 00; 0. .0 0 _ .0 000 z 600 000

:58: 00m 0000

00 00 0 0 00 0 0 0 0 0 020.080 000

00 000 00 00 0.0 00 2 .0 00.0 00.0 00.0 0 _ .0 0: z 600 000

82002 00m 0000

0 0 0 0 00 . 0 0 0 0 0 0-0-0 «moEO

00 000 00 00 0.0 .0 20 00.0 .2 0.0 00.0 00._ 2500 000

80.0002 00m e000

0 0 00 0 0 0 00 0 0 00 0-0-0 000800

00 0.0 00 00 0.0 00 2.0 00.0 00.0 2.0 00.0 00; 2800 000

5.602 00m 0000

00 0 00 0 00 0 00 0 0 00 seam 2m:

00 000 00 s. 0.0 00 2 .0 00.0 00.0 00.0 t .0 000 z 800 000

8:000: mesom

0 0 000 0 0 00 00 0 0 .0 seam am:

00 000 00 0., 0.0 00 0_ .0 _0.0 00.0 00.0 00.0 000 2800 000

E E E Ema 8mm 8mm .x. .x. eel: 0\0 0\0 0\0 0:25.00;
00 02 =2 00 5 m m 02 6 m 0 z

“033,—. EDT—«:2

 

3.8—q Sou mo 0000380 :08
£29.80 Be“ 00 :02: 05 m._ 2:? seem 005905088 802.50% 0:008 cam 00 25 5.50 :3on 33.0.0: 30.: 08 mmmbmfiw 0:03.805 .0 0321

94

.mod n m 3 E9506 banEfiwE Ho: 80 8:0— 2:00 2: B 0032.8 0:32 6808 05.53: :03 08 mac—38 E m_ 3:35:39

 

N—

on

_m

Cw

w_ m_ _.0~ m 3.0— 5.0 wm.w wwd— Nw.N_ mm.0_ .>.U
o:_0>
wv _m 0N mm 2 .o 006 no.0 wo.m 0N6 w0._ 0:002 =05>O

00.0080 0 0.000

95

 

.mod n m 00 0:200:20 >_0:0oE:w_0 Ho: 80 8:0— 2:00 05 5:5 0020> 00.80% 0000 :00 038 E 0_ 00:00£:w_m0

.oQS \v 3 QQQN 00 00.003 05 886 0808030000 0— :0 00000 .080 0030: $003.09
.mQE \v 00 030R .8 00.83 on: wfihfi 83050000 I :0 0.0000 038:: 530$ ow08><x

 

 

 

 

000— 00.2 00.3 00.0—
30 000003 an 000:

00.3 00.00 00.0 00.0
A083 008m 530$ ow0b><

00. 0w 000 0000
08032: 06 0008: Z .0on

z 600 000 z 8000 000 z 600 000 2 E00 000
0-0-0 0w080 0-0-0 0380 0-0-0 0380 0-0-0 0380 0388000000 80E
000000093300 EDS 038.8 50$

 

.2 E8 000 00 000 0-0-0 000:5 003 00000000 000 82008 =00 0000 0

E :2,on 00300028300 .0. 0:0 088.20 30.: :8 080% :_ Ems? 000 0:03 0:0 .080 E 00.0 0030: 0w0b>0 .6088 826: 38,—. .m 0300

6
9

3003 @5953 m3“ OH :25 “mots: mm? 2505-805 .00—90:96 Ho: 00qu .<Z
0020000000 ._00.0 s ._0.0 .000 u 0 00 E00000 000000500 00 0085:0082 000.00....02
000000300800; :

 

 

 

 

 

 

...... *** *** 0.3.. ** **
02 0...... ... ... mz ...... 0008.5 x
02 02 ... ... ... ...... 00.005 0
<2 mZ ... , <2 ... 03.3.. 0002.8qu .m
m2 0.3....” 0.2.. m2 .13." .I. azaxmkzu. .Q
<Z m2 m2 <Z m2 m2 252.005. U
.0 m2 ...... ...: m2 1* 302.8%: .m
... * ...... 3.... ...... 0.... 003.503me3? .m
M: 0.003 0_ 0.003 0 0.003 M: 0003 00 00.03 0 #003 Jam:
00 0002 :0 00002

 

028% 0030:0560 Ems 8.:
80:02:50 50:88 00:0 08 08.00 1003 m:=:800 em: 0:0 .02 .5020 05 00 Um 0:0 I: 0608 8.. 080.8030 00%—0:000 E0885 .0 030...

97

Figure 1. Root media pH of eight species in soilless (O) or 30% soil medium with five water-
soluble fertilizer treatments ( MSU Special ( 19-1 .8-19) a water-soluble fertilizer at 300 (O, 0) ppm
N; Omega 6-6-6 (6-2.6-5) a commercially available blend of organic nutrient sources at 300 (V)
and 600 (v) ppm N; and Fish Emulsion (5-0.4-O. 8) at 300 (I) and 600 (D) ppm N). Data points
are means of two samples of three pots each.

98

pH

pH

pH

pH

 

7.5 ~0-

7_0 4.

6.5 0

6° 4»

551i-

504.

