 

 

.‘shu‘n...

tgqblnl 5,”.
)i. 1%}. 3.13.7.
c‘l
:r .. .5...L
u.
4:.
3.3%:
”ﬁnch . 3‘93!
1. 3 .
.halLﬁ—ﬁlcmﬁtkri
rant-Io!
Jam. ..
it.
. , ‘ Lb 0.‘ ‘
3:}. .. fun. i.

 

THESIS

//

c’\

A

~43

«u

 

LIBP .- nv lllllllllllllllllllllllllllllllllllllllllllllllll

. 23 01766 3042
M.Chl§ ..t 3 ate
University

 

 

 

This is to certify that the

dissertation entitled

Contributions of Roots and Organic Matter
to Soil Aggregate Stabilization
presented by

Djail Santos

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degree in Crop & Soil Sciences

 

 

We

Maj r professor 7

 

Dr. Alvin J.M. Smucker
Date October 23, 1998

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

”M11?!”
' «4'5 ‘1’?
this-3 it 100'
m?

 

 

 

 

 

3&305102304

 

 

 

 

 

 

 

 

 

 

 

 

 

m W14

CONTRIBUTIONS OF ROOTS AND ORGANIC MATTER TO SOIL AGGREGATE
STABILIZATION

By

Djail Santos

A DISSERTATION
Submitted to
Michigan State University
in partial ﬁilﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1998

ABSTRACT

CONTRIBUTIONS OF ROOTS AND ORGANIC MATTER TO SOIL AGGREGATE
STABILIZATION

By

Djail Santos

Soil aggregates and associated porosities control many of the below ground
biological, chemical and physical processes and strongly inﬂuence plant root activities.
This study examines three speciﬁc objectives which address the inter-relationships
between soil aggregates and the retention and spatial distributions of carbon (C) contents
within individual aggregates. First, a ﬁeld study was sampled to assess crop rotation and
conventional- (CT) and no-tillage (NT) modiﬁcations of total C contents and the
concomitant incorporation of 13C into soil macroaggregates. Accumulations of C were
greatest in aggregates sampled at the 0-5 cm depth Analysis of l3C-natural abundance
indicated that 65% of the C in the outer layers of CT aggregates (-26.5960) originated from
alfalfa roots (513C of -27.9%o), while 25% of the NT aggregate C, of the outer layer,
originated ﬁom the alfalfa Aggregate samples taken at 5-15 cm depth had less negative
6'3C values which were similar to the bulk soil. 6'3C values in the inner regions of soil
aggregates were similar at both depths. The second ﬁeld study was designed to investigate
the distribution of original or native C3- and more recent or contemporary C3- and C4-

derived C in aggregates having diﬂ‘erent sized-fractions. The pr0portion of more recent

crop-derived C was 53 to 76% higher in NT soil aggregates at the 0-5 cm depth.
Depositions of crop-derived C in CT soils ranged from 22 to 45% and appeared to not be
affected by soil depth Aggregate stabilities were lower in CT and NT soils than in never-
tilled grassland soils. Never-tilled soils contained proportionally greater quantities of
water-stable aggregates among ﬁactions > 0.5 mm, whereas CT and NT soils contained
more stable aggregates having ﬁ'actions < 0.5 mm. On average for both soil depths, the
mean weight diameter of aggregates (MWD), < 6.5 mm, was 62% lower in CT soil than in
the never-tilled grassland soil. However, following 11 years of NT, the MWD of NT soil
aggregates, were improved by 40% over those ﬁ'om CT. Thirdly, a greenhouse study was
designed to investigate amounts and distributions of recently released root C in soil
aggregates by ryegrass plants under water stressed and non-stressed conditions. Water
availability during a six week growth period signiﬁcantly changed ryegrass root
parameters to depths of 10 cm. Higher C contents were detected in aggregates associated
with non-stressed plants. New inputs of C3-C into whole aggregates, at 0-5 cm depth,
increased by as much as 62% when ryegrass was subjected to drought stress and 25% in
the non-stressed treatment. Aggregates associated with non-stressed ryegrass roots, at the
0-5 cm depth, developed higher MWDs than aggregates from unplanted treatments. Soil
aggregates associated with non-stressed ryegrass roots also had greater MWDs than
aggregates ﬁ'om ryegrass roots under water stress. The soil abrasion techniques for
peeling concentric layers ﬁom aggregates, described in these studies, appear to be useful
for quantifying the sequestration of recently released root C by soil aggregates and
provide a novel approach for further elucidating soil processes associated with the

formation, stability, and ﬁmction of soil aggregates.

To Silvanda and Lucas -

for your love and the joy

you bring me

iv

ACKNOWLEDGMENTS

I wish to express my sincere appreciation and gratitude to my rmjor advisor,
Dr. Alvin Smucker. His friendship, enthusiastic support, and energetic guidance were a
constant throughout my studies at MSU. The opportunity to share his ideas and working
methods has provided me with skills and knowledge that go beyond the degree obtained.
I extend my appreciation to the members of my guidance committee: Dr. George
Merva, Dr. Delbert Mokma, and Dr. Eldor Paul for their time and suggestions to improve
this work. I also thank Dr. Francis Pierce, Dr. Mokma, and Dr. Paul Rieke for providing
laboratory facilities for some of the analyses; Dr. Renate Snyder for her criticism and
valuable suggestions on scientiﬁc writing; and Dr. Eunice Poster for her sincere interest.
The technical expertise of John Ferguson with the TDR and the aggregate tensile
strength equipments, and David Harris and Sven Bohn with the stable isotope analyses, is
gratefully acknowledged. Many thanks to Doug Badgero, Brian Graﬂ‘, Jerry HardweIL,
and Greg Parker for their help and assistance in many instances. I am grateful to Daniel
Rasse for his ﬁiendship and for proofreading portions of this manuscript and providing a
sincere and valuable criticism. The assistance of Jennifer Wagester, Vera Bakic, Sim Loo,
Chad Lynema, Sotaro Oshima, and Helga Scares in several phases of the work is
gratefully appreciated. Their ﬁ'iendship and willingness to help will always be remembered.
I thank my colleagues from the Departament of Agriculture at Universidade
Federal da Paraiba (UFPB) for their support. Among those, I am truly indebted to my

friend Prof. Arinaldo Frazao for his encouragement, unconditional support, and for

representing my interests. A special thanks goes to Carlos and Rita Queiroga for their
genuine ﬁ'iendship and for sharing their time attending many requests.

A special thanks to José Linn and family, for their ﬁiendship and benevolent help
in getting started at MSU. My thanks to many friends from the Brazilian community at
MSU, especially my fellow students in the Departament of Crop and Soil Sciences: Jose’
Lima, Luiz Roberto Guilherme, José Cora, Carlos Paglis, Jorge Rodrigues, Maeli Melloto,
Carlos Alberto Silva, and Dilson Bisognin. Many thanks to my colleagues especially Ralph
DiCosty, Yasemin Kavdir, Tom Mueller, James Sallee and Fagaye Sissoko.

The sponsorship of Conselho Nacional de Desenvolvimento Cientiﬁco e
Tecnolégico (CNPq) and UFPB, made possible the completion of this degree and is
sincerely appreciated. The support provided by the Centro de Formacao de Tecnologos
(CFT) at UFPB, through its directors Prof. Genival Alves de Azeredo (1993-1997) and
Dr. Roberto Germano Costa (1997-present), is gratefully acknowledged.

This research was ﬁmded in part by a USDA Grant (Project no. MICL01884) and
a LTER Graduate Student Grant Research The ﬁnancial support provided by the USDA
grant allowed the completion of the ﬁnal requirements to obtain this degree.

I wish to express my thanks to the Centre for International Cooperation of the
State of Israel (Mashav) and Dr. Reuven Steinhardt (V olcani Center) for providing me
with a scholarship and the opportunity to attend the advanced international course on soil
management and irrigation of drylands. The ﬁnancial support provided through the efforts
of Dr. Donald Isleib (former director of the Inst. of International Agriculture-MSU) and

Dr. Roberto G. Costa (CFT-UFPB), allowed me to attend this training in Israel.

TABLE OF CONTENTS

LIST OF TABLES ................................................... ix
LIST OF FIGURES ................................................... xi
INTRODUCTION .................................................... 1

CHAPTER 1. UNIFORM SEPARATION OF CONCENTRIC SURFACE
LAYERS FROM SOIL AGGREGATES

INTRODUCTION ................................................... 12
MATERIALS AND METHODS ......................................... 13

Equipment description and operation ................................ 13

Soil sample preparation and erosion protocol .......................... 14
RESULTS AND DISCUSSION ......................................... 14
CONCLUSIONS ..................................................... 16
REFERENCES ...................................................... 16

CHAPTER 2. DISTRIBUTION OF ALFALFA ROOT-DERIVED ORGANIC
CARBON IN SOIL AGGREGATES IN TILLED AND NON-

TILLED ROTATIONS
INTRODUCTION ................................................... 17
MATERIALS AND METHODS ......................................... 21
Study site ..................................................... 21
Soil sampling and preparation ...................................... 22
Soil aggregate fractionation ....................................... 23
Aggregate bulk density .......................................... 26
Tensile strength of aggregates ..................................... 26
Total carbon and nitrogen, and carbon isotope ratios .................... 27
Statistical analyses .............................................. 28
RESULTS AND DISCUSSION ......................................... 29
Evaluation of the peeling method ................................... 29
Aggregate bulk density ........................................... 31
Tensile strength of soil aggregates .................................. 32
Total carbon and nitrogen ........................................ 32
Carbon isotope ratios ............................................ 38
CONCLUSIONS ..................................................... 43
REFERENCES ...................................................... 44

CHAPTER 3. MODIFICATIONS OF DISTRIBUTION AND COMPOSITION OF
ORGANIC MATTER IN WATER-STABLE AGGREGATES OF
UNDISTURBED AND CULTIVATED AGROECOSYSTEMS

INTRODUCTION ................................................... 60
MATERIALS AND METHODS ......................................... 63
Study site ..................................................... 63
Sample collection ............................................... 64
Aggregate size distribution and stability .............................. 64
Total carbon and stable isotope analyses .............................. 66
RESULTS AND DISCUSSION ......................................... 68
Aggregate size distribution and stability .............................. 68
Total carbon contents in aggregate size ﬁactions ....................... 73
6'3C values in aggregate size fractions ............................... 79
CONCLUSIONS ..................................................... 83
REFERENCES ...................................................... 84

CHAPTER 4. PLANT ROOT EFFECTS ON SOIL AGGREGATE CARBON
CONTENTS AND PHYSICAL PROPERTIES

INTRODUCTION .................................................. 98
MATERIALS AND METHODS ........................................ 101
Soil ........................................................ 101
Soil colunms (construction and packing) ............................ 101
Planting ..................................................... 102
Water regimes ................................................ 103
Plant, soil and root sampling ..................................... 104
Experimental design ............................................ 109
RESULTS AND DISCUSSION ........................................ 110
Volumetric water content ........................................ 110
Root and shoot dry weight ....................................... 110
Root parameters ............................................... 111
Carbon isotope variations in plant tissues ............................ 113
Soil C and 613C ratio ........................................... 113
Mean weight diameter and aggregate tensile strength ................... 117
CONCLUSIONS .................................................... 1 19
REFERENCES ..................................................... 120
GENERAL CONCLUSIONS .......................................... 139

LIST OF TABLES

Table 1.1. Organic C and exchangeable C and Mg concentrations by two methods for
separating the external and internal layers of soil aggregates ﬁoMtwo soils in Michigan. 15

Table 1.2. Exchangeable cation concentrations in the external and transitional layers and
the internal zones of aggregates from one soil type from Michigan and one from Germany,
separated by the soil aggregate erosion (SAE) method (n = 10). ................. 15

Table 1.3. Tensile strengths of natural and soil aggregate erosion (SAE)-processed soil
aggregates for one soil type ﬁom Michgan and three ﬁom Germany (n = 20). ....... 16

Table 2.1. Peeling time, layer thickness, aggregate surface ﬁ'action (ASF, as mass % of
the whole aggregates), and losses associated with the peeling procedure of outer and
transitional layers of soil aggregates (12.5-16 mm dia) ﬁ'om conventional tillage (CT), no-
tillage (NT), and never-tilled successional grass (NeTSG) treatments at 0-5 and 5-15 cm
depths. Means 3: standard error; n = 4 ...................................... 50

Table 2.2. Bulk densities and tensile strengths of air-dry soil aggregates (12.5-16 mm
dia.) from conventional tillage (CT), no-tillage (NT), and never-tilled successional grass
(NeTSG) treatments at 0-5 and 5-15 cm depths. Means i standard error ........... 51

Table 2.3. Comparison of regression coefﬁcients (a, b, and c) of the calibration equations
of the form (ASF = a + bt + cWi) for aggregate surface ﬁaction (ASF), and the variables
time (t) and aggregate initial weight (Wi) for soil aggregates ﬁom conventional tillage
(CT), no-tillage (NT), and never-tilled successional grass (N eTSG) treatments at 0-5 and
5-15 cm depths ...................................................... 52

Table 3.1. Mean weight diameter (MWD) under conventional tillage (CT), no-tillage
(NT), never-tilled successional grass (N eTSG), and never-tilled successional with maize
(NeTSM) at two sampling depths. (Means :1: standard error, n = 4) ............... 89

Table 4.1. Selected properties of aggregates (3-7 mm dia.) of the Kalamazoo loam soil
used in the greenhouse experiment. Kellogg Biological Station, Hickory Corners, MI.
1996. ............................................................. 124

Table 4.2. Dry weights of leaves, stems, shoots, and roots and total root/shoot ratio of
ryegrass plants grown in a sand (75%) and soil aggregate (25%) matrix for 6 weeks under
water stressed and non-stressed conditions (Means 3: standard error,

n = 8) ............................................................ 125

Table 4.3. Lengths of ﬁve different root width classes of ryegrass plants grown in a sand
(75%) and soil aggregate (25%) matrix for 6 weeks under water stressed and non-stressed
conditions. (Means i standard error; n = 8) ................................ 126

Table 4.4. Percentage of roots in individual root width classes of ryegrass plants grown in
a sand (75%) and soil aggregate (25%) matrix for 6 weeks under water stressed and non-
stressed conditions. (Means i standard error; n = 8) ......................... 127

Table 4.5. Total root dry weight (RDW), length (RL), surface area (RSA), volume (RV),

length density (RLD), and dry weight density (RDWD), of ryegrass plants grown in a sand
(75%) and soil aggregate (25%) matrix for 6 weeks under water stressed and non-stressed
conditions. (Means :h standard error; n = 8) ................................ 128

Table 4.6. 6'3C signatures of shoot (n = 8) and root (n = 4) parts of ryegrass plants grown
in a sand (75%) and soil aggregate (25%) matrix for 6 weeks under stressed and non-
stressed water conditions. (Means :1: standard error). ......................... 129

Appendix A1. Selected characteristics of bulk surface soil samples of conventional tillage
(CT), no-tillage (NT), and never-tilled successional grassland (NeTSG) of the
Agroecosystems Interactions site at WK. Kellogg Biological Station, Hickory Corners,
MI. .............................................................. 142

Appendix A2. Plot layout and sampled plots (shaded) of the Agroecosystems Interactions
site at WK. Kellogg Biological Station. Hickory Corners, MI. ................. 143

Appendix A3. Steps for the progressive and quantitative erosion of concentric layers
ﬁ'om soil aggregates by the mesoSAE chambers (Smucker, A.J.M, J. Wagester, and
D. Santos, unpublished). .............................................. 144

LIST OF FIGURES

Figure 1.1 The pneumatic soil aggregate erosion (SAE) apparatus. Aggregates are placed
in the sample chamber (E). An upward cyclone pattern of air ﬂow is established by passing
an air stream, generated by the motor (A), fan (B) and air-ﬂow controller (C) through the
air foil (D). Eroded soil components are carried by the air stream to the dust collector (F)
and accumulate in the sample bottle (G). The ﬁne-mesh screen (H) ﬁlters soil particles
ﬁ'om the exhaust air (I). Distinct layers can be removed from soil aggregates by stopping
the SAE, changing the sample bottle (G), and repeating the surface-abrasion process . . 13

Figure 1.2. Summation percentage curve, by weight, of a 60-min abrasion process for
removing soil particles ﬁ'om the external layer (ﬁrst linear portion of graph, 0-11min),
transitional layer (second linear portion of graph, 11-32 min), and internal region
(parabolic portion of graph, 32-60 min.) of soil aggregates from the Ap horizon of a
Conover loam (n = 3; 10 aggregates each) .................................. 14

Figure 2.1. Diagramatic representation of the improved soil aggregate erosion (SAE)
apparatus. Aggregates are placed in the removable aggregate chamber. A downward
cyclone pattern of air ﬂow is established by a blade connected to a motor and positioned
on the top of the chamber. Eroded soil particles are deposited in the sample collector and
accumulate in the sample bottle. As in the previous SAE model (Santos et al., 1997),
distinct layers can be separated ﬁom soil aggregates by stopping the SAE, changing the
sample bottle, and repeating the process ................................... 53

Figure 2.2. Total carbon contents of bulk soil samples, outer and transitioml layers,

and inner regions of soil aggregates from conventional tillage (CT) and no-tillage (NT),

at 0-5 and 5-15 cm depths. Values followed by the same uppercase letter within a tillage
treatment and between soil and aggregate samples, and by the same lowercase letter '
within soil and aggregate samples and between tillage treatments are not signiﬁcantly
different at p > 0.05 according to Fisher’s LSD mean separation test .............. 54

Figure 2.3. Total carbon (a), total nitrogen (b), C:N ratio (c), and 613C ratio (d) of bulk
soil samples, outer and transitional layers, and inner regions of soil aggregates ﬁom the
never-tilled successional grass (NeTSG) treatment, at 5-15 cm depth. Means :i: standard
error; n = 4 ......................................................... 55

Figure 2.4. 6‘3C ratios of bulk soil samples, outer and transitional layers, and inner

regions of soil aggregates from conventional tillage (CT) and no-tillage (NT), at 0-5 and
5-15 cm depths. Values followed by the same uppercase letter within a tillage treatment
and between soil and aggregate samples, and by the same lowercase letter within soil and
aggregate samples and between tillage treatments are not signiﬁcantly different at p > 0.05
according to Fisher’s LSD mean separation test .............................. 56

Figure 2.5. Total nitrogen of bulk soil samples, outer and transitional layers, and inner
regions of soil aggregates from conventional tillage (CT) and no-tillage (NT), at 0-5 and
5-15 cm depths. Values followed by the same uppercase letter within a tillage treatment
and between soil and aggregate samples, and by the same lowercase letter within soil and
aggregate samples and between tillage treatments are not signiﬁcantly different at p > 0.05
according to Fisher’s LSD mean separation test .............................. 57

Figure 2.6. C:N ratios of bulk soil samples, outer and transitional layers, and inner regions
of soil aggregates from conventional tillage (CT) and no-tillage (NT), at 0-5 and 5-15 cm
depths. Values followed by the same uppercase letter within a tillage treatment and
between soil and aggregate samples, and by the same lowercase letter within soil and
aggregate samples and between tillage treatments are not signiﬁcantly different at p > 0.05
according to Fisher’s LSD mean separation test .............................. 58

Figure 2.7. Percentage of new C3-derived carbon in bulk soil samples, outer and
transitional layers, and inner regions of soil aggregates from conventional tillage (CT) and
no-tillage (NT), at 0-5 and 5-15 cm depths. ................................. 59

Figure 3.1. Size distribution of intact water-stable aggregate fractions under never-tilled
successional grass (NeTSG), and never-tilled successional with maize (NeTSM) at 0-5 and
5-15 cm depths. Means :t standard error, n = 4 .............................. 90

Figure 3.2. Size distribution of intact water-stable aggregate fractions under conventional
tillage (CT), and no-tillage (NT) at 0-5 and 5-15 cm depths. Means :t standard error,
n = 4 ............................................................... 91

Figure 3.3. Total carbon contents (normalized to a sand-ﬂee basis) of water-stable
aggregate fractions under never-tilled successional grass (N eTSG), and never-tilled
successional with maize (NeTSM) at 0-5 and 5-15 cm depths. Means :i: standard error,

n = 4 .............................................................. 92

Figure 3.4. Total carbon contents (normalized to a sand-free basis) of water-stable
aggregate ﬁ‘actions under conventional tillage (CT), and no-tillage (NT) at 0-5 and 5-15
cm depths. Means :1: standard error, n = 4 .................................. 93

Figure 3.5. 6'3C ratios in water-stable aggregate size fractions under never-tilled
successional grass (NeTSG), and never-tilled successional with maize (NeTSM)at 0-5
and 5-15 cm depths. Means :t standard error, n = 4 ........................... 94

Figure 3.6. 6'3C ratios in water-stable aggregate size fractions under conventional tillage
(CT), and no-tillage (NT) at 0-5 and 5-15 cm depths. Means 3: standard error, n = 4 . . 95

Figure 3.7. Contents of total C from historical and crop sources (normalized to a sand-free
basis) in water-stable aggregate ﬁactions under conventional tillage (CT), no-tillage (NT),
and never-tilled successional with maize (NeTSM) at 0-5 and 5-15 cm depths. Means i

standard error, n = 4 .................................................. 96

Figure 3.8. Percent of the total carbon derived ﬁ'om historical and crop sources in water-
stable aggregate ﬁactions under conventional tillage (CT), no-tillage (NT), and never-
tilled successional with maize (NeTSM) at 0-5 and 5-15 cm depths. Means i standard
error, n = 4. ......................................................... 97

Figure 4.1 . Simpliﬁed representation of the system used for drip irrigation and
measurements of water contents at the depths of 10 and 26 cm in the PVC columns using
an automated and multiplexed time-domain reﬂectometry (TDR) equipment. ....... 130

Figure 4.2. Water characteristic curves for the growth medium (silica sand and soil
aggregates) and pure silica sand used in the greenhouse experiment .............. 131

Figure 4.3. Diagramrnatic representation of the apparatus used for automated
measurements of tensile strength of individual soil aggregates. .................. 132

Figure 4.4. Volumetric water contents in the control and ryegrass treatments, as a function
of water level applied (stressed and non-stressed plants), and position of the TDR probes
(10 and 26 cm depths). Mean values, n = 8. ................................ 133

Figure 4.5. Total carbon contents in whole soil aggregates, outer layers, and inner regions
of soil aggregates from the control and ryegrass treatments in a sand (75%) and soil
aggregate (25%) matrix after 6 weeks under water stressed and non-stressed conditions.
Means i standard error, n = 4. .......................................... 134

Figure 4.6. 513C ratios in whole soil aggregates, outer layers, and inner regions of soil
aggregates ﬁ'om the control and ryegrass treatments in a sand (75%) and soil aggregate
(25%) matrix aﬁer 6 weeks under water stressed and non-stressed conditions. Means :1:
standard error, n = 4. ................................................. 135

Figure 4.7. Percentage of new C3-carbon in whole soil aggregates, outer layers, and inner
regions of soil aggregates the control and ryegrass treatments in a sand (75%) and soil
aggregate (25%) matrix alter 6 weeks under water stressed and non-stressed conditions.
Means 3: standard error, 71 = 4. .......................................... 136

Figure 4.8. Mean weight diameter (MWD) of soil aggregate samples ﬁ'om the control and
ryegrass treatments in a sand (75%) and soil aggregate (25%) matrix after 6 weeks under
water stressed and non-stressed conditions. Means :t standard error, 11 = 4 ......... 137

Figure 4.9. Tensile strength of of soil aggregate samples ﬁ'om the control and ryegrass
treatments in a sand (75%) and soil aggregates (25%) matrix after 6 weeks under water
stressed and non-stressed conditions. Means :1: standard error, n = 4 ............. 138

xiv

INTRODUCTION

Carbon (C) and nitrogen (N) compounds that originate from soil organic rmtter
(SOM) are essential to the functioning of terrestrial ecosystems. Distribution of C and N
compounds within the soil are dependent on soil chemical, physical and biological
properties, root distribution and water inﬁltration (Stewart, 1987). Cycling of C and N
compounds from plant litter, root turnover, microbial decomposition and mineralization
occur within heterogeneous microbial habitats. Soils contain an interstitial mosaic of
microsites within a matrix of solid components and pore spaces where sequestration and
transformation of added chemicals and SOM take place (Bamforth, 1985).

Aggregation results when prinmry soil particles and microaggregates combine
intimately into secondary units of varying sizes and relative stability, cohesive forces and
cementing agents. The structural stability of soil aggregates and the pores between them
inﬂuence and sometimes control important soil processes such as plant emergence, water
and solute inﬁltration, soil aeration and drainage, erosion, plant root and microbiological
activities and crop productivity (Tisdall, 1996). Diﬁ‘erent mechanisms, operating for
diﬁ‘erent size classes of aggregates, were proposed by Tisdall and Oades (1982) and Oades
(1984). Macroaggregates greater than 250 pm in diameter, and specially those greater
than 2 mm in diameter, appear to be held together largely by ﬁne roots and ﬁmgal hyphae.
The stability of microaggregates with diameter of 20-250 um is usually greater than that
of aggregates with larger diameters. Binding agents for this size class appear to be rather

persistent. The stability of microaggregates in the diameter range of 2-20 ,um depends

upon the formation of cation bridges. Units smaller than 2 pm in diameter consist of clay
particles held together by inorganic and organic matter and electrostatic bonds (Oades,
1984; 1993).

Plant roots strongly affect aggregate stabilization. During plant growth, a portion
of root and soil derived C is constantly stabilized in soil through the production of new
microbial biomass. Different plant species and cultivars may signiﬁcantly inﬂuence soil
microbial activity and soil structure formation through variation in supply of available
root-derived C (Juma, 1994) and by a branched network formed by roots, which may
physically enmesh soil microaggregates (Tisdall and Oades, 1982). As soil aggregates are
being formed, the newly added organic C, solutes and particulate matter move through the
larger soil pores and deposit mainly on or near the surfaces of the aggregates. Successive
depositions result in the development of coating materials somewhat richer in those
components compared with concentric inner fractions (Ghadiri and Rose, 1991). The
conventional sieving methods used to separate different particles sizes within aggregates
have provided some information about the effect of soil structure on turnover of soil
organic matter. It may, however, redistribute the soil organic matter (Elliott and
Cambardella, 1991) giving results that sometimes are diﬂicult to interpret. Knowledge of
relationships between C gradients within soil aggregates and soil microbial biomass
activities is important to determine the mechanisms involved in soil aggregate
development and stability and the best approaches to effectively manage the soil for

sustaining intensive agricultural production systems.

In addition to rhizodeposition, large amounts of organic material are returned to
the soil after annual crops are harvested and residues are ploughed back (Haynes and
Francis, 1993). Such incorporation of organic residues may contribute to soil aggregate
formation indirectly by providing a substrate for microorganisms which, in turn, will
produce polysaccharide binding agents (Lynch and Bragg, 1985). The degree of soil
disturbance associated with the incorporation of these residues should inﬂuence the spatial
habitat for biological activity through effects on soil bulk density, pore continuity and size
distribution, and aggregation (Doran, 1987). Yet, limited information exists in the
literature on the effect of soil management, e. g., minimum-tillage or no-tillage, on the
nutrient contents of soil aggregates and their associated organic C (Carnbardella and
Elliott, 1993; Beare et al., 1994).

Aggregation impacts the kinetics of microbial decomposition of organic residues
and inﬂuences the cycling of C in soils. Kilbertus (1980) found that only 6% of the inner
volume of soil aggregates was colonized by microorganisms. Upon soil structure
disruption by cultivation, the organic matter previously less or none accessible to microbial
metabolism, e.g., within pores < 0.5 pm (Emerson et al., 1986) and within aggregates
(Robert and Chenu, 1995), becomes potentially more accessible for microbial activity.
However, Edwards and Bremner (1967) hypothesized that organic material inside
aggregates < 250 ram is physically inaccessible to microorganisms. This agrees with the
observation that microaggregates are not disintegrated with tillage (Tisdall and Oades,
1980; Elliott, 1986). These microhabitats provide protection against predation because the

pores are water-saturated most of the time and subjected to few hydric variations (Robert

and Chenu, 1995). Although survival of microorganisms may occur, the lack of available
02 becomes limiting, inducing anaerobic conditions that decreases their activity and, by
consequence, accessibility of substrates (Robert and Chenu, 1995). Carbon dioxide
production and microbial biomass C was found to be greatest in microaggregates (< 250
am) under aerobic conditions and biomass declined with increasing aggregate diameter
(Seech and Beauchamp, 1988). Extensive predation may occur in frequently unsaturated
large pores, e. g., in sandy soils or on the surfaces of aggregates, with drastic water
potential ﬂuctuations. These pores are essentially aerobic, and more substrates could be
accessible to microbes due to water ﬂow (Robert and Chenu, 1995). .
Conventional tillage practices break up root systems, disrupt soil macroaggregates,
expose more organic matter to microbial attack and decreases the ﬁmgal biomass. The
reduction in macroaggregation may provide a short-term increase in soil fertility through
the release of nutrients ﬁ'om the labile organic matter that binds micro- and
macroaggregates (Elliott, 1986). Prolonged cultivation, will, therefore, result in a rapid
decline in the number of macroaggregates and in labile organic matter (Waters and Oades,
1991) since the incorporation of plant residues results in a more homogeneous distribution
of substrate for the microbes (Coleman and Elliott, 1986). Organic matter may be
protected ﬁom microbial metabolism to some extent by adsorption to clay minerals
(Oades, 1984) and the formation of microaggregates (Edwards and Brenmer, 1967;
Gregorich et al., 1989) by isolation in micropores (Adu and Oades, 197 8) and by physical

protection within stable rmcroaggregates (Elliott, 1986).

A relatively recent tool used to study C dynamics in soils is the analysis of the
stable isotope ‘3C. This technique, also known as the natural abundance of 13C or 5‘3C, is
based on the fact that plants using C3 photosynthesis (Calvin-cycle) incorporate less 13C
than plants with a C4 (Hatch-Slack) photosynthetic cycle and, therefore, diﬂ‘er in their
discrimination between 12C and '3C isotopes (Smith and Epstein, 1971). These 5'3C
differences are transferred to SOM since the 6‘3C values in C3 plants are around -27%o,
while in C, plants they are about -14%o (Balesdent et al., 1987). This natural tracer of in
situ produced organic C has proved to be useful to study C sequestration into different
soil ﬁ'actions and aggregate sizes (V itorello et al., 1989; Martin et al., 1990; Angers and
Giroux, 1996). The 6‘3C technique has been used mainly in long-term studies.

The dynamics of C derived from plants with diﬁ‘erent metabolisms on soil
aggregate stability still need to be tested. Contributions of corn root-derived C to SOM in
bulk soil samples with C3 vegetation history vary with soil management and my not be
detectable at the end of the growing season (Balesdent and Balabane, 1992). After ﬁve
years of continuous corn on a previous tobacco-cultivated soil Ryan and Aravena (1994)
estimated that 55% of the soil microbial biomass under conventional till management was
derived ﬁ'om C4-C, while under no-till management the estimated value was 30%. As
organic C appears to be relatively uniform across aggregate size fractions, the level of
water-stable macroaggregation should provide an estimation of the amount of organic C
stored and physically protected in the soil, especially the labile ﬁ'actions (Angers and
Carter, 1996). Organic matter contained in the macroaggregates was shown to be younger

than that in the microaggregates (Skjemstad et al., 1990). They found, using the (WC

technique, that the microaggregates of an Australian soil contained older C than the
macroaggregates. In an other experiment, larger C contents in stable macroaggregates in a
soil with previous C3 vegetation, were attributed to the young C4-derived organic C
(Puget et aL, 1995). Carbon-l3 natural abundance, however, has not been fully applied to
the study of the spatial and temporal characteristics of C dynamics in the development and
stabilization of soil aggregates which, in turn, may help to elucidate some of these
questions on a short-time basis.

