IMPROVING THE SCHEDULING AND PROFITABILITY OF ANNUAL BEDDING PLANT
PRODUCTION BY MANIPULATING TEMPERATURE, DAILY LIGHT INTEGRAL,
PHOTOPERIOD, AND TRANSPLANT SIZE
By

Tasneem Mohammed Vaid

A THESIS
Submitted to Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Horticulture
2012

ABSTRACT
IMPROVING THE SCHEDULING AND PROFITABILITY OF ANNUAL BEDDING PLANT
PRODUCTION BY MANIPULATING TEMPERATURE, DAILY LIGHT INTEGRAL,
PHOTOPERIOD, AND TRANSPLANT SIZE
By
Tasneem Mohammed Vaid
Volatile fuel prices, an increased emphasis on sustainable production practices, and
declining profit margins have motivated commercial growers of annual bedding plants to reduce
energy inputs and improve crop scheduling accuracy. The objectives of this research were to
quantify the influence of average daily temperature (ADT), photosynthetic daily light integral
(DLI), photoperiod, and transplant size on the flowering characteristics of popular bedding plant
crops. Increasing ADT from 14 to 26 ºC decreased flowering time and plant quality in nearly all
of the 18 species and varieties evaluated in one experiment. In a separate experiment, similar but
variety-specific responses occurred in 16 petunia (Petunia ×hybrida) cultivars grown at 12 to 23
ºC. Linear and non-linear mathematical models were developed under a range of DLI conditions
to predict the influence of ADT on flowering time of these crops, and the base temperatures
(Tmin) at which the flowering rates were zero were estimated. The estimated Tmin ranged from –
3.9 ºC for Diascia barberae to 13.8 ºC for Gomphrena globosa. An additional study quantified
the effects of transplant size and ADT on flowering time and estimated profitability of five
species. Flowering time decreased only when the larger transplants contained more mature
plants, but transplanting a larger propagule was not necessarily a profitable strategy. Finally,
seedling sensitivity and duration of long-day induction, as well as the effects of temperature and
photoperiod interaction on flowering, were determined in three petunia varieties.

DEDICATION
I dedicate this thesis to my loving parents, Farida and Fakhrudin Chitalwala who against all odds
made sure I received the best education possible.

iii

ACKNOWLEDGEMENTS
I would first like to thank my major professor, Dr. Erik Runkle for providing several
opportunities during the past two years for my educational and professional development. His
encouragement and support of my research, faith in my work, and patience with my writing and
working style are sincerely appreciated. I would also like to express my gratitude to members of
my committee, Drs. Jonathan Frantz and Ryan Warner for their advice, expertise, and support of
my research work. Sincere thanks also to Dr. Paul Fisher for his assistance and advice. I would
also like to acknowledge the statistical help provided by the Center for Statistics, and the CANR
statistical group. A special thank you to Rakers Acres for providing the plant material for all my
experiments.
A big thank you to our greenhouse floriculture technician, Mike Olrich for his assistance
with experimental setup, resolving problems, patience, support, and optimism. I would also like
to thank our undergraduate employees for all their help and support in performing the
experiments, particularly Mina and Jon for data collection during my absence; and Heather and
Lesley for their assistance with root washing. Special thanks to our former graduate student, Dr.
Matt Blanchard for helping me transition in to graduate school, setup experiments, develop crop
models, and also for his support and friendship. Thank you also to members of the floriculture
group, and MSU colleagues: Cathy Whitman, Daedre Craig, Heidi Wollaeger, Nate Durussel,
Kristin Getter, Brent Crain, Wei-Kuang Lin, and Xeini Wang for their friendship.
Thank you to my family and friends for their love and support which keeps me going
each and every day. A very big thank you to my loving husband Mohammed for his endless love
and support, and the countless sacrifices that he made to make graduate life easy for me. The
achievement of this degree would not have been possible without him by my side.

iv

TABLE OF CONTENTS
LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
SECTION I
LITERATURE REVIEW ........................................................................................1
Introduction ..............................................................................................................2
Temperature Effects on Plant Development ............................................................4
Leaf Unfolding Rate ....................................................................................8
Flowering ...................................................................................................11
Effect of Temperature on Plant Quality .................................................................15
Effect of Temperature on Flower Number and Quality .............................15
Effect of Temperature on Branch Number ................................................18
Effect of Temperature on Plant Biomass ...................................................18
Effect of Temperature on Plant Height ......................................................20
Daily Light Integral (DLI) .................................................................................... 22
Effect of DLI on Plant Development .....................................................................23
Effect of DLI on Plant Quality...............................................................................26
Effect of DLI on Flower Number and Quality...........................................27
Effect of DLI on Branch number and Biomass .........................................29
Effect of DLI on Plant Height ....................................................................30
Interaction between Temperature and DLI ............................................................32
Temperature and DLI Interaction on Plant Development..........................32
Temperature and DLI Interaction on Plant Quality ...................................34
Conclusions ............................................................................................................35
Literature Cited ......................................................................................................38
SECTION II
EFFECT OF AVERAGE DAILY TEMPERATURE ON FLOWERING TIME AND PLANT
QUALITY OF EIGHTEEN SPECIES AND CULTIVARS OF ANNUAL BEDDING
PLANTS ................................................................................................................50
Abstract ......................................................................................................52
Introduction ................................................................................................53
Materials and Methods...............................................................................57
Results ........................................................................................................63
Discussion ..................................................................................................66
Conclusion .................................................................................................73
Literature Cited ........................................................................................100

v

SECTION III
AVERAGE DAILY TEMPERATURE AND TRANSPLANT SIZE AFFECT FLOWERING
TIME, QUALITY, AND ESTIMATED PROFITABILITY IN FIVE SPECIES OF
ANNUAL BEDDING PLANTS..........................................................................106
Abstract ....................................................................................................108
Introduction ..............................................................................................109
Materials and Methods.............................................................................112
Results ......................................................................................................116
Discussion ................................................................................................119
Literature Cited ........................................................................................133
SECTION IV
DETERMINING THE PHOTOPERIOD-SENSITIVE STAGES OF DEVELOPMENT AND
THE EFFECT OF TEMPERATURE AND PHOTOPERIOD ON FLOWERING IN
THREE PETUNIA CULTIVARS .......................................................................138
Abstract ....................................................................................................140
Introduction ..............................................................................................141
Materials and Methods.............................................................................146
Results ......................................................................................................149
Discussion ................................................................................................152
Literature Cited ........................................................................................162
APPENDIX
MODELING PLANT DEVELOPMENT AND QUALITY PARAMETERS OF 16 PETUNIA
VARIETIES IN RESPONSE TO AVERAGE DAILY TEMPERATURE UNDER TWO
DAILY LIGHT INTERGRAL ENVIRONMENTS ............................................166
Research Objective ..................................................................................167
Materials and Methods.............................................................................167
Results ......................................................................................................170
Literature Cited ........................................................................................190

vi

LIST OF TABLES
Table 1.1. Temperature response categories for popular bedding plant species based on their
estimated Tb and the corresponding percentage delay in flowering time when grown at 15
versus 20 °C ...........................................................................................................37
Table 2.1. Days from seed sow to transplant (TP), plug size and mean node no. at TP, and
characteristics used to determine flowering date for 18 varieties of 16 bedding plant
species in two experimental replicates...................................................................74
Table 2.2. Parameter estimates for linear and non-linear models (Eqns. [1], [3], and [4])
describing the flowering rate response to average daily temperature in 18 bedding plant
species and cultivars. Tmin and Tmax are the species-specific base and maximum
temperatures (ºC), respectively at which the flower development rate is zero. Topt is the
optimum temperature (ºC) at which the flower development rate reaches a maximum
value (Rmax), and C defines the curvature of the function. CI = confidence interval
................................................................................................................................75
Table 2.3. The effect of temperature on the number of nodes below first open flower in 18
species and cultivars of bedding plants. Plants were grown in glass-glazed greenhouse
compartments at five constant temperature setpoints under a 16-h photoperiod and a DLI
–2 –1
–2 –1
of 18 mol·m ·d (spring) or 9 mol·m ·d (fall). When regression slopes were nonsignificant (P ≤ 0.05) between the two replications (rep), data was pooled for statistical
analysis. During the fall replication, five or ten randomly selected plants of each species
were microscopically examined for floral initiation (FI) at transplant. .................78
Table 3.1. Time from seed sow to transplant, mean leaf no. at transplant, and characteristics used
to determine flowering date for five bedding plant species in two experimental replicates.
Seedlings were grown in 36-, 128-, or 288-cell trays. ........................................125
Table 3.2. Estimated production costs and net profits when starting with two transplant sizes
(36-, 128-, or 288-cell trays) in the production of five bedding plant species grown at 17
or 23 °C for first flowering on April 1. Heating costs were calculated using Virtual
Grower software for Grand Rapids, MI. Labor and overhead costs are estimated from
Fisher (2006). See materials and methods for information on production cost and net
profit estimation. Production time is the average days to flower, and assumes a 16-h
−2 −1
photoperiod and a mean DLI of ≥10 mol∙m ∙d . .............................................126
Table 4.1. Days from start of photoperiodic treatments to reciprocal transfer at different stages of
development (leaf number) in three petunia cultivars during the transplant stage ....
..............................................................................................................................157

vii

LIST OF FIGURES
Figure 2.1. The effect of mean daily temperature (°C) on the flowering rate in 9 species of
bedding plants modeled using Eq. [1] (panel B-I) and Eq. [3] (panel A) and parameter
estimates from Table 2.2. for replication 1 (●) and replication 2 (○). When regression
slopes were non-significant between the two replications (P ≤ 0.05), data was pooled (■)
for statistical analysis. Flowering rate was calculated by taking the reciprocal of the
number of days to flower. Each symbol represents the treatment means, and error bars
represent 95% confidence intervals. Tmin (base temperature) is the estimated minimum
2
2
temperature at or below which the rate of progress towards flowering is zero. r or R is
the coefficient of determination generated for linear and non-linear crop models,
respectively. ...........................................................................................................80
Figure 2.2. The effect of mean daily temperature (°C) on the flowering rate in 9 species of
bedding plants modeled using Eq. [1] (panel A, B, D, E, G, H), Eq. [3] (panel C, F), and
Eq. [4] (panel I), and parameter estimates from Table 2.2. for replication 1 (●) and
replication 2 (○). When regression slopes were non-significant between the two
replications (P ≤ 0.05), data was pooled (■) for statistical analysis. Flowering rate was
calculated by taking a reciprocal of the number of days to flower. Each symbol represents
the treatment means, and error bars represent 95% confidence intervals. Tmin (base
temperature) is the estimated minimum temperature at or below which the rate of
2
2
progress towards flowering is zero. r or R is the coefficient of determination generated
for linear and non-linear crop models, respectively. In panel A, the petunia cultivar
‘Bravo Blue’ grown during replication 1 (●) was substituted by ‘Single Frost Blue in
replication 2 (○). Flowering time data from the first replicate did not show a significant
trend (P ≤ 0.05) in impatiens and could not be used for statistical analysis. Tmax and Topt
in panel R was fixed at 35 ºC and 25.1 ºC respectively to allow for precise estimation of
the other model parameters using 50 observations from the second replicate. .....81
Figure 2.3. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and inflorescence diameter (E) in
American marigold ‘Inca II Mix’ for replication 1 (●) and replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was
pooled (■) for statistical analysis. Each symbol represents the treatment means, and error
bars represent 95% confidence intervals. NS = nonsignificant at P ≤ 0.05. Dashed and
solid lines in panel A represent regression equations for replication 1 and 2, respectively.
−2 −1
The mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d ,
respectively. ...........................................................................................................82
Figure 2.4. The effect of mean daily temperature on flower number (A), axillary branch number
(B), height (C), shoot dry weight (D), and flower diameter (E) in cupflower ‘Purple
viii

Robe’ for replication 1 (●) and replication 2 (○). When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical
analysis. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines in panel B represent regression equations for
replication 1 and 2, respectively. The mean daily light integral for replication 1 and 2 was
−2 −1
18 and 9 mol∙m ∙d , respectively.......................................................................83
Figure 2.5. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), stem length (C), shoot dry weight (D), and flower diameter (E) in diascia
‘Diamonte Mix’ for replication 1 (●) and replication 2 (○). When regression slopes were
non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. NS = nonsignificant at P ≤ 0.05. Dashed lines in panel B
represent regression equations for replication 1. The mean daily light integral for
−2 −1
replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.................................84
Figure 2.6. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
flowering tobacco ‘Perfume Deep Purple’ for replication 1 (●) and replication 2 (○).
When regression slopes were non-significant between the two replications (P ≤ 0.05),
data was pooled (■) for statistical analysis. Each symbol represents the treatment means,
and error bars represent 95% confidence intervals. Dashed and solid lines in panel A
represent regression equations for replication 1 and 2, respectively. The mean daily light
−2 −1
integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively..............85
Figure 2.7. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and flower diameter (E) in geranium ‘Pinto
Red’ for replication 1 (●) and replication 2 (○). When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical
analysis. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines in panel A and C represent regression
equations for each parameter in replication 1 and 2, respectively. The mean daily light
−2 −1
integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively..............86
Figure 2.8. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
geranium ‘Ringo 2000 Deep Red’ for replication 1 (●) and replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was
pooled (■) for statistical analysis. Each symbol represents the treatment means, and error
bars represent 95% confidence intervals. Dashed and solid lines in panel A and C
represent regression equations for each parameter in replication 1 and 2, respectively. The
−2 −1
mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.
................................................................................................................................87
ix

Figure 2.9. The effect of mean daily temperature on inflorescence number (A), height (B), shoot
dry weight (C), and inflorescence diameter (D) in gerbera ‘Jaguar Deep Orange’ for
replication 1 (●) or ‘Jaguar Deep Rose’ for replication 2 (○). When regression slopes
were non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. NS = nonsignificant at P ≤ 0.05. Each symbol represents the treatment
means, and error bars represent 95% confidence intervals. Dashed and solid lines in panel
A represent regression equations for replication 1 and 2, respectively. The mean daily
−2 −1
light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively. ....88
Figure 2.10. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), and shoot dry weight (D) in globe amaranth ‘Gnome Purple’ for
replication 1 (●) and replication 2 (○). When regression slopes were non-significant
between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. All
plants died at 14 °C, and therefore data could not be collected. Each symbol represents
the treatment means, and error bars represent 95% confidence intervals. The mean daily
−2 −1
light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively. ....89
Figure 2.11. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and flower diameter (E) in heliotrope ‘Blue
Wonder’ for replication 1 (●) and replication 2 (○).When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical
analysis. NS = nonsignificant at P ≤ 0.05. Each symbol represents the treatment means,
and error bars represent 95% confidence intervals. Dashed and solid lines in panel A and
B represent regression equations for each parameter in replication 1 and 2, respectively.
−2 −1
The mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d ,
respectively. ...........................................................................................................90
Figure 2.12. The effect of mean daily temperature on flower number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
impatiens ‘Blitz 3000 Deep Orange’ for replication 1 (●) and replication 2 (○). NS =
nonsignificant at P ≤ 0.05. Each symbol represents the treatment means, and error bars
represent 95% confidence intervals. Dashed and solid lines in panel A, B and C represent
regression equations for each parameter in replication 1 and 2, respectively. The mean
−2 −1
daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.
................................................................................................................................91
Figure 2.13. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
nemesia ‘Poetry White’ for replication 1 (●) and replication 2 (○).When regression slopes
were non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Dashed and solid lines in panel A and B represent regression
x

equations for each parameter in replication 1 and 2, respectively. The mean daily light
−2 −1
integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively..............92
Figure 2.14. The effect of mean daily temperature on flower number (A), axillary branch
number (B), height (C), shoot dry weight (D), and flower diameter (E) in New Guinea
impatiens ‘Divine Cherry Red’ for replication 1 (●) and replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was
pooled (■) for statistical analysis. NS = nonsignificant at P ≤ 0.05. Each symbol
represents the treatment means, and error bars represent 95% confidence intervals. The
−2 −1
mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.
................................................................................................................................93
Figure 2.15. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and inflorescence diameter (E) in
osteospermum ‘Asti Purple’ for replication 1 (●) and replication 2 (○). When regression
slopes were non-significant between the two replications (P ≤ 0.05), data was pooled (■)
for statistical analysis. NS = nonsignificant at P ≤ 0.05. Each symbol represents the
treatment means, and error bars represent 95% confidence intervals. Dashed and solid
lines in panel A represent regression equations for replication 1 and 2, respectively. The
−2 −1
mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.
................................................................................................................................94
Figure 2.16. The effect of mean daily temperature on flower number (A), axillary branch
number (B), stem length (C), shoot and root dry weight (D and E), and flower diameter
(F) in petunia ‘Bravo Blue’ or ‘Single Frost Blue’ for replication 1 (●) and replication 2
(○). Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. NS = nonsignificant at P ≤ 0.05. Dashed and solid lines in panel A,
B and C represent regression equations for each parameter in replication 1 and 2,
respectively. The mean daily light integral for replication 1 and 2 was 18 and 9
−2 −1
mol∙m ∙d , respectively. ....................................................................................95
Figure 2.17. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and inflorescence diameter
(E) in pot marigold ‘Bon Bon Orange’ for replication 1 (●) and replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was
pooled (■) for statistical analysis. Each symbol represents the treatment means, and error
bars represent 95% confidence intervals. Dashed and solid lines in panel B and C
represent regression equations for each parameter in replication 1 and 2, respectively. The
−2 −1
mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.
............................................................................................................................... 96
Figure 2.18. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), and shoot and root dry weight (D and E) in snapdragon ‘Liberty
xi

Classic Cherry’ for replication 1 (●) and replication 2 (○). Each symbol represents
treatment means, and error bars represent the standard error with 95% confidence levels.
Dashed and solid lines in panel A and B represent regression equations for each
parameter in replication 1 and 2, respectively. The mean daily light integral for
−2 −1
replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.................................97
Figure 2.19. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
stock ‘Hot Cakes Purple’ for replication 1 (●) or ‘Garden Vintage Mix’ for replication 2
(○). When regression slopes were non-significant between the two replications (P ≤
0.05), data was pooled (■) for statistical analysis. Each symbol represents the treatment
means, and error bars represent 95% confidence intervals. NS = nonsignificant at P ≤
0.05. Dashed and solid lines in panel B and C represent regression equations for each
parameter in replication 1 and 2, respectively. The mean daily light integral for
−2 −1
replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.................................98
Figure 2.20. The effect of mean daily temperature on flower number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
torenia ‘Clown Blue’ for replication 1 (●) and replication 2 (○). When regression slopes
were non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Dashed and solid lines in panel A and C represent regression
equations for each parameter in replication 1 and 2, respectively. The mean daily light
−2 −1
integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively..............99
Figure 3.1. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of American marigold ‘Inca II Mix’ grown at 17 or 23 °C. Data were
pooled between two experimental replications when there were no replicate and treatment
interactions (P ≤ 0.05). Error bars represent the standard errors. Means followed by the
same letter are not statistically different (P ≤ 0.05). Mean DLIs during the fall and spring
–2 –1
experiments were 10.0 and 15.0 mol·m ·d , respectively. ..............................128
Figure 3.2. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of geranium ‘Ringo 2000 Deep Red’ grown at 17 or 23 °C. Data were
pooled between two experimental replications when there were no replicate and treatment
interactions (P ≤ 0.05). When the interaction was significant, data are separately
presented for fall (dots) and spring (stripes) replications. Error bars represent the standard
errors. Means followed by the same letter are not statistically different (P ≤ 0.05). Mean
–2 –1
DLIs during the fall and spring experiments were 10.0 and 15.0 mol·m ·d ,
respectively. .........................................................................................................129
Figure 3.3. The effect of starting transplant size (128- or 36-cell trays) on flowering
characteristics of gerbera ‘Jaguar Deep Orange’ grown at 17 or 23 °C. Data were pooled
xii

between two experimental replications when there were no replicate and treatment
interactions (P ≤ 0.05). When the interaction was significant, data are separately
presented for fall (dots) and spring (stripes) replications. Error bars represent the standard
errors. Means followed by the same letter are not statistically different (P ≤ 0.05). Mean
–2 –1
DLIs during the fall and spring experiments were 10.0 and 15.0 mol·m ·d ,
respectively. .........................................................................................................130
Figure 3.4. The effect of starting transplant size (128- or 36-cell trays) on flowering
characteristics of osteospermum ‘Asti Purple’ grown at 17 or 23 °C. Data were pooled
between two experimental replications when there were no replicate and treatment
interactions (P ≤ 0.05). When the interaction was significant, data are separately
presented for fall (dots) and spring (stripes) replications. Error bars represent the standard
errors. Means followed by the same letter are not statistically different (P ≤ 0.05). Mean
–2 –1
DLIs during the fall and spring experiments were 10.0 and 15.0 mol·m ·d ,
respectively. .........................................................................................................131
Figure 3.5. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of snapdragon ‘Rocket Mix’ grown at 17 or 23 °C. Data were pooled
between two experimental replications when there were no replicate and treatment
interactions (P ≤ 0.05). When the interaction was significant, data are separately
presented for fall (dots) and spring (stripes) replications. Error bars represent the standard
errors. Means followed by the same letter are not statistically different (P ≤ 0.05). Mean
–2 –1
DLIs during the fall and spring experiments were 10.0 and 15.0 mol·m ·d ,
respectively. .........................................................................................................132
Figure 4.1. The effect of transferring plants at different stages of development (leaf number)
–1
–1
from long days (LD, 16 h·d ) to short days (SD, 10 h·d ) (●), and from SD to LD (○)
during the transplant stage on flowering characteristics of three petunia varieties. Plants
were transplanted at 12-leaf stage and grown in a greenhouse maintained at a constant
setpoint of 20 °C with a 10 h photoperiod. Each symbol represents the treatment means,
and error bars represent the standard errors. Means followed by the same letter are not
statistically different (P ≤ 0.05). Treatments that had ≤ 30% flowering are not shown,
except for flowering percentage...........................................................................158
Figure 4.2. The effect of transferring plants at different stages of development (leaf number)
–1
–1
from long days (LD, 16 h·d ) to short days (SD, 10 h·d ) (●), and from SD to LD (○)
during the transplant stage on flowering characteristics of three petunia varieties. Plants
were transplanted at 12-leaf stage and grown in a greenhouse maintained at a constant
setpoint of 20 °C with a 10 h photoperiod. Each symbol represents the treatment means,
and error bars represent the standard errors. Means followed by the same letter are not
statistically different (P ≤ 0.05). Treatments that had ≤ 30% flowering are not shown
..............................................................................................................................159
xiii

–1

Figure 4.3. The effect of photoperiod (10 or 16 h·d ) on flowering characteristics of three
petunia varieties grown at 14 or 20 °C. Plants were initially grown at a constant setpoint
of 20 °C with a 16 h photoperiod in an environmental growth chamber, and then
transplanted when they averaged 8 leaves. Error bars represent the standard errors. Means
followed by the same letter are not statistically different (P ≤ 0.05). Treatments that had ≤
30% flowering are not shown, except for flowering percentage. ........................160
–1

Figure 4.4. The effect of photoperiod (10 or 16 h·d ) on flowering characteristics of three
petunia varieties grown at 14 or 20 °C. Plants were initially grown at a constant setpoint
of 20 °C with a 16 h photoperiod in an environmental growth chamber, and then
transplanted when they averaged 8 leaves. Error bars represent the standard errors. Means
followed by the same letter are not statistically different (P ≤ 0.05). Treatments that had ≤
30% flowering are not shown. .............................................................................161
Figure 5.1. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Blue’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................174
Figure 5.2. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Burgundy Star’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................175
Figure 5.3. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Neon Rose’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression

xiv

−2 −1

equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................176
Figure 5.4. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Pink’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................177
Figure 5.5. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Plum Vein’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................178
Figure 5.6. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Red Improved’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................179
Figure 5.7. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Violet’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................180

xv

Figure 5.8. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave White’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................181
Figure 5.9. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Coconut’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................182
Figure 5.10. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Denim’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................183
Figure 5.11. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Purple’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................184
Figure 5.12. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Tidal Wave Silver’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
xvi

of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................185
Figure 5.13. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Blue’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................186
Figure 5.14. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Pink’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................187
Figure 5.15. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Purple Classic’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................188
Figure 5.16. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Purple Improved’ grown under an average daily light integral of 7 mol·m ·d
−2 −1
(open symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the
reciprocal of days to flower. Tmin is the estimated temperature at or below which the rate
of progress towards flowering is zero. Each symbol represents the treatment means, and
error bars represent 95% confidence intervals. Dashed and solid lines represent regression
−2 −1
equations for the 7 and 14 mol·m ·d DLI treatments, respectively. NS =
nonsignificant at P > 0 .05. ..................................................................................189

xvii

SECTION I

LITERATURE REVIEW

1

Introduction
In the United States, bedding and garden plants are the largest contributors (48%) to the
total value of production of all reported floriculture crops, with a 15-state wholesale value of
$1.90 billion in 2011 (USDA, 2012). Michigan ranks third among 15 states, with a reported
wholesale value of > $260 million for bedding and garden plants (USDA, 2012). Bedding plants
are typically grown during some of the coldest months of the year, especially in northern
climates, where heating typically accounts for ≥ 10% of the annual greenhouse production costs
(Bartok, 2001). Volatile fuel prices, an increased emphasis on sustainable production practices,
and declining profit margins have motivated growers to optimize their greenhouse environments
and grow crops as energy-efficiently and as profitably as possible. One way of minimizing the
input costs is to reduce crop production duration in the greenhouse. This can reduce production
cost by minimizing heating and labor costs, fertilizers, water, and other resources needed to
maintain the crops in the greenhouse, and can also potentially increase profitability by enabling
more production cycles in a season.
Crop production time depends primarily on the species and cultivar (Blanchard, 2009),
starting plant size and characteristics (Fisher et al., 2006), desired finish size, average daily
temperature (ADT) (Blanchard and Runkle, 2011a), photosynthetic daily light integral (DLI)
(Blanchard, 2009; Moccaldi and Runkle 2007; Pramuk and Runkle, 2005b), and photoperiod
(Adams et al., 1997; Adams et al., 1998a). Although fast cropping is an appealing concept, it can
result in reduced plant quality. Poor plant quality characteristics, such as tall and weak stems,
few branches, few and small flowers, and reduced biomass, can lower consumer appeal. Plant
quality is mainly determined by the genotype, cultural practices (e.g., root zone management),
and aforementioned environmental conditions (Liu and Heins, 1997). In particular, temperature

2

and light are two environmental factors that primarily influence plant growth and development
and thus, plant quality attributes.
Plant growth, defined as an irreversible increase in plant size, is a function of biomass
production driven by photosynthesis (Heins et al., 2000). Plants use light energy with
wavelengths from approximately 400 to 700 nm, and this waveband is termed photosynthetically
active radiation (PAR). The integrated PAR received during a 24-h period is called the DLI. The
DLI received outdoors depends on the location, time of the year, and regional weather patterns.
−2 −1

For example, the average outdoor DLI in December ranges from 5 to 10 mol·m ·d
−2 −1

Michigan to 20 to 25 mol·m ·d

in
−2 −1

in Florida; in June, the average DLI is 40 to 45 mol·m ·d

in both states (Korczynski et al., 2002). Actual DLI received by plants inside a greenhouse is
reduced substantially (typically by 40 to 60%) by shading from glazing material and obstructions
that block light (Hanan, 1998). Therefore, the actual average DLI received inside a commercial
−2 −1

Michigan greenhouse usually ranges from <5 mol·m ·d

−2 −1

in December to >20 mol·m ·d

in

May and June. High-density planting can further reduce the DLI available to individual plants
due to shading from adjacent plants or ones hanging overhead. Quantum sensors installed at
−2 −1

canopy height can be used to measure the PAR (in µmol·m ·s ) and averages can be recorded
by a datalogger. The DLI can be determined by integrating those values, or by a light meter
specifically designed to measure and calculate the DLI.
Plant development, defined as the process of maturation of plant organs, is a function of
temperature-dependent metabolic processes (Heins et al., 2000). The developmental rate in
plants is mainly controlled by the ADT. ADT can be calculated by measuring the temperature on
a frequent basis (e.g., every 10 s) with a shielded and aspirated thermocouple and then
calculating the daily mean. The rate of plant development is controlled by the plant shoot-tip

3

temperature (Harris and Scott, 1969), which is a function of air temperature, glazing material
temperature, solar radiation, and vapor-pressure deficit during the day (Faust and Heins, 1998;
Shimizu et al., 2004). Faust and Heins (1998) showed that vinca (Catharanthus roseus L.) shoottip temperature was always below air temperature at night, whereas during the day, it was ±2 °C
of air temperature when grown at <25 °C and was 4 to 6 °C lower when grown at >25 °C. In
commercial greenhouses, nearly all environmental control systems measure the air temperature,
and it is extremely uncommon for shoot-tip temperature to be measured.
The present downturn in the economy has led consumers to critically evaluate product
features and prices before they make a purchase (Hall, 2011). This has led to increased market
specifications in terms of plant quality and delivery date, while prices have to be maintained
competitive. Days to flower (DTF) generally decreases as ADT and DLI increase (Blanchard,
2009; Moccaldi and Runkle, 2007; Pramuk and Runkle, 2005b). At the same time, plant quality
of shade-avoiding crops typically improves as ADT decreases and DLI increases (Blanchard,
2009; Pramuk and Runkle, 2005b; Moccaldi and Runkle, 2007). Thus, it is important to quantify
the effects of temperature and DLI on plant growth, development, and quality parameters so that
growers can balance crop timing with quality.

Temperature Effects on Plant Development
Plant developmental rate depends on the rate of biochemical reactions in meristematic
tissues. As temperature increases, activation energy of the reaction molecules driving these
biochemical reactions increases and consequently, the number of ‘active’ molecules increases
according to the Boltzman energy distribution (Jones, 1983). The rate of a biochemical reaction
increases with an increase in the number of molecules with an activation energy equal to or

4

higher than the minimum activation energy required for that particular reaction (Jones, 1983).
Therefore, as the rate of biochemical reactions increase with temperature, plants mature at a
faster rate, at least up to some species-specific maximum temperature. Plant development rate is
mainly controlled by ADT, while plant morphology is influenced by ADT, the difference
between day and night temperatures (DIF), and other environmental factors such as light quality.
Average daily temperature regulates plant development by influencing different events in
a plant’s life cycle such as germination, leaf unfolding, and flowering. The time taken to reach a
particular developmental event as a function of temperature can be expressed as the number of
days to reach that event (n) or as a developmental rate by taking the reciprocal of that number of
days (1/n). Different functions have been used to describe the relationship between the rate of
plant development and ADT. For many floriculture crops, the relationship between the rate of
plant development and ADT has been described by a simple linear function when the ADT is
above the species-specific base temperature (Tb) and at or below the optimum temperature (Topt)
(Adams et al., 1997; Clough et al, 2001; Karlsson et al., 1988; Larsen and Hiden, 1995; Niu et
al., 2001; Park et al, 1998; Whitman et al., 1997; Yuan et al., 1998). For example, the rate of
progress to flowering increased linearly with temperature in the range of 15.2 to 29.8 °C for
sundrop (Oenothera fruticosa L.) (Clough et al, 2001), 14.0 to 29.0 °C for balloon flower
[Platycodon gradiflorus (Jaq.)] cv. Astra Blue (Park et al, 1998), 15.0 to 27.0 °C for tussock
bellflower (Campanula carpatica Jacq). cv. Blue Clips (Niu et al., 2001; Whitman et al., 1997)
and 15.0 to 26.0 °C for largeflower tickseed (Coreopsis grandiflora Hogg ex Sweet),
blanketflower (Gaillardia ×grandiflora Van Houtte), Shasta daisy [Leucanthemum ×superbum
(Bergmans ex J.W. Ingram)] and black-eyed Susan (Rudbeckia fulgida Aiton) (Yuan et al.,

5

1998). Similarly, Karlsson et al. (1988) reported a linear increase in leaf unfolding rate with
temperature between 14.0 and 30.0 °C in ‘Nellie White’ Easter lily (Lilium longiflorum Thunb.).
Within a species-specific temperature range, each degree rise in temperature increases the
rate of plant development by the same incremental amount. This linear relationship between
temperature and plant development rate can be useful in estimating the Tb and the thermal time
(θ) required for a particular developmental event (Roberts and Summerfield, 1987). The base
temperature is the minimum temperature below which no development occurs, and thermal time
is the summation of all degree days above Tb required for a particular developmental event to
occur. A linear relationship is mathematically expressed by the following equation:
1/d = a + bT

(1)

where 1/d is rate of development of a particular event, a = slope, b = intercept, and T = ADT
above Tb and below Topt. Given that the developmental rate at Tb is zero, equation (1) can be
used to calculate Tb:
Tb = –a / b

(2)

The thermal time required for a particular developmental event can be calculated from equation
(1) as:
θ=1/b

(3)

When ADT exceeds the Topt, non-linear functions such as polynomial and exponential
functions have been used for some floricultural species. Karlsson et al. (1991) and Karlsson and
Werner (2001) expressed leaf unfolding rate in hibiscus (Hibiscus rosa-sinensis L.) and
cyclamen (Cyclamen persicum Mill.) as a polynomial-cubic function of temperature. In contrast,
an exponential model was proposed by Larsen (1988) to describe the rate of germination and rate

6

of vegetative and floral development in cineraria (Senecio ×hybridus Hyl.) as a function of
temperature.
Plant developmental rate increases linearly from Tb to Topt, which are specific to each
species. For any species, when temperature is below Tb or exceeds a maximum temperature
(Tmax), the developmental rate becomes zero. Above Tb, plant developmental rate increases to a
maximum value at Topt. Beyond Topt, the rate decreases and ultimately development ceases at
Tmax. Plant injury and death can occur with prolonged exposure to temperatures below Tb or
above Tmax from irreversible damage to the photosynthetic apparatus and cell membranes (Lie
and Huang, 2000; Raison et al., 1980).
The species-specific Tb, Topt, and Tmax are usually determined by the natural origin and
distribution of the species. Tb has been estimated for many floriculture crops and has varied from
≤0.0 °C to ≥10.0 °C (Blanchard, 2009; Clough et al., 2001; Park et al., 1998; Whitman et al.,
1997; Yuan et al., 1998). For example, tussock bellflower, native to the Carpathian mountains of
Eastern Europe, had an estimated Tb for flowering of 0.0 °C (Whitman et al., 1997). In contrast,
for browalia (Browalia speciosa Hook.) and summer snapdragon (Angelonia augustifolia
Benth.), both of tropical origin, the estimated Tb was 9.0 and 10.0 °C, respectively (Blanchard,
2009). Breeding efforts over the years have introduced desirable traits such as early flowering,
frost tolerance, and heat tolerance. Therefore, depending on the breeding objective, cultivars of
the same species may differ in their base and optimum temperatures, as well as thermal time for
a particular event such as flowering. For example, the Tb calculated for petunia (Petunia ×
hybrida Vilm.-Andr.) ‘Dreams Neon Rose’ was 2.7 °C lower than that of ‘Wave Purple’
(Blanchard, 2009). Base and optimum temperatures within a species can also vary with the
developmental process (Clough et al., 2001; Pearson et al., 1995; Yuan et al., 1998; Whitman et

7

al., 1997), physiological process (Van Iersel, 2003) and environmental conditions such as
photoperiod (Adams et al., 1997; 1998a; 1998b) and DLI (Adams et al., 1997; Faust and Heins,
1993; Pramuk and Runkle, 2005b). Blanchard (2009) reported a greater delay in flowering at
lower temperatures for species with a higher Tb as compared to species with lower Tb values
(Table 1). Estimation of Tb and Topt allows one to categorize crops according to their
temperature sensitivity and improve greenhouse crop energy efficiency (Blanchard, 2009). In the
sections below, the effects of temperature on vegetative growth, flowering, and plant quality
characteristics are discussed.
Leaf Unfolding Rate (LUR)
Plant growth can be divided into two developmental phases: vegetative and reproductive.
Flower initiation marks the transition from the vegetative phase to the flowering phase. In some
plants like Easter lily and chrysanthemum, flower initiation terminates leaf production so that a
specific number of leaves (depending on when flower initiation occurs) are formed before
flowering. Plant development can be quantified by the number of leaves formed prior to flower
initiation and the subsequent rate of leaf unfolding (Faust and Heins, 1993). In some
indeterminate plants like begonia (Begonia ×hiemulis Fotsch), flower initiation occurs at a
particular plant size (leaf number) when grown under short days. The ADT affects the leaf
unfolding rate (LUR) such that, as temperature increases, leaves unfold at a faster rate up to Topt,
and thereafter the rate starts to decline (Brøndum and Heins, 1993; Faust and Heins, 1993;
Karlsson and Werner, 2001; Karlsson, 1992; Karlsson et al., 1991; Larsen and Hiden, 1995).
Therefore, knowledge of the leaf number required to unfold prior to flowering in determinate
plants (Easter lily and chrysanthemum) or the minimum leaf count for flower initiation to occur
(begonia) and the LUR at different temperatures allow for precise scheduling of these plants.
8

The effect of ADT on leaf development has been studied for many floriculture crops
including African violet (Saintpaulia ionantha Wendl.) (Faust and Heins, 1993), begonia
(Karlsson, 1992), chrysanthemum (Cockshull et al., 1981; Karlsson et al., 1989b; Larsen and
Hiden, 1995), cyclamen (Karlsson and Werner, 2001), dahlia (Dahlia pinnata Cav.) (Brøndum
and Heins, 1993), Easter lily (Karlsson et al., 1988; Roberts et al., 1983; Wang and Roberts,
1983), hibiscus (Karlsson et al., 1991), pansy (Viola × wittrockiana Gams.) (Adams et al., 1997),
poinsettia (Euphorbia pulcherrima Willd. ex Klotzsch.) (Berghage et al., 1990b) and vinca
(Pietsch et al., 1995). Although LUR response has been frequently described as a function of
average daily air temperature (Adams et al., 1997; Karlsson et al., 1988; Karlsson et al., 1989b;
Karlsson, 1992; Karlsson and Werner, 2001; Larsen and Hiden, 1995; Pietsch et al., 1995),
average hourly air temperatures (Karlsson et al., 1991) and average plant temperatures (Faust
and Heins, 1993) have been used in some studies. In greenhouses where diurnal temperature
fluctuations fall outside the linear range of a plant’s thermal development response, average
hourly temperature predicts LUR more accurately than ADT (Karlsson et al., 1991). Faust and
Heins (1993) reported a greater accuracy (63%) in predicting LUR when average plant
temperatures were used as opposed to average hourly or daily air temperatures. Individual
effects of day and night temperatures on the LUR of Easter lily (Karlsson et al., 1988) and
poinsettia (Berghage et al., 1990b) have also been investigated. However, LUR was a function of
ADT and not only day or night temperature. Wang and Roberts (1983) reported that high soil
temperature (≥24 °C) promoted leaf unfolding in Easter lily when air temperature was lower.
This promotive effect of higher soil temperature on the LUR was probably from an indirect
increase in plant temperature, air temperature surrounding the plant, or both, since the LUR
increased with an increase in air temperature even for unheated soil.

9

Researchers have used different functions, such as linear (Adams et al., 1997; Karlsson et
al., 1988; Karlsson et al., 1989b), quadratic (Karlsson, 1992) or cubic (Karlsson et al., 1991;
Karlsson and Werner, 2001), to quantify the relationship between LUR and temperature. LUR
increased linearly as ADT increased from 0.2 leaves·d
−1

in chrysanthemum, and 1.2 leaves·d

−1

−1

at 10.0 °C to 0.5 leaves·d at 30.0 °C

at 14.0 °C to 2.7 leaves·d

−1

at 30.0 °C in Easter lily

(Karlsson et al., 1988; Karlsson et al., 1989b). Karlsson (1992) reported a quadratic increase in
the LUR with temperature between 13.0 and 28.0 °C in begonia under long days prior to flower
initiation, with Topt at 21.0 °C. Similarly, a cubic-polynomial function was found to best
describe the LUR as a function of temperature between 10.0 and 35.0 °C in hibiscus (Karlsson et
al., 1991) and between 8.0 and 24.0 °C in cyclamen (Karlsson and Werner, 2001). The cubicpolynomial model predicted a maximum rate of 0.23 leaves·d
et al., 1991) and 0.33 leaves·d

−1

−1

at 32.0 °C for hibiscus (Karlsson

at 19.1 °C for cyclamen (Karlsson and Werner, 2001), beyond

which the LUR began to decrease. When temperatures above Topt are included in the
experimental range, the temperature response curve follows an asymmetric peak shape, such that
the absolute value of LUR decreases more rapidly at temperatures greater than Topt as compared
to increase in LUR from Tb to Topt (Brøndum and Heins, 1993; Faust and Heins, 1993; Karlsson
and Werner, 2001; Larsen and Hiden, 1995).
At any given temperature, the LUR varies widely among floriculture crops. For example,
plants at an ADT of 18.0 °C unfolded 0.08 leaves·d
(Pietsch et al., 1995), 0.11 leaves·d

−1

−1

in vinca under supplemental lighting

in begonia (Karlsson, 1992), 0.18 leaves·d

(Berghage et al., 1990b), 0.21 leaves·d
(Faust and Heins, 1993), 0.23 leaves·d

−1

−1

−1

in poinsettia

−2 −1

in African violet under 10.0 mol·m ·d

of light

in hibiscus (Karlsson et al., 1991), 0.33 leaves·d

cyclamen (Karlsson and Werner, 2001), 0.34 leaves·d

10

−1

−1

in chrysanthemum (Karlsson et al.,

in

1989b) and 1.5 leaves·d

−1

in Easter lily (Karlsson et al., 1988). Although differences in the

LUR among these species are largely due to genetic variations in temperature response, to some
extent, differences in experimental conditions such as DLI and the use of air versus plant
temperature to describe the LUR response (Faust and Heins, 1993) probably also influenced
these inter-specific rates. Similarly, Tb for LUR ranged from 1.6 °C for chrysanthemum
(Karlsson et al., 1989b) to 9.8 °C for hibiscus (Karlsson et al., 1991) and Topt ranged from 19.1
°C for cyclamen (Karlsson and Werner, 2001) to 32.0 °C for hibiscus (Karlsson et al., 1991). In
begonia (Karlsson, 1992), cyclamen (Karlsson and Werner, 2001) and hibiscus (Karlsson et al,
1991), the LUR of different cultivars within a species was found to be similar. However,
cultivars or species that have a similar LUR may not flower at the same time due to differences
in the number of leaves formed prior to flower initiation (Karlsson et al., 1991).

