Am.

. .u. ,n
i“

g. ..
”Fawn"
, . .
1.

..

 

a... . . we
asififififi _ «Km/w

. ‘

a
m t
. .5.»

 

 

in:

i fury: Wm
Owl}? t: ..1 .iat

. 6
1415.37»!!- A

. :
dun-.11.
$35.;

 

 

 

 

 

.04 I:
:4 IV" . . Ila. 111$}VJ13 P.‘ .L. , ‘ adIa..K-,.......l :. :53... .9... ;.3ur1«.. 41,

This is to certify that the
thesis entitled

CHANGES IN MICROBIAL COMMUNITY RESPONSES
TO GRADIENTS OF CARBON, NITROGEN, AND
WETTING CYCLES IN CONCENTRIC LAYERS OF
SOIL MACRO-AGGREGATES

presented by

Heather Ann Holdaway

has been accepted towards fulfillment
of the requirements for the

degree In Cr j and Soil Sciences

 

 

WWQMM

Majoflbrofessor’ 5 Signature

W3

Date

MSU is an Afiinnative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p. 15

 

CHANGES IN MICROBIAL COMMUNITY RESPONSES TO
GRADIENTS OF CARBON, NITROGEN, AND WETTING CYCLES IN
CONCENTRIC LAYERS OF SOIL MACRO-AGGREGATES

By

Heather Ann Holdaway

A THESIS

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

MASTER OF SCIENCE

Department of Crop and Soil Sciences

2003

ABSTRACT
CHANGES IN MICROBIAL COMMUNITY RESPONSES TO
GRADIENTS OF CARBON, NITROGEN, AND WETTING CYCLES 1N
CONCENTRIC LAYERS OF SOIL MACRO-AGGREGATES
By

Heather Ann Holdaway

This study examined the effect of carbon (C), nitrogen (N), and wetting cycle gradients
within soil aggregates on microbial communities and populations. New methods of
bacterial extractions and identification were developed for this study. Direct microscopic
counts of bacterial populations were 35% higher in moist than air dry aggregates.
Bacterial populations from contrasting management systems and soil types were native
forest on silt loam soils > conventional-till on silt loam soils 5 native grassland on loam
soils > conventional-till planted to continuous alfalfa on sandy loam soils. A 17% and
32% increase in populations occurred in the interior regions of aggregates exposed to 3
and 6 W/D cycles. Populations in the exterior regions increased 23% with 3 W/D cycles
and decreased 29% with 6 W/D cycles. A decrease in polar tensile strength was observed
in aggregates exposed to 6 W/D cycles. This decrease in strength may be caused by
decreased populations in the exterior regions, leading to fewer microbial by—products
produced and lower aggregate strength. C and N concentrations and other factors appear
to contribute to bacterial populations. Direct relationships exist between aggregate
stability, porosity, and bacterial populations. As greater energy sources are available to
soil bacterial populations, those populations increase in biovolume and density. This
increase in size and population leads to greater bacterial by-products that in turn lead to

increased aggregate formation and stabilization.

This work is dedicated to my parents, Bruce and Kimberly Dopp, for their unending love
and support in all that I do, and instilling in me the confidence to accomplish my goals
and dreams.

iii

ACKNOWLEDGEMENTS

I would like to express my utmost gratitude and appreciation to my mentor and major
professor Dr. Alvin Smucker for his guidance and support during my research. I also
extend my appreciation to Dr. James Tiedje and Dr. Frank Dazzo for their time, advice,
and assistance given while serving on my M.S. Guidance Committee. Special thanks are
given to Dr. Eun-jin Park, Gavin Scott, Joseph Holdaway, Sunjeet Sidhu, Jake Gerst,
Karin Aditjandra, Sharon Stevens, Stefan Nava, Cory McDonald, and Matt VanderHoek
for their assistance with my research.

I want to acknowledge the Michigan Agricultural Experiment Station cooperative
agreement with the Scottish Agricultural College for financial support.

I would also like to acknowledge the NSF/Long Term Ecological Research (LTER) for
Agroecosystems Research at the Kellogg Biological Station (KBS) at Hickory Corners,
MI and the long term site at the Ohio Agricultural Research and Development Center
(OARDC) at Wooster, OH, for providing samples for this research.

Finally, my love, gratitude, and appreciation are extended to my family and friends, for

their support, encouragement, and patience throughout this program.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
INTRODUCTION ............................................................................................................... 1
CHAPTER 1: LITERATURE REVIEW
Soil Aggregate Formation ........................................................................................ 4
Clay and SOM Effects on Aggregation ................................................................... 6
Microbial Effects on Aggregation ........................................................................... 6
Carbon in Aggregates .............................................................................................. 9
Wetting/Drying Cycles .......................................................................................... 10
Organic Matter and Aggregate Stability ................................................................ ll
Concentric Aggregate Layers and Aggregate Stability ......................................... 12
Bacterial Extraction from Soil ............................................................................... l7
Fluorescence Microscopy and Image Analysis ...................................................... 18
REFERENCES .................................................................................................................. 21
CHAPTER 2: DEVELOPMENT AND RESULTS OF METHODS
Soil Dispersion and Bacterial Extraction ............................................................... 25
Staining .................................................................................................................. 31
Microscopy ............................................................................................................ 35
Image Analysis (pre-CMEIAS) ............................................................................. 35
Image Analysis with CMEIAS .............................................................................. 36
Image Analysis (post-CMEIAS) ............................................................................ 39
Conclusion ............................................................................................................. 42
REFERENCES .................................................................................................................. 45

CHAPTER 3: SOIL BACTERLAL COMMUNITY DIVERSITY IN CONCENTRIC
AGGREGATE LAYERS AS INF LUENCED BY SOIL TYPE AND MANAGEMENT
PRACTICE

INTRODUCTION ............................................................................................................. 46
MATERIALS AND METHODS ....................................................................................... 47
Site Description and Soils ...................................................................................... 47
Bacteria] Analyses (Preparation, Microscopy, and Image Analysis) .................... 48
Aggregate Analyses ............................................................................................... 49
Statistical Analysis ................................................................................................. 50
RESULTS AND DISSCUSION ........................................................................................ 50
Bacteria] Populations ............................................................................................. 50
Cell Biovolume Measurements .............................................................................. 52
Cell Biovolume Distributions ................................................................................ 54
Aggregate Stability ................................................................................................ 67

Aggregate Porosity ................................................................................................. 70

C and N Concentrations ......................................................................................... 7O
Bacterial Morphotype Analysis ............................................................................. 73
SUMMARY AND CONCLUSIONS ................................................................................ 75
REFERENCES .................................................................................................................. 77

CHAPTER 4: WETTING AND DRYING MODIFICATIONS OF SOIL MACRO-
AGGREGATE STABILITIES: BIOPHYSICAL MECHANISMS

INTRODUCTION ............................................................................................................. 79
MATERIALS AND METHODS ....................................................................................... 81
Site Description and Soils ...................................................................................... 81
Wetting and Drying Cycles .................................................................................... 82
Bacterial Analyses (Preparation, Microscopy, and Image Analysis) .................... 82
Aggregate Analyses ............................................................................................... 84
Statistical Analysis ................................................................................................. 84
RESULTS AND DISSCUSION ........................................................................................ 85
Aggregate Strength ................................................................................................ 85
Bacterial Populations ............................................................................................. 85
Cell Biovolume Measurements .............................................................................. 89
Cell Biovolume Distributions ................................................................................ 90
C and N Concentrations ......................................................................................... 93
SUMMARY AND CONCLUSIONS ................................................................................ 93
REFERENCES .................................................................................................................. 95
CHAPTER 5: CONCLUSIONS ........................................................................................ 97
APPENDIX ...................................................................................................................... 100

vi

LIST OF TABLES

Table 1.1. Light filter requirements for fluorescent staining ............................................ 19

Table 2.1. Summary results of CMEIAS-l biovolume formula (Dazzo et al., 2003a;
personal communication). Morphotype populations that were tested include cocci, spiral,
U/curved rod, regular rod, unbranched filament, branched filament, rudimentary
branched rod, prosthecate, ellipsoid, and club ................................................................... 37

Table 2.2. Classification accuracy of specific morphotypes using CMEIAS (Lui et al.,
2001). Accuracy test results of the CMEIAS-2 classifier using a large dataset of ground
truth images (from Table 4 of Lui et al., 2001). Rows indicate true class labels; columns
indicate CMEIAS classifier-assigned labels and numbers of correctly and incorrectly
classified cells in each morphotype class. For example, CMEIAS classified 1212 of 1217
cocci correctly (99.6% accuracy), and incorrectly classified 4 cocci as regular rods and 1
cocci as an ellipsoid ........................................................................................................... 38

Table A. 1. Overall summary of data for soils compared ................................................ 103

vii

LIST OF FIGURES

Figure 1.]. Soil aggregate erosion (SAE) chamber .......................................................... 13
Figure 1.2. Rotary shaker containing soil aggregate erosion (SAE) chambers ................ 15
Figure 1.3. Concentric layers of a soil aggregate .............................................................. 16

Figure 2.1. Extracted bacteria per gram of soil at three dilution rates (1:100, 1:150,
1:200) for 6.3-9.5mm native forest (NF) soil aggregates from Hoytville clay loam (Fine,
illitic, mesic Mollic Epiqualfs); n=3 .................................................................................. 27

Figure 2.2. Mean cell biovolume measurements determined by CMEIAS-1 biovolume
formula at three dilution rates (1: 100, 1:150, 1:200 soilzdistilled water) for 6.3-9.5mm
native forest (NF) soil aggregates from Hoytville clay loam (Fine, illitic, mesic Mollic
Epiqualfs); n=3 (aggregates and images) ........................................................................... 29

Figure 2.3. Ratio of bacterial cells to soil particles by direct count for a Kalamazoo loam
(Fine-loamy, mixed, mesic Typic Hapludalf); n=3 ........................................................... 30

Figure 2.4. Mean cell biovolume measurements (A) and cell biovolume distribution (B)
as determined by CMEIAS at 1:200 dilution and two sonication times (55, 105) for 6.3-
9.5mm native forest (NF) soil aggregates from a Hoytville clay loam (Fine, illitic, mesic
Mollic Epiqualfs); n=1 aggregate, n=3 images .................................................................. 32

Figure 2.5. Effects of sedimentation time (15-90 minutes) and 5 second sonication (60 S)
on the ratio of bacterial cells to soil particles for a Kalamazoo loam soil (Fine-loamy,
mixed, mesic Typic Hapludalf); n=3. The 608 is the bacteria to soil ratio after sample
has been exposed to 55 of sonication and sedimentation time is 60 minutes .................... 33

Figure 2.6. Ratio of bacterial cells to soil particles for a Kalamazoo loam (F inc—loamy,
mixed, mesic Typic Hapludalt). Aggregates are of size fraction 4-6.3mm across, stained
with 5-(4,6-dichlorotriazin-2-yl)aminofluorescein (DTAF), acridine orange (A0), and
phenolic aniline blue (PAB). Dilution of 1:200, sedimentation time of 60 min, sonication
time of 53; n=3 aggregates and images .............................................................................. 34

Figure 2.7. Example of CMEIAS data preparation macro; A=frequency distribution,

=% frequency distribution, and C=% cumulative frequency distribution. Cell
biovolume frequency distributions of the exteriors of two field replications of Wooster
native forest (NF) aggregates of size fractions 4-6.3mm; n=l6-17 aggregates, 10 images
each replication .................................................................................................................. 40

Figure 2.8. Example of CMEIAS sampling statistics macro. Relative rank abundance

plot of morphological diversity in Wooster native forest (NF) aggregates of size fractions
4-6.3mm. The letter-number label corresponds to the operational morphological unit

viii

(morphotype size) or morphological diversity classified by CMEIAS (Dazzo et al., 2003);
n=16 aggregates, 10 images ............................................................................................... 41

Figure 2.9. Example of CMEIAS data analyzed in EcoStat; proportional similarity (%)
of interiors (I) and exteriors (E) of Wooster native forest (NF) and conventional-till (CT)
aggregates of size fractions 4-6.3mm across; n=13-17 aggregates, 10 images each
treatment ............................................................................................................................ 43

Figure 3.1. Declining populations of bacteria per gram of exterior and interior soil layers
as soil textures become more coarse. Soil samples from a Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) native forest 01 F ) were sampled from the surface 0-
5cm, a Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native grassland
(NG) succession and a Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) were sampled
from the surface 0-10cm, from soil aggregate size fractions 4-6.3mm across; n=7-21
aggregates, n=20-30 images per layer (dependent on number of field replications),
n=446-1687 cumulative cells sampled. Bars indicate standard deviations ....................... 51

Figure 3.2. Extracted bacteria are significantly lower in cell density from air dry vs.
moist whole aggregates from a Kalamazoo loam (Fine-loamy, mixed, mesic Typic
Hapludalt) native grassland (NG) succession from soil aggregate size fractions 4-6.3mm
across, sampled from the surface O-IOCm; n=3 aggregates, n=20 images, n=446-1687
cumulative cells sampled. Bars indicate standard deviations ........................................... 53

Figure 3.3. Mean cell biovolume measurements determined by CMEIAS-l biovolume
formula for air dry and moist whole aggregates from Kalamazoo loam (Fine-loamy,
mixed, mesic Typic Hapludalt) native grassland (NG) succession from soil aggregate size
fractions 4-6.3mm across, sampled from the surface O-lOCm,; n=3 aggregates, n=20
images, n=750-1014 cumulative cells sampled. Bars indicate standard deviations ......... 55

Figure 3.4. Mean cell biovolume measurements determined by CMEIAS-l biovolume
formula for exterior (E) and interior (1) layers of Wooster silt loam (Fine-loamy, mixed,
mesic Oxyaquic Fragiudalfs) native forest (NF) aggregates, taken from the surface 0-
5cm, Kalamazoo loam (F inc-loamy, mixed, mesic Typic Hapludalf) native grassland
(NG) succession aggregates and Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) aggregates from
soil aggregate size fractions 4-6.3mm across, sampled from the surface 0-10cm; n=7-21
aggregates, n=20-3O images per layer (dependent on number of field replications),
n=446-1687 cumulative cells sampled. Bars indicate standard deviations ....................... 56

Figure 3.5. Increased numbers of cells were observed in moist aggregates, while a larger
biovolume was observed in the air dry aggregates. The distribution of cell biovolumes
were calculated by the CMEIAS data preparation macro for moist and air dry whole
aggregates from Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native
grassland (NG) succession from soil aggregate size fractions 4-6.3mm across, sampled

ix

from the surface 0-10cm; n=3-4 aggregates, n=10 images per treatment, n=750-1014
cumulative cells sampled ................................................................................................... 5 7

Figure 3.6. Increased numbers of cells were observed in the exteriors of aggregates from
3 field replications when compared to their interiors. Biovolumes of these cells showed
no significant differences between regions of the aggregate or field replication. The
distribution of cell biovolumes were calculated by the CMEIAS data preparation macro
for exterior (E) and interior (1) layers of aggregates from 3 field replications (1,2,3) of
Kalamazoo loam (F inc-loamy, mixed, mesic Typic Hapludalf) native grassland (NG)
succession from soil aggregate size fractions 4-6.3mm across, sampled from the surface
O-lOcm; n=16-21 aggregates, n=10 images per replication and layer, n=446-732
cumulative cells sampled ................................................................................................... 58

Figure 3.7. Biovolumes of these cells showed no significant differences between regions
of the aggregate or field replication. The distribution of cell biovolumes were calculated
by the CMEIAS data preparation macro for exterior (E) and interior (1) layers of
aggregates from 2 field replications (2,8) of Ternary sandy loam (Sandy loam, mixed,
frigid Alfic Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) from
soil aggregate size fractions 4-6.3mm across, sampled from the surface O-lOcm; n=7-10
aggregates, n=10 images per replication and layer, n=457-520 cumulative cells sampled59

Figure 3.8. Increased numbers of cells were observed in NF aggregates compared to CT
aggregates, with the exteriors of the aggregates having increased numbers than their
interiors. Cell biovolumes were decreased in the interior region of CT aggregates. The
distribution of cell biovolumes were calculated by the CMEIAS data preparation macro
for exterior (E) and interior (1) layers of aggregates from 2 field replications (2,8) of
Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) native forest (NF)
and conventional-till (CT) from soil aggregate Size fractions 4-6.3mm across, sampled
from the surface 0-5cm; n=l3-l7 aggregates, n=10 images per management system and
layer, n=1091-l687 cumulative cells sampled .................................................................. 61

Figure 3.9. Increased cell biovolumes were observed in the moist vs. air dry whole
aggregates from Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native
grassland (NG) succession from soil aggregate size fraction 4-6.3mm across, sampled
from the surface O-IOCm. Biovolume frequency distributions were calculated using a
constant bin width of0.1um3 by the CMEIAS data preparation macro and by EcoStat;
n=3 aggregates, n=20 images, n=750-1014 cumulative cells sampled .............................. 62

Figure 3.10. Increased cell biovolume frequencies in the exteriors (E) (A) of aggregates
from 3 field replications (1,2,3) were observed following the trend (largest to smallest)
rep3 > repl > rep2, while interiors (I) (B) were rep3 > rep2> repl, from Kalamazoo loam
(Fine-loamy, mixed, mesic Typic Hapludalf) native grassland (NG) succession from soil
aggregate size fraction 4-6.3mm across, sampled from the surface O-lOcm. Biovolume
frequency distributions were calculated using a constant bin width of O. lum3 by the
CMEIAS data preparation macro and by EcoStat; n=16-21 aggregates, n=10 images per
replication and layer, n=446-732 cumulative cells sampled .............................................. 64

Figure 3.11. Increased cell biovolume frequencies were observed in the interiors (I) vs.
exteriors (E) of 2 field replications [2(A),8(B)] from Ternary sandy loam (Sandy loam,
mixed, frigid Alfic Haplorthodes) conventional- till planted to continuous alfalfa (CT- CA)
from soil aggregate size fractions 4- 6. 3mm across, sampled from the surface 0- 10cm.
Biovolume frequency distributions were calculated using a constant bin width of0.1um3
by the CMEIAS data preparation macro and by EcoStat; n=7- 10 aggregates, n=10 Images
per replication and layer, n=457-520 cumulative cells sampled ........................................ 65

Figure 3.12. Increased cell biovolume frequencies were observed in the exteriors (E) of
CT vs. NF aggregates (C), while Slight increases were seen in the interiors (D).
Decreased cell biovolume frequencies were observed in the exteriors vs. interiors (I) of
NF aggregates and increased for CT aggregates from Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) native forest (NF) (A) and conventional- till (CT) (B)
from soil aggregate size fractions 4- 6. 3mm across, sampled from the surface 0- 5cm.
Biovolume frequency distributions were calculated using a constant bin width of0.1um3
by the CMEIAS data preparation macro and by EcoStat; n=13-17 aggregates, n=10
images per management system and layer, 1091-1687 cumulative cells sampled ............ 66