Information on the form and persistence of the aggregating constituents and the
extent over which they are operative is vital for understanding the reasons for the rapid
deterioration of the structure of some soils when continuous row cropping is introduced
(Baldock et al., 1987). This information will help design sustainable agricultural systems
that can decrease or reverse the trends in soil degradation associated with current
cropping practices. This study is an attempt to better understand the relationship between
soil aggregates and the retention and spatial distribution of C and N contents within them.
The ﬁrst objective of this study was to assess the inﬂuence of soil and plant managements
can exert on recent C deposition by contemporary root systems and its contribution to C
gradients within individual soil aggregates. A second objective was to investigate the
distribution of original C3-derived and contemporary C3 and C4-derived C in water-stable
aggregates under diﬁ‘erent tillage management systems and its interactions with aggregate
stability. It was intended to verify whether disturbance (e.g., conventional- vs. no-tillage)
inﬂuence the size distribution of water-stable aggregates and their C contents and

compositions, compared to undisturbed soils. Finally, the effects of contemporary

cropping of a C3 plant species on the amount, composition, and distribution of recently
released root C in soil aggregates under conditions of water-stress and non-stress was
investigated. Crop-water-soil interactions were investigated in terms of their inﬂuences on
the incorporation of C of recent deposition into soil aggregates and on aggregate physical

properties.

REFERENCES

Adu, J .K., and J .M. Oades. 1978. Physical factors inﬂuencing decomposition of organic
materials in soil aggregates. Soil Biol. Biochem. 10:109-115.

Angers, DA, and M. Giroux. 1996. Recently deposited organic matter in soil water-
stable aggregates. Soil Sci. Soc. Am. J. 60:1547-1551.

Angers, DA, and MR. Carter. 1996. Aggregation and organic matter storage in cool,
humid agricultural soils. p.193-211. In M.R. Carter and BA. Stewart (ed.)
Structure and organic matter storage in agricultural soils. Lewis, Boca Raton, FL.
Series: Advances in Soil Science.

Baldock, J.A., B.D. Kay, and M. Schnitzer. 1987. Inﬂuence of cropping treatments on the
monosaccharide content of the hydrolysates of a soil and its aggregate ﬁ'actions.
Can J.Soil Sci. 67:489-499.

Balesdent, J ., and M. Balabane. 1992. Corn root-derived soil organic carbon estirmted by
natural l3C abundance. Soil Biol. Biochem 24:97-101.

Balesdent, J., A. Mariotti, and B. Guillet. 1987. Natural l3C abundance as a tracer for
studies of soil organic matter dynamics. Soil Biol. Biochem. 19:25-30.

Bamforth, SS. 1985. Symposium on ‘Protozoan Ecology’: the role of protozoa in litters
and soils. J. Protozool. 32:404-409.

Beare, M.H., P.F. Hendrix, and DC. Coleman. 1994. Water-stable aggregates and organic
matter in conventional and no-tillage soils. Soil Sci. Soc. Am. J. 58:777-786.

Cambardella, CA, and ET. Elliott. 1993. Methods for physical separation and
characterization of soil organic matter fractions. Geoderma 56:449-457.

Coleman, DC, and ET. Elliott. 1986. Spatio-temporal heterogeneity of organic matter in
agroecosystems: effects on microbial/fauna] interactions. p. 140-148. In Trans.
International Congress of Soil Science. 13th, Hamburg. 1888, Wageningen, The
Netherlands.

Doran, J. 1987. Microbial biomass and mineralizable nitrogen distribution in no-tillage and
plowed soils. Biol. Fert. Soils 5:68-75.

Elliott, ET. 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native
and cultivated soils. Soil Sci. Soc. Am. J. 50:627-633.

Elliott, E.T., and CA. Cambardella. 1991. Physical separation of soil organic matter.
Agric. Ecosys. Environ. 34:407-419.

Edwards, AP, and J.M. Bremner. 1967. Microaggregates in soils. J. Soil Sci. 18:64-73.

Emerson, W.W., RC. Foster, and J .M. Oades. 1986. Organo-mineral complexes in
relation to soil aggregation and structure. p.521-548. In P.M. Huang and M.
Schnitzer (ed.) Interactions of soil minerals with natural organics and microbes.
SSSA Spec. Pub. 17. Madison, WI.

Ghadiri, H., and CW. Rose. 1991. Sorbed chemical transport in overland ﬂow: I. A
nutrient and pesticide enrichment mechanism. J. Environ. Qual. 20:628-633.

Gregorich, E.G., RG. Kachanoski, and RP. Voroney. 1989. Carbon mineralization in soil
size fractions after various amounts of aggregate disruption. J. Soil Sci. 40:649-
659.

Haynes, R.J., and GS. Francis. 1993. Changes in microbial biomass C, soil carbohydrates
composition and aggregate stability induced by growth of selected crop and forage
species under ﬁeld conditions. J. Soil Sci. 44:665-675.

Juma, N.G. 1994. A conceptual ﬁamework to link carbon and nitrogen cycling to soil
structure formation. Agric. Ecosys. Environ. 51:257-267.

Kilbertus, G. 1980. Etude des microhabitats contenus dans les agregats du sol: leur

relation avec la biomasse bacterienne et la taille des procaryotes presents. Rev.
Ecol. Biol Sol 17:543-557.

Lynch, J.M., and E. Bragg. 1985. Microorganisms and soil aggregate stability. Adv. Soil
Sci. 2:133-171.

Martin, A., A. Mariotti, J. Balesdent, P. Lavelle, and R. Vuattoux. 1990. Estimate of
organic matter turnover rate in a savanna soil by 13C natural abundance
measurements. Soil Biol. Biochem 22:517-523.

Oades, J.M. 1984. Soil organic matter and structural stability: mechanisms and
implications for management. Plant Soil 76:319-337.

Oades, J .M. 1993. The role of biology in the formation, stabilization, and degradation of
soil structure. Geoderma 56:377-400.

Puget, P., C. Chenu, and J. Balesdent. 1995. Total and young organic matter distributions
in aggregates of silty cultivated soils. Eur. J. Soil Sci. 46:449-459.

Robert, M., and C. Chenu. 1995. Water potential, soil microhabitats, and microbial
development. p.75-87. In P.M. Huang et al. (ed.) Environmental impact of soil

component interactions: natural and anthropogenic organics. Lewis, Boca Raton,
FL.

Ryan, M.C., and R. Aravena. 1994. Combining l3C natural abundance and fumigation-
extraction methods to investigate soil microbial biomass turnover. Soil Biol.
Biochem. 26:1583-1585.

Seech, A.G., and E.G. Beauchamp. 1988. Denitriﬁcation in soil aggregates of different
sizes. Soil Sci. Soc. Am. J. 52:1616-1621.

Skjemstad, 10., RP. Le Feuvre, and RE. Preble. 1990. Turnover of soil organic matter
under pasture as determined by 13C natural abundance. Aust. J. Soil Res. 28:267-
276.

Smith, B.N., and S. Epstein. 1971. Two categories of 13C/‘2C ratios for higher plants.
Plant Physiol. 47:380-384.

Stewart, J.W.B. 1987. Agronomic signiﬁcance of carbon, nitrogen, phosphorus, and
sulphur interrelationships in soils. INT ECOL Bulletin 1987:25-32.

Tisdall, J .M. 1996. Formation of soil aggregates and accumulation of soil organic matter.
p.57-96. In M.R. Carter and B.A. Stewart (ed.) Structure and organic matter
storage in agricultural soils. Lewis, Boca Raton, FL. Series: Advances in Soil
Science.

Tisdall, J.M., and J.M. Oades. 1980. The management of ryegrass to stabilize aggregates
of a red-brown earth Aust. J. Soil Res. 18:415-422.

Tisdall, J.M., and J .M. Oades. 1982. Organic matter and water-stable aggregates. J. Soil
Sci. 33:141-163.

Vitorello, V.A., C.C. Cerri, F. Andreux, C. Feller, and RL. Victoria. 1989. Organic
matter and natural carbon- 1 3 distribution in forested and cultivated oxisols. Soil
Sci. Soc. Am. J. 53:773-778.

Waters, A.G., and J .M. Oades. 1991. Organic matter in water-stable aggregates. p.163-

174. In W.S. Wilson (ed.) Advances in soil organic matter research; the impact on
agriculture and the environment. Royal Soc. Chem, Cambridge, UK.

10

Reprinted from the Sm! Science Smarty of America Journal
Volume 61. no. 3. May-June I997
677 South Segoe Rd.. Madison. WI 53711 USA

Uniform Separation of Concentric Surface Layers from Soil Aggregates
D. Santos. S. L. S. Murphy, H. Taubner. A. J. M. Smucker. " and R. Horn

11

Uniform Separation of Concentric Surface Layers from Soil Aggregates

D. Santos, S. L. S. Murphy, H. Taubner, A. I. M. Smucker. ‘ and R. Horn

ABSTRACT

ltbwellacceptedthatsoilorgaalcmatterlmpmeesoilstracture:
W.Wmetlodsﬂmitoarknowledgeoﬂhemaayphysleal—
ehedcal mechanisms that control the temporal processes of soil aggre-
gadoa. A soil aggregate erosion (SAE) apparatus was designed to
uniformly remove concentric surface layers of soil aggregates la a
eacloaederosioaehamber.‘l'heSAEwa-medtoteetforthepreaeaee
ofcooceau'atloagradleata withiaweaofdlﬂereatsoiltypee.
'l'heSAEmethodgeaerallylmprovedtherepm-atloaoﬁoaeoaceatra-
do-betweeadteeateraalaadbteraallayeraCoaeeauatioaaofC,
Cm-dMgwereltoDJotoZLIdntoSJ'klowernapeedvelyo
attheexteraallayersofaggrepteaforﬂaplaqaolmrdﬂapladaﬂ
“Removalofsarfaeelayemhadaoeleetoatheteaeﬂestreagth
dmﬂaggreptumnfmmeoadadedthatOeSAEappm-am
meathesepantioaoﬁoagadleataaithlaaatanlaollw
Immmumwmmmmw
propertieeotthialmprovedsepmadoaoleoaeeatrlclayemhomaoﬂ
WrathoaHIeadtonn-thaadvaaeemeatshoarhowledge
ofthe W meetaaiamaamodated with thedevelopmeat and

marital-greater.

Sort. AGGREGATE DEVELOPMENT has been reported to
result from the cementing of smaller aggregates into
larger aggregates containing various soil components
(Dexter et al., 1988). This assembly activity is inﬂuenced
by soil water, ions, minerals, organic matter, plant roots,
and microorganisms, which utilize numerous stabilizing
mechanisms (Lynch and Bragg. 1985). The structural
stability of soil aggregates and the pores between them
affect the inﬁltration. movement. and storage of water;
aeration; erosion; plant root and microbiological activi-
ties; and crop productivity (Tisdall, 1996). During the
formation of aggregates. many components of the soil
are redistributed within the structural unit, developing
gradients within individual soil aggregates. Shrinking
and swelling processes result not only in the well-known
accumulations of ﬁne textural separates but also in vari-
ations in chemical and physical properties within each
aggregate (Horn, 1990).

Horn et al. (1994) demonstrated external surface lay-
ers of aggregates contained more microorganisms than
internal regions. These layers also varied in their chemi-
cal and physical properties. Consequently, different
functions associated with these layered regions within
soil aggregates resulted from specific physical and chem-
ical gradients. Additionally. soil aggregation processes
depend upon the number of wetting and drying events
as well as on the intensity and frequency of shrinking
and swelling. lncrcased numbers of wetting and drying

 

D. Santos and A.J.M Smucker. Dep. of (‘mp and Soil Sciences. Michi-
gan State Univ.. East lensing. Ml 48824; S.LS Murphy. Dep. of
Environmental Science. Rutgers Univ . New Brunswick. NJ 08903:
H. Taubner and R. Horn. lust. for Plant Nutrition and Soil Science.
Olshausenstr. 40. 24118 Kiel. Germany. Received 19 Feb. 1996. ‘Cor-
responding author (lllJZSajsOmsuedul

Published in Soil Sci. Soc. Am. J. 61:720-724 (1907).

events cause more soil particles to be rearranged within
each aggregate. resulting in fewer changes in the hydrau-
lic properties and lower states of free entropy associated

.with soil aggregation (Horn et al., 1994).

Since plant roots are known to grow preferentially
in weak aggregates or in voids between aggregates
(McSweeney and Naussen. 1984; Dexter, 1986), certain
gradients may develop in the presence of roots. In addi-
tion to root growth, preferential flow of soil solutions,
gaseous exchange, microbial colonization, and solute
sequestration at aggregate surfaces enhance conditions
that further develop biological, chemical, and physical
gradients from the external layers to the internal regions
of soil aggregates (Horn et al., 1994). Methods used to
obtain separated layers of soil aggregates include man-
ual peeling (Qureshi and Jenkins, 1987; Horn, 1987;
Ghadiri and Rose, 1991; Priesack and Kisser-Priesack.
1993). dry sieving (Ghadiri and Rose, 1991), combined
freezing and wet sieving (Kayser et al., 1994; Wilcke et
al., 1996), electron probe microanalyses (Qureshi and
Jenkins, 1987), electron microscopy (Foster, 1994), ul-
traviolet photo-oxidation (Ladd et al., 1993), and selec-
tive staining of specific cells within resin-irnpregnated
soil sections (White et al., 1994). Manual peeling of soil
aggregates is a time-consuming and tedious process and
does not provide an opportunity for processing large
quantities of samples and is subject to individual errors.
Although electron microscopy has provided visual evi-
dence of many aggregate components (Foster, 1994),
sample preparation renders them unavailable for fur-
ther analyses. Recent attempts to investigate the inte-
riors of natural soil aggregates using laser confocal mi-
croscopy, by our laboratory, have not been conclusive
(H. Taubner, J. Whallon, and A. Smucker. 1994, unpub-
lished data). Therefore, new methods were tested for
eroding surface layers from soil aggregates. providing
samples for further analyses. A new method needed to
be developed that could uniformly produce separates
from different regions within soil aggregates. Quantifi-
cation of these separate components could enhance our
understanding of several cementing mechanisms associ-
ated with the dynamic soil aggregation responses of
different soil types to specific tillage and cropping rota-
tions. Concentric separation of aggregate layers will also
contribute to our knowledge of the fate and efﬁcacy of
surface-applied materials such as crop residues, pesti-
cide granules and sprays, weed seeds, or fertilizer gran-
ules that are controlled by the breakdown and refor-
mation of soil aggregates. A soil aggregate erosion
(SAE) device could also be used to study the mecha-
nisms of aggregate life expectancies. e.g.. length of time
between the formation and breakdown of soil aggre-
gates. as described by Staricka et al. (1992).

The objectives of this study were (i) to improve a
mechanized SAE device and compare its quantitative

12

SANTOS ET AL: SEPARATION OF CONCENTRIC LAYERS FROM SOIL AGGREGATES 721

efficiency with manual peeling and (ii) verify physical
and chemical concentration gradients within soil aggre-
gates from different soil types.

MATERIALS AND METHODS

Equipment Description and Operation

The SAE apparatus was developed by modifying a seed
cleaning machine (Agriculex, Inc., Guelph. Ontario. Canada).
using the principles proven by an earlier aggregate particle
erosion system reported by Schroeder and Smucker (1988).
The SAE apparatus (Fig.1) includes a squirrel-cage fan and
chamber (Fig. IB) and electric motor (Fig. 1A). A wind-con-
trol cover (Fig.1C) is attached at the air inlet. to manually
control air flow rates. A deflecting air foil (Fig. ID) is located
at the top of the fan chamber, placing it directly below the
base of the removable sample erosion chamber (Fig. 1E). The
airfoil (Fig. 1D) causes vertical flowing wind to be deﬂected
into a revolving cyclone pattern inside the removable sample
erosion chamber (Fig. 1E). Aggregates are retained in the
sample erosion chamber by ﬁne 1.0-mm (16 mesh) nylon
screens glued onto the base of the chamber and conﬁned in
the sample erosion chamber by a similar screen attached to
the lower end of the vertical wind duct. directly above the
removable sample erosion chamber. The interior of the remov-
able sample erosion chamber is lined with abrasive stainless
steel grating materials, assembled from commercially available
kitchen grating utensils. and slides into the removable sample
erosion chamber and is retained stationary by friction during
operation and could be removed from the erosion chamber
for cleaning.

Uniform soil aggregates. ranging in widths from 3 to 15
mm, are placed into the sample erosion chamber (Fig. 1E).
An upward cyclone pattern of air flow, controlled by the
. manually adjusted wind-control cover (Fig. 1C). suspends and
rotates each soil aggregate. Weak soil aggregates are identiﬁed
at the beginning of each erosive run by subjecting natural
aggregates to the SAE apparatus for 30 s. causing some to
split. Split aggregates are discarded. and many of the tips or
acute angles of the unbroken aggregates are abraded during
this abrasive pro-run treatment. Following a complete cleaning
of the SAE. by an air nonle. selected unbroken and uniform
aggregates are comprehensively eroded by exposure to the
cyclone air-flow pattern for the time periods required to re-
move layers of soil materials as concentric shells from the
aggregates. Uniform layers of soil material are eroded from
each soil aggregate as the air-suspended and misting aggre-
gates make contact with abrasive surfaces along the walls of
the sample erosion chamber (Fig. 1E). Removal of each sur-
face layer by the SAE apparatus results in uniform and time-
dependent smaller volumes of the remaining internal portions
of each aggregate.

Air containing eroded soil particles is transported to the
dust collector (Fig. 1F). The dust collector reduces the speed
of air flow. mulling in the deposition of soil particles into
the sample collector (Fig. 10) located at the base of the dust
collector. Air exits from the dust collector through a fine 0.20-
mm (130 mesh) screen via the air vent (Fig. lit) at the base
of the exhaust port (Fig. ll). At the end of each run. the dust
collector is gently tapped to remove soil particles that adhere
to the fine screen at the air exit or on the walls of the dust
collector. The entire SAE apparatus is further cleaned be-
tween samples by directing high-speed jets of compressed air
throughout the chamber.

Run times required to remove soil materials associated with
each region of the aggregates depend upon soil type and soil

K

 

 

 

 

 

 

 

 

Fig.1. ThepneamaricsolaggregateerosloatSAEhppm-ataaAgue-

gatesareplaeedhthesnplechambertﬁ).Aaapwardcycloae
ofairﬂowisestabﬂshedby ' aasirstreem.

bythemotor (A).faa (B),-ddr-llovveaatroler(C) through the
ﬁloﬂtDLErodedsolcompoaeatameem-riedbytheahstre-
tothedustcollector(F)aadacamalate'mthe—plebottie(6).
11elae-meshsu'eeaﬂnlntesasoipmriclesh’omtheeshaamﬁ
(ILDkdacthyemaaberemovedfs'omsoiagpegatesbystoppiag
“SAEch-p'agtheaampiehotﬂetcxmdrepeaﬂngthe-face-
“onprocess.

water content and are determined by weighing soil materials
that accumulate in the sample collector. The desired amount
of soil to be removed from each regional layer is controlled
by the duration of run time as presented by the summation
curve in Fig.2. Experience has demonstrated that the external
surface layers of air-dried soil aggregates erode more rapidly.
producing a linear relationship between run time and summa-
tion percentage for each soil type. Following the removal of
external layers of aggregates. the next region or zone of the
aggregate, which we refer to as the transitional layer. is re-
moved more slowly. producing a nonlinear relationship be-
tween time and the accumulation of eroded soil material. This
phenomenon appears to be the result of a combination of
smaller surface areas of progressively eroded aggregates and
more durable layers within the internal regions of aggregates.
The internal region of soil aggregates erodes even more slowly.
producing a parabolic relationship bemeen run time and sum-
mation curve (th. 2). More than twa concentric layers can
be removed from the surfaces of soil aggregates. as depicted

13

 

 

 

 

722
100 . ‘
b \
s r”
b '0 ’
0 ..........................
>
i
s oo-
.3
.9
0
3 40 r-
a
5 ........
a l
g E
a 20 '- 5
U) : _._ A
’ g 1 Mirror
1» :
o a It A l A; 1'_n J. A 1 a J
0 10 20 so 40 50 60 70
Time (minute)

Fig.2. Semmatioapereeatagecarve.byweight.ofa60-niaabrasion
procau for removing soil particles from the external layer (ﬁrst
hatporﬂoaofpaphb-llmimtdraasitioaallayeuseeondliaear
mamrr—u-suummmrm»
hdMW-hlefnﬂweshmtheApm
ofaCoaoverloa-(u - Jtlﬂaggngateaeaehl.

by the nine data points on the summation graph of Fig. 2.
The thickness of each layer can be estimated by the integral
equation describing the relationship between mass of clay and
the thickness of abraded zones as presented by Kay and Dexter
(1990). Their equation can be rearranged as

[1]

D. - xpCf

where D. is the amount (Mg m ") of eroded material extracted
from each concentric layer. x (tn). 9 is the bulk density (Mg
m") of the aggregate layer. C is the total clay content (Mg
Mg") of the eroded layer from the aggregate. and f is the
fraction of the total clay content from the eroded layer during
SAE run times from t. to :1.

SoﬂSamplePrepalationandI-Zrosioal’rotocol

Field soil samples from the An horizons of a Kalamazoo
loam (fine-loamy. mixed. mesic Typic Hapludalf) and a Col-
wood loam (ﬁne-loamy. mixed. mesic Typic Haplaquoll) were
extracted in larger volumes and transported to the laboratory
in foam packing materials. Large soil blocks were gently bro-
ken into smaller aggregates (widths <20 mm) and air dried.
Chemical compositions of manually peeled internal and exter-
nal layers of soil aggregates were compared with those me-
chanically eroded by the SAE apparatus. Ten aggregates. 3 to
5 mm in diameter. were selected and manually or mechanically
separated by the SAE method into external layers and internal
regions; transitional layers were not analyzed. Aggregates
were separated manually by peeling the same quantity of soil
(using a sharp knife and gloved hands). by weight. from each
aggregate as was eroded by the SAli method of separation.
Composite samples were extracted and combined until at least
5 g of soil materials were accumulated for chemical analyses.
Exchangeable Ca. Mg. K. and Na were extracted by shaking
to equilibrium in 1le soil to solution of l M NH.OAc or
NH.CI. Ion concentrations were assessed by atomic absorption
spectrophotometry. as described by l'lmrnas (I982). Organic

14

SOIL SCI. SOC. AM. 1.. VOL bl. MAY-JUNE I997

C was determined by dry combustion. as described by Nelson
and Sommers (1982). Where necessary. additional soil aggre-
gates were eroded from concentrically smaller soil aggregates
by the SAE method. until similar quantities of soil mass were
accumulated for each layer and region. thus minimizing analyt-
ical errors due to different sample sizes.

In a second experiment. tensile strengths of air-dried whole
or natural aggregates and machine-eroded aggregates con- _
taining only the internal region of each soil aggregate. i.e.. the
transitional layer and internal zone. were measured by the
crushing test of Dexter and Kroesbergen (1985) to determine
whether removal of surface layers would modify aggregate
stability. Samples from the B horizons of the following four
soils were used: a Conover loam (fine-loamy. mixed. mesic
Udollic Ochraqualf). located at the MSU Agronomy Farm.
East Lansing. MI; a Parabraunerde soil (T ypic Hapludalf),
located in Siggen. Germany: a Podsol soil (Haplorthod). lo-
cated in Oberwarmensteinach. Germany; and a Podsolige
Braunerde (Spodic Dystrochrept). located in Wulfersreuth.
Germany. Replicated soil samples for these measurements
were collected in cylinders (12 by 12 cm) at close proximity
to minimize horizontal heterogeneity. Following the removal
of compacted soil adjacent to the cylinder walls. soil volumes
were gently separated into natural aggregates of several size
classes. Twenty natural soil aggregates (13—15 mm diam.) and
twenty aggregates (8—10 mm diam.) whose external layers
had been eroded by the SAE apparatus were measured for
aggregate strength by individually crushing each between two
flat parallel plates. A constantly increasing vertical force was
applied to each aggregate by the top plate. which was con-
nected to a computer-controlled stepper motor. The bottom
ﬂat plate was the platform of an electronic digital balance.
which was connected to a computer. The computer continu-
ously recorded and graphed the applied forces on a monitor.
The graphic peak associated with the first major decline in
force. applied to the aggregate. was identified by solving for
the ﬁrst derivative of force vs. time. Resistance strength. in
kPa. for each aggregate (y). was calculated by the following

eqmﬁom
0.576F
y "' T [2]

where F is the crushing force required to fracture the aggre-
gate. F is determined by

F - bg [3]
where b is the measured balance reading (in kilograms) and
g is the aweleration force due to gravity (9.807 m s"). The
parameter d is the effective spherical diameter of each aggre-
gate and can be calculated as d - (Gm/rip)“, where m is the
mass (in kilograms) of the individual aggregate and p B the
mean aggregate bulk density. Bulk densities of the aggregates
were determined by the paraffin-sealing and water displace-
ment methods (Blake and Hartge. 1986).

RESULTS AND DISCUSSION

Soil layers can be uniformly abraded from individual
or groups of aggregates by eroding from 1 to 15 aggre-
gates in the SAE apparatus (Fig. l) for different periods
of time. Although the times for removing concentric
layers from soil aggregates are different for each soil
type. external layers are removed most easily. producing
a linear portion (e.g., 2.7% per 60 s) of the erosion
summation curve (Fig. 2). Consequently. external layers

smrros ET AL: scrxaartou or CONCENTRIC LAYERS FROM sou. AGGREGATES 723

Tablel. OrgaalcCaadexeh-gedleCandMgconeentratlonsbytwo”Mforseparatiagtheexteraalaadiatenallayersofaol

aggregatesfromtwosoilalaMleltigan.

 

 

 

 

 

 

 

 

 

Organic C Ca Mg

Sol Zone Manual SAE? M-aal SAE M-aal SAE
s is" and. Its“

Ink Hopi-quell Colwood) External 2“ 21.9 12.“ 10.35 2.19 1.70
‘ laser-a1 22.0 23.0 11.72 12.40 2.15 2.20

LSD (0.05) 1.0 1.30 0.34
Typic “aphid (Ki—mo) Easel-a1 17.! 17.2 4.27 4.22 0.47 0.45
late-II 10.3 10.0 4.34 5.. 0.49 050

LS!) (0.05) 2.2 0.39 0.06

 

iSAE-saiaggregaaasresiea.

of aggregates from Conover soils were designated as
the soil removed during the first linear portion (0-11
min) of the erosion summation curve (Fig. 2). The sec-
ond linear region of this erosion summation graph
(ll-32 min.) is designated to be the transitional layer
between the external layer and the internal volume of
aggregates associated with a Conover loam soil (Fig.
2). External and transitional layers of soil eroded from
aggregates by the SAE method and the remaining inter-
nal volumes or regions of aggregates (indicated by the
nonlinear region in the summation curve) were collected
separately and analyzed. In this way. the mass of each
concentric layer for a given soil was nearly uniform.
regardless of the aggregate number and size. It is ac-
knowledged that the uniformity of abraded soil col-
lected at the end of each run time is a function of original
aggregate homogeneity. The relatively small error bars
associated with aggregate erodibility in Fig. 2 suggest a
uniform sampling of soil aggregates. Therefore, the SAE
method appears to overcome much of the heterogeneity
among aggregates peeled manually.

Chemical gradients between the external layer and
internal regions of soil aggregates from two different
soil types in Michigan can be identified by the SAE
method of separation (Table 1). Higher C contents in
the manually separated external layer of the Typic
Haplaquoll soil may have resulted from the uneven cut-
ting or mixing, or both. of samples from the two layers
during manual removal. as C concentrations in both the
external and internal layers were significantly higher
and lower than the same layers separated by the SAE.
However. average C concentrations for both the exter-
nal and internal zones were not significantly different.
23.3 vs. 22.9 g kg". for both methods of separation.
Manual and SAE methods of layer removal demon-
strated very similar C concentrations in both zones

within the aggregates of the Typic Hapludalf soil type
(Table 1). These C deposits. which appear to remain
at or near the surfaces of formed aggregates. result in
somewhat richer coatings of organic C, compared with
inner fractions (Ghadiri and Rose. 1991; Puget et al..
1995). Therefore, differences in the C concentrations
for these two soils. at the external and internal regions
of aggregates. may have resulted from plant cultures
for each site at the time of sampling.

The SAE appeared to be a better method for separat-
ing aggregate layers. as concentrations of both Ca and
Mg were signiﬁcantly greater for internal regions of
aggregates for both soil types separated by this method
(Table 1). Greater differences in the concentration gra-
dients for Ca and Mg. which exist between external and
internal regions of aggregates. separated by the SAE
apparatus. suggest that this method improves the sepa-
ration of aggregate regions containing contrasting chem-
ical gradients. More accurate extraction by the me-
chanical SAE procedure may be due to less cross con-
tamination. a problem often associated with manual
methods of aggregate separation. Discrepancies be-
tween the two methods may have been caused by differ-
ential losses of the ﬁne particles associated with the two
methods of separating aggregate layers. Sample losses
by the SAE method were ~10% by weight. Most of these
loses resulted from either adhesion of fine particles to
the vertical plastic walls of the SAE by static electricity
or via the air exhaust. An improved model of the SAE.
made of metal and recycled air. is presently being de-
signed to overcome these losses.

Concentrations of Ca. Mg. and K within aggregates
of a Udollic Ochraqualf loam were signiﬁcantly higher
in the internal layers, when separated by the SAE
method. In contrast. gradients of K were higher for
external layers of the Typic Hapludalf soil (Table 2).

Table2. WWMMHMWMWhymmmEWWMWam-msﬁl
typefromMiehig-aadoaefromGermany.aeparatedby&esoi1a_g_gregateeroaioa(SAE)-ethod(n-10).

 

 

 

 

 

 

Sol Zena Ca Mg K Na
emol. kg"

Udole W (Conover) Extend 1.“ 0.14 0.05 0..
Tra‘tloaal 1.02 0.15 0405 0.10
later-a1 1.49 0.21 0.07 0..
LSD (.35) 0.25 0.05 0.02 NS?
Typic Haphdalf "Mel Eases-ad 7.0! 3.20 0.44 0.21
Transitional 7.20 3.27 0.“ 0.21
lasaaaal 7.00 3.00 0.35 0.22
1.30 (0.05) NS 0.25 0.04 NS

1 NS - not signiﬁcant.

15

724 sort. scr. soc. AM. 1.. v01. 61. MAY-JUNE I997

Tah1e3. Teasilestreagthsofaataralandsoilaggregateerosioa
(SAE)-proceed soil agpegates for one soil type from Michi-
gan and three from Germany in - 201.

 

 

 

Arms-te-
Sol (series) Natural SAE
kPa —-
Udale (Conover) 00 (32)? 80 (30)
Typic Hapludalf (Pardsraaaarde 009 (331) 997 (2.31)
W (Podaoll 144 (N1 140 (50)
Spade Dystrochrept (Padaalge Ira-aerial SS (46) 105 (56)

 

tNaderslapareashaaasmathast-darddeviatioaofthemeaas.

No signiﬁcant ion concentration gradients could be de-
tected between external and internal layers of soil aggre-
gates for the Haplorthod and Spodic Dystrochrept soils
from Germany (Table 2). Sodium concentration gradi-
ents were both low and not significant for any of the
soils tested. Variations in the degree of aggregation and
differences in the textural separates identiﬁed in aggre-
gate layers have been reported to cause higher concen-
trations of exchangeable cations in the internal regions
of soil aggregates (Horn. 1987).

Tensile or resistance strengths of soil aggregates from
four different natural and SAE-abraded soil types of this
study indicate that aggregates from the Typic Hapludalf
were substantially stronger than the other soil aggre-
gates (Table 3). Tensile strengths of the air-dry soil
aggregates in this study were highly variable and no
differences were observed between noneroded and SAE
treatments of the soil aggregates (Table 3). Therefore.
it appears that when artiﬁcial soil drying rates of aggre-
gates exceed in situ drying rates for corresponding sam-
pling depths. then the analyses of soil aggregates by any
method could provide unrealistic information. Intensive
drying pretreatment and other artificial processing of
soil aggregates may lead to little or no measurable differ-
ences in natural ion gradients.

