Citation
Nitrogen and biomass distribution, and nitrogen and water uptake parameters for citrus

Material Information

Title:
Nitrogen and biomass distribution, and nitrogen and water uptake parameters for citrus
Creator:
Morgan, Kelly Tindel, 1958-
Publication Date:
Language:
English
Physical Description:
xv, 174 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Biomass ( jstor )
Branches ( jstor )
Citrus trees ( jstor )
Crops ( jstor )
Plant roots ( jstor )
Rootstocks ( jstor )
Soil science ( jstor )
Soil water ( jstor )
Soils ( jstor )
Tree trunks ( jstor )
Dissertations, Academic -- Soil and Water Science -- UF
Soil and Water Science thesis, Ph. D
Greater Orlando ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
General Note:
Printout.
General Note:
Vita.
Statement of Responsibility:
by Kelly Tindel Morgan.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022440227 ( ALEPH )
880637297 ( OCLC )
880438696 ( OCLC )

Downloads

This item has the following downloads:


Full Text










NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER
UPTAKE PARAMETERS FOR CITRUS














By

KELLY TINDEL MORGAN












A DISSERTATION PRESENTED TO THE GRADUARE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004














ACKNOWLEDGMENTS

I would like to acknowledge those people without whose help this study would have been impossible. Foremost, I would like to thank my cochairmen, Drs. Thomas Obreza and Johannes Scholberg, not only for financial and moral support beyond my expectations, but most of all for their patience. I would like to thank Dr. Adair Wheaton who shared so willingly his laboratory space and knowledge gained over 40 years of research. Many thanks are owed to Drs. Nick Comerford and Jim Jones for providing much needed insight into root uptake and crop modeling, respectively.

This project would not have been possible without the help of many people during the collection of samples and chemical analysis. The dedication of Tom Graham and Sam Luther for providing essential organization to the collection of more than 4000 plant tissue and 3100 soil samples and Majie Cody and Amanda Myers during many hours of sample preparation and tissue analysis was greatly appreciated. I would be remiss if I did not thank Drs. Harold Browning, Bill Castle, and Larry Parsons for allowing me the time off from my full-time job to pursue this degree. Their patience when the demands of the job and degree delayed the delivery of data or work on their projects was much appreciated. My sincere thanks are also due to the Florida Department of Agriculture and Consumer Services and Cargill, Inc. for providing funding for this project.






ii










Lastly, and most importantly, I would like to thank my wife, Nancy, and two sons, Joshua and Christopher, for their unwavering support and encouragement. Encouragement from my parents, in-laws, and brother sustained me.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS ....................................................................... ii

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

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

ABSTRACT .............................................................. ............ xiv

CHAPTER

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

Ridge Water Quality Project ................................................................ 2
Citrus Best Management Practices ...........................................................4
Decision Support Systems..................................................................... 5
Objectives .................... ..............................................................

2 LITERATURE REVIEW .................................................................. 9

Introduction............................................................................. 9
Citrus Growth Characteristics .............................................................. 10
Citrus Biomass Distribution .......................................................... 11
Citrus Nitrogen Accumulation and Partitioning ............................. ..... 12
Citrus Root Growth Dynamics .....................................................13
Factors Affecting Root Distribution And Root Density ............... ...... 14
Soil Characteristics ...................... .............................................14
Climatic Effects........................................................... ..... 16
Rootstocks ................................................................................ 16
Tree Spacing and Density ........................................................... 17
Fertilization ........................................................................................... 17
Irrigation .......... .............................................. ..................17
Canopy Reduction...................................................................18
Citrus Water Uptake .......................................................................... 18
Factors Affecting ET ..................................................................19
Crop species ..................................................................... 19
Tree size........................................................................20



iv





V


Climate ..................................................................................... 21
Soil characteristics ..................... ......................... ........... .. 21
Soil water content.................................................... .........26
Water table ........................................................................ 27
Soil shading ................. ................................................... 27
Grass and weed growth.......................................................28
Crop Coefficient ......................................................................... 28
Soil Water Depletion Coefficient ..................... ...........................29
Citrus Nitrogen Uptake ..................................................................... 30
Seasonal Nitrogen Uptake........................................................... 31
Nitrogen Uptake Efficiency .......................................................... 33
Seasonal Nitrogen Redistribution ........................ ................... 33
Crop and Environmental Models .........................................................34
Current Citrus Models .............................................................. 34
Environmental Models.............................................................34
Crop Models ..........................................................................36
C onclusions .................... ....... ................................................ .. 36

3 CITRUS BIOMASS AND NITROGEN ACCUMULATION .................... 38

Introduction .................................................................. 38
Methods and Materials ...................................................................... 42
Experiment 1 Mature Citrus Biomass and N Distribution ......................42
Experiment 2 -Biomass and N Accumulation With Increase in Tree Size.....42 Site Description...............................................................................43
Tree Canopy Volume and Trunk Cross Sectional Diameter .....................43
Tree Biomass Fresh Weight .........................................................44
Sample Processing and Nitrogen Analysis ......................................... 45
Leaf Area, Biomass, and N weight Estimation ................................. 46
Statistical Analysis ....................................................................47
Results ............................................................ ............................ 47
Mature Citrus Tree Biomass Distribution Experiment 1 ...................... 47
Biomass Changes with Increase in Tree Size Experiments I and 2..........50
Nitrogen Distribution ............................................................. 60
Mature Citrus Tree N Distribution Experiment 1 ............... .. 60
Nitrogen Balance .......................................................... .........63
Nitrogen Change With Increase in Tree Size Experiments 1 and 2 ..........64
Discussion ..................................................................................... 69
Conclusions ..................................... ............................... 73

4 CITRUS ROOT GROWTH DYNAMICS .............................................74

Introduction ............................................................. ........... 74
Methods and Materials ........................................................................77
Sample Collection.................................................................... 77
Sample Processing and Statistical Analysis ...................................... 77





vi


R esults .......................................................................................... 78
Mature 'Hamlin' Orange Root Distribution ......................................... 78
Root length Density Distribution Changes with Tree Size .......................83
Discussion ................................................................. 87
C onclusions ........................................................................... .....88

5 CITRUS WATER UPTAKE DYNAMICS ..................................... ... 90

Introduction .............................................................. 90
Methods and Materials .......................................... ..........................94
Site Characteristics ....................................................................94
Soil Capacitance Sensor Data Collection .........................................94
Estimation of Daily ETc ..................................... ................. 96
Estimation of Monthly Crop Coefficient (1) ...................................... 97
Estimation of Water Stress Coefficient (K)...................................... 97
Estimation of Soil Water Uptake per Unit Root Length ..........................97
Results ..............................................................................98
Seasonal ETo and ET, Trends ........................................................ 98
Seasonal K K ..................... ....... ...............................100
K, Estimation ..........................................................................100
Soil Water Uptake per unit Root length Density ................................. 109
Discussion .......... ....................................................................... 110
Conclusions .................. .............. ........................... 113

6 CITRUS NITROGEN UPTAKE AND CYCLING ................................... 115

Introduction ...................... ......... ... ............................ ......115
Methods and Materials ............. ............... ...................................119
Site Characteristics ......................................... ...................119
Experiment 1 Nitrogen Uptake Flux ........................... ... 119
Fertilizer rate and application ..... ........................... 119
Soil sampling procedures ................................................... 120
Analytical methods ..........................................................120
Experiment 2 Seasonal Tissue N Concentration ................................ 121
Tissue samples collected ..................... .............. 122
Tissue analysis ................................................................. 122
Experiment 3 Seasonal N Loss ................ ... .. .................... 122
Results .............................................................................. .. 123
N itrogen U ptake ............................................................... ...... 123
Nitrification Estimation .................. .......................................... 130
Seasonal Tissue N Concentration .............................................. 130
Seasonal Fertilizer Use Efficiency ...... ... .. .......................... 136
Seasonal N Losses ........................................ 137
Discussion ......................... ........................... ................. 138
Conclusions ............................ ............. ................. 144






vii


7 SUMMARY AND CONCLUSIONS .................................................... 146

Mature Tree Biomass Distribution .......................................................146
Biomass Vs Tree Size Relationships ................................................. 147
Nitrogen Distribution .............................................. ............. 147
Mature 'Hamlin' Root Distribution ..................................................... 148
Root Length Density Distribution Changes with Tree Size ........................... 149
Seasonal ETo and ET, Trends.............................................................. 149
Seasonal K ............................................................................... ..150
K, Estimation ............. ......... .................... ........ 150
Soil Water Uptake per Unit Root Length ...............................................150
Nitrogen Uptake ........................................................................... 151
Nitrification Estimation ................................................................. 152
Seasonal Tissue Nitrogen Concentration............................................ 152
Seasonal Nitrogen Loss ..................................................................... 153
Citrus Decision Support System ........................................................ 154
Citrus N practices and BMPs .............................................................. 155
APPENDIX

A Equations ........................................................................ ...... 158

LITERITURE CITED......................................................................... 161

BIOGRAPHICAL SKETCH ................................................................. 174













LIST OF TABLES

Table Pg

3-1. Citrus biomass and nitrogen distribution by tree age as reported
by different studies ................................................................... 41

3-2 Mature citrus tree leaf area and leaf area index as a function of tree
canopy volume and trunk cross sectional area ....................................48

3-3. Dry matter accumulation and allocation between tree components for
mature 'Hamlin' orange tree by tree canopy volume as affected by
year of sampling, rootstock, and interaction of year and rootstock .............49

3-4. Nitrogen accumulation and allocation between tree components for
mature 'Hamlin' orange tree by trunk cross sectional area as affected
by year of sampling, rootstock, and interaction of year and rootstock ........51

3-5. Linear regression analysis of dry weight and N accumulation in different
tree components as related to tree canopy volume .................................58

3-6. Linear regression analysis of dry weight and N accumulation in different
tree components as related to trunk cross sectional area ......................... 59

3-7. Mature Citrus tree tissue N concentration as a function of year of
sample and rootstock .......................... ........................................63

4-1. Mean mature fibrous (diameter < 4 mm) root length
density of 'Hamlin' orange tree as affected by rootstock, orientation
distance, and soil depth................................... .......................... 79

4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length
densities for trees ranging from 2years old to > 15 years old ................. 85

4-3. Regression coefficients and statistics root length density as a function
of distance from the tree trunk, and soil depths by canopy volume
using a third order quadratic polynomial model ................... ............. 85

4-4. Regression coefficients and statistics for root length density as a function
of distance from the tree trunk, and soil depths by trunk diameter
using a third order quadratic polynomial model ............. .......................86

viii





ix


5-1. Monthly maximum, minimum, and mean reference evapotranspiration
reported by Florida Automated Weather Network for the Avalon Station and maximum, minimum and mean estimated citrus crop
evapotranspiration betweeen April 2000 and March 2002.................... 99

5-2. Regression analysis of estimated ET, to calculated ETo ratio by
mean soil water content, soil water potential, and available
soil water depletion in soil 1.0 or 0.5 m deep and in either the
irrigated zone or total tree area ......................................... .... ... 104

5-3. Regression analysis of estimated ET to calculated ETo ratio by day
of year for soil water content values greater than 0.070 cmn cm"
(field capacity) using a quadratic function ......................................105

5-4. Estimated soil water coefficient (K,) values for a range of percentage
available soil water depletion (ASWD) using equation 2 found in
Allen et al. (1997) ..................................... ........................... 108

5-5. Regression analysis of estimated soil depletion factor (K,) by mean soil
water potential, and available soil water depletion in soil 1.0 or 0.5 m
deep and in either the irrigated zone or total tree area ......................... 108

5-6. Regression analysis of estimated soil water uptake per unit root length
density on soil water potential in soil at three locations surrounding the
tree and 10, 20, 40 or 80 cm depths using an exponential decay model...... 110

6-1. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in M arch, 2002................................................................ 125

6-2. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in M ay, 2002 ..... ........................................... .................... 126

6-3. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive and active N uptake for samples collected on five consecutive days
in September, 2002 ................................... 127


6-4. Regression equations for estimated maximum N uptake and estimated
active N uptake rates by soil N concentration (mg I") using an
exponential rise to a maximum model ............................................... 129





X



6-5. Seasonal changes in N concentration, size and dry wt. of fruit, flush
leaves, and expanded leaves for 2001 and 2002 seasons........................ 133














LIST OF FIGURES



1-1. Map of Florida with Lake, Polk, and Highlands counties highlighted............... 3

1-2. Plant/soil nitrogen and water balance flow chart. ......................................7

3-1. Tree canopy volume as a function of trunk cross sectional area for
trees from experiments 1 and 2 ........................................... ......... 52

3-2. Leaf area expressed as a function of tree canopy volume (A) and
trunk cross sectional area (B) ................................................. 53

3-3 Leaf area index on a tree basis expressed as a function of tree
canopy volume (A) and trunk cross sectional area (B) ..........................55

3-4. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
canopy volum e ................................................................. ....... 56

3-5. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
trunk cross sectional area ........................................................... 57

3-6. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and root and tap root dry weight accumulation
as a function of canopy volume ....................................................61

3-7. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of trunk cross sectional area .................................... 62

3-8. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of canopy volume.........65


3-9. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of trunk cross
sectional area ............................................ .......... .................. 66


xi





xii



3-10. N weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of canopy volume .................................................... 67

3-11. N weight allocation to total, above ground, and below ground dry
weight, leaf and twig biomass, total branch biomass and root and
tap root accumulation as a function of trunk cross sectional area. ............. 68

4-1. Root length density distribution by depth at 50, 100, 150, and 200 cm
distance form tree trunk between rows of 'Hamlin' orange trees on
Carrizo citrange or Swingle citrumelo rootstocks .................................80

4-2. Root length density distribution at 0-15, 15-30, 30-45, 45-60, 60-75,
and 75-90 cm depth increments by distance from the tree trunk as affected by distance form the tree trunk for 'Hamlin' orange
trees on Carrizo citrange and Swingle citrumelo rootstocks .....................81

4-3. Citrus root distributions by depth below the soil surface and distance
from the tree trunk for trees 2-5 years old, 5-10 years old, 10-15
years old, and > 15 years old ....................................................... 84

5.1. Illustration of EnviroSCAN probe .....................................................95

5-2. Illustration of EnviroSCAN probe layout, and soil surface area used
for determining soil water content for each probe .................................95

5-3. Estimated ET, to calculated ET. ratio as a function of soil water
content in the irrigated zone to a 0.5 m depth, 1 m depth, and
the total tree area to a 1 m depth................................................. 101

5-4. Estimated ETc to calculated ETo ratio as a function of soil water
potential in the irrigated zone to a 0.5 m depth, I m depth, and
the total tree area to a I m depth .................................................. 102

5-5. Estimated ET, to calculated ET, ratio as a function of available
soil water depletion in the irrigated zone to a 0.5 m depth,
I m depth, and the total tree area to a Im depth ................................ 103

5-6. Comparison of estimated crop evapotranspiration (ETc) with calculated
reference evapotranspiration (ETo) ratio which are an approximation
of K. for observations when soil water content values were near
field capacity as a function of day of year (DOY) .............................. 105

5-7. Estimated soil water coefficient K. as a function of soil water
potential in the irrigated zone to a 0.5 m depth, 1 m depth, and





xiii



the total tree area to a I m depth ................................................... 106

5-8. Estimated soil water coefficient K. as a function of available soil
water depletion in the irrigated zone to a 0.5 m depth, 1 m depth,
and the total tree area to a 1 m depth ......................................... 107

6-1. Relationship of estimated maximum N uptake to soil solution
concentration ........................................................................ 128

6-2. Relationship of estimated passive N uptake to soil solution
concentration ....................................................................... 128

6-3. Proportions of nitrate-N, ammonium-N, and total-N from control
pipes as percentage applied during 3 days after application .................... 131

6-4. Seasonal change in N concentration for flush leaves, expanded
leaves, and twigs for 2001 and 2002 .........................................132

6-5. Seasonal change in N concentrations for bark and wood tissue of
small, medium, and large limbs for 2001 and 2002. Iigh N rate
and Low N rate equal to 268.8 and 179.2 kg ha"' yri', respectively ........ 135

6-6. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2001 season ....................................................... .. 139

6-7. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2002 season ............................................................ 140














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

NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER UPTAKE PARAMETERS FOR CITRUS By

Kelly Tindel Morgan

May, 2004

Chair: Thomas Obreza
Cochair: Johannes Scholberg
Major Department: Soil and Water Science

During the last two decades, microirrigation and fertigation have become

commonplace in Florida citriculture. Concurrently, competition for water and nitrate contamination of ground water have become greater concerns. Information on seasonal citrus N demand, water use, and N uptake is needed to minimize water use and nitrate leaching in citrus management. The objectives of this study were to 1) determine longterm and seasonal changes in citrus biomass and N distribution, 2) develop spatial patterns of citrus root length density with increase in tree size, 3) estimate crop water use and soil moisture coefficients, and root water uptake efficiency, and 4) explore seasonal N uptake rates for citrus. Total tree biomass and N distribution were related to canopy volume and trunk cross sectional area. The percentage of total tree N in citrus leaves decreased from 45 to 37% while branch N increased from 6 to 27% as tree canopy volume increased from 5 to 35 m3. Leaf, branch, and root masses comprised 15, 65, and


xiv





xv



20% of total mature tree mass, and accounted for 45, 35, and 20% of total N mass, respectively. Root density increased radially as tree size increased. Trees on Swingle citrumelo rootstock had a higher proportion of fibrous roots near the soil surface than trees on Carrizo citrange. Soil water uptake ranged from 0.8 to 1.1 of ETo. Daily uptake decreased steadily as soil water content decreased. N uptake from the upper 45 cm of soil was greater for trees on Swingle citrumelo compared with Carrizo citrange. N uptake efficiency ranged from 41 to 55% when fertilized at 269 kg N ha' compared with 47 to 70% when fertilized at 179 kg N haT'. Leaf and twig N was highest from August to February and lowest in May. New and improved understanding of citrus water and N dynamics will advance Florida citrus management techniques and decrease environmental impacts.














CHAPTER 1
INTRODUCTION

With a crop value of $640 million in 2002, citrus is one of the most important horticultural crops in Florida. Currently, nearly 2 million ha are under citrus production, with a 1.1 million metric ton annual production accounting for 73 and 18% of US and world production, respectively (Florida Agricultural Statistics Service, 2002). Citrus is typically produced on sandy soils with poor water and nutrient retention capacity. Adequate supply of both irrigation water and fertilizer are therefore required for optimal production. Most ridge soils lack confining soil layers that can prevent fertilizer nitrates from reaching groundwater. Two issues have become greater concerns for citrus production in Florida: 1) increasing competition between agricultural, commercial, and residential use of limited water supplies, and 2) nitrate contamination of some aquifers less than 50 m deep.

Fertilizer application rates and irrigation management practices for citrus rely

upon crude general recommendations that are standardized over large areas and lack the precision needed in today's ecologically conscious and competitive markets. Although fertilizer and irrigation recommendations provide general production guidelines, they do not capture the dynamic nature of processes controlling non-point source pollution associated with citrus production. Therefore, both growers and regulators must be provided with additional tools such as decision support systems to improve water and



I





2

nutrient use efficiencies and assessment of Best Management Practices (BMP) impacts on citrus production and ground water quality.

Ridge Water Quality Study

The U.S. Environmental Protection Agency, in a nation-wide survey, documented widespread nitrate contamination of shallow drinking water wells (Graham and Alva, 1995). In that survey, approximately 55% of wells were found to contain N03-N contamination above the background concentration. Approximately 1.2% and 2.4% of urban and rural drinking water wells, respectively, were found to contain NO3-N concentrations above the Maximum Contamination Level (MCL) of 10 mg L'1 for drinking water. A correlation between drinking water well contamination and areas with higher fertilizer sales and high value crops was established (Graham and Alva, 1995) suggesting that agricultural fertilization practices may have contributed significantly to N03-N contamination of drinking water wells.

Of 3949 drinking water wells surveyed by the Florida Department of Agriculture and Consumer Services (FDACS), 2483 (63%) contained detectible concentrations of N03-N (Graham and Wheaton, 2000). Of these 2483 contaminated wells, 584 (15% of total surveyed) contained N03-N in excess of MCL. The proportion of wells in Florida contaminated with NO3-N was similar to that of the nation-wide survey. However, the proportion of wells contaminated above MCL was an order of magnitude higher, suggesting that the soils of the state of Florida on average are vulnerable to N03-N leaching to groundwater. Eighty-nine percent of wells contaminated above MCL were located in the central Florida counties of Lake, Polk, and Highlands (Fig. 1-1). Portions of these three counties comprise the central Florida ridge. Soils typical of the "ridge" are





3


hyperthermic Entisols composed of uncoated sands with water holding capacities of 0.04 to 0.09 cm3 cm"3, hydraulic conductivities >50 cm h-', cation exchange capacities of 1 to

5 cmol (+) kg'', and depths of more that 10 m.

Long-term monitoring studies and research projects were initiated in 1992 to evaluate the impacts of nutrient and water management practices in citrus on ground water quality. The goals of the projects established by FDACS (Graham and Alva, 1995) were to 1) generate baseline groundwater quality data from several commercial citrus groves in the ridge area; 2) develop recommendations for alternative nutrient and water management practices; and 3) assess the impacts of these alternative management practices on groundwater quality.
























Fig. 1-1. Map of Florida with Lake, Polk, and Iighlands counties highlighted.





4


Results of this study indicated that N removed at harvest accounted for only 30 to 49% of the applied N (Alva and Paramasivam, 1998). Estimates of N added to the biomass of the trees ranged from 18 to 57% of the N applied. The study concluded that additional information was needed for N accumulation with increase in tree size, optimal timing for N application, N uptake parameters, and improved irrigation scheduling (Graham and Wheaton, 2000).

Citrus Best Management Practices

A best management practice (BMP) for any agricultural commodity is an attempt to use the latest scientific data available to reduce the impact of agricultural operations on the environment while maintaining economically viable production. An interim BMP for citrus was established in 1994 that was based on previous N rate studies and current IFAS recommendations. Citrus growers agreeing to abide by the interim BMP would not be held liable by the Florida Department of Environmental Protection for future cost of supplying drinking water to local users as required by Chapter 376.30 (3) (c) F.S. (Graham and Alva, 1995).

The terms of the interim BMP for orange trees 4 years or more of age were quite broad. Annual N applications were restricted to 134 to 269 kg ha' with the stipulation that groves producing less than 50.4 Mg of fruit per ha should apply no more than 202 kg ha-' N annually. A minimum of two applications per year were required for bearing groves receiving up to 168 kg ha1' N. Bearing groves receiving more than 168 kg ha' N per year were required to receive at least three applications. Those groves using fertigation were required to make a minimum of 10 applications. Application of at least half of the annual fertilizer N prior to the rainy season was encouraged. A UF-IFAS








publication (Tucker et al., 1995) was produced to assist growers in determining the rate of N to apply, timing of application, and suggested irrigation scheduling.

In 2002, a revised BMP established rates and timing of N applications based on tree age classes and method of application. The two age classes are 4 to 7 years and >7 years. The methods of application are broadcast only, broadcast and fertigation, and fertigation only. No more than 34 kg ha" N is to be applied at one time, and no more than 34 kg ha' N may be applied from June 15t to September 15*. No fertigation application is to exceed 17 kg ha' N and must be applied at a minimum 1-wk interval.

Decision Support Systems

Important decisions for growers are when and how much fertilizer and irrigation to apply. They need to consider several factors in their decision-making process to determine that the crop value to be gained is greater than the cost of fertilizer and irrigation applied. Fixed fertilizer and irrigation schedules, based on long-term mean climatic conditions, may lead to inefficient use of these inputs due to the large annual variability in atmospheric conditions (Heinemann et al., 2000). Likewise, variations in the amount of rainfall and its distribution may lead to the loss of N from the crop root zone necessitating additional applications. Due to the complexity of the decision making process, researchers have developed computer-based decision support systems (DSS). A DSS can provide information on management options based on local environmental conditions. These systems also provide a means to make the scientific understanding of complex plant, soil, and environmental interactions accessible to decision makers in a concise and interactive manner. Frequently, information from simulation models has formed the foundation for these DSS.





6


Nutrients leached from agricultural soils represent both an economic loss to farmers and a potential environmental pollutant for groundwater. Concerns about the presence of these agricultural chemicals in groundwater and the need for improved understanding of their movement and transport beyond the root zone have increased considerably over the last several decades. Comprehensive mathematical relationships are required to determine crop fertilizer N uptake and to predict the potential impact of NO3N leaching on groundwater quality for various soil and/or environmental conditions. With the exception of insect and disease population and damage dynamics, most modeling work has focused on predicting mineral N transformations, organic C and N transformations, soil water content, water and N uptake, crop yield, and NO3-N leaching (Hoogenboom et al., 1994). Such models help growers manage resources, maximize returns, and reduce impacts on water quality. Current crop simulation models are being used to optimize planting dates and densities (Saseendran et al., 1998), optimize fertilizer and irrigation inputs (Sexon et al., 1998, and Heinemann et al., 2000), maximize profits (Kiniry et al., 1997), and reduce groundwater pollution (Gijsman et al., 2002) in agronomic crops.

Objectives

Robust crop models can provide a scientific basis for improved resource

management in agricultural production. The long-term accumulation of biomass and N with tree development must be understood. Figure 1-2 illustrates the relationships between tree biomass, tree N content, soil N concentration, and soil water content. Change in tree biomass and N content over time impacts total tree N demand. Soil N and water concentrations affect both active and passive N uptake rates. Likewise, changes






7








Fruit







-U
N Ro




N



N Hc






Layer I Soil Lr I N H20zO SoilSoil Lay










Inputs and Outputs State Variables Functions Fig. 1-2. Plant/soil nitrogen and water balance flow chart.








in root distribution with time must be known to understand the effect of these distributions on the rate of water and nutrient uptake and thus soil water and N concentration. An additional complication is that citrus tree scions are grafted onto rootstocks that affect the growth and uptake rates of the resulting tree. The effects of these rootstocks on tree development and uptake rates must also be understood.

To develop a crop model for citrus, detailed field-scale information must be

obtained under local soil and cultural conditions. A review of the existing literature for these relationships was conducted. The result of this review will be presented in Chapter

2. Information gaps in biomass and N accumulation and field-scale uptake rates of mature citrus trees must be determined so that studies can be designed to complement existing information. The central hypotheses tested in this dissertation are 1) generic relationships can be developed that capture changes in citrus dry biomass, N weight, and root length densities with increase in tree size; 2) daily water uptake changes seasonally and is greatly affected by soil water content; 3) seasonal leafN concentration is lowest and tree biomass abscission is highest during periods of rapid tree growth; and 4) fertilizer-N nitrification and uptake are rapid under Florida conditions. Therefore, the objectives of studies in Chapters 3-6 were to 1) determine changes in above-ground citrus biomass and N distribution for trees under recommended N fertilizer management practices across a range of tree sizes; 2) develop relationships that will capture the overall spatial patterns in citrus root length density distribution for different tree sizes; 3) estimate irrigation crop and soil moisture coefficients and soil water use per unit root length density; 4) explore seasonal N uptake rates for citrus; and 5) compare seasonal biomass and N concentration changes for citrus fertigated at two N fertilizer rates.














CHAPTER 2
LITERATURE REVIEW

Introduction

The various species of the genus Citrus are believed to be native to the subtropical and tropical regions of Asia and the Malay Archipelago (Webber and Batchelor, 1943). Citrons were cultivated on the European continent as early as 300 BC. Limon and sweet orange were not known in Europe for some 15 and 17 centuries later, respectively. Sweet orange was first introduced to the North American continent on Columbus' second voyage in 1493. Citrus was grown in coastal settlements of Florida by the mid 16th century and "wild" citrus trees were found on hammocks near lakes or rivers where conditions were particularly favorable for their growth in the second half of the 19th century (Harris, 1875; Adams,1875). Due to freeze damage, citrus production in Florida has moved in the last 120 years from north central Florida to the southern half of the Florida peninsula. Currently 2 million ha of citrus fruit are grown in Florida (Florida Agricultural Statistics Service, 2002). Florida citrus production was 1.1 million metric tons accounting for 73 and 18% of US and world production, respectively.

Vaile (1924) showed that fine sand or sandy loam soils resulted in better growth and production than coarse sands or heavy loams. Subtropical in nature, citrus trees do not exhibit dormancy or shed their leaves during the winter months However, new growth appears in definite cycles, with two to four cycles of growth yearly. The first and usually largest growth starts in the early spring (late February to early March), the second


9





10


from early June to early July, and the third in late summer (August or September). The principal blooming period for all commercial species is early spring and usually lasts approximately 6 weeks (Mid February to late March). The normal period of ripening of most citrus fruits is late fall and winter, preceding the spring bloom. However, late maturing varieties such as 'Valencia' require 12 to 15 months for maturity, which occurs after bloom for the next crop. The period of time when citrus fruit can be harvested is about 9 months.

Citrus production areas in Florida range from upland positions with very sandy "Ridge soils", which are deep and excessively well drained, to relatively low "flatwood" sites that are often flooded in their native state due to the presence of a spodic layer. Flatwoods soils must be drained and bedded before planting to lower the fluctuating water table. Each of these general areas in Florida presents somewhat different challenges for growing citrus trees and fruit production. Although crop growth and nitrogen uptake dynamics are readily available for many agronomic crops, this information is in short supply for citrus. A comprehensive literature review relevant to citrus growth characteristics, root distribution, water and N uptake dynamics, and crop growth models will be presented in this chapter.

Citrus Growth Characteristics

Turrell et al. (1969) proposed citrus tree growth equations based on growing conditions and cultural practices in California. These equations assumed citrus tree growth to be logistic in nature dennding on cultural characteristics such as spacing, pruning, and irrigation and fertilizer scheduling, soil characteristics, and climatic conditions. Several studies measuring citrus biomass in relation to nutrient concentration





11


distribution or uptake have been conducted, but few biomass studies have been conducted for a range of tree sizes grown under similar cultural soil and climatic conditions. Citrus Biomass Distribution

Legaz and Primo-Millo (1981) harvested 4 year-old 'Valencia' orange trees grown outdoors in sand culture in Spain at five different times in a 1-year period. The mean dry mass percentages for leaves, twigs and branches, lateral roots and fibrous roots were 22.5, 28.7, 45.8, and 2.9% respectively.

Cameron and Compton (1945) divided eight year old 'Valencia' trees grown in California into 14 parts, 1) leaves, 2) twigs, 3) shoots, 4) lateral branches 0.75-1.5 cm, 5) tertiary branches 1.5-3.0 cm, 6) secondary branches 3.0-6.0 cm, 7) primary branches > 6.0 cm, 8) trunk, 9) main root, 10) feeder roots, 11) rootlets, 12) small roots 0.3-0.8 cm, 13) intermediate roots 0.8-2.5 cm, and 14) large roots > 2.5 cm. The difference between twigs and shoot was not given, but both had similar N concentrations on a dry mass basis (5.11 and 5.02%/, respectively). Likewise, the sizes of rootlets and feeder roots were not given but had similar total N percentages (1.38 and 1.49%/, respectively). Mean percentage dry weight of leaves, twigs and shoots, branches, trunk, main root, lateral roots and fibrous roots were 16.8, 9.6, 43.2, 1.5, 2.1, 14.0, and 1.7% respectively.

The biomass proportions for young 'Hamlin' orange trees grown under Florida conditions were substantially different. In a study by Mattos (2000), eight categories of plant parts for 7 year-old 'Hamlin' orange trees were used. These categories were 1) summer-fall flush leaves, 2) spring and older leaves, 3) twigs >1.5 cm, 4) twigs <1.5 cm, 5) trunk, 6) roots <0.2 cm, 7) roots 0.2 to 1.0 cm, and 8) roots >1.0 cm. Leaves, branches,





12


trunk, lateral roots, and fibrous roots constituted 13.9, 37.4, 9.0, 10.2 and 14.2% of total dry biomass, respectively.

In a '5N study, Feigenbaum et al. (1987) divided 22 year-old Shamouti orange trees fertilized at two annual N rates into seven components: 1) leaves, 2) twigs <5 mm, 3) branches >5 mm, 4) trunk and main branches, 5) main root (tap), 6) lateral roots >1 mm, and 7) fibrous roots <1 mm. There was no diameter given to differentiate between branches and main branches. Biomass percentages for leaves were 6.4 and 8.4% for the low and high N treatments respectively. Similar differences were found for all categories with the exceptions of main and lateral roots. The ranges in mean percentage of fresh biomass for twigs, branches, trunk and main branches, lateral roots, and fibrous roots were 1.6 to 2.2, 32.5 to 33.4, 24.5 to 29.2, 4.6 to 6.1, and 3.8 to 4.8%, respectively. These

differences between fertilizer-N rates resulted in a greater percentage of biomass below ground for the low nitrogen application (34.1%) compared with that of the higher nitrogen rate (27.8%).

In a similar study, Kato et al. (1984) found different values for 21 year-old Satauma mandarin trees in Japan. In this study, mean percentages of dry biomass for leaves, twigs, branches, trunk, lateral roots and fibrous roots were 15.7, 4.7, 32.2, 23.1, 20.3, and 3.1% respectively.

Citrus Nitrogen Accumulation and Partitioning

Nitrogen balance studies in citrus provide information on physiological tree N requirements and can be used to develop methods that minimize potential losses of N to groundwater and the atmosphere (Feigenbaum et al. 1987). In a study of 8 year-old 'Valencia' trees in California conducted during a period of 2 years (Cameron and





13


Compton, 1945), leaves contained from 40 to 50% of total tree N. Twigs and shoots contained approximately 10% of total N. Trunk and branches contained from 20 to 30% of total tree N. Approximately half of this N was in the bark, whereas this tissue component represented only 5% of the total dry mass. Roots contained from 15 to 20% of the N, half or more of which was in the bark that made up only 5% of the dry mass of the tree.

Seasonal changes in leaves, bark, twig, and root N concentration were greater

than N changes in woody branch tissue. The trees contained more N just before initiation of growth activity during spring than at any other time of year. Maximum bark N content occurred about December 1, followed by a decrease. It was speculated that the reduction between December and February was the result of deposition of starch and possible other carbohydrates in these tissues. A decrease of all tree tissue N concentration occurred during bloom, fruit set, and periods of active growth in the spring and early summer. During the summer and autumn, N concentrations gradually increased to the mid-winter maximum. Mattos (2000) found similar N concentration values for 7 year-old 'Hamlin' trees. Nitrogen concentration was lowest in the trunk and taproot of these trees. The N concentration of leaves (2.1 to 2.6%), twigs (0.4 to 0.8%), and roots (0.6 to 1.7%) varied with tissue age. Younger tissue tended to have greater N concentration compared with older tissues.

Citrus Root Growth Dynamics

Citrus trees are productive and grow well on central Florida deep sandy soils. In some instances, tree size and yield appear to be related to root system characteristics (Castle and Krezdorn, 1975). Citrus fibrous roots are commonly defined as those roots <4





14


mm in diameter. Their dry mass is a relatively small part of the total root system, but their composite length far exceeds that of the woody roots (>4 mm in diameter). These fine roots are considered to be the "functional" part of the root system because of their critical role in water and nutrient uptake. There is some variation among rootstocks in the morphology of fibrous roots (Castle and Youtsey, 1977). Some rootstocks like trifoliate orange [Poncirus trifoliata (L.) Raf.] produce higher specific root length or length/unit mass (Eissenstat, 1991). Fibrous roots are also the most vulnerable part of the root system. Their development, function, and longevity are strongly influenced by soil characteristics, environmental changes, crop species, crop growth stage, and cultural practices.

Factors Affecting Root Distribution and Root Density Soil Characteristics

The distribution of roots is modified by the physical and chemical properties of the soil profile (Hillel, 1971). Widespread root development and high fibrous root concentrations were observed in deep soils of sand texture where there were virtually no impediments to root growth provided that water and nutrients were non-limiting to growth (Ford 1952; 1953a&b; 1954a&b; 1959; 1964; 1972); Ford et al., (1954; 1957). Increased tree size and yield have been related to root system depth and fibrous root mass. The depth of rooting of 'Orlando' tangelo trees on 10 rootstocks growing in deep, sandy soil was correlated with tree height (Castle and Krezdorn, 1975). Although fibrous root distribution was affected by tree height, total fibrous root dry mass measured at the canopy dripline was not correlated with tree height. Ford (1954a; 1964; 1968; 1969; 1972) conducted many studies of citrus trees in poorly drained Spodosols of Florida and





15


concluded that tree size was closely related to fibrous root density. Extensive lateral root development occurred on soils with loamy or clay texture (Boswell et al. 1975; Kaufimann et al., 1972). In these studies, the root systems were shallower than root systems of plants grown in sandy soils with few roots found below a soil depth of 50 to 70 cm (Adriance and Hampton, 1949; Boswell et al., 1975, Cahoon et al., 1956; 1959; 1961; Kimball et al., 1950; Mikhail and EI-Zeflawi, 1978). Furthermore, changes in fibrous root distribution with depth were more gradual compared with sandy soils, and overall fibrous root concentrations were lower (Bielorai, 1977). The lower natural soil fertility (Carlisle et al., 1989), and excessive drainage of sandy soils resulted in higher shoot:root ratios such that fibrous root dry mass densities tended to be lower in sandy soils (Castle, 1978).

Under flatwood conditions where the soil is drained and bedded, virtually all the root mass occurs within 45 cm of the soil surface (Calvert et al., 1967; 1977; Ford, 1954a; Ford, 1972; Reitz and Long, 1955). The quantity of fibrous roots decreases with depth and lateral distance from the trunk. Elezaby (1989) reported lateral fibrous root distribution to a depth of 180 cm of a 10-year-old 'Valencia' tree [Citrus sinensis (L.) Osb.] on 'Volkamer' lemon (C. volkameriana Ten. and Pasq.) grown on a soil with a deep sand profile and spaced at 4.5 m x 6.0 m as: 9% of the fibrous roots between 0 cm and 60 cm from the trunk, 31% between 120 cm and 180 cm, and 21% between 240 cm and 300 cm. The vertical distribution was: 42% of the fibrous roots between 0 cm and 30 cm from the soil surface, and 14% or less at each 30-cm depth increment to 180 cm. In the same study, fibrous root dry mass density (concentration) ranged from 300 g m3 to 1200 g mn. Those data are similar to dry mass densities reported in other Florida studies





16


(Castle, 1978; 1980). In a recent report, data were given as root length densities and ranged from 530 cm m3 for 'Swingle' citrumelo roots [C. paradise Macf. x P. trifoliata

(L.) Rafi to 2020 cm m"3 for trifoliate orange (Eissenstat, 1991). Climatic Effects

Root distribution was studied in 22 mature navel or 'Valencia' orange groves in California (Cahoon et al., 1956). In this study, 500/ were low-yielding while the remaining 50% were high-producing. Fibrous root fresh mass was measured to a depth of 90 cm under the canopy and between rows. Yield was not related to the under-canopy root quantities, but was correlated with the root quantities measured between adjacent rows where soil water contents were typically lower most of the year. Rotstocks

Some citrus roots have been found as deep as 7 m (Ford, 1954b), and in one

instance, roots of mature trees on rough lemon rootstock were discovered 14 m from the tree trunk (Ford, 1970). Castle and Krezdorn (1975) described two general types of root systems, the first characterized as "extensive" featuring extensive lateral and vertical development, and the second as "intensive" with less extensive root expansion and higher fibrous root concentrations mainly confined to the upper soil layers. Trees on rough lemon, 'Volkamer' lemon and 'Palestine' sweet lime (C. limettioides Tan.) rootstocks typified the extensive type of root system where 50% of the fibrous roots occurred below 70 cm in the soil with wider spreading lateral development. Examples of the intensive type were 'Rusk' citrange and trifoliate orange, where few fibrous roots were found below 70 cm, and the root system was less developed laterally. Some rootstocks like sour orange and 'Cleopatra' mandarin were classified as intermediate. Trees on 'Cleopatra'





17


mandarin had a highly developed lateral root system at the surface, but few fibrous roots below the surface. Menocal-Barberena (2000) found no statistically significant differences in vertical or horizontal fibrous root distribution of'Hamlin' orange on 'Cleopatra' mandarin, 'Swingle' citrumelo and 'Carrizo' citrange rootstocks. Vertical and horizontal root distribution were similar to other studies with about 40% of the fibrous roots in the top 30 cm and 9 to 14% at each of the remaining 30 cm depth increments to 180 cm. Few roots were found below 180 cm. Tree Spacing and Density

Due to Florida's rainy season, roots of trees at commercial spacing rapidly

occupy the volume of soil outside the irrigated zone. After canopy closure, they extend into the rootzone of adjacent trees. Elezaby (1989) reported fibrous root concentration in the 0 to 30 cm zone increased from 450 to 1000 g m3 between trees when the in-row distance decreased from 4.5 to 2.5 m. The increased root concentrations in this study were concluded to be the result of overlapping root systems. Trees at the closest spacing showed root concentration increases to depths of 150 cm (Elezaby, 1989). Fertilization

Increases in fertilizer N can increase root growth to a considerable depth, but the largest effects generally occurred near the surface (Ford, 1953b; Ford et al., 1957; Smith, 1956; Smith, 1965).

Irrigation

Irrigation method and scheduling has been shown to change the distribution and/or concentration of citrus fibrous roots. In a California study of trees receiving different irrigation treatments, yield was not correlated with fibrous root density (Cahoon





18


et al., 1964) because trees in the low irrigation rate treatment declined in yield while maintaining root quantities similar to those of the trees in the higher irrigation rate treatment. It was concluded that soil water content was the single most important factor influencing citrus root systems.

In a Florida study, root weight densities were determined under the tree canopy, at the dripline, and in the row middles to a depth of 180 cm for 'Hamlin' orange trees on 'Swingle' citrumelo and 'Carrizo' citrange rootstocks (Menocal-Barberena, 2000). Trees receiving irrigation at a rate of 40 cm yr"' had significantly higher densities than trees receiving 250 cm yr"'. The differences were on the order of 1.3 to 2.3 times greater for the 40 cm yr' treatment at all depths.

Canopy Reduction

Hedging, the annual removal of excess vegetative growth, has become a common method of canopy size control for closely-planted citrus trees. Eissenstat and Duncan (1992) found that within 30 days of canopy reduction, 20% of the fibrous roots between the 9 cm and 35 cm depths were apparently dead. Root length density of these trees recovered within 63 days of canopy reduction. This relatively short-term reduction in fibrous root density adversely affected yield because of fruit abortion.

Citrus Water Uptake

Assuming little or no surface runoff water applied to the soil surface is 1)

retained in the soil, 2) utilized by plants, 3) lost to the atmosphere, or 4) drained below the crop rooting zone. Drainage water may contain substantial quantities of agricultural chemicals and soluble nutrients. Irrigation practices should be aimed at minimizing 1) crop water stress by maintaining sufficient water within the crop rooting zone, 2)





19


pollution of groundwater by leaching, and 3) production costs associated with excessive irrigation, and nutrient and pesticides losses due to leaching.

Mills et al. (1999) reported a significant decrease in citrus stomatal conductance after midday. This decrease was most pronounced for south-facing exterior leaves and increased with increasing evapotranspirational demand (ET.). Soil water use from 2 yearold 'Hamlin' orange trees measured at 0.5-hour intervals using weighing lysimeters indicated that water continued to be removed several hours after the midday decrease in stomatal conductance. Two seemingly opposing theories place control of soil water uptake at the leaf level via leaf water potential (Slatery, 1967) or root via root water potential (Tinker and Nye, 2000). The former assumes that leaf water potential exerts control on stomatal conductance regulating transpiration and thus water uptake. The latter speculates that dehydrating roots, due to low soil water content, indirectly control stomatal conductance through the production of chemical compounds that after translocation to the leaves reduce stomatal aperture. Lafolie et al. (1991) measured decreasing leaf water potential with decreased root water potential until midday. After reduced stomatal conductance at midday, leaf water potential increased without a corresponding decrease in root water potential. This result was given as evidence that stomatal conductance was not controlled by leaf water potential alone. Factors Effecting ET,

Crop species

Citrus are evergreens and therefore require water for transpiration throughout the year. Citrus leaves are thick and waxy, resulting in high cuticular resistance to transpiration (Mills et al. 1999). Koo (1963) and Koo and Sites (1955) stated that water





20


requirements of grapefruit are generally higher than orange or mandarin varieties for trees of equal size. Wiegand and Swanson (1982 a, b, c) and Wiegand et al. (1982) reported that mean daily citrus ET, at Weslaco, Texas ranged from 2.2 to 3.3 mm for Ruby Red grapefruit and 1.9 to 2.7 mm for Marrs oranges from 5 to 10 years of age.

Under similar climatic conditions, citrus trees are known to have lower

transpiration rates compared with other crop plants. Mahrer and Rytwo (1991) reported mean estimated daily crop water use (ETc) rates for cotton in the Hula Valley of Israel of

5.4 mm when irrigated daily, and 4.0 mm during a 14-day period when not irrigated. Likewise, Starr and Paltineanu (1998) reported that daily ETc rate for full canopy corn at Beltsville, MD ranged from 3.8 to 5.0 mm prior to rainfall and 5.2 to 8.0 mm after. Lower citrus transpiration rates are related to lower leaf and canopy conductance (Mills etal. 1999).

Tree size

Large, vigorous, healthire more water than young trees (Tucker et al. 1997). In Florida, large trees at low planting densities (150 to 180 trees per ha) may use 62 to 94 L per day during the winter months and 189 to 219 L per day in July and August (Boman, 1994). Rogers and Bartholic (1976) reported a mean annual ET, of 1210 mm during an 8-year period from a young orange and grapefruit grove on poorly drained soils near the east coast of Florida. These annual ET, values ranged from 820 mm early in the study (tree age 2 years) to 1280 mm at end of the study (tree age 10 years). Linear regressions of annual ET, vs. years during the study resulted in significant (P=. 1) increase in ETc. Mean annual ETc increased at a rate of 19 mm per year or a cumulative increase of approximately 13% in 8 years. Fares and Alva (1999) reported an annual ET,





21

value of 920 mm for 3-yr old 'Hamlin' orange trees grown on deep sandy soils in central Florida. Koo and Harrison (1965), and Koo and Hunter (1969) reported annual ET, values of 1170 mm for mature citrus on the same soil series. Climate

Mean annual ET. for citrus in Florida ranges from 820 to 920 mmn (Rogers and Bartholic 1976; Fares and Alva 1999) for young (<5 years) trees to 1170 to 1280 (Koo 1978, Rogers and Bartholic 1976) for mature (10 years or more) trees. Annual ET, values reported for mature citrus grown in the lower Rio Grande Valley of Texas are similar to those for Florida and ranged from 1044 to 1232 mm (Wiegand et al. 1982). Hoffman et al. (1982) reported annual ET, values for well-irrigated citrus grown in semi arid Arizona of 1470 mm. Lower ET, rates for Florida (humid) compared with Arizona (semi-arid) have been attributed to lower evaporative demand (Rogers et al. 1983, Fares and Alva 1999).

Soil characteristics

Crop water supply must be based on a clear understanding of soil water dynamics. Water in excess of field capacity drains through the vadose zone. Eventually, water that is not taken up by plants or evaporated from soil or plant surfaces makes its way into the ground water and contributes to aquifer recharge (Fares and Alva, 1999). Under-tree sprinklers and drip irrigation systems are designed to deliver water at rates low enough to allow infiltration into the soil without contributing to losses by runoff These systems can be managed in such a manner that the excessive downward drainage through the soil is minimized. The required application amount is governed by the soil-water depletion on a given irrigation date, irrigation efficiency, and the target soil-water level. Most of the





22

terms are not independent. For instance, the amount of applied irrigation water will influence the amount of ET, as well as the amount of drainage (Prajamwong et al., 1997).

In standard irrigation practices, water transport through the soil may be classified into five phases: 1) infiltration during application; 2) redistribution after application ceases; 3) withdrawal by plant roots; 4) evaporation from the soil surface; and 5) drainage of water to lower soil depths. The primary modes of transport of water in soil are 1)

scous flow through liquid-filled pores, and 2) diffusion of vapor through air-filled

res. In principle, both modes contribute to soil water flow. Liquid flow is the dominant ode in saturated to moist soils (Hagan et al., 1967). Vapor flow is of minor importance until soils become quite dry, although the presence of a large temperature gradient favors the contribution of this mechanism. For typical soil water situations, both of these transport modes contribute to a flow rate proportional to potential energy gradients within the soil.

Water is of central importance in the transport of solutes in soils or plants,

whether by diffusion or mass flow (Tinker and Nye, 2000). The concept of potential is fundamental to understanding soil water dynamics. Potential is a measure of the energy state of a chemical compound within a particular system, and hence of the ability of a unit amount of this compound to perform work Difference in potential at different points in a system gives a measure of the tendency of the compound, including water, to move from a region with high potential to a region of lower potential.

Soil water has various forms of potential energy acting on it, all of which

contribute to the total potential. Tinker and Nye (2000) refer to these forms of potential energy as concentration, compression, position in an electrical field, and position in the





23

gravitational field. These same forms of energy are commonly referred to as osmotic, matrix, gravitational and pressure potentials, the sum of which is referred to as total water potential (4). Thus, soil water moves in response to the difference in water potential over a distance. The first published relationship between water flux and energy gradient was obtained empirically in 1856 by Henry Darcy after a study of saturated sand filters (Hagan et al., 1967).

u =-K d /dx Equation 2-1

Where:

u = water flux (cm3 cm2 -1),

K = hydraulic conductivity constant (cm s-),

+ = soil water potential (kPa), and

x = the distance over which the flux is maintained (cm).

The constant of proportionality of Darcy's Law (K) is known as the hydraulic

conductivity, and is a function of both the properties of the medium and the fluid (Tindall and Kunkel, 1999). In saturated soils, K will be constant as long as the structure of the soil remains stable because the water flow pathways will be unchanged. In unsaturated soil, K varies with the water content (0), because the latter defines the total cross-section area for water flow, the effective water-filled pore radius, and the effective pathlength (Tinker and Nye, 2000). A soil with a wide range of pore sizes conducts fluid more rapidly than a soil with small pore sizes (Tindall and Kunkel, 1999). The saturated hydraulic conductivity of soils has a wide range from 10-9 cm s- for clay to 1.0 cm s" or more for sand. Lower values of K for a clay medium (with smaller pore sizes) are likely due to the drag exerted on the viscous fluid by the walls of the pores. Particles of smaller-





24

sized individual grains (such as clays compared with sands) have a larger surface area that increases the drag on water molecules that flow through the soil, reducing permeability and hydraulic conductivity.

As water is lost from the soil, the continuity between water-filled pores also

decreases. A soil with water-filled volume fractions less than 0.1 (0 < 0.1 cm3 cm3) will normally have a very low value for K(0) (Tinker and Nye, 2000). The Poiseuille equation states that the flow rate in a tube increases proportionally to the fourth power of its radius, at a constant pressure gradient. Water in larger soil pores will empty first as the soil dries, effectively reducing the cross sectional diameter of the soil water pathway Therefore, pore size and distribution has a large effect on the flow rate

f= (x r4/ 811) dP/dx Equation 2-2

Where:

f= flow rate in a tube (m3 s"'),

r = radius (in),

r= viscosity (pPa s'), and

dP/dx = pressure gradient.

The driving force for soil-water movement is the difference in matric potential, resulting from a difference in soil water content. Richards postulated that Darcy's Law could be extended to unsaturated states by assuming that the hydraulic conductivity (K), as well as the water content, could be treated as non-hysteretic functions of the pressure head or potential (Slatery, 1967). The matric potential and the water content for a soil are related by the soil-water characteristic curve. By using the slope of the soil characteristic curve (d6/d0), the following equation can be obtained based on Darcy's Law and is





25


known as Richard's equation (Tinker and Nye, 2000). In this equation, flow in unsaturated soil can is expressed in terms of the water content gradient and soil water diffusivity (De).

u = -Ke d/dx = -Ke (d4/dO) (d0/dx) = -Do (dO/dx) Equation 2-3 Where:

u = water flux (cm3 cm2 s-i),

KR = hydraulic conductivity constant at 0 (cm s'),

= soil water potential (kPa),

0 = soil water content (cm3 cm"),

d0/d = slope of the soil characteristic curve, and

x = the distance over which the flux is maintained (cm).

The term diffusivity (Do) is used because the form of equation is the same as that of Fick's law of diffusion (Tinker and Nye, 2000). Furthermore, D is somewhat less convenient than Ko under conditions of hysteresis because Do is discontinuous at each reversal of the direction of potential, while K is continuous and virtually hysteresis-free (Hagan et al., 1967). Experimentally, the effect of hysteresis on Richards' equation has usually been ignored by limiting the soil water potential change to either always drying, or always wetting.

"Field capacity" (Ofc) describes the water content held in the soil after excess water has drained to drier soil layers by redistribution. This equilibrium can be determined in the field by measuring the soil water content as a function of time to determine the value of 0 when de/dt approaches zero. Hillel (1971) noted that the rate at which d/dt approaches zero is dependent on Oi and the depth to which the soil is being





26

wetted. The concept of field capacity is useful in the design of field management schemes for approximating the maximum amount of soil water storage. Field capacity can be used as an upper limit value of 0 within each soil layer such that any water in excess of Of quickly drains to the next deeper soil layer. The soil profile can be described as a vertical sequence of reservoirs with the overflow level for each reservoir corresponding to the value of Ofc for that specific soil layer. During irrigation or rainfall the top reservoir flows over to fill the next lower reservoir until no excess water remains to flow into the next reservoir. With a judicious selection of the depth of each soil layer, this simple analog of the soil profile can be easily modeled.

Soil water content

Estimated annual ET, for a deforested area on the Florida ridge reached 680 mm (Sumner, 1996). This ET rate was attributed to periods of low soil water content because the area was not irrigated. Rogers et al. (1983) reported that growth and fruit yield of citrus trees were greater during a 3-year period for treatments maintained at higher soil water content. During the same period of time, annual ET, averaged 900 and 1210 mm for the lowest and highest soil water content treatments, respectively. Hoffman et al. (1982) reported annual ET, values to be 200 to 500 mm higher than that found by Erie et al. (1965) in Arizona. The lower annual ET, values reported by Erie et al. were attributed to infrequent irrigation resulting in dry soil surfaces and thus increased resistance to water diffusion to the atmosphere. Smajstrla et al. (1986) reported a reduction in growth and ETc with increased available soil water depletion of 2-year-old 'Valencia' orange trees grown in drainage lysimeters. Available soil water depletion setpoints used for irrigation scheduling in this study were 28, 47 and 58%. It was concluded that tree stress




27

occurred at the highest depletion value due to the reduced ability of the soil to transport water to the roots because of reduction in hydraulic conductivity. Fares and Alva (1999) calculated daily ETc for 3-year-old 'Hamlin' orange trees on deep sandy soil in central Florida. Estimated daily ET, values decreased with time after each rainfall or irrigation. Water table depth

Obreza and Admire (1985) concluded that shallow water tables in flatwoods soils could significantly augment water available for root uptake. Graser and Allen (1987) suggested that water-table management by controlling water table depth in the winter and spring could help decrease the need for supplemental irrigation during the dry season. Boman (1994) used drainage lysimeters in which he maintained a water table at 0.61, 0.76, and 0.91 m to measure the effects of water table on ETc, growth, yield and fruit quality of 5-year-old 'Valencia' trees. However, treatment effects were not significant. Soil shading

Castel et al. (1987) estimated soil surface evaporation by comparing water loss from weighing lysimeters in which the soil was covered by plastic with lysimeters that remained uncovered. Mean estimated evaporation was reported as 0.78 mm, greater than 18% of the estimated potential ET of 4.25 mm. Castel and Buj (1992) reported that the percentage of ground shaded by young Clementine trees increased from 10 to 25% during a 4-year period. Evapotranspiration increased by 33% during the same time period. This increase was attributed to the increasing water use by the trees and reduced soil surface evaporation.





28

Grass and weed growth

Smajstrla et al. (1986) used field drainage lysimeters to determine the effect of grass cover on the growth and ET, of 2-year-old 'Valencia' orange trees. Automated covers were installed to cover the lysimeters during rainfall. Soil within the lysimeters was maintained bare or covered completely with bahiagrass. The bare soil lysimeters consistently had the lowest monthly ETC. Measured annual ET, ranged between 1331 to 900 mm for grass-covered lysimeters and 912 to 441 mm for those with bare soil surfaces. Total ET, was 46 to 105% higher per year due to soil grass cover. These results were similar to those reported by Stewart et al. (1969) using non-weighing lysimeters. In their study, estimated annual bare soil evaporation and 2/3 sod cover ETc averaged 68 and 92% of full sod cover, respectively. Tucker et al. (1997) reported reduced soil water use from non-irrigated middles between rows of mature citrus by limiting the height of weed growth by chemical mowing.

Crop Coefient

An estimate of evapotranspiration for a specific crop (ET) is calculated by

multiplying the reference evapotranspiration (ETo) by an empirically determined crop coefficient (K). This coefficient is specific for a crop, growth stage, and growing conditions. The resulting ET, estimates water use of a crop under local or regional climatic conditions.

Rogers et al. (1983) reported monthly measured ET, to calculated ETo ratio values using the mean of four methods of estimating ETo (Penman, Blaney-Criddle, JensenHaise, and Class A pan). The resulting monthly ratios range from 0.90 in January to 1.11 in June. Crop coefficient (Ke) values reported by Doorenbos and Pruitt (1977) after





29


adjustments for humid conditions ranged from 0.9 in March though December to 0.95 in January and February. Castel et al. (1987) estimated monthly K for drip-irrigated mature 'Navel' oranges grown in 'Valencia', Spain. Their K, values were calculated from mean daily ET, estimated from weekly ET values determined by neutron probe measurements. Values ranged from a mean of 0.71 from January through July to 0.90 from August through December. Castel and Buj (1992) suggested these values differed from those reported for Florida due to the lower evaporative demand of the humid Eastern coast of Spain, which has a mean annual ETo of 1166 mm compared with 1400 mm in Florida.

Calculated K values for 3-year-old 'Hamlin' trees grown on sandy soil in central Florida ranged from approximately 1.05 in November through March to 0.85 in May through August (Fares and Alva 1999). Boman (1994) calculated IK values for 5-year-old 'Valencia' orange trees grown in non-weighing lysimeters with water tables maintained at 0.6, 0.75, or 0.9 m from the soil surface. Calculated K, values were at a minimum of

0.6 from December through February and peaked at 1.1 in June and July. Martin et al. (1997) estimated mean daily ETc values for 7-year-old "Redblush" grapefruit in Arizona from soil water content data collected at I to 2 week intervals. Monthly K values were calculated by comparing these estimated daily values with mean daily ETo for the same period. The resulting K ranged from a low of 0.55 to 0.6 in December and January to a high of 1.1 to 1.2 in July.

Soil Water Depletion Coefficient

According to Allen et al. (1998), the water depletion coefficient is defined as the effect of soil water reduction on ET, by reducing the value of IK. It is calculated by





30

multiplying the IK of a given crop by the soil water depletion coefficient (K.) for a given soil water content. Water stress increases as soil water is extracted by evapotranspiration.

Available soil water (ASW) is defined as the difference between drained upper limit (field capacity) and drained lower limit or permanent wilting point. However, the energy expenditure required to extract residual soil water increases as soil water content decreases. Likewise, resistance to water flow increases as residual soil water decreases, reducing water flux to the root boundary. Therefore, crop water uptake is reduced well before wilting point is reached (Allen et al 1998). At field capacity, roots can absorb water fast enough to supply the ETc demand of the atmosphere. However, water becomes more strongly bound to the soil matrix and is more difficult to extract as soil water content decreases. When soil water content drops below a threshold value, water can no longer be transported quickly enough to the roots to supply the transpiration demand of the crop. The fraction of ASW above this threshold is known as readily available water (RAW). For most crops grown on medium and fine textured soils, RAW is as much as 30 to 50% of ASW (Allen et al. 1998). When root zone depletion exceeds this threshold, ET is reduced relative to potential crop ET, and water stress occurs.

Citrus Nitrogen Uptake

Knowledge of the nutritional need of different plant organs as well as the seasonal demand for nutrients is essential in order to establish a physiological basis for crop fertilization (Lagaz and Primo-Millo, 1981). The potential contribution of fertilizer N to the deterioration of ground water quality may be appreciable (Embleton et al., 1978). This impact is especially true in Florida where the combination of high annual rainfall, sandy soils and shallow water tables create conditions that greatly increase the potential





31

for ground water contamination (Alva and Paramasivam, 1999; Calvert and Phung, 1972; Mansell et al., 1980).

Seasonal Nitrogen Uptake

Most N balance studies have been unable to completely account for total N applied to the soil. Some authors attributed this fraction (usually 30 to 50%) to atmospheric loss. Khalaf and Koo (1983) concluded that unaccounted for N was either incorporated into soil organic matter or stored in the tree (Dasberg, 1987), while others made no attempt to fully account for the applied N (Mansell et al., 1980).

Hilgeman (1941) estimated N uptake by grapefruit in Arizona by determining

changes in leaf N concentration seasonally. Maximum N uptake by the trees occurred in March and September relative to January due to higher mean soil temperature. In a 3-year study, Chapman and Parker (1942) determined N removed from solution culture and reported that the months of least N absorption were January and February. Uptake rates increased during the period of late spring through early fall (May to October) with a maximum in July. Roy and Gardner (1946) in Florida reported similar results.

Numerous reports suggest that actively growing tissues act as a sink for N uptake and that the young developing leaves and fruit constitute the strongest sink. Legaz et al. (1982) in Spain studied N distribution in 5-year-old 'Calamondin' trees in sand culture. Trees were labeled with "N for 20 days during flowering, were harvested, and analyzed for N content 0 to 70 days later. Accumulated N was found primarily in fruitlets and newly developed leaves and twigs. About 30% of the labeled N was found in newly formed leaves. In Israel, Feigenbaum et al. (1987) treated 22-year old 'Shamouti' orange trees with '5N labeled fertilizer. Trees had previously been supplied with sufficient N or





32


had been N-depleted to explore the influence of prior fertilization practices on subsequent N uptake. The highest percentage of labeled N occurred in fruit, new leaves and twigs. Only about 20% of the leaf and fruit N originated from the labeled source, suggesting considerable redistribution from stored reserves. Less than 14% of the labeled N was found in roots or large limbs. Dasberg (1987) found that 80% of the N in new growth came from stored rather than applied N, suggesting previous nutrition has significant influence on current season growth and fruit yield. Legaz and Primo-Millo (1988) reported increased N uptake from the beginning of spring flush to bloom. Uptake increased through the spring, reaching a maximum at the summer flush after which uptake declined gradually through winter.

Mooney and Richardson (1992) observed an N concentration gradient between the roots, trunk and branches of citrus trees in New Zealand. High concentrations were found in the branches, with lower concentrations in the roots. Nitrogen concentrations in the trunk were highest at bud break and declined steadily through fruit set and development to a minimum at fruit harvest. Nitrogen concentration for all categories peaked at flowering and then decreased steadily until harvest. Nitrogen uptake efficiencies of 82.0 and 74.1% for ammonium nitrate and urea, respectively, were reported by Mattos (2000). Legaz et al. (1982) reported 50 to 60% of total tree 'SN recovery in above-ground tree parts. Absorption rates increased only slightly from the beginning of growth until flowering, and increased sharply reaching a maximum value at the second growth flush (July) before declining during the fall and winter months. Dasberg (1987) demonstrated that the highest rate of "N uptake by citrus trees occurred during fruit set and the lowest occurred during winter.





33


Nitrogen Uptake Efficiency

Nitrogen uptake efficiency (NUE) is defined as the percentage of applied N taken up by plants (Scholberg et al. 2002). The ability of crop plants to take up and utilize N efficiently is key to providing adequate N for crop growth while reducing N leaching. Mattos (2000) estimated NUE for 6-year old 'Valencia' trees grown in a sandy soil to be 40% and 26% for ammonium nitrate and urea respectively. Feigenbaum et al. (1987) reported that the NUE for a '5N labeled KNO3 applied to 22 year-old 'Shamouti' orange was 40%. Syvertsen and Smith (1996) estimated NUE to be 61% to 83% for 4-year old grapefruit trees grown in lysimeters. Nitrogen uptake efficiency decreased with increased N application rates. Lea-Cox and Syvertsen (1996) reported a similar finding of lower NUE with higher N application rate for greenhouse grown seedlings. The NUE reported ranged from 47% to 60% after an uptake period of 31 days.

Kato et al. (1982) found a 10-fold increase in'5 N uptake of'Satsuma' mandarin during summer (mean temperature 23 C) compared with the winter season (mean minimum temperature 3 C). Scholberg et al. (2002) found N uptake of greenhouse-grown seedlings to be proportional to soil temperature, ETo and canopy biomass. Nitrogen uptake also increased with the time high N concentrations were maintained in the root zone. Increasing the residence time from 2 to 8 hours resulted in an increase in NUE of 95% and 125% for high and low N application rates, respectively. Seasonal Nitrogen Redistribution

Legaz et al. (1982) suggested that at post-blossom, the N concentration in the

spring leaves decreased due to this tissue becoming an N source for the developing fruit. Using 4 year-old 'Valencia' orange trees, daily root N uptake was lower during





34


dormancy, increased during flowering and was highest during fruit set, and later decreased towards the end of the summer and autumn flushes. The greatest accumulation of N absorbed from fertilizer (with respect to the total N absorbed from fertilizer in the whole tree) was found in the young leaves and roots, followed primarily by twigs and stems, then flowers and fruits.

Kato et al (1982) found that total N contents decreased in both bark and wood

during the sprouting period of 21 year-old 'Satsuma' mandarins. Greatest decreases in N were found in parts with higher concentrations of N (i.e. leaves, shoots, and fine roots). It was also concluded that the trunk and large roots are main N reservoirs for new shoot development. The N was reserved mainly as protein, free proline, arginine, and asparagines. Protein decreased in all plant parts in proportion to total N in the plant part. Proline decreased mainly in the leaves and bark, arginine in wood of shoots and asparagines in bark of fine roots.

Crop and Environmental Models
Current Citrus Models

Few predictive models have been developed specifically for use in citrus

production. Most models have been designed for specific applications, with a general user in mind. These models predict population and/or crop damage caused by citrus pathogens (Timmer and Zitko, 1996), and scale insects (Aris and Browning, 1995). Other citrus models are used for irrigation scheduling (Xin et al., 1997), and crop flowering (Bellows and Morse, 1986; Valiente and Albrigo, 2000). Environmental Models

Nitrogen leaching from agricultural soils represents both an economic loss to the farmer and potential groundwater pollution. Mathematical models can be used to assess





35

crop N-fertilizer requirements and to predict effects of N fertilizer management practices on potential nitrate leaching and how it affects groundwater quality. The understanding of solute movement and transport has increased in the last 30 years. Increased environmental concerns pertaining to the runoff and leaching of agricultural chemicals and fertilizer elements in the surface and groundwater has resulted in development and use of computer simulation models to predict transport of potential pollutants in agricultural systems. These models include Nitrate Leaching and Economic Analysis Package (N-LEAP) (Follett et al., 1994), Groundwater Loading Effects of Agricultural Management Systems (GLEAMS) (Reck, 1994; Reyes et al., 1994), Drainage-Modified (DRAINMOD) (Saleh et al, 1994; Verma et al., 1995), Chemicals, Runofl and Erosion from Agricultural Management Systems (CREAMS) (Minkara et al., 1995; Saleh et al., 1994), Leaching Estimation and Chemical Model (LEACHM) (Jemison et al., 1994), and Nitrogen, Carbon, Soil, Water And Plant (NCSWAP) (Jabro et al., 1993).

These models could be applied to citrus production to predict or estimate the depth of N leaching below the crop root zone. Most of these models are deterministic, non-steady state, and comprehensive. They typically require a large number of soil physical, hydraulic, and chemical characteristics for each soil layer, soil N transformation components, weather data, and environmental information to determine N fate and leaching depths. Use of these models for the prediction of N fate under agricultural production conditions has met with mixed results (Kiniry et al., 1997). Jabro et al. (1993) found that neither LEACHM nor NCSWAP successfully predicted nitrate leaching below

1.2 m in a silt loam soil. Jemison et al. (1994) reported accurate predictions using LEACHM in manure fertilized corn crops.





36


Crop Models

Crop-Environment Resource Synthesis (CERES) was developed to model growth and yield of grain crops (Jones and Kirniry, 1987; Kiniry and Bockhot, 1998; Kiniry et al., 1997; Lizaso et al., 2001; Saseendran et al., 1998). CROPGRO was initially developed as a family of crop-specific models for the prediction of legume and vegetable crops (Hoogenboom et al., 1994; Jones et al., 1991; Wagner-Riddle et al., 1997). These are process-oriented models for the simulation of vegetative growth and reproductive development. They predict dry matter growth (Shen et al., 1998), crop development (Batchelor et al., 1994; Batchelor et al., 1997; Piper et al., 1996) and final yield (Batchelor et al., 1996; Heinemann et al., 2000) for a range of agronomic crops. Inputs are daily weather data, soil profile characteristics, and crop management conditions (Gijsman et al., 2002). Crop and soil water (Hoogenboom et al., 1994; Gabrielle et al., 1995; and Xie et al., 2001), N (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995; and Sexton et al., 1998), and C balances are modeled. These models have been combined into the DSSAT (Decision Support System for Agrotechnology Transfer) software (Hoogenboom et al., 1994; Jones and Luyten, 1998).

Conclusions

Considerable research and resources have been devoted to improving our

understanding of how cultural, soil, and environmental factors influence biomass and N accumulation during citrus tree development. However, these studies compared tree component dry weights and N accumulations with tree age and not a measure of tree size. Tree size is not only a function of tree age, but soil, environmental, and horticultural factors as well. Therefore, correlation of dry weights and N accumulations with tree size





37


would provide a better relationship for modeling purposes. Factors affecting citrus root distribution have been studied under Florida soil and environmental conditions. Many of these studies were performed in groves with lower tree densities and different irrigation methods than those currently used in Florida citriculture, and on trees grafted on rootstocks that are no longer in use. Thus, information on the effect of tree size on root length density distribution changes for current production systems are lacking. Likewise, root length density distributions for mature trees on currently used rootstocks grown on Florida sandy soils have not been determined.

Seasonal maximum daily water uptake rates under Florida environmental

conditions have been determined for trees grown on flatwood soils with fluctuating water tables. However, maximum daily water uptake rates for mature citrus trees have not been measured for trees grown on excessively drained "Ridge soils". Likewise, reduction in daily citrus water uptake with decreased soil water potential has not been determined for sandy soils. Citrus N uptake rates have been determined for seedlings and relatively small trees grown in lysimeters. These rates may not reflect uptake rates of mature citrus trees at the field-scale.

Much data on citrus growth, root distribution, water requirements, and N uptake rates are needed to attain the level of crop modeling currently available fbr agronomic crops. Obtaining these data are difficult due to the size of mature citrus trees compared with agronomic crops, and the inability to follow a cohort of trees from planting to maturity. Biomass, N accumulation, and spatial root length density changes as affected by tree size, and water and N uptake dynamics of mature trees under Florida "Ridge" conditions will be presented in the following chapters.














CHAPTER 3
CITRUS BIOMASS AND NITROGEN ACCUMULATION

Introduction

Citrus is native to the subtropical and tropical regions of Asia and the Malay

Archipelago (Webber and Batchelor, 1943). Citrons were introduced into Europe via the Middle East as early as 300 BC, with lemons and sweet oranges following some 15 and 17 centuries later, respectively. Citrus is well adapted to Florida soil and environmental conditions and proliferated in the costal settlements of Florida by the mid 16th century and was in commercial production by the mid 1800s.

Nitrogen application rate studies in citrus have emphasized the effects of timing and amount on increased canopy volume and yield (Sites et al., 1953; Reitz, 1956; Reuther et al., 1957; and Koo, 1979). However, optimum plant growth depends upon maintenance of an efficient balance between roots and shoots (Kramer and Boyer, 1995). Roots are dependent on shoots for carbohydrates, growth regulators, and some organic compounds, while the shoots of a plant are dependent on the roots for water and nutrients. The root to shoot ratio varies widely among species, with age, and with environmental conditions. Understanding the dynamics of root and shoot development with time is essential when determining biomass and N accumulation, and soil water and nutrient uptake dynamics.

Caruso et al. (1999) found that the relative proportion of leaves and twigs to total tree dry weight decreased with tree age. Therefore, relative proportions of total dry matter





39


and N accumulation in different tree components change with age of perennial crops due to the increase in weight of larger branches and trunks of older trees to support the increased tree biomass. Similar changes occur in annual crops with increase in biomass between emergence and harvest. However, as with annual crops, tree size is not dependent on age alone; rootstock (Castle, 1978, 1980), crop nutrition (Feigenbaum et al. 1987), irrigation (Parsons et al., 2001), and restriction of the root system (Mataa and Tomingag, 1998) can limit growth of a citrus tree. Thus, these factors can result in trees of equal age being very different in size, biomass, and N content.

Many crop models such as CERES, CROPGRO, and DSSAT determine the effect of assimilate costs for vegetative and reproductive growth and N budget through C and N balances with increase in crop biomass on water and N uptake, growth, and yield of agronomic crops (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995; and Sexton et al., 1998). Likewise, optimal irrigation and nutrient management is dependent upon the estimation of biomass and N content in citrus trees. Therefore, the relationship of tree size to biomass and N accumulation is needed.

Previous studies have correlated long-term citrus tree biomass and N

accumulation with tree age (Cameron and Appleman, 1935; Cameron and Compton, 1945; Feigenbaum et al., 1987; Kato et al., 1984; Mattos, 2000). Leaves of 3.5, 7, and 15 year-old 'Valencia' trees in California contained from 40 to 50% of total tree N, while twigs and shoots contained approximately 10% of total N (Cameron and Appleman 1935; Cameron and Compton, 1945). Trunk and branches contained from 20 to 30% of total tree N, approximately half of which was in the bark, a tree component that represents only 5% of the total dry mass. The roots contained from 15 to 20% of tree N, half or





40

more of which was in the bark (5% of the dry mass of the tree). The biomass proportions for 7 year-old 'Hamlin' orange trees grown under Florida conditions reported by Mattos (2000) were more similar to the 10 year-old trees cited above than the 3.5 year-old trees (Table 1). Nitrogen concentration was lowest in the trunk and taproot of these trees. The N concentration of leaves (2.1 to 2.6 %), twigs (0.4 to 0.8 %), and roots (0.6 to 1.7 %) varied with tissue age. Younger tissue tended to have greater N concentration compared with older tissues. Kato et al. (1984) and Feigenbaum et al. (1987) harvested older citrus trees (21 and 22 years old, respectively). Leaves comprised a smaller fraction of total biomass in both studies compared with branches and total roots. The leaves of these older trees contained a lower proportion of total tree N than the branches, equal to the proportion of N in the roots (Table 3-1). None of the above studies related biomass or N measurements to tree size parameters such as canopy volume or trunk cross-section area.

Biomass and N distribution relationships based on tree size measurements as opposed to tree age could provide more generic information needed for modeling tree growth and N cycling in citrus production systems. Therefore, the hypotheses to be tested in the following studies were: 1) functional relationships can be defined that correlate biomass and N partitioning of specific tissue categories with tree size using generic growth indicators such as canopy volume or trunk area, and 2) rootstock has a significant effect on citrus biomass and N partitioning. Such relationships can be used to determine citrus N budgets and develop specific fertilizer recommendations that will provide adequate N for growth and production while protecting groundwater from nitrate contamination. Thus, a non-destructive method of estimating an N budget is needed for trees of unknown or mixed ages. Therefore, the objectives of the following studies were





41


Table 3-1. Citrus biomass and nitrogen distribution by tree age as reported by different studies.

Authors Trees Location Cultivar Age Plant Biomass N

(n) (Yrs) Tissue (% Total) (% Total) Cameron and 15 California Val. 3.5 Leaves 30.5 61.9 Appleman Branches 38.4 21.0 (1935) Roots 31.1 17.1

Legaz and 8 Spain Val. 4 Leaves 22.5 30.0 Primo-Millo Branches 28.7 18.5 (1988) Lateral roots 45.8 41.2 Fibrous roots 2.9 10.3

Mattos 6 Florida Ham. 6 Leaves 13.9 35.0 (2000) Branches 46.5 28.Z Lateral roots 25.7 13.4
Fibrous roots 14.1 23.2Cameron and 4 California Val. 10 Leaves 18.5 46.7 Appleman Branches 60.7 39.0 (1935) Roots 20.7 14.2

Cameron and 36 California Val. 15 Leaves 16.8 45.3 Compton Branches 61.4 34.8 (1945) Lateral roots 20.4 17.0 Fibrous roots 1.7 2.9

Kato et al. 1 Japan Sat. 21 Leaves 8.6 27.2 (1984) Branches 65.1 44.6 Lateral roots 19.5 14.3
Fibrous roots 6.8 14.0

Feigenbaum 2 Israel Sham 22 Leaves 7.3 24.6 et al. (1987) Branches 61.0 49.8 Lateral roots 26.5 19.2
Fibrous roots 4.3 3.8


to determine: 1) changes in biomass and N distribution with change in tree size, 2) yearly

changes in biomass and N content of mature citrus trees, and 3) rootstock effect on

mature citrus tree biomass and N distribution. The relationships of canopy volume and

mean trunk diameter to biomass and N content for citrus will form the basis of a

predictive model to estimate the biomass and N distribution based on size measurements.





42

Materials and Methods

Citrus trees of various sizes were measured and dissected into constituent parts during a period of 1 year. Representative tissue samples of constituent parts for each tree were weighed and analyzed to estimate total dry mass, relative percentage dry mass, and N content of the various tree components. These data were used to determine N allocation between different tree constituents. Experiment 1 Mature Citrus Biomass and N Distribution

Two sets of six trees each were dissected in February 2001 and January 2002 at the Water Conserv II site near Winter Garden in western Orange county, Florida. Both sets of trees were 14 year-old 'Hamlin' orange trees planted in 1987 at a spacing of 3 m between trees in the row and 6 m between rows resulting in a tree density of 556 trees ha"

1. Three trees of each set were budded on Swingle citrumelo (Citrsparadwsi Macf x Ponciru& trfoliata (L.) Raf) rootstock, and the remaining three trees of each set on Carrizo citrange (C. sinensis L. Osbeck X P. trifoliata L. Raf.) rootstock. All trees had been fertigated at an annual rate of 240 kg N ha~'. The trees were irrigated (and fertigated) with reclaimed water containing approximately 7 mg N03-N L-'. Experiment 2 Biomass and N Accumulation with Increase Tree Size

A third set of seven 'Valencia' trees on Swingle citrumelo rootstock were

dissected at the K.D. Revell grove owned by Cargill, Inc. near Fort Meade in southern Polk county, Florida. Fresh, dry, and N weights were determined for the same constituent parts as in experiment 1. These trees had been fertilized using dry chemical fertilizer three or more times per year and irrigated with well water.





43


Site Descriptions

The soil series at the Orange county site was Candler fine sand (hyperthermic, uncoated, Typic Quartzipsamment), and at the Polk county site was Zolfo fine sand (sandy siliceous, hyperthermic Grossarenic Entic Haplohumod). These two soils are typical of the central Florida ridge and have a field capacity water content of 0.06 to 0.08 cm cm3 in the upper 1 m. The Candler series consists of excessively drained, very rapidly permeable soils formed from marine deposits. These soils are located in upland areas and typically have slopes of 0-12%. The A and E horizons consist of single-grained fine sand, have a loose texture, and are strongly acidic. A Bt horizon is located at a soil depth of 2 m and includes loamy lamellae of 0.1 to 3.5 cm thick and 5 to 15 cm long. Zolfo series soils are sandy and slightly less well drained than those of Candler. The A horizon consists of fine sand with single-grained, loose texture. The Bh horizon between 4.0 and 5.0 cm consists of fine sand coated with organic matter possessing weak granular to weak fine subangular blocky structure.

Tree Canopy Volume and Trunk Cross-Section Area

Changes in canopy volume have been used in fertilizer rate experiments as measures of tree growth (Whitney et al., 1991). Therefore, tree measurements were determined for the purpose of correlating biomass and N concentration of various tree components to tree size. Canopy diameter of each tree was measured in the row (in-row) and across therow(cross-row) at a height of 1.5 m above ground level. Tree height and hedgerow intercept measurements were made using a 5 m graduated pole. Hedgerow intercept is the height from the ground to the point at which the canopies of two trees meet in the row. These measurements have been used by Whitney et al. (1991) to determine canopy volume based on a spheroid model (Equation 3-1). Trunk diameters 5





44

cm above the ground were determined for each tree by measuring in-row and cross-row dimensions. Trunk cross-sectional areas (TCSA) were determined for each tree assuming an oval shape.


Int 2
(1-(1-(~)))
TCV= Ir rCr Ht* Equation 3-1
4 3

Where:

TCV = Tree canopy volume (m3)

Ir = In-row spacing (m)

Cr = Cross-row spacing (m)

Ht = Canopy height (m)

Int = Canopy intercept height (m)

Tree Biomass Fresh Weight

Fresh weight of the leaves, twigs <7 mm, small branches 7 to 15 mm, medium branches 15 to 30 mm, large branches >30 mm, trunk, tap root, small roots <4 mm, medium roots 4 to 20 mm, and large roots >20 mm were measured in the field. Field weights and three samples of each plant part category were collected for each tree using the following protocol: Twigs less than 7 mm in diameter and attached leaves were cut from the tree with leaves intact. These twigs were placed into a plastic container and weighed on a battery powered field-portable balance. During cutting, one twig out of 20 was placed into a separate container as they were cut and were weighed separately. Leaves were removed from the twigs in this container while still in the field. Fresh weights of these subsamples were measured. Branches 7 mm in diameter and greater were cut into 15 to 30 cm segments, separated into the three size ranges noted above, and





45

weighed separately. Two to three samples equal to 5% of the fresh weight of each size category were removed from each container and placed into labeled plastic bags. Any leaves attached to these branches were removed and weighed prior to weighing the branch segments. The trunk and taproot were cut into pieces and weighed, and three longitudinal slices of each were retained as separate samples.