Figure 3.13. Similar cell biovolume distributions were observed between field
replications (1,2) of exterior (E) and interior (1) regions of NF aggregates (A,B), while
increased variability of field replications was observed for exterior and interior regions of
CT aggregates (C,D) from Wooster silt loam (F inc-loamy, mixed, mesic Oxyaquic
Fragiudalfs) native forest (NF) (A, B) and conventional-till (CT) (C, D) from soil
aggregate size fractions 4- 6. 3mm across, sampled from the surface 0- Scm. Biovolume
frequency distributions were calculated using a constant bin width of 0.1um3 by the
CMEIAS data preparation macro and by EcoStat; n=13- 17 aggregates, n=10 Images per
management system and layer, n=1091-1687 cumulative cell sampled ............................ 68

Figure 3.14. Mean weight diameter (MWD) by wet sieving for Wooster silt loam (F ine—
loamy, mixed, mesic Oxyaquic Fragiudalfs) native forest (NF) aggregates, sampled from
the surface 0-5cm, Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native
grassland (N G) succession aggregates and Ternary sandy loam (Sandy loam, mixed,
frigid Alfic Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA)
aggregates, sampled from the surface O-lOcm. Samples are from soil aggregate size
fractions 4-6.3mm across; n=3. Bars indicate standard errors .......................................... 69

Figure 3.15. Porosity by Saran method for Wooster silt loam (Fine-loamy, mixed, mesic
Oxyaquic Fragiudalfs) native forest (NF) and conventional-till (CT) aggregates, taken
from the surface 0-5cm, and Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA). Soil aggregate
size fractions 4-6.3mm across were sampled from the surface O-lOcm; n=10. Bars
indicate standard errors ...................................................................................................... 71

Figure 3.16. Extracted bacteria per gram soil and C (A) and N (B) concentrations for

exterior (E) and interior (1) layers of Kalamazoo loam (Fine-loamy, mixed, mesic Typic
Hapludalf) native grassland (N G) succession aggregates and Ternary sandy loam (Sandy

xi

loam, mixed, frigid Alfic Haplorthodes) conventional-till planted to continuous alfalfa
(CT-CA). Soil aggregate size fractions 4-6.3mm across were sampled from the surface
O-lOcm; n=7-21 aggregates, n=10 images per management system and layer, n=446-732
cumulative cells sampled, n=2 replications for C and N data. Bars indicate standard
deviations ........................................................................................................................... 72

Figure 3.17. Extracted bacteria per gram soil and C (A) and N (B) concentrations for
exterior (E) and interior (1) layers of Wooster silt loam (Fine-loamy, mixed, mesic
Oxyaquic F ragiudalfs) native forest (NF) and conventional-till (CT). Soil aggregate size
fractions 4-6.3mm across were sampled from the surface O-5cm; n=13—l7 aggregates,
n=10 images per management system and layer, n=109l-l687 cumulative cells sampled,
n=2 replications for C and N data. C and N data from samples in 2000. Bars indicate
standard deviations ............................................................................................................. 74

Figure 4.1. Increased polar tensile strengths (PTS) with 3 W/D cycles and decreased PTS
with 6 W/D cycles for native forest (NF) aggregates from Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) of aggregate size fraction 4-6.3mm across, sampled
from the surface 0-5cm, exposed to O, 3, and 6 wetting-drying (W/D) cycles; n=10
aggregates. Bars indicate standard errors .......................................................................... 86

Figure 4.2. Increased populations of bacteria were observed in the exteriors and interior
of aggregates exposed to 3 W/D cycles. The increase in populations continued in the
interior regions through 6 W/D cycles, while a decrease was observed in the exteriors of
aggregates exposed to 6 W/D cycles. Extracted bacteria per gram soil for exterior and
interior layers of native forest (NF) aggregates from Wooster silt loam (F inc-loamy,
mixed, mesic Oxyaquic Fragiudalfs) of aggregate size fraction 4-6.3mm across, sampled
from the surface 0-5cm, exposed to O, 3, and 6 wetting-drying (W/D) cycles; n=13-16
aggregates, n=10 images per WD cycle and layer, n=7l7-1141 cumulative cells sampled.
Bars indicate standard errors .............................................................................................. 87

Figure 4.3. Increased cell biovolumes were observed in the interior regions of aggregates
exposed to 3 W/D cycles. The mean cell biovolume was calculated by CMEIAS-1
biovolume formula for exterior and interior layers of native forest (NF) aggregates from
Wooster silt loam (F inc-loamy, mixed, mesic Oxyaquic F ragiudalfs) of aggregate Size 4-
6.3mm across, taken from the surface 0-5cm, exposed to 0, 3, and 6 wetting-drying
(W/D) cycles; n=13-16 aggregates, n=10 images per WD cycle and layer, n=717-1 141
cumulative cells sampled. Error bars indicate standard deviations .................................. 89

Figure 4.4. Changes among the % abundances of cell biovolume distributions were
observed with 3 W/D cycles at 0.3 pm3 cell biovolume. Increases were also observed in
the exteriors of O and 6 W/D cycles, and in the interiors of 3 and 6 W/D cycles. The
biovolume frequency distributions were calculated using a constant bin width of 0.1um3
by the CMEIAS data preparation macro and by EcoStat for native forest (NF) aggregates
from Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) of aggregate
size fraction 4-6.3mm across, sampled from the surface 0-5cm, exposed to 0, 3, and 6

xii

wetting-drying (WD) cycles; n=13-l6 aggregates, n=10 images per WD cycle and layer,
n=717-1 141 cumulative cells sampled. ............................................................................. 92

Figure A. 1. Modified bacterial extraction and staining procedure ................................. 101

Figure A.2. Editing procedure for CMEIAS analysis of DTAF stained images ............ 102

xiii

INTRODUCTION

Microorganisms are present in nearly every habitat found in the soil ecosystem.
Soil bacteria are highly numerous and diverse. On average, a gram of soil contains 109
bacteria and some 20,000+ different species (Paul and Clark, 1996; Brady and Weil,
2002). A major role of soil bacteria is the breakdown of organic material, thus releasing
nutrients for uptake by other organisms. Soil bacteria also directly release compounds to
the soil through excretions, which benefit other microbial communities and help bind soil
particles together (Tisdall and Oades, 1982; Kandeler and Murer, 1993; Oades 1993;
Tisdall, 1994; Paul and Clark, 1996; Monreal and Kodama, 1997; Hillel, 1998; Brady and
Weil, 2002).

Microbial diversity within aggregates seems to be as large as the microbial
diversity identified across extensive geographical areas (Grundmann and Normand,
2000). This localized diversity provides a huge resource of multifunctional microbial
communities which may be modified for a specific function (Loeffler et al., 1999).
Communities of microorganisms located in the surface and interior regions of aggregates
can rapidly alter their metabolisms, thereby resulting in completely different microbial
communities (Blackwood, 2001). External more robust bacterial communities are
thought to contribute to the function and strength of soil aggregates through carbon (C)
mineralization, while more C mineralization and sequestration appears to be associated
with microbial populations within internal regions associated with intra-aggregate micro-
porosities (Kandeler and Murer, 1993; Smucker et al., 1996, 1998; Bending et al., 2002;

Dell et al., 2004; Blackwood et al.).

Soil aggregation processes and wetting/drying (W/D) cycles modify microhabitats
by changing pore networks and sequestering C within newly formed microsites
associated with interiors of aggregates. Previous approaches for identifying bacterial
communities and populations include soil dispersion, bacterial extraction, and fluorescent
staining and microscopy. These techniques have been optimized for bulk soil samples
(Bloem, 1995; Paul et al., 1999). Image analysis procedures of bacterial communities
within concentric layers of soil aggregates can be used to quantify bacterial populations
associated with microhabitats located on surfaces and within central regions of macro-
aggregates. Biovolume distributions can also be used to identify gradients of bacterial

biomass from aggregate surfaces to their interiors.

The objectives and hypotheses of this study were:
1. To quantify changes in bacterial populations and communities within concentric
layers of soil aggregates containing measured gradients of C and nitrogen (N)
compounds.

Hypothesis 1: Large concentrations of C and N compounds contained in the

 

surface layers of aggregates contribute to greater aggregate stability by promoting
larger populations of bacterial communities.

Corollary 1.1: The highest positive correlations between C and N concentrations

 

and bacterial communities will be observed in the surface layers of aggregates.
2. To determine if uniform distributions of C and N compounds, generated by more
frequent wetting-drying cycles, within aggregates contribute to more uniform

populations of bacteria and more water stable aggregates.

Hypothesis 2: The absence of C and N gradients will increase the uniformity of
bacterial distributions and communities within aggregates.

Corollary 2.1: More uniform concentrations of total C and N will increase the
water stability of whole aggregates.

Corollary 2.2: Increased concentrations of C and N compounds will be positively
correlated with greater water stability of whole aggregates.

Corollary 2.3: There will be a positive correlation between increased uniformity

of bacterial populations and the increased water stability of whole aggregates.

New methods of bacterial extractions and identification, including the expanded
application of the Center for Microbial Ecology Image Analysis System (CMEIAS), a
computer-assisted image analysis program, for soil bacteria were developed to verify the
above objectives. Combinations of a newly developed soil aggregate erosion (SAE)
method with conventional C and N analyses were integrated to prove the objectives and

hypotheses listed above.

CHAPTER 1

LITERATURE REVIEW

Soil functions as a reservoir for water, nutrients, gases, soil organic matter
(SOM), and microorganisms. Stable soil aggregate formation is crucial for the
sustainability of food production systems, sequestration of carbon (C), and the
remediation of wastes and pollutants (Tisdall and Oades, 1982; Beare et al, 1994;
Monreal and Kodama, 1997; Hillel, 1998; Brady and Weil, 2002). Increased stability of
macro-aggregates increases C sequestration by soil aggregates. This leads to better water
infiltration and aeration porosity, reduced water and wind erosion, and decreased crust
formation on aggregates and the soil surface (Kandeler and Murer, 1993; Edgerton et al,
1995; Monreal and Kodama, 1997; Hillel, 1998; Park and Smucker, 2003). Aggregate
formation and stability results from the interactions of many factors, including C and
nitrogen (N) content, microbial and mesofaunal activities, root exudates, clay minerals,

and wetting/drying (W/D) frequencies.

Soil Aggregate Formation

Understanding what occurs within soil aggregates may explain many of the
processes beneficial for food and fiber production, and how aggregates contribute to
remediation processes that sequester soil C and detoxify anthropogenic pollutants. In
addition to functioning as a reservoir for water, nutrients, gases, microorganisms, and

SOM, soils facilitate massive quantities of micro- and macro-organismal activities.

The soil matrix, a blend of solid and porous phases, results from the flocculation
and cementation of soil textural particles, ions, SOM, and associated microbial and
mesofaunal communities into organized and aggregated structures (Hillel, 1998). These
aggregation processes are dependent on the interactions of C and N compounds,
microbial and mesofaunal activities, root exudates, types of clay minerals, parent
materials, W/D frequencies, time, and other factors (Oades, 1993). Total C, soil bulk
density and porosity, saturated hydraulic conductivity (Ks), and root activities are not
Significantly correlated with increased water stabilities of aggregates (Rasse et al., 2000).
However, they demonstrated that other factors, such as C sources from shoot mulch and
root decomposition, had positive effects on aggregate stability.

Soil management practices have a profound effect on soil structure and aggregate
stability (Tisdall and Oades, 1982; Kandeler and Murer, 1993; Beare at al., 1994; Hillel,
1998; Brady and Weil, 2002). Conventional tillage (CT) practices incorporate crop
residues and fertilizers into the plow layer using a moldboard plow to lift, twist, and
invert the soil. These soils are subjected to extreme changes in soil water content and
temperature. During these extremes, weaker aggregates are frequently broken along
planes of weakness and substantial quantities of SOM are lost to accelerated
mineralization by microbial respiration. Further geochemical decomposition, such as
weathering, leads to reduced SOM, lower particulate organic matter (POM) content
(Beare et al., 1994), and diminished energy resources for soil microbial activities
(Monreal and Kodama, 1997). Reduced or no-tillage (NT) practices use minimal or no
plowing, leaving plant residues on the soil surface. These soils generate higher quantities

of stable soil aggregates at or near the soil surface. These surface soils have nearly

double the amount of total C sequestered within soil aggregates associated with at least
10 years under NT management (Smucker et al., 1996; Dell et al., 2004). Plant residues
on the surfaces of NT soils also protect the soil from erosion by sheltering aggregates

from wind and rain, thereby preserving soil aggregate structure.

Clay and SOM Effects on Aggregation

Clay mineralization ratios and total C are higher in the external regions of
aggregates in CT, NT, and grassland (GL) soils (Smucker et al., 1996). As aggregate size
increased, accumulations of clay in the external increased and clay content of internal
regions of aggregates decreased (Kavdir, 1996). External clay accumulations may be due
to more numerous W/D cycles which align the particles into uniform orientations, leading
to a build up of highly oriented clay particles on the surfaces of the aggregates.
Accumulations may also be due to increased microbial populations in the external layers
of aggregates. Higher levels of C in the external regions of aggregates, which provide
energy sources for microbial communities, and higher aeration in the internal regions

lead to increased microbial populations (Sexstone et al., 1985).

Microbial Effects on Aggregation

Aggregate stability is dependent on soil microorganisms (Tisdall, 1994).
Edgerton et a]. (1995) demonstrated a positive linear correlation between stable macro-
aggregates and microbial biomass C. Changes in the soil environment that affect
microbial communities are also thought to affect soil structure (e.g., aggregation and

aeration) (Kandeler and Murer, 1993; Edgerton et al., 1995). Bacterial biomass C

increased in the presence of macro-aggregates, and decreased where no macro-aggregates
were present (Denef et al., 2001). Wetting and drying cycles of soils with no macro-
aggregate formation led to a reduction in the bacterial activity.

Microbial activities can promote the development of anaerobic regions in soil
aggregates when soil aeration porosities are low, leading to the process of denitrification.
It is expected that greater microbial populations and different microbial communities
would be located in the exterior regions of aggregates from a majority of improved soil
management systems. Results obtained in the Soil Biophysics Lab at Michigan State
University (MSU) showed lower bulk densities in the exterior layers than the interior
layers of aggregates in the 4-6.3mm size fraction across 3 management systems, native
forest (NF), no-till (NT), and conventional—till (CT) for Wooster silt loam soil, sampled in
1998 (Park et al., 2001). These data demonstrated higher porosities in the exterior layers
of aggregates and presumably accelerated the diffusion of atmospheric oxygen (02) and
increased microbial respiration rates of the sequestered C and N compounds. Additional
results comparing bulk densities from three aggregate size fractions (24, 4-6.3, and 6.3-
9.5mm) showed consistent bulk densities through all size fractions for CT systems (Park
et al., 2001). The 2—4mm size fraction of NT systems had higher bulk densities than the
4—6.3 and 6.3-9.5mm fractions. Bulk densities of the forest soil aggregates increased as
the aggregate size fractions decreased (e.g., the lowest bulk density for the 6.3-9.5mm
aggregate fractions and highest bulk density for the 2-4mm fraction).

Microbial diversity within aggregates seems to be as large as the microbial
diversity identified across extensive geographical areas (Grundmann and Normand,

2000). This localized diversity provides a huge resource of multifunctional microbial

communities which may adapt for specific functions (Loeffler et al., 1999).
Communities of microorganisms located in the surface and interior regions of soil
aggregates often alter the net production of metabolic products altering the communities
within aggregates (Blackwood, 2001). Microorganisms detoxify their surroundings
various ways; for example through electron consumption during different processes of
dechlorination (Loeffler et al., 1999), by developing pH and ion concentration gradients
within soil aggregates (Horn, 1990; Smucker et al., 1998), by altering oxygen
concentrations (Sexstone et al., 1985), and by utilizing multiple sources of C and N
compounds within single aggregates (Santos et al., 1997; Smucker et al., 1998; Dell et al.,
2004). These activities result in the biogeochemical processes associated with the
advective-dispersive flow of solutes, the production of specific greenhouse gases (GHG),
C sequestration potentials, plant nutrient sources, and aggregate stability.

Blackwood (2001) reported no change in the composition of microbial
populations due to aggregate size. However, the location and community structures of
the microbial populations appeared to change within individual aggregates. This was
shown by differences in the terminal-restriction fragment length polymorphism (T-RFLP)
composition of microbial populations in aggregates between their internal and external
concentric layers. T-RFLP is a polymerase chain reaction (PCR) based method that
produces electrophoretic bands with lengths representative of the different forms of the
gene of interest (Ogram and Shanna, 2002). Analysis for the dissimilarity of the 16s
rDNA gene determines the diversity of microbial communities in soil.

Sissoko (1997) reported that soluble plant root C and N increased soil microbial

biomass (SMB) and the water stability of soil aggregates. Uniquely different microbial

communities (Blackwood et a1), and specific gradients of C and N (Smucker et al., 1998)
from external layers toward internal regions of soil aggregates have been observed among
aggregates from 3 different soil types.

Microorganisms fulfill many important roles in the soil profile. They promote
soil aggregation processes by providing cementing agents, promote plant grth through
symbiotic relationships (e.g., N-fixing bacteria, mycorrhizal fungi, and plant growth-
promoting rhizobacteria), control root diseases (biocontrol agents), fix N (free-living N-
fixing bacteria), assimilate plant nutrients bound in organic matter (OM) or soil minerals
(P from rock phosphate), and degrade pollutants (Oades, 1993; Tisdall, 1994; Paul and

Clark, 1996).

Carbon in Aggregates

Total C within soil aggregates is higher in the external layers than the internal
layers (Smucker et al., 1996; Horn, 1990; Park et al., 2001; Dell et al., 2004). Most of
this additional C found in the external layers is part of the labile or active C pools
associated with the microbial biomass, or is dissolved organic C (DOC) and
mineralizable POM (Dell et al., 2004). Soil tillage, periodic W/D cycles, oxygen
diffusion, and other abiotic processes are projected to contribute to gradients of C and N
substrates within aggregates (Smucker et al., 1996). Concentrating greater quantities of
C and N compounds, available to microbes in external regions, leads to Specific locations
for accelerated microbial growth and their associated contributions to increased soil

aggregation.

Wetting/Drying Cycles

Aggregate stability and formation is influenced by W/D cycles. Six W/D cycles
have been shown to increase aggregate stability (Utomo and Dexter, 1982). Sissoko
(1997) found aggregate stability, mean weight diameter (MWD), was not affected by
three W/D cycles, but was enhanced with nine W/D cycles when C and N compounds
were added during the wetting phase.

Dissolved organic matter (DOM) and other C and N substrates are delivered to
the internal regions of soil aggregates by repeated W/D cycles of soil aggregates (Utomo
and Dexter, 1982). Each re-hydration of dry aggregates transports and distributes
additional soil organic carbon (SOC) into the intra-aggregate pore networks of stable soil
aggregates. Microbial communities located in the exterior layers of aggregates appear to
be able to alter these C-sources and produce specific metabolites that contribute to
aggregate stability (Christensen et al., 1990).

