Tree wraps and microsprinkler irrigation for freeze protection of young citrus trees

MISSING IMAGE

Material Information

Title:
Tree wraps and microsprinkler irrigation for freeze protection of young citrus trees
Physical Description:
xii, 178 leaves : ill. ; 28 cm.
Language:
English
Creator:
Rieger, Mark, 1960-
Publisher:
s.n.
Publication Date:

Subjects

Subjects / Keywords:
Citrus fruits -- Frost protection   ( lcsh )
Plants -- Winter protection   ( lcsh )
Trees -- Frost protection   ( lcsh )
Horticultural Science thesis Ph. D
Dissertations, Academic -- Horticultural Science -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1987.
Bibliography:
Bibliography: leaves 165-177.
Statement of Responsibility:
by Mark Rieger.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001020618
oclc - 17886188
notis - AFA2032
sobekcm - AA00004846_00001
System ID:
AA00004846:00001

Full Text












TREE WRAPS AND MICROSPRINKLER IRRIGATION FOR
FREEZE PROTECTION OF YOUNG CITRUS TREES









BY

MARK RIEGER


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

UNIVERSITY OF FLORIDA


1987



































This dissertation is dedicated to my family members
and to the loving memory of Suzanne Lynn DiCenzo
















ACKNOWLEDGEMENTS


Deepest appreciation is extended to supervisory committee

chairman, Dr. F. S. Davies, and cochairman, Dr. L. K. Jackson, for

support and aid throughout the course of this project, and L. W.

Rippetoe for technical assistance and valuable moral support.

Special thanks are given to committee members Dr. G. H. Smerage,

Dr. W. J. Wiltbank, and Dr. G. Yelenosky for participating in

examinations and reviewing the manuscript; to Jonathan Crane for acting

as a sounding board and for his constructive criticism of seminars; to

Dr. E. Chen, Prof. G. Cook, Dr. A. Datta, Dr. D. R. Farber, Dr. W.

Huber, and Dr. J. D. Martsolf for valuable assistance with simulation

programs; to Steven Hiss for assistance with graphics; to Dr. G. A.

Couvillon for providing the incentive to finish the dissertation

(nearly) on schedule; and last but not least, to my family members and

friends for support and timely diversion of my attention to the finer

things of life.

















TABLE OF CONTENTS
Page

ACKNOWLEDGEMENTS ........................................ ..... iii

LIST OF TABLES ........................................ ....... vi

LIST OF FIGURES ................................................... vii

ABSTRACT ........................................ ................ xi

CHAPTER

I INTRODUCTION ....................................... .... 1

II REVIEW OF THE LITERATURE ................................. 4

Introduction ........................................ 4
Freezing Injury, Stress, and Cold Acclimation ......... 4
Freezing Injury, Stress, and Cold Acclimation
in Citrus ............................................. 10
Freeze Protection Methodology ......................... 15

III TRUNK TEMPERATURE, FREEZE SURVIVAL AND REGROWTH
OF YOUNG CITRUS TREES AS AFFECTED BY TREE WRAPS
AND MICROSPRINKLER IRRIGATION ............................. 26

Introduction ........................................... 26
Materials and Methods ................................ 27
Results and Discussion ................................. 29
Conclusions ............. ........................... 39

IV MICROCLIMATE OF YOUNG CITRUS TREES PROTECTED BY
MICROSPRINKLER IRRIGATION DURING FREEZE CONDITIONS ........ 43

Introduction ............................. ........... 43
Materials and Methods ................................. 44
Results and Discussion ............................... 46
Conclusions ........................................... 55

V TRUNK TEMPERATURE, LIGHT INTENSITY, AND SPROUTING OF
WRAPPED AND UNWRAPPED YOUNG 'HAMLIN' ORANGE TREES
FOLLOWING A FREEZE ........................................ 56

Introduction ............................................. 56
Materials and Methods .................................. 57
Results and Discussion ................................ 58
Conclusions ........................................ ... 64










VI MODELING AND SIMULATION OF TREE WRAPS AND
MICROSPRINKLER IRRIGATION FOR YOUNG CITRUS
FREEZE PROTECTION ........................................ 65

Introduction ........................................... 65
Model development ...................................... 67
Simulation ............................................ 83

VII THERMAL PROPERTIES AND SIMULATION OF FREEZE
PROTECTION PERFORMANCE OF TREE WRAPS FOR
YOUNG CITRUS TREES ........................................ 88

Introduction ........................................ 88
Materials and Methods .................................. 90
Results and Discussion ................................ 95
Conclusions ............................................ 119

VIII SIMULATION OF PROCESSES INVOLVED IN MICROSPRINKLER
IRRIGATION FOR FREEZE PROTECTION OF YOUNG CITRUS
TREES ................................. *o .. ............. 120

Introduction ....................................... 120
Materials and Methods ................................. 122
Results and Discussion ................................. 125
Conclusions ........................................ 146

IX CONCLUSIONS ............................................... 148

Field Research ......................................... 148
Laboratory and Computer Simulation Analyses ............ 149

APPENDIX

A TREE WRAP SIMULATION PROGRAM .............................. 151

B TREE WRAP AND MICROSPRINKLER IRRIGATION SIMULATION
PROGRAM ................................................... 156

LITERATURE CITED ........... .......................... ........ .... 165

BIOGRAPHICAL SKETCH ...... .............. ......................... 178
















LIST OF TABLES


Page


Table 3-1.



Table 3-2.




Table 3-3.






Table 5-1.



Table 6-1.

Table 7-1.



Table 7-2.




Table 7-3.







Table 7-4.


Effect of microsprinkler irrigation rate and spray
pattern on height of live wood, shoot dry weight,
and number of shoots per trunk on 22 April 1985 ........ 34

Height of live wood of young 'Hamlin' orange trees
on 14 April 1986 as influenced by microsprinkler
irrigation rate during several freezes in Dec.-Jan.
1985-86 .................. .................... ...... 38

Dates of freezes, minimum air temperatures, most
efficient irrigation rates, and coefficients of
determination from quadratic regressions of trunk
heating efficiency vs. irrigation rate during Dec.
-Jan. 1985-86 .......................................... 41

Numbers and dry weights of trunk sprouts on freeze
damaged young 'Hamlin' orange trees on 13 May 1986
as influenced by various trunk wraps ................. 62

Definition of symbols used in model development ........ 87

Thermal conductivity, density, specific heat,
and thermal diffusivity of tree wraps used for
young citrus freeze protection ......................... 96

Minimum trunk and air temperatures of young 'Hamlin'
orange trees wrapped'with fiberglass, fiberglass
with water containers or styrofoam wraps on 4 mild
freeze nights in 1987 .................................. 100

Predicted minimum trunk temperatures for the
20-cm height for simulated freezes when air
temperature decreases from 00 C_ o -5 C in 8 hr
with windspeed of 0.5 + 0.1 m s using either
air temperature or mean radiant in calculation
of radiant heat transfer ............................... 109

Simulated minimum trunk temperatures underneath
fiberglass wraps at various heights from simulations
using different soil temperature regimes and freeze
conditions from 21 Jan. 1985 ........................... 118
















LIST OF FIGURES


Page


Fig. 3-1.



Fig. 3-2.



Fig. 3-3.



Fig. 3-4.






Fig. 3-5.


Fig. 4-1.




Fig. 4-2.





Fig. 4-3.



Fig. 4-4.


Fig. 4-5.


Trunk temperatures of 2-year-old 'Hamlin' orange trees
during an advective freeze of 20 Jan. 1985 as influenced
by microsprinkler irrigation and tree wraps ............. 30

Trunk temperatures of 2-year-old 'Hamlin' orange trees
during a radiative freeze of 26 Jan. 1985 as influenced
by microsprinkler irrigation and trunk wraps .......... 31

Effect of irrigation rate on trunk heating efficiency
of microsprinkler irrigation treatments for young
'Hamlin' orange trees ............................. ..... 33

Trunk temperatures of 2-year-old 'Hamlin' orange
trees and air temperatures during advective freezes
of 25-26 Dec. 1985 aad 27-28 Jan., 1986 for 0, 12,
22, and 38 liter hr microsprinkler irrigation
treatments .......................... ......... ........... 37

Trunk heating efficiency as a function of irrigation
rate for 5 freezes during Dec.-Jan. 1985-86 ............. 40

Air temperature in the canopy of 2-year-old 'Hamlin'
orange trees during a severe advective freeze on
20-21 Jan. 1985 and a radiative freeze on 26-27
Jan. 1985 ............................................ ... 48

Dewpoint temperature in the canopy of 2-year-old
'Hamlin' orange trees during a severe advective
freeze on 20-21 Jan. 1985 and a radiative freeze
on 26-27 Jan. 1985 ...................................... 49

Net radiation above 2-year-old 'Hamlin' orange trees
during a severe advective freeze on 20-21 Jan. 1985
and a radiative freeze on 26-27 Jan. 1985 ............... 51

Net radiation above 2-year-old 'Hamlin' orange trees
during a radiative freeze on 11-12 Jan. 1987 ............ 52

Soil temperature measured 1 cm below the surface and
next to trunks of 2-year-old 'Hamlin' orange trees
during freezes on 25-26 Dec. 1985, 26-27 Dec. 1985,
and 27-28 Jan. 1986 ..................................... 54










Fig. 5-1.



Fig. 5-2.



Fig. 6-1.



Fig. 6-2.


Fig. 6-3.




Fig. 6-4.


Fig. 7-1.


Fig. 7-2.





Fig. 7-3.



Fig. 7-4.






Fig. 7-5.



Fig. 7-6.



Fig. 7-7.


viii


Diurnal trunk temperatures of young 'Hamlin' orange
trees underneath various trunk wraps during a typical
sunny day and a cloudy day in March, 1985 ............... 59

Diurnal photosynthetic photon flux underneath various
tree wraps on 2-year-old 'Hamlin' orange trees during
a sunny day and a cloudy day in March, 1985 ............. 61

Conceptual model of the tree wrap/microsprinkler
irrigation system for young citrus freeze protection
illustrating heat transfer processes .................... 68

Three-dimensional representation of the model system
showing a representative finite region of heat storage ... 72

Schematic diagram of interception of irrigation
water by the tree wrap when microsprinklers were
positioned so that water sprayed above the top of
the wrap, or only on the wrap surface ................... 80

Cross-section of the model system showing 70% coverage
of the wrap surface with water ....................... 81

Schematic representation of the device used to measure
the thermal conductivity of tree wraps in-situ ......... 92

Observed trunk and air temperatures on 20-21 Jan.
1985, 26-27 Jan. 1985, 25-26 Dec. 1985, 26-27
Dec. 1985, and 27-28 Jan. 1986, and predicted trunk
temperatures from simulation of the freezes ......... 102-103

Regression of predicted vs. observed minimum trunk
temperature using 1-dimensional and 2-dimensional
models .................................................. 104

Predicted trunk temperatures for 4 commercially
available wraps, styrofoam, fiberglass, polystyrene,
polyethylene, during a simulated freeze where temperature
drops curvilinearly from 00 to -50 C in 8 hr, and
1
windspeed is 0.5 + 0. m s ........................ 106

Predicted trunk temperatures for air, a wrap with
thermal properties of dry sand, and a fiberglass
wrap ..................... ........ .. ...... ....... ..

Predicted trunk and air temperatures for young
citrus trees with fiberglass wraps during 6 different
simulated freezes .................................. 113

Predicted trunk temperatures as affected by variations
in trunk diameter with 13-cm diameter fiberglass
wraps, and variations in wrap diameter with 2.4-cm
diameter trunks ....................................... 115









Fig. 7-8.




Fig. 8-1.







Fig. 8-2.







Fig. 8-3.







Fig. 8-4.








Fig. 8-5.








Fig. 8-6.


Predicted minimum trunk temperatures of young citrus
tree trunks under fiberglass wraps as influenced by
windspeed when air temperature drops from 0 to -50 C
in 8 hr ................................................. 117

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 38 liters hr microsprinklers
during an advective freeze on 20-21 Jan. 1985,
and a radiative freeze on 26-27 Jan. 1985 ............... 126

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 57 liters hr microsprinklers
during an advective freeze on 20-21 Jan. 1985,
and a radiative freeze on 26-27 Jan. 1985 ............... 127

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 87 liters hr microsprinklers
during an advective freeze on 20-21 Jan. 1985,
and a radiative freeze on 26-27 Jan. 1985 ............... 128

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 22 liters hr microsprinklers
during advective freezes on 25-26 Dec. 1985 and
27-28 Jan. 1986 and a radiative freeze on 26-27
Dec. 1985 ............................................... 130

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 38 liters hr microsprinklers
during advective freezes on 25-26 Dec. 1985 and
27-28 Jan. 1986, and a radiative freeze on 26-27
Dec. 1985 .............................................. 131

Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fiberglass tree
wraps and irrigated with 57 liters hr microsprinklers
during advective freezes on 25-26 Dec. 1985 and
27-28 Jan. 1986, and a radiative freeze on 26-27
Dec. 1985 .............................................. 132









Fig. 8-7.


Fig. 8-8.



Fig. 8-9.


Fig. 8-10.







Fig. 8-11.






Fig. 8-12.







Fig. 8-13.





Fig. 8-14.


Predicted and observed trunk temperatures at the
20-cm height and air temperatures for 2-year-old
'Hamlin' orange trees wrapped with fi erglass tree
wraps and irrigated with 38 liter hr microsprinklers
during mild freezes on 11-12 Jan. 1987 and 23-24
Jan. 1987 ............................................... 133

Predicted minimum trunk temperatures plotted against
those observed at the 20-cm height for all irrigation
rates and dates used for validation ..................... 135

Cross-section of a wrap showing the thickness of
simulated ice on different regions of the wrap surface .. 136

Simulated trunk temperatures at the 20-cm height
for 2 cm-diameter 2-year-old citrus trees with
fiberglass tree wraps under severe advective and
radiative freeze conditions. Simulations were run
using different values of fraction of water intercepted
by the wrap ............................................. 138

Simulated trunk temperatures at the 20-cm height
for 2 cm-diameter 2-year-old citrus trees with
fiberglass tree wraps under severe advective and
radiative freeze conditions. Simulations were run
using different values of water temperature ............. 140

Simulated trunk temperatures at the 20-cm height
for 2 cm-diameter 2-year-old citrus trees with
fiberglass tree wraps under severe advective and
radiative freeze conditions. Simulations were run
using different values of percent coverage of the
wrap surface with water ............................... 142

Simulated minimum trunk temperatures at the 20-cm
height for 2-cm-diameter 2-year-old citrus trees
with fiberglass tree wraps under severe advective
and radiative freeze conditions. Simulations were
run at windspeeds of 0, 1, 4.5, and 10 m s .......... 145

Simulated trunk temperatures at the 20-cm height
for 2 cm-diameter young citrus trees with fiberglass
tree wraps under severe advective and radiative
freeze conditions. Simulations were run using
different values relative humidity ...................... 146










Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

FREEZE PROTECTION OF YOUNG CITRUS TREES
WITH TREE WRAPS AND MICROSPRINKLER IRRIGATION

By

Mark Rieger

August, 1987

Chairman: Frederick S. Davies
Cochairman: Larry K. Jackson
Major Department: Horticultural Science (Fruit Crops)


Tree wraps and microsprinkler irrigation provide a reliable and

economical alternative to soil banks and orchard heaters for freeze

protection of young citrus trees. Field experimentation and computer

simulation were employed to study the freeze protection capabilities of

tree wraps alone and in combination with microsprinkler irrigation.

