Q .
ah ha
‘5.» 9%»

A {1 5,;
’ tygvémw
1.

""ﬁ‘friF
1' IA 5"; v 4;“ ‘ _
I $2.:"‘VTR%E;§I§ £31;
3 ‘- ".FK. v
‘, 5", ﬁgﬁ'lg-‘Hl.-'H L.
mg!" l ﬁfgggi.
é ”iii.

1
'L

,

k

.3} V2“

9' v '
~t 3‘1“":-
z‘c' - ‘t
I '- -

vnge¥:‘”:i“- :.
‘ " l‘dﬁ'éﬁjﬁ? .,.::
V :7: "“f'"; ”q"!
.UA'e-n; ‘Q’, .

‘ 1:”

”T.
‘4‘ ml
jut! ,_‘

mm.
7‘

 

.54. Int.
‘ -. - 47-... .

 

“.
‘ . .. ‘
.:.a~.v.:‘.r ,

’ .
‘ v.1 ’
{ ‘a’ﬁ'ﬂ‘i‘i‘ﬁ' _
. ‘ | a 1_.~
us]. ‘I a; 'P

4

 

“5815

Li” 3’ I‘

J MlCHlGAN STATE UNIV RSITY LIBRARIES

llllll ill \lllllllllll‘li \ l i ll H l ill

 

 

 

 

 

 

 

 

 

ii

3 1293 01409 6246

This is to certify that the
thesis entitled
Modeling Leaf and Inflorescence Development
of the African Violet (Saintpaulia ionantha)
presented by

James Emerson Faust

has been accepted towards fulfillment
of the requirements for

M.S. degree in Horticulture

/ §
x/77K/L/‘7 Am ﬂat/2 1/11
f ' /

Major professor

Date February 26, 1992

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

. ._,_i_._—q'—~.-r --v
Wm?“

._ _ '._q_.,-o_—-

m...— ‘r' d-f'“

AﬂH—I—h-r u

 

 

LIDRARY
Michigan Ettate
University

 

 

 

PLACE ll RETURN BOX to romovo this clackout from your rooord.

TO AVOID FINES return on or baton «to duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wan-9.1

MODELING LEAF AND INFLORESCENCE DEVELOPMENT

OF THE AFRICAN VIOLET (Saintpaulia ionantha WENDL.)

BY

James Emerson Faust

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Horticulture

1992

ABSTRACT

MODELING LEAF AND INFLORESCENCE DEVELOPMENT

OF SAINTPA ULIA IONANYHA WENDL.

By

James Emerson Faust

Models were developed to describe African violet leaf and inﬂorescence
development. Nonlinear functions were used the describe the inﬂuence of temperature
and daily integrated PPF on the rate of leaf unfolding. Time to anthesis was predicted
by integrating two models. The ﬁrst model predicted the time when a reproductive bud
became macroscopically visible in the leaf axil. The second model predicted the time
for an inﬂorescence to develop from visible ﬂower bud to anthesis. The leaf and
inﬂorescence development models were validated in greenhouse experiments with a range
of temperatures from 15 to 30C and daily integrated PPF treatments from 2 to 10
mol m‘2 d“ . The leaf development model predicted leaf number within 11 leaf for 85 96
of observations during an 11 week time period. The inﬂorescence development model

predicted anthesis within i5 days for 71% of the observed inﬂorescences.

DEDICATION

In the country all was dead still. Little stars shone high up; little
stars spread far away in the ﬂood-waters, a ﬁrmament is roused and
stirred for a brief while by the day, but which returns, and will remain at
last eternal, holding everything in its silence and its living gloom. There
was no Time, only Space... Where was he? One tiny upright speck of
ﬂesh, less than an ear of wheat lost in the ﬁeld. He could not bar it. On
every side the immense dark silence seemed pressing him, so tiny a spark
into extinction, and yet, almost nothing, he could not be extinct. Night,
in which everything was lost, went reaching out, beyond stars and sun.
Stars and sun, a few bright grains, went spinning round for terror, and
holding each other in embrace, there in a darkness that outpassed them all,
and left them tiny and daunted. So much, and himself, inﬁnitesimal, at
the core of nothingness, and yet not nothing.

D.H. Iawrence
To Paul Morel, who had the courage to turn and to walk towards the city’s gold

phosphorescence.

To Kimberley Brown and to Bob Freeman, who believed in me and also gave me

courage to turn and to begin to walk in a new direction.

ACKNOWLEDGEMENTS

I would like to thank Dr. Royal Heins for always being available and willing to
discuss my work and for the insight which he continually provided.

I wish also to thank Drs. John Biernbaum and Ken Poff for serving on my
academic committee.

I owe my sanity during the last three years to my wife, Kimberley, who was an
invaluable source of strength and support, and to my friend, Mark Yelanich, who always

managed to keep life at the ofﬁce amusing.

iv

Guidance committee:
The paper format was adopted for this thesis in accordance with departmental and

university regulations. Section I and II are to be submitted to the W
S . EH . l l S . .

TABLE OF CONTENTS

LIST OF TABLE ..................................... ix
LIST OF FIGURES .................................... x
LITERATURE REVIEW Inﬂuence of temperature and irradiance on growth
and development ................................. 1
The African violet ................................ 1
Background ................................ 1
Commercial production ......................... 2
Inﬂuence of the greenhouse environment on growth and
development ............................... 4
Temperature ............................. 5
Irradiance .............................. 6
Carbon dioxide ........................... 9
Relative humidity .......................... 10
Leaf development ................................. 11
Phonology ................................. l 1
Temperature ............................... l4
Irradiance and Photoperiod ...................... 15
Flower development ............................... 16
Temperature ............................... 19
Irradiance ................................. 21
Temperature and Irradiance ...................... 25
Modeling plant development .......................... 26
Degree-day models ........................... 26
Polynomial models ........................... 29
Nonlinear models ............................ 31
Analysis of ﬂuctuating temperature data .............. 33
Plastichron index ............................ 36
Literature cited .................................. 39

vi

SECTION I Modeling leaf development of the African violet

(Saintpaulia ionantha Wendl.) ......................... 49
Abstract ...................................... 50
Materials and Methods ............................. 52
Results ....................................... 57
Discussion ..................................... 5 8
Literature cited .................................. 61

SECTION II Modeling inﬂorescence development of the African violet

(Saintpaulia ionantha Wendl.) ......................... 79
Abstract ...................................... 80
Materials and Methods ............................. 83
Results ....................................... 91
Discussion ..................................... 93
Literature cited .................................. 96

vii

LIST OF TABLES

Table Page
LITERATURE REVIEW

1. Phenological scale describing the development of the sunﬂower
(Schneiter and Miller, 1981) .......................... 15

2. Comparison of Tm, To”, and the interval between the appearance
of leaves at T0,, for several species. ...................... 15

3. The daily integrated PPF at which ﬂower initiation and development
cease to occur for four species ......................... 22

4. Variables used to predict development during different stages of
soybean development (Snyder and Bunce, 1983) .............. 34

5. The rates of development from sowing to ﬂower initiation for seven
treatments. Columns Tl to T, represent the percentage of time each
treatment received the indicated temperature (McPherson et a1. , 1985) 35

SECTION I
1. Inﬂuence of temperature and daily integrated PPF on LUR ....... 75
2. Parameter estimates and 95 % conﬁdence intervals calculated for use
in the LUR model (Eqs. 1-5) .......................... 76
3. Comparison of four methods of using temperature data to predict leaf
number ....................................... 77
SECTION II

1. Phenological description for each stage of inﬂorescence development . 122

viii

Parameter estimates for nonlinear (LER) and polynomial (LBLVB)
submodels of the time to VB model ...................... 123

Parameter estimates for the linear functions describing the number of
days required for each stage of inﬂorescence development ........ 124

The predicted number of days for an inﬂorescence to develop to

anthesis based on the current stage of inﬂorescence development and
temperature (Eq. 6). .............................. 125

ix

Figure

10.

LIST OF FIGURFS

Page

LITERATURE REVIEW

Development of a wheat leaf (leaf number 7) in 0.1 units, 7.0 to 7.9.
The unit designation is determined by the approximate length of the
seventh leaf (on left) relative to the sixth leaf (at right) (Haun, 1973) . 13

Cumulative development of four plantings of Spring wheat
(Haun, 1973) ................................... 13

Developmental stages of a reproductive apex of a rose. Cut surfaces
are shown stippled (Horridge and Cockshull, 1974) ............ 17

Developmental stages of Chrysanthemum ﬂowers. Stages 8 to 10 are
drawn at half the scale of stages 3 to 7 (Cockshull and Hughes, 1972) 18

Bud development meter as computed from a regression equation:
Days to ﬂower = 33.258 - 2.039 * bud length - 0.736 * Temp. +
0.044 * Temp. " bud length (Healy and Wilkins, 1984) ......... 18

Temperature regime for programming tulip bulbs for an early March
sale date (Dehertog, 1989) ........................... 19

Comparison of temperature responses for leaf and ﬂower
development of Chrysanthemum (Karlsson et al., 1989;
Karlsson et al., 1990) .............................. 20

The interaction of temperature and photoperiod on ﬂower
development of Begonia x chiemantha (Heide, 1969) ........... 25

A degree—day model relating rate of deve10pment as a linear function
of temperature ................................... 27

Degree-day model which compensates for reduced development rate at
supraoptimal temperatures (Cross and Zuber, 1972). ........... 28

11.

12.

13.

14.

15.

16.

17.

Four methods of ealculating the accumulated heat units by means of a

sine curve. In each case, the horizontal axis represents 24 hours, the
vertieal axis represents temperature, and the hatched area represents

the accumulated heat units (Baskerville and Emin, 1969) ......... 30

Polynomial function (Y = 0.089 - 0.0237 * Temp. +0.00189 *
Temp.2 - 0.00003177 *Temp.’) describing the LAR of Hibiscus as a
function of temperature (Karlsson et al., 1991) ............... 30

Nonlinear model used to describe rate of development based on four
biologically signiﬁcant parameters Tm, To", Tm, and Ym
(Reed et al., 1976; Landsberg, 1977). .................... 31

A nonlinear model describing the interaction of temperature and
photoperiod on the relative rate of development (Eqs. 5 and 6)
(Coligado and Brown, 1975) .......................... 33

Observed mean rates of development from sowing to ﬂower initiation
of Pigeonpea plotted against mean temperature for each treatment

(McPherson et al., 1985) ............................ 35

Fractional rates of development from sowing to ﬂower initiation of

Pigeonpea (McPherson et a1. , 1985) ..................... 36

An idealized representation of the relationship between leaf length

and time for a single shoot (Lamoreaux et al., 1978) ........... 37
SECTION I

The rate of leaf unfolding for an African violet plant grown at 22C

and a daily integrated PPF of 7 mol m'2 d‘l was 0.244 leaves per day

as determined by the slope of the regression line

Y=0.244*X-0.302 (Rz=.99) .......................... 64

Nonlinear model (Eqs. 1-5) describing LUR as a function of

temperature and daily integrated PPF (R2 = .99). Symbols represent
treatment means .................................. 66

xi

The inﬂuence of daily integrated PPF on A) Top, (Top. = 25.44 -

3.127 * EXP(-0.193 * PPFDQ) and B) LUR.“ (LURm = 0.266 -

0.137 * EXP(-0.418 * PPFDQ). Vertical bars indicate asymptotic

95 % conﬁdence intervals for T0,, and LUR,“ values estimated at

each daily integrated PPF treatment ...................... 68

A comparison between air and plant temperature (A) over the course

of a cloudy day (December 30, 1990), (B) a sunny day

(January 3, 1991), and (C) a cloudy day in which the plants were

growing under high pressure sodium lamps from 0600 to 1800 HR
(December 30, 1990) .............................. 70

Comparison between the predicted (solid line) and observed (0) leaf
number of plants grown in 12 temperature/PPF treatments. ....... 72

Comparison of the observed (0) and the predicted (solid line) leaf
number of all of the plants in the 12 temperature and irradiance
treatments during the 77 days of the greenhouse validation experiment 74

SECTION II

Phenology of the inﬂorescence during development. Each drawing
corresponds to a stage of development described in Table l ....... 99

Correlation between LBL and the percentage of leaf axils forming an
inﬂorescence with respect to leaf number .................. 101

The inﬂuence of daily integrated PPF on the percentage of leaf axils
forming an inﬂorescence with respect to leaf number ........... 103

The number of days to ﬂower shown as a function of average daily
temperature. Data points and error bars represent treatment means
and 95% conﬁdence intervals. ......................... 105

The slepe the regression line (Y = 0.913 * X + 3.68) represents the
rate of leaf extension over time during the ﬁrst 40 mm of leaf
extension for an individual leaf (R2 =0.99) ................. 107

The rate of leaf extension shown as a function of temperature. Data

points represent the treatment means after data from the irradiance
treatments at each temperature were pooled, and error bars display

the 95 96 conﬁdence intervals .......................... 109

xii

10.

11.

12.

A model describing the inﬂuence of daily integrated PPF and
temperature on LBL", .............................. 111

Surface response of the predicted number of days to VB as inﬂuenced
by temperature and measured LBL. ...................... 113

The inﬂuence of temperature on the number of days for an

inﬂorescence to develop through each stage of development. Symbols
represent treatment means and error bars represent the 95 96

conﬁdence intervals. ............................... 115

The observed A) Primary bud diameter and B) inﬂorescence length

at each stage of inﬂorescence development. Data from all temperature

and irradiance treatments were pooled, so data points and error bars
represent the mean values and 95 % conﬁdence intervals for all

treatments at each developmental stage. ................... 117

Comparison between the observed (0) and the predicted (solid line)

number of days from A) the initial leaf measurement to the time of
appearance of VB in the leaf axil, B) the number of days from VB to
anthesis, and C) the number of days from the initial leaf measurement

to anthesis. ..................................... 119

Distribution of the deviation between the observed and predicted

number of days from A) the initial leaf measurement to the time of
appearance of VB in the leaf axil, B) the number of days from VB to
anthesis, and C) the number of days from the initial leaf measurement

to anthesis. ..................................... 121

xiii

LITERATURE REVIEW

LITERATURE REVIEW

Inﬂuence of Temperature and Irradiance
on Plant Growth and Development

The African violet
Background. Saintpaulia ionantha is native to the East African countries of
Kenya and Tanzania which are located at 5 degrees south latitude. The average daily
temperature in these regions ranges from 24 to 28C, while 0.25 cm of rainfall occurs on
approximately 120 days per year. Distribution is limited to a few mountain ranges and
a portion of the coastal plain where Saintpaulia is found growing on rocks and steep
surfaces next to the rain forest epiphytes. Therefore, the natural habitat of the African
violet includes relatively warm temperatures, a low PPF, a 12 hour photoperiod, high
relative humidity, and frequent watering in a well-drained substrate. (Johansson, 1978)
The African violet was ﬁrst taken from Africa to Europe in the late 1800’s.
The genus was named Saintpaulia by Hermann Wendlan to commemorate the German
baron St. Paul von Illaire. Seed was imported into the United States by 1927, and the
ﬁrst plants were sold in 1936 (Kimmins, 1980). By the 19403, Saintpaulia had become
a popular houseplant (Wilson, 1951), and scientiﬁc research had been initiated

(Poesch, 1943; Elliott, 1947). In 1990, 23 million pots were sold in the United States

2
at a total wholesale value of more than 27.6 million dollars (Anonymous, 1990). The

success of Saintpaulia has been due to the relative ease with which it can be grown both
in a commercial greenhouse and in a home environment, and also the many unique
ﬂower and foliage types which both professional and amateur breeders have developed.

The African violet has a rosette growth habit with the leaves growing in a whorl
on a compact stem. The leaves have been bred to display many different shapes which
range from ovate to undulate (Wilson, 1951). The apical meristem grows
indeterminately while the axillary meristems can differentiate into either vegetative shoots
or reproductive inﬂorescences. The ﬁve-petaled ﬂowers can be various shades of violet,
pink, white, or a mixture of these colors, termed bi-color.

Commercial production. Commercial production typically begins with the
pmpagation of leaf cuttings (von Hentig, 1976). Propagation can also be accomplished
by seed or tissue culture. Seed propagation is not commonly used because most of the
varieties do not come true to seed. Much research has been done on micropropagation
(Vazquez and Short, 1978; Start and Cumming, 1976; Bilkey and Cocking, 1981;
Cooke, 1977); however, micropropagation has not had much commercial signiﬁcance due
to economic limitations.

Stock plants supply the leaf cuttings for propagation. Leaves are removed after
they are at least 3 cm long. The petioles are removed, and the leaves are inserted
1 to 2 cm deep into a soilless medium. Roots form endogenously from cells lying
between the leaf traces, followed by shoots which develop exogenously from the
epidermal cells (Naylor and Johnson, 1937). Anywhere from 1 to 10 plantlets will

develop from one leaf cutting (Sanderson and McGuire, 1988). After the plantlets have

3

emerged from the medium, the mother leaves, i.e. the leaves from the cuttings, are torn
off at the medium surface in order to allow more light to reach the developing plantlets.

Growth regulators, speciﬁcally cytokinin (PBA) and Gibberellic acid together,
applied to developing shoots 1 cm long resulted in a 43% increase in the number of
transplantable plantlets which developed from each leaf cutting (Sanderson and
McGuire, 1988). Use of either chemical is not practiced commercially.

Plantlets are separated when they have reached sufﬁcient size for transplanting,
i.e. the stem diameter is 3 mm or greater, and 4 to 5 leaves have unfolded. Two or
three leaves are removed from each plantlet to allow tight spacing between the plug cells.
The roots are then removed in order to increase the ease with which the plantlets are
planted and also to increase the uniformity of development between plants. The plantlet
is inserted into an individual cell (22 cm’), whereupon the roots regenerate from the stem
in about one week. Rooted plantlets are termed plugs. After approximately 8 weeks,
the plugs which have approximately 10 unfolded leaves are transplanted into the ﬁnal
growing container, most typically a 10 cm diameter pot (450 cm’).

