ABSTRACT
Title of dissertation: PEDOGENESIS IN RAIN GARDENS:
THE ROLE OF EARTHWORMS AND
OTHER ORGANISMS IN LONG-TERM
SOIL DEVELOPMENT
Emily Mitchell Ayers, Doctor of Philosophy, 2009
Dissertation directed by: Associate Professor Patrick Kangas
Department of Environmental Science
and Technology
As bioretention comes into widespread use, it has become increasingly important
to understand the development of bioretention soils over time. The objective of this
research is to investigate the development of bioretention soils and the importance of
ecological processes in the performance of rain gardens. The research includes descriptive
studies of pre-existing rain garden soil pro les, laboratory experiments quantifying the
e ect of earthworms on in ltration rates, and a simulation model describing the in uence
of earthworms and soil organic matter on in ltration. Surveys of several di erent rain
gardens of various ages provide the  rst detailed descriptions of rain garden soil pro les.
The study revealed a great deal of biological activity in rain gardens, and evidence of
pedogenesis even in very young sites. The uppermost soil layers were found to be enriched
with organic matter, plant roots, and soil organisms. The  eld sites surveyed showed
no signs of clogging due to the trapping of suspended solids carried in stormwater runo .
Some evidence was found of higher than expected in ltration rates at the  eld sites, which
may be attributable to the e ects of bioturbation by living organisms.
The ability of earthworms to mitigate the e ects of trapped suspended solids on
bioretention soils was assessed in the laboratory. Results show that earthworms are capa-
ble of maintaining the in ltration rate of bioretention soils, but that their e ects have a
high degree of variability. This variability is attributed to soil aggregate instability caused
by the oversimpli cation of the ecosystem. Other organisms play a signi cant role in stabi-
lizing earthworm burrows and casts, and may be essential ingredients in a self-maintaining
bioretention ecosystem.
A simulation model of the action of earthworms on soil in ltration rates was devel-
oped in order to illustrate the physical processes taking place as a result of earthworm
activity. The model was calibrated using data from the  eld study and microcosm exper-
iment.
This research is intended to provide a  rst glimpse into the biological processes at
work in rain garden soils. The research shows that soil organisms are present in rain
gardens, and suggests that their impact on bioretention performance may be signi cant.
PEDOGENESIS IN RAIN GARDENS:
THE ROLE OF EARTHWORMS AND OTHER ORGANISMS
IN LONG-TERM SOIL DEVELOPMENT
by
Emily Mitchell Ayers
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park in partial ful llment
of the requirements for the degree of
Doctor of Philosophy
2009
Advisory Committee:
Associate Professor Patrick Kangas, Chair/Advisor
Professor Allen Davis
Professor Raymond Weil
Professor Adel Shirmohammadi
Assistant Professor Stephanie Lansing
c Copyright by
Emily Mitchell Ayers
2009
Dedication
For my daughters, Evelyn and Violet
ii
Acknowledgments
I would like to express my heartfelt thanks to the many people who have helped me
through this process. Thank you to my advisor, Dr. Patrick Kangas, who has spent the
last  ve years training me as a researcher, laughing only a little bit when I?m startled by
a spider out in the  eld. Your constant support and encouragement have kept me going
even as my life has moved away from the University. To my committee members, Dr.
Allen Davis, Dr. Raymond Weil, Dr. Adel Shirmohammadi, and Dr. David Tilley, for
your excellent instruction in your respective  elds, your valuable comments and questions,
and your generous and patient mentoring. You have challenged me to synthesize the many
disciplines that inform this research in a careful and rigorous manner, and my research
is much the better for it. Thank you to my newest committee member, Dr. Stephanie
Lansing, for your generous willingness to step in for the  nal stages.
Thank you to my husband, Andrew, for your support, patience, constant reassur-
ance, and willingness to sacri ce so that I could achieve my goals. Thank you for your
good humor as our apartment was taken over by soil columns, and the kitchen was com-
mandeered as a laboratory.
Thank you to my family for your love and encouragement, and especially to my
mother, Barbara, for providing the logistical support that allowed me to  nish this docu-
ment while raising my twin girls. I could not have done it without your help. Thank you
Mom and Dad for instilling in me the skills to tackle such a large undertaking, and the
con dence to believe that I could do it.
Thank you to Rebecca Stack, for taking me under your wing, for helping me to  nd
 eld sites and hard-to- nd construction documents, and most of all for your friendship.
Thank you to Dave Blersch, for spending hours and hours listening and helping me to
iii
hash out the ideas underpinning this research. Thank you to Kimberly Monahan for your
logistical support, advice, encouragement, and friendship.
