Nitrogen dynamics in a pecan (Carya illinoensis K. Koch)-cotton (Gossypium hirsutum L.) alley cropping system in the sou...

MISSING IMAGE

Material Information

Title:
Nitrogen dynamics in a pecan (Carya illinoensis K. Koch)-cotton (Gossypium hirsutum L.) alley cropping system in the southern United States
Physical Description:
xvi, 98 leaves : ill. ; 29 cm.
Language:
English
Creator:
Allen, Samuel C ( Samuel Clifton )
Publication Date:

Subjects

Subjects / Keywords:
Intercropping -- Southern States   ( lcsh )
Crops and nitrogen -- Southern States   ( lcsh )
Pecan   ( lcsh )
Cotton   ( lcsh )
Forest Resources and Conservation thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Forest Resources and Conservation -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2003.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Samuel Clifton Allen.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030281464
oclc - 52731565
System ID:
AA00014261:00001


This item is only available as the following downloads:


Full Text











NITROGEN DYNAMICS IN A PECAN (Carya illinoensis K. Koch)-COTTON
(Gossypium hirsutum L.) ALLEY CROPPING SYSTEM IN THE SOUTHERN UNITED
STATES














By

SAMUEL CLIFTON ALLEN


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

UNIVERSITY OF FLORIDA
































Copyright 2003

by

Samuel Clifton Allen































This work is dedicated to the memory of my mother, Harriet Elise Priester Allen
(15 December 1933 24 October 2002).














ACKNOWLEDGMENTS

This project represents a team effort on the part of many people at both the Milton

and Gainesville campuses of the University of Florida. I must first express my sincerest

thanks to my graduate committee chair Dr. Shibu Jose (UF Milton), whose ongoing

direction, encouragement, and financial support via a graduate assistantship made this

study a reality. Dr. P.K. Ramachandran Nair (UF Gainesville), my committee cochair,

was a great source of inspiration during this study of agroforestry as well. I am also

grateful to my committee members Drs. Barry Brecke, Tim Martin and Peter Nkedi-

Kizza for their contributions to this study and for their ongoing support of my academic

progress.

I am deeply grateful for the gracious and professional assistance provided by the

faculty and staff of the West Florida Research and Education Center (WFREC) of the

University of Florida (Milton) during the entire course of this study. Specifically, I wish

to thank Dr. Jeffrey Mullahey (Center Director), Chris Gilmore, Lisa Griswold, Amanda

Larsen, Barbara May, Nik McCue, Sara Merritt, Rick Puckett, Josiah Raymer, Jeanette

Ross and Robin Vickers, and others at WFREC who have contributed to the successful

implementation of this study. I would also like to extend my special thanks to Dr.

Leonard Dunavin, Associate Professor Emeritus, who was never too busy to provide a

listening ear. Very special thanks are due to Drs. Kyehan Lee and Craig Ramsey, two

outstanding postdoctoral researchers who have been my patient friends and mentors over

the past two years. Many fellow graduate students at UF Milton have provided me with








practical help or moral support as well, and these include my friends Maheteme

Gebremedhin, Dawn Henderson, Sanjaya Ranasinghe, Andrew Ruth, Robert Wanvestraut

and Diomides (Diomy) Zamora. Thanks are also due to the many OPS and student

workers at UF Milton who have helped with data collection, sample preparation,

logistical support or other assistance-Chris Adkison, Sean Claypool, Brenda Hahr,

Cathy Hardin, Clay Hayes, Mitch Johnson, Jason Liddle, Jennifer Liddle, Leah McCue,

Chris Payne, Troy Rutherford, David Vaughn, Lewayne White and Andy Whitehurst.

Lastly, I must thank Doug Hatfield (farm manager), Tim Baxley, Sydney Betz, Michael

Dozier, Greg Kimmons, Rex Lawson, Kenny McCreless, Robert Murrell, Joe Nelson,

Thomas Salter III, Vernon Tedder and the other staff of the WFREC Farm (Jay, Florida),

who provided the expert field support that made this study possible.

My appreciation is also expressed to the many people within the School of Forest

Resources and Conservation (SFRC) and the larger Gainesville campus of the University

of Florida, who provided assistance, advice or encouragement during this study. This

includes Dr. Wayne Smith (SFRC Director) and the many faculty under whom I have had

the pleasure of studying since August 1999, including Drs. Henry Gholz, Eric Jokela,

Timothy Martin, P.K. Nair, Donald Rockwood, Robert Schmidt and Timothy White. I

appreciate the key involvement of Drs. Michael Bannister and Sarah Workman of the

Center for Subtropical Agroforestry (CSTAF) as well. Lastly, I want to acknowledge the

assistance of the very capable staff of the SFRC, which includes Cherie Arias, Dawnette

Lauramore, Cindy Love, Marie Meldrum, Scott Sager, Sherry Tucker, Willie Wood and

others.








Thanks are expressed to Drs. Hector Adegbidi, Vimala Nair and Don Graetz (Soil

and Water Science) for their key involvement in the latter phases of the study, and to Dr.

Ken Woodard (Agronomy) for his expert guidance on lysimeter construction and

operation. Thanks are also due to Elizabeth Kennelly, Beverly Welch, and the staff of the

UF Analytical Research Laboratory (ARL) for carrying out the soil and water analyses,

and to Dr. Jason Curtis (Geological Sciences) for conducting the stable isotope analysis.

I wish to give credit to my mentors in the field of forestry and agroforestry, some

of whom have already been cited (Drs. Shibu Jose and P.K. Nair), as well as Drs. Felix

Eslava, Rodel Lasco and Enrique (Ike) Tolentino, of the University of the Philippines at

Los Bafios, and Jeff Palmer and Harold Watson, formerly of the Mindanao Baptist Rural

Life Center, Davao del Sur, Philippines. I also wish to thank the many other friends who

have inspired and befriended me during my graduate student days, including Andrea

Albertin, Michelle Andrianarisata, Dr. Kent Apostol, Dr. Chip Appel, Kathlene Arofio,

Kathryne Arofio, Mario Aviles, Brian Becker, John Bellow, Fred Boltz, Richard

Cardellino, Tammie Coffman, Alyson Dagang, Kristen Davis, Dr. Armando and Ruth De

la Cruz, Mark Drew, Dr. Edward Ellis, Dr. Rico Gazal, Jennifer Hale, Jimmy Knowles,

Dr. Raj and Lena Kumar, Marcus LaPratt, Sichin and Dong Liu, Dr. David May, Robert

McGarvey, Dr. Robert Miller, Soumya Mohan, Abiud Mwale, Nicole Mytyk, Dr. J.L.

and Lynette Peeples, Dr. Stephen Trolove, Dom and Carleta Underwood, Keith

Yearwood and Dr. Yu Xiao. A thank you is also reserved for Daniel and the crew at Jin

Jin Chinese restaurant (Milton) for providing numerous hot, nourishing late-evening

meals to a hungry graduate student.








Lastly, I wish to express my heartfelt thanks to my wife, Carina (Ter) Allen, for

her ongoing moral support and encouragement, and for her gracious help with field work

and sample preparation, in which she always managed to be "beautiful while busy." I am

also thankful to Walter Allen and Kelli Sue Allen, my brother and sister-in-law, for their

great encouragement and prayers, and to my late parents, Dr. James and Harriet Allen,

who always had confidence in me. Most importantly, I give all credit and honor from

this study to God, the Author of all Creation, and to Jesus Christ, who designed our world

in such a beautiful, intricate way, and who desires that all should come to know Him and

to benefit from this earthly inheritance, both now and in future generations.

This study was funded in part by two grants from the USDA Southern Region

Sustainable Agriculture Research and Education (SARE) program (# GS01-009 and

# LS02-136), and by a grant from USDA-IFAFS (# 00-52103-9702) through the Center

for Subtropical Agroforestry of the University of Florida.















TABLE OF CONTENTS
Page

ACKNOWLEDGMENTS .................... ... ......... ....................... .. iv

LIST OF EQ U A TION S .................................................................. ....................... xi

LIST OF TABLES ............................... .................xii

LIST OF FIGURES ............................................... xiii

A B ST R A C T .............................................................................. ...................................xv

CHAPTER

1 IN TR O D U C TIO N ............................................................... ............................

2 NITROGEN MINERALIZATION UNDER A PECAN (Carya illinoensis K.
Koch)-COTTON (Gossypium hirsutum L.) ALLEY CROPPING SYSTEM
IN THE SOUTHERN UNITED STATES ......................... ........................7

Introduction ............................................................................. .......................7
M materials and M methods ............................................................ ....................... 9
Study Area and Configuration..............................................9
Plot Layout and Fertilizer Application....................... ............................10
Sampling M ethods............................ ............................... 1
Data Analyses................. ............ ....................... 12
Results ................................... ........ ........... .......... ..................... 13
Soil W ater Content ....................... .................................. 13
Initial Soil Ammonium and Nitrate...................... ....................... 14
Ammonification ................... .. ....... ......................14
N itrification .................................................................. ...................... 15
M ineralization ................................... ............................15
Fallow Effect and M ineralization................................ ............................. 16
Initial Nitrogen Source and Nitrogen Transformations...............................16
Cotton Lint Yield ................................... ..........................17
D discussion ................................................................................ ...................17
Conclusions ................ .................. ........................... 21








3 GROUNDWATER NITRATE DYNAMICS IN A PECAN (Carya illinoensis
K. Koch)-COTTON (Gossypium hirsutum L.) ALLEY CROPPING SYSTEM
IN THE SOUTHERN UNITED STATES ....................... ..........................34

Introdu action ........................................ ................................................................34
M materials and M ethods .................................. ..... ...... .. .................... 37
Study Area and Configuration...................... ... ......... .............37
Sampling Methods....................... .......... ........................38
W ater Drainage ................................... .............................39
Leaching of Ammonium and Nitrate .................... ......................39
Net Retention Index ..................... ...................................40
Data Analyses....................... ................... .............. ............40
Results and D iscussion........................................................... ...................... 41
Drainage ........................................ ........... ............ ....................... 41
Ammonium Concentrations ................................... ....................41
Nitrate Concentrations................................................42
Ammonium Leached......................... ....... .......................44
N itrate Leached ........................................ ........................ ..................45
Net Retention Index of Ammonium and Nitrate..........................................47
C onclu sions ............................................................................ ............................49

4 COMPETITION FOR NITROGEN IN A PECAN (Carya illinoensis K.
Koch)-COTTON (Gossypium hirsutum L.) ALLEY CROPPING SYSTEM
IN THE SOUTHERN UNITED STATES ..................... ...........................57

Introduction .............................................................................. ..........................57
M materials and M ethods .......................................... ....................60
Study Area and Configuration...................... ... .......................60
Microplots and Fertilizer Application........................ ......................61
Sampling M ethods............................ ............................61
Data Analyses........................... ....... ... ........................64
R results ................................................................................ ................................64
Aboveground Biomass ........................................................64
Nitrogen Concentration and Content......................... ....................... 65
Uptake of Fertilizer Nitrogen.................................. .......................66
Fertilizer Nitrogen Use Efficiency ............................................................67
Atom Percent tN Abundance and Nitrogen Concentration in Soil...............68
Recovery of 'SN in Soil ................... .. ... .............................68
D discussion .......................................................... ............................ ..............69
Conclusions ...........................................................................75

5 SUMMARY AND CONCLUSION...................... ...... .....................84









APPENDIX

LEACHMN PARAMETER INPUT VALUES ..................................................88

LIST OF REFERENCES...................... ..................... ................ ............89

BIOGRAPHICAL SKETCH .......................................................98
















































x















LIST OF EQUATIONS


Equation page

2-1. N et am m unification ........................................................... ........................ 12

2-2. N et nitrification................................................................... ....................... 12

2-3. N et m ineralization............................................................. ......................... 12

3-1. R ichard's equation .............................................................. .......................39

3-2. Leaching of ammonium and nitrate................................... ......................40

3-3. Net retention index (NRI) ...................................... ........................ 40

4-1. Percentage of plant nitrogen derived from fertilizer (NDF) ................................63

4-2. Percentage utilization of fertilizer nitrogen (UFN)............................................. 63

4-3. Percentage recovery of 15N fertilizer in soil (RFNsoi) .........................................63














LIST OF TABLES


Table page

2-1. Surface soil chemical characteristics of Rhodic Paleudult sandy loam soil of
the Jay, FL agroforestry site................................. ....................... 22

2-2. Seasonal and cumulative rates ofN ammonification, nitrification and
mineralization in barrier and non-barrier treatments ...........................................23

3-1. Seasonal and overall averages ofNH4-N and N03-N concentrations in soil
water extracted from 0.3 and 0.9 m depths in barrier and non-barrier
treatm ents.............................. ..... .................................................. 50

3-2. Average monthly rates of NH4-N and N03-N leached at 0.3 and 0.9 m
depths in barrier and non-barrier treatments.................................................51

3-3. Seasonal and cumulative totals ofNH-4-N and NO3-N leached at 0.3 and
0.9 m depths in barrier and non-barrier treatments...............................................52

3-4. Net Retention Index of NH4-N and NO3-N at 0.3 and 0.9 m depths over two
growing seasons and whole study period..................... ..................................53

4-1. Biomass of cotton leaf, stem and seed cotton in barrier and non-barrier
treatm ents............................................................................. ...................... 77

4-2. Nitrogen concentration and nitrogen content in cotton leaf, stem and seed
cotton in barrier and non-barrier treatments ........................ ......................78

4-3. Percentage of nitrogen derived from fertilizer (NDF) and percentage
utilization of fertilizer nitrogen (UFN) for cotton leaf, stem and seed cotton
in barrier and non-barrier treatments ................................... .....................79

4-4. Percentage of nitrogen derived from fertilizer (NDF), percentage utilization
of fertilizer nitrogen (UFN), and other physiological parameters for pecan
trees in barrier and non-barrier treatments..................... ..............................80

4-5. Percentage of 1SN recovery in soil at end of growing season in barrier and
non-barrier treatm ents.......................................... .........................................81

A-1. LEACHMN parameter input values used for modeling field drainage at
the Jay, Florida agroforestry site................................. ......................88














LIST OF FIGURES


Figure page

2-1. Plot layout showing location of in-situ soil incubation bags at the Jay, Florida
agroforestry site.................................................... ... ................................ 24

2-2. Monthly averages for air temperature, soil temperature (at 10 cm depth),
precipitation, and soil water content at the Jay, Florida agroforestry site................25

2-3. Monthly averages of initial soil NH4-N and N03-N levels in alleyways of
barrier and non-barrier treatments...................................................26

2-4. Monthly rates ofN ammonification, nitrification and mineralization in alley
and tree row locations .......................................................... ........................27

2-5. Monthly N mineralization rates in alley and tree row locations as influenced
by initial levels of NH4-N and N03-N ................................... ..........................28

2-6. Monthly N mineralization rates in alley and tree row locations as influenced
by soil temperature (10 cm depth)................................ .......................29

2-7. Monthly N mineralization rates in alley and tree row locations as influenced
by soil water percentage during the 2001 and 2002 growing seasons ...................30

2-8. Seasonal rates ofN ammonification, nitrification and mineralization in
alleyways of barrier and non-barrier treatments as influenced by application
of inorganic fertilizer vs. poultry litter during the 2002 growing season ...............31

2-9. Lint cotton yield in barrier, non-barrier and monoculture treatments during the
2001 and 2002 growing seasons................ ...... .............................32

2-10. Lint cotton yield in barrier and non-barrier treatments as influenced by
application of inorganic fertilizer vs. poultry litter during the 2002 growing
season ............................ ........ ....................... ......... ....................... 33

3-1. Plot layout showing location of suction lysimeters at the Jay, Florida
agroforestry site................................................. .............................................. 54

3-2. Monthly and cumulative drainage values at 0.3 and 0.9 m depths in barrier
and non-barrier treatments as estimated using the LEACHMN model....................55








3-3. Average monthly concentrations of NH4-N and N03-N in soil solution at
0.3 and 0.9 depths in barrier and non-barrier treatments..................................56

4-1. Plot layout showing location of '1N-enriched microplots at the Jay, Florida
agroforestry site............................................ ................................................ 82

4-2. Mean atom %15N abundance and total %N in soil at end of growing season in
barrier and non-barrier treatments....................... .. ................................83














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 DYNAMICS IN A PECAN (Carya illinoensis K. Koch)-COTTON
(Gossypium hirsutum L.) ALLEY CROPPING SYSTEM IN THE SOUTHERN UNITED
STATES


By

Samuel Clifton Allen

May 2003

Chair: Dr. Shibu Jose
Cochair: Dr. P.K. Nair
Major Department: School of Forest Resources and Conservation

A pecan-cotton alley cropping system was established in northwestern Florida in

Spring 2001 to assess tree-crop competition for nitrogen (N) and its effect on

mineralization rates and groundwater nitrate levels, and nitrogen use efficiency.