Many growers purchase plugs to avoid propagation. The ﬂowers usually begin
to appear in the leaf axils near the time when the plants are transplanted from the plug
to the 10 cm pot. The plants will ﬂower and be ready for sale in 7 to 10 weeks. A
saleable plant will typically have 20 leaves and 2 to 3 inﬂorescences each with one or
more open ﬂowers. The wholesale value is approximately $1.25 (Personal
communication, Post Gardens Greenhouse). Total production time from leaf cutting to

sale requires 8 to 10 months.

4
Inﬂuence of the greenhouse environment on growth and development. Research

on the African violet has focused on identifying the inﬂuence of the greenhouse
environment on plant morphology and growth. Environmental factors examined in this
review include temperature, irradiance, carbon dioxide, and relative humidity.
Morphological data usually includes, leaf size, leaf number, ﬂower number, inﬂorescence
number, and ﬂower number per inﬂorescence. Growth is measured as total plant dry
weight and/or fresh weight accumulation.

Temperature. Went concluded in his classical study of plant growth (1957), that
optimal growth, indicated by plant height or time to ﬂower, occurred for most plants
with a day temperature (DT) 5C higher than the night temperature (NT). However, an
exception was noted for Saintpaulia ionantha which ﬂowered earliest at a DT of 14C and
NT of 20 to 23C. This observation has since been published in many botany and
horticulture texts (Leopold and Kriedemann, 1975 ; Larson, 1980; Mastalerz, 1977).
While Dvorska (1979) claimed to conﬁrm Went’s observation, no data were presented
to support the conclusion.

Hildrum and Kristoffersen (1969) examined the effect of DT, NT, and
photosynthetic photon ﬂux density (PPF) on African violet by measuring plant fresh
weight and the number of ﬂowers produced per plant. They found that growth and
ﬂowering were a function of average daily temperature (ADT). Fresh weight and ﬂower
number increased as temperature was increased from 15 to 24C, and then decreased as
temperature was increased further to 27C. Decreasing either the DT or NT resulted in

decreased fresh weight and ﬂower number.

5

Went (1957) also observed an interaction between temperature and illuminance;
i.e. as illuminance increased, the optimal temperature for growth and ﬂowering
decreased. Hildrum and Kristofferson (1969) also reported an interaction between
temperature and illuminance. They found that more ﬂowers were produced on plants
grown at an air temperature of 24 to 27C when the illuminance level was 4 to 8 klx for
16 h (3.1 and 6.2 mol in2 d") from cool white ﬂuorescent (CWF) lamps, then on plants
grown under 12 klx for 16 h (9.3 mol rn'2 d"). The optimal temperature for ﬂowering
was 21 to 24C, while ﬂowering was inhibited at 27C. However, plant temperature was
not reported. It is possible that sufﬁcient heating from thermal radiation occurred on
plants grown at the 9.3 mol m2 (1'1 treatment that the plant temperature was raised above
the optimal temperature, which resulted in decreased ﬂower number.

Vogelezang (1988) used the African violet to study the effect of root-zone heating
on inﬂorescence fresh weight and ﬂower number per plant. She observed that elevating
root-zone temperature from 17 to 25C increased inﬂorescence fresh weight and ﬂower
number. However, her data are inconclusive with respect to determining whether the
observed increase in growth was due to the increased root temperature or the increased
stem or leaf temperature. Plant meristem temperature was more closely correlated to soil
temperature than to either leaf or air temperature. Saintpaulia is a species which has a
relatively large stem diameter and a meristem usually less than 4 cm from the soil
surface; therefore, conductive heat transfer between the stem and the soil can
signiﬁcantly affect meristem temperature and consequently plant development rate.

Being of tropical origin, the African violet is sensitive to low temperature and

chilling injury. Larcher and Neuner (1989) determined the threshold temperature for

6
chilling injury in African violet to be between 9 and 10C. Bodnar and Larcher (1987)

showed African violet organs varied in their degree of sensitivity to cold temperature
(6C for 9 days). The leaves were the most susceptible organ, while the petioles, stem,
and shoot apex were less susceptible. Different tissues in the leaves also varied in their
susceptibility to chilling. The palisade parenchyma cells were the most susceptible while
the spongy parenchyma, upper epidermis, and lower epidermis were less susceptible.
Pollen were much less sensitive than the ovules in the ﬂowers. Preconditioning of the
African violet at 15C DT and 11C NT for one week resulted in a 30-40% decrease in
the amount of irreversible damage to plants placed at 6C for 9 days.

”Ring spot” is the appearance of circular white Spats on the foliage of African
violets. Poesch (1943) showed cold water being splashed on warm leaves to be the eause
of ”ring spot”. Elliott ( 1947) conﬁrmed Poesch’s report and concluded that water which
was 5 to 10C colder than the leaf temperature would cause the collapse of the palisade
parenchyma cells in the leaf. The other leaf tissues remain unaffected.

Irradiance. The African violet is a shade plant (Johansson, 1978); therefore, the
acceptable irradiance level for Saintpaulia is lower than for many other greenhouse
crops. Research has focused on the importance of light on the ﬂower development of
African violets. Stinson and Laurie (1954) microscopically examined crosssections of
leaf axils of plants grown at a range of irradiance levels over a 3 month period of time
to determine the minimum irradiance level required for ﬂower initiation and
development. They found that ﬂoral organs failed to initiate and/or fully develop on
plants grown with less than 500 footcandles (100 umol m" 3") during the peak light

intensity of a day in October through December.

7

Hanchey (1955) measured the number of ﬂowers, leaves, and inﬂorescences on
plants grown under CWF lamps which delivered 100, 300, or 600 footeandles
(15, 44, and 88 umol m’2 s") for 6, 12, or 18 h (0.32 to 5.7 mol rn'2 d“). The results
indieated that ﬂower number, leaf number, and the number of inﬂorescences per plant
were a function of the daily integrated PPF. The manner in which the irradiance was
delivered was not critieal. The number of leaves per plant and the number of ﬂowers
per inﬂorescence increased as the daily integrated PPF increased up to 3.6 mol m2 d‘.
Increasing the daily integrated PPF above 3.6 mol in2 (1'1 did not result in any further
increase in leaf number or the number of ﬂowers per inﬂorescence. The number of
inﬂorescences per plant and the number of ﬂowers per plant increased as daily integrated
PPF was increased up to 5 .7 mol m'2 d’1 . In an experiment conducted under natural
irradiance, daily integrated PPF above an estimated 5 .2 mol m“2 d" resulted in decreased
plant size, e.g. shorter petioles and smaller leaves.

Hildrum and Kristoffersen (1969) observed that fresh weight increased as the
illuminance from CWF lamps increased from 4 to 8 klx for 16 h (3.1 to 6.2 mol m’2 d"),
but increasing the illuminance to 12 klx for 16 h (9.3 mol m'2 d“) did not result in any
further increase in fresh weight. However, ﬂower and bud number increased from
23 to 39 per plant as the daily integrated PPF increased from 3.1 to 9.3 mol m"2 d".

Mortensen (1983) found that increasing the PPF from 38 to 100 umol rn’2 s" for
16 h (2.2 to 5.5 mol rn'2 (1") resulted in increased dry weight, leaf number, and ﬂower
number per plant.

Light quality inﬂuences plant growth and morphology. Cathey et a1. (1978)

examined the effects of seven different ﬂuorescent lamps on African violets. Lamp type

8
signiﬁcantly affected fresh weight, but did not affect leaf length or node number.

Irradiance at canopy level for the different lamps ranged from 14 to 23 W m‘2
(400 to 700 nm) for 16 h (3.7 to 5.8 mol m2 d"). However, daily integrated PPF did
not appear to affect fresh weight or leaf length. Fresh weight was signiﬁcantly affected
by the lamp type. The number of days to ﬂower decreased with the addition of
incandescent lighting with CWF lamps. However, thermal barriers were not placed
above the plants in this experiment; therefore, the increased amount of long-wave
radiation due to the incandescent lamps may have increased plant temperatures. Hanchey
(1955) found that the number of ﬂowers per inﬂorescence was higher, total plant width
was longer, and fewer vegetative lateral shoots developed on plants grown under CWF
lamps compared to sunlight.

Conover and Poole (1981) placed ﬂowering African violets into environments
with illuminances of 0.5, 1.0, and 2.0 klx for 12 h (0.3, 0.6, and 1.2 mol m’2 d")
supplied by CWF lamps. All of the plants in the 0.3 and 0.6 mol rn'2 d“ treatments had
stopped ﬂowering after three months, while 38% of the plants in the 1.2 mol rn'2 d’l
treatment were still ﬂowering. After 9 months, 3, 62, and 100% of the plants were
ﬂowering in the 0.3, 0.6, and 1.2 mol in2 (1'1 treatments, respectively. Conover and
Poole concluded that the African violet has the ability to acclimate and ﬂower under
illuminances as low as 1.0 klx for 12 h (0.6 mol rn'2 d") and to continue vegetative
growth even at illuminances as low as 0.5 ldx for 12 h (0.3 mol m’2 (1"). Different rates
of nitrogen, 0, 6, 12, and 24 mg per pot, were applied to the plants in the post-harvest

environment; however, nitrogen rate did not inﬂuence leaf or ﬂower development.

9
High PPF ean eause destruction of chlorophyll in the leaves of African violets.

Carpenter ( 1969) measured chlorophyll content of plants grown under shade cloth
(maximum illuminance was 1500 footcandles, or 300 umol m’2 s") and direct sunlight
(maximum illuminance was 6685 footcandles, or 1337 umol rn‘2 s") and measured 30 and
7 mg/ 100 g fresh weight for the respective treatments. He also observed that the
breakdown of chlorophyll and foliar burn increased as leaf temperature increased.
Hildrum and Kristoffersen’s (1969) visual observations conﬁrmed that PPF is the main
factor affecting foliar burn and that higher leaf temperature can accentuate the amount
of foliar burn which occurs. Raabe (1957) claimed that temperature difference between
the leaf and applied water, time of exposure to cold water, the wavelength of light, as
well as the PPF could cause chlorophyll degradation in the African violet. Also Raabe
reported that more highly pigmented varieties were more resistent to high light-induced
foliar burn; however, data were not presented. The PPF at which chlorophyll breakdown
begins to occur in the African violet has not yet been identiﬁed.

Carbon dioxide. Greenhouses can become depleted of carbon dioxide especially
during sunny winter conditions when photosynthesis rate is high and greenhouse air is
not being exchanged with outside air (Mastalerz, 1977). Enriched carbon dioxide
environments have been shown to increase plant dry weight gain (Owen et al. 1926;
Wittwer and Robb, 1964). Therefore, supplementary carbon dioxide can be added to the
greenhouse air to improve plant growth.

Mortensen (1983) studied the effect of carbon dioxide enriched air on African
violet and found that increasing carbon dioxide concentration from 330 to 600-900 pl ’ 1‘

while the daily integrated PPF was maintained at 3 mol at2 (1" resulted in earlier

10

ﬂowering, more ﬂowers, more ﬂower buds, and dry matter production was increased
19.7 to 38.6%. Carbon dioxide concentrations above 900 pl 1“ gave no additional effect
on growth. Mortensen (1986) also found that continuous enrichment with carbon dioxide
was superior to intermittently supplying carbon dioxide.

Although Afriean violets respond strongly to an enriched carbon dioxide
atmosphere, the measured beneﬁts in dry weight gain and earlier ﬂowering do not reﬂect
the loss of perceived plant quality. Speciﬁcally, carbon dioxide concentrations above
600 pl 1'1 result in thicker and more brittle leaves. Brittleness increases the amount of
damage which occurs to the plants during shipping and handling. Therefore, a carbon
dioxide concentration of 600 pl 1“ is recommended for greenhouse production
(Fischer, 1989).

Relative humidity. Mortensen (1986) observed that increasing the relative
humidity from 55-60% to 90-95% resulted in a 17 to 36% increase in dry weight. The
increase in dry weight was due to an increase in leaf number and leaf size. Time to
ﬂower was decreased by 2 to 3 weeks, the number of ﬂowers and ﬂower buds per plant
increased by 33%, and dry weight increased 20% as relative humidity was increased.
Similar results have been identiﬁed on many other species: Chrysanthemum, Boston fern,
rose, Hiemalis begonia (Mortensen, 1986), tomato (Mitchell and Hoff, 1977), lettuce

(Tibbitts and Bottenberg, 1976), kale and sugar beet (Ford and Thorne, 1973).

Leaf Development.
Phenology. Development refers to qualitative changes in plant structure, such as

the differentiation of vegetative and reproductive tissue. Plant development can be

11
divided into different developmental stages which are identifiable by morphological

characteristics, such as the third-leaf stage or visible ﬂower bud. Phenology is the study
of these developmental stages as inﬂuenced by the environment and genotype.

Phasic development scales of speciﬁc crops have been created to separate the
different phases of crop development. Scales, such as for the sunﬂower (Table l) and
for wheat (Figures 1 & 2), were developed to identify the status of phenological
development and to provide a numerical deﬁnition to the growth occurring between
stages of development. Phenological scales have been used extensively to study the
effect of the environment on plant development (Baker et al., 1986; Boone et al. , 1990;
Cudney, 1989).

Phenological scales are usually more useful for studying the development of
container-grown ornamental plants than quantitative analysis of growth, i.e. fresh and dry
weight accumulation. The measure of a container-grown ornamental plant is more
closely determined by phenological description, while quantitative analysis of growth is
more useful in modeling the growth of crops in which a portion of the plant is harvested,
e.g. agronomic crops, vegetables, and cut ﬂowers.

Phenological models are usually based upon studies of developmental stages as
isolated processes, i.e. the assumption is made that the preceding phase has no effect on
the current stage of development. However, Karlsson et al. (1989) showed that high
and/or low temperatures occurring during early stages of chrysanthemum development

delayed development in later stages.

12

Table l. Phenological scale describing the development of the sunﬂower

(Schneiter and Miller, 1981).

 

Sunﬂower (Helianthus annuas L.)

 

 

Growth stage Description
Vegetative Emergence: ﬁrst true leaf blade < 4 cm long
VE First true leaf 4 cm long
V1 Second leaf
V2 Third leaf
V3 n number of leaves
Reproductive
R1 Inﬂorescence surrounded by immature bracts becomes
visible
R2 Intemode below base of inﬂorescence elongates 0.5 to
2.0 cm above leaves
R3 Growth of intemode below reproductive bud lifts
inﬂorescence > 2 cm above leaves
R4 Inﬂorescence begins to open
R5 Beginning of anthesis
R6 End of anthesis
R7 Back of inﬂorescence starts to turn light green
R8 Back of head is yellow but bracts remain green
R9 Physiological maturity; bracts become yellow and

brown

 

The optimal temperature for deve10pment can be unique for each stage of

development. Karlsson et al. (1989) identiﬁed the optimal temperature for

chrysanthemum to be 23.1C from the time of disbud to ﬂower bud color and 19.1C from

ﬂower bud color to anthesis.

 

Figure 1. Development of a wheat leaf (leaf number 7) in 0.1 units, from 7.0 to 7.9.
The unit designation is determined by the approximate length of the seventh leaf (on left)
relative to the sixth leaf (at right) (Haun, 1973).

ELONGATION OF CULM

[NRC-ENC! OF SPIKE

GROW! H UNIIS

 

Figure 2. Cumulative development of four plantings of Spring wheat (Haun, 1973).

l4
Angus et al.(1981) showed that the base temperature in wheat was 2.6C from sowing to

emergence and 8.9C from anthesis to maturity. The leaf appearance rate (LAR) of maize
increased from 0.29 to 0.42 leaves (1'1 as the plant age increased from 35 to 43 days after
emergence to 43 to 50 days after emergence (Tollenaar, 1984). Karlsson et al. ( 1988)
did not observe any change in LAR during the development of Easter lily; however, the
LAR of Hibiscus (Karlsson et al., 1991) and wheat (Boone et al., 1990) decreased as the
species changed from vegetative to reproductive development.

Temperature. Temperature is the primary variable used to predict rates of
development (Kiniry et al. , 1991), while photoperiod (Major et al. , 1978), light intensity
(Friend et al. , 1962), water stress (Hodges and French, 1985), and nutrition
(Snyder and Bunce, 1983) have also been included in plant development models.

Vegetative development is often quantiﬁed by measuring LAR. LAR refers to
the reciprocal of the number of days required for one leaf to visually appear, or unfold,
at the apical shoot. LAR varies considerably between species (Table 2) and also between
cultivars of the same species (Tollenaar et al., 1984; Snyder and Bunce, 1983).

LAR has a base temperature or threshold at which increased temperature results
in a linear increase of LAR until an optimum is reached at which point a further increase
in temperature results in a rapid decrease in LAR. The highest rate of development for
any average daily temperature occurs at a constant day and night temperature
(Erwin and Heins, 1990; Coligado and Brown, 1975). Fluctuating temperatures result
in a slower rate of development whenever temperatures occur outside of the linear
temperature range (Erwin and Heins, 1990). Development results from the accumulation

of responses to small time intervals. As a result, more frequent measurements

15

Table 2. Comparison of Tm, To”, and the interval between the appearance of leaves at
T0,.

 

 

 

 

Species Tm. T0,, Interval between leaves appearing at To,
(draw)
Banana 8 27 10.3 (Allen et al., 1988)
Hiemalis Begonia 11 22 6.2 (Faust and Heins, 1990)
Hibiscus 8 31 4.3 (Karlsson et al., 1991)
Wheat 0 25 4.0 (Baker et al., 1986)
Maize 8 32 1.6 (Watts, 1972)

Chrysanthemum 0 30 1.8 (Karlsson et al., 1989)
Sunﬂower - - 1.1 (Rawson and Hindmarsh, 1982)
Sugar beet 1 20+ 1.7 (Milford et al., 1985)

Easter lily l 30 0.4 (Karlsson et al., 1988)

 

of the plant environment should be more useful in predicting development over a longer
period of time (Karlsson et al. 1991). However, Gilmore and Rogers (1958) and Cross
and Zuber (1972) did not observe any improvements in predicting development using 3
hour, and hourly temperatures in degree-day models. The temperature range provided
during these experiments may not have frequently risen above the linear temperature
range.