Flowering
Temperature influences flowering time by affecting flower initiation, development, or
both. Many studies have quantified the influence of temperature on flowering time, and reported
a decrease in DTF with an increase in ADT, within a species-specific range (Adams et al., 1999;
1998b; 1998c; Blanchard and Runkle, 2011a; Clough et al., 2001; Kanellos and Pearson, 2000;
Miller and Armitage, 2002; Moccaldi and Runkle, 2007; Niu et al., 2001; 2000; Pramuk and
Runkle, 2005b; White and Warrington, 1988). A decrease in DTF with an increase in ADT may
be due to earlier flower initiation, which can be observed by repeated apical dissections or
reduced leaf number below the first open flower (Adams et al., 1998b; 1998c; Mattson and
Erwin, 2003), or from faster flower development rate (Brøndum and Heins, 1993; Faust and
Heins, 1994; Karlsson and Werner, 2001; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al.,

11

1998). When plants are subjected to ADT above Topt, the flower development rate (reciprocal of
DTF) begins to decline (Blanchard and Runkle, 2011a; Brøndum and Heins, 1993; Faust and
Heins, 1994; Semeniuk, 1975). Exposure to high temperatures caused a developmental delay in
flowering (greater node number prior to flowering) in calendula, chrysanthemum, impatiens,
pansy, snapdragon and torenia (Warner and Erwin, 2005; 2006; Whealy et al., 1987). A decrease
in flowering rate at temperatures > Topt is known as heat delay, and may be due to a delay in
flower induction, initiation and/or development (Warner and Erwin, 2006). Biochemically,
inhibition of flowering at higher temperatures may be due to peroxidation of membrane
unsaturated fatty acids that causes cell damage (Gutteridge and Halliwell, 1990).
In many plants, a linear function has been used to describe the relationship between
flowering rate and ADT, when ADT is between Tb and Topt (Clough et al., 2001; Niu et al.,
2001; 2000; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al., 1998). In other crops, a
quadratic function best described the effect of ADT on DTF, since the decrease in DTF with an
increase in ADT was not linear between Tb and Topt (Armitage et al., 1981; Clough et al., 2001;
Niu et al., 2001; 2000; Park et al., 1998; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al.,
1998). For example, a greater decrease in flowering time was observed when ADT increased
from 15 to 20 °C compared to when ADT increased from 20 to 25 °C. Although linear and
quadratic equations can be used to simply describe these responses, they lack biological
significance because their constants do not have physiological meaning (Landsberg, 1977).
Several studies have quantified the effect of ADT on flowering time using mathematical
models that contain parameters of biological significance (Tmax, Topt and/or Tb) (Blanchard and
Runkle, 2011a; Blanchard et al., 2011a; Brondum and Heins, 1993; Hiden and Larsen, 1994;
Larsen, 1988; 1989; Larsen and Persson, 1999). Tb and Topt for flower development have been

12

estimated for numerous species and vary among species and in some cases also with the phase of
development. For example, Tb for flower development varied from 0.0 °C in tussock bellflower
(Whitman et al., 1997) to 9.9 °C in summer snapdragon (Blanchard and Runkle, 2011a), and
Topt varied from 19.1 °C in dahlia to 28.0 °C in blue salvia (Salvia farinacea Benth.) (Blanchard
and Runkle, 2011a). Similarly, a few studies have also reported different Tb and Topt for
different phases of flower development in blanketflower, chrysanthemum, Shasta daisy, sundrop,
and tussock bellflower (Clough et al., 2001; Karlsson et al., 1989a; Whitman et al., 1997; Yuan
et al., 1998).
The effect of DIF on flowering development has been evaluated in many plants. When
DT and NT were within the species-specific linear range, flowering time was a function of ADT
and did not vary between DT or NT (Blanchard and Runkle, 2011b; Brøndum and Heins, 1993;
Cockshull et al., 1981; Erwin and Heins, 1990; Lepage et al., 1984; Moe, 1990; Mortensen and
Moe, 1992; Niu et al., 2001; 2000; Pietsch et al., 1995). Since ADT and not DIF controls
flowering time in plants, a +DIF can be used to potentially lower heating costs, since most
greenhouse heating occurs at night (Blanchard and Runkle, 2011b). However, PGR application
or other height control strategies may be required when plants are grown at a +DIF.
A few studies have evaluated the influence of root-zone temperature on flowering time of
a few species (Vogelezang, 1990; 1992; Wai and Newman, 1992). A warmer root zone radiates
heat to the surrounding cool air, which is evident from the increase in air temperature as rootzone temperature increased (Vogelezang, 1990; 1992). Therefore, the promotive effect of higher
root-zone temperature on flowering time can be attributed to an increase in air and thus, shoot-tip
temperature.

13

Although flowering time response to ADT has been extensively studied, the exact
mechanism that induces this response is still not known. However, genetic and molecular
research over the last decade has proposed that an independent thermal induction pathway
(thermosensory) may regulate flowering by affecting the expression of the FT (floral integrator)
gene (Blazquez et al., 2003; Sureshkumar et al., 2006). Blazquez et al. (2003) reported a
reduction in flowering time and a simultaneous increase in FT expression in mouse-ear cress
[Arabidopsis thaliana (L.) Heynh] at 23.0 °C compared to 16.0 °C. Similarly, Sureshkumar et al.
(2006) observed a 10-fold increase in mRNA levels of FT in mouse-ear cress at 27 °C compared
to 23 °C. In addition, changes in ambient temperature had little or no effect on DTF and leaf
number below first open flower in mouse-ear cress plants overexpressing FT (Blazquez et al.,
2003).
Several genes that play an important role in this pathway have already been identified in
mouse-ear cress: FCA and FVE which sense ambient temperature and regulate the flowering
time response (Blazquez et al., 2003), and SVP (short vegetative phase) and FLM (flowering
locus M) which act as negative regulators of FT expression (Lee et al., 2007; Scortecci et al.,
2003). FCA and FVE act separately through an FLC-independent pathway to regulate flowering
time in response to ambient temperature (Blazquez et al., 2003). SVP negatively regulates FT
expression by directly binding to the CArG motifs in the FT gene (Lee et al., 2007). In addition,
svp mutants flowered earlier especially at cooler temperatures (16.0 °C), while plants
overexpressing SVP flowered later especially at warmer temperatures (23.0 °C) (Lee et al.,
2007). Recently, Lee et al. (2010) demonstrated the involvement of miR172, regulated by SVP,
in the thermosensory pathway. Although several components of this pathway have been

14

identified, the sequence of these components, their exact function and how they interact to
influence temperature-dependent flowering in plants requires further research.

Effect of Temperature on Plant Quality
In addition to plant development, temperature (and other factors) influences plant quality
parameters including flower number and size, branch number, plant height, and plant biomass.
Several studies have quantified the influence of temperature on plant quality, and have reported a
decrease in plant quality with an increase in ADT when grown at the same or similar DLI
(Blanchard et al., 2011a; 2011b; Mattson and Erwin, 2003; Moccaldi and Runkle, 2007; Niu et
al., 2001; 2000; Pramuk and Runkle, 2005b). Plant quality is important to growers since buyers
purchase plants that are of the highest quality. The influence of temperature on flower number
and size, branch number, plant height, and plant biomass is discussed in detail below.

Effect of Temperature on Flower Number and Quality
Flower number and size are two of the factors that influence overall plant quality. A high
quality plant would typically have a relatively large number of big flowers and flower buds. In
many shade-avoiding species, ADT influences flower number such that as ADT increases,
flower number decreases (Blanchard et al., 2011a; 2011b; Brøndum and Hein, 1993; Clough et
al., 2001; Mattson and Erwin, 2003; Moccaldi and Runkle, 2007; Niu et al., 2001; 2000; Pramuk
and Runkle, 2005b; Whitman et al., 1997; Yuan et al., 1998). The magnitude of decrease in
flower number with increase in temperature varies widely among species and cultivars. For
example, an increase in ADT from 16.0 to 26.0 °C decreased the flower number of tussock
−2 −1

bellflower by 181% (under 10.8 mol·m ·d

and at 400 umol·mol

15

−1

CO2; Niu et al., 2001),

largeflower tickseed by 80% (Yuan et al, 1998), Shasta daisy by 55% (Yuan et al, 1998),
sundrop by 167% (Clough et al., 2001), Petunia axillaris (Lam.) Britton, et al. by 96% (Warner,
2001), black-eyed Susan by 75% (Yuan et al, 1998) and pansy by 84% (under 10.6
−2 −1

mol·m ·d ) (Niu et al., 2000). Similarly, in ‘Avalanche Pink’, ‘Dreams Rose’ and ‘Wave
Purple’ petunia cultivars, an increase in temperature from 14.0 to 24.0 °C decreased the flower
bud number by 67%, 340% and 28%, respectively (Mattson and Erwin, 2003). Increased flower
number at lower temperatures is sometimes correlated with an increase in leaf number (Mattson
and Erwin, 2003) or branch number (Mattson and Erwin, 2003; Warner and Erwin, 2006).
However, flower bud number increased as temperature increased in some species such as
Chinese lantern lily (Sandersonia aurantiaca Hook.) (Davies et al., 2002), chrysanthemum
(Carvalho et al., 2005), cyclamen (Karlsson and Werner, 2001), and orchid pansy (Achimenes
Pers.; Vlahos et al., 1992). For these crops, increased flower number at higher temperatures was
correlated with an increase in leaf number (Carvalho et al., 2005; Karlsson and Werner, 2001;
Vlahos et al., 1992) or with a reduction in aborted flowers (Davies et al., 2002).
Flower quality is often determined by its size (diameter, area or dry weight) and color
intensity. For many crops, ADT and flower size are inversely related (Brøndum and Heins, 1993;
Carvalho et al., 2005; Clough et al., 2001; Niu et al., 2001; 2000; Moccaldi and Runkle, 2007;
Pearson et al., 1995; Whitman et al., 1997; Yuan et al., 1998). Each degree rise in ADT
decreased flower diameter or area by 0.97 mm in tussock bellflower ‘Blue Chips’ (under 10.8
−2 −1

mol·m ·d ) (Niu et al., 2001), 1.42 cm in dahlia ‘Royal Dahlietta Yellow’ (Brøndum and
Heins, 1993), 0.19 cm in sundrops ‘Youngii-lapsley’ (Clough et al., 2001) and 1.05 cm2 in pansy
‘Universal Violet’ (Pearson et al., 1995). Flower bud development requires carbon import from
the source organs (leaves). Sucrose hydrolysis in the sink organs (flower buds) is necessary to

16

establish a concentration gradient for carbon transport between the source and the sink (Dinar
and Rudich, 1985). High temperatures reduce sucrose hydrolysis and therefore increase its
concentration in flower buds, which reduces or inhibits uptake of carbon by the developing
flower buds (Dinar and Rudich, 1985), and may consequently reduce final flower size. This is
further evident from a reduction in the percentage of dry matter that is partitioned to the flowers
relative to vegetative structures under high temperature stress in chrysanthemum (Karlsson and
Heins, 1992).
Flower color is determined by the spectral properties of light reflected by certain
pigments in petals. The intensity of flower color in some plants, particularly those that are red,
blue, or purple, is greatly influenced by the ADT. A reduction in flower color intensity at higher
temperatures was observed in carnation (Dianthus caryophyllus L.) (Maekawa and Nakamura,
1977), chrysanthemum (Carvalho et al., 2005), rose (Dela et al., 2003) and petunia (Shvarts et
al., 1997). Anthocyanins are plant pigments that impart a red, blue or purple color to leaves and
flowers, and their concentrations are correlated with flower color intensity at different
temperatures. For example, at higher temperatures, a reduction in floral anthocyanin content was
correlated with a reduction in the color intensity (Nozaki et al., 2006; Stiles et al., 2007).
Temperature influences anthocyanin content by regulating its biosynthetic pathway (Shvarts et
al., 1997; Stiles et al., 2007). This may be an adaptive thermoregulatory response in plants; as
temperature increases, anthocyanin pigmentation decreases and the flower color is lighter and
thus more reflective (Stiles et al., 2007). Therefore, more reflection of radiation reduces flower
temperature and thus protects the floral organs from heat damage (Stiles et al., 2007). In
addition, a greater availability of sugars at lower temperatures may also increase the anthocyanin
content by up-regulating the anthocyanin biosynthetic pathway (Solfanelli et al., 2006).

17

Effect of Temperature on Branch Number
Lateral branch number is another parameter of plant quality, since an increase in branch
number is often correlated with an increase in flower number (Mattson and Erwin, 2003; Warner
and Erwin, 2006). Branch number decreased with an increase in ADT in fuchsia (Fuchsia ×
hybrida hort. ex Siebold & Voss) (Erwin et al., 1991), pansy (Warner and Erwin, 2006) and
petunia (Kaczperski et al., 1991; Mattson and Erwin, 2003). In some cases, at higher
temperatures, a decrease in branch number was correlated with a decrease in leaf number below
the first open flower (Mattson and Erwin, 2003).

Effect of Temperature on Plant Biomass
Temperature affects photosynthetic enzyme activity, and consequently it interacts with
other factors (especially light) to influence accumulated plant biomass. For many plants, shoot
dry mass at first flowering increases as ADT decreases, including balloon flower (Park et al.,
1998), calendula (Calendula officinalis L.) (Warner and Erwin, 2005), geranium (Pelargonium ×
hortorum Bailey) (White and Warrington, 1988), impatiens (Impatiens walleriana Hook. f.)
(Pramuk and Runkle, 2005b; Warner and Erwin, 2005), marigold (Tagetes patula L.) (Moccaldi
and Runkle, 2007), mimulus (Mimulus hybridus Hort. Ex Siebert & Voss) (Warner and Erwin,
2005), pansy (Niu et al., 2000; Warner and Erwin, 2006), salvia (Salvia splendens F. Sello ex
Roem & Schult.) (Moccaldi and Runkle, 2007) snapdragon (Antirrhinum majus L.) (Wai and
Newman, 1992; Warner and Erwin, 2005), and Texas firebush (Hamelia patens Jacq.)
(Armitage, 1995). These results are not surprising since plants grown at lower temperatures take
longer to flower (and thus can photosynthesize for a longer time) and are also of higher quality
(increased branch number and flower number and size) compared to plants grown at warmer

18

temperatures. In addition, van Iersel (2003) reported an increase in the net photosynthetic rate
(Pnet) with a decrease in ADT for geranium, marigold, pansy and petunia. This increase in Pnet at
lower temperatures was mainly attributed to a decrease in the dark respiration rate (Rdark), since
gross photosynthesis (Pnet + Rdark) was not greatly influenced by ADT. High temperatures
decrease Pnet by modifying Rubisco kinetics (Brooks and Farquhar, 1985), increasing O2:CO2
solubility (Ku and Edwards, 1977), or both.
A few exceptions to these plant biomass responses to ADT have been reported in celosia
(Celosia argentea L.) (Pramuk and Runkle, 2005b), summer snapdragon (Miller and Armitage,
2002), Texas firebush (Armitage, 1995), and wishbone flower (Torenia fournieri Linden ex E.
Fourn) (Warner and Erwin, 2005), in which plant biomass increased with an increase in ADT.
These plants have a high estimated Tb and therefore develop slowly and accumulate less biomass
at lower temperatures. For example, the estimated Tb for celosia, summer snapdragon and
−2 −1

wishbone flower was 10.2 °C (under a DLI of 15.0 mol·m ·d ) (Pramuk and Runkle, 2005b),
−2 −1

9.9 °C (Blanchard, 2009) and 9.4 °C (under a DLI of 18.0 mol·m ·d ) (Vaid, unpublished
data), respectively.
The effect of DIF has also been evaluated on dry matter partitioning patterns in some
floriculture species. Although the effect of DIF on whole plant biomass is variable, many studies
have reported an increase in stem and leaf dry weight as DIF increases (Hwang et al., 2005;
Karlsson and Heins, 1992; Miller et al., 1993; Myster et al., 1997). In many plants, a +DIF
promotes stem elongation, whereas a –DIF suppresses it. This strategy is commonly employed in
greenhouses to control plant height (Erwin and Heins, 1995).Stem length and dry weight of
Easter lily at anthesis was 79 and 80% greater, respectively when grown under a +8.0 °C DIF
than a –8.0 °C DIF (Miller et al., 1993). In addition, a reduction in photosynthesis due to a

19

decrease in chlorophyll content (Berghage et. al., 1990a) and/or an increase in Rdark (due to a
higher NT) at a –DIF may also contribute to a reduction in stem dry weight. In Easter lily plants
grown under a –DIF, Miller et al. (1993) reported a reduction in leaf and stem total soluble
carbohydrate content by 39 to 46% at visible bud (VB) and anthesis.

Effect of Temperature on Plant Height
Growers and consumers generally prefer short, compact plants since shipping costs can
be lower and the aesthetic value of plants can be increased. Plant growth regulators (PGRs) that
inhibit gibberellin biosynthesis are commonly applied in the bedding plant industry to inhibit
extension growth. However, due to the cost of chemicals and labor, several non-chemical
approaches to suppress plant height have been evaluated. For many plants, height at flowering
decreases as day temperature relative to the night temperature decreases (as the value of DIF
decreases) (Cockshull, 1981; Erwin et. al., 1991; Erwin and Heins, 1990; Grindal and Moe,
1994; Jensen et al., 1996; Karlsson et al., 1989b; Lepage et al., 1984; Moe, 1990; Moe and
Heins, 1990; Myster et al., 1997; Myster and Moe, 1995; Neily et al., 1997; Niu et al., 2001;
2000). For example, as DIF increased from –6.0 to +6.0 °C, plant height increased by 32% in
Chinese lantern lily (Davies et al., 2002), 40% in Easter lily (Erwin and Heins, 1990), 39% in
fuchsia (Erwin et al., 1991), 19% in geranium (Strefeler, 1995), and 9% in Italian bellflower
(Campanula isophylla Moretti; Moe and Mortensen, 1992b). An increase in stem length caused
by a +DIF is primarily from greater cell elongation and not cell division (Erwin et al., 1994).
Elongation of stem parenchyma and epidermal cells increases linearly as DIF increases from –15
to +15 °C, but cell width or number per internode was similar (Erwin et al., 1994). Stem
extension occurs more rapidly when DIF increases from zero to a positive value as compared to

20

when DIF increases from a negative value to zero (Erwin and Heins, 1990). For example, as DIF
increased from –15.0 to 0.0 °C and 0.0 to +15.0 °C, plant height increased by 29 and 61%,
respectively. Although DIF can be used to control plant height, more energy inputs may be
required to maintain a –DIF, especially during the winter (Blanchard and Runkle, 2011b). An
alternative height control strategy using temperature, but with potentially reduced energy
consumption, is a decrease in temperature for 2.0 to 4.0 h towards the end of the night or
beginning of the day (DROP), which is when stem extension is sometimes greatest. This
technique has been effective in controlling plant height in begonia, Easter lily and poinsettia
(Erwin, 1991; Grindal and Moe, 1994; Moe et. al., 1992).
Several studies have investigated the physiological mechanism of DIF. Evidence from
these studies suggests that gibberellins (GA) are involved in the thermoperiodic control of plant
height. Easter lily bulbs soaked in GA3 prior to planting and grown in a –DIF environment were
15% taller than those not treated with GA3 (Zieslin and Tsujita, 1988). In contrast, height
suppression by ancymidol, a GA biosynthesis inhibitor, increased as DIF increased (Erwin et al.,
1989). Grindal et al. (1998) reported a 60% decrease in endogenous GA content of stems in wild
type (WT) garden pea (Pisum sativum L.) plants grown under a –DIF as compared to those
grown under a +DIF. Using a GA-insensitive slender mutant, these authors reported 8% less
stem inhibition in the mutant than in the WT (40 to 50%) when grown under a –DIF.
Involvement of GA in thermoperiodic control of stem elongation is not unexpected, since the
role of GA in promoting cell elongation has long been established.
In addition to DIF, ADT can also influence plant height by affecting the number of nodes
formed prior to flowering. Since increase in ADT developmentally decreases flowering time in
many species, an increase in ADT would typically reduce plant height due to fewer nodes

21

formed prior to flowering. However, several studies have reported an increase in plant height
with an increase in ADT, although plants flowered earlier at higher temperatures (Armitage et
al., 2000; Blanchard et al., 2011a; Kanellos and Pearson, 2000; Miller and Armitage, 2002;
Moccaldi and Runkle, 2007; Pramuk and Runkle, 2005b). Since the number of nodes in these
studies was either variable or not reported, and the actual DIF values were also not provided, the
promotive effect of higher temperatures on plant height can not be clearly understood.

Daily Light Integral (DLI)
Light is electromagnetic radiation and has three characteristics: quantity, quality, and
duration. The spectral distribution of light is referred to as light quality, which acts as a signal for
plant morphogenesis. Blue light and the ratio of red to far-red light (R:FR) are the primary
wavebands that regulate stem extension, and in some cases flower initiation and development
(Runkle and Heins, 2001; Runkle et al., 2001). Duration of light or photoperiod regulates
flowering in many floriculture species (Mattson and Erwin, 2005). Light intensity refers to the
number of photons delivered on either an instantaneous or cumulative basis. As light intensity
within the photosynthetically active waveband (PAR; generally defined as 400 to 700 nm)
increases, the rate of photosynthesis increases until a species-specific saturation point.
The integrated PAR received during a 24-h period is called the DLI. The natural DLI
varies considerably depending on the location, time of year, cloud cover, etc., and this seasonal
variation increases with distance from the equator. DLI received inside a greenhouse depends
upon glazing material, overhead structures, presence of hanging baskets, plant spacing, and
strategies to alleviate heat stress (e.g., shade curtains and whitewash). In regions where annual
DLI varies significantly (e.g., >40 °N lat.) growers use supplemental lighting to increase DLI.

22

Metal halide and high-pressure sodium (HPS) lamps are commonly employed in greenhouses to
provide the supplemental lighting. These lamps differ in their spectral outputs, and therefore
species irradiated with different lamps may vary in growth and morphology, even if light
intensities and photoperiods are similar.

Effect of Daily Light Integral on Plant Development
Under light-limiting conditions, greenhouse growers commonly use supplemental
lighting to increase the total DLI at canopy height. Supplemental lighting can modify the
development rate of LD plants (i.e., decrease flowering time) when it truncates the night length.
In this case, an acceleration of flowering typically causes plants to flower at a lower leaf number.
Supplemental lighting can also accelerate flowering by increasing plant temperature from the
heat radiated by the lamps. Faust and Heins (1998) and Mattson and Erwin (2005) reported an
increase of 1 to 2 °C in plant temperature under supplemental HPS lighting.
Plants can be categorized based on their flowering response to supplemental irradiance:
those with a facultative irradiance (FI) response flower at a lower node number with an increase
in irradiance, whereas those that flower at a similar node number at different DLIs have an
irradiance indifferent (II) response (Erwin and Warner, 2002; Mattson and Erwin, 2005). For
−2 −1

example, increasing DLI from 8.3 to 25.5 mol·m ·d

decreased leaf number at flowering from

26 to 18 (FI response) in rosa del río (Hibiscus cisplatinus St.-Hil.), whereas node number of
monarch rosemallow (H. radiatus Cav.) was unaffected (II response) by DLI under SD (Warner
and Erwin, 2003). DLI and photoperiod interact to influence FI and II responses in some species.
Leaf number at flowering of swamp rosemallow (H. moscheutos L.) decreased as DLI increased
under SD, but was not influenced by DLI when grown under LD (Warner and Erwin, 2003).

23

Several studies have reported a negative correlation between DTF and DLI during the
finish stage for many annual bedding plants (Blanchard et al., 2011a; 2011b; Carvalho et al.,
2006; Currey and Erwin, 2011; Warner and Erwin, 2005; Faust et al., 2005; Oh et al., 2009;
−2 −1

Pietsch et al., 1995). For example, an increase in DLI from 10.5 to 21.8 mol·m ·d

(ADT of

20 °C) reduced DTF from transplant by 21 d in snapdragon, 16 d in calendula, 3 d in impatiens,
4 d in mimulus, and 12 d in wishbone flower (Warner and Erwin, 2005). Providing supplemental
lighting in greenhouses is most effective when the natural DLI is low, since flowering response
to increasing DLI follows a diminishing returns relationship (Blanchard et al., 2011a; Karlsson et
−2 −1

al., 1989b; Pramuk and Runkle, 2005b). For example, increasing DLI by 4.0 mol·m ·d
−2 −1

the natural DLI was 4.0 mol·m ·d

when

accelerated flowering in petunia by 10 d, but only by 2 d

−2 −1

when the natural DLI was 8.0 mol·m ·d

(Blanchard et al., 2011a). Similarly, Gagnon and

Dansereau (1990) observed that supplemental lighting treatments were more effective during
fall-winter experiments when the natural DLIs were lower compared to winter-spring
experiments.
An increase in the DLI during the young plant (liner or plug) stage can reduce flowering
time during the finish stage, mainly due to earlier flower initiation (Pramuk and Runkle, 2005a;
Oh et al., 2010). In addition, supplemental lighting of petunia and marigold during the later,
rather than the early, stage of seedling growth was more effective in reducing DTF, likely
because of the greater leaf area available for light interception (Oh et al., 2010). Increasing DLI
during the seedling stage can be a cost-effective strategy to reduce production time for finish
plants, since lighting costs per transplant are relatively low for young plants, since they are
commercially grown at a high density.

24

Plants have a critical DLI (DLIcrit) below which flowering will not occur and a saturating
DLI (DLIsat) above which the time to flowering response is not further hastened. White and
−2 −1

Warrington (1988) reported a DLIcrit of 3.3 mol·m ·d

for geranium ‘Red Elite’. For many

−2 −1

species, DLIcrit is <5 mol·m ·d , which is the minimum light level tested in most studies
(Fausey et al., 2005; Faust et al., 2005; Moccaldi and Runkle, 2007). In some plants, the critical
DLI required for flowering varies with each phase of the flowering process (Warner and Erwin,
−2 −1

2003). For example, swamp rosemallow under SD required 10 to 12 mol·m ·d
−2 −1

initiation, but at least 14 mol·m ·d

for flower

for successful flower development (Warner and Erwin,
−2 −1

2003). If these plants were grown at a DLI <12 mol·m ·d , they remained vegetative, and if
−2 −1

they were grown at a DLI >12 mol·m ·d

−2 −1

but <14 mol·m ·d , flower buds aborted soon

after they formed (Warner and Erwin, 2003).
DLIsat has also been estimated for many crops, and can vary from <5 to >20
−2 −1

mol·m ·d

depending on the species, cultivar, temperature, carbon dioxide concentration, etc.

(Blanchard, 2009; Hiden and Larsen, 1994; Oh et al., 2009; White and Warrington, 1988). For
example, DLIsat for shade-tolerant plants like vinca and sun-loving plants like spider flower
(Cleome hassleriana Chod.) and treasure flower (Gazania rigens L.) was 4.8 and >20.0
−2 −1

mol·m ·d , respectively (Blanchard, 2009). Similarly, DLIsat for petunia cultivars ‘Dreams
Neon Rose’, ‘Wave Purple’ and ‘Easy Wave Coral Reef’ was estimated at 10.6, 14.1 and 14.4
−2 −1

mol·m ·d , respectively (Blanchard, 2009). Exposing plants to light intensities in excess of
their saturation limit can damage photosystem II, which is the light reaction center for
photosynthesis. This decreases photosynthetic activity, referred to as photoinhibition, which can
considerably reduce plant growth and development under high light intensities. For example,
−2 −1

flower development rate decreased when the DLI was at least 12.0 mol·m ·d

25

for begonia and

−2 −1

vinca, which are both shade-tolerant species with estimated DLIsat of 10.5 and 4.8 mol·m ·d ,
respectively (Blanchard, 2009; Faust et al., 2005).
Several studies in photoperiodic (Corbesier et al., 1998; Roldan et al., 1999) and dayneutral plants (Dielen et al., 2004) have demonstrated that sucrose, a major product of lightdriven photosynthesis, plays an important role in flowering. For example, an increase in leaf
carbohydrate export or apical sucrose concentration during floral induction has been reported in
many plants (Corbesier et al., 1998; Dielen et al., 2001; Lejeune et al., 1993). Since these
changes were detected prior to any morphological signs of flower initiation, these authors
suggested a signaling role for sucrose. In addition, exogenous sucrose could overcome flowering
repression by FLC (Flowering locus C), but could not correct the late-flowering phenotypes of ft
or fwa arabidopsis mutants (Roldan et al., 1999). This suggests that sucrose promotes flowering
by regulating expression of FLC in steps that occur upstream of FT and FWA (Roldan et al.,
1999). Recently, Henry et al. (2011) reported a positive correlation between accumulation of
RhSUC2 (sucrose transporter) and light-induced bud burst in Rosa sp. It is possible that under
higher light levels, increased sucrose production and mobilization represses FLC, thereby turning
on a cascade of genes that ultimately lead to FT expression, and therefore earlier flowering.
However, more research is required to characterize light-induced flowering in plants and the
possible role of sucrose in this pathway.

Effect of Daily light integral on Plant Quality
In addition to ADT, DLI also influences plant quality parameters. Several studies have
quantified the influence of DLI on plant quality characteristics in many plants, and reported an
increase in quality with DLI (Fausey et al., 2005). The effect of DLI on plant quality parameters

26

such as number of flowers and lateral branches, flower size and color, height, and plant biomass
are discussed in detail.

Effect of DLI on Flower Number and Quality
There is a positive correlation between DLI and flower bud number in many species
(Blanchard et al., 2011a; 2011b; Currey and Erwin, 2011; Moccaldi and Runkle, 2007; Pramuk
−2 −1

and Runkle, 2005b). For example, increasing DLI from 10.0 to 20.0 mol·m ·d

(ADT of 20

°C) increased flower bud number by 63% in snapdragon, 56% in impatiens, 61% in mimulus,
and 15% in wishbone flower (Warner and Erwin, 2005). Similarly, in the herbaceous perennial
butterfly gaura (Gaura lindheimeri Engelm. and Gray), lateral inflorescence number per plant
tripled and flower number per inflorescence nearly doubled as DLI increased from 5.0 to 20.0
−2 −1

mol·m ·d

−2 −1

mol·m ·d

(Fausey et al., 2005). In addition, a small increase in DLI from 17.8 to 21.0
reduced the number of blind shoots by almost 50% in several cultivars of cut roses

(Rosa hybrida L.) (Bredmose, 1997). However, seedlings of bedding plants grown under a
higher DLI during the plug stage and then transferred to a common DLI had a lower flower
number at first flowering compared to plugs grown under a lower DLI (Pramuk and Runkle,
2005a; Oh et al., 2010). This could be attributed to the more rapid flowering that occurred with
the higher plug DLI, so that plants had a shorter period to intercept light before flowering. The
high plug DLI also reduced the leaf number below first open flower which may have contributed
to the reduction in axillary inflorescences.
Several studies have reported an increase in flower size with DLI (Karlsson et al., 1989;
Moccaldi and Runkle, 2007; Niu et al., 2000; Pietsch et al., 1995; Warner and Erwin, 2005).
Warner and Erwin (2005) described a linear relationship between DLI and flower size, such that

27

−2 −1

each unit (mol·m ·d ) increase in DLI increased the flower size by 0.09 mm in calendula, 0.59
mm in impatiens, 0.15 mm in mimulus, and 0.25 mm in wishbone flower. Flower bud
development after initiation involves active cell division and elongation, which requires sugar
import to the developing buds from the leaves. Sugar import depends upon sugar mobilization in
the developing buds, and a constant supply of sugars from the leaves. Girault et al. (2010)
reported an increase in RhVI gene transcription under higher light levels in rose, and
consequently an increase in RhVI vacuolar acid invertase activity, an enzyme that participates in
sucrose breakdown in developing rose buds. Similarly, Henry et al. (2011) demonstrated a
correlation between up-regulation of sucrose transporters (RhSUC2) and light-induced bud-burst
in roses. Therefore, under higher light levels, accumulation of sugars due to increased
photosynthesis (Nemali and van Iersel, 2004), stimulation of sugar metabolism in flower buds
(Girault et al., 2010), and up-regulation of sucrose transporters (Henry et al., 2011) may all work
together to promote flower development and expansion.
Anthocyanins are plant pigments that impart a red, blue, or purple pigmentation to leaves
and flowers in many species. Light intensity influences flower pigmentation in plants mainly by
regulating the anthocyanin concentration in flower petals. For example, anthocyanin in flowers
was positively correlated with light intensity in carnation (Maekawa, 1974), lisanthus [(Eustoma
grandiflorum (Raf.) Shinners; Meir et al., 2009)], Oriental hybrid lily (Lilium aurantum Lindl.;
Kawabata et al., 2002), rose (Maekawa et al., 1980), and stock [Matthiola incana (L.) W.T.
Aiton; Kawabata et al., 2002)]. Meir et al. (2009) reported a reduction in anthocyanin
biosynthesis gene expression under light-limiting conditions. These authors suggested that light
intensity influences certain transcription factors that act as master regulators for a number of
anthocyanin biosynthesis genes. In addition, sucrose availability up-regulates the anthocyanin

28

biosynthetic pathway (Solfanelli et al., 2006). Higher light levels increase the synthesis,
mobilization and transport of sugars to the developing flower buds, which may promote
anthocyanin biosynthesis in the flower petals. Therefore, plants grown under higher light have a
brighter and more intense flower color than those grown under low light conditions.

Effect of DLI on Branch Number and Biomass
A few studies have reported an increase in lateral branch number with an increase in DLI
(Carvalho et al., 2006; Faust et al., 2005; Welander, 1983). For example, an increase in DLI from
−2 −1

4.0 to 43.0 mol·m ·d

increased the number of lateral shoots by 7 in ageratum (Ageratum

houstonianum L.) and petunia (Faust et al., 2005). Increased availability of photosynthates at
higher light intensities may promote initiation and development of lateral shoots.
Dry matter accumulation in plants primarily depends upon light, carbon dioxide, and
water for photosynthesis. Many studies have reported an increase in plant biomass in
greenhouse-grown crops with an increase in DLI (Faust et al., 2005; Fausey et al., 2005;
Moccaldi and Runkle, 2007; Warner and Erwin, 2005). For example, as DLI increased from 4.0
−2 −1

to 14.0 mol·m ·d

during the finish stage, shoot dry weight at anthesis increased by 162%,

187%, and 108% in common yarrow (Achillea millefolium L.), butterfly gaura, and lavender
(Lavandula augustifolia Mill.), respectively (Fausey et al., 2005). A similar increase in DLI
during seedling growth linearly increased seedling dry weight by 64% in celosia, 47% in
impatiens, 64% in marigold, and 68% in pansy (Pramuk and Runkle, 2005a).
In addition, some studies have investigated the influence of DLI on root and flower dry
mass (Faust et al., 2005; Torres and Lopez, 2011). When ambient light levels are low, increasing
DLI during propagation increases dry matter accumulation of root and shoot tissues (Lopez and

29

Runkle, 2008; Torres and Lopez, 2011). A denser root mass increases the surface area for
nutrient absorption, and possibly contributes to faster plant growth. Similarly, Faust et al. (2005)
reported an increase in flower dry mass with DLI. Interestingly, under higher light levels, these
authors observed a greater percentage of dry matter partitioned to flowers in plants in the
Asteraceae family compared to others.
When plants are grown under the same DLI, the duration of light interception mainly
influences dry matter accumulation. For example, celosia seedlings grown under 4.0 and 14.0
−2 −1

mol·m ·d

during the plug stage, and then transferred to a common DLI during the finish

stage, flowered in 43 and 33 d, and accumulated 4.5 and 2.9 g of biomass at first flower,
respectively (Pramuk and Runkle, 2005a). The DLI required for maximum dry matter
accumulation is higher than their DLIsat for flowering. For example, the flowering response in
−2 −1

vinca saturated at 4.8 mol·m ·d

−2 −1

(Blanchard, 2009), but it required 21 mol·m ·d

to attain

50% of the maximum total dry mass (Faust et al., 2005). In addition to DLI, light quality also
affects dry matter accumulation. For example, photosynthetic rate (and therefore dry mass
accumulation) is generally highest under blue and especially red light. Therefore, plants grown
under the same DLI but irradiated with lamps having different spectral outputs would probably
vary in dry mass.

Effect of DLI on Plant Height
Under shaded conditions, shade-avoiding plants elongate to increase light interception.
Accordingly, some studies have reported a negative correlation between plant height and DLI
(Armitage, 1995; Blanchard et al., 2011a; Fausey et al., 2005; Miller and Armitage, 2002).
However, the shade-avoidance response in plants is mainly a function of the R: FR (Erwin et al.,

30

2006). In most experiments, higher light levels were created by supplementing ambient daylight
with light from HPS lamps (Blanchard et al., 2011a; Fausey et al., 2005) or metal halide lamps
(Miller and Armitage, 2002) lamps. Since the R: FR ratio of these lamps is greater than that of
sunlight, a reduction in height under higher DLI conditions may be an effect of light quality and
not quantity. However, higher light levels may also contribute to height reduction by reducing
leaf number below first open flower. For example, for most seedlings grown under a higher DLI
(R: FR ratio ≈ 3.6 for all treatments), and then transferred to a common DLI during the finish
stage, a reduction in plant height was accompanied by a lower leaf number below first open
flower (Pramuk and Runkle, 2005a).
Many studies have observed no consistent relationship between plant height and DLI
(Currey and Erwin, 2011; Faust et al., 2005; Moccaldi and Runkle, 2007). In these studies, plant
height responses to DLI varied by species. For example, among the six species of Kalanchoe
studied, plant height decreased, increased, or was unaffected by DLI (similar R: FR ratio across
all treatments) in one, two, or three species, respectively (Currey and Erwin, 2011). This
variability is not unexpected, since in addition to light quality and quantity, many factors in a
greenhouse such as DIF, water and nutrient uptake, and plant spacing can also affect plant
height. Plants grown under a higher DLI develop faster and, if plant density does not change as
plants grow, they may shade each other. Similarly, due to a greater thermal load and faster plant
development under higher light intensities, water and nutrient uptake is generally higher.
Therefore, plant height responses to DLI can sometimes be at least partially attributed to other
variables that influence stem elongation.

31

Interaction between Temperature and DLI
Temperature and DLI are the two primary environmental factors that influence plant
growth, development and quality. In regions where the natural temperature and DLI vary
seasonally, growers commonly use environmental control systems to manipulate ADT and DLI
to ensure precise crop timing and better plant quality. The rate of plant development is mainly a
function of ADT, while DLI primarily controls dry matter accumulation via photosynthesis.
However, these factors interact to influence growth attributes and plant development rates.
Several studies have investigated the interaction of ADT and DLI, and sometimes DIF,
on plant timing and quality parameters (Blanchard et al., 2011a; Moccaldi and Runkle, 2007; Niu
et al., 2000; Pietsch et al., 1995; Pramuk and Runkle, 2005b). Liu and Heins (2002) applied the
−2 −1

concept of photothermal ratio [(PTR); ratio of radiant energy (mol·m ·d ) to thermal energy
(degree days)] to integrate the effects of ADT and DLI on plant timing and quality in poinsettia.
A high PTR is created with a high DLI and low temperature, while a low PTR is created by a
low DLI and high temperature.

Temperature and DLI Interaction on Plant Development
In many species, flowering time decreases as ADT and DLI increase within speciesspecific ranges (Blanchard et al., 2011a, 2011b; Moccaldi and Runkle, 2007; Pramuk and
Runkle, 2005b). Studies have developed mathematical models to predict the interactive effects of
ADT and DLI on flower development rate in celosia, chrysanthemum, cineraria, impatiens,
pansy, petunia, marigold, and salvia (Blanchard et al., 2011a, 2011b; Larsen, 1989; Larsen and
Persson, 1999; Moccaldi and Runkle, 2007; Pramuk and Runkle, 2005b). Increasing DLI has a
greater effect on plant development rate at lower temperatures (Blanchard et al., 2011a;

32

Kaczperski et al., 1991; Pietsch et al., 1995). For example, an increase in DLI from 9.0 to 30.0
−2 −1

mol·m ·d

accelerated flowering time in vinca ‘Grape Cooler’ by 11 d at 20.0 °C, but only by

4 d at 35.0 °C (Pietsch et al., 1995). Similarly, in petunia ‘Wave Purple’, a DLI increase from 4.0
−2 −1

to 14.0 mol·m ·d

hastened flowering by 22 d at 14.0 °C but by only 9 d at 26.0 °C

(Blanchard et al., 2011a). The supplemental lighting treatments may have increased plant
temperature more when plants were grown at lower air temperatures. Also, it is possible that an
increase in ADT and DLI turn on pathways that overlap to induce accumulation of the FT protein
(Blazquez et al., 2003; Roldan et al., 1999). On the other hand, suppression of FT at lower
temperatures may be offset by induction of FT expression by higher irradiance.
Some studies have reported a shift in Tb and Topt with increasing DLI (Adams et al.,
1997; Faust and Heins, 1993; Pramuk and Runkle, 2005b). Increase in DLI from 5.0 to 15.0
−2 −1

mol·m ·d

decreased the estimated Tb of celosia and impatiens by 1.5 and 3.2 °C, respectively

(Pramuk and Runkle, 2005b). Since Tb for celosia and impatiens was estimated using ADT and
not plant temperature, these authors suggested that a decrease in Tb at higher DLI was probably
because of an increase in plant temperature from the higher irradiances. In contrast, Adams et al.
−2 −1

(1997) reported a linear increase in Topt as DLI decreased from 6.0 mol·m ·d . Similarly, in
African violet, Topt for LUR decreased from 25.0 to 23.0 °C as DLI decreased from 10.0 to 1.0
−2 −1

mol·m ·d

(Faust and Heins, 1993). This may be due to a limited supply of photoassimilates at

lower light levels, since a similar shift in Topt for photosynthesis has been demonstrated in
carnation (Enoch and Hurd, 1977). In addition, Karlsson (2002) reported a decrease in DLIsat for
flower initiation with an increase in ADT in native primrose [Primula vulgaris syn. P. acaulis
(L.) Hill]. This may not be entirely surprising, since the promotive effect of increasing DLI is
generally lower at higher temperatures.