Denef et al. (2001) witnessed a continuous cycle in regards to W/D cycles,
aggregate stability, and microbial activity. After a couple W/D cycles the OM became
physically protected and was no longer accessible for microbial attack. This led to
decreased microbial activity, decreased production of binding agents, and decreased
aggregate stability. Upon additional W/D cycles, the less stable aggregate falls apart,
freeing up previously inaccessible C for microbes. This leads to increased microbial
activity, increased production of binding agents, and increased aggregate formation and
stability. When the C source becomes depleted, the cycle of aggregate stability and

microbial activity begins again.

10

Organic Matter and Aggregate Stability

Organic matter is crucial for forming and stabilizing soil aggregates (Brady and
Weil, 2002). Organic compounds act as cementing agents and provide the energy
substrates needed for most bacterial and firngal activities (Tisdall and Oades, 1982). As
organic residues are decomposed, gels and other viscous microbial by-products promote
aggregate formation by acting as gluing agents for particles and OM. These organic
polymers consist of polysaccharides, polyuronides, and glomalin (a glycoprotein
produced by arbuscular mycorrhizal fungi) (Paul and Clark, 1996; Hillel, 1998; Brady
and Weil, 2002). Polysaccharides, polyuronides, and other nonhumic compounds
originating from plants and microorganisms constitute a large group of semi-flexible
molecules capable of forming multiple bonds with several mineral particles
simultaneously. These complex organic polymers chemically bond with particles of
silicate clays and iron and aluminum oxides, orienting and forming bridges between
individual soil particles and cementing them together into water-stable aggregates (Brady
and Weil, 2002). The polymers attach to clay surfaces by means of hydrogen bonding,
anionic and cationic interactions, and van der Waals forces (Hillel, 1998). In cases where
organic polymers are unable to penetrate individual clay micelles, they form protective
films around micro-aggregates. In other cases, solutions of labile organic compounds
penetrate the interior of soil aggregates, forming co-precipitate cementing agents that can
be further mineralized biologically. Increasing the quantity of organic substrate within
the aggregate enables microbial communities to redesign and strengthen the interior

regions of the aggregate.

ll

Functional diversity of soil microbial communities is affected by crop residues
and the quality of SOM (Bending et al., 2002). The quality of SOM directly affects
microbial communities by nutrient substrates available and indirectly by soil aggregation
and aeration. Bending et al. (2002) also found that increased enzyme activity, which
indicates larger microbial communities, increased as native SOM increased. Root
exudates include aliphatic amino and aromatic acids, amides, and sugars (Paul and Clark,
1996). Soil management practices that result in accumulation of organic C (e.g., NT,
manure application, cover crop use), result in increased microbial biomass levels and
changes in microbial community structure (Peacock et al., 2001). Paul and Clark (1996)
reported that additions of simple organic compounds to soils changed their metabolic
profiles and altered the diversity of soil microbial communities. Community response
characteristics vary according to the nature of the compound added (Degens, 1998).
Various applications of manure and N lead to different microbial communities. Peacock
et a1. (2001) found manure applications led to increased microbial biomass, activity, and
diversity, while ammonium nitrate (inorganic N) applications led to decreased microbial
biomass, activity, and diversity. The response of the microbial community to manure
applications is similar to a rhizosphere response where the community is highly active

metabolically and rapidly responding to additions of C (Peacock et al., 2001).

Concentric Aggregate Layers and Aggregate Stability
Soil aggregate erosion (SAE) chambers developed by Smucker et a1. (1998) allow
for surface layers to be eroded from soil aggregates and kept for further soil analyses,

retaining sample purity (Figure 1.1). Individual soil aggregates in single SAE chambers

12

 

Figure 1.1. Soil aggregate erosion (SAE) chamber.

are rotated on a rotary platform shaker (Figure 1.2) leading to the mechanical erosion of
soil materials from the exteriors of each aggregate, as it revolves against the knurled
inside wall of the erosion chamber. The aggregates are weighed at intervals to determine
the amount of the aggregate removed (Figure 1.3).

Soil aggregate stability can be determined by various destructive methods. One is
the wet sieving method (Kemper and Rosenau, 1986). Mean weight diameter (MWD) is
calculated by the following formula, where x, is the mean diameter of the aggregate size
ranges separated by sieving, and w, is the weight fraction based on the total oven dry

weight of the original aggregates that were recovered in that size range:

MWD = Ex,- w,-
i=1

Slow hydration of dry aggregates by nebulization, prevents the breakage of most soil
aggregates during the wetting process (Kemper and Rosenau, 1986).

Another destructive method for determining aggregate Stability is polar tensile strength
(PTS). The PTS of a soil aggregate is the force per unit area that is required to break the
aggregate (Dexter and Kroesbergen, 1985; Hillel, 1998). In this method, the force
applied to the aggregate, required to break the aggregate, is incorporated to the following
equation to calculate PTS, where do is the effective diameter of each aggregate (mm), drn
is the mean aggregate diameter (mm), m is the dry mass of the aggregate (kg), m0 is the
mean mass of the aggregates in the subsample (kg), F is the crushing force required to
break the aggregate (kg m 3'2), b is the measured balance reading (kg), y is the

acceleration force due to gravity (9.807 m 5'2), and Y is the tensile strength (kPa):

l4

 

 

 

Figure 1.2. Rotary shaker containing soil aggregate erosion (SAE) chambers.

15

 

Figure 1.3. Concentric layers of a soil aggregate.
External (E), transitional (T), and internal (1) layers.

dc = dm (m/mo)”3
F = by
Y = 0. 567F/(dc)2

PTS also varies with management practices. Usually, NF aggregates are the highest and
CT are the lowest, and as aggregate size increases, PTS decreases (Aditjandra et al.,

2002)

Bacterial Extraction from Soil

Soil dispersion and slide preparation techniques for extracting bacteria from bulk
soil samples have been developed by Paul et al. (1999) and Bloem (1995). Modifications
of these techniques can provide adequate extraction of bacteria for analysis of air dry and
moist soil aggregates. Lindahl (1996) showed increased soil dispersion and bacterial
counts when pyrophosphate, a chemical dispersing agent, replaced distilled water in the
dispersion and extraction procedure, leading to increased extraction efficiency.
Pyrophosphate disrupts soil aggregates by deflocculating the Clay present in the
aggregate. Pyrophosphate displaces the cations adsorbed to the clay with sodium, which
increases the hydration of the clay micelles leading to repulsion of the micelles (Hillel,
1998). Although the use of pyrophosphate increases soil dispersion and the extraction
efficiency of bacteria there are disadvantages of its use. Fluorescent stains behave
differently with samples dispersed with pyrophosphate, leading to changes in the cell
color (Lindahl, 1996). Pyrophosphate also increases the pH of the soil solution from 6.9
with distilled water to 7.9. This pH change is thought to alter the physiology of bacteria
thus leading to metabolic changes of the bacteria modifying staining results (Lindahl,

1996).

17

Fluorescence Microscopy and Image Analysis

Fluorescent stains and microscopy interfaced with image processing make it
possible to determine the size, shape, and numbers of organisms in the soil matrix (Harris
and Paul, 1994). Several stains may be used for bacterial biomass estimates. 5-(4,6-
dichlorotriazin-2-yl)aminofluorescein (DTAF) is a fluorescent, anionic stain which gives
low levels of background staining, producing images with good contrast (Bloem, 1995).
DTAF binds covalently to neutral amino groups of proteins, specifically to cell surfaces.
One disadvantage of using DTAF is that it is unstable and needs to be prepared fresh
daily. Acridine orange (A0) is a fluorescent, cationic stain which gives high background
staining, especially in clay and loam soils (Bloem, 1995; Paul and Clark, 1996). Phenolic
aniline blue (PAB) is also a fluorescent stain which gives high background staining. For
fluorescent stains, filters need to be used on the light microscope, for the fluorochromes
to fluoresce. The excitation and emission wavelengths for light filters for DTAF, A0,
and PAB are located in Table 1.1 (Paul et al., 1999; Carl Zeiss Microscopy, Germany).

The best method for viewing soil microbes is by soil smears. This is because
there is less background staining and fading of fluorochromes, and the smears are flat
allowing for better image analysis when compared to a filter. Smears can also be stored
for up to a year in the dark at 2°C (Bloem, 1995).

Bacteria can be analyzed by computer-assisted image analysis or manual
observation. Some benefits to computer-assisted image analysis include higher accuracy
in enumeration of bacteria and higher consistency. Reduced photo-bleaching of the

fluorochromes occurs by faster digital acquisition of the image (Paul et al., 1999). The

18

Table 1.1. Light filter requirements for fluorescent staining (Paul et al., 1999; Carl Zeiss
Microscopy, Germany).

 

Stain Excitation (nm) Emission (nm) Color
DTAF 450-490 525 green
AO 440-480 526 orange
PAB 455 600 blue

 

l9

Center for Microbial Ecology at Michigan State University has a computer-assisted
microscopy program, Center for Microbial Ecology Image Analysis System (CMEIAS).
CMEIAS analyses can determine microbial abundance (bacterial cell density, biovolume,
biomass C), morphological diversity (11 predefined cell shapes, 5 undefined cell shapes,
cell size), and spatial distribution (nearest neighbor distances, cells/area, percent
microbial cover) (Lui et al., 2001; Dazzo et al., 2003). It is believed that CMEIAS can be
efficiently used to quantify bacterial morphotypes more accurately with less time

required than by manual analyses.

20

REFERENCES

Aditjandra, K., E]. Park, and A.J.M. Smucker. 2002. Tensile Strength Domains within
Soil Aggregates from Contrasting Ecosystems. ASA Abstracts, Indianapolis, IN.

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.

Bending, G.D, M.K. Turner, and J .E. Jones. 2002. Interactions between crop residue and
soil organic matter quality and the functional diversity of soil microbial communities.
Soil Bio. Biochem. 34: 1073-1082.

Blackwood, CB. 2001. Spatial organization in soil bacterial communities. Ph.D.
Dissertation. Michigan State University, East Lansing.

Blackwood, C.B., C.J. Dell, A.J.M. Smucker, and EA. Paul. Eubacterial community
response to position within soil macroaggregates and soil management. (in preparation).

Bloem, J. 1995. Fluorescent staining of microbes for total direct counts. Molecular
Microbial Ecology Manual 4.1.8: 1-12.

Brady, NC. and RR. Wei]. 2002. The Nature and Properties of Soils. 13th ed. Prentice
Hall, Upper Saddle River, NJ.

Christensen, S., S. Simkins, and J.M. Tiedje. 1990. Spatial variation in denitrification:
Dependency of activity centers on the soil environment. Soil Sci. Soc. Am. J. 54:1608-
1613.

Dazzo, F.B., J. Lui, A. Prabhu, C. Reddy, M. Wadekar, R. Peretz, R. Bollempalli, D.
Trione, B. Marshall, J. Zurdo, H. Hammond, J. Wang, M. Li, D. McGarrell, A. Gore, J.
Maya-Flores, S. Gantner, and N. Hollingsworth. 2003. CMEIAS: Integrative software
package to strengthen microscopy-based approaches for understanding microbial
ecology. Kellogg Biological Station-Long Term Ecological Research All Scientists
Meeting (KBS-LTER ASM), September 12, 2003.
http://lterkbs.msu.edu/Meetings/2003_All_Inv_Meeting/Abstracts/dazzo.htm.

Degens, BR 1998. Microbial functional diversity can be influenced by the addition of
simple organic substrates to soil. Soil Biol. Biochem. 30: 1981-1988.

Dell, C.J., A.J.M. Smucker, and EA. Paul. 2004. Carbon mineralization from concentric
layers of soil aggregates. (submitted).

21

Denef, K., J. Six, H. Bossuyt, S.D. Frey, E.T. Elliot, R. Merckx, and K. Paustian. 2001.
Influence of dry-wet cycles on the interrelationship between aggregate, particulate
organic matter, and microbial community dynamics. Soil Biol. Biochem. 33:1599-1611.

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

Edgerton, D.L., J .A. Harris, P. Birch, and P. Bullock. 1995. Linear relationship between
aggregate stability and microbial biomass in three restored soils. Soil Bio. Biochem.
27:1499-1501.

Grundmann, G.L and P. Normand. 2000. Microscale diversity of the genus Nitrobacter in
soil on the basis of analysis of genes encoding rRNA. Appl. Environ. Microbiol.
66:4543-4546.

Harris, D. and EA. Paul. 1994. Measurement of bacterial growth rates in soil. Appl. Soil
Eco. 1:277-290.

Hillel, D. 1998. Soil Structure and Aggregation. p. 101-125. In Environmental Soil
Physics. Academic Press, San Diego, CA.

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

Kavdir, Y. 1996. Some physical and mineralogical characteristics of soil aggregates.
Special Topic Research Project.

Kandeler, E. and E. Murer. 1993. Aggregate stability and soil microbial processes in a
soil with different cultivation. Geoderrna 56:503-513.

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, Physical and Mineralogical
Methods, 2nd edition. Soil Sci. Soc. Am., Madison, WI.

Lindahl, V. 1996. Improved soil dispersion procedures for total bacterial counts,
extraction of indigenous bacteria and cell survival. J. Microbiol. Methods 25:279-286.

Loeffler, F .E., J.M. Tiedje, and RA. Sanford. 1999. Fraction of electrons consumed in
electron acceptor reduction and hydrogen thresholds as indicators of halorespiratory
physiology. Appl. Environ. Microbiol. 65:4049-4056.

Lui, J., F.B. Dazzo, O.Glagoleva, B. Yu, and AK. Jain. 2001. CMEIAS: A computer

aided system for the image analysis of bacterial morphotypes in microbial communities.
Microbial Ecology 41: 173-194.

22

Monreal, CM. 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. 1993. The role of biology in the formation, stabilization, and degradation of
soil structure. Geodenna 56:377-400.

Ogram, A. and K. Sharma. 2002. Methods of Soil Microbial Community Analysis. p.
554-563. In Manual for Environmental Microbiology, 2nd edition. American Society for
Microbiology, Washington, DC.

Park, E.J. and A.J.M. Smucker. 2003. Intra-aggregate porosity and saturated hydraulic
conductivity of single aggregates in conventional-till, no-till agroecosystems, and native
forest soils. Soil Sci. Soc. Am. J. (in preparation).

Park, E.J., K.L. Aditjandra, and A.J.M. Smucker. 2001. Spatial variability of porosity,
tensile strength, and mineralogy within soil aggregates. ASA Abstracts, Charlotte, NC.

Paul, EA and F .E. Clark. 1996. Methods for studying soil microorganisms. p. 35-68 In
Soil Microbiology and Biochemistry, 2nd edition. Academic Press, San Diego, CA.

Paul, E.A., D. Harris, M.J. Klug, and R.W. Ruess. 1999. The determination of microbial
biomass. p. 291-317. In Standard Soil Methods for Long-Term Ecological Research.
Oxford University Press, New York, NY.

Peacock, A.D., M.D. Mullen, D.B. Ringelberg, D.D. Tyler, D.B. Hedrick, P.M. Gale, and
DC. White. 2001. Soil microbial community responses to dairy manure or ammonium
nitrate applications. Soil Bio. Biochem. 33:1011-1019.

Rasse, D.P., A.J.M. Smucker, and D. Santos. 2000. Alfalfa root and shoot mulching
effects on soil hydraulic properties and aggregation. Soil Sci. Soc. Am. J. 64:725-731.

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

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

Sissoko, F. 1997. Enhancement of soil aggregation by the combined influences of soil
wetting and drying and root-microbial associations. MS. Thesis. Michigan State
University, East Lansing.

Smucker, A.J.M., D. Santos, and Y. Kavdir. 1996. Concentric layering of carbon,
nitrogen, and clay minerals within soil aggregates from tilled and non-tilled
agroecosystems and natural grasslands. Third Eastern Canada Soil Structure Workshop
Proceedings. p. 129-140.

23

Smucker, A.J.M., D. Santos, Y. Kavdir, and E. Paul. 1998. Concentric gradients within
stable soil aggregates. Proceedings of the 16‘h World Congress of Soil Science,
Montpellier, France.

Tisdall, J.M. 1994. Possible role of microorganisms in aggregation of soils. Plant Soil
159:115-121.

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

Utomo, W.H. and AR. Dexter. 1982. Changes in soil aggregate water stability induced
by wetting and drying cycles in non-saturated soil. J. Soil Sci. 33:623-637.

24

CHAPTER 2

DEVELOPMENT AND RESULTS OF METHODS

In recent years, bacterial extraction techniques have included soil dispersion and
preparation of samples on slides for viewing through microscopes. These techniques
focused on bulk soil samples and required a 10g sample (Bloem, 1995; Paul et al., 1999).
When working with soil aggregates, especially concentric layers of aggregates, the
amount of soil per aggregate is limited in quantity (e. g., on average a 4-6.3mm aggregate
weighs 0.15g). This amount is decreased even more when identifying bacterial
populations in concentric layers of the aggregate. Therefore, modifications of the above
techniques are needed to allow for adequate extraction of bacteria from the small sample

sizes associated with the exterior and interior regions of soil aggregates.

Soil Dispersion and Bacterial Extraction

Modifications to the bulk soil bacterial extraction process by Bloem (1995) and
Paul et a1. (1999) which required 10g samples was initiated by determining the minimal
quantity of soil required to extract representative numbers of bacteria from concentric
layers of aggregates. The aggregate size fraction selected for comparing bacterial and
soil C analyses was 4-6.3mm across. Therefore, the average weight of whole aggregates
was approximately 0.15g. Since the minimal weight needed was 10g for conventional
extraction methods, 67 aggregates would be required. The number of aggregates needed
to complete bacterial analyses on concentric aggregate layers (exterior 33% and interior

33%) would be 200 aggregates. Extracting bacteria from these many soil aggregates

25

created difficulties as too many samples, having contrasting and highly variable bacterial
populations would be combined. Combining Samples from large numbers of aggregates
forfeited the possibility of identifying spatial variabilities of bacteria among soil
aggregates.

A reasonable amount of soil from aggregates was defined as the amount that
would give the best representation of the variability of the soil aggregates from that
specific soil type. For instance, assuming that a 4-6.3mm aggregate weighs 0.15g, 3
whole aggregates would be needed to run statistical evaluations on the data collected.
The total amount of soil would therefore be 0.45g. To make calculations simpler the
weight was rounded up to 0.5g. So on average, 3-4 whole aggregates or 9-12 aggregates
sub-divided into 3 concentric layers would supply the amount of soil needed for bacterial
analyses on aggregates.

A dilution study was conducted to determine if the dilution rate of 1:200 of
soil:distilled water was adequate for the smaller amount of soil processed. The efficacies
of dilution rates of 1:100, 1:150, and 1:200 were determined by image analysis. The
quantity of bacteria was used to calculate the extracted bacteria per gram of soil at each
rate. These data are found in Figure 2.1 and Show that the dilution study was linear.
Another important part of the study was to observe the amount of overlapping of cells in
the images acquired. In order for the image analysis system to accurately identify cells,
the image must be segmented so that the individual cells are not touching. This
overlapping of cells will result, for instance, in 2 cells being analyzed as one. To prevent
this, the cells need to be separated during the editing step, before image analysis. This

increases the amount of time spent on editing images. In the 1:200 dilution, little to no

26

8.0E+08

7.0E+08 7*

6.0E+08

 

y = 2E+0§x :45308
R2 = 0.9117

4.0903 - ~- - - f -

5.0E+08 *7 ,7

3.0E+08 ————- —— — ———

Bacteria/g soil

2.0E+08 * 7*

1.0E+08 ,- - - - -~ - - -

 

0.0E+00 e
50 100 150 200 250

Dilution rate (1:x)

 

Figure 2.1. Extracted bacteria per gram of soil at three dilution rates (1:100, 1:150,
1:200) for 6.3-9.5mm native forest (NF) soil aggregates from a Hoytville clay loam (Fine,
illitic, mesic Mollic Epiqualfs); n=3.

