...
Ian»...

Imam.

. «V...(.IOI
\t‘ .

 

 

2:..\\T
.5.
.4

. .
‘ .!:. ) ..
N 2.1.4.!

 

 

. nifhuhuu .
..v $5221
...

a A . . , .. 1 .
.... .. m5. {Li . ....5. .
. , , . ...;wtmwuthy: 3....muls , . . . 5.3.3:. grAfmmmw . , , gaagvnmsﬁtr

 

... .....n.........,. I}
.;o n . I». 75‘

 

W lllllllllllllllllllllllllllllllllllllu

3 1293 01410 1798

This is to certify that the

thesis entitled
Interaction Effects of Soil Nitrogen Levels and
Intercropped Maize (Zea Mays) on Nitrogen Fixation in and
Transfer From Black Locust (Robinia Pseudoacacia)
presented by

Michael Patrick Powers

has been accepted towards fulfillment
of the requirements for

Masters degree in Forestry

 

 

  
  

    
  

' l 0].-“ l '1‘" '-. a

/ Major professor

 

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

 

 

UBRARY
Michigan State
University

 

 

 

puce IN RETURN BOX io roman w. checkout iron. your rocord.

TO AVOID FINES Mum on or bdoro date duo.

 

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I ﬂmdémom
V.
‘Tsuhhm .. .4. — -- A —

 

INTERACTION EFFECTS OF SOIL NITROGEN LEVELS AND
INTERCROPPED MAIZE (ZEA MAYS) ON NITROGEN FIXATION IN AND
TRANSFER FROM BLACK LOCUST (ROBIN/A PSEUDOACACIA)

By

Michael Patrick Powers

A THESIS

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

MASTER OF SCIENCE

Department of Forestry

1 995

Dr. Douglas 0. Lantagne,
Committee Chair

ABSTRACT
INTERACTION EFFECTS OF SOIL NITROGEN LEVELS AND
INTERCROPPED MAIZE (ZEA MAYS) ON NITROGEN FIXATION IN AND
TRANSFER FROM BLACK LOCUST (ROBIN/A PSEUDOACACIA)
BY

Michael Patrick Powers

Inclusion of black locust (Robinia pseudoacacia) in a temperate
agroforestry system could beneﬁt the soil-plant system through nitrogen-ﬁxation
and transfer of this ﬁxed nitrogen to the intercrop. A greenhouse study was
designed to gather initial data on these processes A randomized complete block
design was used with four repetitions. Over three levels of applied liquid
nitrogen, 50, 125 and 200 kg ha'1 equivalent, black locust alone was compared
to black locust intercropped with maize. and maize alone was compared to
maize intercropped with black locust. Nitrogen-ﬁxation showed a dramatic
downward trend with increasing levels of added nitrogen, however ﬁxation was
not completely suppressed. lntercropping had no effect on nitrogen ﬁxation.
Estimates of nitrogen transfer were variable and sometimes negative.
lntercropped maize grew very poorly, presumably due to competition for
phosphorus and potassium or allelopathic effects of black locust. Results
indicated that black locust may well continue to ﬁx atmospheric nitrogen despite
high levels of added fertilizer nitrogen. However, strong competition for nutrients
or allelopathic properties of black locust may be factors to consider in temperate

agroforestry systems.

This research is dedicated to those who believe in the potential of agroforestry
to take advantage of complex ecosystem interrelationships and produce more
products from the same land, and do it in a socially and environmentally

harmonious manner.

 

ACKNOWLEDGMENTS

My parents deserve the ﬁrst and most important recognition. They have
supported me in anything and everything I have ever done, this long four year
masters included. Thanks Mom. Thanks Dad.

I owe most of what I learned to my advisor Dr. Doug Lantagne. His desire
that I explore and try everything related to intercropping and research meant that
I learned more than I expected. He always made time to put aside what he was
doing and help me. At times he worked harder on the thesis than I. Thanks and
respect also go out to my other committee members, Dr. Michael Gold and Dr.
Francis Pierce. Dr. Gold always had good observations, and l beneﬁted from his
world-wide knowledge and strong reputation in temperate agroforestry. Dr.
Pierce challenged me with new ideas each time we talked and he served to
greatly improve my thesis.

Dr. Phu Nguyen deserves special thanks. While not ofﬁcially serving on
my committee, he spent countless hours helping me with a_H my lab work and
counseling me in my outlook and research. He always cleaned off a chair for me
so we could talk no matter what he was doing. He is surely an invaluable

contributor to the Department of Forestry.

Thanks also go out to those awesome folks at the Tree Research Center
where I performed the greenhouse study: Paul (John Denver) Bloese, Randy
Klevickas, and Roy Prentice (formerly of the TRC). These three provided
tremendous practical and academic support for my work. I greatly enjoyed their
wonderful senses of humor. Paul and Randy currently serve as unparalleled
resources for the Department of Forestry.

Finally, I want to recognize and thank all the graduate students with whom
I worked, laughed, studied, and learned. They were a very special group,
combining academic excellence with a cooperative and easy—going spirit. Most
especially I want to thank Julia Parker, Mark Hare, Peggy Payne, and Jill
Huckins-Fisher for their often wacky humor entwined with a strong devotion to

people and the environment.

ﬁ

PREFACE

This thesis is divided into six chapters. The ﬁrst three include the
introduction, literature review, and materials and methods for the entire study.
The results of the study are then divided and discussed in the next two chapters,
N and V. Speciﬁcally, Chapter IV discusses results related to the impact of
maize on the nitrogen-ﬁxation of black locust. Chapter V reviews results on the
potential of nitrogen transfer from the black locust to the maize. The ﬁnal
chapter presents the study’s overall summary and conclusions with
recommendations for future research. Although one of the original goals for this
work was to study root growth, discussion is limited due to the poor growth of

maize intecropped with black locust.

vi

1?

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

List of Tables ix
List of Figures x
Chapter I. INTRODUCTION 1
Chapter II. LITERATURE REVIEW 4
Agroforestry 5
Below-ground Tree/Crop Interaction 7
Direct Nitrogen Transfer In Agroforestry Systems .................................. 11
Black Locust (Robinia pseudoacacia) 16
Nitrogen-Fixation Measurement Techniques 19
Conclusion 27
Chapter III. METHODS 29
Study Site 29
Experimental Design 29
Species Selection 31
Potting Soil 31
Container Construction 32
Fertilization Minus Nitrogen 33
Growth And Transplantation or Black Locust 33

vii

 

Method And Frequency Of Watering 34

 

 

 

 

 

15N Application 35
Maize Planting 36
Measurements and Sampling 36
Tissue And Sell Processing 39
Statistical Analysis 43

 

Chapter IV. 1‘t’N-LABELED FERTILIZER AND MAIZE INTERCROPPING
EFFECTS ON NITROGEN-FIXATION IN BLACK LOCUST

 

 

 

 

 

 

(ROBINIA PSEUDOA CACIA L.) 45
Abstract 45
Introduction 46
Methods 47
Results 49
Discussion and Summary 53

Chapter v. 1“N-LABELED NITROGEN FERTILIZER AND
INTERCROPPING EFFECTS ON NITROGEN TRANSFER
FROM BLACK LOCUST (ROBINIA PSEUDOACACIA L.)

 

 

 

 

 

 

TO MAIZE (ZEA MA YS) 58

Abstract 58
Introduction 59
Methods 61
Results 65
Discussion and Summary 72
Chapter IV. CONCLUSIONS 78

 

List of References 85

viii

Table 1:
Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:

Table 11:

LIST OF TABLES

Terms for 15N Isotope Dilution Technique 21
Generalized ANOVA Table 43

Average Maize Dry Weight and Leaves per Plant and
Average Black Locust Dry Weight 50

Average Black Locust N-Fixation as Measured by %

Biomass N Derived From The Atmosphere (%Ndfa) and mg
Biomass N Derived From The Atmosphere ................................... 63
Data Sets and Assumptions of ANOVA. 63
ANOVA for %""N Excess in Maize Above-Ground Biomass .......... 66
Black Locust Average Root Length Distribution for Two Soil

Depths as Measured in Soil Cores and Average Total Root
Weight and Average Live Nodule Weight 68

 

Selected p—values for Black Locust Root Length, Root
Weight, and Live Nodule Weight Responses to Plant and
Nitrogen Treatments 70

Average Pools of Soil %‘5N Excess and Available Soil N
by Soil Depth 45 Days After Nitrogen Treatments Were
Applied 71

 

Maize Average Root Length Distribution for Two Soil Depths
as Measured in Soil Cores and Average Total Root Weight ......... 71

Nitrogen Transfer from Black Locust to Maize .............................. 72

Figure 1:
Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6:

LIST OF FIGURES

Experimental Design and Arrangement

 

Average Black Locust Dry Weight

 

Average Black Locust N-Fixation as Measured by % Biomass
N Derived From The Atmosphere (%Ndfa) and mg Biomass
N Derived From The Atmosphere

 

Maize %'5N Ratio Excess in Above-Ground Biomass ..................

Black Locust Total Soil Core Root Length, Average Root
Weight, and Average Live Nodule Weight

 

Black Locust Root Length, Root Weight, and Live Nodule
Weight Compared to Maize Soil 15N Ratio Excess ........................

51

52

67

69

75

I. INTRODUCTION

Agroforestry is the term given to systems where agricultural, horticultural,
valuable wood crops, or animals are grown in the same temporal or spatial
context as woody perennials. Agroforestry systems strive to achieve more
production by optimizing the positive biological interactions and land—use
efﬁciency in woody-woody and woody-nonwoody plant combinations.
Incorporation of agroforestry systems into current land-use practice may: 1)
make more productive use of marginal lands or lands in conservation reserves,
including increased production of high value hardwoods; 2) contribute to
reduced soil erosion losses due to increased ground cover; 3) diversify rural
income sources; and 4) reduce nitrogen fertilizer inputs.

Two agroforestry systems appear feasible for Michigan:

o woody-woody intercropping of black locust (Robinia pseudoacacia)

with high value timber species; and

o alley cropping or multi-row windbreak systems in which food crops are

planted in alley-ways created by rows of woody legumes such as black
locust.

This study attempted to provide a measure of black locust's ability to
serve a positive biological role in temperate agroforestry systems. The nitrogen-
ﬁxing ability and fast growth of black locust make it a prime candidate as a
temperate agroforestry tree. Research indicates that nitrogen-ﬁxing woody

species have proven ability for use in forest or agroforestry management

systems in ways similar to that of legumes in agriculture. Legumes ﬁx nitrogen
from the atmosphere and thus reduce the need for nitrogen fertilizer. They
improve soil nitrogen conditions for crops grown with the legume or planted later
in the same ﬁeld. Nitrogen-ﬁxing trees and shrubs also may improve the nutrient
cycling in a ﬁeld. The deep roots of perennials access and take up nutrients
leached beyond the root zones of annual crops. The nitrogen-ﬁxing woody
plants return these nutrients to the surface soil through leaf and stem loss as
well as root turnover and rhizodepositions. It is a widespread practice in many
parts of the world where low-input agriculture is common to plant food crops
simultaneously with perennial woody legumes - often called intercropping or
agroforestry when perennials are used - to take advantage of their nutrient
cycling and nitrogen-ﬁxation characteristics. The soil nitrogen balance is
thereby improved using little to no chemical inputs.

As scientiﬁc and public pressures have demanded more sustainable land-
use systems that also decrease inputs and costs, interest in agroforestry as a
potential altemative to current systems has increased in the United States.
There are many questions to be answered, including what species to use and
what are their potential roles in highly mechanized, high-input agricultural
systems.

In order to study the effect of leguminous plants on the soil-plant N
budget, it is necessary to determine how much nitrogen is being added to the

system. Several techniques have been developed, but the Isotope Dilution

Technique (IDT) is considered the most accurate in terms of its cost and ease of
use. As almost all atmospheric nitrogen is “N and only 0.3663% of atmospheric
N is 1"‘N, this method introduces a source of nitrogen with a higher than normal
1"N percentage. Through analysis of 1"’N:“N ratios, the IDT differentiates
nitrogen ﬁxed from the atmosphere from that already in the soil and that added
as fertilizer N. In this way it is possible to measure the ﬁxed nitrogen that a
legume contributes to a soil-plant system.

Robinia pseudoacacia L. (black locust), a leguminous tree native to the
United States and Michigan, was studied for its multiple uses and promise as a
temperate agroforestry woody perennial. Maize was selected as the intercrop
for its common use and importance in Michigan agriculture.

The study’s hypotheses were:

1. A greater dilution of 15N will occur in the black locust planted with

maize than when black locust is planted alone.

2. The weight of nodules on black locust roots will increase when planted

with maize.

3. A greater dilution of 1"N will also occur in maize planted with black

locust.

4. Total maize root length is reduced when planted with black locust.

I. LITERATURE REVIEW

Introduction

Negative environmental effects and subsequent public outcry directed
toward certain forestry and agricultural practices have increased interest in
developing alternative sustainable production systems in the United States.
Alternative systems would directly embrace the philosophy of resource
stewardship over that of short-term production concerns. Some of the goals of
these new production systems are to decrease farm chemical applications,
shame decrease erosion of our agricultural and forest soils, and increase
agricultural biological diversity. By modifying typical monoculture environments,
farmers may also be able to exploit beneﬁcial plant and animal interactions,
perhaps decreasing reliance on fertilizers and pesticides (Altieri and Leibman,
1986). Diversifying the farm ecology could mean the farmer is producing more
than one product at a time, potentially diversifying farmer income as well
(Arnold, 1987; Gruggenheim and Spears; 1991; Spears, 1987). Research is
currently focused on developing appropriate land-use systems that increase
diversity, build soil resources, and maintain agricultural production. Agroforestry
is a multifaceted, adaptable land-use system that may be able to ﬁll important
roles in North American sustainable production systems. Black locust (Robinia
pseudoacacia L.), a native nitrogen-ﬁxing, fast growing tree, is being studied to

determine its potential role in North American agroforestry.

Agroforestry

Agroforestry is the term given to systems where there are agricultural,
horticultural, valuable woody crops, or animals grown in the same temporal or
spatial context as woody perennials. Agroforestry systems attempt to achieve
more production by: (1) optimizing the positive interactions between the two
plant components of the system, and (2) diversifying output for the land in an
ecologically and socially sustainable way. Incorporation of agroforestry systems
into current land-use practices may also:

0 make more productive use of marginal lands;

0 diversify income sources; and

0 add socioeconomic beneﬁts from revitalization Of rural areas through

production of nontraditional agricultural products.

Some of the more speciﬁc beneﬁts that agroforestry can bring to agricultural
systems include soil erosion control, micro-climate stabilization, water inﬁltration
enhancement, soil moisture retention, inputs to soil organic matter, better light-
use efﬁciency, pest control through increased beneﬁcial insect populations, and
promotion of nutrient cycling (Altieri and Leibman, 1986; Dawson and Pasche,
1991; Jackson et al., 1989; Kang and Wilson, 1987; Giller and Wilson, 1991;
Risch et al., 1983; Sanchez, 1987; Young, 1985).

Agroforestry research indicates that nitrogen-ﬁxing trees (NFTs) can be
used in agroforestry management systems in ways similar to that of legumes in

agriculture (Gordon and Dawson, 1979; Mishra, 1986; McIntyre and Jeffries,

1932; Young, 1989b). In addition, deeper rooted trees and shrubs access
nutrients from soil depths unreachable by annual crop root systems. These
nutrients are returned to the surface soil through leaf and stem abscission and
resulting decomposition (Huxley, 1983). Decreased erosion and soil surface
run-off helps maintain soil N, and improves fertilizer efﬁciency. NFTs and
nitrogen-ﬁxing shmbs are preferred because they biologically ﬁx atmospheric N
and add it to the soil-plant system (Dawson and Paschke, 1991; Khosla and
Toly, 1986; Mulongoy, 1986). Many legumes ﬁx N by forming a symbiotic
relationship with soil bacteria such as Rhizobium. These microorganisms
penetrate the legume roots, exploiting soluble plant carbohydrates to grow and
form nodules. In return they ﬁx N from the atmosphere and makes it available to
the legume (Devlin and Witham, 1983).

Although agroforestry is a relatively new idea to North America, the
inclusion of trees in agricultural systems is not new in many other parts of the
world, and a large and growing body of literature exists covering tropical
agroforestry (Steppler and Nair, 1987). Neither is agroforestry new in the
temperate zones outside North America. In the temperate regions of the
Peoples Republic of China, for example, extensive agroforestry systems exist
that integrate tree (Paulowina/Populus) and wheat production as well as other
agricultural commodities (Chirko, 1993). Farmers in the northern regions of
India intercrop trees with wheat and other cereal crops (Dawson and Pasche,

1991; Sharrna and Geyer, 1990), and a widely used agroforestry system in New

Zealand grazes cattle under fast—growing, managed pine (Pinus radiate) (Gold
and Hanover, 1987).

Though agroforestry in North America is still new, it is becoming more
established. Several national and international conferences on or related to
these subjects have occurred recently (Garret, 1991; Hanover et al., 1992;
Schultz and Colletti, 1994; Williams, 1991 ). Research and farmer experience
indicate that various agroforestry management systems appear feasible for
North America, including Michigan, and include: (1) intercrops of NFTs such as
black locust (Robinia pseudoacacia L.) or autumn olive (Elaeagnus umbellata)
with crops, pasture, or high value timber trees; (2) live fences that provide
forage as well as inexpensive fences or fence posts; (3) windbreaks and
snowdrift fences that help prevent wind erosion and drifting snow; and (4)
placement of food crops, omamentals, or Christmas trees in alley-ways created

by rows or hedgerows of woody legumes or high value timber species.