The trees were planted 3.1 m in row and 6.2 m between rows. Therefore, the roots were cut to a depth of approximately 0.3 m using a shovel in a rectangle 3.1 m in-row x

6.2 m across-row with the tree stump at its center. The bulk of the root system was extracted using a front-end loader equipped with a root rake. All roots were removed from the excavation to a depth of I m, washed, blotted dry, separated into size categories, and weighed in the field.

Sample Processing and Nitrogen Analysis

The leaf area of 50 random leaves from each sample was measured. Each branch segment of each sample was cut into at least five disks of approximately 0.5 to 1 cm thick that facilitated the removal of bark from the wood. Likewise, the bark was removed from each horizontal trunk slice. The bark and wood from the branch and trunk disks were weighed separately to determine the fresh mass proportion of bark to wood for each size category.

Samples were dried at 70 C to a constant weight before recording dry weight.

Total tissue dry weight for each tree was determined by multiplying fresh weights by the respective tissue dry matter content. All tissues were ground using a Cyclotec mill (1093 Sample Mill, Tecator manufacturing, Sweden) for the leaf tissue and Wiley mill model 1 (Arthur Thomas Manufacturing Co., Philadelphia, Pa) for woody tissue. The ground tissues were digested using a Buchi Model K435 12-vessel digestion unit (Buchi





46

Analytical, Inc., New Castle, DE). The digest was analyzed for total Kjeldahl nitrogen using USEPA method 351.2 using a Buchi model B339 steam distillation instrument (Buchi Analytical, Inc., New Castle, DE.).

LeafArea, Biomass and N Weight Estimation

Specific leaf area (cm2 g') were determined for a 50-leaf subsample by dividing total leaf area by leaf dry weight. Total leaf area was estimated by multiplying the mean specific leaf area by the estimated total dry leaf weight for the corresponding tree. Leaf area index was determined for each tree by dividing the leaf area by the corresponding cross sectional canopy area using the in-row and cross-row measurements. Leaf area index was also determined on a per acre basis by dividing the total leaf area of each tree by the land area occupied by the tree (in-row spacing x cross-row distance). Dry weights for each tissue type of individual tree were estimated by multiplying the total field fresh mass by the mean percentage dry mass of three sub samples for each tissue category. Likewise, N accumulation within each tissue type of each tree was estimated by multiplying the total dry weight by the mean N concentration of the three sub samples of each tissue. Total tree dry weight and N accumulation were determined by summing across tissues categories. The above ground dry weight and N accumulations were determined by summing the estimated values for leaf, twig, total branch, and trunk components. Likewise, the below ground biomass and N accumulation was the sum of total root and taproot values. Percentages of total biomass and N weight were determined for each tissue.

Prior to dissecting the second set of mature 'Hamlin' trees in 2002, all fruit was removed and weighed. Representative fruit samples were collected and dried to





47

determine percentage dry weight and analyzed for N concentration to determine total fruit N accumulation. Leaves and twigs collected in 2002 were separated into current year's growth and prior year's growth. Separate samples were collected for dry weight and N concentration determination.

Statistical Analysis

Since the samples were taken during a 14-month period, the samples from the mature (14 year old) 'Hamlin' trees were treated as repeated measures and analyzed accordingly using the SAS general linear models (GLM) procedure (SAS Institute, Cary, NC). Non-linear regression analysis of tissue masses and percentages of all trees were made considering canopy volumes and mean trunk diameters as independent variables.

Results

Mature Citrus Tree Biomass Distribution Experiment 1

Leaf area was significantly different (P=0.05) for TCV and TCSA, but not for rootstocks (Table 3-2). However, neither mean leaf area index on a per tree basis (10.2 and 9.9 for Carrizo and Swingle trees, respectively) nor on a per acre basis (6.4 and 6.2 for Carrizo and Swingle trees, respectively) were significantly different for tree size or rootstock. Total above-ground and below-ground weights increased on both a fresh and dry basis across the range of TCV and TCSA encountered in this study. Maximum total fresh weight was approximately 160 kg tree"' with TCVs ranging from 28 to 38 m3 and TCSAs of 80 to 160 cm2. Maximum dry biomass for the same canopy volumes and trunk cross-sectional areas was approximately 100 kg tree'. Maximum above-ground biomass (leaves, twigs, branches, and trunk) and below-ground (roots and taproot) was approximately 74 and 26 kg tree', respectively.





48





Table 3-2. Mature citrus tree leaf area and leaf area index as a function of TCV and TCSA.
Leaf Area Leaf Area Index (m2 tree') (mn leaf m'2 canopy) Carrizo Swingle Carrizo Swingle Mean 110.0 103.4 10.2 9.9

Standard Deviation 10.3 24.8 1.2 1.8 Statistical Significancez

TCVY * NS NS TCSA * NS NS

Rootstock NS NS NS NS NS = not significant, *= significant P<0.05, and **= significant P<0.01. YTCV = tree canopy volume, TCSA = tree cross-sectional area.




Mature trees on Carrizo rootstocks had significantly greater (P=0.05) mean dry weight (100 kg tree") than those on Swingle (83 kg tree'), whereas, below-ground biomass was not significantly different (Table 3-3). Significantly higher (P0.01) percentage of large branch biomass was found for trees grown on Carrizo citrange (23.8 kg tree") compared with trees grown on Swingle citrange (15.8 kg tree'). Thus, percentage total branch biomass was significantly greater (P=0.05) for the Carrizo









Table 3-3. Dry matter accumulation and allocation between tree components for mature 'Hamlin' orange trees as affected by year of sampling, rootstock, and interaction of year and rootstock.

Dry Weight Accumulation Dry Weight Allocation
Total Above Below Total Total Branches Total Roots
Mass Ground Ground Leaf Twigs Sm. Med. Lrg. Total Trunk Sm. Med. Lrg. Tap (kg dw tree') (% of total weight) Year 1 97.6 69.2 28.3 12.7 5.8 17.0 12.1 19.1 48.2 3.6 5.1 4.9 9.5 10.1
Year 2 87.8 64.9 26.3 12.4 7.3 15.3 9.0 21.1 45.4 3.9 6.1 6.8 6.4 9.2
NSz NS NS NS NS NS NS NS NS NS NS ** NS NS

Carrizo 100.3 77.0 26.6 12.2 7.1 16.0 10.5 23.8 50.3 3.9 5.5 6.1 6.9 6.7
Swingle 82.6 54.7 27.9 12.9 6.0 16.1 10.4 15.8 42.3 3.7 5.8 5.7 8.8 13.1
NS NS NS NS S NS ** NS NS NS NS *

Car YrI 104.2 79.2 25.0 12.9 6.8 15.7 11.2 25.4 52.3 3.9 4.9 5.6 8.5 5.1
CarYr2 96.3 74.8 28.2 11.6 7.5 16.4 9.7 22.2 48.3 3.9 6.2 6.6 5.4 8.3
Swi Yrl 87.6 54.3 33.3 12.4 4.3 18.9 13.5 9.5 41.9 3.2 5.6 3.8 11.1 17.7
Swi Yr2 79.4 55.0 24.4 13.3 7.1 14.2 8.3 20.0 42.5 3.9 6.0 6.9 7.3 10.1
NS NS NS NS NS NS NS ** NS NS NS NS NS NS

zNS = not significant, *= significant P<0.05, and **= significant P<0.01.





50

citrange roostock trees. On a percentage basis, the taproot biomass was significantly greater (P=0.05) for trees on Swingle than those on Carrizo.

Although tree weights appeared to be less in 2002 compared with 2001, total dry weight and dry weights of tree components were not significantly different at the P0.05 level (Table 3-3). Leaf weights represented 12 to 14% of total tree dry weight, while total branch weights (twigs, total branches, and trunk) accounted for 49 to 63% of total tree weight. Dry matter allocation to tree roots was highly variable in all three studies, but averaged in the 19 to 21% range.

Total above-ground N accumulation was significantly affected by TCV and TXA at the P=0.05 level (Table 3-4). Leaf branch, and root N comprised approximately 45, 35, and 20% of total tree N, respectively. Dry weight allocation to leaves, twigs, small branches, medium branches, and trunk were similar for both rootstocks. However, Carrizo trees had almost 50% more N in large branches compared with Swingle. Mean total N accumulation by large roots and taproot was greater for trees on Swingle compared with those on Carrizo. Nitrogen concentrations were not significantly different (P=0.05) within tissues on each rootstock (data not shown), thus differences in percentage of total N mass of each tissue category were due to differences in biomass. Biomass Changes with Increase in Tree Size Experiments 1 and 2

Data from experiments I and 2 were combined to determine relationships

between dry weight and N accumulation and tree size indices such as TCV or TCSA. TCV increased linearly as TCSA increased (Fig. 3-1). Likewise, total leaf area per tree was linearly proportional to both TCV (Fig. 3-2A) and TCSA (Fig. 3-2B). Leaf area









Table 3-4. Nitrogen accumulation and allocation between tree components for mature 'Hamlin' orange trees as affected by year of sampling, rootstock, and interaction of year and rootstock.

Nitrogen Accumulation Nitrogen Allocation
Total Above Below Total Total Branches Total Roots
Mass Ground Ground Leaves Twigs Sm. Med. Lrg. Total Trunk Sm. Med. Lrg. Tap
(kg N tree") (% of total weight)
Year 1 0.81 0.60 0.21 38.3 8.2 10.C 6.0 9.0 24.9 2.01 10.1 6.1 5.5 4.8
Year 2 0.77 0.57 0.20 36.5 8.3 8.6 4.8 10.8 24.2 2.3 10.1 7.0 3.6 4.9
NSZ NS NS NS NS NS NS NS NS NS NS NS ** NS

Carrizo 0.87 0.65 0.22 36.4 8.5 9.2 5.4 11.5 26.1 2.3 10.1 7.4 4.2 3.5
Swingle 0.69 0.50 0.19 38.4 8.0 9.2 5.3 8.2 22.6 2.1 10.1 5.6 4.8 6.4
NS NS NS NS NS NS NS NS NS NS NS NS

Car Yrl 0.87 0.66 0.21 38.1 8.8 9.4 5.6 11.6 26.6 2.2 9.4 7.3 4.9 2.7
CarYr2 0.86 0.64 0.22 34.7 8.1 9.1 5.1 11.4 25.6 2.5 10.7 7.6 3.4 4.4
Swi Yrl 0.71 0.50 0.21 38.6 7.4 10.9 6.5 5.0 22.3 2.0 11.1 4.4 6.4 7.9
Swi Yr2 0.68 0.51 0.17 38.2 8.4 8.0 4.4 10.3 22.7 2.2 9.5 6.4 3.7 5.5
NS NS NS NS NS NS NS NS NS NS NS NS NS


"NS = not significant, = significant P<0.05, and ** = significant P<0.0



tA





52







160 ISO

E so Y = 10.09+ 3.19X R2= 0.87
140

120- A


0 80


40 HamlnlCarzo

2 oD A Experiment 2 I

0 10 20 30 40 50 Canopy Volume (m3)



Fig. 3-1. Tree canopy volume as a function of trunk cross sectional area for trees
from experiments I and 2.





53



160

14 Y =0.96 + 3.18X A
12D = 0.94 A

60

a A
o
R HamliCarr izo
40 a Hamin/Swingle
2 A Experimet 2

0
0 10 20 30 40 50 Canopy Volume (m3)




10D

140 a
Y = -5.30 + 0.83X
12o R I = 0.93






40- HramlirCafrizo

2D A xperiert 2

0 20 40 60 8 100 120 140 160 180 Tnnk Cross Section Area (an Fig. 3-2. Leaf area expressed as a function of tree canopy volume (A) and trunk cross
section area (B).





54

index on a tree basis (LAIr) increased rapidly from 4 to 10 as TCV increased from 2 to 10 m3 (Fig. 3-3A) and TCSA increased from 20 to 80 cm2 (Fig. 3-3B). Little increase in LAla was observed with increasing TCV and TCSA beyond 10 m3 and 80 cm2, respectively. Likewise, leaf area index on an acre basis (LAI,) increased from 1 to 6.3 for the same ranges of TCV and TCSA.

Citrus dry weight accumulation for all tree components increased linearly with increased TCV and TCSA (Fig. 3-4 and Fig. 3-5). Regression coefficients, r2, and root mean square error (RMSE) values for dry weights of all tissues vs. TCV and TCSA are provided in Tables 3-5 and 3-6, respectively. Coefficients of determination (r) were generally higher for each tissue category when compared with TCV than with TCSA. Biomass weights for twig, trunk and root categories varied greatly, resulting in lower r2 and higher RSME values. The medium branch masses varied more than the small or large branch categories, possibly indicating inconsistent and/or incomplete separation of tree components into appropriate diameter ranges. Correlations of trunk and taproot weights with TCV and TCSA were poor compared with those for other tree components. Variation in dry matter allocation to root and tap root were apparently due to differences in root density distribution of the two rootstocks used in this study.

Dry matter accumulation in above-ground biomass increased from 60 to 75% across arange of 5 to 40 mi and 10 to 160 cm2 for TCV (Fig. 3-6A) and TCSA (Fig. 37A), respectively. Leaf biomass declined from approximately 20% of total biomass for trees with TCV of less than 5 m3 and for TCSA values below 20 cm2, to approximately 12% of total biomass for trees with TCV and TCSA values greater than 30 m3 and 160





55



18
Y = -2.32 + 12.98(1-exp O A
14 R2 = 0.63

X 12
V A
Slo
-" 6. Hamlin/Canizo a HaminlSwingle
4 A Experiment 2


0 10 20 30 40 50 Canopy Volume (m3)




14
Y = -0.73 + 11.79(1-exp"X) a
12 R2 = 0.70 A

10

ca *


8
Hamlin/Canizo 4- A a Hamin/Swingle A Experiment 2

0 20 40 60 80 100 120 140 160 180 Trunk Cross Section Area (cm2)
Fig. 3-3. Leaf area index on a tree basis expressed as a function of tree canopy volume
(A)and trunk cross-section area (B).











140 A
120
100
80 o 60
40 20




12
10 o oo

6.

4
2 a
2
S70



450
0
30
20
10





15 a a 10 a
0


0 2D w


Canopy Volume (m3)



Fig. 3-4. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) accumulation
as a function of tree canopy volume for 'Hamlin'/Carrizo- experiment 1
(O),'Hamlin'/Swingle(n),and experiment 2 (A) trees.






57




140
120 A 100



40 A 20


14 B 12 10 0o
8 0


4 a C2a


0
Gor
0 2 0


e5D
040
30
so *



2D o .*


0.



20
10






5
20

0 20 40 60 80 10t 20140160 10 Trunk Cross Section Area (cm2)


Fig. 3-5. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) acunalation as a function of trunk cross section area for 'Hamlin'/Carrizo Experiment 1 (0),
'Hamlin'lSwingle Experiment 1 (), and Experiment 2 (A) trees.





58


Table 3-5. Linear regression analysis of dry weight and N accumulation in different tree components as related to tree canopy volume (TCV)Z.
Yo a R2 RMSEY P

Dry Weight (kg tree')
Total Mass -2.21 2.79 0.97 0.15 <0.0001 Above Ground -2.09 1.95 0.96 0.17 <0.0001 Below Ground 0.21 0.72 0.91 0.27 <0.0001 Leaves 0.09 0.36 0.95 0.20 <0.0001 Twigs 0.26 0.13 0.76 0.49 <0.0001 Sm. Branches -0.57 0.37 0.90 0.29 <0.0001 Med. Branches -0.31 0.23 0.80 0.43 <0.0001 Lg. Branches -.24 0.36 0.82 0.53 <0.0001 Total Branches -2.47 1.25 0.92 0.27 <0.0001 Trunk 0.54 0.07 0.72 0.36 <0.0001 Sm. Roots 0.41 0.13 0.87, 0.27 <0.0001 Med. Roots 0.32 0.14 0.83 0.33 <0.0001 Lg. Roots -0.24 0.18 0.85 0.37 <0.0001 Tap Root 0.06 0.12 0.60 0.57 <0.0001
Nitrogen Weight (g tree')
Total Mass -3.24 23.41 0.97 0.16 <0.0001 Above Ground -7.28 17.52 0.96 0.17 <0.0001 Below round 5.51 5.36 0.92 0.23 <0.0001 Leaves 3.50 8.70 0.92 0.24 <0.0001 Twigs 0.09 1.83 0.85 0.34 <0.0001 Sm. Branches -3.32 2.04 0.96 0.20 <0.0001 Med. Branches -1.28 1.00 0.78 0.45 <0.0001 Lg. Branches -3.46 2.44 0.88 0.32 <0.0001 Total Branches -11.31 5.71 0.94 0.23 <0.0001 Trunk 2.80 0.35 0.65 0.40 <0.0001 Sm. Roots 5.38 1.85 0.86 0.28 <0.0001 Med. Roots 1.82 1.34 0.88 0.29 <0.0001 Lg. Roots -1.10 0.88 0.89 0.32 <0.0001 Tap Root 0.68 0.55 0.63 0.53 <0.0001

z Y =o +aX where X = TCSA, and Yo and a are regression coefficients

Y RMSE dry weight kg dw treel, and N accumulation = g N tree"1





59


Table 3-6. Linear regression analysis of dry weight and N accumulation in different tree components as related to trunk cross-sectional area (TCSAY.
Yo a R RMSEY P
Dry Weight (kg tree')
Total Mass -7.23 0.70 0.93 0.23 <0.0001 Above Ground -5.67 0.50 0.93 0.24 <0.0001 Below Ground -1.20 0.19 0.89 0.29 <0.0001 Leaves -0.65 0.09 0.94 0.22 <0.0001 Twigs -0.04 0.04 0.68 0.47 <0.0001 Sm. Branches -1.13 0.09 0.82 0.40 <0.0001 Med. Branches -0.84 0.06 0.80 0.43 <0.0001 Lg. Branches -.94 0.05 0.78 0.42 <0.0001 Total Branches -5.17 0.33 0.91 0.30 <0.0001 Trunk 0.38 0.02 0.70 0.36 <0.0001 Sm. Roots 0.10 0.04 0.90 0.24 <0.0001 Med. Roots -0.08 0.04 0.91 0.25 <0.0001 Lg. Roots -0.53 0.04 0.78 0.45 <0.0001 Tap Root -0.19 0.03 0.61 0.57 <0.0001
Nitrogen Weight (g tree"')
Total Mass -47.83 6.00 0.94 0.22 <0.0001 Above Ground -41.41 4.51 0.94 0.23 <0.0001 Below round -5.36 1.41 0.92 0.24 <0.0001 Leaves -15.39 2.32 0.93 0.23 <0.0001 Twigs -3.29 0.46 0.82 0.37 <0.0001 Sm. Branches -6.75 0.49 0.89 0.31 <0.0001 Med. Branches -3.52 0.27 0.77 0.47 <0.0001 Lg. Branches -5.69 0.47 0.85 0.38 <0.0001 Total Branches -23.36 1.50 0.92 0.27 <0.0001 Trunk 1.97 0.10 0.82 0.39 <0.0001 Sm. Roots 0.92 0.51 0.89 0.25 <0.0001 Med. Roots -1.48 0.38 0.91 0.25 <0.0001 Lg. Roots -2.65 0.21 0.82 0.40 <0.0001 Tap Root -0.23 0.13 0.58 0.56 0.0002 zy = yo +ax where x = TXA, and yo and a are regression coefficients Y RMSE dry weight kg dw tree"', and N weight = g N tree'





60

c2, respectively (Figs. 3-6B and 3-7B). Likewise, twig biomass decreased from I 11 to 6% of total biomass for trees of the corresponding size categories. Total branch dry weight increased from 15 to 45% as trees matured, while trunk biomass decreased from 12% for young trees to 3% for mature trees (Figs. 3-6C and 3-7C). Few consistent trends were found when comparing root biomass with tree size (Figs. 3-6D and 3-7D), which may be due to problems in recovering all root tissue or differences in root biomass distribution between the two rootstocks used in this study. Nitrogen Distribution

Mature Citrus Tree Nitrogen Distribution Experiment 1

As was the case with dry biomass, total N accumulation was greater for trees on Carrizo rootstock (870 g tree"') than for trees on Swingle (690 g tree''). Total and aboveground N accumluation were significantly (P=0.01) affected by TCV and TCSA, whereas below-ground N accumulation was not (Table 3-4). Rootstock, year, and the interaction of rootstock and year effects were not correlated with TCV or TCSA. Nitrogen concentration was not significantly different compared with TCV, TCSA, rootstock, or year for any of the tissues sampled (Table 3-7). Therefore, N content trends were similar to those for dry weight. Percentage tissue N concentration compared with total tree N weight was not significantly different for tree size, rootstock, or year with the exception of branches greater than 30 mm in diameter. Trees grown on Carrizo citrange had significantly greater (P=0.05) N in larger branches and total branches compared with TCV (Table 3-4).






61

90
80
70 A _._.60
50
40 a 30 a 20 a O 10
0

20* B

15

10
AO .



30
5 0





040

30

20

10 A

0 .
30


20

15- a 10 o A D 00

0 5 10 15 20 25 30 35 40 45 Canopy Volume (m3)

Fig. 3-6. Dry weight allocation to above-ground (dosed), and below-ground (open)
biomass (A); leaf(closed) and twig (open) biomass (B); total branch (closed) and trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D)
as a function of canopy volume for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 (0), and experiment 2 (A) trees.






62


90
80A 70


50
40 AD 0 0 30


10

20 A B 15

10








10A0
0





30

20

20 a
.*
1 "o n 30
20


A A




5 o
0 .

0 20 40 60 80 100 120 140 180 180 Trunk Cross Section Area (c 2) Fig. 3-7. Dry weight accumulation to above-ground (closed), and below-ground (open)
biomass (A); leaf (closed) and twig (open) biomass (B); total branch (closed) and trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D) as a function of tnmk cross section area for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 (0), and experiment 2 (A) trees.





63

Table 3-7. Mature citrus tree tissue N concentration as a function of year sampled and rootstock.

Plant Tissue Year Sampled Rootstock
2001 2002 Carrizo Swingle

Leaves 2.47 2.32 2.47 2.46 Twigs 1.28 1.00 1.09 1.14 Small Branches Bark 1.16 1.24 1.22 1.11 Small Branches Wood 0.30 0.31 0.33 0.28 Med. Branches Bark 1.11 1.28 1.13 1.08 Med. Branches Wood 0.29 0.32 0.30 0.28 Large Branches Bark 1.16 1.33 1.06 1.30 Large Branches Wood 0.38 0.34 0.42 0.33 Trunk Bark 1.29 1.36 1.33 1.14 Trunk Wood 0.42 0.44 0.43 0.42 Fibrous Roots 1.55 1.38 1.61 1.51 Medium Roots 1.04 0.85 1.10 0.97 Large Roots 0.48 0.47 0.50 0.47 Tap Root 0.45 0.45 0.47 0.39

Statistical Significancez
Year NS Rootstock NS Year X Rootstock NS

z NA = No significant difference at the P=O. I level.

Nitrogen Balance

Leaf and twig dry weight accumulation during the previous 12 months was significantly

greater (P=-.05) for trees grown on Carrizo (3963 and 5717 g tree"', respectively)

compared with trees grown on Swingle (3312 and 4525 g tree"', respectively). Total N

content for these tissues were 145 and 127 g tree' for Carrizo and Swingle, respectively.

Mean fruit N accumulations for the two rootstocks were 302 and 258 g tree for trees on

Carrizo and Swingle, respectively. Assuming a conservative 5% increase in N

accumulation in all tissues other than leaves and twigs due to increase in biomass in

2001, the resulting increase in N content was 24 and 18 g tree' for Carrizo and Swingle





64

trees, respectively. Therefore, the total estimated increases in N content for 2001 were 471 and 403 g tree' for trees grown on Canizo and Swingle rootstocks, respectively. The amount of fertilizer N applied in 2001 was approximately 503 g tree"', resulting in apparent fertilizer N uptake efficiencies (FNUE) of 93.6 and 80.1% for Carrizo and Swingle rootstocks, respectively. However, if the 58 g tree"' ofN contained in the 1273 mm ha-1 of reclaimed water applied over the 12-month period is considered, NUE decreases to 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively. Nitrogen Change with Increase in Tree Size Experiments 1 and 2

Nitrogen accumulation increased linearly with increasing TCV and TCSA from 5 to 40 m3 and 20 to 160 cm2, respectively (Figs. 3-8A and 3-9A). Total leaf N mass increased from less than 30 to more than 250 g N tree' across the range of TCV and TCSA measured (Fig. 3-8B and 3-9B). These increases were about 45% of total tree N for trees with TCVs less than 5 m3 to 37% for trees with TCVs greater than 35 m3 (Figs.3-10B and 3-1 1B). Twig N ranged from less than 10 to greater than 50 g tree-1 across the range of trees measured (Figs. 3-8B and 3-9B), but twigs still contained a consistent 9% of total tree N regardless of tree size (Figs. 3-10B and 3-11B). Total N accumulation by branches increased from less than 10 to greater than 200 g N tree- for trees with TCV of less than 5 and greater than 40 m3, respectively (figs. 3-8C and 3-9C), which corresponds to an increase in percentage of total tree N from 6 to 27% for corresponding tree sizes (Fig. 3-10C and 3-11C). The proportion of total tree N in the tnmk decreased from 5 to 3% (Figs. 3-10C and 3-11C). Regression coefficients, R2, RSME, and probability values for dry matter and N accumulation within each tissue type are presented in Tables 3-5 and 3-6.






65




A lo00



800 a o 400

200
A
0
B
400














200





50

0 o 100 o o 000 0 50







200






0 5 10 15 20 25 30 35 40 45 Canopy Volume (m3) Fig. 3-8. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and tunk (open)
N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of canopy volume for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 (0), and experiment 2 (A) trees.






66




lo A 100
WOOW 600A 400 A 200

0
B
400 *







0
0
300 C 200
200 o o o*







150

100 A 50
0
200 **










150 100

50
50 o O o


0 20 40 60 80 100 120 140 160 180 Trunk Cross Section Area (cm2)



Fig. 3-9. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and trunk (open) N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of trunk cross section area for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 (), and experiment 2 (A) trees.






67



A
80 0 so *

60

40
20 0
D O


0

30 W B



20
30











15



0
so D 30 C






25 .
20 A 15
0
30
151


5 on

0 5 10 15 20 25 30 35 40 45

Canopy Volume (m3) Fig. 3-10. Nitrogen allocation to above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (dosed) and taproot (open) (D) as a
function of canopy volume for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 (0), and experiment 2 (A) trees.






68




go"A

70 60
50
40
30, D O 20 a o 10
0


50 A A A
40 30 20








2D0 15





30 D
20





15

10
25 20a



Ao
15 a

0

0 20 40 60 80 100 120 140 160 180 Trunk Cross Section Area (cn2)



Fig. 3-11. Nitrogen allocation of above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (closed) and taproot (open) (D) as a function of trunk cross section area for 'Hamlin'/Carrizo experiment 1 (0),
'Hamlin'/Swingle experiment 1 ([), and experiment 2 (A) trees.





69

Discussion

The mature 'Hamlin' trees used in this study were planted at the same time and received the same horticultural inputs (i.e. fertilizer rates, irrigation schedule, and pest control) for the past 14 years. Total leaf area of each tree differed with tree size, but LAI and LAI, appeared to approach maximum values of 10 and 6.5, respectively. The mean LAl value of 10 is well within the 9 to 11 range found by Syvertsen and Lloyd (1994) for mature citrus trees. These values are much higher that the 3 or less associated with agronomic row crops (Flenet et al., 1996). Citrus is thought to have developed as an understory plant in subtropical rainforests and so has a high tolerance to shade (Syvertsen and Lloyd, 1994) and therefore developed a dense canopy. A large fraction of citrus leaves found within the inner canopy receive 10% or less of recorded on the outermost leaves (Cohen et al., 1987) The observed LAI would indicate that, on average, 10 layers of leaves exist over each unit area of soil under the tree canopy. This finding has significant implication for light interception and photosynthesis, indicating that citrus is an efficient interceptor of light and allows very little of it to strike the soil surface. Leaves in the interior of a citrus tree are adapted for low light levels and tend to be thinner and flatter than exterior leaves (Mills et al., 1999).


Percentages of total biomass and N in leaf, branch and root tissue compared well with 10 and 15 year old trees harvested by Cameron and Appleman (1935) and Cameron and Compton (1945). However, trees grown on Swingle rootstock were significantly smaller than those grown on Carrizo rootstock. Even though tree size was different, dry weight allocation between tree components remained relatively constant. However, there were significant differences in dry weight accumulation in large branch and total branch





70

biomass weights for the two rootstocks in this study. Trees grown on Carrizo citrange rootstocks were larger than those of the same age grown on Swingle citrumelo, therefore the difference in large and total branch weights were correlated with tree size as measured by TCV and TCSA. Hence, the percentages of total biomass for specific tree components were similar for both rootstocks, indicating that above ground biomass is partitioned equally based on the relationship of total biomass to tree size. Therefore, if the relationship of total biomass to tree size is known, and the relationship of component partitioning to tree size is also known, then the biomass of each component part can be, estimated regardless of rootstock effects on tree size.

The biomass associated with individual tree trunks was related to the height of the crotch formed by the main scaffold limbs of the tree. In citrus nurseries, crotch height is relatively uniform. However, the larger trees used in this study were affected by the freezes of the late 1980s, most notably in 1989. The limb structure of the "Hamlin" trees used in the mature tree portion of this study had to be re-grown, in some cases requiring the entire scaffold limbs and a portion of the upper trunk to be removed. Thus the heights of the trunk to the scaffold limbs were not consistent.

Tree size, biomass, N weight, and apparent FNUE were greater for trees grown on Carrizo citrange rootstock compared with trees grown on Swingle citrumelo, indicating that differences in tree size are directly correlated with FNUE. Differences in FNUE may be due to physiology of the rootstocks themselves or due to the distribution of fibrous roots associated with the various rootstocks. The distribution of root length densities and N uptake rates for the two rootstocks in this study will be the subject of Chapters 4 and 6, respectively.





71

Assuming that 30% of the N accumulated in new growth tissues came from

fertilizer (Dasberg, 1987; Feigenbaum et al., 1987; Legaz et al., 1982), it is concluded that 124 and 108 g of accumulated N originated from the fertilizer inputs. The calculated FNUEs of 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively, in this study were similar to the 61 to 83% reported by Syvertsen and Smith (1996) for 4-year old grapefruit trees grown in lysimeters. However, mineralized soil N was not included in the estimation. The contribution of soil organic matter and abscised tree parts will be addressed in Chapter 6. Accounting for these sources of N will reduce overall NUE for the citrus trees in this study.

Citrus trees at the Conserv II site were of the same age and similar in size, but the K.D. Revell grove operated by Cargill, Inc. contained trees of various ages due to past replanting of trees. The water and nutrient holding characteristics of the soil at this site were similar to those of the soil at Conserv II. Therefore, it was assumed that trees grown at this site would follow similar biomass and N partitioning characteristics. Total leaf area for ji bWients increased linearly with TCV and TCSA. LAI increased rapidly with tree size until the trees were approximately 3-4 years old, after which it stabilized at about 10. This information can be used to parameterize light interception functions for a citrus tree photosynthesis and growth model.

Significant relationships were found between total tree fresh and dry biomass and tree size. The ratio of above-ground to below-ground biomass and N content ranged from a low of 3:2 to a high of 3:1 across the range of tree sizes found in this study. Citrus roots of mature citrus trees extend below 1.8 meters (Castle, 1978 and 1980; Elezaby, 1989; Menocal-Barverena, 2000), but the root system in this study was excavated to a depth of





72

only 1 m. Therefore, the ratio of above-ground to below-ground biomass may have remained near 3:2 if roots extending below 1 m were included in the total root biomass measurement.

Citrus trees increase in size with time; branches increase in diameter through

accumulation of xylem tissue, eventually developing a scaffold branch structure of largediameter branches. The relationship of total tree biomass and N weight to TCV and TCSA followed a linear function indicating a constant rate of accumulation with increase in tree size as measured by both TCV and TCSA. This result implies that the partitioning of biomass and N accumulation in all plant parts occurs at rates specific to the tree component. Therefore, the total biomass and/or N weight of a citrus tree can be estimated for any tree size.

Percentage biomass and N weight of woody tree parts (large branches and trunk) increased, while those of leaves and twigs decreased with increase in TCV and TCSA. Caruso et al. (1999) found similar results in peaches where the relative proportion of leaf and twig dry weights decreased with tree age. It can be concluded that to support the increase in total tree weight, the biomass and N content of woody branches and trunk increases at a higher rate compared with leaves and twigs. However, it can be concluded that LAIt is the driving factor in leaf and twig biomass accumulation since the ratio of leaf area to ground area under the canopy remained constant with increase in tree size for medium and large trees. Thus, once the total biomass and N weight of a tree is estimated, the weights of individual tree parts can be estimated based on tree size. Regression equations such as those in Tables 3-5 and 3-6 can be used to simulate biomass and N partitioning in a citrus growth model.





73

Conclusions

Leaf areas of both young and mature citrus trees were correlated with tree size as measured by TCV and TCSA. Leaf area index increased rapidly for young citrus trees and then equilibrated at approximately 10 by age 3 to 4 years. This information is valuable for the estimation of citrus light interception and total photosynthesis. Change in citrus tree dry weight and N content of the two citrus scions in these studies was shown to be a linear function of TCV and TCSA. Partitioning of biomass and N decreased for leaves and twigs, increased for branches, and remained constant for trunk and taproot tissues with increase in tree size. While mature citrus trees grown on Swingle citrumelo rootstock were consistently smaller than trees of similar age grown on Carrizo citrange, mass partitioning of tree parts were similar for both rootstocks. Thus, with the exception of spatial root length density distribution described in chapter 4, the only effect of the two rootstocks and two scions used in these studies was on tree size relative to TCV and TCSA. Therefore, biomass and N partitioning for specific tissues with tree size can be captured in generic linear relationships. The N balance estimated for mature citrus trees in this study indicated an apparent fertilizer N use efficiency of 60 to 70%.














CHAPTER 4
CITRUS ROOT GROWTH DYNAMICS Introduction

While the role of roots in anchoring crop plants, particularly tree crops, to the soil should not be taken for granted, the function of roots as absorbing organs for both water and nutrients can not be overemphasized. The structure of a root system is important in determining the pathway and resistance to water and solute uptake, and the volume of soil accessible to crop plants (Kramer and Boyer, 1995). The entrance of water and nutrients into young roots occurs a few cm behind the root tips because of the lack of a functional xylem at the tip and the suberization of root hypodermis and endodermis tissues with age (Tinker and Nye, 2000). Thus, the larger the length of relatively small diameter fibrous roots a crop root system has, the greater the amount of water and nutrients available to it. Likewise, the larger the soil volume a crop root system occupies, the greater the pool of water and nutrients available for uptake.

The goal of fertilizer application should be the placement of nutrients within the crop root zone to insure the most efficient uptake. Maintenance of adequate water and labile nutrient concentrations within soil zones occupied by the crop root system is essential for optimal nutrient uptake. Understanding the spatial distribution of fibrous roots is essential to ensure proper fertilizer placement to improve nutrient uptake and potentially reduce leaching below the root zone.



74





75

Several Florida studies have demonstrated that tree size and yield were related to fibrous root density and/or distribution (Castle and Krezdorn, 1975; and Ford 1954a; 1964; 1968; 1969; 1972) in the deep sandy soils of central Florida. Since processes that control non-point source pollution are dynamic and greatly affected by genetic traits and environmental conditions, development of models that can integrate interactive effects of spatial and temporal processes will be critical. However, models providing realistic results will need to include accurate information on root growth dynamics. Modeling of citrus root distribution and the determination of water and nutrient uptake parameters can lead to the development of an expert system for the estimation of water and nutrient depletion and uptake by soil depth. Likewise, nutrient leaching can be estimated due to excessive irrigation or heavy rainfall.

Under Florida growing conditions, the quantity of fibrous roots decreased with depth and lateral distance from the trunk (Elezaby, 1989; Menocal-Barberena, 2000). Eighty percent of citrus fibrous roots were found within a 120 cm radius of the tree trunk and 40%0 in the upper 30 cm of well-drained sandy soils. Nearly all citrus roots grow within 45 cm of the soil surface where artificial drainage was provided and/or high water tables occured (Calvert et aL, 1967; 1977; Ford, 1954a; Ford, 1972; Reitz and Long, 1955).

Fibrous root dry mass densities ranged from 300 to 1200 g m3 (Castle, 1978; 1980). Citrus fibrous root length densities ranged from 0.53 cm cm3 for 'Swingle' citrumelo roots to 2.02 cm cm3 for trifoliate orange (Eissenstat, 1991). Elezaby (1989) reported fibrous root concentration in the 0 to 30 cm soil zone increased from 450 to 1000 g m between trees when the in-row distance decreased from 4.5 to 2.5 m due to





76

overlapping root systems.

Castle and Krezdorn (1975) described two general types of root systems, the first characterized by extensive lateral and vertical development, and the second with intensive higher fibrous root density near the soil surface. Trees on rough lemon, 'Volkamer' lemon and 'Palestine' sweet lime (C. limettioides Tan.) typified the extensive type of root system where 50% of the fibrous roots occurred below 70 cm in the soil and produced large, highly-productive trees that dominated the citrus industry in Florida when trees were irrigated less intensively and were set at much lower densities. SUnfortunately, rough lemon has been virtually eliminated as a commercial rootstock due to citrus blight disease. Examples of the intensive-type root system were Carrizo citrange and Swingle citrumelo that had few fibrous roots below 70 cm, and the root system was less developed laterally. These rootstocks now dominate the citrus industry in Florida and are well suited for high-density, intensively irrigated plantings.

The following hypotheses were tested: 1) Root distribution is significantly

affected by rootstock, and 2) Generic relationships can be developed for well-drained soils that describe citrus root densities at various depths from the soil surface and distances from the tree as a function of tree size. To test these hypotheses, the objectives of this study were to: 1) develop information on spatial root length distribution at different soil positions and depths for two citrus rootstocks, and 2) develop functional relationships that define root length densities at various soil positions and depths as a function of tree size. The relationship of vertical and horizontal root length density distribution to tree size resulting from this study can be used to estimate fibrous root densities in various soil layers for citrus water and nutrient uptake models. This





77

relationship will provide a scientific basis for the development of water and nutrient components of an expert system for improved citrus irrigation and N management.

Materials and Methods

Sample Collection

The same 19 'Hamlin' and 'Valencia' orange trees used in the previous biomass and N distribution study (Chapter 3) were used to determine the spatial relationship between citrus root length density and tree size. Soil cores were removed once the trees had been cut to the ground but prior to the excavation of the main root system. Cores were taken with a 7.6 cm diameter bucket auger and roots were sampled at 50, 100, and 150 cm from the tree trunk in the row and 50, 100, 150 and 200 cm between tree rows. Samples were collected at 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 90 cm depths. Each sample was placed into separate plastic bags, sealed, and marked with tree identification, depth and distance from the tree. The samples were placed in a cooler containing ice and were subsequently frozen at -4 C. Sample Processing and Statistical Analysis

Roots were removed from the soil by washing though an 850 an sieve. Any

debris not passing through the sieve was removed manually, and the roots were separated into size categories by diameter. These categories were <4 mm, 4 to 20 mm and >20 mm. Root lengths for roots 0 to 4 mm in diameter were determined prior to drying using the line intersect method (Newman, 1966). Root length density data from samples collected from the 12 trees at the Water Conserv II site (mature tree study) were analyzed by the general linear model procedure of SAS (SAS Institute, Inc., Cary, NC). Root length density data from the soil samples collected from the trees of various sizes at the Cargill





78

grove were analyzed using Proc REG in SAS. Regression equations were determined using SigmaPlot (SPSS, Inc., Chicago IL).

Results

Mature 'Hamlin' Orange Root Distribution

Soil depth and distance from the tree trunk significantly (P = 0.01) affected citrus root length density (Table 4-1). Mean root length density of fine fibrous roots (<4 mm) extracted from soil cores surrounding the 12 mature citrus trees followed a bimodal spatial distribution with depth from the soil surface (Fig. 4-1), and distance from the tree trunk (Fig. 4-2). Mean fine fibrous root density in the upper 15 cm was 1.04 cm cm3. Densities ranged from 1.9 cm cm3 soil at 50 cm from the tree trunk to 0.7 cm cm'3 at 200 cm. Mean densities decreased at the 15 to 30 cm depth to 0.30 cm cm"3 and ranged from

0.5 to 0.07 cm cm"3 at 50 and 200 cm distances, respectively. Mean densities of fine fibrous roots increased at depths below 40 cm to a maximum at the 60 to 75 cm depth (0.28 cm cm3) then declined at the 75 to 90 cm depth (0.27 cm cm'3). Densities at the 60 to 75 cm depth were 0.3, 0.3, 0.3, and 0.03 cm cm"3 at distances of 50, 100,150, and 200 cm from the tree trunk, respectively.

Fine fibrous root densities at the 0 to 15 cm depth were generally greater in the inrow orientation than in the cross-row orientation (data not shown). Mean in-row spatial root length densities (0.41 cm cm"3) were greater, but not significantly different from densities for between-row orientation (0.35 cm cmn3) (Table 4-1), because more overlap in root systems from adjacent trees probably occurred in this orientation.








Table 4-1. Mean fibrous (diameter <4 mm) root length density of mature 'Hamlin' orange tree as affected by rootstock, orientation distance, and soil depth.

Mean Root Length Density (cm cm"3) Rootstock Orientation Distance from tree (cm) Soil depth (cm)
Carrizo Swingle In-row Cross-row 50 100 150 200 0-15 15-30 30-45 45-60 60-75 75-90
0.36 0.41 0.41 0.35 0.49 0.40 0.33 0.17 1.04 0.30 0.16 0.24 0.28 0.27

P Significance" Rootstock 0.290 NS Orientation 0.253 NS Distance 0.002 *** Depth <0.0001 *** Distance*Depth 0.820 NS Rootstock* Distance 0.052 Rootstock*Depth 0.002 *** ZSignificance: NS = not significant, = significant at P=0.1 level, ** = significant at P=0.05 level, and *** = significant at
P-0.0 llevel.






80


0-



20 A



o 40
~ 50 cm Diltnem
--- 100 cm mInc
-r- 150 cm DbMnc
-t- 200cm Diance


80



100
0.0 05 1.0 15 2.0 Root Lenght Density (cm cm")




0



20



40*

S-w- 50 cm Doitance

- 100 cm DiEmnc
eC + 20nm oanc


80 B




0.0o 0.5 1.0 1.5 2.0 Root Length Density (cm cm-)








Fig. 4-1. Root length density distribution by depth at 50, 100, 150 and 200 cm distances
from the tree trunk between rows of 'Hamlin' orange trees on Carrizo citrange
(A) or Swingle citrumelo (B) rootstocks.














2.0

0-15 cm Depth
-- 15-30 cm Depth
-..- -- 30-45 cm Depth
1.5 -o-- 45-60 cm Depth
o --o- 60-75 cm Depth E -- 75-90 cm Depth


1.0




S0.5




0.0
40 60 80 100 120 140 100 180 200 220 Distance (cm)


2.0

0-15 cm Depth
-- 15-30 cm Depth S-~- 30-45 cm Depth
1.5 --- 45-60 cm Depth E--o-- 0-75 cm Depth o --- 75-90 cm Depth


1.0




1 0.5
0
-J




0.0 1
40 80 80 100 120 140 160 180 200 220 Distance (cm)







Fig. 4-2. Root length density distribution at 0-15, 15-30, 30-45, 45-60, 60-75, and 75-90
cm depth increments by distance from the tree trunk as affected by distance from
the tree trunk for 'Hamlin' orange trees on Carrizo citrange (A) and Swingle
citrumelo (B) rootstocks.





82


Root densities at the 50 cm distance in the cross-row orientation decreased more gradually than did densities at 100, 150, and 200 cm distances. Minimum densities occurred at the 45 to 60 cm depth for the 50 cm distance as opposed to the 30 to 45 cm depth for the 100, 150, and 200 cm distances. Similarly, root densities increased at the 60 to 75 cm depth for the 100 and 150 cm distances, and 75 to 90 cm depth for the 50 cm distance. Spatial root distribution differences between rootstocks were not statistically significant (Table 4-1). Mean root length densities at all depths and distances were 0.36 cm cm"3 for trees grown on Carrizo citrange and 0.41 cm cm" for trees grown on Swingle citrumelo. However, the interaction of rootstock and depth was significant at the P=0.01 level. Trees on Swingle had higher root length densities near the soil surface than did trees on Carrizo (Figs.4-1 and 4-2). Conversely, root length densities were greater for trees on Carizo between 15 and 75 cm below the soil surface.

Root length densities for the 0 to 15 cm depth ranged from 2.0 to 0.9 cm cm3 soil at distances of 150 cm or less for trees on Swingle rootstock. Densities ranged from 1.2 to 0.7 cm cm"3 at the same depth and distances for trees on Carrizo rootstock. Root densities decreased for both rootstocks to low values at the 30 to 45 cm depth. Root densities increased for trees on Carrizo at 45 to 60 cm depth, whereas densities for trees on Swingle increased at the 60 to 75 and 75 to 90 cm depths. With the exception of the 50 cm distance from the tree trunk, root densities were greater for trees on Carrizo at 30 to 45 cm than those on Swingle. Likewise, root densities were greater at 45 to 60 cm depth for trees on Carrizo than at the 60 to 75 cm depth for trees on Swingle.





83

Root Length Density Distribution Changes with Tree Size

Citrus root length densities were significantly different at the P=0.01 level for both distance from the tree trunk and depth from the soil surface across a wide range of tree sizes (Table 4-2). Three-dimensional graphical representations of developing root systems are presented in Fig. 4-3. These graphs represent trees approximately 2 to 5 years old (Figs. 4-3A and 4-3B), 5 to 10 years old (Figs. 4-3C and 4-3D), 10 to 15 years old (Figs. 4-3E), and >15 years old (Figs. 4-3F). Root systems were initially concentrated at the surface with few roots deeper than 0.5 m at a distance of 150 cm from the tree trunk. As the citrus trees produced substantial fruit (5 to 10 years of age) root length density increased at the soil surface to the dripline of the tree. Roots eventually extended to the 200 cm distance between tree rows and to a depth of 0.9 m at 150 cm from the trunk. The bimodal nature of the root system can be seen near the tree at depths below 60 cm. By the time the tree reached 10 to 15 years of age and the canopy was nearing full hedgerow dimensions, the bimodality of the root system was fully developed and roots extended past a depth of 1 meter at all distances from the tree.

Tables 4-3 and 4-4 list the regression coefficients for a third order polynomial

relationship of canopy volume and trunk diameter to root length density at all depths and distances. The r2 values were greater, and RSME and P values were lower for most regressions using canopy volume than those using trunk diameter, indicating that canopy volume measurements are a more accurate predictor for assessing root length density compared with trunk cross sectional area.






84



A conXveWnedZmff B caowva-re m



M me ( "O
1.6 1.0






to
000 D
I ** 1E CanpyVoraM -.M7 D Canapvakow-,43ff
" o -.--.. --W of
09an / m












3" A
is I1




0 9


















old (E and > 15 years old (F
3m 00010 m 0000~r



ol () aid > 15 yews old (F ).u# iaartn~





85


Table 4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length densities for trees ranging from 2 to > 15 years old.


RMSE CV R2 P Significance
Orientation 44.66 87.81 0.002 0.291 NS Distance 48.08 42.54 0.12 <0.0001 *** Depth 23.45 51.99 0.17 <0.0001 ***


Table 4-3. Regression coefficients and statistics root length density as a function of
distance from the tree trunk, and soil depths by canopy volume using a third order
quadratic polynomial modelz.

Distance Depth Yo a b c R2 RMSE P
(cm) (cm) (cm cm) 50 0-15 1.29 -0.021 0.002 0.00001 0.84 0.33 0.001 50 15-30 0.99 -0.289 0.027 -0.0006 0.78 0.15 0.005 50 30-45 0.27 -0.059 0.006 -0.0001 0.68 0.09 0.021 50 45-60 0.45 -0.16 0.015 -0.00003 0.83 0.12 0.002 50 60-75 0.29 -0.099 0.009 -0.00002 0.68 0.09 0.022 50 75-90 0.31 -0.14 0.013 -0.0003 0.88 0.06 0.0005 100 0-15 0.73 -0.23 0.023 -0.0005 0.79 0.15 0.005 100 15-30 0.44 -0.13 0.014 -0.0003 0.61 0.15 0.072 100 30-45 0.19 -0.069 0.007 -0.00002 0.48 0.14 0.14 100 45-60 0.20 -0.082 0.008 -0.00002 0.82 0.05 0.002 100 60-75 0.071 -0.068 0.006 -0.00001 0.85 0.04 0.001 100 75-90 0.040 -0.018 0.002 -0.00001 0.51 0.04 0.107 150 0-15 0.17 -0.087 0.011 -0.0001 0.97 0.06 <0.0001
150 15-30 0.10 -0.047 0.005 -0.0001 0.81 0.06 0.003 150 30-45 0.007 -0.004 0.001 -0.00001 0.74 0.02 0.010 150 45-60 0.025 -0.012 0.001 -0.00001 0.80 0.03 0.003 150 60-75 0.007 -0.004 0.0004 -0.000001 0.75 0.01 0.008 150 75-90 0.008 -0.004 0.0003 -0.000001 0.65 0.01 0.030 200 0-15 0.042 -0.020 0.002 -0.0001 0.99 0.01 0.014
200 15-30 -0.013 0.007 -0.001 -0.0001 0.99 0.01 <0.0001
200 30-45 -0.0003 0.001 -0.0001 0.00001 0.99 0.01 <0.0001 200 45-60 -0.002 -0.001 0.0001 0.00001 0.99 0.01 0.004 200 60-75 0.003 -0.001 0.0001 -0.000001 0.99 0.01 0.004 200 75-90 -0.001 0.001 -0.0001 0.00001 0.99 0.01 0.0002

Y=Yo+aX+bX2+cX3 where X = TCV, and Yo, a, b, and c are regression coefficient.




Full Text
154
weights were 18.8 and 21.9 g tree'1 for the same years. Fruit and leaf biomass losses were
greatest in April through May and September through December of each year. Little loss
of either of these tissues occurred from June through August. Cumulative fruit biomass
varied more than bloom and leaf biomass, with 1635 and 946 g tree'' removed in 2001
and 2002, respectively. These fruit biomass values represented a cumulative loss of 30.1
and 16.8 g tree'1 of N for these years. Leaf biomass losses were 2041 and 2806 g tree'1 for
2001 and 2002, respectively. Cumulative annual N loss from leaf fall amounted to 42.2
and 51.4 g tree'1 for the same years.
Citrus Decision Support System
Due to the complexity of grower decision-making processes, researchers have
developed computer based decision support systems (DSS) to provide information on
management options. These DSS store and organize information such as rates of water,
fertilizer, and agrichemicals applied to specific fields and provide information on
predicted future events such as irrigation scheduling and N leaching. Crop models are
used to determine the effect a given management decision will have on the crop such as
growth rate or yield. Crop models such as CROPGRO, CERES and others (Hoogenboom
et al., 1994; Jones et al., 1991; Wagner-Riddle et al., 1997) are process-oriented models
that simulate vegetative growth and reproductive development. The models predict dry
matter growth (Shen et al., 1998), crop development (Batchelor et al., 1994; Batchelor et
al., 1997; Piper et al., 1996) and final yield (Batchelor et al., 1996; Heinemann et al.,
2000) for a range of agronomic crops. Inputs are daily weather data, soil profile
characteristics, and crop management conditions (Gijsman et al., 2002). Such models are
currently being used for a number of purposes such as yield forecasting and long-term


113
rate. Hence, a model of soil water uptake and depletion based on root length densities
would be appropriate for citrus.
The test hypotheses established for this experiment relating water uptake to time
of year, soil water content, and root density has been confirmed. A model based on the
concepts of Kc and Kg to estimate daily citrus water uptake is reasonable. The estimation
of soil water uptake and resulting depletion based on root length density is sound and
would provide a reasonable soil water balance for a nutrient management expert system.
Conclusions
Based on the results from this study it is concluded that ETC can be calculated by
modifying ET values for crop and residual soil moisture conditions using appropriate K
and K* coefficients. Minimum Kc was approximately 0.85 in December and January,
while a maximum of approximately 1.05 occurred during the months of June and July.
Soil water use decreased with soil water content, resulting in Kg values of 1.0 at nearly
15% ASWD to 0.6 at 50% ASWD. With few exceptions, daily soil water uptake per unit
root length density was similar for all soil layers. The best correlation between daily
water use and soil water content was found in the soil volume containing the highest root
length density. Therefore, the hypothesis that soil water uptake relative to calculated
reference evapotranspiration is related to season of year, soil water content and root
length densities was confirmed. Estimation of soil water uptake and resulting soil
depletion based on root length density would allow for a relatively accurate assessment of
soil water depletion, crop water status, and effective soil storage capacity using a layer
soil profile modeling approach. Such approach would allow model users to predict soil
,4
O'
/
c*-


96
between trees (1.5 m). Between row sensors were installed midway between the tree
trunk and the canopy dripline (0.9 m), the canopy dripline (1.8 m), and between the
dripline and midline between tree rows (2.7 m).
Access tubes (5 cm diameter, acrylonitrile butadiene styrene) were installed, and
each sensor was individually normalized following manufacturer recommendations. A
general calibration curve was developed for the soil type at the site using a gravimetric
method described by Morgan et al. (1999). Daily Penman ET0 values from a Florida
Automated Weather Network (FAWN) station located less than 0.4 km away from the
actual field site were recorded.
Estimated Daily ETC
Soil water content in deep sandy soils equilibrates to field capacity within a few
hours after irrigation or rainfall. The net change in 0 over a 24 hour period was calculated
for each sensor using the difference between 0 values recorded at midnight and values
recorded the previous midnight (A0). To avoid hysterisis effects, only 0 data collected on
days receiving no rainfall or irrigation were used for this calculation. Daily soil water
depletion depth was calculated for each sensor by multiplying the A0 by a corresponding
soil depth for that sensor (Fig. 5-1). The resulting soil water depths were then multiplied
by the surface area assigned to the given probe (Fig. 5-2). The resulting soil water
depletion volumes were summed as an estimate of daily soil water depletion (DSWD)
Daily ETC was estimated by dividing DSWD by the area occupied by the tree (18 m2).
Available soil water and daily weighted 0 were determined using the same mehod as that
of DSWD by substituting 0rc and mean daily 0 for A0. Mean daily 4> was estimated from
the daily weighted 0 using the soil water characteristic curve for Candler fine sand
%


127
Table 6-3. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in September, 2002. Rootstocks are Carrizo citrange,
and Swingle citrumelo; high N application rate was 269 kg ha'1 yr'1, low rate was 134 kg
ha'1 yr'1.
Days After
Application
Weighted
Solution
N
(MgL1)
N Uptake
Passive Active Total
(g tree'1 d~J)
Cumulative Total
Cumulative N
Loss
N Uptake Control Soil
(gtree1) (%) (% applied)
Carrizo High rate
1
187.8
5.8
4.9
10.7
10.7
17.8
0.1
17.9
2
131.6
4.6
5.4
10.1
20.8
34.4
4.1
38.0
3
91.9
3.4
3.7
7.1
27.9
46.3
18.5
58.6
Carrizo
- Low rate
1
102.6
3.3
4.8
8.1
8.1
24.9
4.9
29.8
2
61.7
2.2
2.7
4.9
13.0
39.8
10.2
48.5
3
28.7
1.5
1.1
2.7
15.6
48.0
19.0
81.8
Swingle
- High rate
1
201.0
7.8
8.4
16.1
16.2
19.0
22.5
41.5
2
121.1
4.1
5.8
10.0
26.1
30.7
25.1
54.8
3
90.2
3.4
5.6
9.0
35.1
41.3
31.7
68.3
Swingle
- Low rate
1
118.7
4.2
9.1
13.2
13.2
31.4
1.0
32.4
2
77.4
2.6
8.2
10.8
24.1
57.2
-25.0
40.6
3
55.9
2.2
3.4
5.5
29.6
70.3
-7.0
64.7


72
only 1 m. Therefore, the ratio of above-ground to below-ground biomass may have
remained near 3:2 if roots extending below 1 m were included in the total root biomass
measurement.
Citrus trees increase in size with time; branches increase in diameter through
accumulation of xylem tissue, eventually developing a scaffold branch structure of large-
diameter branches. The relationship of total tree biomass and N weight to TCV and
TCSA followed a linear function indicating a constant rate of accumulation with increase
in tree size as measured by both TCV and TCSA. This result implies that the partitioning
of biomass and N accumulation in all plant parts occurs at rates specific to the tree
component. Therefore, the total biomass and/or N weight of a citrus tree can be estimated
for any tree size.
Percentage biomass and N weight of woody tree parts (large branches and trunk)
increased, while those of leaves and twigs decreased with increase in TCV and TCSA.
Caruso et al. (1999) found similar results in peaches where the relative proportion of leaf
and twig dry weights decreased with tree age. It can be concluded that to support the
increase in total tree weight, the biomass and N content of woody branches and trunk
increases at a higher rate compared with leaves and twigs. However, it can be concluded
that LAIt is the driving factor in leaf and twig biomass accumulation since the ratio of
leaf area to ground area under the canopy remained constant with increase in tree size for
medium and large trees. Thus, once the total biomass and N weight of a tree is estimated,
the weights of individual tree parts can be estimated based on tree size. Regression
equations such as those in Tables 3-5 and 3-6 can be used to simulate biomass and N
partitioning in a citrus growth model.


publication (Tucker et al., 1995) was produced to assist growers in determining the rate
of N to apply, timing of application, and suggested irrigation scheduling.
In 2002, a revised BMP established rates and timing of N applications based on
tree age classes and method of application. The two age classes are 4 to 7 years and >7
years. The methods of application are broadcast only, broadcast and fertigation, and
fertigation only. No more than 34 kg ha'1 N is to be applied at one time, and no more than
34 kg ha'1 N may be applied from June 15th to September 15th. No fertigation application
is to exceed 17 kg ha'1 N and must be applied at a minimum 1-wk interval.
Decision Support Systems
Important decisions for growers are when and how much fertilizer and irrigation
to apply. They need to consider several factors in their decision-making process to
determine that the crop value to be gained is greater than the cost of fertilizer and
irrigation applied. Fixed fertilizer and irrigation schedules, based on long-term mean
climatic conditions, may lead to inefficient use of these inputs due to the large annual
variability in atmospheric conditions (Heinemann et al., 2000). Likewise, variations in
the amount of rainfall and its distribution may lead to the loss of N from the crop root
zone necessitating additional applications. Due to the complexity of the decision making
process, researchers have developed computer-based decision support systems (DSS). A
DSS can provide information on management options based on local environmental
conditions. These systems also provide a means to make the scientific understanding of
complex plant, soil, and environmental interactions accessible to decision makers in a
concise and interactive manner. Frequently, information from simulation models has
formed the foundation for these DSS.


Table 3-3. Dry matter accumulation and allocation between tree components for mature Hamlin orange trees as affected by year of sampling,
rootstock, and interaction of year and rootstock.
Dry Weight Accumulation Dry Weight Allocation
Total
Mass
Above
Ground
Below
Ground
Total
Leaf
Total
Twigs
Sm.
Branches
Med. Lrg.
Total
Total
Trunk
Sm.
Roots
Med. Lrg.
Tap
(kg dw tree'1)
(% of total weight)
Year 1
97.6
69.2
28.3
12.7
5.8
17.0
12.1
19.1
48.2
3.6
5.1
4.9
9.5
10.1
Year 2
87.8
64.9
26.3
12.4
7.3
15.3
9.0
21.1
45.4
3.9
6.1
6.8
6.4
9.2
NS*
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
**
NS
NS
Carrizo
100.3
77.0
26.6
12.2
7.1
16.0
10.5
23.8
50.3
3.9
5.5
6.1
6.9
6.7
Swingle
82.6
54.7
27.9
12.9
6.0
16.1
10.4
15.8
42.3
3.7
5.8
5.7
8.8
13.1
NS
*
NS
NS
NS
NS
NS
**
*
NS
NS
NS
NS
*
Car Yrl
104.2
79.2
25.0
12.9
6.8
15.7
11.2
25.4
52.3
3.9
4.9
5.6
8.5
5.1
Car Yr2
96.3
74.8
28.2
11.6
7.5
16.4
9.7
22.2
48.3
3.9
6.2
6.6
5.4
8.3
Swi Yrl
87.6
54.3
33.3
12.4
4.3
18.9
13.5
9.5
41.9
3.2
5.6
3.8
11.1
17.7
Swi Yr2
79.4
55.0
24.4
13.3
7.1
14.2
8.3
20.0
42.5
3.9
6.0
6.9
7.3
10.1
NS
NS
NS
NS
NS
NS
NS
**
NS
NS
NS
NS
NS
NS
*NS = not significant, *= significant P<0.05, and **= significant P<0.01.
VO


56
Fig. 3-4. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) accumulation
as a function of tree canopy volume for HamlinVCarrizo- experiment 1
(0),Hamlin7Swingle(D),and experiment 2 (A) trees.


LIST OF FIGURES
Figure page
1-1. Map of Florida with Lake, Polk, and Highlands counties highlighted 3
1 -2. Plant/soil nitrogen and water balance flow chart 7
3-1. Tree canopy volume as a function of trunk cross sectional area for
trees from experiments 1 and 2 52
3-2. Leaf area expressed as a function of tree canopy volume (A) and
trunk cross sectional area (B) 53
3-3 Leaf area index on a tree basis expressed as a function of tree
canopy volume (A) and trunk cross sectional area (B) 55
3-4. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
canopy volume 56
3-5. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
trunk cross sectional area 57
3-6. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and root and tap root dry weight accumulation
as a function of canopy volume 61
3-7. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of trunk cross sectional area 62
3-8. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of canopy volume 65
3-9. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of trunk cross
sectional area 66
xi


140
Fig. 6-7. Seasonal cumulative dry mass (A) and N content (B) of flowers, fruit,
and leaves collected from catch frames under mature citrus trees during
the 2002 season.


13
Compton, 1945), leaves contained from 40 to 50% of total tree N. Twigs and shoots
contained approximately 10% of total N. Trunk and branches contained from 20 to 30%
of total tree N. Approximately half of this N was in the bark, whereas this tissue
component represented only 5% of the total dry mass. Roots contained from 15 to 20% of
the N, half or more of which was in the bark that made up only 5% of the dry mass of the
tree.
Seasonal changes in leaves, bark, twig, and root N concentration were greater
than N changes in woody branch tissue. The trees contained more N just before initiation
of growth activity during spring than at any other time of year. Maximum bark N content
occurred about December 1, followed by a decrease. It was speculated that the reduction
between December and February was the result of deposition of starch and possible other
carbohydrates in these tissues. A decrease of all tree tissue N concentration occurred
during bloom, fruit set, and periods of active growth in the spring and early summer.
During the summer and autumn, N concentrations gradually increased to the mid-winter
maximum. Mattos (2000) found similar N concentration values for 7 year-old Hamlin
trees. Nitrogen concentration was lowest in the trunk and taproot of these trees. The N
concentration of leaves (2.1 to 2.6%), twigs (0.4 to 0.8%), and roots (0.6 to 1.7%) varied
with tissue age. Younger tissue tended to have greater N concentration compared with
older tissues.
Citrus Root Growth Dynamics
Citrus trees are productive and grow well on central Florida deep sandy soils. In
some instances, tree size and yield appear to be related to root system characteristics
(Castle and Krezdom, 1975). Citrus fibrous roots are commonly defined as those roots <4


98
the same irrigation and fertilization practices) were used as an approximation of root
length desities at the various locations and depths.
The rate of soil water withdrawal is influenced by surface evaporation and is
related to the amount of soil shading. Likewise, transpiration by groundcover species
increases withdrawal. Water withdrawal rates per unit root length at all locations and
depths were compared. Increased water loss at the soil surface that could not be explained
by root density was assumed to be due to one or both of the above factors.
Results
Seasonal ET and ETC Trends
Daily ET0 reported by FAWN for the experimental site ranged from a minimum
of 1.1 mm in December, 2001 to a maximum of 6.5 mm in June, 2000 (Table 5-1). The
standard deviations for ET0 by month were relatively low (<0.4 mm) for all months with
the exception of transition months between seasons (February and March in the spring
and August and September in the fall). This result indicates that weather condtions
related to ETC were relatively stable with the exception of these transition periods.
Monthly maximum, minumum, and mean values for ET0 and ETC were not significantly
different for corresponding months during the 2 years of observations included in this
study. Although generally lower, daily ETC followed the same seasonal patterns as ETC.
Exceptions to this trend occurred during the summer months of June through August, but


20
requirements of grapefruit are generally higher than orange or mandarin varieties for trees
of equal size. Wiegand and Swanson (1982 a, b, c) and Wiegand et al. (1982) reported
that mean daily citrus ETC at Weslaco, Texas ranged from 2.2 to 3.3 mm for Ruby Red
grapefruit and 1.9 to 2.7 mm for Marrs oranges from 5 to 10 years of age.
Under similar climatic conditions, citrus trees are known to have lower
transpiration rates compared with other crop plants. Mahrer and Rytwo (1991) reported
mean estimated daily crop water use (ETC) rates for cotton in the Hula Valley of Israel of
5.4 mm when irrigated daily, and 4.0 mm during a 14-day period when not irrigated.
Likewise, Starr and Paltineanu (1998) reported that daily ETC rate for full canopy com at
Behsville, MD ranged from 3.8 to 5.0 mm prior to rainfall and 5.2 to 8.0 mm after.
Lower citrus transpiration rates are related to lower leaf and canopy conductance (Mills
et al. 1999).
Tree size
Large, vigorous, healthy trees require more water than young trees (Tucker et al.
1997). In Florida, large trees at low planting densities (150 to 180 trees per ha) may use
62 to 94 L per day during the winter months and 189 to 219 L per day in July and August
(Boman, 1994). Rogers and Bartholic (1976) reported a mean annual ETC of 1210 mm
during an 8-year period from a young orange and grapefruit grove on poorly drained soils
near the east coast of Florida. These annual ETC values ranged from 820 mm early in the
study (tree age 2 years) to 1280 mm at end of the study (tree age 10 years). Linear
regressions of annual ETe vs. years during the study resulted in significant (P=0.1)
increase in ETC. Mean annual ETC increased at a rate of 19 mm per year or a cumulative
increase of approximately 13% in 8 years. Fares and Alva (1999) reported an annual ETC


16
(Castle, 1978; 1980). In a recent report, data were given as root length densities and
ranged from 530 cm m"3 for Swingle citrumelo roots [C. parachsi Macf. x P. trjfoliata
(L.) Raf] to 2020 cm m'3 for trifoliate orange (Eissenstat, 1991).
Climatic Effects
Root distribution was studied in 22 mature navel or Valencia orange groves in
California (Cahoon et al., 1956). In this study, 50% were low-yielding while the
remaining 50% were high-producing. Fibrous root fresh mass was measured to a depth of
90 cm under the canopy and between rows. Yield was not related to the under-canopy
root quantities, but was correlated with the root quantities measured between adjacent
rows where soil water contents were typically lower most of the year.
Rootstocks
Some citrus roots have been found as deep as 7 m (Ford, 1954b), and in one
instance, roots of mature trees on rough lemon rootstock were discovered 14 m from the
tree trunk (Ford, 1970). Castle and Krezdom (1975) described two general types of root
systems, the first characterized as extensive featuring extensive lateral and vertical
development, and the second as intensive with less extensive root expansion and higher
fibrous root concentrations mainly confined to the upper soil layers. Trees on rough
lemon, Volkamer lemon and Palestine sweet lime (C. limettioides Tan.) rootstocks
typified the extensive type of root system where 50% of the fibrous roots occurred below
70 cm in the soil with wider spreading lateral development. Examples of the intensive
type were Rusk citrange and trifoliate orange, where few fibrous roots were found
below 70 cm, and the root system was less developed laterally. Some rootstocks like sour
orange and Cleopatra mandarin were classified as intermediate. Trees on Cleopatra


57
Trunk Cross Section Area (cm2)
Fig. 3-5. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) accumulation
as a function of trunk cross section area for Hamlin/Carrizo Experiment 1 (O),
Hamlin/Swingle Experiment 1 (), and Experiment 2 (A) trees.


50
citrange roostock trees. On a percentage basis, the taproot biomass was significantly
greater (P=0.05) for trees on Swingle than those on Carrizo.
Although tree weights appeared to be less in 2002 compared with 2001, total dry
weight and dry weights of tree components were not significantly different at the P=0.05
level (Table 3-3). Leaf weights represented 12 to 14% of total tree dry weight, while
total branch weights (twigs, total branches, and trunk) accounted for 49 to 63% of total
tree weight. Dry matter allocation to tree roots was highly variable in all three studies, but
averaged in the 19 to 21% range.
Total above-ground N accumulation was significantly affected by TCV and TXA
at the P=0.05 level (Table 3-4). Leaf, branch, and root N comprised approximately 45,
i
35, and 20% of total tree N, respectively. Dry weight allocation to leaves, twigs, small
branches, medium branches, and trunk were similar for both rootstocks. However,
Carrizo trees had almost 50% more N in large branches compared with Swingle. Mean
total N accumulation by large roots and taproot was greater for trees on Swingle
compared with those on Carrizo. Nitrogen concentrations were not significantly different
(P=0.05) within tissues on each rootstock (data not shown), thus differences in percentage
of total N mass of each tissue category were due to differences in biomass.
Biomass Changes with Increase in Tree Size Experiments 1 and 2
Data from experiments 1 and 2 were combined to determine relationships
between dry weight and N accumulation and tree size indices such as TCV or TCSA.
TCV increased linearly as TCSA increased (Fig. 3-1). Likewise, total leaf area per tree
was linearly proportional to both TCV (Fig. 3-2A) and TCSA (Fig. 3-2B). Leaf area


75
Several Florida studies have demonstrated that tree size and yield were related to
fibrous root density and/or distribution (Castle and Krezdom, 1975; and Ford 1954a;
1964; 1968; 1969; 1972) in the deep sandy soils of central Florida. Since processes that
control non-point source pollution are dynamic and greatly affected by genetic traits and
environmental conditions, development of models that can integrate interactive effects of
spatial and temporal processes will be critical. However, models providing realistic
results will need to include accurate information on root growth dynamics. Modeling of
citrus root distribution and the determination of water and nutrient uptake parameters can
lead to the development of an expert system for the estimation of water and nutrient
depletion and uptake by soil depth. Likewise, nutrient leaching can be estimated due to
excessive irrigation or heavy rainfall.
Under Florida growing conditions, the quantity of fibrous roots decreased with
depth and lateral distance from the trunk (Elezaby, 1989; Menocal-Barberena, 2000).
Eighty percent of citrus fibrous roots were found within a 120 cm radius of the tree trunk
and 40% in the upper 30 cm of well-drained sandy soils. Nearly all citrus roots grow
within 45 cm of the soil surface where artificial drainage was provided and/or high water
tables occured (Calvert et al., 1967; 1977; Ford, 1954a; Ford, 1972; Reitz and Long,
1955).
r\
Fibrous root dry mass densities ranged from 300 to 1200 g m'3 (Castle, 1978;
1980). Citrus fibrous root length densities ranged from 0.53 cm cm"3 for Swingle
citrumelo roots to 2.02 cm cm"3 for trifoliate orange (Eissenstat, 1991). Elezaby (1989)
reported fibrous root concentration in the 0 to 30 cm soil zone increased from 450 to
1000 g m between trees when the in-row distance decreased from 4.5 to 2.5 m due to


101
Fig. 5-3. Estimated ETC to calculated ET0 ratio as a function of soil water content in the
irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a 1 m
depth (C). The data points shown represent a range of soil water content from
field capacity to approximately 50% available soil water depletion.


53
Fig. 3-2. Leaf area expressed as a function of tree canopy volume (A) and trunk cross
section area (B).


65
Canopy Volume (m3)
Fig. 3-8. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and trunk (open)
N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of canopy volume for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.


87
Discussion
The root distributions of the two rootstocks used in this study appear to lie
intermediate between the extensive and intensive distributions described by Castle and
Krezdom (1975). However, there were some subtle differences between rootstocks.
Citrus trees grown on Swingle citrumelo had greater root length densities in the upper 30
cm than did trees grown on Carrizo citrange. The root density distributions of both
rootstocks were bimodal in nature. However, Carrizo roots tended to grow deeper than
did those of Swingle. Trees grown on Carrizo citrange rootstock had higher fibrous root
length densities at all distances from the tree trunk below 60 cm from the soil surface.
These root length densities indicate that the depth of irrigation and the depth to which
fertilizer N is initially placed should be rootstock specific. Thus, mature trees grown on
Swingle citrumelo should be irrigated to a shallower depth compared with trees grown on
Carrizo citrange. Deep irrigation will waste water and potentially leach soil N below the
soil volume containing the largest proportion of roots, thus potentially decreasing NUE.
Menocal-Barberena (2000) found similar trends using fibrous root mass densities
from root samples collected at the same site. In his and the current studies, Swingle had
greater, but not significantly different, fibrous root concentrations than Carrizo. Mean in
row root concentrations were significantly greater than between-row concentrations.
Mean root concentrations in the upper 30 cm were more than four times greater than for
soil layers below the 30 cm depth.
Under central Florida Ridge conditions, citrus fibrous root densities increase in
two modes. The first mode is the development of a dense root mat just below the soil
surface. This portion of the root system expands in radial manner away from the tree


86
Table 4-4. Regression coefficients and statistics for root length density as a function of
distance from the tree trunk, and soil depths by trunk cross-section area using a third
order quadratic polynomial model2.
Distance
(cm)
Depth
(cm)
Y0
A
b
c
R2
RMSE
(cm cm'3)
P
50
0-15
-0.41
0.001
-0.00005
0.00001
0.71
0.44
0.015
50
15-30
0.24
0.0001
-0.00003
0.00001
0.56
0.21
0.071
50
30-45
0.04
0.0001
-0.00001
0.00001
0.68
0.09
0.022
50
45-60
-0.50
0.0005
-0.00006
0.00003
0.88
0.10
0.001
50
60-75
-0.08
0.0001
-0.00002
0.00002
0.66
0.09
0.027
50
75-90
-0.20
0.0002
-0.00003
0.00001
0.79
0.08
0.004
100
0-15
0.23
0.00001
-0.00001
0.00001
0.67
0.19
0.026
100
15-30
-0.01
0.0001
-0.00001
0.00001
0.36
0.19
0.343
100
30-45
-0.42
0.0004
-0.00003
0.00002
0.52
0.13
0.099
100
45-60
-0.17
0.0002
-0.00002
0.00002
0.82
0.05
0.003
100
60-75
-0.15
0.0002
-0.00002
0.00001
0.90
0.03
0.0002
100
75-90
-0.20
0.0002
-0.00001
0.00001
0.97
0.01
<0.0001
150
0-15
-0.50
0.0004
-0.00001
0.00002
0.96
0.07
<0.0001
150
15-30
-0.30
0.0003
-0.00003
0.00001
0.83
0.06
0.002
150
30-45
-0.11
0.0001
-0.00001
0.00001
0.83
0.02
0.002
150
45-60
-0.15
0.0001
-0.00001
0.00001
0.78
0.03
0.005
150
60-75
-0.01
0.00001
-0.00001
0.00001
0.74
0.01
0.010
150
75-90
-0.05
0.00001
-0.00005
0.00001
0.61
0.02
0.046
200
0-15
-0.28
0.0002
0.00003
0.00001
0.94
0.04
0.086
200
15-30
-0.33
0.0003
-0.00003
0.00001
0.78
0.08
0.311
200
30-45
-0.07
0.0001
-0.00001
0.00001
0.80
0.02
0.282
200
45-60
-0.03
0.00001
-0.00003
0.00001
0.87
0.01
0.192
200
60-75
-0.04
0.00001
-0.00004
0.00001
0.88
0.01
0.180
200
Z V vr
75-90
-0.03
;
0.00001
-0.00002
0.00001
0.79
0.01
0.293
z Y-Yo+aX+bX2+cX3 where X = TXA, and Y0, a, b, and c are regression coefficients


22
terms are not independent. For instance, the amount of applied irrigation water will
influence the amount of ETC as well as the amount of drainage (Prajamwong et al., 1997).
In standard irrigation practices, water transport through the soil may be classified
into five phases: 1) infiltration during application; 2) redistribution after application
ceases; 3) withdrawal by plant roots; 4) evaporation from the soil surface; and 5) drainage
of water to lower soil depths. The primary modes of transport of water in soil are 1)
viscous flow through liquid-filled pores, and 2) diffusion of vapor through air-filled
>ores. In principle, both modes contribute to soil water flow. Liquid flow is the dominant
node in saturated to moist soils (Hagan et al., 1967). Vapor flow is of minor importance
until soils become quite dry, although the presence of a large temperature gradient favors
the contribution of this mechanism. For typical soil water situations, both of these
transport modes contribute to a flow rate proportional to potential energy gradients within
the soil.
Water is of central importance in the transport of solutes in soils or plants,
whether by diffusion or mass flow (Tinker and Nye, 2000). The concept of potential is
fundamental to understanding soil water dynamics. Potential is a measure of the energy
state of a chemical compound within a particular system, and hence of the ability of a
unit amount of this compound to perform work. Difference in potential at different points
in a system gives a measure of the tendency of the compound, including water, to move
from a region with high potential to a region of lower potential.
Soil water has various forms of potential energy acting on it, all of which
contribute to the total potential. Tinker and Nye (2000) refer to these forms of potential
energy as concentration, compression, position in an electrical field, and position in the


81
Fig. 4-2. Root length density distribution at 0-15, 15-30, 30-45,45-60, 60-75, and 75-90
cm depth increments by distance from the tree trunk as affected by distance from
the tree trunk for Hamlin orange trees on Carrizo citrange (A) and Swingle
citrumelo (B) rootstocks.