27

overlapping of cells was observed. However, with the lesser diluted samples (1:100 and
1:150) there were increased numbers of cells overlapping. The last parameter looked at
in the dilution study was the mean biovolume of cells analyzed by CMEIAS. No
significant differences in the standard errors of cell biovolumes per gram were observed
across the three dilution rates (Figure 2.2). Therefore, dilutions of 1:200 were used in the
soil dispersion and extraction of bacteria for identifying the spatial distributions of
bacteria in soil aggregates.

The next step in adapting the soil bacterial extraction procedure of Bloem (1995)
and Paul et al. ( 1999) was to replace the soil dispersion step of using a blender with
another method such as vortexing or sonication. Dispersion of the small volumes of soil
and water was not possible with a blender. Vortexing did a good job of mixing and
suspending the soil particles into solution. However, vortexing did not break up the
macro- and micro- aggregates and therefore did not adequately disperse the soil nor break
apart microbe clusters. Dispersion by sonication was tested using the cup-holder
attachment on the W-385 Sonicator (Heat Systems-Ultrasonics, Inc.). Sonicator settings
were the following: continuous cycle time, 90% duty cycle, and output level 3.
Sonication times of O, l, 10, 20, 30, 40, 505 were tested to evaluate the efficacy of
sonication on soil dispersion. Through conversations with colleagues it was determined
that continuous sonication for 2-3 minutes kills all microorganisms. The maximum
number of whole, intact bacteria, with the least number of soil particles, was the desired
result. By direct, manual counts on the microscope it was determined that the highest
bacterial cell to soil particle ratio (by counting stained bacteria and stained soil particles)

was by sonication for 105 (Figure 2.3). However, upon further observation it appeared

28

6.00 1
5.00
4.00

3.00 ,7

Biovolume (pm’)

2.00 L

1.00 *

   

 

 

 

 

 

 

 

0.00 ' ' .
1:100 1:150 1:200
Dilution rate

Figure 2.2. Mean cell biovolume measurements determined by CMEIAS-1 biovolume
formula at three dilution rates (1:100, 1:150, 1:200 soil:distilled water) for 6.3-9.5mm
native forest (NF) soil aggregates from a Hoytville clay loam (Fine, illitic, mesic Mollic
Epiqualfs); n=3 (aggregates and images).

29

:soil particles

Bacterial cells

 

 

 

o 1 10 20 3o 40 50
Sonication time (seconds)

Figure 2.3. Ratio of bacterial cells to soil particles by direct count for a Kalamazoo loam
(Fine-loamy, mixed, mesic Typic Hapludalf); n=3.

3O

that some of the cells were disrupted. Therefore, an optimized set of data were collected
after 5 and 103 of sonication to determine the highest mean cell biovolume measurements
(Figure 2.4A) and cell biovolume distribution (B) as calculated by CMEIAS. The largest
mean cell biovolume from 5 and 105 of sonication was observed at the SS sonication time.
The smaller mean cell biovolume seen at the 105 sonication time reinforces the idea that
cells undergoing sonication for 108 are disrupted in part. The cell biovolume distribution
is increased overall for the SS sonication time when compared with the 105 sonication.
The next step was conducting a sedimentation study. The purpose of this study
was to determine when the majority of soil particles had settled out of the top 5mL of
solution. According to Stoke’s law, t = 6.97n/g(ps-pl), 64 minutes was required for soil
particles 2pm in diameter to settle 1.55cm in distilled water at 25°C. It was assumed that
the density of bacterial cells was less than the density of soil particles. A range of
sedimentation times (15, 30, 45, 60, 64, 75, 90s) and one sonication time (55) were
evaluated (Figure 2.5). The best bacterial cell to soil ratio was observed at the 60 minute

sedimentation with 55 sonication.

Staining

Several biological stains could be used in this experiment, however, the best stain
requires a low background staining of soil particles and will adequately stain the bacteria
present. Three stains were evaluated for the above characteristics: 5-(4,6-dichlorotriazin-
2-yl)aminofluorescein (DTAF), acridine orange (A0), and phenolic aniline blue (PAB).
All are fluorescent stains. By direct microscopic, manual counts it was determined that

DTAF gave the highest bacterial cell to soil particle ratio (Figure 2.6). The modified

31

30 - - mm, A

25 , -. # I105 sonication , , ,, i, *
I: 55 sonication

20-7 —77’- . - -- --*

Mean cell biovolume (pm’)
5:
I

51—7 -—

 

 

 

60 “ ' >B
+55 sonication

4‘ 10$ sonication
50‘ )7, " *— '5'

 

 

 

 

Cell hinvnluma lum3\

 

 

 

14 16 18 20

8 10 12
Sorted cell number

Figure 2.4. Mean cell biovolume measurements (A) and cell biovolume distribution (B)
as determined by CMEIAS at 1:200 dilution and two sonication times (55, 105) for 6.3-
9.5mm native forest (NF) soil aggregates from a Hoytville clay loam (Fine, illitic, mesic
Mollic Epiqualfs); n=1 aggregate, n=3 images.

32

A—LN
0503C)

...s_s
Nth

Bacterial cellszsoil particles
8

I——— _

 

omtscnoo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

15

30

45

60

54

75

90

608

Sedimentation time (minutes)

Figure 2.5. Effects of sedimentation time (15-90 minutes) and 5 second sonication (60 S)
on the ratio of bacterial cells to soil particles for a Kalamazoo loam soil (Fine-loamy,
mixed, mesic Typic Hapludalf); n=3. The 60S is the bacteria to soil ratio after sample
has been exposed to 55 of sonication and sedimentation time is 60 minutes.

33

u-l—B—l—lN
meooo
II
I
I

Bacterial cells:soil particles
0

 

ON-hODCD

 

Figure 2.6. Ratio of bacterial cells to soil particles for a Kalamazoo loam (Fine-loamy,
mixed, mesic Typic Hapludalf). Aggregates are of size fraction 4-6.3mm across, stained
with 5-(4,6-dichlorotriazin-Z-yl)aminofluorescein (DTAF), acridine orange (A0), and
phenolic aniline blue (PAB). Dilution of 1:200, sedimentation time of 60 min, sonication
time of 55; n=3 aggregates and images.

34

bacterial extraction from soils and staining procedures are located in Appendix A (Figure

A.1).

Microscopy

Digital images were obtained from a fluorescence microscope (Leitz Orthoplan 2,
Leitz, Inc., Germany) with the 13 filter for FTIC (fluorescein) stain and a 100x oil
objective. A charge-coupled device (CCD) camera (Leica DC 500, Leica Microsystems
AG, Germany) was used for acquisition of high resolution images (up to 3090 x 3600
pixels). Adobe Photoshop (Adobe Systems, Inc.) image capturing software on a Dell
Pentium 4 PC, allowed for viewing the image before collection, and then saving the
image as a Tif file before further editing. The image capture settings include the
following: 8 bit color, 1030 x 1300 1 shot, always live, 20-30% brightness, and shutter

speeds of 0.1 to 0.3 seconds.

Image Analysis (pre-CMEIAS)

Before image analysis with CMEIAS can occur, the original microscope image
must undergo an editing process. This process has many steps (e. g., changing brightness
and contrast, making image grayscale, marking and outlining bacterial cells, embedding
10pm bar, increasing pixel resolution) which are outlined in Appendix A (Figure A2).
The amount of time spent on editing the images and preparing them for analysis with
CMEIAS can range from 15 to 30 minutes, depending on the number of bacteria present

in the image.

35

Image Analysis with CMEIAS

The Center for Microbial Ecology Image Analysis System (CMEIAS) is a
computer-assisted microscopy program that can analyze digital images of microbial
communities. CMEIAS v3.0 was used to analyze the images collected for this research.
CMEIAS v3.0 has multiple analyses, some of which include: microbial abundance (e. g.,
cell density, biomass C), morphological diversity (16 different cell shapes: 1] predefined
and 5 undefined), and spatial distribution (e.g., nearest neighbor distances, percent
microbial cover) (Lui et al., 2001; Dazzo et al., 2003a). Recent developments with the
object analysis measurement features include biovolume measurements. For the
morphotypes recognized by CMEIAS v3.0, eight out of 18 microbial biovolume formulas
have been identified by comparison to ground truth data obtained by volume
displacement as most accurate to measure this parameter in all the morphotypes classified
by CMEIAS. The most accurate biovolume formula (called the CMEIAS-l biovolume
formula) is adaptive to microbial cell shape (computed as object roundness) and also
contains cell length, width, and area parameters whose computations are all described by
Liu et al. (2001). Summary results of this formula are located in Table 2.1. CMEIAS
morphotype analysis has an accuracy of 94.1% overall with individual accuracies located
in Table 2.2.

CMEIAS has some requirements that need to be met for the best analyses and
compilations of data. One of these requirements is a minimum of 30 objects per image.
This requirement is to determine if adequate sampling has occurred. Another
requirement is that the minimum pixel density of each object is 30 pixels. This

requirement must be met for correct morphotype analysis. To be sure this requirement

36

Table 2.1. Summary results of CMEIAS-1 biovolume formula (Dazzo et al., 2003a;
personal communication). Morphotype populations that were tested include cocci, spiral,
U/curved rod, regular rod, unbranched filament, branched filament, rudimentary
branched rod, prosthecate, ellipsoid, and club.

 

 

 

 

 

 

 

Accuracy (%)
Overall morphotype pgrulations 92.4
Overall communities with representation of all morphotypes 95.8
Computed overall accuracy 94.1

 

37

 

 

 

 

 

 

 

 

 

 

 

 

 

Eons—E
2: =2 :2 o o o o o o o o o o euaoacm
8m
wanna—Sm
93 a: o mm.— o _ a o o o o _ o o bagueaé
2: e3 0 o 8N o o o o o o o o 28285
4.8 ea c v o .3” v c m o o c o 9.6
3a 5 o o o o :2 o a. o o o m anam
EoEflE
3a m: o o a o o a: a o o a c 85535
c? mm: o o a 2 _ a. S: o 2 c .N 8m aawé
8m
mg a: o m a o o _ _ o «2 _ o o 8&5:
2a a: o o o o o o N a E o o 8m EEO
a; 2:. _ o _ o o a o o a an o Sam
93 :2 o o o o _ o a o o o 22 888
ex; .8832 33 M a H m o a m a o m < 8552.5:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.20850 an we Booo
_ use 38 33mg mm 608 v 35836 388.805 98 A5888 $.de 38880 «88 22 mo 52 353.30 mSm—EU .oEwao
Sm .330 259—808 :08 5 £3 eoEmmm—o 38282: can 380:8 mo £383: 28 £33 woewfimaeomfimm—o mSmEU 8365
8828 ”232 awe—o SE 8365 36% .Coom in 3 :5 mo v 038. 88.: momma: 5E 238w .«o “88% omcfl a wean coEmme—o
N-m<HmEO 65 mo 33.0.2 68 Canoe/x doom in 5 Ed mSmEU wEm: 396399: 2.28% we .3838 :osaoEmmEO Nd 033.

38

was being met, one representative image from each soil type and treatment was analyzed
for object area (pixels) and object sizes were greater than 30 pixels each and all objects

found in each image were analyzed.

Image Analysis (post-CMEIAS)

After the collection of data from image analysis with CMEIAS, the data are
available for statistical analysis. Two CMEIAS macros were used, working in Microsoft
Excel, to prepare and compute the data obtained from the CMEIAS analyses (Dazzo et
al., 2003a). The first macro used was the data preparation macro. This macro is used for
both object analysis and classification data. It prepares the data for ecological statistical
analyses by compiling, concatenating, and graphically displaying the data. Examples of
the output data of this macro include the frequency distribution (Figure 2.7A), %
frequency distribution (B), and % cumulative frequency distribution (C). The second
macro used was the sampling statistics macro. This is the analysis where the 30 objects
per image minimum is required. This macro analyzes datasets each containing multiple
images, to determine whether the sampling sizes (cumulative number of objects) are
sufficient for computing the morphological diversity indices in the microbial
communities. Examples of the output data of this macro include the operational
morphological unit (OMU) richness by % abundance (Figure 2.8), and Shannon’s and
Simpson’s diversity indices (Towner, 1999) and 95% tolerance envelopes (Dazzo et al.,
2003)

The number of bacteria extracted per gram soil can be calculated with a minimum

amount of information from the microscopy and image analysis processes (Paul and

39

Frequency

 

5° I
0 ll _..
0.1 0.5 0.9 1.3 1.7 2.1 2.5 2.9 3.3 3.7 4.1 4.5 4.9 5.3 5.7 6.1 6.5 6.9

Biovolume bln wldth (nm’)

"I. Frequency

 

2% I;
0% iii . __g__ __ . 3

tr fi—r. .rY

0.1 0.5 0.9 1.3 1.7 2.1 2.5 2.9 3.3 3.7 4.1 4.5 4.9 5.3 5.7 6.1 6.5 6.9

Biovolume bln wldth (nm‘)

120% .

 

100%

80%

40% ,

% Cumulatlve Frequency

I
I
I
60% I
20% I

 

 

0% ‘. .
0.1 0.5 0.9 1.3 1.7 2.1 2.5 2.9 3.3 3.7 4.1 4.5 4.9 5.3 5.7 6.1 6.5 6.9

 

Blovolume bin wldth (nm‘)

Figure 2.7. Example of CMEIAS data preparation macro; A=frequency distribution,
B=% frequency distribution, and C=% cumulative frequency distribution. Cell
biovolume frequency distributions of the exteriors of two field replications of Wooster
native forest (NF) aggregates of size fractions 4-6.3mm; n=l6-l7 aggregates, 10 images
each replication.

40

 

 

% Abundance
#
O

30-

 

 

 

 

 

¢)" 00" Q5? Q95 Q96 39 (9” Q5? $05 99°

5 s" a s
Q99 6‘9.I €503“ rt{Q‘S‘zp $29?
0 c

Figure 2.8. Example of CMEIAS sampling statistics macro. Relative rank abundance
plot of morphological diversity in Wooster native forest (NF) aggregates of size fractions
4-6.3mm. The letter-number label corresponds to the operational morphological unit
(morphotype size) or morphological diversity classified by CMEIAS (Dazzo et al., 2003);
n=16 aggregates, 10 images.

41

Clark, 1996). The following formulas were used to calculate the number of bacteria per
gram soil for a 1:200 dilution and microscopy with 1000x total magnification:

Area ofimage: (0.135mm) x (0.106mm) = 0.01431mm2
Area of smear (well): 70:2 = 28.27mm2
Image area/smear (well) area: (0.01431)/(28.27) = 0.00051
Mass of soil/smear (well): (4/1000) x (0.495/99.495) = 0.00002g soil/smear
Mass of soil/image: (0.00002) x (0. 00051) = 1.01 x 108g soil/image
Bacteria/g soil: # per image/1.01 x [0'8

EcoStat, an ecological analysis program (Towner, 1999), analyzes community
diversity of samples from the data computed fiom the above CMEIAS macros. Several
diversity indices can be used when comparing two or more communities (e.g.,
proportional similarity (%), Jaccard coefficient, Sorenson coefficient, Morisita index,
Horn index, Sneath/Sokal distance, Bray/Curtis distance, and Chord distance). The
community similarity index selected for these studies was the proportional similarity (%)
index, based on the OMU richness and the % frequency biovolume distributions of
abundance. The proportional similarity (%) index is a simple measure which
incorporates information on species abundances based on the proportion or percentage of
total community abundance that each species comprises. It gives the percentage of

similarity between the communities under comparison (Figure 2.9).

Conclusion

Extraction of bacteria from concentric aggregate layers include efficient
sequential removal of exterior vs. interior aggregate soil layers by the SAE chamber
system, sonication of soil samples for 5 seconds, sedimentation of soil particles to

remove clay-sized minerals, diluting the bacterial solution to 1:200, and preparing the

42

 

(I)
U1
l

(D
O
.L

 

Proportional Similarity (%)

N
01
l

 

 

 

 

 

 

 

 

 

 

70
NFE vs. NFI CTE vs. CTI NFE vs. CTE NFI vs. CTI

Figure 2.9. Example of CMEIAS data analyzed in EcoStat; proportional similarity (%)
of interiors (I) and exteriors (E) of Wooster native forest (NF) and conventional-till (CT)
aggregates of size fractions 4-6.3mm across; n=l3-l7 aggregates, 10 images each
treatment.

43

slides and staining samples with DTAF for microscopic evaluation. Each of these
parameters was optimized for this study. Additional current innovations with image
analysis systems (CMEIAS) and data analyses (CMEIAS macros) allow for more data to
be acquired and analyzed. After determining that data, similarities or differences of
bacterial communities can be analyzed with ecological statistical programs, such as

EcoStat.

44

REFERENCES

Aditjandra, K., E.J. Park, and A.J.M. Smucker. 2002. Tensile Strength Domains within
Soil Aggregates from Contrasting Ecosystems. ASA Abstracts, Indianapolis, IN.

Bloem, J. 1995. Fluorescent staining of microbes for total direct counts. Molecular
Microbial Ecology Manual 4.1.8: 1-12.

Dazzo, F.B., J. Lui, A. Prabhu, C. Reddy, M. Wadekar, R. Peretz, R. Bollempalli, D.
Trione, E. Marshall, J. Zurdo, H. Hammoud, J. Wang, M. Li, D. McGarrell, A. Gore, J.
Maya-Flores, S. Gantner, and N. Hollingsworth. 2003a. CMEIAS: Integrative software
package to strengthen microscopy-based approaches for understanding microbial
ecology. Kellogg Biological Station-Long Term Ecological Research All Scientists
Meeting (KBS-LTER ASM), September 12, 2003.

ht_tp://lter.kbs.msu.edu/l\/Ieetings/2003 All Inv MeetingZAbstracts/dazzo.htm.

Dazzo, F., J. Liu, B. Yu, 0. Glagoleva, A. Jain, F.I. Liu, E. Marshall, D. Trione, and J.
Zurdo. 2003b. CMEIAS Version 1.27 Operator Manual: Computer-Assisted Microscopy
and Digital Image Analysis Software to Strengthen Microscopy-Based Approaches for
Understanding Microbial Ecology

Lui, J., EB. Dazzo, O.Glagoleva, B. Yu, and AK. Jain. 2001. CMEIAS: a computer-
aided system for the image analysis of bacterial morphotypes in microbial communities.
Microbial Ecology 41: 173-194.

