Sources of variation in juice quality of 'Valencia' sweet orange (Citrus sinensis (L.) Osb.)

MISSING IMAGE

Material Information

Title:
Sources of variation in juice quality of 'Valencia' sweet orange (Citrus sinensis (L.) Osb.)
Physical Description:
xvi, 168 leaves : ill. ; 29 cm.
Language:
English
Creator:
Barry, Graham H., 1965-
Publication Date:

Subjects

Subjects / Keywords:
Orange juice -- Quality -- Florida   ( lcsh )
Orange juice -- Composition   ( lcsh )
Horticultural Sciences thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Horticultural Sciences -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2000.
Bibliography:
Includes bibliographical references (leaves 155-167).
Statement of Responsibility:
by Graham H. Barry.
General Note:
Printout.
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 - 002673505
oclc - 46673698
notis - ANE0705
System ID:
AA00009000:00001

Full Text












SOURCES OF VARIATION IN JUICE QUALITY OF
'VALENCIA' SWEET ORANGE [Citrus sinensis (L.) Osb.]












By

GRAHAM H. BARRY


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


2000



























To my wife Linda,

whose numerous contributions to this dissertation

and to my life have been of immeasurable value.














ACKNOWLEDGMENTS


I thank Capespan (Pty) Ltd. (formerly Outspan International), South Africa, for its

generous financial support and leave-of-absence for me to fulfill my educational aspirations:

Mr. John Stanbury and Mr. Louis von Broembsen are thanked for their support and for

providing me this opportunity; and Mr. Steve Burdette is thanked for believing in me and

lending his support, understanding, and patience, and for being a lifeline during my sojourn

in Florida.

Tropicana Products, Inc., Bradenton, Fla., are thanked for financial assistance towards

research expenses.

I sincerely thank Dr. Bill Castle, chairman of my supervisory committee, and mentor, for

his counsel throughout my graduate experience, but especially for helping me find where I

want to go as a scientist. I also thank Dr. Fred Davies, cochairman of my supervisory

committee, for his sound advice and guidance, and my supervisory committee members, Drs.

Gene Albrigo, Ed Echeverria, and Russ Rouseff, for their scientific inputs, guidance, and

encouragement.

For technical assistance, above-and-beyond the call of duty, I thank Mr. Jim Baldwin and

Ms. Jean Eelman. I also thank numerous people at the Citrus Research and Education

Center, Lake Alfred, Fla., who willingly provided equipment and technical assistance: Drs.








Jim Syvertsen, Larry Parsons, and Hyoung Lee, Ms. Pam Russ, Ms. Debbie van Clief, Ms.

Margie Wendell, Mr. Gary Test, Mr. Kelly Morgan, Mr. Gary Barthe, Mr. Pedro Gonzalez,

Mr. Tommy Long, Mr. Tom Robnett, Mr. Tam Nguyen, and Mr. Bob Hoobin. For statistical

advice and guidance, I thank Dr. Ramon Littell.

I thank the research cooperators who participated in this research: Mr. Bill Barber and

Mr. Craig Griffith of Lykes Bros., Ft. Basinger; Mr. Frank Bouis of Florida Fruit Managers,

Howey-in-the-Hills; Mr. Bryant Cawley ofBarron Collier, Immokalee; Mr. Wayne Douberley

of Kahn Grove Service, Sebring; and Mr. Jay Dugger of Becker Groves, Ft. Pierce.

In particular, I thank my wife Linda, for the many sacrifices made, and the assistance

provided in the field and in the laboratory.














TABLE OF CONTENTS


page

ACKNOW LEDGMENTS ......................... ........ ........... iii

LIST OF TABLES .................................... ........... viii

LIST OF FIGURES ....................................... ........... xi

ABSTRACT ............................................. ......... xv

CHAPTERS

1 INTRODUCTION .......................... ........ ............. 1

2 REVIEW OF LITERATURE ............................ ............ 3
Fruit Anatomy, Growth and Development ........................ ... 3
Fruit A natom y ............................... ............ 3
Fruit Growth and Development .......................... ... 5
Juice Chemical Composition ............................ ..... 8
Sugar Accum ulation ................................. ..... 9
Factors Affecting Juice Quality ............................... 10
Definition of Juice Quality ............................. 10
Climate ............................................... 11
Canopy M icroclimate ................................. 15
Inflorescence Type .................................. 17
Rootstock Selection ....................................... 20
Fruit Size ........ ...................... ............ 24
Variability in Juice Quality and Sample Size Estimation .................. 25
Seasonal V ariation ........................................ 25
Regional V ariation ........................................ 26
W ithin-Tree Variation ............................... 27
Sample Size Estimation .............................. 27

v










3 VARIABILITY IN JUICE QUALITY OF 'VALENCIA' SWEET ORANGE,
AND SAMPLE SIZE ESTIMATION FOR JUICE QUALITY EXPERIMENTS 29
M materials and M ethods .................................... 30
Sites and Plant M material .................................... 30
Experimental Design and Data Collection ................... 31
Statistical Analysis .................................. 33
Sample Size Estimation .............................. 33
R results ..................................................... 34
Juice Q quality ............. ................ ........... 34
Sample Size Estimation .............................. 41
D discussion ....................................... ........... 44

4 VARIATION IN JUICE QUALITY OF 'VALENCIA' SWEET ORANGE
DUE TO MACROCLIMATE AND CANOPY MICROCLIMATE ........... 48
M materials and M ethods ........................................... 51
Sites and Plant M material ................. ................... 51
Experimental Design and Data Collection ................... 51
Statistical A analysis ........................................ 53
R results ..................................................... 53
Fruit Size and Juice Quality in Relation to Macroclimate and
Canopy M icroclimate .............................. 53
Air Temperature and Cumulative Degree-Days ................... 60
Relationship Between Juice Quality and Air Temperature ........... 65
D discussion ........................................ .......... 67

5 JUICE QUALITY OF 'VALENCIA' SWEET ORANGES BORNE ON
DIFFERENT INFLORESCENCE TYPES .............................. 72
M materials and M ethods .................................... 73
Site and Plant M material ..................................... 73
Experimental Design and Data Collection ................... ... 74
Statistical Analysis .................................. 75
R results ..................................................... 76
Fruit Size and Juice Quality at Physiological Fruit Drop ............ 76
Fruit Size and Juice Quality at Maturity .................... 78
D discussion ........................................ .......... 83










6 WHOLE-TREE SOLUBLE SOLIDS PRODUCTION IN 'VALENCIA'
SWEET ORANGE: ROOTSTOCK SELECTION, CROP LOAD, AND


FRUIT SIZE


........ .. ........................ 85


M materials and M ethods ..........................................
Site and Plant M material ....................................
Treatments and Data Collection .............................
Statistical A analysis .......................................
Results ....................................................
Tree Productivity ........................................
Fruit Size D distribution ................................. ..
Juice Q quality ............................... ...........
Soluble Solids Production ............................... .
Variability in Juice Quality .............................. ..
D iscu ssion ..................................... ............

7 ROOTSTOCK SELECTION AND PLANT WATER RELATIONS AFFECT
SOLUBLE SOLIDS ACCUMULATION OF 'VALENCIA' SWEET ORANGE
M materials and M ethods ..........................................
Site and Plant M material ....................................
Treatments and Experimental Design .........................
D ata C collection ................... .................... .
Statistical A analysis .......................................
R results ......................................................
1998-99 Season .........................................
1999-2000 Season .......................................
D iscu ssion ...................................................


8 OVERALL DISCUSSION ........

9 SUMMARY AND CONCLUSIONS

LIST OF REFERENCES ............


. . . 15 3

. .. . . . 15 5


BIOGRAPHICAL SKETCH


. 86
. 86
. 87
. 89
. 89
. 89
. 94
. 94
100
103
105


110














LIST OF TABLES


Table page


3-1. Geographic location and canopy position (SWT: southwest top; NEB:
northeast bottom) effects on juice soluble solids concentration (SSC) and
titratable acidity (TA) of' Valencia' sweet orange fruit harvested Mar. 1998 and
Mar. 1999............ .................................. 35

3-2. Partitioning of variance into component sources of variation as percentage of
total variance for soluble solids concentration (SSC) and titratable acidity (TA)
of 'Valencia' sweet orange fruit harvested Mar. 1998 and Mar. 1999 to
estimate within-tree variation in juice quality .................... 37

3-3. Mean juice quality of 'Valencia' sweet oranges harvested at maturity in 1998,
1999, and 2000 from four geographic locations in Florida (50-fruit samples
analyzed at the Citrus Research and Education Center state test house facility,
Lake Alfred, Fla)............ .............................. 39

3-4. Partitioning of variance into component sources of variation as percentage of
total variance from a split-plot in time analysis for soluble solids concentration
(SSC) and titratable acidity (TA) of'Valencia' sweet orange fruit harvested at
maturity in 1998, 1999, and 2000 from four geographic locations in Florida to
estimate between-tree variation in juice quality ........................ 40

4-1. Means and probability values for geographic location and canopy position
effects, and probability value for location x position interaction on fruit size and
juice quality of'Valencia' sweet oranges at maturity in Florida during the 1998-
99 and 1999-2000 seasons ............... ........................ .. 58

4-2. Mean maximum, minimum, and annual air temperatures, and annual cumulative
degree-days (using hourly temperature measurements and 13 C base
temperature) of four citrus-producing locations in Florida during the 1998-99
and 1999-2000 seasons, and their probability values for geographic location,
canopy position, and location x position interaction ................ 62








4-3. Coefficients of determination for linear relationships between juice quality
variables of 'Valencia' sweet oranges harvested in Mar. 1999 and Mar. 2000
from two canopy positions and four geographic locations, and cumulative
degree-days (DD), using hourly air temperature measurements and 13 C base
temperature, summed over various phenological periods. .................. 66

5-1. Effect of inflorescence type (IT) on fruit size and juice quality of 'Valencia'
sweet orange fruit at physiological fruit drop stage of fruit development (24
June 1998 and 1 July 1999) from southwest top (SWT) and northeast bottom
(NEB) canopy positions (CP) ............ .......................... 77

5-2. Effect of inflorescence type (IT) on fruit size and juice quality of 'Valencia'
sweet orange fruit harvested in Mar. 1999 and Mar. 2000 from southwest top
(SWT) and northeast bottom (NEB) canopy positions (CP). ................ 79

5-3. Effect of inflorescence type on mean soluble solids concentration (SSC) and
titratable acidity (TA) adjusted for fruit diameter (covariate) of 'Valencia'
sweet orange harvested in Mar. 1999 and Mar. 2000 averaged across southwest
top and northeast bottom canopy positions ................... 81

5-4. Partitioning of variance into components for soluble solids concentration (SSC)
and titratable acidity (TA) of 'Valencia' sweet orange fruit harvested Mar. 1999
and Mar. 2000 to estimate inflorescence type effects on juice quality. ......... 82

6-1. Tree canopy volume, yield, and cropping efficiency of' Valencia' sweet orange
trees on Carrizo citrange and rough lemon rootstocks in Mar. 1999 and Mar.
2000 at Ft. Basinger, Fla. .......... .......................... 90

6-2. Split-plot analysis of variance of yield by tree canopy quadrant of 'Valencia'
sweet orange trees on Carrizo citrange and rough lemon rootstocks in Mar.
1999 and Mar. 2000 at Ft. Basinger, Fla ..................... 92

6-3. Analysis of variance of arcsin-transformed fruit size distribution data for two
rootstocks, four tree canopy quadrants, and five fruit size categories in Mar.
1999 or four fruit size categories in Mar. 2000 of'Valencia' sweet orange fruit
from trees on Carrizo citrange and rough lemon rootstocks at Ft. Basinger,
Fla ............... ........................................... 95

6-4. Analysis of variance of juice quality variables for two rootstocks, four tree
canopy quadrants, and four fruit size categories in Mar. 1999 and Mar. 2000
of 'Valencia' sweet orange fruit from trees on Carrizo citrange and rough
lemon rootstocks at Ft. Basinger, Fla .......... ................... 97








6-5. Coefficients of determination for linear relationships between juice quality
variables and fruit size of 'Valencia' sweet orange fruit from trees on Carrizo
citrange and rough lemon rootstocks at Ft. Basinger, Fla. ................. 101

6-6. Mean yield, juice content, soluble solids concentration (SSC), soluble solids
(SS) production per tree and SS production per fruit by fruit size category
(count) of'Valencia' sweet orange trees on Carrizo citrange and rough lemon
rootstocks harvested in Mar. 1999 and Mar. 2000 ................ ..... 102

6-7. Partitioning of variance into components for juice content, soluble solids
concentration (SSC), titratable acidity (TA), and ratio of SSC-to-TA of
'Valencia' sweet orange fruit harvested Mar. 1999 and Mar. 2000 to estimate
rootstock, tree canopy quadrant, and fruit size effects on juice quality. ....... 104

7-1. Means (n=3, 8-tree plots) and probability values of juice quality and sugar
composition of 'Valencia' sweet oranges for rootstock cultivar and irrigation
treatments on 14 Sept. 1999 at Ft. Basinger, Fla ....................... 128

7-2. Means (n=3, 8-tree plots) and probability values of juice quality and sugar
composition of 'Valencia' sweet oranges for rootstock cultivar and irrigation
treatments on 5 Nov. 1999 at Ft. Basinger, Fla ........................ 129

7-3. Means (n=3, 8-tree plots) and probability values of juice quality and sugar
composition of 'Valencia' sweet oranges for rootstock cultivar and irrigation
treatments on 10 Dec. 1999 at Ft. Basinger, Fla ....................... 130

7-4. Means (n=3, 8-tree plots) and probability values of juice quality and sugar
composition of 'Valencia' sweet oranges for rootstock cultivar and irrigation
treatments on 26 Jan. 2000 at Ft. Basinger, Fla ................. 131

7-5. Means (n=3, 8-tree plots) and probability values of juice quality and sugar
composition of'Valencia' sweet oranges for rootstock cultivar and irrigation
treatments on 15 Mar 2000 at Ft. Basinger, Fla ....................... 132














LIST OF FIGURES


Figur page

2-1. (A) Diagram of an equatorial cross-section through a Citrus fruit. Stippled
areas within black area represent juice sacs and stalks. Two ventral or septal
vascular bundles (svb) appear between radial septae and one dorsal vascular
bundle (dvb) is positioned outside the tangential wall of each septum. (B) A
dorsal vascular bundle from grapefruit and a column of juice sacs attached to
it by juice sac stalks. A portion of the septum epidermis remains attached to the
vascular tissue, and the latter appears as the darkened line nearest the juice sac
stalks.............. .................................... 4

2-2. Fruit growth of 'Valencia' sweet orange in Australia showing the three stages
of fruit develop ent ......... ................................ 6

2-3. Seasonal changes in (A) titratable acidity and (B) ratio of TSS-to-TA of juice
from 'Valencia' sweet oranges grown under contrasting climatic conditions. 13

2-4. Types of inflorescences in Citrus with leaves present, rudimentary or absent.
Diagrams on left are examples of leafy inflorescences, and diagrams on right are
examples of leafless inflorescences: A one flower and 5 to 7 leaves, B one
flower and no leaves, C six flowers and 5 to 7 leaves, and D six flowers and
no leaves............. .................................. 18

3-1. Number of fruit per sample required to estimate differences between two means
for soluble solids concentration (SSC) and titratable acidity (TA) for a given
degree of precision at P,0.05 using 1998 and 1999 variances for SSC and TA.
Degree of precision refers to the difference in % SSC or % TA to be
detected between two means ........................................ 42

3-2. Number of trees replicationsns) required to estimate differences between two
means for soluble solids concentration (SSC) and titratable acidity (TA) for a
given degree of precision at P<0.05 using 1998, 1999, and 2000 variances for
SSC and TA. Degree of precision refers to the difference in % SSC or % TA
to be detected between two means .......... ..................... 43








4-1. Seasonal changes in fruit size and juice quality of'Valencia' sweet orange from
early stage II of fruit development (June) through maturity (March) from four
geographic locations in Florida averaged across two canopy positions
(southwest top and northeast bottom) during the 1998-99 season (n=5 trees,
2 canopy positions, 3 blocks). NS, *=nonsignificant or significant at
P<0.05, respectively .............. ............................... 54

4-2. Seasonal changes in fruit size and juice quality of 'Valencia' sweet orange from
early stage II of fruit development (June) through maturity (March) from four
geographic locations in Florida averaged across two canopy positions
(southwest top and northeast bottom) during the 1999-2000 season (n=5 trees,
2 canopy positions, 3 blocks). NS, *=nonsignificant or significant at
P<0.05, respectively. ........... ........................... 55

4-3. Seasonal changes in fruit size and juice quality of'Valencia' sweet orange from
early stage II of fruit development (June) through maturity (March) from two
canopy positions (SWT=southwest top; NEB=northeast bottom) averaged
across four geographic locations in Florida during the 1998-99 season (n=5
trees, 3 blocks, 4 locations). NS, *-nonsignificant or significant at P<0.05,
respectively............. ................................ 56

4-4. Seasonal changes in fruit size and juice quality of 'Valencia' sweet orange from
early stage II of fruit development (June) through maturity (March) from two
canopy positions (SWT=southwest top; NEB=northeast bottom) averaged
across four geographic locations in Florida during the 1999-2000 season (n=5
trees, 3 blocks, 4 locations). NS, *=nonsignificant or significant at P<0.05,
respectively ............. ....................................... 57