Irradiance and Photoperiod. The results from investigations of the effect of PPF
and photoperiod on LAR have varied between experiments. The LAR of sunﬂower
decreased 11% as solar radiation was decreased by 50% (Rawson and Hindmarsh, 1983).
The LAR of maize increased up to 20% at 18C as PPF was increased from 27 to
36 mol in2 d“; however, no change in LAR was observed when the same experiment was

conducted at 28C (Warrington and Kanemasu, 1983). LAR of wheat

16
(Friend et al. , 1962) did not change as photoperiod increased from 8 to 16 hour and a

similar daily integrated PPF was provided to both treatments.

Experiments conducted to study the effect of photoperiod often neglect to deliver
- the same daily integrated PPF to the photoperiod treatments which makes it impossible
to separate the effects of light quantity and photoperiod. Aspinall and Paleg (1964) found
no signiﬁcant change in LAR on barley as photoperiod increased from 10 h and a daily
integrated PPF of 6.2 mol m'2 (1'1 to 16 h and a daily integrated PPF of 20 mol m2 d";
however, Cao and Moss (1989) observed a 20 and 33% increase in LAR of wheat and
barley respectively as photoperiod increased from 8 h and a daily integrated PPF of
12 mol m‘2 d‘1 to 24 h with a daily integrated PPF of 35 mol m‘2 d".

The LAR of soybeans grown under an 8 h photoperiod increased 9 % as PPF was
increased from 500 to 1000 pmol m‘2 s'1 (14.4 to 28.8 mol m2 (1"), while the LAR
increased 14 to 33% when PPF was held constant at 500 pmol m'2 s‘1 and the
photoperiod was increased from 10 to 16 h (18 and 28.8 mol m" d") (Snyder and
Bunce, 1983). An increase in photoperiod from 10 to 14 h resulted in a 25% increase

in LAR of banana (Allen et al., 1988); however, PPF was not reported.

Flower development.

Flower induction, initiation, and development can be positively and negatively
inﬂuenced by several environmental factors, including temperature, PPF, photoperiod,
nutrition, and water stress. This review focuses on the effects of temperature and

irradiance on the ﬂowering process.

17
Flower development ean be quantiﬁed by measuring the number of ﬂowers which

develop per inﬂorescence, stem, or plant, or the time required for a plant to develop to
anthesis. The reciproeal of the time to anthesis provides a measure of rate of
development.

Phenologieal sales ean be used to describe the changes which occur to a
vegetative meristem as reproductive organs differentiate (Figure 3). Scales ean also be

used to describe the visible development of the inﬂorescence (Figure 4).

 

Figure 3. Developmental stages of a reproductive apex of the rose. Cut
surfaces are shown stippled (Horridge and Cockshull, 1974).

l8

   

 

      

r l‘ ‘5
MC.
“23‘ 31%

    

Figure 4. Developmental stages of chrysanthemum ﬂowers. Stages 8 to
lOaredrawnathalfthesealeofstages3to7
(Cockshull and Hughes, 1972).

I'Bud meters' provide a method of predicting anthesis as a function of ﬂower bud size

 

 

 

 

 

 

 

 

 

 

andtemperature (Figure5).
— F' 1' 2' 5 s g a m g
3112’) Y"y‘)’ )1 .‘Ja" I I ‘11] l 3 “”1””?
Jazzy mow ,acr/H m z . mum;
/ 233'” 1‘,){/ )3-7/2’ ll‘ éJ_ f WW" 3
23 if rs'ﬂﬂtla'la",r' ,a'/ /ﬁ 15 A za°cns°n§
\1lzsssa1sa'1011121‘31l413m3

 

 

Figure 5. Bud development meter as computed from regression equation:
Days to ﬂower = 33.258 - 2.039 * Bud length - 0.736 * Temp. + 0.044
“ Temp. " Bud length (Healy and Wilkins, 1984).

19
Temperature. Both relatively warm and/or cool temperatures can be required for

ﬂowering. Lilium longiﬂorum bulbs require four to six weeks of 2.5 to 5 .0C
temperatures to provide the vernalization treatment necessary for ﬂower induction
(Lange and Heins, unpublished data).

Commercial Dutch bulb forcers follow a detailed schedule of temperature regimes
to program the ﬂowering of spring crops (Figure 6). The cool temperature treatment for

tulip bulbs is necessary for breaking the dormancy of the partially developed ﬂower buds.

 

—. d
o N
l L

 

 

Temperature (C)
on

 

 

 

 

 

 

 

 

Sept. Oct. Nov. Dec. Jan. Feb. Mar.
Month

Figure 6. Temperature regime for programming tulip bulbs for an early
March sales date (Dehertog, 1989).

The response of ﬂower development to temperature is similar to the response of
leaf development. Below a threshold temperature ﬂower development fails to proceed.

Above the threshold, development increases linearly with respect to temperature until an

20
optimal temperature is reached. Flower development decreases rapidly at supraoptimal

temperatures.
The optimal temperature for ﬂowering can be considerably different than for leaf
development of the same species. The optimal temperature for ﬂower development of

chrysanthemum was 21C, while the optimal temperature for leaf development was 30C
(Figure 7).

 

0.032

O
at

*0.030

0
UI

r0.028

O
.p.

Leaf Development Rate (leaves/day)
o
. .u . . .
Flower Develop Rate(1/days to flowers)

0
01

"0.026

r0.024

.0
N

L°°f\ ~o.022

‘ Flower

O
_a

”0.020

 

 

 

.0
o

I r f I I 0.018
10 15 20 25 3O 35
Temperature (C)

Figure 7. Comparison of temperature responses for leaf and ﬂower
development of chrysanthemum (Karlsson et al., 1989; Karlsson et al.,
1990).

The optimal temperature for ﬂower development may not be the optimal temperature for
the growth of the ﬂoral parts. For example, geranium inﬂorescences developed most
rapidly at 28C, while the largest ﬂower petals were observed at 18C

(Armitage et al., 1981).

21
Time to anthesis is usually described as a function of ADT; however, high NT

ean affect ﬂowering independent of DT. Poinsettia ﬂower initiation was delayed by 21C
NT in one study (Kofranek and Hackett, 1966) and 23C NT in another study
(Berghage et al. , 1987); however, 29C DT did not delay ﬂower initiation or development
(Berghage, et al., 1987).

The sensitivity of ﬂower development to temperature can change during the
course of development. For example, the number of aborted gladiolus ﬂower buds
increased when exposed to high temperature (50C) immediately after the com was
planted or at the appearance of the ﬁfth leaf onward. The ﬂower bud abortion which
occurred during the later stage was increased under a low relative humidity environment
(40-50%) indicating that the apparent effect of temperature may be caused by water
deﬁcit (Shillo and Halevy, 1976).

Harris and Scott (1969) found that development of carnation ﬂowers was a
function of bud temperature when bud and leaf temperature were varied independently.
Flower buds which were 9C warmer than leaf temperature ﬂowered 26 days sooner than
plants grown with the leaves 9C warmer than the ﬂower buds.

Irradiance. The rate of ﬂower development with respect to irradiance follows a
classical asymptotic response. The threshold daily integrated PPF, below which ﬂower
development is signiﬁcantly delayed or completely inhibited, for a number of different
species are shown in Table 3.

Moe (1972) observed on roses that as illuminance was increased from 1.5 to
12 klx for 24 h (1.8 to 14.4 mol rn‘2 d"), the number of days to ﬂower decreased from

67 to 48, the number of ﬂowers per shoot increased from 1.9 to 3.4, and shoot length

22

Table 3. The daily integrated PPF at which ﬂower initiation
and development cease to occur for four species.

 

 

Species Integrated PPF (mol in2 d")
Fuchsia 3.8 (Sachs and Bretz, 1962)

Geranium 1.9 (Armitage et al., 1981)
Azalea 2.3 (Bodson, 1983)

Chrysanthemum 1.1 (Cockshull and Hughes, 1972)

 

decreased from 27 to 20 cm. Moe and Kristoffersen (1968) [also reported that
increasingilluminance from 2 to 10 klx for 16 h (1.5 to 7.8 mol m‘2 d") did not change
the number of petals per ﬂower, while the number of blind shoots decreased by 58 % .
Horridge and Cockshull (1974) suggested that the effect of light on rose development was
that light released the correlative inhibition of the axillary buds. The number of
ﬂowering shoots was directly related to the number of developing lateral shoots;
therefore, high irradiance conditions increased the number of axillary buds which
developed and thus the number of ﬂowers which developed.

Flower abortion occurs when unfavorable environmental conditions, such as low
irradiance or low temperature, occur during ﬂower development. Tomatoes are most
sensitive to low irradiance conditions at the stage of inﬂorescence development when the
peduncle is lengthening and the ﬂoral organs are growing, i.e. the macroscopic
appearance of the inﬂorescence (Kinet, 1977). In chrysanthemum, low daily integrated
PPF during the ﬁrst two weeks of ﬂower initiation caused the greatest delay in ﬂowering,

but once the receptacle and bracts began to development, the inﬂorescence continued to

23
develop, albeit slower, even when plants were grown at daily integrated PPF at which

ﬂower initiation failed to occur (Cockshull and Hughes, 1972).

In geranium, the number of days from germination to visible ﬂower bud
decreased 33% as the PPF increased from 50 to 375 pmol m‘2 s" during an 18 h
photoperiod (3.2 to 24.1 mol in2 d"); however, the same PPF range did not affect the
time from visible ﬂower bud to anthesis. The high PPF possibly induced the ﬂowering
process by expanding the photosynthate pool; after which, temperature controlled the rate
of development (Armitage et al., 1981).

In gladiolus ’San Souci’, reducing the natural irradiance by 25 % resulted in a
27% decrease in the number of ﬂowering plants and 53% decrease in the number of
ﬂorets per spike. The stage of development from the ﬁfth leaf to spike emergence was
the most sensitive to irradiance reduction, i.e. resulted in the greatest reduction in
percentage of ﬂowering plants and number of ﬂorets per spike (Shillo and Halevy, 1976).

Hughes and Cockshull (1971) found that delivering a constant light intensity for
8 h or gradually increasing the light intensity up to a midday high then back down during
an 8 h photoperiod did not inﬂuence the growth or development of chrysanthemum.
However, Kinet (1977) found that the inﬂorescences of tomatoes developed properly
when the plants were grown under an 8 h photoperiod and an irradiance of 1800 erg cm‘2
s“ (2.4 mol in2 d"), while the inﬂorescences nearly all aborted when the plants were
grown under a 16 h photoperiod and an irradiance of 900 ergs cm’2 s" (2.4 mol in2 d“).

Experiments have shown that young leaves can inhibit ﬂowering. A day neutral
variety of tobacco remained vegetative when grown at 46 pmol m‘2 s" , but ﬂowered at

a lower PPF (28 pmol rn‘2 s“) when young leaves were removed (Wardell, 1976). The

24

two youngest leaves of tomatoes grown at 25C utilized a much higher proportion of the
available assimilates than plants grown at 15C; therefore, at higher temperatures the
young leaves more strongly compete for assimilates (Hussey, 1962). It has been
proposed that ﬂower bud abortion in tomatoes is due to the competition for available
photosynthates between vegetative and reproductive development (Calvert, 1969). The
mobilization of leaf carbohydrates appears to be the mechanism for the effect of
irradiance on ﬂowering (Bodson et al., 1977).

Steffen et al. (1988) showed that sink activity and development rate were
unchanged between ﬂowers which had been covered during the entire inductive
photoperiod versus uncovered ﬂowers. Similar experiments with other species, e.g.
roses (Zeislin and Halevy, 1975) resulted in ﬂower abortion, indicating that irradiance
affects more than photosynthesis.

PPF can modify the photoperiodic response of some plants. For example, high
irradiance under SD conditions (96 W m”, full spectrum, for 8 h, or 11.4 mol m‘2 d“)
can eliminate the LD requirement for ﬂowering in Sinapsis (Bodson et al. , 1977).
Bodson (1983) observed that the rate of ﬂower development from ﬂower initiation to the
presence of ovules in the ovary of azaleas was increased as photoperiod was increased
from 8 to 16 hours and also as PPF was increased from 80 to 160 pmol m’2 3". Flower
bud abortion did not occur when plants were grown at 255 pmol m‘2 s" for 8 h
(7.3 mol m'2 d“); however, 4 to 35% of the ﬂower buds aborted on plants grown at 170
to 85 pmol m‘2 s’l for 8 h, respectively (4.8 and 2.5 mol m2 d"). A night interruption
which delivered 1.5 h of incandescent light at 15 pmol of2 s" eliminated the occurrence

of ﬂower bud abortion regardless of the integrated PPF.

25

Temperature and Irradiance. Temperature can inﬂuence the critieal photoperiod
required for ﬂowering. Chrysanthemum ’Encore’ , a short day plant (SDP), required
a shorter night length for ﬂower initiation and a longer night length for ﬂower
development as night temperature was increased from 10 to 27C (Cathey, 1957). Lang
(1965) observed that shorter night periods were required for the long day plant (LDP)
Hyoscyamus as night temperature increased.

As temperature decreased from 24 to 15C the effect of SD on the SDP
Begonia x chiemantha was reduced (Figure 8). The LD treatment can be completely
replaced by low temperature treatments in the LDP Silene (Wellensiek, 1969) and SDP
PeriIIa (Zeevaart, 1969). Temperatures above 30C can replace the requirement for LD

in the LDP Rudbeckia (Mumeek, 1948).

100
90‘
80*
70‘
60‘
50+
401
30*

20 a H Long days
10 . H Short days

 

Days to appearance of flower buds

 

 

 

12 1'5 1B 2'1 2'4 27
Temperature ( C )

Figure 8. The interaction of temperature and photoperiod on ﬂower
development of Begonia x chiemantha (Heide, 1969).

26
Armitage et al. (1981) observed that as temperature increased from 10 to 27C,

the light saturation point for geraniums increased from 700 to 1000 pmol rn'2 s". As
temperature increased from 10 to 32C, the light compensation point increased from
25 to 75 pmol rn'2 s", and dark respiration increased from 2.7 to 9.0 mg CO2 dm‘2 h".

The combination of a low irradiance and high temperature environment is often
associated with ﬂower bud abortion in Easter lily (Mastalerz, 1965). When natural
irradiance is low during winter months, low temperatures are more effective at producing
greenhouse crops. Low temperatures increase the duration of crop production time, thus
allowing for more efﬁcient usage of the low photosynthate supply
(Harris and Scott, 1969).

Flower bud abortion did not occur on iris grown at 18C and 40 W m’2 for 16 h
(10.6 mol m‘2 d"), while 100% ﬂower bud abortion occurred on plants grown at 18C and
40 W m’2 for 8 h (5.4 mol m‘2 (1"). Abortion did not occur at 40 W m'2 for 8 h
(5.4 mol m'2 d") when plants were grown at 12C yet increasing the temperature to 15C
resulted in 88% abortion. Iris ﬂower buds were also reported to be most sensitive to
high temperature and low light conditions during the period of rapid elongation i.e. when

the bud was ﬁrst visible (Fortanier and Zevenbergen, 1973)

Modeling Plant Development

Degree-day models. A quantitative model is created by ﬁtting a linear or a
nonlinear function to data by means of regression analysis. The ﬁrst models used to
predict crop phenology were degree-day models (Bomalaski 1948). Degree-day models

describe rate of development as a linear function of temperature (Figure 9). The base

27
temperature (TIN), is the threshold temperature at which development begins to occur.

Daily maximum (Ti...) and minimum (Tm) temperatures are often used in calculating
degree-day. Eq. 1 can be used to calculate the number of degree-days accumulated over

a given day using Tm, Tu“, and TM.

Degree-days =(Tm-Tu.)/2-T,,_,, (1)

Summation of degree-days required for a developmental process to occur gives
a measure of the thermal time, or heat units, required for a stage of development to
occur. Heat units have been used to predict budbreak in woody species
(Eisensmith et a1, 1980) and to predict harvest dates of many crops including
blueberries (Carlson and Hancock, 1991), pecans (Sparks, 1989), and cucumbers
(Perry and Wehner, 1990).

 

T Optimum

Rate

T Base

t

 

 

 

Temperature

Figure 9. A degree—day model relating rate of development as
a linear function of temperature.

28
Development rate can be described by the reciprocal of the time (T) required for

the particular stage of development.

Development rate = UT (2)

The degree-day model has improved the accuracy with which development can
be predicted as compared to using chronological time (Gallegher et al. , 1979; and Kirby
and Perry, 1987); however, there are problems associated with its use (Wang, 1960).
The major error in the degree—day model is that the predicted developmental response to
temperature is linear, while the actual developmental response to temperature is
curvilinear when a sufﬁciently wide range of temperatures are examined.

Many alternative formulations of Eq. 1 have been developed to improve the

degree-day model’s accuracy at the high and low temperature extremes (Figure 10).

 

0.6

P
U1
.4

Opt

9
p

Development rate
a
on

 

 

 

0.2l
0.1‘TBose
0.0 . . . . . . .
o 5 1015 20 25 so 35 40

Temperature ( C )

Figure 10. Degree-day model which compensates
for reduced development rate at supraoptimal
temperatures (Cross and Zuber, 1972).

29

Eq. 3 provides an example of a degree-day model which accounts for temperatures
outside the linear range. The rate of development is calculated to be zero when the
temperature is less than T3,... A linear function is calculated when the temperature is

between T3,, and To“. A quadratic function is used to calculate rate of development at

temperatures above To":
T _<_ Tm, :DD = 0
Tm, < Tng :DD = a(T-Tm,) (3)
T > Tm :DD = bCl‘m- T,,,,,)2

where DD represents the number of degree-days.

A sine wave can be generated from the daily minimum and maximum
temperatures and used to calculate the accumulated heat units.
Baskerville and Emin (1969) described several methods for using a sine curve to generate
values for the daily heat units accumulated. Figure 11 depicts four methods of
ealculating accumulated heat units.

Polynomial models. A cubic linear function can be used to represent the
biological response of a plant over a wide range of temperatures. Figure 12 shows a
cubic function used to predict the LAR of Hibiscus as a function of temperature.
Polynomial functions have been used to describe the rate of development in the
Easter lily (Erwin and Heins, 1990), beans (Yourstone and Wallace, 1990), and maize
(Tollenaar et al. , 1979). The advantage of using polynomial functions is the relative ease

with which these models can be generated. The disadvantage is that polynomial

3O
expressions, e.g. temperature’, temperature’, and temperature " photoperiod”, seldom
have any biologieal signiﬁeance.