33

Temperature and DLI Interaction on Plant Quality
Temperature and DLI interact to influence plant quality characteristics such as number of
flowers, plant height, flower size and biomass. Since plant quality generally improves as ADT
decreases and DLI increases, a higher plant quality is generally obtained when ADT is relatively
low and DLI is high. For example, marigold plants grown at 15.0 °C and a DLI of 25.0
−2 −1

mol·m ·d

accumulated 144% more dry mass, had 112% more inflorescences, and 49%
−2 −1

greater inflorescence diameters than plants grown at 25.0 °C and a DLI of 5.0 mol·m ·d

(Pramuk and Runkle, 2005b). Similarly, as PTR increased, flower number, dry weight and,
inflorescence diameter increased for several floriculture crops (Blanchard et al., 2011a; Liu and
Heins, 2002; Niu et al., 2000). For example, increase in PTR from 0.2 to 1.9 mol·m

−2

per degree

day increased flower number by 197% in petunia ‘Easy Wave Coral Reef’ (Blanchard et al.,
2011a). Similarly, Liu and Heins (2002) reported a linear increase in plant dry weight and size of
bracts and cyathia with PTR in poinsettia ‘Freedom’.
The influence of PTR on plant height is variable. For example, Blanchard et al. (2011a)
reported a negative correlation between PTR and plant height in petunia ‘Wave Purple’ and
‘Easy Wave Coral Reef’. In contrast, Liu and Heins (2002) reported that plant height in
poinsettia ‘Freedom’ was unaffected by PTR. Plant height often decreases as DLI, and
sometimes ADT, increases, since both factors influence leaf number below first open flower, and
thus plant height. Since a higher PTR is created by high DLI and lower temperatures, the effects
of DLI and ADT on leaf number below first open flower (and therefore plant height) may
counteract each other at a given PTR. This may result in variability in plant height between
different PTR treatments, and therefore no consistent relationship between PTR and plant height.

34

Some studies suggest an interaction between DIF and DLI to influence stem elongation
(Erwin and Heins, 1995; Niu et al., 2000). Niu et al. (2000) observed a greater stem elongation
response to DIF under lower than higher DLI. In contrast, Erwin and Heins (1995) reported an
increase in stem elongation response with irradiance. These differences in stem elongation
response to DIF could be attributed to differences in light quality, and not DLI. For example, Niu
et al. (2000) used HPS lamps to provide supplemental lighting in the medium and high DLI
treatments. Since light from HPS lamps has a greater R: FR ratio than sunlight, plants grown
under the higher DLI treatments may have been shorter because of the light quality in treatments
and less by DIF.
In general, flowering time is hastened as ADT and DLI increases, while plant quality of
shade-avoiding plants decreases as ADT increases and DLI decreases. Therefore, if plants are
grown at a warmer ADT to accelerate flowering, a reduction in plant quality can be offset by
increasing the DLI (Blanchard et al., 2011a). For example, in marigold plants grown at 25.0 °C,
−2 −1

increasing the DLI from 5.0 to 25.0 mol·m ·d

increased inflorescence number, inflorescence

diameter, and plant dry weight by 47%, 16%, and 93%, respectively.

Conclusions
Temperature and light are the two environmental factors that strongly influence crop
timing and plant quality. Greenhouse growers, especially in temperate regions, manipulate the
ADT and DLI to precisely schedule crops and regulate plant quality. Although the physiological
mechanisms underlying plant responses to ADT and DLI are well understood, a greater
understanding of the molecular mechanisms is desirable. In addition, several other environmental
and cultural factors in a greenhouse influence plant timing and morphology such as DIF, light

35

quality, plant nutrition, carbon dioxide concentration, plant spacing, and watering. Therefore,
when investigating plant responses to ADT and DLI, it is important to consider the possible
influence of these factors on the parameters studied, and when possible, control and monitor
them.

36

Table 1.1. Temperature response categories for popular bedding plant species based on their
estimated Tb and the corresponding percentage delay in flowering time when grown at 15 versus
20 °C.

Species and Cultivar

Tb
(°C)

Delay in
flowering
time

Reference

Cold tolerant (Tb <4 °C)
Campanula carpatica
0.0
29%
Whitman et al., 1997
‘Blue Chips’
Gaillardia ×grandiflora
3.3
31%
Yuan et al., 1998
‘Goblin’
Leucanthemum ×superbum
−3.4
25%
Yuan et al., 1999
‘Snowcap’
Petunia ×hybrida
2.8
28%
Blanchard and Runkle, 2011a
'Dreams Neon Rose'
Platycodon grandiflorus
2.9
38%
Park et al., 1998
‘Astra Blue’
Cold intermediate (4 °C< Tb <7 °C)
Coreopsis grandiflora
‘Sunray’
Oenothera fruticosa
‘Youngii Lapsley’
Petunia ×hybrida
'Wave Purple'
Rudbeckia hirta
‘Toto Rustic’
Tagetes erecta
‘Antigua Primrose’

6.8

44%

Yuan et al., 1998

4.4

50%

Clough et al., 2001

5.5

45%

Blanchard and Runkle, 2011a

4.6

43%

Blanchard and Runkle, 2011a

4.4

33%

Blanchard and Runkle, 2011a

Cold sensitive (Tb >7 °C)
Angelonia augustifolia
‘Serena Purple’
Browallia speciosa
‘Bells Marine’
Pentas lanceolata
‘Graffiti Lavender’
Salvia farnacea
‘Victoria Blue’
Zinnia elegans
‘Dreamland Coral’

9.9

81%

8.9

57%

9.3

53%

9.4

62%

7.8

50%

37

Blanchard and Runkle, 2011a

LITERATURE CITED

38

Literature Cited
Adams, S.R., P. Hadley, and S. Pearson. 1998a. The effects of temperature, photoperiod, and
photosynthetic photon flux on the time to flowering of petunia 'Express Blush Pink'. J.
Amer. Soc. Hort. Sci. 123:577−580.
Adams, S.R., S. Pearson, and P. Hadley. 1997. The effects of temperature, photoperiod and light
integral on the time to flowering of pansy cv Universal Violet (Viola ×wittrockiana
Gams). Ann. Bot. 80:107−112.
Adams, S.R., S. Pearson, and P. Hadley. 1998b. An appraisal of the use of reciprocal transfer
experiments: Assessing the stages of photoperiod sensitivity in chrysanthemum cv.
Snowdon (Chrysanthemum ×morifolium Ramat.). J. Exp. Bot. 49:1405−1411.
Adams, S.R., S. Pearson, and P. Hadley. 1998c. The effect of temperature on inflorescence
initiation and subsequent development in chrysanthemum cv. Snowdon (Chrysanthemum
×morifolium Ramat.). Scientia Hort. 77:59−72.
Adams, S.R., S. Pearson, P. Hadley, and W.M. Patefield. 1999. The effects of temperature and
light integral on the phases of photoperiod sensitivity in Petunia ×hybrida. Ann. Bot.
83:263−269.
Armitage, A.M. 1995. Photoperiod, irradiance, and temperature influence flowering of Hamelia
patens (Texas firebush). HortScience 30:255−256.
Armitage, A.M., W.H. Carlson, and J.A. Flore. 1981. The effect of temperature and quantum
flux density on the morphology, physiology, and flowering of hybrid geraniums. J. Amer.
Soc. Hort. Sci. 106:643−647.
Armitage, A.M., A.M. Miller, and J.M. Garner. 2000. The influence of photoperiod and
temperature on growth and flowering of Angelonia angustifolia Benth. - A preliminary
report. Acta Hort. 541:167−170.
Bartok, J.W., Jr. 2001. Energy conservation for commercial greenhouses. Natural Resource, Agr.
and Eng. Serv. Coop. Ext., Ithaca, NY.
Berghage, R.D., J.A. Flore, R.D. Heins, and J.E. Erwin. 1990a. The relationship between day and
night temperature influences photosynthesis but not light compensation point or flower
longevity of Easter lily, Lilium longiflorum Thunb. Acta Hort. 272:91−95.
Berghage, R.D., R.D. Heins, and J.E. Erwin. 1990b. Quantifying leaf unfolding in the poinsettia.
Acta Hort. 272:243−247.
Blanchard, M. 2009. Manipulating light and temperature for energy-efficient greenhouse
production of ornamental crops. Ph.D. Dissertation, Michigan State Univ., East Lansing.

39

Blanchard, M.G. and E. S. Runkle. 2011a. Quantifying the thermal flowering rates of eighteen
species of annual bedding plants. Scientia Hort. 128:30−37.
Blanchard, M.G., E.S. Runkle, and P.R. Fisher. 2011a. Modelling plant morphology and
development of petunia in response to temperature and photosynthetic daily light integral.
Scientia Hort. 129:313−320.
Blanchard, M.G., E.S. Runkle, and J.M. Frantz. 2011b. Energy-efficient greenhouse production
of Petunia and Tagetes by manipulation of temperature and photosynthetic daily light
integral. Acta Hort. 893:857−864.
Blanchard, M.G.and E.S. Runkle. 2011b. The influence of day and night temperature
fluctuations on growth and flowering of annual bedding plants and greenhouse heating
cost predictions. HortScience 46:599−603.
Blazquez, M.A., J.H. Ahn, and D. Weigel. 2003. A thermosensory pathway controlling flowering
time in Arabidopsis thaliana. Nature Genetics 33:168−171.
Bredmose, N. 1997. Chronology of three physiological development phases of single-stemmed
rose (Rosa hybrida L.) plants in response to increment in light quantum integral. Scientia
Hort. 69:107−115.
Brøndum, J.J.and R.D. Heins. 1993. Modeling temperature and photoperiod effects on growth
and development of dahlia. J. Amer. Soc. Hort. Sci. 118:36−42.
Brooks, A.and G.D. Farquhar. 1985. Effect of temperature on the CO2/O 2 specificity of ribulose1,5-bisphosphate carboxylase/oxygenase and the rate of respiration in the light. Estimates
from gas-exchange measurements on spinach. Planta 165:397−406.
Carvalho, S.M.P., H. Abi-Tarabay, and E. Heuvelink. 2005. Temperature affects chrysanthemum
flower characteristics differently during three phases of the cultivation period. J. Hort.
Sci. Biotech. 80:209−216.
Carvalho, S.M.P., S.E. Wuillai, and E. Heuvelink. 2006. Combined effects of light and
temperature on product quality of Kalanchoe blossfeldiana. Acta Hort. 711:121−126.
Clough, E.A., A. Cameron, R.D. Heins, and W.H. Carlson. 2001. Growth and development of
Oenothera fruticosa is influenced by vernalization duration, photoperiod, forcing
temperature, and plant growth regulators. J. Amer. Soc. Hort. Sci. 126:269−274.
Cockshull, K.E., D.W. Hand, and F.A. Langton. 1981. The effects of day and night temperature
on flower initiation and development in chrysanthemum. Acta Hort. 125: 101−110.
Cockshull, K.E.and A.P. Hughes. 1971. The effects of light intensity at different stages in flower
initiation and development of Chrysanthemum ×morifolium. Ann. Bot. 35:915−926.

40

Corbesier, L., P. Lejeune, and G. Bernier. 1998. The role of carbohydrates in the induction of
flowering in Arabidopsis thaliana: comparison between the wild type and a starchless
mutant. Planta 206:131−137.
Currey, C.J.and J.E. Erwin. 2011. Photosynthetic daily light integral impacts growth and
flowering of several Kalanchoe species. HortTechnology 21:98−102.
Davies, L.J., J.L. Catley, I.R. Brooking, and E.A. Halligan. 2002. Temperature and light
requirements for Sandersonia aurantiaca flowering. Acta Hort. 570:105−112.
Dela, G., E. Or, R. Ovadia, A. Nissim-Levi, D. Weiss, and M. Oren-Shamir. 2003. Changes in
anthocyanin concentration and composition in 'Jaguar' rose flowers due to transient hightemperature conditions. Plant Science 164:333−340.
Dielen, V., V. Lecouvet, S. Dupont, and J.M. Kinet. 2001. In vitro control of floral transition in
tomato (Lycopersicon esculentum Mill.), the model for autonomously flowering plants,
using the late flowering uniflora mutant. J. Exp. Bot. 52:715−723.
Dielen, V., M. Quinet, J. Chao, H. Batoko, A. Havelange, and J.M. Kinet. 2004. UNIFLORA, a
pivotal gene that regulates floral transition and meristem identity in tomato (Lycopersicon
esculentum). New Phytol. 161:393−400.
Dinar, M.and J. Rudich. 1985. Effect of heat stress on assimilate partitioning in tomato. Ann.
Bot. 56:239−248.
Erwin, J., P. Velguth, and R. Heins. 1994. Day/night temperature environment affects cell
elongation but not division in Lilium longiflorum Thunb. J. Exp. Bot. 45:1019−1025.
Erwin, J.E. 1991. Thermomorphogenesis in plants. Ph.D. Dissertation, Michigan State Univ.,
East Lansing.
Erwin, J.E.and R.D. Heins. 1990. Temperature effects on lily development rate and morphology
from the visible bud stage until anthesis. J. Amer. Soc. Hort. Sci. 115:644−646.
Erwin, J.E.and R.D. Heins. 1995. Thermomorphogenic responses in stem and leaf development.
HortScience 30:940−949.
Erwin, J.E., R.D. Heins, and M.G. Karlsson. 1989. Thermomorphogenesis in Lilium longiflorum.
Amer. J. Bot. 76:47−52.
Erwin, J.E., R.D. Heins, and R. Moe. 1991. Temperature and photoperiod effects on Fuchsia
×hybrida morphology. J. Amer. Soc. Hort. Sci. 116:955−960.

41

Erwin, J.E.and R.M. Warner. 2002. Determination of photoperiodic response group and effect of
supplemental irradiance on flowering of several bedding plant species. Acta Hort.
580:95−99.
Fausey, B.A., R.D. Heins, and A.C. Cameron. 2005. Daily light integral affects flowering and
quality of greenhouse-grown Achillea, Gaura, and Lavandula. Hortscience 40:114−118.
Faust, J.E.and R.D. Heins. 1993. Modeling leaf development of the African violet (Saintpaulia
ionantha Wendl.). J. Amer. Soc. Hort. Sci. 118:747−751.
Faust, J.E.and R.D. Heins. 1994. Modeling inflorescence development of the African violet
(Saintpaulia ionantha Wendl.). J. Amer. Soc. Hort. Sci. 119:727−734.
Faust, J.E.and R.D. Heins. 1998. Modelling shoot-tip temperature in the greenhouse
environment. J. Amer. Soc. Hort. Sci. 123:208−214.
Faust, J.E., V. Holcombe, N.C. Rajapakse, and D.R. Layne. 2005. The effect of daily light
integral on bedding plant growth and flowering. HortScience 40:645−649.
Fisher, P., H. Warren, and L. Hydock. 2006. Larger liners, shorter crop time. Greenhouse
Grower 24: 8−12.
Gagnon, S.and B. Dansereau. 1990. Influence of light and photoperiod on growth and
development of gerbera. Acta Hort. 272:145−151.
Girault, T., F. Abidi, M. Sigogne, S. Pelleschi-Travier, R. Boumaza, S. Sakr, and N. Leduc.
2010. Sugars are under light control during bud burst in Rosa sp. Plant, Cell and Environ.
33:1339−1350.
Grindal, G., A. Ernstsen, J.B. Reid, O. Junttila, B. Lindgard, and R. Moe. 1998. Endogenous
gibberellin A 1 levels control thermoperiodic stem elongation in Pisum sativum. Physiol.
Plant 102:523−531.
Grindal, G.and R. Moe. 1994. Effects of temperature-drop and a short dark interruption on stem
elongation and flowering in Begonia × hiemalis Fotsch. Scientia Hort. 57: 123−132.
Gutteridge, J.M.C.and B. Halliwell. 1990. The measurement and mechanism of lipid
peroxidation in biological systems. Trends Biochem. Sci. 15:129−135.
Hall, C. 2011. State of the industry report: 2011 economic outlook. Greenhouse Grower 29:
36−37.
Hanan, J.J. 1998. Greenhouses: Advanced technology for protected horticulture. CRC Press,
Boca Raton, Fla.

42

Harris, G.P.and M.A. Scott. 1968. Studies on the glasshouse carnation: Effects of light and
temperature on the growth and development of the flower. Ann. Bot. 33:143−152.
Heins, R.D., B. Liu, and E.S. Runkle. 2000. Regulation of crop growth and development based
on environmental factors. Acta Hort. 514:13−22.
Henry, C., A. Rabot, M. Laloi, E. Mortreau, M. Sigogne, N. Leduc, R. Lemoine, S. Sakr, A.
Vian, and S. Pelleschi-Travier. 2011. Regulation of RhSUC2, a sucrose transporter, is
correlated with the light control of bud burst in Rosa sp. Plant, Cell and Environ.
34:1776−1789.
Hiden, C.and R.U. Larsen. 1994. Predicting flower development in greenhouse grown
chrysanthemum. Scientia Hort. 58:123−138.
Hwang, S., H. Moon, and B. Jeong. 2005. Effects of difference between day and night
temperatures on growth and flowering of miniature roses. J. Kor. Soc. Hort. Sci.
46:321−328.
Jensen, E., S. Eilertsen, A. Ernsten, O. Junttila, and R. Moe. 1996. Thermoperiodic control of
stem elongation and endogenous gibberellins in Campanula isophylla. J. Plant Growth
Reg. 15:167−171.
Jones, H.G. 1983. Plants and microclimate. Cambridge Univ. Press, New York, NY.
Kaczperski, M.P., W.H. Carlson, and M.G. Karlsson. 1991. Growth and development of Petunia
×hybrida as a function of temperature and irradiance. J. Amer. Soc. Hort. Sci.
116:232−237.
Kanellos, E.A.G.and S. Pearson. 2000. Environmental regulation of flowering and growth of
Cosmos atrosanguineus (Hook.) Voss. Scientia Hort. 83:265−274.
Karlsson, M. 1992. Leaf unfolding rate in Begonia ×hiemalis. HortScience 27:109−110.
Karlsson, M., R.D. Heins, and J.E. Erwin. 1988. Quantifying temperature-controlled leaf
unfolding rates in 'Nellie White' Easter lily. J. Amer. Soc. Hort. Sci. 113:70−74.
Karlsson, M., R.D. Heins, J.O. Gerberick, and M.E. Hackmann. 1991. Temperature driven leaf
unfolding rate in Hibiscus rosa-sinenesis. Scientia Hort. 45:323−331.
Karlsson, M.and J. Werner. 2001. Temperature affects leaf unfolding rate and flowering of
cyclamen. HortScience 36:292−294.
Karlsson, M.G. 2002. Flower formation in Primula vulgaris is affected by temperature,
photoperiod and daily light integral. Scientia Hort. 95:99−110.

43

Karlsson, M.G.and R.D. Heins. 1992. Chrysanthemum dry matter partitioning patterns along
irradiance and temperature gradients. Can. J. Plant Sci. 72: 307−316.
Karlsson, M.G., R.D. Heins, J.E. Erwin, R.D. Berghage, W.H. Carlson, and J.A. Biernbaum.
1989. Temperature and photosynthetic photon flux influence chrysanthemum shoot
development and flower initiation under short-day conditions. J. Amer. Soc. Hort. Sci.
114:158−163.
Kawabata, S., Y. Li, M. Adachi, H. Maruyama, E. Ozawa, and R. Sakiyama. 2002. Role of
photoreceptor- and sugar-mediated reactions in light dependent anthocyanin production
in lily and stock flowers. J. Jpn. Soc. Hort. Sci. 71:220−225.
Korczynski, P.C., J. Logan, and J.E. Faust. 2002. Mapping monthly distribution of daily light
integrals across the contiguous United States. HortTechnology 12:12−16.
Ku, S.B.and G.E. Edwards. 1977. Oxygen inhibition of photosynthesis. 1. Temperature
dependence and relation to O2/CO2 solubility ratio. Plant Physiol. 59:986−990.
Landsberg, J.J. 1977. Some useful equations for biological studies. Experimental Agriculture.
13:273−286.
Larsen, R.U. 1988. A prediction model of the development rate of the pot plant Senecio
×hybridus Hyl. A case study. Acta Hort. 230:381−388.
Larsen, R.U. 1989. A prediction model for floral development of Senecio ×hybridus Hyl. 'Molls
Stam'. Acta Hort. 248:47−254.
Larsen, R.U.and C. Hiden. 1995. Predicting leaf unfolding in flower induced shoots of
greenhouse grown chrysanthemum. Scientia Hort. 63:225−239.
Larsen, R.U.and L. Persson. 1999. Modelling flower development in greenhouse chrysanthemum
cultivars in relation to temperature and response group. Scientia Hort. 80:73−89.
Lee, H., S. Yoo, J. Lee, W. Kim, S. Yoo, H. Fitzgerald, J.C. Carrington, and J. Ahn. 2010.
Genetic framework for flowering-time regulation by ambient temperature-responsive
miRNAs in Arabidopsis. Nucleic Acids Research 38:3081−3093.
Lee, J., S. Yoo, S. Park, I. Hwang, J. Lee, and J. Ahn. 2007. Role of SVP in the control of
flowering time by ambient temperature in Arabidopsis. Genes & Devel. 21:397−402.
Lejeune, P., G. Bernier, M.C. Requier, and J.M. Kinet. 1993. Sucrose increase during floral
induction in the phloem sap collected at the apical part of the shoot of the long-day plant
Sinapis alba L. Planta 190:71−74.

44

Lepage, I., J.d. Jong, and L. Smeets. 1984. Effect of day and night temperatures during short
photoperiods on growth and flowering of Chrysanthemum ×morifolium Ramat. Scientia
Hort. 22:373−381.
Lie, X. and B. Huang. 2000. Heat stress injury in relation to membrane lipid peroxidation in
creeping bentgrass. Crop Sci. 40:503−510.
Liu, B.and R.D. Heins. 1997. Is plant quality related to the ratio of radiant energy to thermal
energy? Acta Hort. 435: 171−182.
Liu, B.and R.D. Heins. 2002. Photothermal ratio affects plant quality in 'Freedom' poinsettia. J.
Amer. Soc. Hort. Sci. 127:20−26.
Maekawa, S. 1974. Studies on the coloration of carnation flowers. II. The effect of light on the
growth and anthocyanin formation in detached petals. J. Jpn. Soc. Hort. Sci.
42:347−352.
Maekawa, S.and N. Nakamura. 1977. Studies on the coloration of carnation flowers. VII. The
effects of temperature on the coloration and pigmentation for the intact flower and plant
growth. J. Jpn. Soc. Hort. Sci. 45:375−382.
Mattson, N.S.and J.E. Erwin. 2003. Temperature Affects Flower Initiation and Development
Rate of Impatiens, Petunia, and Viola. Acta Hort. 624:191−197.
Mattson, N.S.and J.E. Erwin. 2005. The impact of photoperiod and irradiance on flowering of
several herbaceous ornamentals. Scientia Hort. 104:275−292.
Meir, S., B. Kochanek, A. Glick, S. Salim, A. Lers, S. Burd, S. Philosoph-Hadas, and D. Weiss.
2010. Reduced petal pigmentation in lisianthus (Eustoma grandiflorum) flowers under
low light conditions is associated with decreased expression of anthocyanin biosynthesis
genes. Acta Hort. 877:1735−1744.
Miller, A.and A.M. Armitage. 2002. Temperature, irradiance, photoperiod, and growth retardants
influence greenhouse production of Angelonia angustifolia Benth. Angel Mist series.
HortScience 37:319−321.
Miller, W.B., P.A. Hammer, and T.I. Kirk. 1993. Reversed greenhouse temperatures alter
carbohydrate status in Lilium longiflorum Thunb. 'Nellie White'. J. Amer. Soc. Hort. Sci.
118:736−740.
Moccaldi, L.A.and E.S. Runkle. 2007. Modeling the effects of temperature and photosynthetic
daily light integral on growth and flowering of Salvia splendens and Tagetes patula. J.
Amer. Soc. Hort. Sci. 132:283−288.

45

Moe, R. 1990. Effect of day and night temperature alternations and of plant growth regulators on
stem elongation and flowering of the long-day plant Campanula isophylla Moretti.
Scientia Hort. 43:291−305.
Moe, R., N. Glomsrud, I. Bratberg, and S. Vals. 1992. Control of plant height in poinsettia by
temperature drop and graphical tracking, p. 41−48. Acta Hort. 327:41−48.
Moe, R.and R. Heins. 1990. Control of plant morphogenesis and flowering by light quality and
temperature. Acta Hort. 272:81−89.
Moe, R.and L.M. Mortensen. 1992. Thermomorphogenesis in pot plants. Acta Hort. 305:19−25.
Mortensen, L.M.and R. Moe. 1992. Effects of various day and night temperature treatments on
the morphogenesis and growth of some greenhouse and bedding plant species. Acta Hort.
327:77−86.
Myster, J., A. Ernstsen, O. Junttila, and R. Moe. 1997. Thermo- and photoperiodicity and
involvement of gibberellins during day and night cycle on elongation growth of Begonia
×hiemalis Fotsch. J. Plant Growth Reg. 16:189−196.
Myster, J.and R. Moe. 1995. Effect of diurnal temperature alternations on plant morphology in
some greenhouse crops - a mini review. Scientia Hort. 62:205−215.
Neily, W.G., P.R. Hicklenton, and D.N. Kristie. 1997. Temperature and developmental stage
influence diurnal rhythms of stem elongation in snapdragon and zinnia. J. Amer. Soc.
Hort. Sci. 122:778−783.
Nemali, K.S.and M.W.v. Iersel. 2004. Light effects on wax begonia: photosynthesis, growth
respiration, maintenance respiration, and carbon use efficiency. J. Amer. Soc. Hort. Sci.
129:416−424.
Niu, G., R.D. Heins, A. Cameron, and W.H. Carlson. 2001. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of Campanula
carpatica 'Blue Clips'. Scientia Hort. 87:93−105.
Niu, G., R.D. Heins, A.C. Cameron, and W.H. Carlson. 2000. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of pansy (Viola
×wittrockiana). J. Amer. Soc. Hort. Sci. 125:436−441.
Nozaki, K., T. Takamura, and S. Fukai. 2006. Effects of high temperature on flower colour and
anthocyanin content in pink flower genotypes of greenhouse chrysanthemum
(Chrysanthemum ×morifolium Ramat.). J. Hort. Sci. Biotech. 81: 728−734.
Oh, W., I. Cheon, K. Kim, and E.S. Runkle. 2009. Photosynthetic daily light integral influences
flowering time and crop characteristics of Cyclamen persicum. HortScience 44:341−344.

46

Oh, W., E.S. Runkle, and R.M. Warner. 2010. Timing and duration of supplemental lighting
during the seedling stage influence quality and flowering in petunia and pansy.
HortScience 45:1332−1337.
Park, B.H., N. Oliveira, and S. Pearson. 1998. Temperature affects growth and flowering of the
balloon flower [Platycodon grandiflorus (Jacq.) A. DC. cv. Astra Blue]. HortScience
33:233−236.
Pearson, S., A. Parker, S.R. Adams, P. Hadley, and D.R. May. 1995. The effects of temperature
on the flower size of pansy (Viola ×wittrockiana Gams.). J. Amer. Soc. Hort. Sci.
70:183−190.
Pietsch, G.M., W.H. Carlson, R.D. Heins, and J.E. Faust. 1995. The effect of day and night
temperature and irradiance on development of Catharanthus roseus (L.) 'Grape Cooler'.
J. Amer. Soc. Hort. Sci. 120:877−881.
Pramuk, L.A.and E.S. Runkle. 2005a. Photosynthetic daily light integral during the seedling
stage influences subsequent growth and flowering of Celosia, Impatiens, Salvia, Tagetes,
and Viola. HortScience 40:1336−1339.
Pramuk, L.A.and E.S. Runkle. 2005b. Modeling Growth and Development of Celosia and
Impatiens in Response to Temperature and Photosynthetic Daily Light Integral. J. Amer.
Soc. Hort. Sci. 130:813−818.
Raison, J.K., J.A. Berry, P.A. Armond, and C.S. Pike. 1980. Membrane properties in relation to
the adaptation of plants to temperature stress, p. 261−273. In: N.C. Turner and P.J.
Kramer (eds.). Adaptation of plants to water and high temperature stress. Jon Wiley &
Sons, Inc., New York, USA.
Roberts, A.N., Y.T. Wang, and F.W. Moeller. 1983. Effects of pre- and post-bloom temperature
regimes on development of Lilium longiflorum Thunb. Scientia Hort. 18: 363−379.
Roberts, E.H.and R.J. Summerfield. 1987. Measurement and prediction of flowering in annual
crops, p. 17−50. In: J.G. Atherton (ed.). Manipulation of flowering. Butterworths,
London.
Roldan, M., C. Gomez-Mena, L. Ruiz-Garcia, J. Salinas, and J.M. Martinez-Zapater. 1999.
Sucrose availability on the aerial part of the plant promotes morphogenesis and flowering
of Arabidopsis in the dark. Plant J. 20:581−590.
Runkle, E.S.and R.D. Heins. 2001. Specific functions of red, far red, and blue light in flowering
and stem extension of long-day plants. J. Amer. Soc. Hort. Sci. 126: 275−282.
Runkle, E.S., R.D. Heins, A.C. Cameron, and W.H. Carlson. 2001. Photocontrol of flowering
and stem extension of the intermediate-day plant Echinacea purpurea. Physiol. Plant.
112:433−441.

47

Scortecci, K., S.D. Michaels, and R.M. Amasino. 2003. Genetic interactions between FLM and
other flowering-time genes in Arabidopsis thaliana. Plant Mol. Biol. 52: 915−922.
Shimizu, H., E.S. Runkle, and R.D. Heins. 2004. A steady-state model for prediction of
poinsettia plant shoot-tip temperature. J. Amer. Soc. Hort. Sci. 129:303−312.
Shvarts, M., D. Weiss, and A. Borochov. 1997. Temperature effects on growth, pigmentation and
post-harvest longevity of petunia flowers. Scientia Hort. 69: 217−227.
Smith, D.R.and R.W. Langhans. 1961. The influence of day and night temperatures on the
growth and flowering of the Easter lily (Lilium longiflorum Thunb. var Croft). J. Amer.
Soc. Hort. Sci. 80:593−599.
Solfanelli, C., A. Poggi, E. Loreti, A. Alpi, and P. Perata. 2006. Sucrose-specific induction of the
anthocyanin biosynthetic pathway in Arabidopsis. Plant Physiol. 140:637−646.
Stiles, E.A., N.B. Cech, S.M. Dee, and E.P. Lacey. 2007. Temperature-sensitive anthocyanin
production in flowers of Plantago lanceolata. Physiol. Plant. 129: 756−765.
Strefeler, M.S. 1995. Growth response of 20 seed geranium cultivars to three day-night
temperature regimes. HortScience 30:1465−1466.
Sureshkumar, B., S. Sridevi, J. Lempe, and D. Weigel. 2006. Potent induction of Arabidopsis
thaliana flowering by elevated growth temperature. PLoS Genetics 2: e106.
Torres, A.P.and R.G. Lopez. 2011. Photosynthetic daily light integral during propagation of
Tecoma stans influences seedling rooting and growth. HortScience 46:282−286.
U.S. Dept. of Agriculture. 2011. Floriculture crops 2010 summary. Agri. Stat. Board,
Washington, D.C.
van Iersel, M.W. 2003. Short-term temperature change affects the carbon exchange
characteristics and growth of four common bedding plant species. J. Amer. Soc. Hort.
Sci. 128:100−106.
Vlahos, J.C., G.F.P. Martakis, and E. Heuvelink. 1992. Daylength, light quality, and temperature
influence growth and development of Achimenes. HortScience 27: 1269−1271.
Vogelezang, J.V.M. 1990. Effect of root-zone and air temperature on flowering, growth and
keeping quality of Begonia ×hiemalis 'Toran'. Scientia Hort. 44:135−147.
Vogelezang, J.V.M. 1992. Effect of root-zone and air temperature on flowering and growth of
Spathiphyllum and Guzmania minor 'Empire'. Scientia Hort. 49:311−322.

48

Wai, K.S.and S.E. Newman. 1992. Air and root-zone temperatures influence growth and
flowering of snapdragons. HortScience 27:796−798.
Wang, Y.T.and A.N. Roberts. 1983. Influence of air and soil temperatures on the growth and
development of Lilium longiflorum Thunb. during different growth phases. J. Amer. Soc.
Hort. Sci. 108:810−815.
Warner, R.M. 2010. Temperature and photoperiod flowering and morphology of four Petunia
spp. HortScience 45:365−368.
Warner, R.M.and J.E. Erwin. 2003. Effect of photoperiod and daily light integral on flowering of
five Hibiscus sp. Scientia Hort. 97:341−351.
Warner, R.M.and J.E. Erwin. 2005. Prolonged high temperature exposure and daily light integral
impact growth and flowering of five herbaceous ornamental species. J. Amer. Soc. Hort.
Sci. 130:319−325.
Warner, R.M.and J.E. Erwin. 2006. Prolonged high temperature exposure differentially reduces
growth and flowering of 12 Viola ×wittrockiana Gams. cvs. Scientia Hort. 108:295−302.
Welander, N.T. 1983. Effect of temperature and irradiance on vegetative and generative
development in Pelargonium ×hortorum 'Radio'. Scientia Hort. 21:283−290.
Went, F.W. 1944. Plant growth under controlled conditions. II. Thermoperiodicity in growth and
fruiting of the tomato. Amer. J. Bot. 31:135−150.
White, J.W.and I.J. Warrington. 1988. Temperature and light integral effects on growth and
flowering of hybrid geraniums. J. Amer. Soc. Hort. Sci. 113:354−359.
Whitman, C.M., R.D. Heins, A. Cameron, and W.H. Carlson. 1997. Cold treatment and forcing
temperature influences flowering of Campanula carpatica 'Blue Clips'. HortScience
32:861−865.
Yuan, M., W.H. Carlson, R.D. Heins, and A. Cameron. 1998. Effect of forcing temperature on
time to flower for Coreopsis grandiflora, Gaillardia ×grandiflora, Leucanthemum
×superbum, and Rudbeckia fulgida. HortScience 33:663−667.
Zieslin, N.and M.J. Tsujita. 1988. Regulation of stem elongation of lilies by temperature and the
effect of gibberellin. Scientia Hort. 37:165−169.

49

SECTION II

EFFECT OF AVERAGE DAILY TEMPERATURE ON FLOWERING TIME AND
PLANT QUALITY OF EIGHTEEN SPECIES AND CULTIVARS OF ANNUAL
BEDDING PLANTS

50

Effect of Average Daily Temperature on Flowering Time and Plant Quality of Eighteen
Species and Cultivars of Annual Bedding Plants

1

Tasneem M. Vaid and Erik S. Runkle

2

Department of Horticulture, Michigan State University, East Lansing, MI 48824

Received for publication__________. Accepted for publication__________. We gratefully
acknowledge funding by Michigan’s plant agriculture initiative at Michigan State University
(Project GREEEN), the Michigan Agricultural Experiment Station, the USDA-ARS Floriculture
and Nursery Research Initiative, and greenhouse growers providing support for Michigan State
University floriculture research. We thank C. Raker & Sons for donation of plant material and
Mike Olrich for his greenhouse assistance.

1
2

M.S. Candidate
Corresponsing author and to whom reprint requests should be addressed. Email address:

runkleer@msu.edu.

51

Abstract
The effect of average daily air temperature on the flowering time and plant quality was
quantified for 18 cultivars of 16 common bedding plant species. Antirrhinum, Calendula,
Diascia, Gerbera, Gomphrena, Heliotropium, Impatiens, Matthiola, Nemesia, Nicotiana,
Nierembergia, Osteospermum, Pelargonium, Petunia, Tagetes, and Torenia were grown in five
glass greenhouse compartments maintained at constant temperature setpoints of 14, 17, 20, 23 or
26 °C. The 16-h photoperiod consisted of the natural photoperiod with supplemental highpressure sodium lighting from 0600 to 2200 HR. The mean photosynthetic daily light integral
–2 –1

during the two replications of this experiment varied from 9 to 18 mol·m ·d . Days to flower
from transplant (DTF), flower or inflorescence number (FN), flower or inflorescence diameter
(FD), branch number (BN), number of nodes below the first open flower (NN), dry shoot mass
(SM), dry root mass (RM), and plant height or length of the longest lateral branch (PH) were
measured when the first flower opened on each plant. As temperature increased from 14 to 26
°C, DTF decreased for all crops except Impatiens walleriana Hook. Plant quality parameters
(FN, FD, BN, SM, RM, and PH) increased as temperature decreased for 14, 9, 8, 14, 7, and 10
crops, respectively. Linear or non-linear regression analysis was performed on the flowering rate
(reciprocal of days to flower) data to estimate the base temperature (Tmin) for each species,
which is the temperature at which the flowering rate is zero. Tmin ranged from –3.9 ºC for
Diascia barberae Hook.f. to 13.8 ºC for Gomphrena globosa L. A positive correlation was
observed between the estimated Tmin and the percentage delay in flowering time when grown at
17 versus 23 ºC. This information can be used by growers to predict plant timing and crop
quality at different temperatures, and to grow crops with a similar temperature response together
for energy-efficient greenhouse production.

52

Introduction
Bedding plants are typically grown during some of the coldest months of the year,
especially in northern climates, where heating typically accounts for ≥ 10% of the annual
greenhouse production costs (Bartok, 2001). Some growers lower their greenhouse temperatures
in winter in an attempt to save on heating costs. However, potential fuel savings at lower
temperatures are generally offset by an increase in the crop production period (Blanchard et al.,
2011b). Volatile fuel prices, an increased emphasis on sustainable production practices, and
declining profit margins have motivated growers to optimize their greenhouse environments and
grow crops as energy-efficiently yet profitably as possible. Energy-efficient production in
temperate climates requires species-specific information on how average daily temperature
(ADT) influences flowering time and plant quality.
Average daily temperature regulates plant development by influencing different events in
a plant’s life cycle such as germination, leaf unfolding, and flowering. The time taken to reach a
particular developmental event as a function of temperature can be expressed as the number of
days to reach that event (n) or as a developmental rate by taking the reciprocal of that number of
days (1/n). Temperature influences flowering time of crops such that as ADT increases within a
species-specific range, days to flower (DTF) decreases (Adams et al., 1999, 1998b, 1998c;
Blanchard and Runkle, 2011a; Clough et al., 2001; Kanellos and Pearson, 2000; Miller and
Armitage, 2002; Moccaldi and Runkle, 2007; Niu et al., 2000, 2001; Pramuk and Runkle, 2005b;
White and Warrington, 1988). A decrease in DTF with an increase in ADT can be due to earlier
flower initiation, which can be observed by repeated apical dissections or reduced leaf number
below the first open flower (Adams et al., 1998b, 1998c; Mattson and Erwin, 2003), or from a
faster development rate (Brøndum and Heins, 1993; Faust and Heins, 1994; Karlsson and

53

Werner, 2001; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al., 1998). When plants are
grown at an ADT above the optimum temperature (Topt), the flower development rate (reciprocal
of DTF) begins to decline (Blanchard and Runkle, 2011a; Brøndum and Heins, 1993; Faust and
Heins, 1994; Semeniuk, 1975). A decrease in flowering rate at temperatures > Topt is known as
heat delay, and may be due to a delay in flower induction, initiation and/or development (Warner
and Erwin, 2006).
In many plants, a linear function has been frequently used to describe the relationship
between flowering rate and ADT, when ADT is between Tmin and Topt (Clough et al., 2001; Niu
et al., 2000, 2001; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al., 1998). With a linear
function, plant development rate increases by the same incremental amount with each degree rise
in temperature, and this relationship can be useful in estimating the Tmin and the thermal time (θ)
required for a particular developmental event (Roberts and Summerfield, 1987). The base
temperature is the minimum temperature below which no development occurs, and thermal time
is the summation of all degree days above Tmin required for a particular developmental event to
occur. A quadratic function was found to best describe the effect of ADT on DTF of some crops,
since the decrease in DTF with an increase in ADT was not linear between Tmin and Topt
(Armitage et al., 1981; Clough et al., 2001; Niu et al., 2000, 2001; Park et al., 1998; Pietsch et
al., 1995; Whitman et al., 1997; Yuan et al., 1998). For example, a greater decrease in flowering
time was observed when ADT increased from 15 to 20 °C compared to when ADT increased
from 20 to 25 °C. Although linear and quadratic equations can be used to simply describe these
responses, they lack biological significance because their constants do not have physiological
meaning (Landsberg, 1977).