 

Begonia x tworhybrida

 

Borago ofiicinaiis

 

8.0 ~>
7S 4»-
7.0 J
6.5 T

6.0 +

50»
450

.0 .L

Coriandrum satlvum

A
v

i»

‘l

Dianthus superbus

 

so 4»
7 s 1-
7.0 i
6.5 1»
6.0 if
5.5 +
5.0 l
4 s ~~

404»

Phasoolus coccinous

 

Tropaeolum majus

 

8.0 +-
7 5 1*

7o 4»

6.01»
550
50*
454

4.0 0

 

 

Viola tricolor

 

+——o+———+-——+-

L
f

 

l L
Y

4»

 

Viola x wmmckiana

 

T

01*

Somplng Wook

12 18

 

 

-o— Tm1-mmlflimlmm
~01 Tmmz-MMHSUMCMW
~v- 1mm 3- momma-own:
-v-. Tm42mpm0ngam$oillx
+ tms'mmth Emulsion sum
—o-» two-woman: Emulsion soul-x

 

 

J.
v

«r-

12 1O

Samling Woek

 

Figure 2. Root media EC of eight species in soilless (O) or 30% soil medium with five water-
soluble fertilizer treatments (MSU Special ( l 9- 1 .8-19) awater-soluble fertilizer at 300 (O, 0) ppm
N; Omega 6-6-6 (6-2.6-5) a commercially available blend of organic nutrient sources at 300 (v)
and 600 ( v) ppm N; and Fish Emulsion (S-O.4-O. 8) at 300 (I) and 600 (D) ppm N). Data points
are means of two samples of three pots each.

100

EC

EC

EC

EC

 

 

 

 

 

 

 

 

 

 

 

Begonia x luberhybrida BONGO OMCIMIIS
1.50 <r .0.
v
1.25 "L / 1r-
/
/
1.00 *r f. ...—c ..
v.- I
0.75 0 4.
K / W
\ \ / v
000 \ \ // /
o ' F V A / -‘
025 +- M 4» K / ‘ ‘9’ \ '
o
Coriandrum sativum ,c Dianthus suporbus
1.50 " ~4 / /
'7’ \
1.25 0 4» \
1.00 i i \/ x
/
K \ / \
0.75 1i- T O.\ \\\ / / ,7
\
o \ \\ (
k N 00 \,
0.50 ‘l' .4» \ .\ \ \ 8
\
\
0.25 .. 0 .\_2/.
Phasoolus coccinous
, so .. T Tropaeolum majus
1.25 l» 1.
1.00 4* .0
0.75 l) .r
050 4* «>-
025 J» t
1 so t V'°" mm” \ Viola x wiflrockiana
\
1.25 4' +5 \
100 ‘i .4, C\ \ \
'\ \ \/
\ \
o 75 .. 0 9i \
0 50 ‘i r \ '3 \ i
k ‘0 § ‘
025 " dr- \g
6 12 18 5 12 13
Stunting Wook Sampllng Wool:

 

 

-0— Tmm1-mmmsmmux
~07 rumma-mmususmmulm
—v- TWSIMpWWO-C-Oswlx
-v-‘ 1mm4-mppmmO-stux
+ rms-mmmmwm:
-o- two-mmsummsouax

 

 

101

 

SECTION V

POSTHARVEST SHELF LIFE OF FIVE EDIBLE FLOWERS

102

Postharvest shelf life of five edible flowers

Kathleen M. Kelleyl

Michigan Agricultural Experiment Station and Departments of Horticulture and Botany & Plant
Pathology, Michigan State University, Plant and Soil Sciences Bldg, East Lansing, MI 48824-
1325

Arthur C. Cameron2 and John A. Biernbaum3

Michigan Agricultural Experiment Station and Department of Horticulture, Michigan State
University, Plant and Soil Sciences Bldg, East Lansing, MI 48824-1325

Kenneth L. Pot‘fS

Michigan Agricultural Experiment Station and Department of Botany & Plant Pathology,

Michigan State University, Plant Biology Bldg, East Lansing, MI 48824-1325

 

Received for publication . Research conducted at Michigan State University, East
Lansing, Michigan. Use of trade names does not imply endorsement of the products named nor
criticism of similar ones not named.

1Graduate Student

2Professor. To whom reprint requests should be sent. E-mail

Address:cameron@msu.edu

3Professor

103

Additional index words: Storage, visual quality, temperature, Viola ><wittrockiana , Viola
tricolor, Borage officinalis, T ropaeolum majus, and Phaseolus coccineus
Abstract. Five species of edible flowers were stored at -2.5 to 20 °C. Flowers were rated
for visual quality each day for two weeks on a scale of 1-5 (5 being the greatest). Viola
tricolor L. ‘Helen Mount’ (viola), V. ><wittrockiana L. ‘Accord Banner Clear Mixture’
(pansy), and T ropaeolum majus L. ‘Jewel Mix’ (nasturtium) flowers showed similar losses in
quality and were all rated a 5 when stored at O and 2.5 °C after two weeks. Borago
oflicinalis (borage) stored at O to 5 °C were rated marketable, 3 or higher, after one week,
and those stored at -2.5 °C were still marketable after two weeks. The highest ratings
Phaseolus coccineus L. (scarlet runner bean) flowers received were 33 and 45 when stored at
O to 10 °C, after one week. Scarlet runner bean flowers were unmarketable after 10 d at O
and2.5 °C,9dat5 °C, and7dat 10°C.
Introduction

There has been an increased popularity of edible flowers as evidenced by an increase in
the number of edible-flower cookbooks, culinary magazine articles, and television segments
(Rusnak, 1999). Consumers purchase packaged flowers for use in meals as a garnish or
ingredient in salads, soups, entrees, desserts, and drinks (Barash, 1998a; Barash 1998b).
Edible flowers are a niche crop (Rusnak, 1999). To successfully market edible flowers,
consumer preferred colors and species should be included in the packages (Kelley et al.,
2000a; Kelley et al., 2000b; Kelley et al., unpublished data). Growers want to ensure that only
their highest quality product, with a maximum shelf life is marketed to encourage repeat

purchases.

104

 

Edible flowers sold in retail stores are usually placed next to herbs in the refi'igerated
sections of the produce department (personal observation). To protect the flowers from
desiccation and to preserve their frail structure, they are typically packaged in small, rigid,
plastic or plastic wrapped packages (Whitman, 1991). To our knowledge, no standards for
harvest and handling have been published in referred journals.