CONCLUSIONS

The SAE apparatus provides a new approach for re-
moving concentric layers of particles from soil aggre-
gates of different soil types. The SAE can be used to
determine ion concentration gradients for natural condi-
tions or speciﬁc treatments. Extraction of the surface
layers of aggregates can also provide a novel approach
for preparing the internal surfaces of aggregates for
further analyses by microscopy and other techniques
without need for embedding soil volumes. Quantiﬁca-
tion of the chemical gradients and associated mecha-
nisms that control soil aggregate stabilization is essential
to improve management of soil structure and contribute
signiﬁcantly to our understanding of agricultural and
environmental soil physics.

ACKNOWLEDGMENTS

The CNPq Fellowship Program of Brazil to D5. is grate~
fully acknowledged.

16

REFERENCES

Blake. G. R.. and K. H. Hartge. I986. Bulk density. p. 363-375. In A.
Klute (ed )Methods of sorl analysts Part1 2nded. Agron. Monogr.
9. ASA and $SA. Madison. WI.

Dexter. A. R. 1986. Model experiments on the behavior of roots at
the interface between a tilled seed-bed and a compacted sub-soil.
I. Effect of seed-bed aggregate size and subsoil strength on wheat
roots. Plant Soil 95:123—133.

Dexter. A.R.. R. Horn. and W. Kemper. 1988. Two mechanisms of
age hardening. J. Soil Sci. 39:163-175.

Dexter. A.R.. and B. Kroesbergen. 198$. Methodology for determina-
tion of tensile strength of soil aggregates. J. Agric. Eng. Res.
31: 1395147.

Foster. R. C. 1994. Microorganisms and soil aggregates. p. 144-155.
In C.E. Pankhurst et al. (ed. )Soil biota. Management tn sustainable
farming systems. CSIRO. Victoria. Australia.

Ghadiri, H.. and CW. Rose. 1991. Sorbed chemical transport in over-
land flow: 1. A nutrient and pesticide enrichment mechanism. J.
Environ. Qual. 20:628—633.

Horn. R. 1987. Die Bedeutung der Aggregierung fur die Nahstoffsorp-
tion in Baden. Z. Pflanzenernahr. Bodenkd. 150:13-16.

Horn. R. 1990. Aggregate characterization as compared to soil bulk
properties. Soil Tillage Res. 17:265-289.

Horn. R.. W. Stepniewski. T. Wlodarczyk. G. Walenzik. and EEM.
Eckhardt. 1994. Denitrification rate and microbial distribution
with homogeneous soil aggregates. Int. Agrophys. 8:65-74.

Kay. B.D.. and A.R. Dexter. 1990. Influence of aggregate diameter.
surface area and antecedent water content on the dispersibility of
clay. Can. J. Soil Sci. 70:655-671.

Kayser. A.T.. W. Wilcke. M. Kaupenjohann. and JD. Joslin. 1994.
Small scale heterogeneity of soil chemical properties. I. A technique
for rapid aggregate fractionation. Z. Pflanzenernahr. Bodenkd.
157:453—458.

Ladd. J.N.. R.C. Foster. and 1.0. Skjemstad. 1993. Soil structure:
Carbon and nitrogen metabolism. Geoderma 56:401-434.

Lynch. J.M.. and E. Bragg. 1985. Microorganisms and soil aggregate
stability. Adv. Soil Sci. 2:133-171.

McSweeney. K.. and U. Naussen. 1984. Soil structure and associated
mating behavior in minerals. Soil Sci. Soc. Am. J. 48:607-612.
Nelson. D.W.. and LE. Sommers. l982. Total carbon. organic carbon.
and organic matter. p. 539-579. In A.L. Page et al. (ed.) Methods
of soil analysis. Part 2. 2nd ed. Agron. Monogr. 9. ASA and SSSA.

Madison. WI.

Priesack. E.. and G.M. Kisser-Priesack. 1993. Modeling diffusion and
microbial uptake of l’C~glucose in soil aggregates. Geoderma
56:561-573.

Puget. C.. C. Chenu. and J. Balesdent. 1995. Total and young organic
matter distributions in aggregates of silty cultivated soils European
.1. Soil Sci. 46:449-459.

Oureshi. R.R.. and DA. Jenkins. 1987. Concentration of phosphorus
and sulphur at soil ped surfaces. J. Soil Sci. 38:255-265.

Schroeder (Murphy). 5.1... and A.J.M. Smucker. 1988. Separation of
spherical layers from granular soil aggregates. p. 190. In Agronomy
abstracts. ASA. Madison. WI.

Starieka. J.A., R.R. Allmaras. W.W. Nelson. and WE. Larson. 1992.
Soil aggregate longevity as determined by the incorporation of
ceramic spheres. Soil Sci. Soc. Am. 1. 56215914597.

Thomas. G.W. 1982. Exchangeable cations. p. 159-165. In AL. Page
et al. (ed.) Methods of soil analysis. Part 2. 2nd ed. Agron. Monogr.
9. ASA and SSSA. Madison. WI.

Tisdall. J.M. 1996. Formation of soil aggregates and accumulation of
soil organic matter. p. 57-96. In M.R. Carter and B.A. Stewart
(ed.) Structure and organic matter storage in agricultural soils.
CRC Press. Boca Raton. FL.

White. D.. E.A. FitzPatrick. and K. Killham. 1994. Use of stained
bacterial inocula to assess spatial distribution after introduction
into soil. Geoderma 63:245-254.

Wilcke. W.. R. Bilumler. H. Desehauer. M. Kaupenjohann. and W.
Zech. 1996. Small scale distribution of Al. heavy metals. and PAl-Is
in an aggregate alpine Podzol. Geoderma 71:19-30.

CHAPTER 2
DISTRIBUTION OF ALFALF A ROOT-DERIV ED ORGANIC CARBON IN SOIL
AGGREGATES IN TILLED AND NON-TILLED ROTATION S

INTRODUCTION

Soil structure directly inﬂuences most physical, chemical and biological processes.
Rates of certain soil processes have been shown to be spatially distributed within
aggregates (Horn, 1990; Santos et al., 1997). These heterogeneously spaced
accumulations may inﬂuence root development and the availability of resources to plants.
Interaggregate voids, resulting ﬁ'om successive wetting and drying cycles or from the
active plant root growth, constitute the dynamic interface between plant roots, the soil
matrix, and associated soil solution. These voids are the focus of pedogenic activity and
tend to be planar and relatively permanent (Qureshi and Jenkins, 1987).

The composition of external surfaces of aggregates is generally different ﬁ'om that
of the bulk soil. Illuviation of clay, organic matter modiﬁcations, preferential leaching of
acidic solutions, and preferential weathering of aggregate surfaces are some of the
processes taking place at the surfaces of soil aggregates (Kayscr et al., 1994; Smucker et
al., 1996; Wilcke and Kaupenjohann, 1997; Santos et al., 1997). Lower concentrations of
organic C and total N in the center of the aggregates have been suggested to result from
prolonged anaerobic conditions in the inner zones (Dexter, 1988). Micro-environment
variations within and between aggregates inﬂuence SOM dynamics since the water content

may be greater within aggregates but oxygen is often less than between aggregates

l7

(Sextone et al., 1985). These important but here-to-fore neglected processes need to be
considered when designing sustainable soil management practices. Aggregate scale
gradients are not taken into account with bulk soil sampling, resulting in improper
assessment of plant root environments. Implications for plant growth and development are
important as roots will preferentially grow in the nmcropores separating aggregates (Heil
and Buntley, 1965; McSweeney and Jansen, 1984; Logsdon and Allmaras, 1991;
Passioura, 1991). Therefore, the root environment will be conﬁned to aggregate surfaces
rather than the bulk soil (Qureshi and Jenkins, 1987) and the zones of inﬂuence exerted by
roots and root hairs are reduced. Under these conditions, the interiors of soil aggregates
are not accessible to the roots, contributing minimal quantities of nutrients and soil water
(Logsdon and Allmaras, 1991; Passioura, 1991; Amato and Ritchie, 1995). Passioura
(1991) pointed out that, in such circumstances, the time needed for water and nutrients to
ﬂow to the roots may be such a large proportion of the life of a crop that they will not be
used during the growing season. Thus, certain gradients within aggregates may develop.
These gradients can be determined by sampling and analyzing individual or groups of
aggregates. Management practices that result in higher soil compaction are expected to
affect soil-plant interactions at the rhizosphere level and enhance the expression of these
gradients. As the density of each structural unit or ped increases, more roots are conﬁned
to ped surfaces and unable to extract water from the centers of compacted units (Smucker
and Allmaras, 1993).

Much research effort has been made to understand aggregate formation and

stability because of its inﬂuence on soil quality properties and agricultural productivity.

18

Methods used to assess the distribution of soil organic matter (SOM) relative to aggregate
structure, usually combine the impact of dry or wet sieving and abrasion of large masses of
soil. The sampling of surface layers of aggregates has the potential to provide information
on the mechanisms associated with the dynamic soil aggregates responses of different soil
types to speciﬁc tillage and cropping rotations. Santos et a1. (1997) reviewed the methods
used to obtain separated layers of soil aggregates and presented a soil aggregate erosion
(SAE) apparatus that provides quantitative removal of successive surface layers of soil
aggregates by mechanical abrasion. The method is of particular interest to study the
relationship between organic matter content and stable macroaggregates composition. The
extent of surfaces exposed is expected to be related not only to the amount of mechanical
energy applied to the sample but also to diﬂ‘erences in properties of the surfaces exposed
at different time intervals.

The resistance to abrasion of air-dry soil aggregates will be mainly related to clay
mineralogy and content, and nature and amount of cementing agents. There is evidence
that clay content and some chemical properties are different when surfaces and interiors of
aggregates are compared (Horn, 1990; Smucker et al., 1996; Santos et al., 1997). Thus,
the amount of soil material abraded from the surfaces associated with the weakest failure
zones (i.e., the original exterior of aggregates) will be different ﬁom that abraded from
surfaces associated with stronger failure zones (i.e., the interior of original aggregates)
which are exposed only when greater energy is applied (Kay and Dexter, 1990). Changes
in management practices may inﬂuence both exposed aggregate surfaces and its

erodibility. Conventional cultivation accelerates decomposition of roots, fungal hyphae,

l9

which stabilize aggregates larger than 0.25 mm (Oades, 1984).

Few studies have been conducted to determine the inﬂuence of root systems on
individual aggregates. A relatively new approach, based on the difference in l3C/‘ZC ratio
(6'3C) between plants with different photosynthetic pathways has been used to partitioning
soil organic carbon (SOC) according to its origins (Balesdent et al., 1987; Balesdent and
Balabane, 1996). Plants using the C4 photosynthetic pathway have higher 6‘3C (mean
values of -12%o) than C3 plants (-28%o). Delta 13C of SOM is assumed to correspond
closely to the 13C content of the plant nmterial from which it is derived. This technique
was used in the present study to determine the storage of SOM and turnover of plant
residue C in different spatial compartments of soil macroaggregates.

The objectives of this study were: i) to utilize an improved Soil Aggregate Erosion
(SAE) apparatus to mechanically remove external surface layers from groups of soil
aggregates, and ii) to investigate the eﬂect of crop rotations and tillage practices on SOM

content and isotopic composition (613C) across soil aggregates.

20

MATERIALS AND METHODS

Study Site

The soil samples were collected at the Agroecosystem Interactions experimental
site that is part of the Long Term Ecological Research (LTER) project at the WK.
Kellogg Biological Station, located in Hickory Corners, 15 miles northeast of Kalamazoo,
in southwestern Michigan (42°24’N; 85 °24’W). Mean annual temperature is 9°C and
precipitation averages 920 mm yr‘l (URL:http://Iter/kbs.msu.edu/description.html). The
experimental site contains Kalamazoo (ﬁne-loamy, mixed, mesic) and Oshtemo (coarse-
loamy, mixed, mesic) soil series (Typic Hapludalfs). The soils have a loam-textured
surface horizon, a sandy loam lower argillic (Bt) horizon, and a stratiﬁed sand and loamy
sand textured parent material (Whiteside, 1982). These soils, developed on glacial till, are
well-drained with moderate permeability in the upper part and highly permeable in the
lower part of the proﬁle. Original vegetation in the region included oak-hickory (Quercus—
Carya) forests with minor areas of elm-ash (Ulmus-Fraxinus) and beech-maple (Fagus-
Acer) (Kﬁchler, 1964). Selected properties of the soils in the experimental plots are shown
in Appendix A1.

The experiment was initiated in the spring of 1986 in a moldboard plowed area
under corn/soybean (Zea mays L./Glycine max (L.) Merr.) rotation for more than 50
years. Experimental plots (40 x 27 m) were installed in a randomized complete block with
four replications (Appendix A2). Treatments were conventional-tillage (CT) and no-tillage

(NT). Nitrogen fertilizer was broadcast shortly after planting at the rate of 168 kg N ha'1

21

yr". An additional treatment consisting of never-tilled plots under successional grass
vegetation (N eTSG) established in order to compare long-term soil structural changes,
was also evaluated. The inclusion of these plots in the study serves as a comparative value,
as the four replicated plots are not randomized within the main experiment. The never-
tilled successional plots were ﬁrst cleared of timber in the late 50's and have been mowed
annually since, which maintained a largely herbaceous community
(URL:http://lter/kbs.msu.edu/description.html).

The CT and NT treatments were continuously cropped with the following crops
since 1986: corn (‘86-88), soybeans (‘89), corn (‘90-94), and alfalfa (Medicago sativa L.)
(September ‘94 - May 96). Alfalfa hay was harvested three times during the 1995 growing
season (on 12 June, 22 July, and 1 September) with all alfalfa biomass removed from the
plots. Alfalfa plants were spray-killed on 2 May ‘96. Conventional tillage plots were
moldboard plowed in early spring of each year followed by one or two passes of a disk
and/or ﬁeld cultivator. Direct drilling was used in NT plots since the spring of 1985. All
plots have received glyphosate [isopropylamine salt of N-(phosphonomethyl) glycine] at
the rate of 7 L ha" in early spring of each year and were treated with pre- and post-

emergence herbicides.

Soil sampling and preparation
Four soil samples were collected per plot in May 1996 ﬁom the 0-5 and 5-15 cm
layers using a ﬂat shovel. Maximum sampling depth of 15 cm was chosen to avoid large

variations in bulk density of soil aggregates when approaching the plow pan. Previous

22

measurements in these experimental plots indicated bulk densities of 1.25, 1.45 and 1.19
Mg m‘3 at the 0-7.6 cm depth, and of 1.28, 1.44 and 1.46 Mg rn’3 at 7.6-15 cm depth, for
CT, NT and NeTSG treatments (Reinert, 1990).

Subsamples were collected in rows free of wheel traﬁc. Subsamples were pooled
together, manually separated along lines of least resistance, air-dried and gently sieved
through a 20-mm screen, by hand, to remove stones and plant residues. Dry soil materials
were further gently sieved by hand to obtain aggregates ranging from 12.5 to 16 mm
across. This size range was chosen because distinction between intact and broken
aggregates could be diﬂicult with smaller sizes. Aggregates were selected by hand to
avoid inclusion of broken aggregates in the samples, careﬁilly packed on foam containers
and kept at room temperature. Aggregates within this size were not available in the 5-cm

surface soil of NeTSG plots.

Soil aggregate fractionation

Aggregates, 12.5-16 mm in diameter, were mechanically separated into outer and
transitional layers, and the inner region using a modiﬁed version of the SAE apparatus
described by Santos et al. (1997) that was manuﬁactured by Agriculex, Inc. (Guelph,
Ontario, Canada). Brieﬂy, six aggregates from each replicated plot and from each depth
were placed into the sample erosion chamber (Fig. 2.1). Samples were ﬁrst processed for
15 seconds for the aggregates from the NeTSG plots, or 1 minute for the aggregates ﬁ'om
CT and NT treatments. Any broken aggregates were removed from the erosion chamber

and not included in the ﬁnal sample. The aggregates were mechanically eroded by

23

removing concentric layers of soil from each aggregate as they rotated; against an abrasive
stainless steel screen along the walls of the erosion chamber. Removal of surface layers
resulted in smaller spherical aggregates with a diameter depending on abrasion time.
Eroded samples included primary soil particles and microaggregates of two concentric
layers, collected in separate sample bottles, which were placed at the bottom of the ﬁmnel
(Fig. 2.1). The inner regions of the aggregates remained in the aggregate chamber.

To evaluate the efﬁcacy of the SAE method, the calibration procedure of Amelung
and Zech (1996) was used. For calibration, 15 aggregates ﬁ'om each replicated plot and
from each depth, were individually peeled for periods ranging from 0.2 to 25 min.
Aggregates that abraded completely during the assigned peeling time, were not considered
in the cah'bration procedure. The mass percentage of each aggregate layer was expressed
as a linear function of the total aggregate weight, W,, and the peeling time, t, according to

the following equation:

%Aggregate Surface Fraction (ASF) = a + bt + CW,- (1)

where a, b, and c are regression coefﬁcients (Amelung and Zech, 1996).

The eroded layer thickness (mm) of the outer and transitional regions of the
aggregate were inferred from their respective mass percentages. Aggregates were assumed
to have a spherical shape. The mean bulk densities of aggregates from each treatment
(Table 2.1), were measured using two distinct groups of aggregates, before and following

removal of the outer layer. These values were considered for the outer and transitional

24

layers, respectively. The procedure allowed calculation of the mass percentage of the
outer and transitional layers which had been eroded for at least 2 minutes for the NeTSG
plots, and at least 15 minutes for CT and NT plots, [Eq. 2]. Preliminary abrasion tests
indicated lower abrasion resistance for aggregates of the NeTSG plots. The following
multiple regression equation was used to estimate the thickness of the eroded layer:

% ASFW, = 100 - W,“l 41t/3 ( rags - layer 1:)3 100 pugg (2)

where r,“ and pass are the radius (mm) and density (g 1000 mm‘3 = g cm‘3 x 10") of the

aggregate. Layer x refers to the thickness (mm) of the layer to be removed. By inserting

r...= [W.- p..." (am-51'” (3)

and Eq. 2 into Eq. 1, the following equation is obtained

lbw, = b'1 [100- W;1 41t/3 (W, pugs" O.751t")"3 - layer x)3 100 ,0“,gg - a — cWJ (4)

where thy,” is the time needed to remove a layer of thickness x ﬁom the surface of the

aggregates. Equation 4 can also be expressed in terms of layer x:

Layerx = (W, pm“ 0.751r‘)“3 + {[(z b + a + c W, —100)m] / (400 1: pm» (5)

25

The effect of aggregate size on the mass percentage of aggregate layers was minimized by

using an aggregate size fraction as narrow as possible (12.5-16 mm dia.).

Aggregate bulk density

Bulk densities of aggregates before and after peeling were determined by the clod
method (Blake and Hartge, 1986). Each aggregate was suspended by a thread, weighed,
and dipped in saran solution until the sample became water proof. The saran-coated
aggregates were then weighed once in the air and once in water. Aggregate bulk density

(pegs) was calculated using the equation:

p...g = (A. W...)/ [W.. - W..... + W... - (W.. p. / 2.)] (6)

where p, = density of water, g cm”; Wmls == oven-dry mass of aggregate, g; W5, = net mass
of aggregate in the air, g; WSW = net mass of the sample and saran in the water, g; WI,“=

mass of saran coating in the air, g; and p, = density of the saran, g cm".

Tensile stength of soil aggregates.

Tensile strength of 10-15 aggregates was measured by crushing the aggregates
between parallel plates using a pressure gauge (Dexter and Kroesbergen, 1985).
Aggregates were oven-dried at 105 °C for 24 h, removed from the oven and allowed to
cool in a desiccator, thus preventing atmospheric water absorption. Aggregates were

removed one at a time immediately before being crushed. The tensile strength (Y) was

26

calculated for each aggregate using the equation:

Y = 0.576 F / D2 (7)

where D is the aggregate effective spherical diameter and 0.576 is a proportionality
constant relating maximum compressive stress to the tensile strength at failure (Dexter and
Kroesbergen, 1985). F is the force (N) required to fracture the aggregate and is
determined by F = b g, where b is the reading (in kilograms, kg) and g is gravity (9.807

m s'z). The effective spherical diameter of each aggregate was calculated as D = (6m/1tp)“3
where m is the mass (kg) of the individual aggregate and p is the mean aggregate bulk
density (g cm") of a speciﬁc treatment and depth, determined by the methods outlined

above.

Total carbon and nitrogen, and carbon isotope ratios

Replicated soil samples ﬁ‘om the 3 aggregate regions were ground to < 250 am
using a mortar and pestle and analyzed for total C and total N by the dry combustion
technique using a CHN analyzer (Carlo Erba, Italy). Delta (6)”C was determined by
combustion on a Isotope Ratio Mass Spectrometer Model 2020 (Europa Scientiﬁc, UK).
The 6'3C value was calculated from the measured l3C/‘zC ratios (R) of the sample and

standard gases as:

513C (%0) = [(Rsarnple ’ Rsmﬂ/Rsmm] x 103 (8)

27

and expressed relative to the internatioml standard PDB. Soil samples taken in 1994,
before alfalfa replaced corn, were used as the reference. The proportion of C derived ﬁ'om

alfalfa residues for the period of September 94 to May 96 (x) was calculated as:

x = (5.5 ‘ {Ml/(5. - ﬁes) (9)

where 6,5 = 613C value of sample (isolated layer or whole soil from alfalfa planted soil),
6“ = 6'3C value of the com-planted soil of the corresponding sample, and 6, = measured

5'3C value of alfalﬁi residues (-27.9%o).

Statistical analyses

Statistical analyses were performed with the Statistical Amlysis System (SAS
Institute, 1996). Where appropriate, one-way analyzes of variance (ANOVA-GLM) were
run, either comparing bulk samples, outer and transitional layers, and inner regions of
aggregates with each tillage, or across tillage within these sample and aggregate
components. Fisher’s LSD mean separation test was used to determine signiﬁcant
differences among means at 0.05 probability level. A t-test and a F test were used to

compare the regression coefﬁcients of the calibration equations.

28

RESULTS AND DISCUSSION

Evaluation of the peeling method

Thicknesses of the eroded layers ranged ﬁom 1.35 to 2.59 mm for the outer layer
and from 1.30 to 3.81 mm for the transitional layers of aggregates ﬁom the ﬁeld replicated
treatments (Table 2.1). Thicker outer and transitional layers abraded for the same time
period, were associated with soils under NT management. Particle losses ranged ﬁ'om 2.2
to 6.6% by weight. Higher losses were associated with the fractionation of the transitional
layer of NT soil aggregates sampled at 0-5 cm depth. The losses were primarily a result of
soil particles adherence to internal surfaces of the aggregate chamber and collector funnel
of the SAE apparatus, despite a thorough brushing and cleaning between runs for recovery
of particles. Nevertheless, the observed losses were much lower than those observed ( z
10% by weight) by an earlier model of the SAE (Santos et al., 1997). Using a
ﬁ’actionation method based on ﬁ'eezing and sieving, Wilcke and Kaupenjohann (1998)
observed mass percentage variations of 10 to 90% for outer layers of aggregates ranging
2-20 mm in diameter.

For a same abrasion time, the amount of soil removed in the transitional layer
(Table 2.1) was generally lower than in the outer layer but not always directly associated
with higher aggregate bulk density values (Table 2.2). Results might be improved by
peeling a smaller number of aggregates or even individual aggregates (data not reported).
The disadvantage of peeling individual aggregates is tlmt several subsamples might be

required to obtain enough soil for analysis.

29

Peeling times for removal of similar quantities of soil particles ﬁ'om the surfaces of
aggregates ranged ﬁom two (NeTSG plots) to 30 minutes (NT treatment at 0.05-0. 15m
depth) (Table 2.1). Lower aggregate erodibility results in larger amounts of aggregates
being abraded during a speciﬁed peeling time. Soil aggregates ﬁ'om NeTSG plots showed
very low abrasion resistance probably due to a large grass root distribution around and
through them. Aggregate surface fractions corresponding to about one-third of their total
mass were isolated as outer (31.5%) and transitional (34%) layers after peeling times as
short as 2 and 3 minutes (Table 2.1).

Regression coefﬁcients of bivariate regressions of ASF yields vs. peeling time (t)
and aggregate weight (Wi) are shown in Table 2.3. The regression coefﬁcients b and c
indicate the relative contribution of its variables, t and Wi, respectively, to the coefﬁcient
of determination, R2. The regression coeﬂicients are highly signiﬁcant (p < 0.01) for
peeling time but not for aggregate weight (Table 2.3). The only exception was for the CT
treatment in the top soil layer. The high signiﬁcance observed for the regression
coeﬂicients seems to indicate that the narrow range of aggregate diameter used (12.5-16
mm) was eﬂicient in reducing the variation in mass percentage of ASF. A ﬁnal conclusion
on that is precluded since no other aggregate sizes were tested. Values of R2 for CT and
NT treatments ranged from 0.80 to 0.88 and were signiﬁcantly different from zero (p <
0.05). The lowest R2 value (0.60), was observed for the NeTSG treatment (Table 2.3).
This poor relationship was due to the fact that the aggregates ﬁ'om the NeTSG plots 1nd a
very low resistance to abrasion. Since many of these weakly stable aggregates broke down

during the assigned peeling time, they were not included in the calculations.

30

Aggregate bulk density

Whole air-dry aggregates had average bulk densities varying ﬁom 1.62 (NeTSG)
to 1.79 (CT) Mg m‘3 while in the inner region the values ranged ﬁ'om 1.51 (NeTSG) to
1.80 (CT) Mg m'3 (Table 2.2). Aggregate bulk densities were not signiﬁcantly different
between tillage systems in the topsoil layer. Bulk densities of whole aggregates increased
signiﬁcantly with soil depth for CT, probably resulting ﬁom decreased SOM contents.
Cultivation of NeTSG plots substantially increased bulk densities of aggregates, especially
at depths of 5-15 cm. The lower bulk density of NT soil aggregates at 5-15 cm depth
could have been the result of partial incorporation of crop residues into the aggregates and
the undistm'bed organic debris of previous crops’ root systems as suggested by Dao
(1993). Aggregate bulk density has been considered to either directly or inversely
inﬂuence root growth and development. Lower bulk densities of aggregates (15-25 mm
dia.) under a grass sod in Texas were attributed to a larger amount of ﬁne roots in the
surface layers (Zobeck et al., 1995). Conversely, Voorhees et al. (1971) observed that, as
aggregate density decreased, root grth not only increased but roots also became
relatively more effective in proliferating and exploring the soil aggregates. Our aggregate
bulk densities were 11 to 40% higher than values obtained previously by the core method
for these soils (Reinert, 1990). Greater bulk densities of aggregates than from soil cores
are the result of greater interaggregate porosities associated with larger soil volumes

(Blake and Hartge, 1986).

31

Tensile strength of soil aggregates

Large diﬂ‘erences in tensile strength of aggregates were observed between
treatments at each sampling depth (Table 2.2). Tensile strengths were 50 and 40% greater
on aggregates from conventional tilled soils at 0-5 and 5-15 cm depths (Table 2.2). In the
present study, the physical reinforcement of aggregates by alfalfa root system may have
potentially contributed to the increasing strengths of the aggregates during the previous
20 months time period. Increased tensile strengths in CT compared to NT system seems
consistent not only with the increase in aggregate bulk density of whole aggregates but
also with a potential shorter duration of wet periods in moldboard plowed soils (Kay et al.,
1994). The surface layer of moldboard-plowed soils (CT treatment) is expected to
experience larger amplitudes in drying events during wetting/drying cycles (Gregorich et
aL, 1993). Moreover, the closeness of the association between roots and the soil matrix
and the intensity of localized soil water extraction by roots will enhance drying-wetting
cycles leading to changes in aggregate strength and stability (Van Noordwijk et al., 1993;

Materechera et al., 1994).

Total carbon and nitrogen

Within tillage systems, signiﬁcant differences in total C concentration between
aggregate layers within CT and NT tillage systems were only observed for the 0-5 surﬁnce
soil (Fig. 2.2). In the 0-5 cm aggregates, C concentrations decreased from 22.6 to 10.8 g
kg" ﬁ'om the outer layer to the inner region for CT samples and from 29.9 to 9.7 g kg" for

NT samples (Fig. 2.2). For the 5-15 cm aggregates, C contents decreased from 12.8

32

(outer) to 9.8 g kg" (inner) for CT, and ﬁ'om 12.7 to 10.4 g kg" for NT treatments.
Carbon gradients from inner regions to outer layers of soil aggregates ranged ﬁom 14.8 to
21.4 g kg" for NeTSG samples (Fig. 2.2). At the 0-5 cm depth, aggregates ﬁom NeTSG
plots displayed greatest C concentrations in the transitional layer (Fig. 2.3). Absence of
signiﬁcant differences in total C between CT and NT treatments for the bulk soil (Fig.
2.2), is consistent with previous observations conducted on these plots (Daroub, 1994).
Several other studies at various locations, lnve indicated that management effects on soil
C are generally conﬁned to the near surface of soil under NT compared to CT (Blevins et
al., 1983; Havlin et al., 1990; Mahboubi et al., 1993; Ismail et al., 1994; Beare et al.,
1994). Comparing 20 years of CT and NT continuous corn, Ismail et al. (1994) reported
increments of soil organic C ranging from 74% and 3% for 0-5 and 5-15 cm depths,
respectively. The main reason for soil C gradients with depth seems to be that plant
residues are not mechanically incorporated into the plow layer therefore accumulating at
the soil surface. Moreover, tillage increases the rate of organic matter decomposition and
C and N mineralization (Blevins et al., 1977).

Soil C concentration in the outer layers of aggregates increased in comparison to
that of the inner regions of aggregates by as much as 200% for NT and 100% for CT
aggregates collected in the top 5 cm layer (Fig. 2.2). Despite the signiﬁcant enrichment of
C in the outer layer of aggregates of CT and NT soils at the depth of 0-5 cm, a
considerable reduction in the amount of C was observed in these treatments, as compared
to the NeTSG plots (Fig. 2.3a). No C concentration gradients were observed in

aggregates ﬁ'om NeTSG plots, which contrasts with the results reported by Amelung and

33

Zech (1996) and with the idea that better oxygen supply in aggregate surface layers
facilitates SOM turnover by aerobic microorganisms (Sextone et al., 1985). A possible
explanation resides in the key role played by grass root systems in NeTSG soils, as they
favored distribution and mixing of large quantities of organic materials into the soil
(Oades, 1984). This homogeneous distribution of organic materials across aggregates
appears to not limit the activity of microorganisms regardless of their location within the
NeTSG aggregates, resulting in minimal gradients within aggregates.

The transitional layers and inner regions of CT and NT aggregates showed similar
levels of C (:10 g C kg" soil) at the 5-15 cm depth (Fig. 2.2) but much lower than the
levels for the inner regions of the NeTSG aggregates (’4 18 g C kg" soil) (Fig. 2.3a). These
C losses from tilled aggregates can be attributed to the disruption of macroaggregate
structures by cultivation, exposing SOM formerly protected within the aggregate structure
(Tisdall and Oades, 1982; Beare et al., 1994). As a result, the quantity of organic binding
agents, such as carbohydrates, can be expected to decline in cultivated soils with loss of
structural stability (Tisdall and Oades, 1982).