///////// /////x. mm
:/é{/// /////////;//z /////////In Kl

 

 

 

 

 

 

 

 

Multan-[v ﬂuca// /.//m/’./.u] um-iP‘Imé-i‘iau-na-mm]

 

Figure 11. Four methods ofealculating the accumulated heat units by
meansofasinecurve. Ineachcase, thehorizontalaxisrepresent324
hours, the vertical axis represents temperature, and the hatched area
represents the accumulated heat units (Baskerville and Emin, 1989).

A) The daily minimum temperature is above the lower threshold (Kl),
andthereisnoupperthreshold. B)Thedaily minimumisbelowthe
lower threshold (K1) and there is no upper threshold. C) No restriction
on the daily minimum, the daily maximum exceeds the upper threshold
(K3), there is no cutoff at the upper threshold. D) No restriction on
daily minimum, daily maximum exceeds upper threshold (K2), vertical
cutoff at the upper threshold.

 

0.25

0.20 t

0.151

0.10‘

W (“WOO/4w)

005‘
.

 

 

 

Figure 12. Polynomial function (Y = 0.089 - 0.0237 * Temp. +
0.00189 " Temp.2 - 0.0000317? "' Temp?) describing the LAR of
Hibiscus as a function of temperature (Karlsson et al., 1991)

31
Nonlinear models. Nonlinear functions have been used to describe the time from

emergence to tassel initiation of maize (Coligado and Brown, 1975), and LAR and
ﬂowering in spring wheat (Angus et al., 1981). Nonlinear models can contain
biologically signiﬁcant variables. For example, the functions in Eq. 4 can be used to

describe a deve10pmental response to temperature (Reed et al. , 1976).

Y=F(T)=MT-de("I‘m-'1')b
a= YmJGm-TMQCFm-Toa)” (4)
b = am'Toﬂ/(Tom'TMin)

The parameters in this function, Tu, Tu“, To", and Y”, all have biological

signiﬁcance. An example of Eq. 4 is provided in Figure 13.

 

1.0
0.9 4 4—y m
0.8 1
e 0.7 . T Opt
‘6
‘ 0.6 4
50.5 «
0.4 4
0.3 <

02“
TW Tm

2:; \ . . x
0 5 10 15 20 25 30 35 4O
Terrperatue ( C )
Figure 13. Nonlinear model used to describe rate
of development based on four biologieally
signiﬁcant parameters, Tm, To,” Tm, and Y,“
(Reed et al., 1976).

 

 

 

32
Nonlinear models can be combined to predict the effect of two or more variables on

development. The nonlinear model shown in Eq. 5 was used to predict the

developmental response of maize to temperature and photoperiod.

Rate of development = 1I'I‘imeop,*(F,('I‘emperature)*F2(Photoperiod)) (5)

Time“, refers to the time required for a developmental phase to be completed under
optimal conditions. Timeo, can have unique value for different cultivars (Coligado and
Brown, 1975) or groupings of cultivars (Kiniry et al. , 1991). F,(Temperature) and

F2(Photoperiod) are nonlinear exponential models.

F,(Temperature) = l-EXP(-B(I‘-r)) (6)

F2(PhOtOPeﬁ0d) = l-EXP(-0(L-I')) (7)

r and I‘ are parameter estimates for the base temperature and critieal photoperiod,
respectively. B and a are the parameter estimates for the temperature and photoperiod
coefﬁcients, respectively. Whenever the environment is at the optimal temperature and
photoperiod for development, the values of the nonlinear model will be equal to 1; thus,
the development rate will be optimized. If either temperature or photoperiod deviate
from the optimal conditions, then a positive value less than 1 is calculated, and the rate
of development will decrease. Finally, if either temperature or light are below the
threshold values, the model will be equal to 0, then no development will be predicted

(Figure 14).

33

 

p—e
O

 

.8« 16h

 

.7“ 14h

 

12h

 

.2‘ 1011

Relative rate of development
0 o o c: o o o o o o
01
I

 

 

 

0 6 10 1% 20 25 sh 35
Temperature ( C )

Figure 14. A nonlinear model (Eqs. 5 & 6) describing the interaction of
temperature and photoperiod on the relative rate of development (Coligado
and Brown, 1975).

The choice of variables to include in the model depends upon the phenological
stage and the species examined. Table 4 shows the environmental variables used in a
model for different stages of soybean development (Hodges and French, 1985).

Analysis of ﬂuctuating temperature data. Results from experiments which contain
treatments with different temperatures can be difﬁcult to analyze, especially when the
temperatures fall outside the linear range. For example, plants grown in a 30C day
temperature and 10C night temperature would develop at a slower rate than plants grown
at the same ADT of 20C. McNaughton et al.(1985) proposed a method of evaluating
data from changing temperature treatments. Multiple linear regression analysis was used
to create a model which describes the observed rate of development as a function of the

fraction of time (T, to T4) spent at each temperature (Table 5). Eq. 8 calculates the

34

Table 4. Variables used to predict development during
different stages of soybean development
(Hodges and French, 1985).

 

 

Stage Variables
Emergence Temperature, water stress
Juvenile Temperature, water stress
Photoperiod-sensitive Photoperiod
Floral growth Temp. , water stress, photoperiod
Flowering Temp. , water stress, photoperiod
First pod growth Temp. , water stress, photoperiod
Last pod growth Temp. , water stress, photoperiod

 

predicted rate of development under the mixed temperature treatments. The coefﬁcients
in Eq. 8 (b, to b.) reﬂect the expected rate of development had plants been grown at

constant temperatures.

Y=r('I')=bo+b,*T,+...b,*T, (8)

where the intercept bo=0, and the coefﬁcients, bl, b2. . .b,l represent the predicted rates

of development at 15, 20, 25, and 30C respectively.

35

Table 5. The rates of development from sowing to ﬂower initiation for seven treatments.
Columns T, to T, represent the percentage of time each treatment received the indicated
temperature (McPherson et al. , 1985)

 

Temperature (C)

 

 

16 20 24 26 30 (16:12:31;
T1 T2 T3 T4 T5
0.5 0 0.5 0 0 0.021

0 0.5 0.5 0 0 0.025

0 0.5 0 0.5 0 0.024

0 0 0.5 0.5 0 0.025

0 0 0.5 0 0.5 0.020

0 0 1.0 0 0 0.026

0 0 0 1.0 0 0.024

 

Figure 15 shows the same data plotted as a function of average daily temperature, while

Figure 16 shows the rates of development from Table 5 as predicted by McNaughton’s

 

math“. 0.030

 

o
o
N
01
II
II

 

0.020 =

 

Rate of development
53
o
m
I

 

 

 

0.010 . . v . .
12 16 20 24 28 32 36
Temperature (C)

Figure 15. Observed mean rates of development
from sowing to ﬂower initiation of Pigeonpea
plotted against mean temperature for each treatment
(McPherson et al., 1985).

36

0.030

 

.0
o
M
an

 

 

0.020

 

Rate of development

0
C
‘
U"

II

 

 

0.010

 

12 10 20 2'4 20 32 35
Temperature (C)
Figure 16. Fractional rates of development from

sowing to ﬂower initiation of Pigeonpea
(McPherson et al., 1985).

Plastichron index. Another method developed to quantify the developmental
status of plants is the plastichron index (PI). PI can be deﬁned as the interval of time
occurring between the stages of development of successive events such as leaf
development. A method of determining fractional PI’s is illustrated in Figure 17. The
lines labeled n and n+1 represent the rate of elongation of successive leaves. The line
AC represents the leaf length which serves as a reference length (1.“) used for
comparison of leaf development. The line DE represents a speciﬁc point in time.
Assuming that n and n+1 are parallel and AC and DE are perpendicular, then DB/BE
= AB/BC. The length of leaves n and n+1 can be measured, and the plastichron age

of the plant are calculated with the following formula:

P1 = n+(ln(L.)-1n(1m))/(1n(L.)-IH(L.+D (9)

37
where n is the number of leaves longer than the reference length, and L. is the length of

18*
154 /
f€144 D
12- /
,0. A B c /
n % n+2
E

0 215 8101214161820
Time (days)

Figure 17. An idealized representation of the relationship

between leaf length and time for a single shoot

(Lamoreaux et al. , 1978).

the leafn.

 

Leaf length (m

 

 

 

0'0me

 

PI gives a quantitative measurement of the development which has occurred
within a stage. For example, a plant can be described as being at the 3.4 plastichron
interval; therefore, the relative progress between the third and the fourth leaf stage is
quantiﬁed.

PI has been used to model plant development (Hofstra et al., 1977) as well as to
evaluate environmental effects on plant development (Snyder and Bunce, 1983).
However, assumptions must be made that the logarithmic growth curves of successive
developmental stages are linear, parallel and equally spaced. These three assumptions

can usually be made under the constant conditions of a controlled-environment study, but

38

not in the ﬂuctuating environment in greenhouse or ﬁeld studies

(Yourstone and Wallace, 1990).

LITERATURE CITED

Allen, R.N., E.B. Dettmann, G.G. Johns, and D.W. Turner. 1988. Estimation of leaf
emergence rates of bananas. Aust. J. Agric. Res. 39:53-62.

Angus, J.F., D.H. Mackenzie, R. Morten, and CA. Schafer. 1981. Phasic
development in field crops 11. Thermal and photoperiodic responses of spring
wheat. Field Crops Res. 4:269-283.

Anonymous, USDA, Floriculture crops: 1990 Summary. National Agricultural. Statistics
Service, Agricultural Statistics Board, Washington, DC.

Armitage, A.M., W.H. Carlson, and LA. 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(5):643-647.

Aspinall, D., and LG. Paleg. 1964. Effects of day length and light intensity on growth
of barley. Aust. J. Biol. Sci. 17:807-822.

Baker, J.T., P.J. Pinter, Jr., R.J. Reginato, and ET. Kanemasu. 1986. Effects of
temperature on leaf appearance in spring and winter wheat cultivars. Agron. J.
78:605-613.

Baskerville, G.L., and P. Emin. 1969. Rapid estimation of heat unit accumulation from
maximum and minimum temperatures. Ecology 50:514-517.

Berghage, R., R. Heins, W. Carlson, and J. Biembaum. Prevent ﬂower delay.
Greenhouse Grower. August 1987.

Bilkey, P.C. and EC. Cocking. 1981. Increased plant vigor by in vitro propagation of
Saintpaulia ionantha Wendl. from sub-epidermal tissue. Hortsci. 16(5):643-644.

Bodnar, M., and W. Larcher. 1987. Chilling susceptibility of different organs and

tissues of Saintpaulia ionantha and Coﬁea arabica. Angew. Botanik.
61:225-242.

39

40

Bodson, M. 1983. Effect of photoperiod and irradiance on ﬂoral development of young
plants of a semi-early and late cultivar of azalea. J. Amer. Soc. Hort. Sci.
108(3):382-386.

Bodson, M., R.W. King, L.T. Evans, and G. Bemier. 1977. The role of
photosynthesis in ﬂowering of the long—day plant Sinapsis alba. Aust. J. Plant.
Physiol. 4:467-478.

Bomalaski, H.H. 1948. Growing degree days. How to apply this unit to measure
maturity of crops. Food Packer. 29:51-61.

Boone, M.Y.L., R.W. Rickman, and FD. Whisler. 1990. Leaf appearance rates of
two winter wheat cultivars under high carbon dioxide conditions. Agron. J.
82:718-724.

Calvert, A. 1969. Studies on the post-initiation of ﬂower buds of tomato
(Lycopersicon esculentum). J. Hort. Sci. 44:117-126.

Cao, W. , and D.N. Moss. 1989. Daylength effect on leaf emergence and phyllochron
in wheat and barley. Crop Sci. 29:1021—1025.

Carlson, J .D., and LP. Hancock Jr. 1991. A methodology for determining suitable
heat-unit requirements for harvest of Highbush blueberry. J. Amer. Soc. Hort.
Sci. 116(5):774-779.

Carpenter, W.J., and LP. Nautiyal. 1969. Light intensity and air movement effects on

leaf temperatures and growth of shade-requiring greenhouse crops. J. Amer. Soc.
Hort. Sci. 94:212-214.

Cathey, H.M. 1957. The effect of temperature upon the critical photoperiod necessary
for initiation and development of ﬂowers of Chrysanthemum monfalium. Proc.
Amer. Soc. Hort. Sci. 69:485-491.

Cathey, H.M., L.E. Campbell, and R.W. Thimijan. 1978. Comparitive development
of 11 plants grown under various ﬂuorescent lamps and different durations of
irradiation with and without additional ineandescent lighting. J. Amer. Soc. Hort.
Sci. 103(6):?81-791.

Cockshull, K.E., and AP Hughes. 1972. Flower formation in
Chrysanthemum monfolium: The inﬂuence of light level. J. Hortic. Sci.
47: 1 13-127.

Coligado, M.C., and D.M. Brown. 1975. A bio-photo—thermal model to predict tassel
initiation time in corn (Zea mays L.). Agric. Meteor. 15:11-31.

41

Conover, C.A., and R.T. Poole. 1981. Light acclimation of African violet. Hortsci.
16(1):92-93.

Cooke, R.C. 1977. Tissue culture propagation of African violets. Hortsci. 12(6):549.

Cross, 11.2., and M.S. Zuber. 1972. Prediction of ﬂowering dates in maize based on
different methods of estimating thermal units. Agron. J. 64:351-355.

Cudney, D.W., L.S. Jordan, C.J. Corbett, and W.E. Bendixen. 1989. Developmental
rates of wild oats (Avena fatua) and wheat (Triticwn aestivum). Weed Sci.
37:521-524.

DeHertog, A. Holland Bulb Forcer’s Guide. Fourth edition, 1989.

Dvorska, L. 1979. The effect of growth regulators and temperature gradient on
ﬂowering of Saintpaulia ionantha Wendl. Acta Hort. 91:507-510.

Eisensmith, S.P., A.L. Jones, and LA. Flore. 1980. Predicting leaf emergence of
‘Montmorency’ sour cherry from degree-day accumulations. J . Amer. Soc. Hort.
Sci. 105(1):75-78.

Elliott, H.E. 1946. Saintpaulia leaf spot and temperature differential. Proc. Amer.
Soc. Hort. Sci. 47:511-514.

Erwin, J .E. and RD. Heins. 1990. Temperature effects on lily development rate and
morphology from the visible bud stage until anthesis. J. Amer. Soc. Hort. Sci.
115(4):644-646.

Erwin, J .E., R.D. Heins, and M.G. Karlsson. 1989. Thermomorphogenesis in Lilium
longiﬂorum. Amer. J. Bot.76(1):47-52.

Faust, J.E., and RD. Heins. 1991. Unpublished data.
Fisher, R.A. 1989. African violet production. Ohio State University Short Course.

Ford, M.A. , and G.N. Theme. 1974. Effects of atmospheric humidity on plant growth.
Am. J. Bot. 38:441-452.

Friend, D.J., V.A. Helson, J.E. Fischer. 1962. Leaf growth in Marquis wheat, as
regulated by temperature, light intensity, and daylength. Can. J. Bot. 40:1299.

Fortanier, EL, and A. Zevenbergen. 1973. Analysis of the effects of temperature and
light after planting on bud blasting in Iris hollandica. Neth. J. Agr. Sci.
21:145-162.

42

Gallegher, J.N., P.V. Biscoe, and LS. Wallace. 1979. Field studies of cereal leaf
growth. IV. Winter wheat leaf extension in relation to temperature and leaf water
status. J. Exp. Bot. 30:657-668.

Gilmore, E.C., and LS. Rogers. 1958. Heat units as a method of measuring maturity
in corn. Agron. J. 50:611-615.

Hanchey, R.H. 1955. Effects of ﬂuorescent and natural light on vegetative and
reproductive growth in Saintpaulia. J. Amer. Soc. Hort. Sci. 66:379-382.

Harris, GP, and M.A. Scott. 1969. Studies on the glasshouse carnation: Effects of
light and temperature on the growth and development of the ﬂower. Ann. Bot.
33:143-152.

Haun, LR. 1973. Visual quantiﬁcation of wheat development. Agron. J. 65:116-119.

Healy, W.E., and H.F. Wilkins. 1984. Temperature effects on ‘Nellie White’ ﬂower
bud development. Hortsci. 19(6):843—844.

Heide, D.M. 1969. Z. Pﬂanzenphysiol. 61:279.

Hildrum, H. , and T. Kristoffersen. 1969. The effect of temperature and light intensity
on ﬂowering in Saintpaulia ionantha Wendl. Acta Hort. 14:249-255.

Hodges, T. and V. French. 1985. Soyphen: Soybean growth stages modeled from
temperature, daylength, and water availability. Agron. J. 77:500-505.

I-Iofstra, G., J.D. Hesketh, and D.L. Myhre. 1977. A plastichron model for soybean
leaf and stem growth. Can. J. Plant Sci. 57:167-175.

Horridge, J .S. , and K.E. Cockshull. 1974. Flower initiation and development in the
glasshouse rose. Scient. Hortic. 2:273-284.

Hughes, A.P., and KB. Cockshull. 1971. A comparison of the effects of diurnal
variation in light intensity with constant light intensity on growth of
Cluysamhemum mortfolium cvr. Bright Golden Anne. Ann. Bot. 35:927-932.

Hussey, G. 1963. Growth and development in the young tomato. II. The effect of
defoliation on the development of the shoot apex. J. Exp. Bot. 14:326-333.

Johansson, D.R. 1978. Saintpaulias in their natural environment with notes on their
present status in Tanzania and Kenya. Biol. Conserv. 14:45-62.

43

Karlsson, M.G., R.D. Heins, LE. Erwin, and RD. Berghage. 1989. Development rate
during four phases of chrysanthemum growth as determined by preceding and
prevailing temperatures. J. Amer. Soc. Hort. Sci. 114(2):234-240.

Karlsson, M.G., R.D. Heins, J .0. Gerberick, and M.E. Hackmann. 1991. Temperature
driven leaf unfolding rate in Hibiscus rosa-sinensis. Scient. Hortic. 45:323-331.

Karlsson, M.G., R.D. Heins, and LE. Erwin. 1988. Quantifying
temperature-controlled leaf unfolding rates in ’Nellie White’ Easter lily. J. Amer.
Soc. Hort. Sci. 113(1):70-74.