Temperature is usually the most important environmental factor limiting shelf life of
fruits, vegetables, and herbs (Watada and Qi, 1999; Aharoni et al., 1993', and Paul], 1999).
Aharoni et a1. (1993) indicated that several herbs were best stored near 0 °C, though some
herbs like basil (0cimum basilicum L.) are sensitive to chilling injury when stored below 10 °C
(Lange and Cameron, 1994; Cantwell and Reid, 1992). To our knowledge, no information has
been published that identify factors limiting quality and no guidelines have been established for
how edible flowers should be stored. Though refrigeration will most likely extend shelf life of
most edible flowers, some could be sensitive to chilling injury.

Of 100 identified species with edible flowers, 28 were tested for greenhouse
production (unpublished data), and 12 were selected for further study. Five of these edible
flowers were selected for further marketing and storage studies based on availability,
recommendations in cookbooks, and ease of production, harvesting and handling. Seeds of all
five species are available from several sources and can be grown to flower in 20 weeks or less.
Viola tricolor (viola) and V. x Wittrockiana (pansy) are popular compact-growing bedding
plants that are easily grown to flower in containers. Pansy has been hybridized to increase
flowers size and is available in a multitude of colors from white to black to multicolor.

Tropaeolum majus (nasturtium) can be grown in hanging baskets and also has edible foliage.

105

‘nm

Borago officinalis (borage) is an annual herb grown frequently for its medicinal properties, but
is now recognized for its culinary attnbutes also. Phaseolus coccineus (scarlet runner bean) is
grown as an annual with flowers that range in color from white to orange and pink. Each edible
flower chosen for this experiment has its own distinct flavor. Viola and pansy have a
Wintergreen or sweet flavor, while nasturtium can have a taste similar to a radish or a peppery
flavor. Borage has a cucumber flavor and scarlet runner bean flowers taste like fresh beans.

The objective for this research was to determine the effect of storage temperatures on
the rate of loss of visual quality for these five edible flowers.

Materials and Methods

Seeds of viola, pansy, borage, nasturtium, and scarlet runner bean species were sown
between September 25 and October 23, 1998, into 30 (80 ml) cell plug trays with a medium
consisting of 40% (by volume) of peat (Sun Gro Horticulture, Bellevue, WA), 40% coconut
coir (Chrystal Company Gro-Brick, St. Louis, MO), and 20% coarse vermiculite (Strong-lite,
Pine Bluff, AR). Pulverized dolomitic lime (9 kg-m'3 (15 lbs-yd'3); National Lime and Stone
Co., Corey, OH) was added to the mix to raise the pH to 6.3. Plug trays were placed into a
25 °C day and night temperature propagation greenhouse. After germination, plug trays were
placed into a 21 °C day and night temperature greenhouse. Four-hundred-watt high-pressure
sodium (HPS) lamps provided a photosynthetic photon flux (PPF) of 100 j,tmol°m'2-s'l when the
ambient greenhouse PPF dropped below 200 umol-m'z-s'l 1 from 0800 to 2400 HR.
Supplemental lighting was terminated with PPF exceeded 400 umol-m'z-s". Seedlings were
watered as needed and fertilized with fish emulsion (500 mg-L‘l Nitrogen (N), 5N—0.4P-O.8K;

Alaska Fish Fertilizer Co., Renton, WA), to supply nutrients for initial growth and development.

106

From November 19 to December 9, plants were transplanted into a 70% (by volume)
peat and 30% perlite (Strong-lite) mixture with a standard starter nutrient charge (0.6 kg
(llb-yd‘3) IQIO3, 0.3 kg (0.5 lb°yd'3) triple superphosphate (0-19.8-0), and 0.9 kg ( 1.5
lb-yd'3) gypsum; 0.07 kg (0.1 lb-yd'3) fritted trace elements; 0.3 kg (0.5 lb-yd'3) wetting agent
(Aquagro “G”, Aquatrols, Pennsaulken, NJ) per m3 of medium) and1.5 kg-m'3 (2.5 lbs-yd'3) of
dolomitic hydrated lime. The pH of this medium was 5.7 at the time of planting. Growing
containers were 14.5-cm—tall x 15.5-cm-wide (1.5-L) plastic, square containers.

Starting December 21, plants were fertilized as needed, either weekly or every two
weeks, with 200 ml of MSU Special water soluble fertilizer (300 mg-L" N, 19N-1.8P-19K, '
25% NH4-N, Greencare). The volume of fertilizer chosen insured that plants were well
fertilized without any leaching. Species began to flower on October 31. The date of visible
bud, first color, when flowering began, full flower and senescence was recorded for 10 to 15
individual flowers on 10 plants of each species. On December 22, to encourage larger, high
quality flowers the temperature in the greenhouses was lowered to a constant 16 °C day (Niu et
al., 1999).

Flowers were harvested from viola and borage at 1000 HR each week from January 29,
1998 to February 19, 1998. Scarlet runner bean, nasturtium and pansy flowers were harvested
at 1000 HR each week from February 26 to March 5 and from March 19 to March 26. On
each harvesting date, flowers were removed from the plants and placed in unsealed plastic bags
for transporting to the laboratory (=22°C) for packaging (Lange and Cameron, 1994).

Flowers were harvested when completely expanded.

107

Flowers of each species were placed into 2-mi1 low-density polyethylene film bags
(Dow Chemical Co., Midland, Mich.) constructed using a Magneta 620 heat sealer (Packaging
Aids Corp, San Francisco) with four 26.5-gauge needle holes (=04 mm in diameter) in each
corner of the bag to improve 02 and C02 exchange (Lange and Cameron, 1994). Viola,
borage, and scarlet runner bean flowers were placed in 15 x l 1-cm polyethylene bags while
pansy and nasturtium flowers were placed in 20 x 1 l-cm bags. Bags consisted of one of the
following: ten viola flowers, ten borage flowers, five scarlet runner bean flowers, three pansy
flowers, or three nasturtium flowers. Flowers were inspected to ensure there was no damage
prior to being sealed in the bag.

