EXPLORING THE BENEFITS OF COVER CROPS TO AGROFORESTRY TREE
PLANTATIONS: AN ANALYSIS OF DIRECT AND INDIRECT NITROGEN TRANSFER IN
ALLEY CROPPING SYSTEMS
By
Patrick Shults

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Forestry – Master of Science

2017

i

ABSTRACT
Exploring the Benefits of Cover Crops to Agroforestry Tree Plantations:
An analysis of direct and indirect nitrogen transfer in alley cropping systems
By
Patrick Shults
Incorporating species capable of biological nitrogen fixation into agroforestry systems can
improve nutrient cycling and soil health while reducing the need for synthetic nitrogenous
fertilizers. Research has confirmed multiple pathways through which biologically fixed nitrogen
is transferred to non-nitrogen fixing plants, including indirect transfer through decaying biomass
and direct transfer through belowground root exudates and shared fungal networks. I analyzed
the benefits of cover crops to the growth and nutrition of important agroforestry tree species
through experiments in the field and greenhouse. In the field study, I examined the contributions
of Trifolium pratense to the health and growth of hybrid poplar (Populus nigra L.×P.
maximowiczii A. Henry ‘NM6’) in an alley cropping system at varying seed rates and tree row
spacing. Soil and poplar leaf nutrient analysis revealed higher nitrogen levels in treatments with
T. pratense at regardless of seed rate at the wider spacing, while narrow spacing treatments
showed signs of competition. In the greenhouse study, the transfer of N through root exudates
from white clover (Trifolium repens) to a hybrid poplar (Populus nigra L.×P. maximowiczii A.
Henry ‘NM6’) and a hybrid chestnut (Castanea sativa x C. crenata ‘Colossal’) was quantified
via 15N isotope tracing methods. Results revealed that over the course of 9 weeks, white clover
supplied 33.3% (Âą7.4) of chestnut N and 12.1% (Âą1.1) of poplar N through root eudates. White
clover (T. repens) was found to exude 0.048 g N m-2 day-1 and roughly 3.4% of its’ total N over
the course of the 63 day experiment. Together these studies reinforce the potential of nitrogenfixing species to benefit mixed systems through both direct and indirect means.
ii

ACKNOWLEDGEMENTS

For his diligent support of my research and academic success, even as an undergraduate
student, I would like to thoroughly thank my advisor, Dr. Pascal Nzokou. I would also like to
thank my committee members, Dr. David Skole and Dr. Philip Robertson, for their
encouragement and feedback during the completion of this research, as well as the MSU
Department of Forestry for their support.

I would like to extend my deep gratitude to Ismail Koc, for teaching me most of what I
know regarding lab and fieldwork and his valuable additions to this research. In addition I
would like to thank Chad Papa and Bret Foster for sacrificing their time to help me with my lab
and fieldwork.

Lastly, I would like to thank my wife for her endless encouragement and support
throughout the difficult periods of these projects.

iii

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... vi
LIST OF FIGURES ..................................................................................................................... viii
LIST OF EQUATIONS .............................................................................................................. x
1

Introduction ............................................................................................................................. 1

2

Literature Review .................................................................................................................... 5
2.1 Agroforestry ..................................................................................................................... 6
2.1.1
Global Agroforestry Trends ...................................................................................... 8
2.1.2
Common Systems and Characteristics ...................................................................... 9
2.2 Current Obstacles Facing Food and Fiber Systems ....................................................... 11
2.2.1
Population Pressures, Poverty, and Land Limitations ............................................ 12
2.2.2
Effects of Agricultural Fertilizer Use ..................................................................... 15
2.2.3
Biodiversity, Pests, and Disease ............................................................................. 18
2.2.4
Future Effects of Climate Change and Possible Feedbacks .................................... 21
2.2.5
Land and Soil Degradation .................................................................................... 26
2.3 Belowground Species Interactions and Nutrient Cycling in Agroforestry .................... 35
2.3.1
Alley Cropping and Nutrient Cycling ..................................................................... 36
2.3.2
Biological Nitrogen Fixation and Alley Cropping.................................................. 39
2.4 Modes of N-Transfer ...................................................................................................... 41
2.4.1
Indirect Transfer...................................................................................................... 41
2.4.2
Direct Transfer ........................................................................................................ 44

3 Contributions of Trifolium pratense to soil health and hybrid poplar growth in temperate
alley cropping systems .................................................................................................................. 47
3.1 Abstract .......................................................................................................................... 47
3.2 Introduction .................................................................................................................... 48
3.3 Methods .......................................................................................................................... 50
3.3.1
Experimental Design ............................................................................................... 51
3.3.2
Field Methods ......................................................................................................... 52
3.3.3
Sample Analysis...................................................................................................... 54
3.3.4
Data Analysis .......................................................................................................... 55
3.4 Results ............................................................................................................................ 56
3.4.1
Cover Crop Nutrient Concentrations ...................................................................... 56
3.4.2
Cover Crop Nutrient Return.................................................................................... 58
3.4.3
Plant-Available Soil Nitrogen ................................................................................. 60
3.4.4
Poplar Leaf Nutrients .............................................................................................. 62
3.4.5
Poplar Height and Diameter.................................................................................... 64
3.4.6
Inter-Variable Correlations ..................................................................................... 66
3.5 Discussion ...................................................................................................................... 66
iv

3.5.1
Biomass Return ....................................................................................................... 67
3.5.2
Soil .......................................................................................................................... 68
3.5.3
Foliar Nutrient Concentrations ............................................................................... 69
3.5.4
Poplar Growth ......................................................................................................... 70
3.6 Conclusion...................................................................................................................... 72
4 Quantifying belowground nitrogen transfer between Trifolium repens and agroforestry tree
species in temperate alley cropping systems via root exudates .................................................... 74
4.1 Abstract .......................................................................................................................... 74
4.2 Introduction .................................................................................................................... 75
4.3 Materials and Methods ................................................................................................... 78
4.3.1
Experimental Design ............................................................................................... 78
4.3.2
Materials ................................................................................................................. 80
4.3.3
Propagation and Seeding......................................................................................... 81
4.3.4
Watering .................................................................................................................. 82
4.3.5
Isotopic Tracer Application .................................................................................... 83
4.3.6
Photosynthesis and Growth Measurements ............................................................ 84
4.3.7
Sampling ................................................................................................................. 84
4.3.8
Exudate Experiment (EX) Sampling ...................................................................... 86
4.3.9
Laboratory Analysis ................................................................................................ 88
4.3.10 Isotopic Analysis ..................................................................................................... 89
4.3.11 Data and Statistical Analysis .................................................................................. 90
4.4 Results ............................................................................................................................ 92
4.4.1
Nitrogen Transfer .................................................................................................... 92
4.4.1.1 𝛅15N Values....................................................................................................... 92
4.4.1.2 Exuded Nitrogen ................................................................................................ 95
4.4.1.3 Percent N Derived from Transfer ....................................................................... 96
4.4.1.4 Tree N Content Derived from Transfer (mgNdft) .............................................. 97
4.4.1.5 Percent of Exuded N Absorbed (%ExAbs) ........................................................ 98
4.4.2
Tree Nutrients (N, P, K, and Ca) ............................................................................ 99
4.4.3
Soil Nitrogen ......................................................................................................... 104
4.4.4
Photosynthesis....................................................................................................... 106
4.4.5
Height and Diameter ............................................................................................. 107
4.4.6
Other Correlations between Variables .................................................................. 108
4.5 Discussion .................................................................................................................... 109
4.5.1
Clover Exudates .................................................................................................... 110
4.5.2
N Transfer ............................................................................................................. 111
3.5.3
Nutrient Dynamics, Photosynthesis, and Growth ................................................. 114
4.6 Conclusion.................................................................................................................... 116
5

Discussion and Research Implications ................................................................................ 119

6

Conclusion ........................................................................................................................... 125

BIBLIOGRAPHY ....................................................................................................................... 127

v

LIST OF TABLES

Table 1 . Summary and description of common global agroforestry systems 3 2 , 3 3 . ........... 10
Table 2. Average nutrient concentrations returned during cutting from natural vegetation
(NV), red clover at a half seed rate (RCH), and red clover at a full seed rate (RCF). ....... 57
Table 3. Total nutrient return via green manure by the end of the 2015 season from
natural vegetation (NV), red clover at a half seed rate (RCH), and red clover at a full seed
rate (RCF). ................................................................................................................. 60
Table 4. Summary of soil ammonium in treatments prior to cover crop establishment
(09/2014) and through two subsequent growing seasons for natural vegetation (NV), red
clover at a half seed rate (RCH), red clover at a full seed rate (RCF), and control
treatments................................................................................................................... 61
Table 5. Summary of soil nitrate in treatments prior to cov er crop establishment (09/2014)
and through two subsequent growing seasons for natural vegetation (NV), red clover at a
half seed rate (RCH), red clover at a full seed rate (RCF), and control treatments. .......... 62
Table 6. Average poplar leaf nitrogen and phosphorous concentrations between spacing
treatments over the course of the experiment for natural vegetation (NV), red clover at a
half seed rate (RCH), red clover at a full seed rate (RCF), and c ontrol treatments........... 63
Table 7. Average height and diameters of poplar at the end of each growing season for
natural vegetation (NV), red clover at a half seed rate (RCH), red clover at a full seed rate
(RCF), and control treatments. ..................................................................................... 65
Table 8. Average biomass productivity of poplars at tree and stand levels for natural
vegetation (NV), red clover at a half seed rate (RCH), red c lover at a full seed rate (RCF),
and control treatments. ................................................................................................ 65
Table 9. Description of sample compartmentalization for each species in greenhouse
experiment. ................................................................................................................ 86
Table 10. Summary of 𝛅 1 5 N values in treatment IR by species and compartment over the
course of the 63-day experiment. ................................................................................. 94
Table 11. Table summary of parameters from clover exudate experiment (EX). ............. 96
Table 12. Summary of N transfer parameters from isotopic labelling experiment. .......... 96
Table 13. Summary of nutrient content per tree between tr eatments IR and C at the end of
the experiment. ........................................................................................................... 99

vi

Table 14. Summary of average nutrient concentrations between treatments IR and C at the
end of the experiment. ............................................................................................... 100
Table 15. Summary of average soil nitrate concentrations in treatments IR and C. ....... 105
Table 16. Summary of photosynthesis measurements in treatments IR a nd C ............... 106
Table 17. Summary of relative tree growth at the end of the experiment. ..................... 108

vii

LIST OF FIGURES

Figure 1. Chemical process for biological nitrogen fixation occurring in the root nodules
of legumes 1 1 ............................................................................................................... 40
Figure 2. Description of mechanisms behind ro ot exudation of compounds through a plant
cell membrane 1 1 3 . ‘PM’ refers to plasma membrane, ‘TMD’ refers to transmembrane
domain, and ‘NBD’ refers to nucleotide -binding domain. .............................................. 44
Figure 3. Plot layout of field study at the Tree Research Center in East Lansing.
Individual squares correspond to a single tree. Relative width of "blocks" corresponds to
narrow and wide spacing treatments and colors indicate vegetation treatments. .............. 51
Figure 4. Average nutrient concentrations of biomass returned during routine cuttings
from natural vegetation (NV), red clover at a half seed rate (RCH), and red clover at a full
seed rate (RCF). .......................................................................................................... 57
Figure 5. Average nitrogen -to-phosphorous ratios based on nutrient concentrations of
cover crop biomass return from natural vegetation (NV), red clover at a half seed rate
(RCH), and red clover at a full seed rate (RCF). ........................................................... 58
Figure 6. Total dry weight of biomass returned from natural vegetation (NV), red clover at
a half seed rate (RCH), and red clover at a full seed rate (RCF) during the 2015 season. . 59
Figure 7. Total nutrient return in biomass from natural vegetation (NV), red clover at a
half seed rate (RCH), and red clover at a full seed rate (RCF) during the 2015 sea son. ... 59
Figure 8. Soil mineralized nitrogen concentrations after combining spacing treatments
across natural vegetation (NV), red clover at a half seed rate (RCH), red clover at a full
seed rate (RCF), and control treatments. ....................................................................... 62
Figure 9. Greenhouse study experimental model and sampling design. .......................... 79
Figure 10. Average 𝛅 1 5 N values for poplar and chestnut soils in treatment IR over the
course of the experiment. ............................................................................................ 94
Figure 11. Average 𝛅 1 5 N values (line) and N content in milligrams (bars) of white clover
exudate in exudate experiment (EX). Horizontal bars represent standard error. .............. 95
Figure 12. Percent of nitrogen derived from transfer (%Ndft) by species over the course
of the experiment. ....................................................................................................... 97
Figure 13. Nitrogen content derived from transfer by species over the course of the
experiment ................................................................................................................. 98

viii

Figure 14. Percentage of available exudate absorbed by each tree species over the course
of the experiment. ....................................................................................................... 99
Figure 15. Changes in nitrogen content over time in clover and trees within treatment IR.
................................................................................................................................ 102
Figure 16. Average soil nitrate concentrations in treatments IR and C over the course of
the experiment, including initial soil samples from bagged soil ( -28). .......................... 104
Figure 17. Changes in average soil organic nitrogen in treatment IR over time. ........... 105
Figure 18. Photosynthetic activity in treatment IR (dotted line) and C (solid line) ....... 107

ix

LIST OF EQUATIONS

Equation 1. Allometric equation for determining biomass of hybrid poplar 1 3 7 . ............... 52
Equation 2. Equation used by UC -Davis Isotope Lab to determine 𝛅15N values from
isotopic analysis. Here, “atom%15N(sample)” refers the percentage of 𝛅15N atoms in
sample and “atom%15N(atmosphere)” refers to the atmospheric 𝛅15N concentration
(0.3663%). ................................................................................................................. 90
Equation 3. Equation to determine the percent nitrogen derived from transfer (%Ndft) for
each tree species at each sampling period. Where “r” denotes recipient, “d ” denotes
donor, “s” denotes soil, “0” denotes the beginning of the experiment, and “t” denotes the
sample period. ............................................................................................................ 91
Equation 4. Conversion of %Ndft into actual N transfer (mg). Where “%Ndf t(F)” refers
to the final value of %Ndft for each tree at the end of the experiment and “N(Tree)” refers
to the total estimated N content of each tree in milligrams. ........................................... 91
Equation 5. Equation f or calculating the proportion of available exudate from clover
actually absorbed by each tree. Where “N(Transfer) denotes the milligrams of N sourced
from transfer in each tree and “ExN(F) denotes the total N exuded over the course of the
experiment in milligrams. ............................................................................................ 91
Equation 6. Calculation for determining the proportion of clover nitrogen lost through
exudation. Where “N(Transfer)” refers to tree N derived from transfer at the end of the
experiment and “N(Clover)” refers to total N content of the clover at the end of the
experiment. ................................................................................................................ 92

x

xi

1

Introduction

Global demand for food and fiber is expected to increase consistently in the coming decades.
Due to limited remaining productive land and a desire to protect remaining natural forests, much
of the increased demand will have to be met from existing production systems 1. However, the
growing problem of global land degradation poses a threat to the ability to increase future
production to meet future demands 2. Moreover, altered climatic conditions as a result of
human-induced climate change will continue to threaten food and fiber yields and require more
resilient agricultural systems to maintain present and future production needs 3–5. In order to
overcome the issues facing food and fiber systems, producers and landowners must embrace
agroecological concepts that ensure high yields while sustaining the integrity of the land in use.

Agroforestry practices integrate agriculture and forestry into the same land management scheme
in order to protect soil and diversify landowner income while providing multiple ecosystem
services 6. These practices present an important pathway for establishing sustainable production
and resilient farming systems that will be critical to meeting future food and fiber demands. A
large body of research has highlighted the potential for agroforestry to not only increase yields
but also restore soil on degraded lands, improve biological diversity, and protect waterways 2,6,7.
Further, the United Nations has recognized agroforestry as having a critical role in climate
change mitigation because of its efficient carbon sequestration and resiliency to changing
climatic conditions 8.

One of the many ways that agroforestry practices support sustainable production is through
enhanced nutrient cycling. This occurs, in part, because of the ability of trees to capture water
1

and nutrients at deeper layers and subsequently make them available to herbaceous crops with
shallow rooting systems 9,10. However, a perhaps more critical element to nutrient cycling in
these practices is the utilization of biologically fixed nitrogen through the incorporation of
leguminous species. These species have symbiotic relationships with soil microbes that inhabit
their roots and “fix” N2 gas from the atmosphere into plant-available forms in exchange for
energy 11. Since nitrogen is an extremely important nutrient to plant health and productivity,
determining the best methods of utilizing this process for the benefit of food and fiber production
has been the subject of much research. Additionally, the high carbon cost of producing
synthetic nitrogenous fertilizers and associated risks to ecosystem health during production has
made biologically fixed nitrogen an increasingly important form of sustainable fertilizer for
production systems.

Thus far, research has identified three pathways for biologically fixed nitrogen to be transferred
from a legume to a non-nitrogen-fixing species: (1) indirect transfer through decaying legume
biomass, mineralization of organic N, and subsequent uptake by plants, (2) direct transfer of
plant available nitrogen through shared belowground fungal networks, and (3) direct transfer
through the exudation of N-rich compounds (primarily NH4+ and amino acids) from living roots
of legumes to the roots of a recipient plant 12,13. As it stands, indirect transfer through decaying
biomass has been the primary mode of transfer utilized in agriculture systems and is thought to
be the most substantial N transfer pathway 12. However, this process requires the mineralization
of organic N into plant-available forms and, as a result, the actual delivery of biologically fixed
N to recipient plants can be difficult to control. Direct transfer, on the other hand, supplies N
compounds that are readily assimilated by recipient plants. Recent research has highlighted the

2

underestimated importance of direct transfer pathways, particularly through root exudates 14–17.
The slow but steady N transfer through direct pathways could serve as a consistent, low-input,
and sustainable form of nitrogenous fertilizer. Regardless of the transfer pathway, utilizing
biologically fixed nitrogen for production systems could displace the need for synthetic
fertilizers, which are often inefficiently used by cropping systems and result in pollution of soil
and waterways18.

Nitrogen transfer, whether through direct or indirect means, is affected by a number of biotic and
abiotic factors. Water stress, low temperatures, and low light availability, for instance, all inhibit
both the fixation of nitrogen and the movement of exudate through soil 13. Transfer can also be
influenced by the level of root-to-root contact, plant growth stages, and legume to non-legume
density ratios. Roggy et al. (2004) revealed the importance of legume species selection to
transfer potential. Species with high soluble N and low lignin content within their leaves, for
instance, had higher indirect transfer rates 14. Exudation potential can differ greatly between
species even within the same family, as has been recorded for species of Trifolium17.

Research focused specifically on N transfer within agroforestry dates at least as far back as the
1980’s, if not earlier 19,20. However, it wasn’t until after the turn of the century that researchers
started identifying and quantifying between the different transfer mechanisms. The development
of 15N isotope methods, including natural abundance and tracer application studies, has allowed
nitrogen sources to be traced and quantified in such studies 12. Presently, there is a small but
powerful pool of research that has utilized these methods in various species combinations with
great success. However, there are some notable research gaps that have yet to be filled. For
3

example, much of this vein of research has focused on the transfer of nitrogen from leguminous
trees and shrubs to herbaceous crops and grasses. There are also a number of studies that have
focused on transfer from herbaceous legumes to grasses. No research, however, exists on the
potential for herbaceous legumes with high nitrogen-fixing capacity and exudation potential,
such as clovers, to provide directly transferred N to valuable tree species. A number of studies
have illuminated multiple benefits of incorporating species such as clover and alfalfa into tree
plantations, including improved nutrient availability to trees and higher foliar N content 19,21,22.
However, few, if any, have actually quantified direct and indirect N transfer in such systems.

As a result of this research gap, I illuminated the potential of two clovers (Trifolium pratense and
Trifolium repens) to benefit temperate alley cropping systems. These studies will build on
existing research regarding the benefits of incorporating legumes into production systems and the
transfer of biologically fixed nitrogen. The novelty of these experiments is the focus on trees as
recipient of N transfer and herbaceous cover crops as the N donor and comparing transfer
between two species combinations. This research consists of two experiments: a two-year field
study that focus on the observed benefits of T. pratense to a hybrid poplar plantation and a
greenhouse study that quantified direct N transfer from T. repens to two important agroforestry
tree species. The results of these studies will inform land management decisions for agroforestry
and tree plantation practitioners regarding soil health and tree nutrition. A preceding section will
offer a literature review describing the present issues facing food and fiber systems, and how
agroforestry and utilizing the transfer of biologically fixed nitrogen can address them.

4

2

Literature Review

As the worlds’ population continues to increase, global food, fiber, and fuel demands will
intensify accordingly. Land conversion and intensified management of existing agricultural
lands have been historically used to increase production and meet demands but they may not
have a place in an era of sustainability and environmental conscientiousness. Clearing land
results in forest losses, substantial greenhouse gas emissions, and damage to a suite of ecosystem
services critical to human and ecosystem health. On the other hand, intense land management
can result in soil degradation and extensive non-point pollution. In addition, land limitations
pose a complex issue that forces tradeoffs between land-use for forestry, agriculture,
conservation, and urban development. As a result of these concerns, considerable research has
been dedicated to determining how future needs can be met while minimizing the negative
effects of land conversion and degradation.

Agroforestry practices address these concerns by integrating agriculture and forestry onto the
same land to take advantage of natural ecological feedbacks that restore and maintain soil health,
create highly productive systems, and provide a myriad of ecosystem services. This literature
review will cover a basic explanation of agroforestry practices as a whole, followed by an
elaboration on current issues facing food and fiber systems and how agroforestry can address
them. As the focus of this research is on nutrient cycling, particularly nitrogen, most of the focus
will be on soil health. However, the systems proposed in this research have multiple beneficial
bi-products and, in order to make a thorough case for their utilization, these benefits and will also
be detailed.

5

2.1

Agroforestry

Agroforestry offers a path to sustainable production via an array of practices that integrate of
agriculture and forestry onto the same land. The multifunctional properties of agroforestry were
described by the Food and Agriculture Organization of the United Nations as:
“. . . part of the solution to addressing these issues, whether they be environmental, economic or
social. Agroforestry systems include both traditional and modern land-use systems where trees are
managed together with crops and/or animal production systems in agricultural settings. They are
dynamic, ecologically based, natural resource management systems that diversify and sustain
production in order to increase social, economic and environmental benefits for land users at all
scales.” 8

Agroforestry is common among smallholder agriculture properties in tropic and subtropical
regions, where it has been used to diversify income and sustain soil quality for farm-dependent
communities. In temperate regions, the domination of agricultural sectors by conventional
farming methods has made it difficult for agroforestry to gain popularity. Nevertheless, it has
caught on in over five million hectares in the last two decades, showing a slow but global shift
towards integrated agriculture 23.

Many of the issues facing agriculture and forest plantations today stem from growing
populations and climate change. However, an overly simplistic approach to crop and fiber
production may be the primary culprit. Monoculture production of single species can be good
for maximizing yields but ultimately require considerable inputs and will likely degrade land
quality over time. Different approaches that focus on creating sustainable, productive
agroecosystems are needed to effectively address these problems. The synergistic relationships

6

between species and the ecosystem services provided can create systems that require few inputs
and produce multiple products.

To sustain future populations, agriculture and fiber systems must be capable of consistent
production while maintaining soil and land quality. These management practices must focus on
maintaining soil carbon and soil structure, controlling erosion, maintaining or improving soil
quality, increasing nutrient cycling efficiency, reducing export of nutrients, and reducing
required inputs such as fertilizer and pesticides 24. Agroforestry integrates agricultural and
forestry goals and practices to utilize interspecific benefits that maximize nutrient, water, and
light use efficiency 25. This results in more diverse, productive, and profitable agroecosystems
that support sustainable land-use 26. These systems not only encourage sustainable ecological
cycles but also diversify crop production for land owners, making both the land and the owner
more resilient to future challenges.

The World Agroforestry Center (ICRAF) called attention to the potential of agroforestry by
pointing out several ways that its practices can address the Millennium Development Goals set in
2000. Sustainable agroforestry practices can 27: .
1. Help eradicate hunger through basic, pro-poor food production systems in disadvantaged
areas based on agroforestry methods of soil fertility and land regeneration
2. Lift more rural poor from poverty through market- driven, locally led tree cultivation
systems that generate income and build assets
3. Advance the health and nutrition of the rural poor through agroforestry systems;

7

4. Conserve biodiversity through integrated conservation-development solutions based on
agroforestry technologies, innovative institutions, and better policies;
5. Protect watershed services through agroforestry- based solutions that enable the poor to
be rewarded for their provision of these services;
6. Assist the rural poor to better adapt to climate change, and to benefit from emerging
carbon markets, through tree cultivation
7. Build human and institutional capacity in agro- forestry research and development
The term “sustainable management” has surpassed dealing only in terms of productivity, and
now encompasses multiple environmental indicators such as biodiversity, soil health, carbon
sequestration, and water protection, among others 28. Nair (2012) addressed the need for an
“evergreen revolution”, stating that what is most critical now is a global agricultural revolution
that surpasses the production levels of the green revolution of the 1960’s while utilizing
agricultural and forestry technology that promotes environmental sustainability.

2.1.1 Global Agroforestry Trends
Agroforestry practices are arguably millennia old, as similar concepts utilizing the benefits of
trees in agriculture have been found in indigenous agricultural systems. Today, roughly 1.2
billion people depend on agroforestry farming systems in some form around the world 29. It is
practiced primarily by smallholder farmers in developing tropical countries that cannot afford the
expensive input-dependent agricultural methods of the industrial temperate countries.

True measures of land area dedicated to agroforestry are difficult to determine. One method is
measuring the amount of agricultural land that has partial tree cover and assuming these trees

8

were intentionally integrated into production. According to one study, roughly one billion
hectares of global agricultural land has more than 10% tree cover 6. In Southeast Asia and
Central America, much of the agricultural land has tree cover exceeding 30% 6. In the tropics,
the true agroforestry land area is disputed. However, the FAO summed up the agriculture and
forestry trends in the tropics very well when it said, “the number of trees in forests is declining,
but the number on farms is increasing”8. One of the major reasons for the growing popularity of
agroforestry in the tropics is a dwindling wood supply due to deforestation 6. This has
encouraged farmers to meet their wood demands by integrating trees into their agricultural
systems.