4-5. (A) Mean monthly maximum (upper), mean (middle), and minimum (lower) air
temperature profiles, and (B) mean monthly cumulative degree-day profile for
two seasons (Apr. 1998 to Mar. 1999, and Apr. 1999 to Mar. 2000) of four
citrus-producing regions in Florida averaged across two canopy positions
(n=3 blocks, 2 canopy positions) ........... .................... .. 63

4-6. (A) Mean monthly maximum (upper), mean (middle), and minimum (lower) air
temperature profiles, and (B) mean monthly cumulative degree-day profile for
two seasons (Apr. 1998 to Mar. 1999, and Apr. 1999 to Mar. 2000) of two
canopy positions averaged across four citrus-producing regions in Florida
(n=3 blocks, 4 locations) ......... ............................ 64

6-1. Yield (kg/tree quadrant) by tree canopy quadrant (SE=southeast,
SW=southwest, NE=northeast, NW=northwest) of 'Valencia' sweet orange
trees on Carrizo citrange (CC) and rough lemon (RL) rootstocks in Mar. 1999
and Mar. 2000 at Ft. Basinger, Fla. (n=3, two-tree plots). Bars with same
letters are not significantly different (P<0.05; LSD) ...................... 93








6-2. Fruit size distribution (weight of fruit in each size category as a percentage) of
'Valencia' sweet orange trees on Carrizo citrange (CC) and rough lemon (RL)
rootstocks in Mar. 1999 and Mar. 2000 at Ft. Basinger, Fla. (n=3, two-tree
plots). Bars with same letter are not significantly different (P<0.05; LSD) ...... 96

6-3. Juice content and soluble solids concentration of 'Valencia' sweet orange trees
on Carrizo citrange (CC) and rough lemon (RL) rootstocks in Mar. 1999 and
Mar. 2000 at Ft. Basinger, Fla. (n=3, two-tree plots). Bars with the same letter
are not significantly different (P<0.05; LSD) ........................ 98

6-4. Titratable acidity and ratio of 'Valencia' sweet orange trees on Carrizo citrange
(CC) and rough lemon (RL) rootstocks in Mar. 1999 and Mar. 2000 at Ft.
Basinger, Fla. (n=3, two-tree plots). Bars with the same letter are not
significantly different (P<0.05; LSD) ............................ 99

7-1. Soluble solids concentration of 'Valencia' sweet oranges on Carrizo citrange
(CC) and rough lemon (RL) rootstocks at four irrigation levels (W=well-
watered, C=control, X=no irrigation, and M=no irrigation plus mulch) during
stage III of fruit development in the 1998-99 season. Bars with the same letter,
within a date, are not significantly different (P<0.05; n=4, 8-tree plots). ...... 119

7-2. Concentration of non-reducing, reducing, and total sugars ofjuice of 'Valencia'
sweet oranges from two rootstocks (CC=Carrizo citrange and RL=rough
lemon) and four irrigation treatments (W=well-watered, C=control, X=no
irrigation, and M=no irrigation plus mulch) at the termination of the experiment
on 22 Mar. 1999. Bars with the same letter, within a sugar, are not significantly
different (P<0.05; n=4, 8-tree plots) ....................... 120

7-3. Midday stem water potential of 'Valencia' sweet orange trees on Carrizo
citrange (CC) and rough lemon (RL) rootstocks at four irrigation levels
(W=well-watered, C=control, X=no irrigation, and M=no irrigation plus mulch)
during stage III of fruit development in the 1998-99 season. Bars with the
same letter, within a date, are not significantly different (P<0.05; 21 Jan. 1999:
n=4 trees, no. of leaves=4; all other dates: n=4 replications, no. of trees=2, no.
ofleaves=2). ............ ............................... 121

7-4. Gravimetric soil water content of soil samples (0- to 15-cm depth) taken from
the drip-area between two trees of 'Valencia' sweet orange in the middle of
each plot of two rootstocks (CC=Carrizo citrange and RL=rough lemon) at
four irrigation levels (W-well-watered, C=control, X=no irrigation, and M=no
irrigation plus mulch) at the termination of the experiment on 22 Mar. 1999.
Bars with the same letter are not significantly different (P<0.05; n=4, 8-tree
plots)................... ............................. 123








7-5. Seasonal changes in fruit fresh weight, juice content, and soluble solids
concentration of'Valencia' sweet oranges on Carrizo citrange (CC) and rough
lemon (RL) rootstocks at three irrigation levels (W=well-watered, C=control,
and M=no irrigation plus mulch) during stages II and III of fruit development
in the 1999-2000 season. Means within rootstocks with the same letter are not
significantly different (P<0.05, NS=nonsignificant; n=3, 8-tree plots). ........ 125

7-6. Seasonal changes in fructose, glucose, and sucrose concentrations of'Valencia'
sweet oranges on Carrizo citrange (CC) and rough lemon (RL) rootstocks at
two irrigation levels (W=well-watered and M=mulch, no irrigation) during
stages II and III of fruit development in the 1999-2000 season. Means within
rootstocks with the same letter are not significantly different (P<0.05,
NS=nonsignificant; n=3, 8-tree plots) ........................ 127

7-7. Osmotic potential ofjuice of'Valencia' sweet oranges on Carrizo citrange (CC)
and rough lemon (RL) rootstocks at two irrigation levels (W=well-watered and
M=no irrigation plus mulch) during stages II and III of fruit development in the
1999-2000 season. Bars with the same letter, within a date, are not
significantly different (P<0.05; n=3, 8-tree plots) ................. 136

7-8. Midday stem water potential of 'Valencia' sweet orange trees on Carrizo
citrange (CC) and rough lemon (RL) rootstocks at three irrigation levels
(W=well-watered, C=control, and M=no irrigation plus mulch) during stages
II and III of fruit development in the 1999-2000 season. Bars with the same
letter, within a date, are not significantly different (P<0.05; n=4
replications, no. of trees=2, no. of leaves=2) .......................... 137

7-9. Gravimetric soil water content of soil samples taken from the drip-area between
two trees of'Valencia' sweet orange trees on Carrizo citrange (CC) and rough
lemon (RL) rootstocks at three irrigation levels (W=well-watered, C=control,
and M=no irrigation plus mulch) during stages HI and III of fruit development
in the 1999-2000 season. Bars with the same letter, within a date, are not
significantly different (P<0.05, NS=nonsignificant; n=4, 8-tree plots). ........ 139

7-10. Fibrous root density of trees of 'Valencia' sweet orange trees on Carrizo
citrange (CC) and rough lemon (RL) rootstocks at two irrigation levels
(W=well-watered and M=no irrigation plus mulch) in Mar. 2000. Bars with the
same letter are not significantly different (P0.05; n=3, 8-tree plots). ........ 141














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


SOURCES OF VARIATION IN JUICE QUALITY
OF 'VALENCIA' SWEET ORANGE
[Citrus sinensis (L.) Osb.]

By

Graham H. Barry

December 2000


Chairman: Dr. William S. Castle
Cochairman: Dr. Frederick S. Davies
Major Department: Horticultural Sciences


Juice quality is the basis for compensating sweet orange producers in Florida's citrus

processing industry. Therefore, understanding the causes of and managing variation in

soluble solids (SS) accumulation is the primary concern of these producers.

The effects of climate, canopy microclimate, inflorescence type, and rootstock selection,

and their relative contribution to variability in juice quality of 'Valencia' sweet orange in

Florida were studied during the 1997-98 through 1999-2000 seasons.

Samples consisting of 35 fruit are required to detect differences (P 0.05) of 0.3% soluble

solids concentration (SSC) and 0.06% titratable acidity (TA), whereas 20-fruit samples can

be used to detect differences of 0.4% SSC and 0.08% TA. Seven replications are required








to detect differences of 0.5% SSC and 0.1% TA, with small gains in precision when tree

numbers exceed 10.

Relatively uniform macroclimatic conditions among geographic locations resulted in small

differences in juice quality among locations. The relationship between juice SSC and mean

air temperature was weak. Factors other than air temperature may be important determinants

of juice SSC, e.g., rootstock and irrigation management associated with soil conditions.

Juice quality differences between canopy positions were due to canopy microclimate

factors other than direct effects of air temperature. Juice quality was dependent on the effect

of inflorescence type on fruit size, resulting in dilution of SS, rather than direct effects of

inflorescence type on juice quality.

Rootstock selection had a large effect on SSC and SS accumulation (per fruit).

Differences in SSC between fruit from trees on Carrizo citrange and rough lemon rootstocks

were >25 percent. Fruit yield largely determined SS production per tree. Juice SSC and TA

were inversely related to fruit size.

Differences in fruit size resulted in dilution of SS because of the greater juice volume of

large fruit. However, dilution does not explain differences in juice quality of fruit from trees

on different rootstock selections. It is proposed that juice SSC in fruit of Citrus spp. from

trees on different rootstock selections is due to active accumulation of solutes in juice sacs

by osmotic adjustment in response to rootstock characteristics which affect plant water

relations.













CHAPTER 1
INTRODUCTION



Citrus is among the most abundantly produced fruit crops worldwide. Florida is the

second largest citrus-producing region in the world, after SAo Paulo state, Brazil, and

produces 12 million tonnes of citrus annually (Food and Agriculture Organization, 2000).

Sweet orange [Citrus sinensis (L.) Osb.] is the major type of citrus produced in Florida,

where it is grown primarily for processing into juice products. More than 95% of the oranges

produced in 1998-99 were processed (Florida Agricultural Statistics Service, 2000).

Juice quality is the basis for compensating sweet orange producers for their fruit in

Florida's citrus processing industry. Therefore, understanding the causes of and managing

variation in soluble solids accumulation is the primary concern of these producers. In this

study, juice quality is initially defined in terms of soluble solids concentration (SSC), titratable

acidity (TA), ratio of SSC-to-TA, and juice content. Commercially, fruit may be harvested

when established standards for these variables are met. After the juice is extracted, it is then

subjected to additional analyses for color, flavor, and defects (Fellers, 1990).

Juice quality of sweet oranges can vary widely. Certain factors, such as scion and

rootstock cultivar (Cameron and Soost, 1977; Castle, 1995; Wutscher, 1979), climate

(Hodgson, 1967; Reuther, 1973; Webber, 1948), and cultural practices (Wheaton et al., 1999)

have known effects on juice quality. Many of these factors have been quantified in carefully








2

controlled experiments, and their roles in juice quality development have been confirmed by

relatively consistent results from many studies. Other factors such as leaf-to-fruit ratio, fruit

size, and canopy position are known to influence juice quality, but have received less

attention, and are often viewed as having indirect effects. There are also some factors such

as the type of inflorescence on which fruit are borne and row orientation about which little

is known regarding juice quality.

Interactions among the numerous factors affecting juice quality also occur. However,

juice quality research has invariably focused on how one factor, or sometimes two factors,

affect juice quality, and a more comprehensive study on the numerous factors affecting juice

quality of 'Valencia' sweet orange in Florida has not been reported. Moreover, while many

of the effects of a factor on juice quality have been studied, the physiological cause is not

known.

The effects of a factor on variability in juice quality also play an important role in

maximizing juice quality. However, the relative contributions of these factors to variability

in juice quality are largely unknown. Moreover, for experimental design, variability in juice

quality must be quantified to estimate sample sizes required for sampling purposes.

Among a multitude of factors that can affect juice quality in 'Valencia' sweet orange, a

few have been selected for this study. The objectives of this research were 1) to identify the

principal sources of variation in juice quality of 'Valencia' sweet orange in Florida and to

quantify the effects of climate, canopy microclimate, inflorescence type, rootstock selection,

and plant water relations on juice quality; and 2) to determine the relative contribution of

these sources of variation to variability in juice quality.













CHAPTER 2
REVIEW OF LITERATURE


Fruit Anatomy, Growth and Development


Fruit Anatomy

The fruit of Citrus spp. are botanically classified as a hesperidium, a particular type of

berry. The hesperidium berry differs from other true berries such as tomato (Lycopersicon

esculentum Mill.), grape (Vitis vinifera L.), and blueberry (Vaccinium spp. L.) by having a

leathery peel surrounding a single enlarged ovary, the edible portion of the fruit. This type

of fruit is unique to six genera-Citrus, Fortunella, Poncirus, Microcitrus, Eremocitrus, and

Clymenia-that compose the "True Citrus Fruit Group" of the Citrus subfamily

(Aurantioideae) in the Rue family (Rutaceae), and it is not found in other genera of the

Aurantioideae nor in any other plants (Swingle, 1943). Fruit of Citrus spp. have other

unusual characteristics, including the presence of juice sacs, and polyembryonic seeds.

The salient anatomical features ofhesperidium type berries related to sugar accumulation,

and hence juice quality, include the vascular tissues and juice sacs (Fig. 2-1). The latter are

unique anatomical structures of Citrus fruit, and, being the edible portion, confer economic

value to fruit. Juice sacs contain juice cells that become highly vacuolated with fruit

maturation (Koch et al., 1986), and within these vacuoles are essentially all the soluble solids,

including sugars and organic acids (Baldwin, 1993).








































^^ dvb


Fig.2-1. (A) Diagram of an equatorial cross-section through a Citrus fruit. Stippled areas
within black area represent juice sacs and stalks. Two ventral or septal vascular bundles (svb)
appear between radial septae and one dorsal vascular bundle (dvb) is positioned outside the
tangential wall of each septum. (B) A dorsal vascular bundle from grapefruit and a column
of juice sacs attached to it by juice sac stalks. A portion of the septum epidermis remains
attached to the vascular tissue, and the latter appears as the darkened line nearest the juice
sac stalks (Koch et al., 1986).










Fruit Growth and Development

Citrus fruit develop slowly, taking = 5 to 18 months for an ovary to grow and develop

into a mature fruit, depending on the type of fruit and particular cultivar as well as growing

conditions. Bain (1958) conducted a comprehensive study on the morphological, anatomical,

and physiological changes in developing 'Valencia' sweet oranges [C. sinensis (L.) Osb.] in

New South Wales, Australia. Other developmental and anatomical studies of fruit of Citrus

spp. were reported by Ford (1942) working with 'Eureka' lemon [C. limon (L.) Burm. f.],

Button (1969) with 'Marsh' grapefruit (C. paradise Macf), and Holtzhausen (1969) with

'Washington' navel orange, and Schneider's (1968) monograph on Citrus anatomy. These

studies showed that fruit of different Citrus spp. follow the same general pattern of growth

and development, but differ in the length of time of the various stages of development.

Growth and development from anthesis to maturity of most citrus fruit follows a sigmoid

growth pattern which has been divided into three stages, corresponding with changes in

growth rate (Fig. 2-2), and characterized by distinct morphological, anatomical, and

physiological changes (Bain, 1958). However, the three stages of fruit development occur

as a continuum, without abrupt changes, and it is difficult to separate the transition between

stages when considering only a few developmental criteria.

Stage I is the cell division period lasting 4 to 9 weeks after anthesis in Australia (smid-

to late-Mar. through mid- to late-May in Florida). During this stage, fruit volume and weight

increase due to growth of the peel mainly by cell division (and some enlargement), and nearly

all the cells of the mature fruit are produced. Juice sac primordia also differentiate and

develop during this stage (Roth, 1977).




















VOLUME (ML)


I I *0 --
I .0 0 FRESH WEIGHT (G)


** /


100






50 / I
STAGE I STAGE STAGE Ml

A-A ** -A-A -A
A A'" -A RADIUS (MM)
A I
S0,"0 | DRY WEIGHT (G)
BLOSSOM I I I
NOV DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG SEPT. OCT. NOV. DEC. JAN.


Fig. 2-2. Fruit growth of 'Valencia' sweet orange in Australia showing the three stages of
fruit development (Bain, 1958).








7

Stage II is the cell enlargement and differentiation period lasting -29 weeks in Australia

(-late-May through mid- to late-Dec. in Florida). This stage is a period of maximum growth

and rapid changes, mainly due to growth of the pulp segments, resulting in a rapid increase

in fruit size (Fig. 2-2) accompanied by cell enlargement, differentiation, and expansion of the

albedo. Cells differentiate into various tissue types such as juice sacs, albedo, flavedo, etc.

Juice sacs grow rapidly and become enlarged, filling the pulp segments. Water accumulates

in the pulp segments during stage II and is the main cause of increased fresh weight of

developing fruit. As a result, juice content increases in the enlarging cells. During the period

of rapid water accumulation, soluble solids (SS) content increases, but soluble solids

concentration (SSC) decreases. During the latter period of stage II, when the water

accumulation rate decreases, both SS content and SSC increase through maturation.

The duration of stage II varies with fruit type and cultivar from 2 to 3 months for lemons

and limes (C. aurantifolia Swing.) to more than 6 months for sweet oranges and grapefruit

(Davies and Albrigo, 1994). Within a cultivar, the duration of stage H may also vary from

3 to 4 months in lowland tropical conditions to 10 months in coastal subtropical conditions

(Reuther and Rios-Castafio, 1969).