Trunk temperatures of 2-year-old 'Hamlin' orange (Citrus sinensis (L.)

Osb.) trees during freeze conditions were greater for wrapped trees

irrigated with a 900 spray pattern than for those irrigated at the same

rate with a 3600 spray pattern. An irrigation rate of 38 liters hr-1

provided the greatest increase in trunk temperature per unit water

applied during several freezes. Air temperature, dewpoint and net

radiation in the tree canopy were not increased by microsprinkler

irrigation during freeze conditions, although soil temperature was 20 to

6 C higher for irrigated than for unirrigated trees.

Thermal properties of commonly used tree wraps were determined in

the laboratory. Thermal diffusivity was lowest for styrofoam wraps with

water containers, intermediate for 9-cm diameter fiberglass wraps, and

highest for thin-walled polyethylene and polystyrene wraps. Thermal

xi









diffusivity was indicative of freeze protection capability of tree

wraps.

The simulation model of the tree wrap/microsprinkler system

predicted trunk temperatures within 10 C of observed means when

simulating the effect of tree wraps alone, although predictions

generally were 10 to 30 C lower than observed means when simulating the

irrigation process. The rate of air temperature decrease, wrap

thickness and trunk diameter were positively correlated with level of

freeze protection in computer simulations of fiberglass wraps; however,

freeze duration was negatively correlated with level of freeze

protection. Windspeed had little effect on trunk temperature of

unirrigated wrapped trees, but trunk temperature of irrigated wrapped
-1
trees was reduced 50 C as windspeed increased from 0 to 10 m s -

Increasing the amount of water intercepted by the wrap and coverage of

the wrap surface with water increased trunk temperature during simulated

freezes. Temperature of the irrigation water was positively correlated

with trunk temperature for radiative but not advective freeze conditions

in simulations.


xii
















CHAPTER I
INTRODUCTION


Freeze damage of young citrus trees has been a problem since the

establishment of the Florida citrus industry. Young trees are

particularly prone to freeze damage due to their small size and vigorous

growth habit. Citrus growers spend $18 to $42 and 3 to 4 years bringing

a young tree into production (95). Traditionally, soil banks have been

used for freeze protection of young citrus trees (59). Soil banks

provide several degrees C protection to the lower portion of the tree,

but are labor intensive and often result in mechanical or disease damage

to the trunk. To avoid the problems associated with soil banks,

Rohrbaugh designed the first insulative tree wrap, consisting of rock

wool inside a 30-cm cylinder of asphalt felt which surrounded the lower

50 cm of a young tree (80). Results from preliminary tests with similar

wraps appeared promising; however, the types of tree wraps used

currently provide only 3 C protection in some cases, and virtually no

protection in others (68, 121).

Overhead irrigation, petroleum fuel heaters, and wind machines have

been used successfully to protect young and mature citrus trees.

However, overhead irrigation systems often are not designed to apply

water at rates necessary for protection under advective freeze

conditions (46). Heating groves or nurseries with petroleum products is

an effective means of cold protection, but rising fuel costs have

limited the use of this practice to high cash value crops (107). Wind









machines rely on the presence of an inversion layer (45), and therefore

are ineffective under advective freeze conditions.

A reliable and economical alternative to the conventional methods

listed above for freeze protection of young citrus is low-volume

microsprinkler irrigation. This system provides protection to the lower

portion of the tree only, much like a soil bank. Previous studies have

shown that trees irrigated with microsprinkers are usually killed above

a height of 50 to 70 cm during severe freezes (30, 104). However,

freeze-damaged trees often produce several vigorous shoots from the

remaining portion of scion wood and attain heights of 1 to 2 m by the

following autumn (104). The net result is that the grower loses about 6

months' growth on the tree but saves the cost of purchasing and

resetting new trees, which is particularly important considering the

scarcity and high prices of nursery trees due to the recent outbreak of

citrus canker disease in Florida.

Microsprinkler irrigation has been reported to alter the

microclimate of citrus plantings by raising the air temperature within

the canopy (15), creating fog or mist, and/or decreasing radiant heat

loss from the irrigated area (53, 107, 108). However, the effects of

microclimatic changes on young citrus tree temperatures and subsequent

freeze survival are unclear.

Recommendations regarding irrigation rate and spray pattern are

currently lacking, although preliminary studies indicate that 76 to 84
-1
liters hr- sprayed in a 900 pattern will protect young trees under most

freeze conditions (30, 107). Davies et al. (30) suggest that

microsprinkler irrigation combined with an insulative wrap provides

greater protection for young citrus trees than irrigation or wraps used









separately. However, wraps left on freeze-damaged trees may reduce

trunk sprouting and delay canopy reestablishment following a severe

freeze. Further research with tree wraps and microsprinkler irrigation

is needed to determine the effects of irrigation rate, spray pattern,

and wrap type on trunk temperature, survival and sprouting of young

citrus trees.

Modeling has been used successfully to predict overhead irrigation

rates required for freeze protection of tree crops (20, 46, 112).

Hence, modeling and simulation of the tree wrap/microsprinkler

irrigation system of freeze protection may yield useful information on

system design and behavior. Analysis of a model can produce information

unobtainable from field or laboratory experiments due to uncontrolled or

unrepresentative environmental conditions. Furthermore, experiments can

be performed with a simulation model more rapidly and inexpensively than

field experiments, improving the productivity and efficiency of the

freeze protection research program.

Research objectives of this dissertation were

1. to study the effects of tree wraps combined with various

microsprinkler irrigation rates and spray patterns on trunk temperature,

survival and regrowth of young citrus trees;

2. to determine whether microsprinkler irrigation alters the

microclimate of a young citrus tree during freeze conditions;

3. to determine the influence of various tree wraps on trunk

temperature and sprouting of freeze-damaged trees;

4. to model the dynamic behavior of trunk temperature during

freeze conditions as influenced by tree wraps and microsprinkler

irrigation.
















CHAPTER II
REVIEW OF THE LITERATURE


Introduction

Man's efforts to protect crops from freezing injury began at least

2000 years ago (83). In addition, the causes of freezing injury in

plants have been studied and speculated upon for well over 100 years

(124). Impetus for continued research in these areas stems from the

huge economic losses of agricultural commodities that occur annually due

to late spring or early fall frosts, or midwinter freezes. Literature

on freezing injury, protection and related topics is voluminous.

Several reviews (18, 75, 86, 91, 98, 116, 124, 135, 157, 167) and books

edited by Levitt (76) Olien and Smith (99), and Li and Sakai (78) have

covered various aspects of freezing stress and injury, and cold

acclimation and hardiness. The intent of this review is to discuss

general aspects freezing injury, resistance to freezing stress, and cold

acclimation in woody plants with special attention to citrus.

Herbaceous plants, with the exception of winter grasses (82), have

little or no resistance to freezing stress or ability to acclimate and

are therefore omitted. In addition, methods of freeze protection will

be reviewed.

Freezing Injury, Stress, and Cold Acclimation

Freezing Injury

Freezing injury results from ice formation in plant tissues and not

low temperatures per se (92). Manifestations of freezing injury include









sunscald and bark splitting of cree trunks, blackheart in trunks and

stems, death and abscission of leaves, flowers, and fruits, frost rings

on fruits, midwinter kill of dormant flower buds and stem cambia, and

death of the entire vegetative structures of cold tender plants (18,

135, 136). These visual symptoms ultimately result from processes which

occur at the cellular level.

Plant cells are injured or killed by either intracellular or

extracellular ice formation (75). It is widely accepted that

intracellular ice formation is always fatal to plant cells, although how

intracellular freezing kills the cell remains unanswered (124). It

follows that cells which survive freezing do so by tolerating

extracellular ice formation. Therefore, much attention has been devoted

to processes which occur during extracellular ice formation and the

resulting consequences.

Extracellular freezing may be summarized as follows. As

temperature decreases below 00 C, water in plant tissues initially

supercools few to several degrees (18, 86). Heterogenous nucleation

occurs in cell walls and extracellular spaces and ice crystals grow in

these areas at the expense of water inside the cells (91, 92). As long

as exosmosis of intracellular water keeps pace with the reduction in

external vapor pressure with decreasing temperature, equilibrium between

the intracellular solution and the external phase is maintained, and

intracellular freezing is avoided by colligative freezing point

reduction (86). However, supercooling and intracellular freezing are

favored at rapid cooling rates (> 100 C min- ), especially if the

permeability of the plasma membrane to water is low (85).










Extracellular freezing is tolerated to varying degrees in plant

tissues depending on species, time of year, preconditioning, plant part,

freezing rate and water content (135). Several hypotheses attempt to

explain causes of cell death due to extracellular freezing. Levitt (74)

hypothesized that intermolecular disulfide bonds formed among proteins

with exposed sulfhydryl groups during freeze dehydration of cells by

external ice. He stated that this resulted from the close proximity of

proteins due to the reduced volume of cytoplasm. Williams (140)

hypothesized that there was a minimum volume that cells attained during

freeze dehydration below which cell death occurred. This minimum volume

was the same for acclimated and unacclimated cells, but the temperature

at which minimum volume was reached was lower for acclimated cells due

to higher solute and bound water contents. Solution effects (i.e.,

concentration of toxic electrolytes, pH changes, precipitation of

compounds, salting out of proteins) may result in injury to cell

membranes and other cellular components during freeze dehydration

(86, 124). Recently, the behavior of the plasma membrane during

freezing was analyzed in isolated protoplasts (124). Apparently,

vesiculation and deletion of membrane material occurs in unacclimated

protoplasts, which causes altered osmotic behavior and cell lysis upon

thawing, whereas these processes are not observed in acclimated

protoplasts. Therefore, the plasma membrane may have an important role

in cold acclimation and tolerance of extracellular ice. However, none

of the above hypotheses are without limiting weaknesses, and no single

hypothesis applies to all plant cells and all freezing situations.









Resistance to Freezing Stress

Resistance to freezing stress is conferred by avoidance and/or

tolerance of ice formation within the plant (76).

Avoidance. Plants avoid freezing in low temperature environments

by desiccation, colligative freezing point depression, and supercooling

(75). Desiccation occurs in seeds and buds only, and freezing point

depression allows avoidance of freezing to about -40 C at most.

Alternatively, supercooling is a very effective avoidance mechanism, but

usually occurs in plants which also tolerate extracellular ice formation

(2, 42). One exception is shagbark hickory, which avoids freezing

injury to -400 C by supercooling of the entire volume of liquid water in

stems (18).

Tolerance. Reasons for difference in tolerance of extracellular

ice among species are unknown, although apparently, hardier plants

survive freezing of a larger amount of their osmotically active water

(1, 16, 18).

The latitudinal range of a particular plant species is related to

its relative tolerance of extracellular ice formation (17). Species

which tolerate little extracellular ice such as citrus, are limited to

southern latitudes where freeze events are less severe and occur

infrequently. Temperate woody plants such as Vitis, Prunus, and

Rhododendron tolerate extracellular ice in most tissues and avoid ice

formation in flower bud primordia via deep supercooling (4, 42, 142).

These plants are not found at latitudes where temperatures below about

-400 C occur, because deep supercooled cells freeze and die at this

temperature due to homogenous nucleation of intracellular water (17).

Species of the northern boreal forests such as Betula and Populus do not










deep supercool, but tolerate freezing of all their osmotically active

water outside the plasmalemma (18). These species are capable of

surviving immersion in liquid nitrogen (-1960 C) after slow cooling to

-300 C.

Cold Acclimation

Irrespective of freezing tolerance mechanisms, temperate woody

plants acclimate to low temperatures in 2 distinct stages. Stage I

acclimation is a phytochrome-mediated response initiated by short days

and enhanced by warm temperatures (38, 58, 64, 87). Moderate increases

in hardiness occur during stage I acclimation (135). A translocacable

"cold hardiness producing substance" presumed to be abscisic acid (63)

has been shown to be involved in stage I acclimation (37, 66, 125).

Stage II acclimation is initiated by exposure to low, nonfreezing

temperatures or frost, and does not involve translocatable substances

(37, 58). Plants acclimate rapidly and to their fullest potential

during stage II acclimation.

Irving and Lanphear (64, 65) showed that woody plants would

eventually acclimate to normal levels when exposed to long days and low

temperatures, suggesting that low temperature has an overriding

influence on photoperiod in cold acclimation. Furthermore, they showed

that plants acclimate substantially before becoming dormant under long

day, low temperature (100 C) conditions (64). Apparently, dormancy and

cold hardiness are separate processes which occur coincidentally in the

natural environment in woody perennials.

Tumanov and Krasavtsev (127) discovered a third stage of hardiness

which develops during prolonged exposure to subfreezing temperatures,

and is rapidly lost upon rewarming. They stated that prolonged exposure








9

to subfreezing temperatures allows sufficient time for water movement to

extracellular sites which prevents intracellular freezing, whereas

natural cooling rates are more rapid and result in supercooling and

intracellular freezing at higher temperatures. Similarly, Rajashekar

and Burke (118) found that prolonged exposure to subfreezing

temperatures reduced or eliminated low temperature exotherms indicative

of supercooling and intracellular freezing in Prunus flower buds.

Apparently, the permeability of the plasma membrane to water at

subfreezing temperatures and freezing rate dictate the ultimate

hardiness level a plant can achieve (85).

Several physical and biochemical changes occur in plants during

cold acclimation. Increased physical stability of the plasma membrane

observed during acclimation may result in reduced probability of

nucleation of the intracellular solution by external ice (124).

Increases in unsaturation of membrane lipids increases fluidity and

decreases the probability of membrane phase transition and subsequent

dysfunction of membrane-bound proteins (126). Permeability of the

plasma membrane to water increased during acclimation in Red-Osier

dogwood (88), but was similar in acclimated and aonacclimated

protoplasts isolated from rye leaves (34). The ratio of gibberellin to

abscisic acid-like substances decreased during stage I acclimation in

Acer leaves (63). Increases in soluble proteins, RNA, sulphydryl

compounds, sugars, amino acids, and organic acids occur during

acclimation of many plants (86). Water content, root hydraulic

conductivity, and stomatal conductance decreased in Red-Osier dogwood

during cold acclimation (54, 101, 135). As with the hypotheses of







10

freezing injury, none of the above physical and biochemical changes can

singularly explain the resistance to freezing injury in plants.

Freezing Injury, Resistance and Cold Acclimation in Citrus

Freezing Injury

Freezing injury to citrus can be economically devastating because

present and future crops are affected by a single freeze event. For

example, in Texas a single freeze in Dec. 1983 reduced the production of

the 1983-84 season by 70%, and extensive tree damage resulted in no

production the following year (123). In Florida, damage was estimated

at one half billion dollars from the 1962 freeze alone (151), and the

severe freezes of Dec. 1983 and Jan. 1985 reduced commercial citrus

acreage by over 80,000 ha in north-central Florida (141).

Manifestations of freezing injury in citrus range from partial

defoliation and fruit abscission to death of entire trees (151, 153).

Watersoaking appears as dull, dark areas on the leaf surface and is

indicative of ice formation within the leaves (113, 175). However, ice

formation in citrus leaves and stems is not always lethal (147, 175).

Bark splitting is thought to result from large ice masses forming in the

bark, creating pressure sufficient to force the tissues apart (153).