The higher C contents in the outer and transitional layers compared to the inner
regions of aggregates from CT and NT results partly from recent additions of organic
materials from alfalfa roots and root exudates, as indicated by the more negative 6‘3C
ratios (Fig. 2.4). The newly added organic matter seems to have been deposited mainly on
or near the surfaces of the aggregates causing the outer layer to be signiﬁcantly richer in
organic C and N contents compared to the inner regions. Fine root residues of annual

crops, such as corn, tend to be a high C:N ratio at harvest time since the plants generally

34

remobilize N from vegetative storage tissues to the seeds (Van Noordwijk and Brouwer,
1997). By contrast, perennial crops may have a greater root turnover during the growing
season and higher N contents at harvest time. Considering that over 70% of the alfalfa
root biomass is concentrated in the upper 20 cm of the soil proﬁle (Jodari-Karimi et al.,
1983), the high incorporation of alfalfa-derived C into the surfaces of aggregates observed
for the relatively short time period of 20 months, corroborates the observation that the
amount of organic material released by roots is related to the total root mass (Shamoot et
al., 1968; Tisdall and Oades, 1982). In structured soils, plant roots tend to use available
voids which they modify partly or completely by pressure during growth so that contact
between roots and soil matrix is increased. Provided that there is a root anchorage
immediately behind the root tips (V oorhees et al., 1971), ﬁne roots and root hairs would
extend through and around the large aggregates leading to enmeshment of the soil matrix
and improved aggregate stability (Tisdall and Oades, 1982).

Wilcke et al. (1996), suggested that preferential sorption of dissolved forms of C
ﬁ'om overlying organic layers into the aggregate surfaces could be one of the explanations
for increments of soil organic C of 38% to 61% for aggregates collected ﬁ'om the upper
23 cm of a Podzol soil. In the present study, however, contributions ﬁom dissolved
organic materials resulting ﬁom the decomposition of alfalfa shoots were not expected
since all material was removed from the plots upon cutting. Despite showing higher total
C levels, most of the C in the outer layers of aggregates ﬁ'om NT treatment comes ﬁom
the previous corn crop, as indicated by the less negative 6'3C ratio values compared to CT

(Fig. 2.4). Accordingly, previous corn root measurements indicated that the proportions of

35

roots in Ap horizon were greater in NT than CT plots, by 52% in 1991 and 61% in 1992
(Huang, 1995). Higher bulk densities and lower macroporosities, suggesting a reduction
of interaggregate pore space, were also observed in NT plots in November 1989 (Reinert,
1990). Greater root proportions in NT plots can be partly explained by the accumulation
or clustering of roots in the more compacted soil layers along the cracks or planes of
weakness between aggregates (McSweeney and Jansen, 1984; Vepraskas and Wagger,
1990; Staricka et al., 1992; Smucker and Allmaras, 1993). By occupying these large pores
and growing around the larger aggregates, plant roots would greatly enhance C
incorporation in outer (29.9 g kg") and transitional (22.7 g kg") layers of NT aggregates,
compared to the bulk soil sample (13.3 g kg") (Fig. 2.2). In contrast to NT system, soil
loosening caused by moldboard plowing in CT plots, probably facilitated the incorporation
of alfalfa-derived C within aggregates. The observed two-fold increase in total C contents
of outer and transitional layers of CT aggregates, as compared to the inner zone, suggests
that the inner cores of the aggregates at some point in time were aggregates of smaller
diameter before plowing took place. Due to their greater tensile strength, small aggregates
are more resistant to breakdown than large aggregates (Hadas, 1987). Following tilhge,
active incorporation of organic matter would enhance emneshment and deposition of fresh
organic nmterials, roots, and microbial biomass on the surfaces of the initial aggregate
(Angers and Chenu, 1998). On a long run, these meclmnisms would be further enhanced
through root activity and rhizodeposition, associated with fauna] activity, wetting and

drying, and potential breakage and reformation of the aggregates.

36

In general the effect of tillage on total N at both depths (Fig. 2.5), was similar to
that of C (Fig. 2.2.). In the surface 5 cm there was signiﬁcantly more total N in NT, while
below 5 cm no signiﬁcant changes were observed (Fig. 2.5). These data suggest that N
availability was comparable in NT and CT treatments. Since there were no large
differences in N inputs between tillage treatments, the lower N content in the outer layer
of CT aggregates indicates greater net N mineralization in the cultivated soil. Cambardella
and Elliott (1994) reported changes in residue decomposition due to alteration in soil
structure by increased tillage intensity. They observed that no-tillage soils had more labile
N associated with macroaggregates than more intensely tilled soils.

Carbon to N ratios are usually higher for NT than CT because younger organic
matter (with higher C:N ratio) accumulates near the soil surface under NT (Frye and
Blevins, 1997). The mixing of young and old organic matter by plowing would diminish
much of the effect of age of SOM on C:N ratio under CT. Carbon to N ratios were similar
between tillage systems (Fig. 2.6). A signiﬁcantly higher C:N ratio for CT was found in
the outer layer of aggregates in the surface 5 cm. The C:N ratio in the outer and
interrmdiate layers of CT and NT aggregates were consistently wider than in
corresponding bulk soil and inner regions sampled at 0—5 cm depth (Fig. 2.6). The C:N
ratio of alfalfa and corn plant materials is about 14 and 58, respectively (Ajwa and
Tabatabai, 1994). The high C:N ratios suggests that the accumulating SOC in the external
layers of aggregates is not highly processed and probably includes relatively large
quantities of particulate organic matter. High C:N ratios can be attributed to incompletely

hurniﬁed organic nuterial and suggests that active binding agents are root hairs and ﬁrngi

37

(Buyanovsky et al., 1994). A different type of organic matter, highly processed and more
stable than the one associated with more external layers, would be located inside the
cultivated aggregates ﬁ'om cultivated plots (C:N ratios z 10). The interior of the
aggregates is protected from microbial attack by structural organization of the primary
particles into aggregates and by the chemical structure of organic matter as it bonds with
mineral particles (Buyanovisky et al., 1994). For the 5-15 cm sampling depth, the C:N
ratios of CT and NT aggregates did not diﬂ‘er from those of NeTSG plots (Fig. 2.3c).

Soil organic matter associated with the outer and transitional layers may have been
processed by soil organisms to a lesser extent than that of the inner regions and bulk soil
samples. Thus, the ﬂow of microbial by-products may depend on the location within the
aggregate (Monreal and Kodama, 1997). The hypothesis that oxidative decay of SOM is
enhanced at the aggregate surfaces and inhibited in inner regions (Amelung and Zech,
1996), does not explain the lower SOC values in the inner regions of cropped soil

aggregates in the present study.

Carbon isotope ratios

The average 6‘3C of soil under successional grass vegetation across aggregate
fractions and bulk soil, at the 0-15 cm depth, was -25.7%o (Fig. 2.3d), conﬁrming that the
inputs of C have been ﬁom C3 vegetation. Mean values (:2 standard deviation, n = 4) of
6‘3C of plant shoot and root samples ﬁ'om successional grass vegetation were -27.7%o
(i 0.41%o) and -27.6%o (d: 0.31%o), respectively. There was a slight enrichment in I3C of

the soil C compared to the native vegetation. Similar soil l3C enrichments have been

38

reported to result from the discrimination in C isotopes that occur during microbial
breakdown (Balesdent et al., 1987), or ﬁom the presence of some C4-type organic nutter.
The presence of some C4 weed species in NeTSG may be related to this enrichment.

Surface (0-20 cm) bulk soil samples, taken in 1994, before alfalfa was planted,
indicated that, at that time, cropping practices had increased the 5‘3C ﬁ'om -25.7%o
(NeTSG) to -22.7%o (t 10960, n = 2) under CT, and to -23.5%o (i 0.52960, n = 2) under
NT. Until 1985, these soils had been primarily under C, (corn) and C3 (wheat/soybeans)
rotation for over 80 years. The mixture of these previous crop residues was calculated to
have an average 6‘3C of -17.1%o (Collins et al., 1997). Since the difference in 6‘3C content
between tillage systems is probably not signiﬁcant, an average 6'3C value of ~23. 1%» for
CT and NT was calculated. Thus, after 9 years, with estimated productions of 5 Mg C ha"
yr" for corn residues (6 years) and 3 Mg C ha" yr" for other crop residues, the inputs of
total soil C by crops planted before alfalfa were estimated to be as much as 30% (Paul et
al., 1998).

Surface soil samples taken in 1986 across the entire experimental area, before the
experiment was set up, showed an average soil organic C of 7.9 g C kg" (n = 98; Reinert,
1990). At that time, reference soil 0C levels (estimated ﬁom the 5-15 cm NeTSG
treatment, Fig. 2.3a), had decreased by 63%. By the spring of 1989, OC contents for CT
and NT were 7.2 and 8.6 g C kg", respectively (n = 12; Reinert, 1990). These values
represented 34 and 40% of the reference 0C levels. The samples taken in 1994 contained
8.9 and 8.6g C kg" for CT and NT, respectively, indicating no substantial increases in C

levels over the 5-year period.

39

The introduction of alfalfa in the continuous corn cropping system allowed us to
analyze the distribution of the new C3-C in the soil. Though not stratiﬁed by depth, 613C
data ﬁ’om samples collected in 1994 in CT (~22.7%o :1: 1.0%o, n = 2), and NT (-23.5%o :1:
0.52%o, n = 2) plots were used as baselines to estimate inputs of alfalfa-derived SOC in the
bulk soil samples taken in 1996. 6'3C values (i standard deviation, n = 2) for outer,
internmdiate, and inner regions of these soil aggregates were, -23.5%o, -24.3%o, and -
23.7%» for CT and -23.1%o, -23.8%o, and -23.4%o for NT, respectively. After 20 months
of alfalfa cropping, 513C in CT bulk soil samples were similar in the top 5 cm (-22.7%o vs.
-22.8%o ) and at the 5-15 cm depth (-22.7%o vs. -23.5%o). Alfalfa-derived C was not
detected in the bulk samples under NT treatment. Changes in 6‘3C are not generally
detected in such a short time span. However, when the isotopic composition was
measured in the concentrically layered fractions separated from individual
macroaggregates, a consistent pattern of deposition of alfalfa-derived C was observed
between aggregate layers. The NC signature decreased (became more negative) in the
outer and transitional layers of aggregates (Fig. 2.4). On average, for aggregates sampled
in the 0-5 cm soil depth of CT and NT plots, 67 and 43% of the C associated with outer
and transitional layers of macroaggregates, respectively, were derived from alfalfa,
whereas no new C was found in the inner regions of the aggregates (Fig. 2.7). Gradients
of 6'3C with soil depth in CT treatments may have occurred due to the absence of
moldboard plowing for nearly 2 years, during alfalfa growth. The accumulation of alfalfa
root-derived C was surprisingly high. Perennial legume forages have been associated with

increases in aggregate stability and have the potential to favor soil C storage on a long-

40

term basis. A short-term increase in aggregate stability after only two years of alfalfa
cropping did not reﬂect signiﬁcant increases in SOC of bulk soil samples, in neither our
studies (Fig. 2.2) nor those reported by Angers (1992). Studying the incorporation of
com-root C into soils, Balesdent and Balabane (1996), observed a large accumulation of
below ground C, probably resulting from a high below ground production and a slow
biodegradation of root-derived material. In this study, the very active ﬁne root production
by alfalfa, along with a rapid root decomposition rate (Rasse and Smucker, 1998a, b)
potentially explains the high proportion of alfalfa-derived C on aggregate surfaces.

A large proportion of the recently deposited organic material is accumulated
predominantly in macroaggregates (Angers and Giroux, 1996; Puget et al., 1995)
supposedly being responsrble for increased aggregate stability of these ﬁ'actions. This
organic matter is highly labile, easily available to heterotrophic microbes, and consequently
subject to potential decomposition upon macroaggregate disruption (Golchin et al., 1997;
Angers and Carter, 1996). Thus, increased macroaggregate stability is a transient process
and its maintenance strongly depends upon continuous additions of C sources to the soil
matrix. The incorporation of young alfalfa-derived C in outer and transitional aggregate
layers in our study seems to conﬁrm the observation of Besnard et al. (1996) that, with
cultivation, a partial breakdown and reformation of soil macroaggregates may occur. In
the present case, the outer and transitional layers could have been developed over the 20-
month period of alfalfa on pre-existing smaller macroaggregates, which later became inner
regions of larger macroaggregates. The initial steps of this process involve rhizosphere

activity, which enhances microaggregation by extensive root exudation, and cell debris

41

and mucigel production (Domaar and Foster, 1991; Foster, 1994). Residues of mucigel
products and root hair are mainly on the surfaces of already formed aggregates and are
rapidly colonized by associated microbial populations. The microﬂora and its by-products
are strongly adhesive to mineral particles (Angers and Chenu, 1998). This process would
lead to the formation of 20-250um microaggregates (Golchin et al., 1997) which would
then be surrounded and enveloped by fungal hyphae, root residues, illuvial clays, etc., into
concentric layers which form larger macroaggregates of several millimeters in diameter.
The high degree of fragmentation of the particles removed by the ﬁactionation
technique prevented a further ﬁ'actionation of each layer and region of the aggregates.
Techniques to discriminate for the contributions of plant- and microbial-derived C among
different sized fractions, within each layer, will provide additional information about the
origins of organic matter in macroaggregates and provide additional mechanisms
associated with their formation, stabilization, and degradation. Successive measurements
of 13C values in compartments of aggregates of several sizes across crop rotations and
tillage practices will assist in the determination of aggregate life expectancies and possible

microniches of SOM and associated microbial activities.

42

CONCLUSIONS

Quantities of C continuously deposited onto soil aggregates can be considered one
of the major factors relating crop and soil management systems to soil structure. The
results of this experiment indicated greater concentrations of C located in the outer or
surface layers of soil aggregates. Accumulation patterns of recently deposited C in
aggregates were shown to be highly tillage-speciﬁc with differences being reﬂected on the
form and amount of carbon entering the soil and its spatial distribution in the soil
aggregates. 6'3C analyzes of CT and NT soils demonstrated that more C from a
contemporary alfalfa crop was deposited on surface of aggregates from CT than ﬁom NT.
Interior regions of aggregates from NT soils remained less negative than soils ﬁ'om CT
treatments. Accumulations of C3-C were greater for aggregates sampled near the soil
surface. The differences in SOC between layers may be related to differences in vertical
distribution of C resulting from the two diﬂ‘erent tillage systems. No-tillage management
appeared to be more effective at accumulating or slowing the decomposition of newly
incorporated C and N in soil aggregates. The results indicate that, on a short-term basis,
the below-ground plant system has an important role in the site-speciﬁc deposition of SOC
and the dynamics of structure. The novel small scale sampling design used in these studies
provided opportunities for quantifying accumulations of organic matter in soil aggregates

and its dynamic interactions within soil structural units, as affected by crop rotations.

43

REFERENCES

Ajwa, H.A., and MA. Tabatabai. 1994. Decomposition of different organic materials in
soils. Biol. Fertil. Soils 18:175-182.

Amato, M., and IT Ritchie. 1995. Small spatial scale soil water content measurement
with time-domain reﬂectometry. Soil Sci. Soil. Am. J. 59:325-329.

Amelung, W., and W. Zech. 1996. Organic species in ped surface and core ﬁactions along
a climosequence in the prairie, North America. Geoderma 74:193-206.

Angers, DA. 1992. Changes in soil aggregation and organic carbon under corn and
alfalfa. Soil Sci. Soc. Am. J. 56:1244-1249.

Angers, D.A., and M.R. Carter. 1996. Aggregation and organic matter storage in cool,
humid agricultural soils. p. 193-211. In M.R. Carter and B.A. Stewart (ed.)
Structure and organic matter storage in agricultural soils, Advances in Soil
Science, CRC Press, Boca Raton, FL.

Angers, D.A., and C. Chenu. 1998. Dynamics of soil aggregation and C sequestration. p.
199-206. In R. Lal et al. (ed.) Soil processes and the carbon cycle. CRC Press,
Boca Raton, FL.

Angers, DA, and M. Giroux. 1996. Recently deposited organic nutter in soil water-
stable aggregates. Soil Sci. Soc. Am. J. 60:1547-1551.

Balesdent, J ., A. Mariotti, and D. Boisgontier. 1990. Effect of tillage on soil organic
carbon mineralization estimated from 13C abundance in corn ﬁelds. J. Soil Sci.
41:587-596.

Balesdent, J., and M. Balabane. 1996. Major contribution of roots to soil carbon storage
inferred from corn cultivated soils. Soil Biol. Biochem. 28:1261-1269.

Balesdent, J., A. Mariotti, and B. Guillet. 1987. Natural l3C abundance as a tracer for soil
organic matter dynamics studies. Soil Biol. Biochem. 19:25-30.

Beare, M.H., P.F. Hendrix, and DC. Coleman. 1994. Water-stable aggregates and organic
matter fractions in conventional- and no-tillage soils. Soil Sci. Soc. Am. J. 58:777-
786.

Besnard, E., C. Chenu, J. Balesdent, P.Puget, and A. Arrouays. 1996. Fate of particulate
organic matter in soil aggregates during cultivation. Eur. J. of Soil Sci. 47:495-
503.

Blake, GR, and KH. Hartge. 1986. Bulk density. p. 363-375. In A. Klute (ed.) Methods
of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.

Blevins, RL., G.W. Thomas, and PL. Cornelius. 1977. Inﬂuence of no-tillage and
nitrogen fertilization on certain soil properties after ﬁve years of continuous corn.
Agron. J. 69:383-386.

Blevins, R.L., M.S. Smith, G.W. Thomas, and W.W. Frye. 1983. Changes in soil
properties after 10 years continuous non-tilled and conventionally tilled corn. Soil
Tillage Res. 3:135-146.

Buyanovsky, G.A., M. Aslam, and G.H. Wagner. 1994. Carbon turnover in soil physical
ﬁ'actions. Soil Sci. Soc. Am. J. 58:1167—1173.

Cambardella, C.A., and ET. Elliott. 1994. Carbon and nitrogen dynamics of soil organic
matter fractions from cultivated grassland soils. Soil Sci. Soc. Am J. 58:123-130.

Collins, H.P., B.A. Paul, K. Paustian, and ET. Elliott. 1997. Characterization of soil
organic carbon relative to its stability and turnover. p. 51-72. In B.A. Paul et al.
(ed.) Soil organic matter in temperate agroecosystems: Long-term experiments in
North America. CRC Press, Boca Raton, FL.

Dao, TH. 1993. Tillage and winter wheat residue management eﬂects on water
inﬁltration and storage. Soil Sci. Soc. Am. J. 57:1586-1595.

Daroub, SH. 1994. The effect of tillage on phosphorus transformations in soils. Ph.D.
diss. Michigan State Univ., East Lansing. (Diss. Abstr. 56-1154).

Dexter, A.R. 1988. Advances in characterization of soil structure. Soil Till. Res. 11:199-
238.

Dexter, A.R., and B. Kroesbergen. 1985. Methodology for determination of tensile
strength of soil aggregates. J. Agric. Eng. Res. 31:139-147.

Dormaar, IF, and RC. foster. 1991. Nascent aggregates in the rhizosphere of perennial
ryegrass (Lolium perenne L.). Can. J. Soil Sci. 71 :465-474.

45

Foster, RC. 1994. Microorganisms and soil aggregates. p. 144-155. In C.E. Pankhurst et
al. (ed.) Soil biota: Management in sustainable farming systems. CSIRO, East
Melbourne, Australia.

Frye, W.W., and RL. Blevins. 1997. Soil organic matter under long-term no-tillage and
conventional tillage corn production in Kentucky. p.227-234. In E.A. Paul et al.
(ed.) Soil organic matter in temperate agroecosystems: Long-term experiments in
North America. CRC Press, Boca Raton, FL.

Golchin, A., J.A. Baldock, and J .M. Oades. 1997. A model linking organic matter
decomposition, chemistry and aggregate dynamics. p. 245-266. In R. Lal et al.
(ed.) Soil processes and the carbon cycle. CRC Press, Boca Raton, FL.

Gregorich, E.G., D. Reynolds, J.L. Culley, M.A. McGovern, and WE. Cumoe. 1993.
Changes in soil physical properties with depth in a conventionally tilled soil after
no tillage. Soil Tillage. Res. 26:289-299.

Hadas, A. 1987. Long-term tillage practice effects on soil aggregation modes and
strength. Soil Sci. Soc. Am. J. 51:191-197.

Havlin, J.L., D.E. Kissel, L.D. Maddux, M.M. Claassen, and J.H. Long. 1990. Crop
rotation and tillage effects on soil organic carbon and nitrogen. Soil Sci. Soc. Am.
J. 54:448-452.

Heil, RD, and GJ. Buntley. 1965. A comparison of the characteristics of the ped faces
and ped interiors in the B horizon of a Chestnut soil Soil Sci. Soc. Am. Proc.
29:583-587.

Horn, R. 1990. Aggregate characterization as compared to soil bulk properties. Soil
Tillage Res. 17:265-289.

Huang, B. 1995. Tillage modiﬁcations of root and shoot growth responses to soil water
content and nitrogen concentration altered by seasons. Ph.D. diss. Michigan State
Univ., East Lansing. (Diss. Abstr. 56-5233).

Isrmil, 1., RL. Blevins, and W.W. Frye. 1994. Long-term no-tillage effects on soil
properties and continuous corn yields. Soil Sci. Soc. Am. J. 58:193-198.

Jodari-Karimi, F., V. Watson, H. Hodges, and F. Whisler. 1983. Root distribution and

water use efﬁciency of alfalfa as inﬂuence by depth of irrigation. Agron. J. 75 :207-
211.

46

Kay, B.D. and A.R. Dexter. 1990. Inﬂuence of aggregate diameter, surface area and
antecedent water content on the dispersibility of clay. Can. J. Soil Sci. 70:655-671.

Kay, B.D., AR. Dexter, V. Rasiah, and CD. Grant. 1994. Weather, cropping practices
and sampling depth effects on tensile strength and aggregate stability. Soil Tillage
Res. 32:135-148.

Kayser, A.T., W. Wilcke, M. Kaupenjohann, and J.D. Joslin. 1994. Small scale
heterogeneity of soil chemical properties. I. A technique for rapid aggregate
ﬁ‘actionation. Z. Pﬂanzenernahr. Bodenk. 157:159-164.

Ktlchler, A.W. 1964. Potential natural vegetation of the conterminous United States. Am.
Geog. Soc. Spec. Publ. 36. American Geographical Society, New York.

Logsdon, SD, and R.R. Allrnaras. 1991. Corn and soybean root clustering as indicated by
root mapping. Plant Soil 131:169-176.

Mahboubi, A.A., R. La], and NR F aussey. 1993. Twenty-eight years of tillage eﬂ‘ects on
two soils in Ohio. Soil Sci. Soc. Am J. 57:506-512.

McSweeney, K., and U. Jansen. 1984. Soil structure and associated mating behavior in
minesoils. Soil Sci. Soc. Am. J. 48:607-612.

Materechera, S.A., J .M. Kirkby, A.M. Alston, and A.R. Dexter. 1994. Modiﬁcation of soil
aggregation by watering regime and roots growing through beds of large
aggregates. Plant Soil 160:57-66.

Monreal, CM. and H. Kodarm. 1997. Inﬂuence of aggregate architecture and minerals on
living habitats and soil organic mtter. Can. J. Soil Sci. 77:367-377.

Oades, J .M. 1984. Soil organic matter and structural stability mechanisms and implications
for management. Plant Soil 76:319-337.

Passioura, J.B. 1991. Soil structure and plant growth. Aust. J. Soil Res. 29:717-728.

Paul, B.A., H.P. Collins, and S. Haile-Mariam. 1998. Analytical determination of soil C
dynamics. In Proc. World Congress of Soil Science, 1888, Montpellier, France.
20-26 Aug. 1998. 1888, Montpellier, France.

Puget, P., C. Chenu, and J. Balesdent. 1995. Total and young organic matter distributions
in aggregates of silty cultivated soils. Eur. J. Soil Sci. 46:449-459.

47

Qureshi, RH., and DA. Jenkins. 1987. Concentration of phosphorus and sulphur at soil
ped surfaces. J. Soil Sci. 38:255-265.

Rasse, D., and A.J.M. Smucker. 1998a. Root recolonization of previous root channels in
corn and alfalfa rotations. Plant Soil. In Press.

Rasse, D., and A.J.M. Smucker. 1998b. Tillage effects on plant biomass and soil nitrogen
dynamics in a corn, alfalfa and corn succession. J. Environ. QuaL In Press.

Reinert, D. 1990. Soil structural form and stability induced by tillage in a Typic Hapludalf.
Ph.D. diss. Michigan State Univ., East Lansing. (Diss. Abstr. 52:0005).

Santos, D., S.L.S. Murphy, H. Taubner, A.J.M. Smucker, and R. Horn. 1997. Uniform
separation of concentric surface layers ﬁom soil aggregates. Soil Sci. Soc. Am J.
61 :720-724.

SAS Institute, Inc. 1996. SAS/STAT user’s guide. Version 6.11 ed. SAS Inst., Inc.,
Cary, NC.

Sextone, A.J., N.P. Revsbech, T.B. Parkin, and J.M. Tiedje. 1985. Direct measurement of
oxygen proﬁles and denitriﬁcation rates in soil aggregates. Soil Sci. Soc. Am J.
49:645-65 1 .

Sharnoot, S., L. McDonald, and W.V. Bartholomew. 1968. Rhizo-deposition of organic
debris in soil. Soil Sci. Soc. Am Proc. 32:817-820.

Smucker, A.J.M., and R.R. Allrnaras. 1993. Whole plant responses to soil compaction.
p.727-731. In Proc. International Crop Science I. CSSA, Madison, WI.

Smucker, A.J.M., D. Santos, and Y. Kavdir. 1996. Concentric layering of carbon, nitrogen
and clay minerals within soil aggregates ﬁom tilled and non-tilled agroecosystems
and natural grasslands. p.129-140. In J. Caron et al. (ed.) Proc. Eastern Canada
Soil Structure Workshop, Merrickville, Ontario, Canada. 21-22 Aug.

Staricka, J.A., R.R. Allmaras, W.W. Nelson, and WE. Larson. 1992. Soil aggregate
longevity as determined by the incorporation of ceramic spheres. Soil Sci. Soc.
Am J. 56:1591-1597.

Tisdall, J.M., and J .M. Oades. 1982. Organic matter and water-stable aggregates. J. Soil
Sci. 33:141-163.

48

Van Noordwijk, M., and G. Brouwer. 1997. Roots as sinks and sources of nutrients and
carbon in agricultural systems. p. 71-89. In L. Brussaard and R. F errera-Cerrato
(ed.) Soil ecology in sustainable agricultural systems, CRC Press.

Van Noordwijk, M, G. Brouwer, and K. Harmanny. 1993. Concepts and methods for
studying interactions of roots and soil structure. Geoderma 56:351-375.

Vepraskas, M.J., and M.G. Wagger. 1990. Corn root distribution and yield response to
subsoiling for Paleudults having different aggregate sizes. Soil Sci Soc. Am J.
54:849-854.

Voorhees, W.B., M. Amemiya, R.R. Alhnaras, and WE. Larson. 1971. Some effects of
aggregate structure heterogeneity on root growth. Soil Sci. Soc. Am Proc.
35:638-643.

Whiteside, ER 1982. Soil survey of the Kellogg Biological Station. Unpublished
Research Report. WK. Kellogg Biological Station Library, Hickory Comers, MI.

Wilcke, W., R. Baumler, H. Deschauer, M. Kaupenjohann, and W. Zech. 1996. Small-
scale distribution of Al, heavy metals, and PAHs in an aggregated alpine Podzol.
Geoderma 71:19-30.

Wilcke, W. and M. Kaupenjohann 1997. Differences in concentrations and fractions of
aluminum and heavy metals between aggregate interior and exterior. Soil Sci.
162:323-332.

Wilcke, W. and M. Kaupenjohann. 1998. Heavy metal distribution between soil aggregate
core and surface ﬁactions along gradients of deposition from the atmosphere.
Geoderma 83:55-66.

Zobeck, T.M., N.A, Rolong, D.W. Fryrear, J.D. Bilbro, and BL. Allen. 1995. Properties

and productivity of recently tilled grass sod and 70-year cultivated soil. J. Soil
Water Conserv. 50:210-215.

49

ANN 05:5 302638.. .moummouwwe me 802on
5::m pg moemwoewwe 03:? mo 32? 56:0: £3 owns; :0 women 203 m8?— Eommmaab :5 850 no 80520:: 05 mo moegmm a

 

 

 

 

 

 

 

agamé m. ﬁaN ”$6.3” aﬁoﬂwd m «.03.. Wmam. 2 m 2 63m.— N 02.02

mdﬁﬁm moan; odﬁmém 8.33:0. 2 vdid wsummem Sdﬂmﬁm om HZ

NdamN moan; c. We. 3 2 denote..— 2 moan-:2 cdﬁodm 3.322 2 HO
33.34

@366 mﬁmmé 062.2 2.33.2 2 mgﬁd 2;“de 2.32: 2 HZ

vdaﬁm mdﬂQN w. 3:52 mmdﬂmv; 2 odﬁod $93.3 omoawmé 2 HO
543.5%an

..\c ..\o 88 :w: ..\o ES :w:
304 mm< $05.35. 08:. mmoq mm< $052022. 0:5.
mommoq
5825 32:382.
198— Eomummﬁuh +5.?— 830

 

.vu a veto e553 « 9802 .350: So 2..m 3e 2o 6 3:253: 52.076 mmﬂw 33332.5 35:30: v5 556 owe—are:
.38 away 3:259:00 82m Te: 88 2-m.m$ moemwoawm :8 .20 m5»:— R:o£§.: :5 .630 no 9:688: maﬁa: 05 55
2.83083 mammo— eqw AmoﬁmoEw: 20:3 23.8 X. 3:8 mm .mm<v 5323 08.96 owmwouwwe £3533: :93. .083 Even .2.~ 033.

50

Table 2.2. Bulk densities and tensile strengths of air-dry soil aggregates (12.5-16 mm
dia.) ﬁom conventional tillage (CT), no-tillage (NT), and never-tilled successional
grass (NeTSG) treatments at 0-5 and 5-15 cm depths. Means :1: standard error.

 

 

 

Bulk density
Treatment Tensile strength
Whole aggregate Inner region Whole aggregates
n Mg m” n Mg m‘3 n kPa
Licmﬁsmh
CT 17 1.69 (0.02) 15 1.80 (0.02) 39 160 (11.1)
NT 16 1.67 (0.02) 13 1.78 (0.02) 49 107 (6.8)
5_:_15_Qm_depth
CT 16 1.79 (0.02) 15 1.63 (0.02) 45 177 (9.8)
NT 16 1.70 (0.01) 15 1.69 (0.01) 48 126 (7.2)
NeTSG 16 1.62 (0.02) 7 1.51 (0.05) 35 111 (8.5)

 

51

Table 2.3. Comparison of regression coefﬁcients (a, b, and c) of the calibration equations
of the form (ASF = a + bt + cWi) for aggregate surface ﬁaction (ASF), and the
variables time (t) and aggregate initial weight (Wi) for soil aggregates from
conventioml tillage (CT), no-tillage (NT), and never-tilled successional grass (NeTSG)
treatments at 0-5 and 5-15 cm depths.

 

 

Treatment a b c R2 ’ ASFm n
(%)
Limulsnth
CT 13.68 2.78” -2.05“ 0.813 77.05 15
NT 16.92 3.25”. -1.94 0.882 88.80 13
MW
CT 12.88 2.47". -1.87 0.88t 81.13 15
NT 3.77 1.83m 0.69 0.803 67.98 14
NeTSG 37.17 19.85 -9.09 0.60 98.90 7

 

... Signiﬁcant (single-tail t-test) at the p < 0.05 and p < 0.01 levels, respectively, for
evaluating the uselfulness of the variables t and Wi to predict ASF.

*Coeﬂicient of determination.

‘ Signiﬁcantly different from zero (p < 0.05, F test).