Stage III is the maturation period, lasting about 22 weeks in Australia (from late Dec.

in Florida) to "maturity," and is distinguished from stage II by decreased morphological,

anatomical, and physiological changes. The flavedo changes color from green to yellow to

orange (depending on climatic conditions), and the growth rate of the fruit decreases, but

juice sacs continue to enlarge. Changes in juice chemical composition occur slowly and

include mainly an increase in soluble sugars and a continued decrease in the concentration of








8

organic acids from a concentration peak attained earlier in development. This results in a

gradual increase in SSC, and a rapid decrease in titratable acidity (TA) in the juice. The most

rapid changes in chemical composition occur during the latter part of stage II and beginning

of stage III.

Stage III can be further subdivided into a period of rapid changes in juice chemical

composition, a decline in the rate of change, development of secondary metabolites associated

with flavor development, and senescence.


Juice Chemical Composition

Citrus fruit, like other fleshy fruit, are predominantly water, with a water content of

-85% to 90% (w/w) in the edible portion of mature sweet oranges. Juice soluble solids of

sweet oranges are chiefly composed of soluble sugars (z 80%) and organic acids (~ 10%),

which, together, control the sweet and sour sensations of taste. Other soluble components

are present in relatively low concentrations and are important in determining juice flavor. The

major sugars in citrus juice include non-reducing (sucrose) and reducing (fructose and

glucose) sugars, present at = 1:1 ratio (Erickson, 1968). Citric acid is the characteristic

organic acid of sweet oranges (70% to 90% of total) (Sinclair and Ramsey, 1944). Malic acid

is the second most abundant organic acid (Sinclair et al., 1945), with lesser amounts of other acids.

Collison (1913) conducted the first published study on seasonal changes in juice quality

of sweet oranges in Florida, followed by a more comprehensive study by Harding et al.

(1940). Although conducted over only one season, Collison (1913) summarized the seasonal








9

changes of several sweet orange, mandarin (C. reticulata Blanco), and grapefruit cultivars and

selections as a gradual increase of total sugar and gradual decrease of acidity as fruit mature.

Non-reducing and reducing sugars increase from stage II of fruit development, the former

more rapidly than the latter (Collison, 1913; Harding et al., 1940), resulting in an increase in

SSC with fruit development (Harding et al., 1940). Organic acids accumulate rapidly in stage

I of fruit development (Bain, 1958; Bartholomew, 1923), and decrease in concentration

during later stages of fruit development (Harding et al., 1940).


Sugar Accumulation

Juice cell vacuoles are the ultimate site of accumulation of photoassimilates in juice cells

(Echeverria and Valich, 1988), where compartmentalization of sucrose, hexose sugars, and

organic acids (mainly citric and malic) occurs (Echeverria and Valich, 1988; Erickson, 1968).

Increasing sugar concentration (Koch and Avigne, 1990) and osmotic potential (Kaufinann,

1970) gradients occur along the post-phloem assimilate transfer path, i.e., the juice sac stalk.

These gradients are established and maintained via compartmentalization of solutes and

sucrose metabolism. Enzymes of sucrose metabolism, including sucrose synthase, sucrose

phosphate synthase, and soluble and alkaline invertases, have been proposed as key enzymes

in sugar accumulation at different stages of fruit development (Kato and Kubota, 1978; Koch

and Avigne, 1990; Lowell et al., 1989). In addition, nonenzymatic sucrose cleavage under

low vacuolar pH also contributes to sucrose breakdown (Echeverria and Bums, 1989;

Echeverria et al., 1992). During the final third of fruit development, juice sacs accumulate

sugars at a reduced rate (Bain, 1958; Koch and Avigne, 1990).










Factors Affecting Juice Quality


Definition of Juice Quality

Juice quality is difficult to define. It does not consist of a single attribute, but many

factors, some of which are easily measured and used to define quality. Other juice

components are more difficult to measure and their roles in determining juice quality are less

understood. Nevertheless, while quality has subjective elements, attempts have been made

to objectively determine and to quantify components of juice quality (Fellers, 1990;

Wardowski et al., 1995).

Citrus juice quality, as used in this study, is commonly defined in terms of SSC, or total

soluble solids (TSS), as measured by Brix hydrometer or refractometer and corrected for acid,

TA expressed as citric acid equivalents as determined by titration with NaOH, and ratio of

SSC-to-TA (Braddock, 1999; Fellers, 1990; Wardowski et al., 1995). Commercially, fruit

may be harvested when established standards for these variables are met. In Florida, where

most sweet oranges are processed, juice quality is legally defined using a 100-point grading

system for color (40 points), flavor (40 points), and defects (20 points) (Fellers, 1990).

Juice quality is primarily a genetic trait, and is secondarily affected by climatic and other

growing conditions which have a large influence on juice quality when combined (Castle,

1995). The major factors affecting juice quality are scion and rootstock cultivar (Cameron

and Soost, 1977; Castle, 1995; Wutscher, 1979; Wutscher and Shull, 1972), climate

(Hodgson, 1967; Reuther, 1973; Webber, 1948), photosynthetically active radiation (PAR)

(Reitz and Sites, 1948; Syvertsen and Lloyd, 1994; Wheaton et al., 1995), and plant water








11

relations (Koo and Sites, 1955; Sites et al., 1951). Other important factors include nutrition

(Embleton et al., 1967; Koo, 1988), tree age, soil conditions, crop load (Wheaton et al.,

1999), and fruit size (Harding and Lewis, 1941).


Climate

Climate is the sum total of all individual meteorological occurrences (weather processes)

at a given place (Geiger, 1965). It comprises the average conditions and regular sequences

of weather, i.e., the long-term condition of the atmosphere. Weather is the condition of the

atmosphere over a short period of time, i.e., the momentary state of the atmosphere

(Trewartha, 1954). Climate is sometimes arbitrarily divided into micro-, meso-, and

macroclimate (Geiger, 1965). Microclimate is the climate found within a small space, e.g.,

a tree's canopy. Mesoclimate is the transitional stage between micro- and macroclimate

caused by, for example, topographical differences between two locations within a relatively

small geographic range. Macroclimate is synonymous with "climate," and is the average

weather of a particular location.

The elements of climate (and weather) are solar radiation, air temperature, precipitation

and humidity, air pressure, and wind, which are under the climatic controls of latitude,

altitude, land and water masses, topography, and ocean currents (Trewartha, 1954).

Among the secondary factors affecting juice quality, climate has the greatest effect

(Reuther, 1973), and largely dictates where citrus is commercially produced and the quality

of fruit produced (Gat et al., 1997; Hodgson, 1967; Reuther, 1973; Webber, 1948). Climate

is widely recognized as a major factor affecting juice quality as evidenced by the inclusion of








12

climatic adaptation on plant breeders' lists of objectives (Hodgson, 1967; Janick and Moore,

1975). The effects of climate on juice quality of Citrus spp. have been reviewed by Webber

(1948), Hodgson (1967), Reuther (1973), and Gat et al. (1997). The most important

elements of climate related to juice quality are air temperature, rainfall, and solar radiation.

Air temperature is thought to have an overriding effect on fruit growth, development, and

maturation (Reuther, 1973). Although SS accumulation is more rapid under lowland tropical

compared with cool subtropical conditions, fruit from the two regions have similar juice SSC

at maturity (Reuther and Rios-Castafio, 1969; Reuther, 1973). Degradation of organic acids

appears to be affected more by air temperature than is SS accumulation, with fruit from

lowland tropical conditions having lower TA at maturity than fruit grown under subtropical

conditions.

The association between climate and juice quality between and within citrus-producing

regions of the world has been demonstrated in numerous studies (Bain, 1949; Cooper et al.,

1963; Nauer et al., 1974; Newman et al., 1967; Reuther and Rios-Castanio, 1969; Reuther et

al., 1969; Webber, 1948). The effects of climate on fruit maturation were demonstrated in

a classic study by Reuther and Rios-Castaijo (1969) who compared two contrasting climatic

conditions under which citrus is commercially produced, subtropical California and tropical

Colombia. The time taken for 'Valencia' sweet orange to attain a SSC-to-TA ratio of 9:1

was <7 months in Palmira, Colombia (tropical conditions), half that of Santa Paula, Cal. (cool,

subtropical conditions), whereas the time taken in Orlando, Fla., was intermediate (Fig. 2-3)

(Reuther, 1973).












(A)


Santa Paula,Calif.


Palmira,
Colombia


5 6 7 8 9 10 11 12 13 14 15 16 17
Months after anthesis


Palmira,
Colombia


Medellin,
Colombia


Santa Paula,Calif.


5 6 7 8 9


Months after anthesis


Fig. 2-3. Seasonal changes in (A) titratable acidity and (B) ratio ofTSS-to-TA ofjuice from
'Valencia' sweet oranges grown under contrasting climatic conditions (Reuther, 1973).


18 19


o
212

o10

I-f


m








14

Comparisons of the effects of climate on juice quality among citrus-producing regions

within the United States (Cooper et al., 1963; Newman et al., 1967; Reuther et al., 1969) and

within California's citrus-producing regions (Nauer et al., 1974; Webber, 1948) showed that

the principal regional difference in juice quality was the time required to attain a given

maturity index. For example, in Orlando, Fla., fruit from 'Valencia' sweet orange trees on

sour orange rootstock took 10 to 11 months to attain a 9:1 ratio, compared with 8 to 9

months for fruit from trees grown in Weslaco, Texas, and 11 to 12 months for Riverside, Cal.

(Newman et al., 1967). Differences in juice SSC at maturity among these regions were

relatively small compared with differences in TA.

The concept ofheat unit accumulation, or degree-days, has been used for making climatic

comparisons among citrus-producing regions (Gat et al., 1997; Reuther, 1973), and to

determine where different cultivars can be commercially produced (Barry et al., 1996; Ben

Mechlia and Carroll, 1989b). Cumulative degree-days have also been used to predict maturity

with varying success (Ben Mechlia and Carroll, 1989a; 1989b; Kimball, 1984; Lomas et al.,

1970; Newman et al., 1967). Cumulative degree-days (DD) are calculated from daily mean

air temperature using 13 C as a base temperature (Reuther, 1973) according to the equation

(Newman et al., 1967):

DD = (mean monthly air temperature 13) x days in month.

Generally, the cumulative DD method to predict maturity dates is considered to be rather

crude (Spiegel-Roy and Goldschmidt, 1996), since the response of fruit at different

developmental stages to temperature varies (Newman et al., 1967). However, using early

season (from the date of flowering to the end of Apr.) cumulative DD, maturity dates of








15

grapefruit were forecast to within 6 d in Israel's Jordan Valley (Lomas et al., 1970). Kimball

(1984) demonstrated a strong relationship (tr=0.96) between juice SSC-to-TA ratio of

'Washington' navel orange and cumulative maximum air temperatures during the early stages

of fruit development (May to Aug. in Tulare Co., Calif.). Ben Mechlia and Carroll (1989a)

also modeled the fruit maturation process in two stages, the time to 6:1, and the time to 9:1

ratio. Mean daily changes in ratio as a function of mean daily temperature was used to give

a more accurate prediction of maturity than cumulative DD.

The simplest method of forecasting maturity dates is to use the average date of maturity

from historical records (Lomas et al., 1970). In Israel, the standard deviation from the mean

maturity date was +10 d, but the maturity dates of individual seasons varied by up to 20 d.


Canopy Microclimate

Canopy microclimate is the climate within and immediately around a tree's canopy, i.e.,

the leaf and shoot system of a tree (Geiger, 1965), and differs from the above canopy ambient

climate due mainly to the size, shape, arrangement, and density of leaves within the canopy

(Kliewer and Smart, 1989). The importance of canopy microclimate to fruit production and

particularly to fruit quality in many fruit species has been recognized by researchers and

growers for many years (Lakso et al., 1989). In the last three decades, the effects of canopy

microclimate or exposure have been quantified, and reviews by Jackson (1980), Smart (1985),

and Palmer (1989) discussed the relationship of canopy light microclimate to several fruiting

processes in apple (Malus spp.) and grape, including internal fruit quality. Considerably more








16

research on canopy microclimate has been conducted on deciduous tree crops and grapes than

subtropical evergreens, including Citrus.

Studies on canopy microclimatic effects on fruit quality of Citrus spp. include sweet

orange (de Vries and Bester, 1996; Fallahi and Moon, 1989; Randhawa and Dinsa, 1947;

Reitz and Sites, 1948; Sites and Reitz, 1949; 1950a; 1950b; Wallace et al., 1955; Winston,

1947), grapefruit (Fallahi and Moon, 1989; Syvertsen and Albrigo, 1980; Wood, 1938),

mandarin (Cohen, 1988; Fallahi and Moon, 1989; Morales et al., 2000), and lemon (Fallahi

and Moon, 1989). These studies showed a consistent relationship between the position that

fruit were borne in the canopy and juice quality. This relationship was commonly attributed

to canopy microclimate factors. For example, juice SSC of individual 'Valencia' sweet

orange fruit within a single tree varied from 5.9% to 13.5% SSC according to their position

within the canopy (Reitz and Sites,1948). In the northern hemisphere, 'Valencia' sweet

oranges from southern-top canopy sectors, where the greatest net radiation occurs, tended

to have higher juice SSC and juice content than fruit from other canopy sectors (Sites and

Reitz, 1949; 1950a; 1950b). Upper canopy positions have higher maximum air temperatures

(Allen and McCoy, 1979; Morales et al., 2000). Differences in juice quality may be

associated with canopy microclimate and exposure of individual fruit to environmental

stresses (Syvertsen and Albrigo, 1980).

Reitz and Sites (1948) attributed canopy position differences in juice quality to "light

classes", implicating photosynthetically active radiation (PAR) as an important factor affecting

within-tree variation in juice quality, although this effect has not been quantified. However,

not all of this effect is due to enhanced photosynthetic effect on sugar accumulation








17

(Syvertsen and Albrigo, 1980). Some of the effect is due to differences in water stress

resulting in less dilution of soluble solids in canopy areas with higher radiation exposure

(Syvertsen and Albrigo, 1980), while other factors probably also have an effect, such as fruit

size and crop load (Albrigo, 1992).


Inflorescence Type

Citrus inflorescences are classified as leafy and leafless (Reece, 1945), by the number of

flowers borne per floral shoot (Randhawa and Dinsa, 1947), or according to the absence or

presence of leaves, the latter being sub-divided further depending on flower-to-inflorescence

leaf ratio (Lenz, 1966; Sauer, 1951; Fig. 2-4). The term "inflorescence" applies to all

flowering shoots arising from axillary buds (Reece, 1945).

Citrus fruit set research conducted during the 1940s through 1980s demonstrated that

leafy inflorescences set a higher percentage of fruit (Jahn, 1973; Lenz, 1966; Moss et al.,

1972; Reece, 1945; Sauer, 1951) and produce larger fruit (Ehara et al., 1981; Guardiola and

Litzaro, 1987; Lenz, 1966) than leafless inflorescences. The proposed mechanisms to explain

these effects involve increased supply of carbohydrates by inflorescence leaves to developing

ovaries (Monselise, 1986; Moss et al., 1972), phytohormones (Erner and Bravdo, 1983;

Guardiola, 1992), or nitrogenous compounds (Lovatt et al., 1988), and of the possible higher

sink strength of leafy inflorescences as a whole (Erner, 1989). Leafy inflorescences, or

associated ovaries, are strong sinks (Moss et al., 1972), despite one study showing higher sink

strength of ovaries from leafless inflorescences (Erner and Bravdo, 1983).





















































Fig. 2-4. Types of inflorescences in Citrus with leaves present, rudimentary or absent.
Diagrams on left are examples of leafy inflorescences, and diagrams on right are examples of
leafless inflorescences: A one flower and 5 to 7 leaves, B one flower and no leaves, C six
flowers and 5 to 7 leaves, and D six flowers and no leaves (Lenz, 1966).








19

Only two studies exist concerning inflorescence effects on fruit quality in citrus (Ehara

et al., 1981; Lenz, 1966). In the earlier study (Lenz, 1966), there was little difference in juice

quality of 'Valencia' sweet orange fruit harvested from leafy and leafless inflorescences,

although fruit from leafy inflorescences tended to have slightly higher TSS (-0.5 %), TA

(<0.1%), and juice content (<1%), despite 5% larger fruit size. There was little difference

in ratio between fruit from the two inflorescence types.

In the second study (Ehara et al., 1981), the effects of increasing inflorescence leafiness

on fruit quality of satsuma mandarin (C. unshiu Marc.) included z 10% larger fruit size, less

flesh relative to peel, more advanced rind color development, and slightly lower juice TSS and

TA. However, these reported differences were not subjected to statistical analysis.

Furthermore, the slightly lower TSS and TA reported may not have been due to inflorescence

typeper se, but differences in fruit size due to inflorescence type or canopy position (Harding

and Lewis, 1941; Ketsa, 1988).

Juice quality among fruit borne on different inflorescences may vary due to the presence

of inflorescence leaves providing a greater source of photoassimilates, and stronger

competition for metabolites via increased sink strength and preferential photoassimilate

partitioning (Moss et al., 1972), or competition between developing organs (leaves and fruit)

during early stages of fruit development when leaves are net importers of photoassimilates

(Kriedemann, 1969).










Rootstock Selection

Rootstock selection has large effects on juice quality of sweet oranges, and trees on

different rootstocks can produce fruit that differ by as much as 30 percent in juice SSC

(Wutscher, 1988). The relationship between rootstock selection and fruit quality of citrus has

been the subject of two reviews (Castle, 1995; Wutscher, 1988).