At the cellular level, Young and Mann (171) observed destruction of

vacuolar, chloroplast, and mitochondrial membranes in hardened and

unhardened sour orange leaves subjected to freezing at -3.30 and -6.7

C, respectively. They suggested that the nature of freezing injury in

leaves was physical, caused by ice crystal growth and subsequent

cellular disruption. However, they offered no explanation as to why

hardened leaves undergo less cellular disruption at the same freezing

temperatures than unhardened leaves, when both types of leaves contained










ice. Lower water content (167) and/or higher bound water fractions

(122) in hardened leaves may reduce the amount of ice formed at a given

temperature, resulting in less physical disruption than in unhardened

leaves. Among citrus species, however, Anderson et al. (1) found no

relationship between bound water content and freeze tolerance of leaves,

and suggested that hardier species tolerate a greater amount of their

water frozen than non-hardy species.

Resistance to Freezing Stress

Citrus species can survive extracellular ice formation within their

tissues (147, 175), although changes in ice tolerance during acclimation

remain unexplained (157). In addition, citrus avoids freezing injury by

supercooling to temperatures as low as -80 C in leaves and stems (158).

Citrus leaves are killed shortly after freezing occurs in the

supercooled state (69, 158). The extent of supercooling in young citrus

trees increased with cooling rate (164), cold hardening (61), and water

stress (154), and decreased in the presence of external moisture or ice

(155, 166). Also, the extent of supercooling varies with species and

cultivar (157), but not in such a manner as to provide an index of

interspecific or varietal differences in hardiness (158). Supercooling

may not be as important a factor as ice tolerance in citrus freezing

resistance because recent studies indicate that several tree species

supercool only 10 to 20 C before freezing under field conditions (2, 4).

Anderson et al. (1) suggested that negative pressure potentials

arise in citrus leaf cells during freeze plasmolysis, decreasing the

amount of ice formed at any given temperature. This is unlikely because

negative pressure potentials would have to be extremely low (< -10 MPa)

to decrease the amount of ice formed significantly (86), and the only








12

documented measurements of negative pressure potentials are much higher

(-6 kPa) (5).

Cold Acclimation

It has long been recognized that environmental conditions preceding

a freeze influence the freeze survival of citrus (94). Greater freeze

damage in Florida and Texas than in California under similar freeze

conditions has been attributed to higher winter temperatures in Florida

and Texas which reduce cold acclimation (23, 24). Several studies have

demonstrated that citrus tolerates lower freezing temperatures if

preconditioned for several weeks at low, nonfreezing temperatures (147,

163, 165, 173).

Generally, environmental conditions which result in inactivity or

lack of growth promote cold hardiness in citrus (167). Girton (50),

Fawcett (35) and Ivanov (cited in 167) were among the first to report on

the influence of cool (< 130 C) temperatures on growth and cold

acclimation of citrus. Seedlings exposed to alternating temperatures

(warm days and cool nights) developed superior hardiness to those

exposed to constant low (30 C) temperatures (96). Yelenosky (147, 151)

used hardening conditions of 2 weeks at 21.1/100 C (day/night) followed

by 2 weeks at 15.6/4.40 C in laboratory experiments which resulted in

maximum hardiness.

Light is necessary for cold acclimation in citrus (96, 144).

Length of the light period is positively correlated with growth (115,

163) but does not affect cold hardiness (96, 163) or induce bud dormancy

(167). Light allows continued photosynthesis and carbohydrate

accumulation during exposure to low temperature which is necessary for

cold acclimation in leaves and stems (160, 166). Although









photosynthetic carbon uptake decreases during cold acclimation (169),

metabolism of photosynthate also decreases (52, 160) which probably

allows for net carbohydrate accumulation. Furthermore, absence or

inhibited function of leaves disfavors cold acclimation (146), and

starch hydrolysis cannot account for increased levels of sugars in

leaves and stems (160).

Water stress increases both freezing tolerance (150) and avoidance

(154) in citrus. Wilcox (138) reported that low (50 C) root

temperatures induced cold hardening in citrus seedlings by reducing root

hydraulic conductivity and leaf water potential. Davies et al. (28)

reported that 'Orlando' tangelo trees not receiving irrigation in the

fall had less leaf and fruit damage than fall irrigated trees following

a radiative freeze. Conversely, Koo (71) found that trees irrigated

during the fall had less leaf and fruit damage following a freeze.

In the acclimated condition, cold hardiness varies widely among

citrus cultivars (157), but when unacclimated and actively growing,

hardy and nonhardy types are killed at similar temperatures (151). In

the acclimated condition, citrus cultivars/relatives can be ranked for

hardiness in decreasing order as follows: trifoliate orange, kumquat,

sour orange, mandarin, sweet orange, grapefruit, lemon, citron (157).

The hardiness of a scion cultivar is affected by the

species/cultivar used as a rootstock (40, 55, 172), although hardiness

differences due to rootstock are slight by comparison to cultivar

differences (157). Generally, more hardy species/cultivars used as

rootstocks impart more hardiness to a scion cultivar, hence the above

ranking also applies to scions grafted to the species/cultivars listed.










Several biochemical and physiological changes occur in citrus

during cold acclimation. Levels of soluble carbohydrates in leaves

increased rapidly during the first week of acclimation in 'Valencia'

orange, although levels in stems increased gradually over a 6-week

period (149). The principal sugars that accumulated in 'Redblush'

grapefruit were glucose, fructose, and sucrose (166). Increased levels

of soluble carbohydrates may lower the freezing point of the tissue

water colligatively, decrease the amount of ice formed at a given

temperature, and decrease the rate of ice growth through the tissues

(145, 160). All amino acids except proline, valine, and glutamic acid

decreased during cold acclimation at 100 C in 'Valencia' orange (149).

Proline accumulated in unacclimated, water-stressed citrus leaves (77),

hence increases in proline during cold acclimation may be induced by

reductions in water potential which occur during hardening (149).

Proline accumulation did not correlate well with cold hardiness among

citrus cultivars/species (152).

Water soluble proteins decreased in 'Redblush' grapefruit during

hardening, contrary to behavior of woody deciduous plants which

accumulate water soluble proteins during acclimation (166). No new

proteins were identified by gel electrophoresis following cold

acclimation of 'Valencia' orange trees (162). However, protein

denaturation increased with freeze damage, and was greater for

unhardened than hardened leaves (162). Increased levels of

hydroxyproline may result from protein denaturation during cold

acclimation (159).

Levels of unsaturated fatty acids and phospholipids increased

during cold acclimation of citrus leaves, and the magnitude of increase










was positively correlated with the relative cold hardiness of the 3

species analyzed (97). This is consistent with the theory that

increased membrane fluidity is partially responsible for increased

freezing tolerance in plants (86).

Freeze Protection Methodology

Methods of freeze protection were first developed 2000 years ago,

when early viticulturists protected grapevines from freezing with fires

fueled by prunings and dead vines (83). Since then, many different

methods of freeze protection have emerged. The concepts behind several

freeze protection methods are described in the following section with

reference to citrus, although nearly all of the methods can be applied

to other tree crops and cropping systems.

Site Selection

Students of pomology quickly learn that site selection is one of

the most important decisions a grower makes when establishing an orchard

(136). Latitude, topography, and proximity of bodies of water affect

the frequency and severity of freezes at a given site (9). Low-lying

areas or "frost pockets" are often 5* C colder than adjacent hilltops

under radiative freeze conditions, but may be slightly warmer than

hilltops under advective freeze conditions (70, 132). In Florida,

freeze probability and severity decrease with latitude (12, 70), and

sites downwind of large lakes can be 40 C warmer than upwind sites (9).

However, climatic changes can occur and render once productive sites

unsuitable for growth of a particular crop, which has apparently

occurred in citriculture in northern Florida (21).










Cultural Practices

Krezdorn and Martsolf (72) recently reviewed effects of cultural

practices on citrus cold hardiness and orchard temperatures. They

emphasized the importance of weed-free compact soil for increasing soil

heat storage and conduction, proper row orientation for air drainage,

maintenance of a full canopy to intercept radiant heat from the soil,

and practices which keep trees in a healthy condition and maximize cold

hardiness. Leyden and Rohrbaugh (80) observed higher temperatures and

less freezing injury on citrus trees in sites with chemical weed control

than in cultivated or sodded sites.

Orchard Heating

Orchard heating is the oldest and most reliable method of freeze

protection (48, 83, 129). Several different types of fuels and heating

devices have been used (83), although ordinances prohibit the use of

those which produce excessive smoke and residue (129). The high costs of

petroleum fuels preclude their use except for high cash value crops and

in nurseries (107).

Heaters provide protection by raising the air temperature and by

producing radiant heat (48). Smoke particles are not of appropriate

size to retard radiant heat losses from an orchard, thus no added

protection is obtained by smoke (89, 132). Radiant heat production is

particularly important during windy, advective freezes when the

convective component of heating is reduced (8, 137). Gerber (44)

compared several types of heaters and found chat radiant heat output was

highest for those with large metal chimneys or mantles. Heater

placement is an important factor because radiant heat decreases

exponentially with distance (8, 33, 80). Absorption of radiant heat by









trees is maximized by placing heaters within rows rather than between

rows (33, 137).

Pollution and expense are not the only problems associated with

heating. Heaters are often difficult to light and maintain at a

constant burning rate (143). Refueling heaters individually is labor

intensive, and automated pipeline systems for refueling can malfunction

due to clogged valves (8). Furthermore, heaters interfere with routine

orchard operations during the summer (83).

Four models exist which describe theory and concepts involved in

freeze protection with heaters. Martsolf (83) has reviewed the

strengths and weaknesses of these orchard heating models in detail.

Gerber's (43) model underestimates the heating requirement under most

freeze conditions, but is fairly accurate under calm, radiative freeze

conditions. Crawford's (25) model accounts for induced flow of air due

to air temperature (hence buoyancy) differences inside and outside the

orchard, which is lacking in Gerber's model, and predicts heat

requirement within 5% of field observations. The box model of Martsolf

and Panofsky is an embellishment of Crawford's model and is largely used

as a research tool (83). The orchard foliage temperature model (137) is

by far the most rigorous development of the orchard heating problem, but

adds little to the current body of knowledge on orchard heating

practices.

Wind Machines

Wind machines have been used for freeze protection since 1916

(129). Wind machines mix warmer air aloft with colder air in the

orchard when a stable inversion exists, hence they are only useful under

calm, radiative freeze conditions (132).










The amount of protection provided by wind machines depends

primarily upon inversion strength, power of the machine, and distance

from the machine (48). Leyden and Rohrbaugh (80) observed 10 to 3 C

increases in air temperature over a 2.8 ha area on nights when a 40 C

inversion existed. Brooks et al. (14) also reported 10 to 30 C air

temperature increases over an area of 4.2 ha when inversion strength was

7 C. Turrell and Austin (130) observed that wind machines provided

better protection on sites which had good air drainage than on sites

with windbreaks around the perimeter. Reese and Gerber (120) concluded

that on calm, clear nights wind machines can be expected to increase air

temperatures 10 C over a kidney-shaped area of 4 ha.

Wind machines are sometimes combined with heaters for freeze

protection, yielding results superior to those from either method used

alone (84, 130). Wind machines plus 20 heaters per ha produced higher

air temperatures than 40 heaters per ha alone (14).

Foam

Foam is an unstable mixture of liquid and gas which acts as an

insulative blanket when applied over crops (7). Bartholic (6) found a

100 C increase in temperature under foam applied to low-growing crops.

However, he found little or no elevation of leaf and air temperatures of

mature citrus trees covered with foam unless heaters were used in

conjunction with the foam. The additional cost and logistical

difficulties associated with foam application to trees make it

impractical to use (7).

Fog

Water droplets in clouds and fog are of appropriate size to cause

scattering of terrestrial radiation (133). Therefore, radiant heat loss









from orchards on freeze nights is minimized in the presence of fog.

Efforts to generate and maintain fog over crops for freeze protection

began about 20 years ago (129). Mee and Bartholic (90) have reviewed

the history and concepts associated with fog generation for freeze

protection. They point out that fog not only reduces radiant heat loss

from the orchard, but can also add heat through condensation and fusion

of water on plant surfaces.

With mature citrus trees, fog provided about 1.50 C more protection

than heaters during a -50 C freeze, and outperformed wind machines

during a -80 C freeze (89). However, in tests under windy conditions,

temperature increases were only 00 to 3.50 C and decreased with distance

from the fog source (36). Although operating costs are very low (89),

fogging systems are problematic due to the large number of emitters and

purity of water required for reliable operation (90).

Flooding

Flooding orchards either before or during a freeze event can

increase the temperature and heat capacity of the orchard floor (41).

Heat which radiates from the soil or water surface can increase air

temperatures 0.50 to 10 C (Davies personal communication). Air

temperature was 0.5 to 10 C higher when the same volume of irrigation

water was distributed among 3 furrows than when only 1 furrow was

flooded (13). Comparable air temperature increases were obtained in

citrus orchards with combinations of wind machines plus heaters, and

wind machines plus flood irrigation (129).

Growth Regulator Sprays

Attempts to alter the hardiness of citrus trees by spray

application of growth regulators have been largely unsuccessful (151).







20

Maleic hydrazide sprays applied during the fall prevented freeze injury

to grapefruit seedlings during a controlled freeze at -60 C (168), but

in field studies trees treated with maleic hydrazide showed similar

damage to unsprayed trees 4 months following a freeze (19). Ethephon,

which increased hardiness and delayed bloom of cherry buds (117), had no

effect on citrus hardiness (168).

Soil Banks and Tree Wraps

Trunks of young citrus trees are protected from freezing injury by

mounding soil or wrapping insulative materials around them during

periods of cold weather. Following severe freezes, the entire canopy

above the bank or wrap is killed, but vigorous sprouts from the

protected portion of the trunk rapidly reestablish a new canopy (129).

Soil banks

In 1904 Hume (59) wrote "no method of protecting trunks of citrus

trees from cold is more efficacious than banking", and with the possible

exception of microsprinkler irrigation (see next section) Hume's

statement remains true to this day. Hume also discussed disease and

insect problems associated with soil banking, and recommended that banks

be removed from trees as soon as danger of freezing is past to prevent

disease damage to the trunk. Other disadvantages of banking include

high labor costs and mechanical damage to trunks from construction and

maintenance of the banks (68). However, soil banks continue to be used

for freeze protection of young trees due to their superior insulative

properties (31).

Generally, trunk temperatures under soil banks are 30 to 120 C

higher than those of unprotected trees, depending on depth within the

soil bank (31, 143). Rates of trunk temperature decrease during a









freeze are similar at all depths, but temperatures deeper within the

bank are initially higher and start to decline later than temperatures

at shallower depths (31, 143). The low thermal diffusivity and large

volume of a soil bank are responsible for its superior freeze protection

capability (129).

Tree wraps

Tree wraps were introduced into citriculture in the mid-1950's to

circumvent the problems associated with soil banks while still providing

freeze protection for young trees (80). Although tree wraps provide

less protection than soil banks (67, 68, 170), they have been proven

effective during freezes in Florida (121) and Texas (56). Furthermore,

tree wraps provide protection from wind,radiation and rodent damage

(129), and prevent trunk sprouting (121).

Early research in Texas (80, 170) demonstrated that the thickness

of fiberglass insulation was positively correlated with trunk

temperature, and survival of young grapefruit trees was best with the

thickest wraps. These studies and others (121) showed that trunk

temperatures decrease from top to bottom of wraps much as in a soil

bank, but the magnitude of the temperature gradient was lower than in a

soil bank.