Below-ground TreeICrop Interaction

As described earlier, agroforestry is a land-use system that attempts to
achieve more production by optimizing the positive interactions between two
plant components of the system. Below-ground interactions are less studied
than those above-ground, yet they may have more inﬂuence on the system (Ong
et al., 1991). The use of deeper rooted NFTs may increase the total soil volume

used by the agroforestry system and improve the potential for higher production.

Conversely, the beneﬁts of NFTs may be negligible, and harmful conﬂicts may
occur, if mating densities increase in the intercrop’s rooting zone (Huxley, 1983;
Jonsson et al., 1988; Ong et al., 1991; Szott et al., 1991; Young, 1989a), or if
allelopathic incompatibility is found. Tree/tree or tree/crop competition for
moisture and nutrients can result in yield and biomass reductions (Buck, 1986;
Chirko, 1993; Dawson and Pasche, 1991; Huxley, 1983; Rachie, 1983).

One study serves as an interesting example of the competition problems
and paradoxes that agroforestry systems can present. Willey and Reddy (1981)
performed a millet and groundnut experiment that included a sole cropping of
each species, an intercropping/control treatment, and an intercropping with a
polythene below-ground barrier between the intercrops. Total intercropping
yields were better than either sole crop, presumably due to an improvement in
light-use efﬁciency. There was a decrease in millet production in the barrier
intercropping treatment versus the control intercropping while groundnut
performed better in the barrier intercropping compared to the control
intercropping. This suggests a positive below-ground impact of the millet on the
groundnut, but a negative below-ground impact of the groundnut on the millet.
Below-ground impacts, then, are important in order to understand the ecology of
agroforestry systems, and effects on production, yet little research has been
aimed at them.

Soil moisture is an important below-ground factor in any agricultural

system. Roots of intercropped trees can beneﬁt plant/soil moisture relations

through deep root systems that create macropores, thereby enhancing water
inﬁltration and preventing soil compaction. This same root development can
also reduce soil erosion and hence improve soil moisture retention (Gish and
Jury, 1983). Tree roots may also reduce soil moisture to associated crops
through competition from surface roots (Szott at al., 1991; Young, 1989a). In an
agroforestry soil management study on a tropical alﬁsol, both Leucaena- and
Glin'cidie-based systems showed greater available water capacity over no-till and
plow-till no-tree systems. While moisture retention declined over time in all
cropping treatments, the agroforestry systems showed the highest moisture
retention (Lal, 1989a). Cumulative water inﬁltration was greatest for the plow-till
treatment, but was followed closely by the Leucaena and then by the Gliricidia
systems (Lal, 1989b). Alley cropping with perennial hedgerows of Cal/landra
calothyrsus (calliandra) and Glin'cidia sepium (gliricidia) on a tropical lnceptisol
in Western Samoa showed improved water holding capacity over the no-tree
control (Rosecrance et al., 1992). Conversely, many species such as Leucaena
can ﬁercely compete for water with intercrops. When a soil partition was placed
between Leucaena and cowpea and sorghum intercrops, yields of cowpea and
sorghum increased versus the no-barrier treatment (Singh et al., 1989). In a
similar study, the soil barrier between a Leucaena Ieucocephala (leucaena)
hedgerow and a millet intercrop virtually eliminated the loss in millet dry matter
production due to root competition for soil moisture (Ong et al., 1991). A note of

caution is appropriate here: it can be difﬁcult to obtain accurate measurements

 

10

in ﬁeld root competition work using soil barriers. Studies show that roots from
tree-only and tree-intercrop plots have invaded no-tree control treatment plots,
even vmen the soil barriers were present (Hauser and Gichuru, 1994).

The literature appears to support the potential for agroforestry systems to
promote nutrient cycling, another important feature of below-ground interactions.
Tree roots are capable of absorbing nutrients from soil too deep for annual crops
to access, and then returning these nutrients to the soil surface through leaf and
stem abscission as well as through ﬁne root turnover in the rooting zone of the
crop (Young, 1989a). NFTs are also important for their contribution of
atmospheric N to the soil/plant system. Transfer of nutrients via direct root
contact and the hyphae of root mycorrhizae holds potential (Giller and Wilson,
1991; Robinson, 1991; van Noordwijk and Dommergues, 1990). Ewel et al.
(1982) compared root to leaf biomass ratios in three agriculture systems, two
forest systems, and four agroforestry systems in Costa Rica and Mexico. Three
of the four agroforestry systems had greater rootzleaf ratios than the other ﬁve
systems, and all had greater absolute root biomass, representing potentially
large nutrient and organic matter sources as roots turnover.

While the nutrient cycling potential of agroforestry may be great, tree root
distribution and competition for soil nutrients may be costly in terms of intercrop
biomass and yield losses. Competition seems especially likely for more mobile
nutrients such as nitrate N (Gillespie, 1989; Grime at al., 1991). Though it is

often speculated that trees have deep root systems, they can also have root

11

distributions similar to annual crops, as found in one study of ﬁve potential
agroforestry tree species (Jonsson, 1988). In a similar study, root distribution
patterns of four tree species were measured. Alchomea oordifolia, Cassia
siamea, and Gmelina arborea had 73%, 76%, and 74% of their ﬁne roots,
respectively, in the top 20 cm of an acid Ultisol (Ruhigwa et al., 1992). These
root systems, therefore, show potential for nutrient competition with food crops.
In contrast, the fourth species, Acacia barten', had only 49% of its ﬁne roots in
the surface soil. Furthermore, the A. batten” root system was concentrated
around its trunk and had deeper roots than the other three species. Central to
efﬁcient nutrient use between the annual and perennial species is the selection
of species with complementary root behavior patterns, but more research is

needed (Huck, 1983).

Direct Nitrogen Transfer In Agroforestry Systems
A potentially important beneﬁt of below-ground tree/crop interaction is N
transfer to the nonlegume. There are three main pathways of N transfer
considered to be present in agroforestry systems:
o decomposition of fallen above-ground biomass and dead root material
(Agboola et al., 1972; Mulongoy, 1986);
o mycorrhizal hyphal interconnections between legume and nonlegume

root systems (Giller and Wilson, 1991; Robinson, 1991) and

12

0 direct contact of legume and nonlegume roots (van Noordwijk and
Dommergues, 1990; Young, 1989b).

In the absence of manure from grazing animals, decomposition of dead leaves
and stem is considered the major pathway for N transfer from legumes to
nonlegumes in agroforestry and intercropping systems, with little direct transfer
between root systems (Catchpoole and Blair, 1990; Giller et al., 1991; Giller and
\Mlson, 1991; Henzell and Vallis, 1977). Reliable data on root N contributions to
soil and the intercrop are lacking (Giller et al., 1991; Giller and Wilson, 1991), in
part due to the unreliable research methods (Chalk, 1985).

Nodulation and resulting N-ﬁxation are strongly inﬂuenced by below-
ground interactions. They are generally greater in low N soils and decrease with
added fertilizer N. Nodule response can be affected on three levels:

0 complete suppression;

0 reduction in N-ﬁxation as well as mass or number of nodules; or

0 reduction or inhibition of nitrogenase activity of mature nodules (Giller

and Wilson, 1991; Heckmann and Drevon, 1987).
In a greenhouse study with treatments of 220, 22, and 0 mg l'1 nitrate-N fertilizer
applied to sainfoin and lucem, N-ﬁxation and nodulation were both suppressed
at the high N treatment. There were no differences between the control and the
low N treatment (Sheehy and McNeill, 1987).
There is also little research showing the effects of intercropping and root

competition on legume N-ﬁxation and nodulation. Van Noordwijk and

13

Dommergues (1990) postulate that roots of N ﬁxing trees have more nodules,
where N-ﬁxation takes place, when they are in close contact with roots of non-N
ﬁxing plants. This increased N-ﬁxation and nodulation may lead to the direct
transfer of N to the nonnodulating plant. General results point to an increase in
N-ﬁxation and nodulation with intercropping. Using the 15N isotope dilution
technique (IDT), N-ﬁxation in ricebean (Vigna unbellata) intercropped with maize
increased to 97 kg N ha" from 62 kg N ha'1 in the monocropped ricebean.
Further, the proportion of N derived from N-ﬁxation increased from 72% to 90%.
The authors believe that the increase in %N from N-ﬁxation in rice bean was due
to competition and depletion of soil N by the maize, thus increasing rice bean N-
ﬁxation and nodulation (Rerkasem and Rerkasem, 1988). However, caution
must be used when interpreting results from isotope dilution studies as the
differences in the 15N enrichment used to determine the amount of N-ﬁxation can
vary awarding to method and different 15N uptake patterns between the legume
and nonlegume crops (Giller and Wilson, 1991). In addition, above-ground
effects such as shading can decrease nodulation and N-ﬁxation (Nambiar et al.,
1983).

Of current N transfer studies, most have been done using legume—grass
swards, and estimates of N transfer in these systems vary greatly. No transfer
was measured from legumes to grass by Henzell and Vallis (1977), yet in other
studies 20% to 50% of legume N was transferred (Burity et al., 1989; Ta and

Faris, 1987). Using the 1"‘N IDT, Broadbent et al. (1982) found no transfer from

14

clover (Tn'folium replies L.) to ryegrass (Lolium n'gidum L.) in a greenhouse
experiment, but measured a transfer up to 79% of legume N in a ﬁeld trial.
Perhaps there was a time delay: the greenhouse experiment ran for two or four
cuttings (exact time duration not given) while the ﬁeld experiment ran for two
years. Similar contrasting ﬁeld and greenhouse results were found by Ledgard
et al. (1985) between clover (Trifolium subtgerraneum L.) and annual ryegrass
(Lolium n'gidum Groud).

In the case of cereals such as corn, there are few deﬁnitive studies
showing transfer of legume N without additions of above-ground biomass. No
such studies have been performed under ﬁeld conditions (Giller and Wilson,
1991; Ta et al., 1989). Searle et al. (1981) found greater maize grain yield and
residual soil N when maize was intercropped with peanut (Arachis hypogaea L.)
and soybean (Glycine max L.), but apparently no contribution of ﬁxed N from the
legumes to the maize. One attempt to measure N transfer to a cereal resulted in
a series of greenhouse experiments using perilite and montmorilonite plant
mediums, and compared the ‘sN-foliar feeding technique against the 15N IDT.
With foliar feeding, Phaseolus been transferred less than 5% of its N to the
maize, and no N transfer was measured using the IDT (Giller et al., 1991).
There was also no proof of N transfer between maize and Phaseolus in either
greenhouse or ﬁeld trials utilizing the "’N IDT by Reeves (1991). Thus, there is
currently little evidence that direct transfer of N from legumes is generally

important for cereal intercrops.

 

15

After a ﬁeld trial found no signiﬁcant N transfer to a grass (Panicum
maximum c.v. Riversdale) from the leguminous trees Calliandra calothyrsus,
Gliricidia sepium, Leucaena Ieucocephala cv. Cunningham, or Sesbania
grandiﬂora, researchers measured N transfer in a greenhouse using the split-
root technique. Leucaena and Gliricidia trees transferred 4.1% of labeled N
after six weeks and up to 7.6% of labeled N to the intercropped Panicum
maximum after 12 weeks in a 1:1 :1 sand, silt, and clay medium (Catchpoole and
Blair, 1990). There are few other reports investigating N transfer from tree roots
to associated crops (van Noordwijk and Dommergues, 1990) as researchers
have only recently begun testing NFTs.

The nutrient competed for the most is soil N, so the potential of direct
transfer of N from NFTs to associated crops deserves more research with more
species over a wider range of environments and plant combinations.
Researchers must begin to systematically study the role played by NFT roots in
agroforestry systems to control the positive and negative aspects of tree/crop
root interaction (Buck, 1986; Gillespie, 1989; Huck, 1983; Szott etaL, 1991; van
Noordwijk and Dommergues, 1990; Young 1989b), including allelopathy

(Putnam and Duke, 1978).

16

Black Locust (Robinia pseudoacacia L.)

The introduction of NFTs into agricultural systems means we need to
understand the below—ground interactions and the role of roots in N transfer.
Primary to this understanding is selecting potential NFTs with complementary
root dynamics for use in North American agroforestry systems.

Black locust (Robinia pseudoacacia L.) is a prime candidate for
consideration in temperate agroforestry systems. Dawson and Pasche (1991)
calle NFT "the only important nitrogen-ﬁxing tree adapted to cool-temperate
climates.“ Economically, black locust is used for lumber, poles, wood ﬁber/pulp,
fuel, beekeeping, railroad ties, mine timbers, and more (Barrett and Hanover,
1991). Physically, its root system is characterized by a tap root with a broad
lateral root system close to the soil surface that is normally about 1.0 to 1.5
times tree height. It can also produce a deep tap root in arid regions that allows
it to evade droughts (Barrett and Hanover, 1991; Huntley, 1990). Physiological
characteristics include its intolerance of competition and shade in natural
systems (Huntley, 1990), although it is used successfully in agroforestry systems
(Shanna and Geyer, 1990). It is tolerant of both acid and calcareous soils, and
is host to a wide variety of common rhizobial strains, making ﬁeld inoculation
usually unnecessary. In addition, black locust has a high rate of photosynthesis
and Is capable of supporting high rates of N-ﬁxation (Mebrahtu and Hanover,
1991), although estimates vary on its ability to improve soil fertility. Reports of

soil N accretion range from 30-40 kg ha" in natural North American stands

17

(Boring and Swank, 1984; Ike and Stone, 1958; Reinsvold and Pope, 1985) to
109 kg ha" in a dense planting of 2% year-old black locust (Dawson et al.,
1982).

Died locust response to fertilizers is typical of legumes, with nodulation
and N—ﬁxation decreasing with increasing levels of N (Dawson et al., 1992;
Johnson, 1992). Using black locust seedlings, Johnson and Bongarten (1991)
demonstrated that nodule biomass and N-ﬁxation, as measured by acetylene
reduction, were lower in the 5.0 mM nitrate treatment than in the 0.0 or 1.0 mM
treatments. Black locust seedlings grown in sand for 50 days treated with a
plus-N treatment (5 mM NH4N03) and a minus-N treatment (no N) showed
decreased nodulation and lower N-ﬁxation rates in the plus-N treatment as
measured by acetylene reduction (Roberts et al., 1983). A greenhouse study
using a soil, sand, and vermiculite planting medium was conducted to determine
the combined inﬂuence of different levels of added nitrogen (0, 25, 50, and 100
mg N kg" soil as nitrate and ammonium) and phosphorous on nodulation and N-
ﬁxation (Reinsvold and Pope, 1987). Low nodule weight in the low N treatment
was attributed to a lack of starter N and therefore delayed physiological
development at the time of measurement. Nodule dry weight and number of
nodules were inhibited at the high and low N treatments, in contrast to the
middle N treatments, though the results were not signiﬁcant. N-ﬁxation as
measued by acetylene reduction assay was greater in the middle treatments

than in the high N treatment, though again the differences were not signiﬁcant.

18

Research on the effects of black locust in US. agroforestry is still new. In
a greenhouse pot experiment, Ntayombya and Gordon (1993) found that
intercropping black locust with barley decreased barley yield compared to the
sole crop. However, total dry matter production increased and soil fertility
improved with the black locust mulch treatment. In separate one year ﬁeld
experiments, two studies found that an alley cropping system with black locust
hedgerows and corn alleys lowered total corn yield on a per hectare basis.
However, these lowered yields were attributed to decreased corn plant
populations in the agroforestry system (Seiter, 1994; Ssekabembe and
Henderlong, 1991). Ssekambe and Henderlong (1991) used soil partitions to
investigate root interaction between the corn and black locust. They found less
corn total root length and better yields in the corn row closest to the black locust
hedgerow in the soil partition treatment. Reasons for this were unclear as they
could be due to higher soil moisture under the hedgerow, allelopathy, or a
beneﬁcial N effect of the black locust roots.

The multiple economic and biological uses of black locust may make it a
viable option for farmers who are seeking to diversify their income and the
ecology of their ﬁelds. However, the disadvantages of black locust must also be
considered when designing research programs and when incorporating it into
agricultural practices. Black locust has sharp thorns that can make management
more difﬁcult, though they will not puncture shoes and tires. Insect pests are

common in pure stands, though not necessarily in less dense systems of

19

intercropping. Its broad and shallow lateral root system indicates a strong ability
to compete for nutrients and moisture. Finally, black locust has rapid growth
from root suckers and heavy seed production that may allow it to escape
cultivation and become a “weed” (Barrett and Hanover, 1991; Dawson and
Pasche, 1991).

Proper and attentive management of black locust will be key in any
agroforestry system that uses it. Such systems will require trained individuals
(farmers or professionals) not only in tree management, but also in physiology,
pest identiﬁcation and control, and species selection. For farmers and foresters
in the United States, this requires exploration and acceptance of using trees in
nonconventional ways (Gold and Hanover, 1987).

Much research remains to be completed before conﬁrming the role of
black locust in temperate agroforestry. To date, there are no N transfer studies
involving black locust. No studies have been found on the effect of root
competition and intercropping on nodulation and N-ﬁxation in black locust. This
study attempts to address the potential of N transfer from black locust as well as

effects of intercropping on nodulation and N-ﬁxation.

Nitrogen-Fixation Measurement Techniques
Various methods exist to measure the symbiotically fixed atmospheric N2
gas (ﬁxed N) in legumes. The method used in this experiment is often called the

1"N Isotope Dilution Technique (IDT). The IDT differentiates the isotopic

20

composition in the sources of N available for plant growth: soil N, fertilizer N,
and atmospheric N (Chalk, 1985). Special terminology is required to explain this
method (Table 1).