Agroforestry is less popular in temperate climates, partly because industrial agricultural practices
are more achievable by the developed countries that make up these areas. Tree cover on
agricultural lands is described as “moderate” in West and Central Europe, but elsewhere in
temperate climates tree cover is typically less than 10% 29. However, agroforestry does have a
small but significant foothold in many temperate regions. In the United States, for instance,
agroforestry traces its roots back to the Great Depression of the 1930’s, when tree windbreaks
were established to combat the rampant soil erosion of the Dust Bowl 30. Today, agroforestry
has become more popular in the U.S. for managing rangelands and protecting water from the
harmful agricultural runoff with riparian buffers31.

2.1.2 Common Systems and Characteristics
By its nature, agroforestry is susceptible to a varied set of definitions that may not accurately
portray its purpose. It is more than just planting trees on farmland. P.K.R. Nair32 stressed two

9

characteristics to define agroforestry. First is the deliberate growing of woody perennials on the
same unit of land as agricultural crops and/or animals either simultaneously or sequentially 32.
Second, agroforestry requires a significant interaction between woody and non-woody species
within a system 32. Furthermore, agroforestry systems typically produce more than one
marketable product. Trees themselves can offer a number of economically viable products
including firewood, lumber, fruits, and nuts.

These requirements are not difficult to meet, and thus there is a range of methods practiced
around the world that could be defined as agroforestry. Indeed part of its attractiveness comes
from its room for creativity and ingenuity for landowners to utilize such interactions. As a result,
classification of agroforestry systems is an ongoing and relatively informal subject. However,
some systems have been well established and are considered important pillars to agroforestry as
an agricultural philosophy. These common systems are described in the table below.

Table 1 . Summary and description of common global agroforestry systems32,33.
Practice
Alley Cropping

Description
Single or multiple tree rows grown at a predetermined spacing with agricultural or
horticultural crops planted between to grow simultaneously

Forest Farming (or Culltivation of non-timber forest products under the protection of forest canopy that
multi-strata farming) is altered to allow correct shade
Strips of vegetation (including trees, grasses, and shrubs) that are grown alongside
Riparian Buffers
waterways to capture sediment, nutrients, and other agricultural runoff with their
roots and prevent them from polluting water systems
Combining trees, livestock, and forage plants together to provide better
Silvopasture
microclimatic condtitions for livestock while producing timber
Linear plantings of trees designed to benefit soil and water conservation while
Windbreaks
enhancing crop production
Agricultural systems whre land is cleared of natural vegetation, cropped for a few
Shifting Cultivation
years, and then left untended for several years to regenerate natural vegetation and
(swidden agriculture)
soil quality
Growing agricultural crops between rows in forest plantations during early years of
Taungya
growth. Usually done as an agreement between a government agency that owns the
plantation and local communities who farm between tree rows.
An assemblage of plants including trees, shrubs, vines, and herbaceous species that
Homegardens
are grown adjacently and managed for household consumption

10

Many landowners, particularly in tropical regions, have been practicing the integration of trees
into agriculture for years without having a name for it 6. These smallholder farmers have long
recognized the beneficial interactions between different species. More research is needed,
however, to establish and understand the mechanisms behind such benefits.

The set of practices associated with agroforestry have potential to create sustainable nutrient
cycles, restore soil health, and establish a suite of other ecosystem services while remaining
economically viable. The remainder of this section will cover the many issues facing global food
and fiber systems, particularly in terms of soil health, followed by an argument on how
agroforestry can effectively address them.

2.2

Current Obstacles Facing Food and Fiber Systems

Despite the advancements of the modern era, agriculture is still an important source of income
for many around the world and often the only source of food for those in developing nations.
Although considerable strides have been made towards ending world hunger in recent years, one
in eight people still suffer from undernourishment, primarily in developing countries 34.
Moreover, roughly 1.4 billion people are still living off of less than US $1.25 per day, of which
70% live in the rural countryside and depend on farming for at least part, if not all, of their
livelihoods 35. Growing populations, land degradation, environmental pollution from fertilizer
use, and climate change are all threatening the ability of present food systems to meet global
demands. Additionally, planted forests are currently and will continue to play a large part in
sustaining the global wood supply for future generations, which is why it is so important to
address the issues facing them today. Plantations are largely made up of exotic species that are

11

intensively mono-cropped and have a range of effects on the land and people involved in them.
These forests are usually spaced regularly, at an even age, and lack most of the principal
characteristics of natural forests 36. As a result, they generally do not have very high of
resiliency to environmental, or even social, challenges and are often unsustainable to the land
they occupy. The next few sections will detail a few of the more salient concerns facing
agriculture and plantation forestry in relation to sustainable production.

2.2.1 Population Pressures, Poverty, and Land Limitations
The global population is expected to increase to 9.6 billion people by 2050 with much of the
growth occurring in developing nations 37. As a result, agricultural production must increase by
70-100% over the next few decades to meet the food demands of future populations 38. The FAO
estimates that there is approximately 4.4 billion hectares on Earth that are theoretically suitable
for cropping, of which only 1.6 billion are currently cultivated 1. While cultivated land area
grew by 12 percent over the last fifty years, much of it was at the expense of forests and other
environmentally important ecosystems 1. Furthermore, the practicality of converting further land
for agricultural uses poses an issue. Most of the land has low food potential due to low fertility
and ecological fragility, as well as limited profitability due to a lack of existing infrastructure.
Moreover, the cost of damaging or completely removing the ecosystem services provided by the
existing ecology in these areas is incalculable in the context of advanced climate change.

Forest plantations also present a growing source of land competition. Global wood demand has
been positively correlated to the population and will continue to do so in the future. Although
the range of actual forest planation area is disputable (between 3-7% of global forest area), it

12

undisputedly accounted for 35% of global wood supply in 2000 39. This number is projected to
grow consistently into the first half of the 21st century, to 46% by 2044 39. Some experts believe
that with the optimal environment for growth, plantations could contribute two-thirds of the
worlds roundwood supply in the near future 36. This growth, when combined with the expected
agricultural and urban growth will require new innovations for integrating multi-faceted land use
practices into global food and fiber systems 40.

The pressures of growing food and fiber demands and limited land availability ultimately mean
that much of the increased production will have to come from currently cultivated land. It is
estimated that roughly 80% of the future food demand will have to be met through intensified
management of existing global crop systems 1. Historically, the need for increased production
from limited land has been met with intensive, high-input management methods, including
extensive fertilizer and pesticide use. While this form of management is capable of improving
yields, it can also bring about a litany of negative side effects, including increased non-point
source pollution and alteration of soil chemistry, which can lead to degraded soil and limited
production in the long-term 2. Moreover, in developing countries where agriculture is often the
major source of income, access to pesticides and fertilizers can be very limited. In Africa, for
instance, commercial fertilizers can be as much as 2-6 times the cost compared to Europe or Asia
and there affordability has had drastic effects on food production27. In Zambia, studies have
shown that up 90% of maize-dependent families experience hunger for at least three months of
the year, getting much worse during periods of drought 23. In areas like this, increasing yields
through higher inputs is simply not an option.

13

Regardless of location, not having the forethought to engage in sustainable and environmentally
friendly agricultural and forestry practices and instead intensifying global inputs to production
systems would likely trade one problem for another. In order to truly address the problems of
land competition and growing demands, system management must focus on creating healthy
soils with sustainable nutrient cycling systems that do not require large amounts of synthetic
inputs. Taking advantage of biological nitrogen fixation (BNF), for instance, can provide access
to what is an essentially unlimited source of atmospheric nitrogen for plant use. This can be
done with mixed systems approaches that incorporate leguminous species. Practices such as
these are common to agroforestry systems, and utilize ecological concepts to both increase and
diversify production from limited land. For example, a study on African farmers found that
improved yields from incorporating nitrogen-fixing trees reduced household hunger periods by
2-3 months 41. Furthermore, a meta-analysis of studies done on African maize systems found
that the presence of nitrogen-fixing trees systems doubled productivity when compared to
unfertilized fields 42.

Agroforestry as a whole offers a low-cost method of improving yields from limited land and
providing food security, particularly for Africa and other developing areas. In a survey, roughly
43% of farmers who had engaged in agroforestry in developing African countries witnessed a
notable improvement in productivity 43. A study from Western Kenya proved that the presence
of trees on farms provided a safe and stable source of fuel wood for both energy and income 43.
This was particularly beneficial to Kenyan women, who are often responsible for obtaining
fuelwood. Furthermore, in Rwanda it was found that one hectare of agroforestry farmland was
capable of producing 140% of the fuelwood required by a six-person family 43. The excess

14

fuelwood could then serve as source of additional income 44. A study in Mozambique found that
the transition to agroforestry systems helped boost overall household income, largely due to
decreased labor requirements 45. Clearly, agroforestry has enormous potential for catalyzing
sustainable development in rural areas, improve economies in the long-term, alleviate poverty,
and ensure resiliency to climate change despite growing populations and land limitations 46,47.
An increase of production by 1% corresponds to a 1% drop in the number of people living on a
less than $1(US) a day 23. Increased income and food security have cascading effects that
improve the livelihoods of the rural poor. One such effect is improved health conditions.
Agroforestry has been linked either directly or indirectly to health benefits in a number of ways.
The ecosystem services provided by such systems improve the quality of air, soil, and water, all
of which can directly influence the health of individuals who depend on them 6. Agroforestry
has even been linked to improvements in the fight against HIV/AIDS 27. These practices can
generate additional income, improve nutrition, reduced labor demands, and stabilize
communities 48. The foundation of these benefits to human livelihoods lies in the ability of
agroforestry practices to create production systems that are sustainable on multiple fronts. In the
developing areas discussed above, perhaps the most important front to production is enhancing
nutrient cycles. This research will stress the importance of biological nitrogen fixation as source
of nutrients for important food and fiber crops, the results of which can have cascading social
benefits.

2.2.2 Effects of Agricultural Fertilizer Use
A common reaction to higher food and fiber demand from limited land is using more synthetic
fertilizers. However, the use of fertilizers comes at a significant cost to the surrounding

15

ecosystems, particularly riparian systems, and they often do not effectively address soil nutrition
deficiencies. Global fertilizer consumption (measured in N, P2O5, and K2O) reached nearly 187
million metric tons in 2014 and is expected to increased up to over 200 million by 2018 49. Prior
to the mid-twentieth century, the primary source of fertilizer was livestock manure. The
discovery of the Haber-Bosch process for converting nitrous gas into ammonium and improved
methods of extracting rock phosphorous brought chemical fertilizers onto the forefront of
agricultural technologies. They were a major part of the green revolution’s advancements in
improving yields and have been critical to diminishing global hunger. However, the
improvements were not without their drawbacks, such as widespread soil chemical alteration and
degraded water quality 18. Extensive research has shown that the use of agrochemicals can be
extremely harmful to the environment both on-site and off-site 50. A major negative effect of
fertilizer use at the site of application is a growing nutrient imbalance in the soil. For instance,
the heavy application of nitrogenous fertilizers paired with minimal potassium applications has
thrown soil nutrient balance off in many areas 51. Plant nutrient uptake and health requires not
only a sufficient amount of individual nutrients but also an appropriate balance between
nutrients. Thus, fertilizer practices have ironically resulted in extremely low use-efficiencies of
the nutrients added to the system 51.

The on-site nutrient imbalance created by fertilizer use has disastrous potential for future crop
yields and is a major concern for sustainable agricultural production. However, some of the
worst effects of fertilizer use occur outside of the fields on which they are applied. Synthetic
nitrogen and rock phosphates have become increasingly less efficient for fertilizing crops and
more environmentally harmful 52,53. It is estimated that only between 25-50% of nitrogen

16

applied to croplands is subsequently taken up by crops 51,53. Only 20% of phosphorous fertilizer
is estimated to be utilized by plants within the year that it is applied 53. Since there is no process
for creating synthetic phosphorous, this level of inefficiency is particularly cost-ineffective when
considering global prices are rising and supplies are diminishing. Additionally, the inefficient
use of the applied nitrogenous and phosphate fertilizer increases the movement of nutrients into
nearby bodies of water and water tables. The abundance of nutrients in the water leads to the
production of algae blooms, the decomposition of which reduces oxygen levels in the water,
making them toxic to living creatures. This process is called eutrophication, and currently
affects over 60% of the US coastal rivers and bays, and is responsible for the “dead zone” where
the Mississippi River feeds into the Gulf of Mexico 54. Agricultural fertilizers serve as a source
of non-point pollution that is extremely difficult to prevent from reaching bodies of water 55.
This form of pollution is a widespread problem in much of East and Southeast Asia, Europe,
North America, and Central Asia 1. Nitrogenous fertilizers are also responsible for emitting
nitrous oxide, a lesser known greenhouse gas that has a higher global warming potential than
carbon dioxide 56. A study conducted on the carbon footprint of the Tropicana orange juice
company found that as much as 58% of their emissions were a direct result of the application of
nitrogen fertilizers 50. The Intergovernmental Panel on Climate Change (IPCC) reported that a
substantial amount of the global nitrous oxide production is tied to fertilizer use, representing
another negative externality from intensive agriculture 57.

Agroforestry is capable of alleviating fertilizer requirements as well as their pollution pressures
on water sources. Multiple studies have shown the potential for leguminous species to be
utilized as a synthetic nitrogen replacement 20–22, which when combined with other agroforestry

17

practices can enhance nutrient cycling on multiple fronts. This will be discussed in greater detail
in sections 2.3 and 2.4. Moreover, riparian buffers (a common agroforestry practice) present a
prime opportunity to protect waterways and wetlands from agrochemical runoff. Due to the
widespread runoff issues in the modern agricultural systems, primarily in temperate systems,
riparian buffers have grown in popularity in both the United States and in Europe. The deep
roots of trees and shrubs in these agroforestry systems are capable of capturing nutrient and soil
runoff before it enters rivers, streams, and other bodies of water during rain events. In Poland,
for instance, a series of shelter belts and buffers has significantly reduced water pollution from
nitrate runoffs and improved water cycling in the region 26. This has negated considerable
potential damage from continued fertilizer use in the area. The ability of riparian buffers to
reduce water pollution is well documented. However, the studies presented in Chapters 3 and 4
emphasize the ability of mixed systems to not only protect marginal lands (such as riparian
borders), but also to improve their soil health and even make them productive.

2.2.3 Biodiversity, Pests, and Disease
It is a generally held belief that monocultures are more susceptible to damage from pests and
disease when compared with natural ecosystems. This is particularly true for plantation forests,
which often do not receive the same pesticide inputs as agricultural systems 58. This
susceptibility is largely thought to be due to the extreme lack of diversity in most plantations,
which typically only grow one, often non-native, species. In diverse systems, the variety of
plants and animals that occupy an ecosystem result in a series of checks that can prevent any one
species from becoming dominant, including pests 59. An inordinately large number of instances
regarding devastating pest-related losses have been recorded monocultures when compared to

18

more diverse systems, indicating that biodiversity can reduce pest-related damage 59. Although
this concept is largely favored, it is difficult to research. However, studies have isolated a
number of factors that explain the prevalence of pest and disease in plantations 36,39,58,60:

1. Monoculture stands offer a large, uninhibited food source to any pest that specifically
preys on the species. This allows for extreme population bursts.
2. The proximity of trees in plantations allows for contact of branches and roots allow rapid
spread of infections from both insects and disease
3. Lack of genetic variation within single species plantations, particularly when planted by
cuttings or in coppicing rotations, reduces the potential for genetic resistance to pests
4. Single species growing on the same site over long periods of time allows for pests and
disease populations to build up.
5. Use of exotic species in areas means they do not have the populations of natural
controlling agents that help reduce pest populations
6. Poor site selection can lead to tree stress, leaving them susceptible to attack

Some studies have particularly stressed the need for genetic diversity within single species
plantations and suggested it is critical to stand health 36. This notion does not bode well for
many short-rotation tree plantations, which are typically planted from cuttings and coppiced for
several rotations, resulting in extremely low genetic diversity. This form of plantation
establishment is typical to short-rotation woody crops (SRWC’s), which are already very
common in Asia and growing in popularity elsewhere for producing biofuels. Moreover,

19

warming temperatures will likely provide more suitable conditions for pest population growth,
especially in areas where pests are usually suppressed by seasonally cold temperatures 36.
In summation, monoculture forest plantations and agriculture systems lack the natural constraints
that keep pest and pathogen populations in check and provide suitable conditions for quick and
devastating spread of infection 36. Frequent monitoring and applications of pesticides are the
most common tools against pests but may not represent a long-term solution. Some have turned
to integrated pest management (IPM) methods that seek to take advantage of the natural
constraints on pests that are missing in plantations 59. Agroforestry practices, however, offer a
higher level of plant diversity that can lead to a greater number of predators and reduced plant
stress and susceptibility. By its nature, agroforestry combines multiple species of crops, trees,
and animals to operate in niche complementarity 6. Agroforestry realizes that interspecific
competition can often be less harmful than the intraspecific competition for resources found in
monocultures. Perhaps the most exciting benefit of biodiversity in agroforestry is its potential to
offer a low-cost form of pest control. As more species are integrated into a production system,
habitat complexity increases and provides the conditions necessary to increase the presence of
natural predators, which in turn reduce pest61. As a result, the use of pesticides is likely to be far
more infrequent when compared to traditional agriculture. This idea has been solidified by
multiple studies that have shown natural enemy populations to be higher and damage from pests
to be lower in agroforestry systems 62. In a study on coffee plantations, for instance, it was found
that coffee agroforestry bolstered pest-control services 63. In fact, these effects were seen even in
the presence of an adjacent forest, showing the importance of perennial species to natural enemy
habitat. Perennial biodiversity has also been shown to reduce the presence of weeds by

20

increasing shade and improving soil quality, which allows crops to out-compete advantageous
weed species 62.

The systems proposed in these studies are inherently biodiverse and can result in higher system
resiliency. In fact, a similar study on pecan-cover crop alley cropping system in the United
States found that the presence of the cover crop improved predacious species populations, such
as lady beetles, which provided pest control functions for the trees 64. The economic benefits
linked to biodiversity have been estimated at trillions of dollars each year 65. It has direct
economic, cultural, and aesthetic benefits as well as a number of indirect benefits towards other
ecosystem services. Since biodiversity is intrinsic to agroforestry, the benefits of the practices
are bolstered.

2.2.4 Future Effects of Climate Change and Possible Feedbacks
The struggle to feed future populations on existing lands suffering from degradation and
intensive management is further complicated by global climate change. By now there is little
disagreement among experts that it has become a serious threat to future generations. When it
comes to agriculture, a changing climate can have an array of effects on yields and sustainable
production depending largely upon the location of the system at both a global and local scale.
This makes predicting future conditions and subsequent impacts on agriculture extremely
difficult.

Two key components of climate change are increased atmospheric CO2 concentrations and
warmer global climates, both of which to some degree may benefit terrestrial plant ecosystems,

21

including agricultural crops. In fact, it has been suggested that a moderate warming of up to 3°C
could have overall positive effects on agricultural yields in the first half of the 21st century,
provided adequate moisture levels 5. It is likely, however, that any positive effects of global
warming on agriculture in the first half of the century will be vastly outweighed by the second
half as temperatures would continue to warm above 3°C 5. The other predicted effects of climate
change also have potential to negate any positive effects, such as changes in precipitation
patterns and increases in extreme weather events.

What is perhaps more important is that the effects of climate change will not be homogenous
across all nations. While warmer temperatures may extend growing seasons and benefit crops in
temperate regions like the United States, Canada, Scandinavia, and Russia, they will likely have
disastrous effects for the tropics 1,5. The disproportionate negative effects of climate change on
tropical developing nations is troubling, as these countries have high existing poverty rates and
dependence on agriculture. The aggregation of all climate effects is likely to increase the
number of undernourished people by 10-150 million, most of which are expected to come from
these areas 1. Yields of maize, for instance, are expected to decrease 10% by 2055 in Africa and
Latin America on smallholder farms as a result of climate change 66. Many farm-dependent
families in these countries already struggle to survive and are extremely vulnerable due to small
farm sizes, limited access to agricultural technology, and other non-climate stressors 3.
Moreover, the presence of disease, conflicts concerning natural resource scarcity, or ethnic strife
may be intensified by migration away from climate-vulnerable areas and have further impacts on
the rural poor 3,4.

22

The biophysical and biogeochemical effects of climate change on agricultural ecosystems have
tremendous potential to result in positive feedback cycles, primarily through increased
emissions. Agricultural activities currently contribute between 5-6 billion tons of GHG
emissions annually (CO2 equivalent) and together with deforestation account for a third of total
human emissions 1. It is responsible for 25%, 50%, and 75% of the global CO2, N2O, and
methane emissions, respectively 1. Although it is uncertain how these emission levels will
change in the coming decades, the rising demand for food implies a likely increase. In fact, it is
estimated that they could increase by as much as 40% 67, which would likely go on to exacerbate
the effects of climate change, resulting in a positive feedback loop.

Soil degradation as a result of intensified land management can also result in higher respiration
rates of carbon dioxide through depletion of soil organic carbon 28. Land conversion from nonagricultural to agricultural land results in emissions of carbon lost in land clearing, reduced
carbon storage ability, and increased emissions on that land over time 67. BajĹželj and Richards
(2014) identified three major pathways for agriculture-climate change positive feedback loops
regarding land conversion to exist:
1. Warming temperatures and other factors result in reduced yields, which incurs greater
land conversion for agricultural purposes in order to supply global food demand
2. As existing cropland becomes less suitable for agricultural production farmers will
seek out new land
3. Alterations in water supply and demand for irrigated agriculture result in cropland
expansion to make up for reduced yields.

23

The above mentioned feedback loops will be likely to hit the poverty-stricken tropic nations the
hardest. Growing populations in these areas will mean greater food demand in the future, but
reduced yields from warming temperatures and other climate change effects will make it difficult
to supply on their own. As a result, these nations would likely need to begin importing massive
amounts of food. Doing so would require money they simply do not have. Thus, the pressure on
the land would increase, resulting in a vicious cycle of hunger and environmental degradation.
These potentially disastrous effects have inspired multiple governments, non-profits, and even
private industries to seek out ways to mitigate climate change. Carbon sequestration, or the
ability of plants to capture and store carbon from the atmosphere, is perhaps the most important
tool we have for reducing atmospheric carbon and minimizing these negative effects in the long
run 68. As a result, much attention has been paid to the potential for mixed tree-cropping systems
to sequester large amounts of carbon while producing other products, such as food. The
Intergovernmental Panel on Climate Change has stated that agroforestry has the highest potential
for carbon sequestration among land management regimes 6. It offers particularly high potential
in that it can create important synergies between mitigation and adaptation activities 69.
The importance of agroforestry to carbon-based climate change mitigation comes from variety of
factors. Mbow et al. (2014) noted that its enormous potential comes from the combination of the
ability of trees to sequester carbon in their biomass and the vast amount of land area that is
suitable for agroforestry practices, including degraded land. This is true even in the modern
agriculture systems of temperate regions. Recent research has estimated that the United States,
for instance, could offset as much as a third of its carbon emissions by engaging in practices such
as silvopasture and alley cropping on even a small portion of its agricultural land base 64. This
could sequester over 500 Tg of carbon annually. Global estimates suggest that over a 50-year

24

period total carbon sequestration could range between 1.1-2.2 Pg of carbon annually, although
this is based on limited information regarding total agroforestry land area 70. Additionally, it can
reduce extraction and deforestation pressures on existing natural forests 6. This is particularly
relevant to tropical regions.

Since carbon is stored in the biomass of trees for a limited time, there has been debate regarding
the non-permanence of this carbon sink and its legitimacy as a climate change mitigation
technique. However, capturing and storing atmospheric carbon for even a short period of time
can be beneficial to mitigation, particularly if the biomass from these trees are turned into longlasting wood products or burned to offset the use of fossil fuels. Additionally, a substantial effort
has been put forth to understand the role of soil organic carbon (SOC) as an atmospheric carbon
sink in both natural forests and agroforestry regimes. As it turns out, research has found that
these systems are also capable of improving SOC pools. It is estimated that as much as 30-300
Mg of carbon per hectare can be stored in one meter of soil on which agroforestry practices are
utilized 25. This also meets the requirements of 10-15 g/kg of carbon described by Osman (2014)
to feasibly restore soil quality on degraded lands. Additionally, enhanced nutrient cycling in
these systems can result in greater productivity of both above and belowground biomass. Lastly,
the generally low levels of soil disturbance common in agroforestry is likely to promote the
stabilization of SOC for long term storage 25.

In addition, to the enormous carbon sequestration potential of agroforestry, the lack of required
agricultural inputs makes them even more cohesive to climate change mitigation. Perhaps the
most important GHG emitted by agricultural activities is nitrous oxide (N2O). Despite its

25

relatively minor abundance in the atmosphere compared to carbon dioxide, its warming potential
makes it an emission of great concern. A major source of nitrous oxide emissions are soil
microorganisms, which create N2O gas during nitrification and denitrification. Large inputs of
synthetic nitrogenous fertilizers common in modern agricultural systems lead to high emissions
of nitrous oxide. Additionally, frequent tilling can also lead to higher biological activity and
emissions 71. Agroforestry not only reduces the need for fertilizer inputs but inherently involves
minimal tilling activities in order to protect perennial tree roots, thereby mitigating nitrous oxide
emissions.
In summation, there is a very close link between soil degradation, climate change, and poverty 72.
Agroforestry offers a pathway to not only mitigate climate change at the global scale but also to
create locally adapted systems that are resilient to its ongoing effects. The systems designed in
this research can specifically reduce nitrous oxide emissions, which will be discussed in later
sections.

2.2.5

Land and Soil Degradation

Since expanding global agricultural area would be neither sufficient nor sustainable in meeting
future food and fiber demands, enhancing yields on existing croplands is the most promising
option. Over the last 50 years, the world’s agricultural production has grown between 2.5 and 3
times while cultivated area has only increased by 12 percent 1. Unfortunately, present
widespread land degradation as a result of intensive agricultural practices has made improving
yields on existing croplands risky.

26

Global estimates suggest that roughly 38% of the global land area could be considered degraded
2

. Soil resources are critical to reaching stable food production. The degradation of soil by

improper agricultural and forestry practices can be tied to the failure of many past civilizations
73

. Recent developments in agriculture have made intensive and unsustainable land management

more common. These developments include increased availability of synthetic fertilizers,
herbicides, and pesticides, improvements in tillage, planting and harvesting equipment, and
increased use of subsurface drainage and irrigation 28. Although these developments have
improved global crop yields and reduced hunger, they ultimately have led to widespread land
degradation. The responses of soil health to these practices have been particularly overlooked in
recent decades and little concern has been paid to long-term sustainability.