Paul, E.A., D. Harris, M.J. Klug, and R.W. Ruess. 1999. The determination of microbial
biomass. p. 291-317. In Standard Soil Methods for Long-Term Ecological Research.
Oxford University Press, New York, NY.

Towner, H. 1999. EcoStat Ecological Analysis Program for Windows, Ver 1.02, Trinity
Software, Compton, NH, Internet site: http://www.trinitysoftware.com/lifesci/index.html.

45

CHAPTER 3
SOIL BACTERIAL COMMUNITY DIVERSITY IN CONCENTRIC

AGGREGATE LAYERS AS INFLUEN CED BY SOIL TYPE AND
MANAGEMENT PRACTICE

INTRODUCTION

Microorganisms promote soil aggregation processes by providing glue and
cementing agents, promote plant growth through symbiotic relationships (e.g., N-fixing
bacteria, mycorrhizal fungi, and plant growth-promoting rhizobacteria), control root
diseases (biocontrol agents), fix N (free-living N-fixing bacteria), assimilate plant
nutrients bound in organic matter (OM) or soil minerals (P from rock phosphate), and
degrade pollutants. Microbial diversity within aggregates has been reported to be as large
as the microbial diversity identified across extensive geographical areas (Grundmann and
Normand, 2000). This localized diversity provides a huge resource of multifunctional
microbial communities. Soil aggregation processes modify microhabitats located in soil
aggregates which are thought to fiirther modify the microbial communities located in
these regions. Bacterial extraction along with fluorescence microscopy and computer-
assisted image analysis allow for quantification of changes in bacterial populations and
communities associated with the above microhabitats in external and internal regions of
soil aggregates.

The objective of this study was to quantify changes in bacterial populations and
community diversities within concentric layers of soil aggregates from three tillage and

plant management systems.

46

MATERIALS AND METHODS

Site Description and Soils

Samples of soil were collected from surface soils of a native forest (NF) site,
long-term conventional-tillage (CT) site, native grassland (NG) succession site, and a
conventional tillage planted to continuous alfalfa (CT-CA) site. The NF site and long-
term CT site are located on a Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic
Fragiudalfs) at the Ohio Agricultural Research and Development Center (OARDC) near
Wooster, OH. The long-terrn CT site has been under CT management for the past 100
years. The NG succession site is located on a Kalamazoo loam (Fine-loamy, mixed,
mesic Typic Hapludalf) at the Michigan State University Kellogg Biological Station
(KBS), a NSF Long Term Ecological Research Station in Hickory Comers, MI. The CT-
CA site is located on a Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) at the Michigan Agricultural Experiment Station in Chatham, MI.
Samples were taken for the NF and CT sites in the spring of 2003 from the surface 0-
5cm. Samples were taken from the NG and CT-CA sites in the summer of 2002 from the
surface 0-10cm. Additional samples were taken from the NG site in September of 2003
for comparison analyses of moist and air-dry aggregates. All samples were stored field
moist in plastic containers prior to arrival in the lab. Soils were stored at 4°C until
processing occurred. Soils were air-dried and manually sieved prior to further analyses.
Sieved aggregate size fractions include >9.5, 6.3-9.5, 4-6.3, 2-4, 1-2, 0.5-1, <O.5mm in
diameter. The aggregate sizes from all field sites used in this study were from the 4-

6.3mm fraction.

47

Bacterial Analyses (Preparation, Microscopy, and Image Analysis)

Bacterial extraction techniques for bulk soil samples have been developed to
allow for soil dispersion and preparation of bacterial samples on slides for computer-
assisted microscopy and image analysis (Bloem, 1995; Paul et al., 1999). Modifications
of the above techniques were needed before small temporal and spatial changes in
bacterial populations and community diversities, modified by intra-aggregate
microhabitats at external and internal regions within soil macro-aggregates were
quantified (Appendix A, Figure A.1). Soil dispersion and bacterial extraction procedures
were modified by first decreasing the amount of soil used from 10g to 0.5g. Soil
dispersion by blending was replaced with continuous sonication (W-385 Sonicator, Heat
Systems-Ultrasonics, Inc.) for 55. Sedimentation was optimized by Stoke’s law to
determine that after 42 minutes, 2pm soil particles had settled out of the top 5mL of
solution used to prepare microscope slides. It was assumed the density of bacteria was
less than the density of soil particles at the same size. The biological fluorescent stain 5-
(4,6-dichlorotriazin-2-yl)aminofluorescein (DTAF) was determined to give the least
background staining by having the highest bacterial cell to soil particle ratio when
compared with acridine orange (A0) and phenolic aniline blue, also fluorescent stains.
Digital images were obtained from a fluorescence microscope (Leitz Orthoplan 2, Leitz
Inc., Germany) with the 13 filter for FTIC (fluorescein) stain and a 100x oil objective. A
charge-coupled device (CCD) camera (Leica DC 500, Leica Microsystems AG,
Germany) was used for acquisition of high resolution images (up to 3090 x 3600 pixels).
Adobe Photoshop (version 7, Adobe Systems, Inc.) image capturing software on a Dell

Pentium 4 PC allowed for viewing the image before collection, and then saving the image

48

as a Tif file before further editing. The image capture settings include the following: 8
bit color, 1030 x 1300, single shot, always live, at a brightness of 20-30%, and a shutter
speed of 0.1-0.3 seconds. Editing of images in PhotoshOp was required before image
analysis to obtain a binary image with the background being lighter than the foreground
objects (Appendix A, Figure A.2). Image analysis with the Center for Microbial Ecology
Image Analysis System (CMEIAS), a computer-assisted microscopy program, analyses
digital images of microbial communities for individual cell size and shape. Biovolume
measurements were calculated using the CMEIAS-l biovolume formula, which is
adaptive to microbial cell shape (computed as object roundness) and also contains cell
length, width, and area parameters whose computations are all described by Liu et al.
(2001). CMEIAS-1 biovolume formula has an overall accuracy of 94.1% with all
CMEIAS morphotypes represented (Dazzo et al., 2003). The data preparation macro
developed for CMEIAS, working in Microsoft Excel, prepares the data collected for
ecological statistical analyses by compiling, concatenating, and graphically displaying the
data (Dazzo et al., 2003). Output data of this macro includes frequency distribution, %
frequency distribution, and cumulative frequency distribution graphs of the data selected.
For this study, cell biovolume distributions were identified. Extracted bacteria per gram
soil were calculated from microscope images and cell counts from CMEIAS analysis

(Paul and Clark, 1996).

Aggregate Analyses

Soil aggregate analyses parameters for this study include wet sieving water

stability, polar tensile strength (PTS), intra-aggregate porosity, and dry combustible total

49

C and N concentrations. Aggregate stability, mean weight diameter (MWD), was
determined by the wet sieving technique with nebulization of dry aggregates to prevent
the breakage of most aggregates during the wetting process (Kemper and Rosenau, 1986).
Aggregate strength was determined by the polar tensile strength technique. The polar
tensile strength of a soil aggregate is the force per unit area that is required to break the
aggregate (Dexter and Kroesbergen, 1985; Hillel, 1998). Porosity was calculated from
determining the bulk densities of soil aggregates by the Saran method (Blake and Hartge,
1986). Total C and N analyses were determined by combustion at 800°C of small

quantities of soil (30mg soil) in a Carlo-Erba CHN analyzer, model NA1500.

Statistical Analysis

Bacterial community diversity was analyzed using EcoStat (Towner, 1999), an
ecological analysis program that is commonly used to perform many statistical tests
including the evaluation of species diversity and community similarity analyses of many
ecosystems. The input data from CMEIAS included the % frequency distributions of cell

biovolumes.

RESULTS AND DISCUSSION
Bacterial Populations
Greater populations of bacteria (23%, 24%, 18%, and 7%) were observed in the
exterior regions of NF, CT, NG, and CT-CA aggregates when compared with the
corresponding interiors (Figure 3.1). Exterior regions of NF aggregates contained the

highest numbers of bacteria (8.35E+10), while the CT-CA interiors had the least

50

9.E+10 —

8.E+10 ~-

7.E+10 <

 

6.E+10 -»w * "
5.E+10 4% 43‘?”

4.E+10 -_

Bacteria/g soil

3.E+10 we ., 7 _ .

2.E+10 .7. .

1.E+10 I

0.E+00

 

 

 

 

 

 

 

 

 

 

 

can:

 

NF

 

silt loam

Exterior I Interior Exterior I Interior

Exterior i

Interior Exterior I Interior
NG CT-CA

loam sandy loam

Figure 3.1. Declining populations of bacteria per gram of exterior and interior soil layers
as soil textures become more coarse. Soil samples from a Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) native forest (NF) were sampled from the surface 0-
5cm, a Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native grassland
(NG) succession and a Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) were sampled
from the surface 0-10cm, from soil aggregate size fractions 4-6.3mm across; n=7-2l
aggregates, n=20-30 images per layer (dependent on number of field replications),
n=446-1687 cumulative cells sampled. Bars indicate standard deviations.

51

(4.66E+09). No significant differences in the quantities of bacteria were observed
between the exterior and interior regions of CT and NF aggregates. One hundred years of
tillage has reduced bacterial populations to 25% below that of the native forest. It was
expected (and observed) that the largest bacterial populations would be associated with
the soils containing the most clay and highest C concentration (Paul and Clark, 1996;
Brady and Weil, 2002). Less destructive management practices lead to more stable
aggregates, which in turn are able to support larger bacterial populations (Young and
Ritz, 2000; Lupwayi et al., 2001). Extracted bacteria per gram soil were calculated from
microscope images and cell counts from CMEIAS analysis. A 35% increase in bacteria
per gram soil was observed in the moist vs. air dry NG aggregates (Figure 3.2). This

indicates that air drying significantly reduces bacterial populations extracted from soil

aggregates.

Cell Biovolume Measurements

Microbial activity and aggregate stability are linearly related (Edgerton et al,
1995). Microbial biovolume evaluations give insight into the strength and stability of soil
aggregates. Mean cell biovolume measurements were calculated by CMEIAS using
CMEIAS-l biovolume formula developed by Dazzo et a1. (2003). This formula is
adaptive to cell shape using roundness shape measurements and also contains cell length,
width, and area parameters as described by Liu et al. (2001). CMEIAS-1 biovolume
formula has an overall accuracy of 94.1% with all CMEIAS morphotypes represented
(Dazzo et al., 2003). In contrasts to populations, no differences were observed between

the mean cell biovolumes of moist and air dry soil aggregates from NG treatments

52

 

1.E+10-7 7 7 777 77 77 77 77 77 77 I 77 77 7;
1.E+10— - , « _ - - ,

7 I 77 1“ ‘ a
I ¥

6.5+09-77 7 -7 7-7 7 , 777

 

Bacteria/g scil

4.190917 7‘. ... 7_7 77 7 7777

2.54.09- ~77 7 77 7 77 7 7' ,' .777

 

 

 

 

 

 

o.E+oo . * 4
Air Dry Moist

Figure 3.2. Extracted bacteria are significantly lower in cell density from air dry vs.
moist whole aggregates from a Kalamazoo loam (Fine-loamy, mixed, mesic Typic
Hapludalf) native grassland (NG) succession from soil aggregate size fractions 4-6.3mm
across, sampled from the surface 0-10cm; n=3 aggregates, n=20 images, n=446-1687
cumulative cells sampled. Bars indicate standard deviations.

53

(Figure 3.3). These results indicate that sampling air dry aggregates or moist aggregates
at field water contents had minimal affects on the bacterial community when considering
their cell biovolumes. Mean cell biovolumes were largest in the NG aggregates, next
largest in the CT-CA aggregates, and smallest in the NF and CT aggregates (Figure 3.4).
Mean cell biovolumes were larger in the exterior regions than the interior regions of NG
aggregates, while no differences were observed between exterior and interior regions of
CT-CA aggregates and NF aggregates. Mean cell biovolumes were higher in the exterior
regions of CT aggregates than their interior regions. It was expected that the larger mean
cell biovolumes would be associated with the least disruptive management system, (e. g.,

NF), finer soil type, (e. g., clay loam), and exterior regions of soil aggregates.

Cell Biovolume Distributions

The distribution of cell biovolume frequencies present bacterial community
comparisons based on cell size compositions among bacterial communities. Cell
biovolume frequencies are calculated by the CMEIAS data preparation macro, that
compiles, concatenates, and graphically presents the CMEIAS data. Sorted cell
biovolume distributions allow for comparisons of numbers of cells analyzed and the
biovolume of those cells. Increased numbers of cells were Observed in the moist NG
aggregates, while the biovolumes of those cells were smaller than the cell biovolumes
observed in the air-dry NG aggregates (Figure 3.5). Cell numbers varied between the
exterior and interior regions of NG aggregates from 3 field replications (Figure 3.6) and
CT-CA aggregates from 2 field replications (Figure 3.7), though no significant

differences were seen in the corresponding cell biovolumes. Cell biovolumes were

54

.3
N
o

1.00 -

 

 

0.80 ~

_0

A

O
I

0.20 -

Mean cell biovolume (urns)
o
C)
o

   

 

 

 

 

 

9
o
o

 

Air Dry Moist

Figure 3.3. Mean cell biovolume measurements determined by CMEIAS-1 biovolume
formula for air dry and moist whole aggregates from Kalamazoo loam (Fine-loamy,
mixed, mesic Typic Hapludalf) native grassland (NG) succession from soil aggregate size
fractions 4-6.3mm across, sampled from the surface 0—10cm,; n=3 aggregates, n=20
images, n=750-1014 cumulative cells sampled. Bars indicate standard deviations.

55

N
O
O
|
—4
1
|
l

 

 

 

 

 

Biovolume (um3)
A 0
0'1
0

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.00 — 7 7
T
T T T

0.50 .1. 7 ‘- 1 . L- V- 7 -
0.00

ExteriorI Interior ExteriorI Interior Exterior Interior Exterior Interior

NF CT NG CT-CA
Silt loam Loam Sandy loam

 

 

 

Figure 3.4. Mean cell biovolume measurements determined by CMEIAS-1 biovolume
formula for exterior (E) and interior (1) layers of Wooster silt loam (Fine-loamy, mixed,
mesic Oxyaquic Fragiudalfs) native forest (NF) aggregates, taken from the surface 0-
5cm, Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native grassland
(NG) succession aggregates and Ternary sandy loam (Sandy loam, mixed, frigid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) aggregates from
soil aggregate size fractions 4-6.3mm across, sampled from the surface 0-10cm; n=7-21
aggregates, n=20—30 images per layer (dependent on number of field replications),
n=446-1687 cumulative cells sampled. Bars indicate standard deviations.

56

 

 

 

+Air Dry
*~— Moist
7 7 7 7 7 7 _ 7 7 7 7 7
5 7 7 1.5
«.E ‘f
3: a
o 4 » 7 -— 7 I
3 i
g 1"
-- it
:3» -— 7.
0 it
2
1
200 400 600 800 1000 1200

 

Sorted Cell Number

Figure 3.5. Increased numbers of cells were observed in moist aggregates, while a larger

biovolume was observed in the air dry aggregates. The distribution of cell biovolumes
were calculated by the CMEIAS data preparation macro for moist and air dry whole
aggregates from Kalamazoo loam (F inc-loamy, mixed, mesic Typic Hapludalf) native

grassland (NG) succession from soil aggregate size fractions 4-6.3mm across, sampled
from the surface 0-10cm; n=3-4 aggregates, n=10 images per treatment, n=750-1014

cumulative cells sampled.

57

 

 

 

 

8 .
+112
s 11
-~e~~25
,2.
7 «0735777777-7- ——-—7—l77
77.773' : O :
I . *
I
l t
57* 7 7 7 7 7 7 7 7 7 i7 7 . 777 7'77
w r l
r I , .
U .
. O
. 7 7 7 77777 7:477 77
i‘ 7
F i 3..
E I I
3 I
o " f i
E , 7 7 7 7 7 , 7-7777 7 7 7
2 41 > I
2 I I
.2
.0
a 7 7 7 77 7 .7 7
o 3
2 v — 7 W‘s-{Ii} «if, I — *" — ‘—
*1
I} 0.17“
1 7 #1, M
or . ~ . . .
0 100 200 300 400 500 600 700 800

Sorted Cell Number

Figure 3.6. Increased numbers of cells were observed in the exteriors of aggregates from
3 field replications when compared to their interiors. Biovolumes of these cells showed
no significant differences between regions of the aggregate or field replication. The
distribution of cell biovolumes were calculated by the CMEIAS data preparation macro
for exterior (E) and interior (1) layers of aggregates from 3 field replications (1,2,3) of
Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native grassland (NG)
succession from soil aggregate size fractions 4-6.3mm across, sampled from the surface
0-10cm; n=16-21 aggregates, n=10 images per replication and layer, n=446-732
cumulative cells sampled.

58

+2E

 

Cell biovolume (urns)

 

100 200 300 400 500 600
Sorted Cell Number

Figure 3.7. Biovolumes of these cells showed no significant differences between regions
of the aggregate or field replication. The distribution of cell biovolumes were calculated
by the CMEIAS data preparation macro for exterior (E) and interior (1) layers of
aggregates from 2 field replications (2,8) of Ternary sandy loam (Sandy loam, mixed,
frigid Alfic Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA) from
soil aggregate size fractions 4-6.3mm across, sampled from the surface 0—10cm; n=7-10
aggregates, n=10 images per replication and layer, n=457-520 cumulative cells sampled.

59

similar for NF aggregates and the exterior regions of CT aggregates, with the interior
regions of CT aggregates having smaller cell biovolumes (Figure 3.8).

Community diversity statistics include various indices for comparing and
contrasting bacterial communities. For this study proportional similarity (%) was
determined, based on the OMU richness and the % frequency biovolume distributions of
abundance. The proportional similarity (%) index is a simple measure which
incorporates information on species abundances based on the proportion or percentage of
total community abundance that each species comprises.
comparison. Proportional similarity (%) was calculated using the cell biovolume
frequencies calculated by the CMEIAS data preparation macro. Community similarities
around 85% or more were not considered to be significantly different in morphological
diversity. The differences between proportional similarities among compared
communities cannot be labeled as increased or decreased cell biovolumes. For this
interpretation and summary, the bacterial cell biovolume frequencies give greater insight
as bacteria] populations and their potential metabolic activities are compared, as
metabolically active cells are larger.

Bacterial cell biovolume frequency distributions, using percent abundance, give
reasonable estimates of bacterial community comparisons of the most abundant size class
or mode (personal communication with Dr. Dazzo). Cell biovolumes of bacterial
communities were 13.5% greater in moist vs. air dry NG aggregates (Figure 3.9).
Therefore it appears that community diversities determined by cell biovolumes can
possibly be used to identify accumulative or historic changes in soil water content

changes within soil aggregates.

60

9 NE

+NF|

A CTE 4 o
7 CTl 7 7 7 7 7 I, 7 7
6 7 7 ; 7 7

 

 

Cell biovolume lumal

 

 

 

 

0 500 1000 1500 2000
Sorted Cell Number

Figure 3.8. Increased numbers of cells were observed in NF aggregates compared to CT
aggregates, with the exteriors of the aggregates having increased numbers than their
interiors. Cell biovolumes were decreased in the interior region of CT aggregates. The
distribution of cell biovolumes were calculated by the CMEIAS data preparation macro
for exterior (E) and interior (1) layers of aggregates from 2 field replications (2,8) of
Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) native forest (NF)
and conventional-till (CT) from soil aggregate size fractions 4-6.3mm across, sampled
from the surface O-SCm; n=13-l7 aggregates, n=10 images per management system and
layer, n=1091-1687 cumulative cells sampled.