Fucik and Hensz (39) suggested that the ratio of rates of bark to

air temperature decrease was indicative of freeze protection capability

of a tree wrap, and a ratio of 0.55 or less was considered adequate for

freeze protection in most situations. However, trunk and air

temperatures decrease at nearly the same rate with soil banks (ratio of

bark and air temperature decrease is about 1.0) (31), which are superior

to tree wraps with respect to freeze protection. Turrell (129) stated









that the thermal diffusivity was the primary determinant of freeze

protection performance of a tree wrap, and this statement probably holds

more often than the above suggestion of Fucik and Hensz.

Several different types of tree wraps are currently used, each with

somewhat different insulative characteristics. A styrofoam wrap which

has 2 containers of water attached to its inner surface produced higher

trunk temperatures and allowed greater tree survival than other wraps in

field studies (30, 67). In laboratory studies (156), this wrap

maintained trunk temperatures of young 'Valencia' orange trees at 00 C

when air temperatures remained at -100 C for 50 hr. Fiberglass wraps

provided 00 to 30 C protection in several field studies, and the level

of protection appeared to be dependent on the rate of air temperature

decrease (121). Thin-walled polystyrene wraps provided very little

protection, and sometimes allowed trunk temperature to fall slightly

below air temperature (30, 68). The level of protection provided by

other wraps was somewhere between that of the styrofoam wrap with water

containers and the thin-walled polystyrene wrap (30, 67, 68). Tree

wraps made of cornstalks were ineffective for freeze protection of young

citrus trees (143), but have been used to prevent wind or radiation

damage to trunks (129).

Sprinkler Irrigation

Sprinkler irrigation has been used successfully for freeze

protection of many crops and areas of the world, from peas and beans in

the Yukon territory (51) to strawberries (81) and citrus in Florida

(30). Perry (109) provides a comprehensive review of field studies

where sprinkling was used to protect various fruit and vegetable crops.

Among the many studies in her review, there were few instances where








23

sprinkler irrigation was used unsuccessfully. The most notable failure

of sprinkler irrigation was reported by Gerber and Hendershoct (47)

where mature citrus trees were killed to the ground as a result of

insufficient application rates (0.25 cm hr- ) during advective freeze

conditions (windspeed, 2 to 5 m s-; minimum air temperature,-9 C). In

their study, irrigated trees sustained more damage than unirrigated

trees due to evaporative cooling and elevation of tissue killing

temperatures by the presence of moisture (46).

Irrigation for freeze protection is often referred to as a

"two-edged sword" because freezing of water on the plant releases heat,

while evaporation extracts heat (136). The ratio of latent heats of

evaporation and fusion is about 7.5, thus if 7.5 times as much water is

frozen as evaporated, the temperature of the plant part will remain near

ambient (48). To maintain the plant part near 0 C, an ice-water film

must be constantly maintained on the surface (49). The irrigation rate

required to maintain an ice-water film depends on windspeed, air

temperature and humidity (20, 48, 109). Clear ice formation and icicles

are field indicators of sufficient application rates (47). Trees may be

damaged by excessive ice loading and limb breakage even if proper

application rates are supplied under severe freeze conditions (48).

Several sprinkler irrigation models have been developed due to the

need to determine appropriate application rates and maximum off times

for adequate freeze protection (20, 22, 46, 111, 114). In these models

the rate of heat loss due to radiation, convection and evaporation

(sublimation) is set equal to the amount of heat gained by the freezing

process, and the latter is divided by the heat of fusion of water to

obtain the required irrigation rate. Exceptions to this are as follows:










1) the model of Chesness et al. (22) assumes that no freezing occurs,

and application rate is calculated by dividing the sum of the heat loss

terms by the specific heat and temperature difference between the water

and the leaf; 2) the model of Phillips et al. (114) calculates the

maximum time that a bud can be left unsprinkled before it drops to a

critical temperature, and does not calculate or depend on irrigation

rate. Predictions from the Gerber and Harrison (46) and SPAR79 (112)

models have been successfully validated in the field. A BASIC computer

program is available through North Carolina State University to predict

irrigation rates required for freeze protection, which is essentially

the SPAR79 model (110).

High- and low-volume undertree irrigation has been used to protect

young and mature citrus trees (11, 15, 27, 30, 107, 108, 139).

High-volume sprinklers used were the "pop-up" (27) and "impact" types

(105). Low-volume undertree sprinklers are variously referred to as

minisprinklers, misters, foggers, spinners, and microjets (79), but will

be referred to as microsprinklers for this discussion.

High-volume irrigation increased trunk, leaf and air temperatures,

and fruit pack-out in mature 'Orlando' trees in the lower third of the

canopy (15, 139). However, temperatures and tree damage were similar in

the upper 2/3 of the canopy. Although somewhat successful, high-volume

irrigation is only feasible for protecting limited acreages due to high

cost and consumption of water (15).

Observations show that microsprinkler irrigation can ameliorate the

effects of less severe freezes on mature trees when relatively high

irrigation rates are used (100). Microsprinkler irrigation at rates of

38 and 52 liter hr-1 provided marginal leaf temperature increases in the


d 52 liter hr provided marginal leaf temperature increases in the










lower portion of the canopy of 'Orlando' tangelo trees, but provided

almost no fruit protection (15). Parsons et al. (107) obtained similar

results with 38 and 87 liter hr-1 microsprinklers with mature 'Valencia'

and 'Temple' orange trees during 2 relatively mild freezes.

Alternatively, microsprinkler irrigation is very effective for

freeze protection of young citrus trees under a wide variety of

conditions (11, 30, 104, 108). Typically, microsprinkler irrigation

protects the lower 50-70 cm of a tree, and as with soil banks and tree

wraps, new canopies are produced after severe freezes and trees attain

heights of 1 to 2 m by the following autumn (104).

Parsons et al. (108) observed air temperature increases of 0 to

2.80 C and fog in a young tree planting irrigated with microsprinklers

and suggested that partial protection was afforded by microclimate

modification. However, Davies et al. (30) produced opposing evidence

when they measured lower net radiation values over microspinklers that

produced higher rates per unit area than those producing lower rates.

Microsprinkler irrigation at 87 liters hr-1 sprayed in a

wedge-shaped 90 pattern protected young 'Hamlin' orange trees during

one of the most severe freezes in recent history in Florida (30). In

this study, it was found that tree wraps used in conjunction with

microsprinklers provided 10 to 20 C more protection than microsprinklers

alone. In addition, irrigated-wrapped trees had higher trunk

temperatures than irrigated-unwrapped trees when irrigation was

discontinued during the night. Further studies indicated that

intermittent microsprinkler irrigation could be used to protect larger

acreages of wrapped young citrus trees than can be accomplished with

constant irrigation (32).
















CHAPTER III
TRUNK TEMPERATURE, FREEZE SURVIVAL AND REGROWTH OF YOUNG
CITRUS TREES AS AFFECTED BY TREE WRAPS AND MICROSPRINKLER IRRIGATION

Introduction

Thousands of young, nonbearing citrus trees have been killed by

severe freezes in Florida during the last 5 years. Young trees are

particularly prone to freeze damage due to their small size and vigorous

growth habit. Traditionally, soil banks or tree wraps have been used

for cold protection of young trees. Jackson et al. (68) compared

several tree wraps with conventional soil banks and found that wrapped

trunk temperatures were 0to 1.50 C higher than those of unwrapped

trunks, but 40 to 60 C lower than those of banked trunks. Although soil

banks provide effective cold protection, they are labor intensive and

often result in mechanical and/or disease damage to the trunk (30). In

a subsequent study, wraps provided 00 to 2.50 C protection for young

'Hamlin' orange trees during several radiative freezes in Florida in

1982 (121).

Low volume, microsprinkler irrigation may provide an alternative to

wraps and banks for freeze protection of young citrus trees. Davies et

al. (30) successfully protected young 'Hamlin' orange trees during the

advective freeze of Christmas 1983 using a combination of tree wraps and

microsprinkler irrigation applied in a 900 pattern. They also showed

that wraps used in conjunction with irrigation provide greater

protection than either method alone. Microsprinkler irrigation without

tree wraps may be less effective under advective than radiative freeze










conditions because damage was observed on trees irrigated at 38 liter
-1
hr-1 during the 1983 Christmas freeze (102, 103).

Because only one irrigation rate (87 liter hr1 ) and spray pattern

(900) were used in the study by Davies et al. (30), the objective of

this research was to study the effects of different microsprinkler

irrigation rates and spray patterns on trunk temperature of wrapped

trees during freeze conditions, and subsequent tree survival and

regrowth.

Materials and Methods

Winter 1984-85 Studies

A 0.3 ha. planting of 2-year-old 'Hamlin' orange trees (Citrus

sinensis (L.) Osb.) on trifoliate orange rootstock (Poncirus trifoliata

(L.) Raf.) located at Gainesville, Florida, was used in these studies.

Trees were spaced 4.6 x 6.2 m and were ca. 1.5 m tall. The lower 40 cm

of all trees was wrapped with 9 cm (R-11) foil-faced fiberglass

insulation. Irrigation treatments were applied in an incomplete

factorial combination of 4 rates (0, 38, 57, 87 liter hr- ) x 2 spray

patterns (900 and 3600). Microsprinklers were placed 1 m from the trees

on the northwest side because winds are generally from this direction

during advective freezes in Florida.

A randomized complete block design with 3 replications was used

employing 3 trees per treatment per replication. Trunk temperatures of

6 of the 9 trees in each treatment were measured with copper-constantan

T-type thermocouples attached to the trunk under the wrap at a height of

20 cm. Air temperature was measured at several locations throughout the

planting by suspending thermocouples from the canopies of the trees at a

height of about 1 m above the soil surface. Data were logged hourly











during several freezes over the winter of 1984-85. Irrigation was

started when air temperature reached 00 C and discontinued when air

temperature returned to 0 C. Two freezes, an advective freeze

beginning on 20 Jan. 1985, and a radiative freeze beginning on 26 Jan.

1985, were selected for discussion as results from other freezes were

similar.

The efficiency of each irrigation treatment (OC increase in trunk

temperature per liter per hr) was calculated by subtracting the

temperature of an unirrigated wrapped trunk from an irrigated wrapped

trunk at the time of the lowest recorded temperature, and dividing by

the irrigation rate. The lowest possible efficiency was taken as zero,

although negative values could occur if the irrigated trunks were cooled

by evaporation below the temperature of the unirrigated trunk.

Winter 1985-86 Studies

Studies during Dec.-Jan. 1985-86 were similar in many respects to

those described in the last section. Two-year-old 'Hamlin' orange trees

on sour orange (Citrus aurantium L.) rootstock planted at the same site

were used to evaluate irrigation rates of 0, 12, 22, 38, and 57 liter

hr-. On the basis of results of the previous winter, the 87 liter hr-

rate and all 3600 spray pattern treatments were not repeated. A

completely randomized design was employed using 7 trees per treatment,

and all trees were wrapped with foil-faced fiberglass (R-11) tree wraps.

Trunk temperatures and efficiency were measured as described in the last

section on 2 trees in each treatment during several freeze nights. The

maximum height of live wood on all trees was measured on 14 April 1986;

however, regrowth of the trees was not evaluated.









Results and Discussion

Winter 1984-85 Studies

The advective freeze on 20 Jan. 1985 was characterized by clear

skies, minimum temperature near -12.0 C with durations below 00 C of 39

hr, windspeed of 2 to 6 m s with gusts to 10 m s and dewpoint

ranging from -8.40 to -25.6 C. The radiative freeze on 26 Jan. 1985

was characterized by clear skies, minimum temperature of -5.0* C with

durations below 00 C of 12 hr, windspeed less than 1 m s-1 and dewpoint

of -5of0 to -9.0* C.

Trunk temperatures during both types of freezes generally were

higher with the 90* than 3600 spray patterns at all irrigation rates,

particularly when minimum air temperatures were reached (Figs. 3-1,

3-2). Trunk temperatures during the advective freeze (Fig. 3-1) were at

or above -2.50 C for all irrigation treatments, while unirrigated (wrap

only) trunk temperatures decreased to -8.00 C and air temperature was

-12.00 C. Trunk temperatures remained above 00 C in all irrigation

treatments during the radiative freeze (Fig. 3-2) while minimum

temperatures were -2.00 C and -5.00 C for unirrigated trunks and air,

respectively.

Fiberglass tree wraps without irrigation maintained trunk

temperatures 3.00 to 4.0* C higher than air temperature under both

radiative and advective freeze conditions (Figs. 3-1, 3-2). In previous

studies (30, 121) fiberglass wraps provided only 1.00 to 3.0* C

protection. This difference may have been due to the greater trunk

diameter and/or greater rates of air temperature decrease in this study

compared to past studies. Larger trunks provide greater thermal mass

thansmaller ones, thus slowing the rate of decline in temperature.


nes, thus slowing the rate of decline in temperature.







30
4 9 0


0
360
-4


-8 UNIRRIGATED








O 4



360 /


-8

a. -8 UNIRRIGATED

UW ,AIR
-12 b




4 -


090


-4


-8 "UNIRRIGATED


-12 C


20 00 04 08 12 16 20 00 04 08 12

TIME (hr)

Fig. 3-1. Trunk temperatures of 2-year-old 'Hamlin' orange trees during
an advective freeze of 20 Jan. 1985 as influenced by
microsprinkler irrigation and tree wraps.1 Irrigation rates
were (a)_f7 liter hr (b), 57 liter hr and (c), and 38
liter hr Spray pattern is denoted by 90 or 360. Each
point is the mean of 6 measurements.


























































20 22 00 02 04


06 08


TIME (hr)


Trunk temperatures of 2-year-old 'Hamlin' orange trees during
a radiative freeze of 26 Jan. 1985 as influenced by
microsprinkler irrigation and trunk wraps, Irrigation rates
were (a) 87 liter hr (b), 57 liter hr1 and (c) and 38
liter hr Spray pattern is denoted by 90 or 360. Each
point is the mean of 6 measurements.


Fig. 3-2.










Also, rapid decreases in air temperature may result in larger

differences between trunk and air temperatures (121).

Irrigation may have been unnecessary on 26 Jan. because wraps

maintained the trunk temperatures above -2.0 C. Nevertheless, when

irrigation is used, fiberglass wraps increase the surface area for water

interception and ice formation around the trunk, thereby providing a

buffer between the trunk and the environment should the irrigation

system fail. Davies et al. (30) found that large ice masses around tree

wraps maintained trunks well above air temperature for prolonged periods

when irrigation was discontinued. Wraps are particularly important when

using 3600 emitters which produce several discrete streams of water that

could miss a small tree trunk due to improper placement or windy

conditions.

Trunk heating efficiency (OC increase in trunk temperature per

liter per hr), decreased as irrigation rate increased, and was higher

for 900 than 3600 patterns at all irrigation rates (Fig. 3-3).
-1
Therefore, the 90-38 liter hr- treatment was the most efficient

pattern-rate combination. Because pumping capacity is often a limiting

factor when irrigating large acreages on a freeze night, further studies
-1
are needed to determine if rates less than 38 liter hr might maintain

trunk temperatures above critical levels with equal or higher

efficiency. Efficiency of lower irrigation rates is discussed in the

next section.