Land degradation usually refers to damage to a set of ecosystem services provided by a particular
area, which are very often a direct result of soil degradation. In comparison, soil degradation
refers specifically to the processes by which human actions cause a decline in soil quality and as
a result it is less suited for achieving a particular goal, such as crop production 38. Lal (2015)
describes four types of soil degradation: physical, chemical, biological, and ecological.

Physical degradation typically refers to the destruction of soil structure, which can lead to
compaction and increased erosion from wind and water 2,28. This greatly compromises the
ability of soil to support long-term agriculture. Soil compaction is the physical consolidation of
soil through the application of a force which disrupts soil structure, reduces porosity, and limits
movement of water and air within the soil. Compaction results in stunted plant growth,
malformation of roots, formation of large soil clods, and often standing pools of water in

27

agricultural fields 2. Reported yield reductions due to compaction have been as high as 90% in
areas of West Africa and 50% in Europe and North America 74,75. Since it is directly tied to
frequent use of heavy agricultural machinery, compaction is more likely to impact cropping
systems in developed nations. On farm losses from compaction in the United States, for
instance, were estimated at $1.2 billion annually 2.

Erosion is the process by which materials from earth surfaces (such as soil) are forcefully
detached and transported elsewhere by a variety of forces, most notably wind and water.
Currently erosion likely poses the greatest threat to the physical health of agricultural soils
worldwide, mostly due to improper soil management. Land management practices that
contribute to erosion include continuous mono-cropping cycles, cultivation on slopes, and lack of
vegetation cover between growing seasons. While certain levels of erosion are natural and
acceptable, many agricultural areas have levels of erosion well above normal levels. On-site
effects of erosion are numerous and devastating. They include exposure of roots, deformation of
terrain, decline in soil quality, and reduced capability of ecosystem functions. Loss of soil
organic matter and nutrients is a significant contributor to soil quality decline. It is estimated
that a 10mm loss of top soil can mean a loss of as much as 350 kg N, 90 kg of P, and 1000 kg of
K per hectare 76. As a result of this and the other effects of erosion on agricultural soils, yield
reductions of 30-90% have been recorded in West Africa and 20-40% in the US Midwest 2.
Total land area that is prone to erosion is estimated at between 1.65 and 1.96 billion hectares, of
which 50% is severely eroded and likely to be abandoned 2,77. This represents roughly 15% of
Earth’s total land area.

28

Chemical soil degradation is characterized primarily by nutrient depletion, leaching, and
acidification 28. Other sources include contamination with industrial wastes or by-products and
reduced cation exchange capacity (CEC). Estimates from the early 1990’s suggest the total
chemically degraded land area in the world is approximately 240 million hectares 2. Of all the
forms of chemical soil degradation, nutrient depletion and leaching have perhaps the largest
implications for future food production. It is the process by which soil nutrient stocks gradually
shrink as result of nutrient mining, primarily from agricultural activities 51. The losses from
over-cultivation, soil erosion, and leaching combined with the lack of nutrient replenishment
result in severely degraded soils. Although leaching happens even under natural vegetation,
agricultural activities greatly exasperate leaching losses 78

Globally, only half of the nutrients removed by crop production are replaced by fertilizer or other
nutrient addition process 8. This results in an average of 50 kg of nutrients lost per hectare on an
annual basis 2. Nutrient depletion affects 136 million hectares worldwide, 97% of which occurs
on developing and least developed countries 2,51. This is extremely true in Sub-Saharan Africa,
one of the few locations where yields have been unable to substantially increase over the last
century of agricultural improvements. In countries like Rwanda, Tanzania, Mozambique and
Niger, nutrient depletion has been responsible for losses that make up as much as 25% of their
Agricultural Gross Domestic Product 79. For the sub-continent as a whole, nutrient depletion has
reduced AGPD by 7% for a total value of over 4 billion US dollars, an amount that greatly
exceeds the external assistance received for agricultural development from 1990-2000 79. The
continued nutrient depletion in these regions has damaged socio-economic welfare and
environmental quality alongside crop yields 51. Worldwide, deficits in NPK (nitrogen,

29

phosphorous, potassium) resulted in a 27% reduction in yields on average 51. Although the
majority of severe chemical soil degradation from nutrient depletion obviously occurs in
developing countries, at the turn of the last century there was no country that was not affected by
nutrient problems 51.

Although to a much lesser degree, acidification of soil is another contributor to chemical soil
degradation that is exasperated by agricultural activities 2. Soil acidity refers to the concentration
of hydrogen (H+) and hydroxide (OH-) ions in a soil solution and is measured as pH. A pH
higher than 7 is considered basic (higher concentration of hydroxide ions) and a pH lower than 7
is acidic (higher concentration of hydrogen ions). The solubility of micronutrients is greatly
affected by the acidity of soil, making it an important trait to consider for agricultural purposes.
A slightly acidic soil of around 6.5 is optimum for most terrestrial plant species 2. Agricultural
practices can accelerate the acidification of soil as a result of nitrogenous fertilizer applications,
leaching of nitrates, removal of produce, and to some extent, the build-up of organic matter 80.
The additions of ammonia-based nitrogenous fertilizers are particularly good at acidifying soil.
The chemical process that converts ammonium to nitrate in the soil acidifies soil with biproducts. Under normal conditions, plants discharge an alkaline substance as they uptake the
nitrogen compounds and the soil remains relatively balanced 2. However, when nitrate is leached
from the soil, which increases in the presence of agricultural activities, the balance of acidic and
alkaline additions is broken in the favor of acidification 2. There is little information regarding
the direct effects of soil acidification on crop yields, but it is well known that extremely acid
soils (below a pH of 4.8) are toxic and can lower production and even kill plants that are not
adapted to them 81. These soils in general have reduced nutrient recycling, particularly in terms

30

of nitrogen. The nitrogen cycle depends on the activity of soil microbes which are negatively
affected by soil acidity. The Rhizobium bacteria that are a responsible for biologically fixed
nitrogen in legumes are particularly affected by altered reduced soil pH 81.

An important part of soil quality is sustaining biotic habitat. Soil biota are responsible for
numerous biological, biogeochemical, and physical processes on which soil quality heavily
depends. These processes include retention and delivery of nutrients to plants, maintenance of
soil structure, pest and pathogen control, bioremediation of pollutants, and erosion control 24.
The positive link between soil biodiversity and the provision of terrestrial ecosystem services is
well documented 82. Of the three forms of soil degradation discussed thus far (physical,
chemical, and biological), biological soil deficiencies are the most complex and suffer from large
knowledge gaps 53. The existing research suggests, however, that it is extremely important to
sustainable agriculture and soil management. The estimated economic benefit of the ecosystem
services it provides is estimated to be 1.5 quadrillion U.S. dollars globally 83.

Microbial populations are perhaps most critical to soil biological health and agricultural
production as they are extremely important in nutrient provision for crops. Because of their
ability to catalyze redox reactions, soil microbial organisms are critical mediators of the cycling
of carbon, nitrogen, and other nutrients 53. Unfortunately, modern industrial agriculture
primarily focuses on chemical manipulation with fertilizers as opposed to improving soil biotic
health 53. Moreover, agricultural practices such as frequent tillage often have a negative impact
on soil biota and the ecosystem services they can provide 84. Tillage disturbs soil habitat,
decreases soil structure, and temporarily increases the decomposition rates of organic matter on

31

which microbial populations depend. Soil organic matter (SOM) is an important indicator of
biological soil quality. Roughly half of SOM is carbon, and is referred to as soil organic carbon
(SOC). Roughly 2% of SOC mass is attributed to soil organisms 53. Despite the small
proportion they make up, they ultimately determine the fate of most organic compounds. The
depletion of the SOC pool via frequent tillage also results in diminished pools of plant available
nutrients such as N,P, and S 28.

Ecological degradation is more accurately interchangeable with land degradation and refers to
the process by which the ecosystem services provided by healthy soil (and other components of
land) are diminished by human activities. These services include elemental cycling, water
purification, and decline in net primary productivity. What is most alarming from the ecological
perspective is the positive feedbacks between soil quality and diminished ecosystem services.
For instance, depletion of soil organic matter reduces nutrient cycling, which in turn diminishes
plant production of organic matter. Very often it can occur where once the process of soil
degradation is started, it can perpetuate itself into a rapid decline or downward spiral 53.
Agriculture contributes to this through practices discussed in the previous sections such as monocropping, over-cultivation, frequent tilling, and application of pesticides and fertilizers 85.
Globally, ecosystem services have been reduced by as much as 60% as a result of intensive
agriculture and deforestation 86.

Since soil health no longer speaks only in terms of production, it must be approached on multiple
fronts. In order to address physical, chemical, biological, and ecological soil degradation
effectively, a holistic approach utilizing agroecological concepts is required. Mixed systems

32

promote healthy soil structure, chemistry, and biology through inter-specific relationships and
occupation of varying ecological niches, creating highly efficient production systems. Rattan Lal
(2015) described several important strategies to improve soil quality and reduce degradation,
including (1) preventing losses from erosion, (2) encouraging the accrual of SOC, (3) improving
availability of macro and micro-nutrients, (4) increasing soil biodiversity and enhancing
microbial processes, and (5) enhancing rhizospheric processes. Through the integration of
multiple species and enhancement of their relationships, agroforestry creates cascading positive
feedback loops between soil biology, chemistry, and structure that ultimately address all of the
aforementioned soil degradation issues.

A of the common trait of agroforestry systems is reduced tillage, largely due to the presence of
perennial root systems susceptible to damage from such activities. Low-tillage agriculture has
been hailed as a sustainable land management technique. This, when combined with the fact that
tree root systems provide added structure, explain why agroforestry systems are remarkably good
at reducing losses from soil erosion 43. Besides erosion control, low-tillage agroforestry can help
build organic matter and reduce the risk of root compaction from agricultural equipment 2. A
recent study in India found that across all agroforestry systems there was an average of 1.17 –
1.65-fold increase in organic carbon within a few years of establishment 87. Soil organic matter
and carbon content are extremely important to soil health. It is responsible for improving
nutrient cycling efficiency, restoring physical soil quality, and increasing soil moisture, among
multiple other benefits 43.

33

Buildup of organic matter is linked to increased fungal and microbial biomass, which in turn
increase carbon use efficiency and soil carbon sequestration 88. Moreover, soil organic matter
has been proven to encourage arbuscular mycorrhizal fungi populations, which enhance plant
uptake of phosphorous and water, and strengthen disease resistance 89. Thus many of the
benefits of improved soil organic matter content have to do with their effects on microbial and
fungal populations, particularly when it comes to nutrient cycling. Inputs, such as leaf litter from
trees, provide microbial communities with the carbon needed for energy. These microbes then
go on to promote decomposition of nutrients into soil, nitrogen fixation, uptake of phosphorous,
and promotion of plant growth 90. Much of the nutrients in decaying leaf litter inputs from trees
comes from deep layers of the soil that would otherwise be inaccessible to shallow-rooting crops,
thereby adding to the overall nutrient budget of a system. Soil biota are even capable of
influencing erosion by helping form stable aggregates, improving macroporosity, and reducing
runoff 87. Through constant inputs of organic matter from trees in agroforestry systems, both soil
organic carbon and soil biota are significantly improved 91.

In a study in the hill-regions of Nepal, three systems were compared: a traditional monoculture, a
fully developed agroforestry system, and a newly established agroforestry system (two years
old). The results showed that agroforestry systems improved multiple soil health indicators such
as cation exchange capacity, soil pH, organic matter and nitrogen content, and even aluminum
content 92. Data revealed that these indicators improved even in the two year old system, proving
that the soil health improvement could be attributed to the differences in land management 92.
Agroforestry practices improve nutrient supply and cycling in a number of ways. The
mechanisms we’ve discussed thus far, such as organic matter build up, access to nutrients deep

34

in the soil by trees, and improved soil biota populations, all result in enhanced nutrient cycling.
Moreover, the incorporation of nitrogen-fixing trees, which is very common to agroforestry,
provides additional nutrients to the system budget. The transfer of this nitrogen to agricultural or
forestry crops can occur in a few different forms, which will be the topic of discussion in
following sections. In the tropics, the use of nitrogen fixing species (also known as fertilizer
trees) has been critical to improving yields 93. Low availability of nitrogen and phosphorous are
major constraints on degraded land in these areas and utilizing the relationships between
legumes, bacteria, and interspecies interactions are integral to restoring soil quality 86. The
ability of agroforestry practices to restore, improve, and maintain soil nutrient cycling will be the
subject of the remainder of this review, dealing particularly with nitrogen relationships between
species.

2.3

Belowground Species Interactions and Nutrient Cycling in Agroforestry

Agroforestry systems are capable of tremendous impacts on belowground ecosystems and the
rhizospheres of individual species. Studies have shown that the addition of even a single tree
species into agricultural systems can positively affect soil microbial communities and nutrient
cycling 94. Through various mechanisms, trees are capable of increasing soil carbon, capturing
leached nutrients, and improving the physical characteristics of soil 95. They are also capable of
accessing resources in the soil that shallow-rooted agricultural plants are not, which leads to
greater overall resource use efficiency 96.

Although trees have considerable potential to positively impact agricultural systems, there is
perhaps equal potential for negative interactions. Positive relationships between species (both

35

above and belowground) can be beneficial to both (mutualistic), beneficial to one and neutral to
the other (facilitative), or beneficial to one and harmful to the other (parasitic) 96. Competition
refers to a relationships where both species are harmed. While plants can have multiple
relationships in regards to different factors (water, nutrients, light, etc.), the emphasis of a
relationship label is based on the overall effects on growth and yield of the species involved.
Considerable research has concluded the most frequent limitation in agroforestry systems is
competition for water 97,98. However, there have been several instances recorded where
hydraulically lifted water from trees can result in higher water use efficiency in soil 9,96,99,100. As
a result, it is critical to understand the varying niches and needs of the species being incorporated
into the same space. Fernandez et al (2008), for instance, stresses the need for selecting tree
species that are capable of reaching nutrients that adjacent agricultural crops would not be able to
access, thereby reducing competition. Moving forward, we will examine in detail the potential
effects of alley cropping, a common agroforestry practice, on nutrient cycling and use by various
species combinations.

2.3.1 Alley Cropping and Nutrient Cycling
There are several potential mechanisms that trees utilize to alter nutrient cycles in agricultural
systems. Perhaps the most important among these include: (1) accessing nutrients in the soil
profile unavailable to shallow-rooted plants, (2) adding these nutrients to the soil through leaf
litter, (3) adding C and other compounds to the soil through root turnover, (4) serving as
catchment for leached nutrients from higher soil profiles, (5) improving microbial communities,
and (6) improving soil structure 95,96,98. Furthermore, nitrogen-fixing species have high potential
to improve nitrogen contents in soil through a few different processes, which will be discussed in

36

detail later on. All of these mechanisms can positively affect an agroecosystem. However,
without careful planning and a developed understanding of individual species these relationships
could lead to competition for resource use and result in lower yields and growth.

Alley cropping is a common agroforestry practice whereby agricultural crops are grown in the
space (or “alleys”) between organized rows of planted trees spaced at a pre-determined distance
(Table 1) 32. It is a widely common form of integrated agriculture in both temperate and tropical
ecosystems and relies heavily on the idea of complimentary root patterns and resource use
between species. The ability of multiple species to create beneficial relationships between one
another and soil health in this practice has been proven in multiple studies. For instance, a longterm study on a silver maple and corn-soybean alley cropping system found that multiple soil
parameters were signicantly improved over the course of the 20-year rotation, including water
stable soil aggregates, soil C and N, and enzyme activity 101. Additionally, alley cropping
provides a method for year-round biomass production and results in a higher annual production
level per unit area compared to mono-culture systems 10. This is particularly important for
tropical agriculture systems.

Alley cropping is particularly beneficial to soil nutrient cycling in part because of the improved
soil microbial habitat it provides. Abundant carbon exudates from tree roots into the soil provide
sources of food for microbe populations, which in turn improve microbe-dependent nutrient
levels, such as nitrogen 102. Additionally, a study on L. leucocephala hedgerows in Nigeria
showed higher levels of larger decomposers like earthworms as well 103. These changes tend to

37

occur primarily in the higher organic matter layers of the soil, which activates the deposited
nutrients from tree leaf litter, making them available to both trees and crops 104.
Provided that trees and row crops are not competing for the same resources, this practice has
high potential to improve nutrient cycling in agroecosystems. Tree leaf litter and roots help to
improve levels of nutrient-rich soil organic matter and microbial activities to maintain constant
nutrient inputs. An alley cropping system studying multiple different potential trees in Nigeria
showed that in all systems recycled N, P, K, Ca, and Mg were 12, 9, 11, 35, and 24 times higher
than controls, respectively 105. Its capacity for nutrient cycling makes no-tillage alley cropping
an important practice for tropical systems, where fertilizers can be expensive and hard to find 106.
It is important to note that agroforestry and alley cropping in particular provide improved
methods for producing not just agricultural crops but also tree products such as fruits, nuts, and
woody biomass. Although much of the focus of agroforestry to date has been on the crops
grown between alleys with tree-based products generally considered as secondary products, a
trend towards equal focus has been seen in recent research. As mentioned before, it is estimated
that plantation-grown trees will likely contribute to nearly 50% of the total global roundwood
supply by 2050 107. Thus, the intensive production of wood fiber on plantations will take on a
new importance in the years to come. Fortunately, as has been made clear throughout this report,
agroforestry provides a prime opportunity to combine production of livestock, crops, and trees.
Considerable studies have been produced regarding alley cropping in recent years with heavy
emphasis on tree crops. Part of the reason for this is the ability of this practice to reclaim
infertile and marginal lands, thereby restoring soil health and improving total land production for
the farmer. In the Midwestern U.S., alley cropping using trees that produce both high quality
lumber in the long-term and fruits or nuts in the short-term have gained considerable popularity

38

108

. Common trees for this include black walnut, pecan, and chestnuts, all of which produce

valuable nuts that can be sold prior to tree harvest for lumber. A few studies have set out to
explore how herbaceous species planted between tree alleys could benefit the growth, yield, and
nutrient status of the tree crops. For instance, a study on multiple herbaceous ground covers
grown between rows black walnut were shown to improve walnut foliar nitrogen content 19. In a
pecan-kura clover alley cropping system in Missouri, soil organic carbon and soil enzyme
activities were significantly higher when compared to controls 22. Water-stable soil aggregation
was also dramatically improved by the presence of both species 22.

2.3.2 Biological Nitrogen Fixation and Alley Cropping
Biological nitrogen fixation (BNF) refers to the process by which certain species of plants are
able to transform atmospheric nitrogen (N2 gas) into plant available forms for their own use.
This transformation requires the enzyme nitrogenase, which is found only in certain soil
microorganisms 11. To achieve BNF, certain plants create symbiotic relationships with these
bacteria by providing them with sugars for energy in exchange for the transformed nitrogen. The
most common relationship of this kind found in nature is between legumes and Rhizobium
bacteria 11. Rhizobia infect the root hairs of leguminous plants which then form “nodules” where
the bacteria are housed and used for BNF. Figure 1 describes the chemical process by which
legumes transform this nitrogen.

39

Figure 1. Chemical process for biological nitrogen fixation occurring in the root
nodules of legumes11

Biological nitrogen fixation is critical to natural and productive soils. In agroforestry, nitrogen
fixing trees are widely used in alley cropping systems to benefit associated agricultural crops.
This is particularly true in the tropics, where the trees are utilized as wood fuel and for soil
restoration 109. The added carbon inputs through leaf litter in agroforestry systems encourages
greater microbiological activity and diversity which synergizes the BNF process 90. In temperate
ecosystems, there are few options in terms of woody leguminous species and herbaceous species
capable of BNF are more likely to be utilized. Clovers, vetches, and other herbaceous nitrogen
fixers are grown and tilled into fields prior to planting of cash crops in order to improve soil
nitrogen levels and yields of future crops 110. In fact, cover crops have been used for soil
conservation since ancient times in societies like Greece, Rome, and China 2. However,
leguminous cover crops have yet to be widely integrated into agroforestry, temperate or tropical.
Until recently, little research has been done on the potential for herbaceous legumes to benefit
agroforestry systems such as alley cropping. In a study on an alley cropping system of pecan and
kura clover, results suggested that the intercropped legume was able to supplement the nitrogen
needs of the pecan 22. A system such as this is capable of improving soil organic matter,
structure, and nutrient cycling while producing high quality forage for cattle-production.
Furthermore, a study in a tropical maize alley cropping system found that when the maize was
sequentially cropped with cowpea, fertilizer additions were no longer required to maintain yields
40

105

. Despite the lack of research it is fairly clear that herbaceous cover crops have great potential

to substantially agroforestry systems.

2.4

Modes of N-Transfer

Leguminous cover crops are extremely efficient at building soil organic matter which in turn has
multiple benefits on nutrient cycling that can result in positive effects on the associated crops
25,28,53

. However, a perhaps more prominent service legumes provide is the ability to transfer

biologically fixed nitrogen to other species incapable of BNF themselves. This nutrient
relationship can occur through more than one mechanism. Research has identified three
pathways for the transfer of nitrogen from legumes to non-nitrogen-fixing plants: (1) indirect
transfer via accumulation and decomposition of biomass into the soil for future plant uptake, (2)
direct transfer of mineralized nitrogen exudates from living root nodules, and (3) direct transfer
via common mycorrhizal networks in the soil 12. All three of these pathways represent important
opportunities to utilize interspecies relationships in agroforestry.

2.4.1 Indirect Transfer
The process of indirect transfer requires a longer time-scale than direct transfer. It is considered
indirect primarily because it requires the organic nitrogen from decaying plant material to
undergo transformation into plant available forms before it can be received by another plant. An
example of this is the common use of cover crops in modern agriculture. In this process, the
species is grown for a period and then is destroyed, usually via tilling into the soil, where the
organic plant tissue decays and its nutrients are transformed by soil microbes into plant available
forms for the subsequently planted cash crop to use. By using legumes capable of BNF for this

41

process, landowners can access a virtually unlimited pool of atmospheric nitrogen to be used for
crop production. In agroforestry, indirect transfer often takes form in leaf litter from leguminous
trees accruing on the soil where they then follow the same process.

The difficulties of this form of nitrogen transfer are largely due to timing and labor. For
instance, using herbaceous legumes in an alley cropping system requires frequent cutting in order
to continually return the nutrient rich biomass (known also as “green manure”) to the soil. In
tropical alley cropping systems with woody legumes, intensive tree trimming is often practiced
for the same purpose 32. Since the biomass must then undergo multiple chemical transformations
before it becomes available to plants, inputs require timing in order allow plants to efficiently
utilize the added nutrients. Timing, however, is difficult to achieve, as multiple and often
immeasurable factors stand in the way of determining the appropriate time to cut. Nevertheless,
indirect transfer today remains the major pathway through which nitrogen is passed from
nitrogen fixing species to non-nitrogen fixing species. This is largely due to its extremely
efficient ability to build soil organic matter which will help to supply nutrients to associated cash
crops regardless of timing. As a result, this form of transfer may appear small in the short-term
while organic matter is accruing but ultimately result in considerable ongoing transfer over the
period of several years 13.

A considerable amount of research has been done on this form of nitrogen transfer. For instance,
in a tropical alley cropping system of G. sepium and maize it was demonstrated that leaf litter
from nitrogen fixing trees contributed as much as 10% of the maize final N content while
increasing overall soil health through indirect transfer 12. Between 3-102 kg N ha-1 yr-1 can

42

potentially be transferred through the death and decay of leguminous root cells alone 111.
Examples of systems where cover crops were used in alley cropping to benefit tree crops are few
and far between, but the results have shown potential. In a study that integrated five leguminous
covers into a walnut plantation, it was found that cool-season legumes such as hairy vetch were
capable of significantly improving soil nitrate levels and stimulating early walnut growth 19. In a
similar experiment in France, walnut foliar nitrogen content was enhanced when alley cropped
with alfalfa and soinfoin 21. In a study on Abies frasieri plantations in Michigan, it was found
that trees intercropped with alfalfa combined with low fertilization were equally as productive as
conventional systems at the recommended fertilizer rate 112.

The problem with many of these studies, however, is that indirect transfer is rarely isolated from
direct transfer. While it is generally accepted that indirect mechanisms are largely responsible
for transfer in these systems, it is likely that it is not sole source of nitrogen transfer. In a study
on P. avium and alley cropped with various leguminous shrubs, direct transfer from root-to-root
exchange were separated from transfer via leaf litter. The results revealed that the trees were
capable of utilizing between 7.5% and 20% of leaf litter N, depending on the shrub species 14.
More importantly, however, the study indicated that direct nitrogen transfer also represented a
substantial N source for the P. avium tree, which varied depending on the species paring 14. As
research goes forward, it will be important to make the distinction between modes of transfer in
order to fully understand the benefits of species pairings within agroforestry systems. By
gaining this better understanding of N transfer and how management may effect it, landowners
may be able to apply practices that more efficiently utilize this mechanism, which could reduce
labor and inputs.

43

2.4.2 Direct Transfer
While indirect N transfer requires the death and decay of organic plant material followed by the
transformation of nutrients into plant available forms, direct transfer offers pathways for plantavailable nitrogenous molecules to pass from the living root system of a nitrogen-fixing plant to
an adjacent root system of another plant that is not capable of the fixation proccess 12. Direct N
transfer occurs through two belowground pathways: root exudation and shared mycorrhizal
networks.

Root exudation refers to the secretion of compounds from the root systems of living plants.
These compounds vary in chemical structure and have multiple impacts on the immediate
rhizosphere. They can also be assimilated by other plant roots in the immediate area. The
mechanisms behind root exudation remain somewhat poorly understood but existing research has
attributed it to three passive pathways: diffusion, ion channels, and vesicle transport (Figure
2)113.

Figure 2. Description of mechanisms behind root exudation of compounds
through a plant cell membrane113. ‘PM’ refers to plasma membrane, ‘TMD’
refers to transmembrane domain, and ‘NBD’ refers to nucleotide-binding
domain.
44

Most plant species deposit a range of molecules into the rhizosphere through this process,
including organic acids, sugars, secondary metabolites, and allelopathic compounds 114.
However, recent research has focused specifically on the potential for nitrogen-rich exudates
from leguminous roots to benefit other species. These nitrogenous compounds are typically
soluble, low-molecular weight, and primarily occur through diffusion channels 115. Biologically
fixed N is the main source of exuded nitrogen. It’s estimated that about 3% of N fixed during
this process is secreted from the roots through exudation 116. Nitrogenous exudates from
temperate legumes primarily take form as ammonium and amino acid molecules, both of which
are capable of uptake by plants 13,117.