The amount of N in soil or biomass that contains the 19N stable isotope is
often expressed as the proportion of 15N present — %‘5N. Atmospheric N is
almost all “N;, with 15N2 making up only 0.3663%. Any material that has a %""N
greater than that in the atmosphere is said to be enriched with 1”N. A 15N
enrichment above the natural abundance of 0.3663% is the %"‘N excess (Giller

and Wilson, 1991) or 1"’N ratio excess.

An enriched fertilizer can be added to the soil, thus labeling the soil with
15N. IDT experiments performed in soil, versus a N-free media, must then
attempt to discriminate between N derived from the soil, fertilizer, and
atmosphere. A non-ﬁxing system (nfs) reference plant is used to estimate the
“N enrichment of the available soil N (Table 1). This requires that the ﬁxing
system (fs) and nfs both absorb N from soil with the same 15N enrichment (Giller
and Wilson, 1991). The IDT is considered by many researchers to be the most
useful and comparatively accurate method for measuring N-ﬁxation. Only
techniques employing the principle of 15N IDT yield a reproducible and precise
estimate of N-ﬁxation integrated overtime under ﬁeld conditions (Rennie, 1985;
Ta et al., 1989). The IDT requires that 15N be uniformly applied to the soil

(Rennie, 1986). This is vital so the fs and nfs roots grow in soil of very similar

 

Table 1:

21

Tom for 1‘N Isotope Dilution Technique

 

Term

Definition

 

 

atom % 15N (or
%15N)

natural abundance
of the atmosphere

atom %‘5N excess
(or %"N excess)

"‘N-enriched nitrogen
(or ”N-enriched)
Nitrogen-ﬁxing system

, Non-nitrogen-ﬁxing
system

% of N derived from
the atmosphere

% of N derived from
the fertilizer

% of N derived from
the soil

the relative amount of 15N atoms
versus 1‘N atoms as a percentage of
the total.

(Giller & Wilson, 1991)

the atom %"N of the atmosphere
= 0.3663 %‘5N
(Giller & Wilson, 1991)

the percentage of 15N above 0.3663
= %“N - 0.3663
(Giller & Wilson, 1991)

nitrogen with a %‘N above 0.3663
(Giller & Wilson, 1991)

fs
nfs
o 15
%Ndfa=1—[————’f’, Nexcess(fs) x100)
% Nexcess(nfs)

(Rennie, 1986)

%‘5 N excess (plant)
%15 N excess (fertilizer)
(Rennie, 1986)

 

%Ndff =

%Nde of nfs = 100 - %Ndff
offs = 100 - (%Ndff + %Ndfa)
(Rennie, 1986)

 

 

 

22

“N enrichments (Danso, 1988; Giller and Wilson, 1991; Rennie, 1985; Vose et
al., 1982). Liquid fertilizer application of 1"’N is believed to be the most precise
method (Vose et al., 1982). However, not all researchers agree and
experiments have shown signiﬁcant variation in IDT studies due to labeling
technique (Chalk, 1985; Danso, 1988). Witty and Ritz (1984) showed conﬂicting
results for reasons that are unclear. Rennie (1986) compared applications of 15N
enriched liquid fertilizer and applications of 15N enriched legume and cereal
organic residues that theoretically might provide a more consistent source of 15N
over time. No differences between these two techniques were found. Ta ef al.
(1988) found no differences in their work comparing direct 1"’N; exposure, foliar
labeling of 1"’N, and 15N isotope dilution technique.

The discussion of labeling technique is centered around the best method
to obtain an accurate estimate of maximum N-ﬁxation and allow comparison
between legume species and varieties. This experiment, however, is not
designed to estimate maximum N-ﬁxation in black locust but to Observe its
response to added N fertilizer and intercropping, conditions that might be
present in an agroforestry system. Thus liquid labeled fertilizer was utilized in
this study based on the existence of supporting literature and the desire to
observe the interaction between added N fertilizer and intercropping on N-
ﬁxation and transfer.

The choice of the nfs reference plant is critical to the accuracy of the IDT

(Danso, 1988; Giller and Wilson, 1991; Rennie and Rennie, 1983), regardless of

23

the labeling method. It is difﬁcult to ensure that 15N uptake is not affected by
variation in rooting depth, or in the timing of N uptake (Giller and Wilson, 1991).
As the concentration of a fertilizer decreases with greater soil depth, so does the
enrichment of 1"’N. Deep and shallow rooted plants, then, would sample soil N of
different 1”N enrichments, confounding any results. For example, if a legume (fs)
has a deeper rooting pattern than its reference (nfs) plant, it would have access
to a soil with a lower %‘5N. Accessing soil of lower %‘5N not available to the nfs
leads to the overestimate of N-ﬁxation (Giller and Wilson, 1991). In other words,
the difference between the level of 1‘N and 1"N can be due to availability in the
soil, and not the legume’s capacity to absorb and ﬁx 1‘N from the atmosphere.

Another potential source of error is that 15N soil enrichment decreases as
soil microbes immobilize added mineral N over time. Therefore, different N
uptake patterns between the fs and nfs is of concern when using the IDT. For
example, if in the early stages of growth the nfs plant absorbs soil N more
rapidly, with the fs absorbing most of its N later, the nfs will accumulate a higher
"N enrichment. This would lead to an overestimate of N-ﬁxation. Conversely, if
the fs samples soil N more rapidly in the early stages of growth than the nfs, N-
ﬁxation estimates are underestimated (Giller and Wilson, 1991).

Finally, N-ﬁxation in trees is the most difﬁcult to estimate of all legume
types. Appropriate nfs reference plants are extremely difﬁcult to ﬁnd as sources
of variation such as different N uptake proﬁles and unlike rooting patterns are

magniﬁed (Giller and Wilson, 1991).

24

Other techniques to measure N-ﬁxation include indirect methods such as
the Acetylene Reduction Assay (ARA) and the A-Value Method. They are
indirect since they measure indicators of N-ﬁxation, but do not measure ﬁxation
itself. The ARA is an attempt to measure nitrogenase activity. Nitrogenase is a
complex of enzymes in legume nodules that reduce N; to NH; for plant use
(Devlin and Witham, 1983). In the ARA, acetylene (Csz) is reduced by
nitrogenase to ethylene (C2H4) at the exclusion of N2. The theoretical conversion
ratio of acetylene reduced to N2 ﬁxed is 4 moles 02H4 reduced : 1 mole N2 ﬁxed.
The ARA is generally performed by incubating the test material (roots and
nodules or a nodule) in an atmosphere containing 10% acetylene in a closed
vessel. The amount of ethylene produced after a period of incubation is then
measured by gas chromatography. The nitrogenase activity, or ARA, is then
expressed as uM C2H4 produced per plant (or nodule) per hour. The amount of
N2 ﬁxed is calculated using the theoretical conversion ratio of 4:1. One can also
calculate the actual conversion ratio using a direct measurement of the rate of
N-ﬁxation by incorporating 15N2 analysis in parallel with an ARA (Giller and
Vlfllson, 1991). The ARA is considered a sensitive, fast and inexpensive method
for measuring N-ﬁxation (LaRue and Patterson, 1981). However, the ARA
requires many samples for a good ﬁeld estimate, and it is a one-time estimate
that must be done immediately after harvest. Furthermore, there is much plant
to plant variation. Additional error is introduced through the uncertainty of

obtaining the entire root and all the nodules (Giller and Wilson, 1991; LaRue

 

25

and Patterson, 1981 ), as well as from the resulting plant disturbance that can
cause a marked reduction in the rate of nitrogenase activity (Minchon et al.,
1986). Moreover, the 4:1 theoretical conversion factor of moles C2H4 to moles
N2 ﬁxed can vary greatly, often ranging between 5:1 and 7.5:1 (Witty and
Minchon, 1988). The best estimates come from ‘ﬂow-thru' systems made on
undisturbed plants (Giller and Wilson, 1991), making this technique suitable only
for smaller pot studies.

A second indirect method for measuring N-ﬁxation is the A-Value Method.
It refers to the amount of available N in the soil in terms of a standard (Fried and
Dean, 1963). For an N-ﬁxing plant there are three sources of N: A(soil),
A(fertilizer), and A(ﬁxed). A(ﬁxed) is equal to the difference of A(soil + ﬁxed) -

A(soil). The actual amount of N ﬁxed can be calculated as:

N; ﬁxed (kg ha") = A(ﬁxed) x total N yield in crop
total N supply to crop

which then becomes:

N2 ﬁxed (kg ha") = [A-value (fs) - A-value (nfs)]

x ° dff ttal an k ha'1

fertilizer N applied (kg ha' ) to fs

(Vose et al., 1982).

An advantage of this technique is the ability to provide different rates of N

fertilizer to the fs and nfs (Vose et al., 1982), allowing a lower fertilizer rate to be

 

26

applied to the nfs in order to obtain an estimate close to maximum N-ﬁxation.
However, this method is yield dependent. In addition, A-values for reference
plants may vary with increasing rates of applied N; some workers have shown
decreasing A—values with increasing applied N (Chalk, 1985). This also
assumes that the addition of a large amount of N fertilizer to the nfs has no effect
on the availability of soil N, a questionable assumption (Witty, 1983). Finally,
the equation term for the rate of N fertilizer means that any loss of fertilizer N by
denitriﬁcation or volatilization will result in error (Vose et al., 1982).

In addition to IDT, a direct technique for measuring N-ﬁxation is the
Nitrogen Balance Method (NBM). In the NBM, the N yield of a fs and a nfs are

compared. The difference in the nfs, if greater, is attributed to N-ﬁxation.

N yield (fs)—N yield (nfs) x 100
N yield (fs)

%Ndfa =

 

(Rennie, 1985)

It provides a measure of the total amount of N ﬁxed over the length of an
experiment (Giller and Wilson, 1991). However, this method is yield dependent
and is based on the assumption that the fs and nfs assimilate identical amounts
of soil and fertilizer N (Rennie, 1985). Positive or negative values for N-ﬁxation
in the absence of nodulation are sometimes encountered when N-ﬂxation is
estimated using the NBM (Rennie, 1985), rendering its results rather

questionable.

27

Of all methods used to measure N-ﬁxation in legumes, measuring ‘5N2
ﬁxation is the most reliable and direct (Bremener, 1977; Giller and Wilson, 1991;
Vose et al., 1982). Plants are grown in an enriched atmosphere of “N2. After an
appropriate period of time, the plant biomass can be harvested and later
analyzed in a mass spectrometer to measure the %"N excess. In this manner
the amount of N-ﬁxation can be accurately determined for that speciﬁc period of
time. While using an enriched 15N2 atmosphere is the most accurate method
known, sophisticated incubation chambers are required, limiting this method to
small greenhouse studies and nearly eliminating the possibility of ﬁeld studies

(Vose at al., 1982).

Conclusion

Nitrogen-ﬁxing trees and shrubs beneﬁt agroforestry systems through soil
erosion control, micro-climate stabilization, water inﬁltration, pest control, and
promotion of nutrient cycling, especially of nitrogen. Major potential negative
impacts include competition for light, nutrients, and moisture. More research is
needed on below-ground root interactions, most notably on effects of
intercropping on legume nodulation, N-ﬁxation, and N transfer. Black locust
currently holds high potential for use as a temperate agroforestry system
component. However, we have only preliminary data on its potential impacts on
intercrops. We do not have data showing how its high nodulation and N-ﬁxation

are affected by intercropping. In part through use of the IDT, this study attempts

 

28

to provide some applicable date, and contribute towards the goal of determining

the role of black locust in North American agroforestry systems.

III. METHODS

Study Site
The study was located in the north greenhouse at the Michigan State
University (MSU) Department of Forestry Tree Research Center, East Lansing,

Michigan.

Experimental Design

A randomized complete block design (RCBD) with split-plots, four
replications, and three subsamples per experimental unit was used. Figure 1
illustrates the arrangement of the experimental design. The main plots were
plant treatments: one maize plant per container, one black locust plant per
container, and an intercropping of one maize and one black locust. The

following terminology is used to compare plant treatment results:

B -> black locust alone;

BM —) black locust in black locust + maize;
M —> maize alone;

MB —+ maize in black locust + maize.

The BM and MB abbreviations represent the same plant treatment. The ﬁrst

letter in each term denotes the plant being referred to in that plant treatment.

29

30

North
IV III II I 7
M BMIMB
B B
B WI
2
lg
BMIMB < BMIMB
BMIMB B
M M

Key: Roman numerals denote repetitions (blocks).

B a black locust main plot

M —> maize main plot

BMIMB -) black locust + maize main plot
N treatments randomly assigned within each main plot. Three subsamples per N
treatment are used for a total of nine containers per main plot.

Figure 1: Experimental Design and Arrangement

31

The term 'BM' is used to compare intercropped black locust (the “B' in BM) to B
(black locust alone). The term “MB” is used to compare intercropped maize (the
'M“ in MB) to M (maize alone). The split-plots were three nitrogen level

treatments: 50, 125 and 200 kilograms of nitrogen per hectare equivalent.

Species Selection

A black locust (Robinia pseudoacacia) clone # 8165.0879 was provided
by the Michigan Cooperative Tree Improvement Program (MICHCOTIP) located
in the MSU Department of Forestry. A clone was used to reduce experimental
variation. A maize hybrid (Pioneer #3751) was used as the cereal crop. This
hybrid was chosen for its common use in Michigan and its ability to scavenge N

(Brinkman, G. 1993).

Potting Soil

The potting soil was a ﬁne grain, mortar sand. The MSU Soil Testing
Service measured the pH to be 8.1, with calcium making up 97% of the soil
bases, or 1760 kg Ca ha" equivalent. The Cation Exchange Capacity (CEC)
ranged from 4.5 to 4.7 crnol kg". NOa' - N was measured at 0.90 mg kg", while

NH.’ - N was 3.52 mg kg", for a total N of 4.42 mg kg". The average bulk

32

density (80) was measured by packing air dried sand into 140 cm3 metal
containers that had been set in an oven for 48 hrs at 105° C. The potting soil
and metal containers were then placed in an oven for 24 hours at 105° C. B0
was then calculated as dry soil weight (g) I container volume (cm’). Two
containers were used and this was repeated twice for a total of four

measurements to calculate an average BD of 1.58 mg m".

Container Construction

The containers for this experiment were constructed by stacking four
plastic crates and securing them with 5 cm wide duct tape to form a single
container. The bottoms from each of the top three crates were removed before
stacking. The ﬁnal dimension of each container was 30 x 30 x 100 cm., for a
total volume of 90,000 cm3 (Figure 2). Heavy paper inserts (Monarch
Manufacturing, Salida, CO) within the containers added strength and stability. A
black plastic 1 mm x 0.5 mm knit mesh was placed at the bottom of each
container to allow drainage. Once the containers were constructed, they were
partially ﬁlled with sand and then placed on top of wooden pallets supported by
concrete blocks. Containers were then topped off with sand to within 3 to 4 cm
of the top of the container. Foil catchments were placed under the containers to
catch leachate, although no leachate was ever collected. The containers were

arranged in a checkerboard pattern to reduce shading (Figure 2).

33

FertilIzatIon Minus Nitrogen

To ensure an even distribution Of macro- and micro- nutrients, the top 25
cm of sand was removed and mixed with three liters (L) of fertilizer solution. The
solution was deionized water containing 7.81 g MgSOx-7H2 O, 3.56 g K2HPO4,
2.25 g K2804, and 1.89 g STEM micronutrient mix (Grace Sierra Horticultural
Products Company, Richardson, Texas) to provide an equivalent of 75 kg ha‘1
Mg, 140 kg ha" P205 and 300 kg ha" K20 on an area basis as well as
micronutrients S, B, Su, Fe, Mn, Mo, and Zn. Before replacement of the surface
soil, the remaining container soil was excavated an additional 15-20 cm and
compacted by hand along the sides of the container. This was done to favor
percolation through the center of the container when six liters of deionized water
were added to saturate the remaining soil. The goals were to bring each
container to ﬁeld capacity, maintain uniformity among containers, and further
reduce available N through leaching. After saturation, the fertilized surface soil
was placed back into the container. The compacted soil along the sides was not
altered before returning the fertilized top soil. This was an effort to discourage
root growth along the sides of the containers. Black locust seedlings were

transplanted to the containers two days later.

Growth And Transplantation Of Black Locust
The black locust clone was grown in styroblocks, each holding 30

seedlings in a basic greenhouse potting mix. These seedlings were pruned

 

34

back three times before transplanting to the containers. At transplanting,
seedlings were taken from the styroblocks and their length measured from the
fork below the terminal leaf down to the root collar. Seedlings were then root-
pruned to 12.5 cm and the root system washed in deionized water to remove the
potting mix and to minimize the amount of introduced N. Seedlings were then
randomly assigned to containers by blocks according to height. They were
transplanted on a diagonal 12.5 cm from the southeast comer of each container.
The combination of high greenhouse temperatures, pruned root systems, and
washed root systems resulted in the loss of all existing leaves. Seedlings were
then top—pruned to a uniform height of 10 cm. Six seedlings died and were
replaced, two having to be replaced a third time: one of those died and was not

replaced, leaving only two subsamples in that treatment combination.

Method And Frequency Of Watering

All plants were watered as needed with deionized water. Deionized water
was routinely sampled for nitrate-N and ammonium-N levels. All samples
stiowed less than 0.1 mg kg" total N as measured for NH4+ - N and N033 N on a
Technicon Autoanalyzer II System (Technicon, 1977 & 1978). Watering was
initially done by hand using a 2 L Dram plastic watering can with a very ﬁne

water nozzle. Later a trickle irrigation system was installed.