Citrus rootstock selections have well-known effects on tree size, crop load, fruit size, and

various fruit quality factors (Castle, 1987; Castle et al., 1993; Wutscher, 1979; 1988). The

classic example of differences in fruit quality among rootstocks is represented by comparison

of fruit borne on trees on rough lemon (Citrus jambhiri Lush.) with trees on sour orange (C.

aurantium L.) rootstocks (Albrigo, 1977; Cook et al., 1952; Harding, 1947). Fruit from trees

on rough lemon are generally large, and low in SSC and TA compared with fruit from trees

on sour orange which produce superior quality fruit because of their high SSC, and good

flavor and size (Castle et al., 1993).

Citrus rootstocks have been characterized according to the vigor imparted to the scion,

and tree productivity (Castle et al., 1993). Scions budded on vigorous rootstocks, e.g., rough

lemon, Volkamer lemon (C. volkameriana Ten. & Pasq.), and macrophylla (C. macrophylla

Wester), generally produce large fruit with lower SSC and TA. Trees on less vigorous

rootstocks, e.g., many trifoliate orange [Poncirus trifoliata (L.) Raf] selections and some of

its hybrids such as the citranges [P. trifoliata (L.) Raf. x C. sinensis (L.) Osb.] and citrumelos

[P. trifoliata (L.) Raf. x C. paradise Macf ], produce smaller fruit but with higher SSC and

TA (Castle et al., 1993; Reitz and Embleton, 1986).








21

The reasons why trees on different rootstock selections consistently produce fruit with

different sugar and acid concentrations are not well understood (Castle, 1995), and the

physiological mechanisms) of rootstock effect on sugar accumulation in citrus fruit has not

been studied. Gardner (1969) noted the absence of published reports on the mechanism of

rootstock influence on juice quality, a situation that has not changed in the ensuing 30 years.

Though it is unclear how rootstocks exert their influence on juice quality, plant water

relations, mineral nutrition, and phytohormones have been proposed as being among the most

important factors involved (Castle, 1995).

Castle (1995) presented an hypothesis which states that citrus juice quality is closely

related to rootstock effects on plant water relations as evidenced by field trial results (Koo

and Sites, 1955; Sites et al., 1951), sucrose transport (Goldschmidt and Koch, 1996; Koch,

1984), and reciprocal fruit grafting studies (Gardner, 1969). The possibility of rootstock

involvement in juice quality is supported by water-deficit stress experiments (Maotani and

Machida, 1980; Peng and Rabe, 1996; Sugai and Torikata, 1976; Yakushiji et al., 1996;

1998). Conclusions from these studies demonstrated an association between plant water

relations and juice quality, while the physiological cause of rootstock effect on juice quality

remains unknown.

Albrigo (1977) demonstrated the central role played by plant water relations in juice

quality of Citrus. 'Valencia' sweet orange trees on rough lemon were less water stressed

(higher leaf water potential) and had lower juice content and SSC than trees on sour orange

or Carrizo citrange. Albrigo (1977) suggested that rootstocks influence SSC through the

degree of dilution caused by the average water potential a particular rootstock maintains for








22

the scion cultivar. It was later proposed that the majority of rootstock influence on fruit

development is by the rootstock's ability to supply water to the plant, and secondarily through

nutrient uptake (Albrigo, 1992). Since vigorous rootstocks are better extractors of soil water

(Castle and Krezdorn, 1977) and keep the plant under less water stress, Albrigo (1977; 1992)

proposed that this was the main reason why vigorous rootstocks result in lower SSC in

individual fruit.

Another role of plant water relations in explaining differences in sugar accumulation

among trees on different rootstocks is based on the observation that plants under water-deficit

stress accumulate more sugars than unstressed plants (Meyer and Boyer, 1981; Yakushiji et

al., 1996; 1998), a survival mechanism to maintain cell turgor and tolerate water stress. The

concept of osmotic adjustment' was proposed as a mechanism to explain sugar accumulation

in Citrus (Yakushiji et al., 1996; 1998).

Osmotic adjustment involves solute accumulation in cells sufficient to decrease cell

osmotic potential when cell water potential decreases at low water potential so that water can

be absorbed from the water source by cells without losing cell turgor or decreasing cell

volume (Morgan, 1984). In the process, osmotically active solutes accumulate in juice

vesicles. When osmotic adjustment occurs under water-deficit conditions, cell size and cell

turgor should be maintained due to active solute accumulation in cells at low water potentials.

When many plants acclimate to water deficits, the maintenance of cell turgor by osmotic


1 Osmotic adjustment, or osmoregulation, is a physiological function triggered by water-

deficit stress. It is the maintenance of a constant internal osmotic pressure, when external

concentration varies (Meyer and Boyer, 1981).








23

adjustment is an important physiological mechanism which can minimize the detrimental

effects of water-deficit stress (Morgan, 1984), and the drought tolerance of many plants is

largely dependent on their capacity for osmotic adjustment and maintenance of cell turgor

through the accumulation of solutes. Inorganic cations, organic acids, amino acids, and

sugars are known primary osmotica that accumulate after internal synthesis or uptake from

external media (Kramer and Boyer, 1995).

Working on satsuma mandarin in Japan, Yakushiji et al.(1996; 1998) demonstrated that

osmotic adjustment occurred in fruit as a mechanism of accumulating sugars under low water

potentials. When trees were moderately water-stressed by mulch treatment or withholding

irrigation, sugars accumulated in the fruit by active osmotic adjustment. Sugar accumulation

in satsuma mandarin fruit was not caused by dehydration under water stress. Water-deficit

stress increased the concentration of sucrose, glucose and fructose, plus total sugar content

of fruit (Yakushiji et al., 1998).

During osmotic adjustment under water-deficit stress, cells must rely on imported solutes

to lower osmotic potential of cells faster than a decrease in water potential of the surrounding

environment to prevent a decrease in cell volume due to dehydration. Yakushiji et al. (1996)

demonstrated that monosaccharides are active components of citrus fruit for osmotic

adjustment. Sucrose degradation to glucose and fructose in juice vesicles provides the

osmotic potential required (Lowell et al., 1989), while osmotic adjustment actively occurred

in fruit juice vesicles under moderately water-stressed conditions as demonstrated by lowering

of juice vesicle potential but not of turgor, resulting in osmotically active solutes

accumulating, viz., hexose sugars. Differences in C partitioning among treatments were








24

related to translocation capability under different water status resulting in stronger sink

activity, as a result of osmotic adjustment and hence greater SSC and sugar content. It is

possible to sustain growth at low water potentials if sufficient osmotic adjustment takes place

in plants. Therefore, it is possible for satsuma mandarin fruit to sustain growth by

accumulating sugars for osmotic adjustment under moderate water-deficit stress. Sugar

accumulation in the fruit was caused by an increase in the partitioning rate of translocation

of photosynthates into fruit under water-deficit stress (in spite of a presumed decrease in

photosynthesis). Thus, fruit were considered to become a large sink for photosynthate

translocation when satsuma mandarin plants were moderately water-stressed at low water

potentials. Since sucrose is the primary substance for translocation and phloem unloading in

citrus fruit (Garcia-Luis et al., 1991; Kriedemann, 1969; Sawamura et al., 1975), degradation

of sucrose to fructose and glucose must be promoted in juice sacs during the process of active

osmotic adjustment in satsuma mandarin fruit as suggested by Yakushiji et al. (1996).

The recent Japanese work (Yakushiji et al., 1996; 1998) implicates a direct role of water

in osmotic adjustment, which causes differences in sink strength and therefore assimilate

translocation and accumulation. This provides support for the role of plant water relations

in rootstock effect on sugar accumulation in Citrus (Albrigo, 1977; 1992).


Fruit Size

Small fruit tend to have higher SSC and TA than large fruit (Harding and Lewis, 1941;

Ketsa, 1988; Miller, 1990). The inverse relationship between juice SSC and TA, and fruit size

is frequently neglected in juice quality research as noted by Gardner (1969), who








25

recommended that juice quality variables be adjusted for fruit size by regression analysis.

Since larger fruit have greater juice volume than smaller fruit (Harding and Lewis, 1941),

dilution of juice soluble solids of larger fruit is commonly used to explain differences in

treatments affecting juice quality variables (Albrigo, 1977; Sites and Camp, 1955; Ketsa,

1988; Ting and Attaway, 1971). However, fruit size is inversely related to crop load

(Gallasch, 1988; Guardiola, 1988; Syvertsen and Lloyd, 1994). Therefore, the numerous

factors (genetic, environmental, horticultural, physiological, etc.) that affect crop load may

have indirect effects on juice SSC and TA.


Variability in Juice Quality, and Sample Size Estimation


The factors affecting juice quality can also contribute to variability in juice quality, but

little is known regarding the magnitude of their contribution. Additionally, in experimental

design, variability is used to estimate sample sizes required for sampling. Relatively little has

been reported regarding variability in juice quality of citrus fruit. Variability in juice quality

among orchards within a growing location, among trees within an orchard (Appleman and

Richards, 1939), within a tree (Reitz and Sites, 1948), and among fruit (Denny, 1922) all

contribute to total variation in juice quality within a growing location.


Seasonal Variation

Any citrus industry experiences season-to-season variation in juice quality (Fellers, 1985)

in the form of the time taken to achieve a particular maturity standard, which can vary by as

much as 3 to 4 weeks from one season to another (Harding et al., 1940). Seasonal variation








26

in TA is usually greater than seasonal variation in SSC (Harding and Sunday, 1949; Harding

et al., 1959). Some of the factors that may have a bearing on seasonal variation in juice

quality must be primarily linked to prevailing weather conditions during the growing season,

especially air temperature, rainfall (amount, timing, and distribution), and sunshine hours due

to cloud cover (Reuther, 1973), but also other factors, e.g., crop load and associated fruit size

(Albrigo, 1992).

Seasonal variation in juice quality can be greater than variation among treatments within

a season (Harding et al., 1940), highlighting the need to conduct juice quality research over

numerous seasons. In addition, since seasonal variation can be greater than within-season

variation in juice quality, trends in treatment effects over more than one season can be masked

by seasonal variation.


Regional Variation

Juice quality depends on, amongst other factors, the environmental conditions under

which fruit are produced with large differences in juice quality occurring among contrasting

climatic conditions (Reuther and Rios-Castafio, 1969). Conversely, small differences in juice

quality occur among regions with relatively small differences in climatic conditions (Harding

et al., 1940). However, the contribution of growing region to variation in juice quality has

not been reported. Possible causes of between-location variation in juice quality include

within-site variation, and between-tree and within-tree variation in juice quality. Between-

location variation is an integral component of genotype x environment interaction studies.








27
Mesoclimatic differences, due to slope, aspect, topography, and natural heterogeneity of

soil conditions, could contribute to variation in juice quality within a given growing location

(Jackson and Lombard, 1993; Platt, 1973). Variation in juice quality within a location can

be large, and may be caused by between-orchard and between-tree variation in juice quality.

Appleman and Richards (1939) reported that tree-to-tree variation in juice quality of fruit

samples of uniform size from trees carefully selected for uniformity occurred. The cause of

this variation is unknown, but natural heterogeneity in site conditions, resulting in differences

in tree vigor, even under well-managed conditions, could be implicated. Tree-to-tree

variation in SSC and TA may be large, in spite of sampling from trees of similar vigor and

crop load (Appleman and Richards, 1939).


Within-Tree Variation

Variation in juice quality among fruit from individual trees is relatively large (Denny,

1922; Reitz and Sites, 1948; Wallace et al., 1955), and is related to the position fruit are

borne on a tree and associated canopy microclimate, as well as to fruit size, crop load and

leaf-to-fruit ratio (Albrigo, 1992), and, possibly, inflorescence type. The principal causes of

within-tree variation in juice quality may be due to differences in PAR, leaf and fruit

temperature (as a result of air temperature), and VPD in the different canopy positions

(Syvertsen and Albrigo, 1980).


Sample Size Estimation

Collecting representative samples is important in all research (Steel and Torrie, 1980),

including horticulture. The estimation of sample size requires a knowledge of the magnitude








28

and composition of variability (Marini and Trout, 1984; Schultz and Schneider, 1955), and

is an essential prerequisite of experimental design (Sharpe and van Middelem, 1955). The

concept of partitioning total variance into its components is a useful statistical tool for

understanding the relative importance of various sources of variability (Hanson, 1955).

Sample size studies on citrus juice SSC were conducted primarily to establish criteria to

enable representative fruit sampling of commercial lots of citrus for remuneration purposes

(Appleman and Richards, 1939; Bartholomew and Sinclair, 1943; Denny, 1922; Reitz and

Sites, 1948; Wallace et al., 1955), and to estimate the sample size required to ensure

representative sampling for research purposes (Reitz and Sites, 1948).

Based on these sample size studies, sample sizes ranging from 18 to 36 fruit are required

to collect a representative sample and to detect differences of 0.5% to 0.6% SSC in navel and

'Valencia' sweet oranges. Invariably these experiments used only one set of fruit samples to

estimate variability in juice quality, whereas Reitz and Sites (1948) demonstrated how juice

quality varied among all fruit within a single tree. Reitz and Sites (1948) suggested that 20

'Valencia' sweet orange fruit of uniform size, to minimize variation, borne on the outside of

a citrus tree's canopy at a height of 1 to 2 m resulted in a sample representative of the entire

tree. The number of replications required to detect differences between treatment means were

not determined in these studies.

The results of some of these studies have subsequently been used for sampling purposes

to determine when to harvest fruit (de Vries and Bester, 1996; de Vries and Moelich, 1996;

Netterville, 1995; Wardowski et al., 1995).













CHAPTER 3
VARIABILITY IN JUICE QUALITY OF 'VALENCIA' SWEET ORANGE, AND
SAMPLE SIZE ESTIMATION FOR JUICE QUALITY EXPERIMENTS



Juice soluble solids concentration (SSC) and titratable acidity (TA) of citrus are

intrinsically variable. This variability can be large, and is a result of various factors that affect

juice quality. Some factors affect individual fruit (Denny, 1922), whereas most effects on

variation in juice quality are at the whole-tree level (Appleman and Richards, 1939; Reitz and

Sites, 1948; Sites and Reitz, 1949), resulting in variability in juice quality among orchards

within a region, among trees, within a tree, and among fruit.

A knowledge of the factors affecting juice quality, and the magnitude of their effects,

contributes to maximizing juice quality and producing fruit of consistent quality by learning

how to manipulate those factors affecting juice quality. Knowing the amount of variability

in juice quality is also a prerequisite for juice quality research, and plays a pivotal role in

experimental design. The concept of estimating and partitioning variance into its component

sources of variation (Hanson, 1955) has extensive application to horticultural work in the field

of sampling and sample size determination (Sharpe and van Middelem, 1955).

Several citrus sample size studies have been conducted to determine the variability in

juice SSC among individual fruit (Appleman and Richards, 1939; Bartholomew and Sinclair,

1943; Denny, 1922; Reitz and Sites, 1948; Sites and Reitz, 1949; Wallace et al., 1955). The

results were commonly used to enable representative fruit sampling of commercial lots of

29








30

citrus for remuneration purposes. Invariably these experiments used only one set of fruit

samples to estimate variability in juice quality, whereas Reitz and Sites (1948) demonstrated

how juice quality varied within a single tree. The outcome of these studies resulted in sample

size estimates varying from 18 to 36 fruit per sample. However, the number of replications

required to detect differences between treatment means, and the relative contributions of the

factors affecting juice quality to variability in juice quality have not been reported.

As part of a larger study of macroclimate and canopy microclimate effects on juice

quality in 'Valencia' sweet orange [Citrus sinensis (L.) Osb.], a study was conducted to

account for within- and between-tree variation in juice quality. The current study examines

how juice quality varies according to some of the factors known to make a contribution to

total variation in juice quality, including geographic locations, orchards within locations,

between and within trees, and between fruit. The objectives of this study were 1) to quantify

these sources of variation in juice quality of'Valencia' sweet orange, and to determine their

relative contributions to variability in juice quality; and 2) to estimate sample sizes.


Materials and Methods


Sites and Plant Material

Commercial orchards of 'Valencia' sweet orange trees on Carrizo citrange [Poncirus

trifoliata (L.) Raf. x C. sinensis (L.) Osb.] rootstock were selected at four geographic

locations after a preliminary climatic analysis of the major citrus-producing regions in Florida:

Howey-in-the-Hills (northern region; 2844'N, 81 46'W; elev. 23 m; Astatula sand soil [a

hyperthermic, uncoated Typic Quartipsamment of the Entisol order]; trees planted 1987),








31

Sebring (central region; 2730'N, 81 26VW; elev. 44 m; Astatula sand soil; trees planted 1983),

Immokalee (southwest coast; 2628'N, 81 25W; elev. 11 m; Oldsmar fine sand soil [a sandy,

siliceous, hyperthermic Alfic Arenic Haploquod of the Spodosol order]; trees planted 1986),

and Ft. Pierce (east coast; 27032N, 80034W; elev. 7 m; Winder sand soil [a fine-loamy,

siliceous hyperthermic Typic Glossaqualf of the Alfisol order]; trees planted 1987). The

Howey and Sebring sites were in upland ridge landscape positions and the orchards were

unbedded, the Immokalee site was in a flatwoods landscape position, and the Ft. Pierce site

was in a marsh landscape position. The orchards in the latter two sites were bedded in two-

and three-row beds, respectively. The trees were planted in north-south rows with tree

densities at the four sites of -380 trees/ha. The trees in each site were irrigated with

microsprinklers, and fertilized annually with -225 kg N/ha. Trees were not topped, but

alternate sides of trees were hedged annually to maintain a 2.4 m-wide drive middle. Trees

were otherwise cared for according to local cultural practices.