165
Ford, H. W. 1969. Water management of wetland citrus in Florida. Proc. First Int. Citrus
Symp. 3:1759-1770.
Ford, H.W. 1970. Problems in using Milam rootstock as a biological barrier. Proc. Fla.
State Hort. Soc. 83:84-86.
Ford, H.W. 1972. Eight years of root injury from water table fluctuations. Proc. Fla. State
Hort. Sci. 85:65-68.
Ford, H.W., I. Stewart, and C D. Leonard. 1954. The effect of iron chelate on root
development of citrus. Proc. Amer. Soc. Hort. Sci. 63:81-87.
Ford, H.W., W. Reuther, P.F. Smith. 1957. Effect of nitrogen on root development of
Valencia orange trees. Proc. Amer. Soc. Hort. Sci. 70: 237-244.
Gabrielle, B., S. Menasseri, and S. Houot. 1995. Analysis and field evaluation of the
CERES models water balance component. Soil Sci. Soc. Am. J. 59:1403-1412.
Gabrielle, B., and L. Kengni. 1996. Analysis and field-evaluation of the CERES models
soil components: nitrogen transfer and transformation. Soil Sci. Soc. Am. J. 60:
142-149.
Gijsman, A.J., G. Hoogenboom, W.J. Parton, and P C. Kerridge. 2002. Modifying
DSSAT crop models for low-impact agricultural systems using a soil organic
matter-residue module from CENTURY. Agron. J. 94:462-474.
Graham, W.D., and A. Alva. 1995. Ridge citrus water quality project annual progress
Report submitted to the Florida Department of Agriculture and Consumer
Services and the Southwest Florida Water Management Districts.
Graham, W.D., and T.A. Wheaton. 2000. Ridge citrus water quality project annual
progress report submitted to the Florida Department of Agriculture and Consumer
Services and the Southwest Florida Water Management Districts.
Graser, E.A., and L.H. Allen, Jr. 1987. Water relations of 7-year-old containerized citrus
trees under drought and flooding stress. Proc. Fla. State Hort. Soc. 100:126-136.
Hagan, R. M., H. R. Haise, and T.W. Edminster (eds). 1967. Irrigation of Agricultural
Lands. Am. Soc. of Agron. Number 11.
Hamblin, A and D. Tennant. 1987. Root length density and water uptake in cereals and
grain legumes. How well are they correlated. Aust. J. Agri. Res. 38(3) 513-527.
Hams, AJ. 1875. The wild orange groves of Florida. FI Fruit Growers Assoc. Proc 3 27
-29. r-


46
Analytical, Inc., New Castle, DE). The digest was analyzed for total Kjeldahl nitrogen
using USEPA method 351.2 using a Buchi model B339 steam distillation instrument
(Buchi Analytical, Inc., New Castle, DE.).
Leaf Area, Biomass and N Weight Estimation
Specific leaf area (cm2 g'1) were determined for a 50-leaf subsample by dividing
total leaf area by leaf dry weight. Total leaf area was estimated by multiplying the mean
specific leaf area by the estimated total dry leaf weight for the corresponding tree. Leaf
area index was determined for each tree by dividing the leaf area by the corresponding
cross sectional canopy area using the in-row and cross-row measurements. Leaf area
index was also determined on a per acre basis by dividing the total leaf area of each tree
by the land area occupied by the tree (in-row spacing x cross-row distance). Dry weights
for each tissue type of individual tree were estimated by multiplying the total field fresh
mass by the mean percentage dry mass of three sub samples for each tissue category.
Likewise, N accumulation within each tissue type of each tree was estimated by
multiplying the total dry weight by the mean N concentration of the three sub samples of
each tissue. Total tree dry weight and N accumulation were determined by summing
across tissues categories. The above ground dry weight and N accumulations were
determined by summing the estimated values for leaf, twig, total branch, and trunk
components. Likewise, the below ground biomass and N accumulation was the sum of
total root and taproot values. Percentages of total biomass and N weight were determined
for each tissue.
Prior to dissecting the second set of mature Hamlin trees in 2002, all fruit was
removed and weighed. Representative fruit samples were collected and dried to


122
Tissue samples collected
The following tissues were sampled at each growth stage: 1) expanding leaves, 2)
expanded leaves, 3) twigs (<7 mm), 4) small limbs (7 to 15 mm), 5) medium limbs (15 to
30 mm), 6) large limbs (>30 mm), 7) trunk, 8) feeder roots (<4 mm), and lateral roots (>4
mm). Two trees from each of the 12 plots were selected for each sampling period. Fifty
non-expanded leaves per tree were collected from the last flush, and 50 expanded leaves
per tree were also collected. Thirty twigs and small (7 to 15 mm) branch segments of 15-
30 cm in length were collected. Cylinders of tissue 10 mm in diameter and 15 mm long
were removed from branches and trunks greater than 15 mm in diameter using a plug
cutter and battery powered drill. Eight plugs from each of four limbs of 15 to 30 mm and
>30 mm diameter, and two trunks were collected on each sampling date. Branch and
trunk samples were separated into bark and wood components. Twelve fruits were
collected on each sampling date. Three soil cores per tree were taken for root removal
and roots were separated by size and depth, (size = <4 mm and >4 mm; depth = 0 to 15
cm, 15 to 30 cm, 30 to 45 cm, 45 to 60 cm, and >60 cm).
Tissue analysis
All fresh tissue samples were weighed, dried for 3 days at 70 C, reweighed, and
ground for nutrient analysis. Tissues were analyzed for total N using the same grinding
and Kjeldahl methods described in Chapter 3. Fruit diameters and leaf area were
measured prior to drying.
Experiment 3 Seasonal N loss
Catch frames 0.9 m wide x 1.5 m long were placed under one tree in each of the
12 plots used for seasonal N concentration determination. Any citrus plant material


143
lysimeters may be 3.5 times too high, reducing the residual NH4-N amount in bulk soil to
16.6% 1 day after application. If this is the case, then values found in this study are
probably similar to those found by Lea-Cox et al. (2001).
In the same study, Lea-Cox et al. (2001) found that tissue accumulation of 15N
increased for the first 3 to 5 days after application with little additional subsequent
increase, with the exception of trees on Volkamer lemon rootstock. Trees on this
rootstock increased in ,5N accumulation to between days 7 and 15, but only at the highest
N application rate. Young trees on Volkamer lemon grow rapidly and are assumed to
have high N demand compared with the two rootstocks used in the current study. This
result confirms that both N demand and availability control uptake rates (Dasberg, 1987;
Kato et al. 1982, and Scholberg et al., 2002). Lea-Cox et al. (2001) found total N in the
soil was reduced by 71.3 to 95.1 of the applied N on day 8 for trees grown on Volkamer
lemon rootstock. Assuming that the majority of N is removed and accumulated by the
tree in the first 3 to 5 days, a reduction in applied N ranging from 70.1 to 83.6% three
days after application for the two rates in the current study would appear to be
reasonable. Likewise, the estimated NUE that ranged from 50.2 to 84.0% of the 15N
applied after 29 days in the above study was similar to the estimated mean maximum
uptake range of 46.7 to 61.7% for the high and low N application rates used in this study.
Mean estimated cumulative dry biomass losses during the 2-year study from
flowers, fruit and leaves were 646, 1290, and 2423 g tree'1, respectively. Leaf loss
accounted for approximately 33% of the total leaf biomass estimated for mature trees of
the same size. This result indicates that leaves could remain on citrus trees for a many as
3 years, which is 1 year longer than reported by Wallace et al. (1945). Leaf longevity


LIST OF TABLES
Table gage
3-1. Citrus biomass and nitrogen distribution by tree age as reported
by different studies 41
3-2 Mature citrus tree leaf area and leaf area index as a function of tree
canopy volume and trunk cross sectional area 48
3-3. Dry matter accumulation and allocation between tree components for
mature Hamlin orange tree by tree canopy volume as affected by
year of sampling, rootstock, and interaction of year and rootstock 49
3-4. Nitrogen accumulation and allocation between tree components for
mature Hamlin orange tree by trunk cross sectional area as affected
by year of sampling, rootstock, and interaction of year and rootstock 51
3-5. Linear regression analysis of dry weight and N accumulation in different
tree components as related to tree canopy volume 58
3-6. Linear regression analysis of dry weight and N accumulation in different
tree components as related to trunk cross sectional area 59
3-7. Mature Citrus tree tissue N concentration as a function of year of
sample and rootstock 63
4-1. Mean mature fibrous (diameter < 4 mm) root length
density ofHamlin orange tree as affected by rootstock, orientation
distance, and soil depth 79
4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length
densities for trees ranging from 2years old to > 15 years old 85
4-3. Regression coefficients and statistics root length density as a function
of distance from the tree trunk, and soil depths by canopy volume
using a third order quadratic polynomial model 85
4-4. Regression coefficients and statistics for root length density as a function
of distance from the tree trunk, and soil depths by trunk diameter
using a third order quadratic polynomial model 86
viii


I
CHAPTERS
CITRUS WATER UPTAKE DYNAMICS
Introduction
Climate, crop development, soil water status, and competition with other plants
affect total soil water use by crop plants. Stomatal conductance regulates both
transpiration and photosynthesis and therefore directly affects the water use and
productivity of plants (Jones et. al., 1985). Stomata are sensitive to environmental
variables such as light, CO2, vapor pressure deficit (VPD), and plant water status (Jarvis
and McNaughton, 1986). Assuming that light, CO2, and VPD conditions are nearly
constant during short intervals of time, hourly and daily changes in plant water status can
have large impacts on stomatal conductance and thus on transpiration rates and ultimately
productivity. The key to plant water status is soil water availability (Allen et al., 1997),
thus an improved understanding of soil water uptake dynamics is essential in optimizing
both the amount and timing of irrigation required for maximum production. Once
understood, seasonal soil water content and root length density effects on citrus water
uptake can be modeled to provide more accurate irrigation scheduling, which will reduce
negative impacts on ground water quantity and quality due to leaching and over
pumping.
Citrus water requirements vary with climatic conditions, variety, and canopy size.
Lower crop evapotranspiration (ETC) rates for Florida (humid) compared with Arizona
(semi-arid) have been attributed to lower evaporative demand (Rogers et al. 1983, Fares
90


84
Fig. 4-3. Citrus root distributions by depth below the soil surface and distance from the
tree trunk for trees 2-5 years old (A and B), 5-10 years old (C and D), 10-15 years
old (E), and > 15 years old (F).


Table 3-1. Citrus biomass and nitrogen distribution by tree age as reported by different
studies.
41
Authors
Trees
(n)
Location
Cultivar
Age
(Yrs)
Plant
Tissue
Biomass
(% Total)
N
(% Total)
Cameron and
15
California
Val.
3.5
Leaves
30.5
61.9
Appleman
Branches
38.4
21.0
(1935)
Roots
31.1
17.1
Legaz and
8
Spain
Val.
4
Leaves
22.5
30.0
Primo-Millo
Branches
28.7
18.5
(1988)
Lateral roots
45.8
41.2
Fibrous roots
2.9
10.3
Mattos
6
Florida
Ham.
6
Leaves
13.9
35.0
(2000)
Branches
46.5
28.£
Lateral roots
25.7
13.4
Fibrous roots
14.1
23.2-
Cameron and
4
California
Val.
10
Leaves
18.5
46.7
Appleman
Branches
60.7
39.0
(1935)
Roots
20.7
14.2
Cameron and
36
California
Val.
15
Leaves
16.8
45.3 -
Compton
Branches
61.4
34.8
(1945)
Lateral roots
20.4
17.0
Fibrous roots
1.7
2.9
Katoetal.
1
Japan
Sat.
21
Leaves
8.6
27.2
(1984)
Branches
65.1
44.6
Lateral roots
19.5
14.3
Fibrous roots
6.8
14.0
Feigenbaum
2
Israel
Sham
22
Leaves
7.3
24.6
etal. (1987)
Branches
61.0
49.8
Lateral roots
26.5
19.2
Fibrous roots
4.3
3.8
to determine: 1) changes in biomass and N distribution with change in tree size, 2) yearly
changes in biomass and N content of mature citrus trees, and 3) rootstock effect on
mature citrus tree biomass and N distribution. The relationships of canopy volume and
mean trunk diameter to biomass and N content for citrus will form the basis of a
predictive model to estimate the biomass and N distribution based on size measurements.


163
Castel, J.R., I. Bautista, C. Ramos, and G. Cruz. 1987. Evapotranspiration and irrigation
efficiency of mature orange orchards in Valencia (Span). Irrigation and
Drainage Systems 3:205-217.
Castel, J.R., and A Buj. 1992. Growth and evapotranspiration of young, drip-irrigated
Clementine trees. Proc. Int. Citriculture, 651-656.
Castle, W.S. 1978. Citrus root systems: Their structure, function, growth, and relationship
to tree performance. Proc. Int. Soc. Citriculture 1:62-69.
Castle, W.S. 1980. Fibrous root distribution ofPineapple orange trees on rough lemon
rootstock at three tree spacings. J. Amer. Soc. Hort. Sci. 105(3):478-480.
Castle, W.S. and AH. Krezdom. 1975. Effect of citrus rootstocks on root distribution and
leaf mineral content ofOrlando tngelo tree. J. Amer. Soc. Hort. Sci. 100(1): 1-
4.
Castle, W.S. and CO. Youtsey. 1977. Root system characteristics of citrus nursery trees.
Proc. Fla. State Hort. Soc. 90:39-44.
Chapman, H.D., and E.R Parker. 1942. Weekly absorption of nitrate by young, bearing
orange trees growing out-of -doors in solution culture. Plant Phys. 17:366-376.
Cohen, S., M. Fuchs, M S. Moreset and Y. Cohen. 1987. The distribution of leaf area,
radiation, photosynthesis, transpiration, and the effect of row shape and direction.
Agricultural and Forest Meteorology 40:145-162.
Dasberg, S., 1987. Nitrogen fertilization in citms orchards. Plant Soil 100:1-9.
Doorenbos, J., and W.O. Pruitt. 1977. Crop water requirement. FAO irrigation and
Drainage paper No. 24, Rome Italy, 144 pages.
Elezaby, A. A. 1989. Physiological and biological studies on root systems of some citrus
stocks. Ph D. Dissertation. Cairo University, Egypt.
Eissenstat, D.M. 1991. On the relationship between specific root length and the rate of
root proliferation: a field study using citrus rootstocks. New Phytol. 118:63-68.
Eissenstat, D.M. and L.W. Duncan. 1992. Root growth and carbohydrate responses in
bearing citrus trees following partial canopy removal. Tree Physiol. 10:245-257.
Embleton, J.W, M. Mastumura, L.H. Stolzy, D.A. Devitt, W.W. Jones, R. El-Motaium
and L.L. Summers. 1978. Citrus Nitrogen fertilizer management, groundwater
pollution, soil salinity and nitrogen balance. Applied Agr. Res. 1 (l):57-64.


89
is correct and a model for root length density distribution can be made if both tree size
and rootstock are included as variables. These relationships provide a scientific basis for
the development of a spatial root length distribution model needed for a citrus expert
system that will estimate water and nutrient uptake.


Ill
Lastly, and most importantly, I would like to thank my wife, Nancy, and two
sons, Joshua and Christopher, for their unwavering support and encouragement.
Encouragement from my parents, in-laws, and brother sustained me.


85
Table 4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length densities
for trees ranging from 2 to > 15 years old.
RMSE
CV
R2
P
Significance
Orientation
44.66
87.81
0.002
0.291
NS
Distance
48.08
42.54
0.12
<0.0001
***
Depth
23.45
51.99
0.17
<0.0001
***
Table 4-3. Regression coefficients and statistics root length density as a function of
distance from the tree trunk, and soil depths by canopy volume using a third order
quadratic polynomial model2.
Distance
(cm)
Depth
(cm)
Yo
a
b
c
R1
RMSE
(cm cm'3)
p
50
0-15
1.29
-0.021
0.002
0.00001
0.84
0.33
0.001
50
15-30
0.99
-0.289
0.027
-0.0006
0.78
0.15
0.005
50
30-45
0.27
-0.059
0.006
-0.0001
0.68
0.09
0.021
50
45-60
0.45
-0.16
0.015
-0.00003
0.83
0.12
0.002
50
60-75
0.29
-0.099
0.009
-0.00002
0.68
0.09
0.022
50
75-90
0.31
-0.14
0.013
-0.0003
0.88
0.06
0.0005
100
0-15
0.73
-0.23
0.023
-0.0005
0.79
0.15
0.005
100
15-30
0.44
-0.13
0.014
-0.0003
0.61
0.15
0.072
100
30-45
0.19
-0.069
0.007
-0.00002
0.48
0.14
0.14
100
45-60
0.20
-0.082
0.008
-0.00002
0.82
0.05
0.002
100
60-75
0.071
-0.068
0.006
-0.00001
0.85
0.04
0.001
100
75-90
0.040
-0.018
0.002
-0.00001
0.51
0.04
0.107
150
0-15
0.17
-0.087
0.011
-0.0001
0.97
0.06
<0.0001
150
15-30
0.10
-0.047
0.005
-0.0001
0.81
0.06
0.003
150
30-45
0.007
-0.004
0.001
-0.00001
0.74
0.02
0.010
150
45-60
0.025
-0.012
0.001
-0.00001
0.80
0.03
0.003
150
60-75
0.007
-0.004
0.0004
-0.000001
0.75
0.01
0.008
150
75-90
0.008
-0.004
0.0003
-0.000001
0.65
0.01
0.030
200
0-15
0.042
-0.020
0.002
-0.0001
0.99
0.01
0.014
200
15-30
-0.013
0.007
-0.001
-0.0001
0.99
0.01
<0.0001
200
30-45
-0.0003
0.001
-0.0001
0.00001
0.99
0.01
<0.0001
200
45-60
-0.002
-0.001
0.0001
0.00001
0.99
0.01
0.004
200
60-75
0.003
-0.001
0.0001
-0.000001
0.99
0.01
0.004
200
75-90
-0.001
0.001
-0.0001
0.00001
0.99
0.01
0.0002
z Y-Y0+aX+bX2+cX3 where X = TCV, and Yo, a, b, and c are regression coefficient.


60
cm2, respectively (Figs. 3-6B and 3-7B). Likewise, twig biomass decreased from 11 to
6% of total biomass for trees of the corresponding size categories. Total branch dry
weight increased from 15 to 45% as trees matured, while trunk biomass decreased from
12% for young trees to 3% for mature trees (Figs. 3-6C and 3-7C). Few consistent trends
were found when comparing root biomass with tree size (Figs. 3-6D and 3-7D), which
may be due to problems in recovering all root tissue or differences in root biomass
distribution between the two rootstocks used in this study.
Nitrogen Distribution
Mature Citrus Tree Nitrogen Distribution Experiment 1
As was the case with dry biomass, total N accumulation was greater for trees on
Carrizo rootstock (870 g tree'1) than for trees on Swingle (690 g tree'1). Total and above
ground N accumluation were significantly (P=0.01) affected by TCV and TCSA, whereas
below-ground N accumulation was not (Table 3-4). Rootstock, year, and the interaction
of rootstock and year effects were not correlated with TCV or TCSA. Nitrogen
concentration was not significantly different compared with TCV, TCSA, rootstock, or
year for any of the tissues sampled (Table 3-7). Therefore, N content trends were similar
to those for dry weight. Percentage tissue N concentration compared with total tree N
weight was not significantly different for tree size, rootstock, or year with the exception
of branches greater than 30 mm in diameter. Trees grown on Carrizo citrange had
significantly greater (P=0.05) N in larger branches and total branches compared with
TCV (Table 3-4).


34
dormancy, increased during flowering and was highest during fruit set, and later
decreased towards the end of the summer and autumn flushes. The greatest accumulation
of N absorbed from fertilizer (with respect to the total N absorbed from fertilizer in the
whole tree) was found in the young leaves and roots, followed primarily by twigs and
stems, then flowers and fruits.
Kato et al (1982) found that total N contents decreased in both bark and wood
during the sprouting period of 21 year-old Satsuma mandarins. Greatest decreases in N
were found in parts with higher concentrations of N (i.e. leaves, shoots, and fine roots). It
was also concluded that the trunk and large roots are main N reservoirs for new shoot
development. The N was reserved mainly as protein, free proline, arginine, and
asparagines. Protein decreased in all plant parts in proportion to total N in the plant part.
Proline decreased mainly in the leaves and bark, arginine in wood of shoots and
asparagines in bark of fine roots.
Crop and Environmental Models
Current Citrus Models
Few predictive models have been developed specifically for use in citrus
production. Most models have been designed for specific applications, with a general
user in mind. These models predict population and/or crop damage caused by citrus
pathogens (Timmer and Zitko, 1996), and scale insects (Aris and Browning, 1995). Other
citrus models are used for irrigation scheduling (Xin et al., 1997), and crop flowering
(Bellows and Morse, 1986; Valiente and Albrigo, 2000).
Environmental Models
Nitrogen leaching from agricultural soils represents both an economic loss to the
farmer and potential groundwater pollution. Mathematical models can be used to assess


Table 4-1. Mean fibrous (diameter <4 mm) root length density of mature Hamlin orange tree as affected by rootstock,
orientation distance, and soil depth.
Mean Root Length Density (cm cm'3)
Rootstock
Orientation
Distance from tree (cm)
Soil depth (cm)
Carrizo
Swingle
In-row
Cross-row
50 100 150 200
0-15
15-30
30-45 45-60
60-75
75-90
0.36
0.41
0.41
0.35
0.49 0.40 0.33 0.17
1.04
0.30
0.16 0.24
0.28
0.27
P
Significance1
Rootstock
0.290
NS
Orientation
0.253
NS
Distance
0.002
***
Depth
<0.0001
***
Distance*Depth
0.820
NS
Rootstock* Distance
0.052
*
Rootstock*Depth
0.002
***
Significance: NS = not significant, = significant at P=0.1 level, ** = significant at P=0.05 level, and *** = significant at
P=0.011evel.


97
previously determined at this site (Obreza et al., 1997). Percentage daily ASWD was
< A p
determined using DSWD and ASW.
Estimated Monthly Crop Coefficient (K)
Estimated daily tree water use (ETC) was calculated for a 24-month period and
compared with calculated daily ET0. The ratios of estimated daily ETC to calculated daily
ET0 for each of the three trees were averaged to estimate the product Kc Ks in equation
5-1. To eliminate the effects of decreased on water uptake, ratios of ETC to ET0 on days
where mean 0 was near 0fc in both the irrigated and non-irrigated areas (Ks assumed to
be 1) were used to estimate daily Kc. The relationship of these estimated daily Kc values
to day of year (DOY) was determined using regression analysis.
Estimated Water Stress Coefficient (Ks)
Daily ETC to ET0 ratios were calculated throughout the year and compared with
mean daily 0. The ratio of ETC to (ET0*Kc) using the Kc estimated for the day was used to
estimate the value of Ks. The relationship of the ratio of ETC to (ET0*Kc) with 0 and (j) is
typically a logistic response curve with a plateau near 0rc (Allen et al., 1998). Regression
analysis was used to determine the relationship of estimated Ks to ASWD and mean soil
Estimation of Soil Water Uptake per Unit Root Length
Daily estimated ETC on a per unit root length basis for each sensor were
determined for the mean under-canopy, dripline, and between-row locations for soil
depths of 10, 20, 40 and 80 cm. Root length densities determined in Chapter 4 (for trees
on the same rootstock, same approximate tree size, and grown in the same location under


169
Mansell, R.S., J.G.A. Fiskell, D.V. Calvert and J.S. Rogers. 1986. Distributions of labeled
nitrogen in the profile of a fertilized sandy soil. Soil Sci. 141:120-126.
Mansell, RS., H.M. Selim, D.V. Calvert, E.H. Stewart, L.H Allen, D A. Gratez, J.G.A.
Fiskell and J.S. Rogers. 1980. Nitrogen and water distribution in fertilized sandy
soil during irrigation and drainage. Soil Sci. Soc. Amer. J. 44:95-102.
Martin, E.C., A.K. Hla, P.M.Waller, and D.C. Slack. 1997. Heat unit-based crop
coefficient for grapefruit trees. J. Appl. Engr. In Agrie. 13(4):485-489.
Mataa, M. and S. Tominaga. 1998. Effects of root restriction on tree development in
Ponkan mandarin. J. Am. Soc. Hortic. Sci. 123(4) 651-655.
Mattos, D. 2000. Citrus response functions to N, P, and K fertilization. Ph.D.
Dissertation, University of Florida, Gainesville, FI.
Menocal-Barberena, 2000. Effect of rootstock on root distribution of citrus. MS. Thesis.
University of Florida, Gainesville, FI.
Mikhail, E.H. and B.M. El-Zeflawi. 1978. Effect of soil types and rootstocks on root
distribution and leaf composition of citrus trees. Proc. Int. Soc. Citriculture 1:
214-216.
Mills, T.M., K.T. Morgan, and L.R Parsons. 1999. Canopy position and leaf age
affect stomatal response and water use in citrus. J. Crop Production 2(2): 163-
179.
Minkara, M.Y., J.H Wilhoit, C.W. Wood, and K.S. Yoon. 1995. Nitrate monitoring and
GLEAMS simulation for poultry litter application to pine seedlings. Trans, of the
ASAE 38(1): 147-152.
Mooney, P.A., and A.C. Richardson. 1992. Seasonal trends in the uptake and
distribution of nitrogen in Satsuma mandarins. Proc. Int. Soc. Citriculture 593-
597.
Morgan, K. T., L. R. Parsons, T. A. Wheaton, D. J. Pitts, and T. A. Obreza. 1999. Field
calibration of a capacitance water content probe in fine sand soils. J. Soil Sci. Soc.
Am. 63:987-989.
Newman, E.I. 1966. A method of estimating the total length of root in a sample. J. Appl.
Ecology 3:139-145.
Obreza, T.A., and K.E. Admire. 1985. Shallow water table fluctuations in response to
rainfall, irrigations, and evapotranspiration in flatwood citrus. Proc. Fla. State
Hort. Soc. 68:24-29.


31
for ground water contamination (Alva and Paramasivam, 1999; Calvert and Phung, 1972;
Mansell et al., 1980).
Seasonal Nitrogen Uptake
Most N balance studies have been unable to completely account for total N
applied to the soil. Some authors attributed this fraction (usually 30 to 50%) to
atmospheric loss. Khalaf and Koo (1983) concluded that unaccounted for N was either
incorporated into soil organic matter or stored in the tree (Dasberg, 1987), while others
made no attempt to fully account for the applied N (Mansell et al., 1980).
Hilgeman (1941) estimated N uptake by grapefruit in Arizona by determining
changes in leaf N concentration seasonally. Maximum N uptake by the trees occurred in
March and September relative to January due to higher mean soil temperature. In a 3-year
study, Chapman and Parker (1942) determined N removed from solution culture and
reported that the months of least N absorption were January and February. Uptake rates
increased during the period of late spring through early fall (May to October) with a
maximum in July. Roy and Gardner (1946) in Florida reported similar results.
Numerous reports suggest that actively growing tissues act as a sink for N uptake
and that the young developing leaves and fruit constitute the strongest sink. Legaz et al.
(1982) in Spain studied N distribution in 5-year-old Calamondin trees in sand culture.
Trees were labeled with 15N for 20 days during flowering, were harvested, and analyzed
for N content 0 to 70 days later. Accumulated N was found primarily in fruitlets and
newly developed leaves and twigs. About 30% of the labeled N was found in newly
formed leaves. In Israel, Feigenbaum et al. (1987) treated 22-year old Shamouti orange
trees with 15N labeled fertilizer. Trees had previously been supplied with sufficient N or


35
crop N-fertilizer requirements and to predict effects of N fertilizer management practices
on potential nitrate leaching and how it affects groundwater quality. The understanding of
solute movement and transport has increased in the last 30 years. Increased
environmental concerns pertaining to the runoff and leaching of agricultural chemicals
and fertilizer elements in the surface and groundwater has resulted in development and
use of computer simulation models to predict transport of potential pollutants in
agricultural systems. These models include Nitrate Leaching and Economic Analysis
Package (N-LEAP) (Follett et al., 1994), Groundwater Loading Effects of Agricultural
Management Systems (GLEAMS) (Reck, 1994; Reyes et al., 1994), Drainage-Modified
(DRAINMOD) (Saleh et al, 1994; Verma et al., 1995), Chemicals, Runoff, and Erosion
from Agricultural Management Systems (CREAMS) (Minkara et al., 1995; Saleh et al.,
1994), Leaching Estimation and Chemical Model (LEACHM) (Jemison et al., 1994), and
Nitrogen, Carbon, Soil, Water And Plant (NCSWAP) (Jabro et al., 1993).
These models could be applied to citrus production to predict or estimate the
depth of N leaching below the crop root zone. Most of these models are deterministic,
non-steady state, and comprehensive. They typically require a large number of soil
physical, hydraulic, and chemical characteristics for each soil layer, soil N transformation
components, weather data, and environmental information to determine N fate and
leaching depths. Use of these models for the prediction ofN fate under agricultural
production conditions has met with mixed results (Kiniry et al., 1997). Jabro et al. (1993)
found that neither LEACHM nor NCSWAP successfully predicted nitrate leaching below
1.2 m in a silt loam soil. Jemison et al. (1994) reported accurate predictions using
LEACHM in manure fertilized com crops.


144
may also be affected by climatic conditions and the incidence of pest and diseases.
Cumulative N losses for flowers, fruit, and leaves were 20.3,23.4, and 49.3 g tree'1,
respectively. Assuming that all N from senesced plant parts is incorporated into soil
organic matter, and that this rate of N addition is similar to previous years, 93 g or more
of N may be available on an annual basis due to mineralization. Using the N balance
calculated for mature citrus trees grown on Carrizo and Swingle rootstocks in Chapter 3,
600 g N tree'1 could be available to the trees. This amount of N would reduce N uptake
efficiency to 78.5 and 67.2% for trees grown on Carrizo and Swingle rootstocks,
respectively. These values are near the upper limit of NUE values estimated for citrus in
studies by Syvertsen and Smith (1996).
Conclusions
Leaf, twig, and branch bark N concentrations decreased through the spring to
minimums in May and June of each year. This time period corresponds to a period of
high vegetative and reproductive growth rates. High NUE in May compared with October
/indicates that the reduction in tissue N concentration is not due to ability of the tree to
extract available N from the soil, but possibly a redistribution of N from leaf, twig and
branch bark tissues in response to low N supply. Tissue N concentrations recovered by
late summer, approaching winter values. Under Florida conditions, NH-N was rapidly
converted to NO3-N. Nitrogen uptake rates were greater in late spring when soil
temperatures were high and leaf N concentrations were low, compared with late summer
when soil temperatures were similar and leaf N concentrations were higher. This
relationship indicates a correlation between tissue N concentration and N uptake rates.
Such a relationship is fundamental to modeling N uptake in any crop. Tree biomass and


45
weighed separately. Two to three samples equal to 5% of the fresh weight of each size
category were removed from each container and placed into labeled plastic bags. Any
leaves attached to these branches were removed and weighed prior to weighing the
branch segments. The trunk and taproot were cut into pieces and weighed, and three
longitudinal slices of each were retained as separate samples.
The trees were planted 3.1 m in row and 6.2 m between rows. Therefore, the roots
were cut to a depth of approximately 0.3 m using a shovel in a rectangle 3.1m in-row x
6.2 m across-row with the tree stump at its center. The bulk of the root system was
extracted using a front-end loader equipped with a root rake. All roots were removed
from the excavation to a depth of 1 m, washed, blotted dry, separated into size categories,
and weighed in the field.
Sample Processing and Nitrogen Analysis
The leaf area of 50 random leaves from each sample was measured. Each branch
segment of each sample was cut into at least five disks of approximately 0.5 to 1 cm thick
that facilitated the removal of bark from the wood. Likewise, the bark was removed from
each horizontal trunk slice. The bark and wood from the branch and trunk disks were
weighed separately to determine the fresh mass proportion of bark to wood for each size >'
category.
Samples were dried at 70 C to a constant weight before recording dry weight.
Total tissue dry weight for each tree was determined by multiplying fresh weights by the
respective tissue dry matter content. All tissues were ground using a Cyclotec mill (1093
Sample Mill, Tecator manufacturing, Sweden) for the leaf tissue and Wiley mill model 1
(Arthur Thomas Manufacturing Co., Philadelphia, Pa) for woody tissue. The ground
tissues were digested using a Buchi Model K435 12-vessel digestion unit (Buchi


Table 3-4. Nitrogen accumulation and allocation between tree components for mature Hamlin orange trees as affected by year of sampling,
rootstock, and interaction of year and rootstock.
Nitrogen Accumulation Nitrogen Allocation
Total Above Below Total Total Branches Total Roots
Mass Ground Ground Leaves Twigs Sm. Med. Lrg. Total Trunk Sm. Med. Lrg. Tap
(kg N tree'1) (% of total weight)
Year 1
0.81
0.60
0.21
38.3
8.2
10.C
6.0
9.0
24.9
2.01
10.1
6.1
5.5
4.8
Year 2
0.77
0.57
0.20
36.5
8.3
8.6
4.8
10.8
24.2
2.3
10.1
7.0
3.6
4.9
NSZ
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
**
NS
Carrizo
0.87
0.65
0.22
36.4
8.5
9.2
5.4
11.5
26.1
2.3
10.1
7.4
4.2
3.5
Swingle
0.69
0.50
0.19
38.4
8.0
9.2
5.3
8.2
22.6
2.1
10.1
5.6
4.8
6.4
NS
NS
NS
NS
NS
NS
NS
*
NS
NS
NS
*
NS
NS
Car Yrl
0.87
0.66
0.21
38.1
8.8
9.4
5.6
11.6
26.6
2.2
9.4
7.3
4.9
2.7
Car Yr2
0.86
0.64
0.22
34.7
8.1
9.1
5.1
11.4
25.6
2.5
10.7
7.6
3.4
4.4
Swi Yrl
0.71
0.50
0.21
38.6
7.4
10.9
6.5
5.0
22.3
2.0
11.1
4.4
6.4
7.9
Swi Yr2
0.68
0.51
0.17
38.2
8.4
8.0
4.4
10.3
22.7
2.2
9.5
6.4
3.7
5.5
NS
NS
NS
NS
NS
NS
NS
*
NS
NS
NS
NS
NS
NS
*NS = not significant, = significant P<0.05, and ** = significant P<0.0


160
K TAW-fl
8 TAW-RAW
Where:
Ks = Soil water stress coefficient
TAW = 0fc Opwp = Total available water (cm3 cm'3)
0 = Soil water content (cm3 cm'3)
RAW = 0fc 0ra = Readily available water (cm3 cm'3)
6-1. Mechaelis-Menton equation for estimation of nutrient uptake 117
I Imax Cu /(Km+Cu)
Where:
I = inflow flux of nutrient (mol cm'2 s'1),
Imax = maximum active flux (mol cm'2 s'1),
Cl, = nutrient concentration in the soil solution at the root surface (mol cm'3),
Km = Cl, value at \m.J2 (mol cm'3),
6


117
uptake. The active uptake of nutrients across a membrane is enzyme-catalyzed, thus the
relationship of uptake rate to nutrient concentration is hyperbolic with a maximum uptake
rate at the nutrient concentration where available enzymes are saturated. The Michaelis-
Menten equation (Equation 6-1) is often used to estimate uptake rates for crop plants
under given nutrient concentrations.
I Imax Cl, /(Km+CLa)
Equation 6-1
Where:
2 1 A V
I = inflow flux of nutrient (mol cm s ),
Imax = maximum active flux (mol cm'2 s'1),
Cu = nutrient concentration in the soil solution at the root surface (mol cm'3),
Km = Cl, value at Imax/2 (mol cm'3),
Numerous reports suggest that actively growing tissues such as young developing
leaves and fruit constitute the strongest sink for N uptake (Dasberg, 1987; Feigenbaum et
al., 1987; Legaz et al., 1982). Accumulated N was found primarily in fruitlets and newly
developed leaves and twigs. Absorption rates increased from the beginning of growth and
flowering, reached a maximum at the second shoot growth flush (July), and then declined
through dormancy. Only about 20 to 30% of the new leaf and fruit N originated from the
labeled source, suggesting considerable redistribution from stored reserves.
Mooney et al. (1992) observed an N concentration gradient between the roots,
trunk and branches of citrus trees in New Zealand. High concentrations were found in the
branches, with lower concentrations in the roots. Nitrogen concentrations in the trunk
were highest at bud break and declined steadily through fruit set and development with a
minimum at harvest. Kato et al (1982) found that total N content decreased in both bark


156
impacts. This study has improved our understanding of seasonal and long-term N
accumulation by citrus. The information generated will be essential to refine citrus N
BMPs using sound, science-based decision making. Improved BMPs will allow for
sustainable productivity while safeguarding the environment from leaching of excess
NO3-N to ground water.
Current citrus N BMPs base the fertilizer N application rate on the chronological
age of young citnis trees and stress N fertilization timing in the spring and fall of the year
to avoid potential leaching during the summer rainy season. Long-term N accumulation
measured in this study indicate that tree size is a more useful method of determining
potential N demand of young trees than tree age. Likewise, seasonal N demand by mature
citrus trees was greatest during the spring of the year; therefore a citrus N balance will
provide a better basis to accurately determine N fertilizer requirements of mature trees.
Soil water content and labile N concentration in the root zone are keys to N
uptake efficiency of citrus. Floridas sandy soils hold little water and dry quickly,
reducing potential evapotranspiration much more rapidly than finer textured soils. This
decrease in potential uptake impacts passive and active N uptake through reduced water
uptake and N diffusion to the root boundary. Therefore, soil water content and N
concentrations in the soil volume containing the highest root mass must be maintained as
high as practical without forcing N below the root zone. This study showed that two
rootstocks thought to have similar root growth patterns in reality had different root
densities with distance from the tree and with soil depth. The implication of this finding
is that the top 30 cm soil layer may be more critical for water and N fertilizer
management for trees on Swingle citrumelo than for trees on Carrizo citrange.


36
Crop Models
Crop-Environment Resource Synthesis (CERES) was developed to model growth and
yield of grain crops (Jones and Kimiry, 1987; Kiniry and Bockhot, 1998; Kiniry et al.,
1997; Lizaso et al., 2001; Saseendran et al., 1998). CROPGRO was initially developed as
a family of crop-specific models for the prediction of legume and vegetable crops
(Hoogenboom et al., 1994; Jones et al., 1991; Wagner-Riddle et al., 1997). These are
process-oriented models for the simulation of vegetative growth and reproductive
development. They predict dry matter growth (Shen et al., 1998), crop development
(Batchelor et al., 1994; Batchelor et al., 1997; Piper et al., 1996) and final yield
(Batchelor et al., 19%; Heinemann et al., 2000) for a range of agronomic crops. Inputs
are daily weather data, soil profile characteristics, and crop management conditions
(Gijsman et al., 2002). Crop and soil water (Hoogenboom et al., 1994; Gabrielle et al.,
1995; and Xie et al., 2001), N (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995;
and Sexton et al., 1998), and C balances are modeled. These models have been combined
into the DSSAT (Decision Support System for Agrotechnology Transfer) software
(Hoogenboom et al., 1994; Jones and Luyten, 1998).
Conclusions
Considerable research and resources have been devoted to improving our
understanding of how cultural, soil, and environmental factors influence biomass and N
accumulation during citrus tree development. However, these studies compared tree
component dry weights and N accumulations with tree age and not a measure of tree size.
Tree size is not only a function of tree age, but soil, environmental, and horticultural
factors as well. Therefore, correlation of dry weights and N accumulations with tree size


121
and 4.3 g of wet soil were placed into a centrifuge tube and the mass was recorded. Forty
mL of 2 M KC1 was placed into each tube immediately after the soil was weighed. The
tubes were placed in a shaker for 1 h. The solution was filtered into vials that were then
capped. Extracts were refrigerated at 4 C until they could be analyzed for NO3-N and
NH4-N. Extracts were analyzed using a model FS3000 Rapid Flow Analyzer (O I
Analytical, College Station, Texas). USEPA methods 351.2 and 353.2 were used for
ammonium and nitrate analysis, respectively. The remaining soil sample was used to
determine gravimetric soil water content of the soil sample extracted.
The N concentration in the soil on a dry weight basis was determined. Volumetric
soil water content of each sample was used to determine N concentration in the soil
solution. Total N, NH4-N, and NO3-N contents in the soil by sample depth were
estimated using the soil solution N concentration and area of the irrigation emitter. Daily
NH4-N, and NO3-N change for each sample depth was determined by comparing daily N
content estimates for soil inside and outside of the pipes.
Experiment 2 Seasonal Tissue N Concentration
Tissue samples were collected from three replicates of 14-year-old Hamlin
orange on Carrizo citrange and Swingle citrumelo trees fertilized at an annual rate of 179
or 269 kg N ha1. Samples were taken at approximately 6 week intervals corresponding to
specific stages of growth through the year for two seasons. The growth stages used were
1) bloom and spring flush (early March), 2) fruit drop and second vegetative flush (mid
May), 3) first summer flush (early July), 4) second summer flush (mid August), and 5)
pre-harvest (mid October). Tissue N concentrations from these samples were determined
and used to estimate seasonal change in N concentrations.


ACKNOWLEDGMENTS
I would like to acknowledge those people without whose help this study would
have been impossible. Foremost, I would like to thank my cochairmen, Drs. Thomas
Obreza and Johannes Scholberg, not only for financial and moral support beyond my
expectations, but most of all for their patience. I would like to thank Dr. Adair Wheaton
who shared so willingly his laboratory space and knowledge gained over 40 years of
research. Many thanks are owed to Drs. Nick Comerford and Jim Jones for providing
much needed insight into root uptake and crop modeling, respectively.
This project would not have been possible without the help of many people during
the collection of samples and chemical analysis. The dedication of Tom Graham and Sam
Luther for providing essential organization to the collection of more than 4000 plant
tissue and 3100 soil samples and Marjie Cody and Amanda Myers during many hours of
sample preparation and tissue analysis was greatly appreciated. I would be remiss if I did
not thank Drs. Harold Browning, Bill Castle, and Larry Parsons for allowing me the time
off from my full-time job to pursue this degree. Their patience when the demands of the
job and degree delayed the delivery of data or work on their projects was much
appreciated. My sincere thanks are also due to the Florida Department of Agriculture and
Consumer Services and Cargill, Inc. for providing funding for this project.
ii


43
Site Descriptions
The soil series at the Orange county site was Candler fine sand (hyperthermic,
uncoated, Typic Quartzipsamment), and at the Polk county site was Zolfo fine sand
(sandy siliceous, hyperthermic Grossarenic Entic Haplohumod). These two soils are
typical of the central Florida ridge and have a field capacity water content of 0.06 to 0.08
cm cm'3 in the upper 1 m. The Candler series consists of excessively drained, very
rapidly permeable soils formed from marine deposits. These soils are located in upland
areas and typically have slopes of 0-12%. The A and E horizons consist of single-grained
fine sand, have a loose texture, and are strongly acidic. A Bt horizon is located at a soil
depth of 2 m and includes loamy lamellae of 0.1 to 3.5 cm thick and 5 to 15 cm long.
Zolfo series soils are sandy and slightly less well drained than those of Candler. The A
horizon consists of fine sand with single-grained, loose texture. The Bh horizon between
4.0 and 5.0 cm consists of fine sand coated with organic matter possessing weak granular
to weak fine subangular blocky structure.
Tree Canopy Volume and Trunk Cross-Section Area
Changes in canopy volume have been used in fertilizer rate experiments as
measures of tree growth (Whitney et al., 1991). Therefore, tree measurements were
determined for the purpose of correlating biomass and N concentration of various tree
components to tree size. Canopy diameter of each tree was measured in the row (in-row)
and across the row (cross-row) at a height of 1.5 m above ground level. Tree height and
hedgerow intercept measurements were made using a 5 m graduated pole. Hedgerow
intercept is the height from the ground to the point at which the canopies of two trees
meet in the row. These measurements have been used by Whitney et al. (1991) to
determine canopy volume based on a spheroid model (Equation 3-1). Trunk diameters 5


Table 6-5. Seasonal changes in N concentration, size and dry wt. of fruit, flush leaves, and expanded leaves during 2001 and 2002
seasons. There were three replicates of each N rate on two rootstocks (Carrizo citrange and Swingle citrumelo) for a total of 12
measurements per date.
Fruit Flush Leaves Expanded Leaves
Date
Collected
[N]
(%)
Diameter
(mm)
Dry Mass
(g fruit'1)
[N]
(%)
Leaf Area
(cm2 leaf1)
Dry Mass
(gleaf1)
[N]
(%)
Leaf Area
(cm2 leaf1)
Dry Mass
(g leaf1)
179 kgN ha*1 yr'1
3/16/01
2.02
10.0
2.75
33.7
5/18/01
1.33
38.7
6.0
1.96
21.1
0.2
2.02
32.4
0.38
7/2/01
1.29
51.0
6.3
2.35
25.0
0.2
2.29
29.7
0.39
8/13/01
1.13
62.3
15.1
2.38
22.8
0.2
2.24
30.0
0.36
10/13/01
1.11
66.5
21.0
2.11
20.2
0.2
2.29
41.4
0.38
4/4/02
2.18
28.5
0.2
2.14
33.8
0.33
5/20/02
1.18
36.9
5.6
1.90
15.7
0.3
2.00
32.6
0.51
7/15/02
0.99
53.4
11.1
2.19
20.3
0.1
2.27
20.1
0.21
9/9/02
1.02
63.4
15.0
2.50
23.2
0.2
2.19
26.2
0.23
10/25/02
0.95
65.4
18.2
2.54
22.6
0.2
2.24
28.8
0.25
269 kg N ha'1 yr'1
3/16/01
2.75
11.7
2.35
35.3
5/18/01
1.34
39.5
6.2
2.06
20.7
0.1
2.13
29.0
0.31
7/2/01
1.39
51.5
6.5
2,28
24.9
0.2
2.35
31.4
0.40
8/13/01
1.10
62.1
14.9
2.42
20.6
0.1
2.37
35.0
0.36
10/31/01
0.98
64.1
18.9
2.52
23.8
0.2
2.64
49.5
0.55
4/4/02
2.16
29.2
0.2
2.30
32.2
0.34
5/20/02
1.26
35.8
5.1
1.92
17.2
0.3
2.25
31.9
0.51
7/15/02
0.97
54.0
11.5
2.22
20.0
0.1
2.40
22.5
0.24
9/9/02
1.15
63.9
15.1
2.44
23.8
0.2
2.53
28.7
0.25
10/25/02
1.13
67.3
17.8
2.53
22.9
0.2
2.48
30.3
0.25


APPENDIX A
EQUIATIONS
Equation
2-1. Darcys Law water flow through a saturated medium
u =-K d<|) /dx
Where:
u = water flux (cm3 cm'2 s'1),
K = hydraulic conductivity constant (cm s'1),
= soil water potential (kPa), and
x = the distance over which the flux is maintained (cm).
2-2. Poisevilles equation water flow through a tube
f = (tc r4/ 8t)) dP/dx
Where:
f = flow rate in a tube (m3 s'1),
r = radius (m),
q = viscosity (pPa s'1), and
dP/dx = pressure gradient.
2-3. Richards Equation water flow in unsaturated soils
o = -Ke d/dx = -Ke (d/d0) (d0/dx) = -De (d0/dx)
Where:
u = water flux (cm3 cm'2 s'1),
Pace
..23
24
25
158


126
Table 6-2. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in May, 2002. Rootstocks are Carrizo citrange, and
Swingle citrumelo; high N application rate was 269 kg ha1 yr'1, low rate was 134 kg ha1
yr1.
Days After
Application
Weighted
Solution
N
(mgL1)
N Uptake
Passive Active Total
(g tree1 d1)
Cumulative
Total
N Uptake
(g tree1) (%)
Cumulative N
Loss
Control Soil
(% applied)
Carrizo High rate
1
156.1
4.6
5.3
9.9
9.9
17.1
8.6
25.7
2
128.7
3.8
8.1
11.9
21.8
37.5
23.2
46.9
3
88.1
1.3
6.1
7.4
29.3
50.3
36.0
65.2
Carrizo
- Low rate
1
103.3
4.1
17.1
21.2
21.2
42.6
17.7
60.4
2
46.5
1.6
9.7
11.2
32.5
65.2
33.5
89.2
3
20.9
0.3
1.7
2.0
34.4
69.2
26.8
92.5
Swingle High rate
1
195.7
5.6
12.4
18.0
18.0
23.9
-4.5
23.9
2
179.2
4.8
10.5
15.3
33.3
44.1
2.1
59.1
3
139.1
2.2
6.2
8.5
41.8
55.3
5.5
72.0
Swingle -
Low rate
1
101.1
4.0
11.1
15.2
15.2
31.4
13.6
45.0
2
44.3
1.1
4.9
6.0
21.1
43.7
28.6
65.6
3
22.8
0.3
1.1
1.4
22.5
46.6
41.0
72.7


137
N concentration values to estimate the N content of these added biomasses. Estimated
leaf and twig flush gains were 108.8 and 95.2 g tree'1 for trees grown on Carrizo and
Swingle rootstocks, respectively. Fruit N gains amounted to 90.6 and 77.4 g tree'1 for the
same rootstocks. Nitrogen mass associated with blooms and abscised fruit must also be
added and was assumed to be equal to the N loss in blooms and fruitlets. Therefore, the
total estimated tree N content increase between March and May was 227.4 and 200.6 g
tree'1 for trees on Carrizo and Swingle rootstocks, respectively. These N gains would
represent a fertilizer use efficiency of 80.8 and 71.3% for Carrizo and Swingle trees,
respectively.
Hamlin trees gain little leaf and fruit biomass after October of each year.
Therefore, using the same methods of estimation with 100% of annual leaf twig, and
fruit biomass gain from Chapter 3, 90.7 g tree'1 loss due to senescence was balanced by
gains of 471 and 403 g tree'1 for trees on Carrizo and Swingle, respectively (Chapter 3).
Total net seasonal N content gain was approximately 383 and 312.3 g tree'1, or fertilizer
N use efficiencies of 68.3 and 55.7% for trees grown on Carrizo and Swingle,
respectively.
Seasonal N loss
Flower, leaf, and fruit were collected from March, 2001 to December, 2002. A
relatively small amount of decaying twig and branch material was found in the catch
frames at any one time. The majority of these tissues remained attached to the tree until
greatly decomposed. Accurate seasonal timing and amount of biomass and N loss from
twig and branch material could not be determined without removing all dead twig and
branch material at the beginning of the study and periodic removal of dead material from


[N] (%)
132
Fig. 6-4. Seasonal change in N concentration for flush leaves, expanded leaves, and twigs
during 2001 and 2002. High N rate (A) and LowN rate (B) equal to 269 and 179
kg ha'1 yr'1, respectively.


62
Trunk Cross Section Area (cm2)
Fig. 3-7. Dry weight accumulation to above-ground (closed), and below-ground (open)
biomass (A); leaf (closed) and twig (open) biomass (B); total branch (closed) and
trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D)
as a function of trunk cross section area for HamlinVCarrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.


106
c
92
o
¡E
o
O

(O
£
o5
u.
O)
¡
o
W
S
c/>
OI
Soil Water Potential (kPa)
Fig. 5-7. Estimated soil water stress coefficient Ks as a function of soil water potential in
the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a
1 m depth (C). The data points shown represent a range of soil water content
from field capacity to approximately 50% available soil water depletion.


37
would provide a better relationship for modeling purposes. Factors affecting citrus root
distribution have been studied under Florida soil and environmental conditions. Many of
these studies were performed in groves with lower tree densities and different irrigation
methods than those currently used in Florida citriculture, and on trees grafted on
rootstocks that are no longer in use. Thus, information on the effect of tree size on root
length density distribution changes for current production systems are lacking. Likewise,
root length density distributions for mature trees on currently used rootstocks grown on
Florida sandy soils have not been determined.
Seasonal maximum daily water uptake rates under Florida environmental
conditions have been determined for trees grown on flatwood soils with fluctuating water
tables. However, maximum daily water uptake rates for mature citrus trees have not been
measured for trees grown on excessively drained Ridge soils. Likewise, reduction in
daily citrus water uptake with decreased soil water potential has not been determined for
sandy soils. Citrus N uptake rates have been determined for seedlings and relatively small
trees grown in lysimeters. These rates may not reflect uptake rates of mature citrus trees
at the field-scale.
Much data on citrus growth, root distribution, water requirements, and N uptake
rates are needed to attain the level of crop modeling currently available for agronomic
crops. Obtaining these data are difficult due to the size of mature citrus trees compared
with agronomic crops, and the inability to follow a cohort of trees from planting to
maturity. Biomass, N accumulation, and spatial root length density changes as affected
by tree size, and water and N uptake dynamics of mature trees under Florida Ridge
conditions will be presented in the following chapters.


92
Soil water content (0) must be maintained between specific upper and lower
limits such that water availability to the crop does not limit growth or adversely impact
yield or quality. This upper limit of 0 after free drainage occurs is defined as the value of
0 at which redistribution of soil water ceases (HilleL, 1998) and is also known as field
capacity (0fc)- The lower limit or permanent wilting point (0pwp) is the value for 0 at
which a wilted plant can no longer recover turgidity. The range of 0 between 0fc and
0pwp is known as total available water (TAW). These three values for 0 and the soil water
potential () at which they occur are characteristic and relatively constant for any given
soil. If the effects of soil physical characteristics on soil water use are understood, soil
water can be maintained within these limits and the potential for both crop water stress
and environmental contamination can be minimized.
The soil water depletion coefficient (Ks) in equation 5-1 is a measure of the
reduction in ETC caused by reduced soil water uptake due to reduced 1998). Water moves to regions of high from regions of lower as water is removed
from the soil surrounding the root surface. Soil water movement slows as 4> of the bulk
soil decreases and 0 approaches 0pwp. Allen et al. (1997) determined that a 0 exists (0ra)
less than 0fc where water uptake was not limited by . They referred to the range of 0 to
0ra as readily available water (RAW) and used this value to estimate K* as the ratio of
depletion of total available soil water (TAW-0) to the soil water not readily available
(TAW-RAW) where Ks is not greater than unity (Equation 5-2). Therefore, the greater
the RAW for a given soil, the longer water can be withdrawn from that soil before ETC is
limited. A crop and specific depletion in TAW must be determined below which crop
/ >
growth and yield is reduced. Under Florida conditions, Koo(1963) estimated this
L


168
Koo, R.C.J., and D.S. Harrison. 1965. Summary of irrigation research in Florida.
Florida Agricultural Ext. Service Memo. Report 65-8.
Koo, R.C.J., and G.T. Humer, Jr. 1969. Irrigation requirements of citrus grown on
Lakewood fine sand. Proc. Fla. State Hort. Soc 82:69-72.
Koo, R.C.J. and A.G. Smajstrla. 1984. Effects of trickle irrigation and fertigation on fruit
production and juice quality of Valencia oranges. Proc. Fla. State Hort. Soc.
97:8-10.
Kramer,P.J., and J.S. Boyer. 1995. Water Relations of Plants and Soils. Academic Press,
Inc. New York, NY, USA.
Lafolie, F., L. Bruckler, and F. Tardieu. 1991. Modeling root water potential and soil-
root water transport: I. model presentation. Soil Sci. Soc. Am. J. 55:1203-1212.
Lea-Cox, J.D. and J.P. Syvertsen. 1993. Nitrate-N-use efficiency of citrus rootstock
species is affected by salinity stress. Ann. of Botany 72:47-54.
Lea-Cox, J.D., and J.P. Syvertsen. 1996. How nitrogen supply affects growth and
nitrogen uptake use-efficiency, and loss form citrus seedlings. J. Amer. Soc. Hort.
Sci. 121:105-114.
Lea-Cox, J.D. J.P. Syvertsen, D A. Graetz. 2001. Spring-time nitrogen uptake,
partitioning and leaching from young bearing citrus trees of differing nitrogen
status. A. Soc. Hort. Sci. J. 126:242-251.
Legaz, F., and E. Primo-Millo. 1981. Dynamics of 15N labeled nitrogen nutrients in
Valencia orange trees. Proc Int. Soc. Citriculture. 575-582.
Legaz, F. and E. Primo-Millo. 1988. Absorption and distribution of nitrogen-15 applied
to young citrus trees. Proc. Sixth Int. Citrus Congress, Tel Aviv, Israel.
Legaz, F., E. Primo-Millo, E. Primo-Yufera, C.Gil, and J.L. Rubio. 1982. Nitrogen
fertilization in citrus, I: Absorption and distribution of nitrogen in Calamondin
trees (C mitis Bl.) during flowering, fruit set, and initial fruit development
periods. Plant Soil 66:339-351.
Lizaso, J.I., W.D. Batchelor, and S.S. Adams. 2001. Alternate approach to improved
kernel number calculation in CEREX-MAIZE. Trans, of the ASAE 44 1001-
1018.
Mahrer, Y., and G. Rytwo. 1991. Modeling and measuring evapotranspiration in a daily
drip irrigated cotton field. Irrig. Sci. (1991) 12:13-20.


TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ii
LIST OF TABLES viii
LIST OF FIGURES xi
ABSTRACT xiv
CHAPTER
1 INTRODUCTION 1
Ridge Water Quality Project 2
Citrus Best Management Practices 4
Decision Support Systems 5
Objectives 6
2 LITERATURE REVIEW 9
Introduction 9
Citrus Growth Characteristics 10
Citrus Biomass Distribution 11
Citrus Nitrogen Accumulation and Partitioning 12
Citrus Root Growth Dynamics 13
Factors Affecting Root Distribution And Root Density 14
Soil Characteristics 14
Climatic Effects 16
Rootstocks 16
Tree Spacing and Density 17
Fertilization 17
Irrigation 17
Canopy Reduction 18
Citrus Water Uptake 18
Factors Affecting ETC 19
Crop species 19
Tree size 20
IV


139
Date
Fig. 6-6. Seasonal cumulative dry mass (A) and N content (B) of flowers, fruit,
and leaves collected from catch frames under mature citrus trees during
the 2001 season.


83
Root Length Density Distribution Changes with Tree Size
Citrus root length densities were significantly different at the P=0.01 level for
both distance from the tree trunk and depth from the soil surface across a wide range of
tree sizes (Table 4-2). Three-dimensional graphical representations of developing root
systems are presented in Fig. 4-3. These graphs represent trees approximately 2 to 5 years
old (Figs. 4-3A and 4-3B), 5 to 10 years old (Figs. 4-3C and 4-3D), 10 to 15 years old
(Figs. 4-3E), and >15 years old (Figs. 4-3F). Root systems were initially concentrated at
the surface with few roots deeper than 0.5 m at a distance of 150 cm from the tree trunk.
As the citrus trees produced substantial fruit (5 to 10 years of age) root length density
increased at the soil surface to the dripline of the tree. Roots eventually extended to the
200 cm distance between tree rows and to a depth of 0.9 m at 150 cm from the trunk. The
bimodal nature of the root system can be seen near the tree at depths below 60 cm. By the
time the tree reached 10 to 15 years of age and the canopy was nearing full hedgerow
dimensions, the bimodality of the root system was fully developed and roots extended
past a depth of 1 meter at all distances from the tree.
Tables 4-3 and 4-4 list the regression coefficients for a third order polynomial
relationship of canopy volume and trunk diameter to root length density at all depths and
distances. The r2 values were greater, and RSME and P values were lower for most
regressions using canopy volume than those using trunk diameter, indicating that canopy
volume measurements are a more accurate predictor for assessing root length density
compared with trunk cross sectional area.


171
Roy, W.R. and F.E. Gardner. 1946. Seasonal absorption of nutrient ions by orange trees
in sand culture. Proc. Amer. Soc. Hort. Sci. 47: 107-118.
Saleh, A.R., R.L. Benton, and J.L. Fouss. 1994. Performance of the DRAINMOD
CREAMS model with an incorporated nutrient submodel. Trans, of the ASAE
37(4): 1109-1114.
Saseendran, S.A., K.G. Hubbard, K.K. Singh, N. Mendiratta, L.S. Rathore, and S.V.
Singh. 1998. Optimum transplanting dates for rice in Kerala, India, determined
using both CERES v3.0 and ClimProb. Agron J. 90:185-190.
Scholberg, J.M.S., L.R. Parsons, T.A. Wheaton, B.L. McNeal, and K.T. Morgan. 2002.
Soil temperature, nitrogen concentration, and residence time affect nitrogen
uptake efficiency of citrus. J. Environ. Qual. 31:759-768.
Sexton, P.J., W.D. Bachelor, K.J. Boote, and R. Shibles. 1998. Evaluation of
CROPGRO for prediction of soybean nitrogen balance in a Midwestern
environment. Trans, of the ASAE 41(5): 1543-1548.
Shen, J., W.D. Batchelor, J.W. Jones, J.T. Ritchie, R.S. Kanwar, and C.W. Mize. 1998.
Incorporation of a subsurface tile drainage component into a soybean growth
model. Trans, ofthe ASAE 41(5): 1305-1313.
Sites, J.W., I.W. Wander, and E.J. Deszyck. 1953. The effect of fertilizer timing and rate
of application on fruit quality and production ofHamlin oranges. Proc. Fla.
State Hort. Soc. 66:54-62.
Slatery, R.O. 1967. Plant-wat^relationships. Academic Press, London and New York.
Smajstrla, A.G., L.R. Parsons, F.S. Zazueta, G. Vellidis, and K. Aribi. 1986. Water use
and growth of young citrus trees. Paper No. 86-2069. Am Soc. Agrie. Engr.
Summer meeting, San Luis Obispo, CA.
Smith, P.F. 1956. Effect of phosphate fertilization on root growth, soil pH, and chemical
constituents at different depths in an acid sandy Florida soil. Proc. Fla. State Hort.
Soc. 69: 25-29.
Smith, P.F. 1965. Effect of nitrogen source and placement on the root development of
Valencia orange trees. Proc. Fin. State Hort. Soc. 78: 55-59.
Stewart, E.H, J.E. Browning, and E.O Burt. 1969. Evapotranspiration as affected by
plant density and water-table depth. Trans. Am. Soc. Agrie. Engr. 12(5):646-647.
Starr, J.L., and I.C. Paltineanu. 1998. Soil water dynamics using multisensor capacitance
probes in nontraffic Interrow of Com. Soil Sci Soc. Am. J. 62:114-122.


134
apparent for specific leaf weights. Minimum twig N concentrations in May were 0.83%
for both annual N application rates. Maximum twig N concentrations were 0.98 and 1.02
% for the 179 and 269 kg N ha'1 annual application rates, respectively. These mximums
occurred in August.
Branch bark N concentrations remained within a narrow range from 1.0 to 1.3%
during the 2-year period (Fig. 6-5). Mean branch bark N concentrations were 1.04 and
1.17% for low and high N application rates, respectively. Minimum values of 1.00 and
1.11% occurred in May or July of each year. Maximum bark N concentrations were 1.09
and 1.22% for low and high annual application rates, respectively. These maximum
values occurred in October and January. Branch wood N concentrations were lower, but
followed similar trends as those of branch bark tissue (Fig. 6-5). Mean wood N
concentrations were 0.25 and 0.31% for the low and high N application rates. Maximum
wood N concentrations were 0.37 and 0.38% for low and high N application rates,
respectively, and occurred in January and March. Minimum wood N concentrations were
0.23 and 0.29% for low and high N application rates, respectively. These mnimums
occurred in October.
Root N concentrations were greater for roots <4 mm in diameter than for roots >4
mm. Mean N concentrations for roots <4 mm in diameter were 1.35 and 1.34 % for high
and low N application rates, respectively. Mean N concentrations for roots >4 mm in
diameter were 0.85 and 0.89% for high and low N application rates, respectively.
Seasonal trends of N concentration for roots were not as consistent as with other plant
tissues.


15
concluded that tree size was closely related to fibrous root density. Extensive lateral root
development occurred on soils with loamy or clay texture (Boswell et al. 1975;
Kaufimann et al., 1972). In these studies, the root systems were shallower than root
systems of plants grown in sandy soils with few roots found below a soil depth of 50 to
70 cm (Adriance and Hampton, 1949; Boswell et al., 1975, Cahoon et al., 1956; 1959;
1961; Kimball et al., 1950; Mikhail and El-Zeflawi, 1978). Furthermore, changes in
fibrous root distribution with depth were more gradual compared with sandy soils, and
overall fibrous root concentrations were lower (Bielorai, 1977). The lower natural soil
fertility (Carlisle et al., 1989), and excessive drainage of sandy soils resulted in higher
shoot:root ratios such that fibrous root dry mass densities tended to be lower in sandy
soils (Castle, 1978).
Under flatwood conditions where the soil is drained and bedded, virtually all the
root mass occurs within 45 cm of the soil surface (Calvert et al., 1967; 1977; Ford,
1954a; Ford, 1972; Reitz and Long, 1955). The quantity of fibrous roots decreases with
depth and lateral distance from the trunk. Elezaby (1989) reported lateral fibrous root
distribution to a depth of 180 cm of a 10-year-old Valencia tree [Citrus sinensis (L.)
Osb ] on Volkamer lemon (C. volkameriana Ten. and Pasq.) grown on a soil with a
deep sand profile and spaced at 4.5 m x 6.0 m as: 9% of the fibrous roots between 0 cm
and 60 cm from the trunk, 31% between 120 cm and 180 cm, and 21% between 240 cm
and 300 cm. The vertical distribution was: 42% of the fibrous roots between 0 cm and 30
cm from the soil surface, and 14% or less at each 30-cm depth increment to 180 cm. In
the same study, fibrous root dry mass density (concentration) ranged from 300 g m3 to
1200 g m Those data are similar to dry mass densities reported in other Florida studies


8
in root distribution with time must be known to understand the effect of these
distributions on the rate of water and nutrient uptake and thus soil water and N
concentration. An additional complication is that citrus tree scions are grafted onto
rootstocks that affect the growth and uptake rates of the resulting tree. The effects of
these rootstocks on tree development and uptake rates must also be understood.
To develop a crop model for citrus, detailed field-scale information must be
obtained under local soil and cultural conditions. A review of the existing literature for
these relationships was conducted. The result of this review will be presented in Chapter
2. Information gaps in biomass and N accumulation and field-scale uptake rates of mature
citrus trees must be determined so that studies can be designed to complement existing
information. The central hypotheses tested in this dissertation are 1) generic relationships
can be developed that capture changes in citrus dry biomass, N weight, and root length
densities with increase in tree size; 2) daily water uptake changes seasonally and is
greatly affected by soil water content; 3) seasonal leaf N concentration is lowest and tree
biomass abscission is highest during periods of rapid tree growth; and 4) fertilizer-N
nitrification and uptake are rapid under Florida conditions. Therefore, the objectives of
studies in Chapters 3-6 were to 1) determine changes in above-ground citrus biomass and
N distribution for trees under recommended N fertilizer management practices across a
range of tree sizes; 2) develop relationships that will capture the overall spatial patterns in
citrus root length density distribution for different tree sizes; 3) estimate irrigation crop
and soil moisture coefficients and soil water use per unit root length density; 4) explore
seasonal N uptake rates for citrus; and 5) compare seasonal biomass and N concentration
changes for citrus fertigated at two N fertilizer rates.


33
Nitrogen Uptake Efficiency
Nitrogen uptake efficiency (NUE) is defined as the percentage of applied N taken
up by plants (Scholberg et al. 2002). The ability of crop plants to take up and utilize N
efficiently is key to providing adequate N for crop growth while reducing N leaching.
Mattos (2000) estimated NUE for 6-year old Valencia trees grown in a sandy soil to be
40% and 26% for ammonium nitrate and urea respectively. Feigenbaum et al. (1987)
reported that the NUE for a 15N labeled KNO3 applied to 22 year-old Shamouti orange
was 40%. Syvertsen and Smith (1996) estimated NUE to be 61% to 83% for 4-year old
grapefruit trees grown in lysimeters. Nitrogen uptake efficiency decreased with increased
N application rates. Lea-Cox and Syvertsen (1996) reported a similar finding of lower
NUE with higher N application rate for greenhouse grown seedlings. The NUE reported
ranged from 47% to 60% after an uptake period of 31 days.
Kato et al. (1982) found a 10-fold increase in 15 N uptake ofSatsuma mandarin
during summer (mean temperature 23 C) compared with the winter season (mean
minimum temperature 3 C). Scholberg et al. (2002) found N uptake of greenhouse-grown
seedlings to be proportional to soil temperature, ET0 and canopy biomass. Nitrogen
uptake also increased with the time high N concentrations were maintained in the root
zone. Increasing the residence time from 2 to 8 hours resulted in an increase in NUE of
95% and 125% for high and low N application rates, respectively.
Seasonal Nitrogen Redistribution
Legaz et al. (1982) suggested that at post-blossom, the N concentration in the
spring leaves decreased due to this tissue becoming an N source for the developing fruit.
Using 4 year-old Valencia orange trees, daily root N uptake was lower during


66
O 20 40 60 80 100 120 140 160 180
Trunk Cross Section Area (cm2)
Fig. 3-9. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and trunk (open)
N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of trunk cross section area for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.


52
Fig. 3-1. Tree canopy volume as a function of trunk cross sectional area for trees
from experiments 1 and 2.


109
equations presented in Table 5-2. RMSE values in Table 5-4 were generally 25% lower
than those in Table 5-2. The R2 values were greater for equations made using ASWD and
<|> in the irrigated zone to a depth of 0.5 m compared with those using the irrigated area
and total tree area to a 1 m depth. This result indicates that the soil volume with greater
root length densities dried out faster resulting in better correlation of Ks with both ASWD
and 4. Estimated values for K* are approximately 1 at 9fc
Soil Water Uptake per Unit Root Length
Equations resulting from regression analysis of soil water uptake per unit root
length density against mean daily are presented in Table 5-6. An exponential decay
model typically resulted in the best fit. Water uptake per unit root length values were
remarkably similar for the various locations and depths, with the exceptions of the 10 cm
depth between-rows and the 40 and 80 cm depths at the dripline. Maximum water uptake
per unit root length of 0.4 mm3 cm'1 d'1 occured at 9fc or approximately -5 kPa. Water
uptake per unit root length decreased rapidly as 4> decreased to approximately -12 or -13
kPa, then gradually decreased from 0.1 to 0.05 mm3 d1 cm'1 as kPa. A wide scatter in the data along with a long shallow sloping tail resulted in a
relatively low R2 and generally high RMSE values. However, all regressions were
significant at the P=0.01 level and formed an approximation of water uptake at given
bulk 4 within the constratints of the RMSE.
Water uptake per unit root length values of approximately 0.8 mm d'1 cm'1 at 9fc
were found at the 10 cm depth between-rows. This value is double that at other locations
and depths (data not shown). The increase in water uptake could be explained by water


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EVPQH1HJG_QENRIW INGEST_TIME 2015-04-01T19:37:00Z PACKAGE AA00029855_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


Nutrients leached from agricultural soils represent both an economic loss to
farmers and a potential environmental pollutant for groundwater. Concerns about the
presence of these agricultural chemicals in groundwater and the need for improved
understanding of their movement and transport beyond the root zone have increased
considerably over the last several decades. Comprehensive mathematical relationships are
required to determine crop fertilizer N uptake and to predict the potential impact of NO3-
N leaching on groundwater quality for various soil and/or environmental conditions. With
the exception of insect and disease population and damage dynamics, most modeling
work has focused on predicting mineral N transformations, organic C and N
transformations, soil water content, water and N uptake, crop yield, and NO3-N leaching
(Hoogenboom et al., 1994). Such models help growers manage resources, maximize
returns, and reduce impacts on water quality. Current crop simulation models are being
used to optimize planting dates and densities (Saseendran et al., 1998), optimize fertilizer
and irrigation inputs (Sexon et al., 1998, and Heinemann et al., 2000), maximize profits
(Kiniry et al., 1997), and reduce groundwater pollution (Gijsman et al., 2002) in
agronomic crops.
Objectives
Robust crop models can provide a scientific basis for improved resource
management in agricultural production. The long-term accumulation of biomass and N
with tree development must be understood. Figure 1-2 illustrates the relationships
between tree biomass, tree N content, soil N concentration, and soil water content.
Change in tree biomass and N content over time impacts total tree N demand. Soil N and
water concentrations affect both active and passive N uptake rates. Likewise, changes


11
distribution or uptake have been conducted, but few biomass studies have been conducted
for a range of tree sizes grown under similar cultural soil and climatic conditions.
Citrus Biomass Distribution
Legaz and Primo-Millo (1981) harvested 4 year-old Valencia orange trees
grown outdoors in sand culture in Spain at five different times in a 1-year period. The
mean dry mass percentages for leaves, twigs and branches, lateral roots and fibrous roots
were 22.5, 28.7, 45.8, and 2.9% respectively.
Cameron and Compton (1945) divided eight year old Valencia trees grown in
California into 14 parts, 1) leaves, 2) twigs, 3) shoots, 4) lateral branches 0.75-1.5 cm, 5)
tertiary branches 1.5-3.0 cm, 6) secondary branches 3.0-6.0 cm, 7) primary branches >
6.0 cm, 8) trunk, 9) main root, 10) feeder roots, 11) rootlets, 12) small roots 0.3-0.8 cm,
13) intermediate roots 0.8-2.5 cm, and 14) large roots >2.5 cm. The difference between
twigs and shoot was not given, but both had similar N concentrations on a dry mass basis
(5.11 and 5.02%, respectively). Likewise, the sizes of rootlets and feeder roots were not
given but had similar total N percentages (1.38 and 1.49%, respectively). Mean
percentage dry weight of leaves, twigs and shoots, branches, trunk, main root, lateral
roots and fibrous roots were 16.8, 9.6, 43.2, 1.5, 2.1, 14.0, and 1.7% respectively.
The biomass proportions for young Hamlin orange trees grown under Florida
conditions were substantially different. In a study by Mattos (2000), eight categories of
plant parts for 7 year-old Hamlin orange trees were used. These categories were 1)
summer-fall flush leaves, 2) spring and older leaves, 3) twigs >1.5 cm, 4) twigs <1.5 cm,
5) trunk, 6) roots <0.2 cm, 7) roots 0.2 to 1.0 cm, and 8) roots >1.0 cm. Leaves, branches,


77
relationship will provide a scientific basis for the development of water and nutrient
components of an expert system for improved citrus irrigation and N management.
Materials and Methods
Sample Collection
The same 19 Hamlin and Valencia orange trees used in the previous biomass
and N distribution study (Chapter 3) were used to determine the spatial relationship
between citrus root length density and tree size. Soil cores were removed once the trees
had been cut to the ground but prior to the excavation of the main root system. Cores
were taken with a 7.6 cm diameter bucket auger and roots were sampled at 50,100, and
150 cm from the tree trunk in the row and 50, 100, 150 and 200 cm between tree rows.
Samples were collected at 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 90 cm depths.
Each sample was placed into separate plastic bags, sealed, and marked with tree
identification, depth and distance from the tree. The samples were placed in a cooler
containing ice and were subsequently frozen at -4 C.
Sample Processing and Statistical Analysis
Roots were removed from the soil by washing though an 850 an sieve. Any
debris not passing through the sieve was removed manually, and the roots were separated
into size categories by diameter. These categories were <4 mm, 4 to 20 mm and >20 mm.
Root lengths for roots 0 to 4 mm in diameter were determined prior to drying using the
line intersect method (Newman, 1966). Root length density data from samples collected
from the 12 trees at the Water Conserv II site (mature tree study) were analyzed by the
general linear model procedure of SAS (SAS Institute, Inc., Cary, NC). Root length
density data from the soil samples collected from the trees of various sizes at the Cargill


18
et al., 1964) because trees in the low irrigation rate treatment declined in yield while
maintaining root quantities similar to those of the trees in the higher irrigation rate
treatment. It was concluded that soil water content was the single most important factor
influencing citrus root systems.
In a Florida study, root weight densities were determined under the tree canopy, at
the dripline, and in the row middles to a depth of 180 cm for Hamlin orange trees on
Swingle citrumelo and Carrizo citrange rootstocks (Menocal-Barberena, 2000). Trees
receiving irrigation at a rate of 40 cm yr'1 had significantly higher densities than trees
receiving 250 cm yr'1. The differences were on the order of 1.3 to 2.3 times greater for
the 40 cm yr'1 treatment at all depths.
Canopy Reduction
Hedging, the annual removal of excess vegetative growth, has become a common
method of canopy size control for closely-planted citrus trees. Eissenstat and Duncan
(1992) found that within 30 days of canopy reduction, 20% of the fibrous roots between
the 9 cm and 35 cm depths were apparently dead. Root length density of these trees
recovered within 63 days of canopy reduction. This relatively short-term reduction in
fibrous root density adversely affected yield because of fruit abortion.
Citrus Water Uptake
Assuming little or no surface runoff, water applied to the soil surface is 1)
retained in the soil, 2) utilized by plants, 3) lost to the atmosphere, or 4) drained below
the crop rooting zone. Drainage water may contain substantial quantities of agricultural
chemicals and soluble nutrients. Irrigation practices should be aimed at minimizing 1)
crop water stress by maintaining sufficient water within the crop rooting zone, 2)


67
Canopy Volume (m3)
Fig. 3-10. Nitrogen allocation to above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (closed) and taproot (open) (D) as a
function of canopy volume for HamlinVCarrizo experiment 1 (O),
HamlinVSwingle experiment 1 (), and experiment 2 (A) trees.


107
1.2-
0 10 20 30 40 50 60
Available Soil Water Depletion (%)
Fig. 5-8. Estimated soil water stress coefficient Ks as a function of available soil water
depletion in the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total
tree area to a 1 m depth (C). The data points shown represent a range of soil
water content from field capacity to approximately 50% available soil water
depletion.


129
Table 6-4. Regression equations for estimated maximum N uptake and estimated active N
uptake rates by soil N concentration (mg l'1) using an exponential rise to a maximum
model2 and linear modely, respectively.
Y0
a
b R2
Maximum N uptake
RMSE
(g tree'1 d1)
P
-2.44
17.31
0.013 0.46
Active N uptake
3.51
<0.0001
-0.78
0.06
0.63
3.41
0.013
z Y = Y0a(l exp bx) where X = soil N concentration, and a, and b are regression
coefficients
y Y = Y0 + aX where X = soil N concentration, and a, and b are regression coefficients
Passive N uptake was estimated by determining daily water uptake from the
irrigated area to a depth of 45 cm using the water uptake equations presented in Chapter
5. Soil solution N contents were estimated for the same area and depth using the soil N
and gravimetric soil water content values from the daily soil samples. Passive N uptake
was estimated using the assumption that daily passive N uptake was equal to the product
of soil solution N concentration and estimated daily water uptake. Estimated passive N
uptake was subtracted from the estimated maximum N uptake to estimate daily active N
uptake. The regression of estimated active N uptake and soil solution N concentration
was significant at the P=0.05 level (Table 6-4). Due to the compounded error associated
with the estimation of passive uptake using the regression equations for water uptake, the
associated R2 was lower than that for soil overall N uptake (Table 6-4). However, the
RMSE was similar indicating a confidence interval similar to that of the regression of
daily maximum uptake (Fig. 6-1). The relationship of active uptake and soil solution
concentration was linear over the concentration range used in this study (Fig. 6.2), so no
Michaelis-Menten equation constants could be determined.


Discussion
The mature Hamlin trees used in this study were planted at the same time and
received the same horticultural inputs (i.e. fertilizer rates, irrigation schedule, and pest
control) for the past 14 years. Total leaf area of each tree differed with tree size, but LAIt
and LAIa appeared to approach maximum values of 10 and 6.5, respectively. The mean
LAIt value of 10 is well within the 9 to 11 range found by Syvertsen and Lloyd (1994) for
mature citrus trees. These values are much higher that the 3 or less associated with
agronomic row oops (Flenet et al., 1996). Citrus is thought to have developed as an
understory plant in subtropical rainforests and so has a high tolerance to shade (Syvertsen
and Lloyd, 1994) and therefore developed a dense canopy. A large fraction of citrus
leaves found within the inner canopy receive 10% or less of recorded on the outermost
leaves (Cohen et al., 1987) The observed LAI would indicate that, on average, 10 layers
of leaves exist over each unit area of soil under the tree canopy. This finding has
significant implication for light interception and photosynthesis, indicating that citrus is
an efficient interceptor of light and allows very little of it to strike the soil surface. Leaves
in the interior of a citrus tree are adapted for low light levels and tend to be thinner and
flatter than exterior leaves (Mills et al., 1999).
Percentages of total biomass and N in leaf, branch and root tissue compared well
with 10 and 15 year old trees harvested by Cameron and Appleman (1935) and Cameron
and Compton (1945). However, trees grown on Swingle rootstock were significantly
smaller than those grown on Carrizo rootstock. Even though tree size was different, dry
weight allocation between tree components remained relatively constant. However, there
were significant differences in dry weight accumulation in large branch and total branch


120
microsprinkler supply lines. After each uptake study, a quantity of N equal to the N
reduction was applied to those trees receiving the reduced N rate.
Soil sampling procedures
Fifty 1.27 cm diameter polyvinyl chloride (PVC) pipes were inserted to 45 cm
depth beneath the canopy of eight Hamlin on Carrizo and eight Hamlin on Swingle
trees at least 2 weeks prior to each study, assuring that all roots within them would die
prior to fertigation treatment applications. The pipes were arranged in three semicircles
25, 75, and 125 cm from each tree trunk. The ratio of the lengths of these arcs was 1:3:6.
The number of pipes in each arc was proportional to its length, resulting in 5, 15, and 30
pipes in each arc. Soil from these pipes was used as a control to estimate loss of N by
volatilization and immobilization in the absence of roots. Changes in NH4-N and NO3-N
concentrations were used to determine nitrification rates under field conditions.
A composite soil sample consisting of 10 cores taken with a 2 cm diameter auger
were removed from each tree at three depth increments of 15 cm each. Samples were
taken from the same arcs and ratios where the PVC pipes were installed. Samples were
taken 0 h, 1 h, 1 d, 2 d, 3 d, and 4 d after N fertilizer application. The timing of fertilizer
applications corresponded with growth phases of the citrus tree when maximum N uptake
was most likely: early March (bloom and first spring flush), mid May (second spring
flush and fruit expansion), and September (third flush and fruit maturation).
Analytical methods
All soil samples were placed in an insulated cooler containing ice and placed into
a refrigerator at 4 C or less until extractions could be made. Extractions using
approximately 4 g of soil and 40 mL of 2 M KCI were analyzed to determine soil nitrate
(NO3-N) and ammonia (NH4-N) concentrations (Keeney and Nelson, 1987). Between 4.1


Erie, L.J., O.F. French, and K. Harris. 1965. Consumptive use of water by crops in
Arizona. University of Arizona Agri. Exp. Sta. Tech. Bui. No. 169,44 pp.
164
Fares, A., and A. K. Alva. 1999. Estimation of Citrus Evapotranspiration by Soil Water
Mass Balance. Soil Science 164:302-310.
Feigenbaum, S., H. Bielorai, Y. Emer, and S. Dasberg. 1987. The fate of 15N labeled
nitrogen applied to mature citrus trees. Plant and Soil. 97:179-187.
Flenet, F., J.R. Kiniry, J.E. Board, M E. Westgate, and D C. Reicosky. 1996. Row
spacing effects on light extinction coefficients of com, sorghum, soybean and
sunflower. Agron, J. 88(2) 185-190.
Florida Agricultural Statistics Service. 2002. Commercial citrus inventory 1998. Florida
Department of Agriculture and Consumer Services. Tallahassee, FL.
Follett, R.F., M.J. Shaffer, M.K. Brodahl, and G.A. Reichman. 1994. NLEAP simulation
of residual soil nitrate for irrigated and non-irrigated com. J. Soil Water Conser.
49(4):375-382.
Ford, H.W. 1952. The distribution of feeder roots of orange and grapefruit trees on rough
lemon rootstock. Citrus Industry Magazine 14(1 1):22-23.
Ford, H.W. 1953a. Effect of spreading decline disease on the distribution of feeder roots
of orange and grapefruit trees on rough lemon rootstock. Proc. Amer. Soc. Hort.
Sci. 61:68-72.
Ford, H.W. 1953b. Root distribution of chlorotic and iron-chelate-treated citrus trees.
Proc. Fla. State Hort. Soc. 66:22-26.
Ford, H.W. 1954a. Root distribution in relation to the water table. Proc. Fin. State Hort.
Soc. 67:30-33.
Ford, H.W. 1954b. The influence of rootstock and tree age on root distribution of citrus.
Proc. Amef Soc. Hort. Sci. 63:137-142.
I
Ford, H.W. 1959. Growth and root distribution of orange trees on two different rootstocks
as influenced by depth to subsoil clay. Proc. Amer. Soc. Hort. Sci. 74:313-321.
Ford, H.W. 1964. The effect of rootstock, soil and soil pH on citrus root growth in soils
subject to flooding. Proc. Fin. State Hort. Soc. 77:41-45.
Ford, H.W. 1968. Fluctuations of the water table in drained flatwoods groves. Proc Fla
State Hort. Soc. 81:75-79.