All irrigated and wrapped crees had live scion wood as of 27 March,

and only one unirrigated, wrapped tree was killed to the bud union. The

height of live wood increased as irrigation rate increased for the 900

spray pattern (Table 3-1). Additionally, the dry weight and number of











































57 87


IRRIGATION RATE


(liter hr-1)


Effect of irrigation rate on trunk heating efficiency of
microsprinkler irrigation treatments for young 'Hamlin'
orange trees. Efficiency was determined on 21 Jan. 1985 at
0737 hr when air temperature was lowest (-12.0 C). For 900
pattern (A): y= -1.28x10-3x + 0.21, r2= 0.40; for 3600
pattern (o): y= -1.26x10 x + 0.18, r = 0.33. Slopes are not
significantly different, intercepts are significantly
difntrent, 5% level.


0.24


0.20 -


I.

SI-
0

U


4-
uV

z
uj
G


u.
'U


0.16


0.12



0.08


0.04


900 PATTERN








0 3600 PATTERN


8


38


Fig. 3-3.










Table 3-1.


Effect of microsprinkler irrigation rate and spray
pattern on height of live wood, shoot dry weight,
and number of shoots per trunk on 22 April 1985.


Rate
(liters hr )


Height
(cm)


900 3600
23.4


42.4
41.2
54.8


32.4
33.7
30.9


Dry wt.
(g)
Spray Pattern
900 3600
5.3


8.7
12.1
17.3


New shoots
(no.)

900 3600
6.9


9.3
10.8
28.5


8.4
7.8
7.8


ns ns ns


(**), or .05 (*).


SRegression coefficients significant at P< .01










new shoots increased with irrigation rate only for the 900 pattern.

This may be explained by differences in the height of water application

between the 90* and 3600 treatments. The 3600 pattern applied water to

the wrap surface at a height of about 20 cm regardless of irrigation

rate. However, water from the 900 pattern tended to wet the canopy and

form ice at progressively higher levels as irrigation rate increased.

It was obvious from the severity of the advective freeze that the trees

would die-back to wood that was protected by either wrap and/or ice

because air temperatures remained below critical levels (ca -6.7* C) for

several hours. Hence, it is not surprising that trees irrigated with

900 patterns had greater heights of live wood than those irrigated with

the 3600 pattern. Regrowth was greater in the 900 treatment probably

because more wood was present; thus more adventitious buds could be

initiated and undergo development.

Winter 1985-86 Studies

Two advective freezes on 25-26 Dec. 1985 and 27-28 Jan. 1986 were

selected for analysis of trunk temperature because these freezes were

severe enough to provide a rigorous test of the irrigation rates

studied. Freeze conditions on 25-26 Dec. 1985 were characterized by

minimum temperatures of about -7.00 C, windspeeds decreasing throughout

the night from 3 to < 1 m s -1, and dewpoints of about -100 C.

Conditions on 27-28 Jan. 1986 were characterized by minimum temperatures

of about -7.0 OC, windspeeds of 1 to 4 m s-1 and dewpoints of about

-140 C. In addition to the 2 advective freezes, data from 3 radiative

freezes on 22 Dec. 1985, 27 Dec. 1985, and 29 Jan. 1986 were used to

evaluate trunk heating efficiency, when minimum air temperatures of










-3.2, -4.0, and -2.50 C, respectively, occurred under calm, high

dewpoint conditions.

Trunk temperatures during advective freeze conditions increased

with irrigation rate up to 38 liters hr-1, but were similar for the 38

and 57 liter hr-1 treatments (Fig. 3-4). Trunk temperatures for the 38

and 57 liter hr-1 irrigation rates were maintained well above 0 C,
-1
while those for the 12 and 22 liter hr treatments decreased to about

-1.0 and -3.0* C, respectively, on both dates. Unirrigated wrapped

trunks had temperatures just slightly below those for the 12 liter hr-1

rate, hence, this irrigation rate provided little additional trunk

heating compared to the other rates. The relatively poor performance of

the 12 liter hr-1 rate cannot be entirely attributed to the irrigation

rate per se, as droplet size was noticeably different for this

treatment. Rather than the raindrop-sized droplets of higher irrigation

rates, the 12 liter hr-1 emitters produced a fine mist, causing large

amounts of opaque ice to form on the wrap and lower limbs of the trees.

Opaque ice formation is indicative of water freezing before it reaches

the tree (46), which could be expected with mist-sized droplets that

cool rapidly as they travel through the air due to their high surface to

volume ratio and low heat capacity (20). Hence, small droplet size may

have been partially responsible for the poor protection afforded by the
-1
12 liter hr- rate.

Despite differences in trunk temperature due to irrigation rate,

height of live wood was not significantly related (linearly or

quadratically) to irrigation rate (Table 3-2). This is contrary to

results of 1984-85 studies, but may be explained by the less severe

freeze conditions during the 1985-86 winter causing less trunk damage in








a o o o0 0 o o
0 0 0
O O O O


A


U7~Z


o 38 LPH
a 22 LPH
o 12 LPH
* UNIRR
t AIR
a I


W


* a 0


o o


I I


0 0 0 0


I~~ I


tff


19 21


23


1 3 5 7


TIME (hr)


Trunk temperatures of 2-year-old 'Hamlin' orange trees and
air temperatures during advective freezes of 25-26 Dec. 1985
(a), and 27-28 Jqn. 1986 (b) for unirrigated (UNIRR), 12, 22
and 38 liter hr- (LPH) microsprinklerlirrigation treatments.
Trunk temperatures in the 57 liter hr treatment were similar
to those for the 38 liter hr treatment, hence are not shown.
Trunk temperatures were measured at the 20-cm height
underneath fiberglass tree wraps on 2 trees in each treatment.


-1I


1.


0 0


0 0


U
U


a A


* U


-3



-5


-7


Fig. 3-4.


| m


i










Table 3-2. Height of live wood of young 'Hamlin' orange
trees on 14 April 1986 as influenced by
microsprinkler irrigation rate during several
freezes in Dec.-Jan. 1985-86.


Rate
(liters hr )


Height of live wood
(cm)


0 54.7 + 20.12
12 35.9 + 12.5
22 56.9 + 15.6
38 46.1 + 15.6
57 57.3 + 12.8
z
Values of height of live wood are means + SD. Linear
and quadratic regression coefficients were not significant,
P < 0.05. n=7.










general, particularly for unirrigated trees. However, trees in the 12

liter hr-1 treatment were noticeably smaller than all other trees, and

the visually greater canopy damage may have resulted from insufficient

water application under windy conditions, i.e., evaporative cooling.

Trunk heating efficiency was calculated for 5 different freeze

nights and data were fitted most closely by a quadratic equation (Fig.

3-5). Setting the derivative of the equation equal to 0, the relative

maximum point on the curve was obtained which corresponded to the value

of the most efficient irrigation rate (position of the star in Fig.

3-5). Similar to the winter 1984-85 data, trunk heating efficiency

was highest for the 38 liter hr-1 irrigation rate (with a 900 spray

pattern) under the widely varying freeze conditions encountered on 5

nights during the winter of 1985-86 (Table 3-3). This is rather

fortunate because most citrus growers in Florida that have

microsprinkler irrigation systems use the 38 liter hr-1 rate (Davies,

personal communication), and trunk heating efficiency could be maximized

by simply changing from 3600 to 900 spray patterns.

Conclusions

The 900 spray pattern was superior to the 3600 spray pattern with

respect to maintenance of trunk temperature during freeze conditions,

and survival and regrowth of the trees the following spring. A

microsprinkler irrigation rate of 38 liter hr-1 was most efficient for

maintaining trunks of wrapped young citrus trees above damaging

temperatures under a variety of freeze conditions. Twice as much

acreage can be protected using a 38 liter hr-1 than a 87 liter hr-

microsprinkler with a 900 pattern given the same pumping capacity. Use

of irrigation rates < 38 liter hr-1 with 900 spray patterns can provide










































0 15 30 45 60 75


IRRIGATION RATE (liter hr-1)


Trunk heating efficiency as a function of irrigation rate
for 5 freezes during Dec.-Jan. 1985-86. Solid line is a
quadratic regression fitted through the data; the star
denotes the relative maximum on the fitted curve. Each point
is a single determination.


I
" 0.2

10
L.
= 0.1


0.0


Fig. 3-5.










Table 3-3.


Dates of freezes, minimum air temperatures, most
efficient irrigation rates, and coefficients of
determination from quadratic regressions of trunk
heating efficiency vs. irrigation rate during
Dec.-Jan. 1985-86.


- ---- --- -- ----- _- -_--_ _- ___ __--
Date Minimum Most efficient r
air temperaturez irrigation rtey
(C) (liters hr )

Dec. 22 -3.2 38.5 0.76
Dec. 26 -7.0 41.0 0.48
Dec. 27 -4.0 40.4 0.51
Jan. 28 -7.3 38.6 0.54
Jan. 29 -2.5 40.6 0.32

Z Air temperatures are means of 7 measurements.
Most efficient irrigation rate was calculated
by setting the derivative of the quadratic
regression equations equal to zero and solving
for the relative maxima for each date.






42


10 to 40 C additional trunk protection over tree wraps alone and allow

even larger acreages to be protected on any given night. However,

height of live wood and regrowth of trees may be marginally greater

using higher irrigation rates.

















CHAPTER IV
MICROCLIMATE OF YOUNG CITRUS TREES PROTECTED BY
MICROSPRINKLER IRRIGATION DURING FREEZE CONDITIONS


Introduction

Methods of freeze protection that use water include flood

irrigation, fog generation and sprinkler irrigation, and each influences

the microclimate of the orchard. Flooding an orchard during a freeze

can increase soil and air temperatures and upward radiant heat flux from

the orchard floor (13, 41), thereby increasing temperatures of the

trees. Pre-freeze irrigation by either flooding or sprinkling can have

similar effects, but does not present the problem of standing water in

the orchard (41). Fog retards the loss of infrared radiation from the

orchard and raises the dewpoint, which can prevent tree temperatures

from decreasing to damaging levels (89). Therefore, flood irrigation

and fog generation affect temperatures of trees primarily by modifying

the microclimate of the orchard.

Alternatively, sprinkler irrigation provides protection primarily

through direct transfer of latent heat from the ice-water mixture that

coats the tree (20, 46, 109). However, microclimate modification by

high-volume sprinkler irrigation has been observed for mature citrus

trees (15, 27, 108, 139). Leaf and air temperatures were 0 to 30 C

higher for irrigated than unirrigated 'Orlando' tangelo trees (27, 139)

and protection varied with position in the canopy and freeze conditions.

Parsons et al. (107) reported 0 to 10 C differences in air temperature










between irrigated and unirrigated mature citrus canopies using

low-volume microsprinkler irrigation. Fog or mist generation by

microsprinklers may occur under high dewpoint conditions and is thought

to decrease radiant heat loss from the trees and soil surface (108).

However, canopy temperatures at the 2 and 3 m height of irrigated citrus

trees were similar to those of unirrigated trees, despite the presence

of fog or mist in the irrigated block (139). Preliminary studies
-2
indicated that net radiation above young citrus trees was about 8 W m

more negative with high irrigation rates than low rates, but values for

unirrigated trees were not reported (30).

The effect of microsprinklers on microclimate and subsequent freeze

damage of young citrus trees is unclear. The objective of this research

was to determine the effect of microsprinkler irrigation on air

temperature, dewpoint, net radiation and soil temperature around young

citrus trees under various freeze conditions.

Materials and Methods

Plant Material, Freezes and Treatments

A 0.3-ha planting of 126 2-year-old 'Hamlin' orange (Citrus

sinensis (L.) Osb) trees on either trifoliate orange (Poncirus

trifoliata (L.) Raf.) or sour orange (Citrus aurantium L.) rootstock was

used for all experiments. Trees were spaced 4.6 x 6.2 m and were 0.5 to

1.5 m in height. The lower 40 cm of all trees was wrapped with 9 cm

(R-11) foil-faced fiberglass insulation. Microsprinklers were placed 1

m from the trees on the northwest side because winds are generally from

this direction during advective freezes in Florida.

Microclimate measurements during 6 freezes which had widely

variable meteorological conditions were chosen for analysis of treatment











effects. The advective freeze of 20-22 Jan. 1985 was characterized by

minimum temperature near -12.0* C with durations below 00 C of 39 hr,

windspeed of 2 to 6 m s with gusts to 10 m s and dewpoint ranging

from -8.40 to -25.6 C. The radiative freeze on 26-27 Jan. 1985 was
-1
characterized by minimum temperature of -5.0' C, windspeed < 1 m s 1

and dewpoint of -5.3 to -9.0. C. Freeze conditions on 25-26 Dec. 1985

were characterized by minimum temperature of -7.0* C, windspeed of 1 to

3 m s-1, and dewpoint of about -10.0' C. Minimum temperature on 27-28

Jan. 1986 was about -7.0' C, windspeed was 1 to 4 m s-1, and dewpoint

was about -14.0 C. On 26-27 Dec. 1985, minimum temperature was about

-4.00 C, windspeed was 0 to 2 m s-1, and dewpoint averaged -7.0' C.

Minimum temperature on 11-12 Jan. 1987 was -1.5 C, windspeed was 1 to 3
-1
m s and dewpoint was about -4.0 C.

In Jan. 1985, treatments were applied in a factorial combination

of 3 irrigation rates (38, 57, and 87 liters hr-1 ) x 2 spray patterns

(900 and 3600), plus an unirrigated control. Four irrigation rates (12,

22, 38, and 57 liters hr -1) were used with only the 90 spray pattern in

Dec.-Jan. 1985-86, plus an unirrigated control. In Jan. 1987, only a

900 pattern-38 liter hr-1 rate irrigation treatment and an unirrigated

control were used.

Microclimate Measurements

Microclimate around young trees was measured during the freezes

described above for irrigated and unirrigated trees. Air and soil

temperature, dewpoint, and net radiation data were recorded hourly

during the freezes in Jan. 1985. Soil temperature measurements were

repeated during the 3 freezes in Dec.-Jan. 1985-86 because measurements

from the previous year were speculative. Net radiation measurements










were repeated on 11-12 Jan. 1987 to confirm results from 1985.

Observations on fog and mist generation by microsprinklers were made on

all dates. Relative humidity and temperature measurements were made on

all dates inside and outside the planting with a sling psychrometer.

Air temperature was measured with copper-constantan thermocouples

in the canopy of one tree in each treatment at a height of 1 m.

Relative humidity was measured with humidity sensors (Viasala

instruments, Woburn, Mass.) at the same location as air temperature for

one tree in each treatment, and dewpoint was calculated from

simultaneous measurements of air temperature and relative humidity. Net

radiation was measured with Fritschen-type net radiometers placed at a

height of 0.8 m and centered directly over the water spray, between the

tree and the microsprinkler. Preliminary experiments indicated that at

a height of 0.8 m, net radiometers were sensitive enough to measure
-2
differences in upward radiant flux of about 5-10 W m from an area of

about 2.8 m2 (data not shown). One net radiometer per treatment was

used during winters of 1984-85 and 1985-86, but in 1986-87, 4 net

radiometers per treatment were used. Soil temperature was measured with

thermocouples placed approximately 1 cm below the soil surface next to

the trunk of 1 tree in each treatment. This location was chosen because

it was suspected that soil temperature could influence trunk temperature

above the soil surface by conduction of heat vertically along the trunk.