52

.8685 2: 3:832 can .233 £956 2: wEwEEo .m<m 2: wﬁgﬁm .3 83on? :8 Eat 33.38 on :8 £93— 856? A32 :3 .u

828$ ESE m<m $5329 05 E m< .033 0388 05 E 322568 98 860:8 «38mm 05 5 Eamon»... 08 8.03.3: =8 vouch.— .8620

05 .«o a9 2: no @2528 can .908 a 8 385.8 383 a .3 non—£388 m_ 30c an we Eaten 25.98 wasp—Bow < .5950 Samoan
038,28.— oﬁ E 383 2a 83m2ww< damage Am<mv 2:85 Samoan :8 32:95 05 .«o 5355852 £88235 ._.N azur—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

uoaom BEE.
“M.H.,...“ \
“28:8 \1
vacuum
6 ON -
“38.20 owmwouwwax \1v[
0 035%,: : :1

 

 

 

 

 

 

 

.882

 

 

 

 

 

 

 

 

 

 

 

:

 

 

 

53

Total carbon (g kg'l)

Total carbon (g kg‘l)

 

40%
354

 

Conventional tillage
[:1 No-tillage

 

 

 

a

 

a C

B

a a :1 a
AB AB A

Bulk Outer

 

 

 

 

 

 

 

 

O-Scm
a a
B C
S-IScm
,5. . a
a
B
B B
Trans. Inner

Figure 2.2. Total carbon contents of bulk soil samples, outer and transitional layers,
and inner regions of soil aggregates ﬁorn conventional tillage (CT) and no-tillage (NT),
at 0-5 and 5-15 cm depths. Values followed by the same uppercase letter within a tillage
treatment and between soil and aggregate samples, and by the same lowercase letter
within soil and aggregate samples and between tillage treatments are not signiﬁcantly
diﬁ‘erent at p > 0.05 according to Fisher’s LSD mean separation test.

54

 

 

 
  

  

 

 

 

b %m d i m
_ _ a//////////////am . E. m

 

 

A73 3 comet: 38.

55555555
333333

A 73 us 8&8 .83

 

 

 

    

 

 

 

 

 

55

=4.

Figure 2.3. Total carbon (a), total nitrogen (b), C:N ratio (c), and 6'3C ratio (d) of bulk

soil samples, outer and transitional layers, and inner regions of soil aggregates from
the never-tilled successional grass (NeTSG) treatment, at 5-15 cm depth. Means :t

standard error; n

 

 

 

 

 

 

_1 9 d C Conventional tillage 0 - 5 cm
0 No-tillage
'20 1 a
_ C a
E -21 a BC
:5 -22 4 B 33 §
.9 a
a 43 ~ i AB 15.
20 -24 ~ a
°° -25 d a a A
A
-26 - E g E
-27 ..
-28
_19 _ 5 - 15 cm
-20 4 a a
,5. -21 — A A
o\° a
\o/ '22 _l 2 i a a a a A
g -23 1 E A A A A
U i
20 -24 — i E 1‘
oo -25 _
-26 _
-27 ..
-28
Bulk Outer Trans. Inner

Figure 2.4. 6'3C ratios of bulk soil samples, outer and transitional layers, and inner
regions of soil aggregates ﬁom conventional tillage (CT) and no-tillage (NT), at 0-5 and
5-15 cm depths. Values followed by the same uppercase letter within a tillage treatment
and between soil and aggregate samples, and by the same lowercase letter within soil
and aggregate samples and between tillage treatments are not signiﬁcantly different
at p > 0.05 according to Fisher’s LSD mean separation test.

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 O _ Conventional tillage a 0 _ 5 cm
° [:3 No-tillage A
HA a
' AB
on
in 1.5 4 a a a
x; A b A a A a
o A B
an B
.5 1.0 ‘
E
.53
o
E" 0.5 —
0.0 - r ____=.
2.0 _ 5 - 15 cm
:40 a a a a
g A A a A a A
8 A A
8"
1::
'8
S
o
[...
0.0 ~ ‘
Bulk Outer Trans. Inner

Figure 2.5. Total nitrogen of bulk soil samples, outer and transitional layers, and inner
regions of soil aggregates ﬁom conventional tillage (CT) and no-tillage (NT), at 0-5 and
5-15 cm depths. Values followed by the same uppercase letter within a tillage treatment
and between soil and aggregate samples, and by the same lowercase letter within soil
and aggregate samples and between tillage treatments are not signiﬁcantly different at
p > 0.05 according to Fisher’s LSD mean separation test.

57

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35 4 Conventional tillage a 0 - 5 cm
:2] No-tillage A
a
30 - A
25 4
.2
*5 20 —
Z
.. _ a a a
U 15 B B C
10 d
5 _
0 - _ L . __ ”Q
35 r 5-15cm
30 _.
25 r
.2
no a
o 15 — :1 g A A AB A g
10 - T
5 a
0 J
Bulk Outer Trans. Inner

Figure 2.6. C:N ratios of bulk soil samples, outer and transitional layers, and inner regions
of soil aggregates from conventional tillage (CT) and no-tillage (NT), at 0-5 and 5-15
cm depths. Values followed by the same uppercase letter within a tillage treatment and
between soil and aggregate samples, and by the same lowercase letter within soil and
aggregate samples and between tillage treatments are not signiﬁcantly different at
p > 0.05 according to Fisher’s LSD mean separation test.

58

 

80- 0-5cm

60+
40~
204

 

 

 

 

 

-20 _
-40 ~
-60 _
-80 _

 

 

 

New C3-derived C (%)
o
l- bl

l———-l

 

 

80 e Conventional tillage 5 - 15 cm
[:2] No-tillage
60 4

404
20-

 

 

 

 

 

 

 

     
 

-20 -1
.40 «
-60 - _.
-80 ._

 

 

 

 

 

l——~—i

New C3-derived C (%)
o
L

 

 

 

 

Bulk Outer Trans. Inner

Figure 2.7 . Percentage of new C3-derived carbon in bulk soil samples, outer and
transitional layers, and inner regions of soil aggregates from conventional tillage
(CT) and no-tillage (NT), at 0-5 and 5-15 cm depths.

59

CHAPTER 3
MODIFICATIONS OF DISTRIBUTION AND COMPOSITION OF ORGANIC

MATTER IN WATER-STABLE AGGREGATES OF UNDISTURBED AND
CULTIVATED AGROECOSYSTEMS

INTRODUCTION

Organic matter content in soils is the result of a balance between rates of its loss
through biological oxidation and its storage aﬁer decomposition of plant shoot and root
materials incorporated into the soil. Accumulation and preservation of soil organic matter
(SOM) is generally aﬂ‘ected by management practices that result in larger amounts of
biomass such as crop choice, fertilization levels, and climate conditions that enhance
residue decomposition and SOM transformations (Beare et al., 1994; Reicosky et al.,
1995). Disturbed or cultivated soils generally contain greater proportions of SOM in the
ﬁner aggregates and particle ﬁ'actions indicating that macroaggregates are more
susceptible to disruptive forces imposed upon the soil during cultivation (Angers et al.,
1993; Cambardella and Elliott, 1993). Organic C of more recent deposition has been
suggested to be the temporary binding agents between microaggregates and the major
SOM pool depleted upon cultivation of native grassland soils, resulting in 25-50%
reduction in macroaggregation (Elliott, 1986; Gupta and Germida, 1988). Reduced- or
no-tillage management systems have been shown to improve soil aggregation (Carter,
1992; Beare et al., 1994) and reverse the decline in SOM as a result of slower biormss

decomposition and mineralization (Reicosky et al., 1995; Lal and Kimble, 1997).

60

Restoration or maintenance of optimum levels of soil productivity requires a better
understanding of how soil management affects the stabilization and storage of SOM, soil
biology and aggregation processes. In recent years, the potential detrimental eﬁects of
excessive tillage operations on soil properties and agricultural sustainability as they
inﬂuence environmental quality have become a concern No-tillage, reduced tillage, and
other conservation tillage practices have been adopted as alternatives to the conventional
intensive plowing practices. Beneﬁts of conservation tillage practices include maximization
of surface residue retention and enhancement of OM levels near the soil surface, resulting
in reduced evaporative losses, increased rainfall inﬁltration, reduced water runoff and
water erosion hazard (Blevins et al., 1998). Conservation tillage affects C cycling by
increasing soil organic carbon content and its distribution within the soil proﬁle, among
soil aggregate fractions, and the rate of SOM turnover by soil biota (Blevins et al., 1998).
In the long run, conservation tillage practices also promote reduction in energy, labor, and
production costs (Blevins and Frye, 1993). It has been stated that no-tillage management
may be the single most promising technology that can be applied to agroecosystems that
combines economic viability through high crop productivity and increased C storage in
soil, due to lower CO2 emission to the atmosphere (Dick and Durkalski, 1998).

Several studies have been conducted to evaluate the effects of cropping and
conservation tillage mnagement on physical properties. Aggregate stability and size
distribution show signiﬁcant short-term variations with cropping and tillage practices even
during a growing season. Soil aggregation is commonly measured with a wet-sieving

procedure (Kemper and Rosenau, 1986), which provides information on the potential

61

disruption and ﬁ'actionation of aggregates under the combined impact of water and
abrasion. This technique is often used for evaluating the effects of conservation tillage
practices on soil quality because of its relationship to resistance of soil to erosion. Physical
fractionation according to particle size has been used to study the distribution,
composition and stabilization and turnover of SOM as related to its location in the soil
aggregates (Elliott and Cambardella, 1991; Puget et al., 1995; Angers and Giroux, 1996).
Isotopic 13C ﬁ'actionation has been successfully used to investigating the fate of
recent inputs of OM to soil and estimate SOM dynamics associated with whole soils and
aggregate size ﬁ'actions (Balesdent et al., 1990; Puget et al., 1995; Angers and Giroux,
1996; Jastrow et al., 1996). Physiological diﬁerences in enrichments of 13C by C3 and C4
plants allows one to distinguish recently-incorporated SOM by the latest crop ﬁom the
older SOM of previous vegetation provided that diﬂ’erent plants are cultivated in a
sequence (Balesdent and Mariotti, I996). The 13C/ 12C ratio of SOM corresponds closely
to that of the original vegetation ﬁom which it was derived. Based on these C isotope
ratios, organic matter can be either C4-derived (high, or less negative 6‘3C, e.g. -12%o) or
C3-derived (low, or more negative, 6'3C e.g. -28%o) (Boutton, 1996). Relatively little
information is available on the combined effect of tillage practices and cropping variations
on the organic matter distribution and ﬁactionation of soil aggregates. In this study, the
contribution of contemporary C3- and C4-derived tilled crop C in stable aggregates of a
loam soil originally under C3 vegetation was investigated. The amount and fate of recently
deposited C in soil aggregates of an undisturbed grass soil converted to no-till C4 cropland

and its effects on aggregate stability were also assessed.

62

MATERIALS AND METHODS

Study site

Soil samples were collected from the Agroecosystems Interactions Site established
in 1986 at the WK. Kellogg Biological Station (KBS), located in southwest Michigan
(42°24' N, 85°24' W). The soils in the experimental site are well drained [Typic
Hapludalfs either Kalamazoo (ﬁne-loamy) or Oshtemo series (coarse-loamy)], mixed,
mesic developed on glacial outwash (Whiteside, 1982). Mean annual temperature is 9° C
and precipitation averages 920 mm year’l (URL:http://1ter.kbs.msu.edu/description.html).
The area was ﬁrst cultivated during the last decades of the last century and have been
cropped with maize for most of that period (Robertson et al., 1993). Selected soil
properties of the surface horizons are shown in Appendix A1.

The experimental design is a randomized complete block with four treatments and
four replicates. The treatments are: two tillage (conventional and no-tillage) by two
nitrogen fertilization levels (0 and 168 kg N ha"1 yr“). The no-tillage areas have not been
tilled since the spring of 1985. The treatments were applied to large plots of 40 by 27 m
(Appendix A2). Two additional treatments were established on a directly adjacent site
which had not been tilled alter clearing of the native oak-hickory (Quercus-Carya spp.)
forest in the late 1950's. The present successional vegetation in these plots is dominated by
herbaceous perennial dicots and grasses, including T rifolium pratense, Rosa spp.,
Medicago Iupilina, AchiIIea millefolium, Bromus spp. and Poa spp. (Robertson et al.,

1993). The additional treatments were established in a randomized complete block design

63

with four replicated plots of 15 by 40 m with never-tilled successional grass (N eTSG) and
never-tilled successional with maize (N eTSM) treatments. Neither treatment in the never-
tilled area received any type of fertilizer. In the NeTSG treatment vegetation is clipped
once each year (kite fall) with no removal of residues. The CT treatment involved
moldboard plowing to a depth of 20 cm each spring, followed by secondary tillage with a
disc cultivator as needed. For both CT and NT, annual ﬁeld operations included pre- and

post-emergence herbicide treatment, and post-emergence N fertilization.

Sample collection

Soil samples were obtained in the spring of 1997 ﬁ'om the N-fertilized CT and NT
treatments, and ﬁom the NeTSG and NeTSM treatments. Four subsamples were taken
with a ﬂat shovel in each of the four replicated plots per treatment, from the 0-5 and 5-15
cm layers. Samples were pooled per plot, gently broken apart along their planes of
weaknesses, air-dried and gently manually sieved through a 6.3-mm sieve. Subsamples of

air-dry soil were weighed and oven dried at 105 °C for moisture determinations.

Aggregate size distribution and stability

The size distribution of aggregates was measured on air-dry soil samples by wet
sieving (Kemper and Rosenau, 1986) using an apparatus with four stacks of sieves to
collect six aggregate size classes ﬁ'om each sample: 4-6.3, 2-4, 1-2, 0.5-l, 0.25-0.5, and
0.106-0.25 mm, plus microaggregates and colloidal material < 0.106 mm. One-hundred

grams of soil were spread across the surface of the 4-mm sieve and directly immersed in a

64

water bath of distilled water (Angers and Giroux, 1996). Each sieve stack was immersed
in separate containers and the water level was adjusted so that the aggregates on the upper
sieve were just submerged at the highest point of oscillation. The oscillation rate was 33
cycles per minute, the amplitude of the sieving was 3.5 cm, and the sieving duration of
each replication was 20 min.

After wet sieving, the individual water baths were held stationary and
sedimentation was allowed to occur for a period of 48 hours to collect the < 0.106 mm
fraction. F locculation of suspended particles was aided by pouring 20 ml of CaCl2 0.5 M
in each container. Floating organic debris and/or pieces of crop residues were excluded
ﬁ'om the analysis. The soil fractions were washed into beakers, oven-dried at 60°C for 24
hours, and weighed. Samples on each sieve were split. About one-half of the soil in each
ﬁ'action was shaken and dispersed in a solution of Na-hexametaphosphate (5 % w/v) for
20 minutes. They were then passed through a 53-am sieve to collect sand-size mineral
grains, oven-dried and weighed. The distribution of sands accounts for much of the
differences in the conrposition of aggregate size ﬁactions reported in the literature (Elliott
et al., 1991; Cambardella and Elliott, 1993; Beare et al., 1994). The reason is because
some of the sand grains are retained on the ﬁner sieves and included in these size ﬁactions
of aggregates during the sieving process. Thus, the mass of each size fraction was
calculated on a sand-ﬂee (> 53 am) basis to normalize the expression of C distribution in
the size fractions. The mean-weight diameter (MWD) is an index of aggregate stability and

was calculated according to the summation equation:

65

n
MWD=Zx,.w, (1)
l=l

where x,- is the mean diameter of the aggregate size ﬁaction and w, is the proportion of the
total sample weight present in the size ﬁaction (Kemper and Rosenau, 1986). The
remaining halfof the wet-sieved soil in each fraction was ground in a ball mill (< 250 am)

before analysis.

Total carbon and stable isotope analyses

Ground samples were used for the determination of total C and 13C contents.
Samples were then carefully weighed and packed into tin capsules, and wrapped into
pellets so that total soil C and (WC could be determined by combustion on a Isotope Ratio
Mass Spectrometer Model 2020 (Europa Scientiﬁc, UK). The 6'3C is calculated from the
measured l3C/‘zC ratios (R) of the sample and standard gases and expressed relative to the
international standard PDB (as 5'3C, %o). The proportion of C derived ﬁ'om crops (x) was

calculated by the modiﬁed mixing equation (Balesdent and Mariotti, 1996):
x = (5c: ' 6r: ) / (5c' 6rs) (2)

where 6,, = 6‘3C value of sample (isolated fraction from cropped soil), 6,, = 5'3C value of
the reference soil of the corresponding ﬁaction, and 5, = measured 6'3C value of crop

residues. The never-tilled successional grass plots (NeTSG) were used as the reference to

66

calculate the proportion of C derived from crops (both C4 and C3 plants), in each size
fraction, in CT and NT treatments, and the proportion of maize-derived C in the NeTSM
treatment. For CT and NT soils, the pr0portion of C derived from crops was calculated
assuming a mixture residue return of two parts maize to one part non-maize crops (mostly
soybeans) over the lOO-year period these plots had been under cultivation, before the
experiment began. Maize residues return more biomass to the soil than soybean, and
soybean residues decompose faster because of their lower C:N ratio. The residue mixture

was calculated to have an average 6'3C of -17.1%o (Collins et al., 1997).

67

RESULTS AND DISCUSSION

A ggregate size distribution and stability

Long-term cultivation resulted in substantial changes in aggregate size distribution
and a reduction in aggregate stability. There were more stable macroaggregates ranging
from 1 to > 4 mm in the never-tilled successional grass (NeTSG) (Fig. 3.1) than in the
cultivated (CT and NT) soils (Fig. 3.2). High levels of SOM and the continual presence of
roots in grassland soils promote the conditions necessary for microbial and invertebrate
reworking of soil material which develop aggregates with enhanced water stability. Under
these conditions, there would be little or no exposure of the organic material to
decomposition due to absence of disruption by tillage (Cambardella and Elliott, 1993). A
favorable balance of physical forces, such as enmeshment by pressure exerted by growing
roots (Tisdall and Oades, 1982; Douglas and Goss, 1982) and localized drying caused by
root water extraction (Kay, 1990), has been associated with enhanced water-stable
aggregation of undisturbed soils. The effect of roots in physically enmeshing ﬁne particles
of soil into macroaggregates is persisting and usually goes beyond the root death (Tisdall
and Oades, 1982).

When the never-tilled soil was cropped with a no-tillage maize (NeTSM) the
percentage of aggregates > 4 mm in the top 5 cm was higher than in the reference plot
under grass (Fig. 3.1). Reductions in aggregate stability upon cultivation of native soils are
more pronounced in soil macroaggregates whereas the stability of microaggregates usually

remains unchanged (Tisdall and Oades, 1984; Oades, 1984). The increase in the

68

percentage of aggregates > 4 mm under NeTSM may indicate an additive effect of maize
roots on macroaggregate stabilization as roots are thought to play an important role on
aggregation by supplying large quantities of organic material to soils. However, the C
concentration at the depth of 0-5 cm was up to 17% lower in this ﬁaction (> 4 mm) after
12 years of slot maize cropping (Fig. 3.3a).

In recently tilled sod soils changes in aggregate stability often occur before
signiﬁcant declines in total SOM can be detected (Angers et al., 1992; Haynes and Swift,
1990). In the present study, the slot planted maize showed reductions in C levels but
relatively small changes in aggregate stability after 11 years. Based on their ﬁndings during
the ﬁrst four years after the commencement of the present ﬁeld trial, Reinert et al. (1990)
predicted that cropping the never-tilled soil to maize, even when under no-tillage
management, should decrease its stability on a long-term basis. They concluded that the
soils under long-term cultivation were much more sensitive to cyclic variations in weather
ﬂuctuations, i.e., rainfall patterns and associated soil water contents, than undisturbed
soils. Cyclic variations in aggregate stability were more dramatic in the CT treatment
because of the tillage disturbance, whereas there was a considerable similarity between the
cycles in NT and NeTSG soils (Reinert et al., 1990). Annual cycles in aggregate stability
(especially > 4 mm size) followed a trend towards structural stability degradation from fall
to early spring and a restoration action starting early spring and reaching the maximum in
the late summer (Reinert et al., 1990). The cyclical variations observed appeared to be
associated with soil temperature and moisture conditions favorable for biological activity.

In view of these data collected earlier in these plots, it can not be ruled out that the higher

69

percentage of aggregates > 4 mm in the NeTSM, as indicated by an increased MWD value
(Table 3.1), is little more than the result of temporal variations in aggregation than
inﬂuenced by maize root activity.

Overall, there was a greater percentage of water stable aggregates soil in the small
macroaggregates and in the microaggregate size classes (< 0.5 mm) in the cultivated soils.
Lower percentages of soil in the < 0.106 mm size-class were associated with less
disturbance by tillage with CT showing the highest and NeTSG the lowest values. While
tillage intensity is an important factor in aggregate disruption, the relatively lower residue
cover (visual observation), leading to intensive drying/wetting and ﬁeezing/thawing, may
have contributed to the destruction of aggregates into smaller particles. It seems that for
these soils the limit of macro and microaggregation would be somewhere between 0.5 and
0.25 mm. This is because above and below this size the trends in size distribution among
the treatments are in the opposite direction to intensity of cultivation. In general, the
undisturbed (NeTSG and NeTSM) soils had a proportionally higher percentage of soil
aggregates in the fractions > 0.5 mm (Fig. 3.1), whereas CT and NT rmnagement resulted
in more soil in the ﬁ'actions < 0.5 mm (Fig. 3.2). Thus, macroaggregates of rmdisturbed
soils were more stable than of cultivated soils (CT and NT). As discussed by Elliott
(1986), the large shift in fractions distribution greater and smaller than 0.3 mm, and the
small amount of material smaller than < 0.1 mm even under disruption conditions by
slaking would support the model of soil structure proposed by (Tisdall and Oades, 1982)
in which macroaggregates are formed by an assemblage of highly stable microaggregates.

The eﬂea of slaking is to disrupt the soil due to internal forces and results in the

70

production of particles of a more fundamental nature compared to a less aggressive pre-
treatment of samples (Elliott, 1986). It has been suggested that upon wetting, large air-
dried soil aggregates (>1 mm) are susceptible to disruption by slaking breaking down into
aggregates of 0.25 to 0.50 mm in diameter (Angers, 1992). A greater shift in the
aggregate size distribution of a sample towards the smaller size classes upon disruption by
rapid wetting indicates its lower aggregate stability (Kemper and Rosenau, 1986).

The quantity of macroaggregates resistant to slaking was substantially higher under
the undisturbed long-term grass treatments as it can be seen ﬁ'om the small proportion of
microaggregates at both depths (Fig. 3.1). Comparing native pasture soils to adjacent soils
continuously cropped for > 60 years, Skidmore et al. (1975) concluded that the structure
of the sod soil was much more stable than of the tilled soil, which was corroborated by a
large number of roots and cavities in aggregates of sod soils. In the present study, the
effect of NT on macroaggregation (> 0.25 mm), as compared to CT, was clear at both
depths (Fig. 3.2). Greater macroaggregation under NT than under CT was only offset by a
decrease in the 0.25-0.5 mm class at the 0-5 cm depth and similar values for > 4 mm
aggregates at the 5-15 cm depth. The experimental area had been under conventional
tillage for over 100 years before the CT and NT experimental plots were established.
Thus, as opposed to several studies in which CT and NT plots are started out of
undisturbed soils, the aggregate stability data reported here indicate the resilience capacity
of this soil to restore macroaggregation under NT management, over a relatively short
time span. In general, a decrease in aggregate size and stability (i.e., reduction in stable

macroaggregates) with increasing intensity of cultivation is expected and fairly well

71

 

documented in the literature (Oades, 1984; Blevins and Frye, 1993; Beare et aL, 1994;
Franzluebbers and Arshad, 1996; Bajracharya et al., 1997; Monreal and Kodam, 1997).
Aggregate stability of the soils was also evaluated based on their MWD which is
an aggregation index that takes into account the different size particles which breakdown
and are separated during wet sieving. The MWD exhibited a pattern similar to
macroaggregation except for the 0-5 cm depth, where the value for the NeTSM soil was
higher than for NeTSG (Table 3.1). No-tillage had the greatest impact on MWD at the
depthofO-S cmwithanincrease up to 53% comparedwithCT and smallerimpact atthe
depth of 5-15 cm (mean increase of 27% compared with CT). The highest values,
however, were associated with the undisturbed treatments. The mean weight diameter
decreased on cultivation from 2.02 mm in the NeTSG to 0.75 mm (CT) and 1.16 mm
(NT) after over 100 years of cropping. This indicates a decrease in the MWD by as much
as 63% under CT management. Mean weight diameter of NT soils was 55% higher than
that of CT, indicating the beneﬁcial effects of NT practices after 11 years. Studying a ﬁne
sandy loam, Carter (1992) observed tlmt the MWD under no-tillage was 43-53% greater
than under conventional tillage at the depth of 0-5 cm after a 3-5 year period. According
to the classiﬁcation of Le Bissonnais (1996) the MWD values for NeTSG, NeTSM and

NT correspond to very stable soils (MWD > 2.0 mm) and those for CT to soils with low

stability (MWD < 0.8 mm).

72

Total carbon contents in aggregate size ﬁ'actions

Total C concentrations in aggregate size classes were strongly inﬂuenced by the
different proportions of sand in each size class, which agrees with previous reports (Beare
et al., 1994; Elliott, 1986; Puget et al., 1995). Therefore, to describe interaggregate
differences in organic C concentrations, as well as the effects of the treatments and soil
depth, all total C values were normalized on a sand-free basis. Cultivation inﬂuenced the
relative distribution of C in different aggregate size classes. For all aggregate size-classes
and depths, C concentrations were always higher for the never-tilled treatments (N eTSG
and NeTSM) (Fig. 3.3) than for the cultivated soils (CT and NT) (Fig. 3.4) and higher in
the 0-5 cm layer. Higher organic matter levels in undisturbed, steady-state soils are
expected because little vegetation is removed, wetting and drying are slow, and oxidation
is at minimum due to the absence of periodical soil mixing (Dorrnaar and Carefoot, 1996).
In addition, root and crown tissues production is much greater for native grasses than for
cultivated crops, making up a higher proportion of net primary productivity (Sala et al.,
1988)

When total C distribution in the different size ﬁactions of NeTSG (Fig. 3.3) and
CT (Fig. 3.4) soils are analyzed in conjunction with the distribution of water-stable
aggregates (Fig. 3.1 and Fig. 3.2), there seems to be an agreement with the observation
that cultivation results in a smaller proportion of soil containing ‘high-nutrient
macroaggregates’ and larger proportion of soil containing ‘low-nutrient microaggregates’
(Elliott, 1986). Gupta and Germida (1988) and Cambardella and Elliott (1993) also

reported that C concentrations were higher in aggregates of native sod than in those of

73

long-term cultivated grasslands. Organic matter losses are rapid and usually exponential,
with rapid declines during the ﬁrst decades, followed by smaller annual losses, ﬁnally
approaching a new equilibrium characteristic of the particular system considered. For
example, the A horizons of prairies have lost 35 to 50%, based on concentration, or 15 to
30%, based on mass, of their OM since they were broken 100 years ago (Dorrnaar and
Carefoot, 1996).

Carbon concentrations were generally equal to one another in the ﬁve larger
aggregate size classes (> 0.25 mm) and were always higher than in the smaller sin class
(< 0.106 mm) (Fig. 3.3 and Fig. 3.4). Never-tilled grass (NeTSG) soil had higher C
contents in all ﬁactions compared with the never-tilled grass under maize (N eTSM).
Absolute differences in C between size classes were relatively smaller in NeTSM
compared to NeTSG (Fig. 3.3), and in CT compared to NT (Fig. 3.4), at the depth of 0-5
cm. In the deeper layer, the absolute differences were smaller in NeTSG.

There was a considerable decrease in the C concentrations of NeTSG due to maize
cropping (NeTSM) after 11 years, although the decreasing effects of maize in C were not
always coupled with reduction in macroaggregate structure, as discussed earlier. Changes
in C contents in the several size fractions ranged ﬁom 10 to 18 g C kg" soil ﬁ'action at the
depth of 0-5 cm and ﬁ'om 8 to 15 g C kg" at the depth of 5-15 cm (Fig. 3.3). An average
across all water-stable aggregate ﬁactions indicated that C contents under maize
decreased by 1.2 g kg" yr" (0-5 cm) and 0.9 g kg" yr" (5-15 cm). It is important to
mention that in the NeTSM treatment maize has been planted after only a mowing late in

the previous fall, without any fertilization. The biomass production has been very low and

74

the residues returned to soil (mainly ﬁom roots) probably contribute too little to the pool
of organic material. Because the maize is slot planted, one possible explanation for
decreased SOM under NeTSM is the disruption of soil aggregates and increased aeration
by growing maize roots. Even if small, this disturbance still may have provided oxygen
supply and accessibility of new easily oxidizable material, resulting in further substantial
biological oxidation of the residual C in the system (Reicosky et al., 1995).

In order to balance the oxidation of the original SOM and maintain a new level
closer to the original level, large additions of SOM from decomposing residues should be
added to the soil. In an experiment in Indiana, Barber (1979) concluded that in cropping
practices with no incorporation of residue over a period of years following a previous
practice where the residue was returned to the soil, the SOM level would decrease until it
comes to an equihbrium with the new practice. He observed that after 11 years of
continuous maize with no residue return, the SOM level decreased to by 10 percent of its
original level. A higher amount of residues returned to the soil under NeTSM would also
oﬂset the disturbance eﬂ‘ects on physical properties such as the degree of aggregation.
Reid et al. (1982) reported that detrimental effects of maize cropping on soil aggregates
were directly related to maize root mucilages, although their explanation has been
contested (Pojasok and Kay, 1990). During a 28-day growth study of maize, Helal and
Sauerbeck (1989) observed a short-term positive effect on aggregation, which they
attributed to enhanced root exudation and microbial biomass. However, 60 days after
harvest, they veriﬁed that OM levels and aggregate stability were lower in the maize

planted soil than in the control. Despite the fact that for most crops, root decay

75

contributes to the total amount of OM in the soil (Goss, 1987), the beneﬁcial effects are
conﬁned to the topsoil and rapidly disappear when additional organic surfaces are exposed
to microbial attack in root-disrupted soil aggregates (Helal and Sauerbeck, 1989). Root
released materials are more readily available to microbes than are residues ﬁom even the
previous crop and there is less breakdown of these materials in the presence of living roots
than in fallowed soils (Reid and Goss, 1982). Organic matter content in soil aggregates
seems to be dictated by the ability of structural units to pack small-size organic ﬁagments;
it is largely inﬂuenced by the activity of soil microorganisms and plant roots in undisturbed
soils and by tillage intensity in cultivated soils (Monreal and Kodama, 1997). These
authors veriﬁed that aggregates ﬁ'om a forest soil had greater amounts of plant debris,
SOM, larger diameter particles, and higher water stability than those ﬁom a cultivated site.
The lower C contents of CT and NT soils (Fig. 3.4), compared to the NeTSG
plots (Fig. 3.3), is a result of their intensive cultivation since the end of the last century.
The average reduction in C with cultivation, across all the measured aggregate size
ﬁactions in the 0-5 cm layer of CT soil, was in the order of 33 g kg". For the 5-15 cm
layer, C reduction averaged 16 g kg". This illustrates a natural decline in C content of
cultivated soils, particularly when small amounts of crop residue are returned to the soil.
Nearly all the differences in aggregate C contents between CT and NT treatments were
found in the top 5 cm of soil (Fig. 3.4), with an average net increase of 6.6 g kg" in NT.
Coleman et al. (1994) attributed similar effects to the lower abundance and turnover of

macroaggregates in the top soil surface of CT as opposed to NT soils.