Total time for harvesting and packaging the flowers took less than 90 min on each
harvesting date (45 min for harvesting; 45 min for packaging). Six packages of each species
were placed in controlled-temperature chambers at 20, 10, 5, 2.5, 0, and -2.5°C, all within i
1°C. Packages of edible flowers were placed between sheets of black plastic to exclude light
during storage. Each day at 1000 HR for 14 days, flower were rated visually using a five-point
scale (Table 1.). Data were analyzed using Cochran-Mantel-Haenszel Statistics to determine
significance between temperatures for visual quality ratings at one week and two weeks time.

Results and Discussion
Each edible flower showed symptoms of quality decline when stored at temperatures
which did not promote maximum shelf life. Symptoms included water soaking, necrosis,
surface molds, and total tissue collapse. Visual quality ratings for each flower are given in Table

1. A flower which was rated 1 or 2 was considered unmarketable.

108

The pattern of quality loss as a function of temperature at two times for flowers of
nasturtium, viola, and pansy was largely similar (Figure 1). Flowers of these species stored at
0 and 2.5 °C had no visual defects after two weeks and were of marketable quality, (rated 3 or
higher) for one week when stored between -2.5 and 10 °C and for two weeks at -2.5 to 2.5
°C. After two weeks at -2.5, the visual quality of nasturtium flowers declined, though they
were still marginally acceptable. At 10 °C viola flowers were rated unmarketable after 7 d
(marketable for 7 d), while pansy and nasturtium were both rated unmarketable after 11 d.
When stored at 20 °C, viola was rated unmarketable after day two and pansy and nasturtium
were rated unmarketable after day five and six, respectively.

As temperature increased, visual quality of borage flowers decreased rapidly (Figure
2). Borage flowers were only marketable for 1 d when stored at 20 °C, 5 d when stored at 10
°C, 8 d at 5 and 2.5 °C, and 12 d at 0 °C. Only borage flowers stored at temperatures of -2.5
to 5 °C were rated marketable after one week. After two weeks, only flowers stored at -2.5
°C were rated marketable.

After one week of storage, scarlet runner bean flowers had the highest visual quality
when stored at 0 to 10 °C, but average ratings were never higher than 4 (Figure 2). After two
weeks, scarlet runner bean flowers received unmarketable ratings between 1 and 2,
independent of temperature. Scarlet runner bean was rated unmarketable after 2 d at -2.5 °C,
10 dathnd2.5 °C, 9dat5°C,7dat 10 °C,and3 dat20 °C. Whileontheplantina
greenhouse at 16 °C, it took 4 d from the time scarlet runner bean flowered until senescence,
while other flowers tested ranged from 6 d for borage, 8 d for nasturtium and viola and 10 d for

pansy. The longevity of bean flowers both on the plant and when harvested and stored in

109

".-.—.9

controlled-temperature chambers is much shorter than the other species tested. This flower
should only be included in packages expected to be consumed in less than one week.
Conclusion

Based on the results from this research, three of the five flowers, viola, pansy, and
nasturtium, can be stored at 0 and 2.5 °C for two weeks and still have a perfect visual quality.
Viola and pansy flowers, when stored at even lower temperatures of -2.5 °C, received visual
quality ratings above 3. Viola species are often one of the first species blooming in early spring
in Zone 5 (-20 to -10 °C). Perhaps if the viola and pansy flowers are preconditioned during
production to cold temperature similar to those used in this experiment, shelf life might be
extended at below-freezing temperatures.

Borage and scarlet runner bean flowers did not store well for two weeks; only borage
flowers stored at -2.5 °C were acceptable. The longest period of acceptable visual quality for
four of the edible flowers tested existed when they were stored around 0 °C, the benchmark
established by Aharoni et al., (1993) for herbs. Only scarlet runner bean flowers were not
rated marketable when stored at any temperature after two weeks.

LeBlanc et al. (1996) reported that the mean temperatures of refrigerated cases in
grocery stores are 7.6 to 8.4 °C in the winter and surrrrner, respectively. Viola, pansy, and
nasturtium flowers rated in this experiment had no defects when stored for two weeks at 0 and
2.5 °C, which is at least 5 °C lower than the display cases in the grocery store. For all flowers
tested except pansy, it would be expected that the visual quality would not be marketable after
two weeks when stored at temperatures used in the grocery store. Pansy flowers, however,

were marketable after two weeks at 5 °C and slightly below a rating of 3 at 10°C. Pansy

110

 

flowers stored at between 5 and 10 °C may be marketable after two weeks.

111

Literature Cited

Aharoni, N., O. Davir, D. Chalupowic, and Z. Aharon. 1993. Coping with postharvest
physiology of fresh culinary herbs. Acta Hort. 334269-78.

Barash, C.W. 19983. The flavors of flowers. Herb Companion 10:32-37.

Barash, C.W. 1998b. Please eat the flowers. Horticulture 98:36-40.

Cantwell, M. and MS. Reid. 1992. Postharvest handling systems: Fresh herbs. pgs. 211-214.
In postharvest technology of horticultural crops. A.A. Kader, (Ed). Univ. Calif,
Division of Agriculture and Natural Resources.

Kelley, K.M., BK. Behe, J.A. Biernbaum, and KL. Poff. 2000a (accepted for publication)
Consumer and chef perceptions of three edible flower species. HortScience.

Kelley, K.M., B.K. Behe, J .A. Biernbaum, and KL. Poff. 2000b (accepted for publication).
Consumer preference of edible—flower color, container size, and price. HortScience.

Kelley, K.M., B.K. Behe, J .A. Biernbaum, and KL Poff. (In preparation) Consumer ratings
of edible-flower quality, mix, and color.

Lange, DD. and AC. Cameron. 1994. Postharvest shelf life of sweet basil (0cimum
basilicum). HortScience 29: 102-103.