82
Root densities at the 50 cm distance in the cross-row orientation decreased more
gradually than did densities at 100, 150, and 200 cm distances. Minimum densities
occurred at the 45 to 60 cm depth for the 50 cm distance as opposed to the 30 to 45 cm
depth for the 100, 150, and 200 cm distances. Similarly, root densities increased at the 60
to 75 cm depth for the 100 and 150 cm distances, and 75 to 90 cm depth for the 50 cm
distance. Spatial root distribution differences between rootstocks were not statistically
significant (Table 4-1). Mean root length densities at all depths and distances were 0.36
cm cm'3 for trees grown on Carrizo citrange and 0.41 cm cm"3 for trees grown on Swingle
citrumelo. However, the interaction of rootstock and depth was significant at the P=0.01
level. Trees on Swingle had higher root length densities near the soil surface than did
trees on Carrizo (Figs.4-1 and 4-2). Conversely, root length densities were greater for
trees on Carrizo between 15 and 75 cm below the soil surface.
Root length densities for the 0 to 15 cm depth ranged from 2.0 to 0.9 cm cm'3 soil
at distances of 150 cm or less for trees on Swingle rootstock. Densities ranged from 1.2 to
0.7 cm cm'3 at the same depth and distances for trees on Carrizo rootstock. Root densities
decreased for both rootstocks to low values at the 30 to 45 cm depth. Root densities
increased for trees on Carrizo at 45 to 60 cm depth, whereas densities for trees on
Swingle increased at the 60 to 75 and 75 to 90 cm depths. With the exception of the 50
cm distance from the tree trunk, root densities were greater for trees on Carrizo at 30 to
45 cm than those on Swingle. Likewise, root densities were greater at 45 to 60 cm depth
for trees on Carrizo than at the 60 to 75 cm depth for trees on Swingle.


47
determine percentage dry weight and analyzed for N concentration to determine total fruit
N accumulation. Leaves and twigs collected in 2002 were separated into current years
growth and prior years growth. Separate samples were collected for dry weight and N
concentration determination.
Statistical Analysis
Since the samples were taken during a 14-month period, the samples from the
mature (14 year old) Hamlin trees were treated as repeated measures and analyzed
accordingly using the SAS general linear models (GLM) procedure (SAS Institute, Cary,
NC). Non-linear regression analysis of tissue masses and percentages of all trees were
made considering canopy volumes and mean trunk diameters as independent variables.
Results
Mature Citrus Tree Biomass Distribution Experiment 1
Leaf area was significantly different (P=0.05) for TCV and TCSA, but not for
rootstocks (Table 3-2). However, neither mean leaf area index on a per tree basis (10.2
and 9.9 for Carrizo and Swingle trees, respectively) nor on a per acre basis (6.4 and 6.2
for Carrizo and Swingle trees, respectively) were significantly different for tree size or
rootstock. Total above-ground and below-ground weights increased on both a fresh and
dry basis across the range of TCV and TCSA encountered in this study. Maximum total
fresh weight was approximately 160 kg tree'1 with TCVs ranging from 28 to 38 m3 and
TCSAs of 80 to 160 cm2. Maximum dry biomass for the same canopy volumes and trunk
cross-sectional areas was approximately 100 kg tree*1. Maximum above-ground biomass
(leaves, twigs, branches, and trunk) and below-ground (roots and taproot) was
approximately 74 and 26 kg tree'1, respectively.


26
wetted. The concept of field capacity is useful in the design of field management schemes
for approximating the maximum amount of soil water storage. Field capacity can be used
as an upper limit value of 8 within each soil layer such that any water in excess of 0fC
quickly drains to the next deeper soil layer. The soil profile can be described as a vertical
sequence of reservoirs with the overflow level for each reservoir corresponding to the
value of 0fc for that specific soil layer. During irrigation or rainfall the top reservoir flows
over to fill the next lower reservoir until no excess water remains to flow into the next
reservoir. With a judicious selection of the depth of each soil layer, this simple analog of
the soil profile can be easily modeled.
Soil water content
Estimated annual ETC for a deforested area on the Florida ridge reached 680 mm
(Sumner, 1996). This ET rate was attributed to periods of low soil water content because
the area was not irrigated. Rogers et al. (1983) reported that growth and fruit yield of
citrus trees were greater during a 3-year period for treatments maintained at higher soil
water content. During the same period of time, annual ETC averaged 900 and 1210 mm
for the lowest and highest soil water content treatments, respectively. Hoffman et al.
(1982) reported annual ETC values to be 200 to 500 mm higher than that found by Erie et
al. (1965) in Arizona. The lower annual ETC values reported by Erie et al. were attributed
to infrequent irrigation resulting in dry soil surfaces and thus increased resistance to
water diffusion to the atmosphere. Smajstrla et al. (1986) reported a reduction in growth
and ETC with increased available soil water depletion of 2-year-old Valencia orange
trees grown in drainage lysimeters. Available soil water depletion setpoints used for
irrigation scheduling in this study were 28, 47 and 58%. It was concluded that tree stress


166
Heinemann, A.B., G. Hoogenboom, G.A. Georgiev, R.T. de Faria, and J. A. Frizzone.
^ 2000. Center pivot irrigation management optimization of dry beans in humid
areas. Trans. Am. Soc. Agri Engr. 43(6): 1507-1516.
Hilgeman, R.H. 1941. Nitrogen uptake by grapefruit trees in the Salt River Valley. Proc.
Amer. Soc. Hort. Sci. 39:119-124.
Hillel, D. 1971. Soil and water physical principles and processes. Academic Press,
New York, New York.
Hillel, D. 1998. Environmental soil physics. Academic Press. New York, New York.
Hoffman, G.J., J.D. Oster, and W.J. Alves. 1982. Evapotranspiration of mature orange
trees under drip irrigation in an arid climate. Trans. Am. Soc. Agrie. Engr.
25:992-996.
Hoogenboom, G., J.W. White, J.W. Jones, and K.J. Boote. 1994. BEANGRO: A process
oriented dry bean model with a versatile user interface. Agron. J. 86:182-190.
Jabro, J.D., J.M. Jemison, L.L. Lengnick, RH. Fox, and D.D. Fritton. 1993. Field
validation and comparison of LEACHM and NCSWAP. Trans, of the ASAE
36(6): 1651-1657.
Jarvis, P.G. and K.G. McNaughton. 1986. Stomatal control of transpiration: Scaling up
from leaf to region. Advances in Ecological Research 15:1-49.
Jemison, J.M., J.D. Jabro, and R.H. Fox. 1994. Evaluation of LEACHM: n. simulation
of nitrate leaching from nitrogen-fertilized and manured com. Agron. J. 86: 852-
859.
Jones, C.A., and J.R. Kiniry. 1987. CERES-MAIZE: A simulation model of maize
growth and development. Texas A&M Univ. Press, College Station.
Jones, H.G., AN. Lakso, and J.P.Syvertsen. 1985. Physiological control of water status
in temperate and subtropical fruit trees. Horticulutral Reviews 7:301-344.
Jones, J.W., E. Dayan, L.H. Allen, H. Van Keulen, and H. Challa. 1991. A dynamic
tomato growth and yield model (TOMGRO). Trans, of the ASAE 34(2): 663-672.
Jones, J.W, and J.C. Luyten. 1998. Simulation of biological processes. In agricultural
systems modeling and simulation. RM. Peart and R.B. Curry (eds). Marcel
Dekker, Inc. New York, pp 19-62.
Kaufmann, MR., S B. Boswell, and L.N. Lewis. 1972. Effect of tree spacing on root
distribution of 9-year-old Washington Navel oranges. J. Amer. Soc. Hort. Sci.
97:204-206.


114
water depletion throughout a soil profile and assess effective soil water storage capacity
and potential leaching of nutrients associated with rainfall and/or irrigation.


61
Canopy Volume (m3)
Fig. 3-6. Dry weight allocation to above-ground (closed), and below-ground (open)
biomass (A); leaf(closed) and twig (open) biomass (B); total branch (closed) and
trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D)
as a function of canopy volume for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.


28
Grass and weed growth
Smaj stria et al. (1986) used field drainage lysimeters to determine the effect of
grass cover on the growth and ETC of 2-year-old Valencia orange trees. Automated
covers were installed to cover the lysimeters during rainfall. Soil within the lysimeters
was maintained bare or covered completely with bahiagrass. The bare soil lysimeters
consistently had the lowest monthly ETC. Measured annual ETC ranged between 1331 to
900 mm for grass-covered lysimeters and 912 to 441 mm for those with bare soil
surfaces. Total ETC was 46 to 105% higher per year due to soil grass cover. These results
were similar to those reported by Stewart et al. (1969) using non-weighing lysimeters. In
their study, estimated annual bare soil evaporation and 2/3 sod cover ETC averaged 68
and 92% of full sod cover, respectively. Tucker et al. (1997) reported reduced soil water
use from non-irrigated middles between rows of mature citrus by limiting the height of
weed growth by chemical mowing.
Crop Coefficient
An estimate of evapotranspiration for a specific crop (ETC) is calculated by
multiplying the reference evapotranspiration (ET0) by an empirically determined crop
coefficient (Kc). This coefficient is specific for a crop, growth stage, and growing
conditions. The resulting ETC estimates water use of a crop under local or regional
climatic conditions.
Rogers et al. (1983) reported monthly measured ETC to calculated ET0 ratio values
using the mean of four methods of estimating ET0 (Penman, Blaney-Criddle, Jensen-
Haise, and Class A pan). The resulting monthly ratios range from 0.90 in January to 1.11
in June. Crop coefficient (Kc) values reported by Doorenbos and Pruitt (1977) after


102
1.2
1.0 -
0.8
0.6
f0'
0.4 -
fgg, o -a? c'&~c
2 cd 095 m
o co
o<6 /p oo 1 _
o
oB
0.2
12 -
1.0 -
2
% 0.8
¡5,
til 0.6
0.4 -i
£ VW
v--\
8
O O
#>?
O
o^o^o
B
0.2
1.2
1.0 H
0.8
0.6
0.4
&Vrf>
wj$^?D5S
0.2
T 1
12 -10
-20 -18 -16 -14
Soil Water Potential (kPa)
I
-8
i
Fig. 5-4. Estimated ETC to calculated ET0 ratio as a function of soil water potential in the
irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a 1 m
depth (C). The data points shown represent a range of soil water content from
field capacity to approximately 50% available soil water depletion.


CHAPTER 2
LITERATURE REVIEW
Introduction
The various species of the genus Citrus are believed to be native to the subtropical
and tropical regions of Asia and the Malay Archipelago (Webber and Batchelor, 1943).
Citrons were cultivated on the European continent as early as 300 BC. Limn and sweet
orange were not known in Europe for some 15 and 17 centuries later, respectively. Sweet
orange was first introduced to the North American continent on Columbus second
voyage in 1493. Citrus was grown in coastal settlements of Florida by the mid 16th
century and wild citrus trees were found on hammocks near lakes or rivers where
conditions were particularly favorable for their growth in the second half of the 19th
century (Harris, 1875; Adams, 1875). Due to freeze damage, citrus production in Florida
has moved in the last 120 years from north central Florida to the southern half of the
Florida peninsula. Currently 2 million ha of citrus fruit are grown in Florida (Florida
Agricultural Statistics Service, 2002). Florida citrus production was 1.1 million metric
tons accounting for 73 and 18% of US and world production, respectively.
Vaile (1924) showed that fine sand or sandy loam soils resulted in better growth
and production than coarse sands or heavy loams. Subtropical in nature, citrus trees do
not exhibit dormancy or shed their leaves during the winter months However, new
growth appears in definite cycles, with two to four cycles of growth yearly. The first and
usually largest growth starts in the early spring (late February to early March), the second
9


Table 5-1. Monthly maximum, minimum, and mean reference evapotranspiration reported by Florida Automated Weather
Network for the Avalon Station and maximum, minimum and mean estimated citrus crop evapotranspiration.
Reference Evapotranspiration (mm d'1) Standard Crop Evapotranspiration (mm d*1) Standard
Months
Maximum
Minimum
Mean
Deviation
Maximum
Minimum
Mean
Deviation
April,2000
5.0
4.0
4.5
0.31
4.7
3.1
3.9
0.54
May
5.9
4.1
5.0
0.46
5.9
3.9
4.6
0.57
June
6.5
5.0
5.7
0.44
6.5
3.5
4.7
0.72
July
5.8
4.6
5.3
0.35
5.8
3.5
4.6
0.72
August
5.2
4.1
4.7
0.33
5.4
3.0
4.2
0.66
September
5.0
3.3
4.3
0.48
4.8
2.5
3.5
0.70
October
3.8
2.3
3.0
0.38
2.7
1.7
2.2
0.32
November
2.9
1.6
2.3
0.39
2.2
1.3
1.8
0.26
December
2.8
1.5
1.9
0.31
1.9
1.0
1.3
0.24
January, 2001
2.9
1.3
2.0
0.45
2.1
1.0
1.4
0.26
February
3.6
1.2
2.7
0.70
2.6
1.4
1.8
0.29
March
4.3
2.6
3.5
0.38
3.1
2.0
2.8
0.28
April
5.0
3.3
4.5
0.36
5.2
2.5
4.0
0.68
May
5.9
3.6
5.3
0.52
6.2
3.9
4.3
0.60
June
6.4
4.9
5.6
0.35
6.3
3.5
4.8
0.91
July
6.2
5.1
5.5
0.28
6.3
3.6
4.5
0.53
August
6.2
4.6
5.3
0.44
5.9
2.7
4.5
0.63
September
5.5
3.4
4.5
0.62
4.8
1.9
3.4
1.01
October
3.6
2.2
2.9
0.39
3.0
1.5
2.0
0.38
November
2.7
1.4
2.1
0.31
2.8
1.6
1.4
0.47
December
2.2
1.1
1.8
0.27
1.8
1.1
1.3
0.31
January, 2002
3.0
1.2
2.0
0.58
1.9
1.2
1.4
0.22
February
3.2
1.9
2.6
0.29
2.3
1.3
1.9
0.34
March
4.4
2.1
3.6
0.56
3.4
1.6
2.9
0.55
April
5.3
3.3
4.3
0.60
4.6
1.7
3.2
0.74
Month
7 ^ :
NS
NS
Statistical Significance1
NS NS NS
NS
NS
NS
NS = Not significantly different by General linear Model at the p=0.05 level.


30
multiplying the of a given crop by the soil water depletion coefficient (Kg) for a given
soil water content. Water stress increases as soil water is extracted by evapotranspiration.
Available soil water (ASW) is defined as the difference between drained upper
limit (field capacity) and drained lower limit or permanent wilting point. However, the
energy expenditure required to extract residual soil water increases as soil water content
decreases. Likewise, resistance to water flow increases as residual soil water decreases,
reducing water flux to the root boundary. Therefore, crop water uptake is reduced well
before wilting point is reached (Allen et al 1998). At field capacity, roots can absorb
water fast enough to supply the ETC demand of the atmosphere. However, water becomes
more strongly bound to the soil matrix and is more difficult to extract as soil water
content decreases. When soil water content drops below a threshold value, water can no
longer be transported quickly enough to the roots to supply the transpiration demand of
the crop. The fraction of ASW above this threshold is known as readily available water
(RAW). For most crops grown on medium and fine textured soils, RAW is as much as 30
to 50% of ASW (Allen et al. 1998). When root zone depletion exceeds this threshold, ET
is reduced relative to potential crop ETC and water stress occurs.
Citrus Nitrogen Uptake
Knowledge of the nutritional need of different plant organs as well as the seasonal
demand for nutrients is essential in order to establish a physiological basis for crop
fertilization (Lagaz and Primo-Millo, 1981). The potential contribution of fertilizer N to
the deterioration of ground water quality may be appreciable (Embleton et al., 1978).
This impact is especially true in Florida where the combination of high annual rainfall,
sandy soils and shallow water tables create conditions that greatly increase the potential


BIOGRAPHICAL SKETCH
Kelly Tindel Morgan was bom in Columbus, Georgia, in 1958. He grew up and
was educated in Winter Haven, Florida graduating from Winter Haven High School in
1976. Kelly earned an AA degree in chemistry and biology from Polk Community
College in 1978. He married Nancy Greives in 1978 and enrolled at the University of
Florida that fall. He earned his BS degree in 1980 majoring in plant pathology; later he
earned his MS degree in plant pathology in 1982. Kelly worked as Assistant Manager of
a large citrus nursery from 1982 to 1985 during which time his two sons, Joshua and
Christopher, were bom. From 1985 to 1988 he managed citrus groves for investors after
which he worked for the University of Florida at the Citrus Research and Education
Center from 1988 to the present. Currently, he is a Scientific Research Manager directing
horticultural maintenance, treatment application, and data collection on more than 100
acres of citrus research plots associated with the Conserv II reclaimed water project near
Orlando, Florida
174


29
adjustments for humid conditions ranged from 0.9 in March though December to 0.95 in
January and February. Castel et al. (1987) estimated monthly Kc for drip-irrigated mature
Navel oranges grown in Valencia, Spain. Their Kc values were calculated from mean
daily ETC estimated from weekly ET values determined by neutron probe measurements.
Values ranged from a mean of 0.71 from January through July to 0.90 from August
through December. Castel and Buj (1992) suggested these values differed from those
reported for Florida due to the lower evaporative demand of the humid Eastern coast of
Spain, which has a mean annual ET0 of 1166 mm compared with 1400 mm in Florida.
Calculated Kc values for 3-year-old Hamlin trees grown on sandy soil in central
Florida ranged from approximately 1.05 in November through March to 0.85 in May
through August (Fares and Alva 1999). Boman (1994) calculated Kc values for 5-year-old
Valencia orange trees grown in non-weighing lysimeters with water tables maintained
at 0.6,0.75, or 0.9 m from the soil surface. Calculated Kc values were at a minimum of
0.6 from December through February and peaked at 1.1 in June and July. Martin et al.
(1997) estimated mean daily ETe values for 7-year-old Redblush grapefruit in Arizona
from soil water content data collected at 1 to 2 week intervals. Monthly Kc values were
calculated by comparing these estimated daily values with mean daily ET0 for the same
period. The resulting K ranged from a low of 0.55 to 0.6 in December and January to a
high of 1.1 to 1.2 in July.
Soil Water Depletion Coefficient
According to Allen et al. (1998), the water depletion coefficient is defined as the
effect of soil water reduction on ETC by reducing the value of Kc. It is calculated by


124
months of March, May, and September, respectively and were not significantly different
from mean cumulative losses for trees on Swingle citrumelo, which were 73.8, 72.5, and
66.4% for the same months.
Nitrogen losses from control pipes varied greatly across the three studies but were
not significantly affected by application rate, month, or rootstock. Mean cumulative N
loss was 22.6%, ranging from a net gain of 7.0% to a loss of 41.0% for the 3 days after
application. Estimated daily maximum N uptake was determined by subtracting the N
lost from the control pipe from the N lost from the bulk soil. Cumulative daily maximum
N uptake was significantly different at the P=0.01 level by application amount, but not
significantly different by month, and rootstock (Tables 6-1 to 6-3). Maximum uptake as a
percentage of amount applied averaged 46.7 and 61.7% for the high and low application
rates, respectively. Mean cumulative maximum uptake values for Carrizo and Swingle
were 53.9 and 54.4%, respectively.
Estimated total and passive uptake as a function of mean soil solution N concentration is
presented in Figs. 6-1 and 6-2. The equation constants, R2, RMSE, and P values for the
regression of these data are presented in Table 6-4. A great deal of scatter exists in the
data due to the large range in N loss from the control pipes. Therefore, the small R2
indicates that soil solution N concentration explained only 46% of the variation in the
data. While both regressions are significant at the P=0.01 level, the relatively large
RMSE values would result in a large 100(l-a)% confidence interval. Using the estimated
maximum N uptake and soil solution concentration relationship in Figure 6-1, the
Michaelis-Menten equation (Equation 1) constants were approximately 14.5 g N tree'1 d'1
and 60 mg N L'1 for and Km, respectively.


162
Boman, B. J. 1994. Evapotranspiration by young Florida flatwoods citrus trees. J.
Irrigation and Drainage Engineering 120:80-88.
Boswell, S B., C D. McCarty, and L.N. Lewis. 1975. Tree density affects large-root
distribution ofWashington Navel orange trees. HortScience 10:593-595.
Cahoon, G.A., R.B. Harding, and D.B. Miller. 1956. Declining citrus root systems. Calif.
Agrie. 10(9):3-12.
Cahoon, G.A., E.S. Morton, W.W. Jones, and M.J. Garber. 1959. Effects of various types
of nitrogen fertilizers on root density and distribution as related to water
infiltration and fruit yields of Washington Navel oranges in a long-term
fertilizer experiment. Proc. Amer. Soc. Hort. Sri. 74:289-299.
Cahoon, G.A., M.R. Huberty, and M.J. Garber. 1961. Irrigation frequency effects on
citrus root distribution and density. Proc. Amer. Soc. Hort. Sci. 77:167-172.
Cahoon, G.A., L.H. Stolzy, M.J. Garber, and E.S. Morton. 1964. Influence of nitrogen
and water on the root density of mature Washington Navel orange trees. Proc.
Amer. Soc. Hort. Sci. 85:224-231.
Calvert, D.V., RCJ. Koo, and H.W. Ford. 1967. Flood irrigation studies with citrus. Proc.
Fla. State Hort. Soc. 80:79-85.
Calvert, D.V., H.W. Ford, E.H. Stewart, and F.G. Martin. 1977. Growth response of
twelve citrus rootstock-scion combinations on a Spodosol modified by deep
tillage and profile drainage. Proc. Inti. Soc. Citriculture 1:79-84.
Calvert, D.V. and H.T. Phung. 1972. Nitrate-nitrogen movement into drainage lines under
different soil management systems. Soil Crop Sci, Soc. Fla. Proc. 31:229-232.
Cameron, S .H and D. Appleman. 1935. The distribution of total nitrogen in the orange
tree. J. Amer. Soc. Hort. Sri. 30:341-348.
Cameron, S.H., and O.C. Compton. 1945. Nitrogen in bearing orange trees. J. Amer. Soc.
Hort. Sci. 46:60-68.
Carlisle, V.W., F. Sodek, M E. Collins, L.C. Hammond, and W.G. Harris. 1989.
Characterization data for selected Florida soils. University of Florida-Soil Science
Department, Soil Characterization Laboratory, Institute of Food and Agricultural
Sciences in cooperation with U.S. Department of Agriculture-Soil Conservation
Service. Soil Science Research Report No. 89-1. 307 p.
Caruso, T., P. Inglese, F. Sottile, and F.P. Marra. 1999. Effect of planting system on
productivity, dry-matter partitioning and carbohydrate content in above-ground
components ofFlordaprince peach trees. J. Am. Soc. Hortic. Sci. 124(1) 39-45.


68
Trunk Cross Section Area (cm2)
Fig. 3-11. Nitrogen allocation of above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (closed) and taproot (open) (D) as a
function of trunk cross section area for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.


70
biomass weights for the two rootstocks in this study. Trees grown on Carrizo citrange
rootstocks were larger than those of the same age grown on Swingle citrumelo, therefore
the difference in large and total branch weights were correlated with tree size as
measured by TCV and TCSA. Hence, the percentages of total biomass for specific tree
components were similar for both rootstocks, indicating that above ground biomass is
partitioned equally based on the relationship of total biomass to tree size. Therefore, if the
r ^
relationship of total biomass to tree size is known, and the relationship of component
i ...
partitioning to tree size is also known, then the biomass of each component part can be
estimated regardless of rootstock effects on tree size.
The biomass associated with individual tree trunks was related to the height of the
crotch formed by the main scaffold limbs of the tree. In citrus nurseries, crotch height is
relatively uniform. However, the larger trees used in this study were affected by the
freezes of the late 1980s, most notably in 1989. The limb structure of the Hamlin trees
used in the mature tree portion of this study had to be re-grown, in some cases requiring
the entire scaffold limbs and a portion of the upper trunk to be removed. Thus the heights
of the trunk to the scaffold limbs were not consistent.
Tree size, biomass, N weight, and apparent FNUE were greater for trees grown on
Carrizo citrange rootstock compared with trees grown on Swingle citrumelo, indicating
that differences in tree size are directly correlated with FNUE. Differences in FNUE may
be due to physiology of the rootstocks themselves or due to the distribution of fibrous
roots associated with the various rootstocks. The distribution of root length densities and
N uptake rates for the two rootstocks in this study will be the subject of Chapters 4 and 6,
respectively.


76
a
/
overlapping root systems.
Castle and Krezdom (1975) described two general types of root systems, the first
characterized by extensive lateral and vertical development, and the second with
intensive higher fibrous root density near the soil surface. Trees on rough lemon,
Volkamer lemon and Palestine sweet lime (C. limettioides Tan.) typified the extensive
type of root system where 50% of the fibrous roots occurred below 70 cm in the soil and
produced large, highly-productive trees that dominated the citrus industry in Florida
when trees were irrigated less intensively and were set at much lower densities.
Unfortunately, rough lemon has been virtually eliminated as a commercial rootstock due
to citrus blight disease. Examples of the intensive-type root system were Carrizo citrange
and Swingle citrumelo that had few fibrous roots below 70 cm, and the root system was
less developed laterally. These rootstocks now dominate the citrus industry in Florida and
are well suited for high-density, intensively irrigated plantings.
The following hypotheses were tested: 1) Root distribution is significantly
affected by rootstock, and 2) Generic relationships can be developed for well-drained
soils that describe citrus root densities at various depths from the soil surface and
distances from the tree as a function of tree size. To test these hypotheses, the objectives
of this study were to: 1) develop information on spatial root length distribution at
different soil positions and depths for two citrus rootstocks, and 2) develop functional
relationships that define root length densities at various soil positions and depths as a
function of tree size. The relationship of vertical and horizontal root length density
distribution to tree size resulting from this study can be used to estimate fibrous root
densities in various soil layers for citrus water and nutrient uptake models. This


88
trunk while continuing to increase in density near the trunk itself. Expansion continues
through maturity with trees in dense plantings overlapping in both in-row and between-
row directions. This portion of the citrus root system is important for tree stability while
providing adequate roots for water and nutrient uptake for the developing tree. A second
region of root growth develops below 30 cm between 5 and 10 years of age. This region
^ i
of root growth increases in size and density through maturity. The development of a
deeper root system is essential for supplying adequate water and nutrients to the mature
tree from an increasingly large soil volume. These two growth modes result in the
bimodal root distribution of the mature citrus tree. While complex, the development of
citrus root systems with time appears to be predictable and basic trends can be captured
in functional relationships within a citrus model. The data presented here can provide the
root distribution information needed to determine spatial soil water and nutrient uptake
for a citrus growth model.
Conclusions
It is concluded that the root length distribution of trees grown on Swingle and
Carrizo rootstocks followed predictable patterns with increased tree size resulting in
mature trees with bimodal root systems. While both rootstocks developed a dense root
system within the upper 30 cm of the soil surface, the root systems extended beyond 1 m
in depth, so extending sampling to greater soil depths on well-drained soils appears to be
desirable. Trees on Swingle developed higher root length densities near the soil surface,
and lower densities below 30 cm compared with trees on Carrizo. Based on the overall
high r2 values and low RMSE, the functional relationships that were developed in this
study account for most of the variability in root length density. Thus, the test hypothesis


V
Climate 21
Soil characteristics 21
Soil water content 26
Water table 27
Soil shading 27
Grass and weed growth 28
Crop Coefficient 28
Soil Water Depletion Coefficient 29
Citrus Nitrogen Uptake 30
Seasonal Nitrogen Uptake 31
Nitrogen Uptake Efficiency 33
Seasonal Nitrogen Redistribution 33
Crop and Environmental Models 34
Current Citrus Models 34
Environmental Models 34
Crop Models 36
Conclusions 36
3 CITRUS BIOMASS AND NITROGEN ACCUMULATION 38
Introduction 38
Methods and Materials 42
Experiment 1 Mature Citrus Biomass and N Distribution 42
Experiment 2 -Biomass and N Accumulation With Increase in Tree Size 42
Site Description 43
Tree Canopy Volume and Trunk Cross Sectional Diameter 43
Tree Biomass Fresh Weight 44
Sample Processing and Nitrogen Analysis 45
Leaf Area, Biomass, and N weight Estimation 46
Statistical Analysis 47
Results 47
Mature Citrus Tree Biomass Distribution Experiment 1 47
Biomass Changes with Increase in Tree Size Experiments 1 and 2 50
Nitrogen Distribution 60
Mature Citrus Tree N Distribution Experiment 1 60
Nitrogen Balance 63
Nitrogen Change With Increase in Tree Size Experiments 1 and 2 64
Discussion 69
Conclusions 73
4 CITRUS ROOT GROWTH DYNAMICS 74
Introduction 74
Methods and Materials 77
Sample Collection 77
Sample Processing and Statistical Analysis 77


142
lower application rate. Total uptake rates were nearly equal for trees grown on Swingle
and Carrizo at either N application rate.
Numerous studies have been unable to account for 40% or more of total N applied
to soils (Mansell et al., 1986, Dasberg et al., 1987). Some authors attributed this
unaccounted for fraction as being lost to the atmosphere through denitrification of NO3-
N, while others concluded that the unaccounted for N was either incorporated into soil
organic matter or stored in the tree. Using soil of the same type that was used in this
study, Lea-Cox and Syvertsen (1993) could not account for 14.5% of 15N applied to
containers of soil in the greenhouse at 28 C during the first day. This value increased to
67.0% on the seventh day after application. The loss of 15N was assumed to have been
incorporated into microbial biomass because soil pH was too low to account for
appreciable volatilization of NH4-N. These data are similar to the mean of 26.1% N loss
in the control pipes 3 days after application. Nitrogen loss after 7 days totaled 60.9%
assuming the same rate of loss.
In a lysimeter study by Lea-Cox et al. (2001), 4-year-old grapefruit trees grown in
Candler fine sand were fertilized with 2.6, 5.1, and 11.6 g N tree-1 of double-15N labeled
ammonium nitrate. Soil samples contained no NH4-N on either date for the lowest rate,
and a mean of 58.2 and 0% of the highest rate on 1 and 8 days after application,
respectively. This result suggests that the mean of 34.0% NH4-N found 1 day after
application of 85.1 and approximately 43 g N tree'1 in the current study is too low.
However, in the same study, a mean of only 5.7% of NFL-N applied at the 5.1 g N tree1
rate was found in soil under trees not contained in lysimeters compared with 17.9% in
lysimeters at the same application rate. This result implies that the NH4-N amounts in the


150
Seasonal K
The ratios of ETC to ET0 at field capacity were used to estimate Kc. The ETC to
ET0 ratios ranged between 0.81 on DOY 24 and 1.12 on DOY 179. With an R2 of 0.755,
DOY explained more than 75% of the variation in the ETC to ET0 ratios when soil water
content was near field capacity. Therefore, the results of the equation are a good
approximation of the value of Kc for a given DOY.
K, Estimation
The relationship of estimated K to available soil water depletion (ASWD) and
soil water potential was logistic in nature with a value of 1 from field capacity to
approximately 10 to 15% of ASWD. The relationship decreased steadily to
approximately 0.6 at 50% ASWD, indicating a reduction of 40% in ETC between 15 and
50% ASWD.
Soil Water Uptake per Unit Root Length
An exponential decay model resulted in the best fit of unit root length uptake to
soil water potential with a maximum of approximately 0.4 mm3 d'1 cm"1 at field capacity.
Daily ETC per unit root length decreased rapidly with decrease in soil water potential to
-12 or -13 kPa, followed by a more gradual reduction from approximately 0.1 to 0.05
mm3 d"1 cm"1 well past -20 kPa. Higher water uptake per unit root length values at field
capacity were found at the 10 cm depth between rows. This value was double that
observed at other locations and depths. The increase in water uptake could be explained
by water use from non-crop species in the row middles that were not present beneath the
tree canopy.


141
Mean daily soil temperature at the 10 cm depth was 5 C higher in May compared with
March, 2002. Increased soil N losses in May compared with March and September may
be attributed to higher soil temperatures in May.
High N concentrations in leaf, twig and branch tissues have been determined to be
sources of N for developing vegetative and reproductive tissues (Dasberg, 1987; Kato et
al., 1982; and Legaz et al., 1982). Lower expanded leaf, twig and branch bark N
concentrations prior to fertilizer application in May compared with March and September
of2002 agree with results of Kato et al. (1982) and Legaz et al. (1982). This agreement
provides evidence of greater crop demand for N during this period of time. Reduced N
concentrations in expanding flush leaves may be related to N dilution in the tissue during
periods of rapid growth in May. Fertilizer N was added in six equal amounts, four
applications in the spring (February, March, April, and May) and two after the end of the
rainy season in September and October. Leaf N decreased below optimum concentration
prior to the May fertigation application in both years, suggesting that insufficient uptake
occurred from February to April to satisfy the N demand of developing tissues. Increases
in leaf N concentrations after the May fertilizer application in both years implies that
application of all spring N fertilizer should occur prior to May 1 of each year.
Increased active and passive N uptake rates are associated with higher root length
densities (Scholberg et al., 2002). Though trees on Swingle rootstocks have more of their
root mass in the upper 45 cm than do trees on Carrizo (Chapter 4), significant differences
in N uptake by rootstock were not detected. Significantly higher mean daily soil N losses
were found for Hamlin trees growing on Swingle compared with Hamlin trees on
Carrizo at the high N application rate. However, uptake was lower for Swingle at the


XV
20% of total mature tree mass, and accounted for 45, 35, and 20% of total N mass,
respectively. Root density increased radially as tree size increased. Trees on Swingle
citrumelo rootstock had a higher proportion of fibrous roots near the soil surface than
trees on Carrizo citrange. Soil water uptake ranged from 0.8 to 1.1 of ET0. Daily uptake
decreased steadily as soil water content decreased. N uptake from the upper 45 cm of soil
was greater for trees on Swingle citrumelo compared with Carrizo citrange. N uptake
efficiency ranged from 41 to 55% when fertilized at 269 kg N ha'1 compared with 47 to
70% when fertilized at 179 kg N ha'1. Leaf and twig N was highest from August to
February and lowest in May. New and improved understanding of citrus water and N
dynamics will advance Florida citrus management techniques and decrease
environmental impacts.


59
Table 3-6. Linear regression analysis of dry weight and N accumulation in different tree
components as relatad to trunk cross-sectional area (TCSA)Z
Yo
a
R2
RMSEy
P
Dry Weight (kg tree1)
Total Mass
-7.23
0.70
0.93
0.23
<0.0001
Above Ground
-5.67
0.50
0.93
0.24
<0.0001
Below Ground
-1.20
0.19
0.89
0.29
<0.0001
Leaves
-0.65
0.09
0.94
0.22
<0.0001
Twigs
-0.04
0.04
0.68
0.47
<0.0001
Sm. Branches
-1.13
0.09
0.82
0.40
<0.0001
Med. Branches
-0.84
0.06
0.80
0.43
<0.0001
Lg. Branches
-.94
0.05
0.78
0.42
<0.0001
Total Branches
-5.17
0.33
0.91
0.30
<0.0001
Trunk
0.38
0.02
0.70
0.36
<0.0001
Sm. Roots
0.10
0.04
0.90
0.24
<0.0001
Med. Roots
-0.08
0.04
0.91
0.25
<0.0001
Lg. Roots
-0.53
0.04
0.78
0.45
<0.0001
Tap Root
-0.19
0.03
0.61
0.57
<0.0001
Nitrogen Weight (g tree'1)
Total Mass
-47.83
6.00
0.94
0.22
<0.0001
Above Ground
-41.41
4.51
0.94
0.23
<0.0001
Below round
-5.36
1.41
0.92
0.24
<0.0001
Leaves
-15.39
2.32
0.93
0.23
<0.0001
Twigs
-3.29
0.46
0.82
0.37
<0.0001
Sm. Branches
-6.75
0.49
0.89
0.31
<0.0001
Med. Branches
-3.52
0.27
0.77
0.47
<0.0001
Lg. Branches
-5.69
0.47
0.85
0.38
<0.0001
Total Branches
-23.36
1.50
0.92
0.27
<0.0001
Trunk
1.97
0.10
0.82
0.39
<0.0001
Sm. Roots
0.92
0.51
0.89
0.25
<0.0001
Med. Roots
-1.48
0.38
0.91
0.25
<0.0001
Lg. Roots
-2.65
0.21
0.82
0.40
<0.0001
Tap Root
z
-0.23
0.13
0.58
0.56
0.0002
z y = yo +ax where x = TXA, and yo and a are regression coefficients
y RMSE dry weight kg dw tree'1, and N weight = g N tree'1


CHAPTER 6
CITRUS NITROGEN UPTAKE AND CYCLING
Introduction
The optimum timing, frequency, and rate of fertilizer N application for citrus
production under Florida conditions have been explored for nearly 60 years. Nitrogen
best management practices (BMPs) have been established for citrus on the sandy soils of
central Florida. The goals of these practices are to sustain high fruit production and tree
health, and improve N use efficiency of citrus while reducing the impact of N leaching on
ground water quality. The BMPs restrict the annual rate of N fertilizer than can be
applied, and base it on tree age or past production. The timing of N application is
restricted to the drier seasons of the year to reduce potential leaching. A N balance model
for citrus must be developed to predict the effects of these restrictions on citrus
production.
Koo (1979) found a significant N rate and irrigation interaction in high density
planting using low volume microsprinkler irrigation. Koo (1980) later found no yield
difference between dry fertilizer N applications and fertigation through the same low
volume system. In the same experiment, no yield differences were found when 10
fertigation applications were compared with three dry applications per year. Recent
studies by Alva and Paramasivam (1998) and Wheaton (unpublished) have shown similar
results. Koo (1986) and Boman (1993) found no difference in yield when comparing
multiple applications of dry soluble fertilizer with controlled release N sources.
115


110
Table 5-6. Regression analysis of estimated soil water uptake per unit root length density
on soil water potential in soil at three locations and for the 10, 20, 40 or 80 cm depths
using an exponential decay model2.
Location
Depth
(cm)
Yo
a
b
R2
RMSE
(mm d'1 cm1)
P
Under-Canopy
10
0.07
1.08
0.29
0.23
0.14
<0.0001
20
0.43
-0.29
0.05
0.13
0.35
0.0895
40
0.22
15.46
0.50
0.30
0.28
<0.0001
80
0.25
71.70
0.98
0.14
0.30
<0.0001
Dripline
10
0.16
7.89
0.49
0.20
0.26
<0.0001
20
0.14
4.69
0.36
0.24
0.36
<0.0001
40
0.60
4.74
0.15
0.62
0.24
<0.0001
80
0.33
40.75
0.61
0.29
0.30
<0.0001
Between-Rows
10
0.04
1.24
0.08
0.22
0.33
<0.0001
20
0.14
5.28
0.33
0.20
0.33
<0.0001
40
0.03
9.03
0.39
0.35
0.24
<0.0001
80
0.21
13.74
0.30
0.18
0.52
0.0730
2 Y = Y0 + a exp ** where X = root length density, and Yo a, and b are regression
coefficients
use from non-crop species in the row middles that were not present beneath the tree
canopy. However, there was also elevated water use at the 40 and 80 cm depths at the
dripline. These data are assumed to be the result of higher than expected root length
densities at these depths compared with mean root length density data from similar trees.
Therefore, the higher than expected water use per root length at the soil surface between
rows could be a combination of both greater root densities and water use by weed and
grass species.
Discussion
Citrus water uptake followed relatevely consistent patterns during the 2 years of
this study. Daily water withdrawals from the soil followed daily calculated ET, with
higher values occuring during the summer and lower ones in winter. ETC values were
consistently lower than ET0 except during summer months when 0 was near 0fc. These
0?
V
V' V'


Vil
7 SUMMARY AND CONCLUSIONS 146
Mature Tree Biomass Distribution 146
Biomass Vs Tree Size Relationships 147
Nitrogen Distribution 147
Mature Hamlin Root Distribution 148
Root Length Density Distribution Changes with Tree Size 149
Seasonal ET0 and ETC Trends 149
Seasonal Kc 150
Kc Estimation 150
Soil Water Uptake per Unit Root Length 150
Nitrogen Uptake 151
Nitrification Estimation 152
Seasonal Tissue Nitrogen Concentration 152
Seasonal Nitrogen Loss 153
Citrus Decision Support System 154
Citrus N practices and BMPs 155
APPENDIX
A Equations 158
LITERITURE CITED 161
BIOGRAPHICAL SKETCH 174


CHAPTER 3
CITRUS BIOMASS AND NITROGEN ACCUMULATION
Introduction
Citrus is native to the subtropical and tropical regions of Asia and the Malay
Archipelago (Webber and Batchelor, 1943). Citrons were introduced into Europe via the
Middle East as early as 300 BC, with lemons and sweet oranges following some 15 and
17 centuries later, respectively. Citrus is well adapted to Florida soil and environmental
conditions and proliferated in the costal settlements of Florida by the mid 16th century
and was in commercial production by the mid 1800s.
Nitrogen application rate studies in citrus have emphasized the effects of timing
and amount on increased canopy volume and yield (Sites et al., 1953; Reitz, 1956;
Reuther et al., 1957; and Koo, 1979). However, optimum plant growth depends upon
maintenance of an efficient balance between roots and shoots (Kramer and Boyer, 1995).
Roots are dependent on shoots for carbohydrates, growth regulators, and some organic
compounds, while the shoots of a plant are dependent on the roots for water and
nutrients. The root to shoot ratio varies widely among species, with age, and with
environmental conditions. Understanding the dynamics of root and shoot development
with time is essential when determining biomass and N accumulation, and soil water and
nutrient uptake dynamics.
Caruso et al. (1999) found that the relative proportion of leaves and twigs to total
tree dry weight decreased with tree age. Therefore, relative proportions of total dry matter


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
NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER
UPTAKE PARAMETERS FOR CITRUS
By
Kelly Tindel Morgan
May, 2004
Chair: Thomas Obreza
Cochair: Johannes Scholberg
Major Department: Soil and Water Science
During the last two decades, microirrigation and fertigation have become
commonplace in Florida citriculture. Concurrently, competition for water and nitrate
contamination of ground water have become greater concerns. Information on seasonal
citrus N demand, water use, and N uptake is needed to minimize water use and nitrate
leaching in citrus management. The objectives of this study were to 1) determine long
term and seasonal changes in citrus biomass and N distribution, 2) develop spatial
patterns of citrus root length density with increase in tree size, 3) estimate crop water use
and soil moisture coefficients, and root water uptake efficiency, and 4) explore seasonal
N uptake rates for citrus. Total tree biomass and N distribution were related to canopy
volume and trunk cross sectional area. The percentage of total tree N in citrus leaves
decreased from 45 to 37% while branch N increased from 6 to 27% as tree canopy
volume increased from 5 to 35 m3. Leaf, branch, and root masses comprised 15, 65, and
xiv


105
Fig. 5-6. Comparison of estimated crop evapotranspiration (ETC) with calculated
reference evapotranspiration (ET0) ratio. This value represents an approximation
of Kc for observations when soil water content values were near field capacity. Kc
values are expressed as a function of day of year (DOY).
Table 5-3. Regression analysis of estimated ETC to calculated ET0 ratio by day of year for
soil water content values greater than 0.070 cm3 cm'3 (field capacity) using a quadratic
function2.
Yo
a b R2
RMSE
P
DOY
0.71
0.004 -0.00001 0.76
0.056
<0.0001
z Y = Y0 + aX + bX2 where X = ET/ETo, and Y0 a, and b are regression coefficients
the course of a season, daily ETC values were multiplied by the appropriate Kc value
estimated for the DOY. ETc*Kc to ET0 ratios were then calculated to approximate K* and
were plotted against ASWD and (Figs. 5-7 and 5-8). Lines indicating estimations of Ks
using equation 5-2 and 15% depletion as an estimate for RAW are included in Figs. 5-7
and 5-8. Estimated Ks values based on ASWD are presented in Table 5-4. Table 5-5
shows higher R2 values and lower RMSE values compared with ETC to ET0 ratio


54
index on a tree basis (LAIt) increased rapidly from 4 to 10 as TCV increased from 2 to 10
m3 (Fig. 3-3 A) and TCSA increased from 20 to 80 cm2 (Fig. 3-3B). Little increase in
LAIt was observed with increasing TCV and TCSA beyond 10 m3 and 80 cm2,
respectively. Likewise, leaf area index on an acre basis (LAIa) increased from 1 to 6.3 for
the same ranges of TCV and TCSA
Citrus dry weight accumulation for all tree components increased linearly with
increased TCV and TCSA (Fig. 3-4 and Fig. 3-5). Regression coefficients, r2, and root
i
mean square error (RMSE) values for dry weights of all tissues vs. TCV and TCSA are
provided in Tables 3-5 and 3-6, respectively. Coefficients of determination (r2) were
generally higher for each tissue category when compared with TCV than with TCSA.
Biomass weights for twig, trunk and root categories varied greatly, resulting in lower r2
and higher RSME values. The medium branch masses varied more than the small or large
branch categories, possibly indicating inconsistent and/or incomplete separation of tree
components into appropriate diameter ranges. Correlations of trunk and taproot weights
with TCV and TCSA were poor compared with those for other tree components.
Variation in dry matter allocation to root and tap root were apparently due to differences
in root density distribution of the two rootstocks used in this study.
Dry matter accumulation in above-ground biomass increased from 60 to 75%
across a range of 5 to 40 m3 and 10 to 160 cm2 for TCV (Fig. 3-6A) and TCSA (Fig. 3-
7A), respectively. Leaf biomass declined from approximately 20% of total biomass for
trees with TCV of less than 5 m3 and for TCSA values below 20 cm2, to approximately
12% of total biomass for trees with TCV and TCSA values greater than 30 m3 and 160


NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER
UPTAKE PARAMETERS FOR CITRUS
By
KELLY TINDEL MORGAN
A DISSERTATION PRESENTED TO THE GRADUARE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004


103
1.2-
0 10 2D 30 40 50 60
Available Soil Wbter Depletion (%)
Fig. 5-5. Estimated ETC to calculated ET0 ratio as a function of available soil water
depletion in the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total
tree area to a 1 m depth (C). The data points shown represent a range of soil
water content from field capacity to approximately 50% available soil water
depletion.


24
sized individual grains (such as clays compared with sands) have a larger surface area
that increases the drag on water molecules that flow through the soil, reducing
permeability and hydraulic conductivity.
As water is lost from the soil, the continuity between water-filled pores also
decreases. A soil with water-filled volume fractions less than 0.1 (0 < 0.1 cm cm') will
normally have a very low value for K(0) (Tinker and Nye, 2000). The Poiseuille
equation states that the flow rate in a tube increases proportionally to the fourth power of
its radius, at a constant pressure gradient. Water in larger soil pores will empty first as the
soil dries, effectively reducing the cross sectional diameter of the soil water pathway.
Therefore, pore size and distribution has a large effect on the flow rate
f = (7C r4/ 8rj) dP/dx Equation 2-2
Where:
f = flow rate in a tube (m s'),
r = radius (m),
q = viscosity (pPa s'1), and
dP/dx = pressure gradient.
The driving force for soil-water movement is the difference in matric potential,
resulting from a difference in soil water content. Richards postulated that Darcys Law
could be extended to unsaturated states by assuming that the hydraulic conductivity (K),
as well as the water content, could be treated as non-hysteretic functions of the pressure
head or potential (Slatery, 1967). The matric potential and the water content for a soil are
related by the soil-water characteristic curve. By using the slope of the soil characteristic
curve (d/d0), the following equation can be obtained based on Darcys Law and is


108
Table 5-4. Estimated soil water coefficient (Ks) values for a range of percentage available
soil water depletion (ASWD) using Equation 2 as reported by Allen et al., (1997). A
value of 15% ASWD is used for RAW, therefore estimated K* for ASWD less than 15%
are assumed to equal 1.0.
ASWD
(%)
Estimated
Ks
0
1.00
10
1.00
20
0.94
30
0.82
40
0.71
50
0.59
60
0.47
70
0.35
80
0.24
90
0.12
100
0.00
Table 5-5. Regression analysis of estimated soil depletion factor (K*) by mean soil water
potential, and available soil water depletion in soil 0.5 or 1.0 m deep and in either the
irrigated zone or total tree area.
Soil Water Potential Logistic Function2
Yo
Xo
a
b
R2
RSME
P
0.5 m irrigated
0.58
11.50
0.44
2.97
0.53
0.086
<0.0001
1 m irrigated
0.61
11.49
0.41
3.66
0.47
0.091
<0.0001
1 m total
0.65
12.25
0.34
5.11
0.47
0.091
<0.0001
Available Soil Water Depletion Logistic Function2
Yo
Xo
a
b
R2
RMSE
p
0.5 m irrigated
0.40
42.37
0.58
1.98
0.63
0.076
<0.0001
1 m irrigated
0.48
34.85
0.48
1.55
0.39
0.097
<0.0001
1 m total
0.60
28.70
0.38
1.86
0.32
0.103
<0.0001
Z
Y= Y0 +
where X = Ks, and Yo a, and b are regression coefficients


148
accumulated almost 50% more N in large branches compared with trees on Swingle.
Total leaf N weight increased from less than 30 to more than 250 g tree'1 across the range
of canopy volumes and trunk diameters measured. Leaf N accounted for 45% of total N
in trees with canopy volumes less than 5 m3 and 37% of total N in trees greater than 35
m3. Twig N ranged from less than 10 to greater than 50 g tree'1 through the same range of
tree sizes. However, unlike leaves, twig N corresponded to a consistent 9% of total tree
N. Total branch N weight increased from less than 10 to greater than 200 g N tree'1,
which corresponded to an increase in percentage of total tree N from 6 to 27% for the
range of tree sizes measured. Concurrently, the proportion of total tree N in the trunk
decreased from 5 to 3%.
Mature Hamlin Root Distribution
Average root length density of fine fibrous roots surrounding mature citrus trees
followed a bimodal spatial distribution with depth from the soil surface and decreased
with distance from the tree trunk. Mean fine fibrous root density in the upper 15 cm was
1.036 cm cm'3 and ranged from 1.9 cm cm'3 at 50 cm from the tree trunk to 0.7 cm cm'3
at the 200 cm distance. Mean densities decreased at the 15 to 30 cm depth to 0.30 cm cm'
3 ranging from 0.50 to 0.07 cm cm'3 at 50 and 200 cm distances, respectively. Mean
densities of fine fibrous roots increased with subsequent depths to a maximum at the 60
to 75 cm depth and then declined at the 75 to 90 cm depth.
Differences in root spatial distribution between rootstocks were not statistically
significant. Mean root length densities at all depths and distances were 0.36 cm cm'3 for
trees grown on Carrizo citrange and 0.41 cm cm3 for trees grown on Swingle citmmelo.
Trees grown on Swingle had greater root length densities near the soil surface than did


64
trees, respectively. Therefore, the total estimated increases in N content for 2001 were
471 and 403 g tree1 for trees grown on Carrizo and Swingle rootstocks, respectively. The
amount of fertilizer N applied in 2001 was approximately 503 g tree'1, resulting in
apparent fertilizer N uptake efficiencies (FNUE) of 93.6 and 80.1% for Carrizo and
Swingle rootstocks, respectively. However, if the 58 g tree1 of N contained in the 1273
mm ha'1 of reclaimed water applied over the 12-month period is considered, NUE
decreases to 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively.
Nitrogen Change with Increase in Tree Size Experiments 1 and 2
Nitrogen accumulation increased linearly with increasing TCV and TCSA from 5
to 40 m3 and 20 to 160 cm2, respectively (Figs. 3-8A and 3-9A). Total leaf N mass
increased from less than 30 to more than 250 g N tree1 across the range of TCV and
TCSA measured (Fig. 3-8B and 3-9B). These increases were about 45% of total tree N
for trees with TCVs less than 5 m3 to 37% for trees with TCVs greater than 35 m3
(Figs.3-10B and 3-1 IB). Twig N ranged from less than 10 to greater than 50 g tree1
across the range of trees measured (Figs. 3-8B and 3-9B), but twigs still contained a
consistent 9% of total tree N regardless of tree size (Figs. 3-10B and 3-1 IB). Total N
accumulation by branches increased from less than 10 to greater than 200 g N tree1 for
trees with TCV of less than 5 and greater than 40 m3, respectively (figs. 3-8C and 3-9C),
which corresponds to an increase in percentage of total tree N from 6 to 27% for
corresponding tree sizes (Fig. 3-10C and 3-11C). The proportion of total tree N in the
trunk decreased from 5 to 3% (Figs. 3-10C and 3-11C). Regression coefficients, R2,
RSME, and probability values for dry matter and N accumulation within each tissue type
are presented in Tables 3-5 and 3-6.


LITERATURE CITED
Adams, J.S. 1875. Notes on wild orange groves. FI. Fruit Growers Assoc. Proc. 3:24.
Adriance, C.W. and J.E. Hampton. 1949. Root distribution in citrus as influenced by
environment. Proc. Amer. Soc. Hort. Sci. 53:103-108.
Allen, R.G. L.S. Pereira, D. Raes. And M. Smith. 1997. Crop evapotranspiration,
guidelines for computing crop water requirements. FAO Irrigation and Drainage
Paper No. 56. Rome Italy. 300 page.
Alva, A.K., and S. Paramasivam. 1998. Nitrogen management for high yield and quality
of citrus in sandy soils. Soil Sci. Soc. Am. J. 62:1335-1342.
Arias-Reveron, J.M., and H.W. Browning. 1995. Development and mortality of citrus
snow scale under constant temperature and relative humidity. Env. Entomology
24:1189-1195.
Batchelor, W.D., J.W. Jones, K.J. Boote, and G. Hoogenboom. 1994. Carbon-based
model to predict peanut pod detachment. Trans, of the ASAE 37(5): 1639-1646.
Batchelor, W.D., J.W. Jones, and K.J Boote. 1996. Comparisons of methods to compute
peanut seed size distribution by crop growth models. Trans, of the ASAE
39(2):737-744.
Batchelor, W.D., M R. Zeiss, L.P. Pedigo, and R.M. Shibles. 1997. Development of a
model to predict soybean pod color distribution. Trans, of the ASAE 40(1):221-
227.
Bellows, T.S., and J.G. Morse. 1986. Modeling flower development in navel oranges.
Scientifica Hortic. 30:117-126.
Bielorai, H. 1977. The effect of drip and sprinkler irrigation on grapefruit yield, water use
and soil salinity. Proc. Int. Soc. Citriculture 1:99-103.
Bland, W.L. and W.A. Dugas. 1989. Cotton root and soil water extraction. Soil Sci. Soc.
Am. J. 53(6): 1850-1855.
Boman, B.J. 1993. A comparison of controlled-release to conventional fertilizer on
mature Marsh grapefruit. Proc. Fla. State Hort. Soc. 106:1-4.
161


112
Allen et al.(1997) indicated that a region of readily available water exists between
0fc and approximately 30 to 50% ASWD for loam and loamy clay soils where there is
essentially no stress to the crop. This estimate is considerably reduced, however, in the
case of citrus on very sandy soils and may only amount to 10 to 15% of ASWD
Estimates for Ks using equation 5-2 closely approximate measured Ks values presented in
Fig. 5-5. It is therefore concluded that Ks decreased to approximately 0.6 at 50% ASWD,
which translates to a reduction of 40% in ETC between field capacity and 50% ASWD.
Koo (1963,1978) determined that stress associated with soil water depletion greater than
33% during periods of bloom, fruit set, and rapid vegetative growth in the spring months
can reduce potential yield, while depletions of 66% can be tolerated during sumer, fall
and winter months. Thus, crop stress associated with K values of 0.8 and 0.4 should b
used for irrigation scheduling from February through June and from June through
January, respectively, to maximize yields while minimizing water use.
Zaongo et al. (1994) reported close correlations beween water uptake and root
length density in millet and grain sorghum. Bland and Dugas (1989) estimated maximum
water uptake of cotton to be approximately 5 mm3 d'1 cm1 root. Hamblin and Tennqant
(1987) found that mean water uptake of cereals and grain legumes was less than 1 mm3 d
1 cm'1. Thus, soil water uptake per unit root length of 0.1 to 0.4 mm3 d'1 cm'1 root
observed for citrus in this study are similar to published values for other crops, which is
somewhat surprising considering the differences in root morphology of annual versus
perennial crops. Soil water uptake rates were closely related to root length densities, thus
soil regions containing higher root length densities will dry out at a proportionally higher


71
Assuming that 30% of the N accumulated in new growth tissues came from
fertilizer (Dasberg, 1987; Feigenbaum et al., 1987; Legaz et al., 1982), it is concluded
that 124 and 108 g of accumulated N originated from the fertilizer inputs. The calculated
FNUEs of 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively, in this study
were similar to the 61 to 83% reported by Syvertsen and Smith (1996) for 4-year old
grapefruit trees grown in lysimeters. However, mineralized soil N was not included in the
estimation. The contribution of soil organic matter and abscised tree parts will be
addressed in Chapter 6. Accounting for these sources of N will reduce overall NUE for
the citrus trees in this study.
Citrus trees at the Conserv II site were of the same age and similar in size, but the
K.D. Revell grove operated by Cargill, Inc. contained trees of various ages due to past
replanting of trees. The water and nutrient holding characteristics of the soil at this site
were similar to those of the soil at Conserv II. Therefore, it was assumed that trees grown
at this site would follow similar biomass and N partitioning characteristics. Total leaf
area for tree§ju.both experiments increased linearly with TCV and TCSA. LAI increased
rapidly with tree size until the trees were approximately 3-4 years old, after which it
stabilized at about 10. This information can be used to parameterize light interception
functions for a citrus tree photosynthesis and growth model.
Significant relationships were found between total tree fresh and dry biomass and
tree size. The ratio of above-ground to below-ground biomass and N content ranged from
a low of 3:2 to a high of 3:1 across the range of tree sizes found in this study. Citrus roots
of mature citrus trees extend below 1.8 meters (Castle, 1978 and 1980; Elezaby, 1989;
Menocal-Barverena, 2000), but the root system in this study was excavated to a depth of


NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER
UPTAKE PARAMETERS FOR CITRUS
By
KELLY TINDEL MORGAN
A DISSERTATION PRESENTED TO THE GRADUARE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2004

ACKNOWLEDGMENTS
I would like to acknowledge those people without whose help this study would
have been impossible. Foremost, I would like to thank my cochairmen, Drs. Thomas
Obreza and Johannes Scholberg, not only for financial and moral support beyond my
expectations, but most of all for their patience. I would like to thank Dr. Adair Wheaton
who shared so willingly his laboratory space and knowledge gained over 40 years of
research. Many thanks are owed to Drs. Nick Comerford and Jim Jones for providing
much needed insight into root uptake and crop modeling, respectively.
This project would not have been possible without the help of many people during
the collection of samples and chemical analysis. The dedication of Tom Graham and Sam
Luther for providing essential organization to the collection of more than 4000 plant
tissue and 3100 soil samples and Marjie Cody and Amanda Myers during many hours of
sample preparation and tissue analysis was greatly appreciated. I would be remiss if I did
not thank Drs. Harold Browning, Bill Castle, and Larry Parsons for allowing me the time
off from my full-time job to pursue this degree. Their patience when the demands of the
job and degree delayed the delivery of data or work on their projects was much
appreciated. My sincere thanks are also due to the Florida Department of Agriculture and
Consumer Services and Cargill, Inc. for providing funding for this project.
ii

Ill
Lastly, and most importantly, I would like to thank my wife, Nancy, and two
sons, Joshua and Christopher, for their unwavering support and encouragement.
Encouragement from my parents, in-laws, and brother sustained me.

TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ii
LIST OF TABLES viii
LIST OF FIGURES xi
ABSTRACT xiv
CHAPTER
1 INTRODUCTION 1
Ridge Water Quality Project 2
Citrus Best Management Practices 4
Decision Support Systems 5
Objectives 6
2 LITERATURE REVIEW 9
Introduction 9
Citrus Growth Characteristics 10
Citrus Biomass Distribution 11
Citrus Nitrogen Accumulation and Partitioning 12
Citrus Root Growth Dynamics 13
Factors Affecting Root Distribution And Root Density 14
Soil Characteristics 14
Climatic Effects 16
Rootstocks 16
Tree Spacing and Density 17
Fertilization 17
Irrigation 17
Canopy Reduction 18
Citrus Water Uptake 18
Factors Affecting ETC 19
Crop species 19
Tree size 20
IV

V
Climate 21
Soil characteristics 21
Soil water content 26
Water table 27
Soil shading 27
Grass and weed growth 28
Crop Coefficient 28
Soil Water Depletion Coefficient 29
Citrus Nitrogen Uptake 30
Seasonal Nitrogen Uptake 31
Nitrogen Uptake Efficiency 33
Seasonal Nitrogen Redistribution 33
Crop and Environmental Models 34
Current Citrus Models 34
Environmental Models 34
Crop Models 36
Conclusions 36
3 CITRUS BIOMASS AND NITROGEN ACCUMULATION 38
Introduction 38
Methods and Materials 42
Experiment 1 Mature Citrus Biomass and N Distribution 42
Experiment 2 -Biomass and N Accumulation With Increase in Tree Size 42
Site Description 43
Tree Canopy Volume and Trunk Cross Sectional Diameter 43
Tree Biomass Fresh Weight 44
Sample Processing and Nitrogen Analysis 45
Leaf Area, Biomass, and N weight Estimation 46
Statistical Analysis 47
Results 47
Mature Citrus Tree Biomass Distribution Experiment 1 47
Biomass Changes with Increase in Tree Size Experiments 1 and 2 50
Nitrogen Distribution 60
Mature Citrus Tree N Distribution Experiment 1 60
Nitrogen Balance 63
Nitrogen Change With Increase in Tree Size Experiments 1 and 2 64
Discussion 69
Conclusions 73
4 CITRUS ROOT GROWTH DYNAMICS 74
Introduction 74
Methods and Materials 77
Sample Collection 77
Sample Processing and Statistical Analysis 77

VI
Results 78
Mature Hamlin Orange Root Distribution 78
Root length Density Distribution Changes with Tree Size 83
Discussion 87
Conclusions 88
5 CITRUS WATER UPTAKE DYNAMICS 90
Introduction 90
Methods and Materials 94
Site Characteristics 94
Soil Capacitance Sensor Data Collection 94
Estimation of Daily ETC 96
Estimation of Monthly Crop Coefficient (IQ 97
Estimation of Water Stress Coefficient (IQ 97
Estimation of Soil Water Uptake per Unit Root Length 97
Results 98
Seasonal ET0 and ETC Trends 98
Seasonal K; K* 100
Kg Estimation 100
Soil Water Uptake per unit Root length Density 109
Discussion 110
Conclusions 113
6 CITRUS NITROGEN UPTAKE AND CYCLING 115
Introduction 115
Methods and Materials 119
Site Characteristics 119
Experiment 1 Nitrogen Uptake Flux 119
Fertilizer rate and application 119
Soil sampling procedures 120
Analytical methods 120
Experiment 2 Seasonal Tissue N Concentration 121
Tissue samples collected 122
Tissue analysis 122
Experiment 3 Seasonal N Loss 122
Results 123
Nitrogen Uptake 123
Nitrification Estimation 130
Seasonal Tissue N Concentration 130
Seasonal Fertilizer Use Efficiency 136
Seasonal N Losses 137
Discussion.. 138
Conclusions 144

Vil
7 SUMMARY AND CONCLUSIONS 146
Mature Tree Biomass Distribution 146
Biomass Vs Tree Size Relationships 147
Nitrogen Distribution 147
Mature Hamlin Root Distribution 148
Root Length Density Distribution Changes with Tree Size 149
Seasonal ET0 and ETC Trends 149
Seasonal Kc 150
Kc Estimation 150
Soil Water Uptake per Unit Root Length 150
Nitrogen Uptake 151
Nitrification Estimation 152
Seasonal Tissue Nitrogen Concentration 152
Seasonal Nitrogen Loss 153
Citrus Decision Support System 154
Citrus N practices and BMPs 155
APPENDIX
A Equations 158
LITERITURE CITED 161
BIOGRAPHICAL SKETCH 174

LIST OF TABLES
Table gage
3-1. Citrus biomass and nitrogen distribution by tree age as reported
by different studies 41
3-2 Mature citrus tree leaf area and leaf area index as a function of tree
canopy volume and trunk cross sectional area 48
3-3. Dry matter accumulation and allocation between tree components for
mature Hamlin orange tree by tree canopy volume as affected by
year of sampling, rootstock, and interaction of year and rootstock 49
3-4. Nitrogen accumulation and allocation between tree components for
mature Hamlin orange tree by trunk cross sectional area as affected
by year of sampling, rootstock, and interaction of year and rootstock 51
3-5. Linear regression analysis of dry weight and N accumulation in different
tree components as related to tree canopy volume 58
3-6. Linear regression analysis of dry weight and N accumulation in different
tree components as related to trunk cross sectional area 59
3-7. Mature Citrus tree tissue N concentration as a function of year of
sample and rootstock 63
4-1. Mean mature fibrous (diameter < 4 mm) root length
density ofHamlin orange tree as affected by rootstock, orientation
distance, and soil depth 79
4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length
densities for trees ranging from 2years old to > 15 years old 85
4-3. Regression coefficients and statistics root length density as a function
of distance from the tree trunk, and soil depths by canopy volume
using a third order quadratic polynomial model 85
4-4. Regression coefficients and statistics for root length density as a function
of distance from the tree trunk, and soil depths by trunk diameter
using a third order quadratic polynomial model 86
viii

IX
5-1. Monthly maximum, minimum, and mean reference evapotranspiration
reported by Florida Automated Weather Network for the Avalon
Station and maximum, minimum and mean estimated citrus crop
evapotranspiration betweeen April 2000 and March 2002 99
5-2. Regression analysis of estimated ETC to calculated ET0 ratio by
mean soil water content, soil water potential, and available
soil water depletion in soil 1.0 or 0.5 m deep and in either the
irrigated zone or total tree area 104
5-3. Regression analysis of estimated ETC to calculated ET0 ratio by day
of year for soil water content values greater than 0.070 cm3 cm'3
(field capacity) using a quadratic function 105
5-4. Estimated soil water coefficient (Ks) values for a range of percentage
available soil water depletion (ASWD) using equation 2 found in
Allen et al. (1997) 108
5-5. Regression analysis of estimated soil depletion factor (K*) by mean soil
water potential, and available soil water depletion in soil 1.0 or 0.5 m
deep and in either the irrigated zone or total tree area 108
5-6. Regression analysis of estimated soil water uptake per unit root length
density on soil water potential in soil at three locations surrounding the
tree and 10, 20,40 or 80 cm depths using an exponential decay model 110
6-1. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in March, 2002 125
6-2. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in May, 2002 126
6-3. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive days
in September, 2002 127
6-4. Regression equations for estimated maximum N uptake and estimated
active N uptake rates by soil N concentration (mg l'1) using an
exponential rise to a maximum model 129

X
6-5. Seasonal changes in N concentration, size and dry wt. of fruit, flush
leaves, and expanded leaves for 2001 and 2002 seasons
133

LIST OF FIGURES
Figure page
1-1. Map of Florida with Lake, Polk, and Highlands counties highlighted 3
1 -2. Plant/soil nitrogen and water balance flow chart 7
3-1. Tree canopy volume as a function of trunk cross sectional area for
trees from experiments 1 and 2 52
3-2. Leaf area expressed as a function of tree canopy volume (A) and
trunk cross sectional area (B) 53
3-3 Leaf area index on a tree basis expressed as a function of tree
canopy volume (A) and trunk cross sectional area (B) 55
3-4. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
canopy volume 56
3-5. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root dry weight accumulation as a function of
trunk cross sectional area 57
3-6. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and root and tap root dry weight accumulation
as a function of canopy volume 61
3-7. Dry weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of trunk cross sectional area 62
3-8. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of canopy volume 65
3-9. Total, above ground, and below ground, leaf and twig, total branch,
and root and tap root N accumulation as a function of trunk cross
sectional area 66
xi

Xll
3-10. N weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of canopy volume 67
3-11. N weight allocation to total, above ground, and below ground dry
weight, leaf and twig biomass, total branch biomass and root and
tap root accumulation as a function of trunk cross sectional area 68
4-1. Root length density distribution by depth at 50, 100, 150, and 200 cm
distance form tree trunk between rows of Hamlin orange trees on
Carrizo citrange or Swingle citrumelo rootstocks 80
4-2. Root length density distribution at 0-15,15-30, 30-45, 45-60, 60-75,
and 75-90 cm depth increments by distance from the tree trunk
as affected by distance form the tree trunk for Hamlin orange
trees on Carrizo citrange and Swingle citrumelo rootstocks 81
4-3. Citrus root distributions by depth below the soil surface and distance
from the tree trunk for trees 2-5 years old, 5-10 years old, 10-15
years old, and > 15 years old 84
5.1. Illustration of EnviroSCAN probe 95
5-2. Illustration of EnviroSC AN probe layout, and soil surface area used
for determining soil water content for each probe 95
5-3. Estimated ETC to calculated ET0 ratio as a function of soil water
content in the irrigated zone to a 0.5 m depth, 1 m depth, and
the total tree area to a 1 m depth 101
5-4. Estimated ETC to calculated ET0 ratio as a function of soil water
potential in the irrigated zone to a 0.5 m depth, 1 m depth, and
the total tree area to a 1 m depth 102
5-5. Estimated ETC to calculated ET0 ratio as a function of available
soil water depletion in the irrigated zone to a 0.5 m depth,
1 m depth, and the total tree area to a 1 m depth 103
5-6. Comparison of estimated crop evapotranspiration (ETC) with calculated
reference evapotranspiration (ET0) ratio which are an approximation
of Kc for observations when soil water content values were near
field capacity as a function of day of year (DOY) 105
5-7. Estimated soil water coefficient Ks as a function of soil water
potential in the irrigated zone to a 0.5 m depth, 1 m depth, and

Xlll
the total tree area to a 1 m depth 106
5-8. Estimated soil water coefficient Ks as a function of available soil
water depletion in the irrigated zone to a 0.5 m depth, 1 m depth,
and the total tree area to a 1 m depth 107
6-1. Relationship of estimated maximum N uptake to soil solution
concentration 128
6-2. Relationship of estimated passive N uptake to soil solution
concentration 128
6-3. Proportions of nitrate-N, ammonium-N, and total-N from control
pipes as percentage applied during 3 days after application 131
6-4. Seasonal change in N concentration for flush leaves, expanded
leaves, and twigs for 2001 and 2002 132
6-5. Seasonal change in N concentrations for bark and wood tissue of
small, medium, and large limbs for 2001 and 2002. High N rate
and Low N rate equal to 268.8 and 179.2 kg ha'1 yr'1, respectively 135
6-6. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2001 season 139
6-7. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2002 season 140

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
NITROGEN AND BIOMASS DISTRIBUTION, AND NITROGEN AND WATER
UPTAKE PARAMETERS FOR CITRUS
By
Kelly Tindel Morgan
May, 2004
Chair: Thomas Obreza
Cochair: Johannes Scholberg
Major Department: Soil and Water Science
During the last two decades, microirrigation and fertigation have become
commonplace in Florida citriculture. Concurrently, competition for water and nitrate
contamination of ground water have become greater concerns. Information on seasonal
citrus N demand, water use, and N uptake is needed to minimize water use and nitrate
leaching in citrus management. The objectives of this study were to 1) determine long
term and seasonal changes in citrus biomass and N distribution, 2) develop spatial
patterns of citrus root length density with increase in tree size, 3) estimate crop water use
and soil moisture coefficients, and root water uptake efficiency, and 4) explore seasonal
N uptake rates for citrus. Total tree biomass and N distribution were related to canopy
volume and trunk cross sectional area. The percentage of total tree N in citrus leaves
decreased from 45 to 37% while branch N increased from 6 to 27% as tree canopy
volume increased from 5 to 35 m3. Leaf, branch, and root masses comprised 15, 65, and
xiv

XV
20% of total mature tree mass, and accounted for 45, 35, and 20% of total N mass,
respectively. Root density increased radially as tree size increased. Trees on Swingle
citrumelo rootstock had a higher proportion of fibrous roots near the soil surface than
trees on Carrizo citrange. Soil water uptake ranged from 0.8 to 1.1 of ET0. Daily uptake
decreased steadily as soil water content decreased. N uptake from the upper 45 cm of soil
was greater for trees on Swingle citrumelo compared with Carrizo citrange. N uptake
efficiency ranged from 41 to 55% when fertilized at 269 kg N ha'1 compared with 47 to
70% when fertilized at 179 kg N ha'1. Leaf and twig N was highest from August to
February and lowest in May. New and improved understanding of citrus water and N
dynamics will advance Florida citrus management techniques and decrease
environmental impacts.

CHAPTER 1
INTRODUCTION
With a crop value of $640 million in 2002, citrus is one of the most important
horticultural crops in Florida. Currently, nearly 2 million ha are under citrus production,
with a 1.1 million metric ton annual production accounting for 73 and 18% of US and
world production, respectively (Florida Agricultural Statistics Service, 2002). Citrus is
typically produced on sandy soils with poor water and nutrient retention capacity.
Adequate supply of both irrigation water and fertilizer are therefore required for optimal
production. Most ridge soils lack confining soil layers that can prevent fertilizer nitrates
from reaching groundwater. Two issues have become greater concerns for citrus
production in Florida: 1) increasing competition between agricultural, commercial, and
residential use of limited water supplies, and 2) nitrate contamination of some aquifers
less than 50 m deep.
Fertilizer application rates and irrigation management practices for citrus rely
upon crude general recommendations that are standardized over large areas and lack the
precision needed in todays ecologically conscious and competitive markets. Although
fertilizer and irrigation recommendations provide general production guidelines, they do
not capture the dynamic nature of processes controlling non-point source pollution
associated with citrus production. Therefore, both growers and regulators must be
provided with additional tools such as decision support systems to improve water and
1

2
nutrient use efficiencies and assessment of Best Management Practices (BMP) impacts
on citrus production and ground water quality.
Ridge Water Quality Study
The U S. Environmental Protection Agency, in a nation-wide survey, documented
widespread nitrate contamination of shallow drinking water wells (Graham and Alva,
1995). In that survey, approximately 55% of wells were found to contain NO3-N
contamination above the background concentration. Approximately 1.2% and 2.4% of
urban and rural drinking water wells, respectively, were found to contain NO3-N
concentrations above the Maximum Contamination Level (MCL) of 10 mg L'1 for
drinking water. A correlation between drinking water well contamination and areas with
higher fertilizer sales and high value crops was established (Graham and Alva, 1995)
suggesting that agricultural fertilization practices may have contributed significantly to
NO3-N contamination of drinking water wells.
Of 3949 drinking water wells surveyed by the Florida Department of Agriculture
and Consumer Services (FDACS), 2483 (63%) contained detectible concentrations of
NO3-N (Graham and Wheaton, 2000). Of these 2483 contaminated wells, 584 (15% of
total surveyed) contained NO3-N in excess of MCL. The proportion of wells in Florida
contaminated with NO3-N was similar to that of the nation-wide survey. However, the
proportion of wells contaminated above MCL was an order of magnitude higher,
suggesting that the soils of the state of Florida on average are vulnerable to NO3-N
leaching to groundwater. Eighty-nine percent of wells contaminated above MCL were
located in the central Florida counties of Lake, Polk, and Highlands (Fig. 1-1). Portions
of these three counties comprise the central Florida ridge. Soils typical of the ridge are

3
hyperthermic Entisols composed of uncoated sands with water holding capacities of 0.04
to 0.09 cm3 cm'3, hydraulic conductivities >50 cm h*1, cation exchange capacities of 1 to
5 cmol (+) kg'1, and depths of more that 10 m.
Long-term monitoring studies and research projects were initiated in 1992 to
evaluate the impacts of nutrient and water management practices in citrus on ground
water quality. The goals of the projects established by FDACS (Graham and Alva, 1995)
were to 1) generate baseline groundwater quality data from several commercial citrus
groves in the ridge area; 2) develop recommendations for alternative nutrient and water
management practices; and 3) assess the impacts of these alternative management
practices on groundwater quality.
Fig. 1-1. Map of Florida with Lake, Polk, and Highlands counties highlighted.

4
Results of this study indicated that N removed at harvest accounted for only 30 to
49% of the applied N (Alva and Paramasivam, 1998). Estimates of N added to the
biomass of the trees ranged from 18 to 57% of the N applied. The study concluded that
additional information was needed for N accumulation with increase in tree size, optimal
timing for N application, N uptake parameters, and improved irrigation scheduling
(Graham and Wheaton, 2000).
Citrus Best Management Practices
A best management practice (BMP) for any agricultural commodity is an attempt
to use the latest scientific data available to reduce the impact of agricultural operations on
the environment while maintaining economically viable production. An interim BMP for
citrus was established in 1994 that was based on previous N rate studies and current IF AS
recommendations. Citrus growers agreeing to abide by the interim BMP would not be
held liable by the Florida Department of Environmental Protection for future cost of
supplying drinking water to local users as required by Chapter 376.30 (3) (c) F.S.
(Graham and Alva, 1995).
The terms of the interim BMP for orange trees 4 years or more of age were quite
broad. Annual N applications were restricted to 134 to 269 kg ha'1 with the stipulation
that groves producing less than 50.4 Mg of fruit per ha should apply no more than 202 kg
ha'1 N annually. A minimum of two applications per year were required for bearing
groves receiving up to 168 kg ha'1 N. Bearing groves receiving more than 168 kg ha'1 N
per year were required to receive at least three applications. Those groves using
fertigation were required to make a minimum of 10 applications. Application of at least
half of the annual fertilizer N prior to the rainy season was encouraged. A UF-IFAS

publication (Tucker et al., 1995) was produced to assist growers in determining the rate
of N to apply, timing of application, and suggested irrigation scheduling.
In 2002, a revised BMP established rates and timing of N applications based on
tree age classes and method of application. The two age classes are 4 to 7 years and >7
years. The methods of application are broadcast only, broadcast and fertigation, and
fertigation only. No more than 34 kg ha'1 N is to be applied at one time, and no more than
34 kg ha'1 N may be applied from June 15th to September 15th. No fertigation application
is to exceed 17 kg ha'1 N and must be applied at a minimum 1-wk interval.
Decision Support Systems
Important decisions for growers are when and how much fertilizer and irrigation
to apply. They need to consider several factors in their decision-making process to
determine that the crop value to be gained is greater than the cost of fertilizer and
irrigation applied. Fixed fertilizer and irrigation schedules, based on long-term mean
climatic conditions, may lead to inefficient use of these inputs due to the large annual
variability in atmospheric conditions (Heinemann et al., 2000). Likewise, variations in
the amount of rainfall and its distribution may lead to the loss of N from the crop root
zone necessitating additional applications. Due to the complexity of the decision making
process, researchers have developed computer-based decision support systems (DSS). A
DSS can provide information on management options based on local environmental
conditions. These systems also provide a means to make the scientific understanding of
complex plant, soil, and environmental interactions accessible to decision makers in a
concise and interactive manner. Frequently, information from simulation models has
formed the foundation for these DSS.

Nutrients leached from agricultural soils represent both an economic loss to
farmers and a potential environmental pollutant for groundwater. Concerns about the
presence of these agricultural chemicals in groundwater and the need for improved
understanding of their movement and transport beyond the root zone have increased
considerably over the last several decades. Comprehensive mathematical relationships are
required to determine crop fertilizer N uptake and to predict the potential impact of NO3-
N leaching on groundwater quality for various soil and/or environmental conditions. With
the exception of insect and disease population and damage dynamics, most modeling
work has focused on predicting mineral N transformations, organic C and N
transformations, soil water content, water and N uptake, crop yield, and NO3-N leaching
(Hoogenboom et al., 1994). Such models help growers manage resources, maximize
returns, and reduce impacts on water quality. Current crop simulation models are being
used to optimize planting dates and densities (Saseendran et al., 1998), optimize fertilizer
and irrigation inputs (Sexon et al., 1998, and Heinemann et al., 2000), maximize profits
(Kiniry et al., 1997), and reduce groundwater pollution (Gijsman et al., 2002) in
agronomic crops.
Objectives
Robust crop models can provide a scientific basis for improved resource
management in agricultural production. The long-term accumulation of biomass and N
with tree development must be understood. Figure 1-2 illustrates the relationships
between tree biomass, tree N content, soil N concentration, and soil water content.
Change in tree biomass and N content over time impacts total tree N demand. Soil N and
water concentrations affect both active and passive N uptake rates. Likewise, changes

7
C^^Oroundwat
3
Inputs and Outputs
State Variables
txZZH
Functions
Fig. 1-2. Plant/soil nitrogen and water balance flow chart.

8
in root distribution with time must be known to understand the effect of these
distributions on the rate of water and nutrient uptake and thus soil water and N
concentration. An additional complication is that citrus tree scions are grafted onto
rootstocks that affect the growth and uptake rates of the resulting tree. The effects of
these rootstocks on tree development and uptake rates must also be understood.
To develop a crop model for citrus, detailed field-scale information must be
obtained under local soil and cultural conditions. A review of the existing literature for
these relationships was conducted. The result of this review will be presented in Chapter
2. Information gaps in biomass and N accumulation and field-scale uptake rates of mature
citrus trees must be determined so that studies can be designed to complement existing
information. The central hypotheses tested in this dissertation are 1) generic relationships
can be developed that capture changes in citrus dry biomass, N weight, and root length
densities with increase in tree size; 2) daily water uptake changes seasonally and is
greatly affected by soil water content; 3) seasonal leaf N concentration is lowest and tree
biomass abscission is highest during periods of rapid tree growth; and 4) fertilizer-N
nitrification and uptake are rapid under Florida conditions. Therefore, the objectives of
studies in Chapters 3-6 were to 1) determine changes in above-ground citrus biomass and
N distribution for trees under recommended N fertilizer management practices across a
range of tree sizes; 2) develop relationships that will capture the overall spatial patterns in
citrus root length density distribution for different tree sizes; 3) estimate irrigation crop
and soil moisture coefficients and soil water use per unit root length density; 4) explore
seasonal N uptake rates for citrus; and 5) compare seasonal biomass and N concentration
changes for citrus fertigated at two N fertilizer rates.