 

35

"N Application

Ninety-six days after transplanting, random samples consisting of ﬁve
leaﬂets were collected from each tree to provide a baseline for N and 15N levels
in the leaves. The containers were then fertilized with a solution containing
("NHa)2804. Procedures used to determine the amount of N to apply and
preparation of the labeled fertilizer came from Cabrera and Kissel (1989). The
fertilizer enrichment was calculated to produce a uniform soil enrichment of
”UN. This would produce accurate 15N tissue and soil results from the mass
spectrometer (Harris, 1993) . The 15N source in the fertilizer solution consisted
of 10 g 99% “N and 19.22 g 10% “N for a total of 29.22 g of 38.09% “N
enriched source of N as ('5NH.)2SO4. The enriched source was mixed with
1257.27 9 unenriched (NH4)2SO4 and deionized water to make a volume of 3.240
ml of 1"‘N enriched fertilizer concentrate. Ten ml Of this concentrate was
removed for storage and later analysis. Although the target enrichment was
1.20% 1"N, analysis indicated an enrichment of 1.272%. The concentrate, 12,
30, or 48 ml, was mixed with deionized water to a volume of 2L and then applied
to the appropriate randomly selected containers using a 2L Dram watering
container with a ﬁne stream nozzle. This resulted in an N application of 1.0, 2.5,
and 4.0 g of N to approximate 50, 125, and 200 kg N ha" equivalent,
respectively. The entire soil surface was saturated with the solution to provide
an even N distribution. The watering container was rinsed with deionized water

behween applications to avoid cross contamination.

Maize Planting

After fertilization with 15N, maize was planted on a diagonal 13 cm from
the northwest comer of each container. In the intercropped treatment, this was
15 cm from the black locust seedling. Centered at each planting spot was
placed a 5 x 5 x 10 cm planting band 5 cm into the sand. Sand was removed
from the square to a depth of 5 cm and four to ﬁve kernels of maize were then
plead inside and covered with dry sand and a paper bag. When above-ground
maize growth was observed, the bags were removed. At seven to eight days of
growth, the planting bands were carefully removed and the seedlings were

thinned leaving only the tallest.

Measurements and Sampling

Final maize height was measured from the top of the whirl to the soil
surface. Black locust height was taken from the root collar to the fork below the
terminal leaf (of the tallest stem if the seedling had two or more stems).

Diameter measurements were taken of the root collar using digital calipers.

rve ' : 9 ve— nd low round iomass soil
Above-ground biomass was collected from the entire experiment. Each
plant was cut at the root collar. In addition, maize plants were cut at the

midpoint between the top of the whirl and the root collar and the two halves

 

37

placed in separate paper bags. Black locust leaves (including the leaf stem) and

woody stem were harvested separately and placed in paper bags. All plant

biomass was dried for 24 hours at 65° c.

One of the three subsamples of each treatment in each repetition was
randomly selected for soil sampling and root collection. After the above-ground
biomass was harvested, a 5 cm x 5 cm inside diameter soil core was driven into
the center of selected containers to collect root samples for root length and
diameter measurements. Two soil depths were sampled: 0-25 cm and below 25
cm. These samples were placed in separate plastic zipper-type bags and
frozen. Next the containers were tipped onto their sides and the long top panel
of each container was removed. The panels were cut with little disruption to the
soil and the root system. The opened container allowed easy sampling of soil
and roots.

As part of the experiment, “test' containers were employed. Plant and N
treatments were applied to test containers so they could be destructively
sampled during the experiment to view root development. Harvesting of test
containers showed a 25-30 cm plow-like layer resulting from the fertilization
(minus N) procedures. There also appeared to be a wetting front at the 55-60
cm depth. Two soil samples were therefore collected with a standard 2 cm
inside diameter soil sampler from each of the following ranges: 0-25 cm, 25-55
cm, and below 55 cm. Samples were placed in labeled paper bags and allowed

to air dry. After soil sampling, root sub-samples were extracted by hand from the

 

38

entire container and bulked. Maize and black locust roots were carefully
separated by hand to avoid cross contamination. All root samples were placed
in paper bags and dried for 24 hours at 65°C and later analyzed for total N and
%“N.

After soil and root sub-samples were taken, the entire root system was
separated from the soil of each container. Using ﬁne stream nozzles and tap
water, the sand was carefully washed from the root systems onto a 3 mm x 3 mm
inside diameter mesh screen. Individual root systems, as in the maize or black
locust plant treatments, were easily separated from the soil. Separating the two
root systems in the black locust + maize plant treatment was difﬁcult. Main roots
were followed to the root collar for proper identiﬁcation, and ﬁne roots were
traced to the appropriate root collar when feasible. Softer, whiter roots were
attributed to maize, especially below the top 25 cm. Dark brown, coarse roots
were determined to be black locust. Maize root growth in the black locust +
maize treatment was poor, therefore roots in the bottom half of these containers
were conﬁdently identiﬁed as black locust. The one major potential source of
error was in separating the ﬁne roots between black locust and maize in the top

25 cm.

NW

Black locust root nodules were hand picked from each harvested root

system, total N and %‘5N root sub-samples, and length and diameter root sub-

 

39

samples. To determine the presence of nodule N-ﬁxation activity, nodules were
placed into one of two categories: dead or alive. Dead nodules were deﬁned as
having a dark brown or green exterior and/or interior. They were sometimes
flaccid or contained a green liquid. Live nodules were light brown on the outside
and red on the inside, due to the presence of oxygen-carrying leghemoglobin.
Nodules without leghemoglobin cannot ﬁx N (Devlin and Witham, 1983), so a
red nodule is considered a general indicator of N-ﬁxation. Large nodules with
multiple red tips that were green or brown at the nodule base were considered
dead. For root systems in block I, almost all nodules were opened to inspect the
inside color. With experience from block I, nodule activity could be accurately
determined without opening every nodule. Nodules were randomly Opened to
check for accuracy in succeeding blocks if any question of viability was
indicated. After evaluation, collected nodules were stored in paper bags and

refrigerated at 4°C.

Tissue And Soil Processing

All plant tissue (black locust leaves, black locust stem, upper and lower
maize stalks, and roots) were stored in ambient temperatures in paper bags after
drying. All plant tissue except for the entire root systems were dried again at 65°
C for 24 hours before weighing. The biomass samples were then ground in a
Thomas \Nllley Mill model 4 to pass a mesh screen and placed into zipper-type

plastic bags. These ground samples were ground again in a Cyclone Sample

 

40

Mill (Udy Corporation, Fort Collins, CO) to pass a 1 mm screen and stored in
glass vials. Plant N and %“N were then determined from subsamples using a
continuos-ﬂow stable isotOpe ratio mass spectrometer (ANCA-MS, Europe
Scientiﬁc). Subsamples were 4 mg to 10 mg in weight, with black locust leaf
subsamples weighing approximately 4 mg due to the high N content. Though
this method is accurate and comparable to conventional IQeIdahl procedures
(Harris and Paul, 1989; Jensen, 1991), a random sub-sample of 30% of the
above-ground biomass plant N readings were checked for accuracy by a Carlo
Erba Nitrogen Analyzer 1500 Series 2 (Fisions Instruments, Danvers, MA) using
the same weight of subsamples. To calculate N-ﬁxation and N transfer, formulas
for the Isotope Dilution Technique (IDT) were applied using maize as the
nonﬁxing reference plant. The biomass percentage of N derived from the

atmosphere (%Ndfa) was calculated as:

%Ndf a _ %‘5 Nexcess (fs)
- — —,—5—-———x100
% N excess (nfs)
where:
fs ﬁxing system — black locust

nfs non-ﬁxing system — maize

(Rennie, 1986)

 

41

To calculate root length and diameter, the soil cores were thawed In a
cooler and the roots separated from the sand using a hydropnuematic root
elutriator. Recovered roots were immediately placed in whirly pac bags
containing a 100 ml solution of deionized water and 20% alcohol. They were
then stored at 4°C. Black locust and maize roots from the intercropped
treatment were separated by hand. This proved difﬁcult. Observations of roots
from black locust alone and maize alone treatments indicated that black locust
ﬁne roots are not as transparent as maize ﬁne roots. Maize ﬁne roots have an
almost clear translucence. Light clearly goes through the cells to the steele.
Black locust ﬁne roots can have a visible steele as well, but light does not travel
through as well since the cells around the steele appear cloudier and more
ﬁbrous. In addition, the roots of black locust alone in the top 25 cm of the
container were primarily brown and ﬁbrous. Conversely, below 25 cm few maize
roots were located due to the poor growth of the maize intercropped with black
locust. After the root systems were separated, all roots were stained with
malachite green. They were than video taped and computer analyzed to
determine total sample root length and diameter (Smucker at al., 1987).

To determine root dry weight, root samples were ﬁrst manipulated to
remove soil particles. Roots were shaken over a #10 mesh screen and a #250
mesh screen. Fine roots that broke off from the root systems were then
collected from the screens. The roots were then dried for 24 hours at 65°C.

Three 1 g dry weight random samples (a mix of ﬁne and large roots) of each root

42

system were placed in preweighed crucibles. The samples were placed in ovens
at 550°F for four hours and allowed to cool overnight in the ovens. Crucibles
were reweighed and a correction factor determined to account for soil particles
attached to the roots in order to determine the true root dry weight.

To determine soil N content and %“N, 50 ml of 2M KCl was added to 20 g
of soil in a 100 ml ﬂaslc It was then mechanically shaken for one hour. The
resulting solution was poured through Whatrnan #1 ﬁlter paper that had been
rinsed with 0.1M HCI and placed in an oven at 50° C until dry. The resulting
extracts were analyzed colormetrically for NO;' - N and NH“ - N on a Alpkem
RFA 300 (Alpkem Corporation, Clackamas, OR, USA). Diffusions for total N
were performed according to the method of Brooks of al. (1992), with the
following exceptions:

0 ﬁlter papers were leached with 300 ml of deionized water as previous

trials showed this to be adequate for leaching ﬁlter papers of excess
N:

e a 5 ppm N spike as ammonium sulfate was used in the low N soil
extract samples to boost extract N to a range above 40 ug per sample
necessary for accurate measurements of total N and % 15N;

0 all samples were not shaken with glass beads at the beginning of the
six day incubation period but were swirled for 15 seconds each day;

0 reagent and spike N and 15N were accounted for in the ﬁnal N and 15N

estimates.

 

43

Percentage 15N of the diffusions was determined using a continues-ﬂow stable

isotope ratio mass spectrometer (ANCA-MS, Europe Scientiﬁc).

Statistical Analysis
All statistical analysis was performed using analysis of variance (ANOVA)

with linear, quadratic, and logical orthogonal contrasts. A generalized ANOVA
table used in the analysis is presented in Table 2. Assumptions of ANOVA were
tested using formulas provided by Little and Hills (1978). The assumptions
tested were:

0 Error terms are randomly, independently, and normally distributed;

0 The variances of different samples are homogenous; and

o Variances and means of different samples are not correlated.
The fourth assumption, additivity, was not tested as the effects in a design using

split-plots are not additive.

Table 2: Generalized ANOVA Table

 

 

Main Plots
Block (3)
Plant Treatment (PT)
Emma"..- ................................
Sub Plots
Main Plots
Nitrogen Treatment (N)
N r

 

 

 

44

To test the assumptions, 9 table of error terms was constructed for each
data set and analyzed for randomness, independence, and normal distribution.
Homogeneity of variances was tested using the Bartlett’s Test for Homogeneity
of Variances (Little and Hills, 1978). The correlation of variances and means
was determined by graphing each treatment variance and mean over the
replications. A visual inspection determined whether the means and variances
were correlated. If any of these assumptions were not satisﬁed, the data were
transformed using log(10). Some data sets still did not satisfy one or more of the
assumptions after transformation. Violations of the assumptions of ANOVA are

identiﬁed and potential impacts are discussed as results are presented.

IV. “N-LABELED FERTILIZER AND MAIZE INTERCROPPING EFFECTS
ON NITROGEN-FIXATION IN BLACK LOCUST (ROBINIA
PSEUDOACACIA L.)

Abstract
Black locust (Robinia pseudoacacia L.) is a temperate fast-growing,

leguminous tree. As part of the effort to investigate its potential role in North

American agroforestry systems, a two factor, split-plot greenhouse study was

conducted to determine the effect of nitrogen (N) fertilizer and maize

intercropping on the N-ﬁxation ability of black locust. Cloned black locust
seedlings from root cuttings were transplanted to containers of N deﬁcient sand
and grown for 98 days. N and plant treatments were then assigned. N as

('SNHshsoa was applied at 50, 125, and 200 kg ha" equivalent. Maize was

planted by seed to form the following plant treatments: black locust alone, black

locust and maize intercropping, and maize alone. These treatments were
applied for 45 days and the experiment was harvested.
N-ﬁxation was greatly reduced as N application increased (p < 0.001 ).

There was a 60% reduction between the low and medium N application levels,

and an additional 20% reduction from the medium to high levels. lntercropped

and sole black locust derived similar portions of biomass N from the atmosphere

(p = 0.108). There were no interaction effects (p = 0.854), indicating that the

plant treatments had similar response patterns to added N. lntercropped maize

grew poorly, showing nutrient deﬁciencies in phosphorus and potassium,

45

46

therdore results of intercropping on black locust N-ﬁxation are inconclusive.
Black locust may be able to continue to ﬁx atmospheric N despite high
applications of fertilizer N. However, competition for these nutrients as well as
allelopathy may be important limiting factors in black locust agroforestry

systems.

Introduction

N-ﬁxation by NFTs is important in tropical agroforestry systems. It
contributes to nutrient cycling by adding atmospheric N to the soil-plant system
(Young, 1989a), perhaps reducing requirements for added nitrogen fertilizer.
Research on the potential of temperate agroforestry, and what types of systems
will deﬁne it, is still in its infancy. Thus, little research exists about the potential
role of N-ﬁxation in temperate systems. N-ﬁxation is generally reduced with
increasing levels of soil N, but little is known about the effects of intercropping
on N-ﬁxation. It is expected that N-ﬁxation will increase with intercropping, in
part due to depletion of soil N by the nonlegume intercrop (Rerkasem and
Rerkasem, 1988). Conversely, above-ground effects such as shading can
decrease nodulation and N-ﬁxation (Nambiar et al., 1983).

Current evidence points to black locust as a potential leguminous
agroforestry tree because it has multiple uses and fast growth. Though data
suggests that black locust can support a high rate of N-ﬁxation (Dawson and
Johnson, 1992), it is reduced as available soil N increases (Johnson and

47

Bongarten, 1991; Reinsvold and Pope, 1987; Roberts et al., 1983). However,
there are no data describing the effects of intercropping on black locust N-
ﬁxation. It is possible that intercropping will increase black locust N-ﬁxation in
response to the companion crop’s competition for soil nitrogen. Therefore this
study was designed to address the following hypothesis:

. A greater dilution of “N will occur in the black locust when planted

with maize than in black locust planted alone.

The objective was to determine the extent to which black locust increases N-

ﬁxation in response to competition for soil N.

Methods

Cloned root cuttings of black locust were introduced to 30 x 30 x 100 cm
containers of N-deﬁcient sand with fertilizer (minus N) incorporated into the top
25 cm of sand. An error resulted in the application of 75 kg ha" Mg, 140 kg ha"
P205, and 130 kg ha" K20 equivalent, one-half of target levels. The seedlings
were grown for 98 days to allow the internal N demands to equilibrate. This also
provided a root distribution that would approximate that of an already existing
tree in an agroforestry system when maize is planted. On the 99th day, N and
plant treatments were applied. N treatments consisted of liquid (‘sNH4)2SO4
enriched to 1.27% 1"’N applied at 50, 125, and 200 kg N ha" equivalent. Plant
treatments consisted of one black locust, one maize, and one black locust + one

maize. The following terminology is used to compare plant treatment results:

lelle

(bla

1"]

48

B —> black locust alone;

BM —> black locust in black locust + maize;
M -+ maize alone;

MB —> maize in black locust + maize.

The BM and MB terms represent the same intercrop plant treatment. The ﬁrst
letter in each term denotes the plant being referred to in that plant treatment.
The term 'BM' is used to compare intercropped black locust (the “B' in BM) to B
(black locust alone). The term “MB” is used to compare intercropped maize (the
"M’ in MB) to M (maize alone). These treatments were applied for 45 days and
then the experiment was harvested. Leaf and stem biomass were collected
separately, dried for 24 hours at 65° C, and then stored. All biomass was
ground in a Thomas Wiley Mill #4 and then pulverized in a Cyclone Sample Mill
(Udy Corporation, Fort Collins, CO) to pass through a 1 mm screen. The
biomass was then measured for total N and "N in a continuos-ﬂow stable
isotope ratio mass spectrometer (ANCA-MS, Europe Scientiﬁc). Total N
readings were conﬁrmed using a Carlo Erba Nitrogen Analyzer 1500 Series 2
(F isions Instruments, Danvers, MA). N-ﬁxation was estimated by the Isotope
Dilution Technique (IDT. The biomass percentage of N derived from the

atmosphere (%Ndfa) was calculated as:

49

 

%Ndfa _ ( %15 N excess (fs) J
- — 1, x100
% N excess (nfs)
where:
fs = ﬁxing system - black locust
nfs = non-ﬁxing system - maize
(Rennie, 1 986)

Experimental design consisted of randomized complete blocks utilizing split-
plots. Statistical analysis was performed using analysis of variance (ANOVA)
with linear and quadratic orthogonal contrasts. Investigation to determine if the
assumptions of ANOVA were satisﬁed found that the means and variances were
not independent, variances were not homogeneous, and the error terms were
not normally distributed. Transformation of the data to log(10) corrected most of
these assumption violations. Validity of the ANOVA is discussed where

necessary.