LeBlanc, D.I., R. Stark, B. MacNeil, B. Goguen, and C. Beraulieu. 1996. Perishable food
temperature in retail stores. In: New development in refrigeration for food safety and
quality. International Institute Refrigeration Commission C2, 1996-6, 42-57.

Niu, G., A. Franc, R.D. Heins, A. Cameron, and W. Carlson. 1999. Forcing perennials:
Greenhouse temperature affect the timing and quality of herbaceous perennials.

Greenhouse Grower 172130, 132, 134, 136, 138-140.

112

Paull, RE. 1999. Effect of temperature and relative humidity on fresh commodity quality.
Postharvest Biol.Tech. 15 2255-262.

Rusnak, J. 1999. Edible flowers, fresh herbs, baby vegetables: Still room for growth. Produce
Business 15:33-34, 36-37.

Watada, AE and Qi, L. 1999. Quality of fresh-cut produce. Postharvest Biol. Tech.
15(3):201-205.

Whitman, AT. 1991. Edible flowers and culinary herbs: New uses for traditional crops, new

crops for traditional growers. GrowerTalks 54:22-23, 25, 27, 29, 31, 33.

113

 

 

bszm 382258 3238
floomon— oz @250 Be %qu porno 8m mgom 3:3 98 £80.— vow—S, 8m floao—m $53
bfiwzm 388358 begom
38qu 02 3:3 08 28oz 3:3 2a mgom 3:3 8a £56m wow—.5» 8m 836$ 32>
t8 biwzm moo—So mace—Sp 866 98 fine was coo—e8
wooden 02 £83 mo 3:. 2a 253 do 8:. @350 .83 Lo $02 $83 o8 832m Eatammz
can bfiwzm “So—So 881% 860 28 63% was coo—n8
mwoomoa oz 283 mo 8:. 8a £33 mo 8:. 3:3 3on no 532 .683 2a 332m emom 855M Broom
5505 use
:15 Ba mama Jim 9505 can
5% 2m 25E cam 9 0:3 Eob BEE 5% 08 ESE .5505
mwoomon oZ wow—m3 8m £80m BE? 2m £83 28 BE? 2m £80; 28 3:5 8m £80m emacom
m w m N _ 868m

waam

 

.93me 2: mafia 3 m; 938 £030; 03.6.... 95 com £58695“ .3 @8535 mm .mocmtouofimao .9226 33.5 “mogfimom ._ 2an

114

Fig. 1 . The postharvest visual quality assessment for nasturtium viola, and pansy flowers stored
at -2.5 to 20 °C for one and two weeks. Visual quality ratings for one week and two weeks were
significantly different at P =0.001 .

115

 

 

 

 

 

 

 

 

Nasturtium
5» O 0
g 4 .-
Ti
:3 O
G 3 «r O
s
.2 2 -_ 0
>
1 -- o
Viola
5 ._
g 4 ..
To
3
a 3 .—
3
3
S 2 sl-
1 ._
Pansy
5'4 .. -
'6
3
o 3 ._ -
To
3
g 2 + _
1 t _
-2.5 0.0 2.5 5.0 10.0 20.0

Postharvest Temperature (C)

116

-O- After1 Week
—0— After 2 Weeks

Fig. 2. The postharvest visual quality assessment for borage and scarlet runner bean flowers stored
at -2. 5 to 20 °C for one and two weeks. Visual quality ratings for one week and two weeks were
significantly different at P =0. 00 1 .

117

Visual Quality

Visual Quality

 

Borage

 

l 1 l l l l

 

 

Scarlet Runner Bean

1 J l l l

 

l
l l I 1 l

-2.5 0.0 2.5 5.0 10.0 20.0
Postharvest Temperature (C)

-.- After1 Week
-0- After 2 Weeks

118

 

DISSERTATION CONCLUSION AND RESEARCH SUMMARY

119

Dissertation Conclusion and Research Summary

Based on the information presented in this dissertation, summary comments can be made
about marketing, postharvest, and fertilizer experiments. We have learned that there is an
acceptance for certain species of edible flowers. Garden Day and Michigan Chef de Cuisine
participants were more likely to purchase viola and nasturtium flowers and would be very likely to
use them as a salad ingredient, garnish, and as a meal ingredient. Bloomfest participants would also
be more likely to purchase the 16 ounce containers, with all three flower colors offered (yellow,
orange, and blue), at a price of $2 . 99. Garden Day and Master Gardener participants rated flower
quality and would still purchase viola, pansy, and tuberous begonias with visual quality ratings of
3 (ratings below a 3 were considered unmarketable), on a scale of 1-5, while they would only
purchase borage and nasturtium until stage 4.

Viola, pansy, and nasturtium flowers showed similar losses in quality at warmer
temperatures, but were rated a 5 after two weeks of storage at O and 2.5 °C. Borage flowers
received marketable ratings when stored at O to 5 °C for a one week period, with those stored at
-2. 5 °C were still marketable after two weeks. Scarlet runner bean flowers were unmarketable
after2 d at —2.5 °C, 10 dat O and 2.5 °C, 9 dat 5 °C, 7 dat 10 °C, and3 d at 20 °C.

Finally, species of edible flowers were grown using a 30% soil, certifiable organic medium
with amendments. Plants fertilized every two weeks with 300 ppm N, fish emulsion (SN-0.4P-
0.8K at dilution rates of ml/liter), had fresh weights, dry weights, and percent dryweights that were
larger or the same as other treatments used in this experiment. Based on nutrient analysis of dried

whole shoots these plants also had low shoot-tissue nutrient levels, but within recognized or

120

published ranges. All species were successfully growth with one soil and fertilizer treatments.

Other research, which will be published after the completion of the dissertation, focused
on additional aspects of edible flower marketing and production. Three additional marketing
studies were conducted and funded by The Fred C. Gloeckner Foundation, Inc. During Bloomfest
2000, participants answered questions about what species of flowers they would prefer in
containers they would purchase. A telephone survey conducted by Team Telecom in East Lansing,
MI, contacted 448 households in the Detroit Metro Area about their familiaritywith edible flowers.
A final marketing experiment was conducted with four Whole Foods Market stores in the Detroit
Metro Area. For six weeks, 20 packages of edrble flowers were delivered to each store. During
the first week, the price for each container was priced at $3 .99, $2.99 the following week and
$1.99 for the remainder of the experiment.