Results and Discussion

Microsprinkler irrigation rate did not affect air temperature,

dewpoint or net radiation of the trees within either the 900 or 3600

spray pattern during the winter of 1984-85 (data not shown). Hence,

data were classified as either irrigated-900, irrigated-3600 or










unirrigated to simplify discussion of these variables. Soil temperature

data are discussed for the 90 pattern only, because thermocouple

placement for the 3600 spray pattern treatments caused spurious

measurements.

Air Temperature

Air temperature in the tree canopy was similar for irrigated-900,

irrigated-3600 and unirrigated treatments during severe advective and

radiative freezes (Fig. 4-1). Furthermore, air temperatures for all

trees were typically within 0.50 C of those outside the research plot

under both types of freeze conditions (data not shown). Temperature

differences between irrigated 900 and unirrigated treatments approached

1.00 C under radiative conditions for a 3-hr period in the middle of the

night on 26-27 Jan. 1985. This may have been attributable to heat

released by the irrigation water, but was more likely due to random

variation in air temperatures which is frequently observed on radiative

nights. Parsons et al. (108) observed consistent increases in air

temperature of 0.50 to 1.50 C for irrigated young trees, although their

microsprinklers were more closely spaced and freeze conditions were less

severe than in this study.

Dewpoint

Variations in dewpoint were typically less than 10 C among

irrigaced-900, irrigated-3600, and unirrigated treatments under

advective and radiative freeze conditions (Fig. 4-2). As with air

temperature, there was a 3-hr period in the middle of the radiative

freeze night where values were higher for the irrigated-900 than

unirrigated trees (Fig. 4-2b). However, this may be a mathematical

artifact because air temperature was used to calculate dewpoint.

























-8


-12


0


-2


I I I I I I I I I I I


1 1 1 1 1 1 1 1 1 1 1 1


20 22 00 02 04 06
TIME (hr)


Air temperature in the canopy of 2-year-old 'Hamlin' orange
trees during a severe advective freeze on 20-21 Jan. 1985
(a) and a radiative freeze on 26-27 Jan. 1985 (b). Points
are means of 3 values for the irrigated-900 (IRR,90) and
irrigated-3600 (IRR,360) treatments, and single values for
unirrigated (UNIRR) treatments.


Fig. 4-1.



















-10


-15


-

0
A-





















Fig. 4-2.


-2


20 22 00 02 04 06
TIME (hr)


Dewpoint temperature in the canopy of 2-year-old 'Hamlin'
orange trees during a severe advective freeze on 20-21 Jan.
1985 (a) and a radiative freeze on 26-27 Jan. 1985 (b).
Points are means of 3 values for the irrigated-900 (IRR,90)
and irrigated-3600 (IRR,360) treatments, and single values
for unirrigated (UNIRR) treatments. Dewpoint was calculated
using simultaneous air temperature and relative humidity
measurements.










Mist often was observed over irrigated trees only during calm

conditions when dewpoint was close to air temperature, similar to the

observations of Parsons et al. (108). However, humidity measurements

made with a sling psychrometer above unirrigated trees were similar co

those outside the research plot (typically 60-100%), indicating that

microclimate around unirrigated trees within the plot was not influenced

by mist from neighboring irrigated trees.
-2
Net Radiation Net radiation was 10 to 20 W m more negative over

irrigated than unirrigated treatments under advective freeze conditions

(Fig. 4-3a). Higher net radiation values over unirrigated than

irrigated trees were observed again during a less severe freeze on 21-22

Jan. 1985 (data not shown). Net radiation values outside the planting

were comparable to values for unirrigated trees inside the planting.

Hence, the increase in outgoing radiation from the irrigated treatments

was localized, and probably caused by the presence of the relatively

warm irrigation water. If irrigation were favorably modifying the

microclimate of the trees, one would expect net radiation to be less

negative over irrigated trees, but the opposite situation is seen (Fig.

4-3a).

No differences in net radiation occurred between irrigated and

unirrigated treatments or among irrigated treatments under radiative

freeze conditions (Fig. 4-3b). Net radiation was measured above

irrigated (38 liter hr-1-90 microsprinklers) and unirrigated trees

during a mild, radiative freeze on 11-12 Jan. 1987 to verify results

from 1985. Again, net radiation was found to be similar for irrigated

and unirrigated crees (Fig. 4-4), consistent with data in Fig. 4-3b.

Reasons for the lack of differences in net radiation between irrigated


















































00 02
TIME (hr)


Net radiation above 2-year-old 'Hamlin' orange trees during a
severe advective freeze on 20-21 Jan. 1985 (a) and a
radiative freeze on 26-27 Jan. 1985 (b). Points are means
of 3 values for the irrigated-900 (IRR,90) and irrigated-3600
(IRR,360) treatments, and single values for unirrigated
(UNIRR) treatments.


-50

-55

-60

-65

-70

-75



-25


E




O









z-


-65


Fig. 4-3.
































22


00


02


04


06


08


TIME (hr)


Net radiation above 2-year-old 'Hamlin' orange trees during a
radiative freeze on 11-12 Jan. 1987. Points are means of 4
values for irrigated and unirrigated treatments.


'E

*Z

CD





Z
I-




UJL


-56


-58


-60


-62


o UNIRRIGATED
IRRIGATED




0


- oo
o 8 8 2
I I


Fig. 4-4.










and unirrigated trees under radiative conditions are unclear.

Variability in incoming radiation under calm conditions may have masked

any slight differences in upward radiant flux on both occasions.

Davies et al. (30) measured net radiation values of -61.3 and -53.1

W m-2 above trees irrigated with 90-87 liters hr-1 and 360-38 liters
-1
hr- microsprinklers, respectively, under radiative conditions. Their

results could be explained by the presence of greater quantities of 100

to 150 C irrigation water underneath net radiometers for the higher than

the lower irrigation rate. However, it is unclear why differences

between 900 and 3600 spray patterns were not detected under radiative

conditions in this study. Possibly, different net radiometer placement

(1.5 m for Davies et al. (30) and 0.8 m in this study) was responsible

for the conflicting results.

Soil Temperature

Soil temperature generally was highest for the 38 liter hr-1 rate,

intermediate for unirrigated trees, and lowest for the 12 liter hr-1

rate during 3 freezes in Dec.-Jan. 1985-86 (Fig. 4-5). The soil

temperatures for the 22 and 57 liter hr-1 treatments were very similar

to chose for the 12 and 38 liters hr-1 treatments, respectively. The
-l
mist-sized droplets in the 12 liter hr treatment probably cooled co a

greater extent as they traveled through air than larger droplets in

other treatments (46) and consequently reduced soil temperature below

that for unirrigated trees. However, droplet size was comparable for 22
-1
and 38 liters hr1 microsprinklers, and the reduced soil temperature for

the former treatment cannot be explained by greater droplet cooling.

















a


A
.0 0 0 0 0
z 0
a
S 0 0 0 a
!. O O OoooO8


I
b




A
S3 A ,
0 0 0
0 0 0 0 0 0 0 0 0





a 38 LPH
o 12 LPH
I 0 LPH
0
A
o A 0 0
0 A



I -I I II


19 21


23 01


03 05 07


TIME (hr)


Soil temperature measured 1 cm below the surface and next to
trunks of 2-year-old 'Hamlin' orange trees during freezes on
25-26 Dec. 1985 (a), 26-27 Dec. 1985 (b), and 27-28 Jan.
1986 (c). Points are single values for the unirrigated (0
LPH), 12 (12 LPH) and 38 liter hr (38 LPH) treatments.


Fig. 4-5.










Conclusions

Microsprinkler irrigation does not appear to change the

microclimate of a young tree in the same way as reported for mature

trees (15, 107). Mature trees have much greater canopy volume than

young trees, hence a greater capacity to decrease radiation loses and

reduce windspeeds within the canopy, and retain heat released from

irrigation water. Irrigation may, under certain conditions, increase

the long-wave radiant flux from the vicinity of a young tree, but these

effects appear to be localized and do not affect air temperatures.

Irrigation rates > 38 liters hr- elevated soil temperature 20 to 60 C,

while lower irrigation rates reduced soil temperature 20 to 4 C with

respect to the unirrigated condition. Therefore, it is possible that

heat conducted along the trunk from the soil is partially responsible

for elevation of trunk temperatures of wrapped young citrus trees

irrigated with 38 liter hr-1 (or higher rate) microsprinklers. This

possibility is investigated in Chapter VII.
















CHAPTER V
TRUNK TEMPERATURE, LIGHT INTENSITY AND SPROUTING OF WRAPPED AND
UNWRAPPED YOUNG 'HAMLIN' ORANGE TREES FOLLOWING A FREEZE

Introduction

Tree wraps not only provide freeze protection for young citrus

trees during the winter (68, 121, 156) but also influence trunk

sprouting in the spring. Trunk sprouting on freeze-damaged trees may be

beneficial or deleterious, depending on the degree of injury sustained

by the canopy. Following less severe freezes, ample canopy wood may

survive and produce new shoots. In this case trunk sprouting would be

undesirable, occurring at the expense of growth of scaffold limbs.

However, severe freezes may kill a young tree nearly to the bud union,

and the development of trunk sprouts would be necessary for the

reestablishment of the tree canopy. In either case, absence or presence

of a tree wrap and the type of wrap become important factors in the

regrowth and proper training of a young citrus tree.

Sprout inhibition by wraps may be a result of a physical barrier

imposed by a close fitting wrap, and/or modification of the

environmental factors influencing sprout growth such as temperature and

light intensity. Shoot initiation and growth are strongly temperature

dependent in citrus (23, 35, 119, 163), and light is necessary for

shoots to become autotrophic and continue growth once initiated. The

temperature and light regimes underneath wraps have not been studied in

the early spring when sprout initiation and development occur.

Moreover, the influence of various trunk wraps on trunk sprouting has










not been critically examined in citrus. The objective of this study was

to monitor the temperature and light environment of young citrus tree

trunks underneath various wraps during early spring, and to examine the

relationship between these environmental factors and trunk sprouting in

freeze damaged trees.

Materials and Methods

Two-year-old 'Hamlin' orange trees (Citrus sinensis (L.) Osb.) on

trifoliate orange rootstock (Poncirus trifoliata (L.) Raf.) located at

Gainesville, Florida were used in this study. These trees were

protected from severe freezes during the 1984-85 winter with fiberglass

tree wraps and microsprinkler irrigation, but were killed back to a

height of about 60 cm. On 26 Feb. 1985, fiberglass wraps were removed

and 6 wrap treatments applied to 36 trees in a randomized complete block

design as follows: unwrapped (UW), fiberglass (FG), styrofoam (SF),

styrofoam modified to exclude light (MSF), white polystyrene (WP) and

charcoal polystyrene (CP). These treatments were chosen to provide

separation of light and temperature effects on trunk sprouting. The

wraps covered the lower 40 cm of the trunk, of which about 35 cm was

scion wood. Most trees had sprouts above the wrapped portion of the

trunk at this time, but no sprouts were observed on the lower 40 cm.

Trunk temperature was measured with copper-constantan thermocouples

taped to the trunk at a 20-cm height. Air temperatures were measured

with thermocouples suspended on canopy wood 1 m above the ground.

Photosynthetic photon flux (PPF) was measured with Li-Cor quantum

sensors taped to the trunk near the thermocouples. Due to a shortage of

quantum sensors, PPF could only be measured on 3 of the 6 treatments

during any particular day. Therefore, temperature and PPF measurements










were made over a 20 day period in March, 1985, the quantum sensors being

moved among treatments on different days. Temperatures were measured

hourly and PPF at 15 min. intervals.

Sprouts from the lower 40 cm of the trunks were removed and counted

on 13 May 1985, after treatments had been in place for 76 days, and

placed in an oven at 800 C for 24 hr for dry weight determination.

Temperature, sprout number and dry weight data were analyzed for

treatment effects using analysis of variance and Duncan's multiple range

test. Although PPF data could not be analyzed statistically, data for

any particular wrap treatment on different days with similar sky

conditions were comparable.

Results and Discussion

Generally, trunk temperatures under WP and CP wraps were highest

and those under SF and MSF wraps lowest during the daylight hours, with

FG, UW and air temperatures intermediate (Fig. 5-1). However, in the

afternoon on cloudy days, all wrapped trunk temperatures were generally

higher than the air temperature, but differences among wrap treatments

were small and varied from hour-to-hour. In mid-afternoon on sunny

days, differences in trunk temperatures between WP and MSF treatments

approached 170 C, and temperatures under WP wraps reached 410 C. In

addition, temperatures remained above 380 C under WP wraps for 3 to 4

hr, while temperatures under MSF wraps generally remained below 290 C.

During the night, temperatures under the SF and MSF wraps were highest,

lowest for WP, CP, UW and air, and intermediate for FG. By sunrise ,

the temperatures under the MSF wraps were significantly higher (0.50 to

1.00 C) than all other treatments. This was probably a result of the

covering of black tape used to exclude light which made the wrap


























0 0


16- c


L b

38 A WP
a 38 a CP
W A o SF
-A MSF
o UW .
32 A U FG
t AIR

27 s


21 r 0*


16 -


06 10 14 18 22 02 06

TIME (HR)


Fig. 5-1. Diurnal trunk temperatures of young 'Hamlin' orange trees
underneath various trunk wraps during a typical sunny day (a)
and a cloudy day (b) in March, 1985. Each point is the mean
of 6 measurements. Abbreviations are as follows: white
polystyrene (WP), charcoal polystyrene (CP), styrofoam (SF),
modified styrofoam (MSF), unwrapped (UW), and fiberglass
(FG).










airtight, and thus a better insulator than the SF wrap which was

ventilated. Although statistically different, it is questionable

whether a 0.50 to 1.0 C difference in nighttime temperature is of

biological significance to a young tree with respect to trunk sprouting.

The range of trunk temperature encountered in this experiment (160

- 410 C) was above the 130 C threshold which qualitatively controls

budbreak (119, 163) and below the 500 C threshold which can cause tissue

damage in citrus (62, 119). Prolonged periods of relatively high trunk

temperature as observed for some wraps in this study might adversely

affect cambial activity and ultimately the growth of a young tree.

Alternatively, such wraps may induce earlier resumption of cambial

activity in the spring by raising trunk temperature to levels favorable

for growth. Further studies on the influence of tree wraps on growth of

young trees are warranted due to the widespread use of wraps for

protection from freeze, herbicide, and fertilizer damage.

Photosynthetic photon flux varied from ambient on unwrapped trunks

to undetectable levels on trunks of FG and CP treatments. The WP wrap

allowed the most PPF transmission (9-13% of ambient) followed by SF

(4-9% of ambient) and then MSF (less than 0.1% of ambient) (Fig. 5-2).

In the UW, WP and SF treatments, PPF was well above the reported light

compensation point for citrus (73) which is about 1-2% of full sunlight.

Unwrapped trunks had highest sprout numbers and dry weights of all

treatments (Table 5-1). Number of sprouts in the WP treatment was

statistically similar to the UW treatment, but sprouts in the WP

treatment were deformed and had lower dry weights than those in UW

treatment. Apparently this was due to the physical limitations on

sprout growth imposed by the wrap. All other treatments had














0


0
0


0 & A A a A A a


A 0


o
0 0
^ A^ 0


0 0 o
o o 0 o
.0 1 .0 0


0 A
0 _,


aUW
SWP
MSF
*MSF8


0 0


0 o
BA~g0


06 08 10 12 14 16 18

TIME (hr)


Diurnal photosynthetic photon flux (PPF) underneath various
trunk wraps on 2-year-old 'Hamlin' orange trees during a
sunny day (a) and a cloudy day (b) in March, 1985. Each
point represents a single measurement. No PPF was detected
beneath CP and FG wraps. Abbreviations are as follows:
unwrapped (UW), white polystyrene (WP), styrofoam (SF), and
modified styrofoam (MSF).