61

.3383 m=oo 0333an 326mm“: .mowEE on”: .moamwocwwn mu: ufimoom 3 was 8an 78:888.“
3% 92920 2: 3 M53.0 Co £23 :3 2838 a was: Baa—=28 225 30:27:66 55308.7 082065 Eco—é 8me
05 Eat BEES .mmodom EEméé .8323 ON? Bmwoewwm :8 69a 55883 676 98?me 033: Eng—gem 29C. 388 6828
$687.05”: Ems Swans—av. Soc moumwodwwm 2023 be he .m> E08 05 E Bingo 203 882305 :8 703855 6.2m DSwE

$83 25.305 .30

mg n We o 0m m men 7» Wm m
[.gfioeaaeesaélézi I _, I . .

 

aouepunqv %

 

 

 

62

Bacterial cell biovolume distributions in exterior layers were 15%, 18%, and 16%
smaller than interior regions of NG aggregates from three field replications, respectively
(Figure 3.10). Bacterial cell biovolume distributions were 21%, 16%, and 21% different
in external layers between field replications, 1-2, 1-3, and 2-3, respectively, while cell
biovolume diversities for interior layers between field replications were contrasted by
19%, 21%, and 20%. Cell biovolumes located in the exterior regions followed the trend
(largest to smallest) of rep3 > repl > rep2, while cell biovolumes located in the interior
regions followed the trend of rep3 > rep2> repl. Field replication 2 was least similar to
replications 1 and 3. This indicates that the cell size distribution within bacterial
communities could differ considerably across the landscape as indicated by these
contrasts among soil aggregates from three field replications, and supports the idea that
microbial diversity within aggregates are as great as the microbial diversity identified
across extensive geographical areas (Grundmann and Normand, 2000).

Bacterial cell biovolume distributions in exterior layers were the same (e.g., 13%)
as the interior regions of CT-CA aggregates from two field replications, 2 and 8 (Figure
3.11). Bacterial cell biovolume distributions contrasted by 12% within exterior layers
and 16% within interior layers between these two field replications, demonstrating small
spatial variabilities in the bacterial community between field replications.

Bacterial cell biovolume distributions were 23% larger in exterior layers (Figure
3.12A) and not significantly larger (9%) in the interior regions (Figure 3.12B) of CT and
NF aggregates, indicating greater differences in exterior layer communities between
management systems. Bacterial cell biovolume distributions in exterior layers were 20%

smaller than interior regions of NF aggregates (Figure 3.12C) and 11% larger in exteriors

63

 

 

.3383 £00
gum—=83 N963”: cot?— ?8 5:82am: .59 momma: Sn: .moaawocwwa 3-3”: camcom 3 can 8qu 538303 53 mSmEU
2.: 3 macs—.0 mo £23 :5 E838 a was: Bum—=28 22$ mcousnEmG zocoscot 082955 Boo—Io oomtsm 2: 80¢ BEES
.388 EEmdé c283 0N6 Sawflwwa :8 89¢ 55883 670 cam—gum 26a: Eng—am: 29C. 38:. 6228 $887055 88.
ooumfiflmvw Eat .32 A38 A 38 803 Am: 5 €2.82: 223 Name A 32 A much 93:25 9 “mamas can: 05 wage—.8 33030
803 3&5 mcoumozme Eva m Bob 83%:wa mo 73 av Ectouxo 05 E 865329 082329 :8 @8385 .ofim ocswi

Ann—3 2.5.035 =mo
v Wm m 9N N we _. md

  

mg n md o m.m m

o

 

 

 

 

 

III. .I. . I .. .I I . t I _, III... II o
I Cxl ,, am I 8..)uill‘i'. I I I
«I . (IA/«1.4%. {I u 15 IF
I l I I - I I I I I IN %
II -| i I l I In v
I I I I I I I v M
I I | I in W
I I II so w
a II I I g. o
a I. I I l I I I I w a
m :IT Im
$83 222055 =oo
m4. v m6 m md oc
I II I. I 1.... .lI awzyIukP/nbfo .I I F
I |. .l. . II |. 1|. . I. II! I «If IIIN o/o
l lI I i I I II II It. . II In V
I I - II - Iv m
I I II II I I I b w.
w
I I I I I lI I I l 1 Im a
v, II I I I : I I I III I IN. a
I mm .
I: wNIQ. I i Ii III IiI iI i lI i Ii IIIII . w
< I erIIOI. Fa

 

64

1 O I

.3883 £00 03008800 ONmInmvn: .893— 080 80:00:02 800 mowag Sn:
.moummohwwm 3-an ufimoom >0 0:0 08008 80080008 800 mSmZU 08 >0 m83.0 .«0 823 80 88800 a was: 08200.8 803
80:08me 38000.58 083085 .8020 00023 08 80b 00—088 £00800 8800-0 80:08.“ 080 oumwohwwa =00 80¢ $098
0:020 00008800 8 00820 5308880800 800058803 051x 0&5 .0828 .802 30ng 800. >083 b083, 80¢ ”5%.?va
80:00:02 22% N m0 Amv 808026 .m> a 80508 08 8 002030 0.83 020000008.“ 082085 :00 0000205 ._ mm 8&5

$83 080.055 :00
ms N m6 0 m.m m m4. v md m mN N we _. md o

 

 

III?» Fax/"yr. I... .1131... 0:...flmxfimfiwwmk IoIIIIo, II I. . ‘0
I I I I I .I I I .N
o/o
I I I I I I II I .v V
0 m.
I I I I I I I . I u
w
I I I I I I II I I I In w
a . II I I I I I 20
mat. I
m E

 

#83 08:_0>0_m :00
ms N m6 m mh m m4. v m6

 

 

 

[III\I dogsiégodeggol I I. . I... ax .
%

II I I I I I I I I I V
q
H

II I I I I I I U
p
w

IIIIIIII-I IIIIIIIIIII I 3
a

II_NI. I I I I

< MNIQI I

 

65

 

 

 

 

 

 

20
ii A
,I777I 7777,, +NFE
816 I - NFI
: I .
g I
QIHIII_ IIIII~IIIIIIIII
\° I.
. .(I K I I I I I
0 0" -b“fi—‘WWW o—n—Q—do——————~-t-— -—-
0.5 1 1.5 2 2.5 3 3.5 4 4.53 5 5.5 6 6.5 7 7.5
Cell Biovolume (um )
20 '
16 — , I 0 7 f I +CTE B
o ‘ CTI
a 1 . _ _ _ _ I.
I
g 8 . 7 7 7
°\° 4 l ’ if. F N 7 _ -7 II_ _
ok-NMUWM . M0 0 0 I 00-—

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5
Cell Biovolume (um )

 

C
7 7 +NFE

8 ’ CTE
C
N 7 7
'0
I:
3
a __-_ 77 7 7 __
<
°\‘

-"WNM9‘—W~'Ni+° v o I I 0"——-

 

2.5 3 3.5 4 4.5 53 5.5 6 6.5 7 7.5
Cell Biovolume (um )

+NFI D
- CTI

% Abundance

 

 

 

0.5 1 1.5 2 2.5 3 3.5 4 4.535 5.5 6 6.5 7 7.5
Cell Biovolume (pm )

Figure 3.12. Increased cell biovolume frequencies were observed in the exteriors (E) of
CT vs. NF aggregates (C), while slight increases were seen in the interiors (D).
Decreased cell biovolume frequencies were observed in the exteriors vs. interiors (I) of
NF aggregates and increased for CT aggregates from Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) native forest (NF) (A) and conventional—till (CT) (B)
from soil aggregate size fractions 4-6.3mm across, sampled from the surface 0-5cm.
Biovolume frequency distributions were calculated using a constant bin width of 0.1um3
by the CMEIAS data preparation macro and by EcoStat; n=13-17 aggregates, n=10
images per management system and layer, 1091-1687 cumulative cells sampled.

66

vs. interiors of CT aggregates (Figure 3.12D). This indicates that the cell size
distributions reflect modifications in the structure and/or physiological state of these
bacterial communities in situ. These modifications among bacterial communities within
soil aggregates are thought to contribute to the stability of the aggregate. These reported
spatial gradients of changes in the distribution of biovolume diversities of bacteria within
soil aggregates contributes to previously reported contributions by bacteria and their
associated metabolic byproducts (Tisdall, 1994; Paul and Clark, 1996). Native forest
aggregates were observed to have similar cell biovolume distributions between field
replications ( 1,2) of exterior and interior regions (Figure 3.13A and B), while increased
variability of field replications was observed for exterior and interior regions of CT
aggregates (Figure 3.13C and D).

A summary of the bacterial data for these soils is located in Appendix A (Table
A.1). These data include object counts per soil and management practice, extracted
bacteria per gram of soil, and biovolume measurements of mean, median, mode,

skewness, minimum, and maximum.

Aggregate Stability

Mean weight diameter evaluations, by wet sieving, were much higher for soil
aggregates from NF treatments than CT treatments of the Wooster silt loam (Figure
3.14). Therefore it was concluded that 100 years of tillage significantly reduced the
water stability of soil aggregates. Additional wet sieving comparisons of soil aggregates
from native grasslands (NG) on Kalamazoo loam soils demonstrated similar MWD with

NF treatments on Wooster silt loams soils. Conventional tillage (CT) treatments of a

67

20‘

°/o Abundance
0° :3 a;

.5

% Abundance
4" 0° N a: 8

O

20‘

% Abundance
N

O

N
A

% Abundance
N

oeoo

Figure 3.13.

TA 7L

 

 

..L
O)

a:

h

AN
010

 

 

1.5 2

 

i WOO-6W! o o

2.5 3

3.5 4

4.5 5

Cell Biovolume (um3)

2.53

3.54

\WW‘McQMe—MOMw-ooo 9% ~2 =- r

4.55

Cell Biovolume (um3 )

 

2.5 3

3.5 4

. M" ‘."‘3m—' MMoo-u—O-WIW'
4.5 5

Cell Biovolume (uma)

 

2

 

3.5 4

I‘
M'W Morrow-$9 o

2.5 3 4.5 5

Cell Biovolume (m3)

’ Q 1

5.5 6

5.56

5.5 6

6.5

Similar cell biovolume distributions were observed between field
replications (1,2) of exterior (E) and interior (1) regions of NF aggregates (A,B), while
increased variability of field replications was observed for exterior and interior regions of
CT aggregates (C,D) from Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic
Fragiudalfs) native forest (NF) (A, B) and conventional- till (CT) (C, D) from soil
aggregate size fractions 4- 6. 3mm across, sampled from the surface 0- 5cm. Biovolume
frequency distributions were calculated using a constant bin width of 0.1um3 by the
CMEIAS data preparation macro and by EcoStat; n=13-17 aggregates, n=10 Images per

management system and layer, n=109l-l687 cumulative cell sampled.

68

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.6”

0.5—H7 I 7
a 0.4’ 7 .
‘3'" 1
E
.9 0.3"
D
E,
.5 0.2’ ’ ’
3
g 7 ' . . 7
o 0.1
E .

0 7 ‘ ‘ T '. . ' ‘ . '. -
NF CT ‘ NG l CT-CA

Silt loam Loam i Sandy loam

Figure 3.14. Mean weight diameter (MWD) by wet sieving for Wooster silt loam (F ine-
loamy, mixed, mesic Oxyaquic Fragiudalfs) native forest (NF) aggregates, sampled from
the surface O-Scm, Kalamazoo loam (Fine-loamy, mixed, mesic Typic Hapludalf) native
grassland (NG) succession aggregates and Ternary sandy loam (Sandy loam, mixed,
fi'igid Alfic Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA)
aggregates, sampled from the surface O-lOcm. Samples are from soil aggregate size
fractions 0.4-0.63cm across; n=3. Bars indicate standard errors.

69

Wooster silt loam and a Ternary sandy loam (CT-CA) had the lowest MWDs. These data
support the concept that disruptive management decreases aggregate stability (Kandeler
and Murer, 1993; Beare et al., 1994; Hillel, 1998; Brady and Wei], 2002). No differences
among measured polar tensile strengths were observed between whole NF and CT

aggregates (data not reported).

Aggregate Porosity

Aggregate porosities, determined by the Saran method, for tillage and crop
management systems were: NF > CT > CT-CA (Figure 3.15). Native forest aggregates
were 19% more porous than CT aggregates from the Wooster silt loam series.
Aggregates from CT Wooster silt loam were 26% more porous than CT-CA aggregates
from sandy loam soils. Soil aggregates from NF Wooster silt loam were 51% more
porous than CT-CA aggregates from sandy loam soils, indicating that soil type and
management practices together have a large affect on the porosity of aggregates. It has
been reported that as the soil porosity increases, soil bacterial populations also increase

(Young and Ritz, 2000).

C and N Concentrations

A strong positive correlation (R2=0.996, p<0.002) was observed between C and
bacterial populations of CT-CA and NG aggregates (Figure 3.16A) sampled three years
earlier. It is expected that these data represent C and N values in the 2003 aggregates.
As expected C and N concentrations and bacterial populations were positively correlated.

Nitrogen concentrations and bacterial populations of the same aggregates, however, were

70

 

O

1:.
, l
Vie—4

 

-,1——1

 

 

 

 

 

 

 

 

 

o ' l
NF CT CT-CA

l Silt loam 1 Sandy loam

Figure 3.15. Porosity by Saran method for Wooster silt loam (Fine-loamy, mixed, mesic
Oxyaquic Fragiudalfs) native forest (NF) and conventional-till (CT) aggregates, taken
from the surface O-Scm, and Ternary sandy loam (Sandy loam, mixed, fn'gid Alfic
Haplorthodes) conventional-till planted to continuous alfalfa (CT-CA). Soil aggregate
size fractions 4-6.3mm across were sampled from the surface 0-10cm; n=10. Bars
indicate standard errors.

71

8.E+10 I -I

 

7 A
7.E+10IIIIII—— IIIIIIIIIII—I II III
6.E+10II I I I I II I I I _ III III III I I
= NGE
O
"’ '7
$5,510.77 IIIIIII -IIII
g 71 CT_CAI CT-CAE N6!
+ 1,7, __ 777 I 7 I 7 7 7 7 7 7 II
00 4‘ 0 y=2E+10x-3E+10
R2 = 0.9961. p<0.002
3.E+10 _.__._7 II I _ I I _ — _ I 7 7_ I I
2.E+10IIIIIIIIIIIIIII
1.E+10II II I I I I II II I I I I I
0.E+00 -L. . , I I . , . . , ,
0 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1
C(%)
7 B
= ~3E+10x + 6E+10
CT-CA' H... R2 = 0.6053. p<0.22
5,015+10I I I J“ I: IIIIII I I I
= CT-CAE . , I
3 N61 ‘1
m .
R NGE
C
0
‘5
m
m
2.5E+10- 7_ -7 — —
0.0E+OO .I . . . . . . . . .
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
N(%)

 

 

 

 

 

 

 

 

 

Figure 3.16. Extracted bacteria per gram soil and C (A) and N (B) concentrations for
exterior (E) and interior (1) layers of Kalamazoo loam (Fine—loamy, mixed, mesic Typic
Hapludalf) native grassland (NG) succession aggregates and Ternary sandy loam (Sandy
loam, mixed, frigid Alfic Haplorthodes) conventional-till planted to continuous alfalfa
(CT-CA). Soil aggregate size fractions 4-6.3mm across were sampled from the surface
O-lOcm; n=7-21 aggregates, n=10 images per management system and layer, n=446-732

cumulative cells sampled, n=2 replications for C and N data.

deviations.

Bars indicate standard

72

not as strongly correlated nor significant (R2=0.605, p<0.22) (Figure 3.16B). Relatively
high, yet non-significant correlations were observed between C (R2=O.586, p<0.23) and
N (R2=0.568, p<0.25) concentrations and bacterial populations of CT and NF aggregates

(Figure 3.17 A and B).

Bacterial Morphotype Analysis

CMEIAS v3.0 morphotype analyses were conducted on moist and air dry whole
NG aggregates. The proportional similarity (%) of the data concludes that the bacterial
communities were 97.3% similar in morphotype diversity. These data indicate that the
change in environment due to air drying the aggregates, led to minimal changes in the
structure of the bacterial community revealed by this characteristic. The proportional
similarity of morphological diversity for exterior and interior layers of NG and CT-CA
aggregates were 98% and 98%, indicating little to no change in the morphological
diversity of microbial communities throughout the aggregates. The proportional
similarity of morphological diversity for exteriors of NG and CT-CA aggregates was
88% and for interiors was 86%, indicating some contrast among microbial community
diversities between different management practices and soil types. The proportional
similarity of the morphological diversity for the exterior and interior layers of NF
aggregates was 97% similar, indicating only small changes in the characteristic of

bacterial communities throughout aggregate interiors.

73

 

1.E+115 -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I A
'_.§__.4 NFE ‘
= 8.E+10I I II I I I I I I I- I
3 l—f—"l NFI
D
g5.E+10III-- IIIIIIIII
§ y=8E+09x+5E+10
3 R9 = 0.5864. p<0.23 ;
3.5.1.1017, I II IIII I , I I
0aE+OO T T 1 1 T T l
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50
C(%)
1.E+11 — v -~
_ 8.E+10 I II—
8
a. 6.E+1O III II
E
0 “777 . __ __
‘6 4'E+10 y=1E+11x+5E+10
3 R2 = 0.5682, p<0.25 :
2.E+10 I— I I III f
O.E+OO . . . . I
0.00 0.05 0.10 0.15 0.20 0.25
N(°/o)

Figure 3.17. Extracted bacteria per gram soil and C (A) and N (B) concentrations for
exterior (E) and interior (1) layers of Wooster silt loam (Fine-loamy, mixed, mesic
Oxyaquic Fragiudalfs) native forest (NF) and conventional-till (CT). Soil aggregate size
fractions 4-6.3mm across were sampled from the surface 0-5cm; n=l3-l7 aggregates,
n=10 images per management system and layer, n=1091-l687 cumulative cells sampled,
n=2 replications for C and N data. C and N data from samples in 2000. Bars indicate

standard deviations.

74

SUMMARY AND CONCLUSIONS

Soil bacterial communities and populations undergo changes throughout the
season and with changes in soil water contents. Morphologically the community changes
are small, 3%, while cell biovolume changes are larger, 14%, correlating well with
slightly larger abundance size classes in the moist aggregates. Bacterial density was
higher in moist aggregates than air dry aggregates, and no differences were observed in
the mean cell biovolumes.