Results

lntercropping reduced both black locust and maize growth (Table 3).
Maize growth as measured by above-ground dry weight (DW) and number of
leaves per plant decreased dramatically. The reduction in black locust DW is
signiﬁcant at p = 0.026. However since black locust DW variation is not
homogeneous and standard deviation is high, this statistical difference may not

be real (Figure 2). Indeed, the reduction of black locust DW is not reﬂected in

Table 3:

50

Average Maize Dry Weight and Leaves per Plant and Average
Black Locust Dry Weight

Key: M I maize alone; M8 = maize in the maize + black locust intercropping

B - black locust alone; BM 8 black locust In the maize + black locust intercropping
Numbers In parentheses represent one standard deviation.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Maize
Dry Weight (g) s Leaves per Plant
Plant Plant Plant Plant
Treatment Treatment Treatment Tmtment
N Level M MB M M8
N“ 34.50 (3.81) 4.79 (3.87) 12.50 (0.58) 7.5 (1.00)
N13 30.82 (2.88) 12.88 (8.43) 12.25 (0.50) 10.5 (1.29)
N. 25.83 (4.58) 9.52 (4.02) 12.25 (0.50) 9.0 (1.83)
Black Locust
Dry Weight
Plant Plant
Treatment Treatment
N Level B BM
Na. 51.81 (11.91) 48.83 (14.58)
N“. 51.41 (3.28) 45.91 (9.84)
Na. 53.44 (8.98) 43.93 (0.51)

 

 

 

 

 

51

70'

egg

88888

 

 

 

50 125 200
Nitrogen Treatment kg/ha

Key: B 8 black locust alone; BM 8 black locust in the maize + black locust intercropping
Enor bars denote one standard deviation.

Figure 2: Average Black Locust Dry Weight

estimates of N-fixation as measured by % biomass N derived from the
atmosphere nor milligrams of biomass N ﬁxed from the atmosphere (Table 4;
Figure 3). Therefore, no intercrop effect on N-ﬁxation was observed and it was
not affected by the potential DW difference between black locust plant
treatments.

However, N application levels had strong inﬂuences on N-ﬁxation,
decreasing as N fertilizer levels increased (p = 0.001). There was a 60%
decrease between the low and medium N treatments, and a further 20% drop

between the medium and high N treatments (Figure 3). No treatment

Table 4:

Key:

52

Average Black Locust N-Fixation as Measured by $6 Biomass
N Derived From The Atmosphere (%Ndfa) and mg Biomass N
Derived From The Atmosphere

Numbers in parentheses represent one standard deviation.

B-blacklocustaione; BMIblacklocustinthemaize+biacklocustintercropping

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fixed Biomass N (mg) 16 Biomass Ndfa
Plant Treatment Plant Treatment Plant Treatment Plant Treatment
N Level B BM B BM
N50 783 (256) 717 (296) 69.7 (3.3) 69.5 (6.2)
Nm 294 (90) 259 (81) 27.3 (7.2) 28.7 (2.9)
N2.» 259 (84) 223 (21) 21.6 (4.4) 23.1 (4.2)
coo 70 r
MN sec 1 sci
(no) 700 ‘ %Nda
eoo . so -
soo - ‘0 .
coo
300 3 30 f 3"
zoo ; an 2° ' B
‘°° so a. 260 1° 55 Ta 260
Ringer: Trednent m Nitrogen thon ironic
Key: B 8 black locust alone; BM 8 black locust in the maize + black locust intercropping

Error bars denote one standard deviation.

Figure 3:

Average Black Locust N-Fixation as Measured by $6 Biomass
N Derived From The Atmosphere (%Ndfa) and mg Biomass N
Derived From The Atmosphere

 

53

interactions were observed, therefore the change in N-ﬁxation was similar for
each plant treatment as N fertilizer levels changed.

Due to an error, phosphorous and potassium application levels were
lower than expected, perhaps contributing to the severe intercropped maize
(MB) growth reduction. Black locust competition and or allelopathy may also

have restricted MB development.

Discussion and Summary

The reduction in black locust N-ﬁxation with increasing N levels was
expected, and agrees with other work (Johnsen and Bongarten, 1992;
Reinsvold and Pope, 1987; Roberts et al., 1983). in this study, black locust
continued to ﬁx atmospheric N despite the highest N application rate, a rate near
the upper range of the amount applied to ﬁeld maize (Christenson et al., 1992).
it is well known that N-ﬁxation decreases with increasing soil N levels. Still,
black locust may be able to maintain N-ﬁxation despite high soil N
concentrations present in this study. In earlier work, black locust seedlings that
were grown in a greenhouse sand and fed with 5 mM NH4N03 applications for 50
days continued to ﬁx atmospheric N (Roberts et al., 1983). N-ﬁxation in black
locust seedlings supplied with an initial application of o, 25, so, or 100 mg N kg"
and sufficient phosphorus was greater in the middle treatments than in the high
treatment after 60 days. Then after 105 days black locust supplied with 100 mg

N kg" ﬁxed as much N as those supplied with 50 and 25 mg N kg‘1 (Reinsvold

54

and Pope, 1987). Similar increases in N-ﬁxation at the high N treatment over
time was reported by Johnsen and Bongarten (1992). Greenhouse grown black
locust seedlings in sand that received 0.25 mM and 2.00 mM NO; in daily
applications initially showed decreased N-ﬁxation at the high N level. However,
as growth and NW became greater after 93 days, N-ﬁxation at the high N
treatmentnearlyequaledtl'latofthe lothreatment. ltwould appearthataftera
high initial fertilizer application or with stable soil N levels, black locust N-ﬁxation
eventually increases. in addition, black locust shows high initial growth when
supplied with adequate amounts of soil N (Dawson ef al., 1992; Johnsen and
Bongarten, 1992; Reinsvold and Pope, 1987; Roberts et al., 1983). implications
of the high initial growth and continued N-ﬁxation over time despite high soil N
levels may have the following implications: ﬁxed atmospheric N could serve as a
source of additional N in North American agricultural systems; and the regular
addition of ﬁxed atmospheric N through incorporation of senesced or ooppiced
black locust leaves, stems, and roots in managed settings may decrease needed
levels of fertilizer N required to maintain intercrop yields.

it was expected that N-ﬁxation would have been greater in the
intercropped plant treatment due to competition for available soil N. It was also
expected that this difference would be especially apparent at the low N
treatments, resulting in a signiﬁcant plant treatment by N treatment interaction.
However, intercropping had no effect on N-ﬁxation in black locust. It is assumed

that black locust competed effectively for soil N, despite the presence of maize.

55

in studying the rooting habits of both B and M, both were found to have a
substantial portion of their roots near the soil surface. Both species are also
known to be very competitive in N uptake.

However, the shoot and root growth of intercropped maize (M8) were
poor. lf intercropping has an effect on N-ﬁxation, any intercrop difference may
not have been allowed to develop due to the poor growth of intercropped maize.
ltisassurnedthatthepoorgrowthofintercropped maizedidnotprovidethe
anticipated level of N competition required to stimulate additional black locust N-
ﬁxation.

lntercropped maize germinated normally, but later showed serious
symptoms of nutrient stress: burned leaf tips, purple lower leaves, and weak
lower stalks. The symptoms reﬂected phosphorus (P) and potassium (K)
deﬁciencies. Three additional applications of a complete nutrient mix (minus N)
were applied to the surface soil with no improvement. MB did not suffer from N
competition; MB biomass %N increased with increasing N levels. Although the
intercropped maize absorbed more N as it was provided, growth did not improve.
This indicates that other factors limited growth. in this case they appear to be P
and K deﬁciencies, but allelopathic effects of black locust on the intercropped
maize may have played a role as well. The increasing severity of the nutrient
deﬁciency symptoms necessitated harvest at 45 days, 15 days earlier than

originally planned.

56

There is little information available on black locust competition with
intercrops. in a study with similar results. black locust in a greenhouse reduced
theyieldofintercroppedbarleycomparedtotl'lesolecrop, althoughN
concentrations in the intercropped barley increased (Ntayombya and Gordon,
1993). The authors determined that competition for soil resources, especially
moisture, resulted in the lower barley yields. Conversely, two ﬁeld studies
intercropping black locust hedgerows with maize alleyways showed that maize
yields increased at the hedgerow edges (Seiter, 1994; Ssekabembe and
Henderlong, 1991). However, in both ﬁeld studies one year old seedlings were
used and interspecies root interactions most likely did not approach that found in
the barley study or my study.

There is also little data available on allelopathic effects of black locust. A
cursory review of the literature did not uncover any North American studies.
However, an abstract of a study performed in India indicates that leaf Ieachates
of black locust decreased maize (Zea quun'ans) germination and growth after
ﬁve days (Bhardwaj, 1993). Five day growth was reduced, though not
germination, in Tricr‘cum aestivum, Glycine max, Brassica campestn‘s, and Vigna
mango. Abstracts of studies conducted in the former Soviet Union indicate that
black locust leaf exudates adversely affected germination and growth of
numerous herbaceous plants (Matveev, 1975), and crown and root excretions

(phytolines) from black locust adversely affected Oak (species not provided)

57

(Kolesnicenko, 1964). Therefore black locust may have had an allelopathic
affect on intercropped maize in this study.

Evidence of satisfactory maize growing conditions were clearly seen when
maize was grown alone. The sole maize showed no symptoms of P and K
deﬁciencies and its root system explored the entire soil volume. Therefore, poor
intercropped maize growth appeared to be related to black locust competition or
allelopaﬂwy. Black locust grew well in both plant treatments, but better alone.
The competitiveness or allelopathic effects of black locust may be an early
warning regarding the compatibility of this tree species with maize production. in
addition, nutrient deﬁciencies could have been intensiﬁed due to the low P and K
application rates and from black locust absorbing a proportion of the P and K
during the 98 days before maize planting. Therefore, a shortage may have been
created before the maize was planted.

Overall, black locust ﬁxed N regardless of the added N level of the study.
Maize, however, may have difﬁculty competing with established black locust, not
necessarily for N but for P and K Black locust allelopathy may also retard maize
growth. These results may be indicative of future compatibility problems
associated with using black locust as a close intercrop with maize. This may be

especially true for the ﬁrst row of maize in a black locust and maize alley

cropping.

V. "N-LABELED NITROGEN FERTILIZER AND INTERCROPPING
EFFECTS ON NITROGEN TRANSFER FROM BLACK LOCUST
(ROBINIA PSEUDOACACIA L.) TO MAIZE (ZEA MAYS)

Abstract
Black locust (Robinia pseudoacacia L.) is a temperate fast-growing,

leguminous tree. As part of the effort to investigate its potential role in North

Amerimn agroforestry systems, a two factor, split-plot greenhouse study was

conducted to determine the effect of nitrogen (N) fertilizer on transfer of nitrogen

from black locust to intercropped maize. Cloned black locust seedlings from root

cuttings were transplanted to containers of N deﬁcient sand and grown for 98

days. N and plant treatments were then assigned. N as ('sNHt)2SO4 was

applied at 50, 125, and 200 kg ha'1 equivalent. A heavy isotope of nitrogen, 1"‘N,

was used to measure N-ﬁxation and N transfer. Maize was planted by seed to

form the following plant treatments: black locust alone, black locust and maize
intercropping, and maize alone. These treatments were applied for 45 days and
then the experiment was harvested.

lntercropped maize biomass had a signiﬁcantly lower %‘5N excess at the
low N treatment than did maize alone (p < 0.001), and similar 1"‘N levels at the
two higher N treatments. Average intercropped maize soil 1"’N and total
available soil N levels at low N were also lower than for maize alone, however

these plant treatment differences were not signiﬁcant at p = 0.680 and 0.200,

respectively. Black locust planted with maize showed greater live nodule weight

and root length than black locust alone at low N, but these plant treatment

differences were also not signiﬁcant at p = 0.458 and 0.143, respectively.

58

59

In comparing plant treatment means it would appear that the lower %‘5N
excess in lntercropped maize biomass was due to N transfer. Close examination
of the data reveals that either differential maize root access to soil 1"‘N pools or
black locust root interaction with maize roots might explain the 1""‘N dilution in
intercropped maize biomass. A clear interpretation of the data, however, is

difﬁcult due to the poor growth of intercropped maize in this study.

introduction

improved nutrient cycling is an important beneﬁt of agroforestry systems
over traditional agriculture. Deeper perennial root systems access nutrients at
soil depths too deep for agronomic plants and deposit them at the soil surface as
ﬁne root turnover and foliage from leaf and stem abscission (Young, 1985). The
use of nitrogen-ﬁxing trees (NFT s) and shrubs is especially beneﬁcial: they ﬁx N
and add it to the soil-plant system through the nutrient cycling process
(Mulongoy, 1986; Young, 1985). In this way N can be transferred from the NFT
to the intercrop in agroforestry systems. Another potential path for the transfer
of N is by way of direct root contact, direct root mycorrhizae contact, or uptake of
NFT rhizodeposition before being immobilized by soil microorganisms (Agboola
ef al., 1972; Giller and \Mlson, 1991; Mulongoy, 1986; Robinson, 1991; van
Noordwijk and Dommergues, 1990; Young, 1989b). Direct transfer is not
considered a major source of N to companion crops in tropical grass swards,

and even less in legume-cereal intercroppings (Catchpoole and Blair, 1990;

60

Henzell and Vallis, 1977; Giller et al., 1991; Giller and Wilson, 1991). However,
there has been little investigation of direct N transfer from NFT s to crops in
agroforestry systems (van Noordwijk and Dommergues, 1990). No literature or
studies have been found that have investigated the potential of N transfer from
black locust to intercrops in temperate systems. With its broad, shallow root
system and high N-ﬁxation, there is potential for direct transfer. Research in this
area is needed to better understand the beneﬁcial roles black locust may have in
Michigan and US. agricultural systems, and in particular, to determine its ability
to transfer N directly to companion crops.

In order to gain greater knowledge about the potential of black locust to
directly aid the nitrogen status of a companion crop, the following hypotheses
were developed:

0 A greater dilution of 1""N will occur in maize when planted with black

locust than in maize planted alone.

0 Total maize root length is reduced when planted with black locust.
The objectives were to measure N-transfer from the black locust to the
intercropped maize as well as to determine if maize root length is reduced in the
presence of black locust. In an effort to understand the interactions between
black locust and maize, improve experimental validity, and aid interpretation of
results, supporting data were collected. These data included soil "N, total
available soil N (ppm), black locust root length, black locust nitrogen-ﬁxation,

and black locust live nodule weight.

61

Methods

Cloned root cuttings of black locust were introduced to 30 x 30 x 100 cm
containers of N-deﬁcient sand with fertilizer (minus N) incorporated into the top
25 cm of sand. An error resulted in the application of 75 kg ha" Mg, 140 kg ha"
P205, and 130 kg ha" K20 equivalent, one-half of target levels. The seedlings
were grown for 98 days to allow the internal N demands to equilibrate and to
provide a root distribution that would approximate that of an already existing tree
in an agroforestry system when maize is planted. On the 99th day, N and plant
treatments were applied. N treatments consisted of liquid (‘~"Nl-l.),so. (1.27%
“m applied at 50, 125, and 200 kg N ha" equivalent. Plant treatments
consisted of one black locust, one maize, and one black locust + one maize. The

following terminology is used to compare plant treatment results:

B —> black locust alone;

BM —> black locust in black locust + maize;
M —> maize alone;

MB —) maize in black locust + maize.

The BM and MB terms represent the same intercrop plant treatment. The ﬁrst
letter in each term denotes the plant being referred to within that treatment. The
term ‘BM" is used to compare intercropped black locust (the “B” in BM) to B
(black locust alone). The term “MB” is used to compare intercropped maize (the
'M' in MB) to M (maize alone). These treatments were applied for 45 days and
then the experiment was harvested. Leaf and stem biomass were collected
separately and dried for 24 hours at 65° C. These were ground in a Thomas

Wiley Mill if 4 and then pulverized in a Cyclone Sample Mill (Udy Corporation,

62

Fort Collins, CO) and passed through a 1 mm screen. The biomass was then
measured for total N and "N in a continues-ﬂow stable isotope ratio mass
spectrometer (ANCA-MS, Europe Scientiﬁc). Total N readings were conﬁrmed
using a Carlo Erba Nitrogen Analyzer 1500 Series 2 (F isions Instruments,
Danvers, MA). The biomass percentage of N derived from the atmosphere

(%Ndfa) and the amount of N transfer were calculated as :

 

%Ndfa _ [ 96“ Nexcess (fs) J
- — ,5 x100
% N excess (nfs)
where:
%"N excess = the percentage of "‘N above background

levels, or that above 0.3663%.

 

fs = N-ﬁxing system, i.e. black locust

nfs = non N-ﬁxing system, i.e. maize
(Rennie, 1986).

%N in maize derived = R sole maize - R intercrop maize

from ”30" 'OCUSt R sole maize - R intercrop black locust

where R = %‘5N excess.
(Ta and Faris, 1987).

Root and soil harvesting, sampling, and measurement are discussed in detail in
chapter Ill. The experimental design consisted of randomized complete blocks

utilizing split-plots. Statistical analysis was performed by analysis of variance

63

(ANOVA) with linear, quadratic, and logical orthogonal contrasts. Assumptions
of ANOVA were tested using formulas provided by Little and Hills (1978).