Direct N transfer from root exudates occurs belowground, is relatively constant, and eliminates
the need for any additional labor or timing. As a result, exudates from legumes present a
potentially significant source of nitrogen to target species that would be slow, but constant
throughout the growing season. Multiple studies have shown that exudate-based transfer of
nitrogen can provide adjacent crops with a substantial portion of their nitrogen requirements 14–
16,118

. White and red clovers (T. repens and T. pratense), for instance, were found to transfer as

much as 7.5 and 3.6 g-1m-2 N in one growing season, respectively, to adjacent forage grasses
through belowground transfer 17. Existing research on exudate transfer will be discussed in more
detail in Chapter 4.

The second mechanism through which direct transfer can occur is through shared mycorrhizal
networks in the soil. Roughly 70-90% of all plant species are known to form symbiotic

45

relationships with mycorrhizal fungi 119. These fungal species generally show low host
specificity, resulting in belowground web of shared fungal connections through which a variety
of nutrients can pass, including nitrogen 12. Mycorrhizal transfer is bidirectional, meaning that
nutrients can be passed in both directions between connected species 120. However, fungal
mediated transfer of this kind is driven by source-sink relationships and N transfer primarily
occurs from nitrogen-dense species (such as legumes) to less nitrogen-dense species 15. Less
than 10% of N transfer is from non-legumes to legumes, while as much as 80% of transfer occurs
in the reverse direction 121.

Multiple studies have confirmed the transfer of fixed N through shared fungal networks.
Recorded instances of this include berseem clover (Trifolium alexandrinum L.) to maize (Zea
mays L.) where around 4% of maize-N was derived from this form of transfer 122 and pea (Pisum
sativum L.) to barley (Hordeum vulgare L.) where 15% of donor N was transferred to the barley
123

. Fungal relationships have also been recorded between white clover to perennial ryegrass,

soybean to maize, and mung bean to rice 13,124. Jalonen el al. (2009) distinguished mycelial
transfer from exudate transfer in an experiment on G. sepium trees and D. aristatum grass. This
study, similar to other mycorrhizal transfer studies, found that while N transfer through this
mechanism is noteworthy, it is often insignificant in comparison to the other forms of N transfer
12,13,15

.

46

3

Contributions of Trifolium pratense to soil health and hybrid
poplar growth in temperate alley cropping systems

3.1

Abstract

Due to limitations in land availability and a desire to protect remaining natural forests, much of
the future global wood demand will be met by intensively managed plantation forests. These
plantations will also become particularly important to the biofuel industry in the growing years
via short-rotation woody crops (SRWC’s), or fast-growing hybrid trees intended solely for
biomass production. However, while these plantations are effective, they can pose problems to
sustainable land management, primarily regarding nutrient depletion. Agroforestry practices
such as alley cropping provide a feasible solution to issues regarding the sustainability of
SRWC’s. In particular, there is potential for leguminous cover crops to replace the need for
fertilizers and enhance multiple soil health parameters in tree plantations while sustaining or
improving productivity. Here, I examined the potential for red clover (Trifolium. pratense) to
benefit the growth and nutrition of a hybrid poplar (Populus nigra L.×P. maximowiczii A. Henry
(‘NM6’)) as well as soil carbon and nitrogen in a temperate alley cropping system. This was
done using a random-block design with four vegetation treatments: (1) T. pretense at a full
seeding rate, (2) T. pretense at half seeding rate, (3) natural vegetation, and (4) and bare-soil
(maintained via pesticide). Additionally, two spacing treatments were applied to poplar tree
rows: a narrow spacing (1.82 x 1.82m) and a wide spacing (1.82 x 3.05m). The results showed
mixed signs of competition and benefits, with most of the positive effects being seen in the
second year. In the wide spacing, the presence of red clover resulted in higher soil nitrate levels
at both full and half seeding rates by the second growing season of cover cropping. This resulted
47

in higher leaf nitrogen concentrations by the end of the second growing season at both seed rates
in the wide spacing and at the half seed rate in the narrow spacing. However, the positive effects
on poplar and soil nutrition were not reflected in tree growth at either spacing after just two
growing seasons with the cover crop application.

3.2

Introduction

Globally, industrial forest plantation area is estimated to be as much as 264 million hectares
36,125

. While the actual area is disputed based on changing definitions of forest plantations and

methods of estimation, the critical role that they will play in future global fiber production is
certain 126–128. Today the majority of plantation area is focused on creating roundwood for
lumber or pulpwood 36. However, the establishment of short-rotation woody crops (SRWC’s)
such as hybrid poplars and willows for the purpose of producing biofuels has gained
considerable attention in recent years.

Despite the controversy regarding their benefits to climate change mitigation, biofuels do offer
an opportunity to replace fuel requiring a long carbon cycle (coal, natural gas, oil, etc.) with
energy that could be carbon neutral in a much shorter term. As, Ceotto and Candillo (2011) put
it, “. . . society cannot accept the undesirable effects of biofuels when they are done wrong, but
also cannot afford to miss the benefits of biofuels when they are done right” 129. Since these
forest plantations are grown in short cycles (4-10 years) and are extremely low-cost, they are
ideal for marginal lands. In fact, it is estimated that over 5.5 billion gallons of renewable fuels
could be produced on marginal lands in the U.S. Midwest alone 130. Additionally, if agroforestry
practices were applied to these plantations, they could reduce the need for fertilizer inputs while

48

providing multiple other ecosystem services to landowners, including carbon sequestration, soil
restoration, and wildlife habitat, without reducing yields 7,131,132. Practices such as alley cropping
can also provide opportunities to produce crops or livestock between rows of SRWC’s, which
could increase and diversify landowner income using land that was formerly unproductive.
However, doing so without an increase in synthetic nitrogenous fertilizer use is critical, as these
emit high levels of GHG’s during both production and application 49,129.

Cover crops have significant potential for adding nitrogen to the soil as well as improving overall
nutrient cycling 110. Multiple studies have shown the ability of leguminous cover crops to
transfer biologically fixed nitrogen to non-leguminous trees through indirect or direct means 19–
22

. Dupraz et al. (1999), for instance, found that nitrogen content of leaves in black walnut were

improved by the presence of adjacently planted Medicago sativa L. and Onobrychis sativa L.
(both nitrogen-fixers). In a similar study on a black walnut and Vicia villosa Roth. alley crop,
resulted in significantly increased soil nitrate and walnut foliar N concentrations 19.
Furthermore, Waring & Snowdon (1985) also found that nitrogen fixed by clover improved
growth and nutrition of Pinus radiata seedlings after 7 years. In addition to nitrogen transfer, the
ability of cover crops to improve soil organic matter means potentially long-term benefits for
nutrient cycling. Kremer and Kussman (2011), for instance, found that alley cropping kura
clover between rows of pecan increased both soil organic C and enzyme activity after an 8-year
rotation. Some studies, however, have shown signs of significant competition between cover
crops and trees that ultimately reduced yields 133.

49

The mixed results from past studies makes it important to establish further research to examine
the potential of other species combinations. Furthermore, there is little research regarding the
utilization of biological nitrogen fixation to benefit SRWC species. To fill this gap, I
investigated the benefits of alley cropping leguminous cover crops into SRWC plantations. The
species selected for this research was Populus nigra L.×P. maximowiczii A. Henry (‘NM6’), a
hybrid poplar with great potential in biofuel production and documented success in loamy
soils131,134,135 and T. pratense, a common leguminous cover crop. Specifically, I examined the
contributions of T. pratense to soil organic carbon, organic nitrogen, and plant-available
nitrogen, and the subsequent effects on poplar growth and foliar nitrogen. Due to the high
phosphorous requirements for biological nitrogen fixation in legumes 90, foliar phosphorous
levels were also monitored to determine if there is competition between trees and cover crops.
The success of this study will inform the potential for similar alley cropping systems to improve
sustainable production of woody biofuels and other tree products.

3.3

Methods

The field experiment was conducted on a 65 x 78 m plot from 2014 – 2016 for three growing
seasons and was located at Sandhill Farms of the Tree Research Center (long. 42°65’N, lat.
84°42’ E) on Michigan State University’s campus in East Lansing, MI. Average summer and
winter temperatures are 15.5 and -6.6°C, respectively. Annual average precipitation in the area
is 853 mm with rainfall distributed fairly evenly throughout the year. The soil at Sandhill is a
Mariette fine sandy loam, which is classified as moderately well drained with high available
water capacity and medium surface runoff. The species chosen for this experiment were a hybrid

50

poplar (Populus nigra L.×P. maximowiczii A. Henry (‘NM6’)) and red clover (Trifolium
pretense), a common legume and temperate cover crop with high nitrogen-fixing potential 17,136.

3.3.1 Experimental Design
To accomplish the study goals I used a 4 x 2 factorial with a random block design to mitigate any
potential biases in the field. The field was broken into six blocks stretching north-south, each
consisting of four rows of Populus nigra L.×P. maximowiczii A. Henry (‘NM6’). Blocks
alternated between two spacing treatments (1.82 x 1.82 and 1.82 x 3.05 meters) moving from
east to west. Each of the blocks were then divided into four plots for vegetation treatments with
sizes of 9.14 x 5.49 m and 9.14 x 9.14m for the narrow and wide tree spacing, respectively, each
with 24 trees per plot. A buffer space of approximately 5.5 meters was allotted between blocks
and plots in which no treatment was applied to avoid effects between treatments. Four
vegetation regimes were assigned to plots in both spacing treatments: (1) red clover at a full
seeding rate (Treatment RCF), (2) red clover at half seeding rate (Treatment RCH), (3) natural
vegetation (Treatment NV), and (4) a bare soil control (Treatment C).

10'

6 x6
Plot A

BUFFER

10'

6 x 10
Plot B

10'

10'

6 x6
Plot C

RCH

6 x 10
Plot D

C

10'

6 x6
Plot E

RCF

NV

6 x 10
Plot F
*All rows planted with NM-6 hybrid
poplar, besides final row (planted with
DN34)

Figure 3. Plot layout of field study at the Tree Research Center in East Lansing. Individual squares
correspond to a single tree. Relative width of "blocks" corresponds to narrow and wide spacing
treatments and colors indicate vegetation treatments.
51

3.3.2 Field Methods
In May of 2014, Populus nigra L.× P. maximowiczii A. Henry (‘NM6’) cuttings were planted in
rows of 30 at the aforementioned spacing treatments. During the first growing season, no cover
crop treatments were applied. The field was maintained via regular mowing and periodic
herbicide treatments of glyphosate (250ml/L) within two feet of either side of tree rows to allow
trees to establish. The experimental field was surrounded by a three-wire electric fence to
prevent damage from deer browse.

At the end of the first season soil and leaf samples were taken, as well as the height and diameter
of each tree in order to establish baseline data. Soil samples were taken using a stainless steel
tubular soil sampler which was placed into the ground roughly at approximately 15cm in depth.
The soil was collected, bagged, and stored at -18°C until further analysis. Samples were taken at
five random points in each of the six blocks and separated by spacing treatment. Leaves were
sampled by taking one healthy leaf from each tree, which were then bagged, weighed, ovendried, and ground for further analysis. Height of the leading shoot of each tree was taken using a
field ruler and diameter of each shoot was taken at approximately 5cm above its connection to
the cutting with an electronic caliper. Since hybrid poplars are primarily utilized for the creation
of biofuels, diameters were applied to the following equation:

𝐵𝑖𝑜𝑚𝑎𝑠𝑠 (

𝑙𝑏𝑠
) = 𝐵𝑎𝑠𝑎𝑙 𝐴𝑟𝑒𝑎 (𝑠𝑞𝑢𝑎𝑟𝑒 𝑓𝑒𝑒𝑡) 𝑥 562.089
𝑡𝑟𝑒𝑒

Equation 1. Allometric equation for determining biomass of
hybrid poplar137.

52

to provide biomass values, which is the most appropriate value for poplar productivity. This
particular equation was chosen because it was derived using the same poplar hybrid (NM6)
under similar climatic conditions in Escanaba, Michigan 137.

At the beginning of the 2015 season a low mortality rate was observed among the cuttings (less
than 1%) and the vegetation treatments were put in place. Prior to seeding, a motorized rotor
tiller was used to scarify the earth (5-7.5 cm depth) and break up the sod between rows, allowing
for two feet of untilled soil on each side of the trees that would be treated with glyphosate
throughout the season to control weeds. This was done in all plots, including the natural
vegetation and bare earth treatments, in order to kill existing vegetation and maintain similar
starting points within all treatments. After tilling, T. pratense seed was applied to the appropriate
plots (Figure 3) via a hand-operated seed distributor in April of 2015. Red clover was seeded at
13.5 kg/ha for the full rate and 6.73 kg/ha for the half rate. Seed for treatments was sourced from
LaCrosse Seed of Lansing, Michigan. The natural vegetation treatment was thereafter allowed to
grow uninhibited (besides routine mowing), while the bare-earth treatment was treated every six
weeks with glyphosate (7.8 mL/L) to remove vegetation.

Starting in June of 2015, after six weeks to allow the cover crop treatments to effectively
establish, the alley cropped areas were routinely mowed every 21 days in order to regularly
return the biomass to the soil. Prior to each mowing, samples of the biomass from T. pratense
(RCF and RCH) and natural vegetation (NV) treatments were collected to be analyzed for N and
P concentrations as well as total weight to determine total nutrient additions throughout the
season. Samples of cover crop biomass were taken using a push-mower with a bushel bagger

53

attached. A square meter was mowed at three random points in each plot. Each square meter of
biomass was then bagged, dried, weighed, and ground for analysis. The field was mowed and
sampled every three weeks during the growing season on four sample dates (6/11/15, 7/2/15,
7/23/15, and 8/13/15).

Soil samples were taken as in the previous year with a tubular soil sampler by hand at midSummer (July) and early fall (September) in 2015 and 2016. The soil sample pattern was similar
to cover crop sampling with three samples taken at random points in each plot. Soil was bagged
and stored at -18°C until further analysis. At the end of the season, leaf samples were taken by
plucking one healthy leaf from each tree in all plots. Leaves were bagged, dried, weighed, and
ground for nutrient analysis. All tree heights and diameters (excluding trees located inside
buffers) were observed and recorded using a field ruler and electronic caliper as in the previous
season. Cover crops were not tilled and allowed to go to seed at the end of the growing season
of 2015. As a result, similar cover crop densities were observed in the 2016 growing season.
The same management methods were applied, with regular cutting (every 21 days) and herbicide
applications to control treatments and tree rows. Additionally, the same processes for sampling
soil, poplar leaf, height, and diameter was repeated in 2016. However, the cover crop biomass
was not sampled in the second season, and return was assumed to be proportionate, if not the
same as the previous year.

3.3.3 Sample Analysis
Soil samples were mixed thoroughly and a subsample of 5 grams was taken and mixed with 50
mL of potassium chloride (KCl). Samples were then placed in a shaker and mixed for 1 ½ hours

54

for nutrient extraction. After shaking, samples were filtered using Whatman No. 1 filters prewashed in potassium chloride solution to remove soil sediment. The resulting extracted nutrient
solution was analyzed for nitrate and ammonium levels using a SAN++ segmented flow analyzer
(Skalar Inc., Atlanta, GA) measured at 540 and 660 nm, respectively. An additional 10 grams of
soil were placed in an oven at 65°C for 72 hours and thereafter weighed for determining soil
moisture content.

All leaf and cover crop biomass samples dried for 72 hours at 65°C, weighed, and ground to a
fine powder prior to analysis for total N and P. A subsample of 0.3 grams (Âą 0.005) was then
weighed and placed in 100 mL test tube, mixed with 4.5 mL of sulfuric acid (H2SO4) in a vortex
and allowed to sit overnight for pre-digestion. Then, 1.5 mL of hydrogen peroxide (30%) was
added to each tube and then placed in a block digester (AIM600 Block Digesting System) at 340
¹ 10°C for one hour under a programed temperature schedule. After digestion, sample tubes
were filled to 100 mL with deionized water and mixed with a vortex. The digested sample
mixture was then buffered and chlorinated after dialysis to form a chemical complex measured
660 and 880 nm for total N and P, respectively, on a SAN++ segmented flow analyzer (Skalar
Inc.).

3.3.4 Data Analysis
Results from nutrient analysis of poplar leaves were expressed as concentrations (mg/g). To
obtain nutrient return from cover crops, concentrations of N and P from analysis were applied to
the dry biomass weight to find total nutrient content. Results from soil nitrate and ammonium
analysis were returned from analysis in mg/L and were transformed to concentrations (Âľg/mg).

55

The above values, along with values regarding green manure biomass and nutrient return, poplar
nutrient concentrations, height, and diameter were compared for significant differences between
treatments using an analysis of variance or a Student’s T-Test depending on the number of
treatments being compared. Means were then separated using a Tukey’s test of honestly
significant difference. Linear correlations were tested using Pearson’s correlation analysis. All
tests used values of P lower than 0.05 as an indicator of statistically significant difference. All
data analyses were performed in R 3.3.1 software, as was the creation of graphs and tables 138.

3.4

Results

3.4.1 Cover Crop Nutrient Concentrations
Nutrient analysis of the sampled green manure revealed different concentrations of nitrogen and
phosphorous between treatments involving clover (RCF and RCH) and the natural vegetation
treatment (NV) (Table 2). Treatment RCH had significantly higher N concentrations than NV at
both narrow and wide spacings (p = 0.02 and 0.00), while treatment RCF had significantly
higher N concentrations than NV only at the wide spacing (p = 0.0), however the narrow spacing
exhibited strong differences (p = 0.06). No differences in N concentrations were observed
between different spacing treatments in red clover treatments. However, in treatment NV the
biomass in the narrow spacing had higher N concentrations (p = 0.02).

Phosphorous concentrations showed a reverse trend, with natural vegetation plots being
generally higher. Treatment NV had higher P concentrations than treatment RCF (wide and

56

narrow spacing) and treatment RCH (narrow spacing) (p = 0.01, 0.00, 0.00). No differences
between spacing treatments were observed under similar cropping treatments.

Nitrogen/phosphorous ratios were also analyzed to determine potential effects on soil nutrients
and health (Figure 5). Ratios were significantly higher in treatment RCF at both wide and
narrow spacings (p = 0.00 and 0.00). The same was observed in treatment RCH for wide and
narrow spacings (p = 0.00 and 0.00).

Figure 4. Average nutrient concentrations of biomass returned during
routine cuttings from natural vegetation (NV), red clover at a half seed rate
(RCH), and red clover at a full seed rate (RCF).
Table 2. Average nutrient concentrations returned during cutting from
natural vegetation (NV), red clover at a half seed rate (RCH), and red
clover at a full seed rate (RCF).

57

Figure 5. Average nitrogen-to-phosphorous ratios based on nutrient concentrations
of cover crop biomass return from natural vegetation (NV), red clover at a half
seed rate (RCH), and red clover at a full seed rate (RCF).

3.4.2 Cover Crop Nutrient Return
Treatments involving clover (RCH and RCF) had much higher biomass return compared to in the
natural vegetation treatment (Figure 6). At both wide and narrow spacing types, treatments RCH
(p = 0.00 and 0.00) and RCF (p = 0.00 and 0.00) had higher dry weight returns than treatment
NV. Additionally, treatment RCF at the wide spacing had higher biomass production than the
narrow spacing (p = 0.01).

Total nutrient return from green manure over the 2015 season showed similar trends to green
manure nutrient concentrations (Table 3). Treatments RCH and RCF had significantly higher N
content return values at wide (p = 0.00 and 0.00) and narrow spacing treatments (p = 0.00 and
0.00) compared to treatment NV. Treatment RCF also had significantly higher N content at the
wide spacing than at the narrow spacing (p = 0.00). Analysis for phosphorous revealed that
despite lower P concentrations than treatment NV, treatments RCH and RCF showed higher

58

values of total P content at both wide (p = 0.00 and 0.00) and narrow spacings (p = 0.00 and
0.00). No differences were observed between spacing types within vegetation treatments.

Figure 6. Total dry weight of biomass returned from natural
vegetation (NV), red clover at a half seed rate (RCH), and red clover
at a full seed rate (RCF) during the 2015 season.

Figure 7. Total nutrient return in biomass from natural vegetation
(NV), red clover at a half seed rate (RCH), and red clover at a full
seed rate (RCF) during the 2015 season.

59

Table 3. Total nutrient return via green manure by the
end of the 2015 season from natural vegetation (NV),
red clover at a half seed rate (RCH), and red clover at a
full seed rate (RCF).

3.4.3 Plant-Available Soil Nitrogen
There was a notable lack of differences in soil nutrient parameters between spacing treatments
(Tables 4 & 5). The only differences existed in the 2016 mid-season nitrate levels in treatment
RCF where the nitrate in wider spacing was significantly higher than in the narrow spacing (p =
0.00) and RCH where the narrow spacing was higher than the wider spacing (p = 0.00). No
other differences existed between spacing treatments at the other four sampling dates over the
three-year experiment.

Soil nitrate increased in most treatments (including controls) up to the late-season sample in
2016, while soil ammonium showed a decreasing trend after the end of the first growing season
onward. Mid-season estimates of soil nitrate in 2016 (07/2016) showed that treatment RCF had
significantly higher levels than treatments C and NV (p = 0.00 and 0.00) at the wide spacing
while RCH had significantly higher nitrate than the same treatments (p = 0.00 and 0.00) at the
narrow spacing. However, no significant differences were seen in soil nitrate by the end of the
final season. Furthermore, no differences were found between treatments in 2016 at mid or late60

season estimates regarding ammonium. This was due to an apparent lack of soil ammonium in
all treatments.

Due to the lack of differences between spacing treatments, analysis was also done removing
spacing treatments which revealed somewhat similar results (Figure 8). Mid-season 2016 nitrate
levels were significantly higher in both treatments RCH and RCF compared to the control (p =
0.00 and 0.00) and treatment NV (p = 0.00 and 0.00). No significant differences were found
among vegetation treatments by the end of the experiment. Ammonium levels showed no
differences after removing spacing.
Table 4. Summary of soil ammonium in treatments prior to cover crop establishment
(09/2014) and through two subsequent growing seasons for natural vegetation (NV), red
clover at a half seed rate (RCH), red clover at a full seed rate (RCF), and control
treatments.

61

Table 5. Summary of soil nitrate in treatments prior to cover crop establishment
(09/2014) and through two subsequent growing seasons for natural vegetation (NV), red
clover at a half seed rate (RCH), red clover at a full seed rate (RCF), and control
treatments.

Figure 8. Soil mineralized nitrogen concentrations after combining spacing
treatments across natural vegetation (NV), red clover at a half seed rate (RCH), red
clover at a full seed rate (RCF), and control treatments.

3.4.4 Poplar Leaf Nutrients
Samples of poplar leaves were also analyzed for N and P concentrations at the end of all three
seasons (Table 6). No significant differences were found between wide and narrow spacing
treatments in 2014, which was prior to the establishment of cover crop treatments. Additionally,
no significant differences were found between spacing or vegetation treatments by the end of the

62

2015 growing season. There were also no significant differences between spacing treatments
within the same vegetation treatment in either phosphorous or nitrogen. The small sample size in
this experiment likely played a part in the small amount of significant differences.

In 2016, poplar leaves within clover treatments exhibited as much as 25% higher levels of
nitrogen in comparison to bare-soil controls in the wide spacing, and similar results compared to
the natural vegetation control at the narrow spacing. However, these differences were only
significant at the full seed rate (RCF) at the wide spacing (p = 0.04) and at the half seed rate
(RCH) at the narrow spacing (p = 0.05). Concerning phosphorous, vegetation treatments with
clover generally had lower numbers. Red clover at a full seed rate (RCF) was significantly lower
compared to treatments NV and C (p= 0.05 and 0.03) at the wide spacing, while at the narrow
spacing red clover at the half seed rate (RCH) was lower only compared to treatment NV (p =
0.00).
Table 6. Average poplar leaf nitrogen and phosphorous concentrations between spacing
treatments over the course of the experiment for natural vegetation (NV), red clover at a half
seed rate (RCH), red clover at a full seed rate (RCF), and control treatments.

63

3.4.5 Poplar Height and Diameter
Measurements of height and diameter were taken at the end of all three growing seasons. Data
was recorded and averaged by treatment for analysis. Measurements from 2014, prior to cover
crop planting, revealed a slight difference between wide and narrow spacings in both height and
diameter. Poplar heights were roughly 0.1 meter higher and diameters roughly 1 millimeter
larger in the wide spacing compared to the narrow spacing (p = 0.00 and 0.00).

At the end of 2015, heights in treatment RCH were lower than in controls at both wide and
narrow spacing (p = 0.01 and 0.03). Treatments RCF and NV also had lower heights compared
to controls at the wide spacing (p = 0.00 and 0.00). The 2016 growing season showed slightly
different results. At the wide spacing, treatments RCF and NV had lower heights than control
treatments (p = 0.00 and 0.00), as did treatment RCH at the narrow spacing (p = 0.02).
In terms of differences between spacing treatments, within treatment NV, the narrow spacing had
lower average heights than the narrow spacing (p = 0.00) in 2015. In 2016, treatment RCH had
higher heights at the wide spacing than in the narrow spacing (p = 0.00).

Similar trends were seen in the 2015 season in terms of diameter. Treatment RCH had lower
diameters compared to control in both wide and narrow spacing treatments (p = 0.00 and 0.03).
Treatments RCF and NV also had lower diameter than in controls, but only in the narrow
spacing (p = 0.00 and 0.00). More differences between spacing treatments within vegetation
treatments existed in diameters than in heights. Control treatments and treatment RCH both had
larger diameters at the wider spacing (p = 0.00 and 0.00), while treatment NV was higher in the
narrow spacing (p = 0.00). By 2016, the only difference in diameter within the narrow spacing

64

treatment was in treatment RCH, which was lower than controls (p = 0.01). At wide the spacing
RCF and NV were both lower than controls (p = 0.00 and 0.00). Since the biomass
measurements were based on the diameters retrieved from field measures, the same differences
were reflected (Table 8). The only difference occurred when determining biomass in Mg ha-1,
where narrow spacing treatments had higher productivity compared to the wider spacing
treatments, due to a higher number of trees per acre. Interestingly, differences between spacing
treatments were only significant in RCF and NV (p = 0.00 and 0.00) by the end of the
experiment in 2016.
Table 7. Average height and diameters of poplar at the end of each growing season for
natural vegetation (NV), red clover at a half seed rate (RCH), red clover at a full seed rate
(RCF), and control treatments.

Table 8. Average biomass productivity of poplars at tree and stand levels for natural
vegetation (NV), red clover at a half seed rate (RCH), red clover at a full seed rate
(RCF), and control treatments.