Polyethylene root barriers were used to prevent belowground interaction between pecan

and cotton in half the number of test plots, for the duration of the 17-month study (June

2001-October 2002).

The study first examined the effect of tree roots on nitrogen transformations in

soil. It was observed that temporal variations in net ammonification, nitrification and

mineralization were driven primarily by environmental factors (such as soil moisture

content and soil temperature), and by initial ammonium and nitrate levels. In general,

greater nitrification and mineralization rates were observed in the non-barrier treatment

due to higher soil nitrogen. Cotton lint yield reductions were observed in the non-barrier








treatment during both years compared to the barrier treatment, likely due to interspecific

competition for water. In addition, source of N was found to have a significant effect on

cotton yield, with inorganic fertilizer resulting in higher yields in the barrier treatment

compared with organic poultry litter.

The study also examined the "safety net" hypothesis to determine whether tree

roots were able to capture nitrate and ammonium leached below the crop root zone. In

general, the presence of trees in the non-barrier treatment resulted in decreased soil

solution nitrate concentrations and nitrate leaching rates.

Lastly, the results indicated that competition for fertilizer N was minimal because

of differences in temporal patterns of pecan and cotton nitrogen demand, although NDF

may have occurred in unstudied portions of pecan tree tissue. Nitrogen use efficiency of

cotton in barrier treatment was shown to be higher, indicating a greater ability to utilize

the available nitrogen.

Overall, this study reveals that the competitive presence of trees can be utilized to

decrease soil nitrate concentrations and reduce nitrate leaching. This knowledge will

help to improve our understanding of temperate alley cropping systems and to design

systems that utilize the safety net process to maximize nitrogen use efficiency and

minimize groundwater pollution.













CHAPTER 1
INTRODUCTION

Individuals and institutions in the world's temperate regions are increasingly

taking notice of the science and art of alley cropping. This is due in part to growing

concerns over the long-term sustainability of intensive monocultural systems. In the

temperate context, alley cropping involves the planting of timber, fruit or nut trees in

single or multiple rows on agricultural lands, with crops or forages cultivated in the alleys

(Nair 1993, Garrett and McGraw 2000). Major purposes of this type of agroforestry

system include production of tree or wood products along with crops or forage;

improvement of crop or forage quality and quantity by enhancement of microclimatic

conditions; improved utilization and recycling of soil nutrients for crop or forage use;

control of subsurface water levels; and provision of favorable habitats for plant, insect or

animal species beneficial to crops or forage (USDA 1996, Garrett and McGraw 2000).

In the southern United States, pines (Pinus spp.) have been intercropped with row

crops such as cotton (Gossypium spp.), peanuts (Arachis hypogaea), maize (Zea mays

L.), soybean (Glycine max. L. (Merr.)), wheat (Triticum spp.), oats (Avena spp.) and

watermelon (Citrullus lanatus) (Zinkhan and Mercer 1997; Ramsey and Jose 2001).

Pecan (Carya illinoensis L.), an important nut-bearing species, has been intercropped

with soybean, squash (Cucurbitaceae spp.), potatoes (Solanum tuberosum), peaches

(Prunus persica), raspberries (Rubus spp.), various grains, and other crops (Nair 1993;

Williams et al. 1997; Zinkhan and Mercer 1997; Cannon 1999; Long and Nair 1999; Reid

1999; Ramsey and Jose 2001).








As an association of plant communities, alley cropping is deliberately designed to

optimize use of spatial, temporal and physical resources by maximizing positive

interactions (facilitation) and minimizing negative ones (competition) between trees and

crops (Jose et al. 2000a). For example, trees in these systems are capable of improving

soil nutrient status and mineralization patterns (Nair 1993, Palm 1995, Rowe et al. 1999),

thereby improving overall system productivity. Trees are also capable of capturing and

recycling leached fertilizer nutrients and are thus a potential moderating factor in

groundwater pollution caused by leaching of nitrate (N03-N) (Williams et al. 1997,

Garrett and McGraw 2000). In addition, trees on agricultural lands offer landowners the

possibility of accruing carbon via the sequestration of stable carbon stock, an added

incentive for adopting alley cropping (Dixon 1995, Williams et al. 1997, Sampson 2001,

Nair and Nair 2003). However, adoption of alley cropping and other agroforestry

systems has been hampered by a lack of information on how interspecific interactions

affect system management, productivity and sustainability. This is especially true for

temperate agroforestry systems, where research efforts have gained momentum only in

recent years.

The effect of the tree-crop environment on nitrogen (N) ammonification,

nitrification and mineralization patterns in understory soil is an important first

consideration in the design of alley cropping systems. Because nitrogen is generally the

most limiting nutrient in agricultural soils, improving the prediction of nitrogen

transformations in soil is important in being able to maximize usage of natural and

applied nitrogen by crops, and in improving and maintaining site productivity and water

quality (Mary et al. 1999).








This holds true for both organic and inorganic sources of fertilizer N. The use of

poultry litter, for example, as an organic fertilizer for cotton production has been studied

in recent years (Nyakatawa et al. 2000, Malik and Reddy 2001). Hence, understanding

the nutrient release patterns from organic fertilizer compared with conventional inorganic

fertilizer is important in managing organic inputs and optimizing productivity in

agroforestry systems.

Equally important to agricultural system sustainability is the fate of nitrogen

fertilizers and their effect upon groundwater quality. On a national scale, over-

application of N increases the production costs of farmers by millions of dollars each year

(USDA 1998a). Moreover, because nitrates are highly soluble, they are easily

transported through the soil matrix (Aelion et al. 1997), where they may be carried away

by runoff, or leached through the soil profile into the water table (USDA 1998a, Nair et

al. 1999). Such contamination can create conditions for eutrophication and related

ecological disruptions of rivers, lakes, estuaries and aquifers (Johnson and Raun 1995,

USDA 1998a, Bonilla et al. 1999, Ng et al. 2000). From a human health standpoint,

nitrate is a toxin that can find its way into public water supplies or private drinking water

wells. Nitrate-contaminated drinking water has been shown to cause a respiratory

deficiency known as methemoglobinemia ("blue baby syndrome") in infants under six

months of age, and similar problems in elderly adults (Sawyer et al. 1994, Baker 1998,

Bonilla et al. 1999, Ng et al. 2000, Reddy and Lin 2000).

In this regard, the effect of trees in alley cropping systems is of interest due to the

mechanism of nutrient capture, in which deep roots of trees serve as a "safety net" for

capturing nitrates and other nutrients that leach below the root zone of crops (van








Noordwijk et al. 1996, Rowe et al. 1999). At lower depths, tree roots can exploit subsoil

nitrate beyond the rooting depths of crops. A portion of these nutrients that are absorbed

by the trees are later returned to the soil surface through decomposition of fine roots and

litterfall, representing a gain to the soil nutrient pool (Nair 1993, Jose et al. 2000b). This

phenomenon is of importance because it serves as a possible mechanism for nutrient

conservation as well as groundwater clean-up.

From an agronomic standpoint, the effects of tree-crop interactions must

ultimately be considered in light of plant nutrient use efficiency and productivity. In

general, interspecific competition for nitrogen is an important determinant of productivity

since N is generally limiting in such systems. Nitrogen is lost via various

biogeochemical processes such as volatilization, denitrification and leaching. Nitrogen is

also lost when crop biomass is removed from the field following harvest. In addition,

plants of the same species and growth stage can compete heavily for soil nitrogen when

depletion zones of neighboring plants overlap. Moreover, in alley cropping systems,

competitive forces can be even more intense, as most tree species have the bulk of their

fine, feeder roots in the top 30 cm soil layer, thus placing them in a zone of competition

with crop species for water and nutrients (Rao et al. 1993, Lehmann et al. 1998a). Tree-

crop systems must therefore be properly designed and managed in order to maximize

fertilizer use efficiency and minimize deleterious effects of competition on crop yield.

The extent of competition between two species will depend on factors such as

nutrient and water availability, root architecture, rooting depth and proximity to

competing roots, and temporal nutrient demand (Jose et al. 2000a). For example, the

peak intensity of nutrient demand in trees and crops may differ by several months. Trees








tend to exhibit highest nutrient demand in spring during leaf formation, while for crops

such as cotton, the demand would be highest in mid-summer during boll formation.

Pecan-based alley cropping systems offer potential for landowners in the southern

United States, given the large number of pecan orchards in the region, and the possible

environmental and financial benefits that may be accrued from such systems. However,

the fate of nitrogen in such systems remains a critical, largely unstudied factor affecting

system productivity and sustainability. While nitrogen losses cannot be avoided

completely, losses can be minimized through appropriate fertilizer and orchard

management practices and by knowledge of how nitrogen moves in the soil-tree system

(Herrera and Lindemann 2001). Thus, more understanding is needed of the interactive

dynamics of nitrogen in tree-crop systems, in order to maximize fertilizer use efficiency

and optimize production from each component.

A 17-month research project was conducted at the West Florida Research and

Education Center Research Farm of the University of Florida (Jay) to examine the

competitive interactions involving nitrogen in a pecan-cotton alley cropping system. It is

important to have sound knowledge of the nitrogen dynamics in alley cropping systems

so that successful systems can be developed and optimized. This research will aid the

efforts of researchers and landowners to make alley cropping an ecologically viable and

environmentally appealing land use practice.

This study was therefore undertaken with the following three objectives:

* To determine the effect of tree-crop competition on nitrogen ammonification,

nitrification and mineralization patterns;





6

* To determine the degree to which nutrient uptake in trees affects groundwater

ammonium (NH4-N) and nitrate (N03-N) levels and leaching rates; and

* To quantify competition for nitrogen between pecan and cotton.

These objectives comprise chapters 2, 3 and 4 of this dissertation, respectively.

Chapter 5 presents a summary and conclusion.













CHAPTER 2
NITROGEN MINERALIZATION UNDER A PECAN (Carya illinoensis K. Koch)-
COTTON (Gossypium hirsutum L.) ALLEY CROPPING SYSTEM IN THE
SOUTHERN UNITED STATES

Introduction

Nitrogen (N) is generally the most limiting nutrient in agricultural soils.

Improving the prediction ofN ammonification, nitrification and mineralization patterns

in soils is therefore important in being able to maximize usage of natural and applied

nitrogen by crops, and in improving and maintaining site productivity and water quality

(Mary et al. 1999).

Alley cropping systems modify the status of soil nutrients in various ways.

Generally, the inclusion of woody species on farmlands improves soil fertility. For

example, trees help to increase the organic matter content of soil through the addition of

leaf litter and other parts from trees, and they generally provide for more efficient cycling

of nutrients (Nair 1987, Palm 1995). A tree-soil system can also moderate extreme soil

reactions via the increased soil organic matter, improve nutrient release and availability

patterns, and provide a more suitable environment for beneficial microorganisms in the

rooting zone (Nair 1987, Lee and Jose 2003). However, in many instances, tree roots can

lower soil nutrient availability of the associated crop species through competition for

water and nutrients (Jose et al. 2000a,b). The long-term effects of nutrient depletion by

tree roots on soil N mineralization have not been explored in temperate agroforestry

systems.








In addition, the use of organic sources of nitrogen has become an important

practice in low input production systems such as alley cropping and organic farming. For

example, the use of poultry litter as an organic fertilizer for cotton production has been

reported in recent years (Nyakatawa et al. 2000, Malik and Reddy 2001). Poultry litter,

which is high in nitrogen, is viewed as a viable alternative to conventional fertilizers,

provided that it is applied at an appropriate rate and timescale for crop production

(Mitchell and Donald 1999). Understanding nutrient release patterns in poultry litter and

comparing them with those in conventional inorganic fertilizer is an important first step

in managing organic inputs in alley cropping and other agroforestry systems.

Rates of soil N mineralization often vary with the amount and composition of soil

organic matter, soil water availability and soil temperature (Reich et al. 1997, Noble and

Randall 1998, O'Connell and Rance 1999). The rate of N mineralization can be affected

by physical, chemical, biochemical, and microbiological properties of soils (Deng and

Tabatabai 2000). Although several studies have examined the factors affecting soil N

mineralization in tropical agroforestry systems (e.g., Palm and Sanchez 1991, Ikerra et al.

2001, Menezes et al. 2002), such studies have been few in temperate agroforestry systems

(Thevathasan and Gordon 1997, Seiter and Horwath 1999).

The first objective of this study was to examine the effect of belowground root

competition by trees on soil nitrogen mineralization. We hypothesized that tree root

competition would lower ammonium (NH4-N) and nitrate (N03-N) levels in soil in the

non-barrier treatment, which would eventually result in lower rates of N mineralization

and crop yield. The second objective was to compare mineralization rates between

organic poultry litter and regular inorganic fertilizer. Since the alley cropping system








was established in a 49-year-old pecan orchard fallowed for 29 years prior to initiation of

this study, the fallow effect on mineralization and cotton yield was also examined.

Materials and Methods

Study Area and Configuration

This study was conducted at the West Florida Research and Education Center

Farm of the University of Florida, located near Jay in northwestern Florida, USA (30089'

N Lat., 8713' W Long.). The climate is temperate with moderate winters and hot,

humid summers. The soil at the site is classified as a Red Bay sandy loam, which is a

fine-loamy, siliceous, thermic Rhodic Paleudult. The average water table depth is 1.8 m.

The main chemical characteristics of the soil are described in Table 2-1.

For the current study, a pecan-cotton alley cropping system was initiated in

Spring 2001 from an existing orchard of pecan trees planted in 1954. The orchard,

arranged in a 5x20 grid pattern with trees spaced 18.28 m apart, had remained under non-

intensive clover (Trifolium spp.) and ryegrass (Lolium spp.) production for 29 years prior

to the initiation of the current study. For the study, ten plots were demarcated within the

orchard and arranged into five blocks using a randomized complete block design. Each

plot, which consisted of two rows of trees oriented in a north-south direction, was 27.43

m long and 18.28 m wide, with a practical cultivatable width of 16.24 m, and was

separated from its adjacent plot by a buffer zone of the same dimensions.

To assess tree root competition, each block was randomly divided into a barrier

plot and a non-barrier plot. Barrier plots were subjected to a root pruning treatment in

March of 2001 in which a trenching machine was used to dig a 0.2 m wide x 0.9 m deep

trench along both sides of the plot at a distance of 1.5 m from the tree line to separate root








systems of pecan and cotton. A double layer of 0.15 mm-thick polyethylene sheeting was

used to line the ditch prior to mechanical backfilling. The barrier plots thus served as the

tree root exclusion treatment, preventing interaction of tree and cotton roots, while the

non-barrier plots, which did not receive this treatment, served as the tree-root competition

treatment. A sixth block was added in Spring 2002 to create an additional barrier and

non-barrier plot of the same dimensions as the other plots, forming 12 plots total for the

alley cropping system. For control purposes, three additional plots in the orchard were

maintained as a pecan monoculture (no cotton plants), and two plots in an adjacent field

were maintained as a cotton monoculture (no pecan trees).

In addition to the root-barrier study, a second study was initiated in the same alley

cropping system in June 2002 to assess differences in mineralization rates between

regular inorganic fertilizer and organic poultry litter. For both barrier and non-barrier

plots, half (six) of the plots received regular inorganic fertilizer at a rate of 89.6 kg N ha",

and half (six) received organic poultry litter at an equivalent rate.

Plot Layout and Fertilizer Application

For this study, cotton (DP 458 B/RR) was planted in rows 0.91 m apart at 16 rows

per alley in a north-south orientation on 16 May 2001 and 13 May 2002, after disking of

the alleys. A 3-9-18 fertilizer blend was applied to test plots at a rate of 89.6 kg N ha-l on

19 June 2001 and 5 June 2002. For the 2002 application, half of the test plots received

the conventional inorganic fertilizer, and half received organic poultry litter (2-3-2

analysis) (Black Gold Compost Co., Oxford, Florida) that was distributed using a rotary

manure spreader. Application of poultry litter was in split applications, with 2/3 rate








being distributed on 5 June, and 1/3 rate being distributed on 24 June. Conventional

insecticide and herbicide were applied during the growing season as recommended.