CHAPTER 2
LITERATURE REVIEW
Introduction
The various species of the genus Citrus are believed to be native to the subtropical
and tropical regions of Asia and the Malay Archipelago (Webber and Batchelor, 1943).
Citrons were cultivated on the European continent as early as 300 BC. Limn and sweet
orange were not known in Europe for some 15 and 17 centuries later, respectively. Sweet
orange was first introduced to the North American continent on Columbus second
voyage in 1493. Citrus was grown in coastal settlements of Florida by the mid 16th
century and wild citrus trees were found on hammocks near lakes or rivers where
conditions were particularly favorable for their growth in the second half of the 19th
century (Harris, 1875; Adams, 1875). Due to freeze damage, citrus production in Florida
has moved in the last 120 years from north central Florida to the southern half of the
Florida peninsula. Currently 2 million ha of citrus fruit are grown in Florida (Florida
Agricultural Statistics Service, 2002). Florida citrus production was 1.1 million metric
tons accounting for 73 and 18% of US and world production, respectively.
Vaile (1924) showed that fine sand or sandy loam soils resulted in better growth
and production than coarse sands or heavy loams. Subtropical in nature, citrus trees do
not exhibit dormancy or shed their leaves during the winter months However, new
growth appears in definite cycles, with two to four cycles of growth yearly. The first and
usually largest growth starts in the early spring (late February to early March), the second
9

10
from early June to early July, and the third in late summer (August or September). The
principal blooming period for all commercial species is early spring and usually lasts
approximately 6 weeks (Mid February to late March). The normal period of ripening of
most citrus fruits is late fall and winter, preceding the spring bloom. However, late
maturing varieties such as Valencia require 12 to 15 months for maturity, which occurs
after bloom for the next crop. The period of time when citrus fruit can be harvested is
about 9 months.
Citrus production areas in Florida range from upland positions with very sandy
Ridge soils, which are deep and excessively well drained, to relatively low flatwood
sites that are often flooded in their native state due to the presence of a spodic layer.
Flatwoods soils must be drained and bedded before planting to lower the fluctuating
water table. Each of these general areas in Florida presents somewhat different challenges
for growing citrus trees and fruit production. Although crop growth and nitrogen uptake
dynamics are readily available for many agronomic crops, this information is in short
supply for citrus. A comprehensive literature review relevant to citrus growth
characteristics, root distribution, water and N uptake dynamics, and crop growth models
will be presented in this chapter.
Citrus Growth Characteristics
Turrell et al. (1969) proposed citrus tree growth equations based on growing
conditions and cultural practices in California. These equations assumed citrus tree
growth to bejogistic in nature depending on cultural characteristics such as spacing,
pruning, and irrigation and fertilizer scheduling, soil characteristics, and climatic
conditions. Several studies measuring citrus biomass in relation to nutrient concentration

11
distribution or uptake have been conducted, but few biomass studies have been conducted
for a range of tree sizes grown under similar cultural soil and climatic conditions.
Citrus Biomass Distribution
Legaz and Primo-Millo (1981) harvested 4 year-old Valencia orange trees
grown outdoors in sand culture in Spain at five different times in a 1-year period. The
mean dry mass percentages for leaves, twigs and branches, lateral roots and fibrous roots
were 22.5, 28.7, 45.8, and 2.9% respectively.
Cameron and Compton (1945) divided eight year old Valencia trees grown in
California into 14 parts, 1) leaves, 2) twigs, 3) shoots, 4) lateral branches 0.75-1.5 cm, 5)
tertiary branches 1.5-3.0 cm, 6) secondary branches 3.0-6.0 cm, 7) primary branches >
6.0 cm, 8) trunk, 9) main root, 10) feeder roots, 11) rootlets, 12) small roots 0.3-0.8 cm,
13) intermediate roots 0.8-2.5 cm, and 14) large roots >2.5 cm. The difference between
twigs and shoot was not given, but both had similar N concentrations on a dry mass basis
(5.11 and 5.02%, respectively). Likewise, the sizes of rootlets and feeder roots were not
given but had similar total N percentages (1.38 and 1.49%, respectively). Mean
percentage dry weight of leaves, twigs and shoots, branches, trunk, main root, lateral
roots and fibrous roots were 16.8, 9.6, 43.2, 1.5, 2.1, 14.0, and 1.7% respectively.
The biomass proportions for young Hamlin orange trees grown under Florida
conditions were substantially different. In a study by Mattos (2000), eight categories of
plant parts for 7 year-old Hamlin orange trees were used. These categories were 1)
summer-fall flush leaves, 2) spring and older leaves, 3) twigs >1.5 cm, 4) twigs <1.5 cm,
5) trunk, 6) roots <0.2 cm, 7) roots 0.2 to 1.0 cm, and 8) roots >1.0 cm. Leaves, branches,

12
trunk, lateral roots, and fibrous roots constituted 13.9, 37.4, 9.0,10.2 and 14.2% of total
dry biomass, respectively.
In a 15N study, Feigenbaum et al. (1987) divided 22 year-old Shamouti orange
trees fertilized at two annual N rates into seven components: 1) leaves, 2) twigs <5 mm,
3) branches >5 mm, 4) trunk and main branches, 5) main root (tap), 6) lateral roots >1
mm, and 7) fibrous roots <1 mm. There was no diameter given to differentiate between
branches and main branches. Biomass percentages for leaves were 6.4 and 8.4% for the
low and high N treatments respectively. Similar differences were found for all categories
with the exceptions of main and lateral roots. The ranges in mean percentage of fresh
biomass for twigs, branches, trunk and main branches, lateral roots, and fibrous roots
were 1.6 to 2.2, 32.5 to 33.4, 24.5 to 29.2,4.6 to 6.1, and 3.8 to 4.8%, respectively. These
differences between fertilizer-N rates resulted in a greater percentage of biomass below
ground for the low nitrogen application (34.1%) compared with that of the higher
nitrogen rate (27.8%).
In a similar study, Kato et al. (1984) found different values for 21 year-old
Satauma mandarin trees in Japan. In this study, mean percentages of dry biomass for
leaves, twigs, branches, trunk, lateral roots and fibrous roots were 15.7,4.7, 32.2, 23.1,
20.3, and 3.1% respectively.
Citrus Nitrogen Accumulation and Partitioning
Nitrogen balance studies in citrus provide information on physiological tree N
requirements and can be used to develop methods that minimize potential losses of N to
groundwater and the atmosphere (Feigenbaum et al. 1987). In a study of 8 year-old
Valencia trees in California conducted during a period of 2 years (Cameron and

13
Compton, 1945), leaves contained from 40 to 50% of total tree N. Twigs and shoots
contained approximately 10% of total N. Trunk and branches contained from 20 to 30%
of total tree N. Approximately half of this N was in the bark, whereas this tissue
component represented only 5% of the total dry mass. Roots contained from 15 to 20% of
the N, half or more of which was in the bark that made up only 5% of the dry mass of the
tree.
Seasonal changes in leaves, bark, twig, and root N concentration were greater
than N changes in woody branch tissue. The trees contained more N just before initiation
of growth activity during spring than at any other time of year. Maximum bark N content
occurred about December 1, followed by a decrease. It was speculated that the reduction
between December and February was the result of deposition of starch and possible other
carbohydrates in these tissues. A decrease of all tree tissue N concentration occurred
during bloom, fruit set, and periods of active growth in the spring and early summer.
During the summer and autumn, N concentrations gradually increased to the mid-winter
maximum. Mattos (2000) found similar N concentration values for 7 year-old Hamlin
trees. Nitrogen concentration was lowest in the trunk and taproot of these trees. The N
concentration of leaves (2.1 to 2.6%), twigs (0.4 to 0.8%), and roots (0.6 to 1.7%) varied
with tissue age. Younger tissue tended to have greater N concentration compared with
older tissues.
Citrus Root Growth Dynamics
Citrus trees are productive and grow well on central Florida deep sandy soils. In
some instances, tree size and yield appear to be related to root system characteristics
(Castle and Krezdom, 1975). Citrus fibrous roots are commonly defined as those roots <4

14
mm in diameter. Their dry mass is a relatively small part of the total root system, but
their composite length far exceeds that of the woody roots (>4 mm in diameter). These
fine roots are considered to be the functional part of the root system because of their
critical role in water and nutrient uptake. There is some variation among rootstocks in the
morphology of fibrous roots (Castle and Youtsey, 1977). Some rootstocks like trifoliate
orange [Poncirus trifoliata (L.) Raf ] produce higher specific root length or length/unit
mass (Eissenstat, 1991). Fibrous roots are also the most vulnerable part of the root
system. Their development, function, and longevity are strongly influenced by soil
characteristics, environmental changes, crop species, crop growth stage, and cultural
practices.
Factors Affecting Root Distribution and Root Density
Soil Characteristics
The distribution of roots is modified by the physical and chemical properties of
the soil profile (Hillel, 1971). Widespread root development and high fibrous root
concentrations were observed in deep soils of sand texture where there were virtually no
impediments to root growth provided that water and nutrients were non-limiting to
growth (Ford 1952; 1953a&b; 1954a&b; 1959; 1964; 1972); Ford et al., (1954; 1957).
Increased tree size and yield have been related to root system depth and fibrous root
mass. The depth of rooting of Orlando tngelo trees on 10 rootstocks growing in deep,
sandy soil was correlated with tree height (Castle and Krezdom, 1975). Although fibrous
root distribution was affected by tree height, total fibrous root dry mass measured at the
canopy dripline was not correlated with tree height. Ford (1954a; 1964; 1968; 1969;
1972) conducted many studies of citrus trees in poorly drained Spodosols of Florida and

15
concluded that tree size was closely related to fibrous root density. Extensive lateral root
development occurred on soils with loamy or clay texture (Boswell et al. 1975;
Kaufimann et al., 1972). In these studies, the root systems were shallower than root
systems of plants grown in sandy soils with few roots found below a soil depth of 50 to
70 cm (Adriance and Hampton, 1949; Boswell et al., 1975, Cahoon et al., 1956; 1959;
1961; Kimball et al., 1950; Mikhail and El-Zeflawi, 1978). Furthermore, changes in
fibrous root distribution with depth were more gradual compared with sandy soils, and
overall fibrous root concentrations were lower (Bielorai, 1977). The lower natural soil
fertility (Carlisle et al., 1989), and excessive drainage of sandy soils resulted in higher
shoot:root ratios such that fibrous root dry mass densities tended to be lower in sandy
soils (Castle, 1978).
Under flatwood conditions where the soil is drained and bedded, virtually all the
root mass occurs within 45 cm of the soil surface (Calvert et al., 1967; 1977; Ford,
1954a; Ford, 1972; Reitz and Long, 1955). The quantity of fibrous roots decreases with
depth and lateral distance from the trunk. Elezaby (1989) reported lateral fibrous root
distribution to a depth of 180 cm of a 10-year-old Valencia tree [Citrus sinensis (L.)
Osb ] on Volkamer lemon (C. volkameriana Ten. and Pasq.) grown on a soil with a
deep sand profile and spaced at 4.5 m x 6.0 m as: 9% of the fibrous roots between 0 cm
and 60 cm from the trunk, 31% between 120 cm and 180 cm, and 21% between 240 cm
and 300 cm. The vertical distribution was: 42% of the fibrous roots between 0 cm and 30
cm from the soil surface, and 14% or less at each 30-cm depth increment to 180 cm. In
the same study, fibrous root dry mass density (concentration) ranged from 300 g m3 to
1200 g m Those data are similar to dry mass densities reported in other Florida studies

16
(Castle, 1978; 1980). In a recent report, data were given as root length densities and
ranged from 530 cm m"3 for Swingle citrumelo roots [C. parachsi Macf. x P. trjfoliata
(L.) Raf] to 2020 cm m'3 for trifoliate orange (Eissenstat, 1991).
Climatic Effects
Root distribution was studied in 22 mature navel or Valencia orange groves in
California (Cahoon et al., 1956). In this study, 50% were low-yielding while the
remaining 50% were high-producing. Fibrous root fresh mass was measured to a depth of
90 cm under the canopy and between rows. Yield was not related to the under-canopy
root quantities, but was correlated with the root quantities measured between adjacent
rows where soil water contents were typically lower most of the year.
Rootstocks
Some citrus roots have been found as deep as 7 m (Ford, 1954b), and in one
instance, roots of mature trees on rough lemon rootstock were discovered 14 m from the
tree trunk (Ford, 1970). Castle and Krezdom (1975) described two general types of root
systems, the first characterized as extensive featuring extensive lateral and vertical
development, and the second as intensive with less extensive root expansion and higher
fibrous root concentrations mainly confined to the upper soil layers. Trees on rough
lemon, Volkamer lemon and Palestine sweet lime (C. limettioides Tan.) rootstocks
typified the extensive type of root system where 50% of the fibrous roots occurred below
70 cm in the soil with wider spreading lateral development. Examples of the intensive
type were Rusk citrange and trifoliate orange, where few fibrous roots were found
below 70 cm, and the root system was less developed laterally. Some rootstocks like sour
orange and Cleopatra mandarin were classified as intermediate. Trees on Cleopatra

17
mandarin had a highly developed lateral root system at the surface, but few fibrous roots
below the surface. Menocal-Barberena (2000) found no statistically significant
differences in vertical or horizontal fibrous root distribution ofHamlin orange on
Cleopatra mandarin, Swingle citrumelo and Carrizo citrange rootstocks. Vertical and
horizontal root distribution were similar to other studies with about 40% of the fibrous
roots in the top 30 cm and 9 to 14% at each of the remaining 30 cm depth increments to
180 cm. Few roots were found below 180 cm.
Tree Spacing and Density
Due to Floridas rainy season, roots of trees at commercial spacing rapidly
occupy the volume of soil outside the irrigated zone. After canopy closure, they extend
into the rootzone of adjacent trees. Elezaby (1989) reported fibrous root concentration in
the 0 to 30 cm zone increased from 450 to 1000 g m'3 between trees when the in-row
distance decreased from 4.5 to 2.5 m. The increased root concentrations in this study
were concluded to be the result of overlapping root systems. Trees at the closest spacing
showed root concentration increases to depths of 150 cm (Elezaby, 1989).
Fertilization
Increases in fertilizer N can increase root growth to a considerable depth, but the
largest effects generally occurred near the surface (Ford, 1953b; Ford et al., 1957; Smith,
1956; Smith, 1965).
Irrigation
Irrigation method and scheduling has been shown to change the distribution
and/or concentration of citrus fibrous roots. In a California study of trees receiving
different irrigation treatments, yield was not correlated with fibrous root density (Cahoon

18
et al., 1964) because trees in the low irrigation rate treatment declined in yield while
maintaining root quantities similar to those of the trees in the higher irrigation rate
treatment. It was concluded that soil water content was the single most important factor
influencing citrus root systems.
In a Florida study, root weight densities were determined under the tree canopy, at
the dripline, and in the row middles to a depth of 180 cm for Hamlin orange trees on
Swingle citrumelo and Carrizo citrange rootstocks (Menocal-Barberena, 2000). Trees
receiving irrigation at a rate of 40 cm yr'1 had significantly higher densities than trees
receiving 250 cm yr'1. The differences were on the order of 1.3 to 2.3 times greater for
the 40 cm yr'1 treatment at all depths.
Canopy Reduction
Hedging, the annual removal of excess vegetative growth, has become a common
method of canopy size control for closely-planted citrus trees. Eissenstat and Duncan
(1992) found that within 30 days of canopy reduction, 20% of the fibrous roots between
the 9 cm and 35 cm depths were apparently dead. Root length density of these trees
recovered within 63 days of canopy reduction. This relatively short-term reduction in
fibrous root density adversely affected yield because of fruit abortion.
Citrus Water Uptake
Assuming little or no surface runoff, water applied to the soil surface is 1)
retained in the soil, 2) utilized by plants, 3) lost to the atmosphere, or 4) drained below
the crop rooting zone. Drainage water may contain substantial quantities of agricultural
chemicals and soluble nutrients. Irrigation practices should be aimed at minimizing 1)
crop water stress by maintaining sufficient water within the crop rooting zone, 2)

19
pollution of groundwater by leaching, and 3) production costs associated with excessive
irrigation, and nutrient and pesticides losses due to leaching.
Mills et al. (1999) reported a significant decrease in citrus stomatal conductance
after midday. This decrease was most pronounced for south-facing exterior leaves and
increased with increasing evapotranspirational demand (ET0). Soil water use from 2 year-
old Hamlin orange trees measured at 0.5-hour intervals using weighing lysimeters
indicated that water continued to be removed several hours after the midday decrease in
stomatal conductance. Two seemingly opposing theories place control of soil water
uptake at the leaf level via leaf water potential (Slatery, 1967) or root via root water
potential (Tinker and Nye, 2000). The former assumes that leaf water potential exerts
control on stomatal conductance regulating transpiration and thus water uptake. The latter
speculates that dehydrating roots, due to low soil water content, indirectly control
stomatal conductance through the production of chemical compounds that after
translocation to the leaves reduce stomatal aperture. Lafolie et al. (1991) measured
decreasing leaf water potential with decreased root water potential until midday. After
reduced stomatal conductance at midday, leaf water potential increased without a
corresponding decrease in root water potential. This result was given as evidence that
stomatal conductance was not controlled by leaf water potential alone.
Factors Effecting ETC
Crop species
Citrus are evergreens and therefore require water for transpiration throughout the
year. Citrus leaves are thick and waxy, resulting in high cuticular resistance to
transpiration (Mills et al. 1999). Koo (1963) and Koo and Sites (1955) stated that water

20
requirements of grapefruit are generally higher than orange or mandarin varieties for trees
of equal size. Wiegand and Swanson (1982 a, b, c) and Wiegand et al. (1982) reported
that mean daily citrus ETC at Weslaco, Texas ranged from 2.2 to 3.3 mm for Ruby Red
grapefruit and 1.9 to 2.7 mm for Marrs oranges from 5 to 10 years of age.
Under similar climatic conditions, citrus trees are known to have lower
transpiration rates compared with other crop plants. Mahrer and Rytwo (1991) reported
mean estimated daily crop water use (ETC) rates for cotton in the Hula Valley of Israel of
5.4 mm when irrigated daily, and 4.0 mm during a 14-day period when not irrigated.
Likewise, Starr and Paltineanu (1998) reported that daily ETC rate for full canopy com at
Behsville, MD ranged from 3.8 to 5.0 mm prior to rainfall and 5.2 to 8.0 mm after.
Lower citrus transpiration rates are related to lower leaf and canopy conductance (Mills
et al. 1999).
Tree size
Large, vigorous, healthy trees require more water than young trees (Tucker et al.
1997). In Florida, large trees at low planting densities (150 to 180 trees per ha) may use
62 to 94 L per day during the winter months and 189 to 219 L per day in July and August
(Boman, 1994). Rogers and Bartholic (1976) reported a mean annual ETC of 1210 mm
during an 8-year period from a young orange and grapefruit grove on poorly drained soils
near the east coast of Florida. These annual ETC values ranged from 820 mm early in the
study (tree age 2 years) to 1280 mm at end of the study (tree age 10 years). Linear
regressions of annual ETe vs. years during the study resulted in significant (P=0.1)
increase in ETC. Mean annual ETC increased at a rate of 19 mm per year or a cumulative
increase of approximately 13% in 8 years. Fares and Alva (1999) reported an annual ETC

value of920 mm for 3-yr old Hamlin orange trees grown on deep sandy soils in central
Florida. Koo and Harrison (1965), and Koo and Hunter (1969) reported annual ETC
values of 1170 mm for mature citrus on the same soil series.
Climate
Mean annual ETC for citrus in Florida ranges from 820 to 920 mm (Rogers and
Bartholic 1976; Fares and Alva 1999) for young (<5 years) trees to 1170 to 1280 (Koo
1978, Rogers and Bartholic 1976) for mature (10 years or more) trees. Annual ETe values
reported for mature citrus grown in the lower Rio Grande Valley of Texas are similar to
those for Florida and ranged from 1044 to 1232 mm (Wiegand et al. 1982). Hoffinan et
al. (1982) reported annual ETC values for well-irrigated citrus grown in semi arid Arizona
of 1470 mm. Lower ETC rates for Florida (humid) compared with Arizona (semi-arid)
have been attributed to lower evaporative demand (Rogers et al. 1983, Fares and Alva
1999).
Soil characteristics
Crop water supply must be based on a clear understanding of soil water dynamics.
Water in excess of field capacity drains through the vadose zone. Eventually, water that is
not taken up by plants or evaporated from soil or plant surfaces makes its way into the
ground water and contributes to aquifer recharge (Fares and Alva, 1999). Under-tree
sprinklers and drip irrigation systems are designed to deliver water at rates low enough to
allow infiltration into the soil without contributing to losses by runoff. These systems can
be managed in such a manner that the excessive downward drainage through the soil is
minimized. The required application amount is governed by the soil-water depletion on a
given irrigation date, irrigation efficiency, and the target soil-water level. Most of the

22
terms are not independent. For instance, the amount of applied irrigation water will
influence the amount of ETC as well as the amount of drainage (Prajamwong et al., 1997).
In standard irrigation practices, water transport through the soil may be classified
into five phases: 1) infiltration during application; 2) redistribution after application
ceases; 3) withdrawal by plant roots; 4) evaporation from the soil surface; and 5) drainage
of water to lower soil depths. The primary modes of transport of water in soil are 1)
viscous flow through liquid-filled pores, and 2) diffusion of vapor through air-filled
>ores. In principle, both modes contribute to soil water flow. Liquid flow is the dominant
node in saturated to moist soils (Hagan et al., 1967). Vapor flow is of minor importance
until soils become quite dry, although the presence of a large temperature gradient favors
the contribution of this mechanism. For typical soil water situations, both of these
transport modes contribute to a flow rate proportional to potential energy gradients within
the soil.
Water is of central importance in the transport of solutes in soils or plants,
whether by diffusion or mass flow (Tinker and Nye, 2000). The concept of potential is
fundamental to understanding soil water dynamics. Potential is a measure of the energy
state of a chemical compound within a particular system, and hence of the ability of a
unit amount of this compound to perform work. Difference in potential at different points
in a system gives a measure of the tendency of the compound, including water, to move
from a region with high potential to a region of lower potential.
Soil water has various forms of potential energy acting on it, all of which
contribute to the total potential. Tinker and Nye (2000) refer to these forms of potential
energy as concentration, compression, position in an electrical field, and position in the

23
gravitational field. These same forms of energy are commonly referred to as osmotic,
matrix, gravitational and pressure potentials, the sum of which is referred to as total water
potential (4>). Thus, soil water moves in response to the difference in water potential over
a distance. The first published relationship between water flux and energy gradient was
obtained empirically in 1856 by Henry Darcy after a study of saturated sand filters
(Hagan et al., 1967).
u =-K d<|> /dx Equation 2-1
Where:
u = water flux (cm3 cm'2 s'1),
K = hydraulic conductivity constant (cm s'1),
<|> = soil water potential (kPa), and
x = the distance over which the flux is maintained (cm).
The constant of proportionality of Darcys Law (K) is known as the hydraulic
conductivity, and is a function of both the properties of the medium and the fluid (Tindall
and Kunkel, 1999). In saturated soils, K will be constant as long as the structure of the
soil remains stable because the water flow pathways will be unchanged. In unsaturated
soil, K varies with the water content (0), because the latter defines the total cross-section
area for water flow, the effective water-filled pore radius, and the effective pathlength
(Tinker and Nye, 2000). A soil with a wide range of pore sizes conducts fluid more
rapidly than a soil with small pore sizes (Tindall and Kunkel, 1999). The saturated
hydraulic conductivity of soils has a wide range from 10'9 cm s'1 for clay to 1.0 cm s'1 or
more for sand. Lower values of K for a clay medium (with smaller pore sizes) are likely
due to the drag exerted on the viscous fluid by the walls of the pores. Particles of smaller-

24
sized individual grains (such as clays compared with sands) have a larger surface area
that increases the drag on water molecules that flow through the soil, reducing
permeability and hydraulic conductivity.
As water is lost from the soil, the continuity between water-filled pores also
decreases. A soil with water-filled volume fractions less than 0.1 (0 < 0.1 cm cm') will
normally have a very low value for K(0) (Tinker and Nye, 2000). The Poiseuille
equation states that the flow rate in a tube increases proportionally to the fourth power of
its radius, at a constant pressure gradient. Water in larger soil pores will empty first as the
soil dries, effectively reducing the cross sectional diameter of the soil water pathway.
Therefore, pore size and distribution has a large effect on the flow rate
f = (7C r4/ 8rj) dP/dx Equation 2-2
Where:
f = flow rate in a tube (m s'),
r = radius (m),
q = viscosity (pPa s'1), and
dP/dx = pressure gradient.
The driving force for soil-water movement is the difference in matric potential,
resulting from a difference in soil water content. Richards postulated that Darcys Law
could be extended to unsaturated states by assuming that the hydraulic conductivity (K),
as well as the water content, could be treated as non-hysteretic functions of the pressure
head or potential (Slatery, 1967). The matric potential and the water content for a soil are
related by the soil-water characteristic curve. By using the slope of the soil characteristic
curve (d/d0), the following equation can be obtained based on Darcys Law and is

25
known as Richards equation (Tinker and Nye, 2000). In this equation, flow in
unsaturated soil can is expressed in terms of the water content gradient and soil water
diusivity (De).
u = -Ke d/dx = -Ke (d/d0) (d0/dx) = -De (d0/dx) Equation 2-3
Where:
u = water flux (cm3 cm'2 s'1),
Ke = hydraulic conductivity constant at 0 (cm s'1),
= soil water potential (kPa),
a >>
0 = soil water content (cm cm' ),
d0/d x = the distance over which the flux is maintained (cm).
The term diffusivity (De) is used because the form of equation is the same as that
of Ficks law of diffusion (Tinker and Nye, 2000). Furthermore, De is somewhat less
convenient than Ke under conditions of hysteresis because De is discontinuous at each
reversal of the direction of potential, while K is continuous and virtually hysteresis-free
(Hagan et al., 1967). Experimentally, the effect of hysteresis on Richards equation has
usually been ignored by limiting the soil water potential change to either always drying,
i J
or always wetting.
Field capacity (0fc) describes the water content held in the soil after excess
water has drained to drier soil layers by redistribution. This equilibrium can be
determined in the field by measuring the soil water content as a function of time to
determine the value of 0 when d0/dt approaches zero. Hillel (1971) noted that the rate at
which d0/dt approaches zero is dependent on 0¡ and the depth to which the soil is being

26
wetted. The concept of field capacity is useful in the design of field management schemes
for approximating the maximum amount of soil water storage. Field capacity can be used
as an upper limit value of 8 within each soil layer such that any water in excess of 0fC
quickly drains to the next deeper soil layer. The soil profile can be described as a vertical
sequence of reservoirs with the overflow level for each reservoir corresponding to the
value of 0fc for that specific soil layer. During irrigation or rainfall the top reservoir flows
over to fill the next lower reservoir until no excess water remains to flow into the next
reservoir. With a judicious selection of the depth of each soil layer, this simple analog of
the soil profile can be easily modeled.
Soil water content
Estimated annual ETC for a deforested area on the Florida ridge reached 680 mm
(Sumner, 1996). This ET rate was attributed to periods of low soil water content because
the area was not irrigated. Rogers et al. (1983) reported that growth and fruit yield of
citrus trees were greater during a 3-year period for treatments maintained at higher soil
water content. During the same period of time, annual ETC averaged 900 and 1210 mm
for the lowest and highest soil water content treatments, respectively. Hoffman et al.
(1982) reported annual ETC values to be 200 to 500 mm higher than that found by Erie et
al. (1965) in Arizona. The lower annual ETC values reported by Erie et al. were attributed
to infrequent irrigation resulting in dry soil surfaces and thus increased resistance to
water diffusion to the atmosphere. Smajstrla et al. (1986) reported a reduction in growth
and ETC with increased available soil water depletion of 2-year-old Valencia orange
trees grown in drainage lysimeters. Available soil water depletion setpoints used for
irrigation scheduling in this study were 28, 47 and 58%. It was concluded that tree stress

27
occurred at the highest depletion value due to the reduced ability of the soil to transport
water to the roots because of reduction in hydraulic conductivity. Fares and Alva (1999)
calculated daily ETC for 3-year-old Hamlin orange trees on deep sandy soil in central
Florida. Estimated daily ETC values decreased with time after each rainfall or irrigation.
Water table depth
Obreza and Admire (1985) concluded that shallow water tables in flatwoods soils
could significantly augment water available for root uptake. Graser and Allen (1987)
suggested that water-table management by controlling water table depth in the winter and
spring could help decrease the need for supplemental irrigation during the dry season.
Boman (1994) used drainage lysimeters in which he maintained a water table at 0.61,
0.76, and 0.91 m to measure the effects of water table on ETC, growth, yield and fruit
quality of 5-year-old Valencia trees. However, treatment effects were not significant.
Soil shading
Castel et al. (1987) estimated soil surface evaporation by comparing water loss
from weighing lysimeters in which the soil was covered by plastic with lysimeters that
remained uncovered. Mean estimated evaporation was reported as 0.78 mm, greater than
18% of the estimated potential ET of 4.25 mm. Castel and Buj (1992) reported that the
percentage of ground shaded by young Clementine trees increased from 10 to 25% during
a 4-year period. Evapotranspiration increased by 33% during the same time period. This
increase was attributed to the increasing water use by the trees and reduced soil surface
evaporation.

28
Grass and weed growth
Smaj stria et al. (1986) used field drainage lysimeters to determine the effect of
grass cover on the growth and ETC of 2-year-old Valencia orange trees. Automated
covers were installed to cover the lysimeters during rainfall. Soil within the lysimeters
was maintained bare or covered completely with bahiagrass. The bare soil lysimeters
consistently had the lowest monthly ETC. Measured annual ETC ranged between 1331 to
900 mm for grass-covered lysimeters and 912 to 441 mm for those with bare soil
surfaces. Total ETC was 46 to 105% higher per year due to soil grass cover. These results
were similar to those reported by Stewart et al. (1969) using non-weighing lysimeters. In
their study, estimated annual bare soil evaporation and 2/3 sod cover ETC averaged 68
and 92% of full sod cover, respectively. Tucker et al. (1997) reported reduced soil water
use from non-irrigated middles between rows of mature citrus by limiting the height of
weed growth by chemical mowing.
Crop Coefficient
An estimate of evapotranspiration for a specific crop (ETC) is calculated by
multiplying the reference evapotranspiration (ET0) by an empirically determined crop
coefficient (Kc). This coefficient is specific for a crop, growth stage, and growing
conditions. The resulting ETC estimates water use of a crop under local or regional
climatic conditions.
Rogers et al. (1983) reported monthly measured ETC to calculated ET0 ratio values
using the mean of four methods of estimating ET0 (Penman, Blaney-Criddle, Jensen-
Haise, and Class A pan). The resulting monthly ratios range from 0.90 in January to 1.11
in June. Crop coefficient (Kc) values reported by Doorenbos and Pruitt (1977) after

29
adjustments for humid conditions ranged from 0.9 in March though December to 0.95 in
January and February. Castel et al. (1987) estimated monthly Kc for drip-irrigated mature
Navel oranges grown in Valencia, Spain. Their Kc values were calculated from mean
daily ETC estimated from weekly ET values determined by neutron probe measurements.
Values ranged from a mean of 0.71 from January through July to 0.90 from August
through December. Castel and Buj (1992) suggested these values differed from those
reported for Florida due to the lower evaporative demand of the humid Eastern coast of
Spain, which has a mean annual ET0 of 1166 mm compared with 1400 mm in Florida.
Calculated Kc values for 3-year-old Hamlin trees grown on sandy soil in central
Florida ranged from approximately 1.05 in November through March to 0.85 in May
through August (Fares and Alva 1999). Boman (1994) calculated Kc values for 5-year-old
Valencia orange trees grown in non-weighing lysimeters with water tables maintained
at 0.6,0.75, or 0.9 m from the soil surface. Calculated Kc values were at a minimum of
0.6 from December through February and peaked at 1.1 in June and July. Martin et al.
(1997) estimated mean daily ETe values for 7-year-old Redblush grapefruit in Arizona
from soil water content data collected at 1 to 2 week intervals. Monthly Kc values were
calculated by comparing these estimated daily values with mean daily ET0 for the same
period. The resulting K ranged from a low of 0.55 to 0.6 in December and January to a
high of 1.1 to 1.2 in July.
Soil Water Depletion Coefficient
According to Allen et al. (1998), the water depletion coefficient is defined as the
effect of soil water reduction on ETC by reducing the value of Kc. It is calculated by

30
multiplying the of a given crop by the soil water depletion coefficient (Kg) for a given
soil water content. Water stress increases as soil water is extracted by evapotranspiration.
Available soil water (ASW) is defined as the difference between drained upper
limit (field capacity) and drained lower limit or permanent wilting point. However, the
energy expenditure required to extract residual soil water increases as soil water content
decreases. Likewise, resistance to water flow increases as residual soil water decreases,
reducing water flux to the root boundary. Therefore, crop water uptake is reduced well
before wilting point is reached (Allen et al 1998). At field capacity, roots can absorb
water fast enough to supply the ETC demand of the atmosphere. However, water becomes
more strongly bound to the soil matrix and is more difficult to extract as soil water
content decreases. When soil water content drops below a threshold value, water can no
longer be transported quickly enough to the roots to supply the transpiration demand of
the crop. The fraction of ASW above this threshold is known as readily available water
(RAW). For most crops grown on medium and fine textured soils, RAW is as much as 30
to 50% of ASW (Allen et al. 1998). When root zone depletion exceeds this threshold, ET
is reduced relative to potential crop ETC and water stress occurs.
Citrus Nitrogen Uptake
Knowledge of the nutritional need of different plant organs as well as the seasonal
demand for nutrients is essential in order to establish a physiological basis for crop
fertilization (Lagaz and Primo-Millo, 1981). The potential contribution of fertilizer N to
the deterioration of ground water quality may be appreciable (Embleton et al., 1978).
This impact is especially true in Florida where the combination of high annual rainfall,
sandy soils and shallow water tables create conditions that greatly increase the potential

31
for ground water contamination (Alva and Paramasivam, 1999; Calvert and Phung, 1972;
Mansell et al., 1980).
Seasonal Nitrogen Uptake
Most N balance studies have been unable to completely account for total N
applied to the soil. Some authors attributed this fraction (usually 30 to 50%) to
atmospheric loss. Khalaf and Koo (1983) concluded that unaccounted for N was either
incorporated into soil organic matter or stored in the tree (Dasberg, 1987), while others
made no attempt to fully account for the applied N (Mansell et al., 1980).
Hilgeman (1941) estimated N uptake by grapefruit in Arizona by determining
changes in leaf N concentration seasonally. Maximum N uptake by the trees occurred in
March and September relative to January due to higher mean soil temperature. In a 3-year
study, Chapman and Parker (1942) determined N removed from solution culture and
reported that the months of least N absorption were January and February. Uptake rates
increased during the period of late spring through early fall (May to October) with a
maximum in July. Roy and Gardner (1946) in Florida reported similar results.
Numerous reports suggest that actively growing tissues act as a sink for N uptake
and that the young developing leaves and fruit constitute the strongest sink. Legaz et al.
(1982) in Spain studied N distribution in 5-year-old Calamondin trees in sand culture.
Trees were labeled with 15N for 20 days during flowering, were harvested, and analyzed
for N content 0 to 70 days later. Accumulated N was found primarily in fruitlets and
newly developed leaves and twigs. About 30% of the labeled N was found in newly
formed leaves. In Israel, Feigenbaum et al. (1987) treated 22-year old Shamouti orange
trees with 15N labeled fertilizer. Trees had previously been supplied with sufficient N or

32
had been N-depleted to explore the influence of prior fertilization practices on subsequent
N uptake. The highest percentage of labeled N occurred in fruit, new leaves and twigs.
Only about 20% of the leaf and fruit N originated from the labeled source, suggesting
considerable redistribution from stored reserves. Less than 14% of the labeled N was
found in roots or large limbs. Dasberg (1987) found that 80% of the N in new growth
came from stored rather than applied N, suggesting previous nutrition has significant
influence on current season growth and fruit yield. Legaz and Primo-Millo (1988)
reported increased N uptake from the beginning of spring flush to bloom. Uptake
increased through the spring, reaching a maximum at the summer flush after which
uptake declined gradually through winter.
Mooney and Richardson (1992) observed an N concentration gradient between
the roots, trunk and branches of citrus trees in New Zealand. High concentrations were
found in the branches, with lower concentrations in the roots. Nitrogen concentrations in
the trunk were highest at bud break and declined steadily through fruit set and
development to a minimum at fruit harvest. Nitrogen concentration for all categories
peaked at flowering and then decreased steadily until harvest. Nitrogen uptake
efficiencies of 82.0 and 74.1% for ammonium nitrate and urea, respectively, were
reported by Mattos (2000). Legaz et al. (1982) reported 50 to 60% of total tree 15N
recovery in above-ground tree parts. Absorption rates increased only slightly from the
beginning of growth until flowering, and increased sharply reaching a maximum value at
the second growth flush (July) before declining during the fall and winter months.
Dasberg (1987) demonstrated that the highest rate of ,5N uptake by citrus trees occurred
during fruit set and the lowest occurred during winter.

33
Nitrogen Uptake Efficiency
Nitrogen uptake efficiency (NUE) is defined as the percentage of applied N taken
up by plants (Scholberg et al. 2002). The ability of crop plants to take up and utilize N
efficiently is key to providing adequate N for crop growth while reducing N leaching.
Mattos (2000) estimated NUE for 6-year old Valencia trees grown in a sandy soil to be
40% and 26% for ammonium nitrate and urea respectively. Feigenbaum et al. (1987)
reported that the NUE for a 15N labeled KNO3 applied to 22 year-old Shamouti orange
was 40%. Syvertsen and Smith (1996) estimated NUE to be 61% to 83% for 4-year old
grapefruit trees grown in lysimeters. Nitrogen uptake efficiency decreased with increased
N application rates. Lea-Cox and Syvertsen (1996) reported a similar finding of lower
NUE with higher N application rate for greenhouse grown seedlings. The NUE reported
ranged from 47% to 60% after an uptake period of 31 days.
Kato et al. (1982) found a 10-fold increase in 15 N uptake ofSatsuma mandarin
during summer (mean temperature 23 C) compared with the winter season (mean
minimum temperature 3 C). Scholberg et al. (2002) found N uptake of greenhouse-grown
seedlings to be proportional to soil temperature, ET0 and canopy biomass. Nitrogen
uptake also increased with the time high N concentrations were maintained in the root
zone. Increasing the residence time from 2 to 8 hours resulted in an increase in NUE of
95% and 125% for high and low N application rates, respectively.
Seasonal Nitrogen Redistribution
Legaz et al. (1982) suggested that at post-blossom, the N concentration in the
spring leaves decreased due to this tissue becoming an N source for the developing fruit.
Using 4 year-old Valencia orange trees, daily root N uptake was lower during

34
dormancy, increased during flowering and was highest during fruit set, and later
decreased towards the end of the summer and autumn flushes. The greatest accumulation
of N absorbed from fertilizer (with respect to the total N absorbed from fertilizer in the
whole tree) was found in the young leaves and roots, followed primarily by twigs and
stems, then flowers and fruits.
Kato et al (1982) found that total N contents decreased in both bark and wood
during the sprouting period of 21 year-old Satsuma mandarins. Greatest decreases in N
were found in parts with higher concentrations of N (i.e. leaves, shoots, and fine roots). It
was also concluded that the trunk and large roots are main N reservoirs for new shoot
development. The N was reserved mainly as protein, free proline, arginine, and
asparagines. Protein decreased in all plant parts in proportion to total N in the plant part.
Proline decreased mainly in the leaves and bark, arginine in wood of shoots and
asparagines in bark of fine roots.
Crop and Environmental Models
Current Citrus Models
Few predictive models have been developed specifically for use in citrus
production. Most models have been designed for specific applications, with a general
user in mind. These models predict population and/or crop damage caused by citrus
pathogens (Timmer and Zitko, 1996), and scale insects (Aris and Browning, 1995). Other
citrus models are used for irrigation scheduling (Xin et al., 1997), and crop flowering
(Bellows and Morse, 1986; Valiente and Albrigo, 2000).
Environmental Models
Nitrogen leaching from agricultural soils represents both an economic loss to the
farmer and potential groundwater pollution. Mathematical models can be used to assess

35
crop N-fertilizer requirements and to predict effects of N fertilizer management practices
on potential nitrate leaching and how it affects groundwater quality. The understanding of
solute movement and transport has increased in the last 30 years. Increased
environmental concerns pertaining to the runoff and leaching of agricultural chemicals
and fertilizer elements in the surface and groundwater has resulted in development and
use of computer simulation models to predict transport of potential pollutants in
agricultural systems. These models include Nitrate Leaching and Economic Analysis
Package (N-LEAP) (Follett et al., 1994), Groundwater Loading Effects of Agricultural
Management Systems (GLEAMS) (Reck, 1994; Reyes et al., 1994), Drainage-Modified
(DRAINMOD) (Saleh et al, 1994; Verma et al., 1995), Chemicals, Runoff, and Erosion
from Agricultural Management Systems (CREAMS) (Minkara et al., 1995; Saleh et al.,
1994), Leaching Estimation and Chemical Model (LEACHM) (Jemison et al., 1994), and
Nitrogen, Carbon, Soil, Water And Plant (NCSWAP) (Jabro et al., 1993).
These models could be applied to citrus production to predict or estimate the
depth of N leaching below the crop root zone. Most of these models are deterministic,
non-steady state, and comprehensive. They typically require a large number of soil
physical, hydraulic, and chemical characteristics for each soil layer, soil N transformation
components, weather data, and environmental information to determine N fate and
leaching depths. Use of these models for the prediction ofN fate under agricultural
production conditions has met with mixed results (Kiniry et al., 1997). Jabro et al. (1993)
found that neither LEACHM nor NCSWAP successfully predicted nitrate leaching below
1.2 m in a silt loam soil. Jemison et al. (1994) reported accurate predictions using
LEACHM in manure fertilized com crops.

36
Crop Models
Crop-Environment Resource Synthesis (CERES) was developed to model growth and
yield of grain crops (Jones and Kimiry, 1987; Kiniry and Bockhot, 1998; Kiniry et al.,
1997; Lizaso et al., 2001; Saseendran et al., 1998). CROPGRO was initially developed as
a family of crop-specific models for the prediction of legume and vegetable crops
(Hoogenboom et al., 1994; Jones et al., 1991; Wagner-Riddle et al., 1997). These are
process-oriented models for the simulation of vegetative growth and reproductive
development. They predict dry matter growth (Shen et al., 1998), crop development
(Batchelor et al., 1994; Batchelor et al., 1997; Piper et al., 1996) and final yield
(Batchelor et al., 19%; Heinemann et al., 2000) for a range of agronomic crops. Inputs
are daily weather data, soil profile characteristics, and crop management conditions
(Gijsman et al., 2002). Crop and soil water (Hoogenboom et al., 1994; Gabrielle et al.,
1995; and Xie et al., 2001), N (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995;
and Sexton et al., 1998), and C balances are modeled. These models have been combined
into the DSSAT (Decision Support System for Agrotechnology Transfer) software
(Hoogenboom et al., 1994; Jones and Luyten, 1998).
Conclusions
Considerable research and resources have been devoted to improving our
understanding of how cultural, soil, and environmental factors influence biomass and N
accumulation during citrus tree development. However, these studies compared tree
component dry weights and N accumulations with tree age and not a measure of tree size.
Tree size is not only a function of tree age, but soil, environmental, and horticultural
factors as well. Therefore, correlation of dry weights and N accumulations with tree size

37
would provide a better relationship for modeling purposes. Factors affecting citrus root
distribution have been studied under Florida soil and environmental conditions. Many of
these studies were performed in groves with lower tree densities and different irrigation
methods than those currently used in Florida citriculture, and on trees grafted on
rootstocks that are no longer in use. Thus, information on the effect of tree size on root
length density distribution changes for current production systems are lacking. Likewise,
root length density distributions for mature trees on currently used rootstocks grown on
Florida sandy soils have not been determined.
Seasonal maximum daily water uptake rates under Florida environmental
conditions have been determined for trees grown on flatwood soils with fluctuating water
tables. However, maximum daily water uptake rates for mature citrus trees have not been
measured for trees grown on excessively drained Ridge soils. Likewise, reduction in
daily citrus water uptake with decreased soil water potential has not been determined for
sandy soils. Citrus N uptake rates have been determined for seedlings and relatively small
trees grown in lysimeters. These rates may not reflect uptake rates of mature citrus trees
at the field-scale.
Much data on citrus growth, root distribution, water requirements, and N uptake
rates are needed to attain the level of crop modeling currently available for agronomic
crops. Obtaining these data are difficult due to the size of mature citrus trees compared
with agronomic crops, and the inability to follow a cohort of trees from planting to
maturity. Biomass, N accumulation, and spatial root length density changes as affected
by tree size, and water and N uptake dynamics of mature trees under Florida Ridge
conditions will be presented in the following chapters.

CHAPTER 3
CITRUS BIOMASS AND NITROGEN ACCUMULATION
Introduction
Citrus is native to the subtropical and tropical regions of Asia and the Malay
Archipelago (Webber and Batchelor, 1943). Citrons were introduced into Europe via the
Middle East as early as 300 BC, with lemons and sweet oranges following some 15 and
17 centuries later, respectively. Citrus is well adapted to Florida soil and environmental
conditions and proliferated in the costal settlements of Florida by the mid 16th century
and was in commercial production by the mid 1800s.
Nitrogen application rate studies in citrus have emphasized the effects of timing
and amount on increased canopy volume and yield (Sites et al., 1953; Reitz, 1956;
Reuther et al., 1957; and Koo, 1979). However, optimum plant growth depends upon
maintenance of an efficient balance between roots and shoots (Kramer and Boyer, 1995).
Roots are dependent on shoots for carbohydrates, growth regulators, and some organic
compounds, while the shoots of a plant are dependent on the roots for water and
nutrients. The root to shoot ratio varies widely among species, with age, and with
environmental conditions. Understanding the dynamics of root and shoot development
with time is essential when determining biomass and N accumulation, and soil water and
nutrient uptake dynamics.
Caruso et al. (1999) found that the relative proportion of leaves and twigs to total
tree dry weight decreased with tree age. Therefore, relative proportions of total dry matter

39
and N accumulation in different tree components change with age of perennial crops due
to the increase in weight of larger branches and trunks of older trees to support the
increased tree biomass. Similar changes occur in annual crops with increase in biomass
between emergence and harvest. However, as with annual crops, tree size is not
dependent on age alone; rootstock (Castle, 1978, 1980), crop nutrition (Feigenbaum et al.
1987), irrigation (Parsons et al., 2001), and restriction of the root system (Mataa and
Tomingag, 1998) can limit growth of a citrus tree. Thus, these factors can result in trees
of equal age being very different in size, biomass, and N content.
Many crop models such as CERES, CROPGRO, and DSSAT determine the effect
of assimilate costs for vegetative and reproductive growth and N budget through C and N
balances with increase in crop biomass on water and N uptake, growth, and yield of
agronomic crops (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995; and Sexton
et al., 1998). Likewise, optimal irrigation and nutrient management is dependent upon the
estimation of biomass and N content in citrus trees. Therefore, the relationship of tree
size to biomass and N accumulation is needed.
Previous studies have correlated long-term citrus tree biomass and N
accumulation with tree age (Cameron and Appleman, 1935; Cameron and Compton,
1945; Feigenbaum et al., 1987; Kato et al., 1984; Mattos, 2000). Leaves of 3.5, 7, and 15
year-old Valencia trees in California contained from 40 to 50% of total tree N, while
twigs and shoots contained approximately 10% of total N (Cameron and Appleman 1935;
Cameron and Compton, 1945). Trunk and branches contained from 20 to 30% of total
tree N, approximately half of which was in the bark, a tree component that represents
only 5% of the total dry mass. The roots contained from 15 to 20% of tree N, half or

40
more of which was in the bark (5% of the dry mass of the tree). The biomass proportions
for 7 year-old Hamlin orange trees grown under Florida conditions reported by Mattos
(2000) were more similar to the 10 year-old trees cited above than the 3.5 year-old trees
(Table 1). Nitrogen concentration was lowest in the trunk and taproot of these trees. The
N concentration of leaves (2.1 to 2.6 %), twigs (0.4 to 0.8 %), and roots (0.6 to 1.7 %)
varied with tissue age. Younger tissue tended to have greater N concentration compared
with older tissues. Kato et al. (1984) and Feigenbaum et al. (1987) harvested older citrus
trees (21 and 22 years old, respectively). Leaves comprised a smaller fraction of total
biomass in both studies compared with branches and total roots. The leaves of these older
trees contained a lower proportion of total tree N than the branches, equal to the
proportion of N in the roots (Table 3-1). None of the above studies related biomass or N
measurements to tree size parameters such as canopy volume or trunk cross-section area.
Biomass and N distribution relationships based on tree size measurements as
opposed to tree age could provide more generic information needed for modeling tree
growth and N cycling in citrus production systems. Therefore, the hypotheses to be tested
in the following studies were. 1) functional relationships can be defined that correlate
biomass and N partitioning of specific tissue categories with tree size using generic
growth indicators such as canopy volume or trunk area, and 2) rootstock has a significant
effect on citrus biomass and N partitioning. Such relationships can be used to determine
citrus N budgets and develop specific fertilizer recommendations that will provide
adequate N for growth and production while protecting groundwater from nitrate
contamination. Thus, a non-destructive method of estimating an N budget is needed for
trees of unknown or mixed ages. Therefore, the objectives of the following studies were

Table 3-1. Citrus biomass and nitrogen distribution by tree age as reported by different
studies.
41
Authors
Trees
(n)
Location
Cultivar
Age
(Yrs)
Plant
Tissue
Biomass
(% Total)
N
(% Total)
Cameron and
15
California
Val.
3.5
Leaves
30.5
61.9
Appleman
Branches
38.4
21.0
(1935)
Roots
31.1
17.1
Legaz and
8
Spain
Val.
4
Leaves
22.5
30.0
Primo-Millo
Branches
28.7
18.5
(1988)
Lateral roots
45.8
41.2
Fibrous roots
2.9
10.3
Mattos
6
Florida
Ham.
6
Leaves
13.9
35.0
(2000)
Branches
46.5
28.£
Lateral roots
25.7
13.4
Fibrous roots
14.1
23.2-
Cameron and
4
California
Val.
10
Leaves
18.5
46.7
Appleman
Branches
60.7
39.0
(1935)
Roots
20.7
14.2
Cameron and
36
California
Val.
15
Leaves
16.8
45.3 -
Compton
Branches
61.4
34.8
(1945)
Lateral roots
20.4
17.0
Fibrous roots
1.7
2.9
Katoetal.
1
Japan
Sat.
21
Leaves
8.6
27.2
(1984)
Branches
65.1
44.6
Lateral roots
19.5
14.3
Fibrous roots
6.8
14.0
Feigenbaum
2
Israel
Sham
22
Leaves
7.3
24.6
etal. (1987)
Branches
61.0
49.8
Lateral roots
26.5
19.2
Fibrous roots
4.3
3.8
to determine: 1) changes in biomass and N distribution with change in tree size, 2) yearly
changes in biomass and N content of mature citrus trees, and 3) rootstock effect on
mature citrus tree biomass and N distribution. The relationships of canopy volume and
mean trunk diameter to biomass and N content for citrus will form the basis of a
predictive model to estimate the biomass and N distribution based on size measurements.

42
Materials and Methods
Citrus trees of various sizes were measured and dissected into constituent parts
during a period of 1 year. Representative tissue samples of constituent parts for each tree
were weighed and analyzed to estimate total dry mass, relative percentage dry mass, and
N content of the various tree components. These data were used to determine N
allocation between different tree constituents.
Experiment 1 Mature Citrus Biomass and N Distribution
Two sets of six trees each were dissected in February 2001 and January 2002 at
the Water Conserv II site near Winter Garden in western Orange county, Florida. Both
sets of trees were 14 year-old Hamlin orange trees planted in 1987 at a spacing of 3 m
between trees in the row and 6 m between rows resulting in a tree density of 556 trees ha
V Three trees of each set were budded on Swingle citrumelo (iCitrus parodist Macf. x
Poncirus. trjfoliata (L.) Raf) rootstock, and the remaining three trees of each set on
Carrizo citrange (G sinensis L. Osbeck X P. trifoliata L. Raf.) rootstock. All trees had
been fertigated at an annual rate of240 kg N ha1. The trees were irrigated (and
fertigated) with reclaimed water containing approximately 7 mg NO3-N L'1.
Experiment 2 Biomass and N Accumulation with Increase Tree Size
A third set of seven Valencia trees on Swingle citrumelo rootstock were
dissected at the K.D. Revell grove owned by Cargill, Inc. near Fort Meade in southern
Polk county, Florida. Fresh, dry, and N weights were determined for the same constituent
parts as in experiment 1. These trees had been fertilized using dry chemical fertilizer
three or more times per year and irrigated with well water.

43
Site Descriptions
The soil series at the Orange county site was Candler fine sand (hyperthermic,
uncoated, Typic Quartzipsamment), and at the Polk county site was Zolfo fine sand
(sandy siliceous, hyperthermic Grossarenic Entic Haplohumod). These two soils are
typical of the central Florida ridge and have a field capacity water content of 0.06 to 0.08
cm cm'3 in the upper 1 m. The Candler series consists of excessively drained, very
rapidly permeable soils formed from marine deposits. These soils are located in upland
areas and typically have slopes of 0-12%. The A and E horizons consist of single-grained
fine sand, have a loose texture, and are strongly acidic. A Bt horizon is located at a soil
depth of 2 m and includes loamy lamellae of 0.1 to 3.5 cm thick and 5 to 15 cm long.
Zolfo series soils are sandy and slightly less well drained than those of Candler. The A
horizon consists of fine sand with single-grained, loose texture. The Bh horizon between
4.0 and 5.0 cm consists of fine sand coated with organic matter possessing weak granular
to weak fine subangular blocky structure.
Tree Canopy Volume and Trunk Cross-Section Area
Changes in canopy volume have been used in fertilizer rate experiments as
measures of tree growth (Whitney et al., 1991). Therefore, tree measurements were
determined for the purpose of correlating biomass and N concentration of various tree
components to tree size. Canopy diameter of each tree was measured in the row (in-row)
and across the row (cross-row) at a height of 1.5 m above ground level. Tree height and
hedgerow intercept measurements were made using a 5 m graduated pole. Hedgerow
intercept is the height from the ground to the point at which the canopies of two trees
meet in the row. These measurements have been used by Whitney et al. (1991) to
determine canopy volume based on a spheroid model (Equation 3-1). Trunk diameters 5

44
cm above the ground were determined for each tree by measuring in-row and cross-row
dimensions. Trunk cross-sectional areas (TCSA) were determined for each tree assuming
an oval shape.
TCV = Ir Cr Ht *
4
o-o-O
Equation 3-1
Where:
TCV = Tree canopy volume (m3)
Ir = In-row spacing (m)
Cr = Cross-row spacing (m)
Ht = Canopy height (m)
Int = Canopy intercept height (m)
Tree Biomass Fresh Weight
Fresh weight of the leaves, twigs <7 mm, small branches 7 to 15 mm, medium
branches 15 to 30 mm, large branches >30 mm, trunk, tap root, small roots <4 mm,
medium roots 4 to 20 mm, and large roots >20 mm were measured in the field. Field
weights and three samples of each plant part category were collected for each tree using
the following protocol: Twigs less than 7 mm in diameter and attached leaves were cut
from the tree with leaves intact. These twigs were placed into a plastic container and
weighed on a battery powered field-portable balance. During cutting, one twig out of 20
was placed into a separate container as they were cut and were weighed separately.
Leaves were removed from the twigs in this container while still in the field. Fresh
weights of these subsamples were measured. Branches 7 mm in diameter and greater
were cut into 15 to 30 cm segments, separated into the three size ranges noted above, and

45
weighed separately. Two to three samples equal to 5% of the fresh weight of each size
category were removed from each container and placed into labeled plastic bags. Any
leaves attached to these branches were removed and weighed prior to weighing the
branch segments. The trunk and taproot were cut into pieces and weighed, and three
longitudinal slices of each were retained as separate samples.
The trees were planted 3.1 m in row and 6.2 m between rows. Therefore, the roots
were cut to a depth of approximately 0.3 m using a shovel in a rectangle 3.1m in-row x
6.2 m across-row with the tree stump at its center. The bulk of the root system was
extracted using a front-end loader equipped with a root rake. All roots were removed
from the excavation to a depth of 1 m, washed, blotted dry, separated into size categories,
and weighed in the field.
Sample Processing and Nitrogen Analysis
The leaf area of 50 random leaves from each sample was measured. Each branch
segment of each sample was cut into at least five disks of approximately 0.5 to 1 cm thick
that facilitated the removal of bark from the wood. Likewise, the bark was removed from
each horizontal trunk slice. The bark and wood from the branch and trunk disks were
weighed separately to determine the fresh mass proportion of bark to wood for each size >'
category.
Samples were dried at 70 C to a constant weight before recording dry weight.
Total tissue dry weight for each tree was determined by multiplying fresh weights by the
respective tissue dry matter content. All tissues were ground using a Cyclotec mill (1093
Sample Mill, Tecator manufacturing, Sweden) for the leaf tissue and Wiley mill model 1
(Arthur Thomas Manufacturing Co., Philadelphia, Pa) for woody tissue. The ground
tissues were digested using a Buchi Model K435 12-vessel digestion unit (Buchi

46
Analytical, Inc., New Castle, DE). The digest was analyzed for total Kjeldahl nitrogen
using USEPA method 351.2 using a Buchi model B339 steam distillation instrument
(Buchi Analytical, Inc., New Castle, DE.).
Leaf Area, Biomass and N Weight Estimation
Specific leaf area (cm2 g'1) were determined for a 50-leaf subsample by dividing
total leaf area by leaf dry weight. Total leaf area was estimated by multiplying the mean
specific leaf area by the estimated total dry leaf weight for the corresponding tree. Leaf
area index was determined for each tree by dividing the leaf area by the corresponding
cross sectional canopy area using the in-row and cross-row measurements. Leaf area
index was also determined on a per acre basis by dividing the total leaf area of each tree
by the land area occupied by the tree (in-row spacing x cross-row distance). Dry weights
for each tissue type of individual tree were estimated by multiplying the total field fresh
mass by the mean percentage dry mass of three sub samples for each tissue category.
Likewise, N accumulation within each tissue type of each tree was estimated by
multiplying the total dry weight by the mean N concentration of the three sub samples of
each tissue. Total tree dry weight and N accumulation were determined by summing
across tissues categories. The above ground dry weight and N accumulations were
determined by summing the estimated values for leaf, twig, total branch, and trunk
components. Likewise, the below ground biomass and N accumulation was the sum of
total root and taproot values. Percentages of total biomass and N weight were determined
for each tissue.
Prior to dissecting the second set of mature Hamlin trees in 2002, all fruit was
removed and weighed. Representative fruit samples were collected and dried to

47
determine percentage dry weight and analyzed for N concentration to determine total fruit
N accumulation. Leaves and twigs collected in 2002 were separated into current years
growth and prior years growth. Separate samples were collected for dry weight and N
concentration determination.
Statistical Analysis
Since the samples were taken during a 14-month period, the samples from the
mature (14 year old) Hamlin trees were treated as repeated measures and analyzed
accordingly using the SAS general linear models (GLM) procedure (SAS Institute, Cary,
NC). Non-linear regression analysis of tissue masses and percentages of all trees were
made considering canopy volumes and mean trunk diameters as independent variables.
Results
Mature Citrus Tree Biomass Distribution Experiment 1
Leaf area was significantly different (P=0.05) for TCV and TCSA, but not for
rootstocks (Table 3-2). However, neither mean leaf area index on a per tree basis (10.2
and 9.9 for Carrizo and Swingle trees, respectively) nor on a per acre basis (6.4 and 6.2
for Carrizo and Swingle trees, respectively) were significantly different for tree size or
rootstock. Total above-ground and below-ground weights increased on both a fresh and
dry basis across the range of TCV and TCSA encountered in this study. Maximum total
fresh weight was approximately 160 kg tree'1 with TCVs ranging from 28 to 38 m3 and
TCSAs of 80 to 160 cm2. Maximum dry biomass for the same canopy volumes and trunk
cross-sectional areas was approximately 100 kg tree*1. Maximum above-ground biomass
(leaves, twigs, branches, and trunk) and below-ground (roots and taproot) was
approximately 74 and 26 kg tree'1, respectively.

48
Table 3-2. Mature citrus tree leaf area and leaf area index as a function of TCV and
TCSA
Leaf Area
Leaf Area Index
(m2 tree"1)
(m2 leaf m"2
canopy)
Carrizo
Swingle
Carrizo
Swingle
Mean
110.0
103.4
10.2
9.9
Standard Deviation
10.3
24.8
1.2
1.8
Statistical Significance2
TCVy
*
*
NS
NS
TCSA
*
*
NS
NS
Rootstock
NS
NS
NS
NS
*NS = not significant, *= significant P<0.05, and **= significant P<0.01
yTCV = tree canopy volume, TCSA = tree cross-sectional area.
Mature trees on Carrizo rootstocks had significantly greater (P=0.05) mean dry
weight (100 kg tree'1) than those on Swingle (83 kg tree1), whereas, below-ground
biomass was not significantly different (Table 3-3). Significantly higher (P=0.01)
percentage of large branch biomass was found for trees grown on Carrizo citrange (23.8
kg tree"1) compared with trees grown on Swingle citrange (15.8 kg tree"1). Thus,
percentage total branch biomass was significantly greater (P=0.05) for the Carrizo

Table 3-3. Dry matter accumulation and allocation between tree components for mature Hamlin orange trees as affected by year of sampling,
rootstock, and interaction of year and rootstock.
Dry Weight Accumulation Dry Weight Allocation
Total
Mass
Above
Ground
Below
Ground
Total
Leaf
Total
Twigs
Sm.
Branches
Med. Lrg.
Total
Total
Trunk
Sm.
Roots
Med. Lrg.
Tap
(kg dw tree'1)
(% of total weight)
Year 1
97.6
69.2
28.3
12.7
5.8
17.0
12.1
19.1
48.2
3.6
5.1
4.9
9.5
10.1
Year 2
87.8
64.9
26.3
12.4
7.3
15.3
9.0
21.1
45.4
3.9
6.1
6.8
6.4
9.2
NS*
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
**
NS
NS
Carrizo
100.3
77.0
26.6
12.2
7.1
16.0
10.5
23.8
50.3
3.9
5.5
6.1
6.9
6.7
Swingle
82.6
54.7
27.9
12.9
6.0
16.1
10.4
15.8
42.3
3.7
5.8
5.7
8.8
13.1
NS
*
NS
NS
NS
NS
NS
**
*
NS
NS
NS
NS
*
Car Yrl
104.2
79.2
25.0
12.9
6.8
15.7
11.2
25.4
52.3
3.9
4.9
5.6
8.5
5.1
Car Yr2
96.3
74.8
28.2
11.6
7.5
16.4
9.7
22.2
48.3
3.9
6.2
6.6
5.4
8.3
Swi Yrl
87.6
54.3
33.3
12.4
4.3
18.9
13.5
9.5
41.9
3.2
5.6
3.8
11.1
17.7
Swi Yr2
79.4
55.0
24.4
13.3
7.1
14.2
8.3
20.0
42.5
3.9
6.0
6.9
7.3
10.1
NS
NS
NS
NS
NS
NS
NS
**
NS
NS
NS
NS
NS
NS
*NS = not significant, *= significant P<0.05, and **= significant P<0.01.
VO

50
citrange roostock trees. On a percentage basis, the taproot biomass was significantly
greater (P=0.05) for trees on Swingle than those on Carrizo.
Although tree weights appeared to be less in 2002 compared with 2001, total dry
weight and dry weights of tree components were not significantly different at the P=0.05
level (Table 3-3). Leaf weights represented 12 to 14% of total tree dry weight, while
total branch weights (twigs, total branches, and trunk) accounted for 49 to 63% of total
tree weight. Dry matter allocation to tree roots was highly variable in all three studies, but
averaged in the 19 to 21% range.
Total above-ground N accumulation was significantly affected by TCV and TXA
at the P=0.05 level (Table 3-4). Leaf, branch, and root N comprised approximately 45,
i
35, and 20% of total tree N, respectively. Dry weight allocation to leaves, twigs, small
branches, medium branches, and trunk were similar for both rootstocks. However,
Carrizo trees had almost 50% more N in large branches compared with Swingle. Mean
total N accumulation by large roots and taproot was greater for trees on Swingle
compared with those on Carrizo. Nitrogen concentrations were not significantly different
(P=0.05) within tissues on each rootstock (data not shown), thus differences in percentage
of total N mass of each tissue category were due to differences in biomass.
Biomass Changes with Increase in Tree Size Experiments 1 and 2
Data from experiments 1 and 2 were combined to determine relationships
between dry weight and N accumulation and tree size indices such as TCV or TCSA.
TCV increased linearly as TCSA increased (Fig. 3-1). Likewise, total leaf area per tree
was linearly proportional to both TCV (Fig. 3-2A) and TCSA (Fig. 3-2B). Leaf area

Table 3-4. Nitrogen accumulation and allocation between tree components for mature Hamlin orange trees as affected by year of sampling,
rootstock, and interaction of year and rootstock.
Nitrogen Accumulation Nitrogen Allocation
Total Above Below Total Total Branches Total Roots
Mass Ground Ground Leaves Twigs Sm. Med. Lrg. Total Trunk Sm. Med. Lrg. Tap
(kg N tree'1) (% of total weight)
Year 1
0.81
0.60
0.21
38.3
8.2
10.C
6.0
9.0
24.9
2.01
10.1
6.1
5.5
4.8
Year 2
0.77
0.57
0.20
36.5
8.3
8.6
4.8
10.8
24.2
2.3
10.1
7.0
3.6
4.9
NSZ
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
**
NS
Carrizo
0.87
0.65
0.22
36.4
8.5
9.2
5.4
11.5
26.1
2.3
10.1
7.4
4.2
3.5
Swingle
0.69
0.50
0.19
38.4
8.0
9.2
5.3
8.2
22.6
2.1
10.1
5.6
4.8
6.4
NS
NS
NS
NS
NS
NS
NS
*
NS
NS
NS
*
NS
NS
Car Yrl
0.87
0.66
0.21
38.1
8.8
9.4
5.6
11.6
26.6
2.2
9.4
7.3
4.9
2.7
Car Yr2
0.86
0.64
0.22
34.7
8.1
9.1
5.1
11.4
25.6
2.5
10.7
7.6
3.4
4.4
Swi Yrl
0.71
0.50
0.21
38.6
7.4
10.9
6.5
5.0
22.3
2.0
11.1
4.4
6.4
7.9
Swi Yr2
0.68
0.51
0.17
38.2
8.4
8.0
4.4
10.3
22.7
2.2
9.5
6.4
3.7
5.5
NS
NS
NS
NS
NS
NS
NS
*
NS
NS
NS
NS
NS
NS
*NS = not significant, = significant P<0.05, and ** = significant P<0.0

52
Fig. 3-1. Tree canopy volume as a function of trunk cross sectional area for trees
from experiments 1 and 2.

53
Fig. 3-2. Leaf area expressed as a function of tree canopy volume (A) and trunk cross
section area (B).

54
index on a tree basis (LAIt) increased rapidly from 4 to 10 as TCV increased from 2 to 10
m3 (Fig. 3-3 A) and TCSA increased from 20 to 80 cm2 (Fig. 3-3B). Little increase in
LAIt was observed with increasing TCV and TCSA beyond 10 m3 and 80 cm2,
respectively. Likewise, leaf area index on an acre basis (LAIa) increased from 1 to 6.3 for
the same ranges of TCV and TCSA
Citrus dry weight accumulation for all tree components increased linearly with
increased TCV and TCSA (Fig. 3-4 and Fig. 3-5). Regression coefficients, r2, and root
i
mean square error (RMSE) values for dry weights of all tissues vs. TCV and TCSA are
provided in Tables 3-5 and 3-6, respectively. Coefficients of determination (r2) were
generally higher for each tissue category when compared with TCV than with TCSA.
Biomass weights for twig, trunk and root categories varied greatly, resulting in lower r2
and higher RSME values. The medium branch masses varied more than the small or large
branch categories, possibly indicating inconsistent and/or incomplete separation of tree
components into appropriate diameter ranges. Correlations of trunk and taproot weights
with TCV and TCSA were poor compared with those for other tree components.
Variation in dry matter allocation to root and tap root were apparently due to differences
in root density distribution of the two rootstocks used in this study.
Dry matter accumulation in above-ground biomass increased from 60 to 75%
across a range of 5 to 40 m3 and 10 to 160 cm2 for TCV (Fig. 3-6A) and TCSA (Fig. 3-
7A), respectively. Leaf biomass declined from approximately 20% of total biomass for
trees with TCV of less than 5 m3 and for TCSA values below 20 cm2, to approximately
12% of total biomass for trees with TCV and TCSA values greater than 30 m3 and 160

Leaf Area Index Leaf Area Index
55
Canopy Volume (m3)
0 20 40 60 80 100 120 140 160 180
Trunk Cross Section Area (cm2)
Fig. 3-3. Leaf area index on a tree basis expressed as a function of tree canopy volume
(A)and trunk cross-section area (B).

56
Fig. 3-4. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) accumulation
as a function of tree canopy volume for HamlinVCarrizo- experiment 1
(0),Hamlin7Swingle(D),and experiment 2 (A) trees.

57
Trunk Cross Section Area (cm2)
Fig. 3-5. Total (closed), above-ground (open), and below-ground (gray) biomass (A); leaf
(closed) and twig (open) biomass (B); total branch (closed) and trunk (open)
biomass (C); and total root (closed) and taproot (open) biomass (D) accumulation
as a function of trunk cross section area for Hamlin/Carrizo Experiment 1 (O),
Hamlin/Swingle Experiment 1 (), and Experiment 2 (A) trees.

58
Table 3-5. Linear regression analysis of dry weight and N accumulation in different tree
components as related to tree canopy volume (TCV)Z
Yo
a
R2
RMSEy
P
Dry Weight (kg tree'1)
Total Mass
-2.21
2.79
0.97
0.15
<0.0001
Above Ground
-2.09
1.95
0.96
0.17
<0.0001
Below Ground
0.21
0.72
0.91
0.27
<0.0001
Leaves
0.09
0.36
0.95
0.20
<0.0001
Twigs
0.26
0.13
0.76
0.49
<0.0001
Sm. Branches
-0.57
0.37
0.90
0.29
<0.0001
Med. Branches
-0.31
0.23
0.80
0.43
<0.0001
Lg. Branches
-.24
0.36
0.82
0.53
<0.0001
Total Branches
-2.47
1.25
0.92
0.27
<0.0001
Trunk
0.54
0.07
0.72
0.36
<0.0001
Sm. Roots
0.41
0.13
0.87A
0.27
<0.0001
Med. Roots
0.32
0.14
0.83
0.33
<0.0001
Lg. Roots
-0.24
0.18
0.85
0.37
<0.0001
Tap Root
0.06
0.12
0.60
0.57
<0.0001
Nitrogen Weight (g tree'1)
Total Mass
-3.24
23.41
O.97)
0.16
<0.0001
Above Ground
-7.28
17.52
0.96
0.17
<0.0001
Below round
5.51
5.36
0.92
0.23
<0.0001
Leaves
3.50
8.70
0.92
0.24
<0.0001
Twigs
0.09
1.83
0.85
0.34
<0.0001
Sm. Branches
-3.32
2.04
0.96
0.20
<0.0001
Med. Branches
-1.28
1.00
0.78
0.45
<0.0001
Lg. Branches
-3.46
2.44
0.88
0.32
<0.0001
Total Branches
-11.31
5.71
0.94
0.23
<0.0001
Trunk
2.80
0.35
0.65
0.40
<0.0001
Sm. Roots
5.38
1.85
0.86
0.28
<0.0001
Med. Roots
1.82
1.34
0.88
0.29
<0.0001
Lg. Roots
-1.10
0.88
0.89
0.32
<0.0001
Tap Root
0.68
0.55
0.63
0.53
<0.0001
z Y = Yo +aX where X = TCSA, and Yo and a are regression coefficients
y RMSE dry weight kg dw tree'1, and N accumulation = g N tree1

59
Table 3-6. Linear regression analysis of dry weight and N accumulation in different tree
components as relatad to trunk cross-sectional area (TCSA)Z
Yo
a
R2
RMSEy
P
Dry Weight (kg tree1)
Total Mass
-7.23
0.70
0.93
0.23
<0.0001
Above Ground
-5.67
0.50
0.93
0.24
<0.0001
Below Ground
-1.20
0.19
0.89
0.29
<0.0001
Leaves
-0.65
0.09
0.94
0.22
<0.0001
Twigs
-0.04
0.04
0.68
0.47
<0.0001
Sm. Branches
-1.13
0.09
0.82
0.40
<0.0001
Med. Branches
-0.84
0.06
0.80
0.43
<0.0001
Lg. Branches
-.94
0.05
0.78
0.42
<0.0001
Total Branches
-5.17
0.33
0.91
0.30
<0.0001
Trunk
0.38
0.02
0.70
0.36
<0.0001
Sm. Roots
0.10
0.04
0.90
0.24
<0.0001
Med. Roots
-0.08
0.04
0.91
0.25
<0.0001
Lg. Roots
-0.53
0.04
0.78
0.45
<0.0001
Tap Root
-0.19
0.03
0.61
0.57
<0.0001
Nitrogen Weight (g tree'1)
Total Mass
-47.83
6.00
0.94
0.22
<0.0001
Above Ground
-41.41
4.51
0.94
0.23
<0.0001
Below round
-5.36
1.41
0.92
0.24
<0.0001
Leaves
-15.39
2.32
0.93
0.23
<0.0001
Twigs
-3.29
0.46
0.82
0.37
<0.0001
Sm. Branches
-6.75
0.49
0.89
0.31
<0.0001
Med. Branches
-3.52
0.27
0.77
0.47
<0.0001
Lg. Branches
-5.69
0.47
0.85
0.38
<0.0001
Total Branches
-23.36
1.50
0.92
0.27
<0.0001
Trunk
1.97
0.10
0.82
0.39
<0.0001
Sm. Roots
0.92
0.51
0.89
0.25
<0.0001
Med. Roots
-1.48
0.38
0.91
0.25
<0.0001
Lg. Roots
-2.65
0.21
0.82
0.40
<0.0001
Tap Root
z
-0.23
0.13
0.58
0.56
0.0002
z y = yo +ax where x = TXA, and yo and a are regression coefficients
y RMSE dry weight kg dw tree'1, and N weight = g N tree'1

60
cm2, respectively (Figs. 3-6B and 3-7B). Likewise, twig biomass decreased from 11 to
6% of total biomass for trees of the corresponding size categories. Total branch dry
weight increased from 15 to 45% as trees matured, while trunk biomass decreased from
12% for young trees to 3% for mature trees (Figs. 3-6C and 3-7C). Few consistent trends
were found when comparing root biomass with tree size (Figs. 3-6D and 3-7D), which
may be due to problems in recovering all root tissue or differences in root biomass
distribution between the two rootstocks used in this study.
Nitrogen Distribution
Mature Citrus Tree Nitrogen Distribution Experiment 1
As was the case with dry biomass, total N accumulation was greater for trees on
Carrizo rootstock (870 g tree'1) than for trees on Swingle (690 g tree'1). Total and above
ground N accumluation were significantly (P=0.01) affected by TCV and TCSA, whereas
below-ground N accumulation was not (Table 3-4). Rootstock, year, and the interaction
of rootstock and year effects were not correlated with TCV or TCSA. Nitrogen
concentration was not significantly different compared with TCV, TCSA, rootstock, or
year for any of the tissues sampled (Table 3-7). Therefore, N content trends were similar
to those for dry weight. Percentage tissue N concentration compared with total tree N
weight was not significantly different for tree size, rootstock, or year with the exception
of branches greater than 30 mm in diameter. Trees grown on Carrizo citrange had
significantly greater (P=0.05) N in larger branches and total branches compared with
TCV (Table 3-4).

61
Canopy Volume (m3)
Fig. 3-6. Dry weight allocation to above-ground (closed), and below-ground (open)
biomass (A); leaf(closed) and twig (open) biomass (B); total branch (closed) and
trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D)
as a function of canopy volume for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.

62
Trunk Cross Section Area (cm2)
Fig. 3-7. Dry weight accumulation to above-ground (closed), and below-ground (open)
biomass (A); leaf (closed) and twig (open) biomass (B); total branch (closed) and
trunk (open) biomass (C); and total root (closed) and taproot (open) biomass (D)
as a function of trunk cross section area for HamlinVCarrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.

63
Table 3-7. Mature citrus tree tissue N concentration as a function of year sampled and
rootstock.
Plant Tissue
Year Sampled
Rootstock
2001
2002
Carrizo
Swingle
(%)
Leaves
2.47
2.32
2.47
2.46
Twigs
1.28
1.00
1.09
1.14
Small Branches Bark
1.16
1.24
1.22
1.11
Small Branches Wood
0.30
0.31
0.33
0.28
Med. Branches Bark
1.11
1.28
1.13
1.08
Med. Branches Wood
0.29
0.32
0.30
0.28
Large Branches Bark
1.16
1.33
1.06
1.30
Large Branches Wood
0.38
0.34
0.42
0.33
Trunk Bark
1.29
1.36
1.33
1.14
Trunk Wood
0.42
0.44
0.43
0.42
Fibrous Roots
1.55
1.38
1.61
1.51
Medium Roots
1.04
0.85
1.10
0.97
Large Roots
0.48
0.47
0.50
0.47
Tap Root 0.45
Statistical Significance2
Year NS
Rootstock NS
Year X Rootstock NS
0.45
0.47
0.39
z NA = No significant difference at the P=0.1 level.
Nitrogen Balance
Leaf and twig dry weight accumulation during the previous 12 months was significantly
greater (P=0.05) for trees grown on Carrizo (3963 and 5717 g tree1, respectively)
compared with trees grown on Swingle (3312 and 4525 g tree*1, respectively). Total N
content for these tissues were 145 and 127 g tree'1 for Carrizo and Swingle, respectively.
Mean fruit N accumulations for the two rootstocks were 302 and 258 g tree1 for trees on
Carrizo and Swingle, respectively. Assuming a conservative 5% increase in N
accumulation in all tissues other than leaves and twigs due to increase in biomass in
2001, the resulting increase in N content was 24 and 18 g tree1 for Carrizo and Swingle

64
trees, respectively. Therefore, the total estimated increases in N content for 2001 were
471 and 403 g tree1 for trees grown on Carrizo and Swingle rootstocks, respectively. The
amount of fertilizer N applied in 2001 was approximately 503 g tree'1, resulting in
apparent fertilizer N uptake efficiencies (FNUE) of 93.6 and 80.1% for Carrizo and
Swingle rootstocks, respectively. However, if the 58 g tree1 of N contained in the 1273
mm ha'1 of reclaimed water applied over the 12-month period is considered, NUE
decreases to 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively.
Nitrogen Change with Increase in Tree Size Experiments 1 and 2
Nitrogen accumulation increased linearly with increasing TCV and TCSA from 5
to 40 m3 and 20 to 160 cm2, respectively (Figs. 3-8A and 3-9A). Total leaf N mass
increased from less than 30 to more than 250 g N tree1 across the range of TCV and
TCSA measured (Fig. 3-8B and 3-9B). These increases were about 45% of total tree N
for trees with TCVs less than 5 m3 to 37% for trees with TCVs greater than 35 m3
(Figs.3-10B and 3-1 IB). Twig N ranged from less than 10 to greater than 50 g tree1
across the range of trees measured (Figs. 3-8B and 3-9B), but twigs still contained a
consistent 9% of total tree N regardless of tree size (Figs. 3-10B and 3-1 IB). Total N
accumulation by branches increased from less than 10 to greater than 200 g N tree1 for
trees with TCV of less than 5 and greater than 40 m3, respectively (figs. 3-8C and 3-9C),
which corresponds to an increase in percentage of total tree N from 6 to 27% for
corresponding tree sizes (Fig. 3-10C and 3-11C). The proportion of total tree N in the
trunk decreased from 5 to 3% (Figs. 3-10C and 3-11C). Regression coefficients, R2,
RSME, and probability values for dry matter and N accumulation within each tissue type
are presented in Tables 3-5 and 3-6.

65
Canopy Volume (m3)
Fig. 3-8. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and trunk (open)
N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of canopy volume for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.

66
O 20 40 60 80 100 120 140 160 180
Trunk Cross Section Area (cm2)
Fig. 3-9. Total (closed), above-ground (open), and below-ground (gray) N weight (A);
leaf (closed) and twig (open) N weight (B); total branch (closed) and trunk (open)
N weight (C); and total root (closed) and taproot (open) N accumulation (D) as a
function of trunk cross section area for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.

67
Canopy Volume (m3)
Fig. 3-10. Nitrogen allocation to above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (closed) and taproot (open) (D) as a
function of canopy volume for HamlinVCarrizo experiment 1 (O),
HamlinVSwingle experiment 1 (), and experiment 2 (A) trees.

68
Trunk Cross Section Area (cm2)
Fig. 3-11. Nitrogen allocation of above-ground (closed), and below-ground (open) N
weight (A); leaf (closed) and twig (open) N weight (B); total branch (closed) and
trunk (open) N weight (C); and total root (closed) and taproot (open) (D) as a
function of trunk cross section area for Hamlin/Carrizo experiment 1 (O),
Hamlin/Swingle experiment 1 (), and experiment 2 (A) trees.