54

Several studies have quantified the effect of ADT on flowering time using mathematical
models that contain parameters of biological significance (Tmax, Topt and/or Tmin) when
temperatures beyond the species-specific linear limit are included in the experimental range
(Blanchard and Runkle, 2011a; Brøndum and Heins, 1993; Faust and Heins, 1993; Hiden and
Larsen, 1994; Larsen, 1988, 1989; Larsen and Persson, 1999). The species-specific Tmin, Topt,
or Tmax are usually determined by the natural origin and distribution of a particular species. For
example, tussock bellflower (Campanula carpatica Jacq.), native to the Carpathian Mountains of
Eastern Europe, had an estimated Tmin for flowering of 0.0 °C (Whitman et al., 1997). In
contrast, for browalia (Browalia speciosa Hook.) and angelonia (Angelonia augustifolia Benth.),
both of tropical origin, the estimated Tmin was 8.9 and 9.9 °C, respectively (Blanchard, 2009).
Breeding efforts have introduced desirable traits such as early flowering, frost tolerance, and heat
tolerance. Therefore, depending on the breeding objective, cultivars of the same species may
differ in their base and optimum temperatures, as well as thermal time for a particular event such
as flowering. For example, the Tmin calculated for petunia (Petunia ×hybrida Vilm.-Andr.)
‘Dreams Neon Rose’ was 2.7 °C lower than that of ‘Wave Purple’ (Blanchard, 2009).
Base and optimum temperatures within a species can also vary with the developmental
process (Clough et al., 2001; Karlsson et al., 1989a; Pearson et al., 1995; Yuan et al., 1998;
Whitman et al., 1997), physiological process (van Iersel, 2003) and environmental conditions
such as photoperiod (Adams et al., 1997, 1998a, 1998b) and DLI (Adams et al., 1997; Faust and
Heins, 1993; Pramuk and Runkle, 2005b). For example, Topt estimated for four different phases
of growth and development in chrysanthemum (Chrysanthemum ×morifolium Ramat.) varied
from 19.2 to 23.1 °C (Karlsson et al., 1989a). Similarly, Adams et al. (1998a) reported a linear

55

increase in Topt for flower development rate in petunia with an increase in photoperiod, from
20.7 °C under an 8-h photoperiod to 24.3 °C under a 14.4-h photoperiod.
Estimation of Tmin is useful in calculating the thermal time for a particular event and
categorizing plants based on their thermal tolerance (Blanchard, 2009). Blanchard (2009)
reported a greater delay in flowering at lower temperatures for species with a higher Tmin as
compared to species with lower Tmin values. For example, flowering time of petunia ‘Dreams
Neon Rose’ (Tmin = 2.8 ºC) and summer snapdragon (Tmin = 9.9 ºC) was delayed by 28% and
81% when grown at 15 ºC versus 20 ºC, respectively (Blanchard, 2009). Therefore, estimations
of Tmin and Topt allow growers to quickly categorize crops according to their temperature
sensitivity and grow them together for energy-efficient production (Blanchard, 2009).
Although higher temperatures can decrease flowering time, faster crop timing may not
always be desirable, since plant quality parameters like flower bud number, flower size, branch
number, and plant biomass are often inversely related with ADT (Blanchard et al., 2011a, 2011b;
Mattson and Erwin, 2003; Moccaldi and Runkle, 2007; Niu et al., 2000, 2001; Pramuk and
Runkle, 2005b; Warner and Erwin, 2006). For example, as ADT increased from 15 to 25 ºC
–2 –1

(under 15 mol·m ·d ), plant biomass, inflorescence number, and inflorescence diameter in
marigold (Tagetes patula L.) decreased by 35, 53, and 31%, respectively (Moccaldi and Runkle,
2007). Similarly, in petunia ‘Dreams Neon Rose’, each degree rise in ADT decreased the flower
and branch number by 0.9 and 0.3, respectively (Mattson and Erwin, 2003). Therefore, there is
often a trade-off between fast crop timing and high plant quality, especially under light-limiting
conditions.
For many annual bedding plant species, the flowering time response to ADT has been
quantified, but the nature of that response varies widely among species and cultivars (Adams et

56

al., 1998; Armitage et al., 1981; Blanchard and Runkle, 2011a; Moccaldi and Runkle, 2007;
Mattson and Erwin, 2003; Pietsch et al., 1995; Pramuk and Runkle, 2005a). Therefore, researchbased information that describes the effect of ADT on flowering time and plant quality is needed
for crops in which data has not been published. The objective of this research was to quantify the
effect of ADT on flowering time and plant quality of 18 species and cultivars of popular bedding
plants using linear and non-linear mathematical models, and to estimate Tmin for each of the
species and cultivars studied.

Materials and Methods
Plant material: Experimental protocol, data collection and analysis was similar to that
reported by Blanchard and Runkle (2011) and Pramuk and Runkle (2005a). Seedlings of
American marigold (Tagetes erecta L. ‘Inca II Mix’), cup flower (Nierembergia caerulea
(Miers) Millán ‘Purple Robe’], diascia (Diascia barberae Hook.f. ‘Diamonte Mix’), flowering
tobacco (Nicotiana alata Link and Otto ‘Perfume Deep Purple’), geranium (Pelargonium ×
hortorum L.H.Bailey ‘Pinto Red’ and ‘Ringo 2000 Deep Red’), gerbera (Gerbera jamesonii
Bolus ex Hook. f. ‘Jaguar Deep Orange’), globe amaranth (Gomphrena globosa L. ‘Gnome
Purple’), heliotrope (Heliotropium arborescens L. ‘Blue Wonder’), impatiens (Impatiens
walleriana Hook. f. ‘Blitz 3000 Deep Orange’), nemesia (Nemesia foetans Vent. ‘Poetry
White’), New Guinea impatiens (Impatiens hawkeri W. Bull ‘Divine Cherry Red’),
osteospermum [Osteospermum ecklonis (D.C.) Norl. ‘Asti Purple’], petunia ‘Bravo Blue’, pot
marigold (Calendula officinalis L. ‘Bon Bon Orange’), snapdragon (Antirrhinum majus L.
‘Liberty Classic Cherry’), stock [Matthiola incana (L.) W.T. Aiton ‘Hot Cakes Purple’], and
torenia (Torenia fournieri Linden ex E. Fourn. ‘Clown Blue’) were grown in 288-cell (6-mL),

57

128-cell (12-mL) or 36-cell (42.3-mL) plug trays depending on species (Table 2.1) by a
commercial greenhouse (C. Raker & Sons, Litchfield, MI) and received at Michigan State
University (MSU) on 30 March 2010 and 13 April 2010. Gerbera ‘Jaguar Deep Orange,’ petunia
‘Bravo Blue’ and stock ‘Hot Cakes Purple’ grown during the spring experiment were replaced
with ‘Jaguar Deep Rose,’ ‘Single Frost Blue’ and ‘Garden Vintage Mix’ for the fall experiment,
respectively, due to availability.
The experiment was repeated in fall 2010 to record data under different DLI conditions
using the protocol described below. Seedlings were received at MSU on 26 October 2010 and
grown in the controlled environment chambers until the leaf number for each species was similar
to that of the plugs grown in spring 2010. Once the desired leaf count was achieved, the
seedlings were transplanted and grown in greenhouses maintained the five different temperature
set points. Time between seed sow and transplant was within 12 d between replications in time
(Table 2.1).
Growth Chamber Treatments and Environment: Seedlings that were not ready for
transplant on receipt were grown in controlled environment chambers at a constant temperature
–2 –1

setpoint of 20 ºC under a PPF of 180 µmol·m ·s

–2 –1

(DLI ≈ 10 mol·m ·d ) provided by a

combination of cool-white fluorescent (CWF; F96T12CWVHO; Philips, Somerset, NJ) and
incandescent lamps (INC, Philips, Somerset, NJ) with a 16-h photoperiod. The seedlings were
subjectively deemed ready for transplant ≈ 3–5 weeks after seed sow and were accordingly
transplanted. The mean leaf number at transplant was recorded for each species (Table 2.1). The
light intensity in the environment chambers was checked periodically with an instantaneous
quantum sensor (Apogee Instruments Inc., Logan, UT) at canopy height and adjustments were
made by replacing and/or lowering the lamps when required. Plugs were irrigated by hand as

58

necessary with acidified well water (140 mg·L

–1

titratable alkalinity of CaCO3) containing

–1

(mg·L ) 95, 34, and 29 Ca, Mg, and S, and supplemented with a water-soluble fertilizer
–1

providing (mg·L ) 62 N, 6 P, 62 K, 7 Ca, 0.5 Fe, 0.3 Cu, Mn, and Zn, 0.1 B and Mo (MSU
Well Water Special; GreenCare Fertilizers Inc., Kankakee, IL).
Greenhouse Treatments and Environment: Plants that were deemed ready for transplant
were transplanted into 10-cm round containers (480-mL) filled with a peat-based medium
(Suremix, Michigan Grower Products, Galesburg, MI). Nemesia and diascia seedlings were
flowering on receipt, and thus were pinched to 3 to 4 nodes and treated with a 500 mg·L
–2

of ethephon at a volume of 0.2 L·m

–1

spray

(Florel; Bayer CropScience LP, NC) prior to transplant.

The seedlings were thinned to one plant per cell prior to transplant and 10 plants of each species
were grown at constant temperature set points of 14, 17, 20, 23, or 26 °C in separate glass-glazed
greenhouse compartments. Plants were grown under a 16-h photoperiod (0600 to 2200 HR)
created by using the natural photoperiod (lat. 43 °N) and day-extension lighting from high–2 –1

pressure sodium (HPS) lamps that provided a PPF of 100 to 120 µmol·m ·s . In each
greenhouse compartment, a shielded and aspirated 0.13 mm type E thermocouple (Omega
Engineering, Stamford, CT) recorded the air temperature and a line quantum sensor containing
10 photodiodes (Apogee Instruments) placed at canopy height (22 cm above bench height)
recorded the light intensity. A CR10 datalogger (Campbell Scientific, Logan, UT) collected the
environmental data every 10 s and hourly averages were recorded. Actual average air
temperatures during the spring and fall experiments were 18.4, 18.9, 21.4, 23.6, or 26.0 ºC and
14.0, 17.0, 19.6, 23.1, or 25.6 ºC, respectively. Mean DLI during the spring and fall experiments
–2 –1

were 18.0 and 9.0 mol·m ·d , respectively. Average DIF values (day – night temperature)
ranged from 0.5 to 2.0 ºC, and 0.4 to 0.9 ºC in the spring and fall experiments, respectively.

59

Vapor pressure deficit was maintained between 0.8 and 1.1 kPa by steam injection. Plants were
irrigated as necessary with reverse osmosis water supplemented with a water-soluble fertilizer
–1

containing (mg·L ) 125 N, 12 P, 100 K, 65 Ca, 12 Mg, 1.0 Fe and Cu, 0.5 Mn and Zn, 0.3 B,
and 0.1 Mo (MSU RO Special; GreenCare Fertilizers, Inc.).
Data collection and analysis: When each plant flowered according to the individual
characteristics for each species (Table 2.1), the date of flowering was recorded and DTF from
transplant was calculated for each species. Flowering rate was then calculated by taking the
reciprocal of DTF. Flower and flower bud number, axillary branch number, number of nodes
below first open flower on the primary stem, and plant height or length of longest lateral were
also recorded at first open flower. Branch number was not recorded in gerbera. During the fall
experiment, flower or inflorescence diameter at first open flower was also measured with the
exception of globe amaranth and snapdragon. After data collection, plants were cut at the media
surface and the roots were thoroughly washed to remove media particles. The shoots (for all
species) and roots (for 9 species) were separately placed in labeled brown paper bags and dried
in a forced-air oven (Model 630; Precision and Napco, Winchester, VA) at 79 °C for ≥ 3 d and
then weighed on an electrical balance. The experiment was set up as a randomized complete
block design; five temperature treatments were randomly assigned to the experimental units
during the fall and spring replicates (blocks).
SAS 8.0 (SAS Institute, Cary, N.C.) was used to analyze the experimental data. When
linear regression slopes between replications were non-significant (P ≤ 0.05), data was pooled
for statistical analysis. Linear and quadratic regression (REG procedure) analysis was used to
generate equations to describe the effect of ADT on plant quality parameters. Data for leaf
number below first open flower was compared among treatments using Tukey’s honestly

60

significant difference test at P ≤ 0.05. Linear regression analysis was performed on the flowering
rate data, and the slope and intercept values generated from the equations were used to calculate
the base temperatures for each species (Roberts and Summerfield, 1987). Many researchers have
estimated the Tmin, and described the flowering time response as a function of ADT using linear
regression when Tmin < ADT < Topt (Clough et al., 2001; Niu et al., 2000, 2001; Pietsch et al.,
1995; Whitman et al., 1997; Yuan et al., 1998):
1/d to flower = b0 + b1 × ADT

(1)

where 1/d to flower = the rate of progress towards flowering, b1 = slope, b0 = intercept, and
ADT = average daily temperature (°C) above Tmin and below Topt. Given that the
developmental rate at Tmin is zero, equation (1) can be used to calculate Tmin:
Tmin = – b0 / b1

(2)

where Tmin = temperature at or below which the flower development rate is zero.
In gerbera, globe amaranth, and pot marigold, an exponential function was used to
describe the flowering rate response as a function of temperature:
1/d to flower = Rmax × (1–exp (–C × (ADT – Tmin)))

(3)

where Rmax = maximum flower development rate, and C is a constant that defines the curve of
the function. This exponential function incorporates parameters of biological significance (Tmin,
Rmax), and has been used to quantify the effect of ADT on flower development and leaf
unfolding rate in several potted and annual bedding plant species (Blanchard and Runkle, 2011a;
Hidén and Larsen, 1994; Larsen 1988, 1989; Larsen and Hidén, 1995; Larsen and Persson,
1999).
In impatiens, Topt was observed, and therefore a non-linear model was used to describe
the relationship between flower development rate and ADT (Landsberg, 1977; Reed et al., 1976):

61

1/d to flower = A × (ADT – Tmin) × (Tmax – ADT)

B

B

(4)

where A = Rmax / ((Topt – Tmin) × (Tmax – Topt) )

(5)

and B = (Tmax – Topt) / (Topt – Tmin)

(6)

where Topt is the optimum temperature above which the flower development rate starts to
decrease until it reaches zero at Tmax, and B describes the curve of the function. This non-linear
model includes constants that have biological meaning (Tmin, Topt, Tmax, and Rmax), and have
been previously used to model an asymmetrical temperature response curve in African violet
(Saintpaulia ionantha Wendl.; Faust and Heins, 1993), dahlia (Dahlia pinnata Cav.; Brøndum
and Heins, 1993), and several annual bedding plant species (Blanchard and Runkle, 2011a).
Flowering rate data for impatiens in the first replicate did not show a significant trend, and could
not be fit with a linear or non-linear model (Eqns. [1], [3], and [4]). Since we had only 50
observations from the second replicate, Tmax was fixed at 35 ºC (Blanchard and Runkle, 2011a)
and Topt was fixed at 25.1 (estimated from quadratic equations generated from the observed
data) to allow for precise estimates of Tmin and Rmax.
Non-linear regression (NLIN procedure) of SAS was used to generate non-linear models
and estimate the values for Tmin, Rmax, and C for eqns. [3] and [4] using 50 to 100 observations.
Initial parameter estimates for Tmin and Rmax were obtained from eqn. [2] and observed data,
respectively. Eqn. [3] was solved using initial estimates for Tmin and Rmax and a specific value
of temperature to estimate an initial value for C. R2 values for non-linear regression were
generated by linearly regressing predicted values on the observed data (Blanchard and Runkle,
2011a).

62

Results
Flower development rate increased with an increase in ADT from 14 to 26 ºC for all the
species studied except impatiens, but the magnitude of this response varied widely among
species and cultivars (Figures 2.1 and 2.2). For example, the increase in rate of progress towards
flowering with each degree increase in temperature from 14 to 26 ºC ranged from 0.00092 in
diascia to 0.0027 in torenia (Table 2.2). The relationship between flowering rate and ADT was
best described using a linear function in 14 of the 18 species studied, whereas a non-linear
function was used for globe amaranth, gerbera, impatiens, and pot marigold (Figures 2.1 and
2.2). In American marigold, diascia, flowering tobacco, geranium ‘Pinto Red’, impatiens,
petunia, and torenia, the flowering rate response to ADT varied between the spring (replication
1) and fall (replication 2) experiments, and separate functions were used to describe the response
for each replicate. Differences in flowering time response between replications in petunia can
most likely be attributed to cultivar differences, since the petunia cultivar ‘Bravo Blue’ grown
during the spring was replaced with ‘Single Frost Blue’ in the fall replication. The coefficients of
2

2

determination generated for linear (r ) and non-linear models (R ) ranged from 0.33 to 0.94
(Table 2.2). The variability in flowering time response differed between replications of diascia
2

and cultivars of petunia, which considerably influenced the r values. For example, the r2 values
of diascia were 0.65 in replication 2 and 0.33 in replication 1 (Table 2.2).
The estimated Tmin for flower development rate varied among species and cultivars, and
in some cases between replications for the same species (Figures 2.1 and 2.2). For example, Tmin
range from –3.9 ºC for diascia to 13.8 ºC for globe amaranth (Figures 2.1 and 2.2). When Tmin
varied between replications or cultivars within a species, it was generally < 2 ºC. Species and
cultivars in which the estimated Tmin was ≤ 4 ºC were American marigold, diascia, heliotrope,

63

nemesia, New Guinea impatiens, osteospermum, petunia ‘Bravo Blue’, snapdragon, and stock.
Those with an estimated Tmin between 4 to 8 ºC were cupflower, flowering tobacco, geranium
(both cultivars), gerbera, impatiens, petunia ‘Single Frost Blue’, and pot marigold. The only two
crops studied with a Tmin > 8 ºC were globe amaranth and torenia. All of globe amaranth and
20% of torenia plants died when grown at 14 ºC. Although 80% of torenia plants flowered at 14
ºC, they were of poor quality and likely not commercially marketable. There was a positive
correlation (P < 0.0001) between the estimated Tmin and delay in flowering time when plants
were grown at 17 versus 23 ºC. For example, the delay in flowering time when plants were
grown at 17 ºC compared to 23 ºC was 105, 53, and 34% in globe amaranth (Tmin = 13.8 ºC),
geranium ‘Ringo 2000 Deep Red’ (Tmin = 5.7 ºC), and osteospermum (Tmin = –0.5 ºC),
respectively.
Flower or inflorescence number increased with an increase in ADT from 14 to 26 ºC for
14 of the 18 species studied (Figures 2.3-2.9, 2.11-2.13, 2.15, 2.17-2.19, panel A). For example,
an increase in ADT from 14 to 26 ºC increased the predicted inflorescence number in pot
marigold by 24. In globe amaranth, New Guinea impatiens, petunia ‘Single Frost Blue’, and
torenia, flower number increased with temperature up to 20 ºC and then decreased (Figures 2.12,
2.14, 2.16, 2.20, panel B). The response of flower bud number to ADT differed between
replications for impatiens and petunia (Figures 2.12, 2.16). For example, in replication 1,
impatiens flower number showed a quadratic response with an increase in temperature. In
contrast, in replication 2, flower bud number linearly increased with a decrease in temperature.
Average daily temperature had various effects on branch number among the species
studied (Figures 2.3-2.20, panel B). In diascia, flowering tobacco, geranium ‘Pinto Red’ and
‘Ringo 2000 Deep Red’, nemesia, pot marigold, snapdragon, and stock, branch number increased

64

with ADT. There was a positive correlation between flower or inflorescence number and branch
number in these species and cultivars (P < 0.0001). There was no such relationship between
replications in cupflower, diascia, heliotrope, impatiens, petunia, and stock. For example, in
stock, branch number increased as ADT decreased from 26 to 14 ºC in the first replication, but a
significant temperature response was not observed in the second replication. In general, number
of flowers and branches were higher during the spring (when the DLI was higher) compared to
the fall experiment.
A decreasing temperature caused a developmental delay in flowering in cupflower,
geranium ‘Pinto Red’ and ‘Ringo 2000 Deep Red’, gerbera, petunia, and snapdragon; leaf
number below the first flower decreased linearly or quadratically with an increase in ADT (Table
2.3). In contrast, leaf number decreased with a decrease in ADT in American marigold,
heliotrope, impatiens, and torenia. Temperature did not significantly affect leaf number below
first open flower in diascia, flowering tobacco, globe amaranth, nemesia, New Guinea impatiens,
osteospermum, pot marigold, and stock.
As temperature increased from 14 to 26 ºC, plant height decreased in cupflower,
flowering tobacco, geranium ‘Pinto Red’ and ‘Ringo 2000 Deep Red’, gerbera, nemesia,
osteospermum, pot marigold, snapdragon, and stock (Figures 2.4, 2.6-2.9, 2.13, 2.15, 2.17-2.19,
panel C). However, among these species and cultivars, a decrease in leaf number with an
increase in ADT only occurred in cupflower, geranium ‘Pinto Red’ and ‘Ringo 2000 Deep Red’,
gerbera, and snapdragon. On the other hand, plant height decreased with a decrease in ADT in
African marigold, heliotrope, impatiens, and torenia. Leaf number also decreased with ADT in
these species. Heliotrope and impatiens height first increased and then decreased with ADT,

65

while there was no significant effect of temperature on plant height in diascia, flowering tobacco,
globe amaranth, nemesia, New Guinea impatiens, pot marigold, and stock.
Shoot dry mass was inversely related with ADT in cupflower, flowering tobacco,
geranium ‘Pinto Red’ and ‘Ringo 2000 Deep Red’, heliotrope, nemesia, osteospermum, petunia,
pot marigold, snapdragon, and stock (Figures 2.3-2.20, panel D). The opposite trend was
observed in globe amaranth. In American marigold, New Guinea impatiens, and torenia shoot
dry weight first increased with decrease in ADT up to an optimum temperature, and then
decreased. There was no significant relationship between shoot dry mass and ADT in diascia,
gerbera, and impatiens. Among the nine species in which root mass was measured, the
correlation between ADT and root mass was negative in flowering tobacco, geranium ‘Ringo
2000 Deep Red’, nemesia, petunia, pot marigold, snapdragon, and stock; positive in torenia; and
not significant in impatiens (Figures 2.6, 2.8, 2.12-2.13, 2.16-2.20, panel E). For example, as
ADT increased from 14 to 26 ºC, root mass decreased in flowering tobacco, but increased in
torenia. There was a positive correlation between shoot and root mass in all nine species
measured (P < 0.0001). Flower or inflorescence diameter increased with ADT from 14 to 26 ºC
in only nine of the 18 species studied (Figures 2.3-2.20, panel F). Optimum temperature for
flower or inflorescence diameter was observed in diascia, flowering tobacco, impatiens, New
Guinea impatiens, osteospermum, and torenia in the temperature range included in this study.

Discussion
Linear models adequately described the effect of temperature on flowering rate in 14 of
the 18 species studied, since the temperature range (14 to 26 ºC) included in this experiment was
between Tmin and Topt for these species. This temperature range was selected because it

66

encompasses temperatures in which most bedding plants are commercially grown in
greenhouses. Non-linear models were evaluated to describe the flowering rate as a function of
ADT for all the species, but they either failed to converge or did not precisely estimate the model
parameters. Several studies have used a linear function to describe the relationship between
flowering rate and ADT, when ADT is between Tmin and Topt (Clough et al., 2001; Niu et al.,
2000, 2001; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al., 1998). Although linear
models lack constants with biological meaning, they are simple and, within clearly defined
limits, can adequately describe a biological response (Landsberg, 1977). Many researchers have
quantified the effect of ADT on flowering time using mathematical models that contain
parameters of biological significance (Tmin, Tmax, and Topt) (Blanchard and Runkle, 2011a;
Blanchard et al., 2011a; Brøndum and Heins, 1993; Faust and Heins, 1993; Hidén and Larsen,
1994; Larsen, 1988, 1989; Larsen and Persson, 1999). However, these researchers studied plant
responses over a wider temperature range than the one used in this study, and either observed or
approached Tmin, Topt, or both. In our experiments, non-linear models were used in gerbera,
globe amaranth, impatiens, and pot marigold. Although Tmin or Topt was observed only in globe
amaranth and impatiens, non-linear models had a relatively high coefficient of determination,
parameter estimates with an acceptable standard error, and improved predictions, especially at
lower temperatures in gerbera and pot marigold.
The flowering rate response to ADT varied between replications in American marigold,
geranium ‘Pinto Red’, flowering tobacco, petunia, and torenia. Except for flowering tobacco and
petunia, a greater flowering rate and a lower leaf number was observed in the spring (when the
DLI was higher) compared to the fall experiment. Although flowering tobacco and petunia plants
flowered earlier at higher temperatures in fall than in the spring experiment, the difference

67

between replications was ≤ 4 d. Several studies have reported a negative correlation between
DTF and DLI during the finish stage for many annual bedding plants (Blanchard et al., 2011a,
2011b; Carvalho et al., 2006; Currey and Erwin, 2011; Warner and Erwin, 2005; Faust et al.,
2005; Oh et al., 2009; Pietsch et al., 1995). For example, an increase in DLI from 10.5 to 21.8
mol·m–2·d–1 (ADT of 20 °C) reduced DTF from transplant by 21 d in snapdragon, 16 d in pot
marigold, 3 d in impatiens, 4 d in mimulus (Mimulus hybridus Hort. Ex Siebert & Voss), and 12
d in torenia (Warner and Erwin, 2005). A higher DLI can reduce flowering time by initiating
flowering at a lower node, by accelerating the flower development rate due to an increase in
plant temperature, or both. Faust and Heins (1998) and Mattson and Erwin (2005) reported an
increase of 1 to 2 °C in plant temperature under supplemental HPS lighting.
In this experiment, the estimated Tmin ranged from –3.9 to 13.8 °C, which indicates the
range of thermal sensitivity among the different species and cultivars studied. Blanchard and
Runkle (2011a) observed a similar variability in cold tolerance among 18 species of annual
bedding plants. Our previous research on Tmin estimates for the same species (Vaid and Runkle,
2011) has been revised in the current study using observations from two replications performed
under different DLI conditions. The estimated Tmin values here are generally ± 2 °C of those
estimated in previous studies for the same species, but at times different cultivars. For example,
our Tmin estimated for snapdragon and geranium was 2.0 and 0.7 to 1.5 °C higher, respectively,
than that estimated by Blanchard (2009). Pramuk and Runkle (2005b) and Moccaldi and Runkle
(2007) estimated Tmin for impatiens and marigold at 7.5 and –3.0 °C, which was 0.3 higher and
1.7 °C lower than our estimates of 7.2 and –1.3 °C, respectively.
When the flowering rate response varied between replications, Tmin estimated under the
high DLI (spring) was 0.7 to 2.8 °C higher than that estimated under the low DLI (fall), except in

68

flowering tobacco and petunia. Several studies have evaluated the effect of DLI on the estimated
Tmin or Topt for flower development rate, but the results have not been consistent (Adams et al.,
1997; Faust and Heins, 1993; Moccaldi and Runkle, 2007; Pramuk and Runkle, 2005b). For
–2 –1

example, an increase in DLI from 5 to 15 mol·m ·d

decreased the estimated Tmin of celosia

(Celosia argentea L.) and impatiens by 1.5 and 3.2 °C, respectively (Pramuk and Runkle,
2005b), but did not influence the Tmin in marigold and salvia (Salvia splendens F. Sello ex Roem
& Schult.) (Moccaldi and Runkle, 2007). In contrast, Adams et al. (1997), Faust and Heins
(1993), and Pietsch et al. (1995) reported a decrease in Topt or Tmin with a decrease in DLI.
Based on the estimated Tmin, species can be subjectively categorized as cold-tolerant
when Tmin ≤ 4 °C; cold-intermediate when 4 °C < Tmin < 8 °C; and cold-sensitive when Tmin ≥
8 °C. When cold-sensitive species torenia and globe amaranth were grown at 17 versus 23 °C,
the percentage delay in flowering time was 79 to 106. In contrast, in cold-tolerant species, the
delay in flowering time varied from 29 to 41%. Estimations of Tmin enables categorization of
species according to their thermal sensitivity, and therefore can help facilitate energy-efficient
crop production.
As ADT increased from 14 to 26 °C, flower number, branch number, plant height, shoot
mass, and flower diameter decreased in 14, 8, 10, 14, and 9 of the 18 crops studied, respectively.
In addition, root mass decreased as ADT increased in 7 of the 9 crops measured. Plant quality
parameter responses to temperature were consistent with previous studies (Blanchard et al.,
2011a; Blanchard et al., 2011b; Mattson and Erwin, 2003; Moccaldi and Runkle, 2007; Niu et
al., 2001; 2000; Pramuk and Runkle, 2005b). For example, an increase in ADT from 16 to 26 °C
decreased the flower number of tickseed (Coreopsis grandiflora Hogg ex Sweet) by 80%, Shasta
daisy [Leucanthemum ×superbum (Bergmans ex J.W. Ingram)] by 55%, and black-eyed Susan

69

(Rudbeckia fulgida Aiton) by 75% (Yuan et al, 1998). Similarly, each degree rise in ADT
decreased flower diameter or area by 0.97 mm in tussock bellflower ‘Blue Clips’ (under 10.8
mol·m–2·d–1; Niu et al., 2001), 1.42 cm in dahlia ‘Royal Dahlietta Yellow’ (Brøndum and Heins,
1993), 0.19 cm in sundrops (Oenothera fruticosa L.) ‘Youngii-lapsley’ (Clough et al., 2001) and
1.05 cm2 in pansy (Viola × wittrockiana Gams.) ‘Universal Violet’ (Pearson et al., 1995). Flower
bud development requires carbon import from the source organs (leaves). Sucrose hydrolysis in
the sink organs (flower buds) is necessary to establish a concentration gradient for carbon
transport between the source and the sink (Dinar and Rudich, 1985). High temperatures reduce
sucrose hydrolysis and therefore increase its concentration in flower buds, which reduces or
inhibits uptake of carbon by the developing flower buds (Dinar and Rudich, 1985), and may
consequently reduce final flower size. This is further evident from a reduction in the percentage
of dry matter that is partitioned to the flowers relative to vegetative structures under high
temperature stress in chrysanthemum (Karlsson and Heins, 1992).
Branch number was inversely related with ADT in 8 species and cultivars, and an
increase in branch number was positively correlated with an increase in flower number in those
species. Many researchers have reported a decrease in branching with an increase in ADT, such
as in balloon flower [Platycodon grandiflorus (Jacq.) A. DC. ‘Astra Blue’; Park et al., 1998],
fuchsia (Fuchsia × hybrida hort. ex Siebold & Voss; Erwin et al., 1991), pansy (Mattson and
Erwin, 2003; Warner and Erwin, 2006) and petunia (Kaczperski et al., 1991; Mattson and Erwin,
2003). In some of those cases, an increase in lateral branch number was positively correlated
with an increase in flower number (Mattson and Erwin, 2003; Warner and Erwin, 2006). Plant
quality in many of the species in this study was generally higher in the spring than in the fall
replication, which was not unexpected, since the DLI received during spring was almost double

70

that received during the fall. Several studies have reported an increase in quality with DLI
(Fausey et al., 2005; Moccaldi and Runkle, 2007; Warner and Erwin, 2005). For example,
-2 -1

increasing DLI from 10.0 to 20.0 mol·m ·d (ADT of 20 °C) increased flower bud number by
63% in snapdragon, 56% in impatiens, 61% in mimulus, and 15% in wishbone flower (Warner
and Erwin, 2005). Similarly, in the herbaceous perennial butterfly gaura (Gaura lindheimeri
Engelm. and Gray), lateral inflorescence number per plant tripled and flower number per
inflorescence nearly doubled as DLI increased from 5.0 to 20.0 mol·m-2·d-1 (Fausey et al., 2005).
Shoot dry mass at first flowering increased as ADT decreased in 11 of the species and
cultivars studied. Similar responses were observed in balloon flower (Park et al., 1998), pot
marigold (Warner and Erwin, 2005), geranium (White and Warrington, 1988), impatiens
(Pramuk and Runkle, 2005b; Warner and Erwin, 2005), marigold (Moccaldi and Runkle, 2007),
mimulus (Warner and Erwin, 2005), pansy (Niu et al., 2000; Warner and Erwin, 2006), salvia
(Moccaldi and Runkle, 2007), and snapdragon (Wai and Newman, 1992; Warner and Erwin,
2005). These results are not surprising since plants grown at lower temperatures take longer to
flower (and thus can photosynthesize for a longer time before flowering) and are also of higher
quality (increased branch number and flower number and size) compared to plants grown at
higher temperatures. In addition, van Iersel (2003) reported an increase in the net photosynthetic
rate (Pnet) with a decrease in ADT for geranium, marigold, pansy and petunia. High temperatures
decrease Pnet by modifying Rubisco kinetics (Brooks and Farquhar, 1985), increasing O2:CO2
solubility (Ku and Edwards, 1977), or both.
In contrast, shoot biomass increased with increasing temperature in American marigold,
globe amaranth, New Guinea impatiens, and torenia. Similar results were reported in celosia
(Pramuk and Runkle, 2005b), summer snapdragon (Miller and Armitage, 2002), Texas firebush

71

(Hamelia patens Jacq.; Armitage, 1995), and torenia (Warner and Erwin, 2005). Many of these
plants have a high estimated Tmin; for celosia and summer snapdragon, it was 10.2 °C under a
–2 –1

DLI of 15 mol·m ·d

(Pramuk and Runkle, 2005b) and 9.9 °C (Blanchard, 2009), respectively.

In our study, torenia and globe amaranth were identified as both cold-sensitive crops (Tmin > 8
°C). Accumulation of root mass in response to ADT followed a trend similar to that observed
for shoot mass in our study. However, these results could not be compared to previous work
since, to our knowledge, the response of root mass accumulation to ADT has not been
investigated in annual bedding plants.
Among the 18 species studied, plant height decreased in 10 species, but increased in 4
species, with an increase in temperature. Although the greenhouse temperature setpoints in these
experiments were constant, the day was often higher than the night in the lower temperature
treatments, especially in the spring replication, because of high ambient temperatures. For many
plants, height at flowering decreases as day temperature relative to the night decreases (as the
value of DIF decreases) (Erwin et. al., 1991; Erwin and Heins, 1990). For example, as DIF
increased from –6.0 to +6.0 °C, plant height increased by 40% in Easter lily (Erwin and Heins,
1990), 39% in fuchsia (Erwin et al., 1991), 19% in geranium (Strefeler, 1995), and 9% in Italian
bellflower (Campanula isophylla Moretti; Moe and Mortensen, 1992). However, since the DIF
in our experiment was variable between treatments and was generally small (0.5 to 2.0 °C), the
plant height response observed in this study is probably a function of ADT and not DIF. A high
ADT can also decrease plant height by reducing the number of nodes formed prior to flowering
(Clough et al., 2001; Yuan et al., 1998). In contrast, several studies have reported an increase in
plant height with an increase in ADT (Armitage et al., 2000; Blanchard et al., 2011a; Kanellos
and Pearson, 2000; Miller and Armitage, 2002; Moccaldi and Runkle, 2007; Pramuk and Runkle,

72

2005b). Since the number of nodes in these studies was either variable or not reported, and the
actual DIF values were also not provided, the promotive effect of higher temperatures on plant
height can not be clearly understood.

Conclusion
In general, flowering time and plant quality decrease as the ADT increases, especially
under light-limiting conditions. The nature of these responses varies widely among species and
in some cases, also with cultivars. Therefore, there is often a trade-off between faster crop timing
and higher plant quality. Growers can use the temperature response categories and crop models
generated in this study to estimate the consequences of changing the greenhouse temperature on
flowering time and plant quality parameters. In addition to temperature, environmental
conditions such as DLI and photoperiod can also affect the flower development rate (Erwin and
–2 –1

Warner, 2002). Our models assume a 16-h photoperiod and a DLI of ≥ 9 mol·m ·d , and may
not be valid under a lower DLI or if plants with a facultative photoperiodic response are grown
under a shorter photoperiod. In addition, the crop models generated in this study were at
temperatures between 14 and 26 °C, and may not be accurate outside this temperature range.
Finally, DLI greatly influences different plant quality parameters (Fausey et al., 2005; Moccaldi
and Runkle, 2007; Warner and Erwin, 2005), and therefore the linear and quadratic equations
describing the relationship between plant quality parameters and ADT may not be applicable
under DLI conditions other than those reported in this study.

73

Table 2.1. Days from seed sow to transplant (TP), plug size and mean node no. at TP, and
characteristics used to determine flowering date for 18 varieties of 16 bedding plant species in
two experimental replicates.
Days from Plug size
seed sow (mean leaf
Species
to TP
no.) at TP Flowering characteristics
1 inflorescence with ≥50%
African marigold 'Inca II Mix'
21 or 27
288 (6)
of the petals reflexed
Cupflower 'Purple Robe'

34 or 40

288 (10)

Diascia 'Diamonte Mix'

50 or 62

128 (8)

Flowering tobacco 'Perfume Deep
Purple'

32 or 34

288 (7)

Geranium 'Pinto Red'

22 or 26

288 (3)

Geranium 'Ringo 2000 Deep Red'

22 or 26

288 (3)

Gerbera 'Jaguar Deep Orange' or
'Jaguar Deep Rose'

52 or 63

128 (7)

Gerbera 'Revolution Neon Rose'

27

288 (2)

Globe amaranth 'Gnome Purple'

37 or 42

128 (8)

Heliotrope 'Blue Wonder'

44 or 49

128 (10)

Impatiens 'Blitz 3000 Deep
Orange'

27 or 33

288 (6)

1 flower open

Nemesia 'Poetry White'

61 or 73

128 (7)

2 flowers open on an
inflorescence

New Guinea impatiens 'Divine
Cherry Red'

47 or 55

128 (10)

1 flower open

Osteospermum 'Asti Purple'

24 or 39

36 (15 or
18)

1 flower open

Petunia 'Bravo Blue' or 'Single
Frost Blue'

34

288 (6)

1 flower open

Pot marigold 'Bon Bon Orange'

23 or 35

288 (8)

38 or 48

288 (12)

23 or 30

128 (6)

29 or 31

288 (7)

Snapdragon 'Liberty Classic
Cherry'
Stock 'Hot Cakes Purple' or
'Garden Vintage Mix'
Torenia 'Clown Blue'

74

1 flower open
2 flowers open on an
inflorescence
2 flowers open on an
inflorescence
5 flowers open on an
inflorescence
5 flowers open on an
inflorescence
1 inflorescence with all of
the petals reflexed
1 inflorescence with all of
the petals reflexed
1 inflorescence with outer
2 whorls of petals reflexed
2 flowers open on an
inflorescence

1 inflorescence with ≥50%
of the petals reflexed
2 flowers open on an
inflorescence
2 flowers open on an
inflorescence
1 flower open

Table 2.2. Parameter estimates for linear and non-linear models (Eqns. [1], [3], and [4])
describing the flowering rate response to average daily temperature in 18 bedding plant species
and cultivars. Tmin and Tmax are the species-specific base and maximum temperatures (ºC),
respectively at which the flower development rate is zero. Topt is the optimum temperature (ºC)
at which the flower development rate reaches a maximum value (Rmax), and C defines the
curvature of the function. CI = confidence interval.
z
2y
2x
Eq.
Parameter
Estimate
95% CI (±)
No.
R or r
African marigold 'Inca II Mix'
w
[1]
50
0.93
–0.0019
0.0021
Intercept1
Slope1
Intercept2
Slope2

0.0013
w

0.00010

0.0014

0.0024

0.0011
0.00012
Cupflower 'Purple Robe'
–0.011
0.0031
0.0018
0.00015
Diascia 'Diamonte Mix'

[1]

Intercept
Slope

[1]

Intercept1

0.012

0.0038

Slope1

0.00088
0.0035

0.0019

0.87

100

0.86

50

0.33

50

0.65

0.00018

Intercept2

50

[1]

0.00092
0.00010
Slope2
Flowering tobacco 'Perfume Deep Purple'
50
–0.0097
0.0030
Intercept1
Slope1

0.0018

0.00014

Intercept2

–0.014

Slope2
[1]

0.0022
0.00016
Geranium 'Pinto Red'
–0.0081
0.0033

Intercept1
Slope1
Intercept2

0.0012
–0.0049

0.93

0.0032

50

0.94

50

0.85

50

0.94

100

0.91

0.00015
0.0015

0.0010
0.000076
Geranium 'Ringo 2000 Deep Red'
Intercept
–0.0071
0.0016
Slope
0.0012
0.000077
Slope2

[1]

75

Table 2.2. (cont’d).
z
Eq.
Parameter
Estimate
95% CI (±)
No.
Gerbera 'Jaguar Deep Orange' or 'Jaguar Deep Rose'
[3]
100
7.6
3.67
T
min

0.039
0.023
0.055
0.064
Globe amaranth 'Gnome Purple'
13.9
0.45

Rmax
C
[3]

Tmin
Rmax
C

[1]

[4]

Intercept
Slope
Tmin

0.048
0.0080
0.13
0.046
Heliotrope 'Blue Wonder'
–0.0021
0.0027
0.0014
0.00013

Impatiens 'Blitz 3000 Deep Orange'
7.2
2.76

2y

R

or r
y

0.85

y

100

0.86

100

0.84

50

0.54

v

0.038
0.0026
Nemesia 'Poetry White'
100
Intercept
–0.0048
0.0040
Slope
0.0019
0.00019
New Guinea impatiens 'Divine Cherry Red'
97
Intercept
–0.0034
0.0069
Slope
0.0014
0.00033
Osteospermum 'Asti Purple'
64
Intercept
0.00053
0.003
Slope
0.0011
0.00016
Petunia 'Bravo Blue' or 'Single Frost Blue'
50
–0.0050
0.0035
Intercept1

y

Rmax
[1]

[1]

[1]

[1]

Slope1
Intercept2
Slope2
[3]

Tmin

0.0019
–0.014

0.81

0.43

0.78

0.92

0.00016
0.013

0.0026
0.00064
Pot marigold 'Bon Bon Orange'
6.5
2.53

50

0.58

100

0.92

0.053
0.014
0.066
0.040
Snapdragon 'Liberty Classic Cherry'
100
Intercept
–0.0070
0.0028
Slope
0.0017
0.00014

y

Rmax
C
[1]

76

0.87

2x

Table 2.2. (cont’d).
z
Parameter Estimate
95% CI (±)
No.
Eq.
Stock 'Hot Cakes Purple' or 'Garden Vintage Mix'
[1]
100
Intercept
–0.0069
0.0076
Slope
0.0022
0.00037
Torenia 'Clown Blue'
[1]
50
–0.025
0.0050
Intercept1
Slope1
Intercept2

0.0027
–0.019

2y

R

or r

2x

0.59

0.92

0.00023
0.0045

50

0.89

0.0021
0.00022
Slope2
z
Number of observations in data set.
y
Generated by linearly regressing predicted values on the observed data.
x
Generated from linear flowering rate models.
w
Parameter estimates for replicate 1 (spring) and 2 (fall), respectively.
v 1,2
Tmax was fixed at 35 ºC and Topt was fixed at 25.1 ºC. Only data from the second replicate was
included in estimating the model parameters.