Sampling Methods

Monthly ammonification, nitrification, and N mineralization were determined

from July 2001 to October 2002 using the in-situ buried bag technique (Eno 1960) at

specific distances (0, 1.5, 4.2 and 8.4 m from tree) in tree rows and alleys (Figure 2-1).

For each of 10 (later 12) plots, soil was collected from 7 sites per plot: at tree base, on the

first, fourth and eighth rows of cotton, and at commensurate distances in the tree row.

For each sampling site, soil samples were collected in pairs in the top 10 cm of soil using

a hand spade, after which half were removed to the lab for processing and half were

placed into 0.05 mm-thick Fisherbrand polyethylene zipper-seal bags (12.7 cm x 20.32

cm) and returned to the soil profile for month-long incubation. Coarse roots and large

organic debris were removed by hand to avoid extraneous N immobilization during

incubation. Following incubation, the in-situ samples were collected and transported to

the laboratory for processing. All soil samples were kept at 40C until processing.

From each soil sample, a 20 g subsample was separated for ammonium and nitrate

extraction, and an 18-22 g subsample was collected for water content determination. Soil

water content was determined gravimetrically by drying a subsample for 24 hours at

105C. For N extraction, 20 g of soil was mixed with 50 ml of 1 M KCI solution in a 120

ml sample vial, shaken for 1 hr using a Lab-Line Orbit Shaker (Barstead International,

Dubuque, Iowa), allowed to equilibrate for 24 h, and 20 ml of extractant was gravity

filtered and then frozen in 20 ml scintillation vials (Keeney and Nelson 1982).








The samples were analyzed for NH4-N and N03-N content by the Analytical

Research Laboratory of the University of Florida (Gainesville, Florida) using an Alpkem

Flow Solution IV semi-automated spectrophotometer according to EPA methods 350.1

(for NH4-N) and 353.2 (for NO3-N). Values were expressed in mg kg' of dry soil.

Assuming no N losses to plant uptake, leaching or volatilization, monthly net

ammonification, nitrification and mineralization rates were calculated as follows (Hart et

al. 1994, Reynolds et al. 2000):

Net ammonification = ((NH4+-N),+I (NH4+-N)); (2-1)

Net nitrification = ((N03-N)t,+ (NO3-N),); (2-2)

Net mineralization = ((NH4+-N + NO3 -N),+ (NH4+-N + NO3--N),); (2-3)

where

t = initial time of sampling, and

t+1 = final time of sampling.

In addition, cotton was harvested at the end of each growing season to determine

lint yield per row. For both years, two 6.1 m sections were hand harvested from each

row, and lint yield was reported in kg ha-'. Yield of nearby control plots (cotton

monoculture) was also determined by hand harvest. Monthly precipitation, air

temperatures and soil temperatures (at 10 cm depth) for the study period were calculated

from weather data kept on record at the Jay, Florida research station.

Data Analyses

Statistical analyses were performed using SAS 8.2 for Windows (SAS Institute,

Cary, North Carolina) using ANOVA within the framework of a split block design. The

Shapiro-Wilk's test, in combination with ocular inspection of histogram plots, was used








to test all data for normal distribution. Logarithmic [log (x + 1)] transformation was

conducted when such transformation improved normality, and as a result, cotton yield

data were transformed prior to analysis. Differences between means were determined

using the Least Squares Means procedure. Treatment effects were considered significant

at a =0.10.

Results

Soil Water Content

Gravimetric soil water content in the alley and tree row locations was similar

throughout the study period (Figure 2-2). The greatest decline in soil water content was

observed for September-October 2001, during which soil water content decreased from

17.34% to 2.94% in the alley location, and 17.27% to 3.69% in the tree row location.

Soil water content for 2002 was highest in June (12.74% and 9.96% for alley and tree

row locations, respectively) and lowest in September (6.78% and 5.46% for alley and tree

locations, respectively) of that year.

With regard to treatment, barrier and non-barrier treatments were statistically

similar (p=0.6435) in soil water content; however, a seasonal effect was observed for

both treatments (p=0.0293) (Figure 2-2). The non-barrier treatment in particular

exhibited a strong seasonal change in water content from growing season one to the

dormant season (p=0.0135), and from the dormant season to growing season two

(p=0.0160). The water content of barrier treatment tended to be higher in warm summer

months. While not significant for water content, the effect of the barrier treatment was

more pronounced for nitrogen mineralization and cotton yield parameters (discussed

below).








Initial Soil Ammonium and Nitrate

Initial soil ammonium varied by growing season (p=0.0045) but not by treatment

(p=0.2393) (Figure 2-3). Values for initial soil ammonium generally declined during the

study period. Initial soil nitrate exhibited a peak in July 2001 antecedent to fertilizer

application, and showed a similar trend in summer 2002. Nitrate levels differed among

growing seasons (p<0.0001) but not between treatments (p=0.7655).

Ammonification

Monthly rates of ammonification did not differ among treatments (p=0.7776) or

between the alley and tree row locations from July 2001 to September 2002 (Figure 2-4).

Overall, net ammonification was observed mostly during the latter part of the first

growing season (October through December) for both locations and for the barrier and

non-barrier treatments. Net immobilization of NH4-N appears to have occurred during

most of the active growing season and during the winter and spring months. For

example, the average monthly immobilization for the first growing season (July-

November 2001) was 0.10 mg kg"', and the cumulative immobilization for this same

period was 0.41 mg kg"' in the barrier and 0.89 mg kg- 'in the non-barrier treatment,

respectively. Cumulative seasonal rates are presented in Table 2-2. Cumulative

immobilization of NH4-N exhibited no differences among the non-barrier and barrier

treatments and for the seasons.

Regression analysis indicated that ammonification was positively correlated with

time (r2=0.5110), and negatively correlated with initial water content of soil (2=0.2661).

No significant relationship existed between ammonification and initial ammonium

concentration (r2=0.1234) in the soil.








Nitrification

Seasonal differences in nitrification were significant (p=0.0019). Lower rates

were observed during October through February when average temperature was the

lowest (Figure 2-4). Average monthly nitrification rate was 23.1 mg kg-t for the first

growing season compared with 19.84 mg kgl for the second growing season, with no

differences among them (Table 2-2). Nitrification exhibited a significant difference

between the barrier treatments during growing season one, with non-barrier treatment

having a 32% higher cumulative rate than barrier treatment (p-0.1007). Nitrification

rates exhibited significant (p=0.0002) reduction during the dormant season for non-

barrier treatment compared with barrier treatment. However, no difference among

treatments was observed during the second growing season. Further, nitrification rates

did not differ between alley and tree row locations.

Nitrification was weakly but significantly correlated with initial water content

(r2=0.3478) and initial nitrate concentration (r2=0.1635). However, nitrification was

negatively correlated with time (r2=0.2689).

Mineralization

Temporal variation in mineralization followed a pattern similar to that of

nitrification, with lower rates during October through February for both the alley and tree

row locations (Figure 2-4). Average monthly mineralization during growing season one

was 19.78 mg kg-' for barrier treatment and 26.05 mg kg"' for non-barrier treatment, with

a significant difference (p= 0.0750) between them (Table 2-2). No difference in

mineralization was observed for growing season two. However, the non-barrier

treatment had 20% lower (p=0.0156) mineralization rate during the dormant season.








Mineralization rates showed a positive linear relationship with soil temperature

and soil water for both the alley and tree row locations (Figures 2-6 and 2-7). The

relationship was stronger for soil water than for soil temperature. Mineralization

increased substantially with increasing soil water in the barrier treatment (r2=0.8144;

p=0.0004), the non-barrier treatment (r2=0.8215; p=0.0003), and the tree row location

(r2=0.5951; p=0.0090).

Regression analysis revealed that higher levels of initial soil NO3-N resulted in

increased rates of mineralization for the barrier (p=0.0636), but not for the non-barrier

treatment (Figure 2-5). The relationship was also significant (p=0.0225) for the tree row.

No such relationship between mineralization and initial NH4-N was observed for any of

the locations or treatments.

Fallow Effect and Mineralization

A comparison of N mineralization rates for the months of July, August and

September of 2001 and 2002 was carried out to assess carry-over effects of the pre-2001

fallow period on mineralization. For the three-month observation period, mineralization

averaged 33.17 mg kgl' month-' in the first year and 13.21 mg kg1' monthly' in the second

year, representing a significant (p=0.0001) decrease. This second-year decrease in

mineralization was observed in both the barrier (63% decrease) and non-barrier (57%

decrease) treatment areas.

Initial Nitrogen Source and Nitrogen Transformations

The rates ofN ammonification, nitrification and mineralization in the barrier and

non-barrier treatments did not vary in response to the source of nitrogen (Figure 2-8).

Overall, we observed a net ammonification (0.2 mg kg -1; pooled for the root trenching








treatments) in soils under inorganic N and a net immobilization under poultry litter for

the period from June through September. Mean nitrification rates were 71.8 mg kg'l for

the inorganic fertilizer and 79.2 mg kg"' for the poultry litter. Mineralization rates were

also similar, with 72.3 and 77.8 mg kg-' for the inorganic fertilizer and poultry litter,

respectively.

Cotton Lint Yield

Lint yield in 2001 was significantly (p=0.0034) higher in the barrier (712.36 kg

hal') compared to the non-barrier (524.03 kg ha-) treatment, but was similar to the

monoculture (645.56 kg hal) (Figure 2-9). For 2002, cotton in barrier averaged 447.88

kg ha', while that in the non-barrier treatment averaged only 180.40 kg ha',with

significant difference (p<0.0001) among them. The monoculture yield (751.26 kg ha')

was substantially higher than yield from the barrier and non-barrier treatments as well.

Significant differences were also observed between years, with yield in barrier being

37.13% lower in the second year, and yield in non-barrier being 65.58% lower in the

second year, compared to respective 2001 yields.

Although source of nitrogen had no effect (p=0.1197) on lint yield in the non-

barrier treatment, yield was significantly affected (p=0.0111) in the barrier treatment

(Figure 2-10), resulting in an interaction between treatment and source of nitrogen. Lint

yield was 39% higher in the inorganic fertilizer plots (521.22 kg ha") compared to the

poultry litter plots (374.53 kg ha ') for the barrier treatment.

Discussion

During 15 months of sampling, we found significant seasonal variations in the

rates of mineralization, nitrification and ammonification. In both alley and tree row








areas, net mineralization, indicative of plant-available N, peaked during the early growing

season, and decreased gradually towards the end of the year, although fluxes were not

consistent between seasons. For 2001, net mineralization peaked at 42.44 mg kg-1 in July

in the non-barrier treatment while the peak for barrier (31.90 mg kg'') occurred two

months later. Apparently, non-barrier treatment responded to the N fertilization with an

increased rate ofN turnover. Following this was a significant drop in mineralization and

nitrification rates, perhaps due to lower rainfall and lower soil temperatures, which may

have begun to immobilize nitrogen in the soil. For 2002, net mineralization reached

29.37 mg kg"' in May in non-barrier and 25.95 mg kgl' in June in barrier. Mineralization

and nitrification rates observed in our study are within the range reported for both

agricultural (Deng and Tabatabai 2000) and forest soils (P6rez et al. 1998).

Overall, rates of mineralization, nitrification and ammonification appeared to

follow seasonal patterns of temperature and rainfall. In most regions, the highest rates of

N mineralization usually occur in spring and summer (March to July) and rapidly decline

in fall and winter (Bielek 1998). For Florida, which has a longer growing season, we

observed this decline in September or October. The seasonal variations in N

mineralization rates were correlated to soil temperature and soil moisture, as reported

previously (Foster 1989, Goncalves and Carlyle 1994).

In addition to seasonal patterns in mineralization, we observed higher

mineralization during the first growing season (July-September) compared to the second

growing season. One possibility is that the 29-year-long fallow period prior to the

initiation of this trial, and the addition of fertilizer N, may have heavily stimulated the

mineralization process, a process referred to as the "priming effect" (Bielek 1998, Lovell








and Hatch 1998). For example, Bielek (1998) observed that rapid acceleration of

mineralization occurred in more fertile soils than less fertile soils. Our study site has a

nutrient-rich topsoil due to the long-term deposition of pecan leaf biomass, and thus

major additions of inorganic nitrogen to this system could be expected to exhibit a

lowering of C:N ratio of soil (Maithani et al. 1998), which in turn, could accelerate soil N

transformations. Although we supplied inorganic or organic N again during the second

season, the overall C:N ratio may have increased due to build up of organic matter from

cotton residue in the system, or to rapid microbial assimilation ofN (Priha and

Smolander 1999). Data from a companion study support this hypothesis. The C:N ratio

increased from 10.5 during the first growing season to 12.3 during the second growing

season in the same study plots (Lee and Jose, unpublished data). These authors reported

an increase in soil carbon and a decrease in N during the second growing season

compared with the first growing season. It appears that soil fertility during the second

growing season decreased considerably compared to the first. A similar effect has been

reported in studies of tree fallow systems as well (Kwesiga and Coe 1994, Maroko et al.

1998, Ikerra et al. 2001).

Contrary to our expectations, the non-barrier treatment had significantly higher

nitrification and mineralization rates than the barrier treatment during the first growing

season. We postulated that roots of non-barrier trees were acquiring N from the fertilized

alley, resulting in N depletion in the soil. We further hypothesized that this would

eventually lead to reduced rate of N mineralization in the non-barrier treatment compared

to the barrier treatment. In a related study we have shown that N uptake by cotton in the

non-barrier treatment was significantly lower than that in the barrier treatment (Chapter








4). It is possible that lower N demand by cotton plants as a result of competition for

water (Wanvestraut et al. 2003) in the non-barrier treatment has resulted in higher NH4-N

in the soil, leading to greater nitrification and mineralization.

One factor that may affect N transformations in the system is pecan leaf litter.

Kaur et al. (2000) observed that tree mulch build-up can result in higher N availability

under agroforestry systems. Decomposition of pecan tree biomass, which would be

expected to follow an exponential decay curve, could affect N levels in the system, by

supplying varying amounts of nitrogen at varying times during the year, thus altering

litter quality and microbial activity on the orchard floor.

Another consideration is the possibility of chemical inhibition of nitrification by

allelopathic compounds in leaf biomass (Clein and Schimel 1995). While this is

conceivably possible given the long period of pecan leaf and root biomass accumulation

prior to establishment of the system, the effect, if any, would be expected to be uniform

throughout the system. While researchers have looked at this possibility in black walnut,

their results indicated no inhibitory effect of allelochemicals on nitrification and

mineralization in the field (Thevathasan et al. 1999).

Although source of nitrogen did not have any effect on net ammonification,

nitrification and mineralization in our study, lint yield was affected in the barrier

treatment. Cotton plants responded better to inorganic fertilizer than poultry litter when

belowground root competition between pecan and cotton was avoided. Others have

reported similar results for cotton grown with poultry litter and inorganic fertilizer

(Mitchell et al. 1992). Since similar mineralization rates were observed for both the

sources (Figure 2-8), it is unclear why plants responded better to inorganic fertilizer than








poultry litter. As poultry litter is an organic product, the timeframe in which plant

available N becomes available may differ from inorganic fertilizer. A finer scale of

observation might serve to quantify more clearly the differences in N transformations

between the inorganic fertilizer and poultry litter.

Lint yield showed significant decrease (37.1% for barrier and 65.6% for non-

barrier) in 2002 compared to 2001 (Figure 2-9). This may be partially explained by

differences in N mineralization observed during the three months of active vegetative and

reproductive growth of cotton. A direct comparison of N mineralization during this

period (July through September) for the two growing seasons revealed significantly lower

(60%) mineralization during the second year compared to the first year. Competition for

water also appears to be a likely factor for yield reductions, since yield of non-barrier

cotton plants was depressed during both years (Wanvestraut et al. 2003).