Highest bacterial densities were seen in the NF aggregates and the smallest
bacterial populations were observed in the CT-CA aggregates. The largest mean cell
biovolumes were observed in the air dry NG aggregates and the smallest in the air dry NF
aggregates. The highest diversity in the distribution of cell biovolumes between exterior
and interior regions, was observed in the NF aggregates and the lowest in the CT-CA
aggregates. The highest diversity in the distribution of cell biovolumes was observed in
the exteriors of NF vs. CT aggregates and the smallest diversity among bacterial
biovolumes in the interiors of the same aggregates. These differences indicate that the
cell size distributions reflect contrasting structures and/or physiological conditions of
these bacterial communities. Small differences in morphological diversity were seen in
NF, NG and CT-CA aggregates between the aggregate exterior and interior. Larger
differences were observed in exteriors and interiors of aggregates from differing soil
types and management practices.

Carbon and N concentrations appear to contribute to bacterial populations, as

hypothesized. However, since these correlations were relatively small, other factors seem

75

 

to exert greater influences on the population sizes in the communities located in soil
aggregates. Data presented here demonstrate direct relationships between aggregate
stability, porosity, and bacterial populations.

Increased soil bacterial populations were positively related to C and N
concentrations which could be correlated to increased aggregate stability (MWD). As
greater energy sources become available and are readily accessible to soil bacterial
populations, those populations appear to increase in biovolume and density. These
increased sizes and populations lead to greater bacterial by-products that, in turn, lead to

greater formations of more stable aggregates.

76

REFERENCES

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.

Blake, GR. and K.H. Hartge. 1986. Bulk Density. In A. Klute (ed.). Methods of Soil
Analysis: Part l-Physical and Mineralogical Methods. p.363-375. American Society of
Agronomy, Inc. and Soil Science Society of America, Inc., Madison, WI.

Bloem, J. 1995. Fluorescent staining of microbes for total direct counts. Molecular
Microbial Ecology Manual 4.1.8: 1-12.

Brady, NC. and RR. Weil. 2002. The Nature and Properties of Soils. 13th ed. Prentice
Hall, Upper Saddle River, NJ.

Dazzo, F.B., J. Lui, A. Prabhu, C. Reddy, M. Wadekar, R. Peretz, R. Bollempalli, D.
Trione, B. Marshall, J. Zurdo, H. Hammoud, J. Wang, M. Li, D. McGarrell, A. Gore, J.
Maya-Flores, S. Gantner, and N. Hollingsworth. 2003a. CMEIAS: Integrative software
package to strengthen microscopy-based approaches for understanding microbial
ecology. Kellogg Biological Station-Long Term Ecological Research All Scientists
Meeting (KBS-LTER ASM), September 12, 2003.
http://lter.kbs.msu.edu/Meetings/2003_All_Inv_Meeting/Abstracts/dazzo.htm.

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

Edgerton, D.L., J .A. Harris, P. Birch, and P.Bullock. 1995. Linear relationship between
aggregate stability and microbial biomass in three restored soils. Soil Bio. Biochem.
27:1499-1501.

Grundmann, G.L and P. Normand. 2000. Microscale diversity of the genus Nitrobacter in
soil on the basis of analysis of genes encoding rRNA. Appl. Environ. Microbiol.
66:4543-4546.

Hillel, D. 1998. Soil Structure and Aggregation. p. 101-125. In Environmental Soil
Physics. Academic Press, San Diego, CA.

Kandeler, E. and E. Murer. 1993. Aggregate stability and soil microbial processes in a
soil with different cultivation. Geoderrna 56:503-513.

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, Physical and Mineralogical
Methods, 2"d edition. Soil Sci. Soc. Am., Madison, WI.

77

Lui, J., PB. Dazzo, O.Glagoleva, B. Yu, and AK. Jain. 2001. CMEIAS: A computer
aided system for the image analysis of bacterial morphotypes in microbial communities.
Microbial Ecology 41 : 173- 194.

Lupwayi, N.Z., M.A. Arshad, W.A. Rice, and G.W. Clayton. 2001. Bacterial diversity in
water-stable aggregates of soils under conventional and .zero tillage management.
Applied Soil Ecology 16:251-261.

Paul, EA and FE. Clark. 1996. Methods for studying soil microorganisms. p. 35-68 In
Soil Microbiology and Biochemistry, 2nd edition. Academic Press, San Diego, CA.

Paul, E.A., D. Harris, M.J. Klug, and R.W. Ruess. 1999. The determination of microbial
biomass. p. 291-317. In Standard Soil Methods for Long-Term Ecological Research.
Oxford University Press, New York, NY.

Towner, H. 1999. EcoStat Ecological Analysis Program for Windows, Ver 1.02, Trinity
Sofiware, Compton, NH, Internet site: http://www.tn'nitvsoftware.com/lifesci/index.html.

Young, I.M. and K. Ritz. 2000. Tillage, habitat space and function of soil microbes. Soil
Tillage Res. 53:201-213.

78

CHAPTER 4

WETTING AND DRYING MODIFICATIONS OF SOIL MACRO-AGGREGATE
STABILITIES: BIOPHYSICAL MECHANISMS

INTRODUCTION

The soil functions as a reservoir for water, nutrients, gases, soil organic matter
(SOM), and microbial populations. The formation of stable soil aggregates increases
terrestrial C sequestration within the soil profile. Greater aggregate stability, especially
of surface soils, leads to increased water infiltration and aeration porosity, reduced water
and wind erosion, and decreased crust formation. Aggregation processes depend on
several factors, such as: the interaction of carbon (C), nitrogen (N), and SOM, microbial
and mesofaunal activities, wetting/drying (W/D) frequencies, and management practices

(Oades, 1993; Tisdall, 1994; Edgerton et al., 1995).

Organic matter is the primary stabilizing ingredient that is essential for the
formation and stabilization of soil aggregates. SOM is the primary energy substrate
needed for most bacterial activities (Tisdall and Oades, 1982; Brady and Weil, 2002).
Until recently, the measurement of SOM has been limited to bulk soils. However, total C
and N contents have been reported to be higher in the external layers of soil aggregates
than their internal layers (Horn et al., 1990; Smucker et al., 1996). Most of the additional
C located in the external layers is thought to be part of the more labile C pool associated

with the microbial biomass, or dissolved organic C.

Changes in the soil environment that affect microbial communities are also
thought to affect soil structure (e.g., aggregation and aeration) (Kandeler and Murer,

1993; Edgerton et al., 1995). A positive linear correlation between stable macro-

79

aggregates and microbial biomass C was demonstrated by Edgerton et al. (1995).
Bacterial biomass C increased in the presence of macro-aggregates and decreased where
no macro-aggregates were present (Denef et al., 2001). Denef et al. (2001) also
demonstrated a reduction in bacterial activity associated with W/D cycles of soils with no
macro-aggregates present. Microbial diversity within aggregates seems to be as large as
the microbial diversity identified across extensive geographical areas (Grundmann and

Normand, 2000).

Wetting/drying cycles (W/D) influence aggregate formation and stability. Six
W/D cycles have been shown to increase aggregate stability (Utomo and Dexter, 1982).
Sissoko (1997) demonstrated that aggregate stability, as determined by the mean weight
diameters (MWD) from wet sieving data, was not affected with three W/D cycles,
however, aggregate stabilization was enhanced when C and N compounds were added

during the wetting phase of 9 W/D cycles.

Conventional-tillage (CT) management practices incorporate crop residues and
fertilizers into the plow layer using a moldboard plow to lift, twist, and invert the soil.
Surface soils of CT management practices are subjected to extreme changes in water
content and temperature. Extreme changes frequently break or destroy weaker
aggregates exposing additional quantities of SOM to microbial mineralization. Further
weathering leads to continued SOM losses and decreased particulate organic matter
(POM) contents (Beare et al., 1994), reducing energy resources for soil microbial
activities (Monreal and Kodama, 1997). Reduced or no-tillage (NT) practices use
minimal or no plowing, leaving plant residues on the soil surface, which shelter

aggregates from wind, rain, and temperature extremes, preserving aggregate structure.

80

These soils generate higher quantities of stable soil macro-aggregates which have
accumulated nearly doubled the intra—aggregate quantities of total C stored within the
exterior regions of macro-aggregates sampled from loam soils converted from CT to NT
management systems for periods of at least 10 years (Smucker et al., 1996).

The objective of this study was to determine how the frequency of W/D cycles
affects bacterial populations and communities located in the external and internal regions

of soil aggregates sampled from a native forest ecosystem.

MATERIALS AND METHODS

Site Description and Soils

Samples of soil were collected from a native forest (NF) site located on a Wooster
silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) at the Ohio Agricultural
Research and Development Center (OARDC) near Wooster, OH. Samples were taken in
the spring of 2003 from the surface 0-50m. All field samples were transported to the
laboratory in rigid plastic containers to prevent packaging pressures on natural
aggregates. Soils were stored at 4°C until processing. Soil samples were manually
broken into their most stable aggregate size fractions during the air drying and manually
sieved to separate into multiple size fractions without further abrasion of natural
aggregates. Sieved aggregate size fractions included >9.5, 6.3-9.5, 4-6.3, 2-4, 1-2, 0.5-1,
<O.5mm in diameter. Individual size fractions were stored in air-dried conditions within
rigid plastic containers until further analyses. The aggregate size used in this study was

the 4-6.3mm fraction.

81

 

Wetting and Drying Cycles

Ninety uniform soil aggregates, 4.0 to 6.3mm across, were subjected to 0, 3, and 6
wetting and drying (W/D) cycles (16 hours wetting/16 hours drying) before
quantification for polar tensile strength, soil C and N contents, and soil bacterial

populations and biovolume distributions.

Bacterial Analyses (Soil preparation, Microscopy, and Image analysis)

Bacterial extraction techniques for bulk soil samples have been developed to
allow for soil dispersion and preparation of bacterial samples on slides for computer-
assisted microscopy and image analysis (Bloem, 1995; Paul et al., 1999). Modifications
of the above techniques were needed before small temporal and spatial changes in
bacterial populations and community diversities, modified by intra-aggregate
microhabitats at external and internal regions within soil macro-aggregates were
quantified (Appendix A, Figure A.1). Soil dispersion and bacterial extraction procedures
were modified by first decreasing the amount of soil used from 10g to 0.5g. Soil
dispersion by blending was replaced with continuous sonication (W-385 Sonicator, Heat
Systems-Ultrasonics, Inc.) for 58. Sedimentation was optimized by Stoke’s law to
determine that after 42 minutes, 2pm soil particles had settled out of the top 5mL of
solution used to prepare microscope slides. It was assumed that the bacterial cell density
was less than the density of soil particles at the same size. The biological fluorescent
stain 5-(4,6-dichlorotriazin-2-yl)aminofluorescein (DTAF) was determined to give the
least background staining by having the highest bacterial cell to soil particle ratio when

compared with acridine orange (A0) and phenolic aniline blue, also fluorescent stains.

82

 

Digital images were obtained from a fluorescence microscope (Leitz Orthoplan 2, Leitz
Inc., Germany) with the 13 filter for FTIC (fluorescein) stain and a 100x oil objective. A
charge-coupled device (CCD) camera (Leica DC 500, Leica Microsystems AG,
Germany) was used for acquisition of high resolution images (up to 3090 x 3600 pixels).
Adobe Photoshop (version 7, Adobe Systems, Inc.) image capturing software on a Dell
Pentium 4 PC allowed for viewing the image before collection, and then saving the image
as a Tif file before further editing. The image capture settings include the following: 8
bit color, 1030 x 1300, single shot, always live, at a brightness of 20-30%, and a shutter
speed of 0.1-0.3 seconds. Editing of images in Photoshop was required before image
analysis to obtain a binary image with the background being lighter than the foreground
objects (Appendix A, Figure A.2). Image analysis with the Center for Microbial Ecology
Image Analysis System (CMEIAS), a computer-assisted microscopy program, analyses
digital images of microbial communities for individual cell size and shape. Biovolume
measurements were calculated using the CMEIAS-l biovolume formula, which is
adaptive to microbial cell shape (computed as object roundness) and also contains cell
length, width, and area parameters whose computations are all described by Liu et al.
(2001). CMEIAS-l biovolume formula has an overall accuracy of 94.1% with all
CMEIAS morphotypes represented (Dazzo et al., 2003). The data preparation macro
developed for CMEIAS, working in Microsoft Excel, prepares the data collected for
ecological statistical analyses by compiling, concatenating, and graphically displaying the
data (Dazzo et al., 2003). Output data of this macro includes frequency distribution, %
frequency distribution, and cumulative frequency distribution graphs of the data selected.

For this study, cell biovolume distributions were identified. Extracted bacteria per gram

83

soil were calculated from microscope images and cell counts from CMEIAS analysis

(Paul and Clark, 1996).

Aggregate Analyses

Strengths of soil aggregates were measured by the polar tensile strength (PTS)
technique and reported as the force per unit area required to disrupt air dry soil
aggregates. Average aggregate diameters are determined by measuring effective
diameters at three locations around each aggregate, by an extemal-dimension caliber. A
constant force is applied to one polar end of an aggregate placed on a stationary platform.
The applied forces are graphed with significantly large inflection points identified.
Graphic disruptions are viewed and the largest is selected as the force required to break
the major plane of weakness within each aggregate (Dexter and Kroesbergen, 1985;
Hillel, 1998). Another parameter for comparing aggregate stability are the total C and N
contents, determined by the combustion at 800°C of small quantities of soil (30mg) in a

Carlo-Erba CHN analyzer, model NA1500.

Statistical Analysis

Bacterial community diversity was analyzed using EcoStat (Towner, 1999), an
ecological analysis program that is commonly used to perform many statistical tests
including the evaluation of species diversity and community similarity analyses of many
ecosystems. The input data from CMEIAS included the % frequency distributions of cell

biovolumes.

84

RESULTS AND DISCUSSION
Aggregate Strength
Increased polar tensile strengths (PTS) were observed when W/D cycles increased
from 0 to 3 for NF aggregates from Wooster silt loam soils (Figure 4.1). PTS decreased
as frequencies of W/D were increased from 3 to 6 cycles. Although neither of the
increases or decreases in PTS were significant, changes in aggregate strength were

observed.

Bacterial Populations

A 23% increase in bacterial populations was observed in the exterior layers of the
more stable aggregates exposed to 3 W/D cycles, with a 29% decrease in bacterial
populations as W/D cycles increased from 3 to 6. (Figure 4.2). A 17% increase in
bacterial populations was observed in the interior layers of aggregates exposed to 3 W/D
cycles, with an additional 32% increase with 3 more W/D cycles. These data indicate
that bacterial populations in the exteriors of aggregates increase with 3 W/D and then
decrease to lower than initial populations with 6 W/D cycles. In the interiors of
aggregates, however, bacterial populations continue to increase through 6 W/D cycles.
This may be due to increased pore networks with each wetting phase, which increases the
movement of C into interior regions of aggregates. The C is then more accessible to
these p0pulations located in the interior of aggregates. This potential movement of C
from the exterior to interior regions of soil aggregates with 3 to 6 W/D cycles may
explain the decrease in PTS with 6 W/D cycles. The microbial populations in the

exteriors of aggregates exposed to 6 W/D cycles significantly decrease from 3 W/D

85

 

 

 

 

Tensile strength (kPa)

 

 

 

 

 

 

 

 

 

0 7 7#777 .' , ,7, Vi l 777 7 7 7 i. 77 1
l 0 l 3 1

Number of wet/dry cycles

Figure 4.1. Increased polar tensile strengths (PTS) with 3 W/D cycles and decreased PTS
with 6 W/D cycles for native forest (NF) aggregates from Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) of aggregate size fraction 4—6.3mm across, sampled
from the surface 0-5cm, exposed to 0, 3, and 6 wetting-drying (W/D) cycles; n=10
aggregates. Bars indicate standard errors.

86

25.4.10 _ . . , . , 1

2.E+10 III II IIII II I I I II II-II I—I-

1.E+10 -

1.E+10 v

 

1.E+10 «

8.E+09 ~

Bacteria/g soil

6.E+09 <

4.E+09 <

2.E+09 -

0.E+00 -

Exterior Interior Exterior Interior Exterior

 

 

Figure 4.2. Increased populations of bacteria were observed in the exteriors and interior
of aggregates exposed to 3 W/D cycles. The increase in populations continued in the
interior regions through 6 W/D cycles, while a decrease was observed in the exteriors of
aggregates exposed to 6 W/D cycles. Extracted bacteria per gram soil for exterior and
interior layers of native forest (NF) aggregates from Wooster silt loam (Fine-loamy,
mixed, mesic Oxyaquic Fragiudalfs) of aggregate size fraction 4-6.3mm across, sampled
from the surface O-Scm, exposed to 0, 3, and 6 wetting-drying (W/D) cycles; n=13-16
aggregates, n=10 images per WD cycle and layer, n=717-1 141 cumulative cells sampled.
Bars indicate standard errors.

87

cycles. This decrease in population indicates there is less microbial by-products being
produced to help stabilize the aggregate, leading to decreased aggregate strength. These
data clearly support the accentuation of microbial community activities by repeated
hydrations/dehydrations of intra-aggregate regions and support previous reports by Denef

et al. (2001).

Cell Biovolume Measurements

Microbial biovolume data can give insight into the strength and stability of soil
aggregates. Biovolume measurements were calculated using the CMEIAS-l biovolume
formula, which is adaptive to microbial cell shape (computed as object roundness) and
also contains cell length, width, and area parameters whose computations are all
described by Liu et al. (2001). CMEIAS-1 biovolume formula has an overall accuracy of
94.1% with all CMEIAS morphotypes represented (Dazzo et al., 2003). Increased mean
cell biovolumes were observed in the interior regions of NF aggregates from Wooster silt
loam soils, exposed to 3 W/D cycles than other interior regions exposed to O and 6 W/D
cycles or exterior region of aggregates exposed to O, 3, or 6 W/D cycles (Figure 4.3). A
decrease, though not significant, was observed in the exterior regions of aggregates
exposed to 6 W/D cycles. No differences were observed between exterior and interior
regions of NF aggregates exposed to O and 6 W/D cycles. Cell size and numbers of
bacteria both increased in interior regions of aggregates exposed to 3 W/D cycles.
However, cell size decreased in the interior regions of aggregates exposed to 6 WfD
cycles while populations increased. These data support previously reported information

that microbial activity and aggregate stability are linearly

88

 

 

.0
a)

 

P
on
.

Mean cell biovolume (um’)
_o o
w «A

P
N

.0
—l

     

.0
o

 

Exterior i Interior

6W/D

 

Exterior Interior Exterior Interior I

OW/D 3WID

Figure 4.3. Increased cell biovolumes were observed in the interior regions of aggregates
exposed to 3 W/D cycles. The mean cell biovolume was calculated by CMEIAS-l
biovolume formula for exterior and interior layers of native forest (NF) aggregates from
Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) of aggregate size 4-
6.3mm across, taken from the surface O-SCm, exposed to 0, 3, and 6 wetting-drying
(W/D) cycles; n=l3—l6 aggregates, n=10 images per WD cycle and layer, n=717-1141
cumulative cells sampled. Error bars indicate standard deviations.

89

related (Edgerton et al, 1995). Additionally, our data demonstrate that aggregate
stabilities increase as bacteria] diversities increase in exterior regions of aggregates.
Furthermore, aggregate stability declined with increasing W/D cycles transferred

bacterial diversities toward interior regions of macro-aggregates.