65

3.4.6 Inter-Variable Correlations
Extensive analysis was completed to find possible linear correlations between variables that
might give insight into how the cover crop treatments might have affect soil health and poplar
growth and nutrition. However, few valuable inter-variable relationships were revealed. No
correlations were found between green manure N return and poplar height or diameter in either
2015 or 2016. Green manure N return was positively correlated to poplar leaf N in the narrow
spacing of treatment NV (p = 0.01, Adj r2 = 0.62). Dry biomass weight return in the narrow
spacing of treatment RCF was positively linked to soil organic C (p = 0.01, Adj r2 = 0.54).

3.5

Discussion

The objectives of this study were to determine the effects, be they positive or negative, of alley
cropping leguminous cover crops into a hybrid poplar biofuel plantation, and how management
techniques may influence those effects. After three years of data (including a baseline year
without alley cropping), the results gave mixed signs of competition and benefits between T.
pratense and the hybrid poplar, but ultimately the study objectives were achieved. Soil and leaf
nitrogen responses to the alley cropping treatments were clear, although they were not reflected
in growth measurements by the end of the experiment. Additionally, evidence suggests that
management methods can have impacts on mitigating negative competitive effects in such
systems. These details will be discussed in the following sections.

66

3.5.1 Biomass Return
Cover crop analysis revealed significantly higher amounts of biomass, organic N, and organic P
being returned to the soil in T. pratense treatments regardless of seeding rate. This was to be
expected, as red clover is very productive in terms of biomass and is an extremely productive
nitrogen-fixer 139,140. Field observations revealed little visual difference between treatments red
clover seeded at the full rate versus at the half rate (RCF vs RCH), which was reflected in both
biomass and nutrient return. Additionally, the natural vegetation treatment consisted primarily of
forbs and grasses, while treatment C remained free of vegetation.

Higher P concentrations in the alley crop biomass from treatment NV compared to other
treatments could have been due to the lack of competition for P resources within these plots.
This is supported by the slightly smaller margins of difference in P concentrations between
vegetation treatments at the wider spacing (Table 2), where competition would be less
significant. Additionally, lower P concentrations in treatments RCH and RCF reinforce this idea.
Leguminous species require large amounts of phosphorous to sustain biological nitrogen fixation
90

.

Thus, it follows that clover P concentrations might be lower in the absence of sufficient

available phosphorous, particularly between narrow tree rows. However, total phosphorous
return by the end of the 2015 season was higher in treatments RCH and RCF compared to natural
vegetation at both spacing treatments. This likely reflects a greater overall efficiency at utilizing
soil P by red clover plots.

As was expected, nitrogen returns were far greater in red clover treatments compared to
treatment NV. Values of N return did not differ significantly between spacing treatments but a

67

trend of lower N return was seen in the narrow row spacing, which indicates higher competition
for soil N between poplar and clover roots.

3.5.2 Soil
Soil nitrate and ammonium levels were fairly consistent between treatments RCF and RCH,
which falls in line with the trends seen in cover crop nutrient return (Figure 8). Both red clover
treatments had generally higher plant available nitrogen levels compared to other treatments,
particularly in terms of nitrate, although significant differences didn’t occur until 2016. This is
indicative of the “delayed effect” that is associated with indirect transfer, due to the required
decay and chemical transformations that need to take place for the nitrogen to become available
12

. Up to July of 2016, there was a steady rise in soil nitrate in clover treatments as that nitrogen-

rich biomass began to decay and become available to the poplar trees. By the end of the 2016
growing season, however, nitrate levels declined sharply. Since these samples were taken after a
period of heavy rain, the reason for this decline is most likely nitrate leaching. Although there
was no significant difference, there was also generally higher ammonium levels seen in the red
clover treatments compared to controls. This may reflect some form of direct N transfer
occurring, most likely through root exudates. The primary compound released by temperate
herbaceous legumes through this pathway is ammonium, due to its low-molecular weight and
ability to “leak” out of the root membrane, particularly in the early months of its life-cycle 13,141.

Soil parameters in treatment NV reflect the lack of nutrient rich green manure returns, with lower
nitrate levels throughout the experiment, even compared to treatment C. Bare-soil control plots
were regularly treated with herbicide to remove plant growth between rows and the only organic

68

matter inputs to them was leaf litter at the end of the growing season. Thus the higher nitrate
availability in these plots compared to treatment NV would most likely be due to a lack of
competition between trees and forbs.

3.5.3 Foliar Nutrient Concentrations
Differences in soil parameters between treatments were reflected in poplar leaf nutrient
concentrations. Leaf N concentrations in treatments RCF and RCH were not statistically higher
than other treatments until 2016. This is also a strong indication of the “delayed effect” of
indirect nitrogen transfer pathway 12. Transferring N through biomass, which was likely the
primary form of transfer in this study, requires the decomposition of the biomass followed by
mineralization of organic N into plant available forms by microbes. Completing this process and
successfully transferring N indirectly may still take more than one season to complete. The
longer periods required by similar studies to see the benefits of legumes to tree species reflects
this finding 20–22.

Regardless of timescale, the inputs from the previous season (2015) obviously benefited poplar
nutrition, particularly at the wider spacing treatment. At the narrow spacing, only treatment
RCH had higher poplar leaf N concentrations by 2016, indicating higher competition for N
between poplar trees and clover at the full clover seeding rate (RCF). This could be a direct
result of greater root interaction between roots and competition for plant-available N. It could
also be caused by greater competition for P resources, resulting in lower biological nitrogen
fixation by red clover and thus limited indirect (or direct) N transfer. This is greatly reinforced
by the results of poplar leaf P analysis, which were consistently lower than treatments C and NV

69

by 2016. This competition for P resources could hinder growth in the long-term, although other
studies have shown that competition for soil resources tends to decline over time in these
systems 19–22. However, phosphate fertilizers could assist growth in the early years of a rotation.
Additionally, it is possible that some of the competition for phosphorous could be mitigated in
the short term through management methods. This is particularly evident in Table 6, where at
the wide spacing and at a lower seeding density (RCH), phosphorous levels were higher
compared to other red clover treatments. Moreover, these management methods did not result in
any limitations in N transfer. Without using isotopic tracing methods, it is difficult to discern the
relative importance of direct and indirect N transfer within this system. However, higher poplar
leaf N in treatments RCH and RCF immediately after the first season of high inputs from cover
crop biomass indicate that indirect transfer was likely the most important transfer mechanism.

3.5.4 Poplar Growth
Growth measurements revealed puzzling and inconsistent results. Heights and diameters of
poplars within treatments RCH and RCF were not higher than treatment C at any point, but were
higher than treatment NV at the wide spacing in both 2015 and 2016. Natural vegetation
treatments varied widely with spacing while bare-soil controls were relatively steady.
Regardless, there were no observable trends in terms of height or diameter and there were no
linear correlations between nutrient return or soil parameters and growth. Similar results were
seen in biomass, where bare-soil controls were consistently higher than red clover treatments, but
treatment NV ranged largely between wide and narrow spacing. Additionally, there was a lack
of consistent trends within the red clover treatments, with half seed rates being more productive
at the wide spacing compared to the narrow spacing. In summation, neither spacing nor

70

vegetation treatments showed consistent effects on productivity of poplar throughout the
experiment. However, it is certain that the benefits seen in soil and poplar leaves in terms of
nitrogen were not yet reflected in growth measures at the end of two seasons with the cover crop
applications.

This study did not look at soil moisture or competition for water resources in the field. Since
hybrid poplars have high water requirements for growth, competition for water resources
between the cover crop and trees was very possible. The NV plots also had regular vegetation in
the alleys between trees and competition for water should not differ significantly between these
treatments. In bare-soil plots (treatment C), however, it is likely that water competition was
reduced by the lack of vegetation. This likely explains why this treatment had consistently
higher productivity compared to treatments RCH and RCF at both spacing treatments. It is
unclear, however, why treatment NV had such variation, with the highest measurements of any
treatment at the narrow spacing and the lowest measurements at the wide spacing. Nevertheless,
the high water demand by hybrid poplars does make them more susceptible to competition in
alley cropping systems such as these. This can reduce poplar growth, biological nitrogen
fixation, and nutrient return from cover crop biomass. Similar to with phosphorous competition,
it is possible to mitigate some of these competitive effects with management methods. The RCH
treatment (half clover seed-rate) at the wide spacing seemed to do better than the full seed rate,
indicating a possible reduction in resource competition, including water.

The mixed signs of benefits and competition brought by the presence of cover crops, particularly
in red clover treatments, has been demonstrated in similar studies. Wilson et al (2011) found

71

similar signs of competition in a short-term study on cover crops and Abies fraseri trees.
However, other studies have found the benefits of cover crops eventually outweigh symptoms of
competition in the long term 19,20,22. Waring et al (1985) studied the effects of clover on P.
radiate in comparison to urea fertilization (and controls). By the end of the second year,
significant signs of competition were seen between clover and tree seedlings 20. However, by the
third year the competition had subsided. By the end of the seven-year study, seedlings in clover
treatments had significantly higher growth measurements than controls and were the same as
seedlings in the fertilized treatment, indicating that clover could serve as a substitute for heavy
fertilization 20. Since the benefits of red clover to soil and poplar nutrition were seen as early as
the second season in this study, it is reasonable to believe that similar benefits to growth would
be seen over a longer period. Long-term studies such as these are difficult and cumbersome but
ultimately may be needed to clarify the real potential of cover crops to tree plantations over time.

3.6

Conclusion

The purpose of this study was to determine the potential benefits of regularly returned green
manure from red clover (T. pratense) to hybrid poplar (Populus nigra L.×P. maximowiczii A.
Henry (‘NM6’)) growth and soil nutrient parameters at different tree row spacing treatments.
Cover crop nutrient returns of both phosphorous and nitrogen were higher in red clover
treatments at both the full and half seed rate compared to plots with natural vegetation. This
generated a positive response in soil nitrate within the same plots in the subsequent growing
season, indicating a delayed effect to complete indirect nitrogen transfer. The effects of the
cover crop biomass return and soil nutrient response were reflected in poplar leaf nutrition by the
second growing season. At narrow spacing treatments, competition for other soil resources,

72

including water and phosphorous, may have resulted in limited N transfer. As a result, poplar
leaf N concentrations were significantly higher only in plots where clover was seeded at the halfrate by the second year, where competition for soil resources was assumed to be weaker. At the
wide spacing, the presence of red clover resulted in significantly higher poplar leaf N by the
second growing season only at the full seed rate compared to controls. Similar N concentrations
were seen at the half-seed rate, but were not significant due to higher variability and limited
sampling power. However, competition for soil phosphorous was apparent, with generally lower
concentrations found in treatments involving red clover. Additionally, no benefits to poplar
growth were seen by the presence of red clover.

In summation, this study revealed mixed signs of competition and benefits of cover crops to
hybrid poplar trees over the two-year experimental period. Red clover (T. pretense) has potential
to improve soil nitrogen, nitrogen cycling, and poplar leaf N concentrations but could outcompete trees for phosphorous, particularly in the short term. However, evidence suggests that
these competitive effects could be at least partially mitigated through management methods such
as wider tree row spacing and cover crop seeding density. Additionally, this study falls in line
with similar alley cropping experiments where signs of competition were more prominent in
early years but were generally diminished with time as beneficial relationships became more
prominent and positive influences on tree growth were seen. Longer-term studies are needed to
see how these dynamics might change throughout the life-cycle of a plantation.

73

4

Quantifying belowground nitrogen transfer between Trifolium
repens and agroforestry tree species in temperate alley cropping
systems via root exudates

4.1

Abstract

Direct nitrogen transfer via root exudates represents a potentially significant source of plantavailable nitrogen to agroforestry systems. Nitrogen-fixing species, such as legumes, are capable
of capturing and assimilating atmospheric nitrogen. In the process, low molecular-weight
nitrogenous compounds such as ammonium and amino acids are secreted from roots into the
rhizosphere where they can be readily assimilated by the root systems of other species. Utilizing
this source of N transfer could be beneficial to the growth and nutrition of targeted species. To
date, existing research on this subject has focused primarily on transfer of N from nitrogen-fixing
trees to herbaceous crops or herbaceous legumes to associated grasses. To take existing research
further, this experiment was established to quantify how mineralized nitrogen is directly
transferred from herbaceous nitrogen-fixing legumes to recipient tree species in alley cropping
systems. Specifically, this experiment focused on the direct transfer of fixed nitrogen from an
herbaceous legume (Trifolium repens) to two tree species (Castanea sativa x C. crenata
(“Colossal”) and Populus nigra L.×P. maximowiczii A. Henry ‘NM6’) in a simulated alley
cropping system. To accomplish this I utilized the 15N nitrogen tracing method that has been
used in similar studies in a greenhouse setting. In this process, the donor species (nitrogen fixer)
is fertilized with K15N03 via foliar application and the recipient crop species is then analyzed for
nitrogen isotope ratios after a designated growth period. The values of percent nitrogen derived
from transfer in both poplar 12.1% (Âą1.1) and chestnut 33.3% (Âą7.4) revealed the potential
importance of N transfer to alley cropping systems. The majority of transfer was likely to have

74

taken place via the production of nitrogenous root exudates from white clover, of which 2.62
(Âą1.24) mg N was produced daily in each pot (or 0.04g N m-2 day -1). Previous research
indicates that the role of direct N transfer increases in field settings over time. Since this study
took place in a greenhouse, similar studies in the field over longer periods are needed to
understand how this potential is reflected in actual production systems.

4.2

Introduction

There is considerable existing research indicating the indirect benefits provided by legumes to
non-legumes in agroforestry systems. However, until recently it was rare to find studies that
sought to explore the specific mechanisms that make this possible. We know that the presence
of legumes in agricultural and natural systems can often mean the provision of atmospherically
fixed nitrogen to species that cannot achieve this process themselves, either indirectly or directly.
However, there is still much research needed to gain a better understanding of these transfer
mechanisms so that they can be utilized in sustainable agriculture and forestry.

Within the small pool of research that has explored direct nitrogen transfer, the large portion of it
has focused on the pathway of nitrogen from leguminous trees to herbaceous crops in tropical
climates or herbaceous legumes to grasses in temperate zones. For example, in a field
experiment on conventionally and organically grown grass-clover leys in Switzerland, an
average of 51% of total grass nitrogen was found to have been transferred from nearby clover
species, the majority of which was assumed to have been through root exudates 142. Furthermore,
in a long-term field study on a silvopastoral site in Central America, transfer from nitrogen-

75

fixing Gliricidia sepium trees to Dichantium aristatum grass swards via leaf litter (indirect
transfer) was shown to only account for one-third of total nitrogen transfer, implying that the
direct forms of transfer have higher relative importance 143. The same species were later used in
a greenhouse experiment that discerned between direct transfer pathways (root exudates and
common mycorrhizal networks). The results showed that root exudates alone were capable of
supplying as much as 63% of total grass nitrogen 15. Although grass only absorbed roughly 22%
of the available G. sepium exudates, it proved to be an important source of nitrogen15.

Despite a recent growth in research on direct N transfer, there is still little research focusing on
herbaceous legumes as N donors to trees. Dupraz et al. (1999) showed that the presence of
alfalfa planted between rows of black walnut improved walnut foliar nitrogen content but did not
distinguish the mode of transfer. In 2011, a study was published showing that kura clover
benefited soil and tree nitrogen in a pecan alley cropping system, but this study also did not
identify or distinguish how nitrogen was transferred 22. Roggy et al (2004) focused on trees as
the recipient species of nitrogen transfer but did not use herbaceous species as the donor. This
study found that E. augustifolia, was capable of transferring as much as 25% of its fixed nitrogen
to Prunus trees via root exchange, while other species contributed lower amounts 14 To
summarize, some studies exist that focus on trees as the beneficiaries of adjacently planted
herbaceous legumes, but it appears that few, if any, have actually quantified and distinguished
direct nitrogen transfer from indirect transfer in such a system.

Here I quantified direct nitrogen transfer from Dutch white clover (Trifolium repens) to a hybrid
chestnut species (Castanea sativa x C. crenata (“Colossal”)) and a hybrid poplar species

76

(Populus nigra L.×P. maximowiczii A. Henry ‘NM6’) in a greenhouse setting. Dutch white
clover was chosen as the donor species because of its high nitrogen fixing and transfer potential
13,17

. Chestnuts and poplars were chosen as recipient tree species to highlight two important tree

plantation industries: fruit/nut production and biomass energy production. The results of this
research have potential to benefit both industries and improve the sustainability of their
production.

While the distinction between direct and indirect forms of transfer does not mean from the
perspective of the recipient plant, the more efficient utilization of direct transfer mechanisms
could mean less inputs and labor for the landowner. Since indirect transfer requires the turnover
of plant biomass, continuous mowing and re-seeding is required to maintain legumes in cover
cropped tree plantations. However, if direct transfer is found to be an equal or perhaps even
more substantial source of N for the trees, these labor and inputs would be unnecessary.
Furthermore, a landowner could take this a step further by utilizing the above-ground biomass
for raising livestock while continuing substantial N transfer through belowground mechanisms.

The objectives of this study were to (1) determine the nitrogen transferred from T. repens to the
two chosen species, (2) examine the potentially different ability of the two species to absorb
available clover exudates in the soil, (3) determine the exudate potential of T. repens, and (4)
compare the photosynthetic, nutrient dynamics, and growth to trees grown in interaction
treatments to those grown in ideal conditions (controls). These values will be achieved by
growing these species together in a greenhouse setting and utilizing isotopic tracing methods

77

over a short (63 day) experimental timeline. These findings will inform decision making for tree
plantation owners and help lay a foundation for future research.

4.3

Materials and Methods

4.3.1 Experimental Design
This study consisted of two experiments:

1. Transfer Experiment (two treatments):
a. Interaction (IR): Trees and 15N labelled white clover were grown together in
large pots allowing for uninhibited root interaction to study direct nitrogen
transfer
b. Control (C): Clover, chestnut, and poplar grown in individual pots to compare
growth and nutrient dynamics with interaction treatment.
2. Root Exudate Experiment (EX): Strips of white clover (same size as in interaction
treatment) grown individually in pots, labelled with 15N, and then transplanted to
hydroponic tanks for one week to study total exudate potential throughout the course of
the transfer experiment

The interaction treatment was replicated four times for both tree species and were divided into
four sample dates. An initial sample was taken immediately prior to applying the isotopic tracer
(5/2/16) and was followed by three more samples after application in three week increments
(5/23/16, 6/13/16, and 7/4/16). Following the initial sample, samples in the exudate experiment
and Treatment (C) were harvested at the same time as Treatment (IR). Treatment (C) consisted

78

of three replicates of unlabeled white clover, chestnut, and poplar grown alone in pots.
Similarly, pots in the exudate experiment, which consisted of isotopically labelled white clover
grown alone in pots, was replicated three times at each sampling date. All sampling in this
experiment was destructive, meaning the whole plant was harvested and separated by
compartment for analysis. The experimental model and sampling pattern is depicted in Figure 9.

Figure 9. Greenhouse study experimental model and sampling design.

These experiments took place in a greenhouse running east-west, located at the Tree Research
Center (TRC) of Michigan State University’s campus in East Lansing, Michigan (42°40’N,
84°27’W) from April 2016 to early July 2016. The average temperature during the course of the
experiment was 21.1°C (71°F) with maximum temperatures ranging between 21-29°C (70-85°F)
and minimum temperatures ranging between 15-22°C (60-72°F). The greenhouse was well
79

ventilated and mimicked external environmental conditions for the majority of the experimental
period (with the exception of precipitation).

4.3.2 Materials
Soil used for the experiment was a coarse perennial potting mix from Al-Par Peat Company in
Elsie, Michigan and was composed of 33% pine bark fines, 33% reed sedge peat, 3%
vermiculite, and 1-3% sand. Containers used for (IR), (EX), and (C) had soil volumes of 21, 10,
and 14 liters, respectively, with the exception of units containing white clover in (C), which also
used the 10L containers from (EX).

Cuttings of Populus nigra L.×P. maximowiczii A. Henry (‘NM6’) were obtained from 2-year
coppice fields at the Tree Research Center on MSU’s campus. Cuttings were harvested one
month prior to the beginning of the experiment (March 2016) and cold-stored at 18°C prior to
planting. Chestnut saplings (Castanea sativa x C. crenata (“Colossal”)) were obtained from
Nash Nurseries in Owosso, Michigan. Saplings were one year old, propagated by seed in the
spring of 2015 and treated with three applications of 20-10-20 (NPK) solution in the first
growing season but were not fertilized in the nursery during the months of 2016 prior to
purchase. The chestnut saplings were purchased and planted prior to leafing out for the season.
White clover (Trifolium repens) seed was obtained from LaCrosse Seed in Lansing, Michigan
and was not inoculated at purchase. Inoculant to be used prior to seeding was also purchased
from LaCrosse Seed and was the recommended strain for true clovers (Rhizobium
leguminosarum biovar trifolii). The isotopic tracer used for this study was purchased from
Sigma-Aldrich (St. Louis, MO) and contained 98% atom-enriched K15N03.

80

4.3.3 Propagation and Seeding
Saplings, cuttings, and white clover seed were sown or transplanted four weeks prior to the
beginning of the experiment to allow root systems to establish. A few days prior to this planting,
soil in all pots were fertilized with non-nitrogenous fertilizer to avoid nutrient deficiencies
without tampering with nitrogen dynamics. Fertilizers used were 0-45-0 triple super phosphate
(45% P2O5, 15% calcium) and 0-0-50 sulfate of potash (50% potash, 18% sulfate, 2% chloride).
Pots were given fertilizers in amounts proportionate to size. In Treatment (IR) were given 1.5
grams of each fertilizer dissolved in water and pots in Experiment (EX) and Treatment (C) were
given one gram of each.

When transplanting chestnuts, special care was taken to shake off as much pre-existing soil in the
root ball as possible, so as not to influence the soil composition in the experiment. Since the
previous soil was mostly mulch, it was not difficult to shake or brush off the soil. Poplar cuttings
were inserted, buds facing upward, into the soil until only one bud was showing above the
surface to maintain one new stem per planted cutting. In treatment IR, trees were planted in pots
first, prior to seeding the white clover. White clover was seeded in all treatments using the same
method. Prior to sowing, all seeds were manually inoculated with Rhizobium leguminosarum
biovar trifolii. Seeds were spread on a large plastic tray and sprayed gently with a mixture of
whole milk with a small amount of molasses to make sure that the inoculant would stick to the
seed and provide sugar for microbial growth. The inoculant was then sprinkled heavily onto the
seed and mixed by hand. The seeds were spread out on the tray and left in a shaded area to dry

81

for 10-15 minutes before sowing. Once dry, clover was sown at the USDA recommended full
seeding rate of four pounds per acre 144.

To keep the observational units of clover uniform in size and density between both experiments,
seed was spread to create “strips” of clover. One strip was approximately 10 x 30 cm (or 4 x
11”) and required approximately 2.46mL of white clover seed. Seed was spread evenly over the
designated strip area and covered with 0.3-0.5 cm of soil. In Treatment (IR) two strips of clover
were sewn on both sides of the planted tree. In Experiment (EX) and Treatment (C), single strips
were planted in each pot. All pots were well watered after propagation and watered daily until
plants were well established.

4.3.4 Watering
Throughout the course of the experiment pots were watered to minimize competition for
moisture. All water used for the pots was de-ionized to prevent contamination with external
nutrient inputs. Smaller pots, such as in (EX) and (C), typically required watering every day,
especially for pots that had clover in them. Pots in (IR) were watered at least every other day.
After the isotopic tracer had been applied to the clover, all watering going forward was done at
or below the soil surface using a special nozzle for gentle water flow to avoid water coming in
contact with the isotopic label on the clover leaves and contaminating the soil. Additionally,
pots were never watered above the saturation point. By doing this, leaching of the isotopic tracer
was kept to a minimum.

82

4.3.5 Isotopic Tracer Application
Labelling of herbaceous species using a 15N foliar brushing method has been demonstrated in
multiple studies 14,16,142,145,146 and has been proven to be the most effective process for exudate N
labelling 147. Putz et al (2010) effectively labelled red clover (T. pratense) N using a 98atom%
15

N urea and Jalonenen et al (2009) labelled G. sepium trees using 99atom% K15NO3-, both using

foliar application methods. I used the same process to label mobile and immobile nitrogen in the
plant with the 15N tracer. Four weeks after propagation, the isotopic tracer was applied to strips
of clover in Treatments IR and EX after taking the initial sample to ascertain natural 15N
abundance. This marked the beginning of the experimental period, which would last nine weeks
in total, divided into three sampling periods. Tracing methods were modeled after methods
utilized by Jalonen et al. (2009) on G. sepium trees in a similar greenhouse experiment. Total
fertilizer applied to each strip of clover accounted for less than 0.01% of total nitrogen content
after four weeks of growth, which was determined by a pilot study where clover was grown
alone in the same soil and analyzed for nitrogen content.

The tracer was applied in three applications, or every two days over the course of one week. In
each application, each individual “strip” of clover received 2.5 mg of K15N03 diluted in 5mL of
deionized water with 1% v/v Brij-35 (30%) to act as a surfactant and ensure fertilizer was
absorbed onto the leaf surface. The tracer solution was carefully applied to clover leaves
manually with a small paintbrush. This totaled to 15 mg K15N03 in treatment IR pots (two strips
per pot) and 7.5 mg K15N03 in Treatment EX pots over the course of the week. To avoid
contamination of soil or recipient species leaves with the tracer, the soil was lined with a thin
plastic and trees were covered with aerated plastic bags briefly during the application process.

83

4.3.6 Photosynthesis and Growth Measurements
Photosynthesis measurements were conducted on each of the trees in the IR and C treatments
using a LICOR (LI6400XT, Lincoln, NE). The instrument was calibrated before each
measurement and the photosynthesis photon flux density (PPFD) was set at 1800 Âľmol m-2 s-1,
reference CO2 was maintained at 400 Âľmol, and the air flow rate was held at 500 Âľmol s-1.
These measurements were taken three times at the midpoint of each of the three experimental
periods (days 11, 32, and 53).

Growth measurements were taken immediately prior to destructive sampling. Using a standard
meter stick, the height of the tallest living bud in each tree was recorded in centimeters.
Diameters were retrieved using an electronic caliper and recorded in millimeters. In chestnuts,
diameter was taken exactly one inch above the root collar. Since poplars are prolific at sprouting
and they were propagated from cuttings, several samples had more than one shoot. As a result,
diameter was taken from the tallest shoot at exactly one inch above its base in the cutting. Data
for height and diameter in each observational unit were subtracted from the initial measurements
to be displayed as growth during data analysis.