76

The lower C contents in the top 0-5 layer of our CT treatment reﬂects a
redistribution eﬂ‘ect of organic matter by plowing. Beare et al. (1994) reported that
aggregates of CT had a lower capacity to protect SOM than aggregates of NT in the
surface 5 cm of a sandy clay soil. Frequent disruption of aggregates are expected to
reduce the quantity of macroaggregate-protected SOM and increase the quantity of
unprotected SOM in proportion to the total aggregate SOM (Coleman et al., 1994).
Reducing tillage intensity and increasing surface residue cover inhibit loss of organic
matter from soil. Twenty years of bare-fallow, stubble-mulch, and no-till management on a
grassland soil of Nebraska reduced the surface total C by 28, 18, and 12% of the original
level (Cambardella and Elliott, 1992). Plant residues decompose more slowly on the soil
surfacethanwhentheyareincorporated intothe soil,mainlyasaresultofchangesinsoil
temperature and moisture, and availability and location of substrate. At the 5-15 cm depth,
except for the > 4 mm aggregate size, there were no signiﬁcant effects of tillage on
organic C contents (Fig. 3.4b). The higher values of C in macroaggregates > 4 mm under
CT probably resulted ﬁ'om the redistribution or deep burial of C-enriched surface
aggregates by tillage, as indicated by similar concentrations of C at both depths in CT. The
reduction in C with depth was greater in NT and would probably be even greater if a more
stratiﬁed sampling of the top 5 cm had been performed. From whole soil samples taken in
the spring of 1996, it was observed that after 10 years ofCT and NT nmnagement at this
experimental site, C concentrations in NT (13.2 g kg") were somewlmt greater than in CT
(11.5 g kg") (Santos and Smucker, 1998). While repeated cultivation of soils enhances

SOM decomposition and changes the composition of the residual organic matter, the

77

 

 

W;a—EL"P ' ‘

effects of crop rotations and reduced tillage on SOM usually beconre signiﬁcant only after
several years of their adoption. Several studies have shown greater amounts of C in soils
that have been under reduced or no-tillage for long periods compared with soils under
moldboard plow tillage (Havlin et al., 1990; Beare et al., 1994; Franzluebbers and Arshad,
1996). Blevins et al. (1983) reported 37 and 12% more organic C (at 0-5 cm depth) under
NT and CT maize, respectively, than in an unfertilized soil.

Measurements of living ﬁactions of SOM such as soil microbial biomass C
(SMBC), have indicated early changes in SOM for relatively short periods before these
can be detected by chemical analysis (Powlson et al., 1987). Although SMBC evaluations
were out of the scope of the present study, measurements taken twice during 1996
(Sissoko and Smucker, 1998), indicated that the higher the intensity of cultivation the
lower the SMBC values will be. For samples taken in November, SMBC values at 0-20
cm were consistently higher in NeTSG (310 :t 16 pg C g" soil; where n = 3; i standard
deviation ) and NT (210 d: 12 pg C g" soil), than in CT (175 :1: 22 reg C g" soil). Soil
microbial biormss measured in May in CT and NT averaged 32% greater than in
November, whereas seasonal variation in NeTSG was only 13% (Sissoko and Smucker,
1998). The decrease in SMBC in November may have been the result of a long drought
period dining the 1996 growing season. Nevertheless, the observed decrease in SMBC
with CT compared to NT is similar to that reported by Salinas-Garcia et al. (1997). When
the NeTSG soil is compared, the results are similar to that reported by Follett and

Schimel (1989) who found that SMBC decreased in the order of sod soil > NT > CT.

78

6’ 3 C values in aggregate size fractions

The 6'3C of the whole soil under the present successional vegetation, cleared ﬁom
forest in the 1950's, shows a mean value of -25.7 %o (Paul et al., 1998), indicating that the
organic matter in this soil was derived mainly ﬁom C3 vegetation. This 6'3C value is
rehtively similar to that in the surface layer of the native vegetation of deciduous oak-
hickory forest (-26.1%o) (Paul et al., 1998). The 613C values (:1: standard deviation, n = 4)
of random grass plant samples were -27.6 %o (:t 0.41%o) and -27.6 (:1: 0.31%o) for shoots
and roots, respectively. This indicates a slight 6'3C enrichment (1.9 %o) of the SOM. Since
plant residue decomposition in soils is a microbial-mediated process, a discrimination
against the heavier l3C could have resulted in an increase in the 6‘3C value of the soil
(Balesdent and Mariotti, 1996). Also, separate constituents of plant residues are
isotopically distinct from whole plant tissue and decompose at different rates. More rapid
decomposition of cellulose and hemicellulose (1-2 %o enrichment in 13C) than lignin (2-6
%o depletion in 13C) ﬁactions could deplete 13C enrichment of recently added plant
material as it is incorporated into the SOM pool (Boutton, 1996; Huggins et al., 1998).
The presence of some C4 plants in the pool of grass species in the successional vegetation
could be another explanation. Several other causes for this enrichment are discussed in
detail by Boutton (1996).

The 6‘3C values of whole soil samples of CT and NT, taken at the depth of 0-5
cm, were -22.8 :t 0.27 %o (means 1 standard error, n = 4) and -22.2 d: 1.03 %o,
respectively. For the 5-15 cm samples, the values were -23.6 i 0.28%o for CT and -22.1 d:

0.70%o for NT (Santos and Smucker, 1998). Small differences in 6‘3C were observed

79

 

between the several size ﬁactions of the NeTSG soil (Table 3.5), without exceeding F
0.8960. For all aggregate ﬁ'actions at both depths of CT and NT treatments, the NC values
were greater than -25.7%o (Fig. 3.6), indicating an enrichment in ”C relative to the
successional vegetation (NeTSG). Changes in the original 5‘3C signature of the
successional vegetation soil in NeTSM and NT treatments were greater in the surface soil
than in the 5-15 cm soil layer (Fig. 3.5 and 3.6, respectively). In the CT treatment, mixing
of soil and burial of residues by tillage produced a more uniform 6'3C signature across the
0-15 cm depth. In samples from 0-5 cm layer of the NeTSM treatment and ﬁom both
depths of the CT treatment, macroaggregates appeared to be slightly enriched in 13C. This
seems to indicate that SOM is more decomposed in smaller than in larger aggregates.
However, increase in 613C (less negative) with increase in aggregate size, found by Anger
and Giroux (1996), was observed for the NT treatments only in the 5-15 cm layer. For the
0-5 cm layer of NT, inputs of c.,-derived material were apparently greater in
microaggregates and smaller macroaggregates (<0.5 m). If microbial activity is relatively
greater in these size ﬁ'actions (Jastrow et al., 1996), microbial biomass and the residual
substrate remaining in the soil after microbial degradation would be enriched in 13C
relative to the initial composition of the substrate (Jastrow et al., 1996). However, the
high variation in the 513C values for NT at this depth (Table 3.6), probably resulted ﬁom
undecomposed or partly decomposed maize residues, in addition to potential disruption by
slaking during the sieving step, prevents any ﬁnal conclusion. The results indicate a slight
trend for increasing 613C values in the direction of macroaggregates, similar to those

reported by Puget et al. (1995) for a silty soil under NT and superﬁcial tillage.

8O

Assuming that the organic matter in the NeTSG treatment had reached a steady
state by the time the soils were ﬁrst cultivated, its total C content represents the initial
carbon (C3-derived) in the soils (Fig. 3.3). These values were used as a baseline to
calculate the clmnges in the C3-C content as a result of cultivation for over 100 years, as
well as the inputs of C ﬁ'om the crops during this period. The relative contents (Fig. 3.7)
and proportions (Fig. 3.8) of recently deposited C in the CT and NT treatments were
almost always higher in macroaggregates (> 0.25 mm). The opposite trend of higher
percentage of “new” C (ﬁ'om crops) in aggregates < 0.5 mm in the NT treatment at 0-5
cm seems to be related to redistribution of ﬂesh residues to smaller fractions during wet
sieving, as indicated by the high standard errors of the 6'3C values (Fig. 3.6).

The proportions of C4-derived C in the NeTSM treatment after 11 years of maize
cropping ranged ﬁ‘om 9.5 to 14% and from 1 to 3%, for the 0-5 and 5-15 cm depths,
respectively (Fig. 3.8). The proportion of crop-derived C was much higher in NT soil at
the depth of 0-5 cm depth, ranging from 53 to 76% and representing almost twice the
pr0portion in the 5-15 cm depth (Fig. 3.8). For CT soils, these proportions were not
greatly affected by depth and ranged ﬁom 28 to 35% at 0-5 cm and ﬁ'om 22 to 45% at the
5-15 cm depth. The similarity in the proportions in CT indicates the inﬂuence of the
topsoil mixing by annual plowing. Since the organic matter responsible for stabilizing
microaggregates is considered to turn over much more slowly than that of
macroaggregates (Tisdall and Oades, 1982), the higher proportion of new C in the
microaggregates of the NT soil seems to be inconsistent. Jastrow et al. (1996) detailed

several possible explanations suggesting that not all C associated with microaggregates

81

may be biochemically recalcitrant. Some of their hypotheses are summarized here: i) most
of the new-C associated with microaggregates might come from the surfaces of larger
aggregates. In soils with highly stable aggregates, such as those in NT, the
microaggregates released by disruption upon slaking could be previously bormd into
macroaggregates in situ but were not stabilized enough to resist the disruption; ii) a recent
pool of labile material could be associated in signiﬁcant quantities to the surfaces of
microaggregates released ﬁom slaked macroaggregates; and iii) incrustation of organic
debris and clay particles could produce slaking-stable nricroaggregates in which the core
newly incorporated debris are physically protected ﬁom rapid decomposition. The present
ﬁndings and the results reported by Jastrow et al. (1996) suggest that microaggregates are

probably more dynamic than suggested by the current models of soil aggregation.

82

CONCLUSIONS

The results of this study show important differences in water-stable aggregate
distributions between NT and CT soils, and in a recently cultivated soil. These differences
may explain their contrasting retentions of SOM. In comparison to the permanent grass
treatment, 11 years of non-tilled and non-fertilized maize cropping caused a decline in C
levels but the effects were not clear on aggregate stability. In the long-term cultivated soils
(CT and NT), macroaggregation was greatly reduced with a concomitant increase in the
proportion of small macroaggregates and microaggregates. The introduction of no-tillage
greatly increased C levels in the top 5-cm soil, with trends in improvements in stability of
macroaggregates at both depths after 11 years. The development and/or stabilization of
new and old macroaggregates in NT management may indicate that the resilience capacity
of this soil is an important mechanism for storage and protection of SOM that is otherwise
lost under CT. The extent to which aggregation and SOM contents change as a result of
cropping and tillage practices are often diﬁcult to determine based on other reports since
it depends on several site-speciﬁc factors, such as soil properties, climate, type of crop,

and management and fertilization practices.

83

 

REFERENCES

Angers, DA. 1992. Changes in soil aggregation and organic carbon under corn and
alfalfa. Soil Sci. Soc. Am. J. 56:1244-1249.

Angers, D.A., and M. Giroux. 1996. Recently deposited organic matter in soil water-
stable aggregates. Soil Sci. Soc. Am. J. 60:1547-1551.

Angers, D.A., A. N’dayegamiye, and D. Cdté. 1993. Tillage-induced differences in
organic matter of particle-size ﬁactions and microbial biomass. Soil Sci. Soc. Am.
J. 57:512-516.

Angers, D.A., A. Pesant, and J. Vigneux. 1992. Early cropping-induced changes in soil
aggregation, organic matter, and microbial biomass. Soil Sci. Soc. Am 1.56:115-
119.

Bajracharya, RM., R. Lal, and J .M. Kimble. 1997. Long-term tillage effects on soil
organic carbon distribution in aggregates and primary particle ﬁactions of two
Ohio soils. p.113-123. In R Lal et al. (ed.) Management of carbon sequestration in
soil. CRC Press, Boca Raton, FL.

Balesdent, J ., and A. Mariotti. 1996. Measurement of soil organic matter turnover using
”C natural abundance. p.83-112. In T.W. Boutton and S. Yarnasaki (ed.) Mass
spectrometry of soils. Marcel Dekker, New York.

Balesdent, J., D. Boisgontier, and A. Mariotti. 1990. Effect of tillage on soil organic
carbon mineralization estimated from l3C abundance in corn ﬁelds. J. Soil Sci.
41 :587-596.

Barber, S.A. 1979. Corn residue management and soil organic matter. Agron. J. 71 :625-
627.

Beare, M.H., P.F. Hendrix, and DC. Coleman. 1994. Water-stable aggregates and organic
matter ﬁ'actions in conventional and no-tillage soils. Soil Sci. Soc. Am. J. 58:777-
786.

Blevins, KL, and W.W. Frye. 1993. Conservation tillage: an ecological approach to soil
management. Adv. in Agron. 51:33-78.

84

Dr

Blevins, R.L., R. Lal, J.W. Doran, G.W. Langdale, and W.W. Frye. 1998. Conservation
tillage for erosion control and soil quality. p.51-68. In F .J. Pierce and W.W. Frye
(ed.) Advances in soil and water conservation. Ann Arbor Press, Chelsea, MI.

Blevins, R.L., G.W. Thomas, M.S. Smith, W.W. Frye, and PL. Cornelius. 1983. Soil Till.
Res. 3:135-146.

Boutton, T.W. 1996. Stable carbon isotope ratios of soil organic matter and their use as
indicators of vegetation and climate change. p.47-82. In T.W. Boutton and S.
Yarnasaki (ed.) Mass spectrometry of soils. Marcel Dekker, New York.

Cambardella, C.A., and ET. Elliott. 1992. Particulate soil organic-matter changes across a
grassland cultivation sequence. Soil Sci. Soc. Am. J. 56:777-783.

Cambardella, C.A., and ET. Elliott. 1993. Carbon and nitrogen distribution in aggregates
from cultivated and native grassland soils. Soil Sci. Soc. Am. J. 57:1071-1076.

Carter, M.R. 1992. Inﬂuences of reduced tillage systems on organic matter, microbial
biomass, macro-aggregate distribution and structural stability of the surface soil in
a humid climate. Soil Till. Res. 23:361-372.

Coleman, D.C., P.F. Hendrix, M.H. Beare, D.A. Crossley, S. Hu, and RC]. van Vliet.
1994. The impacts of management and biota on nutrient dynamics and soil
structure in sub-tropical agroecosystems: impacts on detritus food webs. p.133-
143. In C.E. Pankhurst et al. (ed.) Soil biota: Management in sustainable farming
systems. CSIRO, East Melbourne, Australia,

Collins, H.P., B.A. Paul, K Paustian, and ET. Elliott. 1997. Characterization of soil
organic carbon relative to its stability and turnover. p.51-72. In B.A. Paul et al.
(ed.) Soil organic matter in temperate agroecosystems: Long-term experiments in
North America. CRC Press, Boca Raton, FL.

Dick, W.A., and J .T. Durkalski. 1998. No-tillage production agriculture and carbon
sequestration in a Typic Fragiudalf soil of Northern Ohio. p.59-71. In R. Lal et al.
(ed.) Mamgement of carbon sequestration in soil. CRC Press, Boca Raton, FL.

Douglas, J.T., and M.J. Goss. 1982. Stability and organic matter content of surface soil
aggregates under diﬂ‘erent methods of cultivation and in grasslands. Soil Till. Res.
2:155—175.

Dormaar, J .F., and J .M. Carefoot. 1996. Implications of crop residue management and
conservation tillage on soil organic matter. Can. J. Soil Sci. 76:627-634.

85

Elliott, ET. 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native
and cultivated soils. Soil Sci Soc. Am. J. 50:627-633.

Elliott, E.T., and C.A. Cambardella. 1991. Physical separation of soil organic matter.
Agric. Ecosys. Environ. 34:407-419.

Elliott, E.T., C.A. Palm, D.E. Reuss, and C.A. Monz. 1991. Organic nurtter contained in
soil aggregates ﬁom a tropical chronosequence: correction for sand and light
ﬁaction. Agric. Ecosys. Environ. 34:443-451.

Follett, RF, and BS. Schimel. 1989. Effect of tillage practices on microbial biomass
dynamics. Soil Sci. Soc. Am. J. 53:1091-1096.

F ranzluebbers, A.J., and M.A. Arshad. 1996. Water-stable aggregation and organic matter
in four soils under conventional and zero tillage. Can. J. Soil Sci. 76:387-393.

Goss, M.J. 1987. The speciﬁc effects of roots on the regeneration of soil structure. p. 145-

155. In G. Monnier and M.J. Goss (ed.) Soil compaction and regeneration. A.A.
Balkema, Rotterdam, The Netherlands.

Gupta, V.V.S.R, and J .J. Germida 1988. Distribution of microbial biomass and its
activity in different soil aggregate size classes as affected by cultivation. Soil Biol.
Biochem. 20:777-786.

Havlin, J.L., and DE. Kissel, L.D. Maddux, M.M. Claassen, and J.H. Long. 1990. Crop
rotation and tillage effects on soil organic carbon and nitrogen. Soil Sci. Soc. Am.
J. 54:448-452.

Haynes, R.J., and SR. Swift. 1990. Stability of soil aggregates in relation to organic
constituents and soil water content. J. Soil Sci. 41 :73-83.

Helal, M. H, and D. Sauerbeck. 1989. Carbon turnover in the rhizosphere. Z.
Pﬂanzenemahr. Bodenk. 152:211-216.

Huggins, D.R., C.E. Clapp, R.R. Allmaras, J.A. Lamb, and M.F. Layese. 1998. Carbon
dynamics in com-soybean sequences as estimated ﬁ'om natural carbon-13
abundance. Soil Sci. Soc. Am J. 62:195-203.

Jastrow, J.D., T.W. Boutton, and RM. Miller. 1996. Carbon dynamics of aggregate-

assoeiated organic nutter estimated by carbon-13 natural abundance. Soil Sci. Soc.
Am. J. 60:801-807.

86

 

Ker

Lal

M(

02

Pa

Kay, B.D. 1990. Rates of change in soil structure under different cropping systems. Adv.
Soil Sci. 12:1-52.

Kemper, W.D., and RC. Rosenau. 1986. Aggregate stability and size distribution. p.425-
442. In A. Klute (ed.), Methods of soil analysis. Part 1, 2nd ed. Agron. Monogr. 9,
ASA and SSSA, Madison, WI.

Lal, R, and J .M. Kimble. 1997. Conservation tillage for carbon sequestration. Nutrient
Cycling in Agroecosystems 49:243-253.

Le Bissonnais, Y. 1996. Aggregate stability and measurement of soil crustability and
erodibility. 1. Theory and methodology. Eur. J. Soil Sci. 47:425-437.

Monreal, C.M., and H. Kodama 1997. Inﬂuence of aggregate architecture and minerals
on living habitats and soil organic matter. Can. J. Soil Sci. 77:367-377.

Oades, J .M. 1984. Soil organic matter and structural stability: mechanisms and
implications for management. Plant Soil 76:319-337.

Paul, B.A. et al. 1998. The carbon cycle in agroecosystems in the US Corn Belt. In press.

Pojasok, , T., and B.D. Kay. 1990. Effect of root exudates ﬁ'om corn and bromegrass on
soil structural stability. Can. J. Soil Sci. 70:351-362.

Powlson, D.S., P.C. Brookes, and B.T. Christensen. 1987. Measurements of soil microbial
biomass provides an early indication of changes in total soil organic matter due to
straw incorporation. Soil Biol. Biochem. 19:159-164.

Puget, P., C. Chenu, and J. Balesdent. 1995. Total and young organic carbon distribution
in aggregates of silty cultivated soils. Eur. J. Soil Sci. 46:449-459.

Reicosky, D.C., W.D. Kemper, G.W. Langdale, C.L. Douglas Jr., and RE. Rasmussen.
1995. Soil organic matter changes resulting ﬁ'om tillage and biomass production. J.
Soil Water Cons. 50:253-261.

Reid, J.B., and M.J. Goss. 1982. Suppression of decomposition of l‘C-labelled plant roots
in the presence of living roots of corn and perennial ryegrass. J. Soil Sci. 33:387-
395.

Reid, J.B., M.J. Goss, and PD. Robertson. 1982. Relationship between the decrease in

soil stability eﬁ‘ected by the growth of corn roots and changes in organically bound
iron and aluminum. J. Soil Sci. 33:397-410.

87

Reinert, D.J., F .J . Pierce, and M.-C. Fortin. 1990. Temporal variation in structural stability
induced by tillage. p.63-72. In J .A. Stone et al. (ed.) Soil structure research in
Eastern Canada. Proc. Eastern Canada Soil Structure Workshop, Guelph, Ontario,
Canada. 10-11 Sept.

Robertson, G.P., J. R. Crum, and B.G. Ellis. 1993. The spatial variability of soil resources
following long-term disturbance. Oecologia 96:451-456.

Sala, O. E., W.J. Parton, L.A. Joyce, and WK. Laurenroth. 1988. Primary production of
the Central Grassland region of the United States. Ecology 69:40-45.

Salinas-Garcia, J .R., F.M. Hons, and J .E. Matocha. 1997. Long-term effects of tillage and
fertilization on soil organic matter dynamics. Soil Sci. Soc. Am. J. 61 :152-159.

Santos, D., and A.J.M. Smucker. 1998. Distribution of alfalfa root-derived organic carbon
in soil aggregates in tilled and non-tilled rotations. In preparation.

Sissoko, F., and A.J.M. Smucker. 1998. Enhancement of soil aggregation by the combined
inﬂuences of soil wetting and drying and root-microbial associations. Agron. J.
(Submitted).

Skidmore, E.L., W.A. Carstenson, and EB. Banbury. 1975. Soil changes resulting from
cropping. Soil Sci. Soc. Am. J. 39:964-967.

Tisdall, J .M., and J .M. Oades. 1982. Organic matter and water stable aggregates in soils.
J. Soil Sci. 33:141-163.

Whiteside, ER 1982. Soil survey of the Kellogg Biological Station. Unpublished
Research Report. WK. Kellogg Biological Station Library, Hickory Comers, MI.

88

 

Table 3.1. Mean weight diameter (MWD) under conventional tillage (CT), no-tillage
(NT), never-tilled successional grass (N eTSG), and never-tilled successional with
maize (NeTSM) at two sampling depths. (Means :t standard error; n = 4).

 

 

 

 

Depth CT NT NeTSG NeTSM
cm MWD (mm)
0 - 5 0.75 :1: 0.07 1.16 i 0.05 2.02 :t 0.09 2.12 :I: 0.08
5- 15 0.75-1:0.09 0.94:1:0.10 1.91:1:0.12 1.75:1:0.11

 

89

 

35 " _ Never-tilled grass - maize 0 " 5 cm

1:] Never-tilled grass

25-

 

20-
154

10-

Water-stable aggregates (%)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35- I S-TSCm

30-

25a

 

20-

15-

10-1

Water-stable aggregates (%)

 

 

 

 

 

 

 

 

 

 

 

 

2-4 1-2 0.5-l 0.25-0.5 0.1-0.25 <0.106

Aggregate size-class (mm)

Figure 3.1. Size distribution of intact water-stable aggregate ﬁ'actions under never-tilled
successional grass (NeTSG), and never-tilled successional with maize (NeTSM) at 0-5
and 5-15 cm depths. Means i standard error, n = 4.

90

 

35 — — Conventional tillage 0 - 5 cm
1:: No—tillage
30 -
25 -
20 —
15 -

10-

Water-stable aggregates (%)

 

 

 

 

 

 

 

 

 

 

30~

25-

20-

15—

 

10-

,_ I

>4 2-4 1-2 0.5-1 0.25-0.5 0.1-0.25 <0.106

Water-stable aggregates (%)

 

 

 

 

 

 

 

 

 

 

 

 

 

Aggregate size-class (mm)

Figure 3.2. Size distribution of intact water-stable aggregate ﬁactions under conventional

tillage (CT), and no-tillage (NT) at 0-5 and 5-15 cm depths. Means 3: standard error,
n = 4.

91

 

60- 0-5 cm

50*

 

40-

 

 

30*

201

10*

Total carbon (g C kg'l soil fraction)

 

 

 

 

 

 

 

 

 

60 - — Never-tilled grass - maize 5 " 15 C
1:! Never-tilled grass
50 -l

40-

 

 

30*

20*

10*

Total carbon (g C kg'l soil ﬁaction)

 

 

 

 

 

 

 

 

   

 

 

 

 

 

>4 2-4 1-2 0.5-l 0.25-0.5 0.1-0.25 <0.106

Aggregate size-class (mm)

Figure 3.3. Total carbon contents (nomralized to a sand-ﬂee basis) of water-stable
aggregate ﬁ'actions under never-tilled successional grass (N eTSG), and never-tilled
successional with maize (NeTSM) at 0-5 and 5-15 cm depths. Means 1 standard error,
n = 4.

92

 

60 - - Conventional tillage 0 -5 cm
[:1 No-tillage

50 *
40 a

30*

Total carbon (g C kg'l soil fraction)

 

 

 

 

 

 

 

 

   

 

60* 5-150m
50*
40*

30*

  

20*

 

10

1

Total carbon (g C kg'l soil fraction)

 

 

 

051.- 0.2.5-05 0.-.10I25 <0.106

 

Aggregate size-class (mm)

Figure 3.4. Total carbon contents (normalized to a sand-free basis) of water-stable
aggregate ﬁ'actions under conventional tillage (CT), and no-tillage (NT) at 0-5 and 5-15
cm depths. Means A: standard error, n = 4.

93

 

-20 _ O Never-tilled grass - maize O -5 cm
El Never-tilled grass
-21 -

-22 z
-23 A I . O t I E
-24 ..

m- E

-26 .1

513C ratio (%0)

 

 

-20 _ 5 - 15 cm
-21 ..
-22 _
-23 a

-24 .

513C ratio (%0)

I I ‘
a» U TH E .U 0: ii E

-26 *

 

 

>4 2-4 1-2 0.5-1 0.25-0.5 0.1-0.25 <0.106

Aggregate size-class (mm)

Figure 3.5. 6'3C ratios in water-stable aggregate size fractions under never-tilled
successional grass (NeTSG), and never-tilled successional with maize (NeTSM) at 0-5
and 5-15 cm depths. Means 3: standard error, n = 4.

94

 

 

 

 

 

 

 

 

 

 

 

-17 4 0 Conventional tillage 0 - 5 cm
El No-tillage T
-18 * __
g -19 - T- __ __ '7 H] E]
.13 -20 d 0:1 q: E]
o 9] E3 -—— *-
SS.‘ -21 - _,_ —— __
to .-n—
-22 - E E in E E E
E
-23 _
-17 _ 5 - 15 CHI
-18 i
:8 -19 -
‘o’
g -20 -
U
:3 -21 - }
°° i
'22 ‘ TE E T E E E E E E
-23 - E 3'!—

 

 

>4 2-4 1-2 0.5-1 0.25-0.5 0.1-0.25 <0.106

Aggregate size-class (mm)

Figure 3.6. 613C ratios in water-stable aggregate size fractions under conventional tillage
(CT), and no-tillage (NT) at 0-5 and 5-15 cm depths. Means 2t standard error, n = 4.

95

 

 

 

40 ‘ :3 CropC CTO-Scm . CTS-lScm
- HistoricalC

30‘

. I

      

 

 

NTO-Scm NTS-lScm
40‘ .

 

 

 

 

       
 

NeTSM 0—5 an 4 NeTSM 5-15 cm

 

Total carbon (g C kg'1 soil fraction)

 

 

 

 

 

20 J
10 ‘
0 A >4 2-4 1.2 0.5-1 0.2-0.5 0.1-0.2 <o.| ' >4 2.4 1.2 0.5-l 0.2-0.5 0.1-02 <0.l
Aggregate size-class (mm) Aggregate size-class (mm)

Figure 3.7. Contents of total C ﬁom historical and crop sources (normalized to a sand-ﬂee
basis) in water-stable aggregate fractions under conventional tillage (CT), no-tillage
(NT), and never-tilled successional with maize (NeTSM) at 0-5 and 5-15 cm depths.
Means :h standard error, n = 4.

96

Percent of total C

 

100 ‘ CTQ-Scm ‘ ‘ ‘ ‘ ‘CTS-lScm
80-‘
60‘
404

20-

 

NT 5456!!)

 

 

 

NeTSM 0-5 cm NeTSM 5-15 cm
100 r a -

 

 

 

 

 

 

>4 2-4 1-2 0.5-l 0.2-0.5 0.1—0.2 <0.l . >4 2—4 l-2 0.5-l 0.2-0.5 0.1-0.2 <0.l

Aggregate size class (mm) Aggregate size-class (mm)

:Z Crop C
— Historical C

Figure 3.8. Percent of the total carbon derived from historical and crop sources in water-
stable aggregate fractions under conventional tillage (CT), no-tillage (NT), and never-

tilled successional with maize (NeTSM) at 0-5 and 5-15 cm depths. Means i stande
error, n = 4.

97

CHAPTER 4
PLANT ROOT EFFECTS ON SOIL AGGREGATE CARBON CONTENTS AND
PHYSICAL PROPERTIES

INTRODUCTION

The primary source of C inputs to soils is through photosynthetic ﬁxation by
plants. The C that is translocated to the roots is present in soils as living roots and
rhizomes, root exudates, mucilages, sloughed root cells and tissues, and as microbial
tissues and metabolites (Qian et al., 1997; Balesdent and Balabane, 1992). Organic C
released ﬁ'om growing roots may stabilize aggregates directly, or indirectly by providing a
source of energy for microorganisms in the rhizosphere which may in turn produce
stabilizing materials (Tisdall and Oades, 1979). Soil aggregation is thought to play an
important role in the C dynamics of cultivated soils since it affects, and is affected by, the
microbial-regulated C transformations and cycling in soils.

Soil aggregate surfaces reﬂect soil structural processes since they provide much of
the contact area between roots and rhizosphere soil. As soil aggregates are being formed
and root systems elongate around or through them, newly added root C, solutes and
particulate organic matter move through the large soil pores and accumulate on or near
the surfaces of aggregates. However, in aggregated soils, the role of structure in soil
organic matter (SOM) dynamics has received little attention (Amelung and Zech, 1996),
despite evidence that physical protection of SOM within aggregates is important in

controlling its decomposition (Elliot, 1986; Gupta and Germida, 1988; Oades, 1988; Van

98

Veen and Kuikman, 1990; Gregorich et al., 1997). Yet, most of the information available
on the amounts of C derived from roots is based on analysis of bulk soil samples through
sieving techniques. One of the limitations of the sieving method is the redistribution of
SOM among aggregate ﬁactions that have been produced by the method itself: As a
result, wet sieving data can be difﬁcult to interpret (Elliott and Cambardella, 1991). The
distribution of organic compounds and minerals within aggregates can be assessed with
greater reliability by sequential removal and analysis of surface and inner fractions of
aggregates (Ghadiri and Rose, 1991; Kayser et al., 1994; Santos et al., 1997). In this way,
surface accumulations of recently deposited C on soil aggregates by contemporary root
systems can be estinmted (Santos and Smucker, 1998).

Recent studies have shown the usefulness of the 6'3C technique for estimating soil
C dynamics in a situation where the original vegetation has been replaced by plants with a
different photosynthetic pathway (Balesdent et al., 1987; Skjemstad et al., 1990). The
natural abundance of 13C is higher in C4 compared to C3 species. Plants with the C3
pathway (mean 6'3C of -28%o), discriminate against l3CO2 during photosynthesis, causing
the 13C/‘ZC ratios of their phytomass to be depleted in 13C relative to those of C4 plants
(mean 6'3C of -12%o) (Smith and Epstein, 1971). By growing C4 plants on soils previously
only exposed to C3 plants, this difference in isotope ratio may be utilized to study C
partitioning in plant-soil systems. This technique can be not sensitive enough for detecting
changes in C composition of disturbed bulk soil samples during and/or aﬁer a single
growing season (Balesdent and Balabane, 1992). Plant C allocation to the soil is known to

be inﬂuenced by physiological factors such as the stage of plant development (Keith et al.,

99

1986) and water availability (Boutton, 1996). To date, most studies using the natural 13C
abundance technique (e.g., Balesdent et al., 1987; Martin et al., 1990; Angers and Giroux,
1996) have dealt with C dynamics on whole disturbed soils and/or aggregate ﬁactions
sieved out of bulk samples. Its use in successive external surfaces of soil aggregates could
estimate, on a short-term basis, the quantity and spatial location of root-derived organic C
recently deposited on or near soil aggregates. Differences among the sequestration rates
are important because the dynamics of C derived from C4 are different than ﬁ'om C3 plants
resulting in contrasting C concentrations in a range of soil particle ﬁactions (Huggins et
aL, 1995). In this case, the ﬁner textured fractions accumulated more total C under C4
(corn) than under C3 (soybean) crops.