To determine within-location variation in juice quality, three blocks were selected within

the large commercial orchard at each geographic location. Within each block, research plots

consisting of 120 trees (10 rows x 12 trees per row) with buffer rows and trees were used.

The trees within each geographic location were healthy, and of similar age, size, and crop load.


Experimental Design and Data Collection

There were two experiments in this study. Experiment 1 was designed to estimate

within-tree variation in juice quality and was conducted over two seasons, and Experiment 2

which was conducted over three seasons to estimate between-tree variation in juice quality.








32

Experiment 1. A 4x2 factorial design (four geographic locations and two canopy

positions), established as a crossed-nested design, was used. The sources of variation were

geographic location, tree nested within location, canopy position, and the interactions,

location x position, and position x tree nested within location. The two canopy positions,

southwest top (SWT) and northeast bottom (NEB), were selected to provide a contrast in

juice quality (Reitz and Sites, 1948). Each canopy sector represented about one-eighth of a

tree's canopy. Five randomly selected trees from one of the three blocks in each of the four

geographic locations were used.

When juice reached a 13:1 ratio (9 Mar. 1998 and 8 Mar. 1999), six randomly selected

fruit, regardless of size, were harvested from each of the SWT and NEB tree canopy

positions. Juice was extracted from individual fruit using a citrus reamer (Sunkist Inc., Los

Angeles). Brix of juice samples was measured using a hand-held temperature-compensated

Brix refractometer (Atago Co., Japan), and TA was determined by titration with 0.3125 N

NaOH and 0.5% phenolphthalein solution. The acid correction factor was added to Brix

measurements to determine SSC (Fellers, 1990).

Experiment 2. A nested design was used with four geographic locations and three

blocks. The sources of variation were geographic location and block nested within location.

In Mar. 1998, ten randomly selected trees in each of the three blocks at the four geographic

locations were used, whereas five different trees were used in Mar. 1999 and Mar. 2000.

When juice reached a 13:1 ratio (9 Mar. 1998, 8 Mar. 1999, and 9 Mar. 2000), 50 fruit

were harvested from around each tree at a height of 1 to 2 m. Juice was extracted and

analyzed at the Citrus Research and Education Center state test house facility, Lake Alfred, Fla.










Statistical Analysis

Experiment 1. Juice quality variables were analyzed using PROC GLM and PROC

MIXED (Littell et al., 1996; SAS Institute Inc., 1996). Location and canopy position were

fixed effects, whereas tree nested within location was a random effect. From the analysis of

variance (ANOVA), the significance of treatment effects was determined, and means were

separated by least significant difference (LSD). To partition total variation into all the possible

component sources of variation, geographic location and canopy position were treated as

random effects, and variance components for all effects were estimated using PROC MIXED.

Experiment 2. Juice quality variables were analyzed using PROC NESTED (SAS

Institute Inc., 1996). Location was a fixed effect, and block nested within location was a

random effect. The significance of treatment effects was determined from the ANOVA, and

means were separated by least significant difference (LSD). To estimate variances of all effects

to partition them into their component sources of variation, geographic location was treated

as a random effect. Seasonal variation was also analyzed using season as the sub-plot in a

split-plot in time analysis.


Sample Size Estimation

A standard sample size equation (Steel and Torrie, 1980) was used to calculate sample

size estimates for the number of fruit per sample or trees replicationss) required to detect a

difference between two means,

n =2-2"S2
d2








34
where n is the number of fruit or trees, t is Student's t-value for the degrees of freedom

associated with S2 at P 0.05 (when a = 0.05 and df= 15 to 27, then t 2.1), S2 is the sample

variance for fruit or trees, and d is the desired degree of precision, or the difference to be

detected between treatment means.

The numbers of fruit per sample and replications required to detect differences between

two treatment means for each juice quality variable, at a desired degree of precision and at

P<0.05, were plotted to generate Figs. 3-1 and 3-2.


Results


Juice Quality

Experiment 1. The overall mean SSC of'Valencia' sweet orange juice at maturity in two

seasons was 12.7%. In 1998, juice SSC of fruit from Howey was significantly lower than the

other three locations at P<0.06 (Table 3-1), but there was no significant difference in SSC

among locations in 1999. Among locations, mean SSC ranged from 12.0% to 13.2% in Mar.

1998, but the range was narrower in Mar. 1999 when mean SSC ranged from 12.4% to

12.8%. Fruit from the SWT canopy position had significantly higher SSC than the NEB

canopy position in both seasons, the difference being 0.7% and 0.6% SSC in 1998 and 1999,

respectively. However, the effect of canopy position on SSC was not consistent at all

locations in 1998 (significant interaction). Fruit from the SWT canopy position had

significantly higher SSC than the NEB canopy position at Sebring and Howey, but there were

no significant differences in SSC between canopy position at Ft. Pierce and Immokalee.












5 0 0
-0o
00o
0o0


-o
0
CA



0




0

4)
0
















4 .




0
.co




41







o U












4)


o|





.2
-f

-o
.0


000


ON
o\ c


0

S0
0 2
'0 0 x
e & e

j x- o o s?
.Ml." i1 *PP
04 .4 0
4)E 0~
4 0 4) E *4


01


o t- en
oo o\ o ao



00ONO00


z
0 0 4 0
oooo s


oor- too oo

0000



00-0

z











-
oo oo Cs



c -i -4 C- -
C14enC14 cl
CIO-


vI


C C





04
.
Q 9











(D
4.















0 ff
z "





* 4-
A g









Fsg
O










.2





$4
^





















-00
4-
S .




















u
0 0


,
'S A


Na
s a




'^ II
iC .
U



gso o
Sg a

. &



- fl 4)
4)'- *
4. 2 '




0 I- C

CQ ci
sI-


oo00
00 o
000








36

Mean TA in Mar. 1998 was 0.84%, and in Mar. 1999 mean TA was 0.91% (Table 3-1).

In 1998, there were no significant differences in TA among locations nor between canopy

positions. The range of mean TA among locations was 0.16%. However, in 1999, TA was

significantly higher in fruit from Immokalee and Howey than Ft. Pierce, whereas TA from

Sebring was lower than that from Immokalee, but not significantly different from the other

locations, with a range of mean TA among locations of 0.15%. No trend in TA among

locations over the two seasons was evident. In 1999, fruit borne in the SWT canopy position

had significantly lower TA than NEB canopy positions by 0.1% TA. The lower SSC and

higher TA of fruit borne in the NEB canopy position in 1999 resulted in fruit of lower SSC-

to-TA ratio compared with fruit from the SWT canopy position.

Variation among fruit, represented by "error" variance, and variation among trees within

locations contributed the most to total variance of SSC and TA (Table 3-2). These two

sources of variation, together, accounted for >55% of total variance in juice SSC, and 72%

(1998) and 63% (1999) of total variance in TA. Variation between canopy positions

contributed an intermediate amount to total variation in SSC (13% in 1998 and 32% in 1999),

and variation among locations contributed the smallest amount to total variation in SSC

(<10%). For TA, variation among locations contributed an intermediate amount to total

variation (7% in 1998 and 18% in 1999), and variation between canopy positions contributed

the least to total variation (0% in 1998 and 10% in 1999). Total variance of juice SSC in

1999 was less than half that in 1998, and, for TA, total variance in 1999 was two-thirds that

of 1998.










Table 3-2. Partitioning of variance into component sources of variation as percentage of total
variance for soluble solids concentration (SSC) and titratable acidity (TA) of'Valencia' sweet
orange fruit harvested Mar. 1998 and Mar. 1999 to estimate within-tree variation in juice
quality.
Variance (% of total varianceY
SSC TA
Source of variation 1998 1999 1998 1999
Location 0.1032 (8.0) 0.0000 (0.0) 0.0021 (6.7) 0.0036 (17.7)
Tree(location) 0.4266 (32.9) 0.1149 (18.1) 0.0078 (25.2) 0.0037 (18.2)
Canopy position 0.1685 (13.0) 0.2037 (32.2) 0.0000 (0.0) 0.0021 (10.3)
Location x Position 0.2017 (15.5) 0.0000X (0.0) 0.0017 (5.5) 0.0005 (2.5)
Position x Tree(location) 0.0626 (4.8) 0.0760 (12.0) 0.0049 (15.8) 0.0012 (5.9)
Errors 0.3349 (25.8) 0.2387 (37.7) 0.0145 (46.8) 0.0092 (45.3)
TOTAL 1.2975 0.6333 0.0310 0.0203
z Total variance partitioned into component sources of variation as estimated by PROC
MIXED (Littell et al., 1996; SAS Institute Inc., 1996). Numbers in parenthesis represent
percentages of total measured variation.
Y Error or residual variation representing fruit-to-fruit variation.
xThe zero variance estimates are apparent statistical anomalies, and are due to the numerator
mean square (MS) being smaller than the denominator MS used in the F-ratio for the source
of variation being tested.








38
Experiment 2. Mean juice SSC of'Valencia' sweet oranges at maturity across the four

locations and three seasons was 13.1%. There was no significant difference in juice SSC

among the three seasons (P=0.2331), and the range in mean SSC among seasons across all

locations was 0.5% SSC (Table 3-3). However, location effects were inconsistent in the three

seasons (significant season x location interaction). In 1998, juice SSC of fruit samples from

Ft. Pierce, Immokalee, and Sebring were not significantly different, but they were significantly

higher than the SSC of fruit samples from Howey (Table 3-3), with a range in SSC among

locations of 1.1% SSC. In 1999 and 2000, there were no significant differences in juice SSC

among locations (Table 3-3), with the narrowest range in SSC among locations of 0.6% SSC

in 1999 and an intermediate range in 2000 of 0.8% SSC. There was no trend in SSC among

locations over the three seasons, there being no difference in SSC among locations in two of

three seasons (Table 3-3).

Mean juice TA at maturity across the four locations and three seasons was 0.96%, being

significantly lower (at P<0.075) in 1998 than 1999 and 2000 (Table 3-3). In all three seasons,

TA did not differ among locations. The range in TA within each season among locations was

0.07% TA in 1998, 0.05% TA in 1999, and 0.16% TA in 2000.

The primary source of variation in juice SSC and TA was tree-to-tree variation,

represented by "error" variance, which contributed -75% of total variance (Table 3-4).

Variation due to seasons was the next largest component of variation, contributing 14% to

21% to total measured variance. Variation among locations contributed only a small amount

to variance (1% to 8%), and block-to-block variation made a negligible contribution to total

measured variance (<5%).










Table 3-3. Mean juice quality of 'Valencia' sweet oranges harvested at maturity in 1998,
1999, and 2000 from four geographic locations in Florida (50-fruit samples analyzed at the
Citrus Research and Education Center state test house facility, Lake Alfred, Fla.).
Geographic location SSCz (%) TAy (%)
1998 1999 2000 1998 1999 2000
Ft. Pierce 13.4 aw 13.3 NS 13.0NS 0.91 NS 0.97 NS 0.93 NS
Howey 12.6 b 12.9 12.9 0.93 1.02 0.95
Immokalee 13.7 a 12.9 12.6 0.86 1.00 1.09
Sebring 13.7 a 12.7 13.4 0.88 0.98 1.01
Mean 13.4 12.9 13.0 0.89 0.99 0.99
P-valuex 0.0010 0.1371 0.4857 0.3627 0.9450 0.2145
SSoluble solids concentration.
Y Titratable acidity.
xProbability value.
w Within-column means with same letter are not significantly different; n=10 (1998), n=5
(1999 and 2000); Ns=nonsignificant; P<_0.05 (LSD).









40
Table 3-4. Partitioning of variance into component sources of variation as percentage of total
variance from a split-plot in time analysis for soluble solids concentration (SSC) and titratable
acidity (TA) of 'Valencia' sweet orange fruit harvested at maturity in 1998, 1999, and 2000
from four geographic locations in Florida to estimate between-tree variation in juice quality.
Source of variation SSC TA

Season 14.3 20.7
Location 8.5 0.9
Block 0.2 4.3
Errors 76.9 74.2
'Error or residual representing tree-to-tree variation.










Sample Size Estimation

The relationships between the number of fruit or replications and the difference to be

detected between treatment means (degree of precision) are represented by asymptotic lines

(Figs. 3-1 and 3-2). The difference to be detected, or degree of precision, becomes smaller

as the number of fruit increases, and the slope of the line decreases with increasing sample

size. With relatively small sample sizes, of less than 13 fruit per sample, relatively large

changes in precision occur for small increases in fruit number, but when sample sizes increases

above 20 fruit per sample, relatively small changes in precision occur with increasing sample

size (Fig. 3-1). Similarly for replications, relatively large changes in precision occur with

increasing number of trees until eight trees, but when the number of replications exceeds 10

trees, relatively small changes in precision occur (Fig. 3-2).

For both SSC and TA, variance was greater in 1998 than the other seasons, resulting in

the 1998 curve being further from the horizontal and vertical axes (Figs. 3-1 and 3-2).

Estimates of sample size from the 1998 curve will result in larger sample sizes than if the

curves from the other seasons are used. Juice TA had greater within- and between-tree

variation than juice SSC. The greater variation in TA than SSC was reflected in larger

coefficients of variation for TA than SSC (- 10% vs. =3%). As a result, it is more difficult

to demonstrate treatment differences in TA than SSC. Therefore, for the estimation of sample

size, a more liberal degree of precision for TA is required than for SSC. The alternative is

impractically large samples are required to estimate small differences in TA.

Using 1998 data, for example, 12 fruit per sample are required to detect a difference of

0.5% SSC between two means (Fig. 3-1), whereas 33 fruit per sample are required to detect










0.8 --------------
0.8 SSC -.- 1998 --a 1999

0.7 ----- -----------------------

0o 0.6 ----- -- ------------------- ---

S0.7 -------- -- ------ ------

& 0.3 ------------------- -
M0.5 -
0

a)043 -- -- --- -


0.2 -----------------------


1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
No. of fruit



0.12 TA -- 1998 -a- 1999



0.10 ------- --- ---------------



0.08 ------- -- ------------



0.06 ---------------------- ----------
0.04


1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
No. of fruit


Fig. 3-1. Number of fruit per sample required to estimate differences between two means
for soluble solids concentration (SSC) and titratable acidity (TA) for a given degree of
precision at P:0.05 using 1998 and 1999 variances for SSC and TA. Degree of precision
refers to the difference in % SSC or % TA to be detected between two means.










SSC -*- 1998 -a- 1999 -a- 2000
0.7 -------- -- ---------

S0.6 --------------------- -----

0 .5 -- -- --- - -
0
u 0.4 -- --------- -----

00.3 ------------ --- ----

0.2 -------- ---- ----------------

0.1 --- 2I I I
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
No. of trees


0.12
-- 1998 --a- 1999 -&- 2000


p 0.10 --- -- -------------
0o


0.08 ------ ----- ------ ---
o

0.06
~0.06 ---------------- ------ -




1 2 3 4 5 6 7 8 9 1011 12 13 14 15 16 17 18 1920
No. of trees


Fig. 3-2. Number of trees replicationsns) required to estimate differences between two
means for soluble solids concentration (SSC) and titratable acidity (TA) for a given degree
of precision at P0.05 using 1998, 1999, and 2000 variances for SSC and TA. Degree of
precision refers to the difference in % SSC or % TA to be detected between two means.








44
a difference of 0.3% SSC. The number of replications (trees) required to detect the same

differences are seven vs. 19 (Fig. 3-2). Similarly for TA, 13 fruit per sample are required to

detect a difference of 0.1% TA between two means using 1998 data (Fig. 3-1), whereas 35

fruit per sample are required to detect a difference of 0.06% TA. The number of replications

required to detect these differences in TA are seven and 18 trees, respectively (Fig. 3-2).

Many possible combinations of numbers of fruit per sample and replications exist, depending

on the degree of precision selected.


Discussion


Seasonal variation in SSC and TA was relatively small, being larger for TA than SSC as

observed by Harding and Sunday (1949) and Harding et al. (1959). Mean annual air

temperature did not differ substantially during the period of the study, resulting in little

seasonal variation in juice quality (see Chapter 4). However, seasonal variation in juice

quality can be significant (Fellers, 1985; Harding et al., 1940), highlighting the importance of

conducting field studies over more than one season. Collecting data over more than one

season provides data about another source of variation thereby allowing more precise

estimates of sample size.

Under comparable conditions of climate, rootstock, tree age, and row orientation,

geographic location had little effect on juice SSC and TA of 'Valencia' sweet orange over

three seasons. This apparently small effect of production region on juice quality is attributed

to the similarity in mean annual air temperatures among locations within the narrow

geographic range from which fruit were sampled (see Chapter 4) and the similarity in general








45
cultural practices of citrus trees grown in the different regions in Florida. The relatively small

differences in climatic conditions within Florida contrast with large differences in climatic

conditions among citrus-producing regions of the world in different climatic zones where

large differences in juice quality occur (Reuther and Rios-Castanio, 1969).

Variation in juice SSC among blocks within a location was intermediate to low, relative

to other sources of variation, and less than variation among locations. Since blocks with trees

of similar age, size, and crop load were selected for the study, low variation in juice SSC

among blocks can be expected.