1200


800


400


200




2000


I
N
E

o
E
5W

a-
0L.

&


1600-


1200


800


400


200


A 0a
0 0
M~ WD 0 .0


Fig. 5-2.


- --










Table 5-1.


Wrap
type


Numbers and dry weights of trunk sprouts on
freeze damaged yi gi 'Ialiiil' orange trees on
13 May 1986 influenced by various trunk wraps.


Trunk
sprouts
(no. )


Dry
wt
(g)


Unwrapped 8.4az 8.5a
White polystyrene 5.8a 3.1b
Modified styrofoam 1.5b 1.4b
Styrofoam 1.3b 2.5b
Fiberglass O.Ob O.Ob
Charcoal polystyrene O.Ob O.Ob


Z Means followed by the same letter are not significantly
different, Duncan's multiple range test, 5%. n=6.










statistically similar sprout number and dry weight, although no sprouts

were found on any of the trunks in the FG and CP treatments. It is

possible chat sprouts had developed somewhat but abscised before 13 May

in the FG and CP treatments due to the lack of PPF. Small etiolated

sprouts have been observed under fiberglass wraps (Davies, unpublished).

Jackson et al. (67) reported that sprouting occurred on 22% and 0% of

freeze damaged trees with SF and FG wraps, respectively, which is

qualitatively similar to the results in this study.

Some PPF (as low as 0.1% of ambient as in the MSF treatment) was

necessary for development of trunk sprouts on freeze damaged trees.

Furthermore, there was a direct correlation between PPF and sprout

growth. This is supported by the fact that trunks receiving the highest

PPF (UW and WP) had the highest number and dry weight of sprouts, trunks

receiving no measurable PPF (FG and CP) had no sprouts, and treatments

receiving intermediate intensities (SF and MSF) were intermediate in

sprout number and dry weight.

It is possible that a PPF-temperature interaction occurred in this

study, although unfortunately, the experiment was not designed for

statistical analysis of interaction. However, some of the results may

be explained better considering a PPF-temperature interaction. For

example, the CP/WP treatment pair and the SF/MSF treatment pair were

designed to create similar trunk thermal regimes but different PPF

regimes. The WP and CP treatments both had relatively high mean

temperatures and different PPF regimes, while the SF and MSF treatments

both had relatively low mean temperatures and different PPF regimes.

Large differences in sprouting occurred between the WP and CP

treatments, suggesting thac PPF influences sprout development given a










relatively high trunk thermal regime. On the Other hand, no statistical

differences in sprouting occurred between the SF an MSF treatments,

suggesting that PPF does not control sprout development given a

relatively low trunk thermal regime. Therefore, the effect of PPF on

sprouting was either strong or nonexistent, depending on the trunk

temperature regime.

Conclusions

In general, PPF was necessary for trunk sprouts to develop and

persist under the various wraps, and both high PPF and high temperature

favored the development of trunk sprouts. In relative terms, if the

temperature was high but PPF low (as for CP), or temperature low and PPF

high (as for SF), or both temperature and PPF were low (as MSF) then the

number and dry weight of sprouts were reduced. Therefore, wraps that

exclude light entirely are probably best for trunk sprout inhibition of

moderately damaged young citrus trees where regrowth will occur on

surviving canopy wood. If trees have been killed to the wrap, however,

removal of the wrap at the earliest reasonable date would encourage

development of trunk sprouts which will produce a new canopy.

















CHAPTER VI
MODELING AND SIMULATION OF TREE WRAPS AND MICROSPRINKLER
IRRIGATION FOR YOUNG CITRUS FREEZE PROTECTION

Introduction

Freeze protection using tree wraps and microsprinkler irrigation

has been the subject of several field studies with young citrus trees

(30, 32, 104, 107, 108). These studies mainly present data on trunk,

leaf and air temperatures during freezes and resulting tree survival and

regrowth. Although these studies demonstrate the effectiveness of tree

wraps and microsprinkler irrigation for freeze protection under the

conditions encountered, one can only speculate on the relative

importance of humidity, air temperature, rate of air temperature

decrease, windspeed, freeze duration, wrap type, water temperature,

etc., on trunk temperature and subsequent tree survival. Freeze

chambers cannot accurately simulate radiation, humidity, temperature and

windspeed regimes observed in the field during freezes, hence laboratory

studies would be neither practical nor useful in yielding accurate

quantitative information. Alternatively, modeling and simulation of the

wrap/microsprinkler system may provide detailed information that is

currently lacking on system behavior, and can hasten the progress of

research on freeze protection of young citrus trees by allowing

investigators to study the problem year-round instead of only a few

nights per year.

Models of overhead sprinkling for freeze protection have been used

to predict irrigation requirements for successful freeze protection of










apple buds (109, 114) and mature citrus foliage (20, 46). Basically,

these models calculate the rate of heat loss from a plant part under

given freeze conditions, then use this to calculate the irrigation rate

required to maintain the plant part above a predetermined critical

temperature. Hence, output from sprinkler rate models consists of a

single value, i.e., the irrigation requirement for freeze protection

under a given set of environmental conditions (110). Practical

applications of these models are limited because most irrigation systems

are not designed to produce variable irrigation rates. However, concepts

and information contained in sprinkler rate models were useful in

developing a model of the tree wrap/microsprinkler irrigation system.

Rather than calculating static values of irrigation requirement, a

more heuristic approach was taken to model the dynamic behavior of trunk

temperature as influenced by the wrap/microsprinkler system given freeze

conditions and other easily determined parameters. Using this approach,

simulation experiments can be run and output, e.g., trunk temperature,

evaluated at any time, under a variety of conditions. Furthermore,

simulation can be used as a screen for hypotheses, reducing the amount

of field experimentation necessary to achieve a particular goal.

Ultimately, information can be derived that can be useful in redesigning

and improving the system.

The primary objective of this research was to develop a

process-based, dynamic heat transfer model of the wrap/microsprinkler

system of young citrus freeze protection and use simulation to study

system behavior. Simulations were run to assess the influence of

parameters and environmental conditions on trunk temperature. Trunk










temperature predicted within one standard error of observed data during

model validation were considered accurate.

Model Development

Description

The system under consideration consists of a young citrus tree

trunk enveloped by an insulative tree wrap 8-14 cm in diameter and 30-40

cm in height (Fig. 6-1). Water is applied by a microsprinkler

positioned 1 m from the tree, which consists of a plastic tube with an

orifice and directional cap on the end that control the flow rate and

spray pattern of the irrigation water. When the irrigation is on, the

wrap surface is partially covered by a thin layer of water and an ice

layer if conditions favor freezing. The wrap/ice mass intercepts a

fraction of the irrigation water, and sensible or latent heat is

transferred from the water to the wrap/ice surface. As ice accretes on

the wrap surface, the fraction of water intercepted becomes greater as

does the surface area over which water is distributed. In the field,

some of the irrigation water often is intercepted by the canopy above

the wrap, which in addition to protecting some of the canopy, can change

the trunk temperature by flowing down the trunk inside the wrap.

Symbols used in model development, their definitions and units are

listed in Table 6-1.

Heat Conduction within the Wrap and Trunk

The tree wrap system can be viewed as 2 main components: a cylinder

of citrus wood (trunk) of variable radius, surrounded by an annulus of

insulative material (wrap) of specified thickness and composition.

Within the confines of the outer wrap surface, temperature varies

spatially and temporally, hence the trunk/wrap composite cylinder is a





























RADIATI


FREEZING







IRRIGATION
r- -
**-~---

..- -
II -- --


CONVECTION,
EVAPORATION


Fig. 6-1. Conceptual model of the tree wrap/microsprinkler irrigation
system for young citrus freeze protection illustrating heat
transfer processes; Q = heat conduction from within the trunk
and wrap to the external boundary.










time-continuous, distributed thermal system. The mathematical model for

the rate of change in temperature with time at any point within the

trunk/wrap composite cylinder is developed for the case of

one-dimensional (radial) heat flow in the following paragraphs.

The heat stored (h) per unit volume of a region within the

trunk/wrap composite cylinder is defined as

h (cal cm-3) = p c T [6-la]

where c is the specific heat, p the density, and T the temperature of

the region. In a time interval At, the change in heat content of a

region of volume= Ar rAG Az (Fig. 6-2) can be written as


Ah AT
Ar rA9 Az = p c Ar rAG Az [6-1b]
At At


where Ar, rAO and Az are the infintesimal distance increments in the

radial, angular, and vertical dimensions. The rate of change in heat

content of the region must be equal in magnitude and opposite in sign to

the change in flux (j) with distance (Ar), according to the principle of

continuity (26)

A(h Ar rAG Az) A(j Ar rAG Az)
= [6-2a]
At Ar


The terms Ar, A0, and Az are independent of time and radius, and r is

independent of time, so [6-2a] can be rewritten as


Ah A(rj)
Ar rA Az = --- r A Az [6-2b]
At Ar


Cancelling like terms and taking the limit as At and Ar approach zero,






70



3h 1 a(rj)
= --- [6-3]
at r ar


The flux in the radial direction j can be defined as


aT
j =-K [6-4]
ar


where aT/3r is the temperature gradient tetween the inner and outer

surfaces of the region, and K is the thermal conductivity of the region.

Substituting for j in [6-3] yields


ah 1 a aT
= (r K -) [6-5]
at r ar ar


Assuming'K is constant with respect to r, and substituting [6-la] for h


aT 1 a aT
= a (r -) [6-6]
at r 3r ar


where a = K/(p c), the thermal diffusivity of the region.

Similar derivations hold for the other 2 dimensions, resulting in

the following form for 3-dimensional heat transfer in cylindrical

coordinates (26)




2 2
at r ar 3r r za az


Thus, the rate of temperature change of any region between the trunk

center and wrap surface is dependent on the thermal diffusivity and

temperature gradients in the radial, circumferencial and vertical

dimensions about the region.










Equation [6-7] was not solved analytically due to its complexity,

and the fact that the parameters K, p, and c change discretely with r at

the trunk/wrap interface. Instead, a finite difference version of [6-7]

was employed to calculate rates of change in temperature, where second

derivatives were approximated by second forward differences (26)

dT (T -2T +T ) 1 (T -T )
dr,0,z (Tr+1,9,z 2Tr,,z Tr-1,e,z) 1 r+1,,z r-1,Qz
dt Ar r 2Ar

1 (T 0 -2T +T )
Sr, 9+lz rOz r,9-lz
2 2
r A9

(T -2T +T )
+ r,e,z+1 r,e,z r,Oz-l [6-8]
Az2


Hence, if a and the temperatures of the 6 surrounding regions are

known or set to assumed initial conditions, the rate of temperature

change of any region within the trunk/wrap cylinder can be calculated.

In order to use [6-8], the distributed tree wrap system was divided

into concentric layers in the radial direction, pie-shaped regions in

the angular direction, and disc-shaped regions in the vertical direction

yielding a discrete model system (Fig. 6-2). Therefore, the model

system had a finite number of regions of heat storage each with a single

temperature at any time t, whereas the real system contains an infinite

number of different points and temperatures. Values of A0 and Az were

set at 0.63 radians (360) and 1.0 cm, respectively, and Ar was either

0.2 or 0.5 cm, which were the maximum achievable without affecting

simulation output, allowing the programs to run in the least amount of

time possible. Using these values there were 5200 (Ar=0.5) or 13000

(Ar=0.2) finite regions of heat storage for a typical trunk/wrap

composite cylinder 13 cm in diameter and 40 cm in height.


linder 13 cm in diameter and 40 cm in height.
























Ar



:----' V-Az




rAe















Fig. 6-2. Three-dimensional representation of the model system showing a
representative finite region of heat storage. For a typical
tree wrap 13 cm in diameter and 40 cm in height there were
5200 regions when Ar=0.5 cm, A9=0.63 radians, and Az=1.0 cm.










The temperatures of most interest are those corresponding to the

cambial region of the trunk, because adventitious buds arise from this

area to produce a new canopy. The cambium of a fully acclimated citrus

tree is killed at temperatures of about -6.7 C (157), and it is

important to know conditions under which the wrap/microsprinkler system

allows the temperature of the cambium to reach critical values.

After rates of change were calculated, state variables were updated

by Euler or rectangular numerical integration with respect to time

Statet = Statet-d + dt(Rate-d ) [6-9]
t-dt t-dt
where subscripts t and t-dt denote current and previous times.

Temperature was the only state variable of interest in simulations

except for ice layer thickness, which appeared in only those runs of the

irrigation simulation which had rather severe simulated environmental

conditions.

The Euler method assumes that the rate of change of a state

variable is constant over the time period dt, hence dt must be small

enough to justify this assumption. The maximum useful time step in heat

transfer simulations is a function of thermal diffusivity and distance

increment (dx), and generally is described as (26)

(dx)2
dt < [6-10]
2 a
where the calculation is for the component in the system with the

highest x and lowest dx. Time steps of 0.3 or 0.8 seconds were used

when simulating freeze protection with wraps and irrigation or wraps

alone respectively, due to the different assumptions made in each case

(see section on simulation). Greater time steps caused unstable

oscillations of temperatures with time that were fatal to program

execution, while lower time steps did not affect output.










Heat Exchange at the System Boundaries

Internal boundary

The internal boundary of the model system is the center of the

trunk, and its rate of temperature change cannot be calculated using

[6-8] because r=O at the internal boundary, and [6-8] requires the

temperature at r-1 for calculating radial heat transfer. Therefore, the

temperature of the internal boundary was assumed to equal the average

temperature of the adjacent regions, (i.e., where r=l). This assumption

is often made with cylindrical thermal systems (26).

External boundary

Energy is exchanged between the system and the environment across

the external system boundary, or outermost layer of the system, by the

following processes: irrigation (IR), radiation (R), evaporation (E),

and convection (CV) (Fig. 6-1). Conduction (Q) allows heat transfer to

and from the external boundary, and therefore provides the linkage

between heat transfer at the surface and that within the system. The

external boundary differs in composition in both space and time, being

either wrap material, water or ice, and it is necessary to determine its

state before calculating the energy balance. However, a general form of

the energy balance of the external boundary is

dT
Ar(p c) s = (Q+IR-R-CV-E) [6-11]
s dt


where the subscript s denotes the external system boundary, and Ar its

thickness. If the irrigation is on, the external boundary is a thin

water layer where water is striking the wrap surface. Field

observations indicated that the thickness of the water layer was about

0.1 cm, and this value was used in all simulations. If the irrigation










is off but an ice layer is present, then IR=0 and E represents

sublimation instead of evaporation. Volume changes upon freezing of

water are assumed negligible, and Ar remains equal to 0.1 cm when the

system boundary is an ice layer. If the system boundary is wrap

material (no irrigation or ice), then IR and E equal 0 and Ar becomes

the thickness of the wrap surface layer, which was either 0.2 or 0.5 cm,

similar to other wrap layers. Use of Ar in [6-11] implicitly assumes

that temperature changes of the system boundary occur uniformly

throughout its volume, which may only be true for small values of Ar.