Conclusions

Temporal variations in net ammonification, nitrification and mineralization were

driven primarily by environmental factors (e.g., soil moisture content and soil

temperature), and by initial ammonium and nitrate levels. However, these and other

factors appear to have exerted a combined influence on N transformations over the study

period. Competitive interactions for resources such as water and nitrogen resulted in a

decreased ability for nitrogen uptake in plants in the non-barrier treatment compared to

those in the barrier treatment. This, in turn, may have resulted in a higher build-up of soil

N in the non-barrier treatment. Effects of the pre-trial fallow period appear to have

diminished by the second growing season.





22


Table 2-1. Surface soil chemical characteristics of Rhodic Paleudult sandy loam soil of
the Jay, FL agroforestry site.
Soil property Value
Soil pH 6.0
Organic matter (%) 3.1
CEC (cmol kg-l) 8.0
Nitrogen concentration (%) 0.1
Phosphorous* (mg kg"1) 46.0
Potassium* (mg kg") 86.0
Calcium* (mg kg') 636.0
Magnesium* (mg kg"l) 168.0
Sulfur* (mg kg l) 16.0
*water soluble form









Table 2-2. Seasonal and cumulative rates ofN ammonification, nitrification and mineralization in barrier and non-barrier treatments.
Observation period Ammonification (mg kg"') Nitrification (mg kg"') Mineralization (mg kg"')
Barrier No barrier Barrier No barrier Barrier No barrier


2001 Growing Season -0.41 a'A2 -0.89 aA 99.29 aA 131.16aA 98.88 aA 130.27
(July-November) (1.88)3 (1.57) (6.26) (9.09) (5.21) (7.72)
2001/2002 Dormant Season -2.02 aA -2.87 aA 76.07 aA 62.10 aB 74.04 aA 59.23 a
(December-April) (1.03) (1.17) (2.97) (2.06) (2.45) (1.19)
2002 Growing Season -2.09 aA -2.76 aA 95.38 aA 103.09 aA 93.29 aA 100.33
(May-September) (0.52) (0.89) (3.03) (3.13) (2.99) (2.50)
Whole Study Period -4.52 -6.52 270.74 296.35 266.21 289.83
(July 2001-September 2002) (0.68) (0.67) (2.40) (3.39) (2.08) (2.97)
SWithin-season values followed by the same lowercase letter are not significantly different at the 0.10 level of probability.
2 Across-season values followed by the same uppercase letter are not significantly different at the 0.10 level of probability.
3 Standard errors of the mean are given in parentheses.


bA

B

aA








Crop alley
Tree row 1 2 3 4 5 6 7 8

.o. .e ... .

1 0.50 m *
to tr : :

1.50 m







to tr ee




Cotton rows Reference pecan tree
0.76 m apart
A In-situ polyethylene
bags for incubation
study
8.40 m
to tiee


Figure 2-1. Plot layout showing location of in-situ soil incubation bags at the Jay, Florida
agroforestry site.










35
N Air temperature (OC)
30 Soil temperature (*C)

V 25

20





400
350






- 300
E












E 250

i 200
S10


5














Barrier
400

350

300
















20 -0- No barrier
250










0


o -----------------
0

0

25














J A S O N D J F M A M J J AS O
-0 Barrier







2001 2002- er




Month
0
J ASONDJ FMAMJJ AS0
2001 2002
Month


Figure 2-2.Monthly averages for air temperature, soil temperature (at 10 cm depth),
precipitation, and soil water content at the Jay, Florida agroforestry site.









10.00
-*- Barrier
8.00 --0-- No barrier

E 6.00
z
I 4.00
z

2.00

000

30.00

S25.00

1 20.00

S16.00

z 1000

5.00
o .oo --- -^y --
0.00
JASON DJ F M A M J J AS
2001 2002
Month

Figure 2-3.Monthly averages of initial soil NH4-N and N03-N levels in alleyways of
barrier and non-barrier treatments.








8 8
8 I--------------- |8 | -----------

6No Barrier
SAlley NBarrier 6 Tree row

0 4 o 4






E E
60 60 -
-2 -2

E -4 E -4


-8 -8


Alley Barrier Tree row
S50s No Barrier C 50

0 40 40

o 30 a 30

20 20

10 10

0110d 0
60 60
Alley EBarrier Tree row
E50 *No Barrier
*x I




20 0
1 40 E L
.30 2 30






JASONDiJFMAMJJAS JASONDIJFMAMJJAS
2001 2002 2001 2002
Month Month

Figure 2-4. Monthly rates ofN ammonification, nitrification and mineralization in alley
and tree row locations.









60


50

01 40


S30


. 20
20





0

60
60


All Barrier
Alley o No barrier



oo



.* .
o
0 0


Treo :o w





Tree row


50 50 *


S40 40


S30 *3 30
S* IS *
S y =2 0860+12 410
20 2 20 R=03402

I *I *
10 10


0 0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 80 0.0 5.0 10.0 15.0 20.0 25.0 30.0

Initial NH4-N (mg kg-1) Initial NO3-N (mg kg-1)


Figure 2-5. Monthly N mineralization rates in alley and tree row locations as influenced
by initial levels of NH4-N and N03-N.


Alley Barrier
Alley o No barrier
-Linear (Bamer)
---- Linear (No barrier)
o

oo
y = 07959x + 14 1121 .-'
S R2= 0 2402,--
0



0 y 03695x + 15 0636
S R 0 1423
* *









60


50


a 40
E
E
0
S30


* 20

i
10


0

60


50

6,
0 40





| 20
E
Iu


0 5 10 15 20 25 30 35

Soil temperature (OC)

Figure 2-6. Monthly N mineralization rates in alley and tree row locations as influenced
by soil temperature (10 cm depth).


Barrier Alley
o No barrier
- Linear (Barrier)
-------- Linear (No barrier) O

00

0
y =1 1177x-59254
R= 03871 -

o **
O y 06741x+25214
.0 oR'2= 02869






Tree row


*

0


y = 0 9159x + 05586
R2 02643


*








60
Barrier Alley
o No Barrier
50 --- Linear (Barrier)
-------- Linear (No Barrier) O
E 40

30 0
a y 20716x+31099 S-
W 0.2r








6o
S20 0 y .3x5 772036 07467






Tree row
S50

E 40





10

a0
~ 0 y ----------------------


0 5 10 15 20
Soil water (%)


Figure 2-7. Monthly N mineralization rates in alley and tree row locations as influenced
by soil water percentage during the 2001 and 2002 growing seasons.






31


4.00
S 3.00 g Barrier
3.00
S 2.00 *Nobarrier
t 2,00
E

1 00
0.00

-1.00
S -2.00
E -3.00

-400

100.00

80.00

S6000




20.00
0.00

12000
Z





9 80.00
0.00




40.00

80.00
1 40.00

S 20.00

0.00
Inorganic Poultry litter

Fertilizer type

Figure 2-8. Seasonal rates ofN ammonification, nitrification and mineralization in
alleyways of barrier and non-barrier treatments as influenced by application
of inorganic fertilizer vs. poultry litter during the 2002 growing season.






32

1400
2001 Barrier
1200 No barrier
o Monoculture
S1000

8 s00

.2 600

S400
_j
200

0
1400
2002
1200

;1000

- 800

600

E 400
-J
200


1 4 8 9 13 16 M
Row


Figure 2-9. Lint cotton yield in barrier, non-barrier and monoculture treatments during
the 2001 and 2002 growing seasons.









t Barrier
f No barrier
I T


Inorganic Poultry litter
Fertilizer type


Figure 2-10. Lint cotton yield in barrier and non-barrier treatments as influenced by
application of inorganic fertilizer vs. poultry litter during the 2002
growing season.













CHAPTER 3
GROUNDWATER NITRATE DYNAMICS IN A PECAN (Carya illinoensis K. Koch)-
COTTON (Gossypium hirsutum L.) ALLEY CROPPING SYSTEM IN THE
SOUTHERN UNITED STATES

Introduction

The fate of nitrogen (N) fertilizers and their effect on groundwater quality is of

direct importance to farmers, researchers, environmentalists and the general public.

Intensive agricultural practices have led to inefficient use of applied nitrogen and to

contamination of surface and subsurface drainage water through nitrate leaching (Bonilla

et al. 1999, Ng et al. 2000). On a national scale, over-application of N increases the

production costs of farmers by millions of dollars each year (USDA 1998a). Moreover,

the negative effects of nitrate leaching on rivers, lakes and rural residential wells are of

increasing public and scientific concern (USDA 1998b, Bonilla et al. 1999, Ng et al.

2000). Agroforestry has been identified as a land use practice with potential for

alleviating some of these problems.

A major form of agroforestry in the USA is alley cropping, which involves the

planting of row crops or pasture in alleys formed by single or multiple rows of trees or

shrubs (Garrett and Buck 1997). Trees in agroforestry systems are capable of recycling

soil nutrients that leach down the crop rooting zone, thereby reducing ground water

contamination and increasing nutrient use efficiency in the system (Rowe et al. 1999).

Nitrogen is a limiting soil nutrient in temperate alley cropping systems. It is lost

via harvests of crop biomass, litterfall by trees, and by volatilization and leaching

processes. In addition, plants of the same species and growth stage compete heavily for








nitrate (NO3-N) and ammonium (NH4-N) due to the high mobility of water and nitrate

ions along the root surface, resulting in depletion zones in the soil that overlap with

neighboring plants. Nitrogen fertilizer supplements are therefore usually needed in order

to maintain favorable plant growth.

Similarly, trees and crops possess a high potential for interspecific competition for

nitrate in the topsoil, depending on factors such as rooting depth, water availability and

tree species phenology (Jose et al. 2000a). Many trees have the bulk of their fine, feeder

roots in the top 30-cm deep topsoil, placing them in a zone of competition with crop

roots. At lower depths, tree roots can exploit subsoil nitrate and other nutrients beyond

the rooting depths of most crops, placing trees at a competitive advantage for acquiring

leached nitrogen. In other cases, interspecific competition may not occur if the two

species utilize different sets of soil resource horizons or their temporal demands for

nitrogen differ. Water availability also affects nutrient competition, directly by

decreasing mass flow and indirectly by reducing crop growth rate. Competition, in turn,

is mitigated in part by the use of chemical or organic N amendments.

Interspecific competition for nutrients, while present, is generally of minor

importance to system productivity in the temperate region. However, certain factors can

lead to increased competition for nutrients. For example, Jose et al. (2000b) observed

that competition for fertilizer nitrogen was minimal in a black walnut-maize alley crop,

since nutrient acquisition was not simultaneous among system components. However,

water availability was observed to be a factor in nutrient competition, as competition for

water by tree roots was responsible for reduction in biomass in intercropped corn,

resulting in decreased efficiency of fertilizer use (Jose et al. 2000b). Similarly, in a








poplar-barley system in southern Ontario, associated trees and crops were found to utilize

different sets of soil nutrient resource horizons (Williams et al. 1997). However,

competition for available nutrients cannot be avoided when N is limited and fertilizer is

not supplied (Chirwa et al. 1994, de Montard et al. 1999, Imo and Timmer 2000).

Intercropped trees can benefit when N is applied to nearby crops, as the nutrient

may find its way to nearby tree roots. This form of nutrient capture could be an

important factor in the growth of pecan trees, for example, which are often below their

potential yielding capacity due to nitrogen deficiency (Arnold and Crocker 1999).

The "safety net" hypothesis of nutrient capture holds that the deep roots of trees

are capable of retrieving nitrate and other nutrients that have leached below the rooting

zone of associated agronomic crops, and of eventually recycling these nutrients as

litterfall and rootfall in the cropping zone. A study in Sweden, for example, focused on

the effects of tree harvesting on natural N levels in soil (Browaldh 1996). The study

found that natural N levels increased in the vicinity of harvested sites, due to the lack of

N uptake by tree roots of harvested trees. In other studies, isotopic tracers such as 15N-

enriched fertilizer have been used to trace movement of applied N in alley cropping

systems (e.g., Jose et al. 2000b). It is unclear, however, to what extent the safety net role

would help in alleviating N leaching in temperate alley cropping systems. Understanding

the dynamics of ground water nitrogen with and without tree-crop root interaction would

be the first step in making alley cropping an environmentally appealing and ecologically

viable land use option for landowners. Hence, the objective of this study was to

determine the role of tree roots on groundwater ammonium and nitrate leaching in a

pecan-cotton alley cropping system.








Materials and Methods

Study Area and Configuration

This study was conducted at the West Florida Research and Education Center

Farm of the University of Florida, located near Jay in northwestern Florida, USA (30089'

N Lat., 87o13' W Long.). The climate is temperate with moderate winters and hot,

humid summers. The soil at the site is classified as a Red Bay sandy loam, which is a

fine-loamy, siliceous, thermic Rhodic Paleudult. The average water table depth is 1.8 m.

A mature pecan-cotton alley cropping system was established in Spring 2001

from an existing orchard of pecan trees that had been planted at a uniform spacing of

18.28 m in 1954 and that had remained under grass cover for 29 years until the initiation

of the current study. For this study, 10 plots were demarcated within the orchard and

arranged into 5 blocks using a randomized complete block design (this number was

reduced to 6 plots in June 2002 due to the establishment of another study in the orchard

(see Chapter 2)). Each plot, which consisted of two rows of trees oriented in a north-

south direction, was 27.43 m long and 18.28 m wide, with a practical cultivatable width

of 16.24 m, and was separated from its adjacent plot by a buffer zone of the same

dimensions.

To assess tree root competition, each block was randomly divided into a barrier

plot and a non-barrier plot. Barrier plots were subjected to a root pruning treatment in

March of 2001 in which a trenching machine was used to dig a 0.2 m wide x 0.9 m deep

trench along both sides of the plot at a distance of 1.5 m from the trees to separate root

systems of pecan and cotton. A double layer of 0.15 mm-thick polyethylene sheeting was

used to line the ditch prior to mechanical backfilling. The barrier plots thus served as the








tree root exclusion treatment, while the non-barrier plots, which did not receive this

treatment, served as the tree-root competition treatment.

Cotton (DP 458 B/RR) was planted in rows 0.91 m apart at 16 rows per alley in a

north-south orientation on 16 May 2001 and 13 May 2002, after disking of the alleys. A

3-9-18 fertilizer blend was applied at a rate of 89.6 kg N ha' on 19 June 2001 and 5 June

2002. Conventional insecticide and herbicide were applied during the growing season as

recommended.

Sampling Methods

Soil solution (free soil water that is not in equilibrium with the soil matrix)

(Weston and Attiwill 1996) was sampled 1-2 times monthly over a 15-month period from

ceramic cup lysimeters installed in pairs at depths of 0.3 and 0.9 m at specific distances

of 1.5, 4.2 and 8.4 m from a reference tree in each plot (Figure 3-1). Lysimeters were

fitted with a highly porous (-45% porosity) ceramic cup (Soil Moisture Equipment Corp.,

Santa Barbara, California) that allowed for collection of soil solution 24-48 hr after

application of a vacuum at 30-50 kPa (Talsma et al. 1979). The samples were collected

in 20 ml scintillation vials and kept frozen until analysis. Samples were analyzed for

NH4-N and N03-N concentrations by the Analytical Research Laboratory of the

University of Florida (Gainesville, Florida) using an Alpkem Flow Solution IV semi-

automated spectrophotometer according to EPA methods 350.1 (for NH4-N) and 353.2

(for N03-N). Values were expressed in mg L1.

Data from each month were averaged across rows to produce one composite

observation per plot at each depth (van Miegroet et al. 1994). This was necessary because

missing data (resulting from dry soil conditions or damaged lysimeters) precluded whole-








study analysis for a row effect. A Hydrosense (Decagon Devices, Pullman, Washington)

reflectometry soil moisture probe was used on a monthly basis within a 12 cm surface

layer to monitor volumetric water content in all test plots.

Water Drainage

A process-based deterministic model, LEACHMN, the nitrogen component of

LEACHM (Leaching Estimation and Chemistry Model; Hutson and Wagenet 1992) was

used to calculate subsurface water drainage in the pecan-cotton system. LEACHMN is a

simulation model that uses data inputs for weather, crops, and soil profile physical and

chemical properties, in determining the movement of water and nitrogen in an

agricultural system (Sogbedji et al. 2000, Ng et al. 2001). LEACHMN parameter input

values are listed in the Appendix. The model uses the Richard's equation:

a / at = (a / az) [K( )((H / az)] U(z,t) (3-1)

where

0= volumetric water content (m3 m3);

t = time (days);

z = depth (mm);

K = hydraulic conductivity (mm day");

H = hydraulic head (mm); and

U= water lost per unit time by transpiration (day-').