Kinet, J .M. 1977 . Effect of defoliation and growth substances on the development of
the inﬂorescence in tomato. Scient. Hort. 6:27-35.

Kinet, J .M. 1977 . Effect of light conditions on the development of the inﬂorescence
in tomato. Scient. Hort. 6:15-26.

Kimmins, R.K. Gloxinias, African violets, and other gesneriads, ed. Larson, R.A.,
Introduction to Floriculture, Academic Press Inc. 1980.

Kiniry, J.R., W.D. Rosenthal, B.S. Jackson, and G. Hoogenboom. Predicting leaf
development of crop plants. In: Hodges, T. Predicting Crop Phenology. CRC
Press. 1991.

Kirby, E.J.M. and M.W. Perry. 1987. Leaf emergence rates of wheat in a
mediterranean environment. Aust. J. Agric. Res. 38:455-464.

Kofranek, A.M., and W.P. Hackett. 1966. The inﬂuence of daylength and night
temperature on the ﬂowering of poinsettia, cultivar ’Paul Mikkelsen’. Proc.
Amer. Soc. Hort. Sci. 87:515-524.

Lamoreaux, R.J., W.R. Chaney, and K.M. Brown. 1978. The plastichron index: A
review after two decades of use. Amer. J. Bot. 65(5):586-593.

Landsberg, LL 1977 . Some useful equations for biological studies. Expl. Agric.
13:273-286.

Lang, A., Encyclopedia of Plant Physiology, Vol. 15 (Part 1), Ruhland, W., Ed.,
MacMillan, Melbourne, 1969. p. 1380.

Lange, N.E., and RD. Heins. 1991. Unpublished data.

Larcher, W. , and G. Neuner. 1989. Cold-induced sudden reversible lowering of in vivo
chlorophyll ﬂuorescence after saturating light pulses. Plant Phys. 89:740-742.

44
Larson, R.A. Introduction to Floriculture. Academic Press, Inc. 1980.

Leopold, A.C., and P.E. Kriedemann. Plant growth and development, Second edit.
McGraw-Hill Book Co., 1975.

Major, D.J., D.R. Johnson, J.W. Tanner, and LC. Anderson. 1975. Effects of
daylength and temperature on soybean development. Crop Sci. 15: 174-179.

Mastalerz, J.W. The greenhouse environment, J. Wiley and Sons, Inc., 1977.

Mastalerz, J.W. 1965. Bud blasting in Lilium longiﬂorum. Proc. Amer. Soc. Hort.
Sci. 87:502-509.

McNaughton, K.G., P.W. Gandar, and H.G. McPherson. 1985. Estimating the effects

of varying temperature on the rate of development of plants. Ann. Bot.
56:579-595.

McPherson, H.G., Warrington, LL, and Tumbull, H.L. 1985. The effects of

temperature and daylength on the rate of development of pigeonpea. Ann. Bot.
56: 1597-1611.

Milford, G.F.J., T.O. Popock, and J. Riley. 1985. An analysis of leaf growth in sugar
beet. I. Leaf appearance and expansion in relation to temperature under controlled
conditions. Ann. Appl. Biol. 106:163-172.

Mitchell, C.A., and J .E. Hoff. 1977. Inﬂuence of ambient humidity and root zone
shielding on tomato seedling growth. J. Amer. Soc. Hort. Sci. 102(5):587-590.

Moe, R. 1972. Effect of daylength, light intensity, and temperature on growth and
ﬂowering in roses. J. Amer. Soc. Hort. Sci. 97(6):796—800.

Moe. R., and T. Kristoffersen. 1968. The effect of temperature and light on growth
and ﬂowering of Rosa ’Baccara’ in greenhouses. Acta Hort. 14:157-166.

Mortensen, L.M. 1986. Effect of relative humidity on growth and ﬂowering of some
greenhouse plants. Scient. Hortic. 29:301-307.

Mortensen, L.M. 1983. Growth responses of some greenhouse plants to environment.
XII. Effect of CO2 and light on photosynthesis and growth of
Saintpaulia ionantha, Kalanchoe blossfeldiana and Nephrolepsis axaltata. Agric.
Univ. of Norway, Report No.277. 62:1-16.

Mortensen, L.M. 1986. Effect of intermittent as compared to continuous CO2
enrichment on growth and ﬂowering of Chrysanthemum x mortfolium Ramat. and
Saintpaulia ionantha H. Wendl. Scient. Hortic. 29:283-289.

45

Murneek, A.E. Vemalization and Photoperiodism, Eds. Murneek, A.E. and R.C.
Whyte. Chronica Botanica, Waltham, Mass., 1948, pg. 39.

Naylor, E.E., and B. Johnson. 1937. A histologieal study of vegetative reproduction
in Saintpaulia ionantha. Amer. J. Bot. 24:673-678.

Owen, 0., T. Small, and P.H. Williams. 1926. Carbon dioxide in relation to
glasshouse crops. Part III. The effect of enriched atmosphere of tomatoes and
cumbers. Ann. Appl. Biol. 13:560-576.

Perry, KB, and Wehner. 1990. Prediction of cucumber harvest date using a heat unit
model. Hortsci. 25:405-406.

Poesch, G.H. 1942. Ring spot on Saintpaulia. Proc. Amer. Soc. Hort. Sci.
41:381-382.

Raabe, RD. 1957 . Physiological breakdown of chlorophyll in the Gesneriaceae.
Phytopath. 47:28.

Rawson, H.M., and J.H. Hindmarsh. 1983. Light, leaf expansion and seed yield in
sunﬂower. Aust. J. Plant Physiol. 10:25-30.

Rawson, H.M., and J.H. Hindmarsh. 1982. Effects of temperature on leaf expansion
in sunﬂower. Aust. J. Plant Physiol. 9:209-219.

Reed, K.L., E.R. Hamerly, B.F. Dinger, and P.G. Jarvis. 1976. An analytical model
for ﬁeld measurement of photosynthesis. J. Appl. Ecol. 13:925-942.

Sachs, R.M., and CF. Bretz. 1962. The effect of daylength, temperature, and
gibberellic acid upon ﬂowering in Fuchsia hybrida. Proc. Amer. Soc. Hort. Sci.
80:581-588.

Sanderson, K.C., and J. McGuire. 1988. Growth regulator sprays during propagation
increase African violet crowns and leaves. Hortsci. 23(6): 1085.

Schneiter, A.A. , and J .F. Miller. 1981. Description of sunﬂower growth stages. Crop.
Sci. 21:901.

Shillo, R., and AH. Halevy. 1976. The effect of various environmental factors on
ﬂowering of gladiolus. 1. Light intensity. Scient. Hortic. 4:131-137.

Shillo, R., and AH. Halevy. 1976. The effect of various environmental factors on
ﬂowering of gladiolus. 111. Temperature and moisture. Scient. Hortic. 4: 147-155.

46

Snyder, P.W., and J .A. Bunce. 1983. Use of the plastichron index to evaluate effects
of light, temperature and nitrogen on growth of soya bean (Glycine max L.
Merr.). Ann. Bot. 52:895-903.

Sparks, D. 1989. Predicting nut maturity of the pecan from heat units. Hortsci.
24(3):454-455.

Start, ND. and B.G. Cumming. 1976. In vitro propagation of Saintpaulia ionantha
Wendl. Hortsci. ll(3):204-206.

Steffen, J.D., K.E. Cockshull, G. Navissano, and R.M. Sachs. 1988. Inﬂorescence
development in Orrysanthemum monfolium as a function of light on the
inﬂorescence. Ann. Bot. 61:409—413.

Stinson, R.F. and A. Laurie. 1954. The effect of light intensity on the initiation and
development of ﬂower buds in Saintpaulia ionantha. Proc. Amer. Soc. Hort. Sci.
64:459-467.

Tibbitts, T.W. , and G. Bottenberg. 1976. Growth of lettuce under controlled humidity
levels. J. Amer. Soc. Hort. Sci. 101(1):70-73.

Tollenaar et.al. 1984. Differences in rates of leaf appearance among maize hybrids and
phases of development. Can. J. Plant Sci. 64:759-763.

Tollenaar, M., T.B. Daynard, and RB. Hunter. 1979. Effect of temperature on rate
of leaf appearance and ﬂowering date in maize. Crop Sci. 19:363-366.

Vazquez, A.M., and K.C. Short. 1978. Morphogenesis in cultured ﬂoral parts of
African violet. J. Exp. Bot. 29:1265-1271.

Vogelezang, J .V.M. 1988. Effect of root-zone heating on potplants. Acta Hortic.
229:421-427.

von Hentig, W.U. 1976. Results of propagation with leaf cuttings of
Saintpaulia ionantha. Acta Hortic. 64:55-63.

Wang, 1960. A critique of the heat unit approach to plant response studies. Ecology
41:785-789.

Wardell, W.L. 1976. Floral activity in solution of deoxyribonucleic acid extracted from
tobacco stems. Plant Phys. 57:855-861.

Warrington, LL, and ET. Kanemasu. 1983. Corn growth response to temperature and
photoperiod II. Leaf-initiation and leaf appearance rates. Agron. J.
75:755-761.

47

Watts, W.R. 1972. Leaf extension in Zea mays. 11. Leaf extension in response to
independent variation of the temperature of the apical meristem, of the air around
the leaves, and of the root-zone. J. Exp. Bot. 23:713-21.

Wellensiek, S.J., The induction of ﬂowering, some case histories. Evans, L.T., Ed.,
MacMillan, Melbourne, 1969. p 350.

Went, F.W. The experimental control of plant growth. Chronica Botanica Co., 1957.

Wilson, H. van Pelt. 1951. The compete book of African violets. Barrows and Co.
Inc., New York, pp.136-143.

Wittwer, S.H., and W. Robb. 1964. Carbon dioxide enrichment of greenhouse
atmosphere for vegetable crop production. Econ. Bot. 18:34.

Yourstone, K.S. , and DR. Wallace. 1990. Application of plastichron index to common
been grown in controlled environments. J. Amer. Soc. Hort. Sci.
115(5):820-823.

Yourstone, K.S., and DH. Wallace. 1990. Effects of photoperiod and temperature on

rate of node development in indeterminate bean. J. Amer. Soc. Hort. Sci.
115(5):824-828.

Zeislin, N., and A.H. Halevy. 1975. Flower bud atrophy in ’Baccara’ roses. II. The
effect of environmental factors. Scient. Hortic. 3:383-391.

Zeevaart, J .A.D., The induction of ﬂowering, some case histories. Evans, L.T., Ed.,
MacMillan, Melbourne, 1969. p 116.

SECTION I

MODELING LEAF DEVELOPMENT OF THE AFRICAN
VIOLET (Saintpaulia ionantha WENDL.)

Modeling Leaf Development of the African Violet (Saintpaulia ionantha Wendl.)

James E. Faustl and Royal D. Heins’

Department of Horticulture, Michigan State University, East Lansing, MI 48824-1325

Received for publication . We acknowledge the support of the Michigan
Agricultural Experiment Station and Express Seed Co. , Oberlin, Ohio. The cost of
publishing this paper was defrayed in part by the payment of page charges. Under postal
regulations, this paper therefore must be hereby marked advertisement solely to indicate

this fact.

‘Graduate Research Assistant

2Professor of Horticulture

49

50
Production and Culture

Modeling Leaf Development of the African Violet (Saintpaulia ionantha Wendl.)

Additional index words. daily integrated photosynthetic photon ﬂux density, irradiance,

nonlinear function, temperature

Abbreviations: LUR, Leaf unfolding rate; PPF, Photosynthetic photon ﬂux density

Abstract. The rate of leaf unfolding was determined for African violet
(Saintpaulia ionantha Wendl.) ‘Utah’ plants grown under 20 combinations of temperature
and PPF. A nonlinear model was used to predict LUR as a function of temperature and
daily integrated PPF. The minimum and maximum temperatures for leaf unfolding were
estimated at 8 and 30.8C, respectively. The maximum predicted LUR was 0.266
leaves day" which occurred at 25C and a daily integrated PPF of 10 mol rn'2 d". The
optimum temperature for leaf unfolding decreased to 23C and the maximum rate
decreased to 0.175 leaves day1 as the daily integrated PPF decreased from 10 to 1
mol m’2 d". A greenhouse validation experiment using 12 combinations of temperature
and daily integrated PPF was conducted to validate the LUR model. Plant temperature
was measured by inserting a hypodermic-needle thermocouple probe into the stem near
the apical meristem. Plant temperatures used in the model predicted leaf development
more accurately than did air temperatures, but using average hourly temperature data was

no more accurate than using average daily temperature data.

51
Saintpaulia ionantha Wendl. , commonly referred to as the Afriean violet, is an

important greenhouse crop in the United States. In 1990, 23 million pots were sold at
a wholesale value of 27.6 million dollars (Anonymous, 1990). Production occurs year
around in the United States and peaks at holidays; Valentine’s Day is the largest
marketing date. Most commercial producers of African violets begin production by
purchasing small plants or “plugs” which usually possess 8 to 12 unfolded leaves. The
plugs are transplanted into 10 cm diameter pots, grown, and then sold when the plants
have approximately 20 to 25 unfolded leaves and ﬁve or more open ﬂowers.

Phasic development scales have been used to identify the status of phenological
development. Vegetative development can be described by leaf number and the rate at
which leaves appear, or unfold. Phenological scales can be useful to the grower for
identifying both the current developmental status of a crop, and the development required
over a future period of time in order for a crop to be at the proper stage of development
at the market date.

Temperature is the primary variable used in models to predict rates of
development (Hodges, 1991). Average hourly temperatures (Karlsson et al. , 1991),
average daily temperatures (Karlsson et al., 1988), and minimum and maximum daily
temperature (Hodges and French, 1985) data have been used in plant-development
models. Air temperatures are most commonly used in leaf-development models;
however, soil temperatures have also been used to predict development of species such
as wheat while the apical meristem is below the soil surface (Swan et al., 1987).

PPF is a variable which is not usually included in leaf-development models;

however, Hanchey (1955) observed that leaf number of African violets decreased from

52
44 to 22 leaves per plant as illuminance decreased from 600 to 100 footcandles for 6 h

per day (1.9 to 0.31 mol m2 d"). The African violet is a shade plant (Johansson, 1978);
therefore, commercial producers grow African violets under shadecloth. Also, the daily
integrated PPF delivered to a crop during cloudy, winter conditions can be below
2 mol rn'2 d" .

Both temperature and PPF inﬂuence development rate of the African violet and
can be computer controlled and monitored in greenhouse environments; therefore,
temperature and PPF need to be considered in the development of Afriean violet
phenology models.

The objectives of our research were twofold: ﬁrst, to describe the inﬂuence of
temperature and PPF on the rate of leaf development of the African violet and second,

to develop a model which would predict leaf development in the greenhouse environment.

Materials and Methods

Model description. LUR, expressed in the number of leaves unfolded per day,
describes the rate at which leaves unfold, or appear, at the apical meristem. A leaf was
considered unfolded when the leaf blade reached 7 mm in length. The slope of a linear
regression line ﬁt to the number of unfolded leaves as a function of time represented the
LUR for a given plant.

The following nonlinear functions (Reed et al. , 1976; Landsberg, 1977) were used

to describe LUR as a function of both temperature and daily integrated PPF:

53

LUR = Aa'TMJCTm'nB (I)
A = wisdom-Tuscan»: (2)
B = crawled-T...) (3)

where Tm. and Tm refer to the minimum temperature and the maximum temperature at
which LUR is zero. T0,, is the temperature at which the optimum LUR occurs for a
given daily integrated PPF. LUR,” refers to the maximum value for LUR at a given
daily integrated PPF. The following nonlinear functions were used to describe LUR,“

and T0,:

T0,, = ao+a1 *EXP(a2*PPFm) (4)

LUR” = b0 + b‘ *EXP(b2*PPFDI) (5)

where a0 and b0 indieate the asymptotic values of the functions, and a,, a2, b,, and b,, are
parameter estimates.

Estimating parameters. Parameter estimates and asymptotic 95 % conﬁdence
limits (Table 1) for the nonlinear functions were estimated with SAS procedure NLIN
(SAS Institute, Inc. , 1989). In previous experiments the minimum temperature for leaf
growth of the African violet was approximately 8C (Faust and Heins, unpublished data);
therefore, the value of Tm. was ﬁxed at 8C.

Experimental design. A split plot design was used with temperature as the main
plot and PPF as the split plot. Four PPF treatments were located in each of ﬁve

temperature treatments. Five plants were grown in each of the 20 temperature/PPF

54

treatments. African violet ‘Utah’ plants possessing 8 to 10 unfolded leaves were received
in 3 cm diameter (22 cm’) cells and transplanted into 10 cm diameter pots (450 cm’)
containing a commercial peat-based medium (Baccto Professional Plant Mix, Michigan
Peat Co. , Houston, TX). Immediately after transplanting, plants were placed into one
of ﬁve (6.3 m’) walk-in growth chambers (Hotpack, Model UWP 3009-2, Philadelphia,
PA). Air temperature was adjusted to maintain plant temperatures at 14, 18, 22, 26, and
30C. Plant temperature was measured by inserting a hypodermic-needle thermocouple
probe (Omega Hypl-30-1/2-T-G-60—SMP-M) into stem, petiole, and leaf tissue. Layers
of neutral-density shadecloth were placed above the plants in each growth chamber to
create PPF of 23, 92, 161, 230 pmol rn'2 s“. PPF was supplied 12 h per day by cool
white ﬂuorescent lamps (Philips VHO F96Tl2/CW/VHO) which resulted in daily
integrated PPF treatments of 1, 4, 7, and 10 mol m'2 day". The range of PPF treatments
delivered within a chamber resulted in a 11C plant-temperature difference between PPF
treatments.

Dates of leaf unfolding were recorded every 2 to 4 days until the ﬁrst ﬂower had
opened on each plant or until day 77 of the experiment. Orthogonal polynomial contrasts
were utilized to determine the trend analyses. SAS procedure GLM
(General Linear Models) (SAS Institute Inc. , 1989) was used for analysis of variance.