From February to J une 2000, an additional nutrient management experiment was
conducted to compare plant growth and development of impatiens grown in a soilless medium vs.
a root medium with compost. This research was supported by a grant from the Organic Farming
Research Foundation. There were 12 fertilizer treatments and two media treatments for a total of
24 treatments. Organic amendments were either incorporated into the media or applied as awater
soluble nutrient source ever two weeks. The response to the organic fertilizer was dependent on
the type of root medium used. Nutrient analysis of shoot samples was completed and media
samples were analyzed every three weeks for pH and EC.

From J une to September 1999, a cost of production model was developed for use with
niche crops grown in a minimally-heated greenhouse. Variable and fixed costs were tabulated for

an entire greenhouse operation including headhouse options, vehicle purchase or lease options, and

121

three levels of costs for pest control, lighting, and cooling. A final spreadsheet allows the user to
total the cost for their operation based on which options they choose.

Cultural information has also been collected on the growth and development of 1 8 annual
and perennial species with edible flowers. Specific production recommendations and expected
times to harvest will be published in a how-to article.

Finally, with the help of David Cappert, Research Associate in the Department of
Entomology, biological control was used over a three year period to control insect populations in
the greenhouse. This research was also partially fimded by the Organic Farming Research
Foundation, Project GREEEN and a grant from 1PM Soils. Information from this component has
been compiled into a research report to be presented to growers. While biological control was
generally successful, the high level of management and high cost of purchasing small amounts of
predators and parasites would not likely make this approach economically feasible. Further

research should include the use of spray materials identified as acceptable for organic certification.

Each paper in this dissertation and those that will be published in the near future have
added to the sparse knowledge about edible flow ers. During the past three years the authors have
been contacted by several small growers, consumers who attended the authors presentation on the
subject of edible flowers, and other individuals who are interested in production or use of edible
flowers. Our future goal is to make this information available to these interested producers and
consumers in a form other than scientific papers. Of the original objectives outlined, all have been
accomplished at a level that would allow us to provide the information necessary for producers to

successfully begin production and marketing organically- grown greenhouse edible flowers. Our

122

goal now must be to communicate this information in a useable format for potential growers.

Though the primary research materials for this dissertation were edible flowers, the
experiments conducted for the published papers can be used as models for other niche greenhouse
crops where little is known about the market potential, production, and postharvest shelf life.
There is great potential for continued study of other crops using the methods described in our

papers.

123

APPENDICES

124

Appendix A

Survey instrument used for an edible flower tasting with
Garden Days participants at Michigan State University
7 August 1998.

Title of Survey: Consumer perceptions of three edible-flower species.

Edible Flower Evaluation, Form A.

Cup A
First, only look at the flowers in the cup and rate them using the 1-7 scale below. Circle the number
which most closely reflects your response. If you have no strong feeling either way, please circle
a “4” to show you have a “neutral” feeling on this item

Please look at the flowers in the cup marked “A” and make your responses on those flowers only.
Just from a visual perspective, how do you feel these flowers look?

Neutral
Unappealing l 2 3 4 5 6 7 Appealing
Undesirable l 2 3 4 S 6 7 Desirable
Not at all interested I 2 3 4 5 6 7 Very interested
in tasting in tasting

Now, please smell the flowers in the cup and circle the response that most accurately reflects how
you feel about the smell of these flowers. If you don’t notice much fragrance, circling a “4” would
reflect this.

Unappealing l 2 3 4 5 6 7 Appealing
Unpleasant l 2 3 4 5 6 7 Pleasant

Next, please taste the flowers in Cup A and circle the response that most accurately reflects their
taste.

Tasteless l 2 3 4 5 6 7 Tasty
Undesirable l 2 3 4 5 6 7 Desirable
Finally, how likely would you be to . . . Very Moderately Very

Unlikely Likely Likely
Purchase this as a garnish for a meal? 1 2 3 4 S 6 7
Purchase this to eat in a salad? 1 2 3 4 5 6 7
Grow and harvest this from your garden? 1 2 3 4 S 6 7

Purchase this if it were available for use in
a salad or as a food garnish? 1 2 3 4 5 6 7
(with a reasonable cost/price)

Serve this to fiiends or family at a meal? 1 2 3 4 5 6 7

126

Please answer as many of the following questions regarding yourself as you want, giving your best
estimate where exact answers are not known. These questions are very important; they will help
us make sure that we are getting a representative sample of people. Your responses are
anonymous; we have no way to connect your response to this form.

1.

5.

6.

In what year were you born?
Are you... Female? or Male?
What is the highest level of education you have completed? Please check one.

some high school high school graduate
some college/technical school college/tech. graduate

Of the following, which category best represents your 1998 household income before
taxes? Please check one.

less than $20,000 $60,000 to $79,999 $120,000 to $139,999
$20,000 to $39,999 $80,000 to $99,999 $140,000 to $159,999
$40,000 to $59,999 $100,000 to $1 19,999 $160,000 or more

What is your family status? Please check one.

_sing1e, dependents—married, dependents
_sing1e, no dependents ______married, no dependents

How many people live in your household, counting yourself as one?
number of people in my household

What is the zip code for your mailing address?

 

127

Appendix B

Survey instrument used for an edible flowers tasting with members
of the Michigan Chef de Cuisine Inc. Association at the
Detroit Athletic Club in Detroit, Mich., 8 March 1999.

Title of Survey: Professional chef perceptions of three edible-flower species.

128

Edible Flower Evaluation, Form A
Cup A
First, only look at the flowers in the cup and rate them using the 1-7 scale below. Circle the

number which most closely reflects your response. If you have no strong feeling either way, please
circle a “4” to show you have a “neutral” feeling on this item.