Leaching of Ammonium and Nitrate

The respective amounts of NH4-N and N03-N that leached below the root zone at

0.3 and 0.9 m depths in barrier and non-barrier treatments was determined using the

following equation (Moreno et al. 1996):








LN = CD (3-2)

where

LN = amount of NH4-N or N03-N leached (in kg ha');

C = concentration of NH4-N or N03-N (in mg L"') in the sampled soil solution; and

D = respective water drainage (in mm month"') at the sampled site (as estimated using the

LEACHMN model).

Net Retention Index

Net retention index (NRI), a measure of the net effect of tree root uptake on soil

NH4-N and N03-N levels in a tree-crop system, was derived from the following equation:

NRI = (NB NNB) / N (3-3)

where

NRI = retention of NH4-N or N03-N (no units);

NB = amount of NH4-N or N03-N leached (in kg ha') in barrier treatment; and

NNB = amount of NH4-N or NO3-N leached (in kg ha"') in non-barrier treatment.

For NRI, a value of 0 indicates complete leaching has occurred (i.e., no safety net effect),

whereas a value of 1 indicates no leaching has occurred (i.e., complete safety net effect),

with values in-between indicating the relative effectiveness of tree roots in intercepting

NH4-N or N03-N, all other factors being equal.

Data Analyses

Statistical analyses were performed using SAS 8.2 for Windows (SAS Institute,

Cary, North Carolina) using ANOVA within the framework of a split block design. The

Shapiro-Wilk's test, in combination with ocular inspection of frequency distributions,

was used to test for normality. All data were transformed logarithmically [log (x + 1)]








prior to analysis. Differences between means were determined using the Least Squares

Means procedure. Treatment effects were considered significant at a = 0.10.

The data set for the 15-month study was analyzed as a whole and in three sub-sets

based on season. These seasons, which are referred to in the study as growing season

one, dormant season, and growing season two, fell within the following months

respectively: June-November 2001, December 2001-April 2002, and May-August 2002.

Results and Discussion

Drainage

Monthly and cumulative field drainage data as estimated using LEACHMN are

shown in Figure 3-2. Average monthly drainage at the 0.3 m depth was 48 mm for

barrier treatment and 41 mm for non-barrier treatment. Average drainage at the 0.9 m

depth was 32 mm and 21 mm, respectively, for the barrier and non-barrier treatments.

Values for cumulative drainage, an expression of the sum effect of monthly drainage,

were 716 mm for barrier and 614 mm for non-barrier treatments, respectively, at the 0.3

m depth. Cumulative drainage at the 0.9 m depth was 476 mm and 322 mm, respectively,

for barrier and non-barrier treatments. A higher drainage in the barrier treatment would

be expected given the lack of water uptake by tree roots in that treatment (Jose et al.

2000a, Wanvestraut et al. 2003).

Ammonium Concentrations

In general, concentrations of NH4-N in soil solution were very low and often close

to the limit of detection (Table 3-1; Figure 3-3), a result similar to other lysimetric studies

(Frazer et al. 1990, van Miegroet et al. 1994). The scarcity of NH4-N in soil solution is

not surprising given that NH4 is an exchangeable ion that is easily adsorbed to clay








particles, and that under favorable conditions NH4-N can be quickly oxidized to NO3-N

by nitrifying soil bacteria (Breitenbeck and Boquet 1992).

When examined over the duration of the 15-month study period (i.e., eliminating

the growing season variable from the analysis), average NH4-N concentrations were not

significant for any factor (i.e., treatment, depth, growing season, block) or combination of

factors (Table 3-1). Mean concentrations ranged from 0.03 mg L-' in non-barrier (0.3 m

depth) to 0.05 mg L"' in non-barrier (0.9 m depth).

Similar results were found when data were analyzed according to season (Table 3-

1). Average NH4-N concentrations were significant for changes in growing season

(p=0.0006) and block (p=0.0732) only. Overall NH4-N concentrations for the first

growing season (2001) were similar to those in the dormant season, with the exception of

the 0.9 m depth in the non-barrier treatment, which increased during the dormant season

and remained higher during the second growing season (2002). Average NH4-N

concentrations for both depths in the barrier treatment also increased in growing season

two compared to levels in previous observation periods.

Nitrate Concentrations

The inorganic N in solution was dominated by N03-N, a result similar to other

studies (van Miegroet et al. 1994). Unlike NH4-N, the presence of significant amounts of

N03-N in soil solution is expected given that N03-N is mobile in soils and can be leached

from surface to underlying soil layers by percolating rainwater (Breitenbeck 1990).

Nitrate concentrations in the soil solution at 0.3 and 0.9 m depths showed strong temporal

variations based on season (Table 3-1; Figure 3-3).








When data were analyzed as a whole over the whole study period, NO3-N

concentrations were significant for depth (p=0.0003) only (Table 3-1). Mean NO3-N

concentrations for barrier treatment at 0.3 and 0.9 m depths were 14.36 mg LU and 10.82

mg L1, respectively, representing a 24.7% decrease with depth. Mean concentrations for

non-barrier at corresponding depths were 10.16 mg L-' and 7.15 mg L-', respectively,

representing a 29.6% decrease.

When examined by season, N03-N concentration was also significant for depth

(p=0.0025) (Table 3-1). For growing season one, N03-N concentration decreased with

depth, exhibiting a 54.5% decrease in the barrier treatment and a 44.7% decrease in the

non-barrier treatment, respectively. A similar trend for depth was seen in growing season

two, where N03-N concentrations decreased by 35.1% in barrier and 29.6% in non-

barrier treatments, respectively. Dormant season trends showed a 91.3% increase with

depth in the barrier treatment, but no increase or decrease with depth in the non-barrier

treatment.

Seasonal NO3-N concentration was also significant for growing season (p<.0001)

and depth x growing season interaction (p<.0001) (Table 3-1). In barrier treatment, the

average N03-N concentrations at 0.3 m depth remained the same for both growing

seasons, but were lower in the dormant season. Non-barrier N03-N concentrations at 0.3

m depth differed for all three seasons, with growing season two being the highest (20.97

mg L') and dormant season being the lowest (3.06 mg L '). For barrier treatment at 0.9

m depth, concentrations for growing season two were higher than those for growing

season one. Nitrate concentrations did not differ by season at 0.9 m depth in the non-

barrier treatment.








Ammonium Leached

When examined over the whole study period, the average monthly rate of NH4-N

leached was not significant for any factor or combination of factors (Table 3-2). This

result is expected given the similar findings for overall NH4-N concentrations in the

previous section. The average rates of NH4-N leached were in the range of 0.01-0.02 kg

ha-' month'.

On a seasonal basis, statistical analysis showed that the average NH4-N leaching

rate was also not significant for any factor or combination of factors (Table 3-2). A

comparison of treatment means at the 0.9 m depth in growing season one indicated that

the average NH4-N leaching rate for barrier treatment was significantly (p=0.0794) higher

than that for non-barrier. However, this treatment difference (0.03 kg ha"1 month"') is on

too minute a scale for drawing strong conclusions as to significance.

Seasonal and cumulative totals of NH4-N leached are shown in Table 3-3.

Analysis of the whole study period showed that total NH4-N leached was not significant

for any factor or combination of factors. The total NH4-N leached for the whole study

period averaged 0.23 kg ha'l in barrier and 0.14 kg ha in non-barrier treatments,

respectively, which did not represent a significant difference.

Examined by season, total NH4-N leached was significant for growing season

(p=0.0032), and there was a treatment x growing season interaction (p=0.0390) (Table 3-

3). A comparison of treatment means at the 0.9 m depth in growing season one indicated

that total NH4-N leached in the barrier treatment was significantly (p=0.0197) higher than

that in the non-barrier. In addition, total NH4-N leaching levels in the barrier treatment

decreased significantly in both depths after the first growing season.








Nitrate Leached

Over the entire 15-month study period, the average rate of NO3-N leaching was

significant for treatment (p=0.0470) and for depth (p<0.0002) (Table 3-2). A comparison

of treatment means at 0.9 m depth showed that the barrier treatment mean (3.40 kg ha-'

month-') was significantly (p=0.0345) higher than the non-barrier mean (1.28 kg ha-'

month-'). With regard to depth effects, for barrier treatment, the mean rate of NO3-N

leaching at 0.3 m depth (9.24 kg ha"' month-') was significantly (p=0.0237) higher than

the rate at 0.9 m. Similarly, for non-barrier treatment, the rate at 0.3 m depth (5.50 kg

ha"' month-') was significantly (p=0.0022) higher than the rate at the deeper depth.

Examination by season showed that the average monthly rate of N03-N leached

was significant for treatment (p=0.0494), depth (p<.0001), growing season (p<.0001),

and depth x growing season (p<.0001) (Table 3-2). With regard to specific treatment

effects, in growing season one, the average NO3-N leaching rate (3.67 kg ha-' month-) at

0.9 m in barrier treatment was significantly higher (p=0.0550) than the rate (1.27 kg ha"'

month- ) observed at the same depth in the non-barrier treatment. For the dormant

season, a treatment effect was observed for 0.9 m depth, where the average N03-N

leaching rate in barrier (3.95 kg ha' month') was higher (p-0.0722) than that for non-

barrier (1.55 kg ha' month"'). A treatment effect for root barrier was not observed in

growing season two.

Average NO3-N leaching was also significant for depth, as stated earlier.

Growing seasons one and two exhibited a depth effect for both treatments, with lower

rates of NO3-N leaching occurring at the lower depth. The dormant season did not exhibit

a depth effect within each treatment.








Across seasons, average N03-N leaching fluctuated in barrier treatment at 0.3 m

depth, with the greatest reduction in N03-N leaching occurring during the dormant season.

A reduction in N03-N leaching during the dormant season also occurred in the non-

barrier treatment at 0.3 m. For both treatments at 0.9 m depth, N03-N leaching rates did

not vary significantly when the two growing seasons were compared.

Seasonal and cumulative totals of N03-N leached are shown in Table 3-3. When

examined over the whole study period, total N03-N leached was significant for treatment

(p=0.0049) and depth (p-0.0010). Specifically, the cumulative amount ofNO3-N

leached during the entire study period at 0.3 m depth in the barrier treatment (121.94 kg

ha ') was higher than that leached in the non-barrier treatment (63.83 kg hal'),

representing a decrease of 47.7%. A similar trend occurred at 0.9 m depth, where

cumulative N03-N leaching in barrier (45.56 kg ha"') was higher than in non-barrier

(13.05 kg ha '), a 71.4% decrease. Cumulative N03-N leaching rates were also lower at

0.9 m depths compared to 0.3 m depths, in both treatments.

Analysis of total N03-N leached by season showed overall significant differences

for treatment (p=0.0051), depth (p<.0001), growing season (p<.0001), and depth x

growing season (0.0006) (Table 3-3). For each growing season, the cumulative amount

of N03-N leached at 0.9 m depth of barrier was greater than three times that for non-

barrier, indicating a significant amount of N03-N being taken up in the non-barrier

treatment at the lower depth. Treatments also varied by depth during both growing

seasons, with lower depth having a reduced N03-N leaching rate compared to the upper

depth. No significant interactions were observed within the dormant season for either

treatment or depth variables.








Various studies have attempted to estimate nitrate leaching rates from agricultural

systems. Moreno et al. (1996) conducted a study of water balance and nitrate leaching in

an irrigated maize crop on a sandy loam in Spain. They found that nitrate leaching

during the 3-year study period totaled 43 kg ha- under a low fertilization regime (170 kg

N ha' year ') and 150 kg ha"' under a high fertilization regime (500 kg N ha' year").

They observed that main periods of leaching occurred during periods of bare soil and

high rain. In another study, Van Miegroet et al. (1994) estimated that 20-40% of the

added fertilizer N leached below 60 cm soil depth in a short-rotation sycamore plantation

fertilized with 450 kg ha' of urea over a 3-year period. The results of our study indicated

that the barrier treatment had the potential to leach 43.87 kg N hal' year' down below 0.9

m depth. The non-barrier treatment, on the other hand, had a potential of leaching only

15.76 kg N ha' year' below 0.9 m. For both treatments N03-N accounted for 99.9% of

total inorganic nitrogen.

Net Retention Index of Ammonium and Nitrate

Net retention index (NRI) for NI-H-N at 0.3 m depth was similar for both growing

seasons and for the whole study period (Table 3-4). NRI for NH4-N was highest at 0.9 m

depth during growing season one. The NRI for NO3-N was higher at 0.9 m depths, with

mid-range values occurring during growing season one and during the whole study period.

The high NRI (0.81) for N03-N at 0.9 m depth during growing season two is indicative of

a significant retention ofNO3-N in the non-barrier treatment compared to the barrier

treatment during that season. These findings suggest the occurrence of a safety net

process on the part of pecan tree roots, particularly at 0.9 m depth.








Similar results demonstrating the safety net role of trees have been reported in the

literature. Browaldh (1995), for example, studied the influence of trees on N dynamics in

a poplar (Populus sp.)-based alley cropping system with oats and barley in Sweden. He

found that the presence of trees resulted in lower N03-N concentrations near the tree, and

concluded that the trees were reducing the potential for nitrate leaching. In another study,

Lehmann et al. (1998b) found that N leaching losses were 53% lower in an Acacia

saligna-Sorghum bicolor alley cropping system compared to sorghum monoculture.

They postulated that reduced leaching could have been due to a higher root abundance

and a higher nutrient uptake-to-leaching ratio in the alley cropping system, resulting in

higher nutrient use efficiency. Higher rooting density for the non-barrier treatment

compared to the barrier treatment has also been reported from our alley cropping system

(Wanvestraut et al. 2003). It is possible that the higher rooting density, specifically at

deeper horizons, increased nutrient uptake-to-leaching ratio in our non-barrier treatment

in comparison to the barrier treatment.

Jama et al. (1998) hypothesized that trees can rapidly root into subsoil and capture

NO3-N that has accumulated in agricultural subsoils. They compared inorganic NO3-N

and NH4-N in soil to a depth of 3.95 m in 11-month-old stands of calliandra (Calliandra

calothyrsus Meissner), sesbania [Sesbania sesban (L.) Merr.] and other woody perennial

species. They found that Calliandra and sesbania reduced soil NO3-N in the top 2 m by

150-200 kg N ha'.

In a recent study by Rowe et al. (1999), the safety net hypothesis was tested by

applying '5N-labeled ammonium sulfate at three depths in the soil between mixed

hedgerows of deep-rooted Peltophorum dasyrrhachis and shallower-rooted Gliricidia








sepium. They estimated that Peltophorum took up 42 kg N ha- and Gliricidia took up 21

kg N hal from beneath the main crop rooting zone. Their results indicated that the

deeper-rooted species had the capacity to act as a safety net, since it was taking up a

significant proportion of its N from deeper layers. Although pecan and cotton roots share

the top 0.3 m soil layer in our non-barrier treatment, the majority of the roots below 0.6 m

have been identified as pecan roots (Wanvestraut et al. 2003). This could be the reason

for the high NRI (0.86 for NH4-N and 0.81 for NO3-N) observed at 0.9 m in our system.

Conclusions

Overall, the results of our study indicate that the competitive presence of trees can

be utilized to decrease soil nitrate concentrations and reduce N leaching in alley cropping

systems. The barrier treatment had the potential to leach 43.87 kg N ha' year-' down

below 0.9 m depth. The non-barrier treatment exhibited a significantly lower potential

for leaching with only 15.76 kg N ha1' year-' below 0.9 m. For both treatments NO3-N

accounted for 99.9% of the total inorganic nitrogen. Further, tree water uptake, in

addition to cotton water uptake in the non-barrier treatment, may have decreased water

drainage in comparison to the barrier treatment, thereby influencing leaching rates.

Overall, the presence of tree roots were observed to be an important factor influencing

the fate of nitrogen in the system. These findings will improve our understanding of

nitrogen dynamics in temperate alley cropping systems, which in turn, will help in

designing systems that can utilize the safety net phenomenon to maximize fertilizer use

efficiency while minimizing nitrate contamination of groundwater.