4.3.7 Sampling
At each sampling date, all plant material was separated by compartment (leaf, stem, shoot, root),
weighed, and stored in a cooler at 4-6°C until they were taken back to the lab and immediately
placed in an oven for drying. All soil samples were bagged and stored at -18°C until further

84

analysis. Special care was taken for samples from (EX) to obtain exudate samples, and will be
described in detail in the next section.

In (IR) where clover and tree species were grown simultaneously, delicate handling was required
to obtain samples, particularly with respect to roots. In pots with chestnuts, trees were first cut at
the base to separate roots from above ground biomass. In pots with poplar, new shoots were
removed from the shoot first. Biomass were then separated into stems and leaves, bagged, and
stored in a cooler. Aboveground biomass of clover was harvested using a pair of hand-shears.
Clover was cut at the base of the stem and the resulting sample was then bagged and stored in a
cooler.

To obtain root samples, the soil (which had become an aggregated root ball by the first
experimental sample) was gently removed from the container and place in a large, shallow
plastic bin. Then, the roots were slowly and gently separated by shaking and brushing off excess
soil. Once separated, the roots of the two clover strips and tree roots in each pot were rinsed
with deionized water to remove any residual soil. Both tree and clover roots were thereafter
bagged and stored in a cooler. Poplar roots were separated from the shoot (cutting) before
bagging.

Obtaining plant samples in treatment C was done in the same way as IR but since all species
were grown alone, root retrieval was considerably easier. All compartmentalization of plant
parts remained the same in all treatments (see Table 9). Soil samples were taken after the
removal of roots from the soil aggregates. Remaining soil in the bin was mixed thoroughly by

85

hand and a plastic bag was filled with approximately 0.16 L of soil, which was then placed in a
cooler until it could be stored in a freezer in the lab.

Table 9. Description of sample compartmentalization for
each species in greenhouse experiment.
Plant Sampling Compartmentalization
Poplar
Leaf, Stem, Shoot, Root
Chestnut
Leaf, Stem, Root
Clover
Leaf and Stem, Root

4.3.8 Exudate Experiment (EX) Sampling
One week prior to each sample date (after the initial sample), three replicates were transplanted
to aerated hydroponic tanks to study total exudate potential. Samples were removed from their
pots and root balls were gently shaken to remove as much soil as possible. To further remove
any soil from the roots of the clover, samples were placed on a wire screen with 1.3cm gaps and
rinsed with deionized water until only bare roots were left. The aerated tanks were filled with
approximately 6 liters of deionized water. Clover strips were laid into the tanks on a plastic
mesh lining that allowed roots to be submerged in water but held the aboveground portion of the
plant above the waterline. Once strips were stable in the tanks, the top of the tank was lined with
a plastic wrap to prevent airborne contaminants from getting into the water and effecting
analysis. Tanks were aerated via a single 0.42 amp air pump (max PSI – 3.48) with multiple
hose attachments. Hoses (0.635cm diameter) were attached to O2 diffusing stones which were
placed beneath the roots in each tank (one for each replicate). Tanks were allowed to sit in the
greenhouse under the same light conditions as the other treatments for one week to allow exudate
to collect in the water solution, and this was assumed to be representative of exudate production
in treatment IR.
86

At the end of the week samples in aerated tanks were harvested alongside samples from the other
two treatments. Strips of clover were gently lifted from aerated tanks and allowed to drain of
water for a few minutes. They were then placed on a plastic tray and cut at the base of the stem
to separate aboveground from belowground biomass (stem/leaves vs roots). Biomass samples
were bagged, weighed, and refrigerated prior to drying and further analysis. All exudate solution
volume was measured and placed into multiple 1-liter glass bottles which were then stored in a
refrigerator at 4°C. To obtain exudate, exudate solution was first filtered using Whatman #1
filters to remove any dead plant cells. The remaining solution was then poured in Pyrex glass
beakers (500 mL) and placed in an oven with plenty of air circulation at 50°C to slowly
evaporate water from the solution until only the dry exudate remained. To avoid volatilization of
NH3, solution in all containers was maintained at a pH of 4. This was done via daily monitoring
with an electrode (calibrated daily) and adding sulfuric acid (H2SO4) with a pipette until pH was
less than 4. Total additions of sulfuric acid to solution amounted to less than 0.5 mL in each
sample. Evaporation of all water from the solution typically took between 8-12 days. Once this
process was finished, dried exudate was collected from Pyrex containers, weighed, ground to a
fine powder using a pestle and mortar, and stored in a 25mL plastic Nalgene bottle at room
temperature until analyzed for 15N and total nitrogen. Since the collected samples only represent
the middle week of each sample period, N content values were multiplied by three to extrapolate
exudation to cover the entire sample period.

87

4.3.9 Laboratory Analysis
Soil samples were mixed thoroughly and a subsample of 5 grams was taken and mixed with 50
mL of potassium chloride (KCl). Samples were then placed in a shaker and mixed for 1 ½ hours
for nutrient extraction. After shaking, samples were filtered using Whatman No. 1 filters prewashed in potassium chloride solution to remove soil sediment. The resulting extracted nutrient
solution was analyzed for nitrate levels using a SAN++ segmented flow analyzer (Skalar Inc.,
Atlanta, GA) measured at 540nm. An additional 10 grams of soil were placed in an oven at
65°C for 72 hours and thereafter weighed to retrieve soil moisture content. Results of the
analysis were expressed in mg/L and were transformed to concentrations (Âľg/mg) using Equation
1.

All compartmentalized plant material obtained from sampling was weighed before and after
drying to determine green and dry weights. Material was dried in a ventilated oven at 65°C for
72 hours and then ground to a fine powder using a coffee grinder. The grinder was cleaned after
each sample using a mixture of 200 proof ethyl alcohol and deionized water (4:1) to avoid crosscontamination. Poplar shoots were too large to grind with a coffee grinder and had to be
processed using an electric wood mill that ground it to 0.2 mm or less. Dried, ground samples
were then funneled into small 25 mL plastic Nalgene vials and stored at room temperature until
further analysis.

To obtain measurements of nutrient concentrations in plant material, a subsample of 0.3 grams (Âą
0.005) was weighed and placed in 100 mL test tube mixed with 4.5 mL of sulfuric acid (H2SO4)
in a vortex and allowed to sit overnight for pre-digestion. Then, 1.5 mL of (30%) hydrogen

88

peroxide (H2O2) was added to each tube and then placed in a block digester (AIM600 Block
Digesting System) at 340 ¹ 10°C for one hour under a programed temperature schedule. After
digestion, sample tubes were filled to 100 mL with deionized water and mixed well with a
vortex. To determine N and P concentrations, the digested sample mixture was then buffered
and chlorinated after dialysis to form a chemical complex measured at 660nm and 880nm on a
SAN++ segmented flow analyzer (Skalar Inc., Atlanta, GA). Measurements of K and Ca were
obtained using the digested solution in an atomic absorption spectrometer (AAnalyst 400, Perkin
Elmer). Values were expressed in parts per million (ppm) or mg/L and were transformed into %
nutrient and nutrient content using the sample dry weights. Measurements of nitrogen and
phosphorous were taken in all treatments (IR, EX, C) in both clover and tree species. Potassium
and calcium measurements were taken only in trees. Nutrient analysis was done by plant
compartment and as plant totals. Values were expressed as both concentrations and content.

4.3.10 Isotopic Analysis
Isotopic analysis of nitrogen content was done at the UC-Davis Stable Isotope Facility in Davis,
California via PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ Europa 20-20
isotpe ratio mass spectrometer (Sercon LTD., Cheshire, UK). Plant samples were combusted at
1000°C in a reactor packed with chromium oxide and silvered copper oxide. After combustion,
the oxides were removed in a reduction reactor (reduced copper at 650°C) and a helium carrier
flowed through a water trap (magnesium perchlorate) and a CO2 trap. N2 and CO2 were then
separated on a Carbosieve GC column (65°C, 65 mL/min) before going into an IRMS 148.

89

Soil samples were analyzed for isotope ratios using an Elementar Vario EL Cube (Elementar
Analysensysteme GmbH, Hanau, Germany) interfaced to a PDZ Europa 20-20 isotope ratio mass
spectrometer (Sercond Ltd., Cheshire, UK). Samples were combusted at 1080°C in a reactor
with copper copper oxide and tungsten (VI) oxide. After combustion, oxides were removed in a
reduction reactor (reduced copper at 650°C) and the helium carrier then flowed through a water
trap (magnesium perchlorate). N2 and CO2 were separated using a molecular sieve adsorption
trap before entering the IRMS 148.

Throughout the analysis are interspersed replicates of no less than two different laboratory
standards that have been calibrated against NIST Standard Reference Materials 148. Final values
are expressed in delta form which are expressed relative to international standards VPDB
(Vienna PeeDee Belemnite) and Air for carbon and nitrogen, respectively. Delta values were
found using Equation 2.

δ15𝑁(𝑠𝑎𝑚𝑝𝑙𝑒) =

𝑎𝑡𝑜𝑚%15𝑁(𝑠𝑎𝑚𝑝𝑙𝑒) − 𝑎𝑡𝑜𝑚%15𝑁(𝑎𝑡𝑚𝑜𝑠𝑝ℎ𝑒𝑟𝑒)
𝑥 100
𝑎𝑡𝑜𝑚%15𝑁(𝑎𝑡𝑚𝑜𝑠𝑝ℎ𝑒𝑟𝑒)

Equation 2. Equation used by UC-Davis Isotope Lab to determine 𝛅15N values from
isotopic analysis. Here, “atom%15N(sample)” refers the percentage of 𝛅15N atoms in
sample and “atom%15N(atmosphere)” refers to the atmospheric 𝛅15N concentration
(0.3663%).

4.3.11 Data and Statistical Analysis

The 𝛅15N values received from isotopic analysis were then applied to Equation 4 to
determine the percent of nitrogen derived from transfer (%Ndft) in each tree. Using this
90

formula, %Ndft can be found in both trees at each sampling period. Since 𝛅15N values were
compartmentalized into the different parts, 𝛅15N were averaged across all compartments in
each observational unit to be applied to the formula.
%𝑁𝑑𝑓𝑡(𝑡) =

δ15𝑁𝑟(0) − δ15𝑁𝑟(𝑡)
δ15𝑁𝑠(0) − δ15𝑁𝑑(𝑡)

Equation 3. Equation to determine the percent
nitrogen derived from transfer (%Ndft) for each
tree species at each sampling period. Where “r”
denotes recipient, “d” denotes donor, “s” denotes
soil, “0” denotes the beginning of the experiment,
and “t” denotes the sample period.

Values of transferred milligrams of nitrogen in each tree, percentage of clover exudate that
was absorbed by each tree, and the percentage of total clover nitrogen content that was exuded
by the end of the experiment, were determined using Equations 4,5, and 6, respectively.

N(Transfer) = %Ndft(F) x N(Tree)
Equation 4. Conversion of %Ndft into actual N
transfer (mg). Where “%Ndft(F)” refers to the
final value of %Ndft for each tree at the end of
the experiment and “N(Tree)” refers to the total
estimated N content of each tree in milligrams.

%ExNAbs(Tree) =

N(Transfer)
ExN(F)

Equation 5. Equation for calculating the
proportion of available exudate from clover
actually absorbed by each tree. Where
“N(Transfer) denotes the milligrams of N
sourced from transfer in each tree and “ExN(F)
denotes the total N exuded over the course of
the experiment in milligrams.

91

%NEx(Clover) =

N(Transfer)
(N(Transfer) + N(Clover))

Equation 6. Calculation for determining the proportion of clover
nitrogen lost through exudation. Where “N(Transfer)” refers to
tree N derived from transfer at the end of the experiment and
“N(Clover)” refers to total N content of the clover at the end of the
experiment.

The above values along with values regarding soil nitrate, plant nutrient concentrations and
content, photosynthetic activity, height, and diameter were compared for significant differences
between treatments using an analysis of variance. Means were then separated using a Tukey’s
test of honestly significant difference. Linear correlations were tested using Pearson’s
correlation analysis. All tests used values of P lower than 0.05 as an indicator of statistically
significant difference. All data analyses were performed in R 3.3.1 software, as was the creation
of graphs and tables 138.

4.4

Results

4.4.1 Nitrogen Transfer
Parameters describing transfer are expressed in 𝛅15N, percent of tree nitrogen derived from
transfer (%Ndft), tree N content (mg) derived from transfer (mgNdft), and percent of
available clover exudate absorbed by trees (%ExAbs).

4.4.1.1

𝛅15N Values

Clover in treatments EX and IR were successfully and uniformly labelled by the K15NO3fertilizer. Table 10 shows the 𝛅15N values of the clover and trees in treatment IR by plant
compartment. Clover leaf 𝛅15N in both poplar and chestnut pots increased significantly three
weeks after the isotope application at day 0 (p = 0.00 and 0.00), as did clover roots (p = 0.02 and
92

0.03) and then decreased over the next six weeks as the 15N was diluted from continued growth,
N assimilation, and N transfer.

In chestnut, 𝛅15N values in roots and leaves differed significantly at the end of the experiment
from Day 0 (p = 0.00 and 0.01). Chestnut stem 𝛅15N increased after the first sample period but
slowly decreased over the next six weeks.
Poplar leaf and root 𝛅15N were also significantly higher at the end of the experiment than the
beginning (p = 0.00 and 0.00).

There were no significant differences in tree leaf or stem 𝛅15N values between poplar and
chestnut at any of the sample periods. Nor were there significant differences between clover
roots or leaves between poplar and chestnut treatments. Clover leaf 𝛅15N in treatment EX were
lower than those in treatment IR at day 21 of the experiment, but did not differ at day 42 or at the
end of the experiment. More importantly, clover roots in treatment EX did not differ from clover
roots in IR.

The 𝛅15N values of both species within treatment IR were positively and significantly correlated
with days since the isotopic labelling of the clover. Poplar trees showed a Pearson Correlation
Index (PCI) of 0.69 and an adjusted r2 of 0.44 (p = 0.00). Chestnut trees had a PCI of 0.69 and
an adjusted r2 of 0.44 (p = 0.00).

93

Table 10. Summary of 𝛅15N values in treatment IR by species and compartment over the course of the
63-day experiment.

Soil 𝛅15N values were higher at day 63 than at day 0 in chestnut pots (p=0.049). In poplar pots,
soil 𝛅15N was higher in each successive sample period than at day 0 (p = 0.02, 0.01, 0.00).
Lastly, exudate 𝛅15N in treatment EX differed significantly from day 21 to day 42 (p=0.02), day
21 to day 63 (p=0.00), and strongly from day 42 to day 63 (p=0.07). Additionally, soil 𝛅15N was
positively correlated with both chestnut and poplar roots (p = 0.00 and 0.00, r2 = 0.58 and 0.72,
and PCI = 0.78 and 0.86).

Figure 10. Average 𝛅15N values for poplar and chestnut soils in treatment IR
over the course of the experiment.

94

Clover root 𝛅15N and exudate 𝛅15N in treatment EX was positively correlated (p=0.00) with an
adjusted r2 = 0.84 and a correlation of 0.93. Clover roots 𝛅15N and tree root 𝛅15N were positively
correlated in both chestnut and poplar pots in treatment IR (p = 0.03 and 0.00). Additionally,
poplar and tree average 𝛅15N were both positively correlated over successive sample periods (p =
0.00 and 0.00).

4.4.1.2

Exuded Nitrogen

Table 11 describes nitrogen parameters recovered from treatment EX. Exuded N amounts
remained steady throughout the experiment and did not differ significantly between periods.
Daily exudation throughout the experiment averaged at 2.62 (Âą1.24) mg N. Total exudation over
63 days amounted to 172.8 (Âą78.5) mg N.

The percent of clover total N content that was exuded did not significantly differ between
experimental periods. Daily exudation averaged at 0.16% (Âą0.04) of total clover N. Total
experimental exudation amounted to 3.4% (Âą0.85) of total clover N.

Figure 11. Average 𝛅15N values (line) and N content in milligrams (bars) of white
clover exudate in exudate experiment (EX). Horizontal bars represent standard error.
95

Table 11. Table summary of parameters from clover
exudate experiment (EX).

4.4.1.3

Percent N Derived from Transfer

The 𝛅15N values were averaged by individual trees within pots and applied to equations 3, 4, and
5 to determine %Ndft, mgNdft, and %ExAbs, which are shown in Table 12.

Table 12. Summary of N transfer parameters from isotopic labelling experiment.

96

Transfer was measured as cumulative throughout the successive sample periods. In chestnuts,
%Ndft values were higher than day 21 at day 63 (p = 0.01). In poplar pots, estimates were not
significantly different from day 21 to the end of the experiment. Total estimated N transfer for
chestnut treatments at the end of the experiment was 33.3% (Âą7.4) of total tree N. In poplars,
estimated N transfer ranged between 12.1% (Âą1.1) of total tree N. Percent transferred N differed
significantly between poplar and chestnut at the end of the experiment (p = 0.04).

Chestnut Ndft% was positively correlated with days since the beginning of the experiment (p=
0.00) with an adjusted r2 of 0.63. The same was true in poplar treatments (p = 0.00), with an
adjusted r2 value of 0.78.

Figure 12. Percent of nitrogen derived from
transfer (%Ndft) by species over the course of the
experiment.

4.4.1.4

Tree N Content Derived from Transfer (mgNdft)

Nitrogen content derived from transfer within trees showed similar patterns to %Ndft, as is to be
expected. However, there were no significant differences. Positive and significant correlations
were found between mgNdft and days since the beginning of the experiment in chestnut (p =
0.01) and poplar (p = 0.00). Total N transfer in chestnut pots was 30.14(Âą12.7) mg N per pot and
97

11.97(Âą1.8) mg N per pot in poplars. No significant differences existed between species
treatments.

Figure 13. Nitrogen content derived from transfer
by species over the course of the experiment

4.4.1.5

Percent of Exuded N Absorbed (%ExAbs)

Percent of total clover-exuded N absorbed into the trees were produced on the assumption that
all transfer occurred via root exudates and had the same trends as N content transfer. There were
no statistically significant difference between sample periods in poplar or chestnut pots. No
significant correlations were found between the percent of exudate absorbed and days since the
beginning of the experiment. Percent exuded absorbed in chestnut pots at the end of the
experiment was 17.4% (Âą8.2) per pot by the end of the experiment and 6.9% (Âą0.7) per pot in
poplars. No significant differences existed between species treatments.

98

Figure 14. Percentage of available exudate absorbed by
each tree species over the course of the experiment.

4.4.2 Tree Nutrients (N, P, K, and Ca)
Table 13. Summary of nutrient content per tree between treatments IR and C
at the end of the experiment.

99

Table 14. Summary of average nutrient concentrations between treatments IR and
C at the end of the experiment.

Nitrogen
Chestnut leaf N concentrations were significantly higher in the control throughout all four
sample periods (p = 0.00, 0.00, 0.00, 0.02). Chestnut roots in the interaction treatment were
higher than in the control at day 0 (p = 0.00). Besides these, no other differences in N
concentrations were observed in chestnut pots. Poplar shoots in the interaction treatment had
higher N concentrations at day 0 (p = 0.00). While poplar leaves in the control treatment had
significantly higher N concentrations at days 21 and 42 (p = 0.00, 0.00), N concentrations did not
differ in any compartment between poplar treatments by day 63. In both tree species, N

100

concentrations between interaction and control treatments differed more in early sample periods
but became more similar towards the end of the experiment.

Chestnut leaf N content was higher in controls at days 21, 42, and 63 (p = 0.00, 0.00, 0.00).
Poplar leaf N content showed the same trend with controls being higher in the last three sample
periods (p = 0.00, 0.00, 0.00). Total tree N content values were higher in chestnut controls at
days 42 and 63 (p = 0.00, 0.00) and higher in poplar controls at days 21, 42, and 63 (p = 0.00,
0.00, 0.00). Poplar N content was higher than chestnut N at day 0 (p = 0.00) but no differences
occurred between species treatments in subsequent sampling periods.

Clover N contents among treatments did not differ at day 0 of the experiment. At day 21, clover
grown in control treatments was higher than in treatment EX (p = 0.02), but did not differ from
clover grown in treatment IR. At day 42, control treatments had higher N content than in
treatment EX, IR (chestnut), and IR (poplar) (p = 0.00, 0.00, 0.00). By day 63, treatment EX had
lower N contents than in C, IR (chestnut), and IR (poplar) (p = 0.00, 0.00, 0.01), but there were
no differences between control and interaction treatments. Clover N content and tree N content
showed no correlation in chestnut pots from treatment IR. However, poplar pots showed a
negative correlation between the two factors (p = 0.04) with an adjusted r2 of 0.19 and a
Pearson’s Correlation Index of (-0.499).

101

Figure 15. Changes in nitrogen content over time in clover and trees
within treatment IR.

Phosphorous
Chestnut leaf P concentrations were higher in controls at days 42 and 63 (p = 0.04 and 0.00).
Root concentrations were higher in controls at days 21, 42, and 63 (p = 0.00, 0.00, and 0.00).
Stem concentrations were higher in controls at days 42 and 63 (p = 0.03 and 0.00).

Poplar leaf P concentrations were higher in controls at day 21, 42, and 63 (p = 0.03, 0.00, and
0.00), poplar stem concentrations were higher at days 42 and 63 (0.03 and 0.00), and poplar roots
had higher P concentrations at day 63 (p = 0.00).

Total chestnut tree phosphorous was significantly higher in controls by day 63 (p = 0.00) and
total poplar tree phosphorous was higher in controls at days 42 and 63 (p = 0.00 and 0.00). No
differences in total P content were observed between poplar and chestnut pots throughout the
experiment.

There were no differences in clover P content between poplar and chestnut interaction treatments
at any point. At the end of the experiment, clover in treatment EX had a higher P content than

102

both interaction treatments (p = 0.00 and 0.00). Clover grown in controls had higher P content
than EX, IR (chestnut), and IR (poplar) (p = 0.046, 0.00, and 0.00).

Clover P content and tree P content showed no linear correlations in chestnut pots from treatment
IR. However, poplar pots in this treatment did show a significant positive correlation between
clover and tree P (p = 0.02) with an adjusted r2 of 0.31 and a Pearson’s Correlation Index of
0.595.

Potassium
Potassium was measured in both chestnuts and poplars from treatments IR and C. Chestnut leaf
K concentrations were higher in control treatments from day 21 until the end of the experiment
(p = 0.01, 0.00, and 0.00). In poplars, both leaves and stems in control treatments had higher
concentrations from day 21 onward (p = 0.00, 0.00, 0.00, and 0.00). Additionally, control poplar
roots were higher at day 63 (p = 0.00). The same differences were observed in both trees in
regards to K content by plant compartment. Total tree K content in chestnut controls were found
to higher at days 42 and 63 (p = 0.00 and 0.00). In poplars, control tree K content was higher at
days 21, 42, and 63 (p = 0.01, 0.00, 0.00)

Calcium
Fewer significant differences were found in terms of calcium concentrations and contents. No
differences in Ca concentrations were found amongst different plant components between
control and interaction treatments in poplar or chestnut plots. However, chestnut and poplar leaf
Ca contents were higher in controls than in interaction treatments at day 63 (p = 0.00 and 0.00).

103

Similarly, both poplar and chestnut total tree Ca contents were higher in controls by the end of
the experiment (p = 0.00 and 0.01).

4.4.3 Soil Nitrogen

In control and interaction treatments, soil nitrate levels declined from day 0 to day 63 of the
experiment, with a levels being significantly lower at the end than at the beginning (see Figure
16).

Figure 16. Average soil nitrate concentrations in treatments IR and C over
the course of the experiment, including initial soil samples from bagged
soil (-28).

At day 0, poplar controls showed higher nitrate levels than chestnut control pots (p = 0.00), this
reversed by day 21, with chestnut controls being higher (p = 0.00). Curiously, chestnut soil
nitrate in control pots rose slightly between days 0 and 42, before declining to levels similar to
other treatments by day 63. As a result, chestnut control pots had higher nitrate levels than
chestnut interaction pots at days 21 and 42 (p = 0.00 and 0.00). However, by the end of the
experiment, there were no significant differences in soil nitrate concentrations between all
treatments.
104

Table 15. Summary of average soil nitrate concentrations in
treatments IR and C.

Soil organic N measurements were taken in treatment IR pots as a result of the isotopic analysis.
An increasing trend was seen from day 0 to day 63 in percent soil N in both chestnut and poplar
pots (see Figure 21). However, no significant differences were found between sample periods or
tree species.

Figure 17. Changes in average soil organic nitrogen in treatment IR over
time.

105

4.4.4 Photosynthesis
Poplars in both treatments IR and C declined in photosynthetic activity from day 0 to day 63, and
did not differ between treatments at any point. Both treatments were significantly and inversely
correlated with days since the beginning of experiment Pearson Correlation Indexes of (-0.56)
and (-0.73) for treatments C and IR, respectively.

Chestnut controls showed higher photosynthetic activity than chestnuts in treatment IR in the
first sample period (p = 0.01) but declined over the next two periods, while treatment IR grew
(see Figure 25). Chestnuts in treatment IR did not have a significant linear correlation, while
controls showed a significant negative correlation (p = 0.00) with a Pearson’s Correlation of (0.74).
Table 16. Summary of photosynthesis measurements in treatments IR and C

106

Figure 18. Photosynthetic activity in treatment IR (dotted line) and C (solid line)

4.4.5 Height and Diameter
Since chestnut trees were purchased as one-year saplings, there was an existing degree of
morphological variability going into the experiment. To eliminate the effect of this, values of
height and diameter were expressed in relative growth compared to the initial sample (shown in
Table 17). Analysis of chestnut height growth showed no significant differences between sample
dates or treatments. In poplar pots, which were started from cuttings, control treatments had both
higher relative height (p = 0.00) and diameter (p = 0.00) compared to the interaction treatment by
the end of the experiment.

107

Table 17. Summary of relative tree growth at the end of the
experiment.

4.4.6 Other Correlations between Variables
Percent nitrogen derived from transfer and N content derived from transfer were positively and
significantly correlated in both chestnuts and poplars (p = 0.00 and 0.00). Furthermore, N
content derived from transfer was also positively and significantly correlated with total tree N
content in both species (p = 0.01 and 0.00)

In both poplar and chestnut pots of treatment IR, there was a negative correlation between soil
nitrate and percent nitrogen derived from transfer (%Ndft). As soil nitrate declined throughout
the experiment, nitrogen transfer continued despite the lack of resources. Chestnuts had a PCI of
(-0.55) and poplars had a PCI of (-0.86) (p = 0.02 and 0.00).