Greenhouse validation of the LUR model. Four 10 m2 glass greenhouses were set
to maintain air temperatures of 15, 20, 25, or 30C from December, 1990, to
February, 1991 . Greenhouse temperatures were controlled by a greenhouse
climate-control computer (Priva, Model CD750, De Lier, Holland). Each greenhouse

was divided into thirds to provide three PPF treatments. Plants in the low-PPF treatment

55
were placed under neutral-density shadecloth to reduce natural PPF by 50% . Plants in

the medium-PPF treatment received the natural-PPF environment. From 0600 to 1800 h
each day (4.3 mol m2 d"), plants in the high—PPF treatment received natural PPF plus
an additional PPF of 100 pmol m‘2 s" supplied by 400 W high-pressure sodium lamps
(Phillips Electronics Ltd. , Model SGH701, Ontario, Canada). Another layer of
50% PPF-reduction neutral-density shadecloth was pulled over all PPF treatments when
the natural PPF increased above approximately 300 pmol m" s". The average daily
integrated PPF were 2.6, 4.5, and 8.8 mol m‘2 (1'1 for the three PPF treatments over the
course of the experiment.

PPF was monitored at canopy level with LI-COR (LI-1909A) quantum sensors.
Plant temperature in each treatment was monitored by inserting hypodermic-needle
thermocouple probes (Omega Hyp1-30-1/2-T-G-60-SMP-M) into the stem near the apieal
meristem. Thirty-minute average plant temperatures and PPF measurements were
recorded with a datalogger (Easy Logger 800, Omnidata International, Logan, Utah,).
Air temperature was monitored 30 cm above the canopy with a shaded and aspirated
thermocouple. Two hour average air temperatures were also recorded with a datalogger
(Digistrip III, Kaye Instruments Co. , New Bedford, Conn.).

All plants were subirrigated with a nutrient solution consisting of 3.6 mmol N and
1.3 mmol K from calcium and potassium nitrate or watered based upon the electrical
conductivity of the medium. Electrical conductivity of the media in the root zone was
maintained between 0.5 and 1.0 m8 throughout the experiment using the 2:1 water/soil
(vol/vol) method (Warncke and Krauskopf, 1983). The electrical conductivity of the

irrigation water was 0.65 m8 and the bicarbonate alkalinity was 310 mg l".

56
0.25 to 0.5 ml phosphoric acid per liter irrigation water was applied to the medium as

needed to maintain media pH between 5.5 and 6.5.

Tests of model prediction. Four methods of averaging actual temperature data
were compared when the LUR model was validated. The methods were based on
1) average hourly plant temperature, 2) average hourly air temperature, 3) average daily
plant temperature and 4) average daily air temperature. The calculated average
temperatures were used along with the daily integrated PPF to predict LUR on either an
hourly or daily basis.

ADT was calculated from measurements taken from 0600 HR one day to
0600 HR the following day. Daily integrated PPF measured during the photoperiod of
one day was used to ealculate LUR starting at the beginning of the photoperiod of that
same day and continuing for 24 h until the beginning of the next photoperiod. For
example, daily integrated PPF measured from 0700 to 1700 HR on January 1 was used
in the LUR model along with the average daily or hourly temperatures from 0700 HR
onJanuary 1 to0600HRofJanuary2tocalculateLUR from0700HRonJanuary l to
0600 HR of January 2.

Two techniques were used to compare the predictive value of the four methods
of using temperature data in the LUR model. First, the absolute deviation between actual
and predicted leaf number was calculated for each recorded leaf number. Second, the
slope of the observed leaf number plotted against the predicted leaf number was
ealculated by linear regression. Perfect prediction of the rate of leaf unfolding would

result in a slope equal to one; therefore, the absolute deviation between the slope of the

57
predicted leaf number and the slope of the observed leaf number provided another

comparison of the methods for entering temperature data into the LUR model.

Results

The leaves of each plant unfolded as a linear function of time (Figure 1).
Temperature, daily integrated PPF, and the interaction between temperature and daily
integrated PPF signiﬁcantly inﬂuenced LUR (Table 1). LUR increased as temperature
increased from 14C to an optimum temperature, and then decreased sharply as
temperature increased above the optimum temperature. LUR also increased at all
temperatures as daily integrated PPF increased from 1 to 7 mol m‘2 d", but did not
increase further at 10 mol m“2 (1'1 (Figure 2).

The statistieally signiﬁcant interaction between temperature and daily integrated
PPF was reﬂected by the increase in T0,, and LUR», as daily integrated PPF increased
with respect to temperature. To, increased from 22.6 to 25.5C (Figure 3A), and LUR”,
increased from 0.175 to 0.27 leaves d" (Figure 33) as the daily integrated PPF increased
from 1 to 10 mol m2 d".

A model was created which predicts LUR based on temperature and daily
integrated PPF data (Figure 2). Estimates for the parameters in Equations 1-5 are shown
in Table 2.

In the greenhouse-validation experiment, average hourly air temperatures were
usually maintained within i1 .5 C of the setpoint temperature, while plant temperature
frequently differed from air temperature. Plant temperature during the photoperiod was

closely correlated to the instantaneous solar radiation. Plant temperature during the

58
daylight hours on cloudy days was often 1 to 3C below air temperatures (Figure 4A),

while plant temperature during the daylight hours on sunny days was up to 3C higher
than air temperature (Figure 48). Plant temperature during the night was typieally 2 to
SC lower than air temperature (Figure 4 A, B, & C). Plant temperature increased 3 to
4C when the high-pressure sodium lamps were used (Figure 4C).

Plant temperature data predicted leaf number more accurately than air
temperature data based on a comparison between actual and predicted leaf numbers. The
precision of the model was not improved by using average hourly temperatures rather
than average daily temperatures (Table 3).

The leaf unfolding model (Eqs. 1-5) accurately predicted leaf unfolding over
77 days of the validation experiment (Figure 5). The predicted leaf number was within
one leaf of the observed leaf number 84% of the time (more than 300 measurements on

48 plants) when average daily plant temperatures were used in the model (Figure 6).

Discussion

The inﬂuence of temperature on leaf development of Afriean violets was similar
to other plant species. In African violets, we determined Tm, to be 8C, T0,. to be
between 23 and 25.5C, and Tm to be 30.8C, temperatures similar to those determined
for other tropical species (Kiniry et al. , 1991). LUR“, varies considerably from
0.10 leaves day" for banana (Allen et al., 1988) to 2.5 leaves day" for the Easter lily
(Karlsson et al., 1988). LUR“ml for African violets was 0.266 leaves day‘.

Leaf development was inﬂuenced by daily integrated PPF. PPF is not typically

used in plant development models beeause the daily integrated PPF at which most crops

59
are produced is sufﬁciently high to saturate the photosynthetic apparatus. However,

African violets are susceptible to physiological damage at high PPF, so growers often
produce African violets at PPF which can limit photosynthesis and leaf development.
Therefore, daily integrated PPF was included as a variable in the LUR model.

No signiﬁeant difference resulted from using average hourly temperatures in the
model as compared to average daily temperatures. Similar results have been observed
by researchers using degree day models (Cross and Zuber, 1972;
Gilmore and Rogers, 1958). The temperature-response curves used in the LUR model
were developed from data collected on plants grown at constant temperatures, but may
not reﬂect the developmental responses which occur during brief exposures to
temperatures outside the linear-response range. Therefore, the model reﬂects
development rates which occur over broader time intervals; thus, average daily
temperatures were the most accurate at predicting leaf development.

Plant temperatures gave a more accurate prediction of leaf number with the LUR
model than air temperatures. The actual developing plant-tissue temperature must be
determined to accurately predict speciﬁc organ or tissue development.
Harris and Scott (1969) observed that independent ﬂuctuations of plant-organ
temperature, e.g. , ﬂower buds and leaves, determined the rate of their development.

Plant temperature is dependent on the energy exchange between the plant and its
environment. During the night period of the greenhouse experiment, plant temperature
was frequently 2 to SC below air temperature. Part of this temperature difference can
be attributed to energy loss to the greenhouse glass via long wave radiation

(Hanan et al., 1978). However, we have also observed a l to 3C drop in temperature

60
of African violets during dark periods in growth chambers. Vogelezang (1988) also

observed that meristem temperature of African violets is more closely correlated to soil
temperature than to air temperature. We hypothesize that the observed difference
between air and plant temperature is due at least in part to conductive heat transfer from
the plant to the soil. The African violet has a rosette growth habit, and the meristem is
typically less than 3 cm above the soil surface. Evaporation of water from the soil
surface would result in evaporative cooling. Conductive heat transfer from the plant
stem to the cooler soil would results in reduced plant temperatures.

In summary, an LUR model based on average daily plant temperatures and daily
integrated PPF accurately predicted leaf development of Afriean violets grown for
77 days in a greenhouse under a range of temperature and PPF conditions. Plant
temperature predicted leaf development more precisely than did air temperature. No
beneﬁt was obtained by using average hourly temperatures rather than average daily

temperatures.

LITERATURE CITED

Allen, R.N., E.B. Dettmann, G.G. Johns, and D.W. Turner. 1988. Estimation of leaf
emergence rates of bananas. Aust. J. Agric. Res. 39:53-62.

Anonymous, USDA, Floriculture crops: 1990 Summary. National Agricultural
Statistics Service, Agricultural Statistics Board, Washington, DC.

Cross, H.Z., and M.S. Zuber. 1972. Prediction of ﬂowering dates in maize based on
different methods of estimating thermal units. Agron. J. 64:351-355.

Gilmore, E.C., and LS. Rogers. 1958. Heat units as a method of measuring
maturity in corn. Agron. J. 50:611-615.

Hanan, J.J., W.D. Holley, and KL. Goldsberry. 1978. Greenhouse Management.
Springer-Verlag Berlin, Heidelberg. pp. 130-151.

Hanchey, RH. 1955. Effects of ﬂuorescent and natural light on vegetative and
reproductive growth in Saintpaulia. J. Amer. Soc. Hort. Sci. 66:379-382.

Harris, GP, and M.A. Scott. 1969. Studies on the glasshouse earnation: Effects of
light and temperature on the growth and development of the ﬂower. Ann. Bot.
33:143-152.

Hodges, T. 1991. Temperature and water stress effects on phenology, p. 7-13. In T.
Hodges (ed.). Predicting crop phenology. CRC Press, Inc.

Hodges, T. and French, V. 1985. Soyphen: Soybean growth stages modeled from
temperature, daylength, and water availability. Agron J. 77 :500-505 .

Johansson, D.R. 1978. Saintpaulias in their natural environment with notes on their
present status in Tanzania and Kenya. Biol. Conserv. 14:45-62.

Karlsson, M.G., R.D. Heins, and IE. Erwin. 1988. Quantifying

temperature-controlled leaf unfolding rates in ‘Nellie White’ Easter lily. J.
Amer. Soc. Hort. Sci. 113(1):70-74.

61

62

Karlsson, M.G., R.D. Heins, J.O. Gerberick, and M.E. Hackmann. 1991.
Temperature-driven leaf unfolding rate in Hibiscus rosa-sinensis. Scient. Hortic.
45 :323-331.

Kiniry, J.R., W.D. Rosanthal, B.S. Jackson, and G. Hoogenboom. 1991. In T. Hodges
(ed.). Predicting crop phenology. CRC Press, Inc.

Landsberg, J .J . 1977 . Some useful equations for biological studies. Expl. Agric.
13:273-286.

Reed, K.L., E.R. Hamerly, R.F. Dinger, and P.G. Jarvis. 1976. An analytical model
for ﬁeld measurement of photosynthesis. J. Appl. Ecol. 13:925-942.

SAS Institute, Inc. 1989. SAS/STAT User’s Guide, Version 6, Fourth edition, Vol.
2, Cary, NC.

Swan, J .B., E.C. Schneider, J.F. Moncrief, W.H. Paulson, and A.E. Peterson. 1987.
Estimating corn growth, yield, and grain moisture from air growing degree days
and residue cover. Agron J. 79:53-60.

Vogelezang, J .V.M. 1988. Effect of root-zone heating on potplants. Acta Hortic.
229:421-427.

Warncke, D., and D. Krauskopf. 1983. Greenhouse growth media: testing and nutrition
guidelines. Mich. State Univ. Ext. Bull. E-1736.

63

Figure 1. The rate of leaf unfolding for an African violet grown at 22C and a daily
integrated PPF of 7 mol m'2 d" was 0.244 leaves (1'1 as determined by the
slope of the regression line Y=0.244*X-0.302 (R’=0.99).

12

 

22C

‘ 10 mol m‘zd”1
10 A

 

 

Iantai
mint?!

 

 

Leaf Number
.p.

 

 

 

 

Days

 

65

Figure 2. Nonlinear model (Eqs. 1-5) describing LUR as a function of temperature
and daily integrated PPF (R2 =0.99). Symbols represent treatment means.

Leaf Unfolding Rate (leaves d“)

0.30

0.25

0.20

0.15

0.10

0.05

0.00

 

010 mol m"’d"

 

 

'3 7 mol m'zd" -
A 4 mol m‘zd"1 A
O 1 mol m‘zd"

 

//\\

 

Z/ \

 

 

 

//
/

 

 

10 15 20 25 30
Temperature (C)

35

Figure 3.

67

The inﬂuence of daily integrated PPF on A) To, (To,=25.44-3.127*
EXP(-0.193*PPFm))andB)LURM,,(LURm=0.266-0.l37‘EXP(-0.418*
PPFDQ). Vertical bars indicate asymptotic 95 % conﬁdence intervals for
Top, and LUR,“ values estimated at each daily integrated PPF treatment.

TOPJC)

—1
(Leaves d )

Max

LUR

68

28

 

26-

 

 

 

 

22-

0.30

 

0.25 -

0.20 -

.0
or

 

 

0.10

 

1 2 3 4 5 6 7 8 910
Daily Integrated PPF (mol m—zd—1)

69

Figure 4. A comparison between air and plant temperature (A) over the course of
a cloudy day (December 30, 1990), (B) a sunny day (January 3, 1991),
and (C) a cloudy day in which the plants were growing under high-
pressure sodium lamps from 0600 to 1800 HR (December 30, 1990).

70.

 

00 Air temp 350
24 ‘ A-A Plant temp. A ' 300
22 - oo PPF . 250

 

 

 

A A
8 '1'

NV)

8 'E

3 _

3'3. 1%. O O

”2;? 0‘3 E

0"" CE)- 3

L1.

.93 a.

a.

 

 

 

 

O 4 8 12 16 20 24
Time of Day (h)

71

Figure 5. Comparison between the predicted (solid line) and observed (0) leaf
number of plants grown in 12 temperature/PPF treatments.

>00
om 8 o... om om 8 o... om om 8 9V om om oo o... om o

 

1 u q u u u u u q - - q u q u to] o
O O O 088
11 11 11 oo 1 m
o
o o
11 o 11 11 o 1 OP
0 o
0% om
o 30.. 0mm 11 o 30.. com 11 26.. on, 1 mp

 

72

 

 

 

 

o 228: 0811 528.2 cow + £282 of - 9
n n L n p . — p F p — 6|. 0
4 1 u 1 + d 4 o A d 4 q
o oo o
1' 1' .l O o I. m
n... 1 1- . cue - o.
o 8 oo
oo o 00
no.1 00m. 11 no... omw 11 1 o no... one 1 m..—

 

 

 

 

 

- b P F b n p n — p b —

 

JeqwnN 108‘]

1.111111... -

Figure 6.

73

Comparison of the observed (0) and the predicted (solid line) leaf number
of all of the plants in the 12 temperature and PPF treatments during the
77 days of the greenhouse-validation experiment.

Observed Leaf Number

74

 

20
18-
161
149
12j
10j

N-P
1.1

 

O

0

 

 

I U

l

r I I

I

' l

2 4 6 8 10
Predicted Leaf Number

12141618 20

75

Table 1. Inﬂuence of temperature and daily integrated PPF on LUR.

 

 

 

 

 

 

 

Daily integrated PPF
(mol in2 (1")
Temperature (C) 1 4 7 10
Leaf unfolding rate (leaves (1")
14 0.115 0.149 0.146 0.137
18 0.139 0.202 0.174 0.213
22 0.181 0.223 0.254 0.241
26 0.161 0.232 0.268 0.263
30 0.086 0.124 0.173 0.176
Source Signiﬁcance
Temperature
Linear NS
Quadratic ***
Cubic *
Daily integrated PPF
Lil-1w #8!!!
Quadratic ***
Cubic NS

Temp. x PPFm

##8

 

' NS, *, *** Nonsigniﬁcant or signiﬁcant at P=.05, .001, respectively.

Table 2. Parameter estimates and 95 % conﬁdence intervals calculated for use in the

LUR model (Equations 1-5).

 

 

76

 

 

. Asymptotic 95 96
Parameter Estimate conﬁdence interval
lliqgatrgng quraatirosns Lower Upper
Tm 30. 83 30.36 31.31
Ton
a0 25.44 23.22 27.66
al -3. 127 -4.861 -1.392
a2 -0. 193 -0.559 0.173
Yam
b0 0.266 0.252 0.280
b, —0. 137 -0. 162 -0. 112
b; -0.418 -0.621 -0.215

 

77

8...... F a... .96.. a. .8 ea... .
6836... 8» 8.. 3.. 5...? 85.... .8 2. .9... ea .8 .82 9.3 .38.... E .
.ncdum Q3 .3 agenda... :32

 

 

 

 

 

 

13331:: as:
In... 8... $4.. 2... z... .2. we. 8... n... R m...
«a... S... R... 3... a... B... 8.. E... z n. n...
on... 8... .e... 8... a... on... 8.. me... a... e... a. on
a... ..... 2... a... 2... a... 8... 8.. N. a. as
8... 2... 8... n... 8... 3.... 8.. 8... a... 8 e...
a... m... a... a... 8... 8... S... a... no. 2. m... a
3... no... 8... 8... E... as... 8... a... a... we no
8... .3 .2... B... N... a... e... 8... a... E e...
8... ..... 8... e... we... «a... a... E... 5.. o. m... on
.2. z... a... 8... an. 8.. en... E. a... .1 m...
on... a... R... 2... m... 8.. a... a... m... .. a...
a... 8... «a... S... «a. 9.... a... mm... as .1 Z n.
.02...
._... 828.. .8...
£38.... a 2.823.. too. a .8 2.... .03.... A... New: .2... dw.
.%.§ue§swﬁ8 thwaﬁezeg an... e... . __.._..a a...
.988 .98. .95. .95. d8... .98. .95.. dB...
ﬁne wan...“ an“... N"mm... he“... MM...“ 2%.... ﬁg...
.03.. .o>< .03.. .25. .o>< .o>< .o>< .o><

 

89...... .8. 6.8... o. 8.... ages 9...... .o 82.8... 58 .e 8.838 .m 2.3.