Table 3-1. Seasonal and overall averages of NH4-N and N03-N concentrations in soil water extracted from 0.3 and 0.9 m depths in
barrier and non-barrier treatments.
Observation period Average NH4-N concentration (mg L-1) Average N03-N concentration (mg L')
Barrier No barrier Barrier No barrier
0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m
2001 Growing Season 0.04 a'A2 0.03 aA 0.03 aA 0.01 aA 18.21 aA 8.29 bcA 12.82 abA 7.09 cA
(June-November) (0.01)3 (0.02) (0.01) (0.00) (2.96) (1.08) (2.16) (3.09)
2001/2002 Dormant Season 0.02 aA 0.02 abA 0.02 aA 0.07 bB 5.86 aB 11.21 bAB 3.06 aB 6.93 abA
(December-April) (0.00) (0.01) (0.00) (0.05) (1.15) (1.80) (0.82) (1.45)
2002 Growing Season 0.11 aB 0.09 aB 0.06 aA 0.10 aB 22.10 aA 14.34 bcB 20.97 abC 8.06 cA
(May-August) (0.04) (0.03) (0.03) (0.02) (5.72) (1.99) (5.64) (3.65)
Whole Study Period 0.04 a 0.04 ab 0.03 ab 0.05 b 14.36 a 10.82 be 10.16 ab 7.15 c
(June 2001-August 2002) (0.01) (0.01) (0.01) (0.03) (1.93) (0.95) (1.63) (1.45)
1 Within-season values followed by the same lowercase letter are not significantly different at the 0.10 level of probability.
2 Across-season values followed by the same uppercase letter are not significantly different at the 0.10 level of probability.
3Standard errors of the mean are given in parentheses.









Table 3-2. Average monthly rates of NH4-N and N03-N leached at 0.3 and 0.9 m depths in barrier and non-barrier treatments.
Observation period Average rate of NH4-N leached (kg ha"' month ') Average rate of N03-N leached (kg ha"' month l)
Barrier No barrier Barrier No barrier
0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m
2001 Growing Season 0.04 a'A2 0.04 aA 0.03 abA 0.01 bA 15.15 aA 3.67 bAB 9.37 aA 1.27 cA
(June-November) (0.02)3 (0.02) (0.02) (0.00) (3.45) (0.68) (2.50) (0.33)
2001/2002 Dormant Season 0.00 aB 0.00 aB 0.00 aA 0.02 aA 2.45 abcB 3.95 bA 1.12 cB 1.55 cA
(December-April) (0.00) (0.00) (0.00) (0.02) (0.58) (0.86) (0.35) (0.41)
2002 Growing Season 0.01 aAB 0.00 aAB 0.01 aA 0.00 aA 9.96 aC 2.10 bcB 7.73 abA 0.39 cA
(May-August) (0.00) (0.00) (0.00) (0.00) (3.76) (0.73) (2.88) (0.36)
Whole Study Period 0.02 a 0.02 a 0.01 a 0.01 a 9.24 a 3.40 b 5.50 ab 1.28 c
(June 2001-August 2002) (0.01) (0.01) (0.01) (0.01) (1.76) (0.45) (1.20) (0.24)
Within-season values followed by the same lowercase letter are not significantly different at the 0.10 level of probability.
SAcross-season values followed by the same uppercase letter are not significantly different at the 0.10 level of probability.
3Standard errors of the mean are given in parentheses.









Table 3-3. Seasonal and cumulative totals of NH4-N and N03-N leached at 0.3 and 0.9 m depths in barrier and non-barrier treatments.
Observation period Total NH4-N leached (kg ha') Total N03-N leached (kg ha"')
Barrier No barrier Barrier No barrier
0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m 0.3 m 0.9 m
2001 Growing Season 0.20 a'A2 0.19 aA 0.11 abA 0.02 bA 81.81 aA 19.08 bA 41.23 abA 5.09 cA
(June-November) (0.07)3 (0.09) (0.08) (0.01) (14.97) (2.51) (11.35) (3.02)
2001/2002 Dormant Season 0.02 aB 0.02 aB 0.02 aA 0.12 aA 12.24 abB 19.76 aA 5.59 bB 7.42 abA
(December-April) (0.00) (0.01) (0.00) (0.11) (2.49) (5.71) (2.03) (2.50)
2002 Growing Season 0.03 aB 0.01 aB 0.01 aA 0.00 aA 27.90 aB 6.72 bB 17.02 abB 0.55 cB
(May-August) (0.01) (0.00) (0.01) (0.00) (9.98) (2.86) (7.32) (0.51)
Whole Study Period 0.25 a 0.21 a 0.14 a 0.14 a 121.94 a 45.56 b 63.83 b 13.05 c
(June 2001-August 2002) (0.03) (0.04) (0.03) (0.04) (9.74) (2.65) (5.79) (1.44)
Within-season values followed by the same lowercase letter are not significantly different at the 0.10 level of probability.
SAcross-season values followed by the same uppercase letter are not significantly different at the 0.10 level of probability.
3Standard errors of the mean are given in parentheses.





53


Table 3-4. Net Retention Index of NH4-N and N03-N at 0.3 and 0.9 m depths over two
growing seasons and whole study period.
Observation period NRI of NH4-N NRI of N03-N
0.3 m 0.9 m 0.3 m 0.9 m
2001 Growing Season 0.33 0.86 0.38 0.65
(June-November)
2002 Growing Season 0.33 0.25 0.22 0.81
(May-August)
Whole Study Period 0.35 0.13 0.40 0.62
(June 2001-August 2002)









Crop alley

1 2 3 4 5


6 7 8


* .....
















Lysimeter pairs
Cotton rows 0.76 m (0.3 and 0.9 m depth)


Figure 3-1. Plot layout showing location of suction lysimeters at the Jay, Florida
agroforestry site.


Reference tree









Tree row












































JJA SONDiJ FMAM J JA JJA SO NDJ FMAM J JA
2001 2002 2001 2002
Month Month

Figure 3-2. Monthly and cumulative drainage values at 0.3 and 0.9 m depths in barrier
and non-barrier treatments as estimated using the LEACHMN model.










0.30
Barrier (0 3 m depth)
0.25 U Barrier (0.9 m depth)
.J O No barrier (0.3 m depth)
0 No barrier (0.9 m depth)
S 0.20


S 015


0
o 0.10
z

Z 0.05


0.00
70.00
Barrier (0.3 m depth)
60.00 N Barrier (0.9 m depth)
S No barrier (0.3 m depth)
0 50.00 0 No barrier (0.9 m depth)
E
40.00


S30.00


S2000

Z 10.00


0.00
J J A S N D J F M A M J J A
2001 2002
Month

Figure 3-3. Average monthly concentrations ofNH4-N and NO3-N in soil solution at 0.3
and 0.9 m depths in barrier and non-barrier treatments.













CHAPTER 4
COMPETITION FOR NITROGEN IN A PECAN (Carya illinoensis K. Koch)-COTTON
(Gossypium hirsutum L.) ALLEY CROPPING SYSTEM IN THE SOUTHERN UNITED
STATES

Introduction

A primary form of agroforestry in North America is alley cropping, which

involves the planting of row crops or pasture in alleys formed by single or multiple rows

of trees or shrubs (Garrett and Buck 1997, Gillespie et al. 2000). Important crops for

alley cropping in the southern United States include cotton (Gossypium spp.), peanut

(Arachis hypogaea), maize (Zea mays L.), soybean (Glycine max. L. (Merr.)), wheat

(Triticum spp.) and oats (Avena spp.), combined with trees such as pines (Pinus spp.) and

pecan. Alley cropping is growing in popularity among landowners who wish to

maximize arable land use, diversify production, and increase off-season farm income.

The effect of trees in these systems is of interest environmentally as well, in part because

trees are capable of capturing and recycling fertilizer nutrients from deeper soil horizons,

and thus may help in improving nutrient use efficiency and in mitigating groundwater

contamination (Rowe et al. 1999).

Nitrogen (N) is generally the most limiting soil nutrient in temperate alley

cropping systems due to various reasons. First, nitrogen is lost via various

biogeochemical mechanisms such as volatilization, denitrification and leaching.

Nitrogen is also lost when crop biomass is removed from the field following harvest. In

addition, plants of the same species and growth stage can compete heavily for nitrogen

when zones of depletion in the soil overlap with neighboring plants. Moreover, in alley








cropping systems, competitive forces can be even more intense, as most tree species have

the bulk of their fine, feeder roots in the top 30 cm soil layer, thus placing them in a zone

of competition with crop species for water and nutrients (Rao et al. 1993, Lehmann et al.

1998a). Thus, tree-crop systems must be properly designed and managed in order to

maximize fertilizer use efficiency and minimize deleterious effects of competition on

crop yield.

The extent of competition between two species will depend on factors such as

nutrient and water availability, root architecture, rooting depth and proximity to

competing roots, and temporal nutrient demand (Jose et al. 2000a). Generally, tree roots

can exploit subsoil nitrogen and other nutrients beyond the rooting depths of crops, a

process that places trees at a competitive advantage for nutrients (Williams et al. 1997).

In addition, the peak intensity of nutrient demand in trees and crops may differ by several

months, as trees tend to exhibit highest nutrient demand in spring during leaf formation,

and crops such as cotton would be at highest demand in mid-summer during boll

formation.

In some cases, intercropped trees can receive benefits when fertilizer is applied to

nearby crops, as some of the nutrients will be intercepted and taken up by tree roots. This

type of secondhand fertilization could be an important factor in the growth of associated

tree species in alley cropping systems. Pecan trees, for example, which are often below

their potential yielding capacity due to nitrogen deficiency (Arnold and Crocker 1999),

could thus benefit from the application of fertilizer to an intercropped row crop.

Pecan-based alley cropping systems offer potential for Southern landowners,

given the large number of pecan orchards in the southeastern USA and the possible








environmental and financial benefits that may be accrued from such systems. However,

the movement of nitrogen in pecan-cotton systems remains an unstudied but critical

factor affecting the growth and productivity of both trees and crops. While nitrogen

losses cannot be avoided completely, losses can be minimized through appropriate

fertilizer and orchard management practices and by knowledge of how nitrogen moves in

the soil-tree system (Herrera and Lindemann 2001). Thus, more understanding is needed

of the interactive dynamics of nitrogen in tree-crop systems, in order to maximize

fertilizer use efficiency and optimize production from each component.

Hence, the present study was conducted to examine competition for nitrogen

between pecan and cotton using 5N-labeled ammonium sulfate fertilizer. Labeled s1N-

enriched fertilizer has been used as a nitrogen-tracing technique in cotton (Freney et al.

1993, Rochester et al. 1993, Bondada et al. 1996, Karlen et al. 1996) and to a very limited

degree in pecan (Kraimer et al. 2001).

The specific objectives of the study were as follows:

* Quantify the total uptake of nitrogen by trees and cotton with and without

interspecific competition;

* Determine if the competition between trees and cotton can change the relative uptake

of fertilizer nitrogen;

* Determine whether fertilizer use efficiency of cotton plants is altered as a result of the

interspecific competition; and

* Quantify the recovery of fertilizer nitrogen in soil at four successive depths with and

without interspecific competition.








Materials and Methods

Study Area and Configuration

This study was conducted at the West Florida Research and Education Center

Farm of the University of Florida, located near Jay in northwestern Florida, USA (30o89'

N Lat., 8713' W Long.). The climate is temperate with moderate winters and hot,

humid summers. The soil at the site is classified as a Red Bay sandy loam, which is a

fine-loamy, siliceous, thermic Rhodic Paleudult. The average water table depth is 1.8 m.

For this study, a pecan-cotton alley cropping system was initiated in Spring 2001

from an existing orchard of pecan trees that had been planted at a uniform spacing of

18.28 m in 1954 and that had remained under grass cover for 29 years until the initiation

of the current study. Ten plots were established within the orchard and arranged into five

blocks using a randomized complete block design. Each plot, which consisted of two

rows of trees oriented in a north-south direction, was 27.43 m long and 18.28 m wide,

with a practical cultivatable width of 16.24 m, and was separated from its adjacent plot

by a buffer zone of the same dimensions.

To assess tree root competition for nitrogen fertilizer, each block was randomly

divided into a barrier plot and a non-barrier plot. Barrier plots were subjected to a root

pruning treatment in March of 2001 in which a trenching machine was used to dig a 0.2

m wide x 0.9 m deep trench along both sides of the plot at a distance of 1.5 m from the

trees to separate root systems of pecan and cotton. A double layer of 0.15 mm-thick

polyethylene sheeting was used to line the ditch prior to mechanical backfilling. The

barrier plots thus served as the tree root exclusion treatment, preventing interaction of








tree and cotton roots, while the non-barrier plots, which did not receive this treatment,

served as the tree-crop competition treatment.

Microplots and Fertilizer Application

For this study, cotton (DP 458 B/RR) was planted in rows 0.91 m apart at 16 rows

per alley in a north-south orientation on 16 May 2001, after disking of the alleys.

Conventional insecticide and herbicide were applied during the growing season as

recommended. In each plot, one microplot (2.60 m x 0.76 m), containing 8-10 plants,

was established on the first, fourth and eighth rows of cotton, respectively (going west to

east in each plot) (Figure 4-1). To quantify nutrient competition, 15N enriched fertilizer

((NH4)2SO4, 5% atom enrichment) was uniformly hand-applied to microplots at a rate of

89.6 kg N ha'l on 19 June 2001 at the same time, rate and formulation as the regular

fertilizer application (Timmons and Cruse 1990). Each microplot was arranged so that

one of the pecan trees in the tree row was in the center and could serve as the target tree

for 15N sampling.

Sampling Methods

Six cotton plants (aboveground portions) from each microplot were sampled for

iSN content in leaf, stem and boll components. Cotton leaf samples were collected on 8

November 2001 prior to leaf senescence. The same plants were later harvested at

physiological maturity on 4 December 2001, and separated into stem and boll

components. In addition, foliar samples from each associated pecan tree were collected

on 10 October 2001. For this purpose, the tree canopy was divided into an upper and

lower half, and leaves were collected via shotgun harvest method from all four cardinal

directions in both the halves to provide one composite leaf sample per tree.








Following collection, all plant tissue samples were dried at 650C for 72 hours. In

preparation for combustion analysis, all green plant tissue samples (cotton leaves and

stems, and tree leaves) were ground with a model 4 Wiley Mill (Arthur H. Thomas

Company, Philadelphia, Pennsylvania) to pass through a 1 mm screen, and then re-

ground using a burr coffee grinder. Both grinders were thoroughly cleaned between

samples to prevent cross-contamination of the 15N plant material. Cotton lint was de-

seeded and manually shredded with scissors in preparation for analysis.

Soil cores, measuring 120 cm in length and 3.5 cm in diameter, were collected in

pairs at random points within each microplot in January 2002, using a tractor-mounted

hydraulic corer and polyethylene collection tubes. The cores were divided into 15 cm

increments to a depth of 120 cm, composite for each microplot depth, air dried, and a

subsample was fine-ground with a mortar and pestle.

For determination of total N and '5N concentrations, subsamples of the ground

plant material and soil samples were analyzed by the University of Florida Geological

Sciences Department (Gainesville, Florida) using a Finnigan-MAT DELTAP'"U isotope

ratio mass spectrometer with a ConFlo III interface attached to a Costech ECS 4010

elemental analyzer (Schepers et al. 1989). Percent nitrogen derived from fertilizer,

percent utilization of fertilizer nitrogen, and percent nitrogen recovery in soil, were

calculated from the enrichment data to determine the degree of interspecific competition

for nitrogen.

Percentage of plant nitrogen derived from fertilizer (NDF), a measure of the

relative amounts ofN a crop obtains from the soil and from applied fertilizer, was

calculated as follows (Wienhold et al. 1995):








NDF (%)= 100* (a -b)/(c-d) (4-1)

where

a = atom % 15N abundance in cotton leaf, stem, seed cotton, or tree leaf;

b = atom % '5N abundance in control cotton leaf, stem, seed cotton, or tree leaf;

c = atom % 15N abundance of fertilizer; and

d = natural atom % 15N abundance.

Percentage utilization of fertilizer N (UFN), a measure of fertilizer use efficiency,

was calculated for cotton plants and pecan foliage as follows (Wienhold et al. 1995,

Barber et al. 1996):

UFN (%) = (%NDF S) / R (4-2)

where

%NDF = percentage of plant nitrogen derived from fertilizer;

S = kg N ha-' in cotton leaf, stem, seed cotton, or tree leaf; and

R = kg N ha- applied.