Photosynthetic activity was not correlated with leaf N content in poplar controls or interaction
treatments. In chestnuts, photosynthesis was positively and significantly correlated with leaf N
content (p = 0.00, r2 = 0.67) but not in the control (p = 0.08).

108

Regardless of exudate or root estimates, photosynthetic activity was positively correlated with N
content derived from transfer (mgNdft) within leaves of chestnuts (p = 0.04, PCI = 0.86, r2 =
0.72). Conversely, poplar trees showed a significant negative correlation for the same variables
(p = 0.00, PCI = (-0.62), r2 = 0.32).

4.5

Discussion

The goal of this study was to establish the existence of and quantify the direct nitrogen transfer
mechanism between herbaceous legumes and tree species. Due to limitations in resources, this
project did not separate nitrogen transfer between mycorrhizal and exudate sources. However,
since existing research has shown that mycorrhizal transfer typically makes up a relatively small
portion of total transfer 15,16, this did not harm the integrity of the study.

The use of 15N leaf-feeding techniques allowed us to discern between the relative contributions
of the soil and clover exudate to tree N content. Labelling techniques in tracing experiments
similar to these have been shown to be reliable in multiple studies 14,16,17,121,141,146. After
labelling, the 𝛅15N values of clover were similar to those found in previous studies on this
species (T. repens), confirming that the clover was effectively labelled 141. The added N to the
clover that occurred through labelling accounted for less than 0.1% of the total clover nitrogen
by the end of the experiment. As a result, leaf labelling did not likely physiologically affect
clover growth or exudation and therefore would not influence estimates of N transfer. The
decrease in 𝛅15N values in clover biomass compartments and exudates over time was a result of
biomass accrual and subsequent 15N dilution.

109

4.5.1 Clover Exudates
Clover N content in the exudate experiment (EX) was lower than that in the interaction treatment
(IR) by the end of the experiment, however this was the only difference seen between these
treatments and could have been due to the week spent in hydroponic tanks. As a result, the lack
of statistically significant differences in either N content or 𝛅15N between these treatments
throughout the majority of the experiment implies that exudate values obtained from treatment
EX were representative of transfer in interaction treatments. It is important to note, however,
that exudation potential is often positively influenced by other root systems within the
rhizosphere 115. Estimates for exudation potential came from clover grown alone, thus, the lack
of interaction with other roots in this treatment could result in different exudate estimates.

The exudate was clearly labelled by the 15N isotope, meaning a portion of the exudate came from
mobile N within the plant. The relatively lower 𝛅15N values seen in the exuded N compared to
clover biomass, particularly towards the end of the experiment, supports the fact that exuded
mobile N are usually recently fixed compounds within root nodules 116,146.

Exudation potential of white clover (T. repens) amounted to roughly 2.74 mg N day-1 pot-1. Few
studies have explored the exudation potential of this species. Høgh-Jensen and Schjoerring et al
(2000) revealed that white clover exuded 4.5-7.5 g N year-1 m-2 to rye grass in mixed clovergrass leys in Copenhagen.

When extrapolated to a square meter, the exudation potential of

white clover in this study would be 0.04 g N m-2 day-1, which would put it within a similar rate,
given similar growing season lengths (around 150 days). Additionally, the percent of total clover
nitrogen that was exuded by the end of the experiment (3.4%) falls in line with estimates from

110

previous research, which suggest about 3% of N captured during biological nitrogen fixation is
secreted from the roots through exudation 116.

4.5.2 N Transfer
Values of nitrogen transfer in this study were retrieved using only exudate 𝛅15N as the “donor”
source in equation 3. However, since the 15N label could be unevenly portioned within the
clover, 𝛅15N values of tree roots and exudates could differ 16. The 𝛅15N values of the clover
roots represent both mobile N and N bound in radical structure within total root N content
15.

These values are used in situations where decaying root cells are the main form of

transfer. Because of this, a similar study by Jalonen et al. (2009) used the same equation to
determine the contributions of N from G. sepium trees to D. aristatum, but also substituted the
𝛅15N values of tree roots to account for this potential difference.

In this study, 𝛅15N values did not differ significantly between exudate and clover roots. Isotopic
analysis of cover in (IR) and (EX) revealed that the clover leaves, roots, and exudates were all
effectively labelled and showed similar patterns to what has been seen in previous 15N transfer
experiments 15,16. Also it is extremely unlikely that the young clover roots (111 days old by the
end of the experiment) would have had any significant issues with senescence or root decay 141.
Clover roots and stolons have high longevity, averaging 290 and 411 days before showing signs
of decay, respectively 149. Moreover, it has been suggested by multiple publications that the
majority of transfer takes place through exudates during early growth stages of legumes 13,141.
Considering also that soil organic N did not significantly rise during the experiment, it can be
safely assumed that the majority of transfer taking place in this experiment was through direct

111

transfer pathways. As a result, this study did not use 𝛅15N values from clover roots in transfer
estimates.

This study did not attempt to differentiate between direct transfer through exudates or through
common mycelial networks as other studies have 146. However, the use of potting soil as the
growth medium in this study would likely have hampered mycelial growth due to a lack of
natural fungal populations. Additionally, other studies, which used natural soils, reported a low
contribution to overall transfer (<1%) through fungal connections 146. As a result, it can then be
assumed the majority of direct transfer in this study likely occurred via living root exudates.

Estimates of nitrogen transfer were statistically higher in chestnuts compared to poplars at the
end of the experiment. Since 𝛅15N values between clover from poplar and chestnut pots in
treatment IR were very similar, differences in transfer could be due to varying abilities to capture
exudate between species or differing rhizospheric interactions between root systems, which can
greatly affect exudation 13. The difference in transfer between species could also be attributed to
differing root patterns. Since concentrations of N exudates are higher closer to the roots, transfer
is greater when roots of the two species come in close proximity 13. No measures of rooting
patterns were taken during the experiment but an observed difference between the two species
revealed that chestnuts likely had a wider-spreading root system and likely came in closer
contact with clover roots. Poplar roots tended toward the bottom of the pots they were in, and
had few lateral roots in contact with clover.

112

The final estimates of transfer for poplar (12%) and chestnut (33%), are within range of previous
estimates in other studies. For instance, a similar study quantifying transfer from clover to grass
had estimates as high as 51% 142. Transfer rates increased throughout the experiment in both
species. Since exudation remained relatively constant throughout the experiment, with no real
differences between sample periods, the observed increases were likely due to an increased
ability to absorb available exudate by the end of the experiment. Figure 15 shows the percent of
exudate absorbed remaining relatively constant for poplars, which explains the lower transfer
values. In chestnuts, however, ability to absorb the available exudate ranged considerably, being
highest at day 63, where the largest boost in transfer was observed. Ultimately, estimates of the
proportion of available exudate absorbed for poplar (7%) and chestnuts (17.5%) were quite low,
although a similar study found similar absorption rates at 22%15. The higher values seen in
chestnuts were likely due to a wider spreading root crown that would be more capable of
capturing exudates. Considering that nitrogen transfer increases in field settings over time
16,21,146,150

, these are promising figures for such a short period. Furthermore, since a considerable

amount of the exudate ends up in the soil, this could facilitate long-term soil nutrient cycling and
greater indirect transfer 146. Root exudates are also known to positively influence soil microbial
activity and can stimulate symbiotic relationships with rhizobia and mycorrhizal fungi, which
could encourage transfer through fungal networks and indirect pathways, creating a positive
feedback between nitrogen fixation, transfer, and exudation 113.

Since these estimates came from a controlled greenhouse study using potting soil and pots, they
may not be entirely representative to what would occur in a field setting over the same period of
time. The nature of the pots likely influenced the root structure of both plants in treatment IR,

113

and could mean there was more root-to-root contact. As a result, transfer estimates from this
study could be higher than what they would be in a field in the same time frame where roots
would have more space to grow and possibly avoid contact. However, as was just mentioned
and proved in other studies, nitrogen transfer in field settings increases with time and root growth
even in field settings, indicating that these estimates could be representative of long-term
nitrogen transfer potential. Additionally, a field setting may actually increase transfer through
mycelial networks due to a higher natural abundance of fungal populations.

3.5.3 Nutrient Dynamics, Photosynthesis, and Growth
Both poplar and chestnut trees had higher concentrations and contents of N, K, and Ca by the
end of the experiment. The largest differences were seen in the potassium contents of trees
between treatments. There was significant evidence for competition over all four of these
nutrients between trees and clover, particularly nitrogen. The decrease in soil nitrate over the
course of the experiment in interaction treatments indicates a priority use of soil N resources
over biologically fixed nitrogen by clover and trees. Additionally, the lack of significant
differences between clover N content in controls and in treatment IR indicates that clover was
able to outcompete the tree for this plant-available soil N. Due to the limited available soil N
sources early on in the experiment, it is reasonable to believe a large portion of the N after that
point was biologically fixed by the clover. Since exuded N is often recently fixed, it is likely that
N transfer became a more important resource for the trees in interaction treatments, as soil
resources were depleted.

114

Despite the obvious signs of competition for nutrients between clover and trees in interaction
treatments, these surprisingly did not seem to affect photosynthetic activity. In poplar pots,
photosynthesis was practically the same between control and interaction treatments, with
treatment IR being slightly higher throughout. Perhaps even more surprising were the results of
chestnut photosynthesis measurements. While values in the control treatment were significantly
higher than in treatment IR at the first sample period, they showed a declining trend throughout
the entire experiment. Meanwhile, chestnuts in treatment IR increased from the beginning to the
end of the experiment. Photosynthesis was linearly correlated with leaf N content in chestnuts
from treatment IR but not in any of the other treatments. Moreover, the steady accrual of 15N
within poplar and chestnut leaves indicates the importance of transfer to foliar nitrogen. It is
possible that leaf nutrient ratios could be partially responsible for this difference in
photosynthesis activity. Ultimately, however, the importance of these values to this research is
that nutrient competition did not seem to affect photosynthetic activity over the period of this
short study.

In terms of growth measurements, no significant differences found between controls in chestnut
pots indicates another variable that was seemingly unaffected by nutrient competition. In
poplars, however, controls showed greater height and diameter growth. Considering that poplars
were started from cuttings and chestnuts were bought as one-year saplings with variable heights
and diameters, poplar growth data may be more reliable. However, the short period of time over
which the experiment took place diminishes the importance of the growth parameters in general.

115

The purpose of establishing the controls was to get an idea of how the presence of clover
affected the overall nutrition, growth, and metabolism of the tree species when grown together.
The circumstances under which these were grown in this study are very different from what
would occur in a field setting. Limited space for root growth in pots as well as limited potential
for nutrient cycling capability and finite nutrient availability in the potting soil that wasn’t
regularly fertilized could explain why trees in controls did so much better. Thus, the results of
the nutrient analysis probably do not accurately represent competition dynamics where roots are
not inhibited by the space limits of a pot and where resources are provided by natural soils.
Trees in the control treatment experienced almost no limitations in nutrients during the 63-day
experiment, which would likely not be experienced in the field over time without external inputs.
Since nitrate levels were the same between interaction and controls by the end of the experiment,
trees in controls would likely begin to suffer from nitrogen limitations. At this point trees in the
interaction treatment would have an advantage over controls based on the availability of
nitrogenous exudate from clover.

4.6

Conclusion

The goal of this study was to establish and quantify the existence of a direct transfer pathway for
nitrogen in a controlled environment. The values of percent nitrogen derived from transfer in
both poplar (12.1%) and chestnut (33.3%) over nine weeks shows the likely importance of this
pathway. Due to the age of the white clover, the generally higher 𝛅15N values in exudate
compared to clover roots, and the use of potting soil, it is extremely likely that the majority of
transfer occurred through root exudates and not through mycelial connections or root decay. The
higher values of transfer, combined with the lack of difference in photosynthetic activity and

116

morphological features between control and interaction treatments, indicate that chestnut root
structure may be better suited to direct nitrogen transfer compared to poplar in the short-term.
This difference falls in line with the influence of species selection on transfer stressed by Roggy
et al (2004).

While nutrient analysis of controls revealed significant signs of competition between clover and
trees, these circumstances are not representative of what would occur in the field. Additionally,
the negative correlation between soil nitrate and transfer in interaction treatments indicates the
ability of trees to persist in environments with low nutrient availability with the help of direct N
transfer. This is reinforced by the fact that exudation is positively influenced by higher
competition for N, typically as a result of increased biological nitrogen fixation by associated
legumes 116,141,146. The additions of N exudates to the soil would also likely benefit the long-term
soil nutrient availability in such systems. It is therefore reasonable to believe that in a field
setting, alley cropping chestnut and poplar plantations with white clover would benefit nutrient
cycling and growth of trees over time. This has been shown to be true in similar studies

20–

22,146,150

Ultimately, this study, combined with previous research, shows a general trend of an increased
importance of direct N transfer in systems. This implies that alley cropping systems such as
these could be beneficial to landowners seeking low-input production, particularly on marginal
lands with low fertility. They could also prove to be very effective at restoring soils, due to
consistent N transfer and additions to the soil. The exudate transfer mechanism for nitrogen
provide an important shortcut to the remineralization process that indirect N transfer requires,

117

and could therefore be very useful to landowners. Similar tracing studies on a longer timeframe
and in the field, preferably over the length of a full production cycle, are required to more fully
understand how exudate-based transfer can benefit tree plantations.

118

5

Discussion and Research Implications

Previous research has suggested that decaying biomass is the transfer pathway that has the most
potential for N transfer. The field study in this research did not discern between modes of
transfer, and is therefore not applicable to the argument of relative pathway importance. The
greenhouse study, however, revealed an important belowground relationship between trees and
herbaceous legumes. White clover (T. repens) directly supplied 12.1% of poplar N and 33.3% of
chestnut N. Due to the age of the white clover and the high exudate potential during early root
growth, it is extremely likely that the majority of transfer occurred via root exudates. Since these
values were derived under controlled conditions, they could differ in a field setting. However,
research has suggested that direct transfer pathways increase with time in the field, as exudates
accrue in the soil 16,98. While these findings do not disprove the relative importance of indirect
transfer, they may suggest that the role of direct transfer has been historically underestimated.
Nevertheless, the role of indirect transfer via green manure in improving soil health and poplar
nutrition in the field study was obvious. Treatments involving red clover (T. pratense) regularly
returned more green biomass and nutrients to the soil during mowing, which elevated soil nitrate
and nitrogen in poplar leaves over the course of just two years. Despite, this there was a lack of
apparent benefits to poplar growth. However, similar studies have shown that this amount of
biomass return benefits tree growth in the long-term 19–22. The results of the field study suggest a
similar relationship is likely between poplar and red clover over the next several growing
seasons.

Ultimately, this research affirms the importance of nitrogen transfer in agroforestry systems and
the need to gain a better understanding of it. It falls in line with several other studies that stress
119

the underrated importance of exudate-based N transfer and the benefits of cover crops to tree
plantations in the field. It also reinforces the importance of species selection to N transfer
dynamics. Nitrogen transfer can differ widely depending on nitrogen-fixing capacity of the
donor, differing root patterns, and differing growth phases 14,136. The notable difference in N
transfer between poplars and chestnuts seen in the greenhouse study supports this theory.

Much more research is needed to understand the mechanisms of N transfer, particularly in the
long-term. As previously mentioned, most longer studies have shown that species relations of
these kind are often competitive in the short term but ultimately end up being beneficial.
However, no long-term field studies have discerned between N transfer mechanisms that result in
beneficial relationships. Additionally, few studies have attempted to analyze the effects of
management methods on direct transfer. Regular cutting of cover crops, for instance, is shown to
increase N transfer through root turnover 149. The effects of cutting on exudate transfer,
however, have not been measured. Similar isotope studies in the field are required to understand
how direct mechanisms change over time and to determine their actual importance at the end of a
production cycle. This research adds to a small, but powerful block of research that has proven
the important role of N transfer dynamics in alley cropping systems. Hopefully, it will assist
future studies that seek to address remaining research gaps regarding the effects of time scales,
physical and biological factors, and species selection on transfer.

Locally, this research could directly benefit the biofuel and tree nut orchard industries in
temperate climates. Clearly, clover-based cover crops have the potential to benefit both poplar
and chestnut plantations either through indirect or direct means of N transfer. In the eastern part

120

of the United States, hybrid chestnut trees have provided a potential comeback for a tree that has
been nearly wiped out by disease. Michigan, for instance, has more chestnut plantation area than
another state in the U.S and a growing chestnut market 151. The chestnut-clover alley cropping
system described in this research has potential to offer a low-input form of production for
landowners. At the small scale, this provides an opportunity for landowners to diversify their
production and become more resilient to socio-economic stresses. At the large scale, this could
aid the growth of a burgeoning industry while providing multiple ecosystem services.

Additionally, biofuels derived from hybrid poplars represent a burgeoning market for low-input
production systems in temperate and tropical climates. Despite the ongoing debate over their
carbon neutrality, the importance of short-rotation woody crops (SRWC’s) intended for biofuels
is growing. The Kyoto Protocol of the UNFCCC stressed the importance of agroforestry and
short-rotation woody crops to providing ecosystems services such as carbon sequestration while
remaining productive 132. Maintaining these systems without further artificial nitrogen inputs is
critical considering the high carbon cost of producing them and the GHG emissions and runoff
risks involved during application 2. This research, as well as others, suggests the potential for
utilizing nitrogen-fixers as a replacement for such fertilizers 20. Moreover, the utilization of
direct N transfer could encourage “no-till” agricultural practices which could reduce emissions of
nitrous oxide and carbon dioxide when applied at a wide scale. Lastly, the relatively higher
biodiversity, soil carbon additions, soil microbial activity, and efficient water use from these
systems can be highly beneficial to production and ecosystem protection.

121

What is perhaps the most intriguing implication of this research is that both of these species,
when incorporated with clover into agroforestry systems could be effectively grown on infertile
and marginal lands. In fact, since N transfer tends to increase under competitive conditions, this
could be the best use for herbaceous legume-tree systems such as these. Since 38% of global
land area is currently considered degraded and that number is likely to grow, this could have
potentially massive implications for production systems. Inputs to the soil from clover both
below and above ground can help to restore soil nutrients on a site while remaining productive.
This could ultimately result in healthier soils by the end of one or several production cycles. At
a wide scale this could mean increased carbon sequestration and soil protection per unit of land
area. Alley cropping systems on marginal lands also provide opportunities for producing other
products, such as livestock. Gamble et al. (2014) studied short-rotation willow and poplar
plantations alley cropped with forage grasses in Minnesota to determine carbon sequestration and
biomass production potential. They found that the poplar hybrid ‘NM6’ was capable of
providing substantial biomass potential while maintaining opportunities for livestock forage 134.
However, a problem posed by SRWC’s is that frequent harvest removes considerable nutrients
from an ecosystem, making them unsustainable without inputs. Here the potential for alley
cropping systems that focus on herbaceous species that could benefit woody species becomes
very apparent.

These systems could also be critical to protecting water sources when serving as riparian buffers
while also improving landowner income at the small scale. A considerable focus of recent
research has been on using SRWC species in riparian buffers, because of their dense root
systems that can capture leaching nutrients and runoff 64. Additionally, other potential

122

ecosystem services provided by these systems are numerous and would likely be paired with an
improved and diversified income for landowners. This would improve landowner resiliency to
climate change as well as future economic and political shifts, which would be particularly
important in tropical climates, where there is a considerable amount of degraded land and higher
dependency on agriculture 2. Many of these areas have already used agroforestry practices with
great success, largely because of their ability to enhance nutrient cycling and reduce the need for
inputs, which are expensive and often difficult to find. While the exact species combinations
used in these experiments are not necessarily feasible in tropical climates, the basic principles
regarding nitrogen transfer dynamics discovered in this research can be paralleled to these
ecosystems with great benefit.

Global consumption of synthetic nitrogenous fertilizer was estimated at 113,100,000 tonnes in
2013 and is expected to grow at 1.4-1.5% into the next decade 49. As mentioned before, there is
a disproportionate global access and use of fertilizer. Many of the landowners in developing
countries cannot get the fertilizers they need to improve production while industrial agriculture
systems in developed countries have greater access to fertilizer and over-use it to the point of
degrading the surrounding ecosystems. A critical element of this research is outlining the
potential for cover crops to replace the need for synthetic nitrogenous fertilizer. This can reduce
costs to the landowner and allow them to engage in more diverse production systems (such as
silvopasture) that would additionally improve landowner income.

Ultimately, effectively utilizing the nitrogen transfer mechanism can be beneficial to both
temperate and tropical landowners and help begin to address many of the issues mentioned in

123

Chapter 2 through the enhancement of agroecological feedback cycles. In both regions, barriers
to the adoption of agroforestry systems most often have to do with fear over direct and incurred
costs. The systems described by this research require minimal inputs and labor, have a high
ability to restore soil health, and are extremely adaptable to marginal and low-fertility lands.
These characteristics could make them more enticing to landowners that are hesitant to change
their agricultural practices.

124

6

Conclusion

The experiments conducted in this research provide further evidence for the existence of a direct
transfer mechanism for nitrogen between white clover (T. repens) and the two chosen tree
species (Castanea sativa x C. crenata (“Colossal”) and Populus nigra L.×P. maximowiczii A.
Henry (‘NM6’)). Relative estimated nitrogen contributions by the donor to the trees were 15.9%
of poplar N and 33.3% of chestnut N. The majority of this transfer occurred via root exudates.
By the end of the experiment, the amount of available N exudate absorbed by the trees was
ranged between 7-17.5%, depending on the species, indicating that exudates are a larger source
of soil N even than estimated by transfer. Disparity in transfer values between tree species was
likely due to varying rooting structures, but the chestnut trees appeared to be more adept to direct
N transfer. Ultimately the estimates under controlled greenhouse condition may not accurately
reflect transfer over the same time period in the field and could be overestimated. However,
since direct N transfer typically increases in the field over time, these estimates may actually
represent long-term N transfer in field settings. Furthermore, the field study proved that the
presence of red clover (T. pretense) in hybrid poplar plantations (Populus nigra L.×P.
maximowiczii A. Henry (‘NM6’)) significantly improved poplar leaf and soil nitrogen
concentrations after just two years at wider spacings. The benefits seen in clover treatments
were not reflected in poplar growth. However, previous research suggests that positive effects
on growth would likely be seen in ongoing seasons. Spacing treatments in this study showed
that narrow row gaps incurred greater signs of competition, while wider row spacing showed
signs of beneficial nitrogen transfer.

125

The results of these studies reveal a high potential for clover species (T. pretense and T. repens)
to benefit nitrogen cycling and relationships in tree plantations. Longer and more in-depth field
studies are needed to determine how nitrogen transfer changes over the course of a full
production cycle and the relative importance of indirect and direct transfer pathways. Ultimately
these cover crops can present a cheap and low-labor solution to maintaining and restoring soil
fertility in long-term productive agroforestry systems, the implications of which could be great
for both the environment and landowners.