Variation in juice SSC and TA among trees was large, in spite of sampling from trees of

similar vigor and crop load. Appleman and Richards (1939) also observed tree-to-tree

variation in juice quality of samples of uniformly sized fruit from trees carefully selected for

uniformity. The cause of this variation is unknown. Despite the use of clonally propagated

citrus trees and uniform cultural practices applied across all trees within an orchard, natural

heterogeneity in site conditions may be sufficient to result in subtle differences in tree

behavior, even under well-managed conditions. Nevertheless, the inherent tree-to-tree

variation observed in this study highlights the need to sample fruit from sufficient numbers of

trees to account for tree-to-tree variation.

Canopy position had a large effect on juice SSC, and fruit borne in upper, sun-exposed

canopy positions had higher SSC and lower TA than fruit borne in lower, semi-shaded

positions. This effect was attributed to differences in incident radiation in the different tree

canopy positions (Reitz and Sites, 1948), although this has not been quantified. Data from

the current study reinforce the importance of canopy microclimate as a source of within-tree








46
variation in juice quality. However, in addition to the affect of canopy microclimate on juice

quality, the position where fruit are borne in citrus trees also has a meaningful contribution

to variability in SSC. The cause of this variability may be due to fruit-to-fruit variation within

canopy positions. Therefore, to take a representative sample from a tree and to avoid

canopy-related bias, fruit should be sampled from all canopy positions. In contrast to SSC,

canopy position effects on TA were less important.

Variation in juice SSC and TA among fruit was relatively large. Similar observations

were made by Denny (1922), Reitz and Sites (1948), and Wallace et al. (1955). Inherent

fruit-to-fruit variation in juice quality may be due to variation in fruit size, differences in leaf-

to-fruit ratio, differences in exposure of bearing shoots to solar radiation, or other unknown

factors, and has implications in ensuring that fruit samples consist of adequate numbers of

fruit to be representative of the unit being sampled.

Results from previous sample size studies (Appleman and Richards, 1939; Wallace et al.,

1955) showed that =30 fruit were required to detect differences of 0.5% SSC, whereas Sites

and Reitz (1949) proposed that 20 fruit of uniform size sampled from the four cardinal points

from the outside of a tree's canopy at 1 m height were required to detect a difference of

0.5% SSC. From the current study, a difference of 0.5% SSC could be detected by sampling

as few as 12 fruit. Considering the relatively large contributions of trees and fruit as sources

of variation in juice quality of 'Valencia' sweet oranges, samples consisting of 35 fruit are

required to detect differences between means of 0.3% SSC and 0.06% TA at P<0.05,

whereas 20-fruit samples can be used to detect differences of 0.4% SSC and 0.08% TA. The

sample size selected depends on the differences to be detected for both SSC and TA, as well








47

as practical factors such as the level of detection of the instrument used to measure the

variable of interest, cost of sampling and transportation, and time taken to conduct

measurements.

Figure 3-2 can be used as a guideline to select the number of replications required, with

small gains in precision when tree numbers exceed 10. The value of Fig. 3-2 lies in the user's

ability to either specify a preferred degree of precision and then to determine the number of

replications required, or to select a practical number of replications and then to assess whether

the corresponding degree of precision is acceptable to the researcher's needs.

When too few replications are used in an experiment, true differences among treatments

may not be detected. In such cases, type II errors are made, i.e., fail to reject the null

hypothesis when it is false, and conclude that there are no differences among treatments when

true differences do exist. For example, previous studies on rootstock effects on juice quality

were unable to detect differences in SSC of 1.0% and TA of -0.1% among fruit samples

from trees on different rootstocks (Castle and Phillips, 1980; Castle et al., 2000). In these

examples too few replications were used, n=3 and n=5, respectively.














CHAPTER 4
VARIATION IN JUICE QUALITY OF 'VALENCIA' SWEET ORANGE
DUE TO MACROCLIMATE AND CANOPY MICROCLIMATE



Macroclimate has a large effect on juice quality of Citrus spp., and is the principal factor

determining where citrus cultivars are commercially produced (Gat et al., 1997; Hodgson,

1967; Reuther, 1973; Webber, 1948). The major climatic factors affecting the development

of citrus fruit are air temperature and soil water (Reuther, 1973). Air temperature has an

overriding effect on fruit growth, development, and maturation.

Reuther and Rios-Castafto (1969) compared the contrasting climatic conditions of

subtropical California with tropical Colombia, and demonstrated the effects of climate on fruit

maturation. 'Valencia' sweet orange [C. sinensis (L.) Osb.] attained a soluble solids

concentration (SSC)-to-titratable acidity (TA) ratio of 9:1 within 7 months under lowland,

tropical conditions, half that of cool, subtropical conditions. Comparisons of the effects of

climate on juice quality among citrus-producing regions within the United States (Cooper et

al., 1963) and within California's citrus-producing regions (Nauer et al., 1974; Webber, 1948)

showed that the principal difference among regions was the time required to attain a given

maturity index.

The concept of heat unit accumulation, or degree-days, has been used to compare climate

among citrus-producing regions (Gat et al., 1997; Reuther, 1973), and to determine where

different cultivars can be commercially produced (Barry et al., 1996; Ben Mechlia and Carroll,

48








49

1989b). Cumulative degree-days have also been used to predict maturity with varying success

(Kimball, 1984; Lomas et al., 1970; Newman et al., 1967). Using 13 C as a base

temperature, 'Valencia' sweet orange fruit achieve market acceptance when produced in

regions that have 1300 to 3500 annual cumulative degree-days (DD). This is a relatively

broad range compared with other cultivars, e.g., 'Clementine' mandarin (C. reticulata

Blanco), which requires 1600 to 2200 DD (Barry et al., 1996).

Juice quality of citrus is also influenced by fruit position within a citrus tree's canopy, and

associated microclimate (Morales et al., 2000; Reitz and Sites, 1948; Syvertsen and Albrigo,

1980). Reitz and Sites (1948) measured juice quality of individual 'Valencia' sweet orange

fruit from a tree on rough lemon rootstock. Juice SSC of individual fruit varied from 5.9%

to 13.5% SSC. In the northern hemisphere, 'Valencia' sweet oranges from southern-top

canopy sectors, where the greatest net radiation occurs, tend to have higher juice SSC and

juice content than fruit from other canopy sectors (Sites and Reitz, 1949; 1950a; 1950b).

Differences in juice quality may be associated with canopy microclimate and exposure of

individual fruit to environmental stresses (Syvertsen and Albrigo, 1980). Reitz and Sites

(1948) attributed canopy position differences in juice quality to "light classes", implicating

photosynthetically active radiation (PAR) as an important factor affecting within-tree variation

in juice quality, although this effect has not been quantified. However, not all of this effect

is due to a presumed enhancement of photosynthetic effect on sugar accumulation (Syvertsen

and Albrigo, 1980). Some of the effect is due to differences in water stress resulting in less

dilution of soluble solids in canopy areas with higher radiation exposure (Syvertsen and








50

Albrigo, 1980), while other factors probably also have an effect, such as fruit size and crop

load (Albrigo, 1992).

Florida's citrus-producing regions fall into the subtropical climatic zone of the south-

eastern United States (Jackson and Davies, 1999). Historical meteorological data ofFlorida's

citrus-producing regions were reviewed in a preliminary study to identify geographic locations

in Florida with dissimilar climatic conditions (National Oceanic and Atmospheric

Administration, 1996). The analysis of long-term mean monthly air temperatures for different

locations in the Florida citrus-production region revealed relatively small diurnal and seasonal

differences in air temperature among locations, with the largest differences occurring during

the winter. Due to these relatively small differences, compared with other citrus-producing

regions of the world, e.g., California and South Africa, it was unknown whether there were

sufficient differences in air temperature among Florida's citrus-producing regions to cause

differences in juice quality among locations.

Little is known about the direct relationships between air temperature and juice quality

among citrus-producing regions in Florida, and among fruit from different canopy positions

within a citrus tree. Also, the relationship between cumulative degree-days among citrus-

producing regions in Florida and juice quality have not been quantified.

The objectives of this research were 1) to determine and quantify possible macroclimate

and canopy microclimate effects on fruit growth and juice quality of'Valencia' sweet orange

in the major citrus-producing regions of Florida; and 2) to relate juice quality to temperature

and cumulative degree-days.










Materials and Methods


Sites and Plant Material

This experiment was conducted in the four geographic locations in Florida with three

blocks per location used in the study to estimate sample size (see Chapter 3 for detailed

description of Sites and Plant Material). Briefly, 'Valencia' sweet orange trees on Carrizo

citrange [Poncirus trifoliata (L.) Raf. x C. sinensis (L.) Osb.] rootstock planted between

1983 and 1987 were used.


Experimental Design and Data Collection

The study was conducted during the 1998-99 and 1999-2000 seasons. A 4x2 factorial

design (four geographic locations and two canopy positions), established as a crossed-nested

design, was used, where the sources of variation were geographic location, block nested

within location, and canopy position, and the interaction effect, location x position.

Southwest top (SWT; upper, exposed) and northeast bottom (NEB; lower, partially-shaded)

canopy positions were selected, representing two contrasting canopy microclimates (Reitz and

Sites, 1948). Each canopy sector represented approximately one-eighth of a tree's canopy.

Fruit samples were taken monthly from the end of stage I/beginning of stage II (mid-

June) through stage III (March) of fruit development. During each month of sampling, five

randomly selected trees from each of the three blocks in the four geographic locations were

used, from which 10 fruit were harvested from each of the SWT and NEB tree canopy

positions. Fruit were randomly sampled within a canopy position, regardless of fruit size.








52

Different trees were used at each sampling time in the 10 x 12-tree research plot as the quantity

of fruit available in each canopy position was limited.

Fruit equatorial diameter and fruit fresh weight (FW) were measured. To extract juice

from fruit <55 mm in diameter, fruit were hand-peeled, homogenized, and then strained

through a nylon cloth to separate juice from pulp. For fruit >55 mm in diameter, juice was

extracted using a citrus reamer (Sunkist Inc., Los Angeles), and juice weight was measured

and juice content (w/w) calculated. Brix of juice samples was measured using a hand-held

temperature-compensated Brix refractometer (Atago Co., Japan), and TA was determined

by titration with 0.3125 N NaOH and 0.5% phenolphthalein solution. The acid correction

factor was added to Brix measurements to determine SSC (Fellers, 1990), and the SSC-to-TA

ratio was determined.

Air temperature was recorded hourly using remote dataloggers (Spectrum Technologies,

Plainfield, Ill.) from 1 Apr. 1998 through 31 Mar. 2000. A datalogger was mounted in each

of the two canopy positions of one tree within each of the three blocks and four geographic

locations, i.e., a total of 24 dataloggers. These data were summarized into daily, monthly,

and annual intervals for further analysis. Degree-days (DD) were calculated from daily mean

air temperatures using 13 C as a base temperature (Newman et al., 1967) according to the

equation, DD = (mean monthly air temperature 13) x days in month.

The DD calculations were based on hourly air temperatures. These calculations tend to

under-estimate cumulative DD (by -10%) when DD calculations are derived from mean

monthly maximum and minimum air temperature data.










Statistical Analysis

Juice quality variables were analyzed using PROC GLM and PROC MIXED (Littell et

al., 1996; SAS Institute Inc., 1996). Location and canopy position were fixed effects, and

tree nested within location was a random effect. The significance of treatment effects at each

sample date was determined by analysis of variance (ANOVA) and means were separated by

least significant difference (LSD). Repeated measures analysis was used to analyze seasonal

trends in juice quality development, using monthly juice quality data.

Mean monthly air temperature and DD data were analyzed by repeated measures

analysis. Juice quality variables were then regressed against climatic variables to determine

any relationships between the two groups of variables.


Results


Fruit Size and Juice Quality in Relation to Macroclimate and Canopy Microclimate

During early stage II of fruit development (June through September), SSC decreased

before increasing. Titratable acidity increased until July, before decreasing through maturity,

but it decreased more gradually after December (Figs. 4-1 through 4-4).

Season. Across all treatments, fruit diameter was marginally smaller in Mar. 1999, with

lower SSC and TA, and as a result, higher ratio than in Mar. 2000 (Table 4-1).

Geographic location. At maturity in Mar. 1999, there were no differences in fruit

diameter and juice quality among geographic locations, although fruit from Ft. Pierce had the

highest SSC compared with the three other locations (Table 4-1 and Fig. 4-1). However,

there were some seasonal differences in juice quality among locations when analyzed at each



































uouoo sP!lOS olqnlo\ s
(%) uoiio) spqos aiqnups


1 1






4'







:I.
0








-8







C0







0





4'

8


z I I I




I
I






I I
oI I I I
z I I I I I
I I I I I
I I I I I I







I I I 0 0 I
I I I "I


I II
I II Ill

11I1I II
I I IIl II

l I l I I
z




I I 1 1 1 I

I I I I I
I I I I l lI
I II I l
I I I I l
I I I i l l


I I I I



I I I
zI I

I
z
t 0 00. '


t-- o 10 .-
(mnu) ijolvaerp 1iuj4


o 0






09





o o0





ZE
0 01.-*









a o

















S o0
$ -; J2,
.tg 0






































()t -n TO4 0 O N
-4 1 1- -
(o/0) uouoo spqjos aiqnloS


0 0 0
00


e'i 0CC ~O
- -
OTW)J


C6o t
0





S-.



v 7


0
1uin


I I I
I I I I
J~7III I I
fflt~iIj I I I I
I I I
vx& I I
I I I
I I F I

0 a~ 0 *r~ 0 'f~
~


55





0 .




(30
oS 0



0 I- a























0..
5 o



03









o 0
o













Cd,
. i

io
me0


I I



/ I


! !








































- c-


I I
I I
I I I I I
I I
I I I
I I I
I I I I
I I I I
I I I
I I I
I I I I
I I I
I I I I
I I I
I I I
I I I I -
I I I I I
I I I I
I I I I
I I I
I I I
I I I I
I I I
I I I I
I I I I
I I I -
I I I
I I
I I
I I
I I I I I


o i
_


(%) uouoo spilos olqnlos


I I *ao I Il
(wuz) im~uwi~p 1!fl


I I I I I
I I I I I
I I I
I I I
I I I I I
I I I
I I I I I
I I I I I
I I I I
I I I I I
I I I
I I I I
I I I I I
I I I
I I I I I
I I I I
I I I I I
I I I I I
I I I I
I I I I I
I I I I I
I I I I I I -
I I I I I I
I I I I I I
I I I I I
I I I I -
I I I I
I I I I
I I I I I
I I I
I I I
I I I I I
I I I I I 2
I I I I I I


I i I I I I I
> T 0 00 10 I T C


1 1 1 I




I l l I








I l l I
I I I I
I I I I I
I I I I I I
I I I I I I


I I I I I
I I 1 I I
1 II I I I
I I o


F
It-


2o


I.~4

Ih




-o







CD


0 *


a' a.

Ib'
o











04-*. s
I 4-


0







4) 0b
O









o






00





.6 .o ii
4-g g aS
)g
>

-, E i '























r:II
0.
.0 Go
















00










01
02 b ...


z

0

to
4)
-8s


0&
0


*iD




































m N 0 0\ 00
(%) uouoo splos oiqnlos


(urn) j.iaurmip liui


I I I
I I









11
I I I

















Ilill
Ilill
lilll
I I


















I II
I I I I
I I I



I I I I
I I I
I Il
I l l
I l l II
I I I


S I I I
I I I I







I I I I

* I I I I


Il l I
I I I I










I I I I
I I I Iq
* I I I I I
I I I I I

I I I I
I I I I I
I I I
I I I I
I I I I

I I

I I I I


F-'




(#


0
o 0 ,








4-.
>n







0
*s .t












_o

0.0

B O"0













0 0



W 1











o

.0
x '
S0


I-
as




0


0 0







o 0














> o



I-
.0 0
- .S


z
00 00








----



z
ooo 0
----


m
z
e-0 m oo
*^ M 00
t~r-~ I --r~


4))
4) 4W
-a > 0
t4 op a a
( D > >o
0
.^ o g-0 o
fc a oi0
u


z
- 7s (7s
2--


o
00 0
o o o


-- o



-- o






S0





r-o 0
o\ o











0
0 0-\ 0






ooi o
00e~ 0





0o
-- 0







o
- < o0
-- 6



0

o
r1r- 6~


0
U

0

W o
-.9 .
so
.0
0 )
4


0
Wo

cs 9
- l
SI I


.2

0
x


0 0


ri~








59

sample date. During early fruit development (June through September) in the 1998-99

season, fruit from Sebring and Immokalee were significantly smaller than fruit from Howey

and Ft. Pierce, and from August to December, fruit from Sebring had significantly higher SSC

(Fig. 4-1). Acidity was significantly higher for fruit from Sebring and Immokalee than Howey

and Ft. Pierce at the July and August sampling dates. However, this effect was reversed by

January, implying that the rate of acid degradation was greater in fruit from Sebring and

Immokalee than Howey and Ft. Pierce. There were no differences in ratio among locations.

At maturity in Mar. 2000, fruit from Sebring were significantly smaller than fruit from

Immokalee and Howey, and fruit diameter at Ft. Pierce was intermediate (Table 4-1 and Fig. 4-2).