Please look at the flowers in the cup marked “A” and make your responses on those flowers
only.Just from a visual perspective, how do you feel these flowers look?

Neutral
Unappealing l 2 3 4 5 6 7 Appealing
Undesirable l 2 3 4 5 6 7 Desirable

Not at all interested I 2 3 4 5 6 7 Very interested
in tasting in tasting

Now, please smell the flowers in the cup and circle the response that most accurately reflects how
you feel about the smell of these flowers. If you don’t notice much fragrance, circling a “4” would
reflect this. ”

Unappealing l 2 3 4 S 6 7 Appealing
Unpleasant l 2 3 4 5 6 7 Pleasant

Describe the fragrance of the flower

 

Next, please taste the flowers in Cup A and circle the response that most accurately reflects their
taste.

Tasteless 1 2 3 4 5 6 7 Tasty
Undesirable l 2 3 4 5 6 7 Desirable

Describe the taste of the flower

 

Finally, how likely would you be to . . . Very Moderately Very
Unlikely Likely Likely

Purchase this flower for use in a meal? 1 2 3 4 5 6 7

Purchase this to eat in a salad? 1 2 3 4 5 6 7

More likely to purchase if grown organically? 1 2 3 4 5 6 7

Purchase this if 10% of the flower had

129

insect damage? 1 2 3 4 5 6 7

What food item would you most likely use this flower in?

 

Do you currently use this flower in your presentations, if not would you?

 

How many days during a week would you use this flower in food presentations?

 

What would you be willing to pay for a dozen of these flowers?

 

Please answer as many of the following questions regarding yourself as you want, giving your best
estimate where exact answers are not known. These questions are very important; they will help
us make sure that we are getting a representative sample of people. Your responses are
anonymous; we have no way to connect your response to this form.

 

1. What is the zip code of your establishment?

2. Are you... _Female? or _Male?

3. How many meals per week do you serve at your establishment last week?
4. Least expensive entree on your menu?

 

5. Years employed as a chef?

 

6. Are you certified?

 

7. If so, what is your certification level?

 

Thank you for your time. Any additional comments that you could express in writing would be
appreciated.

130

 

Appendix C

Survey instrument used at Bloomfest at C obo Hall,
Detroit, Mich., 9 and 10 April 1999.

Title of Survey: Consumer preference of edible-flower color,
container size, and price.

131

Dear Participant:

Several Michigan State University researchers are investigating consumer perceptions of edible
flowers. We would like you to take a few minutes (less than 10), to help us evaluate the pictures of
the containers of edible flowers. Your response is anonymous. We have no way to meet you,
as an individual, to this corrpleted survey form. You are flee to not answer any question you choose,
but please try to answer every question We are not able to use incomplete responses. Thank you
for your time.

Please look at the corresponding board of containers of edible flowers. Please consider the following
situation: You are buying a container of edible flowers to use in a meal you are preparing for family
and fiiends. Using a 100 point system, please assign points to the containers of flowers (A through
AA), giving your favorite the most points and continuing to use the points until you do not have any
more. The more you like a package the more points you should allocate to it.

EDIBLE FLOWER BOARD

A C D E F Ni.-

SI

1. Have you ever eaten edible flowers before? yes __no
If yes, list the names if you can

 

 

 

 

2. Have you ever purchased edible flowers before? If yes, where did you purchase them?

3. How much would you pay for a 9 count container at Kroger? Merchant of
Vino?

4. How much would you pay for a 18 count container at Kroger? Merchant of
Vino?

5. What is the name of the store where you shop for ingredients for special dinners?

132

O\

. How many prepackaged salad mixes have you purchased in the last month?
number of packages.

7. Are you more or less likely to eat a salad during the summer months (circle one)?

8. Are you more likely to purchase a prepackaged salad nix or separate salad ingredients
(circle one)?

9. How many meals did you cook at home last week? How many times did you eat
out?

10. How many hours a week, on the average, do you spend in your garden during the spring and
summer months?

 

11. What percentage of your garden is flowers? vegetables? lawn?

12. How likely would you be to

Very Moderately Very
Purchase edible flowers because Unlikely Likely Likely
they were grown pesticide free? 1 2 3 4 5 6 7
Purchase edrble flowers as a
garnish for a meal? 1 2 3 4 5 6 7
Pm'chase edible flowers to
eat in a salad? 1 2 3 4 5 6 7
How likely would you be to purchase edible
flowers if they had 10% insect darmge? l 2 3 4 S 6 7

Please answer as many of the following questions regarding yourself as you want, giving your best estimate
where exact answers are not known. These questions are very important; they will help us make sure that we
are getting a representative sample of people. Your responses are anonymous; we have no way to connect
your response to this form.

1. In what year were you born? 2. Are you. __ Female? Or _ Male?

3. What is the highest level of education you have completed? Please check one.
_some high school _high school graduate _some college/technical school _college/tech. grad.

4. Of the following, which category best represents your 1998 household income before taxes? Please
check one.

__ less than $20,000 _ $20,000 to $39.99 _ $40,000 to $59,999

__ $60,000 to $79,999 _ $80,000 to $99,999 __ $100,000 to $119,999

___$120,000 to $139,999 _ $140,000 to $159,999 _ $160,000 or more

5. What is your family status? Please check one.

_ single, dependents _ single, no dependents
_ married, dependents _ married, no dependents

133

6. How many people live in your household, counting yourself as one? number of people.

7. What is the zip code for your mailing address?

 

134

Appendix D

Survey instrument used at a Master Gardener Conference 29 June 1999
and at Garden Days at Michigan State University,
5 and 6 August 1999

Title of Survey: Consumer ratings of edible-flower quality, mix, and color.