Process descriptions

The heat transfer processes in the right hand side of [6-11]

(Q, IR, E, CV, R) require specific mathematical descriptions, and all
-2 -1
must be in the same units (cal cm s ) to be summed as shown yielding

the rate of temperature change of the external system boundary.

The irrigation process is given by

IR= [RATE cH20 (TH20 T )/AREA [6-121

where RATE is the amount of water intercepted by the wrap in liters per

second, AREA is the surface area over which the water is uniformly

distributed, TH20 and cH20 are temperature and specific heat of water

respectively, and Ts is the temperature of the external boundary. In

[6-12] it is evident that as TH20 approaches Ts, IR approaches 0

regardless of RATE.

Conduction (Q) of heat to the external boundary from within is

given by (128)

Q= K(T in-T ) [rs In(rs/rin] [6-13]

where Tin and rin are the temperature and radius of a specified region

internal to the external boundary. For example, if the outer system










boundary is wrap material, then T. and r. are the temperature and

radius of the wrap layer beneath the wrap surface layer. Alternatively,

if the external boundary is water, then T. and r. would correspond to

the temperature and radius of the ice layer or the wrap surface layer,

depending on presence or absence of ice.

Radiative heat transfer is given by
3
R= 4UT (T -T ) [6-14]
a s a
3
where 4GT is the radiative heat transfer coefficient (46, 111). The
a
emmissivity of the surface is assumed to equal 1, and is therefore

omitted in [6-14]. Some of the consequences of this assumption are

discussed in chapter VII.

Evaporative heat flux is calculated by multiplying the mass flux of

water vapor from the external boundary by the lament heat of

vaporization (or sublimation)

E= Lv [Dwv Sh(W -W )] [6-15]

where D is the water vapor diffusivity of air, L is the latent heat
wv v
of vaporization (or sublimation), W and W are water vapor
s a
concentrations of the external boundary and air, respectively, and Sh is

the Sherwood number which embodies windspeed and wrap dimensions (93).

Water vapor concentration of the air (W ) is calculated from air
a
temperature and relative humidity using the psychrometric equation. The

air adjacent to the external boundary is assumed to be saturated with

water vapor at the surface temperature (Ws) if the external boundary is

water or ice. Evaporative heat flux equals 0 if the external boundary

is wrap material.

Convective heat transfer was calculated using Monteith's (93)

equation for cylinders









Cv = (K Nu)/d (T -T ) [6-16]
a s a
where (KaNu)/d is the convective heat transfer coefficient (93, 109),

and Nu is the Nusselt number embodying windspeed, wrap dimensions, and

kinematic viscosity of air.

Freezing of water on the external boundary invokes a boundary

condition where the temperature at the ice-water interface is 00 C. The

radius of the ice-water interface (r i) is therefore important, and it
i-w
is determined by calculating the rate of ice accretion, dr. /dt. The
--w
rate of heat conduction away from an ice-water interface (Q i) is (61)

dr.
--w
Qi-w = L (i ) [6-17]
dt
where L is the latent heat of fusion, and pi is the density of ice.

Latent heat produced at the interface can move in 2 directions, either

out (Qout) into the water layer or in (Qin) towards the wrap surface,

hence

Q. = Q. + Q [6-18]
Qi-w ia out
and,

Qin= K (Ti-273)/(riw ln(r.i_/r)) [6-19]

Qout=K0 (273-Tw)/rwl n(rwl/ri ) (r wl/r i [6-20]

where T, ri and T wl rwl are the temperatures, radii of the ice layer

and water layer adjacent to the interface, respectively. Substituting

[6-19] and [6-20] into [6-17] and rearranging



dr.
-W = (Qout- Q )/(Lf p) [6-21]
dt










The radius of the ice-water interface is obtained by numerical

integration of [6-21] yielding ice thickness, and adding to it the

radius of the wrap.

Forcing Functions

The forcing functions or inputs required by the model are soil and

air temperatures, windspeed, and relative humidity as functions of time.

Forcing functions are independent variables that "force" the system to

change state, and their magnitudes are not affected by the state of the

system. With the exception of soil temperature, environmental variables

are readily available in the form of weather forecasts or can be

measured with hygrothermographs and anemometers. Soil temperatures must

be measured with thermocouples or similar devices near the trunk and

just below the soil surface. However, excluding the terms for vertical

heat transfer in [6-8] allows for independence of the model from soil

temperature data, which is desirable provided model behavior is not

greatly altered.

Simulated air and soil temperatures decreased with time from

initial values to desired minima by linear or negative exponential

functions. Relative humidity was either held constant, or allowed to

increase or decrease linearly with time as is often observed during a

freeze. Wind velocity oscillated around a mean value as a sinusoidal

function of time with a given amplitude and a period of about 30 min to

approximate the stochastic nature of wind velocity observed on freeze

nights.

Parameters

Most of the parameters used in the model were obtained from

handbooks (128, 134), and parameters that could not be obtained from the










existing literature were estimated through laboratory or field

experimentation. The latter included the thermal properties of tree

wraps and some of the characteristics of the irrigation process. The

determination of K, p, c and a of various tree wraps was rather

involved, and is left for discussion in Chapter VII. Parameters used in

the simulation programs are listed in appendices A and B.

Four characteristics of the irrigation process were required prior

to simulation: water temperature, fraction intercepted by the wrap,

amount of the wrap surface covered with water and distribution of the

water on the wrap surface.

Temperature of the water striking the wrap surface was estimated

with thermocouples on the surface of fiberglass wraps directly in the

path of the water spray. Wraps were placed around trunks of 6,

2-year-old 'Hamlin' orange trees and temperatures were recorded hourly

as trees were irrigated with 38 liter hr-I emitters during 4 mild freeze

nights in Jan.-Feb. 1987. Similar data were collected during Dec.-Jan.

1985-86 on wrapped 'Hamlin' orange trees irrigated with 38 and 57 liter
-1
hr- microsprinklers. Water temperature striking the surface ranged

from 100 to 150 C and was relatively constant over time and freeze

conditions.

To estimate the fraction of water intercepted, a fiberglass wrap

placed around 2.5 cm polyvinyl chloride pipe was hung over a bucket, and

irrigation water applied to the wrap for 20 to 30 min using 38, 57, and

87 liter hr-1 microsprinklers. Water that dripped from the wrap was

collected and its volume measured in a graduated cylinder. The volume

intercepted was divided by the total volume of water leaving the

microsprinkler during the same time period to give the fraction




























INTERCEPTION


Schematic diagram of interception of irrigation water by the
tree wrap when microsprinklers were positioned so that water
sprayed above the top of the wrap (left side), or only on the
wrap surface (right side). Hatching indicates areas not
covered by water.


Fig. 6-3.















0.05



0.095


.25


1 I II
' 0.31 '
t 1 / tI


\ II
\ I I



\\ I / /
\ is





MICROSPRINKLER


Cross-section of the model system showing 70% coverage of the
wrap surface with water (sections 1-4,8-10). The numbers
associated with each section indicate the relative amount of
intercepted water distributed uniformly over the surface area
of the section.


0.05


0.095


Fig. 6-4.










intercepted. Values ranged from 0.05-0.11 depending on orientation of

the microsprinkler, being lowest when some water sprayed above 40 cm

(top of the wrap) and highest when water sprayed no higher than 40 cm

(Fig. 6-3). Interception values were adjusted accordingly in

simulations as ice accreted on the wrap and produced a larger surface

for interception.

The percentage of wrap surface covered by water and/or ice and the

distribution of water around the wrap surface are closely related.

These parameters vary with freeze conditions and canopy morphology, and

are difficult to measure in the field. The surface layer of the wrap

was equally divided into 10 sections, and each section was treated

separately in order to vary the coverage and distribution of water

around the wrap surface (Fig. 6-4). On the surface towards the

microsprinkler, all sections were assumed to be covered with water, with

distribution was highest for the section 1, intermediate for sections 10

& 2, and lowest for peripheral sections 9 & 3. Values of distribution

on the surface towards the microsprinkler are proportional to the area

of a plane normal to the streamlines of the irrigation water in each

section. Water can reach the surface opposite the microsprinkler

(sections 4-8) after being deflected or dripped from the canopy,

allowing some or all of these sections to be covered with water and/or

ice. Field observations suggested that only sections 8 and 4 received

any water except during severe freezes when all sections could be

covered with ice and water. Therefore, the total amount of wrap surface

covered by water/ice was estimated to be 70% for most simulations, but

provisions were made in simulation programs for varying this value

between 50-100% in increments of 10%. The volume of water distributed










on the wrap surface away from the microsprinkler was uncertain as well,

but like percent coverage, it could be varied in simulation programs.

Water deflected or dripped from the canopy onto the wrap surface

creates variability in characteristics of the irrigation water by

increasing interception, changing distribution, and lowering water

temperature due to contact with lower temperature objects and longer

distances traveled. While this may not be a factor on the wrap surface

towards the microsprinkler where direct interception prevails, it is an

important consideration for the wrap surface opposite the microsprinkler

because the only water reaching this area is that deflected or dripped

from the canopy. Additionally, water striking the canopy has been

observed to flow down the trunk underneath wraps in the field and reduce

trunk temperature by as much as 60 C at the 20-cm height on mild freeze

nights. The same process probably increases trunk temperatures on severe

freeze nights when trunk temperature normally would drop below 00 C in

the absence of liquid water inside the wrap. There is no provision in

the model to account for water flowing down the trunk and changing its

temperature. This factor, along with the degree of uncertainty with

which the above characteristics of irrigation water are known, decreases

the likelihood of accurate trunk temperature predictions in irrigation

simulations.

Simulation

Simulation programs for the model system were written in FORTRAN

and run on a Digital Vax-ll computer. Separate programs were developed

for tree wraps alone and for irrigation and wraps together, because

different simplifying assumptions were made in each case.










Simulation of Freeze Protection with Tree Wraps

Heat transfer in the angular direction was disregarded when

simulating tree wraps alone, which assumes that temperature does not

vary with position around the system at any given radius r and height z.

This is a reasonable assumption because temperatures at various

positions around the trunk and wrap surfaces at a given height are

observed to be nearly equivalent in the field. Additionally, it was

found that trunk surface temperatures predicted when considering only

radial heat transfer were within + 0.3* C of those predicted for the

20-cm height when considering both radial and vertical heat transfer

(See Fig. 7-3). Trunk temperatures at the 20-cm height were roughly

averages of those for the 1 and 40-cm heights, and disregarding vertical

heat transfer resulted in prediction of a single average trunk

temperature. Disregarding the vertical dimension allowed independence

from soil temperature data if trunk temperatures corresponding to the

20-cm height only were desired, and the simplified program for

1-dimensional (radial) heat transfer ran in 95% less computer time than

the 2-dimensional (radial and vertical) program.

In both the 1- and 2-dimensional simulations, the rate of

temperature change at the wrap surface was calculated using equation

[6-11] (disregarding terms for irrigation and evaporation) after

obtaining values of Q, CV, and R. The terms for angular heat transfer

(and vertical heat transfer for 1-dimensional simulation) were deleted

from [6-8] to calculate rates of temperature change throughout the wrap

and trunk. Distance increments in the radial (Ar) and vertical (Az)

directions were set as high as possible without sacrificing accuracy,










which after many simulation trials, were determined to be 0.2 cm and 1.0

cm, respectively.

Simulation of Freeze Protection with Microsprinkler Irrigation and Tree

Wraps

Heat transfer in the vertical direction was disregarded when

simulating wraps and irrigation together, which assumes that vertical

heat transfer has no effect on temperature in the model system. This is

a reasonable assumption because omitting vertical heat transfer in the

wrap simulation resulted in prediction of an average trunk temperature,

or that for the 20-cm height (see Chapter VII). Furthermore, the

3-dimensional heat transfer simulation used 8 hr of computer time for

each hr of time simulated, hence heat transfer in one dimension had to

be disregarded to allow the simulation to run in a managable time

period. The vertical dimension was the obvious choice for omission

because temperature gradients were lower and distances greater than the

other dimensions, and simulations ran about 50 times as fast without

vertical heat transfer calculations.

The outer system boundary was either water, ice or wrap material,

depending on water distribution and the point on the wrap surface being

considered. The rate of temperature change of the system boundary was

calculated using [6-11], omitting terms E and IR for areas of the wrap

surface not covered by water. Heat transfer in the ice and water layers

was assumed to occur in only the radial direction which simplified

calculations of energy transfer for the outer system boundary. A

subroutine for ice accretion using equations [6-17]-[6-21] was accessed

if the temperature of the water layer was < 00 C or if an ice layer was

already present.






86



Equation [6-8] without terms for vertical heat transfer was used to

calculate rates of temperature change throughout the wrap and trunk.

The radial (Ar) and angular (ae) distance increments were set at 0.5 cm

and 0.63 radians (360) respectively, which were the maximum achievable

without affecting model output. As for the tree wrap simulations, Ar

and AO were determined after many simulation trials to increase speed of

program execution.










Table 6-1. Definition of symbols used in model development.


Symbol


AREA
c
Cv
d
D
E
h
IR
j
K
L
Nu
Q
R
RATE
r
Sh
T
t
W
z


Subscripts


Definition
2
surface area of wrap (cm )
specific heat (cal g oK- )
convective heat flux (cal cm s )
characteristic dimension jcm)
diffusion coefficient (cm s 1
evaporative heat flux (cal cm s )
-3
heat content (cal cm )
sensible heat flux from irrigation water (cal cm s )
heat flux (cal cm s ) 1
-1 -1 K-1)
thermal conductivity (cal cm s K
latent heat (cal g )
Nusselt number (dimensionless)
heat conduction (cal cm s ) -1
radiative heat flux (cal cm s
irrigation rate (liter s )
radius (cm)
Sherwood number dimensionlesss)
temperature (OK or OC)
time (s)
water vapor concentration (g cm-3
vertical distance (cm)

2 -1
thermal diffusivity (cm s )
density (g cm ) -
Stephan-Boltzmann constant (cal cm s K )
angular distance (radians)
"change in"



air
fusion
water
ice
interior region
ice-water interface
outer or exterior region
radial coordinate
system boundary
time
vaporization
water layer
water vapor
vertical coordinate


angular coordinate
















CHAPTER VII
THERMAL PROPERTIES AND SIMULATION OF FREEZE PROTECTION
PERFORMANCE OF TREE WRAPS FOR YOUNG CITRUS TREES

Introduction

Traditionally, young citrus trees have been protected from freezing

injury by banking soil around trunks during periods of cold weather

(59). Following severe freezes, the tree canopy above the soil bank is

killed, but after removal of the bank trees sprout from the surviving

wood and produce new canopies. Although an effective means of freeze

protection, construction and maintenance of soil banks often causes

mechanical or disease damage to trunks, and are labor intensive. For

these reasons, tree wraps made of inert insulating materials were

introduced in Texas in the mid-1950's (80) and are now used in other

citrus growing areas. It was thought that wraps could provide freeze

protection similar to soil banks, but unlike soil banks could be left in

place for 3-4 years, until the tree was better able to survive freezes

by itself. However, trunk temperatures under soil banks typically

typically 80 C higher than air temperature, but only 0 to 40 C higher

than air temperature under tree wraps (67, 68, 121, 170).

Despite lower potential for freeze protection, wraps are widely

used because they provide sprout inhibition and protection from

herbicide, fertilizer and mechanical injury. Wraps composed of dried

plant materials have been used in California to protect trunks from

wind, rodent and radiation damage (129). Furthermore, wraps augment the