There were no significant differences in SSC among locations at maturity, although fruit from

Immokalee (largest fruit) had the lowest SSC, and fruit from Sebring (smallest fruit) had the

highest SSC. Fruit from Immokalee had significantly higher TA, resulting in significantly lower

ratio, than the other locations. During fruit development in the 1999-2000 season, significant

differences in fruit diameter through fruit development were due to larger fruit diameter for fruit

from Immokalee and Howey, and smaller fruit from Sebring, whereas fruit diameter at Ft. Pierce

was intermediate (Fig. 4-2). Significant differences in SSC among locations were evident from

November through February, when fruit from Sebring and Howey had significantly higher SSC

than fruit from Ft. Pierce and Immokalee (Fig. 4-2). There were no differences in TA and ratio

among locations through the season until the final sampling at maturity in March.

Canopy position. Differences in fruit diameter and juice quality between canopy

positions appeared early in fruit development, being apparent from the August sampling date

through maturity (Figs. 4-3 and 4-4). In Mar. 1999, fruit from the SWT canopy position








60
were significantly smaller than fruit from the NEB canopy position (by = 1 mm), but in Mar.

2000 there were no differences in fruit diameter between canopy positions (Table 4-1, and

Figs. 4-3 and 4-4). However, the relationship between fruit diameter and canopy position was

not consistent at all locations (significant location x position interaction; Table 4-1). In Mar.

1999 and in Mar. 2000, fruit diameter was larger in the SWT than NEB canopy position at

Howey and Immokalee, respectively (data not shown).

In both seasons, fruit borne in the SWT canopy position had significantly higher SSC (by

0.6% SSC in 1999 and 0.8% SSC in 2000), lower TA (by 0.1% TA in both seasons), and

higher ratio (by 2.2 points in 1999 and 2 points in 2000) than fruit borne in the NEB position

(Table 4-1, and Figs. 4-3 and 4-4). However, in Mar. 2000, the relationship between juice

quality variables and canopy position was not consistent at all locations (significant location

x position interaction; Table 4-1). There was no difference in SSC between canopy positions

at Ft. Pierce, there was no difference in TA between canopy positions at Howey, and the

difference in ratio between canopy positions at Sebring and Ft. Pierce was of greater

magnitude than at Howey (data not shown).


Air Temperature and Cumulative Degree-Days

Season. The 100-year mean annual (Jan. to Dec.) air temperature for Florida is 21.4 C

(Southeast Regional Climate Center, 2000), resulting in 3066 annual cumulative degree-days

(DD). These data are based on standard meteorological station measurements, averaged

across the state of Florida. During the two seasons of the experiment, mean annual air

temperature was higher than the long-term mean by 1.2 C and 0.4 C in the 1998-99 and

1999-2000 seasons, respectively. The higher mean annual air temperature resulted in 3504








61
and 3212 DD in the two seasons, 12% and 5% higher annual cumulative degree-days than

the long-term mean for Florida. The 1998-99 season was one of only three seasons in the

past 100 years with mean annual air temperature above 22.5 C.

Mean annual (Apr. to Mar.) air temperature, at the four locations used in this study and

based on hourly measurements within a tree's canopy, was 22.3 C and 21.5 'C in the 1998-

99 and 1999-2000 seasons, respectively (Table 4-2). During these periods, annual cumulative

degree-days were 3408 and 3124 DD, a seasonal difference of =9%.

Geographic location. There were significant differences in mean maximum air temperature

(T.) among locations in the two seasons (Table 4-2 and Fig. 4-5). In both seasons, T. was

highest for Immokalee and lowest for Howey. The two locations differed by 1.2 C and 1.0 C

in the 1998-99 and 1999-2000 seasons, respectively, with T. for Ft. Pierce and Sebring being

intermediate. During spring and early autumn, T. did not differ among locations. During

summer 1999, T. at Howey tended to be higher than the other locations, and during late autumn

and winter, Tm. was 2 to 3 C higher at Immokalee compared with Howey (Fig. 4-5). There

were no significant differences among locations in mean minimum air temperature (T, .), mean

monthly air temperature (Tm.n), and, therefore, monthly cumulative DD.

Canopy position. Mean T. was significantly higher in the SWT than the NEB canopy

position throughout both seasons (Table 4-2 and Fig. 4-6), and at all locations. The

difference in mean T. between canopy positions was 1.6 C and 1.3 C in the 1998-99 and

1999-2000 seasons, respectively. The SWT canopy position had significantly lower mean Tr,,

than the NEB canopy position by 0.2 oC, but this canopy position effect was significant at

only Immokalee and Sebring, and at Ft. Pierce and Howey there were no differences in T,











4 OC

00
0

410)

O0~
a ,



~00
4 a -



> 02


o -



x
C d
&. o












0














Ed
0 d
.- 3 -8















ao 0.
4 -
0i












0
*oo
















1 .0
.u *



NE a

a s







U,*
CO^
c 82


o
4r.
0

00
04-4
x

4 4
0
S5
*"-4


^Ira
e!fe*


-I,.


0



o


"T 0



ow 0






0







NM o



ao o 9
erc 0f


0
S




m






t
i-
I-



.0


o



.4- 4)
















.
a S
e M
ll






-
4 II





ias

*- o.*











Io-' i 0
-. 3 0


e.1.
c^ C^ C)^







N555
cololl
g S *5
i^ wIc.


U)
z


"Cr
Ci Cl Cl Cl


U)
z






C% o- 0





z

U)
Z
\> >6 ^ t-~
---- \ ^


fc s co












430 -i--------------------------i
40
A
35 --- ----------------------- ---------------






15 -- -
20 --- ------- ------ -- ------ -------

15 ------------ --------- --------- ---- -



-B
10


Apr98 Jun98 Aug98 Oct98 Dec98 Feb99 Apr99 Jun99 Aug99 Oct99 Dec99 Feb00
Ft. Pierce Howey Immokalee -- Sebring

500 -


400 ---- ----- ------------------ --------- --


g 300 -- ------(-- --------- ---- --- ----- -


o 200 -- -----


100 -- ------- ----- -- ----


0 I I I I I I I I I I I I I I I I I I I I I I I
Apr98 Jun98 Aug98 Oct98 Dec98 Feb99 Apr99 Jun99 Aug99 Oct99 Dec99 Feb00
--- Ft. Pierce -- Howey Immokalee v Sebring



Fig. 4-5. (A) Mean monthly maximum (upper), mean (middle), and minimum (lower) air
temperature profiles, and (B) mean monthly cumulative degree-day profile for two seasons
(Apr. 1998 to Mar. 1999, and Apr. 1999 to Mar. 2000) of four citrus-producing regions in
Florida averaged across two canopy positions (n=3 blocks, 2 canopy positions).

















25 --- --------- --- ------- ------ -----

t2-- --- -

10
15 ------ -------- -------- ------------- ---- --

10 -- ----- ---- ---- ------ --------


Apr98 Jun98 Aug98 Oct98 Dec98 Feb99 Apr99 Jun99 Aug99 Oct99 Dec99 Feb00
-- NEB --- SWT

500 -
spring I summer I autumn I winter I spring I summer I autumn I winter


400 -- ------ --------- ---- --- ---- -----


300-- ------------------- --------


0 200 --------- -------- ------- -------- ------- -------- Q
200 -------------------- ------------ ---- ------






Apr98 Jun98 Aug98 Oct98 Dec98 Feb99 Apr99 Jun99 Aug99 Oct99 Dec99 Feb00
NEB -- SWT


Fig. 4-6. (A) Mean monthly maximum (upper), mean (middle), and minimum (lower) air
temperature profiles, and (B) mean monthly cumulative degree-day profile for two seasons
(Apr. 1998 to Mar. 1999, and Apr. 1999 to Mar. 2000) of two canopy positions averaged
across four citrus-producing regions in Florida (n=3 blocks, 4 locations).








65

between canopy positions. The SWT canopy position had a significantly higher T_, but this

difference between canopy positions, which occurred during autumn and winter (Table 4-2

and Fig. 4-6), was small (0.2 C). In addition, this difference in T___ between canopy

positions was only significant at Howey. Canopy position effect on annual cumulative DD

was the same as for T., since DD is directly related to T..


Relationship Between Juice Quality and Air Temperature

The relationship between SSC and annual cumulative DD within a citrus-growing season

among geographic locations was weak and inconsistent (r2=0.24 to 0.41; Table 4-3). Whereas,

TA was closely related to annual cumulative DD in 1999 (r~0.94), but not in 2000 (r?0.00).

Additional analysis revealed that the low r2 in 2000 was due to the relatively high TA at

Immokalee, and the lack of canopy position effect on TA at Howey, the reasons for which

are unknown. When these two factors were removed from the regression model, r?=0.93.

Soluble solids concentration and ratio were positively correlated with annual cumulative DD,

and TA was negatively correlated with annual cumulative DD (1999 only). Coefficients of

determination for analyses of possible relationships between juice quality variables (SSC, TA,

and ratio) and cumulative degree-days (based on annual, monthly, and various phenological

periods) were highest when annual cumulative degree-days was the dependent variable, but

only for fruit harvested in Mar. 1999. All other relationships were nonsignificant.

When comparing juice quality between citrus-growing seasons, SSCs were higher in the

season with lower annual cumulative DD (1999-2000 season). No relationship between SSC

and annual cumulative DD for the two seasons and four locations was apparent.










Table 4-3. Coefficients of determination for linear relationships between juice quality
variables of 'Valencia' sweet oranges harvested in Mar. 1999 and Mar. 2000 from two
canopy positions and four geographic locations, and cumulative degree-days (DD), using
hourly air temperature measurements and 13 C base temperature, summed over various
ohenological periods.


Linear relationship
Independent variable Z Dependent variable y 1999 2000

SSC Apr. to Mar. DD 0.41 0.24 NS
TA Apr. to Mar. DD 0.94 0.00 NS
Ratio Apr. to Mar. DD 0.81 0.06 NS
SSC Jun. to Aug. DD 0.04 NS 0.02 NS
TA Jun. to Aug. DD 0.00 NS 0.17 NS
Ratio Jun. to Aug. DD 0.00 NS 0.10 NS
SSC Nov. to Mar. DD 0.24 NS 0.05 NS
TA Nov. to Mar. DD 0.36 NS 0.03 NS
Ratio Nov. to Mar. DD 0.32 NS 0.00 NS


SSoluble solids concentration (SSC), titratable acidity (TA), and SSC-to-TA ratio.
Y Cumulative degree-days (DD) summed by growing season of'Valencia' sweet orange used
in this study (1 Apr. through 31 Mar.), summer coinciding with an increase in SSC (1 June
through 31 Aug.), and maturation period (1 Nov. through 31 Mar.).
NS, Nonsignificant or significant at P0.05, respectively.










Discussion


Seasonal changes in SSC and TA are due to a combination of metabolism and dilution,

through increased fruit size and juice volume, of soluble juice components (Goldschmidt and

Koch, 1996). Few studies have addressed early developmental changes in juice quality, but

Richardson et al. (1997) demonstrated similar seasonal changes in SSC and TA during stages

II and III of fruit development. The initial decrease in SSC, during early stage II of fruit

development (June through September), occurred in spite of sugar accumulation taking place

during this period (Garcia-Luis et al., 1991; Lowell et al., 1989; Richardson et al., 1997).

This apparent dilution of sugars, and resultant decrease in SSC, is attributed to a greater

water influx rate than sugar accumulation (Huang et al., 1992) during this stage of rapid fruit

expansion (Bain, 1958). However, during the second half of stage II of fruit development

(October through December), when fruit growth rate is relatively slow (Bain, 1958), the rapid

increase in SSC is a result of a shift in balance towards greater accumulation of soluble solids

in the juice of fruit (Richardson et al., 1997), and less dilution due to reduced water influx

(Huang et al., 1992). During stage III, SSC continued to increase at a similar rate to that of

stage II, even though sugars accumulate at a reduced rate (Richardson et al, 1997). These

changes are likely due to decreased water influx (Huang et al., 1992) and associated

decreased fruit growth rate (Bain, 1958). The initial increase in TA during stage II is

attributed to organic acid synthesis (Hirai and Ueno, 1977; Marsh et al., 2000; Richardson

et al., 1997), in spite of rapid fruit growth (Bain, 1958). This is followed by a decrease in TA

caused by degradation of organic acids (Richardson et al., 1997) together with dilution

through increasing fruit size. The lower rate of acid degradation during stage III (from -mid-








68

December) coincides with decreased fruit growth rate, indicating that less dilution of organic

acids occurs, but acid degradation continues (Richardson et al., 1997).

Macroclimatic conditions among Florida's citrus-producing regions were relatively

uniform, with relatively small diurnal and seasonal differences in air temperature. As a result,

annual cumulative degree-days for the two seasons (3408 and 3124 DD, respectively) did not

differ among locations, and were at the upper-end of the proposed range of degree-days

required for 'Valencia' sweet orange (Barry et al., 1996). These data are comparable with

annual cumulative degree-days for the lower Rio Grande Valley, Texas, Imperial Valley,

Calif. (Reuther, 1973), Hainan Province, China (Huang, 1993), and the low-lying parts of

Northern Province, South Africa.

These small differences in air temperature among locations are supported by the narrow

geographic range within which citrus is produced in Florida, by comparison with other citrus-

producing regions of the world. Among the four geographic locations used in this study, the

northern location is <300 km from the southern location, a difference of only 2 15' latitude.

The difference in elevation among locations is <40 m, with little topographical difference in

peninsular Florida to affect climate. As a result, there was no consistent trend among

locations in juice quality of 'Valencia' sweet orange on Carrizo citrange rootstock, and

differences in juice quality among locations were small. The narrow range in SSC among

locations (=0.6% SSC) was of similar magnitude to that from a comparison made between

Merritt Island and Central Florida (Harding et al., 1940), and for Florida Agricultural

Statistics Services (2000) data. The relatively small difference in juice quality between

seasons is also supported by Florida Agricultural Statistics Services (1999) data.








69

The time taken from anthesis to a 9:1 ratio was = 10 months, and differed by only 10 to

14 d among locations. This was <1 month shorter than the time taken in a previous study

using the same maturity index for sour orange (C. aurantium L.) rootstock (Reuther, 1973).

Therefore, harvest date can be affected by relative differences in the time taken to a given

ratio, and may have economic advantage for fresh fruit depending on market conditions

(Florida Agricultural Statistics Services, 1999).

While annual cumulative DD were lower in the 1999-2000 season compared with the

1998-99 season, fruit were larger with higher SSC, and reached a 9:1 ratio sooner in the

1999-2000 season than in the 1998-99 season. However, TA was higher in the 1999-2000

season, with a more gradual decline in TA during stage In of fruit development. The

relationship between juice SSC of 'Valencia' sweet orange and mean air temperature, or

annual cumulative DD, was weak. This lack of relationship is supported by Newman et al.

(1967) who proposed that cumulative DD was not related to juice SSC of'Valencia' sweet

orange, a premise also supported by Reuther (1973). Where relatively large differences in air

temperature occur, e.g., among citrus-producing regions of the world, air temperature had

larger effects on TA and ratio than on juice SSC (Reuther and Rios-Castafio, 1969).

However, macroclimatic differences among Florida's citrus-producing regions were too small

to affect SSC of 'Valencia' sweet orange. Therefore, there was no evidence to support a

relationship between juice SSC and mean air temperature among Florida's citrus-producing

regions.

Florida's mean annual air temperature during the two seasons of the experiment were

higher than the long-term mean annual air temperature, with the 1998-99 season approaching









70
a record high. The lower SSC in the 1998-99 season than the 1999-2000 season may be due

to dark respiration losses of photoassimilates, as speculated by Reuther (1973), implying that

the plateau of the temperature-response curve for 'Valencia' sweet orange was within the

temperature range of this experiment. It is proposed that a mean annual air temperature

higher than that experienced in this study, e.g., in more tropical growing conditions, would

result in 'Valencia' sweet oranges with relatively lower SSC compared with a mean annual

air temperature closer to the long-term mean for Florida.

'Valencia' sweet orange has a broad climatic range within which relatively high quality

fruit are produced (Hodgson, 1967). Therefore, 'Valencia' sweet orange appears to be

insensitive to relatively small differences in air temperature, as shown here among locations

and seasons. Apparently, within the Florida citrus-producing region with its narrow range

in air temperature, and apparent lack of climatic diversity under which the experiment was

conducted, factors other than air temperature are more important to variation in juice quality

of 'Valencia' sweet orange. These factors may include rootstock selection (see Chapters 6

and 7; Castle et al., 1993; Wutscher, 1979) and plant water relations (Koo and Sites, 1955;

Sites et al., 1951). For example, differences as large as 30 percent in SSC occur among fruit

from trees on different rootstocks (Castle, 1995; Wutscher, 1988). Other factors that may

contribute to differences in juice quality among locations are soil conditions and associated

soil preparation practices (Jackson and Davies, 1999; Tucker et al., 1995), tree age (Wheaton

et al., 1999), and cultural practices such as fertilization (Embleton et al., 1967; Koo, 1988)

and pruning (Wheaton et al., 1995).








71

Fruit harvested from the upper, exposed canopy position were slightly smaller, with

higher SSC, lower TA, higher ratio, earlier maturity or time to 13:1 ratio, and possibly a more

rapid rate of fruit development than fruit from the lower, partially shaded canopy position.