After log(10) transformation, some data sets still did not satisfy the
assumptions of ANOVA Table 5 presents all the data sets reported in this
thesis and whether they meet the assumptions of ANOVA as normal or
transformed data. it is believed that the ANOVA is robust enough that the
problems of normality did not compromise statistical analysis given the data sets
to be evaluated (Ries, 1995). Other violations must be analyzed and described.
For example, the black locust live nodule weight variance was not independent
of the means. A greater variance at the low N treatment (high nodulation/low
nodule death) than at high N (low nodulation/high nodule death) presents a

danger that true mean differences at the high N treatments were masked. In this

Table 5: Data Sets and Assumptions of ANOVA.

 

 

 

 

 

 

 

 

 

Data Set Assumptions Not Satisﬁed

Black locust Root Length Error Term Normality

Live Nodule Weight Independence of Means and
Variances

Error Term Normality
Maize Biomass “N Ratio Homogeneity of Variances
Excess

Maize Biomass %N None

Sell ”it Ratio Excess None

Soil N in ppm Independence of Means and
Variances

Homogeneity of Variances

 

 

64

particular case the possibility that true mean differences were masked appears
remote.

Another example is in the maize biomass %"N ratio excess data. The
intercropped maize (MB) had greater variance than M, especially at the high N
treatment. This is not surprising in light of its poor growth. When pooled with
other variance terms to calculate mean square error terms and f-statistics, the
inclusion of the high MB could affect the ﬁnal f-statistic and mask real
differences in other treatments. This again does not appear likely: the low N
mean of the M8 was signiﬁcantly less than the other treatments, even with its
high variance. The remaining means were so close that real differences likely
do not exist. Even the highest variance was only 0.4% of its mean, so some
differences among variances seem large in part because the variances
themselves were so small.

Finally, in the total soil N data set, there was a tremendous amount of
variation. Plant treatment means varied greatly, as do mean rankings over the
three N treatments. Soil N response to N applications was signiﬁcant and this
was consistent across the measured N variables, therefore, N treatment
differences should be accurate. It is interesting that in M and MB plant
treatments, soil N variation in the top soil section was low compared to the
middle and bottom sections, whereas in B it was comparatively high. The means
were independent of this pattern. Overall, trends were hard to identify due to the

65

large number of treatments and variance, so simple caution is suggested while

interpreting total soil N data set.

Results

Black locust and maize dry weight (DW) are reported in chapter N. MB
growth and development was severely reduced while BM growth showed some
decline (Table 3). In maize above-ground biomass %“N excess, both plant
treatment (PT) and nitrogen (N) level responses were highly signiﬁcant (p <
0.001). However, the PT x N interaction was also signiﬁcant at p = 0.001,
indicating that M and MB biomass "’N were affected differently by N (Table 6).
Logical orthogonal contrasts indicate that at low N, MB above-ground biomass
had a signiﬁcantly lower %‘5N excess than did M (Table 6, Figure 4).
Conversely, M and M8 were not different at the greater N treatments. In the
analysis of N-ﬁxation by black locust, N fertilization affected both plant
treatments equally. Speciﬁc comparisons found both 8 and BM nitrogen
responses to be linear (p = 0.001) and quadratic (p = 0.002). These results
indicate that a deﬁnable negative relationship between available soil N and N-
ﬁxation was found in this study, but that no interaction was shown between
intercropping and N-ﬁxation at p = 0.854. Similar responses to added N were
observed in analysis of black locust live nodule weight. However, 8 and BM root

length actually increased at the highest N treatment level while B root weight

66

 

 

 

Table 6: ANOVA for %"N Excess in Maize Above-Ground Biomass
R’ - 0.913351 c.v. - 3.531855
Source DF Sums of Mean Square F Value p
Squares
Main Plots 7 0.01745
Block (B) 3 0.00055 0.00018 0.21 0.8843
Plant Treatment (PT) 1 0.01349 0.01349 15.73 0.0019
in 3 0.00342 0.00114
Sub Plots 23 0.11873
Main Plots 7 0.01745
Nitrogen Treatment (N) 2 0.05928 0.03454 40.41 0.0001
N linear 0.05155 0.05155 50.15 0.0001
N quadratic 0.01772 0.01772 20.57 0.0007
01'
Nm v N—m < 0.00001 < 0.00001 0.00 0.9532
N50 v (th x N200) 0.05928 0.05928 80.82 0.0001
N x PT 2 0.02159 0.01085 12.55 0.0011
N lin x PT 0. 01147 0. 01147 13.38 0.0033
N qua x PT 0.01022 0.01022 11.93 0.0048
Ol‘
("125 v N200) x PT 0.00115 0.00115 1.35 0.2678
[N50 v (N125 x Nzooll x PT 0.02053 0.02053 23.95 0.0004
Errorm 12 0.01029 0.00085

 

N50 - 50 kg N ha" equivalent

Nm - 125 kg N ha" equivalent
N200 - 200 kg N ha" equivalent

67

0.950
1615M Ratio

0.850”

_‘

0.750 ’

 

 

0'550 50 125 200
Nitrogen Treatment Ito/ha

Key: M I maize alone; MB I maize lntercropped with black locust
Enor bars denote one standard deviation.

Figure 4: Maize %“N Ratio Excess in Above-Ground Biomass

decreased and BM root weight remained stable (Table 7; Figure 5). Trend
analysis and contrasts determined that intercropping maize with black locust had
no statistical effect on root weight, root length, or live nodule weight, and that
there were no differential plant treatment responses to applied N in these data
sets (Table 8). The black locust root weight data set had high standard
deviations and a low correlation of variation at 0.470. Root weight variation may
simply be high and more samples would be needed to measure any treatment
differences if they exist.

There were signiﬁcant increases in available soil 1"’N and total available

soil N as N increased (p < 0.001). Maize PT x N interaction terms in both data

Table 7:

Key: Na, - 50 kg N ha"; N135 =125 ltg N ha"; N200 = 200 kg N ha"

68

Black Locust Average Root Length Distribution for Two Soil

Depths as Measured in Soil Cores and Average Total Root
Weight and Average Live Nodule Weight

Btblecltlocustalone; BMItheblacklocustintheblacklocusHmaizelnterclopping

The numbers in parentheses are standard deviations.

 

 

 

 

 

 

 

Root Length (m) Root Weight (g) Live Nodule Weight
Plant Treatment Plant Treatment Plant Treatment
N Soil
um Depth a BM a BM a BM
9.13 13.50
" ' 2‘ c'“ (4.28) (5.38)
Na. 18.87 13.14 3.10 4.55
25 cm ‘ “a 5.42 (7.32) (2.90) (2.78) (2.80)
below (1.34) (1.59)
0 - 25 cm 5.70 9.37
(3.57) (2.78)
N... 15.08 12.97 0.53 0.85
25 cm a 5.10 4.75 (2.28) (3.57) (0.25) (0.54)
below (1.75) (2.35)
22.55 13.34
° ' 2‘ ‘m (5.51) (7.01)
Na. 13.75 12.81 0.34 0.33
25 cm ‘ 9.75 13.31 (3.81) (2.54) (0.09) (0.15)
below (2.53) (10.22)

 

 

 

 

 

 

 

 

 

69

Tdaimmwmmw

 

 

”r
B
m)-
M 94
(ml ,0. "l
B
10:
s 1 l 1
50 125 200
NlrogenTreeeneulrgihe
AverageuadtlocthootWem

 

 

 

r. t
10 h- 4

 

 

50 125 200
NlrogenTreetmentlrgfhe

AverageBlackLocuetleeNoduleWelgl'lt

50m

LN

 

0.50 r

 

 

0'10 50 125 200
Nlrogen Treetrnent ltgihe

Key: B 8 black locust alone; BM = black locust in the maize + black locust intercropping
Error bars denote one standard deviation
Y-axes use log)10) scale.

Figure 5: Black Locust Total Soil Core Root Length, Average Root
Weight, and Average Live Nodule Weight

70

Table 8: Selected p-values for Black Locust Root Length, Root Weight,
and Live Nodule Weight Responses to Plant and Nitrogen

 

 

 

 

Treatments

Key: PT 8 plant treatment

N linear 8 linear trend contrast for nitrogen treatment

N quadratic 8 quadratic trend contrast for nitrogen treatment

PTxN 8plenttreatmentbynitrogentreatmentinteraction

p-vaiues

Source Root Length Root Weight Live Nodule Wt.
PT 0.441 0.093 0.280
N linear 0.010 0.236 < 0.001
N quadratic 0.013 0.949 0.071
PT x N 0.242 0.562 0.530

 

 

sets were not signiﬁcant, and contrasts conﬁrmed the lack of statistically
different plant treatment responses to added N. Logical contrasts indicated that
both soil “N and total soil N were signiﬁcantly greater in the middle section of
the containers (p = 0.014 and 0.003, respectively) and statistically equal at the
top and bottom depths (Table 9). MB total root length and root weight were
greatly reduced compared to M (Table 10). However, poor top growth of
intercropped maize confounded the treatment effects so statistical analysis was
inappropriate. The differential root growth likely resulted in differential sampling
of soil "N pools.

lntercropped maize above-ground biomass %"N excess levels, when
analyzed by block, were higher than M at the medium N application and

sometimes at the high N application level. Positive N transfer estimates depend

 

71

Table 9: Average Pools of Soil %"N Excess and Available Soil N by Soil
Depth 45 Days After Nitrogen Treatments Were Applied

Averages were calculated using over all N treatments and all plant treatments. Soil 15N used
weiglaed means to account for different soil weights by depth.

 

 

 

 

 

soil Depth soil-“N Ratio Excess Available Soil N (ppm)
0 - 25 cm 0.242 0.960
25 cm - 55 cm 0.419“ 2.538“
55 cm and below 0.253 1.177
Wgreater than the top and ”greater than the top and
bottom depths at p 8 0.014 bottom depths at p 8 0.025

 

Table 10: Maize Average Root Length Distribution for Two Soil Depths
as Measured in Soil Cores and Average Total Root Weight

Key: Na,- 50 ltg N ha"; N125 =125 kg N ha"; N200: 200 ltg N ha"
M 8 maize alone; MB 8 the maize in the maize + black locust intercropping
The numbers in parentheses are standard deviations.

 

 

 

 

 

 

 

 

Root Length (m) Root Weight (g)
Plant Treatment Plant Treatment
N Level Soil Depth M MB M MB
0 - 25 cm 5.33 (1.85) 0.65 (0.52)
Ng. 18.45 (3.81) 1.53 (0.73)
25 cm a
“'0‘” 12.34 (2.77) 0.37 (0.18)
0 - 25 cm 3.19 (1.24) 2.00 (1.97)
N.” 13.26 (2.34) 3.57 (1.80)
25 cm 8.
below 8.60 (5.26) 1.66 (1.48)
0 - 25 cm 3.12 (1.64) 0.57 (0.37)
Na. 12.44 (4.02) 2.05 (2.12)
25 cm 8
below 5.06 (1.79) 0.55 (0.78)

 

 

 

 

 

 

 

72

Table 11: Nitrogen Transfer from Black Locust to Maize

Key: N50 - l50 kg N ha" equivalent
Nm - 125 kg N ha" equivalent
Nu - 200 kg N be" equivalent

 

 

 

 

 

B 8 black locust
Replic‘lon NLevel %BN %MaizeN TotalmgN
in Maize from B Transfen'ed
l N50 1.02 13.30 10.08
ll N50 0.39 25.64 5.64
lll N50 0.14 7.45 1.94
N N50 1 .22 42.41 8.73
l N95 -4.53 -11.22 -54.31
ll N125 -0.99 -5.25 -10.08
lll N135 -1.16 -7.09 43.72
IV N125 -1.07 -6.16 -9.82
l N200 5.17 20.12 66.36
ll N200 -1.19 -4.75 46.05
lll N200 13.21 41.04 141.97
IV Nm 0.30 3.13 3.03

 

 

 

 

 

 

on the intercropped nonlegume biomass having a lower %‘5N excess than the
nonlegume planted alone. Because the opposite occurred, N transfer estimates
were often negative. Transfer estimates at low N were positive, though variable

(Table11).

Discussion and Summary

With a cursory view of the data, the ﬁrst hypothesis, that a greater dilution
of 1“N will occur in MB versus M, would be accepted at the low N application.
MB had a signiﬁcantly lower biomass %"N excess than M, but only at the low N
treatment (Figure 4). However, a critical review of the data does not support
acceptance of this hypothesis. Hypothesis number two, that MS root length

73

would be reduced versus M due to the presence of black locust, was not
addressed due to the confounding effects ofthe poor MB growth.

There are at least two potential explanations for the dilution of MB
biomass 15N. First, the differential root growth patterns between M and MB
resulted in differential access to soil “N pools. Soil pools of 1""N do not vary by
plant treatment, but they do vary by soil depth, with the middle (25-55 cm) being
signiﬁcantly greater than the top (0-25 cm) and bottom (55 cm and below) (Table
9). M explored the entire soil volume and MB did not. Soil core sampling
revealed that well over one-half of the MB root length was located in the top 25
cm of soil whereas M incorporated over one-half of its root length 25 cm and
below. Over all three N treatments, M had much more root length than MB
below 25 cm, and greater total root weight throughout the container (Table 10),
therefore M and MB did not equally explore the soil pools of 15N. This was
conﬁrmed through visual inspection of opened containers.

Another potential explanation for the signiﬁcantly lower MB above-ground
biomass %"N excess at low N is the presence of greater BM root length and live
nodule weight than 8 (Figure 5). High root and nodule concentrations in the root
zone of nonlegumes may lead to direct N transfer (van Noordwijk and
Dommergues, 1990). in this case, black locust always maintained at least one-
half of its root length in the top 25 cm of soil, more than half at the low and
middle N treatments (Table 7). Competition for soil N may have increased BM

root length and live nodule weight, while BM root weight decreased. The greater

74

BM versus 8 root length and live nodule weight may explain the lower MB soil
"N than M over all N treatment levels, especially at low N, although this soil 1"’N
difference is not signiﬁcant at p = 0.344 (Figure 6).

in this study it is evident that there is a trend showing an interaction
between plant treatments and N levels on black locust root length, nodulation,
and soil "N levels. Though not statistically signiﬁcant, the presence of more BM
versus 5 root length and nodules coincide with the lower MB soil "‘N at low N.
The lower MB soil 1"N presumably resulted from the lower N application in
combination with competition for soil N between maize and black locust. It may
be that the lower MB biomass %‘5N excess at low N resulted from the ensuing
lower soil "N as well as greater maize and black locust root contact that allowed
maize to take up black locust rhizodeposition. Still, it is difﬁcult to separate the
potential effect of differential M and MB rooting patterns and resulting differential
"N uptake from the potential effect of black locust roots and nodules in contact
with maize roots. While it is plausible that MB received some its N directly from
black locust, the difference in biomass 1"‘N ratios may simply have resulted from
differential access to 1"’N.

The difﬁcult interpretation of the data is due to the poor growth of
intercropped maize. The poor growth was presumably caused by soil nutrient

competition, most likely for phosphorous (P) and potassium (K). Physical

Kl

75

Tummsclcuemuw Amwwmw

0r 5 3° W 3 l, ..
30 ’ Weigh .l ,
\

10:

 

 

 

 

 

 

 

so 125 a. 50 135 sec
”TM“ NlrogenTreatmerditgfhe

 

.
b
0.50:
.

 

 

.1 A 1 l
0 0 50 125 200
NlrogenTreetmenilqihe

AverqelhlzeSoil'NRdloElceas

0M :
0.50 ;
15NRdo

0.10 E

 

.
0““
D

 

 

 

0.01 50 125 m

Megan Treatment loihe

Key: M 8 maize alone; MB 8 the maize in maize + black locust intercropping;
B 8 black locust alone: BM 8 the black locust in maize + black locust intercropping
Error bars denote one standard variation.
Y-axes use log(10) scale.

Figure 6: Black Locust Root Length, Root Weight, and Live Nodule
Weight Compared to Maize Soil 1"N Ratio Excess

 

76

manifestations included brown, burned-looking leaf tips, dark purple lower
leaves, and weak and broken stalks. Nutrient deﬁciencies could have been
intensiﬁed due to the low P and K application rates and from black locust
absorbing a proportion of the P and K during the 98 days before maize planting.
Since MB biomass %N increased with increasing N fertilizer with no
improvement in MB health status, it is apparent that N was not limiting.

Related effects of black locust on companion crops vary. In a separate
greenhouse experiment, black locust reduced the yield of intercropped barley
compared to the sole crop (Ntayombya and Gordon, 1993). Conversely, two
ﬁeld studies intercropping black locust hedgerows with maize alleyways showed
that maize yields increased at the hedgerow edges (Seiter, 1994; Ssekabembe
and Henderlong, 1991). However, Ssekabembe and Henderlong (1991)
reported results with one year old seedlings, so interspecies root interactions
most likely did not approach that found in the barley study or in my study. It will
be very interesting to see if the positive edge effects seen in the maize continue
with time as interaction with black locust roots increases. In the Seiter study
workers chisel plow between the black locust hedgerow and the ﬁrst row of

maize, greatly reducing the potential for root competition. After four years of

 

FIT.

intercropping these species no negative affects have been observed in the
maize (Seiter, 1 995)
Shading of the intercropped maize may have prevented adequate growth

as well. As this was not a treatment and shading was not expected to be a

77

problem, light readings at the maize level were not taken. However, black locust

lacks a dense canopy, therefore shading should not have been a major factor in

decreasingMBgrowth.