Discussion
The mature Hamlin trees used in this study were planted at the same time and
received the same horticultural inputs (i.e. fertilizer rates, irrigation schedule, and pest
control) for the past 14 years. Total leaf area of each tree differed with tree size, but LAIt
and LAIa appeared to approach maximum values of 10 and 6.5, respectively. The mean
LAIt value of 10 is well within the 9 to 11 range found by Syvertsen and Lloyd (1994) for
mature citrus trees. These values are much higher that the 3 or less associated with
agronomic row oops (Flenet et al., 1996). Citrus is thought to have developed as an
understory plant in subtropical rainforests and so has a high tolerance to shade (Syvertsen
and Lloyd, 1994) and therefore developed a dense canopy. A large fraction of citrus
leaves found within the inner canopy receive 10% or less of recorded on the outermost
leaves (Cohen et al., 1987) The observed LAI would indicate that, on average, 10 layers
of leaves exist over each unit area of soil under the tree canopy. This finding has
significant implication for light interception and photosynthesis, indicating that citrus is
an efficient interceptor of light and allows very little of it to strike the soil surface. Leaves
in the interior of a citrus tree are adapted for low light levels and tend to be thinner and
flatter than exterior leaves (Mills et al., 1999).
Percentages of total biomass and N in leaf, branch and root tissue compared well
with 10 and 15 year old trees harvested by Cameron and Appleman (1935) and Cameron
and Compton (1945). However, trees grown on Swingle rootstock were significantly
smaller than those grown on Carrizo rootstock. Even though tree size was different, dry
weight allocation between tree components remained relatively constant. However, there
were significant differences in dry weight accumulation in large branch and total branch

70
biomass weights for the two rootstocks in this study. Trees grown on Carrizo citrange
rootstocks were larger than those of the same age grown on Swingle citrumelo, therefore
the difference in large and total branch weights were correlated with tree size as
measured by TCV and TCSA. Hence, the percentages of total biomass for specific tree
components were similar for both rootstocks, indicating that above ground biomass is
partitioned equally based on the relationship of total biomass to tree size. Therefore, if the
r ^
relationship of total biomass to tree size is known, and the relationship of component
i ...
partitioning to tree size is also known, then the biomass of each component part can be
estimated regardless of rootstock effects on tree size.
The biomass associated with individual tree trunks was related to the height of the
crotch formed by the main scaffold limbs of the tree. In citrus nurseries, crotch height is
relatively uniform. However, the larger trees used in this study were affected by the
freezes of the late 1980s, most notably in 1989. The limb structure of the Hamlin trees
used in the mature tree portion of this study had to be re-grown, in some cases requiring
the entire scaffold limbs and a portion of the upper trunk to be removed. Thus the heights
of the trunk to the scaffold limbs were not consistent.
Tree size, biomass, N weight, and apparent FNUE were greater for trees grown on
Carrizo citrange rootstock compared with trees grown on Swingle citrumelo, indicating
that differences in tree size are directly correlated with FNUE. Differences in FNUE may
be due to physiology of the rootstocks themselves or due to the distribution of fibrous
roots associated with the various rootstocks. The distribution of root length densities and
N uptake rates for the two rootstocks in this study will be the subject of Chapters 4 and 6,
respectively.

71
Assuming that 30% of the N accumulated in new growth tissues came from
fertilizer (Dasberg, 1987; Feigenbaum et al., 1987; Legaz et al., 1982), it is concluded
that 124 and 108 g of accumulated N originated from the fertilizer inputs. The calculated
FNUEs of 84.0 and 71.8% for Carrizo and Swingle rootstocks, respectively, in this study
were similar to the 61 to 83% reported by Syvertsen and Smith (1996) for 4-year old
grapefruit trees grown in lysimeters. However, mineralized soil N was not included in the
estimation. The contribution of soil organic matter and abscised tree parts will be
addressed in Chapter 6. Accounting for these sources of N will reduce overall NUE for
the citrus trees in this study.
Citrus trees at the Conserv II site were of the same age and similar in size, but the
K.D. Revell grove operated by Cargill, Inc. contained trees of various ages due to past
replanting of trees. The water and nutrient holding characteristics of the soil at this site
were similar to those of the soil at Conserv II. Therefore, it was assumed that trees grown
at this site would follow similar biomass and N partitioning characteristics. Total leaf
area for tree§ju.both experiments increased linearly with TCV and TCSA. LAI increased
rapidly with tree size until the trees were approximately 3-4 years old, after which it
stabilized at about 10. This information can be used to parameterize light interception
functions for a citrus tree photosynthesis and growth model.
Significant relationships were found between total tree fresh and dry biomass and
tree size. The ratio of above-ground to below-ground biomass and N content ranged from
a low of 3:2 to a high of 3:1 across the range of tree sizes found in this study. Citrus roots
of mature citrus trees extend below 1.8 meters (Castle, 1978 and 1980; Elezaby, 1989;
Menocal-Barverena, 2000), but the root system in this study was excavated to a depth of

72
only 1 m. Therefore, the ratio of above-ground to below-ground biomass may have
remained near 3:2 if roots extending below 1 m were included in the total root biomass
measurement.
Citrus trees increase in size with time; branches increase in diameter through
accumulation of xylem tissue, eventually developing a scaffold branch structure of large-
diameter branches. The relationship of total tree biomass and N weight to TCV and
TCSA followed a linear function indicating a constant rate of accumulation with increase
in tree size as measured by both TCV and TCSA. This result implies that the partitioning
of biomass and N accumulation in all plant parts occurs at rates specific to the tree
component. Therefore, the total biomass and/or N weight of a citrus tree can be estimated
for any tree size.
Percentage biomass and N weight of woody tree parts (large branches and trunk)
increased, while those of leaves and twigs decreased with increase in TCV and TCSA.
Caruso et al. (1999) found similar results in peaches where the relative proportion of leaf
and twig dry weights decreased with tree age. It can be concluded that to support the
increase in total tree weight, the biomass and N content of woody branches and trunk
increases at a higher rate compared with leaves and twigs. However, it can be concluded
that LAIt is the driving factor in leaf and twig biomass accumulation since the ratio of
leaf area to ground area under the canopy remained constant with increase in tree size for
medium and large trees. Thus, once the total biomass and N weight of a tree is estimated,
the weights of individual tree parts can be estimated based on tree size. Regression
equations such as those in Tables 3-5 and 3-6 can be used to simulate biomass and N
partitioning in a citrus growth model.

73
Conclusions
Leaf areas of both young and mature citrus trees were correlated with tree size as
measured by TCV and TCSA. Leaf area index increased rapidly for young citrus trees
and then equilibrated at approximately 10 by age 3 to 4 years. This information is
valuable for the estimation of citrus light interception and total photosynthesis. Change in
citrus tree dry weight and N content of the two citrus scions in these studies was shown to
be a linear function of TCV and TCSA. Partitioning of biomass and N decreased for
leaves and twigs, increased for branches, and remained constant for trunk and taproot
tissues with increase in tree size. While mature citrus trees grown on Swingle citrumelo
rootstock were consistently smaller than trees of similar age grown on Carrizo citrange,
mass partitioning of tree parts were similar for both rootstocks. Thus, with the exception
of spatial root length density distribution described in chapter 4, the only effect of the two
rootstocks and two scions used in these studies was on tree size relative to TCV and
TCSA Therefore, biomass and N partitioning for specific tissues with tree size can be
captured in generic linear relationships. The N balance estimated for mature citrus trees
in this study indicated an apparent fertilizer N use efficiency of 60 to 70%.

CHAPTER 4
CITRUS ROOT GROWTH DYNAMICS
Introduction
While the role of roots in anchoring crop plants, particularly tree crops, to the soil
should not be taken for granted, the function of roots as absorbing organs for both water
and nutrients can not be overemphasized. The structure of a root system is important in
determining the pathway and resistance to water and solute uptake, and the volume of
soil accessible to crop plants (Kramer and Boyer, 1995). The entrance of water and
nutrients into young roots occurs a few cm behind the root tips because of the lack of a
functional xylem at the tip and the suberization of root hypodermis and endodermis
tissues with age (Tinker and Nye, 2000). Thus, the larger the length of relatively small
diameter fibrous roots a crop root system has, the greater the amount of water and
nutrients available to it. Likewise, the larger the soil volume a crop root system occupies,
the greater the pool of water and nutrients available for uptake.
The goal of fertilizer application should be the placement of nutrients within the
crop root zone to insure the most efficient uptake. Maintenance of adequate water and
labile nutrient concentrations within soil zones occupied by the crop root system is
essential for optimal nutrient uptake. Understanding the spatial distribution of fibrous
roots is essential to ensure proper fertilizer placement to improve nutrient uptake and
potentially reduce leaching below the root zone.
74

75
Several Florida studies have demonstrated that tree size and yield were related to
fibrous root density and/or distribution (Castle and Krezdom, 1975; and Ford 1954a;
1964; 1968; 1969; 1972) in the deep sandy soils of central Florida. Since processes that
control non-point source pollution are dynamic and greatly affected by genetic traits and
environmental conditions, development of models that can integrate interactive effects of
spatial and temporal processes will be critical. However, models providing realistic
results will need to include accurate information on root growth dynamics. Modeling of
citrus root distribution and the determination of water and nutrient uptake parameters can
lead to the development of an expert system for the estimation of water and nutrient
depletion and uptake by soil depth. Likewise, nutrient leaching can be estimated due to
excessive irrigation or heavy rainfall.
Under Florida growing conditions, the quantity of fibrous roots decreased with
depth and lateral distance from the trunk (Elezaby, 1989; Menocal-Barberena, 2000).
Eighty percent of citrus fibrous roots were found within a 120 cm radius of the tree trunk
and 40% in the upper 30 cm of well-drained sandy soils. Nearly all citrus roots grow
within 45 cm of the soil surface where artificial drainage was provided and/or high water
tables occured (Calvert et al., 1967; 1977; Ford, 1954a; Ford, 1972; Reitz and Long,
1955).
r\
Fibrous root dry mass densities ranged from 300 to 1200 g m'3 (Castle, 1978;
1980). Citrus fibrous root length densities ranged from 0.53 cm cm"3 for Swingle
citrumelo roots to 2.02 cm cm"3 for trifoliate orange (Eissenstat, 1991). Elezaby (1989)
reported fibrous root concentration in the 0 to 30 cm soil zone increased from 450 to
1000 g m between trees when the in-row distance decreased from 4.5 to 2.5 m due to

76
a
/
overlapping root systems.
Castle and Krezdom (1975) described two general types of root systems, the first
characterized by extensive lateral and vertical development, and the second with
intensive higher fibrous root density near the soil surface. Trees on rough lemon,
Volkamer lemon and Palestine sweet lime (C. limettioides Tan.) typified the extensive
type of root system where 50% of the fibrous roots occurred below 70 cm in the soil and
produced large, highly-productive trees that dominated the citrus industry in Florida
when trees were irrigated less intensively and were set at much lower densities.
Unfortunately, rough lemon has been virtually eliminated as a commercial rootstock due
to citrus blight disease. Examples of the intensive-type root system were Carrizo citrange
and Swingle citrumelo that had few fibrous roots below 70 cm, and the root system was
less developed laterally. These rootstocks now dominate the citrus industry in Florida and
are well suited for high-density, intensively irrigated plantings.
The following hypotheses were tested: 1) Root distribution is significantly
affected by rootstock, and 2) Generic relationships can be developed for well-drained
soils that describe citrus root densities at various depths from the soil surface and
distances from the tree as a function of tree size. To test these hypotheses, the objectives
of this study were to: 1) develop information on spatial root length distribution at
different soil positions and depths for two citrus rootstocks, and 2) develop functional
relationships that define root length densities at various soil positions and depths as a
function of tree size. The relationship of vertical and horizontal root length density
distribution to tree size resulting from this study can be used to estimate fibrous root
densities in various soil layers for citrus water and nutrient uptake models. This

77
relationship will provide a scientific basis for the development of water and nutrient
components of an expert system for improved citrus irrigation and N management.
Materials and Methods
Sample Collection
The same 19 Hamlin and Valencia orange trees used in the previous biomass
and N distribution study (Chapter 3) were used to determine the spatial relationship
between citrus root length density and tree size. Soil cores were removed once the trees
had been cut to the ground but prior to the excavation of the main root system. Cores
were taken with a 7.6 cm diameter bucket auger and roots were sampled at 50,100, and
150 cm from the tree trunk in the row and 50, 100, 150 and 200 cm between tree rows.
Samples were collected at 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 90 cm depths.
Each sample was placed into separate plastic bags, sealed, and marked with tree
identification, depth and distance from the tree. The samples were placed in a cooler
containing ice and were subsequently frozen at -4 C.
Sample Processing and Statistical Analysis
Roots were removed from the soil by washing though an 850 an sieve. Any
debris not passing through the sieve was removed manually, and the roots were separated
into size categories by diameter. These categories were <4 mm, 4 to 20 mm and >20 mm.
Root lengths for roots 0 to 4 mm in diameter were determined prior to drying using the
line intersect method (Newman, 1966). Root length density data from samples collected
from the 12 trees at the Water Conserv II site (mature tree study) were analyzed by the
general linear model procedure of SAS (SAS Institute, Inc., Cary, NC). Root length
density data from the soil samples collected from the trees of various sizes at the Cargill

78
grove were analyzed using Proc REG in SAS. Regression equations were determined
using SigmaPlot (SPSS, Inc., Chicago IL).
Results
Mature Hamlin Orange Root Distribution
Soil depth and distance from the tree trunk significantly (P = 0.01) affected citrus
root length density (Table 4-1). Mean root length density of fine fibrous roots (<4 mm)
extracted from soil cores surrounding the 12 mature citrus trees followed a bimodal
spatial distribution with depth from the soil surface (Fig. 4-1), and distance from the tree
trunk (Fig. 4-2). Mean fine fibrous root density in the upper 15 cm was 1.04 cm cm'3.
Densities ranged from 1.9 cm cm'3 soil at 50 cm from the tree trunk to 0.7 cm cm'3 at 200
cm. Mean densities decreased at the 15 to 30 cm depth to 0.30 cm cm'3 and ranged from
i
0.5 to 0.07 cm cm' at 50 and 200 cm distances, respectively. Mean densities of fine
fibrous roots increased at depths below 40 cm to a maximum at the 60 to 75 cm depth
(0.28 cm cm'3) then declined at the 75 to 90 cm depth (0.27 cm cm 3). Densities at the 60
to 75 cm depth were 0.3, 0.3, 0.3, and 0.03 cm cm'3 at distances of 50, 100,150, and 200
cm from the tree trunk, respectively.
Fine fibrous root densities at the 0 to 15 cm depth were generally greater in the in
row orientation than in the cross-row orientation (data not shown). Mean in-row spatial
root length densities (0.41 cm cm'3) were greater, but not significantly different from
densities for between-row orientation (0.35 cm cm3) (Table 4-1), because more overlap
in root systems from adjacent trees probably occurred in this orientation.

Table 4-1. Mean fibrous (diameter <4 mm) root length density of mature Hamlin orange tree as affected by rootstock,
orientation distance, and soil depth.
Mean Root Length Density (cm cm'3)
Rootstock
Orientation
Distance from tree (cm)
Soil depth (cm)
Carrizo
Swingle
In-row
Cross-row
50 100 150 200
0-15
15-30
30-45 45-60
60-75
75-90
0.36
0.41
0.41
0.35
0.49 0.40 0.33 0.17
1.04
0.30
0.16 0.24
0.28
0.27
P
Significance1
Rootstock
0.290
NS
Orientation
0.253
NS
Distance
0.002
***
Depth
<0.0001
***
Distance*Depth
0.820
NS
Rootstock* Distance
0.052
*
Rootstock*Depth
0.002
***
Significance: NS = not significant, = significant at P=0.1 level, ** = significant at P=0.05 level, and *** = significant at
P=0.011evel.

80
Fig. 4-1. Root length density distribution by depth at 50, 100, 150 and 200 cm distances
from the tree trunk between rows of Hamlin orange trees on Carrizo citrange
(A) or Swingle citrumelo (B) rootstocks.

81
Fig. 4-2. Root length density distribution at 0-15, 15-30, 30-45,45-60, 60-75, and 75-90
cm depth increments by distance from the tree trunk as affected by distance from
the tree trunk for Hamlin orange trees on Carrizo citrange (A) and Swingle
citrumelo (B) rootstocks.

82
Root densities at the 50 cm distance in the cross-row orientation decreased more
gradually than did densities at 100, 150, and 200 cm distances. Minimum densities
occurred at the 45 to 60 cm depth for the 50 cm distance as opposed to the 30 to 45 cm
depth for the 100, 150, and 200 cm distances. Similarly, root densities increased at the 60
to 75 cm depth for the 100 and 150 cm distances, and 75 to 90 cm depth for the 50 cm
distance. Spatial root distribution differences between rootstocks were not statistically
significant (Table 4-1). Mean root length densities at all depths and distances were 0.36
cm cm'3 for trees grown on Carrizo citrange and 0.41 cm cm"3 for trees grown on Swingle
citrumelo. However, the interaction of rootstock and depth was significant at the P=0.01
level. Trees on Swingle had higher root length densities near the soil surface than did
trees on Carrizo (Figs.4-1 and 4-2). Conversely, root length densities were greater for
trees on Carrizo between 15 and 75 cm below the soil surface.
Root length densities for the 0 to 15 cm depth ranged from 2.0 to 0.9 cm cm'3 soil
at distances of 150 cm or less for trees on Swingle rootstock. Densities ranged from 1.2 to
0.7 cm cm'3 at the same depth and distances for trees on Carrizo rootstock. Root densities
decreased for both rootstocks to low values at the 30 to 45 cm depth. Root densities
increased for trees on Carrizo at 45 to 60 cm depth, whereas densities for trees on
Swingle increased at the 60 to 75 and 75 to 90 cm depths. With the exception of the 50
cm distance from the tree trunk, root densities were greater for trees on Carrizo at 30 to
45 cm than those on Swingle. Likewise, root densities were greater at 45 to 60 cm depth
for trees on Carrizo than at the 60 to 75 cm depth for trees on Swingle.

83
Root Length Density Distribution Changes with Tree Size
Citrus root length densities were significantly different at the P=0.01 level for
both distance from the tree trunk and depth from the soil surface across a wide range of
tree sizes (Table 4-2). Three-dimensional graphical representations of developing root
systems are presented in Fig. 4-3. These graphs represent trees approximately 2 to 5 years
old (Figs. 4-3A and 4-3B), 5 to 10 years old (Figs. 4-3C and 4-3D), 10 to 15 years old
(Figs. 4-3E), and >15 years old (Figs. 4-3F). Root systems were initially concentrated at
the surface with few roots deeper than 0.5 m at a distance of 150 cm from the tree trunk.
As the citrus trees produced substantial fruit (5 to 10 years of age) root length density
increased at the soil surface to the dripline of the tree. Roots eventually extended to the
200 cm distance between tree rows and to a depth of 0.9 m at 150 cm from the trunk. The
bimodal nature of the root system can be seen near the tree at depths below 60 cm. By the
time the tree reached 10 to 15 years of age and the canopy was nearing full hedgerow
dimensions, the bimodality of the root system was fully developed and roots extended
past a depth of 1 meter at all distances from the tree.
Tables 4-3 and 4-4 list the regression coefficients for a third order polynomial
relationship of canopy volume and trunk diameter to root length density at all depths and
distances. The r2 values were greater, and RSME and P values were lower for most
regressions using canopy volume than those using trunk diameter, indicating that canopy
volume measurements are a more accurate predictor for assessing root length density
compared with trunk cross sectional area.

84
Fig. 4-3. Citrus root distributions by depth below the soil surface and distance from the
tree trunk for trees 2-5 years old (A and B), 5-10 years old (C and D), 10-15 years
old (E), and > 15 years old (F).

85
Table 4-2. Regression analysis of citrus fibrous (diameter < 4 mm) root length densities
for trees ranging from 2 to > 15 years old.
RMSE
CV
R2
P
Significance
Orientation
44.66
87.81
0.002
0.291
NS
Distance
48.08
42.54
0.12
<0.0001
***
Depth
23.45
51.99
0.17
<0.0001
***
Table 4-3. Regression coefficients and statistics root length density as a function of
distance from the tree trunk, and soil depths by canopy volume using a third order
quadratic polynomial model2.
Distance
(cm)
Depth
(cm)
Yo
a
b
c
R1
RMSE
(cm cm'3)
p
50
0-15
1.29
-0.021
0.002
0.00001
0.84
0.33
0.001
50
15-30
0.99
-0.289
0.027
-0.0006
0.78
0.15
0.005
50
30-45
0.27
-0.059
0.006
-0.0001
0.68
0.09
0.021
50
45-60
0.45
-0.16
0.015
-0.00003
0.83
0.12
0.002
50
60-75
0.29
-0.099
0.009
-0.00002
0.68
0.09
0.022
50
75-90
0.31
-0.14
0.013
-0.0003
0.88
0.06
0.0005
100
0-15
0.73
-0.23
0.023
-0.0005
0.79
0.15
0.005
100
15-30
0.44
-0.13
0.014
-0.0003
0.61
0.15
0.072
100
30-45
0.19
-0.069
0.007
-0.00002
0.48
0.14
0.14
100
45-60
0.20
-0.082
0.008
-0.00002
0.82
0.05
0.002
100
60-75
0.071
-0.068
0.006
-0.00001
0.85
0.04
0.001
100
75-90
0.040
-0.018
0.002
-0.00001
0.51
0.04
0.107
150
0-15
0.17
-0.087
0.011
-0.0001
0.97
0.06
<0.0001
150
15-30
0.10
-0.047
0.005
-0.0001
0.81
0.06
0.003
150
30-45
0.007
-0.004
0.001
-0.00001
0.74
0.02
0.010
150
45-60
0.025
-0.012
0.001
-0.00001
0.80
0.03
0.003
150
60-75
0.007
-0.004
0.0004
-0.000001
0.75
0.01
0.008
150
75-90
0.008
-0.004
0.0003
-0.000001
0.65
0.01
0.030
200
0-15
0.042
-0.020
0.002
-0.0001
0.99
0.01
0.014
200
15-30
-0.013
0.007
-0.001
-0.0001
0.99
0.01
<0.0001
200
30-45
-0.0003
0.001
-0.0001
0.00001
0.99
0.01
<0.0001
200
45-60
-0.002
-0.001
0.0001
0.00001
0.99
0.01
0.004
200
60-75
0.003
-0.001
0.0001
-0.000001
0.99
0.01
0.004
200
75-90
-0.001
0.001
-0.0001
0.00001
0.99
0.01
0.0002
z Y-Y0+aX+bX2+cX3 where X = TCV, and Yo, a, b, and c are regression coefficient.

86
Table 4-4. Regression coefficients and statistics for root length density as a function of
distance from the tree trunk, and soil depths by trunk cross-section area using a third
order quadratic polynomial model2.
Distance
(cm)
Depth
(cm)
Y0
A
b
c
R2
RMSE
(cm cm'3)
P
50
0-15
-0.41
0.001
-0.00005
0.00001
0.71
0.44
0.015
50
15-30
0.24
0.0001
-0.00003
0.00001
0.56
0.21
0.071
50
30-45
0.04
0.0001
-0.00001
0.00001
0.68
0.09
0.022
50
45-60
-0.50
0.0005
-0.00006
0.00003
0.88
0.10
0.001
50
60-75
-0.08
0.0001
-0.00002
0.00002
0.66
0.09
0.027
50
75-90
-0.20
0.0002
-0.00003
0.00001
0.79
0.08
0.004
100
0-15
0.23
0.00001
-0.00001
0.00001
0.67
0.19
0.026
100
15-30
-0.01
0.0001
-0.00001
0.00001
0.36
0.19
0.343
100
30-45
-0.42
0.0004
-0.00003
0.00002
0.52
0.13
0.099
100
45-60
-0.17
0.0002
-0.00002
0.00002
0.82
0.05
0.003
100
60-75
-0.15
0.0002
-0.00002
0.00001
0.90
0.03
0.0002
100
75-90
-0.20
0.0002
-0.00001
0.00001
0.97
0.01
<0.0001
150
0-15
-0.50
0.0004
-0.00001
0.00002
0.96
0.07
<0.0001
150
15-30
-0.30
0.0003
-0.00003
0.00001
0.83
0.06
0.002
150
30-45
-0.11
0.0001
-0.00001
0.00001
0.83
0.02
0.002
150
45-60
-0.15
0.0001
-0.00001
0.00001
0.78
0.03
0.005
150
60-75
-0.01
0.00001
-0.00001
0.00001
0.74
0.01
0.010
150
75-90
-0.05
0.00001
-0.00005
0.00001
0.61
0.02
0.046
200
0-15
-0.28
0.0002
0.00003
0.00001
0.94
0.04
0.086
200
15-30
-0.33
0.0003
-0.00003
0.00001
0.78
0.08
0.311
200
30-45
-0.07
0.0001
-0.00001
0.00001
0.80
0.02
0.282
200
45-60
-0.03
0.00001
-0.00003
0.00001
0.87
0.01
0.192
200
60-75
-0.04
0.00001
-0.00004
0.00001
0.88
0.01
0.180
200
Z V vr
75-90
-0.03
;
0.00001
-0.00002
0.00001
0.79
0.01
0.293
z Y-Yo+aX+bX2+cX3 where X = TXA, and Y0, a, b, and c are regression coefficients

87
Discussion
The root distributions of the two rootstocks used in this study appear to lie
intermediate between the extensive and intensive distributions described by Castle and
Krezdom (1975). However, there were some subtle differences between rootstocks.
Citrus trees grown on Swingle citrumelo had greater root length densities in the upper 30
cm than did trees grown on Carrizo citrange. The root density distributions of both
rootstocks were bimodal in nature. However, Carrizo roots tended to grow deeper than
did those of Swingle. Trees grown on Carrizo citrange rootstock had higher fibrous root
length densities at all distances from the tree trunk below 60 cm from the soil surface.
These root length densities indicate that the depth of irrigation and the depth to which
fertilizer N is initially placed should be rootstock specific. Thus, mature trees grown on
Swingle citrumelo should be irrigated to a shallower depth compared with trees grown on
Carrizo citrange. Deep irrigation will waste water and potentially leach soil N below the
soil volume containing the largest proportion of roots, thus potentially decreasing NUE.
Menocal-Barberena (2000) found similar trends using fibrous root mass densities
from root samples collected at the same site. In his and the current studies, Swingle had
greater, but not significantly different, fibrous root concentrations than Carrizo. Mean in
row root concentrations were significantly greater than between-row concentrations.
Mean root concentrations in the upper 30 cm were more than four times greater than for
soil layers below the 30 cm depth.
Under central Florida Ridge conditions, citrus fibrous root densities increase in
two modes. The first mode is the development of a dense root mat just below the soil
surface. This portion of the root system expands in radial manner away from the tree

88
trunk while continuing to increase in density near the trunk itself. Expansion continues
through maturity with trees in dense plantings overlapping in both in-row and between-
row directions. This portion of the citrus root system is important for tree stability while
providing adequate roots for water and nutrient uptake for the developing tree. A second
region of root growth develops below 30 cm between 5 and 10 years of age. This region
^ i
of root growth increases in size and density through maturity. The development of a
deeper root system is essential for supplying adequate water and nutrients to the mature
tree from an increasingly large soil volume. These two growth modes result in the
bimodal root distribution of the mature citrus tree. While complex, the development of
citrus root systems with time appears to be predictable and basic trends can be captured
in functional relationships within a citrus model. The data presented here can provide the
root distribution information needed to determine spatial soil water and nutrient uptake
for a citrus growth model.
Conclusions
It is concluded that the root length distribution of trees grown on Swingle and
Carrizo rootstocks followed predictable patterns with increased tree size resulting in
mature trees with bimodal root systems. While both rootstocks developed a dense root
system within the upper 30 cm of the soil surface, the root systems extended beyond 1 m
in depth, so extending sampling to greater soil depths on well-drained soils appears to be
desirable. Trees on Swingle developed higher root length densities near the soil surface,
and lower densities below 30 cm compared with trees on Carrizo. Based on the overall
high r2 values and low RMSE, the functional relationships that were developed in this
study account for most of the variability in root length density. Thus, the test hypothesis

89
is correct and a model for root length density distribution can be made if both tree size
and rootstock are included as variables. These relationships provide a scientific basis for
the development of a spatial root length distribution model needed for a citrus expert
system that will estimate water and nutrient uptake.

I
CHAPTERS
CITRUS WATER UPTAKE DYNAMICS
Introduction
Climate, crop development, soil water status, and competition with other plants
affect total soil water use by crop plants. Stomatal conductance regulates both
transpiration and photosynthesis and therefore directly affects the water use and
productivity of plants (Jones et. al., 1985). Stomata are sensitive to environmental
variables such as light, CO2, vapor pressure deficit (VPD), and plant water status (Jarvis
and McNaughton, 1986). Assuming that light, CO2, and VPD conditions are nearly
constant during short intervals of time, hourly and daily changes in plant water status can
have large impacts on stomatal conductance and thus on transpiration rates and ultimately
productivity. The key to plant water status is soil water availability (Allen et al., 1997),
thus an improved understanding of soil water uptake dynamics is essential in optimizing
both the amount and timing of irrigation required for maximum production. Once
understood, seasonal soil water content and root length density effects on citrus water
uptake can be modeled to provide more accurate irrigation scheduling, which will reduce
negative impacts on ground water quantity and quality due to leaching and over
pumping.
Citrus water requirements vary with climatic conditions, variety, and canopy size.
Lower crop evapotranspiration (ETC) rates for Florida (humid) compared with Arizona
(semi-arid) have been attributed to lower evaporative demand (Rogers et al. 1983, Fares
90

91
and Alva 1999). Rogers and Bartholic (1976) determined that annual ETC increased at a
rate of 19 mm per year as trees grew, leading to a cumulative increase of approximately
13% in an 8-year period.
Soil water content can also be reduced by evaporation from the soil surface and
transpiration from non-crop species (Allen et al., 1998). Generally, soils lose the ability
to transport water to the surface as they dry (Hillel, 1998). Citrus ETC decreases as the
fraction of the soil surface receiving full sun decreases and the canopy shades an
increasingly larger ground area (Castel and Buj, 1992). Soil water use or apparent ETC
increases with increased ground coverage by non-crop species (Smajstrla et al., 1986).
The above factors combine to limit ET for a given crop under given conditions.
Allen et al. (1998) proposed that ETC can be derived from calculated ET0 as follows:
ETC = ET0 Kc Ks Equation 5-1
Where:
ETC = Crop evapotranspiration (mm d'1)
ET0 = Potential evapotranspiration (mm d'1)
Kc = Crop coefficient
Ks = Soil stress coefficient
The crop coefficient (Kc) is defined as the ratio of ETC to ET0 at field capacity (0fc) In
this case Ks is assumed to be equal to unity. This coefficient is indicative of climatic
and/or developmental effects on ETC compared with ET0 when water uptake is not limited
by soil water depletion. Estimates of Kc for citrus range from a minimum of 0.6 in the fell
and winter to a maximum of 1.2 during the summer months (Boman, 1994; Fares and
Alva, 1999; Martin et al., 1997; Rogers et al., 1983).

92
Soil water content (0) must be maintained between specific upper and lower
limits such that water availability to the crop does not limit growth or adversely impact
yield or quality. This upper limit of 0 after free drainage occurs is defined as the value of
0 at which redistribution of soil water ceases (HilleL, 1998) and is also known as field
capacity (0fc)- The lower limit or permanent wilting point (0pwp) is the value for 0 at
which a wilted plant can no longer recover turgidity. The range of 0 between 0fc and
0pwp is known as total available water (TAW). These three values for 0 and the soil water
potential () at which they occur are characteristic and relatively constant for any given
soil. If the effects of soil physical characteristics on soil water use are understood, soil
water can be maintained within these limits and the potential for both crop water stress
and environmental contamination can be minimized.
The soil water depletion coefficient (Ks) in equation 5-1 is a measure of the
reduction in ETC caused by reduced soil water uptake due to reduced 1998). Water moves to regions of high from regions of lower as water is removed
from the soil surrounding the root surface. Soil water movement slows as 4> of the bulk
soil decreases and 0 approaches 0pwp. Allen et al. (1997) determined that a 0 exists (0ra)
less than 0fc where water uptake was not limited by . They referred to the range of 0 to
0ra as readily available water (RAW) and used this value to estimate K* as the ratio of
depletion of total available soil water (TAW-0) to the soil water not readily available
(TAW-RAW) where Ks is not greater than unity (Equation 5-2). Therefore, the greater
the RAW for a given soil, the longer water can be withdrawn from that soil before ETC is
limited. A crop and specific depletion in TAW must be determined below which crop
/ >
growth and yield is reduced. Under Florida conditions, Koo(1963) estimated this
L

93
delpetion in TAW to be 33% from February to June and 66% from July to January. These
values were determined for relatively low density plantings with overhead irrigation. The
depletion amounts for high density plantings irrigated with under-tree low volume
mircrosprinklers have not been determined.
K TAW-fl
* TAW-RAW
Equation 5-2
Where:
Ks = Soil water stress coefficient
TAW = 0fc pwp = Total available water (cm3 cm'3)
0 = Soil water content (cm3 cm'3)
RAW = 0fc 0ra = Readily available water (cm3 cm'3)
The hypotheses to be tested for the following study on citrus soil water dynamics
are: 1) seasonal changes in maximum daily water uptake under non-limiting soil water
conditions follows predictable patterns relative to ETo, 2) water uptake decreases with
soil water content, and 3) soil water uptake is greatest in soil volumes containing the
highest root length densities. The objectives of this study were to: 1) estimate mature tree
daily ETC during a 2-year period, 2) calculate monthly Kc values based on the relationship
Kc = ETC / (ET0*Kg), 3) determine the relationship of estimated ETC to soil water content
to determine Ks values over a range of , and 4) evaluate ETC per unit root length density.
The resulting relationships will provide critical information required for the development
of predictive models for citrus water uptake and soil water depletion over time. Such
models can provide a basis to protect Floridas water resources though better irrigation
scheduling and appropriate water application rates.
/

94
Materials and Methods
Site Characteristics
Hamlin orange (Citrus sinensis L.) grafted on Carrizo citrange (C. sinensis L.
Osbeck X Poncirus trifoliata L. Raf.) rootstock trees planted 3.1 m in the row and 6.1 m
between rows (used in Experiment 1 of Chapter 3) were used in this study. The trees had
been pruned each of the past 3 years and had formed a hedgerow approximately 3.8 m
wide and 5.9 m tall. Herbicides were applied as needed to maintain a nearly weed-free
strip 3.5 to 4.0 m wide beneath the tree canopies. Soil type at the site was Candler fine
sand (hyperthermic, uncoated Typic Quartzipsamments) with a field capacity of
approximately 0.08 cm3 cm"3 (Obreza et al., 1997). Irrigation was applied to the tree row
using one microsprinkler per tree with a flow rate of approximately 60.5 L hr"1 and a 360
degree spray pattern with a diameter of approximately 3.7 m. The equivalent mean
precipitation rate of the sprinkler was 0.58 cm hr"1. Reclaimed municipal-waste water
provided by the Water Conserv II Project was used as the source of irrigation water.
Soil Capacitance Sensor Data Collection
Soil water content at 10, 20, 40, and 80 cm depths was recorded at 30 minute
intervals during a 2-year period in the irrigated and non-irrigated areas under three
mature (14-year-old) citrus trees. These increments represent soil depths of 0 to 15,15 to
30, 30 to 60, and 60 to 100 cm, respectively (Fig. 5-1). These 0 data were obtained using
EnviroSCAN (Sentek Pty. Ltd., South Australia) capacitance sensors installed two
distances away from the tree trunk in the row and three distances perpendicular to the
row (Fig. 5-2). In-row sensors were placed 0.75 m from the tree trunk and at the midpoint

95
Fig. 5.1. Illustration of EnviroSCAN probe. Number on sensor indicates depth of sensor.
^ Tree trunk
O EnviroSCAN probe
Fig. 5-2. Illustration of EnviroSCAN probe layout, and soil surface area used for
determining soil water content for each probe.

96
between trees (1.5 m). Between row sensors were installed midway between the tree
trunk and the canopy dripline (0.9 m), the canopy dripline (1.8 m), and between the
dripline and midline between tree rows (2.7 m).
Access tubes (5 cm diameter, acrylonitrile butadiene styrene) were installed, and
each sensor was individually normalized following manufacturer recommendations. A
general calibration curve was developed for the soil type at the site using a gravimetric
method described by Morgan et al. (1999). Daily Penman ET0 values from a Florida
Automated Weather Network (FAWN) station located less than 0.4 km away from the
actual field site were recorded.
Estimated Daily ETC
Soil water content in deep sandy soils equilibrates to field capacity within a few
hours after irrigation or rainfall. The net change in 0 over a 24 hour period was calculated
for each sensor using the difference between 0 values recorded at midnight and values
recorded the previous midnight (A0). To avoid hysterisis effects, only 0 data collected on
days receiving no rainfall or irrigation were used for this calculation. Daily soil water
depletion depth was calculated for each sensor by multiplying the A0 by a corresponding
soil depth for that sensor (Fig. 5-1). The resulting soil water depths were then multiplied
by the surface area assigned to the given probe (Fig. 5-2). The resulting soil water
depletion volumes were summed as an estimate of daily soil water depletion (DSWD)
Daily ETC was estimated by dividing DSWD by the area occupied by the tree (18 m2).
Available soil water and daily weighted 0 were determined using the same mehod as that
of DSWD by substituting 0rc and mean daily 0 for A0. Mean daily 4> was estimated from
the daily weighted 0 using the soil water characteristic curve for Candler fine sand
%

97
previously determined at this site (Obreza et al., 1997). Percentage daily ASWD was
< A p
determined using DSWD and ASW.
Estimated Monthly Crop Coefficient (K)
Estimated daily tree water use (ETC) was calculated for a 24-month period and
compared with calculated daily ET0. The ratios of estimated daily ETC to calculated daily
ET0 for each of the three trees were averaged to estimate the product Kc Ks in equation
5-1. To eliminate the effects of decreased on water uptake, ratios of ETC to ET0 on days
where mean 0 was near 0fc in both the irrigated and non-irrigated areas (Ks assumed to
be 1) were used to estimate daily Kc. The relationship of these estimated daily Kc values
to day of year (DOY) was determined using regression analysis.
Estimated Water Stress Coefficient (Ks)
Daily ETC to ET0 ratios were calculated throughout the year and compared with
mean daily 0. The ratio of ETC to (ET0*Kc) using the Kc estimated for the day was used to
estimate the value of Ks. The relationship of the ratio of ETC to (ET0*Kc) with 0 and (j) is
typically a logistic response curve with a plateau near 0rc (Allen et al., 1998). Regression
analysis was used to determine the relationship of estimated Ks to ASWD and mean soil
Estimation of Soil Water Uptake per Unit Root Length
Daily estimated ETC on a per unit root length basis for each sensor were
determined for the mean under-canopy, dripline, and between-row locations for soil
depths of 10, 20, 40 and 80 cm. Root length densities determined in Chapter 4 (for trees
on the same rootstock, same approximate tree size, and grown in the same location under

98
the same irrigation and fertilization practices) were used as an approximation of root
length desities at the various locations and depths.
The rate of soil water withdrawal is influenced by surface evaporation and is
related to the amount of soil shading. Likewise, transpiration by groundcover species
increases withdrawal. Water withdrawal rates per unit root length at all locations and
depths were compared. Increased water loss at the soil surface that could not be explained
by root density was assumed to be due to one or both of the above factors.
Results
Seasonal ET and ETC Trends
Daily ET0 reported by FAWN for the experimental site ranged from a minimum
of 1.1 mm in December, 2001 to a maximum of 6.5 mm in June, 2000 (Table 5-1). The
standard deviations for ET0 by month were relatively low (<0.4 mm) for all months with
the exception of transition months between seasons (February and March in the spring
and August and September in the fall). This result indicates that weather condtions
related to ETC were relatively stable with the exception of these transition periods.
Monthly maximum, minumum, and mean values for ET0 and ETC were not significantly
different for corresponding months during the 2 years of observations included in this
study. Although generally lower, daily ETC followed the same seasonal patterns as ETC.
Exceptions to this trend occurred during the summer months of June through August, but

Table 5-1. Monthly maximum, minimum, and mean reference evapotranspiration reported by Florida Automated Weather
Network for the Avalon Station and maximum, minimum and mean estimated citrus crop evapotranspiration.
Reference Evapotranspiration (mm d'1) Standard Crop Evapotranspiration (mm d*1) Standard
Months
Maximum
Minimum
Mean
Deviation
Maximum
Minimum
Mean
Deviation
April,2000
5.0
4.0
4.5
0.31
4.7
3.1
3.9
0.54
May
5.9
4.1
5.0
0.46
5.9
3.9
4.6
0.57
June
6.5
5.0
5.7
0.44
6.5
3.5
4.7
0.72
July
5.8
4.6
5.3
0.35
5.8
3.5
4.6
0.72
August
5.2
4.1
4.7
0.33
5.4
3.0
4.2
0.66
September
5.0
3.3
4.3
0.48
4.8
2.5
3.5
0.70
October
3.8
2.3
3.0
0.38
2.7
1.7
2.2
0.32
November
2.9
1.6
2.3
0.39
2.2
1.3
1.8
0.26
December
2.8
1.5
1.9
0.31
1.9
1.0
1.3
0.24
January, 2001
2.9
1.3
2.0
0.45
2.1
1.0
1.4
0.26
February
3.6
1.2
2.7
0.70
2.6
1.4
1.8
0.29
March
4.3
2.6
3.5
0.38
3.1
2.0
2.8
0.28
April
5.0
3.3
4.5
0.36
5.2
2.5
4.0
0.68
May
5.9
3.6
5.3
0.52
6.2
3.9
4.3
0.60
June
6.4
4.9
5.6
0.35
6.3
3.5
4.8
0.91
July
6.2
5.1
5.5
0.28
6.3
3.6
4.5
0.53
August
6.2
4.6
5.3
0.44
5.9
2.7
4.5
0.63
September
5.5
3.4
4.5
0.62
4.8
1.9
3.4
1.01
October
3.6
2.2
2.9
0.39
3.0
1.5
2.0
0.38
November
2.7
1.4
2.1
0.31
2.8
1.6
1.4
0.47
December
2.2
1.1
1.8
0.27
1.8
1.1
1.3
0.31
January, 2002
3.0
1.2
2.0
0.58
1.9
1.2
1.4
0.22
February
3.2
1.9
2.6
0.29
2.3
1.3
1.9
0.34
March
4.4
2.1
3.6
0.56
3.4
1.6
2.9
0.55
April
5.3
3.3
4.3
0.60
4.6
1.7
3.2
0.74
Month
7 ^ :
NS
NS
Statistical Significance1
NS NS NS
NS
NS
NS
NS = Not significantly different by General linear Model at the p=0.05 level.

100
Seasonal Kc*Kg
The ratio of estimated daily ETC to calculated daily ET0 is an approximation of the
quantity KC*KS in Equation 5-1. The ETC to ET0 ratios (= KC*KS) were plotted against the
weighted 0, ASWD, and for the irrigated area to depths of 0.5 and 1 m, or for the total
land area allocated to the tree area to a 1 m depth (Figs. 5-3 to 5-5). Regression equations,
R2, RMSE, and P values are provided in Table 5-2. The R2 values for the equations are
generally smaller than 0.5 due to the wide scatter of data points as indicated by the
relatively large RMSE values. However, all relationships were significant at the P=0.01
level. The trends were particularly strong for regressions against soil water potential.
Theoretically, the Ks value should be approximately 1 when 0 is near 0fc (0.075
to 0.08 cm3 cm"3). Therefore, the ETC to ET0 ratios should approximate Kc at 0fc. Fig. 5-
6 illustrates the ETC to ET0 ratios by day of year (DOY) when mean 0 in the irrigated
zone was between 0.07 and 0.085 cm3 cm"3 to a depth of 1 m. These ratios ranged from
0.81 on DOY 24 (January) to 1.12 on DOY 179 (June). The regression equation for this
relationship is given in Table 5-3. With an R2 of 0.76, DOY explaines more than 76% of
the variation in the ETC to ET0 ratios when 0 was between 0.07 and 0.085 cm3 cm"3.
Therefore, the equation provides a good approximation of the value of Kc for a given
DOY.
K* Estimation
The ETC to ET0 ratios for 0 values less than 0fc would approximate the Kg value
assuming the Kc is 1. Since we have demonstrated that Kc values do not equal 1 during

101
Fig. 5-3. Estimated ETC to calculated ET0 ratio as a function of soil water content in the
irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a 1 m
depth (C). The data points shown represent a range of soil water content from
field capacity to approximately 50% available soil water depletion.

102
1.2
1.0 -
0.8
0.6
f0'
0.4 -
fgg, o -a? c'&~c
2 cd 095 m
o co
o<6 /p oo 1 _
o
oB
0.2
12 -
1.0 -
2
% 0.8
¡5,
til 0.6
0.4 -i
£ VW
v--\
8
O O
#>?
O
o^o^o
B
0.2
1.2
1.0 H
0.8
0.6
0.4
&Vrf>
wj$^?D5S
0.2
T 1
12 -10
-20 -18 -16 -14
Soil Water Potential (kPa)
I
-8
i
Fig. 5-4. Estimated ETC to calculated ET0 ratio as a function of soil water potential in the
irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a 1 m
depth (C). The data points shown represent a range of soil water content from
field capacity to approximately 50% available soil water depletion.

103
1.2-
0 10 2D 30 40 50 60
Available Soil Wbter Depletion (%)
Fig. 5-5. Estimated ETC to calculated ET0 ratio as a function of available soil water
depletion in the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total
tree area to a 1 m depth (C). The data points shown represent a range of soil
water content from field capacity to approximately 50% available soil water
depletion.

104
Table 5-2. Regression analysis of estimated ETC to calculated ET0 ratio by mean soil
water content, soil water potential, and available soil water depletion in the upper 0.5 and
1.0 m of soil for either the irrigated zone or total allocated tree area.
Soil Water Content Cubic Function2
Yo
a
b
c
R2
RMSE
P
0.5 m irrigated
8.32
-398.0
6626
-35004
0.52
0.12
<0.0001
1 m irrigated
3.02
-145.6
2695
-14903
0.36
0.14
<0.0001
1 m total
4.15
-209.6
3876
-21806
0.42
0.13
<0.0001
Soil Water Potential
- Exponential Decay Functiony
Yo
a
b
R2
RMSE
P
0.5 m irrigated
0.51
1.12
0.14
0.44
0.13
<0.0001
1 m irrigated
0.41
1.16
0.11
0.38
0.13
<0.0001
1 m total
0.37
1.33
0.10
0.46
0.13
<0.0001
Available Soil Water Depletion Logistic Function*
Yo
Xo
A
B
R2
RMSE
p
0.5 m irrigated
0.55
26.06
0.40
3.44
0.51
0.12
<0.0001
1 m irrigated
0.02
81.10
0.93
0.92
0.25
0.15
<0.0001
1 m total
0.34
34.70
0.69
1.09
0.27
0.15
<0.0001
z Y = Y0 +aX+bX2 +cX3
where X =
- ETc/ETo, and Y0, a, b, and c are regression
coefficients
y Y = Y0 + aexp bX where X = ET x Y=Y0 +
a
1 +
where X = ETc/ETo, and Y0 a, and b are regression coefficients

105
Fig. 5-6. Comparison of estimated crop evapotranspiration (ETC) with calculated
reference evapotranspiration (ET0) ratio. This value represents an approximation
of Kc for observations when soil water content values were near field capacity. Kc
values are expressed as a function of day of year (DOY).
Table 5-3. Regression analysis of estimated ETC to calculated ET0 ratio by day of year for
soil water content values greater than 0.070 cm3 cm'3 (field capacity) using a quadratic
function2.
Yo
a b R2
RMSE
P
DOY
0.71
0.004 -0.00001 0.76
0.056
<0.0001
z Y = Y0 + aX + bX2 where X = ET/ETo, and Y0 a, and b are regression coefficients
the course of a season, daily ETC values were multiplied by the appropriate Kc value
estimated for the DOY. ETc*Kc to ET0 ratios were then calculated to approximate K* and
were plotted against ASWD and (Figs. 5-7 and 5-8). Lines indicating estimations of Ks
using equation 5-2 and 15% depletion as an estimate for RAW are included in Figs. 5-7
and 5-8. Estimated Ks values based on ASWD are presented in Table 5-4. Table 5-5
shows higher R2 values and lower RMSE values compared with ETC to ET0 ratio

106
c
92
o
¡E
o
O

(O
£
o5
u.
O)
¡
o
W
S
c/>
OI
Soil Water Potential (kPa)
Fig. 5-7. Estimated soil water stress coefficient Ks as a function of soil water potential in
the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total tree area to a
1 m depth (C). The data points shown represent a range of soil water content
from field capacity to approximately 50% available soil water depletion.

107
1.2-
0 10 20 30 40 50 60
Available Soil Water Depletion (%)
Fig. 5-8. Estimated soil water stress coefficient Ks as a function of available soil water
depletion in the irrigated zone to a 0.5 m depth (A), 1 m depth (B), and the total
tree area to a 1 m depth (C). The data points shown represent a range of soil
water content from field capacity to approximately 50% available soil water
depletion.

108
Table 5-4. Estimated soil water coefficient (Ks) values for a range of percentage available
soil water depletion (ASWD) using Equation 2 as reported by Allen et al., (1997). A
value of 15% ASWD is used for RAW, therefore estimated K* for ASWD less than 15%
are assumed to equal 1.0.
ASWD
(%)
Estimated
Ks
0
1.00
10
1.00
20
0.94
30
0.82
40
0.71
50
0.59
60
0.47
70
0.35
80
0.24
90
0.12
100
0.00
Table 5-5. Regression analysis of estimated soil depletion factor (K*) by mean soil water
potential, and available soil water depletion in soil 0.5 or 1.0 m deep and in either the
irrigated zone or total tree area.
Soil Water Potential Logistic Function2
Yo
Xo
a
b
R2
RSME
P
0.5 m irrigated
0.58
11.50
0.44
2.97
0.53
0.086
<0.0001
1 m irrigated
0.61
11.49
0.41
3.66
0.47
0.091
<0.0001
1 m total
0.65
12.25
0.34
5.11
0.47
0.091
<0.0001
Available Soil Water Depletion Logistic Function2
Yo
Xo
a
b
R2
RMSE
p
0.5 m irrigated
0.40
42.37
0.58
1.98
0.63
0.076
<0.0001
1 m irrigated
0.48
34.85
0.48
1.55
0.39
0.097
<0.0001
1 m total
0.60
28.70
0.38
1.86
0.32
0.103
<0.0001
Z
Y= Y0 +
where X = Ks, and Yo a, and b are regression coefficients

109
equations presented in Table 5-2. RMSE values in Table 5-4 were generally 25% lower
than those in Table 5-2. The R2 values were greater for equations made using ASWD and
<|> in the irrigated zone to a depth of 0.5 m compared with those using the irrigated area
and total tree area to a 1 m depth. This result indicates that the soil volume with greater
root length densities dried out faster resulting in better correlation of Ks with both ASWD
and 4. Estimated values for K* are approximately 1 at 9fc
Soil Water Uptake per Unit Root Length
Equations resulting from regression analysis of soil water uptake per unit root
length density against mean daily are presented in Table 5-6. An exponential decay
model typically resulted in the best fit. Water uptake per unit root length values were
remarkably similar for the various locations and depths, with the exceptions of the 10 cm
depth between-rows and the 40 and 80 cm depths at the dripline. Maximum water uptake
per unit root length of 0.4 mm3 cm'1 d'1 occured at 9fc or approximately -5 kPa. Water
uptake per unit root length decreased rapidly as 4> decreased to approximately -12 or -13
kPa, then gradually decreased from 0.1 to 0.05 mm3 d1 cm'1 as kPa. A wide scatter in the data along with a long shallow sloping tail resulted in a
relatively low R2 and generally high RMSE values. However, all regressions were
significant at the P=0.01 level and formed an approximation of water uptake at given
bulk 4 within the constratints of the RMSE.
Water uptake per unit root length values of approximately 0.8 mm d'1 cm'1 at 9fc
were found at the 10 cm depth between-rows. This value is double that at other locations
and depths (data not shown). The increase in water uptake could be explained by water

110
Table 5-6. Regression analysis of estimated soil water uptake per unit root length density
on soil water potential in soil at three locations and for the 10, 20, 40 or 80 cm depths
using an exponential decay model2.
Location
Depth
(cm)
Yo
a
b
R2
RMSE
(mm d'1 cm1)
P
Under-Canopy
10
0.07
1.08
0.29
0.23
0.14
<0.0001
20
0.43
-0.29
0.05
0.13
0.35
0.0895
40
0.22
15.46
0.50
0.30
0.28
<0.0001
80
0.25
71.70
0.98
0.14
0.30
<0.0001
Dripline
10
0.16
7.89
0.49
0.20
0.26
<0.0001
20
0.14
4.69
0.36
0.24
0.36
<0.0001
40
0.60
4.74
0.15
0.62
0.24
<0.0001
80
0.33
40.75
0.61
0.29
0.30
<0.0001
Between-Rows
10
0.04
1.24
0.08
0.22
0.33
<0.0001
20
0.14
5.28
0.33
0.20
0.33
<0.0001
40
0.03
9.03
0.39
0.35
0.24
<0.0001
80
0.21
13.74
0.30
0.18
0.52
0.0730
2 Y = Y0 + a exp ** where X = root length density, and Yo a, and b are regression
coefficients
use from non-crop species in the row middles that were not present beneath the tree
canopy. However, there was also elevated water use at the 40 and 80 cm depths at the
dripline. These data are assumed to be the result of higher than expected root length
densities at these depths compared with mean root length density data from similar trees.
Therefore, the higher than expected water use per root length at the soil surface between
rows could be a combination of both greater root densities and water use by weed and
grass species.
Discussion
Citrus water uptake followed relatevely consistent patterns during the 2 years of
this study. Daily water withdrawals from the soil followed daily calculated ET, with
higher values occuring during the summer and lower ones in winter. ETC values were
consistently lower than ET0 except during summer months when 0 was near 0fc. These
0?
V
V' V'

Ill
trends follow reported ETC values for citrus under both humid and arid climatic
conditions (Boman, 1994; Castel et al., 1987; Doorenbos and Pruitt, 1977; Martin et al.,
1997; and Rogers et al., 1983). Reported IQ values for central Florida ranged from
approximately 0.6 in the winter to 1.1 in summer (Boman, 1994; Fares and Alva, 1999;
Rogers et. al., 1983). Estimated mean daily IQ values ranged from 0.55 to 1.2 for citrus
under semi-arid to arid conditions (Doorenbos and Pruitt, 1977; Hoffman et al., 1982;
Martin et al., 1997; and Wiegand et al., 1982). Thus, seasonal mean ETC to ET0 ratios
reported for this study fall in the range of values documented in the literature.
Allen et al. (1997) refered to estimated Kc values from soil water content
measured several days apart as time-averaged Kc and stated that these values are affected
by the evaporative power of the atmosphere. They further stated that the higher the
evaporative power of the atmosphere, the faster the soil will dry between water
applications, and the smaller the time-averaged IQ will be. The reduction in ETC with
lower 0 and reflected by Ks decreasing from 1 to 0.6 as ASWD increased from 10 to
50% seems rather extreme. However, Rogers et al. (1983) suggested that lower estimated
Kc values in the spring were caused by low rainfall and low 0 outside the irrigated zone.
Their reported Kc values of0.77, 0.72, and 0.95 for March, April and May are 81.3, 71.3,
and 89.6% of the Kc values estimated using the regression equation in Table 5-3.
Estimated IQ during the rainy season months of June and July were 101.8 and 92.2% of
calculated values using the same equation, indicating that IQ values esi mated from
monthly water balances can lead to lower estimates of IQ during periods of little rainfall
and high evaporative demand.

112
Allen et al.(1997) indicated that a region of readily available water exists between
0fc and approximately 30 to 50% ASWD for loam and loamy clay soils where there is
essentially no stress to the crop. This estimate is considerably reduced, however, in the
case of citrus on very sandy soils and may only amount to 10 to 15% of ASWD
Estimates for Ks using equation 5-2 closely approximate measured Ks values presented in
Fig. 5-5. It is therefore concluded that Ks decreased to approximately 0.6 at 50% ASWD,
which translates to a reduction of 40% in ETC between field capacity and 50% ASWD.
Koo (1963,1978) determined that stress associated with soil water depletion greater than
33% during periods of bloom, fruit set, and rapid vegetative growth in the spring months
can reduce potential yield, while depletions of 66% can be tolerated during sumer, fall
and winter months. Thus, crop stress associated with K values of 0.8 and 0.4 should b
used for irrigation scheduling from February through June and from June through
January, respectively, to maximize yields while minimizing water use.
Zaongo et al. (1994) reported close correlations beween water uptake and root
length density in millet and grain sorghum. Bland and Dugas (1989) estimated maximum
water uptake of cotton to be approximately 5 mm3 d'1 cm1 root. Hamblin and Tennqant
(1987) found that mean water uptake of cereals and grain legumes was less than 1 mm3 d
1 cm'1. Thus, soil water uptake per unit root length of 0.1 to 0.4 mm3 d'1 cm'1 root
observed for citrus in this study are similar to published values for other crops, which is
somewhat surprising considering the differences in root morphology of annual versus
perennial crops. Soil water uptake rates were closely related to root length densities, thus
soil regions containing higher root length densities will dry out at a proportionally higher

113
rate. Hence, a model of soil water uptake and depletion based on root length densities
would be appropriate for citrus.
The test hypotheses established for this experiment relating water uptake to time
of year, soil water content, and root density has been confirmed. A model based on the
concepts of Kc and Kg to estimate daily citrus water uptake is reasonable. The estimation
of soil water uptake and resulting depletion based on root length density is sound and
would provide a reasonable soil water balance for a nutrient management expert system.
Conclusions
Based on the results from this study it is concluded that ETC can be calculated by
modifying ET values for crop and residual soil moisture conditions using appropriate K
and K* coefficients. Minimum Kc was approximately 0.85 in December and January,
while a maximum of approximately 1.05 occurred during the months of June and July.
Soil water use decreased with soil water content, resulting in Kg values of 1.0 at nearly
15% ASWD to 0.6 at 50% ASWD. With few exceptions, daily soil water uptake per unit
root length density was similar for all soil layers. The best correlation between daily
water use and soil water content was found in the soil volume containing the highest root
length density. Therefore, the hypothesis that soil water uptake relative to calculated
reference evapotranspiration is related to season of year, soil water content and root
length densities was confirmed. Estimation of soil water uptake and resulting soil
depletion based on root length density would allow for a relatively accurate assessment of
soil water depletion, crop water status, and effective soil storage capacity using a layer
soil profile modeling approach. Such approach would allow model users to predict soil
,4
O'
/
c*-

114
water depletion throughout a soil profile and assess effective soil water storage capacity
and potential leaching of nutrients associated with rainfall and/or irrigation.

CHAPTER 6
CITRUS NITROGEN UPTAKE AND CYCLING
Introduction
The optimum timing, frequency, and rate of fertilizer N application for citrus
production under Florida conditions have been explored for nearly 60 years. Nitrogen
best management practices (BMPs) have been established for citrus on the sandy soils of
central Florida. The goals of these practices are to sustain high fruit production and tree
health, and improve N use efficiency of citrus while reducing the impact of N leaching on
ground water quality. The BMPs restrict the annual rate of N fertilizer than can be
applied, and base it on tree age or past production. The timing of N application is
restricted to the drier seasons of the year to reduce potential leaching. A N balance model
for citrus must be developed to predict the effects of these restrictions on citrus
production.
Koo (1979) found a significant N rate and irrigation interaction in high density
planting using low volume microsprinkler irrigation. Koo (1980) later found no yield
difference between dry fertilizer N applications and fertigation through the same low
volume system. In the same experiment, no yield differences were found when 10
fertigation applications were compared with three dry applications per year. Recent
studies by Alva and Paramasivam (1998) and Wheaton (unpublished) have shown similar
results. Koo (1986) and Boman (1993) found no difference in yield when comparing
multiple applications of dry soluble fertilizer with controlled release N sources.
115

116
Syvertsen and Smith (1996) reported that leaching losses were generally small (2
to 9%) from low and medium N rate treatments in a lysimeter study, except when N
application coincided with frequent and/or intensive rainfall events. They also concluded
that 28% of applied N might have been lost in planted ly si meters due to volatilization
and/or denitrification. Immobilization into organic matter by soil microbes was not
considered a significant mechanism of N removal due to the very low organic matter
content of the Entisol used in the lysimeters. Estimated N uptake was 61% for the high N
application rate of 1.6 kg tree1 yr"1 and 83% for the lowest application rate of 0.3 kg tree"
1 yr"1. These values are similar to a previously estimated NUE of 68% (Syvertsen et al.
1993). Lea-Cox and Syvertsen (1996) and Scholberg et al. (2002) reported similar
findings of lower NUE with higher N application rate in a greenhouse studies.
Scholberg et al. (2002) found N uptake of greenhouse-grown seedlings to be
proportional to soil temperature, potential ET, and canopy biomass. Overall N uptake
increased with residence time in the root zone. Alva and Paramasivam (1998) reported
improved N use efficiency (0.36 to 0.39 Mg*1 fruit per kg N applied) of field grown
citrus, which was substantially greater than that reported by Koo and Smajstrla (1984)
(0.23 Mg'1 fruit per kg N supplied). This improvement was attributed to incorporation of
dry and fertigated nutrients under the canopy with light irrigation, applying no fertilizer
during the rainy season (between June and August), and maintaining adequate but not
excessive soil water content to 90 cm depth.
There is a positive relationship between the concentration of a nutrient in the soil
solution at the root surface and its uptake rate by plants. Passive nutrient uptake is
defined as the amount of a nutrient taken into a plant as solute associated with water

117
uptake. The active uptake of nutrients across a membrane is enzyme-catalyzed, thus the
relationship of uptake rate to nutrient concentration is hyperbolic with a maximum uptake
rate at the nutrient concentration where available enzymes are saturated. The Michaelis-
Menten equation (Equation 6-1) is often used to estimate uptake rates for crop plants
under given nutrient concentrations.
I Imax Cl, /(Km+CLa)
Equation 6-1
Where:
2 1 A V
I = inflow flux of nutrient (mol cm s ),
Imax = maximum active flux (mol cm'2 s'1),
Cu = nutrient concentration in the soil solution at the root surface (mol cm'3),
Km = Cl, value at Imax/2 (mol cm'3),
Numerous reports suggest that actively growing tissues such as young developing
leaves and fruit constitute the strongest sink for N uptake (Dasberg, 1987; Feigenbaum et
al., 1987; Legaz et al., 1982). Accumulated N was found primarily in fruitlets and newly
developed leaves and twigs. Absorption rates increased from the beginning of growth and
flowering, reached a maximum at the second shoot growth flush (July), and then declined
through dormancy. Only about 20 to 30% of the new leaf and fruit N originated from the
labeled source, suggesting considerable redistribution from stored reserves.
Mooney et al. (1992) observed an N concentration gradient between the roots,
trunk and branches of citrus trees in New Zealand. High concentrations were found in the
branches, with lower concentrations in the roots. Nitrogen concentrations in the trunk
were highest at bud break and declined steadily through fruit set and development with a
minimum at harvest. Kato et al (1982) found that total N content decreased in both bark

118
and wood during the sprouting period of 21 year-old Satsuma mandarins. Greatest
decreases in N were found in parts with higher concentrations of N (e g. leaves, shoots,
and fine roots). They also concluded that the trunk and large roots were main N reservoirs
for new shoot development.
Due to the perennial nature of citrus, leaf twig and branch biomass accumulated
in previous years periodically abscises. Wallace et al. (1945) estimated that citrus leaves
function on the tree up to 18 to 24 months before senescence. They found an average of
18.1 kg tree1 dry matter loss per year from mature Valencia orange trees grown in
California. Dry matter losses were 9.1,4.0, and 4.9 kg tree'1 for leaves, twigs, and
branches, respectively.
Information on N uptake rates and N cycling for the development of seasonal N
balance under Florida conditions on a field scale is lacking and will be critical to assess N
application quantity, frequency and timing decisions. In order to improve our
understanding of the underlying processes, the following hypotheses were tested: 1)
seasonal N uptake rates are related to leaf N status, 2) fertilizer-N is rapidly converted
into NO3-N that can be readily leached from typical ridge soils, 3) leaf N
concentrations are lowest during periods of high growth rate due to N dilution in the dry
matter, 4) changes in tree N reserves account for the majority of N in new leaves, and 5)
tree biomass and N senescence follow predictable seasonal patterns. The main goal of the
current studies was to provide critical information needed for a citrus N budget for a
citrus production system under Florida conditions. The objectives of this study were to 1)
determine seasonal changes in N uptake rates for citrus, 2) quantify changes in residual
soil N and nitrification with time in the absence of citrus roots, 3) measure seasonal

119
changes in plant tissue N concentration, and 4) determine cumulative tree biomass and N
losses for citrus during a 2-year period.
Materials and Methods
Site Characteristics
Fourteen-year-old Hamlin orange on Carrizo citrange and Swingle citrumelo
rootstocks at the same location as Experiment 1 in Chapter 3 were used for the three
experiments presented in this chapter. The trees had been fertigated at an annual rate of
179 or 269 kg N ha'1 at approximately monthly intervals from February to October using
equal split applications for the 3 years prior to the start of this experiment. Irrigation was
applied by an automated irrigation system using switching tensiometers to trigger
irrigations. Irrigation was applied when soil water potential in the upper 30 cm dropped
below -10 kPa during the bloom and fruit set period of February to May and -15 kPa for
the remainder of the year. Reclaimed water containing 7 mg L'1 or less of NO3-N was
used for all irrigations. The soil type at the site was Candler fine sand (hyperthermic,
uncoated Quartzipsamments) with water holding capacity of 0.05 to 0.08 cm3 cm'3 and
cation exchange capacity of less than 5 cmol kg'1.
Experiment 1 Nitrogen Uptake Flux
Fertilizer rates and applications
The N fertilizer rates used in this study were approximately 50 and 100% of the
monthly rate based on 269 kg N ha'1 yr'1 in 6 monthly applications, or 45 kg N ha'1 per
application. The 100% rate is equivalent to 500 g N tree'1 yr'1 or 83 g N tree'1 per
application. The reduction in N application rate was accomplished by reducing the
application time to four representative trees of each rootstock using valves in the

120
microsprinkler supply lines. After each uptake study, a quantity of N equal to the N
reduction was applied to those trees receiving the reduced N rate.
Soil sampling procedures
Fifty 1.27 cm diameter polyvinyl chloride (PVC) pipes were inserted to 45 cm
depth beneath the canopy of eight Hamlin on Carrizo and eight Hamlin on Swingle
trees at least 2 weeks prior to each study, assuring that all roots within them would die
prior to fertigation treatment applications. The pipes were arranged in three semicircles
25, 75, and 125 cm from each tree trunk. The ratio of the lengths of these arcs was 1:3:6.
The number of pipes in each arc was proportional to its length, resulting in 5, 15, and 30
pipes in each arc. Soil from these pipes was used as a control to estimate loss of N by
volatilization and immobilization in the absence of roots. Changes in NH4-N and NO3-N
concentrations were used to determine nitrification rates under field conditions.
A composite soil sample consisting of 10 cores taken with a 2 cm diameter auger
were removed from each tree at three depth increments of 15 cm each. Samples were
taken from the same arcs and ratios where the PVC pipes were installed. Samples were
taken 0 h, 1 h, 1 d, 2 d, 3 d, and 4 d after N fertilizer application. The timing of fertilizer
applications corresponded with growth phases of the citrus tree when maximum N uptake
was most likely: early March (bloom and first spring flush), mid May (second spring
flush and fruit expansion), and September (third flush and fruit maturation).
Analytical methods
All soil samples were placed in an insulated cooler containing ice and placed into
a refrigerator at 4 C or less until extractions could be made. Extractions using
approximately 4 g of soil and 40 mL of 2 M KCI were analyzed to determine soil nitrate
(NO3-N) and ammonia (NH4-N) concentrations (Keeney and Nelson, 1987). Between 4.1

121
and 4.3 g of wet soil were placed into a centrifuge tube and the mass was recorded. Forty
mL of 2 M KC1 was placed into each tube immediately after the soil was weighed. The
tubes were placed in a shaker for 1 h. The solution was filtered into vials that were then
capped. Extracts were refrigerated at 4 C until they could be analyzed for NO3-N and
NH4-N. Extracts were analyzed using a model FS3000 Rapid Flow Analyzer (O I
Analytical, College Station, Texas). USEPA methods 351.2 and 353.2 were used for
ammonium and nitrate analysis, respectively. The remaining soil sample was used to
determine gravimetric soil water content of the soil sample extracted.
The N concentration in the soil on a dry weight basis was determined. Volumetric
soil water content of each sample was used to determine N concentration in the soil
solution. Total N, NH4-N, and NO3-N contents in the soil by sample depth were
estimated using the soil solution N concentration and area of the irrigation emitter. Daily
NH4-N, and NO3-N change for each sample depth was determined by comparing daily N
content estimates for soil inside and outside of the pipes.
Experiment 2 Seasonal Tissue N Concentration
Tissue samples were collected from three replicates of 14-year-old Hamlin
orange on Carrizo citrange and Swingle citrumelo trees fertilized at an annual rate of 179
or 269 kg N ha1. Samples were taken at approximately 6 week intervals corresponding to
specific stages of growth through the year for two seasons. The growth stages used were
1) bloom and spring flush (early March), 2) fruit drop and second vegetative flush (mid
May), 3) first summer flush (early July), 4) second summer flush (mid August), and 5)
pre-harvest (mid October). Tissue N concentrations from these samples were determined
and used to estimate seasonal change in N concentrations.

122
Tissue samples collected
The following tissues were sampled at each growth stage: 1) expanding leaves, 2)
expanded leaves, 3) twigs (<7 mm), 4) small limbs (7 to 15 mm), 5) medium limbs (15 to
30 mm), 6) large limbs (>30 mm), 7) trunk, 8) feeder roots (<4 mm), and lateral roots (>4
mm). Two trees from each of the 12 plots were selected for each sampling period. Fifty
non-expanded leaves per tree were collected from the last flush, and 50 expanded leaves
per tree were also collected. Thirty twigs and small (7 to 15 mm) branch segments of 15-
30 cm in length were collected. Cylinders of tissue 10 mm in diameter and 15 mm long
were removed from branches and trunks greater than 15 mm in diameter using a plug
cutter and battery powered drill. Eight plugs from each of four limbs of 15 to 30 mm and
>30 mm diameter, and two trunks were collected on each sampling date. Branch and
trunk samples were separated into bark and wood components. Twelve fruits were
collected on each sampling date. Three soil cores per tree were taken for root removal
and roots were separated by size and depth, (size = <4 mm and >4 mm; depth = 0 to 15
cm, 15 to 30 cm, 30 to 45 cm, 45 to 60 cm, and >60 cm).
Tissue analysis
All fresh tissue samples were weighed, dried for 3 days at 70 C, reweighed, and
ground for nutrient analysis. Tissues were analyzed for total N using the same grinding
and Kjeldahl methods described in Chapter 3. Fruit diameters and leaf area were
measured prior to drying.
Experiment 3 Seasonal N loss
Catch frames 0.9 m wide x 1.5 m long were placed under one tree in each of the
12 plots used for seasonal N concentration determination. Any citrus plant material

123
falling onto the frames was collected at approximately 2-week intervals for two seasons
(2001 and 2002). The plant material was separated into 1) flowers, 2) fruit, 3) twigs, and
4) leaves. Material in each of the four categories was counted, dried, weighted, and
analyzed for total Kjeldahl N using the same grinding and analytical procedures
described in chapter 3.
Assuming the material collected in the catch frames was proportional to the
amount of material under the entire canopy, the biomass and N concentration of each
tissue was multiplied by the ratio of the area under the canopy to the area of the catch
frame. Cumulative seasonal N loss was determined for each tissue.
Results
Nitrogen Uptake
Irrigation after sampling on the third day after fertilizer application likely leached
a portion of the N below the 45 cm sampling depth. Therefore, soil N losses were
calculated for the control pipes and bulk soil for 1,2, and 3 days after application.
Differences in mean cumulative N loss from the soil during the 3 days after application
were significant at the P=0.05 level for month of year, with greatest loss occurring in
May. Soil N losses by rootstock and application rate are presented in Tables 6-1 to 6-3
for the months of March, May, and September, respectively. Mean percentage of total
soil N loss during the 3-day period for the high application rate of approximately 83 g
tree1 were 60.7, 68.6, and 63.4% for March, May, and September, respectively. These
means were significantly different at the P=0.05 level from mean percentage N losses for
the lower rate which had respective values of 83.6, 82.6, and 73.1%. Mean cumulative N
soil losses for Hamlin trees on Carrizo citrange were 70.5, 78.8, and 70.1% for the

124
months of March, May, and September, respectively and were not significantly different
from mean cumulative losses for trees on Swingle citrumelo, which were 73.8, 72.5, and
66.4% for the same months.
Nitrogen losses from control pipes varied greatly across the three studies but were
not significantly affected by application rate, month, or rootstock. Mean cumulative N
loss was 22.6%, ranging from a net gain of 7.0% to a loss of 41.0% for the 3 days after
application. Estimated daily maximum N uptake was determined by subtracting the N
lost from the control pipe from the N lost from the bulk soil. Cumulative daily maximum
N uptake was significantly different at the P=0.01 level by application amount, but not
significantly different by month, and rootstock (Tables 6-1 to 6-3). Maximum uptake as a
percentage of amount applied averaged 46.7 and 61.7% for the high and low application
rates, respectively. Mean cumulative maximum uptake values for Carrizo and Swingle
were 53.9 and 54.4%, respectively.
Estimated total and passive uptake as a function of mean soil solution N concentration is
presented in Figs. 6-1 and 6-2. The equation constants, R2, RMSE, and P values for the
regression of these data are presented in Table 6-4. A great deal of scatter exists in the
data due to the large range in N loss from the control pipes. Therefore, the small R2
indicates that soil solution N concentration explained only 46% of the variation in the
data. While both regressions are significant at the P=0.01 level, the relatively large
RMSE values would result in a large 100(l-a)% confidence interval. Using the estimated
maximum N uptake and soil solution concentration relationship in Figure 6-1, the
Michaelis-Menten equation (Equation 1) constants were approximately 14.5 g N tree'1 d'1
and 60 mg N L'1 for and Km, respectively.

125
Table 6-1. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in March, 2002. Rootstocks are Carrizo citrange, and
Swingle citrumelo; high N application rate was 269 kg ha'1 yr'1, low rate was 134 kg ha'1
yr1.
Days After
Weighted
N Uptake
Cumulative
Total
Cumulative N
Loss
Application
Solution
N
Passive Active Total
N Uptake
Control Soil
(mg L'1)
(g tree'1 d'1)
(gtree1) (%)
(% applied)
Carrizo-
High rate
1
165.7
5.1
5.1
10.2
10.2
13.9
12.4
26.3
2
152.7
4.5
5.2
9.8
20.0
27.3
27.4
40.8
3
115.8
2.2
7.8
10.0
30.0
41.1
37.3
59.4
Carrizo-
Low rate
1
80.0
2.9
12.0
14.9
14.9
41.5
0.9
42.4
2
50.6
1.8
5.6
7.4
22.3
62.2
19.6
73.9
3
31.0
0.6
1.8
2.4
24.7
68.8
24.4
81.7
Swingle -
High rate
1
183.8
5.7
9.0
14.7
14.7
17.9
7.1
25.0
2
181.3
5.4
4.2
9.6
24.3
29.7
7.8
47.3
3
167.1
3.0
10.4
13.4
37.7
46.0
5.4
62.1
Swingle -
-Low rate
1
163.5
5.3
13.8
19.1
19.1
27.2
0.7
28.1
2
124.1
4.5
11.6
16.1
35.2
50.5
16.6
62.6
3
65.8
1.6
10.2
11.8
47.0
67.4
32.7
85.6

126
Table 6-2. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in May, 2002. Rootstocks are Carrizo citrange, and
Swingle citrumelo; high N application rate was 269 kg ha1 yr'1, low rate was 134 kg ha1
yr1.
Days After
Application
Weighted
Solution
N
(mgL1)
N Uptake
Passive Active Total
(g tree1 d1)
Cumulative
Total
N Uptake
(g tree1) (%)
Cumulative N
Loss
Control Soil
(% applied)
Carrizo High rate
1
156.1
4.6
5.3
9.9
9.9
17.1
8.6
25.7
2
128.7
3.8
8.1
11.9
21.8
37.5
23.2
46.9
3
88.1
1.3
6.1
7.4
29.3
50.3
36.0
65.2
Carrizo
- Low rate
1
103.3
4.1
17.1
21.2
21.2
42.6
17.7
60.4
2
46.5
1.6
9.7
11.2
32.5
65.2
33.5
89.2
3
20.9
0.3
1.7
2.0
34.4
69.2
26.8
92.5
Swingle High rate
1
195.7
5.6
12.4
18.0
18.0
23.9
-4.5
23.9
2
179.2
4.8
10.5
15.3
33.3
44.1
2.1
59.1
3
139.1
2.2
6.2
8.5
41.8
55.3
5.5
72.0
Swingle -
Low rate
1
101.1
4.0
11.1
15.2
15.2
31.4
13.6
45.0
2
44.3
1.1
4.9
6.0
21.1
43.7
28.6
65.6
3
22.8
0.3
1.1
1.4
22.5
46.6
41.0
72.7

127
Table 6-3. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in September, 2002. Rootstocks are Carrizo citrange,
and Swingle citrumelo; high N application rate was 269 kg ha'1 yr'1, low rate was 134 kg
ha'1 yr'1.
Days After
Application
Weighted
Solution
N
(MgL1)
N Uptake
Passive Active Total
(g tree'1 d~J)
Cumulative Total
Cumulative N
Loss
N Uptake Control Soil
(gtree1) (%) (% applied)
Carrizo High rate
1
187.8
5.8
4.9
10.7
10.7
17.8
0.1
17.9
2
131.6
4.6
5.4
10.1
20.8
34.4
4.1
38.0
3
91.9
3.4
3.7
7.1
27.9
46.3
18.5
58.6
Carrizo
- Low rate
1
102.6
3.3
4.8
8.1
8.1
24.9
4.9
29.8
2
61.7
2.2
2.7
4.9
13.0
39.8
10.2
48.5
3
28.7
1.5
1.1
2.7
15.6
48.0
19.0
81.8
Swingle
- High rate
1
201.0
7.8
8.4
16.1
16.2
19.0
22.5
41.5
2
121.1
4.1
5.8
10.0
26.1
30.7
25.1
54.8
3
90.2
3.4
5.6
9.0
35.1
41.3
31.7
68.3
Swingle
- Low rate
1
118.7
4.2
9.1
13.2
13.2
31.4
1.0
32.4
2
77.4
2.6
8.2
10.8
24.1
57.2
-25.0
40.6
3
55.9
2.2
3.4
5.5
29.6
70.3
-7.0
64.7

128
Fig. 6-1. Relationship of estimated total N uptake as a functrion of soil solution
concentrations. Total N uptake is the sum of passive and active N uptake components.
Dashed lines denote Michaelis-Menten equation constants of Imax and Km.
Soil Solution N Concentration (mg L'1)
Fig. 6-2. Relationship of estimated active N uptake to soil solution concentration.

129
Table 6-4. Regression equations for estimated maximum N uptake and estimated active N
uptake rates by soil N concentration (mg l'1) using an exponential rise to a maximum
model2 and linear modely, respectively.
Y0
a
b R2
Maximum N uptake
RMSE
(g tree'1 d1)
P
-2.44
17.31
0.013 0.46
Active N uptake
3.51
<0.0001
-0.78
0.06
0.63
3.41
0.013
z Y = Y0a(l exp bx) where X = soil N concentration, and a, and b are regression
coefficients
y Y = Y0 + aX where X = soil N concentration, and a, and b are regression coefficients
Passive N uptake was estimated by determining daily water uptake from the
irrigated area to a depth of 45 cm using the water uptake equations presented in Chapter
5. Soil solution N contents were estimated for the same area and depth using the soil N
and gravimetric soil water content values from the daily soil samples. Passive N uptake
was estimated using the assumption that daily passive N uptake was equal to the product
of soil solution N concentration and estimated daily water uptake. Estimated passive N
uptake was subtracted from the estimated maximum N uptake to estimate daily active N
uptake. The regression of estimated active N uptake and soil solution N concentration
was significant at the P=0.05 level (Table 6-4). Due to the compounded error associated
with the estimation of passive uptake using the regression equations for water uptake, the
associated R2 was lower than that for soil overall N uptake (Table 6-4). However, the
RMSE was similar indicating a confidence interval similar to that of the regression of
daily maximum uptake (Fig. 6-1). The relationship of active uptake and soil solution
concentration was linear over the concentration range used in this study (Fig. 6.2), so no
Michaelis-Menten equation constants could be determined.