SECTION II

MODELING INFLORESCENCE DEVELOPMENT OF THE
AFRICAN VIOLET (Saintpaulia ionantha WENDL.)

Modeling Inflorescence Development of the African violet (Saintpaulia ionantha
Wendl.)

James E. Faust‘ and Royal D. Heinsz

Department of Horticulture, Michigan State University, East Lansing, MI 48824-1325

Received for publication . We acknowledge the support of the
Michigan Agricultural Experiment Station and Express Seed Co. , Oberlin, Ohio. The
cost of publishing this paper was defrayed in part by the payment of page charges.
Under postal regulations, this paper therefore must be hereby marked advertisement

solely to indicate this fact.

1Graduate Research Assistant

2Professor of Horticulture

79

80
Production and Culture.

Modeling Inﬂorescence Development of the African violet (Saintpaulia ionantha Wendl.)

Additional index words. leaf extension rate, photosynthetic photon ﬂux density,

temperature.

Abbreviations: ADT, average daily temperature; DT, day temperature; NT, night
temperature; LER, leaf extension rate; LBL, leaf blade length; PPF, photosynthetic

photon ﬂux density; VB, visible ﬂower bud

Abstract. The effects of temperature and PPF on ﬂower initiation and
development were quantiﬁed to provide a basis for an inﬂorescence development model.
The percentage of leaf axils forming an inﬂorescence increased as the daily integrated
PPF increased from 1 to 4 mol rn‘2 (1'l , while the rate of inﬂorescence development was
a function of average daily temperature. The appearance of a visible bud in the leaf axil
was correlated with leaf blade length of the subtending leaf. A polynomial function was
used to describe leaf blade length at the time of visible bud as a function of temperature
and daily integrated PPF. A nonlinear function was used to describe the inﬂuence of
temperature on the rate of leaf extension. The time of visible appearance of an
inﬂorescence in the leaf axil was then predicted by measuring the current leaf blade
length and estimating the time required for the leaf blade to extend to the length required

for VB appearance

81
The number of days required for an inﬂorescence to develop from visible bud to

anthesis was inﬂuenced by temperature. A phasic development scale was used to
describe the developmental status of an inﬂorescence. A model was then created which
predicted time to anthesis based upon temperature and the current stage of inﬂorescence
development. The number of days from the time of leaf unfolding to anthesis for the
inﬂorescence which will be initiated in the leaf axil decreased from 87 to 57 days as

temperature increased from 18 to 26C.

Commercial producers of container-grown ﬂowering plants are required to
produce plants for speciﬁc market dates; therefore, the ability to predict the date of
anthesis for a crop is essential. Prediction requires the grower be able to identify the
developmental status of a crop and then properly adjust the greenhouse environment so
that the crop is ﬂowering at the market date.

Accurate scheduling of African violet (Saintpaulia ionantha Wendl.) production
for speciﬁc market dates can be very difﬁcult for three reasons. First, the African violet
is a day neutral species with respect to ﬂower initiation and development (Halevy, 1985);
therefore, the species does not possess a precise mechanism which can be used to induce
ﬂower initiation on a speciﬁc date. Second, the apical meristem grows indeterminately
and inﬂorescences develop in leaf axils. Not all leaf axils produce an inﬂorescence and
there is not a method available to determine which leaf axils will produce an
inﬂorescence. Third, once initiation has occurred, quantitative data are not available
relating the greenhouse environment to the rate of inﬂorescence development after

initiation has occurred.

82
ADT is the primary factor inﬂuencing the rate of plant development

(Hodges, 1991), not the relationship between DT and NT (Karlsson et al. 1988;
Berghage, 1990). However, the literature conﬂicts on whether ﬂower development of
the Afriean violet is a function of ADT or relationship between DT and NT. The
African violet has been classiﬁed as a species which ﬂowers faster when grown with
cooler DT than NT (Leopold and Kriedemann, 1975; Kimmins, 1980; Mastalerz, 1965)
based on data presented by Went (1957) which showed that African violets ﬂowered
more quickly when grown at 14C day temperature (DT) and 20C night temperature (NT)
than when grown at constant 20C. However, Hildrum and Kristoffersen (1969) observed
that time to ﬂower, the number of ﬂowers and buds per plant, inﬂorescences per plant,
and ﬂowers and buds per inﬂorescence were inﬂuenced by ADT, and not the relationship
between DT and NT.

Initiation and development of an inﬂorescence in the leaf axil is inﬂuenced by the
capacity of the leaves to export photosynthates . The transition from importing to
exporting earbohydrates from a leaf is related to leaf expansion (Turgeon, 1989). As a
result, ﬂowering of some species is inﬂuenced by PPF (Kinet, 1977) and plant leaf area
(Ramina et al., 1979). Likewise, a similar ﬂowering response has been observed on
African violet. Hildrum and Kristoffersen (1969) observed the number of ﬂowers and
buds per plant, inﬂorescences per plant, and ﬂowers and buds per inﬂorescence increased
as illuminance increased from 4 to 12 klux for 16 h (3.1 to 9.3 mol in2 d"). Stinson and
Laurie (1954) reported that ﬂower initiation and development of African violets were

inhibited when plants were grown in greenhouses and given an illuminance of no greater

83
than 300 footcandles during the brightest part of the day (< 2 mol m‘2 d“). Research

on the African violet which relates ﬂowering to leaf area has not been conducted.

The objective of our research was to develop a model which would predict
ﬂowering of the African violet. Our approach was to divide the model into three
components. First, a method was developed to predict which leaves formed an
inﬂorescence in the leaf axil. Second, a model was developed which related the
appearance of a visible ﬂower bud in the leaf axil with the length of the subtending leaf
blade. Third, quantitative relationships were developed which described inﬂorescence

development rate from VB to anthesis as a function of temperature.

Materials and methork

Predicting the occurrence of inﬂorescence development in a leaf axil. Modeling
inﬂorescence development of the African violet requires the ability to accurately predict
which leaves will have an inﬂorescence develop in the leaf axil. An experiment was
conducted to develop a method to predict which leaves would produce an inﬂorescence
in the leaf axil.

One hundred twenty Afriean violet ‘Utah’ plants possessing 8 to 10 leaves were
transplanted into 10 cm diameter pots and were placed into four 10 m2 glass greenhouses
which were set to maintain air temperatures of 15, 20, 25, and 30C from Dec. 1990
through Feb. 1991. Each greenhouse was divided into three PPF treatments. Plants in
the low PPF treatment had neutral density shade cloth placed above the plants to reduce
the natural PPF by 50% . Plants in the medium PPF treatment received the natural PPF

environment. The high PPF treatment had natural PPF levels plus 100 pmol rn‘2 s" from

84
0600 to 1800 h each day (4.3 mol m2 d“) provided from 400W high pressure sodium

lamps. One layer of 50% PPF reduction neutral density shade cloth was pulled over
plants in all PPF treatments when the natural PPF increased above approximately
300 umol m'2 s". The average daily integrated PPF were 2.6, 4.5 , and 9.5 mol rn‘2 d’1
for the three PPF treatments over the course of the experiment. PPF was monitored at
eanopy level with quantum sensors (Ll-COR, LI-1909A).

The leaf axil of each leaf on a plant was examined for the presence of an
inﬂorescence greater than 2 mm long, and the LBL of each leaf was measured after
10 ﬂowers had opened on the plant. The percentage of leaf axils possessing an
inﬂorescence greater than 2 mm long was calculated for each leaf number.

Leaves were numbered so that the most recently unfolded leaf at the time of
transplanting was designated as leaf number zero, and each successive leaf to unfold was
designated leaf number 1, 2, 3, etc. . . Leaves which unfolded prior to the start of the
experiment were numbered, -1, -2, -3 etc..., from the second most recently unfolded to
the oldest leaf.

Model development - Time to VB. The date of ﬂower initiation in the African
violet is uncertain, therefore, some physical measure other than time is necessary to
predict when an inﬂorescence will be macroscopically visible (2 mm long) in the leaf
axil. We choose to relate the time to VB to the LBL of a subtending leaf on a pre-VB
plant; therefore, leaf blade measurements could be used to predict VB. Assuming that

an inﬂorescence developed in the leaf axil of a given leaf, the time to VB would be:

85
where LBL“, represents the leaf blade length at VB, and LERT represents the rate of leaf

blade extension in mm d‘ at the particular temperature (T).
The following nonlinear function (Reed et al., 1976; Landsberg, 1977) was used

to describe LER as a function of temperature:

LER = aCF-Tudﬂm- ’ (2)
a = LERm/(Foa-Tudﬂm-Toa)’ (3)
b = (rm-Two/(rW-Tm (4)

where Tm, and Tm refer to the minimum and the maximum temperature at which LER
is zero. T0,, is the temperature at which the optimum LER occurs, T represents the plant
temperature, and LER,,“ refers to the maximum value for LER at To“.

A polynomial function was used to describe the LBL“, as a function of

temperature and daily integrated PPF:

LBL", = B, + arr + 13,111“2 + B3*PPFm (5)

where T represents temperature and PPFDI represents the daily integrated PPF.

Model development - Time from VB to anthesis. Once VB has appeared in the
leaf axil, the time until anthesis is determined by the rate of inﬂorescence development.
Inﬂorescence development was quantiﬁed with a phasic development scale which
identiﬁed nine stages of development. As the inﬂorescence of an African violet grows

through the leaf canopy, the pedicel and peduncle curve to protect the primary ﬂower

86
bud. The degree of curvature of the peduncle and pedicel was used to identify the stages

of the phasic development scale (Figure l & Table 1).
The number of days required for each stage of inﬂorescence development was

linearly related to temperature and was predicted with the following equation:
Number of days at stagex = Sx * T + Ix (6)

where Sx equals the slope of the regression, and Ix is the intercept. T represents
temperature, and X is the stage number.

The number of days required for an inﬂorescence to develop from any stage of
development to anthesis ean be predicted by integrating the number of days required for

each individual stage.

Days from stagex 8
toﬁrstopenﬂower =E(S,*T+I,) (7)

r=X
Comprehensive ﬂowering model. Functions (1) and (6) were combined to create
a comprehensive model which predicted the number of days until the ﬁrst open ﬂower

based on measurement of the LBL, daily integrated PPF, and temperature.

Days to ﬁrst Eq. 1 + Eq. 6 ...VB not present
open ﬂower = (8)

Eq. 6 ...VB present

87

Parameter estimates. Linear regression was used to describe the increase in LBL
from the time of leaf unfolding until the time the leaf blade extended to 40 mm. The
slope of the linear regression function was calculated for leaf numbers -1 to 2.

Parameters for the nonlinear function describing LER were estimated (Table 2)
using SAS procedure NLIN (SAS Institute, Inc., 1989).

Parameters for the multiple linear regression used to quantify the effect of
temperature and PPF on LBL", (Table 2) and parameters for the linear regression used
to quantify the time required for each stage of inﬂorescence development were estimated
(Table 3) using SAS procedure GLM (SAS Institute, Inc., 1989).

Experimental designs : The influence of PPF on inﬂorescence initiation and
development. One hundred African violet ‘Utah’ plant with 8 to 10 unfolded leaves were
transplanted into 10 cm diameter pots (450 cm’). Plants were placed into four PPF
treatments in each of two walk-in growth chambers (HotPack, Model UWP 3009-2,
Philadelphia, PA) which maintained plant temperatures at 2611C DT and 23i1C NT.
Plant temperature was measured by inserting a hypodermic needle thermocouple probe
(Omega Hypl-30-l/2-T-G-60-SAP-M) into stem, petiole, and leaf tissue. Plant
temperature was calculated as the mean temperature of the plant tissues. Cool white
ﬂuorescence lamps delivered a PPF of 23, 46, 92, or 181, pmol m’2 s’1 for 12 h (1'1
resulting in daily integrated PPF of l, 2, 4, or 8 mol at2 d". Layers of neutral density
shade cloth were placed above the plants to produce the PPF treatments within each
temperature treatment.

The leaf axil of each leaf on a plant was examined for the presence of an

inﬂorescence greater than 1 cm long after 10 ﬂowers had opened on the plant. The

88
percentage of leaf axils possessing an inﬂorescence greater than 1 cm long was calculated

for each leaf number. Analysis of variance was performed to determine the inﬂuence
of PPF on each leaf number.

Influence of the relationship between DT and NT on time to ﬂower. Eighty
African violet ‘Utah’ plants were placed into four walk-in growth chambers (Hotpack,
Model UWP 3009-2, Philadelphia, PA) with air temperature setpoints at 15, 20, 25, and
30C. Five plants per treatment were moved between chambers at 0800 and 2000 h each
day in order to create a total of 16 DT/NT treatment combinations. Cool white
ﬂuorescent lamps (Philips VHO F96T12/CW/VHO) delivered a PPF of 167 pmol m‘2 s‘1
for 12 h (1“ resulting in a daily integrated PPF of 7 mol m“2 d".

The dates of anthesis were recorded for the ﬁrst ﬁve ﬂowers of each plant. A
ﬂower was considered open when the ﬁve petals formed a planar surface. Plant
temperature was frequently 1 to 2C higher than air temperature during the photoperiod
and 1 to 2C lower than air temperature during the dark period. Actual plant
temperatures were used in regression analysis.

Inﬂuence of temperature and PPF on the time ﬁom transplant to VB.
Twenty-seven African violet ‘Sparkle’ plants possessing 8 to 10 leaves were transplanted
into 10 cm diameter pots and were placed in three growth chambers (Conviron, Model
E-15) which were set to maintain plant temperature at 18, 22, and 26C. Each growth
chamber was divided into three PPF treatments. Cool white ﬂuorescence lamps
(F72T12/CW/VHO) delivered a PPF of 50, 100, or 200 pmol in2 s‘1 for 12 h (1'1

resulting in daily integrated PPF of 2.2, 4.3, and 8.7 molm'2 d". The actual

89
temperature of the plants in the 2.2 mol m'l d‘I treatment within the 18C growth chamber

was approximately 16C.

LBL was measured and the leaf axil was examined for the appearance of VB on
all leaves every 2 to 3 days over a period of 64 days. Data were statistically analyzed
as a split plot design with temperature as the main plot and PPF as the split plot. The
SAS GLM procedure was used to for analysis of variance (SAS Institute, 1989).

Inﬂuence of temperature and PPF on the rate of inﬂorescence development from
VB to anthesis. Thirty-six African violet ‘Sparkle’ plants with four or more developing
inﬂorescences were placed into three growth chambers (Conviron, Model E-15,
Asheville, NC) set to maintain plant temperature at 18, 22, and 26;:1C. Each growth
chamber was divided into three PPF treatments. Cool white ﬂuorescence lamps
(F72T12/CW/VHO) and 60W incandescent bulbs delivered a PPF of 50, 100, or
200 pmol m'2 s“ for 12 h cl'l resulting in daily integrated PPF of 2.2, 4.3, and
8.7 mol rn‘2 d".

The stage of inﬂorescence development, inﬂorescence length, and the diameter
of the primary ﬂower bud were measured every 2 to 3 days for 30 days. The time
required for an inﬂorescence to develop through a stage was measured on the
4 to 5 inﬂorescences present at the start of the experiment and also on the ﬁrst
2 to 3 inﬂorescences which appeared during the experiment. Inﬂorescence length was
measured from the point of stem attachment to the uppermost part of the inﬂorescence.
Data were statistically analyzed as a split plot design with temperature as the main plot
and PPF as the split plot. The SAS GLM procedure was used to for analysis of variance

(SAS Institute, 1989).

90
Greenhouse validation of the comprehensive ﬂowering model. Twenty-four

African violet ‘Sparkle’ plugs were placed in two greenhouses which were set to maintain
air temperatures of 20 and 25C from March to May 1991. Each greenhouse was divided
to provide two PPF treatments. The plants in the high PPF treatment received natural
PPF, while the low PPF treatment had two layers of the neutral density cloth placed
above the plants which reduced natural PPF by 75 96 . One layer of neutral density shade
cloth (50% reduction) was pulled over the plants in both treatments when the natural PPF
exceeded approximately 300 pmol m'2 s“ . The average daily integrated PPF for the high
and low PPF treatments over the time of the experiment was 3.2 and 8.2 mol in2 d”,
respectively. The average plant temperature in the two greenhouses over the time of the
experiment was 21.5 and 25.0C.

LBL, stage of inﬂorescence development, and inﬂorescence length were measured
three times per week.

Data collected during the validation were used to test the model’s prediction of
the time of VB and anthesis for individual inﬂorescences. LBL at the time of transplant
was used to predict the number of days to VB of leaf numbers -3, -2, -1, and 0. LBL
at the time of leaf unfolding was used to predict the number of days to VB of the ﬁrst
ﬁve leaves which unfolded during the experiment, i.e. leaf numbers 1 through 5. The
date of VB was used to predict the number of days to anthesis for each inﬂorescence
developing in the axils of leaf numbers -3 to 5.

General procedures. The plants grown in all of the experiments were subirrigated
with a nutrient solution consisting of 3.6 mmol N and 1.3 mmol K from calcium and

potassium nitrate. Electrical conductivity of the media in the root zone was maintained

91
between 0.5 and 1.0 1118 using the 2:1 water/soil (vol/vol) method (Warncke and

Krauskopf, 1983). Phosphoric acid was included in the nutrient solution as needed to

maintain media pH between 5.5 and 6.5.

Results

The occurrence of an inﬂorescence developing in the lenf axil. The occurrence
of an inﬂorescence developing in the axil of a leaf was related to the age of the leaf at
transplanting (Figure 2). Inﬂorescences did not developed in the axils of leaf numbers
-12 to -7. The percentage of leaves forming an inﬂorescence increased from 0 to 100%
as leafnumber increased from -6 to -1, was 100% for leaves -1 to 2, and then decreased
back to 0% as leafnumber increased from 2 to 9.