77

Table 2.3. The effect of temperature on the number of nodes below first open flower in 18
species and cultivars of bedding plants. Plants were grown in glass-glazed greenhouse
compartments at five constant temperature setpoints under a 16-h photoperiod and a DLI of 18
–2 –1
–2 –1
mol·m ·d (spring) or 9 mol·m ·d (fall). When regression slopes were non-significant (P ≤
0.05) between the two replications (rep), data was pooled for statistical analysis. During the fall
replication, five or ten randomly selected plants of each species were microscopically examined
for floral initiation (FI) at transplant.
Temperature set point (ºC)
Species

20

23

26

14.2b

y

13.6b

15.1b

15.3b

17.9a

15.9a

16.4a

16.6a

16.9a

17.4a

Cupflower 'Purple Robe'

78.1a

74.6a

74.7a

72.7a

76.2a

Diascia 'Diamonte Mix'

16.0

16.9

16.0

17.1

17.0

1
2

Flowering tobacco
'Perfume Deep Purple'

11.0

10.3

9.9

10.3

9.9

NS

1

9.6a

9.9a

9.9a

9.5a

7.5b

L1

2

10.9ab

11.0ab

11.3ab

11.7a

9.5b

L1

9.5a

9.2ab

8.9abc

8.1bc

7.7c

L

37.6a

38.7a

34.7a

30.5a

25.1a

L

‒

v

5.1

4.9

5.3

5.0

12.4

12.3

13.0

13.8

12.4

z

NS

Geranium 'Pinto Red'

14

Trend

17

African marigold 'Inca II
Mix'

Rep

Geranium 'Ringo 2000
Deep Red'
Gerbera 'Jagura Deep
Orange' or 'Jaguar Deep
Rose'
Globe amaranth 'Gnome
Purple'
Heliotrope 'Blue Wonder'
Impatiens 'Blitz 3000
Deep Orange'

1

x***

x***

Q2

**

*

L1 Q2

NS *

L

Q

***
NS

Q2

Q2

***

NS

*** ***

Q

*** ***

Q

NS
NS1

***

2

10.4b

10.8b

12.2b

11.6b

16.3a

L2

1

5.7c

6.9bc

7.2abc

8.8a

8ab

L1

2

7.4

8.8

7.7

8.2

8.1

12.2

12.0

12.2

12.4

11.8

NS

10.0

9.6

10.1

9.2

10.0

NS

18.9

18.9

19.6

20.1

‒

u

NS

1

13.7ab

15.2a

12.3bc

11.2bc

10.0c

2

17.3

17.6

18.7

18.7

16.4

***

NS2

Nemesia 'Poetry White'
New Guinea impatiens
'Divine Cherry Red'
Osteospermum 'Asti
Purple'
Petunia 'Bravo Blue' or
'Single Frost Blue'

L1

78

L1

***

***

Q2
Q1

Q1

NS2

***

***

Table 2.3. (cont’d).
Temperature set point (ºC)
Species
Pot marigold 'Bon Bon
Orange'
Snapdragon 'Liberty
Classic Cherry'
Stock 'Hot Cakes Purple'
or 'Garden Vintage Mix'
Torenia 'Clown Blue'

Rep

14

17

20

23

26

Trend

15.2

15.5

16.2

15.4

15.9

NS

13.5a

12.2ab

11.3b

10.6b

11.4b

L Q

16.7

17.2

17.1

17.3

18.4

NS

6.2a

6.5a

5.8a

6.3a

6.8a

z

z

** *

NS **

L

Linear and quadratic trends describing the leaf number response to ADT
Means within rows followed by the same letter are not significantly different by Tukey’s
honestly significant difference test at P ≤ 0.05.
x
1,2Linear and quadratic trends in replicate 1 (spring) and 2 (fall), respectively.
w
Plants within the column not examined for floral initiation.
v
Treatment not included in analysis because 100% of plants died.
u
Treatment not included in the experiment.
NS,*,**,***
Nonsignificant or significant at P ≤ 0.05, 0.01, or 0.001, respectively.
y

79

Q

0.06
0.05

A

Pot marigold

C

Stock

B

Nemesia

0.04
0.03

Flowering rate (1/d to flower)

D

Heliotrope

Tmin = 2.5 °C

r2 = 0.59

R2 = 0.92

0.01
0.00

Tmin = 3.0 °C

Tmin= 6.5 °C

0.02

E

Flowering tobacco

r2 = 0.81

F

Snapdragon

0.05
Tmin = 6.7 °C

0.04

r2 = 0.94

0.03
0.02

Tmin = 1.5 °C

0.00
0.035
0.030
0.025
0.020
0.015
0.010

G

Tmin = 5.4 °C

Tmin = 4.0 °C

r2 = 0.82

0.01

r2 = 0.93

r2 = 0.87

American marigold

H

Diascia

I

Osteospermum

r2 = 0.33

Tmin = 1.5°C
r2 = 0.93
Tmin = -1.3 °C

Tmin = -3.9 °C

r2 = 0.87

r2 = 0.65

Tmin = -0.5 °C
r2 = 0.78

0.000
12 14 16 18 20 22 24 26 12 14 16 18 20 22 24 26 12 14 16 18 20 22 24 26 28

Mean daily temperature (°C)

Figure 2.1. The effect of mean daily temperature (°C) on the flowering rate in 9 species of
bedding plants modeled using Eq. [1] (panel B-I) and Eq. [3] (panel A) and parameter estimates
from Table 2.2. for replication 1 (●) and replication 2 (○). When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis.
Flowering rate was calculated by taking the reciprocal of the number of days to flower. Each
symbol represents the treatment means, and error bars represent 95% confidence intervals. Tmin
(base temperature) is the estimated minimum temperature at or below which the rate of progress
2
2
towards flowering is zero. r or R is the coefficient of determination generated for linear and
non-linear crop models, respectively.

80

0.07
0.06
0.05
0.04

Petunia

A

B

Torenia

C

Globe amaranth

Tmin = 9.4°C

Tmin = 5.5°C

r2 = 0.92

r2 = 0.58

0.03
0.02

Tmin = 2.6 °C

Flowering rate (1/d to flower)

0.01
0.00

Tmin = 8.7 °C

r2 = 0.92
D

Geranium 'Ringo'

0.030

Tmin = 13.8 °C

r2 = 0.89
E

Geranium 'Pinto Red'

R2 = 0.86
F

Gerbera

Tmin = 6.5 °C

0.025

r2 = 0.85

0.020
0.015

0.000
0.045

Tmin = 4.8 °C

r2 = 0.91
G

Cupflower

H

Tmin = 7.6 °C

r2 = 0.94

Tmin = 5.7 °C

0.010

R2 = 0.85

New Guinea impatiens

I

Impatiens

0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.000

Tmin = 6.2 °C

Tmin = 2.4 °C

r2 = 0.86

r2 = 0.43

Tmin = 7.2 °C
R2 = 0.54

12 14 16 18 20 22 24 26 12 14 16 18 20 22 24 26 12 14 16 18 20 22 24 26 28

Mean daily temperature (°C)

Figure 2.2. The effect of mean daily temperature (°C) on the flowering rate in 9 species of
bedding plants modeled using Eq. [1] (panel A, B, D, E, G, H), Eq. [3] (panel C, F), and Eq. [4]
(panel I), and parameter estimates from Table 2.2. for replication 1 (●) and replication 2 (○).
When regression slopes were non-significant between the two replications (P ≤ 0.05), data was
pooled (■) for statistical analysis. Flowering rate was calculated by taking a reciprocal of the
number of days to flower. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Tmin (base temperature) is the estimated minimum temperature at or
2
2
below which the rate of progress towards flowering is zero. r or R is the coefficient of
determination generated for linear and non-linear crop models, respectively. In panel A, the
petunia cultivar ‘Bravo Blue’ grown during replication 1 (●) was substituted by ‘Single Frost
Blue in replication 2 (○). Flowering time data from the first replicate did not show a significant
trend (P ≤ 0.05) in impatiens and could not be used for statistical analysis. Tmax and Topt in
panel R was fixed at 35 ºC and 25.1 ºC respectively to allow for precise estimation of the other
model parameters using 50 observations from the second replicate.
81

American marigold 'Inca II Mix'

40
20

B

NS

16

Branch number

D

Shoot dry weight (g)

60

0

14
12
10
8
0

C

30

Height (cm)

16

y = 0.309x2 - 17.182x + 248.130
y = -2.766x + 78.390

A

Inflorescence diameter (cm)

Inflorescence number

80

y = -0.039x2 + 2.286x - 6.877

y = -0.065x2 + 2.578x - 15.197

E

NS

14
12
10
8
6
0
10
8
6
4
2
0
12

14

16

18

20

22

24

26

28

27
24
21
18
15
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.3. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and inflorescence diameter (E) in American
marigold ‘Inca II Mix’ for replication 1 (●) and replication 2 (○). When regression slopes were
non-significant between the two replications (P ≤ 0.05), data was pooled (■) for statistical
analysis. Each symbol represents the treatment means, and error bars represent 95% confidence
intervals. NS = nonsignificant at P ≤ 0.05. Dashed and solid lines in panel A represent regression
equations for replication 1 and 2, respectively. The mean daily light integral for replication 1 and
−2 −1
2 was 18 and 9 mol∙m ∙d , respectively.

82

Cupflower 'Purple Robe'
3.5
A

y = -5.005x + 150.770

80
60
40
20
0

Branch number

40

y = 0.539x2 - 23.537x + 276.980
y = -0.502x + 28.339

35
30
25
20
15
10
0

y = -0.185x + 5.225

E

y = -0.064x + 4.177

2.5
2.0
1.5
1.0
0.5

3.6
3.2
2.8
2.4
2.0
0.0

C

26

D

3.0

0.0

B

Flower diameter (cm)

Flower number

100

Shoot dry weight (g)

120

y = 0.075x2 - 2.752x + 42.492

12

14

16

18

20

22

24

26

28

Height (cm)

24
22
20
18
16
14
12
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.4. The effect of mean daily temperature on flower number (A), axillary branch number
(B), height (C), shoot dry weight (D), and flower diameter (E) in cupflower ‘Purple Robe’ for
replication 1 (●) and replication 2 (○). When regression slopes were non-significant between the
two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. Each symbol represents
the treatment means, and error bars represent 95% confidence intervals. Dashed and solid lines in
panel B represent regression equations for replication 1 and 2, respectively. The mean daily light
−2 −1
integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.

83

Diascia 'Diamonte Mix'
20

A

y = -2.115x + 80.795

80
60
40
20

Shoot dry weight (g)

Inflorescence number

100

0

25
20
15
10
5

Stem length (cm)

0

NS

E

y = -0.0054x2 + 0.193x + 0.290

15
10
5
0

y = 0.080x2 - 4.125x + 64.520
NS

B

Flower diameter (cm)

Branch number

30

D

2.4
2.2
2.0
1.8
1.6
1.4
0.0

C

12

NS

14

16

18

20

22

24

26

28

40
30
20
10
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.5. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), stem length (C), shoot dry weight (D), and flower diameter (E) in diascia ‘Diamonte
Mix’ for replication 1 (●) and replication 2 (○). When regression slopes were non-significant
between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. Each symbol
represents the treatment means, and error bars represent 95% confidence intervals. NS =
nonsignificant at P ≤ 0.05. Dashed lines in panel B represent regression equations for replication
−2 −1
1. The mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.

84

Flowering tobacco 'Perfume Deep Purple'
16
A

y= -0.115x2 + 2.485x + 28.043
y = 0.052x2 - 2.626x + 50.779

50

Shoot dry weight (g)

Inflorescence number

60

40
30
20
10
0

Root dry weight (g)

18
16
14
12
10

y = -0.095x2 + 2.579x + 27.248

50

Height (cm)

10
8
6
4

3.0

40
30
20
0

E

y = -0.130x + 4.323

2.5
2.0
1.5
1.0
0.5
0.0

C

Flower diameter (cm)

Branch number

y = -0.236x + 20.694

y = 0.020x2 - 1.344x + 26.923

D

12

0
B

20

0

14

6.5

F

y = -0.0097x2 + 0.343x + 2.284

6.0
5.5
5.0
4.5
4.0
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.6. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
flowering tobacco ‘Perfume Deep Purple’ for replication 1 (●) and replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was pooled
(■) for statistical analysis. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Dashed and solid lines in panel A represent regression equations for
replication 1 and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and
−2 −1
9 mol∙m ∙d , respectively.

85

Geranium 'Pinto Red'
40
y = -0.169x + 7.238
y = 0.032x2 - 1.657x + 25.404

10
8
6
4
2
0
14

Branch number

y = -0.055x2 + 1.534x + 2.468

B

12
10
8
6
4
0
32

y = 0.094x2 - 5.594x + 92.976

E

y = -0.0077x2 + 0.191x + 4.368

30
25
20
15
10
5
0
6.0
5.6
5.2
4.8
4.4
4.0
3.6
0.0

C

y = -1.034x + 46.245
NS

30

Height (cm)

D

35

Shoot dry weight (g)

A

Flower diameter (cm)

Inflorescence number

12

12

14

16

18

20

22

24

26

28

28
26
24
22
20
18
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.7. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and flower diameter (E) in geranium ‘Pinto Red’
for replication 1 (●) and replication 2 (○). When regression slopes were non-significant between
the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. Each symbol
represents the treatment means, and error bars represent 95% confidence intervals. Dashed and
solid lines in panel A and C represent regression equations for each parameter in replication 1
and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and 9
−2 −1
mol∙m ∙d , respectively.

86

Geranium 'Ringo 2000 Deep Red'
30

12

y = 0.028x2 - 1.393x + 20.538
y = 0.031x2 - 1.781x + 28.764

10
8
6
4
2
0

Height (cm)

Root dry weight (g)

Branch number

10
8
6

30

20
15
10
5

y = -0.319x + 16.161

12

y = -0.725x + 35.352
y = -0.030x2 + 0.792x + 19.607

27
24
21
18
15
0

E

y = -0.099x + 3.048

2.0
1.5
1.0
0.5
0.0

C

y = -1.442x + 43.004

25

0

B

14

0

D

Shoot dry weight (g)

A

Flower diameter (cm)

Inflorescence number

14

6.0

F

y = -0.0085x2 + 0.231x + 4.063

5.6
5.2
4.8
4.4
4.0
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.8. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in
geranium ‘Ringo 2000 Deep Red’ for replication 1 (●) and replication 2 (○). When regression
slopes were non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines in panel A and C represent regression equations for
each parameter in replication 1 and 2, respectively. The mean daily light integral for replication 1
−2 −1
and 2 was 18 and 9 mol∙m ∙d , respectively.

87

Gerbera 'Jaguar Deep Orange' or 'Jaguar Deep Rose'
y= -0.160x + 6.795
y = -0.493x + 18.041

14
12
10
8
6
4
2
30
0
B

y = -0.261x + 19.818

Height (cm)

25
20
15
10
5
0
12

14

16

18

20

22

24

26

Shoot dry weight (g)

16

25

A

Inflorescence diameter (cm)

Inflorescence number

18

28

C

NS

20
15
10
5
0
D

NS

12
10
8
6
4
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.9. The effect of mean daily temperature on inflorescence number (A), height (B), shoot
dry weight (C), and inflorescence diameter (D) in gerbera ‘Jaguar Deep Orange’ for replication 1
(●) or ‘Jaguar Deep Rose’ for replication 2 (○). When regression slopes were non-significant
between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. NS =
nonsignificant at P ≤ 0.05. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Dashed and solid lines in panel A represent regression equations for
replication 1 and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and
−2 −1
9 mol∙m ∙d , respectively.

88

Globe amaranth 'Gnome Purple'
16
A

y = -0.141x2 + 5.421x - 44.794

8
6
4

C

y = -0.060x2 + 3.052x - 27.540

12
10
8
6
4

2
0

0

B

y = -0.100x2 + 4.042x - 28.161

16

Branch number

14

Height (cm)

10

Shoot dry weight (g)

Inflorescence number

12

14
12
10
8
0

D

y = 0.309x - 4.314

4
3
2
1
0

16

18

20

22

24

26

28

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.10. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), and shoot dry weight (D) in globe amaranth ‘Gnome Purple’ for
replication 1 (●) and replication 2 (○). When regression slopes were non-significant between the
two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. All plants died at 14 °C,
and therefore data could not be collected. Each symbol represents the treatment means, and error
bars represent 95% confidence intervals. The mean daily light integral for replication 1 and 2
−2 −1
was 18 and 9 mol∙m ∙d , respectively.

89

Heliotrope 'Blue Wonder'

18
15
12
9
6
3
0
30

Branch number

y = -0.076x2 + 2.224x - 3.158
y = 0.094x2 - 4.327x + 55.425

NS
y= 0.073x2 - 2.574x + 41.752

28
26
24
20
18

30

D

y = 0.013x2 - 0.846x + 16.442

10
8
6
4
2
E

NS

1.5
1.0

22

0

12

0

B

0.5

16

Height (cm)

Shoot dry weight (g)

21

14

A

Flower diameter (cm)

Inflorescence number

24

0.0

C

12

NS

14

16

18

20

22

24

26

28

25
20
15
10
5
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.11. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and flower diameter (E) in heliotrope ‘Blue
Wonder’ for replication 1 (●) and replication 2 (○).When regression slopes were non-significant
between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. NS =
nonsignificant at P ≤ 0.05. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. Dashed and solid lines in panel A and B represent regression
equations for each parameter in replication 1 and 2, respectively. The mean daily light integral
−2 −1
for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.

90

Impatiens 'Blitz 3000 Deep Orange'
5

A

y = -3.167x2 + 136.540x - 1332.800
y = -2.505x + 124.030

200

Shoot dry weight (g)

Flower number

240

160
120
80
40

10
9
8
7
6
5
0

Height (cm)

12

NS

4
3
2
1
0

E

NS

F

y = 0.0096x2 - 0.466x + 9.940

1.5
1.0
0.5
0.0

C

y = -0.085x2 + 3.817x - 33.112
NS

10
8
6
4
2

Flower diameter (cm)

Branch number

11

y = -0.068x2 + 3.032x - 25.507
NS

B

Root dry weight (g)

0

D

0

6.0
5.5
5.0
4.5
4.0
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.12. The effect of mean daily temperature on flower number (A), axillary branch number
(B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in impatiens ‘Blitz
3000 Deep Orange’ for replication 1 (●) and replication 2 (○). NS = nonsignificant at P ≤ 0.05.
Each symbol represents the treatment means, and error bars represent 95% confidence intervals.
Dashed and solid lines in panel A, B and C represent regression equations for each parameter in
replication 1 and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and
−2 −1
9 mol∙m ∙d , respectively.

91

Nemesia 'Poetry White'
A

y = -0.290x2 + 7.816x + 7.809
y = 0.204x2 - 10.084x + 150.650

80
60
40
20

3.0
D

y = -0.204x + 11.979
y = 0.028x2 - 1.100x + 16.169

10
8
6
4
0

F

y = 0.0043x2 - 0.228x + 4.023

2.0
1.5
1.0
0.5

0.5
0.4
0.3
0.2
0.1
0.0

C

y = -0.236x + 21.569

21
18
15
12
9

Flower diameter (cm)

Branch number

12

Height (cm)

y = 0.0026x2 - 0.139x + 1.945

2.5

0.0
B

Root dry weight (g)

0

y = 0.011x2 - 0.561x + 7.627

E

Shoot dry weight (g)

Inflorescence number

100

0

2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.13. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in nemesia
‘Poetry White’ for replication 1 (●) and replication 2 (○).When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis.
Each symbol represents the treatment means, and error bars represent 95% confidence intervals.
Dashed and solid lines in panel A and B represent regression equations for each parameter in
replication 1 and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and
−2 −1
9 mol∙m ∙d , respectively.

92

New Guinea impatiens 'Divine Cherry Red'

Flower number

120

7

A

y = -0.995x2 + 38.226x - 293.340

Shoot dry weight (g)

140

100
80
60
40
20

Branch number

14

NS

12
10
8
6
4
0

y = -0.049x2 + 1.882x - 14.723

E

y = -0.029x2 + 1.081x - 4.236

4
3
2
1

6.5
6.0
5.5
5.0
4.5
4.0
3.5
0.0

y = -0.041x2 + 1.606x - 8.087

C
12

Height (cm)

D

5

0

B

Flower diameter (cm)

0

6

12

14

16

18

20

22

24

26

28

10
8
6
4
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.14. The effect of mean daily temperature on flower number (A), axillary branch number
(B), height (C), shoot dry weight (D), and flower diameter (E) in New Guinea impatiens ‘Divine
Cherry Red’ for replication 1 (●) and replication 2 (○). When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis.
NS = nonsignificant at P ≤ 0.05. Each symbol represents the treatment means, and error bars
represent 95% confidence intervals. The mean daily light integral for replication 1 and 2 was 18
−2 −1
and 9 mol∙m ∙d , respectively.

93

Osteospermum 'Asti Purple'
14

50
40
30
20
10
0

B

35

Branch number

y = -1.890x + 59.473
y = 0.383x2 - 17.626x + 210.840

NS

30
25
20
15
10
0

y = 0.069x2 - 2.985x + 52.310

C
28

Height (cm)

D

Shoot dry weight (g)

A

Inflorescence diameter (cm)

Inflorescence number

60

y = -0.328x + 13.423

E

y = -0.031x2 + 1.157x - 4.990

12
10
8
6
4
0

6.5
6.0
5.5
5.0
4.5
0.0
12

14

16

18

20

22

24

24
20
16
12
0
12

14

16

18

20

22

24

Mean daily temperature (°C)

Figure 2.15. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot dry weight (D), and inflorescence diameter (E) in osteospermum
‘Asti Purple’ for replication 1 (●) and replication 2 (○). When regression slopes were nonsignificant between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis.
NS = nonsignificant at P ≤ 0.05. Each symbol represents the treatment means, and error bars
represent 95% confidence intervals. Dashed and solid lines in panel A represent regression
equations for replication 1 and 2, respectively. The mean daily light integral for replication 1 and
−2 −1
2 was 18 and 9 mol∙m ∙d , respectively.

94

Petunia 'Bravo Blue' or 'Single Frost Blue'
60

Shoot dry weight (g)

Flower number

50

12

y = -0.123x2 + 6.383x - 47.859
y = -0.309x2 + 10.869x - 66.640

A

40
30
20
10

Branch number

22

y = 0.203x + 13.082
NS

20
18
16
14
12

20

Stem length (cm)

6
4
2

y = 0.409x + 1.087
NS

16
12
8
4
0

E

y = -0.053x + 1.504

1.5
1.0
0.5
0.0

C

Flower diameter (cm)

0

y = -0.031x2 + 0.936x - 2.040

8

0

B

Root dry weight (g)

0

D

10

8.5

F

y = -0.160x + 10.235

8.0
7.5
7.0
6.5
6.0
5.5
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.16. The effect of mean daily temperature on flower number (A), axillary branch number
(B), stem length (C), shoot and root dry weight (D and E), and flower diameter (F) in petunia
‘Bravo Blue’ or ‘Single Frost Blue’ for replication 1 (●) and replication 2 (○). Each symbol
represents the treatment means, and error bars represent 95% confidence intervals. NS =
nonsignificant at P ≤ 0.05. Dashed and solid lines in panel A, B and C represent regression
equations for each parameter in replication 1 and 2, respectively. The mean daily light integral
−2 −1
for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.

95

Pot marigold 'Bon Bon Orange'

35
30
25
20
15
10

20

y = -0.492x + 24.282
y = 0.020x2 - 0.991x + 24.825

16
14
12
10

25

Height (cm)

y = 0.048x2 - 2.419x + 33.222

E

y = 0.0079x2 - 0.389x + 5.064

F

y = -0.324x + 11.875

8
6
4
2

1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0

C

20
15
10

y = 0.045x2 - 2.034x + 35.497
y = 0.144x2 - 5.465x + 66.695

5
0
12

D

10

0

B

18

0

Shoot dry weight (g)

y = -2.031x + 70.868

40

0

Branch number

12

A

Root dry weight (g)

45

14

16

18

20

22

24

26

Inflorescence diameter (cm)

Inflorescence number

50

28

8
6
4
2
0

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.17. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and inflorescence diameter (E) in
pot marigold ‘Bon Bon Orange’ for replication 1 (●) and replication 2 (○). When regression
slopes were non-significant between the two replications (P ≤ 0.05), data was pooled (■) for
statistical analysis. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines in panel B and C represent regression equations for
each parameter in replication 1 and 2, respectively. The mean daily light integral for replication 1
−2 −1
and 2 was 18 and 9 mol∙m ∙d , respectively.

96

Snapdragon 'Liberty Classic Cherry'
10

y = -0.196x2 + 7.349x - 52.023
y = 0.037x2 - 2.028x + 28.312

20

Shoot dry weight (g)

25

A

15
10
5
0
50
45
40
35
30
25
20
15
10

y = 0.172x2 - 8.971x + 135.130
y = 0.092x2 - 4.994x + 83.907

D

y = 0.028x2 - 1.648x + 24.765

E

y = 0.0079x2 - 0.425x + 5.886

8
6
4
2
0

B

1.6

Root dry weight (g)

Branch number

Inflorescence number

30

1.4
1.2
1.0
0.8
0.6
0.4
0.2

0

0.0

C

50

y = -1.085x + 58.133

12

14

16

18

20

22

24

26

28

Height (cm)

45
40
35
30
25
20
0
12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.18. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), and shoot and root dry weight (D and E) in snapdragon ‘Liberty Classic
Cherry’ for replication 1 (●) and replication 2 (○). Each symbol represents treatment means, and
error bars represent the standard error with 95% confidence levels. Dashed and solid lines in
panel A and B represent regression equations for each parameter in replication 1 and 2,
−2 −1
respectively. The mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d ,
respectively.

97

Stock 'Hot Cakes Purple' or 'Garden Vintage Mix'

16
12
8
4
y = -0.074x2 + 2.635x - 8.963
NS

18
16
14
12
10
8
0

4
3
2
1
E

y = -0.109x2 + 3.379x - 2.320
NS

25
20
15
10
0

y = 0.0033x2 - 0.164x + 2.090

0.5
0.4
0.3
0.2
0.1
0.0

C

30

Height (cm)

y = -0.181x + 5.638

5

0

B

Root dry weight (g)

20

D

Shoot dry weight (g)

20

0

Branch number

y = 0.039x2 - 2.351x + 42.739

A

24

Flower diameter (cm)

Inflorescence number

28

4.5

F

y = -0.086x + 5.202

4.0
3.5
3.0
2.5
2.0
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.19. The effect of mean daily temperature on inflorescence number (A), axillary branch
number (B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in stock
‘Hot Cakes Purple’ for replication 1 (●) or ‘Garden Vintage Mix’ for replication 2 (○). When
regression slopes were non-significant between the two replications (P ≤ 0.05), data was pooled
(■) for statistical analysis. Each symbol represents the treatment means, and error bars represent
95% confidence intervals. NS = nonsignificant at P ≤ 0.05. Dashed and solid lines in panel B and
C represent regression equations for each parameter in replication 1 and 2, respectively. The
−2 −1
mean daily light integral for replication 1 and 2 was 18 and 9 mol∙m ∙d , respectively.

98

Torenia 'Clown Blue'
8
y = -19.157x + 503.320
y = -3.990x2 + 153.090x - 1218.200

300
200
100

Branch number

20

y = 0.275x + 9.286

18
16
14
12
10
8
0

Height (cm)

21

7

NS
y = -0.084x2 + 4.055x - 31.874

18
15
12
9
6
0

y = -0.056x2 + 2.444x - 21.891

E

y = -0.0074x2 + 0.333x - 3.135

F

y = -0.010x2 + 0.355x - 0.236

5
4
3
2
1

1.0
0.8
0.6
0.4
0.2
0.0

C

D

6

0

B

Root dry weight (g)

0

Flower diameter (cm)

Flower number

A

Shoot dry weight (g)

400

3.2
3.0
2.8
2.6
2.4
2.2
2.0
1.8
0.0

12

14

16

18

20

22

24

26

28

12

14

16

18

20

22

24

26

28

Mean daily temperature (°C)

Figure 2.20. The effect of mean daily temperature on flower number (A), axillary branch number
(B), height (C), shoot and root dry weight (D and E), and flower diameter (F) in torenia ‘Clown
Blue’ for replication 1 (●) and replication 2 (○). When regression slopes were non-significant
between the two replications (P ≤ 0.05), data was pooled (■) for statistical analysis. Each symbol
represents the treatment means, and error bars represent 95% confidence intervals. Dashed and
solid lines in panel A and C represent regression equations for each parameter in replication 1
and 2, respectively. The mean daily light integral for replication 1 and 2 was 18 and 9
−2 −1
mol∙m ∙d , respectively.

99

LITERATURE CITED

100

Literature Cited
Adams, S.R., P. Hadley, and S. Pearson. 1998a. The effects of temperature, photoperiod, and
photosynthetic photon flux on the time to flowering of petunia 'Express Blush Pink'. J.
Amer. Soc. Hort. Sci. 123:577–580.
Adams, S.R., S. Pearson, and P. Hadley. 1997. An analysis of the effects of temperature and light
integral on the vegetative growth of pansy cv. Universal Violet (Viola ×wittrockiana
Gams.). Ann. Bot. 79:219–225.
Adams, S.R., S. Pearson, and P. Hadley. 1998b. An appraisal of the use of reciprocal transfer
experiments: Assessing the stages of photoperiod sensitivity in chrysanthemum cv.
Snowdon (Chrysanthemum ×morifolium Ramat.). J. Exp. Bot. 49:1405-1411.
Adams, S.R., S. Pearson, and P. Hadley. 1998c. The effect of temperature on inflorescence
initiation and subsequent development in chrysanthemum cv. Snowdon (Chrysanthemum
×morifolium Ramat.). Scientia Hort. 77:59–72.
Adams, S.R., S. Pearson, P. Hadley, and W.M. Patefield. 1999. The effects of temperature and
light integral on the phases of photoperiod sensitivity in Petunia ×hybrida. Ann. Bot.
83:263–269.
Armitage, A.M. 1995. Photoperiod, irradiance, and temperature influence flowering of Hamelia
patens (Texas firebush). HortScience 30:255–256.
Armitage, A.M., W.H. Carlson, and J.A. Flore. 1981. The effect of temperature and quantum
flux density on the morphology, physiology, and flowering of hybrid geraniums. J. Amer.
Soc. Hort. Sci. 106:643–647.
Armitage, A.M., A.M. Miller, and J.M. Garner. 2000. The influence of photoperiod and
temperature on growth and flowering of Angelonia angustifolia Benth. - A preliminary
report. Acta Hort. 541:167–170.
Bartok, J.W., Jr. 2001. Energy conservation for commercial greenhouses. Natural Resource, Agr.
Eng. Serv. Coop. Ext., Ithaca, NY.
Blanchard, M. 2009. Manipulating light and temperature for energy-efficient greenhouse
production of ornamental crops. Ph.D. Dissertation, Michigan State Univ., East Lansing.
Blanchard, M.G. and E.S. Runkle. 2011. Quantifying the thermal flowering rates of eighteen
species of annual bedding plants. Scientia Hort. 128:30–37.
Blanchard, M.G., E.S. Runkle, and P.R. Fisher. 2011a. Modelling plant morphology and
development of petunia in response to temperature and photosynthetic daily light integral.
Scientia Hort. 129:313–320.

101

Blanchard, M.G., E.S. Runkle, and J.M. Frantz. 2011b. Energy-efficient greenhouse production
of Petunia and Tagetes by manipulation of temperature and photosynthetic daily light
integral. Acta Hort. 893:857–864.
Brøndum, J.J. and R.D. Heins. 1993. Modeling temperature and photoperiod effects on growth
and development of dahlia. J. Amer. Soc. Hort. Sci. 118:36–42.
Brooks, A. and G.D. Farquhar. 1985. Effect of temperature on the CO2/O2 specificity of
ribulose-1,5-bisphosphate carboxylase/oxygenase and the rate of respiration in the light.
Estimates from gas-exchange measurements on spinach. Planta 165:397–406.
Carvalho, S.M.P., S.E. Wuillai, and E. Heuvelink. 2006. Combined effects of light and
temperature on product quality of Kalanchoe blossfeldiana. Acta Hort. 711:121–126.
Clough, E.A., A. Cameron, R.D. Heins, and W.H. Carlson. 2001. Growth and development of
Oenothera fruticosa is influenced by vernalization duration, photoperiod, forcing
temperature, and plant growth regulators. J. Amer. Soc. Hort. Sci. 126:269–274.
Currey, C.J. and J.E. Erwin. 2011. Photosynthetic daily light integral impacts growth and
flowering of several Kalanchoe species. HortTechnology 21:98–102.
Dinar, M. and J. Rudich. 1985. Effect of heat stress on assimilate partitioning in tomato. Ann.
Bot. 56:239–248.
Erwin, J.E. and R.D. Heins. 1990. Temperature effects on lily development rate and morphology
from the visible bud stage until anthesis. J. Amer. Soc. Hort. Sci. 115:644–646.
Faust, J.E. and R.D. Heins. 1993. Modeling leaf development of the African violet (Saintpaulia
ionantha Wendl.). J. Amer. Soc. Hort. Sci. 118:747–751.
Faust, J.E. and R.D. Heins. 1994. Modeling inflorescence development of the African violet
(Saintpaulia ionantha Wendl.). J. Amer. Soc. Hort. Sci. 119:727–734.
Faust, J.E. and R.D. Heins. 1998. Modelling shoot-tip temperature in the greenhouse
environment. J. Amer. Soc. Hort. Sci. 123:208–214.
Faust, J.E., V. Holcombe, N.C. Rajapakse, and D.R. Layne. 2005. The effect of daily light
integral on bedding plant growth and flowering. HortScience 40:645–649.
Hidén, C. and R.U. Larsen. 1994. Predicting flower development in greenhouse grown
chrysanthemum. Scientia Hort. 58:123–138.
Kaczperski, M.P., W.H. Carlson, and M.G. Karlsson. 1991. Growth and development of Petunia
×hybrida as a function of temperature and irradiance. J. Amer. Soc. Hort. Sci. 116:232–
237.

102

Kanellos, E.A.G. and S. Pearson. 2000. Environmental regulation of flowering and growth of
Cosmos atrosanguineus (Hook.) Voss. Scientia Hort. 83:265–274.
Karlsson, M. and J. Werner. 2001. Temperature affects leaf unfolding rate and flowering of
cyclamen. HortScience 36:292–294.
Karlsson, M.G.and R.D. Heins. 1992. Chrysanthemum dry matter partitioning patterns along
irradiance and temperature gradients. Can. J. Plant Sci. 72:307–316.
Karlsson, M.G., R.D. Heins, J.E. Erwin, R.D. Berghage, W.H. Carlson, and J.A. Biernbaum.
1989. Temperature and photosynthetic photon flux influence chrysanthemum shoot
development and flower initiation under short-day conditions. J. Amer. Soc. Hort. Sci.
114:158–163.
Ku, S.B. and G.E. Edwards. 1977. Oxygen inhibition of photosynthesis. 1. Temperature
dependence and relation to O2/CO2 solubility ratio. Plant Physiol. 59:986–990.
Landsberg, J.J. 1977. Some useful equations for biological studies. Exp. Agric. 13:273–286.
Larsen, R.U. 1988. A prediction model of the development rate of the pot plant Senecio
×hybridus Hyl. A case study. Acta Hort. 230:381–388.
Larsen, R.U. 1989. A prediction model for floral development of Senecio ×hybridus Hyl. 'Molls
Stam'. Acta Hort. 248:247–254.
Larsen, R.U. and L. Persson. 1999. Modelling flower development in greenhouse
chrysanthemum cultivars in relation to temperature and response group. Sci. Hort. 80:73–
89.
Mattson, N.S. and J.E. Erwin. 2003. Temperature affects flower initiation and development rate
of Impatiens, Petunia, and Viola. Acta Hort. 624:191–197.
Mattson, N.S. and J.E. Erwin. 2005. The impact of photoperiod and irradiance on flowering of
several herbaceous ornamentals. Scientia Hort. 104:275–292.
Miller, A. and A.M. Armitage. 2002. Temperature, irradiance, photoperiod, and growth
retardants influence greenhouse production of Angelonia angustifolia Benth. Angel Mist
series. HortScience 37:319–321.
Moccaldi, L.A. and E.S. Runkle. 2007. Modeling the effects of temperature and photosynthetic
daily light integral on growth and flowering of Salvia splendens and Tagetes patula. J.
Amer. Soc. Hort. Sci. 132:283–288.
Moe, R. and L.M. Mortensen. 1992. Thermomorphogenesis in pot plants. Acta Hort. 305:19–25.

103

Niu, G., R.D. Heins, A. Cameron, and W.H. Carlson. 2001. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of Campanula
carpatica 'Blue Clips'. Scientia Hort. 87:93–105.
Niu, G., R.D. Heins, A.C. Cameron, and W.H. Carlson. 2000. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of pansy (Viola
×wittrockiana). J. Amer. Soc. Hort. Sci. 125:436–441.
Oh, W., I. Cheon, K. Kim, and E.S. Runkle. 2009. Photosynthetic daily light integral influences
flowering time and crop characteristics of Cyclamen persicum. HortScience 44:341–344.
Park, B.H., N. Oliveira, and S. Pearson. 1998. Temperature affects growth and flowering of the
balloon flower [Platycodon grandiflorus (Jacq.) A. DC. cv. Astra Blue]. HortScience
33:233–236.
Pearson, S., A. Parker, S.R. Adams, P. Hadley, and D.R. May. 1995. The effects of temperature
on the flower size of pansy (Viola ×wittrockiana Gams.). J. Amer. Soc. Hort. Sci.
70:183–190.
Pietsch, G.M., W.H. Carlson, R.D. Heins, and J.E. Faust. 1995. The effect of day and night
temperature and irradiance on development of Catharanthus roseus (L.) 'Grape Cooler'.
J. Amer. Soc. Hort. Sci. 120:877–881.
Pramuk, L.A. and E.S. Runkle. 2005a. Photosynthetic daily light integral during the seedling
stage influences subsequent growth and flowering of Celosia, Impatiens, Salvia, Tagetes,
and Viola. HortScience 40:1336–1339.
Pramuk, L.A. and E.S. Runkle. 2005b. Modeling growth and development of Celosia and
Impatiens in response to temperature and photosynthetic daily light integral. J. Amer.
Soc. Hort. Sci. 130:813–818.
Reed, R.L., E.R. Hamerly, B.E. Dinger, and P.G. Jarvis. 1976. An analytical model for field
measurement of photosynthesis. J. Appl. Ecol. 13:925–942.
Roberts, E.H. and R.J. Summerfield. 1987. Measurement and prediction of flowering in annual
crops, p. 17–50. In: J.G. Atherton (ed.). Manipulation of flowering. Butterworths,
London.
Semeniuk, P. 1975. Effect of temperature on growth flowering and fruiting in blue-troll
Browallia. HortScience 10:480–481.
Strefeler, M.S. 1995. Growth response of 20 seed geranium cultivars to three day-night
temperature regimes. HortScience 30:1465–1466.

104

Vaid, T.M. and E.S. Runkle. 2011. Categorizing nineteen bedding plant varieties into
temperature response groups based on their estimated base temperatures. HortScience
46:S111.
van Iersel, M.W. 2003. Short-term temperature change affects the carbon exchange
characteristics and growth of four common bedding plant species. J. Amer. Soc. Hort.
Sci. 128:100–106.
Wai, K.S. and S.E. Newman. 1992. Air and root-zone temperatures influence growth and
flowering of snapdragons. HortScience 27:796–798.
Warner, R.M. and J.E. Erwin. 2003. Effect of photoperiod and daily light integral on flowering
of five Hibiscus sp. Scientia Hort. 97:341–351.
Warner, R.M. and J.E. Erwin. 2005. Prolonged high temperature exposure and daily light
integral impact growth and flowering of five herbaceous ornamental species. J. Amer.
Soc. Hort. Sci. 130:319–325.
Warner, R.M. and J.E. Erwin. 2006. Prolonged high temperature exposure differentially reduces
growth and flowering of 12 Viola × wittrockiana Gams. cvs. Scientia Hort. 108:295–302.
White, J.W. and I.J. Warrington. 1988. Temperature and light integral effects on growth and
flowering of hybrid geraniums. J. Amer. Soc. Hort. Sci. 113:354–359.
Whitman, C.M., R.D. Heins, A. Cameron, and W.H. Carlson. 1997. Cold treatment and forcing
temperature influences flowering of Campanula carpatica 'Blue Clips'. HortScience
32:861–865.
Yuan, M., W.H. Carlson, R.D. Heins, and A. Cameron. 1998. Effect of forcing temperature on
time to flower for Coreopsis grandiflora, Gaillardia ×grandiflora, Leucanthemum
×superbum, and Rudbeckia fulgida. HortScience 33: 663–667.

105

SECTION III

AVERAGE DAILY TEMPERATURE AND TRANSPLANT SIZE AFFECT
FLOWERING TIME, QUALITY, AND ESTIMATED PROFITABILITY IN FIVE
SPECIES OF ANNUAL BEDDING PLANTS

106

Average Daily Temperature and Transplant Size Affect Flowering time, Quality, and
Estimated Profitability in Five Species of Annual Bedding Plants

1

Tasneem M. Vaid and Erik S. Runkle

2

Department of Horticulture, Michigan State University, East Lansing, MI 48824

Received for publication__________. Accepted for publication__________. We gratefully
acknowledge funding by Michigan’s plant agriculture initiative at Michigan State University
(Project GREEEN), the Michigan Agricultural Experiment Station, the USDA-ARS Floriculture
and Nursery Research Initiative, and greenhouse growers providing support for Michigan State
University floriculture research. We thank C. Raker & Sons for donation of plant material, Mike
Olrich for his greenhouse assistance, Dr. Paul Fisher (University of Florida) for his advice on
economic analysis.