Objectives of the present study were to assess the effects of a C3 crop on the
quantity, composition and distribution of recently released root C to soil aggregates from
a fallow soil dominated by C3 grasses for the past 40 years. Using a short-term plant
growth experiment, C originating from both a C3 vegetation site and from contemporary
ryegrass plants were traced and the plant root effects on some soil physical and chemical

properties were investigated.

100

MATERIALS AND METHODS

Soil

The experimental soil was collected ﬁom the A horizon (0-20 cm) of an
uncultivated site, under C3-grass vegetation for at least 40 years, at the Kellogg Biological
Station, near Hickory Corners, MI. Bulk samples of the soil, a Kalamazoo loam (ﬁne
loamy, mixed, mesic, Typic Hapludalt), were sieved to remove root materials and pebbles,
and obtain aggregates in the size range of 3-7 mm in diameter. Selected properties of the

soil aggregates are shown in Table 4.1.

Soil columns (construction and packing)

Polyvinyl chloride (PVC) columns, 20 cm diameter and 34 cm length, were split in
halfhorizontally and reassembled with duct tape (Fig. 4.1). The bottom end of each
column was covered with stainless steel wire screen (1 mm opening size) and eight layers
of cheese-cloth seemed with duct tape. Colunms were placed on a 1-cm layer of steam-
sterilized (220°C for 24 hours) pea gravel contained in a 30-cm wide aluminum foil plate
with a drainage hole in the center. Two pairs of holes, 3 mm in diameter and 5 cm apart,
were drilled in one side of the columns, 10 and 26 cm from the top. Two pairs of l9-cm
long stainless steel waveguides, connected to a coaxial cable and embedded in resin at one
end, were inserted through the holes in the PVC soil columns to monitor volumetric water
content in each column by the Time Domain Reﬂectometry (TDR) technique (Topp et al.,

1980).

101

Eight kg of air dry white silica sand was packed in layers of 2 kg at the bottom lulf
of each column. The top half of each column was then ﬁlled with 30 alternating layers of
previously weighed amounts of aggregates (94 g layer“) and sand (390 g layer“),
respectively, so that the surface was about 2 cm ﬁom the top of the column. The columns
were topped with 0.5 cm sand to minimize evaporation. The use of a mixture of sand and
aggregates as grth media took into consideration that: (i) penetration resistance of sand
has been reported to vary little with water content across a wide range of water contents
(Cook et al., 1997); (ii) during a preliminary trial, a desirable quantity of preferential root
growth occurred around soil aggregates; and (iii) plant roots can be easily washed free
ﬁ'om the sand to ensure the entire root system is removed. Water characteristic curves of
the growth medium (silica sand plus aggregates) and of the silica sand were determined on

a ceramic pressure plate apparatus (Klute, 1986) and are reported in Fig. 4.2.

Planting

Columns were saturated with a half-strength Hoagland’s nutrient solution
(Hoagland and Amon, 1950) at a rate of 10 L per column and allowed to drain for 4 days.
They were then saturated with distilled water and allowed to drain for approximately 1
day when the water content of the soil column was considered to be at ‘container
capacity’. On July 5 1997, ryegrass seeds (Lolium perenne L. ‘Dimension’), with an
average weight of 0.18 g per 100 seeds, were planted at a rate to provide about 4 plants
percrn2 ofthecolunmsurfacearea. Theexperimentwasset upandmaintainedina

greenhouse at 24°C :1: 2 day/night ambient temperature. Natural daylight was

102

supplemented by high-pressure sodium vapor lights at 36,000 it candles (387.5 1x), placed
at a height of 1.5 m above the top of the columns, for a 16-h photoperiod. For the ﬁrst
two days after planting the top of the columns were covered with aluminum foil to reduce
evaporative losses. On July 15, a copper sulphate (CuSO4.5H20) solution (5% w/v) was

sprayed on the surface sand to eliminate and prevent algae development.

Water regimes

One week after planting, columns were irrigated four times daily with a half-
strength Hoagland’s nutrient solution using 3 drip emitters per column, equally spaced,
with a delivery rate of 10 mL min". An automated irrigation system was set up to deliver
the appropriate solution amounts to the treatments at 10:00, 14:00, 18:00, and 22:00 h.
The inﬂuence of drought stress on allocation of root C to soil aggregates was studied by
dividing the columns into two random groups, after initial wetting. On July 25, water
stress was randomly imposed on one group of columns at the one-tiller stage by applying
one-fourth (1 minute irrigation time) of the amount of water in the adequately watered
columns (4 minutes irrigation time). Then , the number of drip emitters in the unplanted
control treatment was changed to two per column. Irrigation was applied at 5:30, 9:30,
13:30, and 17:30 h. After July 31, the irrigation was exclusively distilled water in an
attempt to reduce mineral toxicity symptoms in corn plants. On August 8, the number of
emitters in the unplanted control treatment was reduced to one per column.

The amount of water applied was monitored in each column by the TDR

technique. The two pairs of waveguides from each eolunm were connected to a series of

103

multiplexers by 2-m length 50-0 coaxial cables which were then connected to a Metallic
TDR cable tester, model Tektronix 1502C (Tektronix, Inc., Redmond, OR 97756). The
cable tester was connected to an IBM-compatible PC (Fig. 4.1). The waveforms obtained
on the cable tester were analyzed and converted to volumetric water content using a
software written in BASIC language (J. Ferguson, unpublished) which utilizes the linear

cahbration equation of Lidieu et al. (1986).

Plant, soil and root sampling

The total growth period was for 6 weeks following planting. At harvest, plant
shoots were cut at the soil surface and oven dried (65 °C) for determining dry matter and
6'3C. The columns were then left without irrigation for a few days so that the soil columns
would not be too wet when disassembled. Upon sampling, the soil was moist but not too
wet so that aggregates were not excessively sheared during the procedure. A wooden
piston apparatus was used to push the mass of sand plus aggregates out of each PVC
column. The cylindrical segments were then cut longitudinally into 1/4 sections. Two 1/4
sections of the growth media in the top halves of the columns were kept for root analysis
and the other two combined for soil aggregate analysis. In order to collect these samples,
the top halves of the columns were further divided into three 5-em layers to evaluate the
inﬂuence of root density on C allocation to aggregates. The two 1/4 sections designated
for soil analyzes were air-dried. The sand was gently separated ﬁom the soil aggregates
and associated roots by brief manual sieving through 2-mm (15 cm wide) sieve. Large

roots and associated soil aggregates were picked out by hand. Care was taken to separate

104

those aggregates enmeshed by roots and discarding those with roots growing through
them and those ﬁ'ee of roots. Root samples were taken for determination of their 613C
ratios.

The other two 1/4 sections of the growth media designated for root analyzes, plus
the roots in the bottom halves of the columns, were bagged and stored at 4°C and later
washed from sand by the hydropneumatie elutriation method (Smucker et al., 1982).
Roots present in the bottom halves of each column were collected to investigate the
potential effects of varying water content on root growth in pure sand. Washed roots were
stored in whirlpack plastic bags containing about 200 mL 15% (v/v) methanol solution
and preserved at 4°C. Root samples were dyed with Malachite green oxalate by injecting
5 mL of 1% (w/v) dye solution into the plastic storage bag (Smucker, 1990). Stained
roots were rinsed with distilled water on an ultra ﬁne (25 am) nylon screen and uniformly
distributed on a clear glass tray where the roots were covered with a thin water ﬁlm for
image video recording. The glass tray was illuminated ﬁ'om below by a translucent light
table (to provide contrast between the roots and the background) and video recorded by a
computer controlled robotic camera on VHS video tapes (Smucker, 1990). To avoid
overlapping of roots, samples with large amounts of roots were split into 2 or more trays
for the video recording step and then recombined for dry weight measurements. Roots
were then dried at 60°C for 48 hours and their weights recorded. Video recorded root
images (100 per tray samples) were processed by a Sun Ultra-based computer algorithm in
the Root Image Processing Laboratory (RIPL), at Michigan State University

(URL:http://www.rootdig.ess.msu.edu) (Pietola and Smucker, 1998). The version of the

105

program for washed roots (WR-RIPL version 3) was used for measurements of root
length by classifying roots into 5 classes, based on their diameter, using an image
resolution of 276 pixels per 1 cm by the image processing system Average root width
classes were Class 1 (0.2 mm), Class 2 (0.5 mm), Class 3 (0.9 mm), Class 4 (1.4 mm), and
Class 5 (2.1 mm). Root length was measured for each class and a summation of roots in
all classes was also obtained. All data reported are for roots without debris (Dowdy et al.,
1998). Root surface area (RSA) was calculated as a sum of all image proﬁle areas in all
width classes, assuming a uniform cylindrical root shape, with the formula A = 2 1: r L,
where r is root radius and L total root length Root volumes (RV) were calculated as
V=nFL

Subsamples of intact aggregates of each class were analyzed for physical and
chemical properties. Water-stability of air-dry aggregates was assessed with the wet
sieving technique (Kemper and Rosenau, 1986). The tensile strength of dried aggregates
was measured by the crushing test (Dexter and Kroesbergen, 1985). Soil aggregates were
ﬁrst air-dried and then oven dried at 105 °C for 24 hours and cooled in a desiccator before
being crushed. Each aggregate was measured for its strength by a vertical force applied
between two ﬂat parallel plates (Fig. 4.3). A constamly increasing vertical force was
applied to each aggregate by the top plate, which was connected to a computer-controlled
stepper-motor. The bottom ﬂat plate was the platform of a digital balance, whose output
was recorded by a computer. The computer continuously recorded and graphed the
applied forces on a monitor. The graphic peak associated with the ﬁrst major decline in

force, applied to the aggregate, was identiﬁed by solving for the ﬁrst derivative of applied

106

force vs. time using a program written in BASIC (John Ferguson, unpublished). The

tensile strength of each aggregate was calculated by the equation

Y = 0.576 (F/D") (1)

where D is the aggregate effective spherical diameter and 0.576 is a proportionality
constant relating maximum compressive stress to tensile strength at failure (Dexter and
Kroesbergen, 1985). F is the force (N) required to ﬁacture an aggregate and was

calculated by

F=bg (2)

where b is the measured balance reading (in kilograms, kg) and g is the acceleration force
due to gravity (9.807 m s‘z). The mean sieve diameter, d , of all the aggregates to be
crushed (40 aggregates per treatment) was determined ﬁ'om the average of the sieve sizes
(d = (0.003 + 0.007)/2) = 0.005 m). The effective spherical diameter (D) of each
aggregate was then determined according to their individual masses, m, in kilograms. The
adjustment was done on the assumption that all aggregates within each treatment had the
same density, p (Dexter and Kroesbergen, 1985). If m, is the mean mass of a batch of

aggregates, then

p=(6mO/7ra°)=(6m/7tD3) (3)

107

which results in

D = d (m/mo)”3 (4)

Groups of soil aggregates of each class were mechanically separated into external
and internal fractions using a modiﬁed version of the soil aggregate erosion (SAE) method
described by Santos et al. (1997). In the present study, soil aggregates were abraded using
the stainless steel (SS) mesoSAE chambers, 2.5 cm in diameter, developed by the Soil
Biophysics Laboratory, at Michigan State University. A detailed abrasion protocol is given
in Appendix A3. Brieﬂy, each of 8-10 aggregates was placed in individual SS mesoSAE
chambers and rotated at 300 rpm in a reciprocal shaker, having a rotational diameter of
2.5 cm. The shaker was stopped when 33% (:t 2%) of the original aggregate mass was
abraded (outer layer) and passed through the 356 ,um screen inside the chamber. Soil
material ﬁom each aggregate passing through the screen was then thoroughly mixed and
stored in glass vials while the sand particles (> 356 am) remaining on the screen were
discarded. Another run was conducted so that a transitional layer of aggregates,
representing about 33% of the remaining aggregate mass, could be collected. The
percentages of the mass remaining at the time the outer and transitional layers were
collected was considered to be the inner region of the aggregate (Appendix A3).
Replicated soil samples ﬁom each concentric layer and whole aggregates were ground
using a mortar and pestle. Total C was determined by dry combustion in a CHN analyzer

(Carlo Erba, Italy) and 513C was determined by mass spectrometry (Europa Scientiﬁc,

108

UK). The NC value was expressed relative to the international standard PDB (as 6'3C,
%o). The ﬁaction of organic C derived from ryegrass plants was calculated using the

modiﬁed form of the mixing equation (Balesdent and Mariotti, 1996):

New C3-C (%) = (as - egg/(a,ye - 5,3) x 100 (5)

where 6, is 6'3C value for an aggregate layer or whole aggregate cultivated with ryegrass,
63, is the 5‘3C value for the corresponding aggregate layer or whole aggregate ﬁ'om the
grass soil (unplanted control treatment), and <3rye is the 6'3C value of ryegrass root

residues.

Experimental design

A factorial experiment (3 x 2) was arranged in a split plot design with 8
replications per treatment (32 columns total). The main treatments were ryegrass, maize
(data not reported), and an unplanted control, split into two water treatments (water-

stressed and non-stressed control).

109

RESULTS AND DISCUSSION

Volumetric water content

The watering treatments produced diﬂ‘erences in the water contents of the planted
columns. Beyond 17 days after planting, soil water contents of the stressed treatments
were always lower than that of the non-stressed treatment (Fig. 4.4). Top portions of the
non-stressed columns of the unplanted control, had an average soil water content in the
range of 0.17 cm3 cm'3 while in the ryegrass cohrmns it varied ﬁom 0.10 to 0.16 cm3 cm".
Soil water contents in both top and bottom portions of the ryegrass columns of both
stressed and non-stressed treatments dropped sharply, particularly in the last two weeks of
the experimental period, indicating a very proﬁrse growth of the ryegrass root system.
Twenty-two days after planting, all treatments were watered to container capacity in a
ﬂow test to check for drainage limitations in any of the columns, as evidenced by the rise

in water content seen on the graphs (Fig. 4.4). No limitations in drainage were observed.

Root and shoot dry weight

Differences in water content affected the accumulation of dry matter in ryegrass
plants. Shoot and root dry weight, and rootzshoot ratio decreased signiﬁcantly with
imposition of water stress (Table 4.2). These results differ ﬁom an earlier report
concerning ryegrass’s response to drought occurring under greenhouse environments.
Jupp and Newman (1987) reported no signiﬁcant differences in shoot dry weight and

rootzshoot ratio between L. perenne plants under severe and long-term water stressed and

110

non-stressed treatments. However, they reported signiﬁcant increases in lateral root
initiation and elongation under prolonged low water potential, with total length of lateral
roots increasing between three to ﬁve times compared to non-stressed plants. Also, the
increase in rootzshoot ratio of ryegrass plants with a decrease in water availability in loam

soils, observed by Davidson (1969), was not veriﬁed in the present study using 75% sand.

Root parameters

Water deﬁcit signiﬁcantly reduced ryegrass total root length to the depth of 10 cm
(Table 4.3). The distribution of the total RL across the several root width classes indicated
an even higher proportion of roots in the Classes 1 and 2 (< 0.5 mm dia.). In the top of the
colurrms, the percent of roots < 0.5 mm was 99% of the total RL for both stressed and
non-stressed plants. In the lower column, the proportion of roots in the Classes 1 and 2
was even higher, but similar for both water treatments, approaching nearly 100% of the
total RL (Table 4.4). From these results it is clear that ﬁne roots (< 0.2 mm), rmke the
largest contribution to the total root length Low water contents had a deleterious effect
on root dry weights at the depths of 0-5 and 5-10 cm, possibly because many of these
roots had grown before drought conditions were initiated and may have died.

Ryegrass root surface area (RSA), volume (RV), length density (RLD) and dry
weight density (RDWD) parameters were signiﬁcantly lower in the stressed treatment at
the 0-5 and 5-10 cm depth (Table 4.5). Root length density, an index of water uptake
capacity, was comparable to values of 6.9 and 2.4 cm em", reported by Haynes and

Francis (1993) for ﬁeld-cultivated perennial ryegrass at the depths of 0-10 and 10-20 cm.

111

However, RLD values were much lower than values of 27.6 and 23.6 cm cm'3 for a sandy
loam and a silty loam soil, reported by Reid and Goss (1981) for 42-d old ryegrass plants
grown in a greenhouse. Although the present study did not include a pure soil aggregate
bed for comparisons, these variations could be a result of the less favorable conditions for
root growth in the sand plus soil aggregates mixture used in the present experiment.

Root LD in the bottom portion of the columns was 9 and 15 times lower than in
the top portions, for water-stressed and non-stressed treatments, respectively. Similarly,
RDWD values were 10 and 20 times lower in the bottom columns, respectively. Reduction
in root parameters in the top 10 cm of the column was a direct effect of reduced root
growth of plants under water deﬁcit conditions. Higher RLD and RDWD for non -stressed
plants were expected due to the ryegrass characteristic ﬁne root networks and large
branching and root length densities. The lack of signiﬁcant differences between water
treatments for the 10-15 and 17-34 cm depths, indicates that plant roots responded to
decreasing soil water contents by greater growth in the wetter region of deeper soil. In this
regard, severe water stress would cause damage to the root system while moderate stress
would enhance its growth (Eghball and Maranville, 1993; Smucker et al., 1991). From
Fig. 4.4 it can be seen that the imposed stress may have been too severe towards the end
of the experimental period, especially in the top 10 cm, causing the rapid death of root
hairs and cortical cells (Jupp and Newman, 1987). As there is some indication that roots of
perennial ryegrass are relatively short-lived with a large proportion of roots living for less
than 14 days at 27°C (Forbes et al., 1997), recovery following an imposed stress would

have a great effect on root turnover and regrowth.

112

Carbon isotope variation in plant tissues

Differences in 5‘3C of both shoot and root of ryegrass were observed in plants
grown under water stressed conditions. Signiﬁcantly higher (less negative) 61°C were
observed in plants subjected to water stress (Table 4.6). The mean 613C value across water
treatments for ryegrass was -29.8%o for shoots, and -26.3%o and -28.1%o for stressed and
non-stressed roots, respectively, across the three sampling depths. Differences in soil
water availability and soil fertility, have a greater impact on stomatal conductance and
photosynthetic capacity, respectively, of C3 compared to C4 plants (Balesdent and
Mariotti, 1996; Boutton, 1996). Plants under adequate water levels would have open
stomata to maximize the rate of CO2 diffusion into the leaves. This results in a relatively
high ratio of intercellular CO2 concentration in the leaves (p,) to the partial pressure of
atmospheric CO2 (Pa) and a smaller (PC (or Larger A vahre, as expressed in the plant
ecophysiological literature) (F arquhar et al., 1989; Boutton, 1996). The p/p, ratio
indicates the rate at which CO2 diffuses into the leaves as opposed to the rate at which it is
utilized by photosynthesis. Conversely, plants under conditions of water stress would
decrease stomatal aperture to prevent water losses. Closure of stomata would also limit
CO2 diffusion into the leaves, resulting in low p/p, and larger 513C (or smaller A values)

(Boutton, 1996).

Soil C and 13C ratio
Total C contents (0-5 cm depth) were higher for whole aggregates, outer layer and

inner regions of soil aggregates cultivated with ryegrass under well-watered conditions

113

compared to the unplanted soil (Fig. 4.5). For the water-stressed treatment, differences
were found only when whole aggregates were analyzed indicating that the reduced root
dry weights and lengths of plants under water stress explored lesser volumes of soil. For
the 5-10 cm depth, C contents of whole aggregates and their outer layers, under non-
stressed ryegrass differed signiﬁcantly from the unplanted soil. Carbon contents in soil
aggregates under plants with adequate water supply was always higher than for those
under water deﬁcit conditions in the top 5 cm (Fig. 4.5). Diﬂ‘erences in C contents
between water treatment and aggregates of the unplanted soil were generally not distinct
for any of the depths. However, there was a trend of increased C content ﬁ'om 0-5 to 5-10
cm depth for whole aggregates and their outer layers. This could be an indication of partial
leaching of the original C ﬁ'om the aggregates in the top 0-5 cm.

The higher C contents in soil aggregates cultivated with ryegrass and located in the
top layers is likely to be a result of additions of organic materials to the soil from roots and
root exudates. The higher proportion of ryegrass roots at 0-5 than at 5-10 cm (Table 4.5)
corroborate this inference. Materechera et al. (1992) reported 27 to 30% increases of C in
2—4 rmn aggregates alter the soil was cropped to a different ryegrass species (Lolium
rigidum) for 4-5 months. They attributed this increase to be the result of the high root
mass and length density. The addition of C through senescence of root segments and root
hairs growing around and within aggregates may have occurred in the present study. The
more negative 6‘3C values seem to conﬁrm these effects in the inner regions of aggregates
especially for plants under non-stressed conditions at 0-5 cm depth (Fig. 4.6). The changes

in 6‘3C even in the inner regions of aggregates are in agreement with our concept of

114

aggegate invasions by root hairs of ryegrass and associated fungi which invade the
interior of larger aggregates. Trends in (WC values in aggregates ﬁ'om the ryegrass
colunms under water stress were not as clear. These effects were evident due to the high
root length density and root mass of ryegrass, consistent with the observation of Tisdall
and Oades (1979, 1982) and Haynes and Francis (1993). These researchers reported that
the amount of organic material released by roots is related to their length and mass.
Greater root growth means greater possrbilities for C production by root exudates as well
as greater physical entanglement of particles by roots (Monroe and Kladivko, 1987; Reid
and Goss, 1981). Thus the ability of crop roots to provide CC to soil structural units is an
important contribution for long-term soil stability.

The cultivation of soil aggregates originally with an average initial 6‘3C of -24.3%o
with ryegrass (6”C of roots ranging from -28.1 to -26.3%o) provided another opportunity
for determining new C3-C contributions to the soil system, using Eq. 5. After 6 weeks of
ryegrass growth, 5‘3C values for the whole aggregates became more negative by ~ 1960 at
both sampling depths with similar values for water-stressed and non-stressed treatments.
This decrease strongly suggests that much of the C present in surfaces of aggregates
originates from the current ryegrass plants. 6'3C values of outer layers and inner regions of
aggregates decreased to a greater extent under non-stressed than in stressed conditions,
with values similar to those in the whole aggregates (Fig. 4.6). In the 5-10 cm depth, 6‘3C
values were similar for outer layer and inner regions, regardless of irrigation treatment. At
both depths, the greatest variations in 613C values were associated with aggregates ﬁ'om

water stressed treatments, as values ranged from 1 to nearly 2 %o units. This seems to be a

115

result of the high variability in the isotopic composition of plants under enviromnental
stress (Balesdent and Mariotti, 1996) and/or an indication that the root-aggregate contact
were more erratic under water stress conditions.

New C3-C in whole aggregates was 62% (4.39 g kg“) for stressed ryegrass and
25% (2.07 g kg“) for the non-stressed treatment in soil aggregates located in the 0-5 cm
region (Fig. 4.7). External layers and inner aggregate regions had 20% (0.88 g kg“) and
8% (0.39 g kg“) of new C3-C, respectively. The sum of C3-C contents in external layers
and internal regions of aggregates was less than the values observed for whole aggregates
because the transitional region (representing 1/3 of the aggregate mass) was not reported.
This newly sequestered C that became part of the soil organic matter pool in the
aggregates may have included some root debris that had been located within the
aggregates. Except for these debris, all other visible root material, speciﬁcally very ﬁne
roots growing on the aggregate surfaces, had been carefully removed with the aid of a
magniﬁer lens, prior to the aggregate peeling procedure.

Soil aggregates from the unplanted control columns showed a slight variation in
6'3C values with an increase of about on-half%o unit. Since the unplanted control samples
were used as background to obtain the proportion of new C3-C (Fig. 4.7) these variations
were reﬂected in the estimated inputs. Further peeling of the original aggregates used in
the columns could provide some insights into these variations. In addition, variations in the
calculated proportions between stressed and non-stressed treatments for similar initial and
ﬁnal 6‘3C values (e.g. outer layer of aggregates) were due to the different 613C assigned to

roots when using Eq. 5. On average for all depths, 6‘3C values of roots of water stressed

116

plant were 1 to 2 %o units less negative than those of non-stressed plants (Table 4.6),
leading to differences of 37% in the calculated new C3-C inputs between water treatment
(Fig. 4.7). This indicates the importance of the characterization of a speciﬁc root 6'3C to

correct for differences in C inputs by plants under physiological stresses.

Mean weight diameter and aggregate tensile strength

Plants with greater root mass and root length density are usually expected to
increase soil aggregate stability (Reid and Goss, 1980, 1981; Haynes and Francis, 1993).
However, a signiﬁcant effect of ryegrass on MWD was observed only for the 0-5 cm
depth of non-stressed plants compared to stress treatment (Fig. 4.8). This effect is
expected to persist for several months because roots decompose slowly (Tisdall and
Oades, 1980). The lower MWD values in the ryegrass soil under water stress conditions
may result ﬁ'om the fact that the aggregates used in this study were under meadow for at
least 40 years, with relatively low soil C contents (Table 4.1). Partly because of their low
C contents, these aggregates are not strongly bound together and drying causes acute
development of incipient ﬁacture faults, while rewetting, resulting in greater slaking
(Haynes and Swift, 1990; Haynes and Beare, 1997). More ﬁequent wetting and drying
events seemed to inﬂuence in the MWD of aggregates in the unplanted control treatment
with an indication of higher values towards the surface layer (Fig. 4.8). The apparent no
change in MWD with depth for non-stressed conditions corroborate this observation.
Fewer roots of ryegrass at the depth of 10-15 cm resulted in lower MWD values,

compared to the unchanged MWD in the control treatment.

117

Factors such as the intensity and frequency of wetting and drying, the previous
management of the soil, the proportion of clay, and continued wetting and drying cycles
decrease the percentage of water-stable aggregates and tensile strength of stabilized
aggregates by tilth mellowing (Utomo and Dexter, 1981). The high RLD and RDWD of
ryegrass probably resulted in very ﬁne roots penetrating the aggregates which, in turn,
increased the chance of aggregate breakage. Although tensile strength measurements were
conducted on visibly intact aggregates, the values for aggregates under ryegrass at the top
5 cm were signiﬁcantly lower than at 5-10 cm (Fig. 4.9). For the non-stressed ryegrass
treatment, tensile strength values were similar in the top 10 em while at 5-15 cm showed

the lowest value compared to the unplanted control.

118

CONCLUSIONS

The experiment showed important inﬂuences of water availability on growth of
plant roots and its effects on properties of soil aggregates. There were signiﬁcant
diﬂ‘erences in sheet and root dry weights of ryegrass plants in response to water stress.
The majority of total RL was comprised by roots < 0.5 mm diameter indicating the
importance of ﬁne root contribution to the total RL. Water availability signiﬁcantly
changed ryegrass rooting patterns and root activity to the depth of 10 cm. In general,
RSA, RLD, and RDWD of ryegrass decreased with water stress. After 6 weeks of
ryegrass growth, C contents were higher in the whole aggregates and their outer and inner
regions under well-watered conditions. For samples subjected to water stress higher C
contents were observed only when the whole aggregates were analyzed. Regardless of
water treatment, 613C values for whole aggregates became ~1%o more negative in relation
to the initial 6'3C value of -24.3%o for this soil. The aggregate abrasion technique
described was useﬁrl in quantifying the amount of recently released root C into soil
aggregates. Inputs of new C3-C in whole aggregates were estimated to be 62%

(4.39 g kg“) and 25% (2.07 g kg“) of the total C in stressed and non-stressed ryegrass
treatments at the 0-5 cm depth. For external layers and inner regions of aggregates
associated with stressed plants new C3-C accounted for 20% (0.88 g kg“) and 8% (0.39 g
kg“) of the total C. Signiﬁcant inﬂuences of water and plant treatments on soil aggregate
stability and tensile strength were not evident from this experiment, probably as a result of

the relatively short time.

119

REFERENCES

Angers, D.A., and M. Giroux. 1996. Recently deposited organic matter in soil water-
stable aggregates. Soil Sci. Soc. Am J. 60:1547-1551.

Amelung, W., and W. Zech. 1996. Organic species in ped surfaces and core ﬁactions
along a climosequence in the prairie, North America. Geoderma 74:193-206.

Balesdent, J ., and M. Balabane. 1992. Corn root-derived soil organic carbon estimated by
natural l3C abundance. Soil Biol. Biochem. 24:97-101.

Balesdent, J., and A. Mariotti. 1996. Measurement of soil organic matter turnover using
13C natural abundance. p. 83-112. In T.W. Boutton and S. Yarnasaki (ed.) Mass
spectrometry of soils. Marcel Dekker, New York, NY.

Balesdent, J., A Mariotti, and B. Guillet. 1987. Natural ”C abundance as a tracer for soil
organic matter dynamics studies. Soil Biol. Biochem. 19:25-30.

Boutton, T.W. 1996. Stable carbon isotope ratios of soil organic matter and their use as
indicators of vegetation and climate change. p. 47-182. In T.W. Boutton and S.
Yarnasaki (ed.) Mass spectrometry of soils. Marcel Dekker, New York, NY.

Cook, A., C.A. Marriott, W. See], and CE. Mullins. 1997. Does the uniform packing of
sand in a cylinder provide a uniform penetration resistance? Plant Sci] 190:279-
287.

Davidson, RL. 1969. Eﬂ‘ects of soil nutrients and moisture on root/shoot ratios in Lolium
perenne L. and Trifolium repens L. Ann Bot. (London) 33:571-577.

Dexter, A.R., and B. Kroesbergen. 1985. Methodology for determination of tensile
strength of soil aggregates. J. Agric. Eng. Res. 31:139-143.

Dowdy, R.H., A.J.M. Smucker, M.S. Dowlan, and J.C. Ferguson. 1998. Automated
image analysis for separating plant roots ﬁom soil debris elutriated ﬁom soil cores.
Plant Soil 200:91-94.

Eghball, B., and J.W. Maranville. 1993. Root development and nitrogen inﬂux of corn

genotypes grown under combined drought and nitrogen stresses. Agron. J. 85:147-
152.

120

Elliott, ET. 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native
and cultivated soils. Soil Sci. Soc. Am. J. 50:627-633.

Elliott, E.T., and C.A. Cambardella. 1991. Physical separation of soil organic matter.
Agric. Eeosyst. and Environ. 34:407-419.

Farquhar, G.D., J .R. Ehleringer, and KT. Hubick. 1989. Carbon isotope discrirrrination
and photosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40:503-537.

Forbes, P.J., K.E. Black, and J.E. Hooker. 1997. Temperature-induced alteration to root
longevity in Lolium perenne. Plant Soil 190:87-90.

Ghadiri, H., and CW. Rose. 1991. Sorbed chemical transport in overland ﬂow: I. A
nutrient and pesticide enrichment mechanism. J. Environ. Qual. 20:268-633.

Gregorich, E.G., C.F. Drury, B.H. Ellert, and BC. Liang. 1997. Fertilization effects on
physically protected light fraction organic matter. Soil Sci. Soc. Am. J. 61 :482-
484.

Gupta, V.V.S.R., and J.J. Germida. 1988. Distribution of microbial biomass and its
activity in different soil aggregate size classes as affected by cultivation. Soil Biol.
Biochem. 20:777-786.

Haynes, R.J., and M.H. Beare. 1997. Inﬂuence of six crop species on aggregate stability
and some labile organic matter ﬁactions. Soil Biol. Biochem. 29:1647-1653.

Haynes, R1, and GS. Francis. 1993. Changes in rrrierobial biomass C, soil carbohydrate
composition and aggregate stability induced by growth of selected crop and forage
species under ﬁeld conditions. J. Soil Sci. 44:665-675.

Haynes, R1, and RS. Swift. 1990. Stability of soil aggregates in relation to organic
constituents and soil water content. J. Soil Sci. 41 :73-83.