Vi. CONCLUSIONS

The broad objective of this study was to aid in efforts to determine black
locust's role in temperate agroforestry systems and contribute to the debate on
agroforestry's role in safer and more sustainable Michigan and US.
agroecosystems. Speciﬁcally, this work attempted to collect initial data that
would elucidate this NFT’s ability to contribute to the soil-plant N cycle through
N-ﬁxation and N transfer to an intercrop.

An important aspect of this study was validation of the assumptions of the
lDT to assure that the estimates of N-ﬁxation are accurate. it is important that
both the nfs reference plant (maize) and fs (black locust) plants have similar N
uptake patterns. However, this is not critical when soils are low in carbon
sources, such as organic matter, as was the case in this study. Initially, maize
has low demand for soil N, greatly increasing soon after initiation of the V6
growth stage (i.e., presence of six leaves) while black locust seedlings scavenge
for N and increase growth throughout their development. Considering this and
the fact that black locust had a well-developed root system when the ('5NH4)2$O..
was applied, it is assumed that N-ﬁxation was underestimated. However, the
fact that the soil medium was a sand low in organic matter and had a low cation
exchange capacity (CEC), the immobilization of applied "N overtime would
have been minimal. Furthermore, visual inspection of the B and M root systems

showed them to be similarly distributed, indicating equal access to soil 15N at the

78

79

time at harvest. In addition, analysis of soil total N and 1“N at three levels of
depth showed no signiﬁcant differences by plant treatment. Considering the low
soil N immobilization, equal root distributions at harvest, and similar soil %"N
exmss levels across plant treatments, the discrepancy in ﬁnal biomass %‘5N
excess at the time of harvest caused by low initial maize N uptake should result
in only a slight underestimate of N-ﬁxation, and at least allow a conﬁdent ranking
of the treatments.

This study presented data on N-ﬁxation, nodulation, above-ground
biomass %‘5N excess, root length, root weight, available soil N, and soil 1"’N.
Added fertilizer N signiﬁcame suppressed N-ﬁxation and live nodule weight,
while available soil N, soil 1"’N, and above—ground %‘5N excess increased.
lntercropping showed one signiﬁcant difference in these data: maize grown with
black locust had lower biomass %‘5N excess at low N than did sole maize.
Similar trends due to intercropping at low N were also noted in black locust root
length and live nodule weight, though these intercrop differences were not
signiﬁcant. lntercropping also greatly reduced maize growth which resulted in
much less above- below-ground biomass. This repressed growth was caused by
nutrient competition for P and K, allelopathy, or a combination of these.
Competition may have been intensiﬁed due to an error in nutrient applications
(minus N) where the experiment received one-half the target amounts of P and

K in addition, intercropped black locust absorbed P and K for 98 days before

80

the introduction of maize. N was provided in amounts as called for in the
experimental proposal and no N competition was observed.

it was expected that there would be lower biomass %"N excess at low N
in intercropped maize. The presumption was that the presence of black locust
would absorb soil N in competition with the maize companion crop thereby
reducing available soil N and soil "N. The reduced soil N would require black
locust to increase N-fixation as well as root length and nodulation in the
presence of maize roots. The high amount of root interaction and direct maize
root contact with black locust roots and rhizospheres would aid N-transfer. This
N-transfer would be observed as lower biomass %“N in the intercropped maize.
Indeed, there was high black locust/maize root interaction and contact at the low
N application, and intercropped maize had a signiﬁcantly lower %“N excess.
However, the poor growth of the intercropped maize introduced a second
hypothesis explaining the lower biomass %‘5N - dissimilar access to pools to
soil “N that resulted in the different biomass “N levels. Thus the poor
intercropped maize growth confounded the results and no deﬁnitive conclusions
can be made at this time regarding the potential of black locust to transfer N to a
companion crop.

Still, the author ﬁnds it intriguing that at low N intercropped black locust
root length and live nodule weight were greater than sole black locust. It is
conceivable that if the intercropped maize had not grown poorly and the

experiment had been longer in duration, these differences would have been

81

greater. N-ﬁxation did not follow this trend, but a time delay between the
development of greater root length and increased live nodule weight and
ensuing lower %‘5N excess in the plant tissue used to estimate N-ﬁxation may
provide an explanation for this result. Only further studies can validate or reject
this hypothesis.

Despite the poor growth of intercropped maize, and also in light of it,
there are some interesting observations to be made from this work. The
presence of one-half of black locust root length in the top 25 cm of soil, and
presumably most of the nodule weight, conﬁrms the potential of N-transfer, and
the results of this study allude to this potential. The normal growth of maize
alone, black locust alone, and intercropped black locust, despite less than
optimum soil nutrient levels, indicate that black locust competes well for
nutrients. While N competition does not appear to be important, P and K
competition deserves study. lntercropped maize P and K deﬁciencies indicated
that black locust will out compete maize for these nutrients, perhaps because
they are key for maintaining high N-ﬁxation rates (Reinsvold and Pope, 1987).
This points to the importance of maintaining ideal P and K levels in systems
incorporating black locust.

In this study, as in others (Dawson at al., 1992; Johnson and Bongarten,
1992; Reinsvold and Pope, 1987), black locust continued to ﬁx N under the
highest N application rate. This could mean that black locust will continue to ﬁx

N despite the high fertilizer N applications necessary for maize. In temperate

82

agroecosystems this may result in the addition of atmospheric N, and in
combination with the N cycling ability of black locust, the additional N introduced
into the system may eventually reduce conventional rates of expensive N
fertilizer.

Finally, to conﬁrm the assumptions under which the IDT is used, it is vital
to collect and analyze root and soil data, in addition to plant biomass, as done in
this study. Without this data it would have been tempting to claim solid evidence
of N-transfer as a result of the lower biomass %“N at low N in intercropped
maize. The additional data, though time consuming to collect, provided an
important check on the validity of the results and allowed the ﬁndings to reﬂect a
realistic interpretation. Readers should question similar studies that do not
provide such important collaborative data.

If this work was to be repeated, a few changes are warranted. Though a
simple error calculating the correct P and K applications necessary for this work
may have played a role in the nutrient deﬁciency and poor growth observed in
the intercropped maize, managing nutrients in a nutrient-less medium such as
sand can be difﬁcult. The researcher must have a good understanding of how to
develop an appropriate balance and avoid problems such as high salt
concentrations. An alternative to a nearly sterile sand is to use a soil a little
higher in N but sufﬁcient in the other nutrients, such as a Boyers Field Sand.
This would provide a soil allowing for study of reactions to added fertilizer N, but

would simplify nutrient management. Another change would be to place the

83

experiment outside a greenhouse. Black locust seedlings could be started in a
sand, and when weather permitted, transplanted into outside containers.
Atmospheric deposition of N and "N from air and rain could be reduced by
covering the containers at the root collar of the plants, allowing only the drip
irrigation tube for the deionized water. With this done, environmental conditions
would be a bit more realistic while still providing a closed and accessible system
for biomass and soil analysis. Incorporating minirhizotrons into the containers to
observe intercrop root interaction would enable another measurement of
interaction as well as root and nodule turnover. Ultimately, repeated above- and
below-ground measurements should be taken to provide a time series of results.
This would provide a greater understanding of the development of interactions
and treatment effects.

For an initial study such as this, the methods used were appropriate to
validate and satisfy the assumptions of the IDT. However, this would be much
more difﬁcult in the ﬁeld. The rooting behavior of maize and black locust was
similar in this greenhouse study using sand, but in the ﬁeld their root systems
are not similar, making maize a less suitable reference plant for calculating N-
ﬁxation in, and N transfer from, black locust. A proper ﬁeld study would have to
give a great amount of consideration to a satisfactory reference plant. If
greenhouse studies provide results that encourage further study, then the time

and resources necessary needed for a ﬁeld experiment might then be

34

appropriate. Reeves (1991) provided an excellent example of the process of
shifting from greenhouse to ﬁeld studies to investigate N-ﬁxation and N transfer.
The scientiﬁc and practitioner communities both recognize the need to
develop agroecosystems that are safer and more sustainable. Rural
communities look to diversify their products from agriculture to survive.
Agroforestry is a land—use system gaining recognition as one alternative to help
fulﬁll these parallel goals, and black locust has surfaced as a native temperate
nitrogenoﬁxing woody perennial with great potential as an agroforestry tree. It is
hoped that this study contributes to the ongoing investigation into the suitability

of agroforestry and the usefulness of black locust.

LIST OF REFERENCES

LIST OF REFERENCES

Agboola, AA and AAA Fayemi. 1972. Fixation and Excretion Of Nitrogen By
Tropical Legumes. Agroforestry Journal, 64: 409-412.

Altieri, MA, and M. Leibman, 1986. Insect Weed and Plant Disease
Management in Multiple Cropping Systems. In: Multiple Cropping
Systems, pp. 183-218. C.A Francis (ed). Macmillan Publishing
Company.

Arnold, J.E.M. 1987. Economic Considerations In Agroforestry. In:
Agroforestry: A Decade Of Development, pp. 173-190. HA Steppler and
PKR. Nair (eds). International Council for Research in Agroforestry,
ICRAF House, off Limuru Road, Gigiri. PO. Box 30677, Nairobi, Kenya.

Barrett, RP. and J.W. Hanover. 1991. Robinia pseudoacacia: A possible
temperate zone counterpart to Leucaena? in: The 2nd Conference on
Agroforestry in North America, pp. 27-41. H.E. 'Gene' Garrett (ed.).
School of Natural Resources, University of Missouri, Columbia, Missouri,
6521 1.

Bhardwaj, SD. 1993. Effect of Leaf Leachate of Robinia pseudoacacia on Seed
Germination and Growth of Some Agricultural Crops. Indian Journal of
Forestry, 16(3): 285-286.

Boring, LR. and WT. Swank. 1984. The Role of Black Locust (Robinia
pseudoacacia) in Forest Succession. Journal of Ecology, 72: 749-766.

Broadbent, F .E., T. Nakashima, and G.Y. Chang. 1982. Estimation of Nitrogen
Fixation by Isotope Dilution in Field and Greenhouse Experiments.
Agronomy Journal, 74: 625-628.

 

Bremner, J.M. 1977. Use of Nitrogen-Tracer Techniques for Research on :
Nitrogen Fixation. In: Biological Nitrogen Fixation in Farming Systems of --
the Tropics. A Ayanaba and P.J. Dart (eds). John Wiley and Sons.

Brinkman, G. 1993. Pioneer Seed Company. Personal communication.
Brooks, P., D. Herman, and L. Halverson. 1992. 1"’N Diffusioin Technique for

Soil KCI Extracts. Department of Soil Science, 108 Hilgard Hall,
University of California, Berkeley, CA 94720.

85

 

86

Buck, MG. 1986. Concepts of resource sharing in agroforestry systems.
Agroforestry Systems, 4(3): 191 -204.

Burity, H.A, T.C. Ta, M.A. Paris, and BE. Coulman. 1989. Estimation of
nitrogen ﬁxation and transfer from alfalfa to associated grasses in mixed
swards under ﬁeld conditions. Plant and Sell, 114: 249-255.

Cabrera, ML, and DE. Kissel. 1989. Review and Simpliﬁcation of Calulations
in "N Tracer Studies. Ferlizer Research, 20: 11-15.

Catchpoole, D.W. and G.J. Blair. 1990. Forage Tree Legumes. II.
Investigation of Nitrogen Transfer to an Associated Grass Using 5 Split-
Root Technique. Australian Journal of Agricultural Research, 40: 531-
537.

Chalk, P.M.. 1985. REVIEW: Estimation of N2 Fixation By Isotope Dilution: An
Appraisal of Techniques Involving "‘N Enrichment and their Application.
Soil Biology and Biochemistry, 17(4): 389-410.

Chirko, GP. 1993. PaulowniaNVinter Wheat lntercropping Quantifying The
Relationship Between Photosynthetically Active Radiation (PAR) And
Yield. Ph.D. Dissertation. Michigan State University, Department of
Forestry, East Lansing, Michigan.

Christenson, D.R., DD. Wamcke, M.L. Vitosh, L.W. Jacobs, and JG. Dahl.
1992. Fertiziler Recommendations for Field Crops in Michigan.
Extension Bulletin E-550A. Cooperative Extension Service, Michigan
State University, East Lansing, Michigan.

Danso, SKA 1988. The Use of "’N Enriched Fertilizers for Estimating Nitrogen
Fixation in Grain and Pasture Legumes. In: Nitrogen Fixation by
Legumes in Medeterranean Agriculture, pp. 354-359. D.P. Beck and LA.
Materon (eds). Martinus Nijhoff Publishers.

Dawson, J.O., T.A., White, P.J. Dzialowy, and RA Herendeen. 1982. Nitrogen
Accretion in Surﬁcial Soil of a Young Plantation of Robinia pseudoacacia.
Illinois Agricultural Experiment Station Forestry Research Report 82-7.
Department of Forestry, University of Illinois at Urbana-Champaign. As
reported in: Dawson, J.O, C.S. Vogel, and KH. Johnsen. 1992.
Nitrogen Relations in Black Locust. In: Black Locust: Biology, Culture,
and Utilization, pp. 170—183. J. W. Hanover, K Miller, and S. Plesko
(eds). Department of Forestry, Michigan State University, East Lansing,
MI 48824.

 

ir—

Dawson, J.O., M.W. Pasche. 1991. Current and potential uses of nitrogen-
ﬁxing trees and shrubs in temperate agroforestry systems. In: The 2nd

87

Conference on Agroforesz in North America, pp. 183-209. H.E. 'Gene'
Garret (ed.). School of Natural Resources, University of Missouri,
Columbia, Missouri, 6521 1.

Dawson, J.O, C.S. Vogel, and KH. Johnson. 1992. Nitrogen Relations in Black
Locust In: Black Locust: Biology, Culture, and Utilization, pp. 170-183.
J. W. Hanover, K Miller, and S. Plesko (eds). Department of Forestry,
Michigan State University, East Lansing, MI 48824.

Devlin, RM. and F .H. Witham. 1983. Plant Physiology, Fourth Edition.
Wadsworth Publishing Company.

Ewel, J., F. Benedict, C. Berish, B. Brown, S. Gliessman, M. Arnador, R.
Berrnudez, A Martinez, R. Miranda, and N. Price. 1982. Leaf Area, light
transmission, roots and leaf damage in nine tropical plant communities.
Agra-Ecosystems, 7: 305-326.

Fried, M., L. A. Dean. 1953. A Concept Concerning the Measurement of
Available Soil Nutrients. Soil Sciences, 73: 263-271.

Garrett, H.E. 'Gene' (ed). 1991. The 2nd Conference on Agroforestry in North
America. School of Natural Resources, University of Missouri, Columbia,
Missouri, 6521 1

Giller, KE., J. Orrnesher, and F .M. Awah. 1991. Nitrogen Transfer from
Phaseolus Bean to lntercropped Maize Measured Using 1"’N-Enriched and
1""N-lsotope Dilution Methods. Soil Biology and Biochemistry, 13(4): 339-
346.

Giller, KE. and KF. Wilson (eds). 1991. Nitrogen Fixation in Tropical
Cropping Systems. CAB. lntemational.

Gillespie, AR. 1989. Modelling Nutrient Flux and Interspecies Root
Competition in Agroforestry lnterplantings. Agroforestry Systems, 8(3):
257-265.

"'J-r‘~_,_' e‘ ".- -_ _ ‘

 

Gish, T.J., and WA Jury. 1983. Effect of Plant Roots and Root Canals on
Solute Transport. Transactions of the ASAE, American Society of
Agricultural Engineers. ASAE, Order Department, 2950 Niles Road, St.
Joseph, MI 49085.

Gold, MA and J.W. Hanover, 1987. Agroforestry Systems for the Temperate
Zone. Agroforestry Systems, 5: 109-121.

88

Gordon, J.C., J.O. Dawson. 1979. Potential uses of nitrogen-ﬁxing trees and
shrubs in commercial forestry. Botanical Gazette, 140(Supplement): 888-
S90

Grime, J.P., B.D. Campbell, J.M.L Mackey, and JG. Crick. 1991. Root
Plasticity, Nitrogen Capture and Competitive Ability. In: Plant Root
Growth: An Ecological Perspective, pp. 381 -397. D. Atkinson (ed).
Blackwell Scientiﬁc Publications.

Gruggenheim, S. and J. Spears. 1991. Sociological And Environmental
Dimensions Of Sociological Forestry Projects. In: Putting People First:
Sociological Variables In Rural Development, pp. 304-339. M.M. Cemea
(ed). Oxford University Press.

Hanover, J.W., K Miller, and S. Plesko (eds). 1992. Proceedings: lntemational
Conference on Black Locust: Biology, Culture, & Utilization. Department
of Forestry, Michigan State University, East Lansing, Michigan.

Harris, 0 and EA Paul. 1989. Automated Analysis of "N and “c in Biological
Samples. Communications in Soil Science and Plant Analysis, 20(9810):
935-947.

Harris, D. 1993. Personal communication. Department of Plant and Soil
Sciences, Michigan State University, East Lansing, Michigan.

Hauser, S. and MP. Gichuru, 1994. Root Distribution of Daciyladenai (Acioa)
barten' and Senna (Cassia) siamea in Alley Cropping on Ultisol. ll.
impact on Water Regime and Consequences for Experimental Design.
Agroforestry Systems, 26(1): 9-22. :

Heckmann, MO, and Drevon, J.J. 1987. Inhibition of Symbiotic Nitrogen i
Fixation By Nitrate. In: Proceedings of an lntemational Conference on ;
the Physiological Limitations and the Genetic Improvement of Symbiotic
Nitrogen Fixation, pp. 97-105. F. O’Gara, S. Manian, and J.J. Drevon
(eds). Cork, Ireland, Sept. 1-3, 1987. Kluwer Academic Publishers.