130
Nitrification Estimation
Changes in NO3-N and NH4-N content inside the control pipes were used to
estimate N loss and nitrification rates with time. Soil NO3-N, NH4-N, and total N as a
percentage of N applied inside the control pipes are shown as a function of time in Fig.
6-3. Mean NO3-N content in the upper 45 cm of soil increased 24 h after application to
150.6% of NO3-N applied. The mean NO3-N content decreased during the next 48 h to
125.8% of total NO3-N applied. Content of NH4-N decreased to 34.0% of that applied
after 24 h and steadily declined afta- application to 16.1% of NH4-N applied on day 3.
The sum of NO3-N and NH4-N (total N) declined throughout the period to 73.9% of total
N applied. This result indicated that approximately 26.1% of N was lost during the 3-day
period due to volatilization of NH4-N, microbial activity, or incorporation of N into
organic matter. Whether some of the immobilized N would become available at a later
time is unclear. Part of the N immobilization may be associated with recently decayed
root biomass in the control pipes which had relatively low C:N ratio. The length of time
that N would be lost at this rate is unclear and would certainly be greater if the fertilizer
were not incorporated with water. This loss would be emphasized if dry N fertilizer
sources were used. The nitrification rate in this soil was rapid, with a mean of more than
50% of the NH4-N converted to NO3-N within the first 24 h. Nearly 85% of applied NH4-
N was converted to NO3-N in 3 days, assuming all of the N loss was NH4-N.
Seasonal Tissue N Concentration
Leaf and twig N concentrations followed a cyclic pattern during the 2 years of periodic
sampling. Nitrogen values were significantly different at the P=0.05 level by month of
year and application rate, but not by rootstock. Therefore, values for specific N rates were

131
Fig. 6-3. Proportions of nitrate-N, ammonium-N, and total-N from control pipes as
percentage applied during 3 days after application. Initial increase in nitrate-N is
assumed to be from nitrification of ammonium-N. Low soil pH limits
volatilization, thus loss in total-N was assumed to be due to microbial processes.
averaged across rootstocks. Respective N values were near maximum from August to
February with minimums in May of each year (Fig. 6-4). Minimum leaf N concentrations
were 1.9 and 2.0% for the flush and expanded leaves, respectively, for trees fertilized
with 179 kg N ha'1 yr'1. Minimum leaf N concentration from trees fertilized with 269 kg
N ha'1 yr'1 were 2.0 and 2.2% for flush and expanded leaves, respectively. Maximum leaf
N concentrations were 2.5 and 2.7% for 179 and 269 kg N ha'1 annual N application
rates, respectively. Leaf areas for both flush and expanded leaves were generally greater
for leaves from trees receiving the higher annual N rate (Table 6-5). No such trend was

[N] (%)
132
Fig. 6-4. Seasonal change in N concentration for flush leaves, expanded leaves, and twigs
during 2001 and 2002. High N rate (A) and LowN rate (B) equal to 269 and 179
kg ha'1 yr'1, respectively.

Table 6-5. Seasonal changes in N concentration, size and dry wt. of fruit, flush leaves, and expanded leaves during 2001 and 2002
seasons. There were three replicates of each N rate on two rootstocks (Carrizo citrange and Swingle citrumelo) for a total of 12
measurements per date.
Fruit Flush Leaves Expanded Leaves
Date
Collected
[N]
(%)
Diameter
(mm)
Dry Mass
(g fruit'1)
[N]
(%)
Leaf Area
(cm2 leaf1)
Dry Mass
(gleaf1)
[N]
(%)
Leaf Area
(cm2 leaf1)
Dry Mass
(g leaf1)
179 kgN ha*1 yr'1
3/16/01
2.02
10.0
2.75
33.7
5/18/01
1.33
38.7
6.0
1.96
21.1
0.2
2.02
32.4
0.38
7/2/01
1.29
51.0
6.3
2.35
25.0
0.2
2.29
29.7
0.39
8/13/01
1.13
62.3
15.1
2.38
22.8
0.2
2.24
30.0
0.36
10/13/01
1.11
66.5
21.0
2.11
20.2
0.2
2.29
41.4
0.38
4/4/02
2.18
28.5
0.2
2.14
33.8
0.33
5/20/02
1.18
36.9
5.6
1.90
15.7
0.3
2.00
32.6
0.51
7/15/02
0.99
53.4
11.1
2.19
20.3
0.1
2.27
20.1
0.21
9/9/02
1.02
63.4
15.0
2.50
23.2
0.2
2.19
26.2
0.23
10/25/02
0.95
65.4
18.2
2.54
22.6
0.2
2.24
28.8
0.25
269 kg N ha'1 yr'1
3/16/01
2.75
11.7
2.35
35.3
5/18/01
1.34
39.5
6.2
2.06
20.7
0.1
2.13
29.0
0.31
7/2/01
1.39
51.5
6.5
2,28
24.9
0.2
2.35
31.4
0.40
8/13/01
1.10
62.1
14.9
2.42
20.6
0.1
2.37
35.0
0.36
10/31/01
0.98
64.1
18.9
2.52
23.8
0.2
2.64
49.5
0.55
4/4/02
2.16
29.2
0.2
2.30
32.2
0.34
5/20/02
1.26
35.8
5.1
1.92
17.2
0.3
2.25
31.9
0.51
7/15/02
0.97
54.0
11.5
2.22
20.0
0.1
2.40
22.5
0.24
9/9/02
1.15
63.9
15.1
2.44
23.8
0.2
2.53
28.7
0.25
10/25/02
1.13
67.3
17.8
2.53
22.9
0.2
2.48
30.3
0.25

134
apparent for specific leaf weights. Minimum twig N concentrations in May were 0.83%
for both annual N application rates. Maximum twig N concentrations were 0.98 and 1.02
% for the 179 and 269 kg N ha'1 annual application rates, respectively. These mximums
occurred in August.
Branch bark N concentrations remained within a narrow range from 1.0 to 1.3%
during the 2-year period (Fig. 6-5). Mean branch bark N concentrations were 1.04 and
1.17% for low and high N application rates, respectively. Minimum values of 1.00 and
1.11% occurred in May or July of each year. Maximum bark N concentrations were 1.09
and 1.22% for low and high annual application rates, respectively. These maximum
values occurred in October and January. Branch wood N concentrations were lower, but
followed similar trends as those of branch bark tissue (Fig. 6-5). Mean wood N
concentrations were 0.25 and 0.31% for the low and high N application rates. Maximum
wood N concentrations were 0.37 and 0.38% for low and high N application rates,
respectively, and occurred in January and March. Minimum wood N concentrations were
0.23 and 0.29% for low and high N application rates, respectively. These mnimums
occurred in October.
Root N concentrations were greater for roots <4 mm in diameter than for roots >4
mm. Mean N concentrations for roots <4 mm in diameter were 1.35 and 1.34 % for high
and low N application rates, respectively. Mean N concentrations for roots >4 mm in
diameter were 0.85 and 0.89% for high and low N application rates, respectively.
Seasonal trends of N concentration for roots were not as consistent as with other plant
tissues.

135
1.4 -
12 -
1.0 -
0.8 -
0.6
0.4 -
0.2-
X o.o
2-1.4
1.2 -I
1.0
0.8-
0.6 -
0.4 -
0.2-
0.0
Small limbs Bark
Small limbs Wbod
Medium limbs Bark
Medium limbs Wbod
large limbs Bark
Large limbs Wax!
Small Limbs Bark
Small limbs Wbod
Med urn limbs Bark
Medium limbs Wbod
Large Limbs Bark
large limbs Wood
ri 1
2/1/01 5/1/01
-1i
8/1/01
-ii
11/1/01
-iii*f
2/1/02 5/1/02
B
8/1/02 11/1/02
Date
Fig. 6-5. Seasonal changes in N concentrations for bark and wood tissue of small,
medium, and large limbs during 2001 and 2002. High N rate (A) and Low N rate
(B) equal to 269 and 179 kg ha'1 yr'1, respective!

136
Fruit N concentration was highest in May of each year at 1.25 and 1.30% for low
and high N application rates, respectively (Table 6-5). Nitrogen concentrations decreased
through the season to means of 1.03 and 1.06% for low and high N application rates just
prior to harvest. Mean fruit diameter did not differ due to N application rate. However,
fruit dry masses were greater for the lower application rates.
Seasonal Fertilizer Use Efficiency
Seasonal changes in tree N content from March to May were estimated using the
mean biomass values for trees grown on Carrizo and Swingle rootstocks in chapter 3
multiplied by the mean reduction in N for the period. Nitrogen reduction for leaves was
49.0 and 42.6 g tree'1 for Carrizo and Swingle trees, respectively. This reduction was
relatively large compared with N reduction in twigs and branches of 26.4 and 18.4 g tree*
1 for trees on the same rootstocks. Thus, total estimated N reduction due to reduction in N
concentration from March to May was 75.4 and 61.0 g tree'1 for trees grown on Carrizo
and Swingle rootstocks, respectively. The amount of tree N gained from spring flush and
developing fruit must be added to these values. However, the amount of biomass gained
through spring flush was not measured in these studies. It is the authors observation that
more biomass is gained in the spring of the year compared with summer vegetative
flushes, but the amount of biomass has not been determined. To estimate it, 75% of the
reported biomass gained for these trees in 2001 (chapter 3) was used. Biomass values in
Table 6-5 indicate that approximately 30% of fruit biomass is accumulated prior to the
May sampling. Therefore, 30% of the mean fruit biomass collected prior to the 2002
destructive tree sampling was used as an estimate of fruit biomass gain between March
and May. These estimated biomass values were multiplied by the May periodic sampling

137
N concentration values to estimate the N content of these added biomasses. Estimated
leaf and twig flush gains were 108.8 and 95.2 g tree'1 for trees grown on Carrizo and
Swingle rootstocks, respectively. Fruit N gains amounted to 90.6 and 77.4 g tree'1 for the
same rootstocks. Nitrogen mass associated with blooms and abscised fruit must also be
added and was assumed to be equal to the N loss in blooms and fruitlets. Therefore, the
total estimated tree N content increase between March and May was 227.4 and 200.6 g
tree'1 for trees on Carrizo and Swingle rootstocks, respectively. These N gains would
represent a fertilizer use efficiency of 80.8 and 71.3% for Carrizo and Swingle trees,
respectively.
Hamlin trees gain little leaf and fruit biomass after October of each year.
Therefore, using the same methods of estimation with 100% of annual leaf twig, and
fruit biomass gain from Chapter 3, 90.7 g tree'1 loss due to senescence was balanced by
gains of 471 and 403 g tree'1 for trees on Carrizo and Swingle, respectively (Chapter 3).
Total net seasonal N content gain was approximately 383 and 312.3 g tree'1, or fertilizer
N use efficiencies of 68.3 and 55.7% for trees grown on Carrizo and Swingle,
respectively.
Seasonal N loss
Flower, leaf, and fruit were collected from March, 2001 to December, 2002. A
relatively small amount of decaying twig and branch material was found in the catch
frames at any one time. The majority of these tissues remained attached to the tree until
greatly decomposed. Accurate seasonal timing and amount of biomass and N loss from
twig and branch material could not be determined without removing all dead twig and
branch material at the beginning of the study and periodic removal of dead material from

138
many trees. Therefore, only blooms, leaves and fruit were included in this study.
Seasonal cumulative dry masses and N masses for each of these tissues as a function of
time are shown in Figs. 6-6 and 6-7 for 2001 and 2002, respectively.
Bloom dry mass accumulated in March and April of each year. Cumulative
senesced flower weights were 529 and 764 g tree'1 for 2001 and 2002, respectively.
Cumulative flower N masses were 18.8 and 21.9 g N tree'1 for the same years. Fruit and
leaf biomass losses were greatest from April through May and September through
December of each year. Little loss of either of these tissues occurred from June through
August.
Cumulative weight loss associated with fruit drop varied more than bloom and
leaf biomass, with 1635 and 946 g removed in 2001 and 2002, respectively. These fruit
biomass values represented a cumulative loss of 30.1 and 16.8 g of N tree"1 for these
years. Leaf biomass losses were 2041 and 2806 g tree'1 for 2001 and 2002, respectively.
Cumulative annual N loss from leaf fall amounted to 42.2 and 51.4 g tree'1 for the same
years.
Discussion
Seasonal N uptake varied by month of year, with May having the highest uptake rate of
the 3 months studied. This result was consistent with other uptake studies using young
citrus trees in lysimeters (Syvertsen and Smith, 1996; Lea-Cox et al., 2001), and
greenhouse studies (Lea-Cox and Syvertsen, 1996; Scholberg et al., 2002). Increased
uptake rates could have been due to higher soil temperature (Scholberg et al., 2002) or
greater tree N demand (Dasberg, 1987; Feigenbaum et al., 1987; and Legaz et al., 1982).

139
Date
Fig. 6-6. Seasonal cumulative dry mass (A) and N content (B) of flowers, fruit,
and leaves collected from catch frames under mature citrus trees during
the 2001 season.

140
Fig. 6-7. Seasonal cumulative dry mass (A) and N content (B) of flowers, fruit,
and leaves collected from catch frames under mature citrus trees during
the 2002 season.

141
Mean daily soil temperature at the 10 cm depth was 5 C higher in May compared with
March, 2002. Increased soil N losses in May compared with March and September may
be attributed to higher soil temperatures in May.
High N concentrations in leaf, twig and branch tissues have been determined to be
sources of N for developing vegetative and reproductive tissues (Dasberg, 1987; Kato et
al., 1982; and Legaz et al., 1982). Lower expanded leaf, twig and branch bark N
concentrations prior to fertilizer application in May compared with March and September
of2002 agree with results of Kato et al. (1982) and Legaz et al. (1982). This agreement
provides evidence of greater crop demand for N during this period of time. Reduced N
concentrations in expanding flush leaves may be related to N dilution in the tissue during
periods of rapid growth in May. Fertilizer N was added in six equal amounts, four
applications in the spring (February, March, April, and May) and two after the end of the
rainy season in September and October. Leaf N decreased below optimum concentration
prior to the May fertigation application in both years, suggesting that insufficient uptake
occurred from February to April to satisfy the N demand of developing tissues. Increases
in leaf N concentrations after the May fertilizer application in both years implies that
application of all spring N fertilizer should occur prior to May 1 of each year.
Increased active and passive N uptake rates are associated with higher root length
densities (Scholberg et al., 2002). Though trees on Swingle rootstocks have more of their
root mass in the upper 45 cm than do trees on Carrizo (Chapter 4), significant differences
in N uptake by rootstock were not detected. Significantly higher mean daily soil N losses
were found for Hamlin trees growing on Swingle compared with Hamlin trees on
Carrizo at the high N application rate. However, uptake was lower for Swingle at the

142
lower application rate. Total uptake rates were nearly equal for trees grown on Swingle
and Carrizo at either N application rate.
Numerous studies have been unable to account for 40% or more of total N applied
to soils (Mansell et al., 1986, Dasberg et al., 1987). Some authors attributed this
unaccounted for fraction as being lost to the atmosphere through denitrification of NO3-
N, while others concluded that the unaccounted for N was either incorporated into soil
organic matter or stored in the tree. Using soil of the same type that was used in this
study, Lea-Cox and Syvertsen (1993) could not account for 14.5% of 15N applied to
containers of soil in the greenhouse at 28 C during the first day. This value increased to
67.0% on the seventh day after application. The loss of 15N was assumed to have been
incorporated into microbial biomass because soil pH was too low to account for
appreciable volatilization of NH4-N. These data are similar to the mean of 26.1% N loss
in the control pipes 3 days after application. Nitrogen loss after 7 days totaled 60.9%
assuming the same rate of loss.
In a lysimeter study by Lea-Cox et al. (2001), 4-year-old grapefruit trees grown in
Candler fine sand were fertilized with 2.6, 5.1, and 11.6 g N tree-1 of double-15N labeled
ammonium nitrate. Soil samples contained no NH4-N on either date for the lowest rate,
and a mean of 58.2 and 0% of the highest rate on 1 and 8 days after application,
respectively. This result suggests that the mean of 34.0% NH4-N found 1 day after
application of 85.1 and approximately 43 g N tree'1 in the current study is too low.
However, in the same study, a mean of only 5.7% of NFL-N applied at the 5.1 g N tree1
rate was found in soil under trees not contained in lysimeters compared with 17.9% in
lysimeters at the same application rate. This result implies that the NH4-N amounts in the

143
lysimeters may be 3.5 times too high, reducing the residual NH4-N amount in bulk soil to
16.6% 1 day after application. If this is the case, then values found in this study are
probably similar to those found by Lea-Cox et al. (2001).
In the same study, Lea-Cox et al. (2001) found that tissue accumulation of 15N
increased for the first 3 to 5 days after application with little additional subsequent
increase, with the exception of trees on Volkamer lemon rootstock. Trees on this
rootstock increased in ,5N accumulation to between days 7 and 15, but only at the highest
N application rate. Young trees on Volkamer lemon grow rapidly and are assumed to
have high N demand compared with the two rootstocks used in the current study. This
result confirms that both N demand and availability control uptake rates (Dasberg, 1987;
Kato et al. 1982, and Scholberg et al., 2002). Lea-Cox et al. (2001) found total N in the
soil was reduced by 71.3 to 95.1 of the applied N on day 8 for trees grown on Volkamer
lemon rootstock. Assuming that the majority of N is removed and accumulated by the
tree in the first 3 to 5 days, a reduction in applied N ranging from 70.1 to 83.6% three
days after application for the two rates in the current study would appear to be
reasonable. Likewise, the estimated NUE that ranged from 50.2 to 84.0% of the 15N
applied after 29 days in the above study was similar to the estimated mean maximum
uptake range of 46.7 to 61.7% for the high and low N application rates used in this study.
Mean estimated cumulative dry biomass losses during the 2-year study from
flowers, fruit and leaves were 646, 1290, and 2423 g tree'1, respectively. Leaf loss
accounted for approximately 33% of the total leaf biomass estimated for mature trees of
the same size. This result indicates that leaves could remain on citrus trees for a many as
3 years, which is 1 year longer than reported by Wallace et al. (1945). Leaf longevity

144
may also be affected by climatic conditions and the incidence of pest and diseases.
Cumulative N losses for flowers, fruit, and leaves were 20.3,23.4, and 49.3 g tree'1,
respectively. Assuming that all N from senesced plant parts is incorporated into soil
organic matter, and that this rate of N addition is similar to previous years, 93 g or more
of N may be available on an annual basis due to mineralization. Using the N balance
calculated for mature citrus trees grown on Carrizo and Swingle rootstocks in Chapter 3,
600 g N tree'1 could be available to the trees. This amount of N would reduce N uptake
efficiency to 78.5 and 67.2% for trees grown on Carrizo and Swingle rootstocks,
respectively. These values are near the upper limit of NUE values estimated for citrus in
studies by Syvertsen and Smith (1996).
Conclusions
Leaf, twig, and branch bark N concentrations decreased through the spring to
minimums in May and June of each year. This time period corresponds to a period of
high vegetative and reproductive growth rates. High NUE in May compared with October
/indicates that the reduction in tissue N concentration is not due to ability of the tree to
extract available N from the soil, but possibly a redistribution of N from leaf, twig and
branch bark tissues in response to low N supply. Tissue N concentrations recovered by
late summer, approaching winter values. Under Florida conditions, NH-N was rapidly
converted to NO3-N. Nitrogen uptake rates were greater in late spring when soil
temperatures were high and leaf N concentrations were low, compared with late summer
when soil temperatures were similar and leaf N concentrations were higher. This
relationship indicates a correlation between tissue N concentration and N uptake rates.
Such a relationship is fundamental to modeling N uptake in any crop. Tree biomass and

145
N senescence patterns were relatively consistent across seasons providing a relationship
for seasonal tree N balance for citrus. Understanding seasonal N uptake rates and cycling
is necessary to develop of crop models. Such models provide a scientific basis for
improved water and N management in agricultural production. Providing citrus growers
with tools like an expert system to increase NUE and avoid nitrate leaching is essential,
especially in Florida due to the low water and nutrient holding capacity of the sandy soils
there.

CHAPTER 7
SUMMARY AND CONCLUSIONS
Simulating plant growth is complex due to the interactions of biological
processes, soil physical characteristics, and environmental factors (Jones and Luyten,
1998). Biological processes and soil physical characteristics define the crop system and
include photosynthesis, respiration, transpiration, biomass accumulation, soil water and
nutrient uptake, and nitrogen leaching. Environmental factors such as temperature,
radiation, wind, humidity, rainfall, and irrigation influence these processes and are
typically required as model inputs. Modeling these processes requires the estimation of
many state variables with time through the use of linear or non-linear functions and
associated parameters and constants. The goals of this study were to 1) determine
changes in above ground citrus biomass and N distribution for trees under recommended
N fertility practices over a range of tree sizes, 2) collect information on spatial
distribution patterns of citrus root length density for different tree sizes, 3) estimate
evapotranspiration crop and soil moisture coefficients, and soil water use per unit root
length density, 4) explore seasonal N uptake rates for citrus, and 5) compare seasonal
biomass and N concentration changes for citrus fertigated at two annual N fertilizer rates.
Mature Tree Biomass Distribution
Leaf biomass represented 12 to 15% of total tree biomass, while total branch
weights were 50 to 65% of total tree weight. Total root biomass was highly variable, but
146

147
averaged 20-24% of total tree weight. Mature trees on Carrizo citrange rootstocks were
significantly larger than those on Swingle citrumelo. The proportion of large branch
biomass with respect to total tree biomass was significantly greater for trees grown on
Carrizo compared with trees grown on Swingle. On a percentage basis, taproot biomass
was significantly greater for trees on Swingle than those on Carrizo. Mean total biomass
of large roots and taproots was greater for trees on Swingle than for those on Carrizo.
Biomass Vs Tree Size Relationships
Total leaf area per tree was proportional to both tree canopy volume and average
trunk diameter. Maximum total fresh and dry weights for trees with canopy volumes
ranging from 28 to 38 m3 and trunk cross sectional area of 130 to 150 mm were
approximately 160 and 100 kg tree'1, respectively. Above-ground and below ground dry
weights of large citrus trees were approximately 74 and 26 kg, respectively. Leaf biomass
declined from approximately 20% of total biomass for trees with canopy volumes of less
than 5 m3 and trunk cross sectional area of less than 60 mm to approximately 12% of
total biomass for trees with canopy volumes greater than 30 m3 and trunk cross sectional
area greater than 120 mm. Likewise, twig biomass decreased from 11% of biomass to
6% for trees in the same size categories. Total branch dry biomass increased from 15 to
45% across the range of small to large trees, while trunk biomass concurrently decreased
from 12 to 3%.
Nitrogen Distribution
As with dry biomass, estimated total N mass was greater for trees on Carrizo
rootstock compared with trees on Swingle. Mean leaf, branch, and root N were
approximately 45, 35, and 20% of total N for mature citrus trees. Trees grown on Carrizo

148
accumulated almost 50% more N in large branches compared with trees on Swingle.
Total leaf N weight increased from less than 30 to more than 250 g tree'1 across the range
of canopy volumes and trunk diameters measured. Leaf N accounted for 45% of total N
in trees with canopy volumes less than 5 m3 and 37% of total N in trees greater than 35
m3. Twig N ranged from less than 10 to greater than 50 g tree'1 through the same range of
tree sizes. However, unlike leaves, twig N corresponded to a consistent 9% of total tree
N. Total branch N weight increased from less than 10 to greater than 200 g N tree'1,
which corresponded to an increase in percentage of total tree N from 6 to 27% for the
range of tree sizes measured. Concurrently, the proportion of total tree N in the trunk
decreased from 5 to 3%.
Mature Hamlin Root Distribution
Average root length density of fine fibrous roots surrounding mature citrus trees
followed a bimodal spatial distribution with depth from the soil surface and decreased
with distance from the tree trunk. Mean fine fibrous root density in the upper 15 cm was
1.036 cm cm'3 and ranged from 1.9 cm cm'3 at 50 cm from the tree trunk to 0.7 cm cm'3
at the 200 cm distance. Mean densities decreased at the 15 to 30 cm depth to 0.30 cm cm'
3 ranging from 0.50 to 0.07 cm cm'3 at 50 and 200 cm distances, respectively. Mean
densities of fine fibrous roots increased with subsequent depths to a maximum at the 60
to 75 cm depth and then declined at the 75 to 90 cm depth.
Differences in root spatial distribution between rootstocks were not statistically
significant. Mean root length densities at all depths and distances were 0.36 cm cm'3 for
trees grown on Carrizo citrange and 0.41 cm cm3 for trees grown on Swingle citmmelo.
Trees grown on Swingle had greater root length densities near the soil surface than did

149
trees on Carrizo. Conversely, root length densities were greater for trees on Carrizo
between 15 and 75 cm below the soil surface. Root length densities increased for trees on
Carrizo at the 45 to 60 cm depth, whereas densities for trees on Swingle increased at the
60 to 75 and 75 to 90 cm depths.
Root Length Density Distribution Changes with Tree Size
Distance from the tree trunk and depth from the soil surface significantly affected
citrus root length density across a wide range of tree sizes. Root systems of young trees
were initially concentrated at the soil surface, with few roots deeper than 0.5 m at a
distance of 150 cm from the tree trunk. As the citrus trees began to produce fruit (5 to 10
years of age) root length density increased at the soil surface to a distance equal to the
dripline of the tree. Roots extended to the 200 cm distance between tree rows and to a
depth of 0.9 m at 150 cm from the trunk. The bimodal nature of the root system was
observed close to the trunk at depths below 60 cm. By the time trees reached 10 to 15
years of age and the canopy was nearing a full hedgerow, the bimodality of the root
system was fully developed and roots extended below lm at all distances from the tree.
Seasonal ET0 and ETC Trends
Daily ET0 reported by FAWN for the experimental site ranged from a minimum
of 1.12 mm in December 2001 to a maximum of 6.48 mm in June 2001. Monthly
maximum, minimum, and mean ET0 and ETC were not significantly different for the same
months during the 2 years of this study. Although generally lower, daily ETC followed the
same seasonal patterns as ET0. The exception to this correlation occurred during the
summer months of June through August, and then only when soil was near field capacity.

150
Seasonal K
The ratios of ETC to ET0 at field capacity were used to estimate Kc. The ETC to
ET0 ratios ranged between 0.81 on DOY 24 and 1.12 on DOY 179. With an R2 of 0.755,
DOY explained more than 75% of the variation in the ETC to ET0 ratios when soil water
content was near field capacity. Therefore, the results of the equation are a good
approximation of the value of Kc for a given DOY.
K, Estimation
The relationship of estimated K to available soil water depletion (ASWD) and
soil water potential was logistic in nature with a value of 1 from field capacity to
approximately 10 to 15% of ASWD. The relationship decreased steadily to
approximately 0.6 at 50% ASWD, indicating a reduction of 40% in ETC between 15 and
50% ASWD.
Soil Water Uptake per Unit Root Length
An exponential decay model resulted in the best fit of unit root length uptake to
soil water potential with a maximum of approximately 0.4 mm3 d'1 cm"1 at field capacity.
Daily ETC per unit root length decreased rapidly with decrease in soil water potential to
-12 or -13 kPa, followed by a more gradual reduction from approximately 0.1 to 0.05
mm3 d"1 cm"1 well past -20 kPa. Higher water uptake per unit root length values at field
capacity were found at the 10 cm depth between rows. This value was double that
observed at other locations and depths. The increase in water uptake could be explained
by water use from non-crop species in the row middles that were not present beneath the
tree canopy.

151
Nitrogen Uptake
Differences in mean cumulative N loss from the soil 3 days after fertilizer
application were significantly influenced by month of year, with the greatest loss in May.
Mean percentage of total soil N loss during the 3 day period for the high (83 g N tree'1)
application rates were 60.7, 68.6, and 63.4% for March, May, and September;
respectively. These means were significantly different from mean percentage losses for
the lower N rate, which were 83.6, 82.6 and 73.1% for the same three months. Mean
cumulative N soil losses for Hamlin trees on Carrizo citrange were 70.5, 78.8, and
70.1% of N applied for the months of March, May, and September, respectively, and
were not significantly different from mean cumulative losses for trees on Swingle
citrumelo, which were 73.8, 72.5, and 66.4% of total N applied for the same months.
Nitrogen losses from the control treatment varied greatly in the three studies and
were not significantly different by application rate, month, or rootstock. Mean cumulative
N loss was 22.6%, and ranged from a net gain of 7.0% to a loss of 41.0% during the 3
days after fertilizer application. Cumulative daily maximum N uptake was significantly
different by application amount, but not significantly different by month and rootstock.
Maximum uptake as a percentage of amount applied averaged 46.7 and 61.7 for the high
and low application rates, respectively. Mean cumulative N uptakes for Carrizo and
Swingle rootstocks were 53.9 and 54.4%, respectively. Estimates of the Michaelis-
Menten equation constants I max and Km for active N uptake were 8.5 g tree'1 d1 and 45
mg L1, respectively.

152
Nitrification Estimation
Mean NO3-N content in the upper 45 cm of soil increased 24 h after application to
151% of NO3-N applied. The mean NO3-N content decreased during the next 48 h to
126% of total NO3-N applied. Content of NH4-N decreased to 34% of that applied after
24 h and steadily declined after application to 16% of NH4-N applied on day 3. The sum
of NO3-N and NH4-N (total mineral N) declined throughout the period to 73.9% of total
N applied on day 3, indicating that approximately 26.1% N was lost in 3 days due to
volatilization of NH4-N, microbial activity, or incorporation in organic matter. The
nitrification rate in this soil was very rapid. With an average of more than 50% of the
NH4-N converted to NO3-N within the first 24 h, nearly 85% of applied NH4-N was
converted to NO3-N in 3 days assuming all of the N loss was NH4-N.
Seasonal Tissue N Concentration
Leaf and twig N percentages were near maximum from August to February and at
minimum in May of each year. Minimum leaf N concentrations were 1.9% and 2.0% for
the flush and expanded leaves, respectively, on trees fertilized with 179 kg N ha'1 yr'1.
Minimum leaf N concentration for trees fertilized with 269 kg N ha'1 yr'1 were 2.0 and
2.2% for flush and expanded leaves, respectively. Maximum leaf N concentrations were
2.5 and 2.7% for 179 and 269 kg ha'1 annual application rates, respectively. The areas of
both flush and expanded leaves were generally greater for trees receiving the high annual
N rate. Minimum twig N concentrations occurred in May and were 0.83% for both annual
N application rates. Maximum twig N concentrations of 0.98 and 1.02% for the 179 and
269 kg N ha'1 annual application rates, respectively, occurred in August.

153
Branch bark N concentrations were consistently in a narrow range of 1.0 to 1.3%
during the 2-year period, with means of 1.04 and 1.17% for low and high N application
rates, respectively. Minimum bark N concentrations of 1.00 and 1.11% occurred in May
or June of each year, while maximum concentrations of 1.09 and 1.22% occurred in
October and January for low and high annual N application rates, respectively. Branch
wood N concentrations were lower, and followed trends similar to those of branch baric
tissue. Mean wood N concentrations were 0.25 and 0.31% for the low and high N
application rates, respectively. Maximum wood N concentrations were 0.37 and 0.38%
for the same rates, respectively, and occurred in January and March. Minimum wood N
concentrations were 0.23 and 0.29% for low and high N application rates, respectively,
and occurred in October.
Root N concentrations were greater for roots <4 mm in diameter compared with
larger roots. Mean N concentrations were 1.34 and 1.35% for low and high N application
rates, respectively, for roots <4 mm in diameter. Mean N concentrations for roots >4 mm
in diameter were 0.85 and 0.89% for the same rates. Fruit N concentration was highest in
May of each year (1.25 and 1.30% for low and high N application rates, respectively).
Nitrogen concentrations decreased through the season to means of 1.03 and 1.06% for
low and high N application rates, respectively, just prior to harvest. Mean fruit diameter
was not constantly different by N application rate. However, fruit dry weights were
greater for the lower application rate.
Seasonal Nitrogen Loss
Flower dry weight accumulated in March and April of each year with cumulative
weights of 529 and 764 g tree-1 for 2001 and 2002, respectively. Cumulative flower N

154
weights were 18.8 and 21.9 g tree'1 for the same years. Fruit and leaf biomass losses were
greatest in April through May and September through December of each year. Little loss
of either of these tissues occurred from June through August. Cumulative fruit biomass
varied more than bloom and leaf biomass, with 1635 and 946 g tree'' removed in 2001
and 2002, respectively. These fruit biomass values represented a cumulative loss of 30.1
and 16.8 g tree'1 of N for these years. Leaf biomass losses were 2041 and 2806 g tree'1 for
2001 and 2002, respectively. Cumulative annual N loss from leaf fall amounted to 42.2
and 51.4 g tree'1 for the same years.
Citrus Decision Support System
Due to the complexity of grower decision-making processes, researchers have
developed computer based decision support systems (DSS) to provide information on
management options. These DSS store and organize information such as rates of water,
fertilizer, and agrichemicals applied to specific fields and provide information on
predicted future events such as irrigation scheduling and N leaching. Crop models are
used to determine the effect a given management decision will have on the crop such as
growth rate or yield. Crop models such as CROPGRO, CERES and others (Hoogenboom
et al., 1994; Jones et al., 1991; Wagner-Riddle et al., 1997) are process-oriented models
that simulate vegetative growth and reproductive development. The models predict dry
matter growth (Shen et al., 1998), crop development (Batchelor et al., 1994; Batchelor et
al., 1997; Piper et al., 1996) and final yield (Batchelor et al., 1996; Heinemann et al.,
2000) for a range of agronomic crops. Inputs are daily weather data, soil profile
characteristics, and crop management conditions (Gijsman et al., 2002). Such models are
currently being used for a number of purposes such as yield forecasting and long-term

155
effects of crop sequencing under given management inputs and weather conditions. Crop
and soil water status (Hoogenboom et al., 1994; Gabrielle et al., 1995; and Xie et al.,
2001) and N and C balances (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995;
and Sexton et al., 1998) have been modeled. These models can combine with crop
developmental models of irrigation scheduling, fertilizer N fate determination, and nitrate
leaching estimates. A DSS can provide the framework for storing of information needed
by the model and can include a user-friendly output to assess production options.
Citrus production would benefit greatly from such a DSS. Data collected and
relationships determined in this dissertation as well as literature sources can provide
information required to model seasonal and temporal citrus N demand, root development
processes, seasonal water and N uptake, and seasonal N distribution. These models can
lead to a DSS capable of providing irrigation scheduling, fertilizer requirements and
rates, and environmental impacts of nitrate leaching. Verification of best management
practices and grower compliance with such practices can be determined using a DSS. The
ultimate goal of such a DSS would be the improved allocation of water and fertilizer
resources for optimal yield with minimal environmental impact.
Citrus N Practices and BNPs
Floridas population has increased from 3 million in 1950 to more than 16 million
at present, creating severe competition between agricultural, commercial, residential, and
environmental users for Floridas limited water resources. As a result of increased
demand, consumption and quality of water has become highly scrutinized by regulators.
Improvements in both N uptake efficiency and timing of N applications to coincide with
N demand will balance the needs of the citrus tree while minimizing water quality

156
impacts. This study has improved our understanding of seasonal and long-term N
accumulation by citrus. The information generated will be essential to refine citrus N
BMPs using sound, science-based decision making. Improved BMPs will allow for
sustainable productivity while safeguarding the environment from leaching of excess
NO3-N to ground water.
Current citrus N BMPs base the fertilizer N application rate on the chronological
age of young citnis trees and stress N fertilization timing in the spring and fall of the year
to avoid potential leaching during the summer rainy season. Long-term N accumulation
measured in this study indicate that tree size is a more useful method of determining
potential N demand of young trees than tree age. Likewise, seasonal N demand by mature
citrus trees was greatest during the spring of the year; therefore a citrus N balance will
provide a better basis to accurately determine N fertilizer requirements of mature trees.
Soil water content and labile N concentration in the root zone are keys to N
uptake efficiency of citrus. Floridas sandy soils hold little water and dry quickly,
reducing potential evapotranspiration much more rapidly than finer textured soils. This
decrease in potential uptake impacts passive and active N uptake through reduced water
uptake and N diffusion to the root boundary. Therefore, soil water content and N
concentrations in the soil volume containing the highest root mass must be maintained as
high as practical without forcing N below the root zone. This study showed that two
rootstocks thought to have similar root growth patterns in reality had different root
densities with distance from the tree and with soil depth. The implication of this finding
is that the top 30 cm soil layer may be more critical for water and N fertilizer
management for trees on Swingle citrumelo than for trees on Carrizo citrange.

157
Determining root system distribution patterns of the various rootstocks on which citrus is
propagated is essential to irrigation scheduling for improved fertilizer N uptake
efficiency.
While current N BMPs are believed to reduce N leaching below the citrus root
zone, more accurate methods of determining annual fertilizer N rates, application timing,
and irrigation scheduling are essential to improve uptake efficiency and enhance
environmental protection. This study confirmed that the basic principles used to derive
citrus N BMPs are sound, but also generated data suggesting that refining and fine-tuning
BMPs may be necessary to make them totally effective. Adoption of the N balance
approach to fertilization evapotranspiration-based irrigation scheduling, and rootstock
specific N fertilizer management are essential to sustainable productivity and improved
ground water quality.

APPENDIX A
EQUIATIONS
Equation
2-1. Darcys Law water flow through a saturated medium
u =-K d<|) /dx
Where:
u = water flux (cm3 cm'2 s'1),
K = hydraulic conductivity constant (cm s'1),
= soil water potential (kPa), and
x = the distance over which the flux is maintained (cm).
2-2. Poisevilles equation water flow through a tube
f = (tc r4/ 8t)) dP/dx
Where:
f = flow rate in a tube (m3 s'1),
r = radius (m),
q = viscosity (pPa s'1), and
dP/dx = pressure gradient.
2-3. Richards Equation water flow in unsaturated soils
o = -Ke d/dx = -Ke (d/d0) (d0/dx) = -De (d0/dx)
Where:
u = water flux (cm3 cm'2 s'1),
Pace
..23
24
25
158

159
Ke = hydraulic conductivity constant at 0 (cm s'1),
0 = soil water content (cm cm' ),
d0/d = slope of the soil characteristic curve, and
x = the distance over which the flux is maintained (cm).
3-1. Equation for tree canopy volume estimation
K
TCV = Ir Cr Ht *
4
d-(i-(^)2))
Where:
TCV = Tree canopy volume (m3)
Ir = In-row spacing (m)
Cr = Cross-row spacing (m)
Ht = Canopy height (m)
Int = Canopy intercept height (m)
5-1. Crop evapotranspiration (ETC) estimation
ETc = ET0*Kc*K8
Where:
ETC = Crop evapotranspiration (mm d1)
ET0 = Potential evapotranspiration (mm d'1)
Kc = Crop coefficient
K = Soil stress coefficient
44
91
5-2. Soil water stress Coefficient (Kc) estimation
93

160
K TAW-fl
8 TAW-RAW
Where:
Ks = Soil water stress coefficient
TAW = 0fc Opwp = Total available water (cm3 cm'3)
0 = Soil water content (cm3 cm'3)
RAW = 0fc 0ra = Readily available water (cm3 cm'3)
6-1. Mechaelis-Menton equation for estimation of nutrient uptake 117
I Imax Cu /(Km+Cu)
Where:
I = inflow flux of nutrient (mol cm'2 s'1),
Imax = maximum active flux (mol cm'2 s'1),
Cl, = nutrient concentration in the soil solution at the root surface (mol cm'3),
Km = Cl, value at \m.J2 (mol cm'3),
6

LITERATURE CITED
Adams, J.S. 1875. Notes on wild orange groves. FI. Fruit Growers Assoc. Proc. 3:24.
Adriance, C.W. and J.E. Hampton. 1949. Root distribution in citrus as influenced by
environment. Proc. Amer. Soc. Hort. Sci. 53:103-108.
Allen, R.G. L.S. Pereira, D. Raes. And M. Smith. 1997. Crop evapotranspiration,
guidelines for computing crop water requirements. FAO Irrigation and Drainage
Paper No. 56. Rome Italy. 300 page.
Alva, A.K., and S. Paramasivam. 1998. Nitrogen management for high yield and quality
of citrus in sandy soils. Soil Sci. Soc. Am. J. 62:1335-1342.
Arias-Reveron, J.M., and H.W. Browning. 1995. Development and mortality of citrus
snow scale under constant temperature and relative humidity. Env. Entomology
24:1189-1195.
Batchelor, W.D., J.W. Jones, K.J. Boote, and G. Hoogenboom. 1994. Carbon-based
model to predict peanut pod detachment. Trans, of the ASAE 37(5): 1639-1646.
Batchelor, W.D., J.W. Jones, and K.J Boote. 1996. Comparisons of methods to compute
peanut seed size distribution by crop growth models. Trans, of the ASAE
39(2):737-744.
Batchelor, W.D., M R. Zeiss, L.P. Pedigo, and R.M. Shibles. 1997. Development of a
model to predict soybean pod color distribution. Trans, of the ASAE 40(1):221-
227.
Bellows, T.S., and J.G. Morse. 1986. Modeling flower development in navel oranges.
Scientifica Hortic. 30:117-126.
Bielorai, H. 1977. The effect of drip and sprinkler irrigation on grapefruit yield, water use
and soil salinity. Proc. Int. Soc. Citriculture 1:99-103.
Bland, W.L. and W.A. Dugas. 1989. Cotton root and soil water extraction. Soil Sci. Soc.
Am. J. 53(6): 1850-1855.
Boman, B.J. 1993. A comparison of controlled-release to conventional fertilizer on
mature Marsh grapefruit. Proc. Fla. State Hort. Soc. 106:1-4.
161

162
Boman, B. J. 1994. Evapotranspiration by young Florida flatwoods citrus trees. J.
Irrigation and Drainage Engineering 120:80-88.
Boswell, S B., C D. McCarty, and L.N. Lewis. 1975. Tree density affects large-root
distribution ofWashington Navel orange trees. HortScience 10:593-595.
Cahoon, G.A., R.B. Harding, and D.B. Miller. 1956. Declining citrus root systems. Calif.
Agrie. 10(9):3-12.
Cahoon, G.A., E.S. Morton, W.W. Jones, and M.J. Garber. 1959. Effects of various types
of nitrogen fertilizers on root density and distribution as related to water
infiltration and fruit yields of Washington Navel oranges in a long-term
fertilizer experiment. Proc. Amer. Soc. Hort. Sri. 74:289-299.
Cahoon, G.A., M.R. Huberty, and M.J. Garber. 1961. Irrigation frequency effects on
citrus root distribution and density. Proc. Amer. Soc. Hort. Sci. 77:167-172.
Cahoon, G.A., L.H. Stolzy, M.J. Garber, and E.S. Morton. 1964. Influence of nitrogen
and water on the root density of mature Washington Navel orange trees. Proc.
Amer. Soc. Hort. Sci. 85:224-231.
Calvert, D.V., RCJ. Koo, and H.W. Ford. 1967. Flood irrigation studies with citrus. Proc.
Fla. State Hort. Soc. 80:79-85.
Calvert, D.V., H.W. Ford, E.H. Stewart, and F.G. Martin. 1977. Growth response of
twelve citrus rootstock-scion combinations on a Spodosol modified by deep
tillage and profile drainage. Proc. Inti. Soc. Citriculture 1:79-84.
Calvert, D.V. and H.T. Phung. 1972. Nitrate-nitrogen movement into drainage lines under
different soil management systems. Soil Crop Sci, Soc. Fla. Proc. 31:229-232.
Cameron, S .H and D. Appleman. 1935. The distribution of total nitrogen in the orange
tree. J. Amer. Soc. Hort. Sri. 30:341-348.
Cameron, S.H., and O.C. Compton. 1945. Nitrogen in bearing orange trees. J. Amer. Soc.
Hort. Sci. 46:60-68.
Carlisle, V.W., F. Sodek, M E. Collins, L.C. Hammond, and W.G. Harris. 1989.
Characterization data for selected Florida soils. University of Florida-Soil Science
Department, Soil Characterization Laboratory, Institute of Food and Agricultural
Sciences in cooperation with U.S. Department of Agriculture-Soil Conservation
Service. Soil Science Research Report No. 89-1. 307 p.
Caruso, T., P. Inglese, F. Sottile, and F.P. Marra. 1999. Effect of planting system on
productivity, dry-matter partitioning and carbohydrate content in above-ground
components ofFlordaprince peach trees. J. Am. Soc. Hortic. Sci. 124(1) 39-45.

163
Castel, J.R., I. Bautista, C. Ramos, and G. Cruz. 1987. Evapotranspiration and irrigation
efficiency of mature orange orchards in Valencia (Span). Irrigation and
Drainage Systems 3:205-217.
Castel, J.R., and A Buj. 1992. Growth and evapotranspiration of young, drip-irrigated
Clementine trees. Proc. Int. Citriculture, 651-656.
Castle, W.S. 1978. Citrus root systems: Their structure, function, growth, and relationship
to tree performance. Proc. Int. Soc. Citriculture 1:62-69.
Castle, W.S. 1980. Fibrous root distribution ofPineapple orange trees on rough lemon
rootstock at three tree spacings. J. Amer. Soc. Hort. Sci. 105(3):478-480.
Castle, W.S. and AH. Krezdom. 1975. Effect of citrus rootstocks on root distribution and
leaf mineral content ofOrlando tngelo tree. J. Amer. Soc. Hort. Sci. 100(1): 1-
4.
Castle, W.S. and CO. Youtsey. 1977. Root system characteristics of citrus nursery trees.
Proc. Fla. State Hort. Soc. 90:39-44.
Chapman, H.D., and E.R Parker. 1942. Weekly absorption of nitrate by young, bearing
orange trees growing out-of -doors in solution culture. Plant Phys. 17:366-376.
Cohen, S., M. Fuchs, M S. Moreset and Y. Cohen. 1987. The distribution of leaf area,
radiation, photosynthesis, transpiration, and the effect of row shape and direction.
Agricultural and Forest Meteorology 40:145-162.
Dasberg, S., 1987. Nitrogen fertilization in citms orchards. Plant Soil 100:1-9.
Doorenbos, J., and W.O. Pruitt. 1977. Crop water requirement. FAO irrigation and
Drainage paper No. 24, Rome Italy, 144 pages.
Elezaby, A. A. 1989. Physiological and biological studies on root systems of some citrus
stocks. Ph D. Dissertation. Cairo University, Egypt.
Eissenstat, D.M. 1991. On the relationship between specific root length and the rate of
root proliferation: a field study using citrus rootstocks. New Phytol. 118:63-68.
Eissenstat, D.M. and L.W. Duncan. 1992. Root growth and carbohydrate responses in
bearing citrus trees following partial canopy removal. Tree Physiol. 10:245-257.
Embleton, J.W, M. Mastumura, L.H. Stolzy, D.A. Devitt, W.W. Jones, R. El-Motaium
and L.L. Summers. 1978. Citrus Nitrogen fertilizer management, groundwater
pollution, soil salinity and nitrogen balance. Applied Agr. Res. 1 (l):57-64.

Erie, L.J., O.F. French, and K. Harris. 1965. Consumptive use of water by crops in
Arizona. University of Arizona Agri. Exp. Sta. Tech. Bui. No. 169,44 pp.
164
Fares, A., and A. K. Alva. 1999. Estimation of Citrus Evapotranspiration by Soil Water
Mass Balance. Soil Science 164:302-310.
Feigenbaum, S., H. Bielorai, Y. Emer, and S. Dasberg. 1987. The fate of 15N labeled
nitrogen applied to mature citrus trees. Plant and Soil. 97:179-187.
Flenet, F., J.R. Kiniry, J.E. Board, M E. Westgate, and D C. Reicosky. 1996. Row
spacing effects on light extinction coefficients of com, sorghum, soybean and
sunflower. Agron, J. 88(2) 185-190.
Florida Agricultural Statistics Service. 2002. Commercial citrus inventory 1998. Florida
Department of Agriculture and Consumer Services. Tallahassee, FL.
Follett, R.F., M.J. Shaffer, M.K. Brodahl, and G.A. Reichman. 1994. NLEAP simulation
of residual soil nitrate for irrigated and non-irrigated com. J. Soil Water Conser.
49(4):375-382.
Ford, H.W. 1952. The distribution of feeder roots of orange and grapefruit trees on rough
lemon rootstock. Citrus Industry Magazine 14(1 1):22-23.
Ford, H.W. 1953a. Effect of spreading decline disease on the distribution of feeder roots
of orange and grapefruit trees on rough lemon rootstock. Proc. Amer. Soc. Hort.
Sci. 61:68-72.
Ford, H.W. 1953b. Root distribution of chlorotic and iron-chelate-treated citrus trees.
Proc. Fla. State Hort. Soc. 66:22-26.
Ford, H.W. 1954a. Root distribution in relation to the water table. Proc. Fin. State Hort.
Soc. 67:30-33.
Ford, H.W. 1954b. The influence of rootstock and tree age on root distribution of citrus.
Proc. Amef Soc. Hort. Sci. 63:137-142.
I
Ford, H.W. 1959. Growth and root distribution of orange trees on two different rootstocks
as influenced by depth to subsoil clay. Proc. Amer. Soc. Hort. Sci. 74:313-321.
Ford, H.W. 1964. The effect of rootstock, soil and soil pH on citrus root growth in soils
subject to flooding. Proc. Fin. State Hort. Soc. 77:41-45.
Ford, H.W. 1968. Fluctuations of the water table in drained flatwoods groves. Proc Fla
State Hort. Soc. 81:75-79.

165
Ford, H. W. 1969. Water management of wetland citrus in Florida. Proc. First Int. Citrus
Symp. 3:1759-1770.
Ford, H.W. 1970. Problems in using Milam rootstock as a biological barrier. Proc. Fla.
State Hort. Soc. 83:84-86.
Ford, H.W. 1972. Eight years of root injury from water table fluctuations. Proc. Fla. State
Hort. Sci. 85:65-68.
Ford, H.W., I. Stewart, and C D. Leonard. 1954. The effect of iron chelate on root
development of citrus. Proc. Amer. Soc. Hort. Sci. 63:81-87.
Ford, H.W., W. Reuther, P.F. Smith. 1957. Effect of nitrogen on root development of
Valencia orange trees. Proc. Amer. Soc. Hort. Sci. 70: 237-244.
Gabrielle, B., S. Menasseri, and S. Houot. 1995. Analysis and field evaluation of the
CERES models water balance component. Soil Sci. Soc. Am. J. 59:1403-1412.
Gabrielle, B., and L. Kengni. 1996. Analysis and field-evaluation of the CERES models
soil components: nitrogen transfer and transformation. Soil Sci. Soc. Am. J. 60:
142-149.
Gijsman, A.J., G. Hoogenboom, W.J. Parton, and P C. Kerridge. 2002. Modifying
DSSAT crop models for low-impact agricultural systems using a soil organic
matter-residue module from CENTURY. Agron. J. 94:462-474.
Graham, W.D., and A. Alva. 1995. Ridge citrus water quality project annual progress
Report submitted to the Florida Department of Agriculture and Consumer
Services and the Southwest Florida Water Management Districts.
Graham, W.D., and T.A. Wheaton. 2000. Ridge citrus water quality project annual
progress report submitted to the Florida Department of Agriculture and Consumer
Services and the Southwest Florida Water Management Districts.
Graser, E.A., and L.H. Allen, Jr. 1987. Water relations of 7-year-old containerized citrus
trees under drought and flooding stress. Proc. Fla. State Hort. Soc. 100:126-136.
Hagan, R. M., H. R. Haise, and T.W. Edminster (eds). 1967. Irrigation of Agricultural
Lands. Am. Soc. of Agron. Number 11.
Hamblin, A and D. Tennant. 1987. Root length density and water uptake in cereals and
grain legumes. How well are they correlated. Aust. J. Agri. Res. 38(3) 513-527.
Hams, AJ. 1875. The wild orange groves of Florida. FI Fruit Growers Assoc. Proc 3 27
-29. r-

166
Heinemann, A.B., G. Hoogenboom, G.A. Georgiev, R.T. de Faria, and J. A. Frizzone.
^ 2000. Center pivot irrigation management optimization of dry beans in humid
areas. Trans. Am. Soc. Agri Engr. 43(6): 1507-1516.
Hilgeman, R.H. 1941. Nitrogen uptake by grapefruit trees in the Salt River Valley. Proc.
Amer. Soc. Hort. Sci. 39:119-124.
Hillel, D. 1971. Soil and water physical principles and processes. Academic Press,
New York, New York.
Hillel, D. 1998. Environmental soil physics. Academic Press. New York, New York.
Hoffman, G.J., J.D. Oster, and W.J. Alves. 1982. Evapotranspiration of mature orange
trees under drip irrigation in an arid climate. Trans. Am. Soc. Agrie. Engr.
25:992-996.
Hoogenboom, G., J.W. White, J.W. Jones, and K.J. Boote. 1994. BEANGRO: A process
oriented dry bean model with a versatile user interface. Agron. J. 86:182-190.
Jabro, J.D., J.M. Jemison, L.L. Lengnick, RH. Fox, and D.D. Fritton. 1993. Field
validation and comparison of LEACHM and NCSWAP. Trans, of the ASAE
36(6): 1651-1657.
Jarvis, P.G. and K.G. McNaughton. 1986. Stomatal control of transpiration: Scaling up
from leaf to region. Advances in Ecological Research 15:1-49.
Jemison, J.M., J.D. Jabro, and R.H. Fox. 1994. Evaluation of LEACHM: n. simulation
of nitrate leaching from nitrogen-fertilized and manured com. Agron. J. 86: 852-
859.
Jones, C.A., and J.R. Kiniry. 1987. CERES-MAIZE: A simulation model of maize
growth and development. Texas A&M Univ. Press, College Station.
Jones, H.G., AN. Lakso, and J.P.Syvertsen. 1985. Physiological control of water status
in temperate and subtropical fruit trees. Horticulutral Reviews 7:301-344.
Jones, J.W., E. Dayan, L.H. Allen, H. Van Keulen, and H. Challa. 1991. A dynamic
tomato growth and yield model (TOMGRO). Trans, of the ASAE 34(2): 663-672.
Jones, J.W, and J.C. Luyten. 1998. Simulation of biological processes. In agricultural
systems modeling and simulation. RM. Peart and R.B. Curry (eds). Marcel
Dekker, Inc. New York, pp 19-62.
Kaufmann, MR., S B. Boswell, and L.N. Lewis. 1972. Effect of tree spacing on root
distribution of 9-year-old Washington Navel oranges. J. Amer. Soc. Hort. Sci.
97:204-206.

167
Kato, T.S., S. Kubota, and S. Bambang. 1982. Uptake of 15N-itrogen by citrus trees in
winter and repartitioning in spring. J. Japan Soc. Hort. Sci. 50:421-426.
Kato, T., Y. Makoto, and S. Tsukahara. 1984. Storage forms and reservoirs of nitrogen
used for new shoot development in Satsuma mandarin trees. J. Japan Soc. Hort.
Sci. 52:393-398.
Keeney, D. R and D. W. Nelson. 1987. Nitrogen-Inorganic Forms, sec. 33-3, extraction
of exchangeable ammonium, nitrate, and nitrite, pp.648-9. In A. L. Page et al.,
eds., Methods of soil analysis: Part 2, Chemical and microbiological properties.
Agronomy, a series of, monographs, no.9 pt.2, Soil Science Society of America,
Madison, Wisconsin USA.
Khalaf, H.A., and RC.J. Koo. 1983. The use of controlled-release nitrogen on container
grown citrus seedlings. Citrus Vegt. Mag. 46:10-32.
Kimball, M.H., A. Wallace, and R.T. Mueller. 1950. Changes in soil and citrus root
characteristics with non-tillage. California Citrograph 35:432-433.
Kiniry, J.R., J.R. Williams, RL. Vanderlip, J.D. Atwood, D C. Reicosky, J. Mulliken,
W.J. Cox, H.J. Mascagni, S.E. Hoolinger, and W.J. Wiebold. 1997. Evaluation of
two maize models for nine U.S. Locations. Agron. J. 89:421-426.
Kiniry, J.R, and AJ. Bockholt. 1998. Maize and sorghum simulation in diverse Texas
environments. Agron. J. 90:682-687.
Koo, RC.J. 1963. Effects of frequency of irrigation on yield of orange and grapefruit.
Proc. Fla. State Hort. Soc. 74:1-5.
Koo, RC.J. 1978. Response of densely planted Hamlin orange on two rootstocks to
low volume irrigation. Proc. Fla. State Hort. Soc. 91:8-10.
Koo, RC.J. 1979. The influence ofN, K, and irrigation on tree size and fruit production
ofValencia orange. Proc. Fla. State Hort. Soc. 95:10-13.
Koo, RC.J. 1980. Results of citrus fertigation studies. Proc. Fla. State Hort. Soc. 93:33
36.
Koo, RC.J. 1986. Controlled-release sources of nitrogen for bearing citrus. Proc. Fla.
State Hort. Soc. 99:46-48.
Koo, RC.J., and J.W. Sites. 1955. Results of research and response of citrus to
supplemental irrigation. Proc. Soil Sci. Soc. Fla. 15:180-190.

168
Koo, R.C.J., and D.S. Harrison. 1965. Summary of irrigation research in Florida.
Florida Agricultural Ext. Service Memo. Report 65-8.
Koo, R.C.J., and G.T. Humer, Jr. 1969. Irrigation requirements of citrus grown on
Lakewood fine sand. Proc. Fla. State Hort. Soc 82:69-72.
Koo, R.C.J. and A.G. Smajstrla. 1984. Effects of trickle irrigation and fertigation on fruit
production and juice quality of Valencia oranges. Proc. Fla. State Hort. Soc.
97:8-10.
Kramer,P.J., and J.S. Boyer. 1995. Water Relations of Plants and Soils. Academic Press,
Inc. New York, NY, USA.
Lafolie, F., L. Bruckler, and F. Tardieu. 1991. Modeling root water potential and soil-
root water transport: I. model presentation. Soil Sci. Soc. Am. J. 55:1203-1212.
Lea-Cox, J.D. and J.P. Syvertsen. 1993. Nitrate-N-use efficiency of citrus rootstock
species is affected by salinity stress. Ann. of Botany 72:47-54.
Lea-Cox, J.D., and J.P. Syvertsen. 1996. How nitrogen supply affects growth and
nitrogen uptake use-efficiency, and loss form citrus seedlings. J. Amer. Soc. Hort.
Sci. 121:105-114.
Lea-Cox, J.D. J.P. Syvertsen, D A. Graetz. 2001. Spring-time nitrogen uptake,
partitioning and leaching from young bearing citrus trees of differing nitrogen
status. A. Soc. Hort. Sci. J. 126:242-251.
Legaz, F., and E. Primo-Millo. 1981. Dynamics of 15N labeled nitrogen nutrients in
Valencia orange trees. Proc Int. Soc. Citriculture. 575-582.
Legaz, F. and E. Primo-Millo. 1988. Absorption and distribution of nitrogen-15 applied
to young citrus trees. Proc. Sixth Int. Citrus Congress, Tel Aviv, Israel.
Legaz, F., E. Primo-Millo, E. Primo-Yufera, C.Gil, and J.L. Rubio. 1982. Nitrogen
fertilization in citrus, I: Absorption and distribution of nitrogen in Calamondin
trees (C mitis Bl.) during flowering, fruit set, and initial fruit development
periods. Plant Soil 66:339-351.
Lizaso, J.I., W.D. Batchelor, and S.S. Adams. 2001. Alternate approach to improved
kernel number calculation in CEREX-MAIZE. Trans, of the ASAE 44 1001-
1018.
Mahrer, Y., and G. Rytwo. 1991. Modeling and measuring evapotranspiration in a daily
drip irrigated cotton field. Irrig. Sci. (1991) 12:13-20.

169
Mansell, R.S., J.G.A. Fiskell, D.V. Calvert and J.S. Rogers. 1986. Distributions of labeled
nitrogen in the profile of a fertilized sandy soil. Soil Sci. 141:120-126.
Mansell, RS., H.M. Selim, D.V. Calvert, E.H. Stewart, L.H Allen, D A. Gratez, J.G.A.
Fiskell and J.S. Rogers. 1980. Nitrogen and water distribution in fertilized sandy
soil during irrigation and drainage. Soil Sci. Soc. Amer. J. 44:95-102.
Martin, E.C., A.K. Hla, P.M.Waller, and D.C. Slack. 1997. Heat unit-based crop
coefficient for grapefruit trees. J. Appl. Engr. In Agrie. 13(4):485-489.
Mataa, M. and S. Tominaga. 1998. Effects of root restriction on tree development in
Ponkan mandarin. J. Am. Soc. Hortic. Sci. 123(4) 651-655.
Mattos, D. 2000. Citrus response functions to N, P, and K fertilization. Ph.D.
Dissertation, University of Florida, Gainesville, FI.
Menocal-Barberena, 2000. Effect of rootstock on root distribution of citrus. MS. Thesis.
University of Florida, Gainesville, FI.
Mikhail, E.H. and B.M. El-Zeflawi. 1978. Effect of soil types and rootstocks on root
distribution and leaf composition of citrus trees. Proc. Int. Soc. Citriculture 1:
214-216.
Mills, T.M., K.T. Morgan, and L.R Parsons. 1999. Canopy position and leaf age
affect stomatal response and water use in citrus. J. Crop Production 2(2): 163-
179.
Minkara, M.Y., J.H Wilhoit, C.W. Wood, and K.S. Yoon. 1995. Nitrate monitoring and
GLEAMS simulation for poultry litter application to pine seedlings. Trans, of the
ASAE 38(1): 147-152.
Mooney, P.A., and A.C. Richardson. 1992. Seasonal trends in the uptake and
distribution of nitrogen in Satsuma mandarins. Proc. Int. Soc. Citriculture 593-
597.
Morgan, K. T., L. R. Parsons, T. A. Wheaton, D. J. Pitts, and T. A. Obreza. 1999. Field
calibration of a capacitance water content probe in fine sand soils. J. Soil Sci. Soc.
Am. 63:987-989.
Newman, E.I. 1966. A method of estimating the total length of root in a sample. J. Appl.
Ecology 3:139-145.
Obreza, T.A., and K.E. Admire. 1985. Shallow water table fluctuations in response to
rainfall, irrigations, and evapotranspiration in flatwood citrus. Proc. Fla. State
Hort. Soc. 68:24-29.

170
Obreza, T. A, D.J. Pitts, L.R. Parsons, T.A. Wheaton, and K.T. Morgan. 1997. Soil
water-holding characteristics affects citrus irrigaiton scheduling strategy. Proc.
Fla. State Hort. Soc. 110:36-39.
Paramasivam, S., A.K. Alva, O. Prakash, and S.L. Cui. 1999. Denitrification in the
vadose zone and in surficial groundwater of a sandy Entisol with citrus
production. Plant and Soil 208:307-319.
Parsons, L.R., T.A. Wheaton, and S.W. Castle. 2001. High application rates of
reclaimed water benefit citrus tree growth and fruit production. Hort. Science
36:1273-1277.
Prajamwong, S., G. P. Merkley, and R. G. Allen. 1997. Decision Support Model for
Irrigation Water Management. J. Irrig. And Drain. Engrg., ASAE, 122(2): 106-
113.
Piper, E.L., K.J. Boote, J.W. Jones, and S.S. Grimm. 1996. Comparison of two penology
models for prediction flowering and maturity date of soybean. Crop Sci. 36:1606
1614.
Quemada, M., and M.L. Cabrera. 1995. CERES-N model predictions of nitrogen
mineralized from cover crop residues. Soil Sci. Soc. Am. J. 59:1059-1065.
Reck, W.R. 1994. GLEAMS modeling of BMPs to reduce nitrate leaching in middle
Suwannee River area. Environmentally Sound Agriculture: Proceedings of
Second Conference 20-22 April 1994.
Reitz, H.J. and W.T. Long. 1955. Water table fluctuation and depth of rooting of citrus
trees in the Indian River area. Proc. Fla. State Hort. Soc. 68:24-29.
Reitz, H.J. 1956. Timing fertilization of citrus in the Indian River area. Proc. Fla. State
Hort. Soc. 69:58-64.
Reuther, W., P.F. Smith, G.K. Scudder, and G. Hmciar. 1957. Responses ofValencia
orange trees to timing, rates, and ratios of nitrogen fertilization. Proc. Amer. Soc.
Hort. Soc. 70:223-236.
Reyes, M R., R.L. Bengtson, and J.L.Fouss. 1994. GLEAMS-WT hydrology submodel
modified to include subsurface drainage. Trans, of the ASAE 37(4): 1115-1120.
Rogers, J.S., and J.F. Bartholic. 1976. Estimated evapotranspiration and irrigation
requirements for citrus. Proc. Soil and Crop Sci. Soc. Fla. 35:111-117.
Rogers, J.S., L.H. Allen, and D. V. Calvert. 1983. Evapotranspiration from a developing
citrus grove in himid climate. Trans. Am. Soc. Agrie. Eng. 26:1778-1783.

171
Roy, W.R. and F.E. Gardner. 1946. Seasonal absorption of nutrient ions by orange trees
in sand culture. Proc. Amer. Soc. Hort. Sci. 47: 107-118.
Saleh, A.R., R.L. Benton, and J.L. Fouss. 1994. Performance of the DRAINMOD
CREAMS model with an incorporated nutrient submodel. Trans, of the ASAE
37(4): 1109-1114.
Saseendran, S.A., K.G. Hubbard, K.K. Singh, N. Mendiratta, L.S. Rathore, and S.V.
Singh. 1998. Optimum transplanting dates for rice in Kerala, India, determined
using both CERES v3.0 and ClimProb. Agron J. 90:185-190.
Scholberg, J.M.S., L.R. Parsons, T.A. Wheaton, B.L. McNeal, and K.T. Morgan. 2002.
Soil temperature, nitrogen concentration, and residence time affect nitrogen
uptake efficiency of citrus. J. Environ. Qual. 31:759-768.
Sexton, P.J., W.D. Bachelor, K.J. Boote, and R. Shibles. 1998. Evaluation of
CROPGRO for prediction of soybean nitrogen balance in a Midwestern
environment. Trans, of the ASAE 41(5): 1543-1548.
Shen, J., W.D. Batchelor, J.W. Jones, J.T. Ritchie, R.S. Kanwar, and C.W. Mize. 1998.
Incorporation of a subsurface tile drainage component into a soybean growth
model. Trans, ofthe ASAE 41(5): 1305-1313.
Sites, J.W., I.W. Wander, and E.J. Deszyck. 1953. The effect of fertilizer timing and rate
of application on fruit quality and production ofHamlin oranges. Proc. Fla.
State Hort. Soc. 66:54-62.
Slatery, R.O. 1967. Plant-wat^relationships. Academic Press, London and New York.
Smajstrla, A.G., L.R. Parsons, F.S. Zazueta, G. Vellidis, and K. Aribi. 1986. Water use
and growth of young citrus trees. Paper No. 86-2069. Am Soc. Agrie. Engr.
Summer meeting, San Luis Obispo, CA.
Smith, P.F. 1956. Effect of phosphate fertilization on root growth, soil pH, and chemical
constituents at different depths in an acid sandy Florida soil. Proc. Fla. State Hort.
Soc. 69: 25-29.
Smith, P.F. 1965. Effect of nitrogen source and placement on the root development of
Valencia orange trees. Proc. Fin. State Hort. Soc. 78: 55-59.
Stewart, E.H, J.E. Browning, and E.O Burt. 1969. Evapotranspiration as affected by
plant density and water-table depth. Trans. Am. Soc. Agrie. Engr. 12(5):646-647.
Starr, J.L., and I.C. Paltineanu. 1998. Soil water dynamics using multisensor capacitance
probes in nontraffic Interrow of Com. Soil Sci Soc. Am. J. 62:114-122.

172
Sumner, D M. 1996. Evapotranspiration from successional vegetation in a deforested
area of the Lake Wales Ridge, Florida. Water-Resources Investigation. 96-4244.
Syvertsen, J.P., and J.J. Lloyd. 1994. Citrus. In. Handbook of Environmental Physiology
of Fruit Crops, Volume H: Sub-tropical and Tropical Crop. Schaffer, B. and P C.
Andersen (eds.) Boca Raton, FI, CRC Press Inc., USA.
Syvertsen, J.P., M L. Smith, and B.J. Boman. 1993. Tree growth, mineral nutrition and
nutrient leaching losses from soil of salinized citrus. Agr. Ecosyst. Environ.
45:319-334.
Syvertsen, J.P., and M.L. Smith. 1996. Nitrogen uptake efficiency and leaching losses
from lysimeter-grown citrus trees fertilized at three nitrogen rates. J. Amer. Soc.
Hort. Sci. 121:57-62.
Timmer, L.W., and S.E. Zitko. 1996. Evaluation of a model for prediction of post bloom
fruit drop of citrus. Plant and Disease 80:380-383.
Tindall, J. A., and J. R. Kunkel. 1999. Unsaturated Zone Hydrology for Scientists and
Engineers. Prentice-Hall, Upper Saddle River, New Jersey.
Tinker, P. B., and P. H. Nye. 2000. Solution Movement in the Rhizosphere. Oxford
University Press, New York, New York.
Tucker, D. P., C. G. Erickson, and K. T. Morgan. 1997. Middles Management Methods
in Citrus Affect Soil Moisture Retention and Vegetation Species. Proc. Fla. State
Hort. Soc. 110: 39-43.
Tucker, D.P.H., AK. Alva, L.K. Jackson, and T.A. Wheaton. 1995. Nutrition of Florida
citrus trees. Univ. of Fla. Coop. Ext. Serv. SP169.
Turrell, F.M., M.J. Garber, W.W. Jones, W.C. Cooper, and RH. Young. 1969. Growth
equations and curves for citrus trees. Hilardia 39(16):429-445.
Vaile, R.S. 1924. A survey of orchard practices in the citrus industry of southern
California. Cal. Agrie. Exper. Sta. Bul.374. 40 pp.
Valiente, J. L. and Albrigo, L. G. 2000. Modeling flowering date of sweet orange
(Citrus sinensis (L.) Osbeck) trees in central Florida based on historical weather
records Proceedings of the International Society of Citriculture.
Verma, A., R.S. Kanwar, and U.S. Tim. 1995. Modification of DRAINAGE model by
using the nitrogen component form the GLEAMS model. Trans, of the ASAE
38(3):717-724.

173
Wagner-Riddle, C., T.J. Gillespie, LA Hunt, and C.J. Swanton. 1997. Modeling a rye
cover crop and subsequent soybean yield. Agron. J. 89:208-218.
Wallace, R.R., and F.E. Gardner. 1945. Seasonal absorption of nutrient ions by orange
trees in sand culture. Proc. Fla. State Hort. Soc. 58:25-36.
Webber, H.J., and L.d. Batchelor. 1943. The Citrus Industry Volume I Histroy, Botany
and Breeding. University of California Press. Berkely, C A, USA.
Wiegand, C.L., and W.A Swanson. 1982a. Citrus responses to irrigation: I. Irrigation
requirements; daily, monthly, and annual evapotranspiration amounts; and water
management recommendations. J. Rio Grande Valley Hort. Soc. 35:73-85.
Wiegand, C.L., and W.A. Swanson. 1982b. Citrus responses to irrigation: EL Fruit yield,
size, and number. J. Rio Grande Valley Hort. Soc. 35:87-95.
Wiegand, C.L., and W.A Swanson. 1982c. Citrus responses to irrigation: HI. Tree trunk
and canopy growth. J. Rio Grande Valley Hort. Soc. 35:97-107.
Wiegand, C.L., W.A. Swanson, and R.R. Cruse. 1982. Marrs, Valencia, and Ruby Red
juice quality as affected by irrigation plus rainfall. J. Rio Grande Valley Hort.
Soc. 35:109-120.
Whitney, J.D., A Elezaby, W.S. Castle, T.A. Wheaton and R.C. Littell. 1991. Citrus tree
spacing effects on soil water use, root density and fruit yield. Transactional ASAE
34(1). 129-134.
Xie, Y., J.R. Kiniry, V. Nedbalek, and W.D. Rosenthal. 2001. Maize and sorghum
simulations with CERES-Maize, SORKAM, and ALMANAC under water-
limiting conditions. Agron. J. 93:1148-1155.
Xin, J., F.S. Zazueta, AG. Smajstrla, T.A Wheaton, J.W. Jones, P.H. Jones, and D.D.
Dankel. 1997. CIMS an integrated real-time computer system for citrus
microirrigation management. Applied Engineering in Agriculture 13:785-790.
Zaongo, C.G.L.,L.R. Hossner, and C.W. Wendt. 1994. Root distribution, water use, and
nutrient uptake of millet and grain sorghum. Soil Sci. 157(6):379-388.
Zur, B., J.W. Jones, K.J. Boote, and L.C. Hammond. 1982. Total Resistance to water
flow in field soybeans II. Limiting soil moisture. Agron. J. 74(1):99-105.

BIOGRAPHICAL SKETCH
Kelly Tindel Morgan was bom in Columbus, Georgia, in 1958. He grew up and
was educated in Winter Haven, Florida graduating from Winter Haven High School in
1976. Kelly earned an AA degree in chemistry and biology from Polk Community
College in 1978. He married Nancy Greives in 1978 and enrolled at the University of
Florida that fall. He earned his BS degree in 1980 majoring in plant pathology; later he
earned his MS degree in plant pathology in 1982. Kelly worked as Assistant Manager of
a large citrus nursery from 1982 to 1985 during which time his two sons, Joshua and
Christopher, were bom. From 1985 to 1988 he managed citrus groves for investors after
which he worked for the University of Florida at the Citrus Research and Education
Center from 1988 to the present. Currently, he is a Scientific Research Manager directing
horticultural maintenance, treatment application, and data collection on more than 100
acres of citrus research plots associated with the Conserv II reclaimed water project near
Orlando, Florida
174

I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
(X .
Thomas A. Obreza, Chair
Professor of Soil and Water Science
I certify that 1 have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Johannas M. S. Scholberg, Cochair
Assistant Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
es W. Jones,
distinguished Professor of Agricultural and
Biological Engineering
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Nicholas B. Comerford *
Professor of Soil and Water Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Thomas Adair Wheaton
Professor Emeritus of Horticultural Science
This dissertation was submitted to the Graduate Faculty of the Collage of Agricultural and Life
Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements
for the degree of Doctor of Philosophy.
May 2004 v ^ v. \ ^
Dean, College of Agricultural and LifejSciences
Dean, Graduate School



152
Nitrification Estimation
Mean NO3-N content in the upper 45 cm of soil increased 24 h after application to
151% of NO3-N applied. The mean NO3-N content decreased during the next 48 h to
126% of total NO3-N applied. Content of NH4-N decreased to 34% of that applied after
24 h and steadily declined after application to 16% of NH4-N applied on day 3. The sum
of NO3-N and NH4-N (total mineral N) declined throughout the period to 73.9% of total
N applied on day 3, indicating that approximately 26.1% N was lost in 3 days due to
volatilization of NH4-N, microbial activity, or incorporation in organic matter. The
nitrification rate in this soil was very rapid. With an average of more than 50% of the
NH4-N converted to NO3-N within the first 24 h, nearly 85% of applied NH4-N was
converted to NO3-N in 3 days assuming all of the N loss was NH4-N.
Seasonal Tissue N Concentration
Leaf and twig N percentages were near maximum from August to February and at
minimum in May of each year. Minimum leaf N concentrations were 1.9% and 2.0% for
the flush and expanded leaves, respectively, on trees fertilized with 179 kg N ha'1 yr'1.
Minimum leaf N concentration for trees fertilized with 269 kg N ha'1 yr'1 were 2.0 and
2.2% for flush and expanded leaves, respectively. Maximum leaf N concentrations were
2.5 and 2.7% for 179 and 269 kg ha'1 annual application rates, respectively. The areas of
both flush and expanded leaves were generally greater for trees receiving the high annual
N rate. Minimum twig N concentrations occurred in May and were 0.83% for both annual
N application rates. Maximum twig N concentrations of 0.98 and 1.02% for the 179 and
269 kg N ha'1 annual application rates, respectively, occurred in August.


58
Table 3-5. Linear regression analysis of dry weight and N accumulation in different tree
components as related to tree canopy volume (TCV)Z
Yo
a
R2
RMSEy
P
Dry Weight (kg tree'1)
Total Mass
-2.21
2.79
0.97
0.15
<0.0001
Above Ground
-2.09
1.95
0.96
0.17
<0.0001
Below Ground
0.21
0.72
0.91
0.27
<0.0001
Leaves
0.09
0.36
0.95
0.20
<0.0001
Twigs
0.26
0.13
0.76
0.49
<0.0001
Sm. Branches
-0.57
0.37
0.90
0.29
<0.0001
Med. Branches
-0.31
0.23
0.80
0.43
<0.0001
Lg. Branches
-.24
0.36
0.82
0.53
<0.0001
Total Branches
-2.47
1.25
0.92
0.27
<0.0001
Trunk
0.54
0.07
0.72
0.36
<0.0001
Sm. Roots
0.41
0.13
0.87A
0.27
<0.0001
Med. Roots
0.32
0.14
0.83
0.33
<0.0001
Lg. Roots
-0.24
0.18
0.85
0.37
<0.0001
Tap Root
0.06
0.12
0.60
0.57
<0.0001
Nitrogen Weight (g tree'1)
Total Mass
-3.24
23.41
O.97)
0.16
<0.0001
Above Ground
-7.28
17.52
0.96
0.17
<0.0001
Below round
5.51
5.36
0.92
0.23
<0.0001
Leaves
3.50
8.70
0.92
0.24
<0.0001
Twigs
0.09
1.83
0.85
0.34
<0.0001
Sm. Branches
-3.32
2.04
0.96
0.20
<0.0001
Med. Branches
-1.28
1.00
0.78
0.45
<0.0001
Lg. Branches
-3.46
2.44
0.88
0.32
<0.0001
Total Branches
-11.31
5.71
0.94
0.23
<0.0001
Trunk
2.80
0.35
0.65
0.40
<0.0001
Sm. Roots
5.38
1.85
0.86
0.28
<0.0001
Med. Roots
1.82
1.34
0.88
0.29
<0.0001
Lg. Roots
-1.10
0.88
0.89
0.32
<0.0001
Tap Root
0.68
0.55
0.63
0.53
<0.0001
z Y = Yo +aX where X = TCSA, and Yo and a are regression coefficients
y RMSE dry weight kg dw tree'1, and N accumulation = g N tree1


25
known as Richards equation (Tinker and Nye, 2000). In this equation, flow in
unsaturated soil can is expressed in terms of the water content gradient and soil water
diusivity (De).
u = -Ke d/dx = -Ke (d/d0) (d0/dx) = -De (d0/dx) Equation 2-3
Where:
u = water flux (cm3 cm'2 s'1),
Ke = hydraulic conductivity constant at 0 (cm s'1),
= soil water potential (kPa),
a >>
0 = soil water content (cm cm' ),
d0/d x = the distance over which the flux is maintained (cm).
The term diffusivity (De) is used because the form of equation is the same as that
of Ficks law of diffusion (Tinker and Nye, 2000). Furthermore, De is somewhat less
convenient than Ke under conditions of hysteresis because De is discontinuous at each
reversal of the direction of potential, while K is continuous and virtually hysteresis-free
(Hagan et al., 1967). Experimentally, the effect of hysteresis on Richards equation has
usually been ignored by limiting the soil water potential change to either always drying,
i J
or always wetting.
Field capacity (0fc) describes the water content held in the soil after excess
water has drained to drier soil layers by redistribution. This equilibrium can be
determined in the field by measuring the soil water content as a function of time to
determine the value of 0 when d0/dt approaches zero. Hillel (1971) noted that the rate at
which d0/dt approaches zero is dependent on 0¡ and the depth to which the soil is being


131
Fig. 6-3. Proportions of nitrate-N, ammonium-N, and total-N from control pipes as
percentage applied during 3 days after application. Initial increase in nitrate-N is
assumed to be from nitrification of ammonium-N. Low soil pH limits
volatilization, thus loss in total-N was assumed to be due to microbial processes.
averaged across rootstocks. Respective N values were near maximum from August to
February with minimums in May of each year (Fig. 6-4). Minimum leaf N concentrations
were 1.9 and 2.0% for the flush and expanded leaves, respectively, for trees fertilized
with 179 kg N ha'1 yr'1. Minimum leaf N concentration from trees fertilized with 269 kg
N ha'1 yr'1 were 2.0 and 2.2% for flush and expanded leaves, respectively. Maximum leaf
N concentrations were 2.5 and 2.7% for 179 and 269 kg N ha'1 annual N application
rates, respectively. Leaf areas for both flush and expanded leaves were generally greater
for leaves from trees receiving the higher annual N rate (Table 6-5). No such trend was


12
trunk, lateral roots, and fibrous roots constituted 13.9, 37.4, 9.0,10.2 and 14.2% of total
dry biomass, respectively.
In a 15N study, Feigenbaum et al. (1987) divided 22 year-old Shamouti orange
trees fertilized at two annual N rates into seven components: 1) leaves, 2) twigs <5 mm,
3) branches >5 mm, 4) trunk and main branches, 5) main root (tap), 6) lateral roots >1
mm, and 7) fibrous roots <1 mm. There was no diameter given to differentiate between
branches and main branches. Biomass percentages for leaves were 6.4 and 8.4% for the
low and high N treatments respectively. Similar differences were found for all categories
with the exceptions of main and lateral roots. The ranges in mean percentage of fresh
biomass for twigs, branches, trunk and main branches, lateral roots, and fibrous roots
were 1.6 to 2.2, 32.5 to 33.4, 24.5 to 29.2,4.6 to 6.1, and 3.8 to 4.8%, respectively. These
differences between fertilizer-N rates resulted in a greater percentage of biomass below
ground for the low nitrogen application (34.1%) compared with that of the higher
nitrogen rate (27.8%).
In a similar study, Kato et al. (1984) found different values for 21 year-old
Satauma mandarin trees in Japan. In this study, mean percentages of dry biomass for
leaves, twigs, branches, trunk, lateral roots and fibrous roots were 15.7,4.7, 32.2, 23.1,
20.3, and 3.1% respectively.
Citrus Nitrogen Accumulation and Partitioning
Nitrogen balance studies in citrus provide information on physiological tree N
requirements and can be used to develop methods that minimize potential losses of N to
groundwater and the atmosphere (Feigenbaum et al. 1987). In a study of 8 year-old
Valencia trees in California conducted during a period of 2 years (Cameron and


17
mandarin had a highly developed lateral root system at the surface, but few fibrous roots
below the surface. Menocal-Barberena (2000) found no statistically significant
differences in vertical or horizontal fibrous root distribution ofHamlin orange on
Cleopatra mandarin, Swingle citrumelo and Carrizo citrange rootstocks. Vertical and
horizontal root distribution were similar to other studies with about 40% of the fibrous
roots in the top 30 cm and 9 to 14% at each of the remaining 30 cm depth increments to
180 cm. Few roots were found below 180 cm.
Tree Spacing and Density
Due to Floridas rainy season, roots of trees at commercial spacing rapidly
occupy the volume of soil outside the irrigated zone. After canopy closure, they extend
into the rootzone of adjacent trees. Elezaby (1989) reported fibrous root concentration in
the 0 to 30 cm zone increased from 450 to 1000 g m'3 between trees when the in-row
distance decreased from 4.5 to 2.5 m. The increased root concentrations in this study
were concluded to be the result of overlapping root systems. Trees at the closest spacing
showed root concentration increases to depths of 150 cm (Elezaby, 1989).
Fertilization
Increases in fertilizer N can increase root growth to a considerable depth, but the
largest effects generally occurred near the surface (Ford, 1953b; Ford et al., 1957; Smith,
1956; Smith, 1965).
Irrigation
Irrigation method and scheduling has been shown to change the distribution
and/or concentration of citrus fibrous roots. In a California study of trees receiving
different irrigation treatments, yield was not correlated with fibrous root density (Cahoon


138
many trees. Therefore, only blooms, leaves and fruit were included in this study.
Seasonal cumulative dry masses and N masses for each of these tissues as a function of
time are shown in Figs. 6-6 and 6-7 for 2001 and 2002, respectively.
Bloom dry mass accumulated in March and April of each year. Cumulative
senesced flower weights were 529 and 764 g tree'1 for 2001 and 2002, respectively.
Cumulative flower N masses were 18.8 and 21.9 g N tree'1 for the same years. Fruit and
leaf biomass losses were greatest from April through May and September through
December of each year. Little loss of either of these tissues occurred from June through
August.
Cumulative weight loss associated with fruit drop varied more than bloom and
leaf biomass, with 1635 and 946 g removed in 2001 and 2002, respectively. These fruit
biomass values represented a cumulative loss of 30.1 and 16.8 g of N tree"1 for these
years. Leaf biomass losses were 2041 and 2806 g tree'1 for 2001 and 2002, respectively.
Cumulative annual N loss from leaf fall amounted to 42.2 and 51.4 g tree'1 for the same
years.
Discussion
Seasonal N uptake varied by month of year, with May having the highest uptake rate of
the 3 months studied. This result was consistent with other uptake studies using young
citrus trees in lysimeters (Syvertsen and Smith, 1996; Lea-Cox et al., 2001), and
greenhouse studies (Lea-Cox and Syvertsen, 1996; Scholberg et al., 2002). Increased
uptake rates could have been due to higher soil temperature (Scholberg et al., 2002) or
greater tree N demand (Dasberg, 1987; Feigenbaum et al., 1987; and Legaz et al., 1982).