Percentage recovery of '5N fertilizer in soil (RFNsoil), a measure of the applied

5N remaining in soil, was determined using the following equation (De Mattos 2000):

RFNsoi (%) = 100 ((a c) / (b c)) (Np / Nf) (4-3)

where

a = atom % 15N abundance in fertilized soil material;

b = atom % 1SN abundance in labeled N fertilizer;

c = atom % 15N abundance in non-fertilized soil (average background level);

Np = total N of soil sample (in g); and

Nf= total amount of 15N applied to the soil as labeled fertilizer (g).








Data Analyses

Statistical analyses were performed using SAS 8.2 for Windows (SAS Institute,

Cary, North Carolina) using the Proc Mixed procedure within the framework of a split

block design. The Shapiro-Wilk's test, in combination with ocular inspection of

frequency distributions, was used to test all data for normal distribution. Logarithmic

[log (x + 1)] or arcsin transformation was conducted on data when such transformation

improved normality. Differences between means were determined using the Least

Squares Means procedure. Treatment effects were considered significant at a = 0.05.

Results

Aboveground Biomass

The barrier treatment resulted in a 58.4% increase in total aboveground biomass

of cotton plants compared to the non-barrier treatment (5750.26 kg ha' vs. 3629.09 kg

ha"', respectively) (Table 4-1). Total biomass also varied significantly by row (p=0.0257)

in the non-barrier treatment. Cotton foliage biomass varied significantly by treatment

(p=0.0474), with leaf biomass in the barrier treatment being 38.3% higher than that in the

non-barrier treatment (1126.48 kg ha"' vs. 814.69 kg ha ', respectively). Within the

barrier treatment, no significant difference among rows was observed for leaf biomass;

however, leaf biomass increased significantly at Row 8 (p=0.0180) in the non-barrier

treatment, representing a 66.1% increase over Row 1.

For cotton stem, biomass varied by treatment (p=0.0135) and by row (p=0.0109)

(Table 4-1). Stem biomass was 66.4% higher in the barrier than the non-barrier treatment

(3815.28 kg haJ' vs. 2293.19 kg ha ', respectively). Within each treatment, stem biomass








increased significantly with increasing distance from the tree, occurring at Row 4 in

barrier (p=0.0304) and Row 8 in non-barrier (p=0.0112) treatments.

Seed cotton biomass was significantly different between treatments (p=0.0493)

and rows (p=0.0020) (Table 4-1). Overall, seed cotton biomass in barrier treatment was

55.1% greater than that in non-barrier treatment (808.51 kg hal vs. 521.21 kg ha',

respectively). For the barrier treatment, Row 1 had significantly higher seed cotton

biomass than Rows 4 and 8. Seed cotton biomass in the non-barrier treatment showed no

significant variation among rows.

Average dry leaf biomass of pecan trees was similar for both treatments, with

2516.40 kg ha' for barrier treatment and 2580.60 kg ha' for non-barrier treatment,

respectively, based on litter trap collection data obtained from a separate study at the site

(Zamora and Jose, unpublished data).

Nitrogen Concentration and Content

For each type of plant tissue analyzed, nitrogen concentrations were not affected

by treatment or row, except for cotton leaf, which had a higher N concentration in Rows

4 and 8 in barrier treatment (Table 4-2). Nitrogen content, however, was affected by both

treatment and row. In particular, the barrier treatment resulted in a 66.8% increase in

total aboveground nitrogen content in cotton (83.02 kg ha ) compared to the non-barrier

(49.77 kg ha ') treatment.

Nitrogen concentration of cotton foliage in the barrier treatment was 14.1% higher

than in the non-barrier treatment, although not a statistically significant increase (Table 4-

2). However, leaf nitrogen content was 56% higher in the barrier treatment (p=0.0409)

than the non-barrier treatment (%N=42.04 and 26.95, respectively). Within the non-








barrier treatment, average leaf nitrogen content was significantly higher for Row 8

(%N=33.37) compared to Row 1 (%N=20.97).

Stem nitrogen concentrations averaged 0.99% as a whole, and did not differ

across rows (Table 4-2). However, stem nitrogen content was 80.7% greater in the

barrier treatment (p=0.0073), and showed a significant increase in Rows 4 and 8 of the

barrier treatment.

Seed cotton nitrogen concentration showed no significant difference across rows,

averaging 0.24% and 0.23% N for barrier and non-barrier treatments, respectively (Table

4-2). Seed cotton nitrogen content, although not significant between treatments, was

significantly higher in Row 1 than in Row 4 of the barrier treatment.

For pecan foliage, percent nitrogen concentration was 2.16% in barrier and 2.27%

in non-barrier treatments, respectively, which did not represent a significant difference

(Table 4-4). Nitrogen concentration of senesced leaves was 1.88%, based on samples

obtained from a separate control area. Canopy nitrogen content was also similar for both

treatments, with 54.36 kg hal' for barrier and 58.58 kg ha-' for non-barrier treatments,

respectively. Leaf litter nitrogen content was estimated to be 47.31 kg ha' for barrier and

48.52 kg ha' for non-barrier treatments, representing the potential nitrogen deposition

from leaf litter.

Uptake of Fertilizer Nitrogen

The nitrogen derived from fertilizer (NDF) in cotton leaf, stem and seed cotton

showed significant differences between treatments (Table 4-3). Overall, NDF in cotton

leaf was significantly (p=0.0019) lower in the barrier treatment (15.82%) compared to the

non-barrier treatment (20.40%), representing a 22.5% decrease. However, NDF








increased with increasing distance from trees in the non-barrier treatment, becoming

significantly higher for Row 8 (p=0.0073).

A similar trend was observed for cotton stem, with significantly (p=0.0048) lower

NDF for stems in the barrier treatment (17.30%) compared to those in the non-barrier

treatment (21.15%) (Table 4-3). The NDF for cotton stem also increased across rows in

the non-barrier treatment, becoming significantly higher at Row 8 (p=0.0038). The NDF

for seed cotton, which did not vary significantly according to row or treatment, was

14.76% and 16.95%, respectively, for the barrier and non-barrier treatments. The NDF

for pecan foliage did not vary significantly between treatments, averaging -0.026% for

barrier treatment and 0.063% for non-barrier treatment, respectively (Table 4-4). In this

instance, a negative value for NDF and UFN in the barrier treatment was possible

because the observed 15N abundance was minutely lower than the background average.

Fertilizer Nitrogen Use Efficiency

Total fertilizer N use efficiency for leaf, stem and seed cotton combined was

15.13% for barrier plants and 11.70% for non-barrier plants, respectively, representing an

overall increase in total UFN of 29.3% (p=0.0340) (Table 4-3). Total UFN increased

from outer to inner rows and was significant for Row 8 in both barrier (p=0.0147) and

non-barrier (p=0.0006) treatments. Overall, utilization of fertilizer nitrogen was similar

in both cotton leaf and stem, and higher than UFN in seed cotton, for both treatments.

For cotton foliage, UFN was similar for both barrier (7.37%) and non-barrier

(6.25%) treatments (Table 4-3). However, UFN of cotton leaf increased with increasing

distance from tree base, becoming significantly higher at Row 8 (p=0.0051) in the non-

barrier treatment. A significantly higher UFN for stem was observed for barrier (7.46%)








plants compared to non-barrier (5.22%) plants. In addition, UFN for stem varied by

treatment and row, being significantly higher at Row 8 for both barrier (p=0.0122) and

non-barrier (p=0.0069) treatments. The UFN of seed cotton was significantly higher in

the barrier (0.29%) than the non-barrier (0.23%) treatment (p=0.0348), with no

significant row effects. The UFN of pecan foliage was not significantly different

between treatments, averaging -0.025% for barrier treatment and 0.037% for non-barrier

treatment, respectively (Table 4-4).

Atom Percent L5N Abundance and Nitrogen Concentration in Soil

Overall, atom %15N abundance in soil was found to be similar (p=0.4869) in

barrier and non-barrier treatments based on isotopic analysis of the 15 cm soil core

increments (Figure 4-2). Rows were also not significant (p=0.6402). However, depth

was determined to be significant (p<0.0001), and there was an interaction between

treatment and depth (p=0.0308). In the non-barrier treatment, the %15N distribution was

uniform across rows and depths. However, in barrier treatment, %15N increased with

depth, with values for upper layers (0-60 cm) being similar to those in the non-barrier

treatment, and those for lower layers (60-105 cm) being higher.

Nitrogen concentration in soil did not vary significantly by treatment (p=0.8820)

or by row (p=0.2404) (Figure 4-2). However, N concentration decreased significantly

with depth (p<0.0001) at 15-30 cm depth and below.

Recovery of 15N in Soil

A comparison of 30 cm soil increments (0-30 cm, 30-60 cm, 60-90 cm, and 90-

120 cm depths) showed that percent recovery of 15N at the end of the growing season did

not differ between treatments (p=0.3251) (Table 4-5). However, depth was significant








(p<0.0001), and the interaction between treatment and depth was significant (p=0039).

Specifically, the treatment x depth interaction at the 90-120 cm depth showed a decrease

(p=0.0207) in '5N recovery in the non-barrier treatment, where the rate of recovery was

2.08% compared to 3.93% in the barrier treatment. Total %'5N remaining in soil at the

end of the growing season for all depths combined was 18.94% for barrier treatment and

14.55% for non-barrier treatment, respectively.

Discussion

Overall, we observed that cotton plants in the barrier treatment had a 58.45%

higher total aboveground biomass compared to the non-barrier treatment. Cotton foliage

biomass, for example, was 38.30% higher in the barrier than the non-barrier treatment.

Stem and seed cotton biomass were also significantly higher in the barrier treatment.

Similar results have been observed for other studies employing belowground root

barriers in alley cropping systems. Miller and Pallardy (2001) studied belowground

competition in a maize-silver maple (Acer saccharinum L.) alley cropping system in

north-central Missouri. In their study, which was located on a Mexico series soil (a fine,

montmorillonitic, mesic, Udollic Ochraqualf), they observed a 27.77% higher maize

grain yield in plots with root barriers compared to plots without them. Similarly, Jose et

al. (2000b) observed a 67.33% higher grain yield and a 37.21% higher stover biomass in

root barrier plots, in their study of a black walnut (Juglans nigra L.)-maize alley

cropping system on a Parke Silt Loam soil (Ultic Hapludalf) soil in Indiana. Singh et al.

(1989) observed that root barriers in a sorghum (Sorghum bicolor Moench)-leucaena

(Leucaena leucocephala Lam. de Wit) alley cropping system in semiarid India, raised

sorghum grain yields from 0.42 Mg hal' to 1.63 Mg ha'. However, in each of these








studies, competition for water was identified as the primary factor causing biomass

reductions, rather than competition for nutrients, since water deficits resulted in smaller

crop size and reduced fertilizer use efficiency. Findings by Wanvestraut et al. (2003),

who conducted a companion study to the one reported here, confirmed the occurrence of

water competition in the current study site. Cotton lint yield in their study was reduced

by 21% due to belowground competition for water in the absence of a root barrier.

For pecan trees, average tree diameter and dry leaf biomass did not differ between

treatments. However, a reduction in tree growth and yield would be expected in response

to the root pruning treatment. Miller and Pallardy (2001), for example, observed a 20%

reduction in annual radial growth of silver maple at the end of the first growing season

after root barrier installation. These parameters were not assessed in our study.

Although nitrogen concentrations in cotton tissues were not affected by treatment,

nitrogen content was affected by both treatment and row. Cotton plants in the barrier

treatment had a 56% higher nitrogen content than those in the non-barrier treatment

because of the greater plant biomass in the barrier treatment. For both treatments, total

nitrogen content increased with increasing distance from the tree, with Row 8 being

significantly higher than Row 1 for both treatments. This row effect is not fully

understood, but differences in nitrogen content may have resulted from more favorable

soil moisture or light conditions in the mid-alley rows.

The root barrier had no significant effect on pecan leaf nitrogen concentration or

canopy nitrogen content. These findings are in contrast to the results of Jose et al.

(2000b), who observed leaf nitrogen concentration in 15-year-old black walnut to be

15.5% lower in barrier trees than non-barrier trees. In that study, the non-barrier tree








roots apparently acquired N from the fertilized alley, producing a higher nitrogen

concentration in the black walnut leaves. The reasons for the differing results are not

clear, although they could be attributed to tree age and root system development. The

roots of older trees (as in our study) would be expected to have access to a wider and

deeper pool of soil nutrients compared to younger trees, and would therefore be less

dependent on near-surface fluctuations in soil nitrogen availability.

Nitrogen uptake in crops can be supplemented via the mineralization of tree leaf

and root litterfall, which is dictated largely by the litter quality, which differs among tree

species and management (Handayanto et al. 1997). Pecan trees in our study were

estimated to have an average of 56.47 kg ha' in total canopy nitrogen content, which,

allowing for resorption, results in an estimated 47.91 kg ha- of nitrogen being deposited

to the orchard floor as leaf litter on a yearly basis (Table 4-4). This build-up and

decomposition of litterfall, combined with rootfall and deposition of other organic debris,

may serve to augment soil N supplies and thus reduce the amount of nitrogen fertilizer

required for cotton production (Sanginga et al. 1990, Seiter and Horwath 1999).

Rhoades et al. (1998), for example, in their study of a sorghum-mimosa (Albizia

julibrissin) alley cropping system on a highly-weathered Ultisol (Typic Hapludult) in

Georgia, found that tree-mulch additions enhanced crop biomass production and N

uptake by 2 to 3.5 times in both high and low moisture conditions. In another study,

Seiter and Horwath (1999) demonstrated that soil organic matter could increase by 4 to

7% in alley cropping systems with alder (Alnus sinuta) and maize in comparison to maize

monocultures following four years of cropping on a Chehalis soil series (a fine silty,

mesic, Ultic Haploxeroll) in Oregon.








Overall, our 12-25% NDF range in cotton plants was lower than the 33-40% NDF

reported for cotton in other studies (Rochester et al. 1994, Gibb et al. 2002). Our non-

NDF range of 75-88% was higher than the 67-68% rate reported in other studies

(Rochester et al. 1993, Gibb et al. 2002). Thus, use of indigenous soil N in our study

appears to have been generally higher than average. This is perhaps due to an abundance

of soil nitrogen mineralized from the long-term deposition of pecan leaf and root litter in

the orchard.

With regard to the specific effects of treatment on NDF, NDF in cotton leaf and

stem was significantly lower in the barrier treatment compared to the non-barrier

treatment, representing a decrease of 22.5% and 18.20%, respectively. Apparently,

cotton plants in the barrier treatment took up more of their nitrogen from the nitrogen

already present in the soil than the applied fertilizer nitrogen. One of the factors that

could influence nitrogen allocation patterns in intercropped cotton is the synchrony of

crop and tree nutrient demand (Xu et al. 1993, Handayanto et al. 1997, Rowe et al. 1999).

While initial nutrient demand in cotton is low prior to fruiting or flowering (Ayala and

Doerge 2001, Crozier 2003), demand intensifies about 45 days after emergence, with a

prolonged peak about two weeks after first bloom, when flower production, boll filling,

and boll maturation are most active (Crozier 2003). In contrast, pecan trees in our study

exhibited a pattern of early leaf development, indicating heavy nutrient demand early in

the year. Thus, tree-root acquisition of soil N may have occurred well before cotton

stand establishment, leaving soil in the non-barrier treatment less rich in N than in the

barrier treatment. This may explain the higher dependency on fertilizer N exhibited in

non-barrier plants compared to barrier plants.








The NDF level in pecan leaves was minimal but existent (0.063%) for non-barrier

trees, and minimally negative for barrier trees. Pecan has been shown to recover as much

as 19.5% of applied fertilizer, if applied early in the growing season (Kraimer et al.

2001). However, as explained earlier, the bulk of nitrogen taken up by the pecan trees in

our study was obtained prior to the June application of fertilizer. It may be possible that

there is substantial NDF stored elsewhere in the tree, such as in roots, stems or branches;

however, these components were not measured in our study.