Cell Biovolume Distributions

The distribution of cell biovolume frequencies present bacterial community
comparisons based on cell size compositions among bacterial communities. Cell
biovolume frequencies are calculated by the CMEIAS data preparation macro, that
compiles, concatenates, and graphically presents the CMEIAS data.

Community diversity statistics include various indices for comparing and
contrasting bacterial communities. For this study proportional similarity (%) was
determined, which is a simple measure which incorporates information on species
abundances based on the percentage of total community abundance that each species
comprises. Proportional similarity (%) was calculated using the cell biovolume
frequencies calculated by the CMEIAS data preparation macro. Community similarities
around 85% or more were not considered to be significantly different in morphological
diversity. The differences between proportional similarities among compared
communities cannot be labeled as increased or decreased cell biovolumes. For this
interpretation and summary, the bacterial cell biovolume frequencies give greater insight
as bacterial populations and their potential metabolic activities are compared, as

metabolically active cells are larger. Bacterial cell biovolume frequency distributions,

90

 

using percent abundance, give reasonable estimates of bacterial community comparisons
of the most abundant size class or mode (personal communication with Dr. Dazzo).

Bacterial cell biovolume distributions in exterior layers were not significantly
different (9%, 15%, and 8%) than interior regions of NF aggregates exposed to 0, 3, and
6 W/D cycles. The cell biovolumes in the exterior regions of aggregates exposed to 3
W/D cycles were actually smaller than the interior cell biovolumes. The proportional
similarity (%) indicates that bacterial communities become more diverse following 3
short term W/D cycles and less diverse with 6 W/D cycles. A significant increase in the
% difference of % abundance in cell biovolume distribution was observed with 3 W/D
cycles for biovolume size 0.3pm3. Bacterial cell biovolume distributions were 8%
smaller, 12% smaller, and 8% smaller in exterior layers for 0 vs. 3, 0 vs. 6, and 3 vs. 6
W/D cycles, and 12% larger, 8% smaller, and 16% smaller in interior regions. This
indicates that W/D cycles influence microbial communities in both the exterior and
interior regions of aggregates. The diversity of exterior regions of aggregates, increased
with both 3 and 6 W/D cycles; however, the diversity of interior regions increased with 3
W/D cycles and decreased with 6 W/D cycles. A significant increase in the % difference
of % abundance in cell biovolume distributions was observed with 3 W/D cycles at
0.3um3 cell biovolume (Figure 4.4). Increases were also observed in the exteriors of O
and 6 W/D cycles, and in the interiors of 3 and 6 W/D cycles. These data show that W/D
cycles do affect cell size composition of bacterial communities in soil aggregates.

A summary of the bacterial data for these soils is located in Appendix A (Table

A. 1). These data include object counts per soil and management practice, extracted

91

 

 

i
i
i
i

 

 

a.
l
i

 

 

NNwPAém
. or.
i

 

—L
o;

 

 

_L

 

 

°/o difference of cell biovolume 0.3um3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I
:3
i

 

 

 

(OW/D ‘ 3WID l 6W/D 3vs.0l 6vs.3l 6vs.0‘ 0vs.3i 6vs.3 i 6vs.0
I .

Exterior vs. Interior Exterior Interior

Figure 4.4. Changes among the % abundances of cell biovolume distributions were
observed with 3 W/D cycles at 0.3um3 cell biovolume. Increases were also observed in
the exteriors of O and 6 W/D cycles, and in the interiors of 3 and 6 W/D cycles. The
biovolume frequency distributions were calculated using a constant bin width of 0.1um3
by the CMEIAS data preparation macro and by EcoStat for native forest (NF) aggregates
from Wooster silt loam (Fine-loamy, mixed, mesic Oxyaquic Fragiudalfs) of aggregate
size fraction 4-6.3mm across, sampled from the surface O-Scm, exposed to O, 3, and 6

wetting-drying (WD) cycles; n=l3-16 aggregates, n=10 images per WD cycle and layer,
n=717-1141 cumulative cells sampled.

92

 

bacteria per gram of soil, and biovolume measurements of mean, median, mode,

skewness, minimum, and maximum.

C and N Concentrations
Carbon and nitrogen concentrations from the soil aggregates exposed to 0, 3, and

6 W/D cycles are currently being analyzed.

SUMMARY AND CONCLUSION

Soil bacterial communities and populations undergo changes throughout the
season. Wetting-drying cycles influence bacterial communities in the exterior and
interior regions of soil aggregates. These influences include desiccation and osmotic
stresses. The cell biovolumes of bacterial communities were observed to become more
diverse with 3 W/D cycles and less diverse with 6 W/D cycles. The diversity of cell
biovolume distributions in exterior regions of aggregates, increased with both 3 and 6
W/D cycles; however, the diversity in interior regions increased with 3 W/D cycles and
decreased with 6 W/D cycles. Significant changes to the bacterial community
composition seem to occur with 3 W/D cycles. Mean cell biovolumes were larger in the
interior regions of aggregates exposed to 3 W/D cycles than any other exterior region of
aggregates exposed to O, 3, and 6 W/D cycles, or interior region of aggregates exposed to
O and 6 W/D cycles.

Bacterial density increased in the exterior regions of aggregates with 3 W/D

cycles and decreased with 3 additional W/D cycles. Bacterial density increased in the

93

 

interior regions of aggregates through 3 and 6 W/D cycles. Differences in the PTS of soil
aggregates exposed to 0, 3, and 6 W/D cycles were observed though not significant. PTS
increased with 3 W/D cycles and decreased with 6 W/D cycles. This difference may be
explained by changes in soil bacterial populations as aggregates were exposed to
increasing W/D cycles. The bacterial populations in soil aggregates increased in the
interior regions of aggregates through 6 W/D cycles. However, the populations increased
only through 3 W/D cycles for the exterior regions and then decreased with 6 W/D
cycles. This may be due to increased pore networks with each wetting phase, which
increases the movement of C into interior regions of aggregates. The C is then more
accessible to these populations located in the interior of aggregates. This potential
movement of C from the exterior to interior regions of soil aggregates with 3 to 6 W/D
cycles may explain the decrease in PTS with 6 W/D cycles. The microbial populations in
the exteriors of aggregates exposed to 6 W/D cycles were significantly lowered as W/D
cycles increased from 3 to 6. This decrease in population indicates there are fewer
microbial by-products produced with increasing W/D cycles, leading to decreased

strength among soil aggregates.

94

 

REFERENCES

Aditjandra, K., E.J. Park, and A.J.M. Smucker. 2002. Tensile Strength Domains within
Soil Aggregates from Contrasting Ecosystems. ASA Abstracts, Indianapolis, IN.

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.

Bloem, J. 1995. Fluorescent staining of microbes for total direct counts. Molecular
Microbial Ecology Manual 4.1.8: 1-12.

Brady, NC. and RR. Weil. 2002. The Nature and Properties of Soils. 13‘h ed. Prentice
Hall, Upper Saddle River, NJ.

Dazzo, F.B., J. Lui, A. Prabhu, C. Reddy, M. Wadekar, R. Peretz, R. Bollempalli, D.
Trione, E. Marshall, J. Zurdo, H. Hammoud, J. Wang, M. Li, D. McGarrell, A. Gore, J.
Maya-Flores, S. Gantner, and N. Hollingsworth. 2003. CMEIAS: Integrative software
package to strengthen microscopy-based approaches for understanding microbial
ecology. Kellogg Biological Station-Long Term Ecological Research All Scientists
Meeting (KBS-LTER ASM), September 12, 2003.
http://lter£bs.msu.edu/Meetings/2003_All_Inv_Meeting/Abstracts/dazzo.htm.

Denef, K., J. Six, H. Bossuyt, S.D. Frey, E.T. Elliot, R. Merckx, and K. Paustian. 2001.
Influence of dry-wet cycles on the interrelationship between aggregate, particulate
organic matter, and microbial community dynamics. Soil Biol. Biochem. 33:1599-1611.

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

Edgerton, D.L., J.A. Harris, P. Birch, and P.Bullock. 1995. Linear relationship between
aggregate stability and microbial biomass in three restored soils. Soil Bio. Biochem.
27:1499-1501.

Grundmann, G.L and P. Normand. 2000. Microscale diversity of the genus Nitrobacter in
soil on the basis of analysis of genes encoding rRNA. Appl. Environ. Microbiol.

66:4543-4546.

Hillel, D. 1998. Soil Structure and Aggregation. p. 101-125. In Environmental Soil
Physics. Academic Press, San Diego, CA.

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

95

 

 

Kandeler, E. and E. Murer. 1993. Aggregate stability and soil microbial processes in a
soil with different cultivation. Geoderma 56:503-513.

Lui, J., F.B. Dazzo, O.Glagoleva, B. Yu, and AK. Jain. 2001. CMEIAS: A computer
aided system for the image analysis of bacterial morphotypes in microbial communities.
Microbial Ecology 41: 173-194.

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

Paul, EA and F.B. Clark. 1996. Methods for studying soil microorganisms. p. 35-68 In
Soil Microbiology and Biochemistry, 2nd edition. Academic Press, San Diego, CA.

Paul, E.A., D. Harris, M.J. Klug, and R.W. Ruess. 1999. The determination of microbial
biomass. p. 291-317. In Standard Soil Methods for Long-Term Ecological Research.
Oxford University Press, New York, NY.

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

Sissoko, F. 1997. Enhancement of soil aggregation by the combined influences of soil
wetting and drying and root-microbial associations. MS. Thesis. Michigan State
University, East Lansing.

Smucker, A.J.M., D. Santos, and Y. Kavdir. 1996. Concentric layering of carbon,
nitrogen, and clay minerals within soil aggregates from tilled and non-tilled
agroecosystems and natural grasslands. Third Eastern Canada Soil Structure Workshop
Proceedings. p. 129-140.

Tisdall, J.M. 1994. Possible role of microorganisms in aggregation of soils. Plant Soil
159:115-121.

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

Towner, H. 1999. EcoStat Ecological Analysis Program for Windows, Ver 1.02, Trinity
Software, Compton, NH, Internet site: http://www.trinitysoftware.com/lifesci/index.html.

Utomo, W.H. and AR. Dexter. 1982. Changes in soil aggregate water stability induced
by wetting and drying cycles in non-saturated soil. J. Soil Sci. 33:623-637.

96

 

CHAPTER 5

CONCLUSIONS

Soil bacterial communities and populations undergo changes throughout the
season and with increased cycling soil water contents. This study has shown that
increased bacterial populations and proportionally larger cell biovolumes exist in moist
soil aggregates than in air dry aggregates. Therefore, it seems that water stresses
(desiccation and osmotic) in the soil adversely affect microbial communities and
populations. Wetting and drying cycles influence bacterial communities in the exterior
and interior regions of soil aggregates. Bacterial populations increased in both the
exterior and interior regions of aggregates exposed to 3 W/D cycles, however,
populations decreased in the exterior regions of aggregates exposed to 6 W/D cycles
while continuing to increase in the interiors. Polar tensile strength of aggregates
increased with 3 W/D cycles and decreased with 6 W/D cycles. This change in aggregate
strength may be explained by the changes in soil bacterial populations with W/D cycles.
The increase in bacterial populations in the interiors of aggregates exposed to 6 W/D
cycles may be due to increased pore networks with each wetting phase, which increases
the movement of C into interior regions of aggregates. The C is then more accessible to
these populations located in the interior of aggregates, hence the increase in bacterial
populations in the interior regions and decrease in exterior regions. This potential
movement of C from the exterior to interior regions of soil aggregates with 3 to 6 W/D
cycles may also explain the decrease in polar tensile strength observed with 6 W/D

cycles. The decrease of bacterial populations in the exterior regions of aggregates

97

 

indicates there is less microbial by-products being produced to help stabilize the
aggregate, leading to decreased aggregate strength.

Soil type and management practices influence soil bacterial communities and
populations. The largest bacterial populations were associated with the soils containing
the most clay and highest C concentration (silt loam > loam > sandy loam), with the
exterior regions of aggregates having increased populations over the interior regions.
Increased bacterial populations were observed with soil aggregates under the least
destructive management practice. These aggregates also had higher stabilities, leading
them to be able to support the larger populations of bacteria observed. The native
grassland loam aggregates had larger biovolumes than the conventional-till planted to
continuous alfalfa sandy loam aggregates, which in turn had larger biovolumes than the
native forest and conventional-till silt loam aggregates. This indicates that cell
biovolumes were not significantly influenced by soil type (clay content) or management
practice. The highest diversity in the distribution of cell biovolumes was observed in the
exteriors of native forest vs. conventional-till aggregates and the smallest diversity in the
interiors of the same aggregates. These differences indicate that the cell size distributions
reflect differences in the structure and/or physiological conditions of these bacterial
communities.

Carbon and N concentrations appear to contribute to bacterial populations,
however, since some correlations are relatively low, other factors seem to be contributing
to the population estimates of the communities located in soil aggregates. Results from
this study demonstrate direct relationships between aggregate stability, porosity, and

bacterial populations. Increased soil bacterial populations were positively related to C

98

 

and N concentrations and increased aggregate stability as determined by MWD. As
greater energy sources are made available and accessible to soil bacterial populations,
those populations increase in biovolume and density. This increase in size and
population leads to greater bacterial by-products that in turn lead to increased aggregate

formation and stabilization.

99

 

 

APPENDIX

Figure A. 1. Modified bacterial extraction and staining procedure ................................. 101
Figure A.2. Editing procedure for CMEIAS analysis of DTAF stained images ............ 102
Table A. 1. Overall summary of data for soils compared ................................................ 103

100

 

Figure A. 1. Modified bacterial extraction and staining procedure.

Direct Microscopy Counts for Bacteria using Soil Smears and DTAF

(Bloem, 1995; Paul at al., 1999; modified by Dopp, 2003)

Reagents

l.

2.
3.

Buffer: Phosphate Buffered Saline — 0.05M NazHPO4 (7.8g 1") and 0.15M NaCl (8.8
g l") adjusted to pH 9.0

Stain: DTAF (5-(4,6-dichlorotriazin-2-yl) aminofluorescein) — 2mg in 10ml buffer
Water: filtered, distilled

Filter buffer, stain, and water through 0.22pm membrane. Stain solution is good for 1
day.

Vortex for 10 sec per time.

Sample and Slide preparation

1.
2.
3.

4.

£1!

Combine 0.5g eroded soil and 10mL filtered distilled water in sedimentation tube.
Sonicate for 5 sec continuous with cup holder attachment.

Vortex and allow coarse particles to sediment for 42 minutes (Stoke ’5 law). Transfer
top 5mL sample to disposable centrifuge tube.

Increase dilution from I :20 to 1:200 (soil:water) (vortex and transfer 0.5m] and add
4. 5ml water).

Add 0.1mL formaldehyde. (can stop here and store at 4°C until slide preparation)
Vortex sample before subsampling to prevent further sedimentation for replicate
slides.

Place 4p.L of soil suspension onto 6mm wells on slides. Spread suspension around
well with pipette tip without touching the tip to the slide surface.

8. Allow smears to dry completely (e.g., overnight)

Staining

1. Flood each well with 8uL of stain.

2. Store slide(s) in covered container with wet tissue to prevent drying.

3. Stain for 30 minutes.

4. Place slide(s) in staining jar containing buffer for 30 minutes, three times.

5. Place slide(s) in staining jar containing distilled water for final rinse for 30 minutes.
6. Remove slide(s) from final rinse and air dry.

7. Add small drop of VECTASHIELD Mounting medium (Vector Laboratories) to each

smear and cover with large (50mm) cover glass. Add immersion oil to top of cover
slip before microsc0py.

Counting
Observe bacteria on a fluorescence microsc0pe using a 10x ocular, a 100x objective and

“I 3 ” filter.

101

 

Figure A.2. Editing procedure for CMEIAS analysis of DTAF stained images. LJ

1.

 

Open image in Adobe Photoshop

2. Resize image and increase pixel resolution (Image<image size, uncheck resample

#99:“

10.
11.

12.

l3.
14.

15.

l6.
17.

image box, increase resolution to 300 pixels/inch — size automatically adjusts)
Decrease brightness to 15; increase contrast to 15
(Image<adjust<brightness/contrast)

Duplicate image (Image<duplicate); change name (delete 1 and add e)

Change to grayscale (Image<mode<grayscale)

Invert image(Image<adjust<invert)

Increase brightness to 5-10 (depends on image — increase to reduce background
color and noise, but don’t want to lose cells)

Mark each cell (pencil tool) with a black dot; constantly look between original
image (color) and edited image

Zoom in (300-400%) and outline cells with pencil tool; may need to compare with
original image

Use paint tool to fill in cells

Use eraser (change color so white is on top) to fill in edges of cells; change colors
back to black on top

Adjust levels to decrease background to white (move almost all the way to the left
of the box) (Image<adjust<levels)

Move cells that are close together apart or are at edge of image in

Decrease noise of the image (smoothes out cells some — decreases size slightly)
(Filter<noise<median)

Add bar scale (10pm) from microscope calibration scale acquired from same
magnification

Flatten image (Layer<flatten image)

Save edited image

102

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.... 00.0 00.0 ~n0é ~00 00.0 00v.0 0:000... _ 930 “.2
3.0 3.0 2.0 N030 00.0 00.0 N500 00+m~00 m Q30 ”.2
00 0th :20 N0...0 50.0 00.0 000.0 00+MON0 _ D\>>0 “.2

3:. 00.0 00.0 005.0 0N0 0V0 0N0.0 0I+mvoé w Q20 “.2
n: 050 00.0 000;.‘ 00.0 Nvd 50.0 00+m0_..~ _ D\>>0 ”.2
000 0NN :0 00N.0 :0 Nvd 0N0.0 09000 w 930 ”.2

50—. 05v 00.0 00w.0 5N0 0V0 0000 03.900 _ ._.O

000? 0N8 3.0 50.0 00.0 _.0.0 :00 0.4050 w H0

02:. 00.0 00.0 000.0 00.0 Nvd N000 0I+w 90.0 _ “.2

500—. 0NN 0...0 036 5.0 3.0 000.0 0_.+M00.0 w ”.2
000 00K :0 n0_..N 00.0 00.0 0_.N.I 00+w00.v _ n5

NO0F 5.5 00.0 0V0.N N00 00.0 00W? 00+w00.v 0 n5

000p 0IN 3.0 INQN 00.0 0.; F004. 00+MNNO _ 00v.

n00w 00K 3.0 0N0._. 00.? 00... 005; 00+w0w0 w 00v.

30? 00.0 3.0 000.N 0.0 00.0 00.0 0I+m _.0._. E08 00x
005 50.0 3.0 V000. n00 05.0 0v0.0 00+MVO$ >6 :0 00v.

“C300 «00.50 83::me EDEEF: wmmcgmxw GUOE cmfiwe came :00 O\m_._0uomm
mED_O>0_m

 

 

 

 

 

88388 88 80 8% .00 08883 =Eo>O ._.< 203.

103

   

lliliilll‘ijiljliiiiijiifiilijiljllil