1
2

M.S. Candidate
Corresponsing author and to whom reprint requests should be addressed. Email address:

runkleer@msu.edu.

107

Abstract
Volatile fuel costs and declining profit margins have made it necessary for growers,
especially in temperate climates, to improve their production efficiency. A shorter finish
production time can potentially reduce the heating, labor and overhead costs. However, shorter
crop timing is sometimes offset by reduced crop quality. We grew American marigold (Tagetes
erecta L.), geranium (Pelargonium × hortorum L.H. Bailey), gerbera (Gerbera jamesonii Bolus
ex Hook. f.), osteospermum [Osteospermum ecklonis (D.C.) Norl.], and snapdragon
(Antirrhinum majus L.) in glass-glazed greenhouse compartments to quantify how temperature
and transplant size influence flowering time and quality parameters. This information was then
used to estimate the net profit per pot and per square meter week for each species. Seedlings
from two different transplant size trays (128- versus 288-cell size or 36- versus 128-cell size)
were transplanted in 10-cm pots and grown at constant temperature setpoints of 17 or 23 ºC and
under a 16-h photoperiod provided by high-pressure sodium lamps. Days to flower from
transplant, inflorescence number, inflorescence diameter, branch number, number of nodes
below the first open flower, and plant height were measured at first flowering. Flowering time
decreased as average daily temperature increased in all five species. American marigold,
geranium and snapdragon flowered earlier when grown from a larger compared to a smaller
transplant, whereas transplant size did not influence flowering time in gerbera and
osteospermum. Inflorescence number and diameter was greater at 17 than 23 ºC in three and four
species, respectively, whereas transplant size had little or no influence on these parameters.
Estimated net profit per pot was greater at 23 ºC and with the smaller transplant size in all five
species. Therefore, transplanting a larger transplant or lowering the greenhouse temperature to
save on heating costs is not necessarily a profitable strategy.

108

Introduction
In the United States, bedding and garden plants are the largest contributors (48%) to the
total value of production of all reported floriculture crops, with a 15-state wholesale value of
$1.90 billion in 2011 (USDA, 2012). In temperate regions, bedding plants scheduled for spring
sales are typically grown beginning in late winter, when high energy inputs are required to
maintain greenhouse temperature setpoints. The cost of heating greenhouses in these climates,
not surprisingly, accounts for a major share of the indirect cost of greenhouse production
(Bartok, 2001). A general assumption is that lowering the greenhouse temperature during the
winter can reduce the heating bills, which may not always be true, since the crop production
period is lengthened at lower temperatures (Blanchard et al., 2011b).
A shorter crop production cycle in a greenhouse can reduce heating, labor and overhead
expenses on a per-crop basis, and can potentially increase profitability by enabling more
production cycles in a season. Although fast cropping is an appealing concept, it can result in
reduced plant quality, since plants photosynthesize for a shorter duration before flowering. Poor
ornamental plant quality characteristics, such as tall and weak stems, few branches, few and
small flowers, and reduced biomass, can lower consumer appeal and potentially bring a lower
price. Crop production time and in some cases, plant quality, can depend on the species and
cultivar (Blanchard, 2009), starting plant size and characteristics (Fisher, 2006), desired finish
size, average daily temperature (ADT) (Blanchard and Runkle, 2011), photosynthetic daily light
integral (DLI) (Blanchard, 2009; Moccaldi and Runkle, 2007; Pramuk and Runkle, 2005b), and
photoperiod (Adams et al., 1997, 1998a). Assuming a plant is grown under an inductive
photoperiod, temperature is the primary environmental factor that controls plant development
rate and thus, the period of time in which plants can harvest light before flowering.

109

The number of days (n) to reach a particular developmental event as a function of
temperature can be converted to a rate by taking a reciprocal of that number of days (1/n).
Average daily temperature influences flowering time by affecting flower initiation, development,
or both. Many studies have quantified the influence of temperature on flowering time, and
reported a decrease in days to flower (DTF) with an increase in ADT, when ADT is between the
species-specific Tmin (base temperature at or below which the flower development rate is zero)
and Topt (optimum temperature at which the flower development rate is maximum) (Adams et
al., 1998b, 1998c, 1999; Blanchard and Runkle, 2011; Clough et al., 2001; Kanellos and Pearson,
2000; Miller and Armitage, 2002; Moccaldi and Runkle, 2007; Niu et al., 2000, 2001; Pramuk
and Runkle, 2005b). For example, an increase in ADT from 14 to 24 °C decreased flowering
time by 47 d in angelonia (Angelonia angustifolia Benth.), 18 d in dianthus (Dianthus chinensis
L.), and 24 d in verbena (Verbena × hybrida Groenl. & Ruempl.) (Blanchard and Runkle, 2011).
A decrease in DTF with an increase in ADT is usually from faster flower development rate
(Brøndum and Heins, 1993; Faust and Heins, 1994; Karlsson and Werner, 2001; Pietsch et al.,
1995; Whitman et al., 1997; Yuan et al., 1998), although some plants initiate flowers at a lower
node count at higher temperatures (Adams et al., 1998b, 1998c; Mattson and Erwin, 2003).
When plants are grown at an ADT above Topt, the flower development rate (reciprocal of DTF)
begins to decline until it reaches zero at the species-specific maximum temperature (Tmax)
(Blanchard and Runkle, 2011; Brøndum and Heins, 1993; Faust and Heins, 1994; Semeniuk,
1975).
In addition to plant development, temperature also influences plant quality parameters
including flower number and size, branch number, and plant height. Several studies have
reported a decrease in quality parameters of ornamental crops with an increase in ADT when

110

grown at the same or similar average DLI (Blanchard et al., 2011a, 2011b; Mattson and Erwin,
2003; Moccaldi and Runkle, 2007; Niu et al., 2000, 2001; Pramuk and Runkle, 2005b). For
example, as ADT increased from 16 to 26 °C inflorescence number decreased by 80% in
largeflower tickseed (Coreopsis grandiflora Hogg ex Sweet), 55% in Shasta daisy
[Leucanthemum ×superbum (Bergmans ex J.W. Ingram)] and 75% in black-eyed Susan
(Rudbeckia fulgida Aiton) (Yuan et al., 1998). Similarly, each degree increase in ADT linearly
decreased flower diameter by 0.97 mm in tussock bellflower (Campanula carpatica Jacq) ‘Blue
–2 –1

Clips’ (under 10.8 mol·m ·d ) (Niu et al., 2001), 1.42 cm in dahlia (Dahlia pinnata Cav.)
‘Royal Dahlietta Yellow’ (Brøndum and Heins, 1993), and branch number by 0.3 in petunia
(Petunia × hybrida Vilm.-Andr.) ‘Dreams Neon Rose’ (Mattson and Erwin, 2003).
Many studies have investigated the influence of two or more environmental factors and
their interaction on flowering time and quality of annual bedding plants (Adams et al., 1997,
1998a; Moccaldi and Runkle 2007; Pramuk and Runkle, 2005b). For example, Moccaldi and
Runkle (2007) and Pramuk and Runkle (2005b) developed crop models to predict the effect of
ADT and DLI on DTF and plant quality parameters such as flower number and diameter, shoot
mass, and plant height. The effect of environmental factors such as ADT, DLI, and photoperiod
on flowering time and plant quality parameters has been reported for numerous bedding plants,
but there is little published information on the influence of transplant size and temperature on
these parameters.
Bedding plants are commercially sold in different transplant sizes that commonly range
from 512-cell (3.1-mL) to 36-cell (42.3-mL) trays, depending on the species and grower
preferences. Practical experience and some initial research data (Fisher, 2006) indicate that a
larger, more mature plant flowers earlier than a smaller, less-developed plant. A shorter crop

111

production period from transplant has many potential advantages: lower labor, heating and
overhead costs; delayed transplanting in the winter, which can reduce heating costs if outdoor
temperatures later in the year are higher; and increased number of production cycles in a season.
However, faster flowering from use of a larger transplant is at least partly offset by the higher
price of the larger transplant (Fisher, 2006). In addition, the influence of starting size of plants on
plant quality parameters has not been published to our knowledge in annual bedding plants when
grown under the same environmental conditions. Therefore, research-based information was
needed to quantify the differences in flowering time and plant quality, as well as further improve
the reliability of scheduling bedding plants. The objectives of this research were 1) to study the
effect of ADT and transplant size on flowering time and plant quality of five species of popular
bedding plants, and 2) to estimate the potential change in net profit per pot and per square meter
week in growing a larger versus a smaller transplant at two temperature treatments.

Materials and Methods
Plant material. Experimental protocol, data collection and data analysis were similar to
that reported by Blanchard and Runkle (2011) and Pramuk and Runkle (2005b). Seedlings of
American marigold (Tagetes erecta L. ‘Inca II Mix’), geranium (Pelargonium ×hortorum
L.H.Bailey ‘Ringo 2000 Deep Red’), gerbera (Gerbera jamesonii Bolus ex Hook. f. ‘Jaguar
Deep Orange’), osteospermum [Osteospermum ecklonis (D.C.) Norl. ‘Asti Purple’], and
snapdragon (Antirrhinum majus L. ‘Rocket Mix’) were grown in two different tray sizes [128cell (12 mL) and 288-cell (6 mL) or 36-cell (42 mL) and 128-cell], depending on species (Table
3.1), by a commercial greenhouse (C. Raker & Sons, Litchfield, MI) and received at Michigan
State University (MSU) on 26 October 2010. Production was scheduled by the commercial

112

producer such that the seedlings were ready for transplant when received at MSU. On receipt the
plants were immediately transplanted and grown in greenhouses maintained at two different
temperature set points.
The experiment was repeated in spring 2011 to generate data under different DLI
conditions. Seedlings in this replicate were received at MSU on 3 March 2011 and were grown
in a glass-glazed greenhouse at a constant temperature setpoint of 20 °C under the natural
photoperiod (lat. 43 °N) until the leaf number for each species and for each transplant size was
similar to that of the transplants grown previously. Once the desired leaf count was achieved
(Table 3.1), the seedlings were transplanted and grown in separate glass-glazed greenhouse
compartments maintained at different temperature set points.
Greenhouse environments. Seedlings were transplanted into 10-cm round containers
(480-mL) filled with a peat-based medium (Suremix, Michigan Grower Products, Galesburg,
MI). The seedlings were thinned to one plant per cell prior to transplant and 10 plants of each
species were grown at constant temperature set points of 17 or 23 °C in separate glass-glazed
greenhouse compartments. Plants were grown under a 16-h photoperiod (0600 to 2200 HR)
created by using the natural photoperiod (lat. 43 °N) and day-extension lighting from high–

pressure sodium (HPS) lamps from 0600 to 2200 HR that provided a PPF of 80 to 100 µmol·m
2 –1

·s . In each greenhouse compartment, a shielded and aspirated 0.13 mm type E thermocouple

(Omega Engineering, Stamford, CT) recorded the air temperature and a line quantum sensor
containing 10 photodiodes (Apogee Instruments) placed at canopy height (22 cm above bench
height) measured the PPF. A CR10 datalogger (Campbell Scientific, Logan, UT) collected the
environmental data every 10 s and hourly averages were recorded. Actual average air
temperatures during the fall and spring experiments were 17.0 or 23.1 and 18.6 or 22.7 ºC,

113

–

respectively. The mean DLIs during the fall and spring experiments were 10.0 and 15.0 mol·m
2 –1

·d , respectively. Average DIF values (day – night temperature) ranged from 0.3 to 0.7 ºC and

0.7 to 2.3 ºC in the fall and spring experiments, respectively. Vapor pressure deficit was
maintained between 0.7 and 1.2 kPa by steam injection at night. Plants were irrigated as
necessary with reverse osmosis water supplemented with a water-soluble fertilizer containing
–1

(mg·L ) 125 N, 12 P, 100 K, 65 Ca, 12 Mg, 1.0 Fe and Cu, 0.5 Mn and Zn, 0.3 B, and 0.1 Mo
(MSU RO Special; GreenCare Fertilizers, Inc., Kankakee, IL).
Data collection and analysis. When each plant flowered according to the individual
characteristics for each species (Table 3.1), the date of flowering was recorded and DTF from
transplant was calculated for each species. Inflorescence number, inflorescence diameter,
axillary branch number, number of nodes below the first open flower on the primary stem, and
plant height (from media surface to tallest inflorescence) were also recorded at first open flower.
Branch number and inflorescence diameter were not recorded in gerbera and snapdragon,
respectively. The experiment was set up as a split-plot design; two temperature treatments
(whole plot factor) were randomly assigned to the greenhouse compartments, and two transplant
size (sub-plot factor) treatments were randomly assigned to the experimental units (plants) within
each temperature treatment. SAS (SAS Institute, Cary, N.C.) was used to analyze the
experimental data. Mean separation using Tukey’s honestly significant difference test at P ≤ 0.05
was performed on all the data to compare differences between treatments. Data were pooled
when there were no significant replicate and treatment interactions.
Cost estimation. Costs were estimated for a finish market date of April 1 using the
procedure described by Fisher et al. (2006). Production time was calculated based on the average
days to flower for each treatment, and assumed a 16-h photoperiod and an average DLI of ≥10

114

−2 −1

mol∙m ∙d . Greenhouse heating costs were estimated for Grand Rapids, MI using Virtual
Grower software version 2.51 (Frantz et al., 2010; USDA−ARS, 2009). Greenhouse
characteristics used for heating cost estimations were the same as those used by Blanchard et al.
(2011b): eight spans 34.1 × 7.3 m, arched 3.7-m roof, 2.7-m gutter, polyethylene double layer
roof, polycarbonate bi-wall ends and sides, forced air unit heaters burning natural gas at US$
−1

0.36∙m˗3, 50 % heater efficiency, no energy curtain, an air infiltration rate of 1.0 h , and
constant temperature setpoint of 17 and 23 ºC. Price of individual transplant sizes was obtained
from a commercial producer (C. Raker & Sons), although prices could vary depending on
volume, shipping and handling, and distributors. Each 10-cm pot was assumed to occupy 0.023
2

m of space, which is a typical spacing used by commercial growers. A greenhouse space use
efficiency of 80% was assumed, and labor and overhead expenses were estimated from Fisher
˗2

˗1

(2006). The estimated labor and overhead cost of $3.25∙m ∙week was calculated by updating
˗2

˗1

the original estimate of $2.90∙m ∙week from Fisher (2006) using the Consumer Price Index
increase of 12% between 2006 and 2011 (USDL, 2012). This figure was further updated to per
2

m per week by dividing $3.25 by the space use efficiency (0.80) to give the final labor and
overhead (not including heating and lighting) costs of $4.06. To estimate the net profit per pot
and per square meter week, the wholesale sales price for each 10-cm pot was assumed to be
$1.25 for American marigold, geranium, and snapdragon (USDA, 2012), and $1.50 for gerbera
and osteospermum [the starting material was generally more expensive (Table 3.2), and these
plants often have a greater commercial value]. Net profit per pot was calculated by subtracting
the total cost from the wholesale price, and net profit per square meter week was calculated by
dividing the net profit per square meter (number of pots per square meter × net profit per pot) by
the production time in weeks. Lighting and other variable costs (pots, media, and tags), the

115

potential change in net profit from additional crop turns or differences in plant quality (and thus,
sales price) were not considered.

Results
American marigold. Flowering time decreased by 7 or 8 d for larger (128-cell) compared
to smaller (288-cell) transplants; and by 12 or 13 d when plants were grown at 23 versus 17 ºC
(Figure 3.1A). Inflorescence number and diameter at flowering were similar between transplant
sizes, but were 68 to 70 % and 21 to 32 % greater at 17 than 23 ºC, respectively. Temperature
and transplant size interacted to influence plant height, branch number, and leaf number at first
flowering. Transplant size had no influence on plant height, and number of branches and leaves
at 17 ºC, but smaller transplants grown at 23 ºC were taller, had greater branching, and
developed more leaves than larger transplants (Figure 3.1C, D). Plant height at flowering was
greater when grown at 23 ºC, irrespective of transplant size. ADT did not affect branch number
and leaf number below first open flower in larger transplants, but smaller transplants formed
10% and 24% fewer branches and leaves, respectively when grown at 17 than 23 ºC.
Geranium. Temperature had a greater influence on flowering time than transplant size
(Figure 3.2A). Plants grown from a 128-cell tray (with 6 leaves) flowered 1 to 5 d earlier
compared to plants grown from a 288-cell tray (with 4 leaves). On the other hand, plants grown
at 23 °C flowered 17 to 20 d earlier than those grown at 17 °C. Inflorescence number was similar
between transplant sizes when grown at 23 ºC, but larger transplants developed significantly
more inflorescences than smaller transplants when grown at 17 ºC (Figure 3.2B). Similarly, plant
height was similar between transplant sizes when grown at 17 ºC, but 128-cell transplants were
10% taller than 288-cell transplants when grown at 23 ºC (Figure 3.2C). Both transplant sizes

116

had 45 to 47% fewer inflorescences and were 10 to 16% shorter at flowering when grown at 23
compared to 17 ºC. Transplant size did not significantly influence inflorescence diameter,
whereas plants grown at 23 ºC had 7 to 9% smaller inflorescences than those grown at 17 ºC
(Figure 3.2E). Temperature and transplant size did not influence leaf number below the first open
flower (Figure 3.2F). Branch number was not significantly affected by transplant size in the first
replication, but larger transplants had greater branching in the second replication (Figure 3.2D).
Similarly, ADT did not significantly influence branch number in smaller transplants, whereas
lower compared to higher temperature promoted greater branching in larger transplants in the
first replication.
Gerbera. Flowering time was 10 or 11 d shorter at 23 versus 17 ºC, but was not
influenced by transplant size (Figure 3.3A). Temperature and transplant size did not influence
inflorescence number and leaf number below the first open inflorescence (Figure 3.3B, E),
whereas these parameters interacted to influence plant height at first flowering (Figure 3.3C).
Plant height was greater at 23 °C for larger (36-cell) transplants, but was similar when starting
from smaller (128-cell) transplants. Inflorescence diameter was greater in plants grown at 17°C
than 128-cell transplants grown at 23 °C (Figure 3.3D).
Osteospermum. Transplant size had no effect on flowering time, but the higher
temperature accelerated flowering time by 8 and 11 d in smaller (128-cell) and larger (36-cell)
transplants, respectively (Figure 3.4A). Temperature and transplant size did not affect
inflorescence number and branch number at first flowering (Figure 3.4B, D). Larger transplants
grown at 23 ºC were 20% shorter that those grown at 17 ºC (Figure 3.4C). In contrast, ADT did
not influence plant height at flowering in smaller transplants. Inflorescence diameter was similar
between transplant sizes, but plants grown at 17 ºC had 11 to 23% larger inflorescences than

117

those grown at 23 ºC (Figure 3.4E). Plants from 36-cell transplants and grown at 23 °C
developed more leaves before flowering than either transplant size grown at 17 °C (Figure 3.4F).
Snapdragon. The higher temperature and larger (128-cell) transplant size accelerated
flowering time compared to lower temperature and smaller (288-cell) transplant size (Figure
3.5A). Plants grown at 23 ºC flowered 13 or 14 d earlier than those grown at 17 ºC, whereas the
larger transplants flowered 7 or 8 d earlier than the smaller transplants. Inflorescence number
was variable among all treatments, but 288-cell transplants grown at 17 °C developed more
inflorescences than either transplant size grown at 23 °C (Figure 3.5B). Plant height was similar
between transplant sizes at 17 ºC, but smaller transplants were significantly taller than larger
transplants at 23 ºC (Figure 3.5C). Plants grown at 17 ºC had 13 to 26% more branching than
those grown at 23 ºC (Figure 3.5D). At 17 ºC, larger transplants formed significantly fewer
branches than the smaller transplants, whereas branch number was similar between transplant
sizes grown at 23 ºC. Larger transplants developed significantly fewer nodes before flowering
compared to smaller transplants, irrespective of temperature (Figure 3.5E). ADT did not
influence leaf number in larger transplants, whereas smaller transplants developed more leaves
when grown at 17 compared to 23 ºC.
Economic analysis. On a per-pot basis, the estimated net profit was usually higher (or the
net loss was lower) when starting with smaller transplants and growing at the higher temperature
(Table 3.2). For example, in American marigold, maximum estimated net profit per pot was
162% greater when smaller transplants were grown at 23 ºC than when larger transplants were
grown at 17 ºC. In gerbera and osteospermum, larger transplants resulted in a net loss per pot
regardless of the growing temperature. In geranium, a net loss per pot was estimated for both
transplant sizes when grown at 17 versus 23 ºC. Interestingly, in American marigold and

118

snapdragon, the estimated net profit per square meter week was similar between transplant sizes
at 23 ºC, but was 8 to 12% lower for larger compared to smaller transplants grown at 17 ºC.

Discussion
Among the five species studied, transplanting a larger versus a smaller transplant
decreased flowering time in American marigold, geranium (only at 17 ºC), and snapdragon. In
these species, the initial leaf number at transplant and transplant production period (time from
seed sow until plants were ready for transplanting) were positively correlated, whereas there was
a negative correlation between flowering time and initial leaf number (P < 0.0001). For example,
in snapdragon, larger transplants flowered 7- or 8-d earlier, but had a 5- or 6-d longer transplant
production period compared to smaller transplants. Therefore, the total time from seed sow to
flowering was similar between transplant sizes in these species. Although the total production
period was not affected by transplant size, transplanting a larger compared to a smaller transplant
may be more cost-effective, since production costs per transplant are relatively low for young
plants given their high density. Consistent with these results, Fisher (2006) reported a 2- to 3week delay in flowering time when transplanting a smaller transplant (84- versus 50-cell tray).
The magnitude of difference in flowering time between the two transplant sizes reported by
Fisher (2006) was greater than that observed in our study, possibly due to a greater difference in
development between the two transplant sizes (initial leaf number not reported). Interestingly, in
gerbera and osteospermum, initial leaf number was similar between transplant sizes (128-cell
versus 36-cell) and therefore, transplant size did not influence DTF in these species.
All plants grown at 23 ºC flowered faster than those grown at 17 ºC. This increase in
temperature had a greater effect on flowering time than increasing the transplant size. For many

119

floriculture crops, an increase in ADT linearly increased the flowering rate, when ADT was
above the species-specific Tmin and at, or below, Topt (Adams et al., 1997; Clough et al, 2001;
Whitman et al., 1997; Yuan et al., 1998). For example, the rate of progress towards flowering
increased linearly as temperature increased from 15 to 26 °C for largeflower tickseed,
blanketflower (Gaillardia ×grandiflora Van Houtte), Shasta daisy, and black-eyed Susan (Yuan
et al., 1998). An increase in flowering time from a decrease in ADT may be due to a
developmental delay in flowering when flowering is induced at a lower node (Adams et al.,
1998b, 1998c; Mattson and Erwin, 2003), but typically it is from a slower rate of flower
development (Blanchard and Runkle, 2011; Brøndum and Heins, 1993; Faust and Heins, 1994;
Karlsson and Werner, 2001; Pietsch et al., 1995).
In our study, the lower temperature (17 ºC) developmentally delayed flowering time in
snapdragon only in the smaller transplants. In contrast, osteospermum transplants grown at 17 ºC
formed fewer nodes before flowering than those grown at 23 ºC, although flowering occurred
earlier at 23 ºC. A similar response was observed in American marigold, but only in the smaller
transplants. Consistent with our results, Blanchard (2009) reported a decrease in leaf number
below the first open inflorescence with a decrease in ADT in African marigold ‘Antigua
Primrose’. In some species, an increase in leaf number from an increase in ADT could be
attributed to an increase in leaf development rate, and not delayed flower initiation (Adams et al.,
1998c). In addition, larger transplants formed significantly fewer leaves than smaller transplants
in American marigold (only at 23 ºC) and snapdragon. This and previous results suggest that
larger transplants of American marigold and snapdragon could have already initiated flowers.
Transplant size had little or no effect on inflorescence number and diameter in this study.
On the other hand, inflorescence number and diameter were significantly greater at 17 ºC in 3

120

and 4 species, respectively. These responses to ADT are consistent with previous studies on
bedding plants (Blanchard et al., 2011a, 2011b; Mattson and Erwin, 2003; Moccaldi and Runkle,
2007; Niu et al., 2000, 2001). For example, celosia (Celosia argentea L.), impatiens (Impatiens
walleriana Hook. f.), and petunia ‘Wave Purple’ and ‘Easy Wave Coral Reef’ plants grown at 17
–2 –1

versus 23 ºC and under a DLI of 10.0 mol·m ·d

had 7%, 63%, and 8 to 18% greater

inflorescences, respectively (Blanchard and Runkle, 2011; Pramuk and Runkle, 2005a).
Similarly, flower size was 33 to 53% greater in pansy (Viola ×wittrockiana Gams.) when plants
were grown at 17 ºC compared to 23 ºC and under a range of DLI conditions (Niu et al., 2000).
An increase in flower number and diameter at lower temperatures could be attributed to the
longer time that plants harvested light before flowering (Pramuk and Runkle, 2005a; Oh et al.,
2010), higher net photosynthetic rate (van Iersel; 2003), greater import of photosynthates into the
developing buds (Dinar and Rudich, 1985) and/ or an increase in leaf number (and thus branch
number) below the first open flower (Mattson and Erwin, 2003; Warner and Erwin, 2006).
Smaller transplants had significantly greater branching only in American marigold grown
at 23 ºC and snapdragon at 17 ºC. Therefore, in most instances, transplant size had no effect on
lateral branching. Branch number was greater in plants grown at 17 ºC than 23 ºC in geranium
(only in the first replication) and snapdragon, whereas American marigold transplants formed
comparatively fewer branches when grown at 17 ºC. Many researchers have reported an increase
in branch number with a decrease in ADT in several species, including balloon flower
[Platycodon grandiflorus (Jacq.) A. DC.] ‘Astra Blue’ (Park et al., 1998), fuchsia (Fuchsia ×
hybrida hort. ex Siebold & Voss) (Erwin et al., 1991), pansy (Mattson and Erwin, 2003; Warner
and Erwin, 2006) and petunia (Mattson and Erwin, 2003). In some cases, at higher temperatures,
a decrease in branch number was correlated with a decrease in leaf number below the first open

121

flower (Mattson and Erwin, 2003). Among the species we studied, a decrease in branch number
was positively correlated (P < 0.001) with a decrease in leaf number below the first open
inflorescence in American marigold and snapdragon. In addition, a shorter production period in
larger transplants and plants grown at 23 ºC could potentially reduce plant quality in these
species.
In American marigold and snapdragon, smaller transplants were significantly taller than
larger transplants when grown at 23 ºC. Plant height of geranium, snapdragon, and
osteospermum (larger transplants only) at flowering was greater at 17 ºC compared to 23 ºC,
whereas an opposite occurred in gerbera (larger transplants only) and American marigold. ADT
and DIF (day – night temperature) can influence plant height by affecting leaf number below the
first flower and internode length, respectively (Clough et al., 2001; Erwin et. al., 1991; Erwin
and Heins, 1990; Mattson and Erwin, 2003). Higher temperatures developmentally accelerate
flowering in some species (Mattson and Erwin, 2003), which can potentially decrease plant
height (Clough et al., 2001) due to fewer nodes formed prior to flowering. Among the five
species we studied, a decrease in plant height was positively correlated with a decrease in leaf
number below the first open inflorescence in American marigold and snapdragon. Similarly, as
DIF increased from –6.0 to +6.0 °C, plant height increased by 39% in fuchsia (Erwin et al.,
1991), 19% in geranium (Strefeler, 1995), and 9% in Italian bellflower (Campanula isophylla
Moretti) (Moe and Mortensen, 1992b). In our study, although the greenhouse temperature
setpoints were constant, the day temperatures were sometimes higher than the night temperatures
and the magnitude was generally greater at 17 than at 23 °C. The mean DIF values were
generally small in this experiment (0.3 to 2.3 °C), but occasional exposure to high daytime
temperatures, especially during the spring replication, could have contributed to an increase in

122

plant height at lower temperatures (Whitman et al., 1997). The effects of transplant size on
flowering characteristics cannot be compared to previous work, since to our knowledge this has
not been reported in annual bedding plants.
The net profit was greater, or the net loss per pot was less, when plants were grown at 23
compared to 17 °C irrespective of transplant size. Heating cost estimations by Blanchard and
Runkle (2009) indicate that in some cases, less energy for heating is consumed on a per-crop
basis by growing at a higher temperature because production time is shorter. For example,
petunia ‘Wave Purple’ flowered 17 d earlier at 23 °C than at 17 °C, and for a greenhouse in
Grand Rapids, MI and a market date of April 1, 22% less energy for heating was predicted by
growing the crop at the higher temperature (Blanchard and Runkle, 2009). In addition, a shorter
production period can reduce labor and overhead expenses and can enable the production of
more crops, assuming they can be marketed.
In addition, growing a larger transplant (128- versus 288-cell or 36- versus 128-cell size)
was either less profitable or resulted in a net loss/ greater net loss per pot and per square meter
week, except for American marigold and snapdragon grown at 23 °C. Although larger
transplants reduced flowering time in some species, the decrease in heating, labor and overhead
costs was generally offset by the higher price of the larger transplant. In contrast to our results,
Fisher et al. (2006) reported that using a larger transplant was more profitable. However, in their
study, the larger transplant (50-cell size) accelerated flowering by 2 to 3 weeks compared to the
smaller one (84-cell size), whereas in our study the difference in flowering time was ≤ 8 d
between transplant sizes. Therefore, although a larger transplant may flower earlier, it may not
always be profitable. This economics depends upon individual grower and transplant
characteristics such as, the magnitude of difference in maturity (leaf number) and flowering time

123

between transplant sizes, plant spacing, demand for greenhouse space, heating labor and
overhead costs, and higher price/shipping costs of the larger transplant. The influence of
transplant size on plant development and quality depends on the magnitude of difference in
development between transplant sizes, and also the environmental conditions during transplant
production (Pramuk and Runkle, 2005a). Further location-specific research is needed to identify
the minimum difference in development between transplant sizes that can justify the higher price
of the larger transplant, since this economics is so situational. In addition, more research is
required on how environmental conditions such as DLI and photoperiod during transplant
production influence flower initiation in different transplant sizes, and therefore subsequent
flowering characteristics.

124

Table 3.1. Time from seed sow to transplant, mean leaf no. at transplant, and characteristics used to determine flowering date for five
bedding plant species in two experimental replicates. Seedlings were grown in 36-, 128-, or 288-cell trays.
Time from seed sow to
Mean leaf no. at
transplant (d)
transplant
288
128
36
288 128
36
Species
Flowering characteristics
1 inflorescence with ≥50% of
z
American marigold 'Inca II Mix'
21 or 22 28 or 29
4
6
‒
‒
the ray flowers reflexed
5 flowers open on an
Geranium 'Ringo 2000 Deep Red' 33 or 35 39 or 46
‒
4
6
‒
inflorescence
1 inflorescence with all of
Gerbera 'Jaguar Deep Orange'
‒
45
45
‒
7
7
the ray flowers reflexed
1 inflorescence with ray
−
Osteospermum 'Asti Purple'
‒
35 or 46 15 or 36
9
9
flowers reflexed
2 flowers open on an
Snapdragon 'Rocket Mix'
33 or 42 39 or 47
‒
6
9
‒
inflorescence
z
Transplant sizes not included in the experiment.

125

Table 3.2. Estimated production costs and net profits when starting with two transplant sizes (36-, 128-, or 288-cell trays) in the
production of five bedding plant species grown at 17 or 23 °C for first flowering on April 1. Heating costs were calculated using
Virtual Grower software for Grand Rapids, MI. Labor and overhead costs are estimated from Fisher (2006). See materials and
methods for information on production cost and net profit estimation. Production time is the average days to flower, and assumes a 16−2 −1
h photoperiod and a mean DLI of ≥10 mol∙m ∙d .
Average
temperature Production
(°C)
time (d)
Transplant
size:
17
23
Transplant
size:
17
23
Transplant
size:
17
23
Transplant
size:
17
23

288

128

57
44

49
37

288

128

70
49

65
48

128

36

57
46

56
46

128

36

51
43

52
41

Heating
288

128

Estimated cost per pot (US$)
Labor and
Transplants
overhead
Total cost
American marigold 'Inca Mix II'
288

128

288

128

288

128

$0.15 $0.12 $0.17 $0.35 $0.76 $0.65 $1.08 $1.12
$0.15 $0.12 $0.17 $0.35 $0.59 $0.49 $0.91 $0.96
Geranium 'Ringo 2000 Deep Red'
288

128

288

128

288

128

288

128

$0.19 $0.18 $0.21 $0.41 $0.93 $0.87 $1.33 $1.46
$0.16 $0.16 $0.21 $0.41 $0.65 $0.64 $1.02 $1.21
Gerbera 'Jaguar Deep Orange'
128

36

128

36

128

36

128

36

$0.15 $0.15 $0.54 $0.83 $0.76 $0.75 $1.45 $1.73
$0.15 $0.15 $0.54 $0.83 $0.61 $0.61 $1.30 $1.59
Osteospermum 'Asti Purple'
128

36

128

36

128

36

128

36

$0.13 $0.13 $0.54 $0.83 $0.68 $0.69 $1.35 $1.65
$0.14 $0.13 $0.54 $0.83 $0.57 $0.55 $1.25 $1.51

126

Net profit per pot

Net profit per
square meter
week

288

128

288

128

$0.17
$0.34

$0.13
$0.29

$0.91
$2.35

$0.81
$2.38

288

128

288

128

–$0.08
$0.23

–$0.21
$0.04

–$0.35
$1.43

–$0.98
$0.25

128

36

128

36

$0.05
$0.20

–$0.23
–$0.09

$0.27
$1.32

–$1.47
–$0.60

128

36

128

36

$0.15
$0.25

–$0.15
–$0.01

$0.90
$1.77

–$0.89
–$0.07

Table 3.2. (cont’d).
Average
temperature Production
(°C)
time (d)
Transplant
size:
17
23

288

128

56
42

48
35

Heating
288

128

Estimated cost per pot (US$)
Labor and
Transplants
overhead
Total cost
Snapdragon 'Rocket Mix'
288

128

288

128

288

128

$0.15 $0.12 $0.15 $0.33 $0.75 $0.64 $1.05 $1.09
$0.14 $0.11 $0.15 $0.33 $0.56 $0.47 $0.85 $0.91

127

Net profit per pot

Net profit per
square meter
week

288

128

288

128

$0.20
$0.40

$0.16
$0.34

$1.09
$2.90

$1.01
$2.96

American marigold 'Inca II Mix'
70
288
128

A
B

50

C
D

40
30
20

16

0
8

A

A

B

30

B

20

B

5
4
3
2
1
0

C
A

B

C

C

15
10

Leaf number

Height (cm)

B

B

6

0

F

25
20

A
B

B

B

15
10
5

5
0

A

E

7

10

20

A

A

8

Diameter (cm)

Inflorescence number

B

40

25

B

4

50

30

B

12

10
0

D

20

Branch number

Days to flower

60

24

A

17

0

23

17

23

Mean daily temperature (°C)

Figure 3.1. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of American marigold ‘Inca II Mix’ grown at 17 or 23 °C. Data were pooled
between two experimental replications when there were no replicate and treatment interactions
(P ≤ 0.05). Error bars represent the standard errors. Means followed by the same letter are not
statistically different (P ≤ 0.05). Mean DLIs during the fall and spring experiments were 10.0
–2 –1
and 15.0 mol·m ·d , respectively.

128

Geranium 'Ringo 2000 Deep Red'
90
A

288
128

B

60

C

C

45
30

18

Branch number

Days to flower

75

21

A

15

C
C

2

Inflorescence diameter (cm)

B

4

C
AB

bc
D D

0
7

E

6

A

A

5

B

B

A

A

4
3
2
1
0

F

10

A

B

C

20
16
12
8

8

A

A

6
4
2

4
0

BC b

6

Leaf number

Inflorescence number

A

6

24

Height (cm)

B

8

28

CD

9

3

10

0

a

a

12

15
0

D

17

23

0

17

23

Mean daily temperature (°C)

Figure 3.2. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of geranium ‘Ringo 2000 Deep Red’ grown at 17 or 23 °C. Data were pooled
between two experimental replications when there were no replicate and treatment interactions
(P ≤ 0.05). When the interaction was significant, data are separately presented for fall (dots) and
spring (stripes) replications. Error bars represent the standard errors. Means followed by the
same letter are not statistically different (P ≤ 0.05). Mean DLIs during the fall and spring
–2 –1
experiments were 10.0 and 15.0 mol·m ·d , respectively.

129

Days to flower

70

A
A

60

128
36

A

B

B

50
40
30
20
10

Inflorescence number

0

B

12

D
10

8

A

6

A

A
4

A

A

A
B

8

AB

6
4
2
0
80

E

70

Leaf number

80

Inflorescence diameter (cm)

Gerbera 'Jaguar Deep Orange'

60

A
A

A

50

A

40
30
20

2

10
0

0

C

17

23

Height (cm)

25
20

A

AB
B

15

AB

10
5
0

17

23

Mean daily temperature (°C)

Figure 3.3. The effect of starting transplant size (128- or 36-cell trays) on flowering
characteristics of gerbera ‘Jaguar Deep Orange’ grown at 17 or 23 °C. Data were pooled between
two experimental replications when there were no replicate and treatment interactions (P ≤ 0.05).
When the interaction was significant, data are separately presented for fall (dots) and spring
(stripes) replications. Error bars represent the standard errors. Means followed by the same letter
are not statistically different (P ≤ 0.05). Mean DLIs during the fall and spring experiments were
–2 –1
10.0 and 15.0 mol·m ·d , respectively.

130

Osteospermum 'Asti Purple'

Days to flower

60

A
A

A

50

35

128
288
B

B

40
30
20

30

Branch number

70

A

A
A

15
10
5

C

A

15
10

AB

A

0
7

AB

15
10

B

E

6

B

B

5
4
3
2
1
0
30

20

25

F
B

B

AB

A

20
15
10

5
0

A

A

A

20

Inflorescence diameter (cm)

A

Leaf number

Inflorescence number
Height (cm)

B

20

25

A

5

25

0

A

25

10
0

D

5
17

0

23

17

23

Mean daily temperature (°C)

Figure 3.4. The effect of starting transplant size (128- or 36-cell trays) on flowering
characteristics of osteospermum ‘Asti Purple’ grown at 17 or 23 °C. Data were pooled between
two experimental replications when there were no replicate and treatment interactions (P ≤ 0.05).
When the interaction was significant, data are separately presented for fall (dots) and spring
(stripes) replications. Error bars represent the standard errors. Means followed by the same letter
are not statistically different (P ≤ 0.05). Mean DLIs during the fall and spring experiments were
–2 –1
10.0 and 15.0 mol·m ·d , respectively.

131

Snapdragon 'Rocket Mix'

Days to flower

60

A

50
288
128

A
B

50

C
D

40
30
20

Branch number

70

10

B

B

10
5

80

BC

C

20
10

40

15

B

30

0
AB

A

40

A

20

0

Height (cm)

B

Leaf number

Inflorescence number

0

D

E
A

30

C

B
C

20
10

C

0
A

AB

17

23

B
C

60
40
20
0

17

23

Mean daily temperature (°C)

Figure 3.5. The effect of starting transplant size (288- or 128-cell trays) on flowering
characteristics of snapdragon ‘Rocket Mix’ grown at 17 or 23 °C. Data were pooled between two
experimental replications when there were no replicate and treatment interactions (P ≤ 0.05).
When the interaction was significant, data are separately presented for fall (dots) and spring
(stripes) replications. Error bars represent the standard errors. Means followed by the same letter
are not statistically different (P ≤ 0.05). Mean DLIs during the fall and spring experiments were
–2 –1
10.0 and 15.0 mol·m ·d , respectively.

132

LITERATURE CITED

133

Literature cited

Adams, S.R., P. Hadley, and S. Pearson. 1998a. The effects of temperature, photoperiod, and
photosynthetic photon flux on the time to flowering of petunia 'Express Blush Pink'. J.
Amer. Soc. Hort. Sci. 123:577–580.
Adams, S.R., S. Pearson, and P. Hadley. 1997. The effects of temperature, photoperiod and light
integral on the time to flowering of pansy cv Universal Violet (Viola ×wittrockiana
Gams). Ann. Bot. 80:107–112.
Adams, S.R., S. Pearson, and P. Hadley. 1998b. An appraisal of the use of reciprocal transfer
experiments: Assessing the stages of photoperiod sensitivity in chrysanthemum cv.
Snowdon (Chrysanthemum ×morifolium Ramat.). J. Exp. Bot. 49:1405–1411.
Adams, S.R., S. Pearson, and P. Hadley. 1998c. The effect of temperature on inflorescence
initiation and subsequent development in chrysanthemum cv. Snowdon (Chrysanthemum
×morifolium Ramat.). Scientia Hort. 77:59–72.
Adams, S.R., S. Pearson, P. Hadley, and W.M. Patefield. 1999. The effects of temperature and
light integral on the phases of photoperiod sensitivity in Petunia ×hybrida. Ann. Bot.
83:263–269.
Bartok, J.W., Jr. 2001. Energy conservation for commercial greenhouses. Natural Resource, Agr.
and Eng. Serv. Coop. Ext., Ithaca, NY.
Blanchard, M. 2009. Manipulating light and temperature for energy–efficient greenhouse
production of ornamental crops. Ph.D. Dissertation, Michigan State Univ., East Lansing.
Blanchard, M. and E. Runkle. 2009. Energy-efficient annuals part 6: Petunias. Greenhouse
Grower 27:36–41.
Blanchard, M.G. and E.S. Runkle. 2011. Quantifying the thermal flowering rates of eighteen
species of annual bedding plants. Scientia Hort. 128:30–37.
Blanchard, M.G., E.S. Runkle, and P.R. Fisher. 2011a. Modelling plant morphology and
development of petunia in response to temperature and photosynthetic daily light integral.
Scientia Hort. 129:313–320.
Blanchard, M.G., E.S. Runkle, and J.M. Frantz. 2011b. Energy-efficient greenhouse production
of Petunia and Tagetes by manipulation of temperature and photosynthetic daily light
integral. Acta Hort. 893:857–864.
Brøndum, J.J. and R.D. Heins. 1993. Modeling temperature and photoperiod effects on growth
and development of dahlia. J. Amer. Soc. Hort. Sci. 118:36–42.