Total UFN for leaf, stem and seed cotton combined was 15.13% for barrier plants

and 11.70% for non-barrier plants, respectively, representing an overall decrease in total

UFN of 22.67%. Here, plants in the non-barrier treatment showed reduced growth, and

so were unable to utilize the available fertilizer efficiently, a result observed for maize

plants in previous research (Wienhold et al. 1995, Jose et al. 2000b). Total UFN

increased from outer to inner rows and was significant for Row 8 in both barrier and non-

barrier treatments. Utilization of fertilizer nitrogen was similar in both cotton leaf and

stem, and much higher than UFN in seed cotton, for both treatments. The UFN for pecan

in non-barrier was 0.037%, indicating minimal uptake and utilization of the fertilizer N.

When evaluating soil N status at the end of the growing season, we observed that

soil N concentration followed a predictable pattern of decline at 15 cm depth and below

in both treatments. This trend would be expected given the shallow root zone of cotton

and the susceptibility of N to leaching in the sandy soil of the study site.

We also observed that atom %15N concentration remained uniform for the non-

barrier treatment, while that for the barrier treatment increased at 60 cm depth and below.

A similar treatment x depth interaction was observed for %15N recovery, where tree root








uptake at the 90-120 cm depth apparently reduced the rate of recovery to 2.08% in the

non-barrier treatment compared to 3.93% in the barrier treatment. Overall, for all depths

combined, %15N recovery in soil was 18.94% for barrier and 14.55% for non-barrier

treatments, respectively, somewhat lower than the 20-25% recovery range reported by

other researchers (Karlen et al. 1996, Gibb et al. 2002).

Various factors in our study may have affected recovery of fertilizer-sN in the

soil. Tree roots may have accessed a portion of the fertilizer nitrogen in the non-barrier

treatment, thus reducing recovery in the soil. In addition, a substantial amount of N may

have been lost to biogeochemical mechanisms. For example, Kraimer et al. (2001)

reported that '5N levels in soil under pecans were highest at soil depths just above the

water table (280 cm), indicating losses due to leaching in the five months after

application. In another study, Karlen et al. (1996) speculated that 50% of 15N applied to

their study was lost via volatilization, denitrification or leaching below the crop root

zone. Likewise, Rochester et al. (1993) noted an exponential decline of 'N recovery

with time, which they attributed to biological immobilization and possibly denitrification.

Because the 15N-labeled fertilizer in our study was necessarily applied as a side-dress in a

single application, a significant portion may have leached out of the system, given the

high rainfall observed in July (139.19 mm) and August (217.42 mm) of 2001, the two

months following application.

Overall, these findings illustrate the importance of proper nutrient management in

alley cropping systems in order to achieve most efficient use of nitrogen inputs. As with

any cotton production system, appropriate fertilizer management should be followed, in

order to improve fertilizer-N efficiency, supply crop needs for nitrogen, and reduce losses








of nitrate (NO3-N) due to leaching (Ayala and Doerge 2001). This includes use of soil

tests, monitoring of petiole nitrate levels, and split applications of fertilizer by use of

side-dress and/or fertigation applications. The use of commercial nitrification inhibitors

is also suggested as a means to increase fertilizer nitrogen recovery and lint yield, as a

delay in nitrification has been shown to improve total nitrogen uptake in cotton,

especially during the boll maturation period (Freney et al. 1993, Rochester et al. 1994).

In addition, conservation tillage (e.g., no-tillage or strip-till) and winter cover

management practices may be applied, to maximize accumulation of plant surface mulch

and thus conserve water and increase soil fertility (Varco et al. 1999). Taken together,

these measures can help to achieve more efficient synchrony of the nutrient, water and

light demand functions of cotton and pecan during the growing season.

Conclusions

Although nitrogen is shared between cotton and pecan according to the

competitive abilities of the two species, existing soil nutrient levels, water availability,

and temporal plant nutrient demands are critical factors in this process. Our results

indicate that cotton plants are subject to competition for nitrogen; companion studies

have indicated that they are subject to competition for water also. Competition for

nitrogen was alleviated to a great extent by the application of fertilizer nitrogen, and

possibly by existing soil nitrogen reserves from the deposition of pecan litterfall. Further,

nitrogen uptake and allocation patterns in both pecan and cotton were influenced largely

by temporal differences in N demand and apparently by the abundance of mineralized

nitrogen in soil. We observed increases in nitrogen content of cotton in the presence of

root barrier, although the barrier had no significant effect on pecan leaf nitrogen








concentration or canopy nitrogen content. The NDF was lower for cotton in barrier

plants, indicating that cotton in this treatment was taking up a higher percentage of its

nitrogen from nitrogen already present in the soil. The NDF in pecan was minimal,

indicating an early and substantial uptake of N prior to the cotton season and fertilizer

application. However, it may be possible that there is substantial NDF stored elsewhere

in the tree, such as in roots, stems or branches. Total UFN was higher in barrier cotton

plants, indicating a greater ability to utilize the available fertilizer efficiently. In soil,

depth was the primary factor influencing fertilizer N concentrations and recovery rates.

However, a treatment x depth interaction was observed at the lowest depth in the non-

barrier treatment, where the '5N recovery rate was lower than that in the barrier treatment.

Apparently, fertilizer nitrogen in the non-barrier treatment was taken up by tree roots

from this deeper horizon. Overall, the alley cropping system in this study exhibits

potential for nutrient capture and increased fertilizer use efficiency, given the apparent

ability of pecan trees to intercept nitrogen fertilizer at lower depths and to provide

litterfall to the cropping zone.








Table 4-1. Biomass of cotton leaf, stem and seed cotton in barrier and non-barrier
treatments.
Treatment Row Biomass (kg ha ')
Leaf Stem Seed cotton Total

Barrier 1 1166.71 a' 2946.95 a 1147.25 a 5260.91 a
(102.32)2 (530.21) (174.08)
4 1066.13 a 4073.43 b 625.76 b 5765.32 a
(105.97) (513.59) (59.69)
8 1146.59 a 4425.45 b 652.51 b 6224.55 a
(136.25) (468.53) (34.47)
Mean3 1126.48 3815.28 808.51 5750.26
(114.85) (504.11) (89.41)
No barrier 1 623.59 a 1689.72 a 528.42 a 2841.73 a
(123.80) (227.47) (37.94)
4 784.51 ab 2092.03 a 486.24 a 3362.78 a
(113.12) (195.81) (49.71)
8 1035.96 b 3097.82 b 548.97 a 4682.75 b
(58.65) (370.44) (11.27)
Mean3 814.69 2293.19 521.21 3629.09
(98.52) (264.57) (32.97)
p value4 0.0474 0.0135 0.0493 0.0076
SWithin-treatment values followed by the same lowercase letter are not significantly
different at the 0.05 level of probability.
2 Standard errors of the mean are given in parentheses.
3 Mean indicates the treatment mean.
4 p value indicates the significance between treatment means.








Table 4-2. Nitrogen concentration and nitrogen content in cotton leaf, stem and seed
cotton in barrier and non-barrier treatments.
Treatment Row Nitrogen concentration (%) Nitrogen content (kg ha"')
Leaf Stem Seed Leaf Stem Seed Total
cotton cotton

Barrier 1 3.45 a 0.95 a 0.21 a 40.19 a 28.0 a 2.40 a 70.59 a
(0.12)2 (0.06) (0.02) (3.72) (5.12) (0.44)
4 3.90 b 1.04 a 0.25 a 41.78 a 42.95 b 1.60 b 86.33 ab
(0.06) (0.07) (0.02) (4.54) (7.07) (0.25)
8 3.83 b 1.07 a 0.26 a 44.16 a 46.20 b 1.77 ab 92.13 b
(0.12) (0.07) (0.04) (5.63) (3.35) (0.36)
Mean3 3.73 1.02 0.24 42.04 39.05 1.92 83.02
(0.10) (0.07) (0.03) (4.63) (5.18) (0.35)
No barrier 1 3.26 a 0.97 a 0.23 a 20.97 a 17.13 a 1.22 a 39.32 a
(0.20) (0.10) (0.02) (4.80) (3.94) (0.06)
4 3.33 a 0.90 a 0.22 a 26.52 ab 18.99 a 1.06 a 46.57 ab
(0.22) (0.05) (0.03) (4.61) (2.49) (0.10)
8 3.21 a 0.95 a 0.25 a 33.37 b 28.71 a 1.34 a 63.42 b
(0.09) (0.07) (0.04) (2.25) (2.18) (0.20)
Mean3 3.27 0.94 0.23 26.95 21.61 1.21 49.77
(0.17) (0.07) (0.03) (3.89) (2.87) (0.12)
p value4 0.0942 0.3540 0.6811 0.0409 0.0073 0.0554 0.0122
SWithin-treatment values followed by the same lowercase letter are not significantly
different at the 0.05 level of probability.
2 Standard errors of the mean are given in parentheses.
3 Mean indicates the treatment mean.
4 p value indicates the significance between treatment means.








Table 4-3. Percentage of nitrogen derived from fertilizer (NDF) and percentage
utilization of fertilizer nitrogen (UFN) for cotton leaf, stem and seed cotton in
barrier and non-barrier treatments.
Treatment Row NDF (%) UFN (%)
Leaf Stem Seed Leaf Stem Seed Total
cotton cotton

Barrier 1 15.37 a' 17.60 a 12.75 a 6.86 a 5.41 a 0.31 a 12.58 a
(0.79)2 (1.24) (2.18) (0.66) (0.88) (0.05)
4 15.31 a 16.43 a 18.23 a 7.08 a 7.74 ab 0.31 a 15.13 ab
(1.40) (0.95) (4.11) (0.87) (1.12) (0.06)
8 16.77 a 17.88 a 13.31 a 8.17 a 9.24 b 0.26 a 17.67 b
(1.91) (2.00) (1.61) (1.22) (1.17) (0.06)
Mean 15.82 17.30 14.76 7.37 7.46 0.29 15.13
(1.37) (1.40) (2.63) (0.92) (1.06) (0.06)
No barrier 1 18.26 a 18.98 a 16.19 a 4.37 a 3.66 a 0.22 a 8.25 a
(0.76) (0.66) (2.92) (1.12) (0.87) (0.04)
4 19.32 a 19.44 a 13.96 a 5.60 a 4.05 a 0.16 a 9.81 a
(0.76) (1.15) (1.27) (0.85) (0.42) (0.02)
8 23.61 b 25.04b 20.69a 8.79 b 7.94 b 0.31 a 17.04 b
(1.04) (1.18) (1.37) (0.69) (0.45) (0.05)
Mean3 20.40 21.15 16.95 6.25 5.22 0.23 11.70
(0.85) (1.00) (1.85) (0.89) (0.58) (0.04)
value4 0.0019 0.0048 0.1333 0.2580 0.0161 0.0348 0.0340
' Within-treatment values followed by the same lowercase letter are not significantly
different at the 0.05 level of probability.
2 Standard errors of the mean are given in parentheses.
Mean indicates the treatment mean.
4 p value indicates the significance between treatment means.








Table 4-4. Percentage of nitrogen derived from fertilizer (NDF), percentage utilization of
fertilizer nitrogen (UFN), and other physiological parameters for pecan trees
in barrier and non-barrier treatments.
Parameter Treatment p value'
Barrier No barrier
Average tree dbh (cm) 216.62 (9.47)2 221.35 (5.99) 0.6837
Average dry leaf biomass (kg tree-') 83.88 (10.18) 86.02 (8.36) 0.8749
Leaf nitrogen concentration (fresh) (%) 2.16(0.05) 2.27 (0.07) 0.3846
Leaf 15N concentration (fresh) (%) 0.366 (0.002) 0.370 (0.001) 0.2754
Canopy nitrogen content (kg tree ') 1.81 (0.22) 1.95 (0.19) 0.6513
Canopy nitrogen content (kg ha-) 54.36 (6.60) 58.58 (5.69) 0.6406
Leaf NDF (fresh) (%)3 -0.026 (0.04) 0.063 (0.02) 0.0880
Leaf UFN (fresh) (%)3 -0.025 (0.03) 0.037 (0.01) 0.0975
Leaf litter nitrogen concentration (%) 1.88 (0.03) 1.88 (0.03)
Leaf litter nitrogen content (kg tree-') 1.58 (0.19) 1.62 (0.16)
Leaf litter nitrogen content (kg ha') 47.31 (5.74) 48.52 (4.71)
Sp value indicates the significance between treatment means. Since senesced leaf litter
samples were obtained from a control area, p values for these parameters are not shown.
2 Standard errors of the mean are given in parentheses.
3A negative value for NDF or UFN is possible when observed '5N abundance fluctuates
below the background average.








Table 4-5. Percentage of 5N recovery in soil at end of growing season in barrier and
non-barrier treatments.
Treatment Row %15N recovery in soil
0-30 cm 30-60 cm 60-90 cm 90-120 cm Total

Barrier 1 9.39 1.24 4.13 4.78 19.55
(1.81)1 (0.08) (2.74) (2.05) (1.11)
4 6.01 1.70 6.29 2.09 16.08
(1.62) (0.42) (3.73) (0.54) (1.07)
8 10.87 1.70 3.70 4.92 21.18
(3.48) (0.32) (0.70) (2.07) (1.23)
Mean2 8.76 a 1.55 b 4.71 c 3.93 be 18.94
(1.42) (0.18) (1.48) (0.98) (0.65)
No barrier 1 6.26 1.67 2.39 1.48 11.80
(0.72) (0.65) (1.19) (0.73) (0.59)
4 7.96 1.35 2.74 2.96 15.00
(0.90) (0.18) (0.93) (2.55) (0.87)
8 10.70 2.30 2.05 1.79 16.84
(2.35) (0.18) (0.61) (1.46) (1.08)
Mean2 8.31 a 1.77 b 2.39 b 2.08 c 14.55
(0.94) (0.24) (0.51) (0.95) (0.50)
p value 4 0.8683 0.8418 0.2474 0.0207 0.3251
Standard errors of the mean are given in parentheses.
2 Mean indicates the treatment mean.
3Means for a given treatment followed by the same lowercase letter are not significantly
different at the 0.05 level of probability.
4 p value indicates the significance between treatment means.









Crop alley


2 3 4 5 6 7 8
* .
* S S S S
* U S B S S
I-7-- 1--C-^-1_r-1-- 7-^-2-"^""l-----
I


I---
* SI S


* .
I
I *

I I *

S I H I
S I I *
M H *

S *


K'


2.60 m

2.60 m


STarget pecan tree

S15N (NH4)2SO4

Regular (NH4)2SO4


Figure 4-1. Plot layout showing location of 15N-enriched microplots at the Jay, Florida
agroforestry site.


Tree row


0.76 m


I'


Cotton rows






83



Soil N atom %16N


Barrier
0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48


0-15

15-30

S30-45


?" 45-60

S 60-75


75-90


90-105


--e Row 1
--- Row 4
-- Row 8


No Barrier
0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48


Soil N concentration (%)

Barrier No Barrier
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

0-15

15-30


E 30-45

S45-60

O 60-75

Row 1
75-90 R
-B-- Row 4
-- Row 8
90-105


Figure 4-2. Mean atom %"5N abundance and total %N in soil at end of growing season in
barrier and non-barrier treatments.













CHAPTER 5
SUMMARY AND CONCLUSION

This study has examined nitrogen (N) dynamics in a pecan-cotton alley cropping

system in the southern United States, specifically nitrogen mineralization patterns

(Chapter 2), groundwater nitrate dynamics (Chapter 3), and tree-crop competitive effects

(Chapter 4). Findings from these chapters are briefly summarized below.

In Chapter 2, we studied the effect of tree roots on nitrogen transformations in

soil. We observed that temporal variations in net ammonification, nitrification and

mineralization were driven primarily by environmental factors (e.g., soil moisture content

and soil temperature), and by initial ammonium and nitrate levels. However, these and

other factors appear to have exerted a combined influence on N transformations over the

study period. Mineralization varied by treatment during the first growing season, when

non-barrier exhibited a higher mineralization rate than barrier treatment. During the

, winter-season dormant period, however, mineralization in non-barrier treatment

experienced a significant decrease. Nitrification and mineralization rates in season two

were the same for both barrier and non-barrier treatments. Lint yield reductions were

observed in non-barrier treatment during both years compared to barrier treatment, likely

due to interspecific competition for water. However, for 2002, differences in yield were

more pronounced than the previous year, and are indicative of a short-lived fallow effect,

which appears to have enhanced 2001 crop yields but then diminished by the second crop

year. Source of N (chemical fertilizer vs. chicken manure) was found to have a

significant effect on cotton yield as well, with inorganic fertilizer resulting in moderately