126

BIBLIOGRAPHY

127

BIBLIOGRAPHY

1. FAO. The State of the World’s Land and Water Resources for Food and Agriculture:
Managing Systems at risk. 1st ed. Milton Park, Abingdon; 2011.
2. Osman KT. Soil degradation, conservation and remediation. 2014. 1-237 p.
3. Morton JF. The impact of climate change on smallholder and subsistence agriculture.
Proceedings of the National Academy of Sciences of the United States of America.
2007;104(50):19680–5.
4. Hertel TW, Rosch SD. Climate change, agriculture, and poverty. Applied Economic
Perspectives and Policy. 2010;32(3):355–385.
5. Rosenzweig C, Tubiello FN. Developing Climate Change Impacts and Adaptation Metrics for
Agriculture Produced for the Global Forum on Sustainable Development on the Economic
Benefits of Climate Change Policies. The Integrated Assessment Journal. 2008;8 (1)(July):165–
184.
6. Nair PKR. Agroforestry -The Future of Global Land Use Advances in Agroforestry. 2012. 541
p.
7. Jose S. Agroforestry for ecosystem services and environmental benefits: An overview.
Agroforestry Systems. 2009;76(1):1–10.
8. FAO. Agroforestry. 2016 [accessed 2016 Nov 13].
http://www.fao.org/forestry/agroforestry/en/
9. Hirota I, Sakuratani T, Sato T, Higuchi H, Nawata E. A split-root apparatus for examining the
effects of hydraulic lift by trees on the water status of neighbouring crops. Agroforestry Systems.
2004;60(2):181–187.
10. Kang BT. Alley cropping—soil productivity and nutrient recycling. Forest Ecology and
Management. 1997;91(1):75–82.
11. FAO. Technical paper 2: Biological nitrogen fixation. FAO Corporate Document Repository.
[accessed 2017 Jan 16]. http://www.fao.org/wairdocs/ilri/x5546e/x5546e05.htm
12. Munroe JW, Isaac ME. N2-fixing trees and the transfer of fixed-N for sustainable
agroforestry: a review. Agronomy for Sustainable Development. 2013;34(2):417–427.
13. Thilakarathna MS, McElroy MS, Chapagain T, Papadopoulos YA, Raizada MN.
Belowground nitrogen transfer from legumes to non-legumes under managed herbaceous
cropping systems. A review. Agronomy for Sustainable Development. 2016;36(4).
128

14. Roggy JC, Moiroud a., Lensi R, Domenach a. M. Estimating N transfers between N2-fixing
actinorhizal species and the non-N2-fixing Prunus avium under partially controlled conditions.
Biology and Fertility of Soils. 2004;39:312–319.
15. Jalonen R, Nygren P, Sierra J. Transfer of nitrogen from a tropical legume tree to an
associated fodder grass via root exudation and common mycelial networks. Plant, cell &
environment. 2009;32(10):1366–76.
16. Sierra J, Daudin D, Domenach a. M, Nygren P, Desfontaines L. Nitrogen transfer from a
legume tree to the associated grass estimated by the isotopic signature of tree root exudates: A
comparison of the 15N leaf feeding and natural 15N abundance methods. European Journal of
Agronomy. 2007;27(2–4):178–186.
17. Høgh-Jensen H, Schjoerring JK. Below-ground nitrogen transfer between different grassland
species: Direct quantification by 15N leaf feeding compared with indirect dilution of soil 15N.
Plant and Soil. 2000;227:171–183.
18. Potter P, Ramankutty N, Bennett EM, Donner SD. Characterizing the spatial patterns of
global fertilizer application and manure production. Earth Interactions. 2010;14(2).
19. Van Sambeek J., Ponder F, Rietveld W. Legumes increase growth and alter foliar nutrient
levels of black walnut saplings. Forest Ecology and Management. 1986 [accessed 2016 Sep
7];17(2–3):159–167. http://linkinghub.elsevier.com/retrieve/pii/037811278690109X
20. Waring, H.D; Snowdon P. Clover and urea as sources of nitrogen fore the establishment of
Pinus radiata. Australian Forest Research. 1985;15:115–121.
21. Dupraz C, Simorte V, Dauzat M, Bertoni G, Bernadac A, Masson P. Growth and nitrogen
status of young walnuts as affected by intercropped legumes in a Mediterranean climate.
Agroforestry Systems. 1998 [accessed 2016 Sep 7];43(1/3):71–80.
http://link.springer.com/10.1023/A:1026499103899
22. Kremer RJ, Kussman RD. Soil quality in a pecan-kura clover alley cropping system in the
Midwestern USA. Agroforestry Systems. 2011;83(2):213–223.
23. Garrity D. Agroforestry and the Future of Global Land Use. In: Nair PKR, Garrity D, editors.
Agroforestry - The Future of Global Land Use. Nairobi, Kenya: World Agroforestry Centre
(ICRAF); 2012. p. 21–31.
24. Wall DH. Sustaining biodiversity and ecosystem services in soils and sediments. Washington
D.C.: Island Press; 2004. 1-12 p.
25. Lorenz K, Lal R. Soil organic carbon sequestration in agroforestry systems. A review.
Agronomy for Sustainable Development. 2014;34(2):443–454.
26. Kedziora A. The Network of Shelterbelts As An Agroforestry System Controlling the Water
Resources and Biodiversity in the Agricultural Landscape. Papers on Global Change IGBP.
2015;22(1):63–82.
129

27. Garrity DP. Agroforestry and the achievement of the Millennium Development Goals. The
journal of Agroforestry systems. 2004;61(1–3):5–17.
28. Lal R. Restoring soil quality to mitigate soil degradation. Sustainability (Switzerland).
2015;7(5):5875–5895.
29. Zomer RJ, Neufeldt H, Xu J, Ahrends A, Bossio D, Trabucco A, van Noordwijk M, Wang
M, Bondeau A, Feddema J, et al. Global Tree Cover and Biomass Carbon on Agricultural Land:
The contribution of agroforestry to global and national carbon budgets. Scientific Reports.
2016;6(April):29987.
30. National Agroforestry Center. Alley Cropping: an Agroforestry Practice. 2014.
31. Idassi JO. Profitable Farms and Woodlands: A Practical Guide in Agroforestry for
Landowners, Farmers and Ranchers. 2012. 108 p. http://nac.unl.edu/profitable_farms.htm
32. Nair PK. Classification of agroforesty systems. 1993. 21-38 p.
33. National Agroforestry Center U. A brief overview of Agroforestry in the United States. 2014.
34. WFP, IFAD, FAO. The State of Food Insecurity in the World, 2013: The Multiple
Dimensions of Food Security. Rome, Italy; 2013.
35. Van der Ploeg JD. Poverty Alleviation and Smallholder Agriculture: The Rural Poverty
Report 2011. Development and Change. 2012;43(1):439–448.
36. FSC. Strategic Review on the Future of Forest Plantations. 2012:121.
37. United Nations. World population projected to reach 9.6 billion by 2050 | UN DESA | United
Nations Department of Economic and Social Affairs. United Nations - Department of Economic
and Social Affairs. 2013 [accessed 2016 Nov 14].
http://www.un.org/en/development/desa/news/population/un-report-world-population-projectedto-reach-9-6-billion-by-2050.html
38. Bindraban PS, van der Velde M, Ye L, van den Berg M, Materechera S, Kiba DI, Tamene L,
RagnarsdĂłttir KV, Jongschaap R, Hoogmoed M, et al. Assessing the impact of soil degradation
on food production. Current Opinion in Environmental Sustainability. 2012;4(5):478–488.
39. FAO, CAB. Planted Forests: Uses, Impacts, and Sustainability. 1st ed. Evans J, editor. Rome:
Food and Agriculture Organization of the United Nations; 2009.
40. Bauhus J, Van Der Meer PJ, Kanninen M. Ecosystem Goods and Services from Plantation
Forests. 2011. 249-250 p.
41. Franzel S. Symposium 4: Agroforestry Systems in Africa. In: World Congress of
Agroforestry. Nairobi, Kenya; 2009.
42. Sileshi G, Akinnifesi FK, Ajayi OC, Place F. Meta-analysis of maize yield response to
130

woody and herbaceous legumes in sub-Saharan Africa. Plant and Soil. 2008;307(1–2):1–19.
43. Thorlakson T, Neufeldt H. Reducing subsistence farmers’ vulnerability to climate change:
evaluating the potential contributions of agroforestry in western Kenya. 2012. 15 p.
44. Quinion A, Chirwa PW, Akinnifesi FK, Ajayi OC. Do agroforestry technologies improve the
livelihoods of the resource poor farmers? Evidence from Kasungu and Machinga districts of
Malawi. Agroforestry Systems. 2010;80(3):457–465.
45. Lasco RD, Delfino RJP, Catacutan DC, Simelton ES, Wilson DM. Climate risk adaptation by
smallholder farmers: The roles of trees and agroforestry. Current Opinion in Environmental
Sustainability. 2014;6(1):83–88.
46. Challinor A, Wheeler T, Garforth C, Craufurd P, Kassam A. Assessing the vulnerability of
food crop systems in Africa to climate change. Climatic Change. 2007;83(3):381–399.
47. Islam KK, Hoogstra M, Ullah MO, Sato N. Economic contribution of participatory
agroforestry program to poverty alleviation: A case from Sal forests, Bangladesh. Journal of
Forestry Research. 2012;23(2):323–332.
48. Rahman SA, Imam MH, Snelder DJ, Sunderland T. Agroforestry for Livelihood Security in
Agrarian Landscapes of the Padma Floodplain in Bangladesh. Small-scale Forestry.
2012;11(4):529–538.
49. FAO. World fertilizer trends and outlook to 2018. 2015. 66p p.
50. Shorette K. Outcomes of global environmentalism: Longitudinal and cross-national trends in
chemical fertilizer and pesticide use. Social Forces. 2012;91(1):299–325.
51. Tan ZX, Lal R, Wiebe KD. Global Soil Nutrient Depletion and Yield Reduction. Journal of
Sustainable Agriculture. 2005;26(1):123–146.
52. Zhu ZL, Chen DL. Nitrogen fertilizer use in China - Contributions to food production,
impacts on the environment and best management strategies. Nutrient Cycling in
Agroecosystems. 2002;63(2–3):117–127.
53. Lehman RM, Cambardella CA, Stott DE, Acosta-Martinez V, Manter DK, Buyer JS, Maul
JE, Smith JL, Collins HP, Halvorson JJ, et al. Understanding and enhancing soil biological
health: The solution for reversing soil degradation. 2015. 988-1027 p.
54. National Oceanic and Atmospheric Administration. NOAA’s National Ocean Service
Education: Estuaries. 2008 [accessed 2016 Nov 18].
http://oceanservice.noaa.gov/education/kits/estuaries/media/supp_estuar09b_eutro.html
55. Schaffner M, Bader HP, Scheidegger R. Modeling the contribution of point sources and nonpoint sources to Thachin River water pollution. Science of the Total Environment.
2009;407(17):4902–4915.

131

56. Snyder CS, Bruulsema TW, Jensen TL, Fixen PE. Review of greenhouse gas emissions from
crop production systems and fertilizer management effects. Agriculture, Ecosystems and
Environment. 2009;133(3–4):247–266.
57. IPCC. IPCC Guidelines for National Greenhouse Gas Inventories Volume - IV Agriculture,
Forestry and other Land use. Institute of Global Environmental Strategies (IGES), Hayama,
Japan. 2006:1–5.
58. Evans J. Forest Plantations. Forests and Forest Plants. 2009;I.
59. FAO. Protecting plantations from pests and diseases - Report based on the work of W.M.
Ciesla. 2001;(March):19.
60. FAO, Asia-Pacific Forestry Commission. Asia-Pacific Forestry Sector Outlook Study II:
Southeast Asia Subregional Report - Part 3. 2011.
61. Iverson AL, MarĂ­n LE, Ennis KK, Gonthier DJ, Connor-Barrie BT, Remfert JL, Cardinale
BJ, Perfecto I. Do polycultures promote win-wins or trade-offs in agricultural ecosystem
services? A meta-analysis. Journal of Applied Ecology. 2014;51(6):1593–1602.
62. PumariĂąo L, Sileshi GW, Gripenberg S, Kaartinen R, Barrios E, Muchane MN, Midega C,
Jonsson M. Effects of agroforestry on pest, disease and weed control: A meta-analysis. Basic and
Applied Ecology. 2015;16(7):573–582.
63. Karp DS, Mendenhall CD, SandĂ­ RF, Chaumont N, Ehrlich PR, Hadly EA, Daily GC. Forest
bolsters bird abundance, pest control and coffee yield. Ecology Letters. 2013;16(11):1339–1347.
64. Udawatta RP, Jose S. Agroforestry strategies to sequester carbon in temperate North
America. Agroforestry Systems. 2012;86(2):225–242.
65. Pisupati B, Warner E. Biodiversity and the Millennium Development Goals. 2003:1–50.
66. Jones PG, Thornton PK. The potential impacts of climate change on maize production in
Africa and Latin America in 2055. Global Environmental Change. 2003;13(1):51–59.
67. BajĹželj B, Richards K. The Positive Feedback Loop between the Impacts of Climate Change
and Agricultural Expansion and Relocation. Land. 2014;3(3):898–916.
68. FAO. Forestry for a low-carbon future. Rome: Food and Agriculture Organization of the
United Nations; 2016.
69. Verchot L V., Van Noordwijk M, Kandji S, Tomich T, Ong C, Albrecht A, Mackensen J,
Bantilan C, Anupama K V., Palm C. Climate change: Linking adaptation and mitigation through
agroforestry. Mitigation and Adaptation Strategies for Global Change. 2007;12(5):901–918.
70. Dixon RK. Agroforestry systems: sources of sinks of greenhouse gases? Agroforestry
Systems. 1995;31(2):99–116.

132

71. Keller M, Varner R, Dias JD, Silva H, Crill P, de Oliveira Jr RC, Asner GP. Soil–atmosphere
exchange of nitrous oxide, nitric oxide, methane, and carbon dioxide in logged and undisturbed
forest in the Tapajós National Forest, Brazil. Earth Interactions. 2005;9(23):1–28.
72. Lal R. Climate Change and Soil Degradation Mitigation by Sustainable Management of Soils
and Other Natural Resources. Agricultural Research. 2012;1(3):199–212.
73. Montgomery DR. Dirt: The Erosion of Civilizations. Berkeley, CA: University of California
Press; 2012.
74. Eriksson J, Hakansson I, Danfors B. The effect of soil compaction on soil strucutre and crop
yields. Swedish Institute of Agricultural Engineering. 1974;354.
75. Charreau C. Problmes Poses Par L’utilization Agricole Des Sols Tropicaux Par Des Cultures
Annuelles. Agronomie Tropicale. 1972;27:905–929.
76. Hicks DH, Anthony T. Soil Conservation Technical Handbook. 2001. 205 p.
77. Lal R. Soil degradation by erosion. Land Degradation & Development. 2001;539(November
2001):519–539.
78. Havlin J. Soil fertility and fertilizers: an introduction to nutrient management. Upper Saddle
River, New Jersey: Person Prentice Hall; 2005.
79. Drechsel P, Gyiele L, Kunze D, Cofie O. Population density, soil nutrient depletion, and
economic growth in sub-Saharan Africa. Ecological Economics. 2001;38(2):251–258.
80. Upjohn B, Fenton G, Conyers M. Soil Acidity and Liming. Soil Science. 2005;105(4):283.
81. Slattery JF, Coventry DR, Slattery WJ. Rhizobial ecology as affected by the soil
environment. Australian Journal of Experimental Agriculture. 2001;41(3):289–298.
82. Rey Benayas JM, Newton AC, Diaz A, Bullock JM. Ecosystem Services by Ecological
Restoration: A Meta-Analysis. Science. 2009;325(August):1121–1124.
83. Pimentel D, Wilson C, McCullum C, Huan R, Dwen P, Flack J, Tran Q, Saltman T, Cliff B.
Economic and environmental benefits of biodiversity. BioScience 47. 1997;11:747–757.
84. Kladivko EJ. Tillage systems and soil ecology. Soil and Tillage Research. 2001;61(1–2):61–
76.
85. Gomiero T. Soil degradation, land scarcity and food security: Reviewing a complex
challenge. Sustainability (Switzerland). 2016;8(3):1–41.
86. LeĂłn JD, Osorio NW. Role of litter turnover in soil quality in tropical degraded lands of
Colombia. The Scientific World Journal. 2014;2014.
87. Saha R, Ghosh PK, Mishra VK, Majumdar B, Tomar JMS. Can agroforestry be a resource
conservation tool to maintain soil health in the fragile ecosystem of north-east India? Outlook on
133

Agriculture. 2010;39(3):191–196.
88. Bailey VL, Smith JL, Bolton H. Fungal-to-bacterial ratios in soils investigated for enhanced
C sequestration. Soil Biology and Biochemistry. 2002;34(7):997–1007.
89. Davinic M, Moore-Kucera J, Acosta-MartĂ­nez V, Zak J, Allen V. Soil fungal distribution and
functionality as affected by grazing and vegetation components of integrated crop-livestock
agroecosystems. Applied Soil Ecology. 2013;66:61–70.
90. Araujo ASF, Leite LFC, De Freitas Iwata B, De Andrade Lira M, Xavier GR, Do Vale
Barreto Figueiredo M. Microbiological process in agroforestry systems. A review. Agronomy for
Sustainable Development. 2012;32(1):215–226.
91. Bauhus J, Schmerbeck J. Silvicultural options to enhance and use forest plantation
biodiverssity. In: Ecosystem Goods and Services from Plantation Forests. Earthscan; 2011.
92. Schwab N, Schickhoff U, Fischer E. Transition to agroforestry significantly improves soil
quality: A case study in the central mid-hills of Nepal. Agriculture, Ecosystems and
Environment. 2015;205:57–69.
93. Kaczan D, Arslan A, Lipper L. Climate Smart Agriculture? A review of current practice of
agroforestry and conservation agriculture in Malawi and Zambia. 2013;(13):pp 1-62.
94. Mitchell RJ, Campbell CD, Chapman SJ, Cameron CM. The ecological engineering impact
of a single tree species on the soil microbial community. Journal of Ecology. 2010;98(1):50–61.
95. Schroth G. A review of belowground interactions in agroforestry, focussing on mechanisms
and management options. Agroforestry Systems. 1999;43:5–34.
96. FernĂĄndez ME, Gyenge J, Licata J, Schlichter T, Bond BJ. Belowground interactions for
water between trees and grasses in a temperate semiarid agroforestry system. Agroforestry
Systems. 2008;74(2):185–197.
97. Ong CK, Corlett JE, Singh RP, Black CR. Above and below ground interactions in
agroforestry systems. Forest Ecology and Management. 1991;45(1–4):45–57.
98. Jose S, Gillespie AR, Pallardy SG. Interspecific interactions in temperate agroforestry.
Agroforestry Systems. 2004;61:237–255.
99. Bayala J, Heng LK, van Noordwijk M, Ouedraogo SJ. Hydraulic redistribution study in two
native tree species of agroforestry parklands of West African dry savanna. Acta Oecologica.
2008;34(3):370–378.
100. Burgess SSO, Adams MA, Turner NC, Ward B. Characterisation of hydrogen isotope
profiles in an agroforestry system: Implications for tracing water sources of trees. Agricultural
Water Management. 2000;45(3):229–241.
101. Udawatta RP, Kremer RJ, Nelson KA, Jose S, Bardhan S. Soil quality of a mature alley
134

cropping agroforestry system in temperate North America. Communications in Soil Science and
Plant Analysis. 2014;45(19):2539–2551.
102. Bardhan S, Jose S, Udawatta RP, Fritschi F. Microbial community diversity in a 21-yearold temperate alley cropping system. Agroforestry Systems. 2013;87(5):1031–1041.
103. Hauser S. Distribution and activity of earthworms and contribution to nutrient recycling in
alley cropping. Biology and Fertility of Soils. 1993;15(1):16–20.
104. Nii-Annang S, GrĂźnewald H, Freese D, HĂźttl RF, Dilly O. Microbial activity, organic C
accumulation and 13C abundance in soils under alley cropping systems after 9 years of
recultivation of quaternary deposits. Biology and Fertility of Soils. 2009;45(5):531–538.
105. Kang BT, Caveness FE, Tian G, Kolawole GO. Longterm alley cropping with four
hedgerow species on an Alfisol in southwestern Nigeria - Effect on crop performance, soil
chemical properties and nematode population. Nutrient Cycling in Agroecosystems.
1999;54(2):145–155.
106. Das Chagas Ferreira Aguiar A, Bicudo SJ, Sobrinho JRSC, Martins ALS, Coelho KP, de
Moura EG. Nutrient recycling and physical indicators of an alley cropping system in a sandy
loam soil in the pre-Amazon region of Brazil. Nutrient Cycling in Agroecosystems.
2010;86(2):189–198.
107. Kanninen M. Plantation Forests: Global Perspectives. In: Ecosystem Goods and Services
from Plantation Forests. EarthScan; 2011.
108. Kort J, Poppy L, Gordon A, Caron L. Temperate agroforestry: When trees and crops get
together. Agriculture, Ecosystems and Environment. 2009;131(1–2):1–3.
109. Tassin J, Rangan H, Kull CA. Hybrid improved tree fallows: Harnessing invasive woody
legumes for agroforestry. Agroforestry Systems. 2012;84(3):417–428.
110. Snapp SS, Swinton SM, Labarta R, Mutch D, Black JR, Leep R, Nyiraneza J, O’Neil K.
Evaluating Cover Crops for Benefits, Costs, and Performance within Cropping System Niches.
2003;(November).
111. Ledgard S., Steele KW. Biological nitrogen fixation in mixed legume / grass pastures.
1Plant and Soil. 1992;141(1):137–153.
112. Lin, Y.; Wilson, A.R.; Nzokou P. Combining Legume Groundcovers and Low Nitrogen
Fertilization in a Short-rotation Fraser fir (Abies fraseri) Cropping System: Effect on Prouctivity,
Soil Fertility, and Nutrient Leaching. HortScience. 2011;46(3):1–6.
113. Badri D V., Vivanco JM. Regulation and function of root exudates. Plant, Cell and
Environment. 2009;32(6):666–681.
114. Walker TS, Bais HP, Grotewold E, Vivanco JM. Update on Root Exudation and
Rhizosphere Biology Root Exudation and Rhizosphere Biology. 2003;132(May):44–51.
135

115. Haichar F el Z, Santaella C, Heulin T, Achouak W. Root exudates mediated interactions
belowground. Soil Biology and Biochemistry. 2014;77:69–80.
116. Ta TC, MacDowell FDH, Faris MA. Excretion of nitrogen assimilated from N2 fixed by
nodulated roots of alfalfa (Medicago sativa)’. 1986;(20).
117. Nasholm T, Knut K, Ganeteg U. Uptake of organic nitrogen by plants. Blackwell
Publishing Ltd; 2009.
118. Arcand MM, Knight JD, Farrell RE. Estimating belowground nitrogen inputs of pea and
canola and their contribution to soil inorganic N pools using 15N labeling. Plant and Soil.
2013;371(1–2):67–80.
119. Parniske M. Arbuscular mycorrhiza: the mother of plant root endosymbioses. Nature
Review Microbiology. 2008;6(10).
120. He X-H, Critchley C, Bledsoe C. Nitrogen Transfer Within and Between Plants Through
Common Mycorrhizal Networks ( CMNs ). Critical Reviews in Plant Sciences. 2003;22(6):531–
567.
121. He X, Xu M, Qiu GY, Zhou J. Use of 15N stable isotope to quantify nitrogen transfer
between mycorrhizal plants. Journal of Plant Ecology. 2009;2(3):107–118.
122. Frey B, Schueppe H. Transfer of Symbiotically Fixed Nitrogen from Berseem ( Trifolium
alexandrinum L .) to Maize via Vesicular-Arbuscular Mycorrhizal Hyphae. 2008;122(3):447–
454.
123. Johansen A, Jensen ES. Transfer of N and P from intact or decomposing roots of pea to
barley interconnected by an arbuscular mycorrhizal fungus. Soil Biology and Biochemistry.
1996;28(1):73–81.
124. Haystead A, Malacjzuk N, Grove TS. Underground transfer of nitrogen between pasture
plants infected with vesicula arbuscular mycorrhizal fungi. New Phytologist. 1988;108(4):417–
423.
125. FAO. Planted Forests. Food and Agriculture Organization of the United Nations. 2012
[accessed 2016 Nov 22]. http://www.fao.org/forestry/plantedforests/en/
126. ITTO. Encouraging industrial forest plantations in the tropics - Report of a Global Study.
2009. 143 p.
127. Garforth M, Landell-Mills N, Mayers J. Plantations, Privatisation, Poverty and Power Changing Ownership and Management of State Forests. London: Earthscan; 2005. 3-21 p.
128. Kröger M, Kröger M. Global tree plantation expansion : a review Global tree plantation
expansion : a review. 2012;(3).
129. Ceotto E, Candilo M Di. Sustainable Bioenergy Production, Land, and Nitrogen use. In:
136

Biodiversity, Biofuels, Agroforestry, and Conservation Agriculture. Springer Science; 2011. p.
101–123.
130. Gelfand I, Sahajpal R, Zhang X, Izaurralde RC, Gross KG, Robertson PG. Sustainable
bioenergy production from marginal lands in the U.S. Midwest. Letter Research. 2013;493.
131. Gruenewald H, Brandt BK V, Schneider BU, Bens O, Kendzia G, HĂźttl RF. Agroforestry
systems for the production of woody biomass for energy transformation purposes. Ecological
Engineering. 2007;29(4):319–328.
132. Calfapietra C, Gielen B, Karnosky D, Ceulemans R, Scarascia Mugnozza G. Response and
potential of agroforestry crops under global change. Environmental Pollution.
2010;158(4):1095–1104.
133. Wilson AR, Nzokou P, Cregg B. Ground covers in fraser fir (Abies fraseri [pursh] poir.)
production systems: Effects on soil fertility, tree morphology and foliar nutrient status. European
Journal of Horticultural Science. 2010;75(6):269–277.
134. Gamble JD, Johnson G, Sheaffer CC, Current DA, Wyse DL. Establishment and early
productivity of perennial biomass alley cropping systems in Minnesota, USA. Agroforestry
Systems. 2014;88(1):75–85.
135. Lamerre J, Schwarz KU, Langhof M, von WĂźhlisch G, Greef JM. Productivity of poplar
short rotation coppice in an alley-cropping agroforestry system. Agroforestry Systems.
2015;89(5):933–942.
136. Pirhofer-Walzl K, Rasmussen J, Høgh-Jensen H, Eriksen J, Søegaard K, Rasmussen J.
Nitrogen transfer from forage legumes to nine neighbouring plants in a multi-species grassland.
Plant and Soil. 2012;350(1–2):71–84.
137. Miller RO. DEVELOPING AN ALGORITHM TO PREDICT SINGLE-TREE BIOMASS
WEIGHT FROM STEM DIAMETER MEASUREMENTS IN YOUNG HYBRID POPLAR
ENERGY PLANTATIONS IN MICHIGAN . 2016;2016(July):4–6.
138. R Core Team. R: A Language and Environment for Statistical Computing. 2016.
139. Purwin C, Sienkiewicz S, Pysera B, Lipiński K, Fijałkowska M, Piwczyński D, Puzio N.
Nitrogen fractions and amino acid content in alfalfa and red clover immediately after cutting and
after wilting in the field. Journal of Elemntology. 2012:723–734.
140. Penn State University. Hairy Vetch as a Crop Cover — Crops and Soils — Penn State
Extension. PSU Extension Services. 2016 [accessed 2017 Jan 25].
http://extension.psu.edu/plants/crops/soil-management/cover-crops/hairy-vetch-as-a-cover-crop
141. Lesuffleur F, Salon C, Jeudy C, Cliquet JB. Use of a 15N2 labelling technique to estimate
exudation by white clover and transfer to companion ryegrass of symbiotically fixed N. Plant
and Soil. 2013;369(1–2):187–197.

137

142. Oberson a., Frossard E, Bßhlmann C, Mayer J, Mäder P, Lßscher a. Nitrogen fixation and
transfer in grass-clover leys under organic and conventional cropping systems. Plant and Soil.
2013;371(1–2):237–255.
143. Sierra J, Nygren P. Transfer of N fixed by a legume tree to the associated grass in a tropical
silvopastoral system. Soil Biology and Biochemistry. 2006;38:1893–1903.
144. St. John L, Ogle D. Plant Guide - White Clover. 1998. 5-8 p.
145. Putz B, Drapela T, Wanek W, Schmidt O, Frank T, Zaller JG. A simple method for in situlabelling with 15N and 13C of grassland plant species by foliar brushing. Methods in Ecology
and Evolution. 2011;2(3):326–332.
146. Jalonen R, Nygren P, Sierra J. Transfer of nitrogen from a tropical legume tree to an
associated fodder grass via root exudation and common mycelial networks. Plant, Cell and
Environment. 2009;32(10):1366–1376.
147. Hertenberger G, Wanek W. Evaluation of methods to measure differential 15N labeling of
soil and root N pools for studies of root exudation. Rapid Communications in Mass
Spectrometry. 2004;18(20):2415–2425.
148. University of California - Davis. Stable Isotope Facility. 2016 [accessed 2016 Sep 16].
http://stableisotopefacility.ucdavis.edu/13cand15n.html
149. Sturite I, Henriksen TM, Breland TA. Longevity of White clover (Trifolium repens) leaves,
stolons and roots, and consequences for nitrogen dynamics under northern temperate climatic
conditions. Annals of Botany. 2007;100(1):33–40.
150. Van Sambeek JW, Ponder F, Rietveld WJ. Legumes increase growth and alter foliar
nutrient levels of black walnut saplings. Forest Ecology and Management. 1986;17(2–3):159–
167.
151. Bricault B. Chestnuts: Growing a food crop and an industry in Michigan | MSU Extension.
MSU Extension Services. 2013 [accessed 2017 Feb 15].
http://msue.anr.msu.edu/news/chestnuts_growing_a_food_crop_and_an_industry_in_michigan

138