134

Clough, E.A., A. Cameron, R.D. Heins, and W.H. Carlson. 2001. Growth and development of
Oenothera fruticosa is influenced by vernalization duration, photoperiod, forcing
temperature, and plant growth regulators. J. Amer. Soc. Hort. Sci. 126:269–274.
Dinar, M.and J. Rudich. 1985. Effect of heat stress on assimilate partitioning in tomato. Ann.
Bot. 56:239–248.
Erwin, J.E. and R.D. Heins. 1990. Temperature effects on lily development rate and morphology
from the visible bud stage until anthesis. J. Amer. Soc. Hort. Sci. 115:644–646.
Erwin, J.E., R.D. Heins, and R. Moe. 1991. Temperature and photoperiod effects on Fuchsia
×hybrida morphology. J. Amer. Soc. Hort. Sci. 116:955–960.
Faust, J.E.and R.D. Heins. 1994. Modeling inflorescence development of the African violet
(Saintpaulia ionantha Wendl.). J. Amer. Soc. Hort. Sci. 119:727–734.
Fisher, P., H. Warren, and L. Hydock. 2006. Larger liners, shorter crop time. Greenhouse
Grower 24:8–12.
Frantz, J.M., B. Hand, L. Buckingham, and S. Ghose. 2010. Virtual grower: Software to
calculate heating costs of greenhouse production in the United States. HortTechnology
20:778–785.
Kanellos, E.A.G.and S. Pearson. 2000. Environmental regulation of flowering and growth of
Cosmos atrosanguineus (Hook.) Voss. Scientia Hort. 83:265–274.
Karlsson, M.and J. Werner. 2001. Temperature affects leaf unfolding rate and flowering of
cyclamen. HortScience 36:292–294.
Mattson, N.S.and J.E. Erwin. 2003. Temperature affects flower initiation and development rate
of Impatiens, Petunia, and Viola. Acta Hort. 624:191–197.
Mattson, N.S.and J.E. Erwin. 2005. The impact of photoperiod and irradiance on flowering of
several herbaceous ornamentals. Scientia Hort. 104:275–292.
Miller, A.and A.M. Armitage. 2002. Temperature, irradiance, photoperiod, and growth retardants
influence greenhouse production of Angelonia angustifolia Benth. Angel Mist series.
HortScience 37:319–321.
Moccaldi, L.A.and E.S. Runkle. 2007. Modeling the effects of temperature and photosynthetic
daily light integral on growth and flowering of Salvia splendens and Tagetes patula. J.
Amer. Soc. Hort. Sci. 132:283–288.
Moe, R.and L.M. Mortensen. 1992. Thermomorphogenesis in pot plants. Acta Hort. 305:19–25.

135

Niu, G., R.D. Heins, A. Cameron, and W.H. Carlson. 2001. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of Campanula
carpatica 'Blue Clips'. Scientia Hort. 87:93–105.
Niu, G., R.D. Heins, A.C. Cameron, and W.H. Carlson. 2000. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of pansy (Viola
×wittrockiana). J. Amer. Soc. Hort. Sci. 125:436–441.
Oh, W., E.S. Runkle, and R.M. Warner. 2010. Timing and duration of supplemental lighting
during the seedling stage influence quality and flowering in petunia and pansy.
HortScience 45:1332–1337.
Park, B.H., N. Oliveira, and S. Pearson. 1998. Temperature affects growth and flowering of the
balloon flower [Platycodon grandiflorus (Jacq.) A. DC. cv. Astra Blue]. HortScience
33:233–236.
Pietsch, G.M., W.H. Carlson, R.D. Heins, and J.E. Faust. 1995. The effect of day and night
temperature and irradiance on development of Catharanthus roseus (L.) 'Grape Cooler'.
J. Amer. Soc. Hort. Sci. 120:877–881.
Pramuk, L.A.and E.S. Runkle. 2005a. Photosynthetic daily light integral during the seedling
stage influences subsequent growth and flowering of Celosia, Impatiens, Salvia, Tagetes,
and Viola. HortScience 40:1336–1339.
Pramuk, L.A.and E.S. Runkle. 2005b. Modeling growth and development of Celosia and
Impatiens in response to temperature and photosynthetic daily light integral. J. Amer.
Soc. Hort. Sci. 130:813–818.
Semeniuk, P. 1975. Effect of temperature on growth flowering and fruiting in blue-troll
Browallia. HortScience 10:480–481.
Strefeler, M.S. 1995. Growth response of 20 seed geranium cultivars to three day-night
temperature regimes. HortScience 30:1465–1466.
U.S. Dept. of Agriculture. 2012. Floriculture crops 2011 summary. Agri. Stat. Board.
Washington, D.C.
U.S. Dept. of Labor. 2012. Consumer Price Index. Bureau of Labor Statistics. Washington, D.C.
van Iersel, M.W. 2003. Short-term temperature change affects the carbon exchange
characteristics and growth of four common bedding plant species. J. Amer. Soc. Hort.
Sci. 128:100–106.
Warner, R.M.and J.E. Erwin. 2006. Prolonged high temperature exposure differentially reduces
growth and flowering of 12 Viola ×wittrockiana Gams. cvs. Scientia Hort. 108:295–302.

136

Whitman, C.M., R.D. Heins, A. Cameron, and W.H. Carlson. 1997. Cold treatment and forcing
temperature influences flowering of Campanula carpatica 'Blue Clips'. HortScience
32:861–865.
Yuan, M., W.H. Carlson, R.D. Heins, and A. Cameron. 1998. Effect of forcing temperature on
time to flower for Coreopsis grandiflora, Gaillardia ×grandiflora, Leucanthemum
×superbum, and Rudbeckia fulgida. HortScience 33:663–667.

137

SECTION IV

DETERMINING THE PHOTOPERIOD-SENSITIVE STAGES OF DEVELOPMENT
AND THE EFFECT OF TEMPERATURE AND PHOTOPERIOD ON FLOWERING IN
THREE PETUNIA CULTIVARS

138

Determining the Photoperiod-Sensitive Stages of Development and the Effect of
Temperature and Photoperiod on Flowering in Three Petunia Cultivars

1

Tasneem M. Vaid and Erik S. Runkle

2

Department of Horticulture, Michigan State University, East Lansing, MI 48824

Received for publication__________. Accepted for publication__________. We gratefully
acknowledge funding by Michigan’s plant agriculture initiative at Michigan State University
(Project GREEEN), the Michigan Agricultural Experiment Station, the USDA-ARS Floriculture
and Nursery Research Initiative, and greenhouse growers providing support for Michigan State
University floriculture research. We thank C. Raker & Sons for donation of plant material, Mike
Olrich for his greenhouse assistance, Dr. Paul Fisher (University of Florida) for his advice on
economic analysis.

1
2

M.S. Candidate
Corresponsing author and to whom reprint requests should be addressed. Email address:

runkleer@msu.edu.

139

Abstract
Petunia ×hybrida Vilm.-Andr. is a facultative long day (LD) plant and is typically
commercially produced in temperate climates beginning in late winter, when the natural
photoperiod is short. Greenhouse growers in these climates often use low-intensity, LD lighting
to promote flowering in this species. Limiting LD lighting to periods of photoperiod sensitivity
can promote flowering, while potentially reducing lighting costs and undesirable stem extension.
We performed reciprocal transfer experiments with petunia ‘Easy Wave Neon Rose’, ‘Improved
Wave Purple’ and ‘Classic Wave Purple’, in which plants with 4, 6, 8, 10, or 12 leaves were
moved from inductive LDs to non-inductive short days (SDs), and vice versa, to determine when
these varieties become sensitive to photoperiod. In a separate experiment, the effects of average
daily temperature and photoperiod on the flowering of petunia were quantified. Plants were
grown in glass-glazed greenhouses at 14 or 20 ºC and with a 10- or 16-h photoperiod created by
a combination of a truncated 9-h natural day extended with light from incandescent lamps.
Developmentally, ‘Easy Wave Neon Rose’ became receptive to photoperiod after unfolding ≤4
leaves, whereas ‘Classic Wave Purple’ and ‘Improved Wave Purple’ became photoperiodsensitive after unfolding >8 leaves. Flowering was fastest in plants grown under LDs and at 20
ºC for all petunia varieties. For example, ‘Easy Wave Neon Rose’, ‘Classic Wave Purple’ and
‘Improved Wave Purple’ plants flowered 31, 49, or 59 d earlier, respectively when grown at 20
ºC with LDs compared to 14 ºC with SDs. However, plants that flowered earliest had fewer
flower buds (except in ‘Classic Wave Purple’) and branches. In ‘Easy Wave Neon Rose’ and
‘Improved Wave Purple’, LDs had less of a promotive effect when grown at 14 ºC than at 20 ºC.

140

Introduction
Annual bedding and garden plants that are commercially grown in temperate regions
have a narrow sales period in late spring, which necessitates precise scheduling of flowering in
these species. Flower initiation and development is a complex function of average daily
temperature (ADT), daily light integral (DLI), and for photoperiodic crops, photoperiod (Adams
et al., 1998a; Blanchard et al., 2011; Kacsperski et al., 1991). Several studies have investigated
the influence of two or more of these environmental factors on the flowering characteristics of
annual bedding plants (Adams et al., 1996, 1997, 1998; Moccaldi and Runkle 2007; Pramuk and
Runkle, 2005b). For example, Adams et al. (1997) reported a linear increase in the rate of
progress towards flowering in pansy (Viola ×wittrockiana Gams.) with an increase in ADT, DLI,
and photoperiod up to an optimum. Interestingly, DLI and particularly ADT can influence crop
responses to photoperiod by modifying the length of the juvenile phase (Adams et al., 1999).
Leaves of photoperiodic plants perceive changes in daylength to induce flowering (Thomas and
Vince-Prue, 1997), while ADT mainly influences the plant development rate (Blanchard and
Runkle, 2011).
Photoperiodic floral induction depends upon the length of the juvenile phase and the
minimum number of inductive cycles necessary for floral initiation (Warner, 2009). The juvenile
phase is the early stage of plant development during which flowering can not be induced
regardless of the environmental conditions (Thomas and Vince-Prue, 1997). The length of the
juvenile phase, and the minimum number of inductive cycles required for floral induction in
mature plants, depends upon the species, cultivar, and growing temperature (Adams et al., 1999,
2003; Verheul et al., 2005; Warner, 2009). For example, at 20 ºC, the juvenile phase lasted from
9 to 12 d after emergence in celosia (Celosia argentea L.; Warner, 2009), and 25 to 40 d in

141

snapdragon (Antirrhinum majus L.; Adams et al., 2003) depending on the cultivar. In addition,
there is considerable variation in the minimum number of inductive cycles required for flowering
among species. For example, some short day (SD) plants such as, whiteedge morning-glory
[Pharbitis nil (L.) Roth], and red goosefoot (Chenopodium rubrum L.) can be induced to flower
by a single inductive cycle (Thomas and Vince-Prue, 1997), whereas others such as strawberry
(Frageria L.) require at least 21 SDs for floral induction (Verheul et al., 2005).
Some photoperiodic species exhibit a photoperiod-sensitive phase during development,
which comes after the juvenile phase and before a second photoperiod-insensitive phase during
the later stages of development (Adams et al., 1999; Warner, 2009). For example, in celosia and
petunia (Petunia ×hybrida Vilm.-Andr.) ‘Express Blush Pink’ grown at ≈ 20 ºC, the photoperiod
sensitive phase was ≈ 9 to 45 d and 16 to 56 d after emergence, respectively (Adams et al., 1999;
Warner, 2009). The phases of photoperiod sensitivity can be determined by reciprocal transfer
experiments, in which plants are transferred from long days (LDs) to SDs, and vice versa, at
different stages in plant maturity (Adams et al., 1999, 2003; Munir et al., 2010). This technique
has been used in several field and ornamental crops, such as maize (Zea mays L.; Kiniry et al.,
1983), barley (Hordeum vulgare L.; Roberts et al., 1988), soybean [Glycine max (L.) Merr.;
Wilkerson et al., 1989], opium poppy (Papaver somniferum L.; Wang et al., 1997), celosia
(Warner, 2009), chrysanthemum (Dendranthema ×grandiflorum Kitam.; Adams et al., 1998),
petunia (Adams et al., 1999), and snapdragon (Adams et al., 2003; Munir et al., 2010).
Many researchers have analyzed the plastochron index (leaf number before first open
flower) data when flowering was terminal, in addition to flowering time data, to separate the
effects of photoperiod on flower induction and flower development (Adams et al., 1998b, 2003).
Although photoperiod primarily influences flower induction, in some species, it also affects

142

flower bud development rate. For example, snapdragon ‘Bells’ transferred from LDs to SDs
during early flower development flowered 4 d later than those grown under continuous LDs
(Adams et al., 2003). Similarly, transferring chrysanthemum plants from SDs to LDs during
early flower development delayed flowering and resulted in distorted, abnormal flowers (Adams
et al., 1998b). Interestingly, some species with visible flower buds, such as garden balsam
(Impatiens balsamina L.), soybean, and perilla (Perilla L.), revert to vegetative growth when
transferred from inductive to non-inductive conditions (Han et al., 2006; Lam and Leopold,
1961; Pouteau et al., 1997). This reversion is primarily due to a lack of meristem commitment
and a rapidly diminishing floral signal in previously induced leaves. This is evident from a
positive correlation between the amount of induction and the timing of reversion (Pouteau et al.,
1997), and also from a decrease in the number of flowers initiated under a non-inductive
photoperiod (Lam and Leopold, 1961).
In addition to the flowering time, plant quality parameters (e.g., flower bud number,
branch number or shoot mass) can be influenced by the timing and duration (number) of
inductive photoperiods in some species (Damann and Lyons, 1993; Imamura et al., 1966;
Warner, 2009). For example, increasing the duration of SD exposure from 5 to 30 d when celosia
plants had developed only one true leaf, increased the inflorescence number by 175% (Warner,
2009). A similar trend occurred in celosia plants with 2 or 3 leaves, whereas inflorescence
number in plants with 4 or 5 leaves was not influenced by the duration of SDs. In contrast,
flower number, branch number and shoot mass decreased as the duration of SDs increased from
1 to 12 d in celosia plants exposed to SD before expansion of the first leaf pair. Knowledge of
the effects of timing and duration of photoperiod on flowering characteristics of photoperiodic
species can allow commercial growers to balance crop timing with acceptable plant quality.

143

Petunia ×hybrida, derived from a cross between P. axillaris and P. integrifolia
(Stehmann et al., 2009) is among the top ten bedding plants produced in the United States, with a
reported 15-state wholesale value of $132 million in 2011 (USDA, 2012). There are over 360
petunia cultivars available in the United States (Kelly et al., 2007), most of which are sold in
flats, hanging baskets, or pots (USDA, 2012). Several studies have categorized petunia as a
facultative long day (LD) plant (Adams et al., 1998a; Piringer and Cathey, 1960; Warner, 2010),
since plants flowered earlier when grown under LDs compared to short days (SDs). For example,
Adams et al. (1998) reported a linear increase in rate of progress towards flowering (reciprocal of
days to flower, DTF) of petunia ‘Express Blush Pink’ with an increase in photoperiod from 8 to
–1

14.4 h·d . Similarly, all four petunia species evaluated by Warner (2010) flowered 8 to 21 d
earlier, and developed 14 to 22 fewer leaves prior to flowering, when grown under LDs versus
SDs. In temperate climates, petunias are typically grown beginning in late winter, when the
ambient photoperiods are short, which necessitates the use of LD lighting to promote flowering
in this species. Limiting LD lighting to periods of photoperiod sensitivity can promote flowering,
while potentially reducing lighting costs and undesirable stem extension (Adams et al., 2001).
However, plant quality is often positively correlated with flowering time, and sometimes with
the leaf number below first open flower (Blanchard and Runkle, 2011; Pramuk and Runkle,
2005a, Warner, 2009). Previous studies investigating photoperiod sensitivity in petunia ‘Express
Blush Pink’ indicates that this variety has a distinct photoperiod-senstive phase. However, some
petunia ‘Wave’ cultivars can reportedly revert to vegetative growth when grown under noninductive conditions during early flower development (S. Padhye, pers. comm.).
In addition to photoperiod, temperature also influences the flowering time in petunia,
such that as ADT increases within a species-specific range, rate of progress towards flowering

144

linearly increases (Adams et al., 1998a; Blanchard et al., 2011). The flowering rate of petunia
‘Easy Wave Coral Reef’ and ‘Wave Purple’ increased by 0.003 and 0.002, respectively with
each degree rise in temperature from 14 to 26 ºC (Blanchard et al., 2011). Some researchers have
studied the interaction effects of temperature and photoperiod on photoperiod receptivity, Topt
for flowering, and flowering characteristics in a few petunia varieties (Adams et al., 1996, 1998a,
1999). In some petunia varieties ADT can influence seedling sensitivity to photoperiod by
modifying the length of the juvenile phase; an increase in ADT from 13.7 to 22.3 ºC decreased
the juvenile phase of petunia ‘Express Blush Pink’ from 21 to 13 d (Adams et al., 1999).
Similarly, an increase in photoperiod from 8 h·d

–1

–1

to 14.4 h·d

linearly increased Topt for

flowering in petunia ‘Express Blush Pink’ from 20.7 ºC to 23.4 ºC (Adams et al., 1998a). Adams
et al. (1996) reported a temperature and photoperiod interaction on flowering time and
morphogical characteristics in trailing petunias ‘Malve’ and ‘White’. The effect of photoperiod
on flowering time of these trailing petunia varieties was greater when plants were grown at a
higher versus a lower temperature. Similarly, the effect of temperature on the rate of progress
towards flowering was greater when plants were grown under LDs than SDs. However, this
response to temperature and photoperiod may vary between petunia varieties, since Adams et al.
(1998a) observed little or no interaction in the effects of temperature and photoperiod on the
flowering characteristics in petunia ‘Express Blush Pink’. Therefore, the objectives of this study
were: 1) to determine the precise timing and minimum duration of inductive LDs to achieve
100% flowering in petunia, and 2) to study the effect of temperature and photoperiod interaction
on petunia flowering characteristics.

145

Materials and Methods
Growth chamber treatments and culture. Experimental protocol, data collection and
analysis were similar to that reported by Blanchard and Runkle (2011). Seeds of petunia ‘Easy
Wave Neon Rose’, ‘Classic Wave Purple’, and ‘Improved Wave Purple’ were sown in 36-cell
(42.3-mL) plug trays on 27 December 2011 by a commercial greenhouse (C. Raker & Sons,
Litchfield, MI) and received at Michigan State University (MSU) on 4 January 2012. On receipt,
the seedlings were unfolding their first leaf and were thinned to one plant per cell. They were
then grown in a controlled environment chamber at a constant temperature setpoint of 20 ºC ± 2
ºC under a 16-h long day (LD; 0700 to 2300 HR) at a photosynthetic photon flux (PPF) of 140
–2 –1

µmol·m ·s

, or for a 10-h short day (SD; 0700 to 1700 HR) at a PPF of 220 µmol·m–2·s–1

provided by a combination of cool-white fluorescent (CWF; F96T12CWVHO; Philips,
Somerset, NJ) and incandescent lamps (INC, Philips, Somerset, NJ). The different light
intensities were used so that the daily light integral (DLI) for both photoperiodic treatments was
–2 –1

≈ 8 mol·m ·d

. The light intensity and plant temperature were checked periodically with an

instantaneous quantum sensor (Apogee Instruments Inc., Logan, UT) at canopy height and a
hand-held infrared thermometer (Tel-Fast 2, Tel-Tru, Rochester, NY), respectively. The desired
light intensities were maintained by replacing and/or lowering the lamps when required. The
temperature setpoint was lowered by 1 to 2 ºC in the environment chamber with the 10-h
photoperiod when plant temperature was > 20 ºC. Plugs were irrigated by hand as necessary with
acidified well water (140 mg·L

–1

–1

titratable alkalinity of CaCO3) containing (mg·L ) 95, 34, and
–1

29 Ca, Mg, and S, and supplemented with a water-soluble fertilizer providing (mg·L ) 62 N, 6
P, 62 K, 7 Ca, 0.5 Fe, 0.3 Cu, Mn, and Zn, 0.1 B and Mo (MSU Well Water Special; GreenCare
Fertilizers Inc., Kankakee, IL).

146

Expt 1. Reciprocal photoperiodic transfers at different stages of development on
flowering. Sixteen plants of each cultivar were randomly selected and transferred from LD to
SD, and vice versa, when they averaged 4, 6, 8, and 10 leaves, which occurred 13 to 26 d after
the start of photoperiodic treatments (Table 4.1). A leaf was included when it had uncurled and
was at least half expanded. In addition, 16 plants per cultivar were grown under continuous LD
or SD until they attained 12 leaves. When plants in all the treatments attained an average of 12
leaves (24 to 30 d after start of treatments), they were transplanted into 10-cm round containers
(480-mL) filled with a peat-based medium (Suremix, Michigan Grower Products, Galesburg,
MI). Ten plants per treatment were selected for uniformity and grown at a constant set point of
20 °C with a 10-h photoperiod in a glass-glazed greenhouse. The actual average air temperature
–2 –1

and DLI in the greenhouse during the experiment were 21.1 ºC and 10.4 mol·m ·d ,
respectively. Treatments were completely randomized between and within greenhouse benches.
The experiment was arranged as a split-plot design; two photoperiods (whole plot factor) were
randomly assigned to the controlled environment chambers, and five leaf number (sub-plot
factor) treatments were randomly assigned to the experimental units (plants) within each
photoperiod.
Expt 2. Temperature and photoperiod interaction on flowering. Fifty plants per cultivar
were grown in a controlled environment chamber at 20 ºC under LDs as previously described.
When the plants attained an average of 8 leaves, forty plants were selected for uniformity and
transplanted into 10-cm round pots (480-mL) and grown at constant temperature set points of 14
or 20 °C in separate glass-glazed greenhouses under a SD or LD. Ten plants were randomly
assigned to each combination of temperature and photoperiod treatment. The experiment was
arranged as a split-plot design; two temperatures (whole plot factor) were randomly assigned to

147

the greenhouses, and two photoperiods (sub-plot factor) were randomly assigned to the
experimental units (plants) within each temperature treatment. The actual average air
–2 –1

temperatures were 15.9 and 20.7 ºC and the mean DLI was 10.0 mol·m ·d

in the greenhouses

during the experiment.
Greenhouse culture and environments. All plants were treated with a 250 ml
paclobutrazol (Piccolo, Fine Americas Inc., Walnut Creek, CA) drench at 5 mg·L

–1

7 to 10 d

after transplant to inhibit extension growth. In the greenhouse, plants were grown under a 10-h
SD or a 16-h LD. Under all lighting regimens, plants were grown under a 9-h (0800 to 1700 HR)
truncated natural photoperiod (lat. 43 °N) using blackcloth and supplemental lighting (at 100 to
–2 –1

120 µmol·m ·s ) from high-pressure sodium lamps when the outside light intensity was < 290
−2 −1

–

µmol·m ·s . The 9-h photoperiod was extended by incandescent lamps (at 1 to 3 µmol·m
2 –1

·s ) to create the SD (1700 to 1800 HR) and LD (1700 to 2400 HR) photoperiod. The air

temperature and PPF in each greenhouse were recorded by a shielded and aspirated 0.13 mm
type E thermocouple (Omega Engineering, Stamford, CT), and a line quantum sensor containing
10 photodiodes (Apogee Instruments) placed at canopy height (22 cm above bench height),
respectively. A datalogger (CR-10, Campbell Scientific, Logan, UT) collected the environmental
data every 10 s and hourly averages were recorded. Plants were irrigated as necessary with
–1

reverse osmosis water supplemented with a water-soluble fertilizer containing (mg·L ) 125 N,
12 P, 100 K, 65 Ca, 12 Mg, 1.0 Fe and Cu, 0.5 Mn and Zn, 0.3 B, and 0.1 Mo (MSU RO Special;
GreenCare Fertilizers, Inc., Kankakee, IL).
Data collection and analysis: Plants were considered in flower when the first flower fully
opened on each plant. At first flowering, the date was recorded and days to flower from
transplant were calculated. Flowering position (terminal or lateral), flower and flower bud

148

number, axillary branch number, and length of the longest lateral were also recorded at first open
flower. When flowering was on the primary stem, the number of leaves below the first open
flower was recorded, whereas when the first flower opened on a lateral branch, the total leaf
number (flowering lateral + primary stem below flowering lateral) was counted. In addition,
flower and flower bud number two weeks after flowering were recorded in Expt. 2. Plants that
did not flower within 90 d (Expt. 1) or 100 d (Expt. 2) of transplant were considered
nonflowering. Some plants demonstrated abnormal flowering characteristics, and were not
included in the statistical anaysis to allow for an unbiased treatment comparison. SAS (SAS
Institute, Cary, N.C.) was used to analyze the experimental data. Mean separation using Tukey’s
honestly significant difference test at P ≤ 0.05 was performed on all the data to compare
differences between treatments. Data from treatments that had ≤30% flowering is not presented,
except flowering percentage.

Results
Expt 1. The effect of reciprocal photoperiodic transfers at different stages of development
on flowering. Essentially all ‘Easy Wave Neon Rose’ and ‘Improved Wave Purple’ flowered
within 90 d after transplant in all the treatments (Figures 4.1A, C). As leaf number of ‘Classic
Wave Purple’ at photoperiodic transfer increased, flowering percentage decreased from 100 to 0
when plants were transferred from SD to LD, but increased from 0 to 40% during the reciprocal
transfer (Figure 4.1B). Although ‘Classic Wave Purple’ plants that flowered in this experiment
had 6 to 8 flower buds at first flowering, additional open flowers were not observed after the first
flower opened (personal observation). For the parameters measured in this study, significant
treatment effects occurred only in ‘Easy Wave Neon Rose’ (Figures 4.1, 4.2). However,

149

reciprocal photoperiodic transfers at different stages of development did not influence stem
length (Figure 4.2D) and flowering position (data not presented) in this variety.
‘Easy Wave Neon Rose’ plants that unfolded 10 or 12 leaves under SDs flowered
significantly later than plants transferred to LDs after unfolding 4 leaves, or initially grown under
LDs until they attained 6 or more leaves (Figure 4.1D). In general, flowering time increased with
increase in leaf number at transfer from SD to LD, but not for the reciprocal transfer. Similarly,
leaf number below the first open flower and lateral branch number generally increased with an
increase in leaf number at transfer from SD to LD for ‘Easy Wave Neon Rose’, but not for the
reciprocal transfer (Figures 4.2A, G). For example, plants grown under SDs until they attained 4,
10, or 12 leaves developed 13, 20, or 21 leaves and 17, 22, or 21 branches, prior to flowering. In
addition, ‘Easy Wave Neon Rose’ plants transferred from SD to LD after the 4-leaf stage had
greater branching and formed more leaves than those subjected to the reciprocal transfer at a
similar stage of development. Flower number of ‘Easy Wave Neon Rose’ was similar within
each photoperiodic treatment, but plants that unfolded 10 leaves under SDs had a significantly
greater flower number than those initially grown under LDs until they attained 4 to 8 leaves
(Figure 4.1G).
Expt 2. Temperature and photoperiod interaction on flowering. All plants of all three
varieties flowered when grown under a 16-h LD (Figures 4.3A-C). In contrast, under a 10-h SD,
flowering percentage was 100%, 20 to 60%, or 90 or 100% in ‘Easy Wave Neon Rose’, ‘Classic
Wave Purple’, and ‘Improved Wave Purple’, respectively. Flowering was delayed the most in all
the petunia varieties when grown at 14 ºC and under a 10-h photoperiod (Figures 4.3D-F). For
example, ‘Easy Wave Neon Rose’, ‘Classic Wave Purple’ and ‘Improved Wave Purple’ plants
flowered 31, 49, or 59 d earlier, respectively, when grown at 20 ºC with LDs compared to 14 ºC

150

with SDs. Among the petunia varieties studied, temperature and photoperiod interacted to
influence the flowering position only in ‘Easy Wave Neon Rose’ (data not presented). For
example, lateral flowering occurred in 90 or 100% of ‘Easy Wave Neon Rose’ plants grown
under SDs. In contrast, the first flower opened terminally in 50% or 100% of ‘Easy Wave Neon
Rose’ plants when grown under LDs at 14 ºC or 20 ºC, respectively.
Temperature did not influence leaf number below the first open flower in all three
varieties, irrespective of photoperiod (Figures 4.3G-I). Except in ‘Easy Wave Neon Rose’, there
was a developmental delay in flowering under SDs, irrespective of ADT. Branching was greater
in all three varieties under SDs compared to LDs, regardless of ADT (Figures 4.4A-C).
Temperature did not influence branch number at flowering in ‘Classic Wave Purple’ and
‘Improved Wave Purple’ under either photoperiod; and in ‘Easy Wave Neon Rose’ under SDs.
However, ‘Easy Wave Neon Rose’ plants grown LDs had 30% greater branching at 14 ºC
compared to 20 ºC.
The effect of temperature and photoperiod on flower number at first flowering and two
weeks after flowering varied among varieties (Figures 4.4D-I). ‘Easy Wave Neon Rose’ and
‘Improved Wave Purple’ plants grown under SDs generally had a greater flower number than
plants under LDs, whereas the opposite occurred in ‘Classic Wave Purple’. Flower number
recorded at flowering and two weeks after flowering was similar between temperature treatments
for ‘Easy Wave Neon Rose’ grown under SDs, and ‘Improved Wave Purple’ grown under LDs.
In contrast, ‘Easy Wave Neon Rose’ and ‘Classic Wave Purple’ plants grown under LDs, and
‘Improved Wave Purple’ grown under SDs, had a greater flower number (at and 2 weeks after
flowering) at 14 ºC compared to 20 ºC. Except in ‘Classic Wave Purple’ plants grown at 20 ºC
with SDs, all the petunia varieties studied had 51 to 226% more flower buds 2 weeks after

151

flowering than those recorded at first flowering. Flower bud abortion was observed in all the
petunia varieties grown under SDs in this study, but data was not recorded.

Discussion
Petunia ‘Easy Wave Neon Rose’ became receptive to inductive LDs very early in
development (≤ 4 leaves), since the leaf number below the first open flower was similar for
plants grown under continuous LDs until flowering (at 20 ºC), and those transferred to SDs after
unfolding 4 to 12 leaves under LDs (Figures 4.2 and 4.3G). A similar leaf number below the first
open flower in ‘Easy Wave Neon Rose’ grown under different photoperiods further demonstrates
that these plants were induced prior to unfolding 8 leaves. Flowering time was greater in plants
grown under SDs after induction compared to those grown under continuous LDs until
flowering, although they all flowered at a similar leaf number (Figures 4.1 and 4.3B). We
speculate that flower bud abortion or delayed development under non-inductive photoperiod may
have delayed flowering in previously induced plants. Flower bud abortion or delayed
development under non-inductive photoperiods has been reported in some crops such as
chrysanthemum, butterfly flower (Asclepias tuberosa L.), dahlia (Dahlia pinnata Cav.), and
snapdragon (Adams et al., 1998b; 2003; Albrecht and Lehmann, 1991; Brøndum and Heins,
1993). Brøndum and Heins (1993) observed complete flower bud abortion in the facultative SD
plant dahlia when grown at 25 ºC under non-inductive photoperiods (>14 h).
‘Improved Wave Purple’ and ‘Classic Wave Purple’ became sensitive to photoperiod
after unfolding 8 leaves, since plants transplanted at the 8-leaf stage and grown under LDs
developed significantly fewer leaves prior to flowering than those grown under SDs, irrespective
of ADT. Although leaf number and DTF were not significantly influenced by the reciprocal

152

transfer treatments in ‘Improved Wave Purple’, plants transferred to LDs earlier in development
(before unfolding 10 leaves) or transferred to SDs later in development (with 10 or 12 leaves)
generally formed fewer leaves (35 to 39) compared to the remaining treatments (43 to 48).
‘Improved Wave Purple’ plants with 10 leaves received 4 LDs (Table 1), which was probably
not sufficient for floral induction. Incomplete floral induction may have caused the variability
observed in leaf number and DTF data recorded for this variety.
‘Classic Wave Purple’ demonstrated a strong photoperiodic response in this study. For
example, nearly all of the plants transferred to LDs after unfolding 4 or 6 leaves flowered,
whereas only 40% of those unfolding 12 leaves under LDs flowered (Figures 4.1B).
Interestingly, Adams et al. (1998) reported that chrysanthemum plants exposed to LDs prior to
SDs (inductive) required fewer SDs for flowering compared to those grown under continuous
SDs. For example, chrysanthemum plants grown under 53 LDs and then transferred to inductive
SDs flowered after 49 SDs, whereas those grown under continuous SDs required 57 SDs for
flowering. This may be due to increased sensitivity to inductive conditions with plant age, since
7-d old arabidopsis (Arabidopsis thaliana Heynh.) plants needed 5 LDs to induce flowering,
whereas only a single inductive cycle was sufficient to induce flowering in 20-d old plants
(Mozley and Thomas, 1995). The progressive decrease in flowering percentage with an increase
in plant age when transferred to LDs, and vice versa, suggests that similar to ‘Improved Wave
Purple’, plants may not have received a sufficient number of LDs for floral induction. In
addition, ‘Classic Wave Purple’ plants that flowered under SDs (at 20 ºC) in both experiments
did not develop additional flowers (data not presented), which suggests that induced plants
reverted to a vegetative state under non-inductive conditions.

153

Flowering was delayed at 14 ºC compared to 20 ºC in all the petunia varieties studied,
irrespective of photoperiod. This was expected, since several studies have reported a decrease in
days to flower (DTF) with an increase in ADT in petunia (Adams et al., 1996; 1998a; Blanchard
et al., 2011; Kaczsperski et al., 1991; Warner, 2010). For example, Petunia ×hybrida ‘Snow
Cloud’ and ‘Mitchell’ flowered 22 d and 35 d earlier, respectively, at 20 ºC than at 14 ºC when
–2 –1

grown under a 16- or 18-h photoperiod and a DLI of 13 or 16.5 mol·m ·d

(Kaczsperski et al.,

1991; Warner, 2010). Similarly, Blanchard et al. (2011) reported a 24 or 25 d delay in flowering
time in petunia ‘Easy Wave Coral Reef’ and ‘Wave Purple’ when grown at 14 ºC versus 20 ºC
–2 –1

and under a DLI of 10 mol·m ·d

.

Interestingly, temperature modified the flowering response to photoperiod in all the
petunia varieties studied. Temperature had a greater effect on flowering time under LDs than
SDs, whereas the LD response decreased at 14 ºC compared to 20 ºC in this study. Growing
plants at 14 versus 20 ºC delayed flowering by 18% or 95% in ‘Easy Wave Neon Rose’ and by
15% or 110% in ‘Improved Wave Purple’ when grown under SDs or LDs, respectively.
Similarly, LDs promoted flowering by 18% or 95% in ‘Easy Wave Neon Rose’ and by 66% or
202% in ‘Improved Wave Purple’ when grown at 14 or 20 ºC, respectively. This response was
particularly strong in ‘Classic Wave Purple’, which demonstrated a nearly obligate LD response
at 20 ºC, whereas a majority of plants flowered under SDs at 14 ºC, and nonflowering plants had
visible buds. These results are consistent with those reported by Adams et al. (1996) in trailing
petunias; growing petunia ‘Malve’ at 14 versus 20 ºC delayed the flowering time by 30% under
an 8-h photoperiod and 72% under a 16-h photoperiod. Similarly, 16-h LDs compared to 10-h
SDs decreased the predicted flowering time of this variety by 88% at 20 ºC and by 42% at 14 ºC.

154

Except for flower number in ‘Classic Wave Purple’ and leaf number in ‘Easy Wave Neon
Rose’ in Expt. 2, treatments that promoted flowering generally had a lower quality (flower and
branch number) and leaf number at flowering. Similarly, a positive correlation between
flowering time, leaf number below the first open flower, and plant quality was reported by
Warner (2009) in celosia. Faster flowering is sometimes offset by reduced plant quality in many
species, since plants intercept light for a shorter duration (Pramuk and Runkle, 2005a; Warner,
–2 –1

2009). A decrease in DTF from 43 d under 4 mol·m ·d

–2 –1

to 33 d under 14 mol·m ·d ,

decreased the leaf number, flower number, and dry weight by 39, 84, and 56%, respectively
(Pramuk and Runkle, 2005a). In addition, branch number, flower number at flowering and 2
weeks after flowering was greater for plants grown at 14 ºC compared to 20 ºC, regardless of
photoperiod. These responses to ADT are consistent with previous studies in petunia (Adams et
al., 1996; Blanchard et al., 2011; Kaczperski et al., 1991). For example, petunia ‘Wave Purple’
–

and ‘Easy Wave Coral Reef’ plants grown at 14 versus 20 ºC and under a DLI of 10.0 mol·m
2 –1

·d

had 17 and 6% greater inflorescences, respectively (Blanchard et al., 2011). Similarly,

Kaczperski et al. (1991) reported a quadratic decrease in the number of lateral shoots with an
increase in ADT from 10 to 30 ºC.
In general, the length of the juvenile phase, the minimum number of cycles required for
floral induction, and flowering responses to temperature and photoperiod varied among the
petunia varieties studied. Previous studies in celosia, snapdragon, and whiteedge morning-glory
have reported a similar varietal difference in photoperiodic behavior (Adams et al., 2003;
Imamura et al., 1966; Warner, 2009). For example, the juvenile stage in snapdragon varied from
9.9 leaves in ‘Chimes’ to 28.8 leaves in ‘Annabel’ (Adams et al., 2003). Similarly, 1, 2, and 10

155

SDs were required to induce flowering in whiteedge morning-glory ‘Kidachi’, ‘Nepal’, and
‘Africa’, respectively (Imamura et al., 1966).
Petunia ‘Easy Wave Neon Rose’ became sensitive to photoperiod after unfolding ≤4
leaves, whereas ‘Classic Wave Purple’ and ‘Improved Wave Purple’ became sensitive after
unfolding >8 leaves. However, the duration of the juvenile phase may be confounded with the
number of inductive cycles required for flowering in this study, especially since plant age may
influence sensitivity to photoperiod in some species (Mozley and Thomas, 1995). Further
research using limited inductive photoperiod at regular intervals during development may help
identify the precise juvenile phase and minimum number of inductive cycles needed for
complete floral induction. In addition, flower bud abortion observed under non-inductive
conditions in this study suggests that flower bud development in the petunia varieties studied is
sensitive to photoperiod, although some petunia varieties may have a distinct photoperiodsensitive phase (Adams et al., 1999). Experiments using reciprocal transfers until flowering may
be performed on additional popular petunia varieties to identify the photoperiod sensitive phases.

156

Table 4.1. Days from start of photoperiodic treatments to reciprocal transfer at different stages
of development (leaf number) in three petunia cultivars during the transplant stage.
Leaf number at photoperiodic transfer
Petunia cultivar
4
6
8
10
12
Easy Wave Neon Rose
13
16
19
22
25
Classic Wave Purple
13
19
22
26
30
Improved Wave Purple
13
16
19
21
24

157

Classic Wave Purple

Flowering percentage

Easy Wave Neon Rose
B

C

E

A

100

Improved Wave Purple

F

80
60
40
20

SD to LD
LD to SD

0
D

Days to flower

70

a

a

a

50
ab

ab

40

b

30
0

ab

a

ab abc

a

a

a

a

a

a

a

a
a

I

H
abc

120

a

a

a

a

a

b

b

b

b

a

a

G

140

Flower number

a

60

a
abc

100
80

c

60

bc

bc

abc

abc

40

a
a

20

a

a

4

6

8

10

12

2

4

6

8

a
a

a

a

0
2

a

10

12

2

4

6

8

10

12

14

Leaf number at photoperiodic transfer

Figure 4.1. The effect of transferring plants at different stages of development (leaf number)
–1
–1
from long days (LD, 16 h·d ) to short days (SD, 10 h·d ) (●), and from SD to LD (○) during
the transplant stage on flowering characteristics of three petunia varieties. Plants were
transplanted at 12-leaf stage and grown in a greenhouse maintained at a constant setpoint of 20
°C with a 10 h photoperiod. Each symbol represents the treatment means, and error bars
represent the standard errors. Means followed by the same letter are not statistically different (P
≤ 0.05). Treatments that had ≤ 30% flowering are not shown, except for flowering percentage.

158

Classic Wave Purple

Easy Wave Neon Rose

Branch number

30

B

A

25
cde bcd

20
15

C

a

bc

a

ab

e

e

de

de

a

a

a

10

a

a

a

a

a
a

a

SD to LD
LD to SD

0

D

60

F

E

a
a

Col 1 vs Col 2
Col 4 vs Col 5

a

50

30
20

a
a

0

a

a

a

a

a

a

I
a

a

40

cd

10

d

d

d

d

a

a

a

4

6

a

a

a

30
ab

a

a

50

b

a

a

H

bc

a
a

a

G

20

a

a

a

a

a

a

a

40

10

Leaf number

a

a

a

de

5

Stem length (cm)

Improved Wave Purple

a
a

a
a

a
a

a

d

0
2

4

6

8

10

12

2

8

10

12

2

4

6

8

10

12

14

Leaf number at photoperiodic transfer

Figure 4.2. The effect of transferring plants at different stages of development (leaf number)
–1
–1
from long days (LD, 16 h·d ) to short days (SD, 10 h·d ) (●), and from SD to LD (○) during
the transplant stage on flowering characteristics of three petunia varieties. Plants were
transplanted at 12-leaf stage and grown in a greenhouse maintained at a constant setpoint of 20
°C with a 10 h photoperiod. Each symbol represents the treatment means, and error bars
represent the standard errors. Means followed by the same letter are not statistically different (P
≤ 0.05). Treatments that had ≤ 30% flowering are not shown.