Thank you to Aisha Sexton, Yin-Phan Tsang, and Frank Koh for your help with
the  eld work. Thank you to Heidi McMillen, Nicholas Regalia, Laura Senkowsky, Katie
Struder, and Kyle Wagner for helping to perform the preliminary earthworm microcosm
survivorship study as your capstone project for NRMT 470. Thank you to Dr. Martin
Rabenhorst for allowing me the use of your lab for sieve analysis, and to Danielle Baldu 
for teaching me to use the equipment.
Thank you to those who granted me access to their sites: Krista Grigg at Washington
Navy Yard, John Hilley at UMCP, Mike Wagner at Beltway Plaza Mall, Donald Vann at
Laurel Regional Hospital, John Bois at Northwestern High School, Principal Elizabeth
Livingston at Mary Harris \Mother" Jones Elementary School, Wilma Linares and James
Ayree at Chevy Chase Bank, Paul DeSousa at Prince George?s County DER, and Laura
Baxter at the Chesapeake Bay Foundation.
Thank you to Dr. Nelson Beyer at Patuxent Wildlife Research Center for your
advice on earthworm preservation, to Jerry Maldonado at Prince George?s County DER
for help in tracking down construction documents, and to Neil Weinstein at the Low
Impact Development Center for helping me to locate construction documents and other
historical data.
iv
Table of Contents
List of Tables ix
List of Figures xv
List of Abbreviations xxi
1 Introduction and Literature Review 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Pedogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Pedogenesis in rain gardens . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.3 The role of biology in pedogenesis . . . . . . . . . . . . . . . . . . . 12
1.1.4 In uence of earthworms on soil structure . . . . . . . . . . . . . . . 14
1.1.5 Self-organization and ecological succession . . . . . . . . . . . . . . . 15
1.1.6 Use of microcosms in ecological research . . . . . . . . . . . . . . . . 17
1.2 Summary of Literature Findings and Research . . . . . . . . . . . . . . . . 18
1.2.1 The history of bioretention design . . . . . . . . . . . . . . . . . . . 18
1.2.2 Bioretention research . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.2.3 Use and value of chronosequences . . . . . . . . . . . . . . . . . . . . 22
1.2.4 Pedogenesis in abandoned farmland and reclaimed mines . . . . . . 24
1.2.5 Earthworms and pedogenesis . . . . . . . . . . . . . . . . . . . . . . 25
1.3 Goals of This Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2 Field Study 31
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1.1 Background on  eld sites . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1.1.1 University of Maryland (UMCP) . . . . . . . . . . . . . . . 33
2.1.1.2 Washington Navy Yard (WNY) . . . . . . . . . . . . . . . 35
2.1.1.3 Mary Harris \Mother" Jones Elementary School (MJES) . 37
2.1.1.4 Chesapeake Bay Foundation Headquarters (CBF) . . . . . 39
2.1.1.5 Northwestern High School (NWHS) . . . . . . . . . . . . . 41
2.1.1.6 Inglewood Center III (PP) . . . . . . . . . . . . . . . . . . 43
2.1.1.7 Claggett Farm (CF) . . . . . . . . . . . . . . . . . . . . . . 45
2.1.1.8 Chevy Chase Bank (CC) . . . . . . . . . . . . . . . . . . . 47
2.1.1.9 Beltway Plaza Mall (BP) . . . . . . . . . . . . . . . . . . . 49
2.1.1.10 Laurel Regional Hospital (LRH) . . . . . . . . . . . . . . . 51
2.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.1 Field sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.2 Laboratory analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.3.1 University of Maryland (UMCP) . . . . . . . . . . . . . . . . . . . . 56
2.3.2 Washington Navy Yard (WNY) . . . . . . . . . . . . . . . . . . . . . 60
2.3.3 \Mother" Jones Elementary School (MJES) . . . . . . . . . . . . . . 63
2.3.4 Chesapeake Bay Foundation Headquarters (CBF) . . . . . . . . . . 66
2.3.5 Northwestern High School (NWHS) . . . . . . . . . . . . . . . . . . 69
2.3.6 Inglewood Center III (PP) . . . . . . . . . . . . . . . . . . . . . . . . 72
2.3.7 Claggett Farm (CF) . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
v
2.3.8 Chevy Chase Bank (CC) . . . . . . . . . . . . . . . . . . . . . . . . 79
2.3.9 Beltway Plaza Mall (BP) . . . . . . . . . . . . . . . . . . . . . . . . 83
2.3.10 Laurel Regional Hospital (LRH) . . . . . . . . . . . . . . . . . . . . 87
2.3.11 Results of in ltration tests . . . . . . . . . . . . . . . . . . . . . . . 91
2.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.4.1 Extent of biological activity . . . . . . . . . . . . . . . . . . . . . . . 93
2.4.1.1 Root systems . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.4.1.2 Soil animals . . . . . . . . . . . . . . . . . . . . . . . . . . 95