Hoagland, DR, and DJ. Arnon. 1950. The water-culture method for growing plants
without soil. California Agric. Exp. Stn. Circ. 347.

Huggins, D.R., C.E. Clapp, R.R. Allmaras, and J .A. Lamb. 1995. Carbon sequestration in
com-soybean agroecosystems. p. 61-68. In R Lal et al. (ed.) Soil management and
greenhouse effect. Lewis Publ., Boca Raton, FL.

Jupp, AP, and El Newman. 1987. Morphological and anatomical effects of severe
drought on the roots of Lolium perenne L. New Phytol. 105:393-402.

121

Kayser, A.T., W. Wilcke, M. Kaupenjohann, and J.D. JoslirL 1994. Small scale
heterogeneity of soil chemical properties. I. A technique for rapid aggregate
fractionation. Z. Pﬂanzenernahr. Bodenk. 157:159-164.

Keith, H., J.M. Oades, and 1K Martin. 1986. Input of carbon to soils from wheat plants.
Soil biol. Biochem. 18:445-449.

Kemper, W.D., and RC. Rosenau. 1986. Aggregate stability and size distribution. p.425-
442. In A. Klute (ed.) Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9.
ASA and SSSA, Madison, WI.

Klute, A. 1986. Water retention: Laboratory methods. p. 635-662. In A. Klute (ed.)
Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA,
Madison, WI.

Lidieu, J., P. de Ridder, P. de Clerck, and S. Dautrebande. 1986. A method of measuring
soil moisture by time-domain reﬂectometry. J. Hydrol. (Amsterdam) 88:319-328.

Martin, A, A. Mariotti, J. Balesdent, P. Lavelle, and R Vuattoux. 1990. Estimate of
organic matter turnover rate in a savanna soil by 13C natural abundance
measurements. Soil Biol. Biochem. 22:517-523.

Materechera, S.A., A.R. Dexter, and AM. Alston. 1992. Formation of aggregates by
plant roots in homogeneised soils. Plant Soil 142:69-79.

Monroe, CD, and E]. Kladviko. 1987. Aggregate stability of a silt loam soil as affected
by roots of corn, soybeans and wheat. Commun. Soil Sci. Plant Anal. 18:1077-
1087.

Oades, J .M. 1988. The retention of organic matter in soils. Biogeochemistry 5:35-70.

Pietola, L., and AJ.M. Smucker. 1998. Fibrous carrot root responses to irrigation and
compaction of sandy and organic soils. Plant Soil 200:95-105.

Qian, J.H., J.W. Doran, and D. T. Walters. 1997. Corn plant contributions to root zone
available carbon and microbial transformations of nitrogen. Soil Biol. Biochem.
29:1451-1462.

Reid, J.B., and M.J. Goss. 1980. Changes in the aggregate stability of a sandy loam

effected by growing roots of perennial ryegrass (Lolium perenne). J. Sci. Food
Agric. 31:325-328.

122

Reid, J.B., and M.J. Goss. 1981. Effect of living roots of different plant species on the
aggregate stability of two arable soils. J. Soil Sci. 32:521-541.

Santos, D., S.L.S. Murphy, H. Taubner, A.J.M. Smucker, and R. Horn. 1997. Uniform
separation of concentric surface layers ﬁ'om soil aggregates. Soil Sci. Soc. Am. J.
61 :720-724.

Santos, D., and A.J.M. Smucker. 1998. Distribution of alfalfa root-derived organic carbon
in soil aggregates in tilled and non-tilled rotations. In preparation.

Skjemstad, J .O., R. P. Le Feuvre, and RE. Preble. 1990. Turnover of soil organic matter
under pasture as determined by 13C natural abundance. Aust. J. Soil Res. 28:267-
276.

Smith, B.N., and S. Epstein. 1971 . Two categories of 13C/ 12C ratios for higher plants.
Plant Physiol. 47:3 80-3 84.

Smucker, A.J.M., S.L. McBurney, and AK. Srivastava. 1982. Quantitative separation of

roots ﬁom compacted soil proﬁle by the hydro pneumatic elutriation system.
Agron. J. 74:500-503.

Smucker, A.J.M. 1990. Quantiﬁcation of root dynamics in agroecological systems.
Remote Sensing Review 5:237-248.

Smucker, A.J.M., A. Nunez-Barrios, and J.T. Ritchie. 1991. Root dynamics in drying soil
environments. Belowground Ecol. 2:4-5.

Tisdall, J.M., and J .M. Oades. 1979. Stabilization of soil aggregates by the root systems of
ryegrass. Aust. J. Soil Res. 17:429-441.

Tisdall, J.M., and J.M. Oades. 1980. The management of ryegrass to stabilize aggregates
of a red-brown earth. Aust. J. Soil Res. 18:415-422.

Tisdall, J.M., and J .M. Oades. 1982. Organic matter and water-stable aggregates. J. Soil
Sci. 33:141-163.

Topp, G.C., J.L. Davis, and AP. Annan. 1980. Electromagnetic determination of soil
water content: Measurements in coaxial transmission lines. Water Resour. Res.
16:574-582.

Utomo, W.H., and A.R. Dexter. 1981. Tilth mellowing. J. Soil Sci. 32:187-201.

Van Veen, J.A., and R]. Kuikman. 1990. Soil structural aspects of decomposition of
organic matter by micro-organisms. Biogeochemistry 11:213-33.

123

Table 4.1. Selected properties of aggregates (3-7 mm dia.) of the
Kalarmzoo loam soil used in the greenhouse experiment. Kellogg
Biological Station, Hickory Corners, MI. 1996.

 

 

Property Status

pH (water) 6.5

CEC 6.3 cmol kg“
Total C 6.3 g kg“
Total N 1.6 g kg“
6'3C signature -23.4 %o

P 0.12 g kg“
K 0.18 g kg“
Ca 0.91 g kg“
Mg 0.15 g kg"
Zn 4 mg kg“
Mn 24 mg kg“

 

124

Table 4.2. Dry weights of leaves, stems, shoots, and roots and shoot/root ratio of
ryegrass plants grown in a sand (75%) and soil aggregate (25%) matrix for 6
weeks under water stressed and non-stressed conditions (Means i standard error,

 

 

 

n = 8).
Leaf Stem Shoot Root1 Total R/S ratio
3
Ryegrass
Stress 10.29 :t 0.95 4.15 :i: 0.41 14.44 :1: 1.27 4.63 :t 0.61 0.32 :t 0.03
Non-stress 14.41 i 1.44 5.53 :i: 0.59 19.94 :1: 1.47 7.23 :h 0.64 0.37 :l: 0.03

 

1 Root data are the sum of all the roots in the two 1/4 sections of the top portions (0-15
cm) of each column, multiplied by two, plus the roots in the bottom (17-34 cm) portion
of each colunm.

125

 

 

 

 

 

 

3 a 3m 85 86 So a 3o 3 a an no a NS @582
3 a Nam 85 So a :3 So a o; S a E. 3 a 3m sea
a: a 2:. See a 33 god a :3 85 a 25 S a 2. 2: a an. aaaaoz
M: a mom Sod a Sod 85 a 86 2.0 a m; we a 3 S a 3K $25
3313
S a ”.2 Sod a wood :3 a :3 85 a :.o we a we we a 2K aeaaoz
we a can see a 33 So a :3 So a 35 no a 3 3. a as seam
€8.8qu
a: a go: :3 a ”86 23 a mud to a f on a q: of a 58 mmaaeoz
a.” a MS wood a 83c 36 a :5 no a 2 we a me 3 a :3 seem
a3
a5 amen 8cm
ﬂea as Ea a: A55 as as: we Ea «.8
m was a was m was N was _ ago
3.:
maﬁa as? see

 

.3 n e Hobo €3.85 um £836 .meomﬁcaoo commando: can @0mech $83 825 8.33 0
8m x55: Aﬁcmmv oumwoewwe :8 Ea Aﬁcmhv 65% a E :3on $53 $3on mo momma—o 523 88 Engage 26 mo mﬁwaoq .mé £an

126

 

 

 

 

127

 

2:. od 3 2 a a: 2 a 0.3 @582

So a 86 S a S no a to S a n: S a New anew
533$

85 3 do a to 3 a «a 2 a v.8 aoaaoz

Se a 85 ed 3 a I. Z a o.» 2 a as scam
3

:3 ed 2 ed a 2 ed a n; aoaaoz

85 a 3o 3 a do go a 3 3 a 3. 2 a 9% anew
€353

.ch a 3.3 3 a no no a 3 S a 4.2 2 a «a aoeaoz

Se in Se 3 a 3 we a .2 I a ta 2 a 3” seem

53343
a
ES 5 Es 4.: Ass we Ea we ES 2:
m ago 4 was m was N was _ £5

 

momma—o 623 “com

.8 n 2 vote @3983 a £836 .mcomﬁaoo commence: can 332% e8?» .598 5.83 0 com x59:
3°va ocmwocwwm :8 93 Agent 38 e a :3on 353 mmﬁwob .«o 3330 £6.59 88 33363 E .308 we omﬁcoocom .vé 05m...

 

 

~N.: H :50 m mo H 3.0 %d H cod 8.3 H madam 2.: H 3.3. cod H hmd mmcbmaoz
mo.wm H 8.? Ed H mmd cod H on; mmdw H mmcmm mm.: H Edm 3.0 H ovd mmobm
:.S H magma E; H 2.0 wmd H omN cmdm H dem 3%: H 3.: mod H omd mmobmaoz
$.63 H 3.30 2: H Nod 93 H Cl mcﬁm H 3&3 mm: H Quom m _.o H avd 38%
3
3.: H 3.36 36 H and emd H bmd ﬂm._m H $.ch as H 3.3 cod H mud £05202
3.? H $.on mod H Rd omd H mad ww.mm H woém _ mwé H vw.mm mod H wed mmobm
again
the? H madman mm: H 8.: mod H mos C.Mwo— H no.3» mm.EH mod: mmd H mmd mmobmcoz
modov H 3.32 v: H ond 36 H ewe 3.3 H modem mm.» H omdc ond H 3; $85
agaﬂdd
p.80 wed p.80 83 $83 p.88 AEV @
03% SAM >M <mm AM 33M

 

Aw n a veto @8986 H 332v .92:on mommy—mac: use 332% 833 .595
383 c com €88 AXES 8&8wa :8 use AﬁcmD 38 a E gem $.83 38ch mo mommﬂo £23 88 =m com .93,va humane
Bees 5 as. .546 bacon is. .3: 85.9 .29: 8a 8&3 .36 ems... .593 amp; be 82 Bob .3. 3.3

128

Table 4.6. 6‘3C signatures of shoot (n = 8) and root (n = 4) parts of ryegrass plants grown
in a sand (75%) and soil aggregate (25%) matrix for 6 weeks under stressed and non-
stressed water conditions. (Means 3: standard error).

 

 

 

 

Stressed Non-stressed
%o

Sham
-29.6 i 0.20 -30.1 :i: 0.23

R991

Ilium

0-5 cm -25.8 :t 1.40 -27.9 t 0.66
5-10 cm -26.0 :t 1.72 -28.1 :f: 0.72
10-15 cm -27.3 :t 0.71 -28.3 a: 0.79
Mean (n=12) -26.3 d: 0.73 -28.1 :t 0.38

 

129

.Eoﬁqmaco AMPS $0830an €8-08: onamxsﬁ was $8835 5 maﬁa maﬁa—co U>m 05 3 So cm was 3 .«o
gave 2: as 323:8 83>» .«o 3580.332: Ea :32“me a3. 8m vow: 889$ 2: mo cotaﬁomoaou vegaﬁm . _ .v oBmE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80 cm
Eu 0
“263252
80 ~ ~
Eu m
:8 S
B3280 aoumwt:
BSQEOoéoS

. x
- - o .
HIM E.— o \i
O .

 

 

 

 

.568 038 MO...

 

 

 

 

\

130

 

 

 

 

 

 

----- Sand+Aggregates
0.4 H; _._ Sand

'8

o

E 0.3 .-

E

8

c:

o

o

3

+5 0.2 -

3

.o

g

'6' 0.1 d ,

> s.

L; -------- -= .......... ., ................................. ..
0.0 I I I I I l I I 1 T l I I I I l I I r I
0 50 100 150 200

Matric potential (-kPa)

Figure 4.2. Water characteristic curves for the growth medium (silica sand and soil
aggregates) and pure silica sand used in the greenhouse experiment.

13]

.8838? :8

352388 mo ﬁwaohm 0:88 we 88808808 38833 8m wow: magnum 08 mo 83888082 28888385 .3» 0885

838885822

 

 

 

 

 

 

 

 

 

 

 

8:888 8282

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
 

_/8§_B 385

r

8&2wa zom v ‘

 

 

F

 

 

 

 

02% 3“: ll

 

 

H

 

ﬂ .l

.888 5&on lv

 

 

 

 

"L

 

 

132

Vol. water content (cm3 cm'3)

Vol. water content (cm3 cm'3)

0.35

 

 

 

 

 

 

 

 

............... Stress Control
0.30 .— Non-stress 10 cm
0.25 .
0.20 « .
0.15 4...... ,.. MW:
0.10 4
0.05 ‘
0.00 T I I I r I f I I
8121620242832 36 4044
0.35
0.30 ‘ W
0.25 ‘
0.20 .
0.15 4
0.10 I
0.05 ‘ Control
26cm
0.00 I I I I I r j I I
8 12 1620 2428 32 364044
Days after planting

 

 

 

 

 

 

    

 

0.35
Ryegrass

0.30 . 10 cm
0.25 -
0.20-
0.15 9 ‘
0.104
0.05 I \\ ;
0.00 I I T I I I I

81216 20 24 28 32 36 4044
0.35

Ryegrass
0.30 . .- ‘ 26 cm
0.25 ~
0.20 ~
i l,
0.15 ‘ llx‘
\\
0.104 x
1‘2“.»

0.05< 132"“.
0.00 I I T I I I I r I

8 12 16 202428 32 364044

Days after planting

Figure 4.4. Volumetric water contents in the control and ryegrass treatments, as a function

133

of water level applied (stressed and non-stressed plants), and position of the TDR
probes (10 and 26 cm depths). Mean values, n = 8.

 

 

12.5 12.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Stress Control Ryegrass
:1 Non-stress O - 5 cm 0 - 5 cm
.7“ 10.0 4 10.0 -
34°
39 7.5 — 7.5 J
I:
.8
§ 5.0 J 5.0 -
E
a 2.5 - 2.5 7
0.0 - 0.0 -
12.5 12.5
Control Ryegrass
5 - 10 cm 5 - 10 cm
.7" 10.0 - 10.0 -
5°40
on
v 7.5 4 W 7.5 -
c:
s a
a _ / _
o 5.0 % 5.0
g 2 5 - % 2 5 .
0.0 4 %i 0.0 -
Whole Outer Inner Whole Outer Inner

Figure 4.5. Total carbon contents in whole soil aggregates, outer layers, and inner regions
of soil aggregates ﬁ'om the control and ryegrass treatments in a sand (75%) and soil
aggregate (25%) matrix aﬁer 6 weeks under water stressed and non-stressed
conditions. Means 3: standard error, n = 4.

134

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 Stress Control Ryegrass
'22 4 El Non-stress 0 - 5 cm '22 d 0 - 5 cm
1°?
2., -23 4 1 -23 - T
.2 E I E
a I E __ E
MD -24 4 -24 - . T
"'00
I 4 E E
-25 4 -25 4 ‘—
Control Ryegrass
'22“ 5-10cm '22“ 5-10cm
’2? 3i
:5 -23 4 I -23 —
.2 I
e 1 I E E I
no -24 - -24 7 E E
"30
-25 7 -25 -
Whole Outer Inner Whole Outer Inner

Figure 4.6. 6‘3C ratios in whole soil aggregates, outer layers, and inner regions of soil
aggregates from the control and ryegrass treatments in a sand (75%) and soil aggregate
(25%) matrix alter 6 weeks under water stressed and non-stressed conditions. Means i
standard error, n = 4.

135

 

     

 

 

 

 

 

 

 

 

 

 

    
 

 

 

 

 

 

Stress 0 - 5 cm
100 - :3 Non-stress

g 80 4
8
5 5
'm 40 -
o ’ j-
E 20 - —- ‘ E
Z

0 _

100 _ 5 - 10 cm

3 80 -
8
g 60 -
§ , 7
la 40 -
U
3
a) 20 - T
z I -— % I

O —1

 

 

 

 

 

 

 

 

 

 

Whole Outer Inner

Figure 4.7. Percentage of new C3-carbon in whole soil aggregates, outer layers, and inner
regions of soil aggregates the control and ryegrass treatments in a sand (75%) and soil
aggregate (25%) matrix alter 6 weeks under water stressed and non-stressed
conditions. Means i: standard error, n = 4.

136

 

 

 

 

 

 

 

 

 

 

 

 

        

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5 5 II: Control O-Scmﬁ
Ryegrass
4 -
3 7 I. T
_L
24
12
A 0
E 5 - 5-10cm
I”;
.... 3T T
'3 E '//7// —L
rib 2‘
o
3 17
5
a)
2 0
5 . lO-lScm
4 a
3. I
T
2 . _l_
1.
0
Stress Non-stress

Figure 4.8. Mean weight diameter (MWD) of soil aggregate samples from the control and
ryegrass treatments in a sand (75%) and soil aggregate (25%) matrix after 6 weeks
under water stressed and non-stressed conditions. Means :1: standard error, 7: = 4.

137

 

300 7 :3 Control 0 ‘ 5 cm

250 4 Ryegrass

200 7
150 4 I
100 7
50 -

 

 

H

 

 

 

 

 

 

 

300 - 5 - 10 cm
250 7

200 7 I
150 7
100 7
50 7
0
300 7

 

 

Tensile strength (kPa)
1+1

 

 

 

 

 

 

 

10-15 cm
250 7
200 7
150 7
100 7
50 -

 

H
H

 

 

 

 

 

 

 

 

 

Stress Non-stress

Figure 4.9. Tensile strength of of soil aggregate samples from the control and ryegrass
treatments in a sand (75%) and soil aggregates (25%) matrix after 6 weeks under water
stressed and non-stressed conditions. Means 2!: standard error, n = 4.

138

GENERAL CONCLUSIONS

The improved version of the Soil Aggregate Erosion (SAE) apparatus, designed
and built for this study based on an earlier version (Chapter 1), was a useful tool for
peeling soil aggregates. This novel approach provided opportunities for quantifying the
accumulation of C and ions within speciﬁc regions of soil aggregates. Aggregate abrasion
measurements by the SAE apparatus appear to also provide a method for determining dry
soil aggregate stability.

Deposition of plant root C to soil aggregates was evaluated by locating recent
additions of C3-C within their concentric SAE-peeled layers using the 513C technique. The
results demonstrated greater concentrations of C accumulated on external layers than in
the interior regions of aggregates sampled ﬁ'om 0-5 cm depths. Approximately 70% of
root—deposited C accumulated on external layers of conventional tillage (CT) aggregates
during 20 months of alfalfa (C3-C) growth. Though total C inputs were higher in no—tillage
(NT), deposition of QC on external layers was lower than in aggregates ﬁ'om CT soils.

No-tillage treatments caused signiﬁcant increases in the stability of aggregates
> 0.5 mm. Following 11 years of NT, aggregate C concentrations increased with
improvements in macroaggregate stability at depths of 0-5 and 5-15 cm. Eleven years of
NT maize without N fertilization caused a reduction in aggregate C contents with little
modiﬁcation of aggregate stability. In soils under long-term cultivation (CT and NT),
macroaggregation was greatly reduced with a concomitant increase in the proportion of

microaggregates. NT showed clear trends in the sequestration of C during the

139

development of new macroaggregates and/or stabilization of old macroaggregates.
Therefore NT of this soil might be an important approach for increasing the storage and
protection of SOM which may be lost during CT.

Drought stresses changed ryegrass rooting patterns and activity as indicated by
decreases in RDW, RL, RSA, RV, RLD, and RDWD to depths of 10 cm. After 6 weeks
of ryegrass growth in a sand (75%) and soil aggregate (25%) matrix, C contents were
higher in whole soil aggregates, in both their outer and inner regions, when ryegrass was
irrigated. Under drought stress conditions, higher soil C contents were observed only
when whole aggregates were analyzed. Inputs of new C in whole aggregates were
estimated to be 62% in drought-stressed and 25% in non-stressed ryegrass treatments at
the 0-5 cm depth. New C accounted for 20 and 8% of the total C in the external layers and
inner regions of aggregates. Several constraints and difﬁculties are associated with studies
of plant drought stress under controlled conditions: i) Regulating water-stress regimes in
sand for some crops may be a limiting factor when using sand growth media instead of
soil; ii) To simulate ﬁeld processes drought stress in the greenhouse should develop
gradually; and iii) Soil volumes occupied by root systems and associated water availability
in the rooting zone are conﬁned to the column vs. fractions of the root system in ﬁeld-
grown plants.

Several new questions were raised by this study which deserve ﬁirther
investigation: i) NT soils appeared to accumulate more C from the previous maize crop
and/or slow the decomposition of the newly deposited C from alfalfa; ii) Developments in

aggregate abrasion techniques and soil sampling provide a new processing approach for

140

peeling aggregates of several sizes for further conﬁrmation of C deposition patterns within
soil structural units; iii) Wet sieving methods have some limitations as redistribution of
SOM among the sieved aggregate ﬁactions my be produced by the method itself, iv)
optimization of the test conditions, e.g., pretreatment of soil aggregates and duration of
sieving are required before wet sieving can be universally applied to C studies for all soil
types; and v) Additional research using different soils and crop species, accompanied by
wetting/drying cycles and sequential soil sampling would provide further insights into the
factors aﬁ°ecting root C deposition into aggregates and the effects of soil C on the

chemical and physical processes of soil aggregate stability.

141

APPENDICES

Appendix A1. Selected characteristics of bulk surface soil samples of conventional tillage
(CT), no-tillage (NT), and never-tilled successional grassland (NeTSG) of the
Agroecosystems Interations site at WK. Kellogg Biological Station, Hickory
Corners, MI.

 

 

 

 

 

 

Particle size
Treatment Org. C N WSA pH’ CEC’
Sand Silt Clay
g kg" % cmolc kg'l
CT 421 431 147 12.4 1.1 78.5 6.3 9.0
NT 422 443 135 11.3 1.0 87.6 5.5 10.0
NeTSG 391 468 141 18.2 1.6 90.3 5.3 8.9

 

WSA = water-stable aggregates (> 250 um).

’ From Daroub (1994).

142

40m

Historically tilled area Never-tilled plots

 

 

.—‘—-—q——.—-

4 8 12 5 16 18 20 22 24
CT NT NT CT _ NeT NeT NeT NeT.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 7 ll 15 17 19 21 23
NT CT CT NT NeT NeT NeT NeT.
NF F NF F LS: SG. SM 3
2 6 10 14 km
CT NT NT CT CT Conventional tillage
NF NF F 1 F NT No—tillage

F Fertilized

 

 

NF Non-fertilized

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 5 9 13 NeT Never-tilled
CT NT NT CT SG Successional grass
vegetation
F F NF NF
SM Successional grass
with maize
I 27 m I

Appendix A2. Plot layout and sampled plots (shaded) of the Agroecosystems Interactions
site at W.K. Kellogg Biological Station. Hickory Corners, MI.

143

Appendix A3. Steps for the progressive and quantitative erosion of concentric layers

ﬁom soil aggregates by the mesoSAE chambers. (Smucker, A.J.M., J. Wagester, and
D. Santos, unpublished).

Pre-cleam'ng

1.

Place 10-20 grams of coarse sand particles, which can be retained on the screen of the
SAE chambers to pre-clean them. Place the chambers into the bases, cap the chambers
with plastic bottle tops or rubber stoppers, enclose in foam supports, and place them
onto a rotational shaker. Run for 30 minutes at 300 rpm.

Note: In order to secure the mesoSAE chambers onto the shaker it may be necessary
to place them into beakers. Try placing 100mL glass beakers into the clamps on the
rotational shaker. Then cut pieces of foam about 25 cm long and 5 cm wide. Back the
foam with pieces of duct tape. Place the mesoSAE chambers in the middle of the foam
with the taped side to the outside. Wrap the foam around the mesoSAE chamber and
then place them into the 100mL beakers which are secured to the clamps on the
rotational shaker.

Discard the sand and thoroughly brush the knurled walls of the SAE chambers with a
wire brush or toothbrush. Clean the screen of the SAE clmrnbers by turning them
upside down and brushing the screen with a steel wire brush. First, use a stiff painting
brush to remove any material from the base and then compressed air to blow the SAE
chambers and the bases clean. Inspect for trapped material and remove it.

Separately weigh the clean bases and the clean SAE chambers to the nearest 0.1 mg.

Record the weights in a permanent ﬁle.

144

Preliminary Erosion Data

4. Select a minimum of 25 uniform air-dry soil aggregates ﬁom the same sample.
Aggregates 4-5 mm in diameter are recommended, however, aggregates ﬁom 0.2 mm
to 18 m can be abraded in the mesoSAE chambers. Weigh each aggregate to the
nearest 0.1 mg and record the weight. Gently place it onto the screen of the SAE
chamber. Weigh the SAE chamber with the aggregate inside of it and weigh the base.
Record the weights. The weight of the SAE chamber with the aggregate inside of it
and the weight of the base will be the weights for time 0. This will expedite the
collection of preliminary soil erosion data. Repeat this procedure for each of the 25
selected aggregates.

5. Cap the 25 mesoSAE chambers and rotate them at 300 rpm for separate runs of 1, 2,
3, 4, 5, 6, 9, 12, 24, 48, and 60 minutes. Note: Each time includes the sum of the
previous times, i.e., a time of 12 minutes is 9 of the minutes already run with three
additional minutes of rotation. If one was to calculate the amount of minutes of actual
rotation it would be the following: 1, 1, 1, 1, 1, 1, 3, 3, 12, 24, and 12 minutes. After
each run, weigh the SAE chamber with the aggregate and the particles inside of it, and
weigh the base with the eroded materials in it. For maximum accuracy, the sum of the
SAE chamber with the aggregate inside of it and the base at time 0 should always
equal the sum of the SAE chamber with the aggregate in it and the base with the
eroded particles in it for all timed runs.

6. Once all of the timed runs have been completed, examine the contents of each SAE

chamber. If the center of the aggregate remains, separately weigh the aggregate center,

145

 

the contents of the SAE chamber, and the contents of the base. To weigh an aggregate
or an aggregate center, extract the aggregate or aggregate center and place it onto a
piece of tared weighing paper that is on a balance. Record the weight to the nearest
0.1 mg. Remove the weighing paper with the aggregate or aggregate center on it. Re-
tare the balance with a new sheet of weighing paper. Use this weighing paper for the
contents of the SAE chamber. Firmly slam the SAE chamber (upside down) onto the
weighing paper, which should be placed on a hard, smooth surface. While inverted on
the paper, brush the screen clean with a steel wire brush so that the dislodged particles
are added to the SAE chamber sample. Weigh the sample. Re-tare the balance with a
new piece of weighing paper. Slam the base (upside down) onto the tarred weighing
paper to remove its contents. Brush out any remaining material from the base into the
sample using a stiff painting brush. Weigh the sample.

. If an aggregate center remains, place it into the mesoSAE chamber and rotate it for an
additional 60 minutes to obtain a 120 minute run. Separately weigh the aggregate
center (if it remains), the contents of the SAE chamber, and the contents of the base.
Repeat the weighing procedure outlined in step 6. The sum of the contents of the SAE
chamber, the contents of the base, and the remaining aggregate center (if it remains)
should equal the weight of the aggregate center at time 60. If the aggregate center
remains after 120 minutes, rotate it in the mesoSAE chamber for an additional 60
minutes to create a 180 minute run. Repeat the weighing procedure used for the 120

minute run.

146

 

8. For each run of l, 2, 3, 4, 5, 6, 9, 12, 24, 48, 60, 120, and 180 calculate an erosion
weight per minute (mg per minute). Determine how long each run should be to extract
the external layer, the transitional layer, and the internal layer. Each layer is 1/3 of the
weight of the aggregate at time 0. For the external layer, calculate the amount of time
it took to erode 1/3 of the weight of the aggregate at time 0. For the transitional layer
calculate the time to erode 2/3 of the weight of the aggregates at time 0. The internal
layer would be the aggregate center remaining after the transitional layer has been
removed by abrasion. It should be 1/3 of the aggregate weight at time 0.

9. An erosion weight per minute can also be determined by using inﬂection points, which
are points where the aggregate begins to erode more slowly (i.e., the external [outer]
layer, transitional layer, and internal layer [region] described by the two inﬂection
points in Fig. 2 (Santos et al., Soil Science Society of America Journal, 61 :720-724,
1997). On Kalamazoo loam soils these two points are between 2 and 6 minutes and
12 and 24 minutes. The soil erosion rate declines ﬁom milligrams to 0.1 milligrams per
minute between 2 and 6 minutes and then ﬁom 0.1 milligrams to 0.001 milligrams per
minute between 12 and 24 minutes. Once the erosion times for separating soil
aggregates into three distinct components (i.e. layers) have been established, rotation
of the mesoSAE chambers only needs to be stopped at these previously established
times.

Sample Aggregate Erosion
10. Select 12 closely uniform aggregates from each sample (use more aggregates if more

soil is needed for the amlyses of each layer). Weigh each aggregate on a piece of tared

147

ll.

12.

weighing paper and then place it into the bottom of the SAE chamber. Place the SAE
chamber into the base and cap the top. Rotate the mesoSAE chambers on the shaker at
300 rpm for the amount of time to erode 1/3 of the weight of each aggregate. Note:
Layers are eroded at different rates for each aggregate.

Weigh the aggregate to determine how much has been eroded by abrasion. Weigh the
sand-grain sized contents of the SAE chamber and store it in a labeled glass
scintillation vial if desired. Weigh the contents of the base and store them in a glass
scintillation vial. The contents of the base are the external layer. Clean both the base
and the SAE chamber as done in step 2. Place the remaining aggregate into the SAE
chamber, put the SAE chamber into the base, cap the top, and rotate for the amount of
time to erode halfof the weight of the remaining aggregate (e.g., the second 1/3 of the
weight at time 0). This would be the time that was calculated to erode 2/3 of the
aggregate at time 0. Therefore, at the end of the run the remaining aggregate should be
1/3 of the weight of the aggregate at time 0. Note: If any aggregates become broken
during the abrasion of either the external or transitional layers, discard the sample and
begin the abrasion process again with new aggregates. Figure 1 shows the location of
the externaL transitional, and internal layers of a soil aggregate.

Weigh the aggregate to determine how much has been eroded by abrasion. Weigh the
contents of the SAE chamber and store them in a labeled glass scintillation vial if
desired. Weigh the contents of the base and store them in a labeled glass scintillation
vial. The contents of the base are the transitional layer. Clean both the base and the

SAE chamber as done in step 2. Place the remaining aggregate, which is the internal

148

 

13.

14.

15.

layer, into the SAE chamber, place the SAE chamber into the base and cap the top.
Rotate it until the internal layer has been completely eroded so that all of the ﬁne
particles have passed through the screen into the base.

Note: The abrasion of the internal layer (region) may take several hours or even days
of rotation.

Weigh the contents of the SAE chamber and store it in a labeled glass scintillation vial
if desired. Weigh the contents of the base and store it in a labeled glass scintillation
vial. The contents of the base are the internal layer. Clean both the base and the SAE
chamber as outlined in step 2.

Select new aggregates and repeat the procedure as outlined in steps 10 — 13. Note:
When using diﬂ‘erent soil types or samples that have received very different treatments,
repeat the procedure as outline in steps 4-13. This will provide greater accuracy in the
collection of aggregate layers.

Process each layer or region of these peeled soil aggregates according to the

experimental protocol.

149

    

HICHIGRN STQTE UNIV LIBRARIES

13,1113 I“ H 157141174111

1293017663042