135

Dear Gardener:

Several Michigan State University researchers are investigating consumer perceptions of edible flowers. We
would like you to participate in this survey concerning edible flower quality. Your response is anonymous.
We have no way to connect you, as an individual, to this completed survey form. You are free to not answer
any question you choose, but please try to answer every question. We are not able to use incomplete
responses. Thank you for your time.

Please look at each picture that will be shown on the slide project screen. After looking at the quality of each
flower, please mark either acceptable or not acceptable.

Slide Number

1 Acceptable Not Acceptable

tab” . . . L _ NM Accemableitiir(12.12..'fiji;ii::.-;';i5;,

 

3 Acceptable Not Acceptable

we. N...

5 Acceptable —,N9_t_ Acceptable

Not A .131??-  :,: _

 

Not Acceptable

 

“Meme!
9 Acceptable Not Acceptable
10 Ampere mime.

1 1 Acceptable ' Not Acceptable

   

‘ "N... m

13 Acceptable Not Acceptable
; ' ...... “ New»

15 Acceptable Not Acceptable

ii 16 Amelie.” '_ . . _ Not Acceptable

17 Acceptable Not Acceptable

* * 18 mm Not Acceptable
19 Acceptable Not Acceptable

N" Acceptable;

 

21 Acceptable Not Acceptable

   

23 Acceptable Not Acceptable

24 I I Awepebie ~~~~ ’ > Not Acceptable

136

25 Acceptable Not Acceptable

Next, please look at each of the slides of Viola x wittrockiana (Pansy). Based on the color of the flower,
how likely would you be to eat this pansy/.7

Very Unlikely Moderately Likely Very Likely
Yellow Pansy 1 2 3 4 5 6 7
Orange Pansy 1 2 3 4 5 6 7

Blue Pansy l 2 3 4 5 6 7

How more likely would you be to...
1. Purchase edible flowers if

grown organically.) 1

[J
b.)
33
UI
ON
\1

I0

. Purchase flowers with 10%

insect damage? 1 2 3 4 5 6 7
3. Have you ever eaten edible flowers? yes no
4. Have you eaten edible flowers in the last 3 months? yes no

5. Have you purchased edible flowers in the last year? If yes, where did you purchase them?
6. Would you ever be likely to buy edible flowers sold in a market? yes no
7. Do you have edible flowers growing in your garden for consuming this year? yes no
8. How many hours each week do you typically spend in your garden?
9. Of the 21 meals in a week how may did you cook at home? How many did you eat out?
10. In what year were you born?
1 1. Are you... Female? Or Male?
12. What is the highest level of education you have completed? Please check one.
_____some high school high school graduate
some college/technical school college/tech. graduate
13. Of the following, which category best represents your 1998 household income before taxes? Please
check one.
_less than $20,000 _$20,000 to 39,999 _ $40,000 to $59,999 __$60,000 to $79,999
_$80.000 to $99,999 _$100,000 to $119,999 __$120,000 to $139,999 ___$140,000 to $159,999
__$160,000 or more

14. What is your family status? Please check one.
_sing1e, dependents _sing1e, no dependents _married, dependents _married, no dependents

15. How many people live in your household, counting yourself as one? number of people.

16. What is the zip code for your mailing address?

137

 

Appendix E

The University Committee on Research Involving
Human Subjects approval form.

Research title: Consumers evaluate edible flowers and culinary herbs.

138

 

OFFICE OF
RESEARCH
AND
GRADUAIE
STUDflfii

mMnnmnmmmm«uu
Research Involving
mmunfimhdx
amean)

IMW¢m9mHmwumy
246 Admrnistratnm Building
asUnqummmm
«BRJOQ

517f355-2180
FAX 517/432-1171

Diem-12m Sure Unmly
flfluhmMaWOMmW
Emmmkaon

MSU Is I! firmiwm.
amuqnnnommumr

MICHIGAN STATE
U N I v E R s l T Y

 

July 22, 1998

TO: Bridget Behe _ .
A216 Plant & 80il SCI. Bldg

RE: IRE”: 98-447
TITLE: CONSUMERS EVALUATE EDIBLE FLOWERS AND CULINARY
HERBS
REVISION REQUESTED: N/A
CATEGORY: l-G
APPROVAL DATE: 07/21/98

The University Committee on Research Involving Human Subjects'(UCRIHS)
review of this project is complete.. I am pleased to adVise that the
rights and welfare of the human subjects appear to be adequately
protected and methods to obtain informed consent are appropriate.

Egrefore, the UCRIHS approved this project and any reViSIOns listed
a ve.

RENEWAL: UCRIHS approval is valid for one calendar year, beginning with
the approval date shown above. Investigators planning to
continue a project beyond one year must use the green renewal
form (enclosed with t e original agproval letter or when a
project is renewed) to seek u date certification. There is a
maXimum of four such expedite renewals ssible. Investigators
wishing to continue a project beyond tha time need to submit it
again or complete reView. .

REVISIONS: UCRIHS must review any changes in procedures involving human
subjects, rior to initiation of t e change. If this is done at
the time o renewal, please use the green renewal form. To
reVise an approved protocol at ann other time during the year‘
send your written request to the CRIHS Chair, requesting reVised
approval and referencing the project's IRB # and title. Include
in our request a description of the change and any revised
ins ruments, consent forms or advertisements that are applicable.

paosnms/

CHANGES: Should either of the followin arise during the course of the
work, investigators must noti UCRIHS romptly: 11) roblems
(unexpected Side effects, comp aints, e c.) involv1ng uman
subjects or (2) changes in the research environment or new
information indicating greater risk to the human sub ects than
existed when the protocol was previously reviewed an approved.

If we can be of any future help, please do not hesitate to contact us
at (517)355-2180 or FAX (517)4 2-1171.

rely,

Mailed)

avid E. Wright, Ph.D.
UCRIHS Chair

DEW:bed

  

cc: John Biernbaum
Kenneth Poff
Kathleen Kelly

139

   

1131:1;31111111115111?1