170
Obreza, T. A, D.J. Pitts, L.R. Parsons, T.A. Wheaton, and K.T. Morgan. 1997. Soil
water-holding characteristics affects citrus irrigaiton scheduling strategy. Proc.
Fla. State Hort. Soc. 110:36-39.
Paramasivam, S., A.K. Alva, O. Prakash, and S.L. Cui. 1999. Denitrification in the
vadose zone and in surficial groundwater of a sandy Entisol with citrus
production. Plant and Soil 208:307-319.
Parsons, L.R., T.A. Wheaton, and S.W. Castle. 2001. High application rates of
reclaimed water benefit citrus tree growth and fruit production. Hort. Science
36:1273-1277.
Prajamwong, S., G. P. Merkley, and R. G. Allen. 1997. Decision Support Model for
Irrigation Water Management. J. Irrig. And Drain. Engrg., ASAE, 122(2): 106-
113.
Piper, E.L., K.J. Boote, J.W. Jones, and S.S. Grimm. 1996. Comparison of two penology
models for prediction flowering and maturity date of soybean. Crop Sci. 36:1606
1614.
Quemada, M., and M.L. Cabrera. 1995. CERES-N model predictions of nitrogen
mineralized from cover crop residues. Soil Sci. Soc. Am. J. 59:1059-1065.
Reck, W.R. 1994. GLEAMS modeling of BMPs to reduce nitrate leaching in middle
Suwannee River area. Environmentally Sound Agriculture: Proceedings of
Second Conference 20-22 April 1994.
Reitz, H.J. and W.T. Long. 1955. Water table fluctuation and depth of rooting of citrus
trees in the Indian River area. Proc. Fla. State Hort. Soc. 68:24-29.
Reitz, H.J. 1956. Timing fertilization of citrus in the Indian River area. Proc. Fla. State
Hort. Soc. 69:58-64.
Reuther, W., P.F. Smith, G.K. Scudder, and G. Hmciar. 1957. Responses ofValencia
orange trees to timing, rates, and ratios of nitrogen fertilization. Proc. Amer. Soc.
Hort. Soc. 70:223-236.
Reyes, M R., R.L. Bengtson, and J.L.Fouss. 1994. GLEAMS-WT hydrology submodel
modified to include subsurface drainage. Trans, of the ASAE 37(4): 1115-1120.
Rogers, J.S., and J.F. Bartholic. 1976. Estimated evapotranspiration and irrigation
requirements for citrus. Proc. Soil and Crop Sci. Soc. Fla. 35:111-117.
Rogers, J.S., L.H. Allen, and D. V. Calvert. 1983. Evapotranspiration from a developing
citrus grove in himid climate. Trans. Am. Soc. Agrie. Eng. 26:1778-1783.


94
Materials and Methods
Site Characteristics
Hamlin orange (Citrus sinensis L.) grafted on Carrizo citrange (C. sinensis L.
Osbeck X Poncirus trifoliata L. Raf.) rootstock trees planted 3.1 m in the row and 6.1 m
between rows (used in Experiment 1 of Chapter 3) were used in this study. The trees had
been pruned each of the past 3 years and had formed a hedgerow approximately 3.8 m
wide and 5.9 m tall. Herbicides were applied as needed to maintain a nearly weed-free
strip 3.5 to 4.0 m wide beneath the tree canopies. Soil type at the site was Candler fine
sand (hyperthermic, uncoated Typic Quartzipsamments) with a field capacity of
approximately 0.08 cm3 cm"3 (Obreza et al., 1997). Irrigation was applied to the tree row
using one microsprinkler per tree with a flow rate of approximately 60.5 L hr"1 and a 360
degree spray pattern with a diameter of approximately 3.7 m. The equivalent mean
precipitation rate of the sprinkler was 0.58 cm hr"1. Reclaimed municipal-waste water
provided by the Water Conserv II Project was used as the source of irrigation water.
Soil Capacitance Sensor Data Collection
Soil water content at 10, 20, 40, and 80 cm depths was recorded at 30 minute
intervals during a 2-year period in the irrigated and non-irrigated areas under three
mature (14-year-old) citrus trees. These increments represent soil depths of 0 to 15,15 to
30, 30 to 60, and 60 to 100 cm, respectively (Fig. 5-1). These 0 data were obtained using
EnviroSCAN (Sentek Pty. Ltd., South Australia) capacitance sensors installed two
distances away from the tree trunk in the row and three distances perpendicular to the
row (Fig. 5-2). In-row sensors were placed 0.75 m from the tree trunk and at the midpoint


2
nutrient use efficiencies and assessment of Best Management Practices (BMP) impacts
on citrus production and ground water quality.
Ridge Water Quality Study
The U S. Environmental Protection Agency, in a nation-wide survey, documented
widespread nitrate contamination of shallow drinking water wells (Graham and Alva,
1995). In that survey, approximately 55% of wells were found to contain NO3-N
contamination above the background concentration. Approximately 1.2% and 2.4% of
urban and rural drinking water wells, respectively, were found to contain NO3-N
concentrations above the Maximum Contamination Level (MCL) of 10 mg L'1 for
drinking water. A correlation between drinking water well contamination and areas with
higher fertilizer sales and high value crops was established (Graham and Alva, 1995)
suggesting that agricultural fertilization practices may have contributed significantly to
NO3-N contamination of drinking water wells.
Of 3949 drinking water wells surveyed by the Florida Department of Agriculture
and Consumer Services (FDACS), 2483 (63%) contained detectible concentrations of
NO3-N (Graham and Wheaton, 2000). Of these 2483 contaminated wells, 584 (15% of
total surveyed) contained NO3-N in excess of MCL. The proportion of wells in Florida
contaminated with NO3-N was similar to that of the nation-wide survey. However, the
proportion of wells contaminated above MCL was an order of magnitude higher,
suggesting that the soils of the state of Florida on average are vulnerable to NO3-N
leaching to groundwater. Eighty-nine percent of wells contaminated above MCL were
located in the central Florida counties of Lake, Polk, and Highlands (Fig. 1-1). Portions
of these three counties comprise the central Florida ridge. Soils typical of the ridge are


7
C^^Oroundwat
3
Inputs and Outputs
State Variables
txZZH
Functions
Fig. 1-2. Plant/soil nitrogen and water balance flow chart.


42
Materials and Methods
Citrus trees of various sizes were measured and dissected into constituent parts
during a period of 1 year. Representative tissue samples of constituent parts for each tree
were weighed and analyzed to estimate total dry mass, relative percentage dry mass, and
N content of the various tree components. These data were used to determine N
allocation between different tree constituents.
Experiment 1 Mature Citrus Biomass and N Distribution
Two sets of six trees each were dissected in February 2001 and January 2002 at
the Water Conserv II site near Winter Garden in western Orange county, Florida. Both
sets of trees were 14 year-old Hamlin orange trees planted in 1987 at a spacing of 3 m
between trees in the row and 6 m between rows resulting in a tree density of 556 trees ha
V Three trees of each set were budded on Swingle citrumelo (iCitrus parodist Macf. x
Poncirus. trjfoliata (L.) Raf) rootstock, and the remaining three trees of each set on
Carrizo citrange (G sinensis L. Osbeck X P. trifoliata L. Raf.) rootstock. All trees had
been fertigated at an annual rate of240 kg N ha1. The trees were irrigated (and
fertigated) with reclaimed water containing approximately 7 mg NO3-N L'1.
Experiment 2 Biomass and N Accumulation with Increase Tree Size
A third set of seven Valencia trees on Swingle citrumelo rootstock were
dissected at the K.D. Revell grove owned by Cargill, Inc. near Fort Meade in southern
Polk county, Florida. Fresh, dry, and N weights were determined for the same constituent
parts as in experiment 1. These trees had been fertilized using dry chemical fertilizer
three or more times per year and irrigated with well water.


147
averaged 20-24% of total tree weight. Mature trees on Carrizo citrange rootstocks were
significantly larger than those on Swingle citrumelo. The proportion of large branch
biomass with respect to total tree biomass was significantly greater for trees grown on
Carrizo compared with trees grown on Swingle. On a percentage basis, taproot biomass
was significantly greater for trees on Swingle than those on Carrizo. Mean total biomass
of large roots and taproots was greater for trees on Swingle than for those on Carrizo.
Biomass Vs Tree Size Relationships
Total leaf area per tree was proportional to both tree canopy volume and average
trunk diameter. Maximum total fresh and dry weights for trees with canopy volumes
ranging from 28 to 38 m3 and trunk cross sectional area of 130 to 150 mm were
approximately 160 and 100 kg tree'1, respectively. Above-ground and below ground dry
weights of large citrus trees were approximately 74 and 26 kg, respectively. Leaf biomass
declined from approximately 20% of total biomass for trees with canopy volumes of less
than 5 m3 and trunk cross sectional area of less than 60 mm to approximately 12% of
total biomass for trees with canopy volumes greater than 30 m3 and trunk cross sectional
area greater than 120 mm. Likewise, twig biomass decreased from 11% of biomass to
6% for trees in the same size categories. Total branch dry biomass increased from 15 to
45% across the range of small to large trees, while trunk biomass concurrently decreased
from 12 to 3%.
Nitrogen Distribution
As with dry biomass, estimated total N mass was greater for trees on Carrizo
rootstock compared with trees on Swingle. Mean leaf, branch, and root N were
approximately 45, 35, and 20% of total N for mature citrus trees. Trees grown on Carrizo


173
Wagner-Riddle, C., T.J. Gillespie, LA Hunt, and C.J. Swanton. 1997. Modeling a rye
cover crop and subsequent soybean yield. Agron. J. 89:208-218.
Wallace, R.R., and F.E. Gardner. 1945. Seasonal absorption of nutrient ions by orange
trees in sand culture. Proc. Fla. State Hort. Soc. 58:25-36.
Webber, H.J., and L.d. Batchelor. 1943. The Citrus Industry Volume I Histroy, Botany
and Breeding. University of California Press. Berkely, C A, USA.
Wiegand, C.L., and W.A Swanson. 1982a. Citrus responses to irrigation: I. Irrigation
requirements; daily, monthly, and annual evapotranspiration amounts; and water
management recommendations. J. Rio Grande Valley Hort. Soc. 35:73-85.
Wiegand, C.L., and W.A. Swanson. 1982b. Citrus responses to irrigation: EL Fruit yield,
size, and number. J. Rio Grande Valley Hort. Soc. 35:87-95.
Wiegand, C.L., and W.A Swanson. 1982c. Citrus responses to irrigation: HI. Tree trunk
and canopy growth. J. Rio Grande Valley Hort. Soc. 35:97-107.
Wiegand, C.L., W.A. Swanson, and R.R. Cruse. 1982. Marrs, Valencia, and Ruby Red
juice quality as affected by irrigation plus rainfall. J. Rio Grande Valley Hort.
Soc. 35:109-120.
Whitney, J.D., A Elezaby, W.S. Castle, T.A. Wheaton and R.C. Littell. 1991. Citrus tree
spacing effects on soil water use, root density and fruit yield. Transactional ASAE
34(1). 129-134.
Xie, Y., J.R. Kiniry, V. Nedbalek, and W.D. Rosenthal. 2001. Maize and sorghum
simulations with CERES-Maize, SORKAM, and ALMANAC under water-
limiting conditions. Agron. J. 93:1148-1155.
Xin, J., F.S. Zazueta, AG. Smajstrla, T.A Wheaton, J.W. Jones, P.H. Jones, and D.D.
Dankel. 1997. CIMS an integrated real-time computer system for citrus
microirrigation management. Applied Engineering in Agriculture 13:785-790.
Zaongo, C.G.L.,L.R. Hossner, and C.W. Wendt. 1994. Root distribution, water use, and
nutrient uptake of millet and grain sorghum. Soil Sci. 157(6):379-388.
Zur, B., J.W. Jones, K.J. Boote, and L.C. Hammond. 1982. Total Resistance to water
flow in field soybeans II. Limiting soil moisture. Agron. J. 74(1):99-105.


value of920 mm for 3-yr old Hamlin orange trees grown on deep sandy soils in central
Florida. Koo and Harrison (1965), and Koo and Hunter (1969) reported annual ETC
values of 1170 mm for mature citrus on the same soil series.
Climate
Mean annual ETC for citrus in Florida ranges from 820 to 920 mm (Rogers and
Bartholic 1976; Fares and Alva 1999) for young (<5 years) trees to 1170 to 1280 (Koo
1978, Rogers and Bartholic 1976) for mature (10 years or more) trees. Annual ETe values
reported for mature citrus grown in the lower Rio Grande Valley of Texas are similar to
those for Florida and ranged from 1044 to 1232 mm (Wiegand et al. 1982). Hoffinan et
al. (1982) reported annual ETC values for well-irrigated citrus grown in semi arid Arizona
of 1470 mm. Lower ETC rates for Florida (humid) compared with Arizona (semi-arid)
have been attributed to lower evaporative demand (Rogers et al. 1983, Fares and Alva
1999).
Soil characteristics
Crop water supply must be based on a clear understanding of soil water dynamics.
Water in excess of field capacity drains through the vadose zone. Eventually, water that is
not taken up by plants or evaporated from soil or plant surfaces makes its way into the
ground water and contributes to aquifer recharge (Fares and Alva, 1999). Under-tree
sprinklers and drip irrigation systems are designed to deliver water at rates low enough to
allow infiltration into the soil without contributing to losses by runoff. These systems can
be managed in such a manner that the excessive downward drainage through the soil is
minimized. The required application amount is governed by the soil-water depletion on a
given irrigation date, irrigation efficiency, and the target soil-water level. Most of the


Ill
trends follow reported ETC values for citrus under both humid and arid climatic
conditions (Boman, 1994; Castel et al., 1987; Doorenbos and Pruitt, 1977; Martin et al.,
1997; and Rogers et al., 1983). Reported IQ values for central Florida ranged from
approximately 0.6 in the winter to 1.1 in summer (Boman, 1994; Fares and Alva, 1999;
Rogers et. al., 1983). Estimated mean daily IQ values ranged from 0.55 to 1.2 for citrus
under semi-arid to arid conditions (Doorenbos and Pruitt, 1977; Hoffman et al., 1982;
Martin et al., 1997; and Wiegand et al., 1982). Thus, seasonal mean ETC to ET0 ratios
reported for this study fall in the range of values documented in the literature.
Allen et al. (1997) refered to estimated Kc values from soil water content
measured several days apart as time-averaged Kc and stated that these values are affected
by the evaporative power of the atmosphere. They further stated that the higher the
evaporative power of the atmosphere, the faster the soil will dry between water
applications, and the smaller the time-averaged IQ will be. The reduction in ETC with
lower 0 and reflected by Ks decreasing from 1 to 0.6 as ASWD increased from 10 to
50% seems rather extreme. However, Rogers et al. (1983) suggested that lower estimated
Kc values in the spring were caused by low rainfall and low 0 outside the irrigated zone.
Their reported Kc values of0.77, 0.72, and 0.95 for March, April and May are 81.3, 71.3,
and 89.6% of the Kc values estimated using the regression equation in Table 5-3.
Estimated IQ during the rainy season months of June and July were 101.8 and 92.2% of
calculated values using the same equation, indicating that IQ values esi mated from
monthly water balances can lead to lower estimates of IQ during periods of little rainfall
and high evaporative demand.


Leaf Area Index Leaf Area Index
55
Canopy Volume (m3)
0 20 40 60 80 100 120 140 160 180
Trunk Cross Section Area (cm2)
Fig. 3-3. Leaf area index on a tree basis expressed as a function of tree canopy volume
(A)and trunk cross-section area (B).


157
Determining root system distribution patterns of the various rootstocks on which citrus is
propagated is essential to irrigation scheduling for improved fertilizer N uptake
efficiency.
While current N BMPs are believed to reduce N leaching below the citrus root
zone, more accurate methods of determining annual fertilizer N rates, application timing,
and irrigation scheduling are essential to improve uptake efficiency and enhance
environmental protection. This study confirmed that the basic principles used to derive
citrus N BMPs are sound, but also generated data suggesting that refining and fine-tuning
BMPs may be necessary to make them totally effective. Adoption of the N balance
approach to fertilization evapotranspiration-based irrigation scheduling, and rootstock
specific N fertilizer management are essential to sustainable productivity and improved
ground water quality.


136
Fruit N concentration was highest in May of each year at 1.25 and 1.30% for low
and high N application rates, respectively (Table 6-5). Nitrogen concentrations decreased
through the season to means of 1.03 and 1.06% for low and high N application rates just
prior to harvest. Mean fruit diameter did not differ due to N application rate. However,
fruit dry masses were greater for the lower application rates.
Seasonal Fertilizer Use Efficiency
Seasonal changes in tree N content from March to May were estimated using the
mean biomass values for trees grown on Carrizo and Swingle rootstocks in chapter 3
multiplied by the mean reduction in N for the period. Nitrogen reduction for leaves was
49.0 and 42.6 g tree'1 for Carrizo and Swingle trees, respectively. This reduction was
relatively large compared with N reduction in twigs and branches of 26.4 and 18.4 g tree*
1 for trees on the same rootstocks. Thus, total estimated N reduction due to reduction in N
concentration from March to May was 75.4 and 61.0 g tree'1 for trees grown on Carrizo
and Swingle rootstocks, respectively. The amount of tree N gained from spring flush and
developing fruit must be added to these values. However, the amount of biomass gained
through spring flush was not measured in these studies. It is the authors observation that
more biomass is gained in the spring of the year compared with summer vegetative
flushes, but the amount of biomass has not been determined. To estimate it, 75% of the
reported biomass gained for these trees in 2001 (chapter 3) was used. Biomass values in
Table 6-5 indicate that approximately 30% of fruit biomass is accumulated prior to the
May sampling. Therefore, 30% of the mean fruit biomass collected prior to the 2002
destructive tree sampling was used as an estimate of fruit biomass gain between March
and May. These estimated biomass values were multiplied by the May periodic sampling


39
and N accumulation in different tree components change with age of perennial crops due
to the increase in weight of larger branches and trunks of older trees to support the
increased tree biomass. Similar changes occur in annual crops with increase in biomass
between emergence and harvest. However, as with annual crops, tree size is not
dependent on age alone; rootstock (Castle, 1978, 1980), crop nutrition (Feigenbaum et al.
1987), irrigation (Parsons et al., 2001), and restriction of the root system (Mataa and
Tomingag, 1998) can limit growth of a citrus tree. Thus, these factors can result in trees
of equal age being very different in size, biomass, and N content.
Many crop models such as CERES, CROPGRO, and DSSAT determine the effect
of assimilate costs for vegetative and reproductive growth and N budget through C and N
balances with increase in crop biomass on water and N uptake, growth, and yield of
agronomic crops (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995; and Sexton
et al., 1998). Likewise, optimal irrigation and nutrient management is dependent upon the
estimation of biomass and N content in citrus trees. Therefore, the relationship of tree
size to biomass and N accumulation is needed.
Previous studies have correlated long-term citrus tree biomass and N
accumulation with tree age (Cameron and Appleman, 1935; Cameron and Compton,
1945; Feigenbaum et al., 1987; Kato et al., 1984; Mattos, 2000). Leaves of 3.5, 7, and 15
year-old Valencia trees in California contained from 40 to 50% of total tree N, while
twigs and shoots contained approximately 10% of total N (Cameron and Appleman 1935;
Cameron and Compton, 1945). Trunk and branches contained from 20 to 30% of total
tree N, approximately half of which was in the bark, a tree component that represents
only 5% of the total dry mass. The roots contained from 15 to 20% of tree N, half or


104
Table 5-2. Regression analysis of estimated ETC to calculated ET0 ratio by mean soil
water content, soil water potential, and available soil water depletion in the upper 0.5 and
1.0 m of soil for either the irrigated zone or total allocated tree area.
Soil Water Content Cubic Function2
Yo
a
b
c
R2
RMSE
P
0.5 m irrigated
8.32
-398.0
6626
-35004
0.52
0.12
<0.0001
1 m irrigated
3.02
-145.6
2695
-14903
0.36
0.14
<0.0001
1 m total
4.15
-209.6
3876
-21806
0.42
0.13
<0.0001
Soil Water Potential
- Exponential Decay Functiony
Yo
a
b
R2
RMSE
P
0.5 m irrigated
0.51
1.12
0.14
0.44
0.13
<0.0001
1 m irrigated
0.41
1.16
0.11
0.38
0.13
<0.0001
1 m total
0.37
1.33
0.10
0.46
0.13
<0.0001
Available Soil Water Depletion Logistic Function*
Yo
Xo
A
B
R2
RMSE
p
0.5 m irrigated
0.55
26.06
0.40
3.44
0.51
0.12
<0.0001
1 m irrigated
0.02
81.10
0.93
0.92
0.25
0.15
<0.0001
1 m total
0.34
34.70
0.69
1.09
0.27
0.15
<0.0001
z Y = Y0 +aX+bX2 +cX3
where X =
- ETc/ETo, and Y0, a, b, and c are regression
coefficients
y Y = Y0 + aexp bX where X = ET x Y=Y0 +
a
1 +
where X = ETc/ETo, and Y0 a, and b are regression coefficients


19
pollution of groundwater by leaching, and 3) production costs associated with excessive
irrigation, and nutrient and pesticides losses due to leaching.
Mills et al. (1999) reported a significant decrease in citrus stomatal conductance
after midday. This decrease was most pronounced for south-facing exterior leaves and
increased with increasing evapotranspirational demand (ET0). Soil water use from 2 year-
old Hamlin orange trees measured at 0.5-hour intervals using weighing lysimeters
indicated that water continued to be removed several hours after the midday decrease in
stomatal conductance. Two seemingly opposing theories place control of soil water
uptake at the leaf level via leaf water potential (Slatery, 1967) or root via root water
potential (Tinker and Nye, 2000). The former assumes that leaf water potential exerts
control on stomatal conductance regulating transpiration and thus water uptake. The latter
speculates that dehydrating roots, due to low soil water content, indirectly control
stomatal conductance through the production of chemical compounds that after
translocation to the leaves reduce stomatal aperture. Lafolie et al. (1991) measured
decreasing leaf water potential with decreased root water potential until midday. After
reduced stomatal conductance at midday, leaf water potential increased without a
corresponding decrease in root water potential. This result was given as evidence that
stomatal conductance was not controlled by leaf water potential alone.
Factors Effecting ETC
Crop species
Citrus are evergreens and therefore require water for transpiration throughout the
year. Citrus leaves are thick and waxy, resulting in high cuticular resistance to
transpiration (Mills et al. 1999). Koo (1963) and Koo and Sites (1955) stated that water


32
had been N-depleted to explore the influence of prior fertilization practices on subsequent
N uptake. The highest percentage of labeled N occurred in fruit, new leaves and twigs.
Only about 20% of the leaf and fruit N originated from the labeled source, suggesting
considerable redistribution from stored reserves. Less than 14% of the labeled N was
found in roots or large limbs. Dasberg (1987) found that 80% of the N in new growth
came from stored rather than applied N, suggesting previous nutrition has significant
influence on current season growth and fruit yield. Legaz and Primo-Millo (1988)
reported increased N uptake from the beginning of spring flush to bloom. Uptake
increased through the spring, reaching a maximum at the summer flush after which
uptake declined gradually through winter.
Mooney and Richardson (1992) observed an N concentration gradient between
the roots, trunk and branches of citrus trees in New Zealand. High concentrations were
found in the branches, with lower concentrations in the roots. Nitrogen concentrations in
the trunk were highest at bud break and declined steadily through fruit set and
development to a minimum at fruit harvest. Nitrogen concentration for all categories
peaked at flowering and then decreased steadily until harvest. Nitrogen uptake
efficiencies of 82.0 and 74.1% for ammonium nitrate and urea, respectively, were
reported by Mattos (2000). Legaz et al. (1982) reported 50 to 60% of total tree 15N
recovery in above-ground tree parts. Absorption rates increased only slightly from the
beginning of growth until flowering, and increased sharply reaching a maximum value at
the second growth flush (July) before declining during the fall and winter months.
Dasberg (1987) demonstrated that the highest rate of ,5N uptake by citrus trees occurred
during fruit set and the lowest occurred during winter.


10
from early June to early July, and the third in late summer (August or September). The
principal blooming period for all commercial species is early spring and usually lasts
approximately 6 weeks (Mid February to late March). The normal period of ripening of
most citrus fruits is late fall and winter, preceding the spring bloom. However, late
maturing varieties such as Valencia require 12 to 15 months for maturity, which occurs
after bloom for the next crop. The period of time when citrus fruit can be harvested is
about 9 months.
Citrus production areas in Florida range from upland positions with very sandy
Ridge soils, which are deep and excessively well drained, to relatively low flatwood
sites that are often flooded in their native state due to the presence of a spodic layer.
Flatwoods soils must be drained and bedded before planting to lower the fluctuating
water table. Each of these general areas in Florida presents somewhat different challenges
for growing citrus trees and fruit production. Although crop growth and nitrogen uptake
dynamics are readily available for many agronomic crops, this information is in short
supply for citrus. A comprehensive literature review relevant to citrus growth
characteristics, root distribution, water and N uptake dynamics, and crop growth models
will be presented in this chapter.
Citrus Growth Characteristics
Turrell et al. (1969) proposed citrus tree growth equations based on growing
conditions and cultural practices in California. These equations assumed citrus tree
growth to bejogistic in nature depending on cultural characteristics such as spacing,
pruning, and irrigation and fertilizer scheduling, soil characteristics, and climatic
conditions. Several studies measuring citrus biomass in relation to nutrient concentration


63
Table 3-7. Mature citrus tree tissue N concentration as a function of year sampled and
rootstock.
Plant Tissue
Year Sampled
Rootstock
2001
2002
Carrizo
Swingle
(%)
Leaves
2.47
2.32
2.47
2.46
Twigs
1.28
1.00
1.09
1.14
Small Branches Bark
1.16
1.24
1.22
1.11
Small Branches Wood
0.30
0.31
0.33
0.28
Med. Branches Bark
1.11
1.28
1.13
1.08
Med. Branches Wood
0.29
0.32
0.30
0.28
Large Branches Bark
1.16
1.33
1.06
1.30
Large Branches Wood
0.38
0.34
0.42
0.33
Trunk Bark
1.29
1.36
1.33
1.14
Trunk Wood
0.42
0.44
0.43
0.42
Fibrous Roots
1.55
1.38
1.61
1.51
Medium Roots
1.04
0.85
1.10
0.97
Large Roots
0.48
0.47
0.50
0.47
Tap Root 0.45
Statistical Significance2
Year NS
Rootstock NS
Year X Rootstock NS
0.45
0.47
0.39
z NA = No significant difference at the P=0.1 level.
Nitrogen Balance
Leaf and twig dry weight accumulation during the previous 12 months was significantly
greater (P=0.05) for trees grown on Carrizo (3963 and 5717 g tree1, respectively)
compared with trees grown on Swingle (3312 and 4525 g tree*1, respectively). Total N
content for these tissues were 145 and 127 g tree'1 for Carrizo and Swingle, respectively.
Mean fruit N accumulations for the two rootstocks were 302 and 258 g tree1 for trees on
Carrizo and Swingle, respectively. Assuming a conservative 5% increase in N
accumulation in all tissues other than leaves and twigs due to increase in biomass in
2001, the resulting increase in N content was 24 and 18 g tree1 for Carrizo and Swingle


167
Kato, T.S., S. Kubota, and S. Bambang. 1982. Uptake of 15N-itrogen by citrus trees in
winter and repartitioning in spring. J. Japan Soc. Hort. Sci. 50:421-426.
Kato, T., Y. Makoto, and S. Tsukahara. 1984. Storage forms and reservoirs of nitrogen
used for new shoot development in Satsuma mandarin trees. J. Japan Soc. Hort.
Sci. 52:393-398.
Keeney, D. R and D. W. Nelson. 1987. Nitrogen-Inorganic Forms, sec. 33-3, extraction
of exchangeable ammonium, nitrate, and nitrite, pp.648-9. In A. L. Page et al.,
eds., Methods of soil analysis: Part 2, Chemical and microbiological properties.
Agronomy, a series of, monographs, no.9 pt.2, Soil Science Society of America,
Madison, Wisconsin USA.
Khalaf, H.A., and RC.J. Koo. 1983. The use of controlled-release nitrogen on container
grown citrus seedlings. Citrus Vegt. Mag. 46:10-32.
Kimball, M.H., A. Wallace, and R.T. Mueller. 1950. Changes in soil and citrus root
characteristics with non-tillage. California Citrograph 35:432-433.
Kiniry, J.R., J.R. Williams, RL. Vanderlip, J.D. Atwood, D C. Reicosky, J. Mulliken,
W.J. Cox, H.J. Mascagni, S.E. Hoolinger, and W.J. Wiebold. 1997. Evaluation of
two maize models for nine U.S. Locations. Agron. J. 89:421-426.
Kiniry, J.R, and AJ. Bockholt. 1998. Maize and sorghum simulation in diverse Texas
environments. Agron. J. 90:682-687.
Koo, RC.J. 1963. Effects of frequency of irrigation on yield of orange and grapefruit.
Proc. Fla. State Hort. Soc. 74:1-5.
Koo, RC.J. 1978. Response of densely planted Hamlin orange on two rootstocks to
low volume irrigation. Proc. Fla. State Hort. Soc. 91:8-10.
Koo, RC.J. 1979. The influence ofN, K, and irrigation on tree size and fruit production
ofValencia orange. Proc. Fla. State Hort. Soc. 95:10-13.
Koo, RC.J. 1980. Results of citrus fertigation studies. Proc. Fla. State Hort. Soc. 93:33
36.
Koo, RC.J. 1986. Controlled-release sources of nitrogen for bearing citrus. Proc. Fla.
State Hort. Soc. 99:46-48.
Koo, RC.J., and J.W. Sites. 1955. Results of research and response of citrus to
supplemental irrigation. Proc. Soil Sci. Soc. Fla. 15:180-190.


128
Fig. 6-1. Relationship of estimated total N uptake as a functrion of soil solution
concentrations. Total N uptake is the sum of passive and active N uptake components.
Dashed lines denote Michaelis-Menten equation constants of Imax and Km.
Soil Solution N Concentration (mg L'1)
Fig. 6-2. Relationship of estimated active N uptake to soil solution concentration.


100
Seasonal Kc*Kg
The ratio of estimated daily ETC to calculated daily ET0 is an approximation of the
quantity KC*KS in Equation 5-1. The ETC to ET0 ratios (= KC*KS) were plotted against the
weighted 0, ASWD, and for the irrigated area to depths of 0.5 and 1 m, or for the total
land area allocated to the tree area to a 1 m depth (Figs. 5-3 to 5-5). Regression equations,
R2, RMSE, and P values are provided in Table 5-2. The R2 values for the equations are
generally smaller than 0.5 due to the wide scatter of data points as indicated by the
relatively large RMSE values. However, all relationships were significant at the P=0.01
level. The trends were particularly strong for regressions against soil water potential.
Theoretically, the Ks value should be approximately 1 when 0 is near 0fc (0.075
to 0.08 cm3 cm"3). Therefore, the ETC to ET0 ratios should approximate Kc at 0fc. Fig. 5-
6 illustrates the ETC to ET0 ratios by day of year (DOY) when mean 0 in the irrigated
zone was between 0.07 and 0.085 cm3 cm"3 to a depth of 1 m. These ratios ranged from
0.81 on DOY 24 (January) to 1.12 on DOY 179 (June). The regression equation for this
relationship is given in Table 5-3. With an R2 of 0.76, DOY explaines more than 76% of
the variation in the ETC to ET0 ratios when 0 was between 0.07 and 0.085 cm3 cm"3.
Therefore, the equation provides a good approximation of the value of Kc for a given
DOY.
K* Estimation
The ETC to ET0 ratios for 0 values less than 0fc would approximate the Kg value
assuming the Kc is 1. Since we have demonstrated that Kc values do not equal 1 during


145
N senescence patterns were relatively consistent across seasons providing a relationship
for seasonal tree N balance for citrus. Understanding seasonal N uptake rates and cycling
is necessary to develop of crop models. Such models provide a scientific basis for
improved water and N management in agricultural production. Providing citrus growers
with tools like an expert system to increase NUE and avoid nitrate leaching is essential,
especially in Florida due to the low water and nutrient holding capacity of the sandy soils
there.


73
Conclusions
Leaf areas of both young and mature citrus trees were correlated with tree size as
measured by TCV and TCSA. Leaf area index increased rapidly for young citrus trees
and then equilibrated at approximately 10 by age 3 to 4 years. This information is
valuable for the estimation of citrus light interception and total photosynthesis. Change in
citrus tree dry weight and N content of the two citrus scions in these studies was shown to
be a linear function of TCV and TCSA. Partitioning of biomass and N decreased for
leaves and twigs, increased for branches, and remained constant for trunk and taproot
tissues with increase in tree size. While mature citrus trees grown on Swingle citrumelo
rootstock were consistently smaller than trees of similar age grown on Carrizo citrange,
mass partitioning of tree parts were similar for both rootstocks. Thus, with the exception
of spatial root length density distribution described in chapter 4, the only effect of the two
rootstocks and two scions used in these studies was on tree size relative to TCV and
TCSA Therefore, biomass and N partitioning for specific tissues with tree size can be
captured in generic linear relationships. The N balance estimated for mature citrus trees
in this study indicated an apparent fertilizer N use efficiency of 60 to 70%.


CHAPTER 4
CITRUS ROOT GROWTH DYNAMICS
Introduction
While the role of roots in anchoring crop plants, particularly tree crops, to the soil
should not be taken for granted, the function of roots as absorbing organs for both water
and nutrients can not be overemphasized. The structure of a root system is important in
determining the pathway and resistance to water and solute uptake, and the volume of
soil accessible to crop plants (Kramer and Boyer, 1995). The entrance of water and
nutrients into young roots occurs a few cm behind the root tips because of the lack of a
functional xylem at the tip and the suberization of root hypodermis and endodermis
tissues with age (Tinker and Nye, 2000). Thus, the larger the length of relatively small
diameter fibrous roots a crop root system has, the greater the amount of water and
nutrients available to it. Likewise, the larger the soil volume a crop root system occupies,
the greater the pool of water and nutrients available for uptake.
The goal of fertilizer application should be the placement of nutrients within the
crop root zone to insure the most efficient uptake. Maintenance of adequate water and
labile nutrient concentrations within soil zones occupied by the crop root system is
essential for optimal nutrient uptake. Understanding the spatial distribution of fibrous
roots is essential to ensure proper fertilizer placement to improve nutrient uptake and
potentially reduce leaching below the root zone.
74


172
Sumner, D M. 1996. Evapotranspiration from successional vegetation in a deforested
area of the Lake Wales Ridge, Florida. Water-Resources Investigation. 96-4244.
Syvertsen, J.P., and J.J. Lloyd. 1994. Citrus. In. Handbook of Environmental Physiology
of Fruit Crops, Volume H: Sub-tropical and Tropical Crop. Schaffer, B. and P C.
Andersen (eds.) Boca Raton, FI, CRC Press Inc., USA.
Syvertsen, J.P., M L. Smith, and B.J. Boman. 1993. Tree growth, mineral nutrition and
nutrient leaching losses from soil of salinized citrus. Agr. Ecosyst. Environ.
45:319-334.
Syvertsen, J.P., and M.L. Smith. 1996. Nitrogen uptake efficiency and leaching losses
from lysimeter-grown citrus trees fertilized at three nitrogen rates. J. Amer. Soc.
Hort. Sci. 121:57-62.
Timmer, L.W., and S.E. Zitko. 1996. Evaluation of a model for prediction of post bloom
fruit drop of citrus. Plant and Disease 80:380-383.
Tindall, J. A., and J. R. Kunkel. 1999. Unsaturated Zone Hydrology for Scientists and
Engineers. Prentice-Hall, Upper Saddle River, New Jersey.
Tinker, P. B., and P. H. Nye. 2000. Solution Movement in the Rhizosphere. Oxford
University Press, New York, New York.
Tucker, D. P., C. G. Erickson, and K. T. Morgan. 1997. Middles Management Methods
in Citrus Affect Soil Moisture Retention and Vegetation Species. Proc. Fla. State
Hort. Soc. 110: 39-43.
Tucker, D.P.H., AK. Alva, L.K. Jackson, and T.A. Wheaton. 1995. Nutrition of Florida
citrus trees. Univ. of Fla. Coop. Ext. Serv. SP169.
Turrell, F.M., M.J. Garber, W.W. Jones, W.C. Cooper, and RH. Young. 1969. Growth
equations and curves for citrus trees. Hilardia 39(16):429-445.
Vaile, R.S. 1924. A survey of orchard practices in the citrus industry of southern
California. Cal. Agrie. Exper. Sta. Bul.374. 40 pp.
Valiente, J. L. and Albrigo, L. G. 2000. Modeling flowering date of sweet orange
(Citrus sinensis (L.) Osbeck) trees in central Florida based on historical weather
records Proceedings of the International Society of Citriculture.
Verma, A., R.S. Kanwar, and U.S. Tim. 1995. Modification of DRAINAGE model by
using the nitrogen component form the GLEAMS model. Trans, of the ASAE
38(3):717-724.


153
Branch bark N concentrations were consistently in a narrow range of 1.0 to 1.3%
during the 2-year period, with means of 1.04 and 1.17% for low and high N application
rates, respectively. Minimum bark N concentrations of 1.00 and 1.11% occurred in May
or June of each year, while maximum concentrations of 1.09 and 1.22% occurred in
October and January for low and high annual N application rates, respectively. Branch
wood N concentrations were lower, and followed trends similar to those of branch baric
tissue. Mean wood N concentrations were 0.25 and 0.31% for the low and high N
application rates, respectively. Maximum wood N concentrations were 0.37 and 0.38%
for the same rates, respectively, and occurred in January and March. Minimum wood N
concentrations were 0.23 and 0.29% for low and high N application rates, respectively,
and occurred in October.
Root N concentrations were greater for roots <4 mm in diameter compared with
larger roots. Mean N concentrations were 1.34 and 1.35% for low and high N application
rates, respectively, for roots <4 mm in diameter. Mean N concentrations for roots >4 mm
in diameter were 0.85 and 0.89% for the same rates. Fruit N concentration was highest in
May of each year (1.25 and 1.30% for low and high N application rates, respectively).
Nitrogen concentrations decreased through the season to means of 1.03 and 1.06% for
low and high N application rates, respectively, just prior to harvest. Mean fruit diameter
was not constantly different by N application rate. However, fruit dry weights were
greater for the lower application rate.
Seasonal Nitrogen Loss
Flower dry weight accumulated in March and April of each year with cumulative
weights of 529 and 764 g tree-1 for 2001 and 2002, respectively. Cumulative flower N


14
mm in diameter. Their dry mass is a relatively small part of the total root system, but
their composite length far exceeds that of the woody roots (>4 mm in diameter). These
fine roots are considered to be the functional part of the root system because of their
critical role in water and nutrient uptake. There is some variation among rootstocks in the
morphology of fibrous roots (Castle and Youtsey, 1977). Some rootstocks like trifoliate
orange [Poncirus trifoliata (L.) Raf ] produce higher specific root length or length/unit
mass (Eissenstat, 1991). Fibrous roots are also the most vulnerable part of the root
system. Their development, function, and longevity are strongly influenced by soil
characteristics, environmental changes, crop species, crop growth stage, and cultural
practices.
Factors Affecting Root Distribution and Root Density
Soil Characteristics
The distribution of roots is modified by the physical and chemical properties of
the soil profile (Hillel, 1971). Widespread root development and high fibrous root
concentrations were observed in deep soils of sand texture where there were virtually no
impediments to root growth provided that water and nutrients were non-limiting to
growth (Ford 1952; 1953a&b; 1954a&b; 1959; 1964; 1972); Ford et al., (1954; 1957).
Increased tree size and yield have been related to root system depth and fibrous root
mass. The depth of rooting of Orlando tngelo trees on 10 rootstocks growing in deep,
sandy soil was correlated with tree height (Castle and Krezdom, 1975). Although fibrous
root distribution was affected by tree height, total fibrous root dry mass measured at the
canopy dripline was not correlated with tree height. Ford (1954a; 1964; 1968; 1969;
1972) conducted many studies of citrus trees in poorly drained Spodosols of Florida and


23
gravitational field. These same forms of energy are commonly referred to as osmotic,
matrix, gravitational and pressure potentials, the sum of which is referred to as total water
potential (4>). Thus, soil water moves in response to the difference in water potential over
a distance. The first published relationship between water flux and energy gradient was
obtained empirically in 1856 by Henry Darcy after a study of saturated sand filters
(Hagan et al., 1967).
u =-K d<|> /dx Equation 2-1
Where:
u = water flux (cm3 cm'2 s'1),
K = hydraulic conductivity constant (cm s'1),
<|> = soil water potential (kPa), and
x = the distance over which the flux is maintained (cm).
The constant of proportionality of Darcys Law (K) is known as the hydraulic
conductivity, and is a function of both the properties of the medium and the fluid (Tindall
and Kunkel, 1999). In saturated soils, K will be constant as long as the structure of the
soil remains stable because the water flow pathways will be unchanged. In unsaturated
soil, K varies with the water content (0), because the latter defines the total cross-section
area for water flow, the effective water-filled pore radius, and the effective pathlength
(Tinker and Nye, 2000). A soil with a wide range of pore sizes conducts fluid more
rapidly than a soil with small pore sizes (Tindall and Kunkel, 1999). The saturated
hydraulic conductivity of soils has a wide range from 10'9 cm s'1 for clay to 1.0 cm s'1 or
more for sand. Lower values of K for a clay medium (with smaller pore sizes) are likely
due to the drag exerted on the viscous fluid by the walls of the pores. Particles of smaller-


125
Table 6-1. Estimated cumulative N losses from control pipes and bulk soil, estimated
cumulative maximum N uptake, and estimates of passive and active N uptake for samples
collected on five consecutive days in March, 2002. Rootstocks are Carrizo citrange, and
Swingle citrumelo; high N application rate was 269 kg ha'1 yr'1, low rate was 134 kg ha'1
yr1.
Days After
Weighted
N Uptake
Cumulative
Total
Cumulative N
Loss
Application
Solution
N
Passive Active Total
N Uptake
Control Soil
(mg L'1)
(g tree'1 d'1)
(gtree1) (%)
(% applied)
Carrizo-
High rate
1
165.7
5.1
5.1
10.2
10.2
13.9
12.4
26.3
2
152.7
4.5
5.2
9.8
20.0
27.3
27.4
40.8
3
115.8
2.2
7.8
10.0
30.0
41.1
37.3
59.4
Carrizo-
Low rate
1
80.0
2.9
12.0
14.9
14.9
41.5
0.9
42.4
2
50.6
1.8
5.6
7.4
22.3
62.2
19.6
73.9
3
31.0
0.6
1.8
2.4
24.7
68.8
24.4
81.7
Swingle -
High rate
1
183.8
5.7
9.0
14.7
14.7
17.9
7.1
25.0
2
181.3
5.4
4.2
9.6
24.3
29.7
7.8
47.3
3
167.1
3.0
10.4
13.4
37.7
46.0
5.4
62.1
Swingle -
-Low rate
1
163.5
5.3
13.8
19.1
19.1
27.2
0.7
28.1
2
124.1
4.5
11.6
16.1
35.2
50.5
16.6
62.6
3
65.8
1.6
10.2
11.8
47.0
67.4
32.7
85.6


135
1.4 -
12 -
1.0 -
0.8 -
0.6
0.4 -
0.2-
X o.o
2-1.4
1.2 -I
1.0
0.8-
0.6 -
0.4 -
0.2-
0.0
Small limbs Bark
Small limbs Wbod
Medium limbs Bark
Medium limbs Wbod
large limbs Bark
Large limbs Wax!
Small Limbs Bark
Small limbs Wbod
Med urn limbs Bark
Medium limbs Wbod
Large Limbs Bark
large limbs Wood
ri 1
2/1/01 5/1/01
-1i
8/1/01
-ii
11/1/01
-iii*f
2/1/02 5/1/02
B
8/1/02 11/1/02
Date
Fig. 6-5. Seasonal changes in N concentrations for bark and wood tissue of small,
medium, and large limbs during 2001 and 2002. High N rate (A) and Low N rate
(B) equal to 269 and 179 kg ha'1 yr'1, respective!


27
occurred at the highest depletion value due to the reduced ability of the soil to transport
water to the roots because of reduction in hydraulic conductivity. Fares and Alva (1999)
calculated daily ETC for 3-year-old Hamlin orange trees on deep sandy soil in central
Florida. Estimated daily ETC values decreased with time after each rainfall or irrigation.
Water table depth
Obreza and Admire (1985) concluded that shallow water tables in flatwoods soils
could significantly augment water available for root uptake. Graser and Allen (1987)
suggested that water-table management by controlling water table depth in the winter and
spring could help decrease the need for supplemental irrigation during the dry season.
Boman (1994) used drainage lysimeters in which he maintained a water table at 0.61,
0.76, and 0.91 m to measure the effects of water table on ETC, growth, yield and fruit
quality of 5-year-old Valencia trees. However, treatment effects were not significant.
Soil shading
Castel et al. (1987) estimated soil surface evaporation by comparing water loss
from weighing lysimeters in which the soil was covered by plastic with lysimeters that
remained uncovered. Mean estimated evaporation was reported as 0.78 mm, greater than
18% of the estimated potential ET of 4.25 mm. Castel and Buj (1992) reported that the
percentage of ground shaded by young Clementine trees increased from 10 to 25% during
a 4-year period. Evapotranspiration increased by 33% during the same time period. This
increase was attributed to the increasing water use by the trees and reduced soil surface
evaporation.


CHAPTER 7
SUMMARY AND CONCLUSIONS
Simulating plant growth is complex due to the interactions of biological
processes, soil physical characteristics, and environmental factors (Jones and Luyten,
1998). Biological processes and soil physical characteristics define the crop system and
include photosynthesis, respiration, transpiration, biomass accumulation, soil water and
nutrient uptake, and nitrogen leaching. Environmental factors such as temperature,
radiation, wind, humidity, rainfall, and irrigation influence these processes and are
typically required as model inputs. Modeling these processes requires the estimation of
many state variables with time through the use of linear or non-linear functions and
associated parameters and constants. The goals of this study were to 1) determine
changes in above ground citrus biomass and N distribution for trees under recommended
N fertility practices over a range of tree sizes, 2) collect information on spatial
distribution patterns of citrus root length density for different tree sizes, 3) estimate
evapotranspiration crop and soil moisture coefficients, and soil water use per unit root
length density, 4) explore seasonal N uptake rates for citrus, and 5) compare seasonal
biomass and N concentration changes for citrus fertigated at two annual N fertilizer rates.
Mature Tree Biomass Distribution
Leaf biomass represented 12 to 15% of total tree biomass, while total branch
weights were 50 to 65% of total tree weight. Total root biomass was highly variable, but
146


119
changes in plant tissue N concentration, and 4) determine cumulative tree biomass and N
losses for citrus during a 2-year period.
Materials and Methods
Site Characteristics
Fourteen-year-old Hamlin orange on Carrizo citrange and Swingle citrumelo
rootstocks at the same location as Experiment 1 in Chapter 3 were used for the three
experiments presented in this chapter. The trees had been fertigated at an annual rate of
179 or 269 kg N ha'1 at approximately monthly intervals from February to October using
equal split applications for the 3 years prior to the start of this experiment. Irrigation was
applied by an automated irrigation system using switching tensiometers to trigger
irrigations. Irrigation was applied when soil water potential in the upper 30 cm dropped
below -10 kPa during the bloom and fruit set period of February to May and -15 kPa for
the remainder of the year. Reclaimed water containing 7 mg L'1 or less of NO3-N was
used for all irrigations. The soil type at the site was Candler fine sand (hyperthermic,
uncoated Quartzipsamments) with water holding capacity of 0.05 to 0.08 cm3 cm'3 and
cation exchange capacity of less than 5 cmol kg'1.
Experiment 1 Nitrogen Uptake Flux
Fertilizer rates and applications
The N fertilizer rates used in this study were approximately 50 and 100% of the
monthly rate based on 269 kg N ha'1 yr'1 in 6 monthly applications, or 45 kg N ha'1 per
application. The 100% rate is equivalent to 500 g N tree'1 yr'1 or 83 g N tree'1 per
application. The reduction in N application rate was accomplished by reducing the
application time to four representative trees of each rootstock using valves in the


155
effects of crop sequencing under given management inputs and weather conditions. Crop
and soil water status (Hoogenboom et al., 1994; Gabrielle et al., 1995; and Xie et al.,
2001) and N and C balances (Gabrielle and Kengni, 1996; Quemada and Cabrera, 1995;
and Sexton et al., 1998) have been modeled. These models can combine with crop
developmental models of irrigation scheduling, fertilizer N fate determination, and nitrate
leaching estimates. A DSS can provide the framework for storing of information needed
by the model and can include a user-friendly output to assess production options.
Citrus production would benefit greatly from such a DSS. Data collected and
relationships determined in this dissertation as well as literature sources can provide
information required to model seasonal and temporal citrus N demand, root development
processes, seasonal water and N uptake, and seasonal N distribution. These models can
lead to a DSS capable of providing irrigation scheduling, fertilizer requirements and
rates, and environmental impacts of nitrate leaching. Verification of best management
practices and grower compliance with such practices can be determined using a DSS. The
ultimate goal of such a DSS would be the improved allocation of water and fertilizer
resources for optimal yield with minimal environmental impact.
Citrus N Practices and BNPs
Floridas population has increased from 3 million in 1950 to more than 16 million
at present, creating severe competition between agricultural, commercial, residential, and
environmental users for Floridas limited water resources. As a result of increased
demand, consumption and quality of water has become highly scrutinized by regulators.
Improvements in both N uptake efficiency and timing of N applications to coincide with
N demand will balance the needs of the citrus tree while minimizing water quality


Xll
3-10. N weight allocation to total, above ground, and below ground, leaf
and twig, total branch, and total root and tap root accumulation
as a function of canopy volume 67
3-11. N weight allocation to total, above ground, and below ground dry
weight, leaf and twig biomass, total branch biomass and root and
tap root accumulation as a function of trunk cross sectional area 68
4-1. Root length density distribution by depth at 50, 100, 150, and 200 cm
distance form tree trunk between rows of Hamlin orange trees on
Carrizo citrange or Swingle citrumelo rootstocks 80
4-2. Root length density distribution at 0-15,15-30, 30-45, 45-60, 60-75,
and 75-90 cm depth increments by distance from the tree trunk
as affected by distance form the tree trunk for Hamlin orange
trees on Carrizo citrange and Swingle citrumelo rootstocks 81
4-3. Citrus root distributions by depth below the soil surface and distance
from the tree trunk for trees 2-5 years old, 5-10 years old, 10-15
years old, and > 15 years old 84
5.1. Illustration of EnviroSCAN probe 95
5-2. Illustration of EnviroSC AN probe layout, and soil surface area used
for determining soil water content for each probe 95
5-3. Estimated ETC to calculated ET0 ratio as a function of soil water
content in the irrigated zone to a 0.5 m depth, 1 m depth, and
the total tree area to a 1 m depth 101
5-4. Estimated ETC to calculated ET0 ratio as a function of soil water
potential in the irrigated zone to a 0.5 m depth, 1 m depth, and
the total tree area to a 1 m depth 102
5-5. Estimated ETC to calculated ET0 ratio as a function of available
soil water depletion in the irrigated zone to a 0.5 m depth,
1 m depth, and the total tree area to a 1 m depth 103
5-6. Comparison of estimated crop evapotranspiration (ETC) with calculated
reference evapotranspiration (ET0) ratio which are an approximation
of Kc for observations when soil water content values were near
field capacity as a function of day of year (DOY) 105
5-7. Estimated soil water coefficient Ks as a function of soil water
potential in the irrigated zone to a 0.5 m depth, 1 m depth, and


151
Nitrogen Uptake
Differences in mean cumulative N loss from the soil 3 days after fertilizer
application were significantly influenced by month of year, with the greatest loss in May.
Mean percentage of total soil N loss during the 3 day period for the high (83 g N tree'1)
application rates were 60.7, 68.6, and 63.4% for March, May, and September;
respectively. These means were significantly different from mean percentage losses for
the lower N rate, which were 83.6, 82.6 and 73.1% for the same three months. Mean
cumulative N soil losses for Hamlin trees on Carrizo citrange were 70.5, 78.8, and
70.1% of N applied for the months of March, May, and September, respectively, and
were not significantly different from mean cumulative losses for trees on Swingle
citrumelo, which were 73.8, 72.5, and 66.4% of total N applied for the same months.
Nitrogen losses from the control treatment varied greatly in the three studies and
were not significantly different by application rate, month, or rootstock. Mean cumulative
N loss was 22.6%, and ranged from a net gain of 7.0% to a loss of 41.0% during the 3
days after fertilizer application. Cumulative daily maximum N uptake was significantly
different by application amount, but not significantly different by month and rootstock.
Maximum uptake as a percentage of amount applied averaged 46.7 and 61.7 for the high
and low application rates, respectively. Mean cumulative N uptakes for Carrizo and
Swingle rootstocks were 53.9 and 54.4%, respectively. Estimates of the Michaelis-
Menten equation constants I max and Km for active N uptake were 8.5 g tree'1 d1 and 45
mg L1, respectively.


VI
Results 78
Mature Hamlin Orange Root Distribution 78
Root length Density Distribution Changes with Tree Size 83
Discussion 87
Conclusions 88
5 CITRUS WATER UPTAKE DYNAMICS 90
Introduction 90
Methods and Materials 94
Site Characteristics 94
Soil Capacitance Sensor Data Collection 94
Estimation of Daily ETC 96
Estimation of Monthly Crop Coefficient (IQ 97
Estimation of Water Stress Coefficient (IQ 97
Estimation of Soil Water Uptake per Unit Root Length 97
Results 98
Seasonal ET0 and ETC Trends 98
Seasonal K; K* 100
Kg Estimation 100
Soil Water Uptake per unit Root length Density 109
Discussion 110
Conclusions 113
6 CITRUS NITROGEN UPTAKE AND CYCLING 115
Introduction 115
Methods and Materials 119
Site Characteristics 119
Experiment 1 Nitrogen Uptake Flux 119
Fertilizer rate and application 119
Soil sampling procedures 120
Analytical methods 120
Experiment 2 Seasonal Tissue N Concentration 121
Tissue samples collected 122
Tissue analysis 122
Experiment 3 Seasonal N Loss 122
Results 123
Nitrogen Uptake 123
Nitrification Estimation 130
Seasonal Tissue N Concentration 130
Seasonal Fertilizer Use Efficiency 136
Seasonal N Losses 137
Discussion.. 138
Conclusions 144


X
6-5. Seasonal changes in N concentration, size and dry wt. of fruit, flush
leaves, and expanded leaves for 2001 and 2002 seasons
133


123
falling onto the frames was collected at approximately 2-week intervals for two seasons
(2001 and 2002). The plant material was separated into 1) flowers, 2) fruit, 3) twigs, and
4) leaves. Material in each of the four categories was counted, dried, weighted, and
analyzed for total Kjeldahl N using the same grinding and analytical procedures
described in chapter 3.
Assuming the material collected in the catch frames was proportional to the
amount of material under the entire canopy, the biomass and N concentration of each
tissue was multiplied by the ratio of the area under the canopy to the area of the catch
frame. Cumulative seasonal N loss was determined for each tissue.
Results
Nitrogen Uptake
Irrigation after sampling on the third day after fertilizer application likely leached
a portion of the N below the 45 cm sampling depth. Therefore, soil N losses were
calculated for the control pipes and bulk soil for 1,2, and 3 days after application.
Differences in mean cumulative N loss from the soil during the 3 days after application
were significant at the P=0.05 level for month of year, with greatest loss occurring in
May. Soil N losses by rootstock and application rate are presented in Tables 6-1 to 6-3
for the months of March, May, and September, respectively. Mean percentage of total
soil N loss during the 3-day period for the high application rate of approximately 83 g
tree1 were 60.7, 68.6, and 63.4% for March, May, and September, respectively. These
means were significantly different at the P=0.05 level from mean percentage N losses for
the lower rate which had respective values of 83.6, 82.6, and 73.1%. Mean cumulative N
soil losses for Hamlin trees on Carrizo citrange were 70.5, 78.8, and 70.1% for the


48
Table 3-2. Mature citrus tree leaf area and leaf area index as a function of TCV and
TCSA
Leaf Area
Leaf Area Index
(m2 tree"1)
(m2 leaf m"2
canopy)
Carrizo
Swingle
Carrizo
Swingle
Mean
110.0
103.4
10.2
9.9
Standard Deviation
10.3
24.8
1.2
1.8
Statistical Significance2
TCVy
*
*
NS
NS
TCSA
*
*
NS
NS
Rootstock
NS
NS
NS
NS
*NS = not significant, *= significant P<0.05, and **= significant P<0.01
yTCV = tree canopy volume, TCSA = tree cross-sectional area.
Mature trees on Carrizo rootstocks had significantly greater (P=0.05) mean dry
weight (100 kg tree'1) than those on Swingle (83 kg tree1), whereas, below-ground
biomass was not significantly different (Table 3-3). Significantly higher (P=0.01)
percentage of large branch biomass was found for trees grown on Carrizo citrange (23.8
kg tree"1) compared with trees grown on Swingle citrange (15.8 kg tree"1). Thus,
percentage total branch biomass was significantly greater (P=0.05) for the Carrizo


3
hyperthermic Entisols composed of uncoated sands with water holding capacities of 0.04
to 0.09 cm3 cm'3, hydraulic conductivities >50 cm h*1, cation exchange capacities of 1 to
5 cmol (+) kg'1, and depths of more that 10 m.
Long-term monitoring studies and research projects were initiated in 1992 to
evaluate the impacts of nutrient and water management practices in citrus on ground
water quality. The goals of the projects established by FDACS (Graham and Alva, 1995)
were to 1) generate baseline groundwater quality data from several commercial citrus
groves in the ridge area; 2) develop recommendations for alternative nutrient and water
management practices; and 3) assess the impacts of these alternative management
practices on groundwater quality.
Fig. 1-1. Map of Florida with Lake, Polk, and Highlands counties highlighted.


IX
5-1. Monthly maximum, minimum, and mean reference evapotranspiration
reported by Florida Automated Weather Network for the Avalon
Station and maximum, minimum and mean estimated citrus crop
evapotranspiration betweeen April 2000 and March 2002 99
5-2. Regression analysis of estimated ETC to calculated ET0 ratio by
mean soil water content, soil water potential, and available
soil water depletion in soil 1.0 or 0.5 m deep and in either the
irrigated zone or total tree area 104
5-3. Regression analysis of estimated ETC to calculated ET0 ratio by day
of year for soil water content values greater than 0.070 cm3 cm'3
(field capacity) using a quadratic function 105
5-4. Estimated soil water coefficient (Ks) values for a range of percentage
available soil water depletion (ASWD) using equation 2 found in
Allen et al. (1997) 108
5-5. Regression analysis of estimated soil depletion factor (K*) by mean soil
water potential, and available soil water depletion in soil 1.0 or 0.5 m
deep and in either the irrigated zone or total tree area 108
5-6. Regression analysis of estimated soil water uptake per unit root length
density on soil water potential in soil at three locations surrounding the
tree and 10, 20,40 or 80 cm depths using an exponential decay model 110
6-1. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in March, 2002 125
6-2. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive
days in May, 2002 126
6-3. Estimated cumulative N losses from control pipes and bulk soil,
estimated cumulative maximum N uptake, and estimates of passive
and active N uptake for samples collected on five consecutive days
in September, 2002 127
6-4. Regression equations for estimated maximum N uptake and estimated
active N uptake rates by soil N concentration (mg l'1) using an
exponential rise to a maximum model 129


Xlll
the total tree area to a 1 m depth 106
5-8. Estimated soil water coefficient Ks as a function of available soil
water depletion in the irrigated zone to a 0.5 m depth, 1 m depth,
and the total tree area to a 1 m depth 107
6-1. Relationship of estimated maximum N uptake to soil solution
concentration 128
6-2. Relationship of estimated passive N uptake to soil solution
concentration 128
6-3. Proportions of nitrate-N, ammonium-N, and total-N from control
pipes as percentage applied during 3 days after application 131
6-4. Seasonal change in N concentration for flush leaves, expanded
leaves, and twigs for 2001 and 2002 132
6-5. Seasonal change in N concentrations for bark and wood tissue of
small, medium, and large limbs for 2001 and 2002. High N rate
and Low N rate equal to 268.8 and 179.2 kg ha'1 yr'1, respectively 135
6-6. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2001 season 139
6-7. Seasonal cumulative dry mass and N content of flowers, fruit,
and leaves collected from catch frames under mature citrus
trees for the 2002 season 140


91
and Alva 1999). Rogers and Bartholic (1976) determined that annual ETC increased at a
rate of 19 mm per year as trees grew, leading to a cumulative increase of approximately
13% in an 8-year period.
Soil water content can also be reduced by evaporation from the soil surface and
transpiration from non-crop species (Allen et al., 1998). Generally, soils lose the ability
to transport water to the surface as they dry (Hillel, 1998). Citrus ETC decreases as the
fraction of the soil surface receiving full sun decreases and the canopy shades an
increasingly larger ground area (Castel and Buj, 1992). Soil water use or apparent ETC
increases with increased ground coverage by non-crop species (Smajstrla et al., 1986).
The above factors combine to limit ET for a given crop under given conditions.
Allen et al. (1998) proposed that ETC can be derived from calculated ET0 as follows:
ETC = ET0 Kc Ks Equation 5-1
Where:
ETC = Crop evapotranspiration (mm d'1)
ET0 = Potential evapotranspiration (mm d'1)
Kc = Crop coefficient
Ks = Soil stress coefficient
The crop coefficient (Kc) is defined as the ratio of ETC to ET0 at field capacity (0fc) In
this case Ks is assumed to be equal to unity. This coefficient is indicative of climatic
and/or developmental effects on ETC compared with ET0 when water uptake is not limited
by soil water depletion. Estimates of Kc for citrus range from a minimum of 0.6 in the fell
and winter to a maximum of 1.2 during the summer months (Boman, 1994; Fares and
Alva, 1999; Martin et al., 1997; Rogers et al., 1983).


80
Fig. 4-1. Root length density distribution by depth at 50, 100, 150 and 200 cm distances
from the tree trunk between rows of Hamlin orange trees on Carrizo citrange
(A) or Swingle citrumelo (B) rootstocks.


CHAPTER 1
INTRODUCTION
With a crop value of $640 million in 2002, citrus is one of the most important
horticultural crops in Florida. Currently, nearly 2 million ha are under citrus production,
with a 1.1 million metric ton annual production accounting for 73 and 18% of US and
world production, respectively (Florida Agricultural Statistics Service, 2002). Citrus is
typically produced on sandy soils with poor water and nutrient retention capacity.
Adequate supply of both irrigation water and fertilizer are therefore required for optimal
production. Most ridge soils lack confining soil layers that can prevent fertilizer nitrates
from reaching groundwater. Two issues have become greater concerns for citrus
production in Florida: 1) increasing competition between agricultural, commercial, and
residential use of limited water supplies, and 2) nitrate contamination of some aquifers
less than 50 m deep.
Fertilizer application rates and irrigation management practices for citrus rely
upon crude general recommendations that are standardized over large areas and lack the
precision needed in todays ecologically conscious and competitive markets. Although
fertilizer and irrigation recommendations provide general production guidelines, they do
not capture the dynamic nature of processes controlling non-point source pollution
associated with citrus production. Therefore, both growers and regulators must be
provided with additional tools such as decision support systems to improve water and
1


93
delpetion in TAW to be 33% from February to June and 66% from July to January. These
values were determined for relatively low density plantings with overhead irrigation. The
depletion amounts for high density plantings irrigated with under-tree low volume
mircrosprinklers have not been determined.
K TAW-fl
* TAW-RAW
Equation 5-2
Where:
Ks = Soil water stress coefficient
TAW = 0fc pwp = Total available water (cm3 cm'3)
0 = Soil water content (cm3 cm'3)
RAW = 0fc 0ra = Readily available water (cm3 cm'3)
The hypotheses to be tested for the following study on citrus soil water dynamics
are: 1) seasonal changes in maximum daily water uptake under non-limiting soil water
conditions follows predictable patterns relative to ETo, 2) water uptake decreases with
soil water content, and 3) soil water uptake is greatest in soil volumes containing the
highest root length densities. The objectives of this study were to: 1) estimate mature tree
daily ETC during a 2-year period, 2) calculate monthly Kc values based on the relationship
Kc = ETC / (ET0*Kg), 3) determine the relationship of estimated ETC to soil water content
to determine Ks values over a range of , and 4) evaluate ETC per unit root length density.
The resulting relationships will provide critical information required for the development
of predictive models for citrus water uptake and soil water depletion over time. Such
models can provide a basis to protect Floridas water resources though better irrigation
scheduling and appropriate water application rates.
/


118
and wood during the sprouting period of 21 year-old Satsuma mandarins. Greatest
decreases in N were found in parts with higher concentrations of N (e g. leaves, shoots,
and fine roots). They also concluded that the trunk and large roots were main N reservoirs
for new shoot development.
Due to the perennial nature of citrus, leaf twig and branch biomass accumulated
in previous years periodically abscises. Wallace et al. (1945) estimated that citrus leaves
function on the tree up to 18 to 24 months before senescence. They found an average of
18.1 kg tree1 dry matter loss per year from mature Valencia orange trees grown in
California. Dry matter losses were 9.1,4.0, and 4.9 kg tree'1 for leaves, twigs, and
branches, respectively.
Information on N uptake rates and N cycling for the development of seasonal N
balance under Florida conditions on a field scale is lacking and will be critical to assess N
application quantity, frequency and timing decisions. In order to improve our
understanding of the underlying processes, the following hypotheses were tested: 1)
seasonal N uptake rates are related to leaf N status, 2) fertilizer-N is rapidly converted
into NO3-N that can be readily leached from typical ridge soils, 3) leaf N
concentrations are lowest during periods of high growth rate due to N dilution in the dry
matter, 4) changes in tree N reserves account for the majority of N in new leaves, and 5)
tree biomass and N senescence follow predictable seasonal patterns. The main goal of the
current studies was to provide critical information needed for a citrus N budget for a
citrus production system under Florida conditions. The objectives of this study were to 1)
determine seasonal changes in N uptake rates for citrus, 2) quantify changes in residual
soil N and nitrification with time in the absence of citrus roots, 3) measure seasonal


40
more of which was in the bark (5% of the dry mass of the tree). The biomass proportions
for 7 year-old Hamlin orange trees grown under Florida conditions reported by Mattos
(2000) were more similar to the 10 year-old trees cited above than the 3.5 year-old trees
(Table 1). Nitrogen concentration was lowest in the trunk and taproot of these trees. The
N concentration of leaves (2.1 to 2.6 %), twigs (0.4 to 0.8 %), and roots (0.6 to 1.7 %)
varied with tissue age. Younger tissue tended to have greater N concentration compared
with older tissues. Kato et al. (1984) and Feigenbaum et al. (1987) harvested older citrus
trees (21 and 22 years old, respectively). Leaves comprised a smaller fraction of total
biomass in both studies compared with branches and total roots. The leaves of these older
trees contained a lower proportion of total tree N than the branches, equal to the
proportion of N in the roots (Table 3-1). None of the above studies related biomass or N
measurements to tree size parameters such as canopy volume or trunk cross-section area.
Biomass and N distribution relationships based on tree size measurements as
opposed to tree age could provide more generic information needed for modeling tree
growth and N cycling in citrus production systems. Therefore, the hypotheses to be tested
in the following studies were. 1) functional relationships can be defined that correlate
biomass and N partitioning of specific tissue categories with tree size using generic
growth indicators such as canopy volume or trunk area, and 2) rootstock has a significant
effect on citrus biomass and N partitioning. Such relationships can be used to determine
citrus N budgets and develop specific fertilizer recommendations that will provide
adequate N for growth and production while protecting groundwater from nitrate
contamination. Thus, a non-destructive method of estimating an N budget is needed for
trees of unknown or mixed ages. Therefore, the objectives of the following studies were


159
Ke = hydraulic conductivity constant at 0 (cm s'1),
0 = soil water content (cm cm' ),
d0/d = slope of the soil characteristic curve, and
x = the distance over which the flux is maintained (cm).
3-1. Equation for tree canopy volume estimation
K
TCV = Ir Cr Ht *
4
d-(i-(^)2))
Where:
TCV = Tree canopy volume (m3)
Ir = In-row spacing (m)
Cr = Cross-row spacing (m)
Ht = Canopy height (m)
Int = Canopy intercept height (m)
5-1. Crop evapotranspiration (ETC) estimation
ETc = ET0*Kc*K8
Where:
ETC = Crop evapotranspiration (mm d1)
ET0 = Potential evapotranspiration (mm d'1)
Kc = Crop coefficient
K = Soil stress coefficient
44
91
5-2. Soil water stress Coefficient (Kc) estimation
93


78
grove were analyzed using Proc REG in SAS. Regression equations were determined
using SigmaPlot (SPSS, Inc., Chicago IL).
Results
Mature Hamlin Orange Root Distribution
Soil depth and distance from the tree trunk significantly (P = 0.01) affected citrus
root length density (Table 4-1). Mean root length density of fine fibrous roots (<4 mm)
extracted from soil cores surrounding the 12 mature citrus trees followed a bimodal
spatial distribution with depth from the soil surface (Fig. 4-1), and distance from the tree
trunk (Fig. 4-2). Mean fine fibrous root density in the upper 15 cm was 1.04 cm cm'3.
Densities ranged from 1.9 cm cm'3 soil at 50 cm from the tree trunk to 0.7 cm cm'3 at 200
cm. Mean densities decreased at the 15 to 30 cm depth to 0.30 cm cm'3 and ranged from
i
0.5 to 0.07 cm cm' at 50 and 200 cm distances, respectively. Mean densities of fine
fibrous roots increased at depths below 40 cm to a maximum at the 60 to 75 cm depth
(0.28 cm cm'3) then declined at the 75 to 90 cm depth (0.27 cm cm 3). Densities at the 60
to 75 cm depth were 0.3, 0.3, 0.3, and 0.03 cm cm'3 at distances of 50, 100,150, and 200
cm from the tree trunk, respectively.
Fine fibrous root densities at the 0 to 15 cm depth were generally greater in the in
row orientation than in the cross-row orientation (data not shown). Mean in-row spatial
root length densities (0.41 cm cm'3) were greater, but not significantly different from
densities for between-row orientation (0.35 cm cm3) (Table 4-1), because more overlap
in root systems from adjacent trees probably occurred in this orientation.


130
Nitrification Estimation
Changes in NO3-N and NH4-N content inside the control pipes were used to
estimate N loss and nitrification rates with time. Soil NO3-N, NH4-N, and total N as a
percentage of N applied inside the control pipes are shown as a function of time in Fig.
6-3. Mean NO3-N content in the upper 45 cm of soil increased 24 h after application to
150.6% of NO3-N applied. The mean NO3-N content decreased during the next 48 h to
125.8% of total NO3-N applied. Content of NH4-N decreased to 34.0% of that applied
after 24 h and steadily declined afta- application to 16.1% of NH4-N applied on day 3.
The sum of NO3-N and NH4-N (total N) declined throughout the period to 73.9% of total
N applied. This result indicated that approximately 26.1% of N was lost during the 3-day
period due to volatilization of NH4-N, microbial activity, or incorporation of N into
organic matter. Whether some of the immobilized N would become available at a later
time is unclear. Part of the N immobilization may be associated with recently decayed
root biomass in the control pipes which had relatively low C:N ratio. The length of time
that N would be lost at this rate is unclear and would certainly be greater if the fertilizer
were not incorporated with water. This loss would be emphasized if dry N fertilizer
sources were used. The nitrification rate in this soil was rapid, with a mean of more than
50% of the NH4-N converted to NO3-N within the first 24 h. Nearly 85% of applied NH4-
N was converted to NO3-N in 3 days, assuming all of the N loss was NH4-N.
Seasonal Tissue N Concentration
Leaf and twig N concentrations followed a cyclic pattern during the 2 years of periodic
sampling. Nitrogen values were significantly different at the P=0.05 level by month of
year and application rate, but not by rootstock. Therefore, values for specific N rates were


116
Syvertsen and Smith (1996) reported that leaching losses were generally small (2
to 9%) from low and medium N rate treatments in a lysimeter study, except when N
application coincided with frequent and/or intensive rainfall events. They also concluded
that 28% of applied N might have been lost in planted ly si meters due to volatilization
and/or denitrification. Immobilization into organic matter by soil microbes was not
considered a significant mechanism of N removal due to the very low organic matter
content of the Entisol used in the lysimeters. Estimated N uptake was 61% for the high N
application rate of 1.6 kg tree1 yr"1 and 83% for the lowest application rate of 0.3 kg tree"
1 yr"1. These values are similar to a previously estimated NUE of 68% (Syvertsen et al.
1993). Lea-Cox and Syvertsen (1996) and Scholberg et al. (2002) reported similar
findings of lower NUE with higher N application rate in a greenhouse studies.
Scholberg et al. (2002) found N uptake of greenhouse-grown seedlings to be
proportional to soil temperature, potential ET, and canopy biomass. Overall N uptake
increased with residence time in the root zone. Alva and Paramasivam (1998) reported
improved N use efficiency (0.36 to 0.39 Mg*1 fruit per kg N applied) of field grown
citrus, which was substantially greater than that reported by Koo and Smajstrla (1984)
(0.23 Mg'1 fruit per kg N supplied). This improvement was attributed to incorporation of
dry and fertigated nutrients under the canopy with light irrigation, applying no fertilizer
during the rainy season (between June and August), and maintaining adequate but not
excessive soil water content to 90 cm depth.
There is a positive relationship between the concentration of a nutrient in the soil
solution at the root surface and its uptake rate by plants. Passive nutrient uptake is
defined as the amount of a nutrient taken into a plant as solute associated with water


4
Results of this study indicated that N removed at harvest accounted for only 30 to
49% of the applied N (Alva and Paramasivam, 1998). Estimates of N added to the
biomass of the trees ranged from 18 to 57% of the N applied. The study concluded that
additional information was needed for N accumulation with increase in tree size, optimal
timing for N application, N uptake parameters, and improved irrigation scheduling
(Graham and Wheaton, 2000).
Citrus Best Management Practices
A best management practice (BMP) for any agricultural commodity is an attempt
to use the latest scientific data available to reduce the impact of agricultural operations on
the environment while maintaining economically viable production. An interim BMP for
citrus was established in 1994 that was based on previous N rate studies and current IF AS
recommendations. Citrus growers agreeing to abide by the interim BMP would not be
held liable by the Florida Department of Environmental Protection for future cost of
supplying drinking water to local users as required by Chapter 376.30 (3) (c) F.S.
(Graham and Alva, 1995).
The terms of the interim BMP for orange trees 4 years or more of age were quite
broad. Annual N applications were restricted to 134 to 269 kg ha'1 with the stipulation
that groves producing less than 50.4 Mg of fruit per ha should apply no more than 202 kg
ha'1 N annually. A minimum of two applications per year were required for bearing
groves receiving up to 168 kg ha'1 N. Bearing groves receiving more than 168 kg ha'1 N
per year were required to receive at least three applications. Those groves using
fertigation were required to make a minimum of 10 applications. Application of at least
half of the annual fertilizer N prior to the rainy season was encouraged. A UF-IFAS


95
Fig. 5.1. Illustration of EnviroSCAN probe. Number on sensor indicates depth of sensor.
^ Tree trunk
O EnviroSCAN probe
Fig. 5-2. Illustration of EnviroSCAN probe layout, and soil surface area used for
determining soil water content for each probe.


44
cm above the ground were determined for each tree by measuring in-row and cross-row
dimensions. Trunk cross-sectional areas (TCSA) were determined for each tree assuming
an oval shape.
TCV = Ir Cr Ht *
4
o-o-O
Equation 3-1
Where:
TCV = Tree canopy volume (m3)
Ir = In-row spacing (m)
Cr = Cross-row spacing (m)
Ht = Canopy height (m)
Int = Canopy intercept height (m)
Tree Biomass Fresh Weight
Fresh weight of the leaves, twigs <7 mm, small branches 7 to 15 mm, medium
branches 15 to 30 mm, large branches >30 mm, trunk, tap root, small roots <4 mm,
medium roots 4 to 20 mm, and large roots >20 mm were measured in the field. Field
weights and three samples of each plant part category were collected for each tree using
the following protocol: Twigs less than 7 mm in diameter and attached leaves were cut
from the tree with leaves intact. These twigs were placed into a plastic container and
weighed on a battery powered field-portable balance. During cutting, one twig out of 20
was placed into a separate container as they were cut and were weighed separately.
Leaves were removed from the twigs in this container while still in the field. Fresh
weights of these subsamples were measured. Branches 7 mm in diameter and greater
were cut into 15 to 30 cm segments, separated into the three size ranges noted above, and


149
trees on Carrizo. Conversely, root length densities were greater for trees on Carrizo
between 15 and 75 cm below the soil surface. Root length densities increased for trees on
Carrizo at the 45 to 60 cm depth, whereas densities for trees on Swingle increased at the
60 to 75 and 75 to 90 cm depths.
Root Length Density Distribution Changes with Tree Size
Distance from the tree trunk and depth from the soil surface significantly affected
citrus root length density across a wide range of tree sizes. Root systems of young trees
were initially concentrated at the soil surface, with few roots deeper than 0.5 m at a
distance of 150 cm from the tree trunk. As the citrus trees began to produce fruit (5 to 10
years of age) root length density increased at the soil surface to a distance equal to the
dripline of the tree. Roots extended to the 200 cm distance between tree rows and to a
depth of 0.9 m at 150 cm from the trunk. The bimodal nature of the root system was
observed close to the trunk at depths below 60 cm. By the time trees reached 10 to 15
years of age and the canopy was nearing a full hedgerow, the bimodality of the root
system was fully developed and roots extended below lm at all distances from the tree.
Seasonal ET0 and ETC Trends
Daily ET0 reported by FAWN for the experimental site ranged from a minimum
of 1.12 mm in December 2001 to a maximum of 6.48 mm in June 2001. Monthly
maximum, minimum, and mean ET0 and ETC were not significantly different for the same
months during the 2 years of this study. Although generally lower, daily ETC followed the
same seasonal patterns as ET0. The exception to this correlation occurred during the
summer months of June through August, and then only when soil was near field capacity.


I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
(X .
Thomas A. Obreza, Chair
Professor of Soil and Water Science
I certify that 1 have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Johannas M. S. Scholberg, Cochair
Assistant Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
es W. Jones,
distinguished Professor of Agricultural and
Biological Engineering
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Nicholas B. Comerford *
Professor of Soil and Water Science
I certify that I have read this study and that in my opinion it conforms to acceptable
standards of scholarly presentation and is full adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Thomas Adair Wheaton
Professor Emeritus of Horticultural Science
This dissertation was submitted to the Graduate Faculty of the Collage of Agricultural and Life
Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements
for the degree of Doctor of Philosophy.
May 2004 v ^ v. \ ^
Dean, College of Agricultural and LifejSciences
Dean, Graduate School