The percentage of leaves with inﬂorescences was also correlated with leaf size at
the time of data collection (Figure 2). Leaf blades which grew to 40 mm in length or
more formed inﬂorescences in more than 60% of the leaf axils. The six oldest leaves
at transplant failed to grow to more than 30 mm in length after transplant, and
inﬂorescences did not develop in their leaf axils. LBL increased from 30 to 52 mm as
leaf number increased from -6 to 0, and the occurrence of inﬂorescences increased from
0 to 100%. LBL decreased from 52 to 12 mm as leaf number increased from
1 to 13 and the occurrence of inﬂorescences decreased from 100% to 0%.

The inﬂuence of PPF on inﬂorescence initiation and development. The percentage
of inﬂorescences developing in the leaf axils increased as the daily integrated PPF
increased from 1 to 4 mol m’2 (1'1 (Figure 3). Increasing the daily integrated PPF from

4 to 8 mol m2 (1" did not result in any further increase in ﬂower initiation and

92
development. Leaf number -1 was the ﬁrst leaf to consistently produce an inﬂorescence

in 100% of the leaf axils.

The eﬂ‘ect of the DT and NT relationship on time to ﬂower. The number of days
from transplant to ﬂower was a function of ADT (Figure 4), not the relationship between
DT and NT; therefore, ADT was used to develop the ﬂowering model. Leaves of plants
grown at 30C DT were chlorotic, and inﬂorescences did not develop.

Time of VB appearance in the leaf axil. LBL was a linear function of time from
7 mm to approximately 40 mm in length for all experimental treatments (Figure 5). The
slope of the linear regression line represented the LER for a given leaf. The maximum
rate of leaf extension was estimated at 1.26 mm dayr1 which occurred at 24.0C
(Figure 6). Tm, and Tm for LER were estimated at 13.8 and 29.0C, respectively.

LBL“ was inﬂuenced by temperature and daily integrated PPF (Figure 7). A
polynomial function was used to describe the inﬂuence of temperature and daily
integrated PPF on LBL“. LBL“, increased as temperature increased from 18 to 22, then
decreased as temperature was increased further to 26C. Final LBL displayed similar
response to temperature (data not shown). LBLVB decreased as the daily integrated PPF
increased from 2.2 to 8.7 mol m" d".

A model based on measured LBL and LER was created to predict the time
required before the appearance of VB in the leaf axil (Eq. 1). The predicted number of
days before VB appeared in the axil a leaf which had a 6 mm long leaf blade decreased
from 47 to 29 days as temperature increased from 18 to 26C (Figure 8).

Timeﬁvm VB to anthesis. A phasic development scale was developed to describe

inﬂorescence development from VB to anthesis (Table l & Figure 1). Linear functions

93
were utilized to describe the number of days required for an inﬂorescence to pass

through each developmental stage (Figure 9). A model based on current stage of
development and temperature was created to predict time to anthesis from any stage of
development (Eq. 7). The predicted number of days from VB to anthesis decreased from
40 to 28 as temperature increased from 18 to 26C (Table 4).

Inﬂorescence length and bud diameter increased with respect to the developmental
stage (Figure 10 A & B). Temperature and PPF did not inﬂuence inﬂorescence height
or bud diameter at each stage of development.

Greenhouse model validation. The comprehensive ﬂowering model accurately
predicted the time to VB and time from VB to anthesis (Figure 11A, B, & C). A
distribution of the deviation between the observed and the predicted number of days show
that the model had the tendency to overpredict the actual time required

(Figure 12A, B, C).

Dbcussion

Leaf area and PPF inﬂuence the ability of a leaf to become a strong source of
photosynthates. Transplanting a plant from an environment where there is strong
competition with surrounding plants for photons and the container capacity of the media
is relatively low to an environment where the competition between neighboring plants for
photons has been removed and the container capacity has increased tenfold inﬂuences leaf
expansion and the capacity of the leaf to intercept photons. Consequently, leaves which
unfold and expand prior to transplant do not expand to their full size, while the youngest

leaves at the time of transplant and the leaves which expand after transplant will expand

94

to their full size. As a result, the youngest leaves at the time of transplant will be the
ﬁrst leaves to have the capacity to become a strong source of photosynthates, thus a high
percentage of these leaves will have an inﬂorescence develop in the leaf axil.

Most commercial producers of African violets purchase small plants from a
propagator. The small plants are then transplanted and grown until they ﬂower, usually
3 to 4 months after transplant. The developmental status of a small plant when received
by the grower can vary due to the age of the plants and the methods of production by the
propagator. The model presented in this paper will be useful for predicting anthesis of .
a crop at the time of transplant and for tracking the progress towards anthesis. The use
of the inﬂorescence development model requires estimation of the ﬁrst leaf to form an
inﬂorescence. The proposed method of determining the ﬁrst leaf to form an
inﬂorescence, i.e. using leaf number -1 , is based on the assumption that ﬂower initiation
has not occurred prior to transplanting. Destructive examination of the leaf axils prior
to transplant occasionally reveals that ﬂower initiation has already occurred on some of
the plants. Consequently, a model must be developed which will predict the ﬁrst leaf
to produce an inﬂorescence based on the developmental status of the plants at the time
of transplant.

Once an inﬂorescence is visible in the leaf axil, the rate at which the inﬂorescence
develops to anthesis was a function of temperature. A phasic development scale was
used in our model to predict anthesis. However, inflorescence height and bud diameter
could have also been used. Our observations indicate that inﬂorescence height and bud
diameter are inﬂuenced by cultivar, thus a model based on inﬂorescence length and bud

diameter would not accurately predict anthesis for many cultivars.

95
Plants grown in growth chambers with ﬂuorescent and incandescent lamps have

leaves which are less pliable than leaves of plants grown under natural PPF with the
same PPF. As a result, the inﬂorescence is met with greater resistance as it penetrates
the leaf canopy. We observed that the number of days required for an inﬂorescence to
pass through Stage 5 during the model development experiment performed in a growth
chamber appeared to be higher than in the greenhouse validation experiment. This
potential error may explain the tendency of the ﬂowering model to overpredict the
number of days form VB to anthesis.

In summary, the time required for an African violet to ﬂower was divided into
two steps: 1) the time from transplant to the appearance of a VB in the leaf axil of the
plants 2) the time from VB to anthesis. Assuming that ﬂower initiation has not occurred
prior to transplanting, an inﬂorescence will develop in the second most recently unfolded
leaf at the time of transplanting when the plant is provided a daily integrated PPF of
22 mol in2 d“. An inﬂorescence will be macrOSCOpically visible in the leaf axil when
the leaf blade of this leaf extends to approximately 39 to 45 mm long, depending on the
temperature and PPF environment; therefore, the time required to the appearance of VB
in the leaf axil is dependent on the rate at which the leaf extends, which a function of
temperature. A phasic development scale was created to describe the current
developmental status of an inﬂorescence. The number of days required for an
inﬂorescence to develop through each stage of inﬂorescence development was described

as a linear function of temperature from 18 to 26C.

LITERATURE CITED

Berghage, R.D., 1989. Modeling stem elongation in the poinsettia. PhD Diss.,
Michigan State Univ., East Lansing.

Hildrum, H., and T. Kristoffersen. 1969. The effect of temperature and light intensity
on ﬂowering in Saintpaulia ionantha Wendl. Acta Hort. 14:249-255.

Hodges, T. 1991. Temperature and water stress effects on phenology, p. 7-13. In T.
Hodges (ed.). Predicting crop phenology. CRC Press, Inc.

Karlsson, M.G., 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(1):70-74.

Kinet, J .M. 1977. Effect of light conditions on the development of the inﬂorescence
in tomato. Scient. Hort. 6:15-26.

Kimmins, R.K. Gloxinias, African violets, and other gesneriads, ed. Larson, R.A.,
Introduction to Floriculture, Academic Press Inc. 1980.

Landsberg, J .J . 1977 . Some useful equations for biological studies. Expl. Agric.
13:273-286.

Leopold, A.C. , and P.E. Kriedemann. Plant growth and development, Second edit.
McGraw-Hill Book Co. , 1975 .

Mastalerz, J .W. The greenhouse environment, J. Wiley and Sons, Inc., 1977.

Ramina, A., W.P. Hackett, and R.M. Sachs. 1979. Flowering in Bougainvillea. Plant
Physiol. 64:810-813.

Reed, K.L., E.R. Hamerly, R.F. Dinger, and P.G. Jarvis. 1976. J. Appl. Ecol.
13:925.

96

97

SAS Institute Inc. 1989. SAS/STAT User’s Guide, Version 6, Fourth edition, Vol. 2,
Cary, NC.

Stinson, R.F. and A. Laurie. 1954. The effect of light intensity on the initiation and

development of ﬂower buds in Saintpaulia ionantha. Proc. Amer. Soc. Hort.
Sci. 64:459—467.

Turgeon, R., 1989. The sink-source transition in leaves. Ann. Rev. Plant Phys. Plant
Mol. Biol. 40:119-138.

Warncke, D., and D. Krauskopf. 1983. Greenhouse growth media: testing and nutrition
guidelines. Mich. State Univ. Ext. Bull. E-1736.

Went, F.W. 1957. The experimental control of plant growth. Chronica Botanica, Co.

98

Figure l. Phenology of the inﬂorescence development. Each drawing corresponds
to a stage of development described in Table 1.

.aoanEEQ 8:88.33. ..o owﬁm

 

 

 

 

 

m N. c w v m N
w

. . a a. ..

 

 

 

 

 

100

Figure 2. Correlation between LBL and the percentage of leaf axils forming an
inﬂorescence with respect to leaf number.

101

a Gav oocoomoaocE
ca 5.2. m__x< ..oo1_

O O O O
8 6 4 2 O

. rrﬁﬂ.
n1-
.. .

O O O O O
6 5 4 3 2

o AEEV .820... .0 1_m1.

100

 

 

 

marrﬁrxr'fxvr'

 

 

 

 

 

 

 

0

1

-5 0 51015

—15-10

Leaf Number

102

Figure 3. The inﬂuence of daily integrated PPF on the percentage of leaf axils
forming an inﬂorescence with respect to leaf number.

Pct. led axils with an inﬂorescence

$

 

 

 

 

 

103

    

 

/ ............................................

/ ......................................

/ ......................

/ .................

/ .......

/ .....

/

/

M

-8-7-6-5-4-3-2-1012345678910 PPF
Leainumber

(mol m-2 d-1)

104

Figure 4. The number of days to ﬂower shown as a function of average daily
temperature. Data points and error bars represent treatment means and
95% conﬁdence intervals.

Days to Flower

105

120

 

 

 

 

 

l
100 -
l C)
80
60 -
. 4‘4
40 -
A DT<NT
* o DT=NT
20 4
a DT>NT
16 20 24 28

Average Daily Temperature (C)

106

Figure 5. The slope of the regression line (Y=0.913*X+3.68) represents the rate
of leaf extension over time during the ﬁrst 40 mm of leaf extension for
an individual leaf (R2=0.99).

107

 

 

 

 

O O O O O
5 4 3 2 1

AEEV £9.61. 2.2m. Be.

20 30 40 50 60 70

10

Days

108

Figure 6. The rate of leaf extension shown as a function of temperature. Data
points represent the treatment means after the data from the PPF
treatments were pooled at each temperature, and error bars display the
95 % conﬁdence intervals.

109

 

 

 

 

4.

2 0. 8 6. 4. 2
.l 1 O O O O
ATU EEV Bow. c0688.”. .81.

0.
0

16 18 20 22 24 26 28
Average Daily Temperature (C)

14

110

Figure 7. A model describing the inﬂuence of daily integrated PPF and temperature
on LBLVB.

111

50

 

 

(I) 45 ‘ FN
: 40 q V _U-\
0 35 .
:53
O" 30 -
C
3 25 ~
<1)
20 ~
3 A 2.2 mol m‘zd‘1
CD 15 1 0 4.3 mol m""d‘1
g 10 1 U 8.7 mol 111-2d-1
_1
5 .
Q - . -

 

 

 

16 18 20 22 24 26 28
Average Daily Temperature (C)

 

112

Figure 8. Surface response of the predicted number of days to VB as inﬂuenced by
temperature and measured LBL.

113

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- ..
41: ......
see .4 $1110...

 

 

 

 

 

xx

\\\
888%... s
\\\\\\\\§§ .1:

\xxxxxxeeeef.
\xxxxese
S§S-

xxxx
axe:
\\\\\\ \:

s“... .

S s ......

~ ...
......sm...
aw aw aw. am am a)“.

 

11111!

114

Figure 9. The inﬂuence of temperature on the number of days for an inﬂorescence
to develop through each stage of development. Symbols represent
treatment means and error bars represent the 95 % conﬁdence intervals.

115

A8 830.qu3

 

 

®N¢NNNONwP ®N¢NNNONmF ©N¢NNNONQF ®N¢NNNON®F
- P P h b n b - n n P P P b — n h p b b O
r 11.1 . .. - . - m
. 4 .. .
r 11 VII - T 1 e
u it gr L
- -- - - - m.
. .. .. .
.1m 3on ...1 5 003m .10 macaw mm 0...on 1 w
. u . ... . n . . . . . + . . n . . . . o
r ... A
r. ...... J l I N
v [I 1. a +
- -- - . .4
. .. . .
,. ... .. ...
.1 4 82m ..H. n 83m 1 N 83m .1 . 88m .. m

 

 

 

 

 

 

SADQ

116

Figure 10. The observed A) primary bud diameter and B) inﬂorescence length at
each stage of inﬂorescence development. Data from all temperature and
PPF treatments were pooled. Data points and error bars represent the
mean values and 95 % conﬁdence intervals for all treatments.

117

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

=0. 1.... .1
r}. 1 70.. 1

u 1 .0. r

u 1 I 1

a 1 n 1

a 1 x r

.1 ... 1

o1 .1

A .18 o1
w w m ...1. m 5 0mwmmmm0

AEEV 5.9.55 cam 508....

AEEV £96.. mocmomoaozs

23456789

1

Inflorescence Stage

Figure 11.

118

Comparison between the observed (0) and the predicted (solid line)
number of days from A) the initial LBL measurement to the time of
appearance of VB in the leaf axil, B) the number of days from VB to
anthesis, and C) the number of days from the initial LBL measurement
to anthesis.

Predicted

119

80 l I l I I l

 

7O ‘ Days to V8
60 -
50 -
40 q
30 - 0°
20 -
1O -
80 t

-1—
-
-

 

-
-
--
-

70 ‘V8 to Anthesis
60 ~

50‘
404
30- an a:
201
10-
80 %%1L%%%

db

 

7O -
60 q Days to Anthesis
50 1 08'“
4O - o
30 -
20 .
10 d
O l 1 1 1 l l

 

 

 

0 1o 20 3o 4 50 6
Observed

7O 80

Figure 12.

120

Distribution of the deviation between the observed and the predicted
number of days from A) the initial LBL measurement to the time of
appearance of VB in the leaf axil, B) the number of days from VB to
anthesis, and C) the number of days from the initial LBL measurement
to anthesis.

Number of Observations

121

 

_s
l\)

. Days to V8 i

—L
03030

Ad

ON-P-CDOJONON-h-OUCDONON-ﬁ

Days to Anthesis

—3—.\

 

 

—10—8—6—4-2O 2 4 6 810
Observed — Predicted Days

Days from VB‘ to Anthesis

 

 

 

I r f‘rﬁ

122

Table l. Phenological description for each stage of inﬂorescence development.

 

Stage

Lasso->5).—

Phenological description of inﬂorescence development
Reproductive bud becomes visible (2 mm) in the leaf axil
The peduncle subtending the primary bud becomes visible
The peduncle begins to curve
The pedicel has curved at a 90° angle with respect to the peduncle

The pedicel has curved at an angle less than 90° with respect to the
peduncle, and the secondary buds are located at the highest point of
the inﬂorescence

The primary bud and pedicel have begun to increase the angle with
respect to the peduncle, the pedicel is located at the highest point of
the inﬂorescence, and the primary bud begins to push through the leaf

canopy

The upper half of the pedicel and the primary bud are at a 90° angle
with respect to the lower half of the pedicel

The upper half of the pedicel has begun to straighten out, and the
petals have begun to unfold.

All ﬁve petals are reﬂexed and are at a 180° angle with respect to
each other

 

123

Table 2. Parameter estimates for nonlinear (LER) (Eqs. 2—4) and polynomial
(LBLvn) (Eq. 5) submodels of the time to VB model.

 

Asymptotic 95 %

Model Parameter Estimate conﬁdence interval

 

 

 

 

Lower Upper
LER T111. 13.8 11.6 15.9
Tm 29.0 16.99 41.02
T0,, 23.96 22.8 25.12
LER,“, 1.26 1.10 1.42
LBL“, Std. Error of Estimate
b0 -18.46 20.39
b, 6.38 1.83
b2 -0.155 0.04

b, -O.564 0.118

 

124

Table 3. Parameter estimates for the linear functions describing the
number of days required for each stage of inﬂorescence development (Eq. 6).

 

95 96 Conﬁdence Intervals

 

 

Stage Parameter Estimate Lower Upper
1 S, 0.23 0.11 0.35
I, 9.16 6.57 11.75

2 S; 0.26 0.11 0.41
I2 10.31 6.70 13.93

3 83 0.25 0.14 0.36
I3 9.93 6.80 13.06

4 S4 0.18 0.09 0.27
I, 7.59 4.93 10.25

5 S, 0.23 0.11 0.35
I, 12.08 8.32 15.84

6 S, 0.32 0.22 0.42
I, 11.69 8.31 15.07

7 S7 0.06 -0.03 0.15
I, 4.90 2.28 7.52

8 S, 0.04 0.01 0.07
I, 3.23 2.43 4.03

 

125

Table 4. The predicted number of days for an inﬂorescence to develop to anthesis
based on the current stage of inﬂorescence development and temperature (Eq. 7)

 

 

 

Temp. Stages of inﬂorescence development
(C) 3 4 5 6 7
18 40 35 30 24 20 12
20 37 32 28 22 18 l 1
22 34 30 26 21 17 10
24 31 27 23 19 16
26 28 25 21 18 15 9

 

p—s
N

 

UIMQGG
NNNNNW
COOOOO

 

   

"lllllllllllllllf