Canopy position effects on juice quality are well-documented (Morales et al., 2000; Reitz and

Sites, 1948; Syvertsen and Albrigo, 1980). Reitz and Sites (1948) attributed these effects to

"light classes", although they did not quantify PAR. In this study, the difference in SSC

between canopy positions appeared to be independent of fruit size, and may be due to some

other within-canopy factorss.

Upper canopy positions reportedly experience higher air temperatures than lower canopy

positions (Allen and McCoy, 1979; Morales et al., 2000), and differences in canopy

temperature have been shown to elevate leaf and fruit surface temperatures (Allen and

McCoy, 1979; Syvertsen and Albrigo, 1980). On warm clear days, sunlit leaf temperatures

may be as much as 8 to 10 C above air temperature (Syvertsen and Albrigo, 1980), whereas

shaded leaf temperatures generally follow ambient air temperature (Syvertsen and Lloyd,

1994). However, daily heat stress in the most exposed canopy positions was not limiting to

juice quality of grapefruit (Syvertsen and Albrigo, 1980).

It appears that canopy microclimatic factors other than direct effects of air temperature

may play important roles in causing differences in juice SSC between canopy positions. These

factors are likely PAR (Reitz and Sites, 1948), and water stress (Syvertsen and Albrigo,

1980).













CHAPTER 5
JUICE QUALITY OF 'VALENCIA' SWEET ORANGES
BORNE ON DIFFERENT INFLORESCENCE TYPES



Citrus inflorescences' are classified as leafy and leafless (Reece, 1945), by the number

of flowers borne per floral shoot (Randhawa and Dinsa, 1947), or according to absence or

presence of leaves, the latter being sub-divided further depending on flower-to-inflorescence

leaf ratio (Lenz, 1966; Sauer, 1951; Fig. 2-4).

Citrus fruit set research conducted during the 1940s through 1980s demonstrated that

leafy inflorescences set a higher percentage of fruit (Jahn, 1973; Lenz, 1966; Moss et al.,

1972; Reece, 1945; Sauer, 1951) and produce larger fruit (Ehara et al., 1981; Guardiola and

Lizaro, 1987; Lenz, 1966) than leafless inflorescences. However, there are only two studies

concerning inflorescence effects on fruit quality in citrus (Ehara et al., 1981; Lenz, 1966).

The results of these studies are conflicting, and inconclusive.

In the earlier study (Lenz, 1966), there was little difference in juice quality of 'Valencia'

sweet orange [Citrus sinensis (L.) Osb.] fruit harvested from leafy and leafless inflorescences,

although fruit from leafy inflorescences tended to have slightly higher total soluble solids

(TSS) (=0.5 %), titratable acidity (TA) (<0.1%), and juice content (<1%), despite = 5% larger

fruit size. There was little difference in ratio between fruit from the two inflorescence types.



1 "Inflorescence" applies to all flowering shoots arising from axillary buds (Reece, 1945).

72








73

In the second study (Ehara et al., 1981), the effects of increasing inflorescence leafiness

on fruit quality of satsuma mandarin (C. unshiu Marc.) included 10% larger fruit size, less

flesh relative to peel, more advanced rind color development, and slightly lower juice TSS and

TA. However, these reported differences were not subjected to statistical analysis.

Furthermore, the slightly lower TSS and TA reported may not have been due to inflorescence

type per se, but differences in fruit size due to inflorescence type or canopy position (Harding

and Lewis, 1941; Ketsa, 1988).

Besides possible inflorescence effects on juice quality, the contribution of inflorescence

type to within-tree variation in juice quality is unknown. Therefore, the objectives of this

study were to determine whether juice quality of'Valencia' sweet orange was affected by the

type of inflorescence on which fruit are borne, and to determine the contribution of

inflorescence type to within-tree variation in juice quality.


Materials and Methods


Site and Plant Material

'Valencia' sweet orange trees on Carrizo citrange [Poncirus trifoliata (L.) Raf. x C.

sinensis (L.) Osb.] rootstock planted in 1987 at 380 trees/ha in a north-south row orientation

at Howey-in-the-Hills (lat. 2844'N, long. 81 46'W; elev. 23 m) were used. The experimental

site is located in part of the subtropical, inland area of peninsula Florida, representative ofthe

northern Florida Ridge citrus production region. The soil type is Astatula sand (hyperthermic,

uncoated Typic Quartzipsamment of the Entisol order), an excessively drained sandy soil with

low organic matter content, typical of the region. Trees with uniform crop load were selected








74
from within a commercial citrus orchard, and were well cared for according to local cultural

practices. The trees were irrigated with microsprinklers, and fertilized annually with 225 kg

N/ha. Trees were not topped, but alternate sides of trees were hedged annually to maintain

a drive middle =2.4 m-wide.


Experimental Design and Data Collection

The experiment was conducted during the 1998-99 and 1999-2000 seasons. A 2x2

factorial design (inflorescence type x canopy position) was used to estimate inflorescence type

effects on variation in juice quality, where both factors were fixed. Leafy (LF) and leafless

(LS) inflorescence types, equivalent to Lenz's (1966) inflorescence types C and D (Fig. 2-4),

were selected for the study. Each inflorescence selected bore a single fruit that did not

necessarily originate from a single-flowered inflorescence, and LF inflorescences had three

to seven leaves, whereas LS inflorescences had no inflorescence leaves. To provide different

conditions under which to test inflorescence effects on juice quality, two contrasting canopy

microclimates were used, viz. southwest top (SWT: upper, exposed) and northeast bottom

(NEB: lower, partially-shaded) canopy positions (Reitz and Sites, 1948).

Four bearing shoots per treatment combination per replication (single-tree replications)

were tagged in late June after physiological fruit drop. At that stage, inflorescence types

could be easily distinguished for identification when harvesting fruit samples at maturity.

After physiological fruit drop, samples of four fruit each were harvested from 10 trees (24

June 1998; N=40 observations) or five trees (1 July 1999; N=20 observations). At maturity

(5 Mar. 1999 and 6 Mar. 2000), all remaining tagged fruit were harvested and samples of








75
three or four fruit per treatment combination from five (1999) or six (2000) trees were used

for juice quality analysis (N=20 in 1999 and N=24 in 2000).

Fruit equatorial diameter and weight were measured. Juice was extracted from individual

fruit using a citrus reamer (Sunkist Inc., Los Angeles), and juice weight was measured to

determine juice content (w/w). Brix of juice samples was measured using a hand-held

temperature-compensated Brix refractometer (Atago Co., Japan), and TA was determined

by titration with 0.3125 N NaOH and 0.5% phenolphthalein solution. The acid correction

factor was added to Brix measurements to determine soluble solids concentration (SSC)

(Fellers, 1990). Ratio of SSC-to-TA was determined.


Statistical Analysis

Fruit size and juice quality variables were subjected to analysis of variance using PROC

GLM (SAS Institute Inc., 1996), followed by mean separation using least significant

difference (LSD). Additional analysis included analysis of covariance with fruit size and juice

quality variables as covariates, and estimation and partitioning of variance into its components

to determine the contribution of the sources of variation to total measured variation using

PROC MIXED (Littell et al., 1996; SAS Institute Inc., 1996). Canopy position and

inflorescence type were fixed factors, but were treated as random factors to estimate their

variances.










Results


Fruit Size and Juice Quality at Physiological Fruit Drop

Fruit size. On 24 June 1998, mean fruit diameter was -44 mm, and there was no

difference in fruit size between inflorescence types (Table 5-1). The effect of canopy position

on fruit size was not consistent. In the NEB canopy position fruit borne on LF inflorescences

were significantly larger than fruit borne on LS inflorescences, whereas in the SWT canopy

position there was no significant difference in fruit size among fruit borne on the two

inflorescence types.

On 1 July 1999, mean fruit diameter was z 50 mm, and fruit borne on LF inflorescences

were significantly larger than fruit borne on LS inflorescences by 1.5 mm (= 3 percent

difference). Fruit borne in the SWT canopy position were significantly larger by 3.3 mm ( 7

percent difference) than fruit borne in the NEB canopy position.

Soluble solids concentration. In June 1998, mean SSC was 11.0% (Table 5-1). Fruit

borne onLF inflorescences had significantly higher SSC than fruit borne on LS inflorescences,

and fruit borne in the SWT part of the tree had significantly higher SSC than fruit borne in the

NEB. However, there was no difference in SSC between inflorescence types in fruit borne

in the NEB canopy position.

In July 1999, mean SSC was 8.8%, and there was no difference in SSC for fruit from the

two inflorescence types, whereas fruit borne in the SWT canopy position had significantly

higher SSC than fruit borne in the NEB.









O
S0 0 oo 77
W) CoN c 0 00- 0

m e 66 ooo
S 00 00 00


CON N
00 V>) t, c; C5M


S0 "*N0 0
CA 14 r- 1~


e~e~ o o o
Or-r rl


NT t- 0(


~~C 000 ON 0
*2 ~ ~ ~ c 2 t- n--

r I




~0,
000 00
a co a
o on 0r
'0~t C) C 4

go oo ooc oo O N v-O
f 00 00 C
( a 0 o "o 'o



~00 00 00 00
.- \



-o



C14C10
o. c en a ro






00 0 en
o4-
. oi 4) o


o4 l S


O,4 cN 00


F .00 0
g1- g2- 2 )





(U -; 000





8 'CI ooo


*0 a '0,







4x)
E!00- Q.g0.


Ho0 T C- N0
uq~ Cli 46 0\C S ( >


vi e % .^. 0

2^ 1- ^ c
W)T ^ ^^o
'g~ 00! ^ s o
.5~~ ~ 0-o rT- T-ce








78

The difference in SSC between canopy positions (1.9% SSC in 1998 and 0.4% SSC in

1999) was greater than the difference in SSC between inflorescence types (1.0% SSC in 1998

and 0.1% SSC in 1999).

Titratable acidity. On 24 June 1998 and 1 July 1999, mean TA was 3.30 and 2.82 %,

respectively (Table 5-1). In both seasons, TA did not differ between the types of

inflorescence on which fruit were borne, but TA was significantly higher in the NEB than

SWT canopy position in 1998. There was no difference in TA between canopy positions in

1999.


Fruit Size and Juice Quality at Maturity

Fruit size. In Mar. 1999, mean fruit diameter was = 70 mm, and LF inflorescences bore

significantly larger fruit than LS inflorescences (Table 5-2), but by only 2 mm (<3 percent

difference). Fruit borne in the NEB canopy position were significantly larger (by 6 mm) than

fruit borne in the SWT canopy position. In Mar. 2000, mean fruit diameter was 76 mm, and

fruit size was not affected by inflorescence type, but fruit borne in the SWT canopy position

were significantly larger (by 2.6 mm) than those born in the NEB canopy position.

Soluble solids concentration. In Mar. 1999, mean SSC was 12.4%, and fruit borne on

LF inflorescences had significantly lower SSC than fruit from LS inflorescences, a difference

of 0.4% SSC (Table 5-2). The opposite effect occurred in fruit harvested in Mar. 2000; fruit

borne on LF inflorescences had significantly higher SSC, by 0.4% SSC.

Canopy position effect on SSC was consistent over the two seasons. Fruit borne in SWT

canopy positions had significantly higher SSC than those borne in NEB canopy positions, a










"5 i oo oi 679

oo o nr >



N 6,-- o


8 ooo a ooo
roo 0


) 0 00
S0 0 0
S- 0 -, 0
CA 00
o o o-c -. com













'- en en
00 0

1 6 0O 0
r-r- r- 00 e,
C)




a 1 0 b 00
0 r a)

o o- I-.







C14 N
a~ N N ,1
H a3 > '~i co i0

a\00 0t|li
a| Ci NN N ^








80

difference of 1.4% SSC in Mar. 1999 and 1.0% SSC in Mar. 2000. The difference in SSC

between canopy positions was three times larger than the difference in SSC between

inflorescence types.

Titratable acidity. In Mar. 1999 and Mar. 2000, mean TA was 1.01% and 0.97%,

respectively (Table 5-2), and neither inflorescence type nor canopy position affected TA.

Ratio. In Mar. 1999 and Mar. 2000, SSC-to-TA ratio ofthe juice was 12.4:1 and 13.4:1,

respectively, and in both seasons, ratio was unaffected by inflorescence type.

Additional analyses. Within the SWT canopy position, significant differences in fruit size

and SSC between inflorescence types occurred in both seasons (Table 5-2); LF inflorescences

bore significantly larger fruit of significantly lower SSC in Mar. 1999, but, in Mar. 2000, the

opposite occurred when LF inflorescences bore significantly smaller fruit of significantly

higher SSC. There were no differences in fruit size and SSC between inflorescence types in

the NEB canopy position.

When the apparent relationship between juice quality variables and fruit size was

subjected to analysis of covariance, the adjusted means of SSC and TA were not significantly

different between inflorescence types (Table 5-3).

Variability in juice quality. For SSC, only 3% to 5% of total measured variance was due

to inflorescence type (Table 5-4). In contrast, canopy position made a large contribution to

variance of SSC in the two seasons (46% to 83%). Tree-to-tree variation (33% of total

variance in Mar. 2000) and sample-to-sample variation, represented by "error", (12% to 17%

of total variance) made intermediate contributions to total variance.









81

Table 5-3. Effect of inflorescence type on mean soluble solids concentration (SSC) and
titratable acidity (TA) adjusted for fruit diameter (covariate) of 'Valencia' sweet orange
harvested in Mar. 1999 and Mar. 2000 averaged across southwest top and northeast bottom
canopy positions.
SSC (%) TA (%)
Inflorescence type 1999 2000 1999 2000

Leafy 11.9 13.1 0.88 0.94
Leafless 12.1 12.9 0.88 0.98

P-value 0.6650 0.1539 0.9600 0.1680










Table 5-4. Partitioning of variance into components for soluble solids concentration (SSC)
and titratable acidity (TA) of 'Valencia' sweet orange fruit harvested Mar. 1999 and Mar.
2000 to estimate inflorescence type effects on juice quality.
Variance (% of total variance)'
SSC TA
Source of variation 1999 2000 1999 2000
Tree 0.0000x (0.0) 0.3869 (33.5) 0.0043 (29.4) 0.0104 (58.2)
Canopy position 0.9487 (83.0) 0.5345 (46.2) 0.0011 (2.7) 0.0004 (2.1)
Inflorescence type 0.0617 (5.4) 0.0345 (3.0) 0.0014 (9.4) 0.0000x (0.0)
Error' 0.1323 (11.6) 0.2004 (17.3) 0.0079 (53.5) 0.0071 (39.6)
TOTAL 1.1426 1.1563 0.0147 0.0179
z Total variance partitioned into component sources of variation as estimated by PROC
MIXED (Littell et al., 1996; SAS Institute Inc., 1996). Numbers in parenthesis represent
percentages of total measured variation.
Y Error or residual variation represents sample-to-sample variation.
xThe zero variance estimates are apparent statistical anomalies, and are due to the numerator
mean square (MS) being smaller than the denominator MS used in the F-ratio for the source
of variation being tested.








83

For TA, <10% of total measured variance was due to inflorescence type, whereas

variation among trees (29% to 58% of total variance) and variation among samples (40% to

53% of total variance) contributed the most to total variance.


Discussion


The type of inflorescence on which fruit were borne in a 'Valencia' sweet orange tree had

only a minor effect on juice quality of fruit, and there was no direct association between

inflorescence type and juice quality. Acid content and ratio of SSC-to-TA were not related

to inflorescence type.

Juice SSC was associated with the effect of inflorescence type on fruit size, as small fruit

tended to have higher SSC than large fruit, regardless of the type of inflorescence on which

fruit were borne. The relatively small difference in SSC between fruit borne on leafy and

leafless inflorescences (-3 percent of mean SSC) was, therefore, an indirect result of fruit size

(Harding and Lewis, 1940; Sites and Camp, 1955). Lenz (1966) and Ehara et al. (1981) also

reported relatively small differences in juice quality associated with inflorescence type. Fruit

borne on leafy inflorescences tended to produce larger fruit (Ehara et al., 1981; Guardiola and

Lizaro, 1987; Lenz, 1966) with concomitant lower SSC (Ehara et al., 1981).

The possible advantage of inflorescence leaves being in close proximity to developing

fruit (Koch, 1984), and a related increase in sink strength (Erner, 1989), did not result in

higher SSC of those fruit. Photoassimilates derived from adjacent leaves may be partitioned

to fruit (Koch, 1984), but allocated to cell wall components rather than to soluble constituents

in juice cells, providing the advantage to leafy inflorescences to produce fruit of larger size








84

(Ehara et al., 1984; Lenz, 1966). Alternatively, the early advantage of the presence of

inflorescence leaves on increased fruit set (Moss et al., 1972) is not maintained through fruit

development. Furthermore, dilution of juice soluble solids may occur in larger fruit during

fruit development.

Regardless of fruit size, fruit harvested from the upper, exposed canopy position had

consistently higher juice SSC than fruit from the lower, partially shaded canopy position, a

well-known effect (Morales et al., 2000; Reitz and Sites, 1948; Syvertsen and Albrigo, 1980;

Chapters 3 and 4). Moreover, canopy position was the major contributor to within-tree

variation in juice SSC, whereas the type of inflorescence on which fruit were borne made only

a nominal contribution to variability in juice SSC. This provides support for factors other

than inflorescence type being important components of within-tree variation in juice SSC.