 

Henzell, BF. and l. Vallis. 1977. Transfer of Nitrogen Between Legumes and
Other Crops. In: Biological Nitrogen Fixation in Farming Systems of the
Tropics, pp. 73-86. A Ayanaba and P.J. Dart (eds). John Wiley 8 Sons.

Huck, MG. 1983. Root Distribution, Growth, and Activity With Reference to
Agroforestry. In: Plant Research and Agroforestry, p 527-540. PA
Huxley (ed). lntemational Council For Research In Agroforestry (ICRAF),
PO Box 30677, Nairobi, Kenya.

89

Huntley, JG 1990. Robinia pseudocacia L. Black Locust. In: Silvics of North
America, Volume 2. Hardwoods, pp. 755-761. R.M. Burns and SH
Honkala (eds). Forest Service, United States Department of Agriculture.

Huxley, PA 1983. The Role of Trees In Agroforestry: Some Comments. In:
Plant Research and Agroforestry, pp. 257-270. Peter A. Huxley (ed).
lntemational Council For Research In Agrofroestry (ICRAF). PO. Box
30677, Nairobi, Kenya.

Ike, AF ., and EL. Stone. 1958. Soil Nitrogen Accumulation Under Black
Locust. Soil Science Society Proceedings, 22: 346-349.

Jackson, J.E., P.J.C. Hamer, and 8. Jackson. 1989. Water-Balance And Soil
Water Relations Studies In A Mixed Tree/Grass/Bare-Soil System. In:
Meteorology and Agroforestry, pp. 431 -442. lntemational Council for
Research in Agroforestry (ICRAF), ICRAF House, off Limuru Road, Gigiri,
PO Box 30677, Nairobi, Kenya.

Jensen, ES. 1991. Evaluation of Automated Analysis of "N and Total N in
Plant Material and Soil. Plant and Soil, 133: 83-92.

Johnsen, KH., and BC. Bongarten. 1991. Allometry of Acetylene Reduction
and Nodule Growth of Robinia pseudoacacia families Subjected to varied
Root Zone Nitrate Concentrations. Tree Physiology, 9: 507-522.

Johnsen, KH. 1992. Nitrogen Fertilization and Growth of Black Locust In:
Black Locust: Biology, Culture, and Utilization, pp. 184-197. J.W.
Hanover, K Miller, and S. Plesko (eds). Department of Forestry,
Michigan State University, East Lansing, Michigan.

Johnsen, KH. and BC. Bongarten. 1992. Relationships between Nitrogen
Fixation and Growth in Robinia pseudoacacia Seedlings: a Functional
Growth Analysis Approach Using 1"’N. Physiologia Plantarum, 85: 77-84.

Jonsson, K, L.. F idjeland, JA Maghembe, and P. Hogberg. 1988. The
Vertical Distribution of Fine Roots in Five Tree Species and Maize in
Morogoro, Tanzania. Agroforestry Systems, 6: 63-69.

Kang, B.T., G.F. Wilson. 1987. The development of alley cropping as a
promising agroforestry technology. In: Agroforestry: a Decade of
Development, pp. 227-244. HA Steppler and P.KR. Nair, (eds).
lntemational Council for Research in Agroforestry. ICRAF House, off
Limuru Road, Gigiri. PO. Box 30677, Nairobi, Kenya.

90

Khosla, PK and GP. Toly. 1986. Renewed scientiﬁc interest in agroforestry.
In: Agroforestry Systems: A New Challenge, pp. 7-21. PK Khosla, S
Puri, and DK Khurana (eds). Natraj Publishers, Edhra Dun, India.

Kolesnicenko, M.V. 1964. The Need to Consider Biochemical Inﬂuences of
Trees. Lesn. Hoz. 17(9): 15-19.

Lal, R. 19899. Agroforestry Systems and Soil Surface Management of 9
Tropical Alﬁsol. IV. Effects on Soil Physical and Mechanical Properties.

Agroforesz Systems, 5(3): 197-215.

Lal, R. 1989b. Agroforestry Systems and Soil Surface Management of a
Tropical Alﬁsol. V. Water Inﬁltrability, Transmissivity, and Soil Water
Sorptivity. Agroforestry Systems, 8(3): 217-238.

LaRue, TA, and T.G. Patterson. 1981. How Much Nitrogen Do Legumes Fix?
Advances in Agronomy, 34: 15—38.

Ledgard, S.F., J.R. Freney, and JR. Simpson. 1985. Assessing Nitrogen
Transfer From Legumes to Associated Grasses. Soil Biology and.
Biochemistry, 17(4): 575-577.

Little, MT, and F.J. Hills, 1978. Agricultural Experimentation: Design and
Analysis. John Wiley and Sons.

Matveev, NM, G.N. Krisanov, and II. Lyzhenko. 1975. The Role of Plant
Exudates in Formation of Herbaceous Cover Under Robinia pseudoacacia
and Sumac Protective Tree Strips in the Steppe Zone. Nauchnye
Doklady Vysshei Shkoly, Biologicheskie Nauki, No. 10, 80-84.

McIntyre. AC., and CD. Jefferies. 1932. The effect of black locust on soil
nitrogen and growth of catalpa. Journal of Forestry, 30: 22-28.

Mebrahtu, T. and J.W. Hanover. 1991. Family Variation in Gas Exchange,
Growth, and Leaf Traits of Black Locust Half-Sib Families. Tree
Physiology, 8: 185-193.

Minchon, F.R., Sheehy, J.E. and Witty, J.F. 1986. Further errors in the
acetylene reduction assay: effects of plant disturbance. Journal of
Expeimental Botany, 37: 1 581 -1 591 .

Mishra, V.K, 1986. Scope of nitrogen ﬁxing plants. In: Agroforestry Systems:
A New Challenge, pp. 213-219. PK Khosla, S Puri, and DK Khuransa
(eds). Natraj Publishers, Dehra Dun, India.

 

pma

91

Mulongoy, K 1986. Nitrogen cycling in alloy cropping systems. IITA Research
Briefs, 7(4): 3,5.

Nambiar, P.T.C., MR. Reddy, C.N. Floyd, P.J. Dart, and R.W. Willey. 1983.
Effect of intercropping on nodulation and N2-Fixation by Groundnut.
Experimental Agriculture, 19: 79-86

Ntayombbya, P. and AM. Gordon. 1994. Effects of Black Locust on Productivity
and Nitrogen Nutrition of lntercropped Barley. In: Opportunities for
Agroforestry in the Temperate Zone Worldwide: Proceedings of the Third
North American Agroforestry Conference, pp. 69-76. R.C. Schultz and
JP. Collettii (eds). Department of Forestry, College of Agriculture, Iowa
State University, Ames, Iowa.

Orig, C.K, J.E. Corlett, RP. Singh, and CR. Black. 1991. Above and below
ground interactions in agroforestry systems. In: Agroforestry Principles
and Practice, pp. 45-57. P.G. Jarvis, (ed.). Elsevier, Amsterdam, the
Netherlands.

Putnam, AR., and W3. Duke. 1978. Allelopathy In Agroecosystems. Annual
Review of Phytopathology, 16: 431-451.

Rachie, KO. 1983. lntercropping Tree Legumes With Annual Crops. In: Plant
Research and Agroforestry, pp. 103-115. PA Huxley (ed). lntemational
Council For Research in Agroforestry (ICRAF).

Reeves, M. 1991. Nitrogen-Mediated Interactions in Legume-Nonlegume
lntercropping Systems. Ph.D. Dissertation, Department of Biology,
University of Michigan, Ann Arbor, MI.

Reinsvold, R.J. and PE. Pope. 1985. Association of Robinia pseudoacacia and
Rhizobium: Potential Nitrogen Accretion. In: Proceedings: Fifth Central
Hardwood Forest Conference, pp. 245-250. J.O. Dawson and KA.
Mauerus (eds). University of Illinois, Urbana-Champaign, Illinois.

Reinsvold, R.J. and PE. Pope. 1987. Combined Effect of Soil Nitrogen and
Phosphorous on Nodulation and Growth of Robinia pseudoacacia.
Canadian Journal of Forest Research, 17: 964-969.

Rennie, R.J. and DA Rennie. 1983. Techniques for Quantifying N2 Fixation in
Association with Nonleguems Under Field and Greenhouse Conditions.
Canadian Journal of Microbiology, 29: 1022-1035.

Rennie, R.J. 1985. Nitrogen Fixation in Agriculture in Temperate Regions. In:
Current Plant Science And Biotechnology In Agriculture, pp. 659-665.

92

H.J. Evans, P.J. Bottomley, & W.E. Newton (eds). Martinus Nijhoff
Publishers.

Rennie, R.J. 1986. Comparison of Methods of Enriching a Soil with Nitrogen-15
to Estimate Dinitrogen Fixation by Isotope Dilution. Agronomy Journal,
78: 158-163.

Rerkasem, K and B. Rerkasem. 1988. Yields and Nitrogen Nutrition of
lntercropped Maize and Ricebean (Vlgna unbellata (Thumb) Ohwi and
Ohashi). Plant and Soil, 108: 151 -162.

Rios, S. K 1995. Personal communication. Professor Emeritus, Department of
Horticulture, Michigan State University.

Risch, S.J., D. Andow, and MA Altieri. 1983. Agroecosystem Diversity and
Pest Control: Data, Tentative Conclusions, and New Research
Directions. Environmental Entomology, 12: 625-629.

Roberts, C.R., R.W Zimmerman, J.W. Stringer, and SB. Carpenter. 1983. The
Effects of Combined Nitrogen on Growth, Nodulation, and Nitrogen
Fixation of Black Locust seedlings. Canadian Journal of Forest
Research, 17: 171-179.

Robinson, D. 1991. Roots and Resource Fluxes in Plants and Communities.
In: Plant Root Growth: An Ecological Perspective, pp. 103-130. D.
Atkinson (ed). Blackwell Scientiﬁc Publications.

Rosecrance, R.C., S. Rogers, and M. Toﬁnga, 1992. Effects of Alley Cropped
Calliandra calofhyrsus and Clin‘cidia sipium Hedges on Weed Growth, Soil
Properties, and Taro Yields in Western Samoa. Agroforestry Systems,
19(1): 57-66.

Ruhigwa, BA, M.P. 616110111, 8. Mambani, and NM. Tariah. 1992. Root
Distribution of Acioa barren, Alchomea cordifolia, Cassia siamea, and
Gmelina arborea in an Acid Ultisol. Agroforestry Systems, 19(1): 67-78.

Sanchez, PA 1987. Soil productivity and sustainability in agroforestry
systems. In: Agroforestry: a Decade of Develooment, pp. 205-226.
Steppler, HA and Nair, PKR (eds). International Council for Research in
Agroforestry. ICRAF House, off Limuru Road, Gigiri. PO Box 30677,
Nairobi, Kenya.

Schultz, R.C., and JP. Collettie (eds), 1994. Opportunities for Agroforestry in
the Temperate Zone Worldwide: Proceedings of the Third North

93

Amerimn Agroforestry Conference, August 16-18, 1993. Department of
Forestry, College of Agriculture, Iowa State Universtiy, Ames, Iowa.

Searle, P.G.E., Y. Comudom, D.C. Shedden, and RA Nance. 1981. Effect of
Maize + Legume lntercropping Systems and Fertilizer Nitrogen on Crop
Yields and Residual Nitrogen. Field Crops Research, 4: 133-145.

Seiter, S. 1994. Crop and Tree Response to Planting Patterns in Alley
Cropping Systems. MS. Thesis. Department of Horticulture, Oregon
State University, Corvallis, Oregon.

Seiter, S. 1995. Personal Communication. Department of Horticulture, Oregon
State University, Corvallis, Oregon.

Sharma, GK, and WA Geyer. 1990. Comparative Growth of Agroforestry
Trees in the Midhill Region of Himachal Pradesh, India. Agroforestry
Systems, 6: 101-111.

Sheehy, J.E., and A McNeill. 1987. Factors Controlling Legume Nitrogen
Fixation and Their Measurement in the Field. In: Proceedings of an
lntemational Conference on the Physiological Limitations and the Genetic
Improvement of Symbiotic Nitrogen Fixation, pp. 89-95. F. O’Gara, S.
Manian, and J.J. Drevon (eds). Cork, Ireland. Kluwer Academic
Publishers.

Singh, R.P., C.K Ong, and N. Saharan. 1989. Above and below ground
interactions in alley-cropping in semi-arid India. Agroforestry Systems,
9(3): 259-274.

Smucker, AJ.M., J.C. Ferguson, W.P. DeBruyn, R.K Belford, and J.T. Ritchie.
1987. Image Analysis of Video-Recorded Plant Root Systems.
Minirhizotron Observation Tubes: Methods and Applications for
Measuring Rhizosphere Dynamics, ASA Special Publication No. 50.
American Society of Agronomy, Crop Science Society of America, and
Soil Science Society of America, 677 S. Segoe Rd., Madison, WI 53711
USA

Spears, J. 1987. Agroforestry: A Development—Bank Perspective. In:
Agroforestry: A Decade Of Development, pp. 5366. HA Steppler and
P.KR. Nair (eds). lntemational Council for Research in Agroforestry.
ICRAF House, off Limuru Road, Gigiri. PO. Box 30677, Nairobi, Kenya.

Ssekabembe, OK and PR. Henderlong. 1991. Below-Ground Interactions in
Alley Cropping: Appraisal of First Year Observations on Maize Grown in
Black Locust (Robinia pseudoacacia) Alleys. In: Proceedings of the 2nd

94

Conference on Agroforestry in North America, pp. 58-73. H.E. 'Gene'
Garret (ed). School of Natural Resources, University of Missouri,
Columbia, Missouri. 65211.

Steppler, HA, and P.KR. Nair (eds). 1987. Agroforestry: A Decade of
Developement. lntematinal Council for Research in Agroforestry, ICRAF
House, PO. Box 30677, Nairobi, Kenya.

Szott, L.T., E.C.M. Femandes, and PA Sanchez. 1991. Soil-plant interactions
in agroforestry systems. In: Agroforestry Principles and Practice, pp.
127-152. P.G. Jarvis (ed). Elsevier, Amsterdam, the Netherlands.

Ta, T.G. and MA Paris. 1987. Effects of Alfalfa Proportions and Clipping
Frequencies on Timothy-Alfalfa Mixtures. II. Nitrogen Fixation and
Transfer. Agronomy Journal, 79: 820-824. As reported in Giller, KE., J.
Ormesher, and FM. Awah. 1991. Nitrogen Transfer from Phaseolus
Bean to lntercropped Maize Measured Using 1"’N-Enriched and 15N-
lsotope Dilution Methods. Soil Biology and Biochemistry, 13(4): 339-346.

Ta, T.C., MA Paris, and F.D.H. MacDowall. 1988. Evaluation of "N Methods
To Measure Nitrogen Transfer From Alfalfa To Companion Timothy.
Plant and Sell, 114: 243-247.

Technicon. 1977. Nitrate and nitrite in water and seawater. Industrial Method
Number 158-71 W. Technicon Industrial Systems, Tarrytown, New York,
USA

Technicon. 1978. Ammonia in water and seawater. Industrial Method Number
334-74wlb. Technicon Industrial Systems, Tarrytown, New York, USA.

van Noordwijk, M. and Y.R. Dommergues. 1990. Root nodulation: the twelfth
hypothesis. Agroforestry Today, 2(2): 9-10.

Vose, P.B., AP. Ruschel, R.L. Victoria, S.M.T. Saito, and E. Matsui. 1982. 15N
As A Tool In Biological Nitrogen Fixation Research. In: Biological
Nitrogen Fixation Technology for Tropical Agriculture, pp. 575-592. P.H.
Graham and SC. Harris (eds). Papers Presented at a Workshop Held at
CIAT, March 9—13, 1981. Call, Columbia. Centro lntemacional de
Agricultura Tropical, CIAT.

Willey, R.W. and MS. Reddy. 1981 A Field Technique For Separating Above-
and Below-Ground Interactions In lntercropping: An Experiment With
Pearl Millet/Groundnut. Experimental Agriculture, 17: 257-264.

95

Williams, P. (ed) 1991. Proceedings of the First Conference on Agroforestry in
North America. Department of Environmental Biology, Ontario
Agricultural College, University of Guelph, Guelph, Canada.

Witty, J.F.. 1983. Estimating Nz-Fixation in the Field Using "'N-Labelled
Fertilizer. Some Problems and Solutions. Soil Biology and Biochemistry,
15: 631-639.

Vlﬁtty, J.F. and K Ritz. 1984. Slow-release 15N Fertlizer Formulations to
Measure N2-Fixation by Isotope Dilution. Soil Biology and Biochemistry,
16: 657-661.

Witty, J.F. and FR. Minchon. 1988. Measurement of Nitrogen Fixation by the
Acetylene Reduction Assay; Myths and Mysteries. In: Nitrogen Fixaton
by Legumes in Mediterranean Agriculture, pp. 331-344. Beck DP. and
Materon, LA (eds). Martunus Nijhoff, Dordrecht.

Young, A 1985. The Potential of Agroforestry as a Practical Means of
Sustaining Soil Fertility. Working Paper No. 34. Nairobi: lntemational
Council For Research in Agroforestry (ICRAF).

Young, A, 19898. Agroforestry For Soil Conservation. Science and Practice of
Agroforestry No.4. Nairobi: lntemational Council on Research in
Agroforestry (ICRAF).

Young, A 1989b. Ten hypotheses for soil-agroforestry research. Agroforestry
Today, 1(1): 13-16.