159

Flowering percentage

Easy Wave Neon Rose

Classic Wave Purple

A

10 h B
16 h

C

D

E

F

100
80
60
40
20
0

a

100

Days to flower

Improved Wave Purple

a
b

80
a

60

b

b

b

c
c

40

d

c

20
0

Leaf number

60

H

G

I
a

50

a
a

40
30
20

b
b ab

14

20

b

ab a

b

b

10
0

14

20

14

20

Mean daily temperature (°C)
–1

Figure 4.3. The effect of photoperiod (10 or 16 h·d ) on flowering characteristics of three
petunia varieties grown at 14 or 20 °C. Plants were initially grown at a constant setpoint of 20 °C
with a 16 h photoperiod in an environmental growth chamber, and then transplanted when they
averaged 8 leaves. Error bars represent the standard errors. Means followed by the same letter
are not statistically different (P ≤ 0.05). Treatments that had ≤ 30% flowering are not shown,
except for flowering percentage.

160

Branch number

10 h
16 h

20
a

15

B

C

a

a

b

ab

a

b

b

c

bc

10

c

5
0

140

Flower number
at first flowering

A

Improved Wave Purple

Classic Wave Purple

Easy Wave Neon Rose
25

E

D
a

120

F

a

a

100
b

80

b

a

60

b

c

40

b

bc

b

20
0

G

H

I

Flower number
at 2 weeks

300

a

250
200
150

ab

a
a

b

b

100

c

b

b
b

c

50
0

14

20

14

20

14

20

Mean daily temperature (°C)
–1

Figure 4.4. The effect of photoperiod (10 or 16 h·d ) on flowering characteristics of three
petunia varieties grown at 14 or 20 °C. Plants were initially grown at a constant setpoint of 20 °C
with a 16 h photoperiod in an environmental growth chamber, and then transplanted when they
averaged 8 leaves. Error bars represent the standard errors. Means followed by the same letter
are not statistically different (P ≤ 0.05). Treatments that had ≤ 30% flowering are not shown.

161

LITERATURE CITED

162

Literature cited

Adams, S.R., P. Hadley, and S. Pearson. 1998a. The effects of temperature, photoperiod, and
photosynthetic photon flux on the time to flowering of petunia 'Express Blush Pink'. J.
Amer. Soc. Hort. Sci. 123:577–580.
Adams, S.R., M. Munir, V.M. Valdes, F.A. Langton, and S.D. Jackson. 2003. Using flowering
times and leaf numbers to model the phases of photoperiod sensitivity in Antirrhinum
majus L. Ann. Bot. 92: 689–696.
Adams, S.R., S. Pearson, and P. Hadley. 1996. The effects of temperature and photoperiod on the
flowering and morphology of trailing petunias. Acta Hort. 435:65–75.
Adams, S.R., S. Pearson, and P. Hadley. 1997. The effects of temperature, photoperiod and light
integral on the time to flowering of pansy cv Universal Violet (Viola ×wittrockiana
Gams). Ann. Bot. 80:107–112.
Adams, S.R., S. Pearson, and P. Hadley. 1998b. An appraisal of the use of reciprocal transfer
experiments: Assessing the stages of photoperiod sensitivity in chrysanthemum cv.
Snowdon (Chrysanthemum ×morifolium Ramat.). J. Exp. Bot. 49:1405–1411.
Adams, S.R., S. Pearson, and P. Hadley. 1998c. The effect of temperature on inflorescence
initiation and subsequent development in chrysanthemum cv. Snowdon (Chrysanthemum
×morifolium Ramat.). Scientia Hort. 77:59–72.
Adams, S.R., S. Pearson, and P. Hadley. 2001. Improving quantitative flowering models through
a better understanding of the phases of photoperiod sensitivity. J. Exp. Bot. 52: 655–662.
Adams, S.R., S. Pearson, P. Hadley, and W.M. Patefield. 1999. The effects of temperature and
light integral on the phases of photoperiod sensitivity in Petunia ×hybrida. Ann. Bot.
83:263–269.
Albrecht, M.L. and J. T. Lehmann. 1991. Daylength, cold storage, and plant-production method
influence growth and flowering of Asclepias tuberosa. HortScience 26:120–121.
Blanchard, M.G. and E. S. Runkle. 2011. Quantifying the thermal flowering rates of eighteen
species of annual bedding plants. Scientia Hort. 128:30–37.
Blanchard, M.G., E.S. Runkle, and P.R. Fisher. 2011. Modeling plant morphology and
development of petunia in response to temperature and photosynthetic daily light integral.
Scientia Hort. 129:313–320.
Brøndum, J.J. and R.D. Heins. 1993. Modeling temperature and photoperiod effects on growth
and development of dahlia. J. Amer. Soc. Hort. Sci. 118:36–42.

163

Damann, M.P. and R.E. Lyons. 1993. Juvenility, flowering, and the effects of a limited inductive
photoperiod in Coreopsis grandiflora and C. lanceolata. J. Amer. Soc. Hort. Sci.
118:513–518.
Han, T., C. Wu, Z. Tong, R.S. Mentreddy, K. Tan, and J. Gai. 2006. Postflowering photoperiod
regulates vegetative growth and reproductive development of soybean. Environ. Exp.
Bot. 55:120–129.
Imamura, S., M. Muramatsu, S.Kitajo, and A. Takimoto. 1966. Varietal difference in
photoperiodic behavior of Pharbitis nil. Bot. Mag. Tokyo 79:714–721.
Kaczperski, M.P., W.H. Carlson, and M.G. Karlsson. 1991. Growth and development of Petunia
×hybrida as a function of temperature and irradiance. J. Amer. Soc. Hort. Sci. 116:232–
237.
Kelly, R.O., Z. Deng, and B.K. Harbaugh. 2007. Evaluation of 125 petunia cultivars as bedding
plants and establishment of class standards. HortTechnology 17:386–396.
Kiniry, J.R., J.T. Ritchie, R.L. Musser, E.P. Flint, and W.C. Iwig. 1983. The photoperiod
sensitive interval in maize. Agron. J. 75:687–690.
Lam, S.L. and A.C. Leopold. 1961. Reversion and reinduction of flowering in perilla. Amer. J.
Bot. 48:306–310.
Mattson, N.S. and J.E. Erwin. 2005. The impact of photoperiod and irradiance on flowering of
several herbaceous ornamentals. Scientia Hort. 104:275–292.
Moccaldi, L.A. and E.S. Runkle. 2007. Modeling the effects of temperature and photosynthetic
daily light integral on growth and flowering of Salvia splendens and Tagetes patula. J.
Amer. Soc. Hort. Sci. 132:283–288.
Moccaldi, L.A. and E.S. Runkle. 2007. Modeling the effects of temperature and photosynthetic
daily light integral on growth and flowering of Salvia splendens and Tagetes patula. J.
Amer. Soc. Hort. Sci. 132:283–288.
Mozley, D. and B. Thomas. 1995. Developmental and photobiological factors affecting
photoperiodic induction in Arabidopsis thaliana Heynh. Landsberg erecta. J. Exp. Bot.
46:173–179.
Piringer, A.A. and M.M. Cathey. 1960. Effect of photoperiod, kind of supplemental light and
temperature on the growth and flowering of petunia plants. Proc. Amer. Soc. Hort. Sci.
76:649–660.
Pouteau, S., D. Nicholls, F. Tooke, E. Coen, and N. Battey. 1997. The induction and
maintenance of flowering in Impatiens. Development 124: 3343–3351.

164

Pramuk, L.A. and E.S. Runkle. 2005a. Photosynthetic daily light integral during the seedling
stage influences subsequent growth and flowering of Celosia, Impatiens, Salvia, Tagetes,
and Viola. HortScience 40:1336–1339.
Pramuk, L.A. and E.S. Runkle. 2005b. Modeling Growth and Development of Celosia and
Impatiens in response to temperature and photosynthetic daily light integral. J. Amer.
Soc. Hort. Sci. 130:813–818.
Roberts, E.H., R.J. Summerfield, J.P. Cooper, and R.H. Ellis. 1988. Environmental control of
flowering in barley (Hordeum vulgare L.). 1. Photoperiod-insensitive phases and effects
of low-temperature and short day vernalization. Ann Bot. 62:127–144.
Pramuk, L.A.and E.S. Runkle. 2005b. Modeling growth and development of Celosia and
Impatiens in response to temperature and photosynthetic daily light integral. J. Amer.
Soc. Hort. Sci. 130:813–818.
Thomas, B. and D. Vince-Prue. 1997. Photoperiodism in plants. Academic Press, Inc., San
Diego, CA.
U.S. Dept. of Agriculture. 2012. Floriculture crops 2011 summary. Agri. Stat. Board,
Washington, D.C.
Verhuel, M.J., A. Sønsteby, and S.O. Grimstad. 2006. Interactions of photoperiod, temperature,
duration of short day treatment and plant age on flowering of Fragaria ×ananassa Duch.
cv. Korona. Scientia Hort. 107:164–170.
Wang, Z., M.C. Acock, and B. Acock. 1997. Photoperiod sensitivity during flower development
of opium poppy (Papaver somniferum L.). Ann. Bot. 79:129–132.
Warner, R.M. 2010. Temperature and photoperiod influence flowering and morphology of four
Petunia spp. HortScience 45:365–368.
Warner, R.M. 2009. Determination of photoperiod-sensitive stages of development of the shortday plant celosia. HortScience 44:328–333.
Wilkerson, G.G., J.W. Jones, K.J. Boote, and G.S. Buol. 1989. Photoperiodically sensitive
interval in time to flower of soybean. Crop Sci. 29:721–726.

165

APPENDIX

MODELING PLANT DEVELOPMENT AND QUALITY PARAMETERS OF 16
PETUNIA VARIETIES IN RESPONSE TO AVERAGE DAILY TEMPERATURE
UNDER TWO DAILY LIGHT INTERGRAL ENVIRONMENTS

166

Research Objective
The objectives of this study were to quantify and model the influence of average daily
temperature (ADT) on flowering characteristics of 16 Wave varieties of petunia under two
photosynthetic daily light integral (DLI) conditions.

Materials and Methods
Plant material. Experimental protocol and data collection and analysis were similar to
that reported by Blanchard and Runkle (2011) and Pramuk and Runkle (2005). On 10 January
2011, seeds of petunia (Petunia ×hybrida Vilm.-Andr.) ‘Easy Wave Blue’, ‘Easy Wave
Burgundy Star’, ‘Easy Wave Neon Rose’, ‘Easy Wave Pink’, ‘Easy Wave Plum Vein’, ‘Easy
Wave Red Improved’, ‘Easy Wave Violet’, ‘Easy Wave White’, ‘Shock Coconut’, ‘Shock
Denim’, ‘Shock Wave Purple, ‘Tidal Wave Silver’, ‘Wave Blue’, ‘Wave Pink’, ‘Wave Purple
Classic’, and ‘Wave Purple Improved’ were sown in 288-cell (6-mL) plug trays by a commercial
greenhouse (C. Raker & Sons, Litchfield, MI). After germination, seedlings were received at
Michigan State University (MSU) on 20 January 2011 and grown in controlled environment
chambers until deemed ready for transplant.
Growth chamber treatments and environment. Seedlings were grown in the controlled
–

environment chambers at a constant temperature setpoint of 20 ºC under a PPF of 180 µmol·m
2 –1

·s

–2 –1

DLI ≈ 10 mol·m ·d ) provided by a combination of cool-white fluorescent (CWF;

F96T12CWVHO; Philips, Somerset, NJ) and incandescent lamps (INC, Philips, Somerset, NJ)
with a 16-h photoperiod. The light intensity was checked periodically with an instantaneous
quantum sensor (Apogee Instruments Inc., Logan, UT) at canopy height and adjustments were
made by replacing and/or lowering the lamps when required. Plugs were hand-irrigated as

167

necessary with acidified well water (140 mg·L

–1

titratable alkalinity of CaCO3) containing

–1

(mg·L ) 95, 34, and 29 Ca, Mg, and S, respectively, and supplemented with a water-soluble
–1

fertilizer providing (mg·L ) 62 N, 6 P, 62 K, 7 Ca, 0.5 Fe, 0.3 Cu, Mn, and Zn, and 0.1 B and
Mo (MSU Well Water Special; GreenCare Fertilizers Inc., Kankakee, IL).
Greenhouse treatments and environment. After 28 d from seed sowing, seedlings with six
to eight leaves were transplanted into 10-cm round (480-mL) containers filled with a peat-based
medium (Suremix, Michigan Grower Products, Galesburg, MI). The mean leaf number at
transplant was recorded for each variety. Seedlings were thinned to one plant per cell prior to
transplant and 20 plants of each species were grown in separate glass-glazed greenhouse
compartments at constant temperature set points of 12, 15, 18, 21, or 24 °C and under a 16-h
photoperiod (0600 to 2200 HR) created by using the natural photoperiod (lat. 43 °N) and dayextension lighting from high-pressure sodium (HPS) lamps. Plants were treated with a 250 ml
paclobutrazol (Piccolo, Fine Americas Inc., Walnut Creek, CA) drench at 4 mg·L

–1

7 to 10 d

after transplant to inhibit extension growth. In each temperature treatment, plants were grown
under two DLIs with 10 plants under each temperature and DLI combination. A low DLI was
created by using ambient light with 50% shade curtains (OLS 50; Ludvig Svensson Inc.,
Charlotte, NC) and supplemental lighting from HPS lamps that provided a PPF of 25 to 50
–2 –1

µmol·m ·s

at canopy height (22 cm above bench height). Plants under the high DLI treatment

were grown under ambient light without a shade curtain and with supplemental HPS lighting that
–2 –1

provided a PPF of 80 to 100 µmol·m ·s

at canopy height. The HPS lamps were programmed
–2 –1

to turn on when outside light intensity was <290 µmol·m ·s
2 –1

·s

–

and turn off at >580 µmol·m

by an environmental computer system (Priva Intégro 724; Priva, Vineland Station, Ontario,

Canada). In each greenhouse compartment, a shielded and aspirated 0.13 mm type E

168

thermocouple (Omega Engineering, Stamford, CT) recorded the air temperature and a line
quantum sensor containing 10 photodiodes (Apogee Instruments) placed at canopy height (22 cm
above bench height) measured light intensity. A CR10 datalogger (Campbell Scientific, Logan,
UT) collected the environmental data every 10 s and hourly averages were recorded. Actual
average air temperatures in the greenhouse during the experiment were 12.7, 15.0, 18.1, 21.0, or
–2 –1

23.4 ºC. Mean DLI under the low and high DLI treatments were 7.1 and 13.7 mol·m ·d ,
respectively. Vapor pressure deficit was maintained at 1.2 kPa by steam injection at night (2200
to 0800 HR). Plants were irrigated as necessary with reverse osmosis water supplemented with a
–1

water-soluble fertilizer containing (mg·L ) 125 N, 12 P, 100 K, 65 Ca, 12 Mg, 1.0 Fe and Cu,
0.5 Mn and Zn, 0.3 B, and 0.1 Mo (MSU RO Special; GreenCare Fertilizers, Inc., Kankakee, IL).
Data collection and analysis: Plants were considered in flower when the first flower fully
opened on each plant. At first flowering, the date was recorded and days to flower from
transplant were calculated. Flower and flower bud number, axillary branch number, and length of
the longest lateral were also recorded at first open flower. When flowering was on the primary
stem, the number of nodes below the first open flower was recorded, whereas when the first
flower opened on a lateral branch, the total node number (flowering lateral + primary stem below
flowering lateral) was counted. The experiment was arranged as a split-plot design; five
temperatures (whole plot factor) were randomly assigned to the greenhouse compartments, and
two DLI treatments (sub-plot factor) were randomly assigned to the experimental units (plants)
within each temperature treatment.
SAS (SAS Institute, Cary, N.C.) was used to analyze the experimental data. Linear and
quadratic regression (REG procedure) analysis was used to generate equations to describe the
effect of ADT on plant quality parameters under low and high DLI conditions. Linear regression

169

analysis was independently performed on the flowering rate data generated under the two DLIs,
and the slope and intercept values generated from these equations were used to calculate the
estimated base temperatures (Tmin) for each species (Roberts and Summerfield, 1987). Many
researchers have estimated the Tmin and described the flowering time response as a function of
ADT using linear regression when Tmin < ADT < Topt (optimum temperature) (Clough et al.,
2001; Niu et al., 2000, 2001; Pietsch et al., 1995; Whitman et al., 1997; Yuan et al., 1998):
1/d to flower = b0 + b1 × ADT

(1)

where 1/d to flower = the rate of progress towards flowering, b1 = slope, b0 = intercept, and
ADT = average daily temperature (°C) above Tmin and below Topt. Given that the
developmental rate at Tmin is zero, equation (1) can be used to calculate Tmin:
Tmin = – b0 / b1

(2)

where Tmin = temperature at or below which the flower development rate is zero.

Results
Flower development rate linearly increased with an increase in ADT from 12 to 24 ºC for
all the petunia Wave varieties included in this study, and the nature of this response varied
among the varieties (Figures 5.1-5.16A). For example, each degree increase in temperature from
12 to 24 ºC increased the rate of progress towards flowering from 0.0014 in ‘Easy Wave Pink’ to
0.0024 in ‘Shock Wave Purple’ (Figure 5.4 and 5.11). Linear models adequately described the
effect of temperature on flowering rate in all the varieties studied, since the temperatures used in
this experiment (12 to 24 ºC) were between Tmin and Topt for these varieties. The coefficients of
2

determination generated for the linear models (r ) ranged from 0.55 to 0.95 (data not presented).
In some varieties, the variability in flowering time response differed between DLI treatments,

170

2

2

which considerably influenced the r values. For example, the r values of ‘Shock Wave Denim’
varied from 0.81 under high DLI to 0.63 under low DLI mainly due to greater variability under
the low DLI treatment (Figure 5.10A). The higher DLI accelerated flowering compared to the
lower DLI in all the varieties irrespective of ADT (Figures 5.1-5.16A), and the absolute
magnitude of this response was generally greater at lower temperatures. For example, ‘Shock
Wave Denim’ plants grown under the higher DLI flowered 9 d and 29 d earlier at 24 ºC and 12
ºC, respectively (Figure 5.10A).
The estimated Tmin for flower development rate varied among the petunia Wave
varieties, and in some cases, also with the DLI (Figures 5.1-5.16A). For example, Tmin ranged
from 1.7 ºC to 7.7 ºC under the high DLI and from 3.5 to 7.5 under the low DLI for ‘Easy Wave
Neon Rose’ and ‘Wave Pink’, respectively (Figures 5.3 and 5.14A). Except in ‘Wave Pink’ and
‘Wave Purple Improved’, Tmin was 0.5 to 3.8 ºC higher under the low DLI conditions (Figures
5.1–5.16A). Petunia varieties in which the estimated Tmin was ≤ 4 ºC regardless of the DLI were
‘Easy Wave Blue’, ‘Easy Wave Neon Rose’, ‘Easy Wave Pink’, and ‘Wave Blue’. Those with
an estimated Tmin between 4 to 8 ºC were ‘Easy Wave Red Improved’, ‘Shock Wave Coconut’,
‘Shock Wave Purple’, ‘Tidal Wave Silver’, ‘Wave Pink’, ‘Wave Purple Classic’, and ‘Wave
Purple Improved’. The estimated Tmin for some varieties, such as ‘Easy Wave Burgundy Star’,
‘Easy Wave Plum Vein’, ‘Easy Wave Violet’, ‘Easy Wave White’, and ‘Shock Wave Denim’,
was ≤ 4 ºC under high DLI, but between 4 to 8 ºC under low DLI. There was a positive
correlation (P < 0.0001) between the estimated Tmin and delay in flowering time when plants
were grown at 12 versus 24 ºC. For example, the percentage delay in flowering time when plants
were grown at 12 ºC compared to 24 ºC was 177 to 181% in ‘Wave Pink’ (Tmin = 7.5 to 7.7 ºC),
and 90 to 108% in ‘Easy Wave Pink’ (Tmin = 2.0 to 3.8 ºC).

171

In all petunia varieties grown under the low DLI treatment, flower number increased as
ADT decreased from 24 to 12 ºC (Figures 5.1-5.16B). For example, as ADT decreased from 24
to 12 ºC, flower bud number increased by 121 and 143% in ‘Tidal Wave Silver’ and ‘Easy Wave
Violet’, respectively. However, under the higher DLI, a similar decrease in ADT increased
flower number for only 6 of the 16 varieties studied (Figures 5.3, 5.6, 5.7, 5.14-5.16). An
optimum temperature for flower bud number was observed under the high DLI for ‘Easy Wave
Blue’, ‘Easy Wave Burgundy Star’, ‘Easy Wave Pink’, ‘Easy Wave Plum Vein’, ‘Easy Wave
White’, ‘Shock Wave Denim’, and ‘Wave Blue’ (Figures 5.1, 5.2, 5.4, 5.5, 5.8, 5.10, 5.13B).
There was no significant relationship between ADT and flower number in ‘Shock Wave
Coconut’, ‘Shock Wave Purple’, and ‘Tidal Wave Silver’ when grown under the high DLI
environment (Figures 5.9, 5.11, 5.12B).
Lateral branch number was inversely related with ADT in 13 and 15 of the 16 petunia
varieties grown under the high and low DLI conditions, respectively (Figures 5.1-5.16C). For
example, within the temperature range studied, branch number in Easy Wave Violet linearly
increased by 25 and 45% under the high and low DLI conditions, respectively. In ‘Easy Wave
Burgundy Star’ grown under the low DLI, and ‘Easy Wave Plum Vein’, ‘Easy Wave White, and
‘Shock Wave Denim’ grown under high DLI, branch number increased as temperature decreased
until an optimum temperature, beyond which it decreased (Figures 5.2, 5.5, 5.8, 5.10C). In
general, the high DLI treatment had a promotive effect on flower bud number and branch
number, particularly at the moderate and higher temperatures, in most petunia varieties.
As ADT increased from 12 to 24 ºC, stem length decreased in ‘Easy Wave Blue’, ‘Easy
Wave Burgundy Star’, ‘Easy Wave Pink’, ‘Easy Wave Plum Vein’, ‘Easy Wave Violet’, ‘Easy
Wave White’, and ‘Shock Wave Coconut’ when grown under the low DLI (Figures 5.1, 5.2, 5.4,

172

5.5, 5.7, 5.8, 5.9). Under the same DLI conditions, stem length of ‘Easy Wave Red Improved’,
‘Shock Wave Denim’, ‘Shock Wave Purple’, ‘Tidal Wave Silver’, ‘Wave Pink’, and ‘Wave
Purple Classic’ first decreased and then increased with ADT. In contrast, there was no significant
effect of ADT on stem length in ‘Easy Wave Neon Rose’, ‘Wave Blue’, and ‘Wave Purple
Improved’ (Figures 5.3, 5.6, 5.10-5.16D). ADT had varying effects on stem length among the
petunia varieties grown under the high DLI (Figures 5.1-5.16D). For example, stem length was
inversely correlated with ADT in ‘Easy Wave Improved Red’ and ‘Shock Wave Coconut’,
whereas an opposite trend was observed in ‘Easy Wave Neon Rose’. In ‘Easy Wave Plum Vein’
and ‘Wave Pink’, stem length first decreased with ADT, and then increased. In contrast, there
was no significant relationship between ADT and stem length in the remaining 11 petunia
varieties grown under the high DLI treatment. Plants grown at ≤14 ºC and under the low DLI had
the longest lateral stems at flowering compared to all other treatments.
There was a developmental delay in flowering time (i.e., plants developed more nodes)
with a decrease in temperature in ten petunia varieties grown under the high DLI and six
varieties under the low DLI (Figures 5.1-5.16E). In contrast, ADT did not significantly influence
node number in ‘Easy Wave Plum Vein’, ‘Easy Wave Red Improved’, ‘Easy Wave Violet’, and
‘Shock Wave Coconut’, regardless of the DLI (Figures 5.5-5.7, 5.9E). Similarly, there was no
significant relationship between node number and ADT in ‘Easy Wave Burgundy Star’ and
‘Shock Wave Denim’ when grown under the low DLI; and ‘Easy Wave Blue’, ‘Easy Wave Neon
Rose’, ‘Easy Wave Pink’, ‘Easy Wave White’, ‘Wave Blue’, and ‘Wave Purple Improved’ when
grown under the high DLI (Figures 5.1-5.4, 5.8, 5.10, 5.13, 5.16E).

173

0.035

30

A

D
25

0.030

y = 0.0017x - 0.0058
Tmin = 3.5 °C

0.025

20
15

0.020
y = 0.0015x - 0.0061
Tmin = 4.0 °C

0.015
0.010

10
NS
y = -0.347x + 24.442

5

0.000

0
B

E

60

Flower number

14

50

12

40

10

30

8
y = -0.180x2 + 5.566x + 8.782
y = -1.376x + 65.561

20

y = -0.339x + 16.705
NS

6

0

0
C

11

Branch number

Stem length (cm)

0.040

Node number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Blue'

10

12

14

16

18

20

22

24

26

10
9
8
7
6

y = -0.104x + 11.459
y = -0.011x2 + 0.243x + 7.283

5
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.1. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Blue’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or 14
−2 −1
mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to flower.
Tmin is the estimated temperature at or below which the rate of progress towards flowering is
zero. Each symbol represents the treatment means, and error bars represent 95% confidence
−2 −1
intervals. Dashed and solid lines represent regression equations for the 7 and 14 mol·m ·d
DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

174

0.05
0.04

25

y = 0.0017x - 0.0050
Tmin = 2.9 °C

0.03

20
15

0.02

10
y = 0.0018x - 0.0116
Tmin = 6.7 °C

0.01
0.00

E

y = -0.337x2 + 11.037x - 50.573
y = -4.376x + 118.44

14

80

12

60

10

40

8
NS

20
0
14

Branch number

5

y = -0.075x2 + 1.415x + 16.125

0

B

100

Flower number

NS

Stem length (cm)

D

A

6

y = -0.045x2 + 1.503x - 0.385

Node number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Burgundy Star'

0

C

10

12

14

16

18

20

22

24

26

12
10
8
6
4

NS

2

y = -0.099x2 + 3.311x - 17.716

0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.2. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Burgundy Star’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

175

A

18

D

16

0.04
y = 0.0016x - 0.0028
Tmin = 1.7 °C

0.03

14
12

0.02
0.01
0.00

10

y = 0.0015x - 0.0052
Tmin = 3.5 °C

8

y = 0.023x2 - 0.656x + 15.738
NS

B

0

E

16

Flower number

75

14

60

12
45

10

30

8

Branch number

0

y = -0.168x + 14.731
NS

y = -1.743x + 80.562
y = -2.591x + 93.953

15

Stem length (cm)

0.05

Node number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Neon Rose'

6
0

C

10

12

14

16

18

20

22

24

26

16
12
8
4

y = -0.424x + 20.188
y = -0.753x + 25.211

0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.3. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Neon Rose’ grown under an average daily light integral of 7 mol·m ·d (open symbols)
−2 −1
or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to
flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

176

D

A

20

0.04

18

y = 0.0015x - 0.0031
Tmin = 2.0 °C

0.03

16
14
12

0.02

0.00

NS
y = -0.383x + 20.585

8
0

B

E
16

60

12

40

8
y = -0.263x2 + 7.223x + 8.842
y = -0.259x2 + 5.919x + 27.403

20

14

y = -0.351x + 16.431
NS

Node number

80

0

Branch number

10

y = 0.0014x - 0.0051
Tmin = 3.8 °C

0.01

Stem length (cm)

0.05

Flower number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Pink'

4
0

C

10

12

14

16

18

20

22

24

26

12
10
8
6
y = -0.426x + 17.732
y = 0.025x2 - 1.168x + 20.607

4
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.4. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Pink’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or 14
−2 −1
mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to flower.
Tmin is the estimated temperature at or below which the rate of progress towards flowering is
zero. Each symbol represents the treatment means, and error bars represent 95% confidence
−2 −1
intervals. Dashed and solid lines represent regression equations for the 7 and 14 mol·m ·d
DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

177

26

D

24

y = 0.0017x - 0.0058

22

Tmin = 3.5 °C

0.03

20
18

0.02

16
14

y = 0.0018x - 0.0092
0.01
0.00

Tmin = 5.2 °C

B

0
14
12

60

10
40

8

20
0

NS
NS

6
0

C

10

12

Branch number

10

E

y = -0.116x2 + 2.498x + 38.407
y = -2.100x + 76.720

80

12

y = 0.085x2 - 3.184x + 44.453
y = -0.340x + 25.049

Stem length (cm)

A

Node number

0.04

Flower number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Plum Vein'

12

14

16

18

20

22

24

26

10
8
6
y = -0.030x2 + 0.940x + 2.553
y = -0.295x + 12.740

4
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.5. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Plum Vein’ grown under an average daily light integral of 7 mol·m ·d (open symbols)
−2 −1
or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to
flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

178

0.04

18

D

A

y = -0.166x + 12.658
y = 0.176x2 - 6.943x + 73.533

y = 0.0018x - 0.0081
Tmin = 4.4 °C

0.03

16
14
12
10
8

0.02
0.01
0.00

B

4
2
0

E

y = -0.133x2 + 3.376x + 12.475
y = 0.162x2 - 8.055x + 116.210

50

Flower number

6

y = 0.0017x - 0.0091
Tmin = 5.5 °C

18
15

40

12

30

9

20
10

3
0

C

10

15

Branch number

6

NS
NS

0

Stem length (cm)

0.05

Node number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Red Improved'

12

14

16

18

20

22

24

26

12
9
6
y = -0.216x + 12.454
y = -0.450x + 16.165

3
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.6. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Red Improved’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

179

A

24

D

20

0.04

y = 0.0018x - 0.0053
Tmin = 3.0 °C

0.03

16
12

0.02

8

y = 0.0015x - 0.0070
Tmin = 4.6 °C

0.01

NS

0.00

Flower number

0

E

B

16

80

14
12

60

10
8

40

6

20

18

4

NS
NS

y = 0.247x2 - 12.536x + 188.070
y = -1.711x + 68.580

0

Branch number

4

y = 0.089x2 - 3.807x + 52.221

Stem length (cm)

0.05

Node number

Flowering rate (1/d to flower)

Petunia 'Easy Wave Violet'

2
0

C

10

12

14

16

18

20

22

24

26

16
14
12
10
8
y = -0.338x + 18.660
y = -0.373x + 18.341

6
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.7. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave Violet’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or
−2 −1
14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to
flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

180

0.04

20

D

A

18

y = 0.0018x - 0.0058

16

Tmin = 3.3 °C

0.03

14
0.02

12

y = 0.0017x - 0.0084
0.01

0

Branch number

22

10

y = 0.079x2 - 3.212x + 45.716

B

0

E

14
12
10
8
y = -0.370x2 + 11.974x - 30.726
y = 0.610x2 - 25.321x + 303.970

Node number

0.00
100
90
80
70
60
50
40
30
20

NS

Tmin = 4.9 °C

Stem length (cm)

0.05

Flower number

Flowering rate (1/d to flower)

Petunia 'Easy Wave White'

6

y = -0.425x + 18.683
NS

0

C

10

12

14

16

18

20

22

24

26

20
18
16
14
12
10

y = -0.063x2 + 1.897x + 2.328
y = 0.079x2 - 3.429x + 49.147

8
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.8. The effect of average daily temperature on flowering characteristics of petunia ‘Easy
−2 −1
Wave White’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or
−2 −1
14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to
flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

181

24

A

D

0.05

20
y = 0.0022x - 0.0105
T
= 4.7 °C
min

0.04

16

0.03

12

0.02

8

y = 0.0021x - 0.013
T
= 6.1 °C
min

0.01
0.00

y = -0.325x + 18.415
y = -0.566x + 23.639

B

4
0
13

E

Flower number

50

12
11

40

10
30

9
8

20
NS

10
0

7

NS
NS

y = -0.114x2 + 2.822x + 24.763

6
0

C

10

14

Branch number

Stem length (cm)

0.06

Node number

Flowering rate (1/d to flower)

Petunia 'Shock Wave Coconut'

12

14

16

18

20

22

24

26

12
10
8
y = -0.204x + 14.515
y = -0.331x + 15.601

6
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.9. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Coconut’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

182

0.04

16

D

A

14

y = 0.0019x - 0.0057
Tmin = 3.0 °C

12
10

0.03

8
0.02

6

0.00

Flower number

80

NS

B

2
0

E

y = -0.343x2 + 10.982x - 47.961
y = 0.217x2 - 11.175x + 162.440

70

4

y = 0.098x2 - 3.941x + 45.353

y = 0.0018x - 0.0117
Tmin = 6.4 °C

0.01

16
14

60
50

12

40

10

30

8

20

NS
y = -0.366x + 18.561

10

6

0

0
C

18

Branch number

Stem length (cm)

0.05

Node number

Flowering rate (1/d to flower)

Petunia 'Shock Wave Denim'

10

12

14

16

18

20

22

24

26

15
12
9
6

y = -0.065x2 + 1.981x - 1.313
y = -0.033x2 + 0.381x + 16.366

3
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.10. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Denim’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

183

A

0.05

26
24
22
20
18
16
14
12
10
8
6

D
y = 0.0024x - 0.0104
Tmin = 4.4 °C

0.04
0.03
0.02

y = 0.0022x - 0.0117

NS

Tmin = 5.3 °C

0.01

y = 0.138x2 - 5.141x + 61.310

0.00

0
B

Flower number

45

E

12
11

35

10

30

9

25

16

8

y = -0.032x2 + 0.848x + 6.417
y = -0.042x2 + 1.377x - 0.206

NS
y = -0.684x + 45.938

20

Branch number

13

40

0

Stem length (cm)

0.06

C

10

12

14

16

18

20

22

24

Node number

Flowering rate (1/d to flower)

Petunia 'Shock Wave Purple'

7
0
26

14
12
10
8

y = 0.030x2 - 1.322x + 26.090
y = 0.045x2 - 2.105x + 32.871

6
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.11. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Shock Wave Purple’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

184

35

D

A

30

0.04

25

y = 0.0022x - 0.0144
Tmin = 6.6 °C

0.03

20
15

0.02

0.00

NS

5

y = 0.204x2 - 8.275x + 100.290

B

80

Flower number

10

y = 0.0020x - 0.0145
Tmin = 7.4 °C

0.01

0
20
18
16
14
12
10
8
6
4

E

70
60
50
40
30

NS

20

y = -0.548x + 22.471
y = 0.077x2 - 3.541x + 51.836

y = 0.139x2 - 8.589x + 158.990

0

0
10

15

Branch number

Stem length (cm)

0.05

Node number

Flowering rate (1/d to flower)

Petunia 'Tidal Wave Silver'

12

14

16

18

20

22

24

26

12
9
6
y = -0.038x2 + 0.898x + 8.985
y = -0.049x2 + 1.084x + 8.311

3
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.12. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Tidal Wave Silver’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

185

20

D

0.035
0.030

18

y = 0.0016x - 0.0040
Tmin = 2.5 °C

16

0.025

14

0.020

12

0.015
0.005
0.000

10

y = 0.0015x - 0.0068
Tmin = 4.4 °C

0.010

NS
NS

0

E

B

16

80

14

60

12
10

40

8
20

y = -0.357x2 + 11.705x - 43.295
y = -3.129x + 99.513

0

y = -0.053x2 + 1.483x + 3.479
NS

C

10

12

Branch number

8

Stem length (cm)

A

0.040

12

14

16

18

20

22

24

Node number

0.045

Flower number

Flowering rate (1/d to flower)

Petunia 'Wave Blue'

6
0
26

10
8
6
y = -0.019x2 + 0.360x + 10.403
y = 0.025x2 - 1.296x + 23.066

4
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.13. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Blue’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or 14
−2 −1
mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to flower.
Tmin is the estimated temperature at or below which the rate of progress towards flowering is
zero. Each symbol represents the treatment means, and error bars represent 95% confidence
−2 −1
intervals. Dashed and solid lines represent regression equations for the 7 and 14 mol·m ·d
DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

186

40

A

0.03

D

35

y = 0.0020x - 0.0157
Tmin = 7.7 °C

30
25

0.02

20
y = 0.0018x - 0.0133
Tmin = 7.5 °C

0.01
0.00

15

y = 0.116x2 - 4.584x + 62.161
y = 0.268x2 - 10.842x + 129.590

10
0

E

B

20

Flower number

100

18
16

80

14
60

12
10

40

Branch number

16

8

y = -0.836x + 28.260
y = 0.083x2 - 3.409x + 45.627

y = -4.680x + 144.530
y = -3.788x + 115.910

20
0

Stem length (cm)

0.04

C

10

12

14

16

18

20

22

24

Node number

Flowering rate (1/d to flower)

Petunia 'Wave Pink'

6
0
26

14
12
10
8
6
4

y = -0.355x + 18.247
y = -0.290x + 13.998

2
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.14. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Pink’ grown under an average daily light integral of 7 mol·m ·d (open symbols) or 14
−2 −1
mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of days to flower.
Tmin is the estimated temperature at or below which the rate of progress towards flowering is
zero. Each symbol represents the treatment means, and error bars represent 95% confidence
−2 −1
intervals. Dashed and solid lines represent regression equations for the 7 and 14 mol·m ·d
DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

187

50

A

D

45

y = 0.0017x - 0.0098
Tmin = 5.7 °C

0.03

40
35
30

0.02

25

0.00

B

NS

0

E

20
18

80

16

60

14
12

40

10
y = -0.450x + 21.534
y = 0.063x2 - 2.625x + 41.152

20
0

8
0

C

10

12

Branch number

15

y = 0.392x2 - 15.624x + 177.980

y = 0.188x2 - 9.564x + 163.720
y = 0.666x2 - 28.309x + 327.740

100

Flower number

20

y = 0.0016x - 0.0107
Tmin = 6.8 °C

0.01

Stem length (cm)

0.04

Node number

Flowering rate (1/d to flower)

Petunia 'Wave Purple Classic'

12

14

16

18

20

22

24

26

10
8
6
y = -0.027x2 + 0.628x + 7.499
y = 0.024x2 - 1.172x + 20.009

4
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.15. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Purple Classic’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

188

0.03

45

D

A

40
35

y = 0.0020x - 0.0123
Tmin = 6.2 °C

30
25

0.02

20
y = 0.0015x - 0.0082
Tmin = 5.6 °C

0.01

15
NS
NS

0.00

Flower number

16

140

14

120

12

100

10

80

8

60

6

40
0
13

4

y = -0.032x2 + 0.651x + 9.430
NS

y = -0.267x2 + 4.679x + 74.587
y = -6.381x + 188.010

20

Branch number

0

E

B

160

10

Stem length (cm)

0.04

Node number

Flowering rate (1/d to flower)

Petunia 'Wave Purple Improved'

2
0

C

10

12

14

16

18

20

22

24

26

12
11
10
9
8
7

y = -0.143x + 12.656
y = -0.262x + 13.649

6
0
10

12

14

16

18

20

22

24

26

Average daily temperature (°C)

Figure 5.16. The effect of average daily temperature on flowering characteristics of petunia
−2 −1
‘Wave Purple Improved’ grown under an average daily light integral of 7 mol·m ·d (open
−2 −1
symbols) or 14 mol·m ·d (solid symbols) Flowering rate was calculated as the reciprocal of
days to flower. Tmin is the estimated temperature at or below which the rate of progress towards
flowering is zero. Each symbol represents the treatment means, and error bars represent 95%
confidence intervals. Dashed and solid lines represent regression equations for the 7 and 14
−2 −1
mol·m ·d DLI treatments, respectively. NS = nonsignificant at P > 0 .05.

189

LITERATURE CITED

190

Literature cited

Blanchard, M.G. and E.S. Runkle. 2011. Quantifying the thermal flowering rates of eighteen
species of annual bedding plants. Scientia Hort. 128:30–37.
Clough, E.A., A. Cameron, R.D. Heins, and W.H. Carlson. 2001. Growth and development of
Oenothera fruticosa is influenced by vernalization duration, photoperiod, forcing
temperature, and plant growth regulators. J. Amer. Soc. Hort. Sci. 126:269–274.
Niu, G., R.D. Heins, A. Cameron, and W.H. Carlson. 2001. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of Campanula
carpatica 'Blue Clips'. Scientia Hort. 87:93–105.
Niu, G., R.D. Heins, A.C. Cameron, and W.H. Carlson. 2000. Day and night temperatures, daily
light integral, and CO2 enrichment affect growth and flower development of pansy (Viola
×wittrockiana). J. Amer. Soc. Hort. Sci. 125:436–441.
Pietsch, G.M., W.H. Carlson, R.D. Heins, and J.E. Faust. 1995. The effect of day and night
temperature and irradiance on development of Catharanthus roseus (L.) 'Grape Cooler'.
J. Amer. Soc. Hort. Sci. 120:877–881.
Pramuk, L.A. and E.S. Runkle. 2005. Modeling growth and development of Celosia and
Impatiens in response to temperature and photosynthetic daily light integral. J. Amer.
Soc. Hort. Sci. 130:813–818.
Roberts, E.H. and R.J. Summerfield. 1987. Measurement and prediction of flowering in annual
crops, p. 17-50. In: J.G. Atherton (ed.). Manipulation of flowering. Butterworths,
London.
Whitman, C.M., R.D. Heins, A. Cameron, and W.H. Carlson. 1997. Cold treatment and forcing
temperature influences flowering of Campanula carpatica 'Blue Clips'. HortScience
32:861–865.
Yuan, M., W.H. Carlson, R.D. Heins, and A. Cameron. 1998. Effect of forcing temperature on
time to flower for Coreopsis grandiflora, Gaillardia ×grandiflora, Leucanthemum
×superbum, and Rudbeckia fulgida. HortScience 33:663–667.

191