2.4.1.3 Earthworms . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.4.1.4 Soil organic matter enrichment . . . . . . . . . . . . . . . . 100
2.4.2 Baseline physical data . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.4.2.1 Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.4.2.2 In ltration rate . . . . . . . . . . . . . . . . . . . . . . . . 103
2.4.2.3 Evidence of pedogenesis . . . . . . . . . . . . . . . . . . . . 105
2.4.2.4 Changes in root biomass over time . . . . . . . . . . . . . . 106
2.4.2.5 Changes in earthworm abundance over time . . . . . . . . 110
2.4.2.6 Colonization by soil animals . . . . . . . . . . . . . . . . . 112
2.4.3 Confounding factors . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
2.4.3.1 E ect of soil texture on earthworm populations . . . . . . . 113
2.4.3.2 E ect of plant cover and litter layer . . . . . . . . . . . . . 115
2.4.3.3 Proximity to natural habitat . . . . . . . . . . . . . . . . . 115
2.4.3.4 E ect of management . . . . . . . . . . . . . . . . . . . . . 115
2.4.4 Sources of error in animal sampling . . . . . . . . . . . . . . . . . . . 116
2.4.4.1 Time of year . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.4.4.2 Outdoor temperature . . . . . . . . . . . . . . . . . . . . . 117
2.4.4.3 Antecedent weather conditions . . . . . . . . . . . . . . . . 119
2.4.5 In ltration tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.4.6 Diversity among rain gardens . . . . . . . . . . . . . . . . . . . . . . 121
2.4.6.1 Di erences in history, location . . . . . . . . . . . . . . . . 121
2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
3 Microcosm Study 125
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.2.1 In ltration testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.2.2 Earthworm feeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.2.3 Storm simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.2.3.1 Suspended solids loading . . . . . . . . . . . . . . . . . . . 134
3.2.3.2 Statistical analyses . . . . . . . . . . . . . . . . . . . . . . . 136
3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.3.1 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.3.2 Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.3 Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.3.4 E ect of treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
3.3.5 Earthworm survivorship . . . . . . . . . . . . . . . . . . . . . . . . . 148
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
3.4.1 Sources of error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
3.4.2 Comparison to Li and Davis (2008b) . . . . . . . . . . . . . . . . . . 163
vi
3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
4 Model 167
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.1.1 The energy systems language . . . . . . . . . . . . . . . . . . . . . . 168
4.1.2 RAINGARDEN model . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
4.2.1 Model run 1: organic matter (O) and earthworms (W) held constant 177
4.2.2 Model run 2: no earthworms (W = 0) . . . . . . . . . . . . . . . . . 179
4.2.3 Model run 3: less cohesive organic matter (increased k4 from 448.35
to 5,000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
5 Conclusions 183
5.1 Implications for Rain Garden Design . . . . . . . . . . . . . . . . . . . . . . 188
5.1.1 Design with self-organization in mind . . . . . . . . . . . . . . . . . 188
5.1.2 Limit pesticide use in the rain garden and its watershed . . . . . . . 188
5.1.3 Avoid disturbance of topsoil and mulch . . . . . . . . . . . . . . . . 189
5.1.4 Limit the sand content of the Bioretention Soil Medium (BSM) . . . 190
5.1.5 Consider seeding rain gardens with earthworms . . . . . . . . . . . . 191
5.1.6 Consider seeding rain gardens with mycorrhizal fungi . . . . . . . . 192
5.1.7 Plant densely . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
5.1.8 Mow to encourage high productivity . . . . . . . . . . . . . . . . . . 193
5.2 Suggestions for future research . . . . . . . . . . . . . . . . . . . . . . . . . 193
A Additional Background on Field Sites 196
A.1 University of Maryland (UMCP) . . . . . . . . . . . . . . . . . . . . . . . . 196
A.1.1 Washington Navy Yard (WNY) . . . . . . . . . . . . . . . . . . . . . 198
A.1.2 Mary Harris \Mother" Jones Elementary School (MJES) . . . . . . 200
A.1.3 Chesapeake Bay Foundation Headquarters (CBF) . . . . . . . . . . 202
A.1.4 Northwestern High School (NWHS) . . . . . . . . . . . . . . . . . . 202
A.1.5 Inglewood Center III (PP) . . . . . . . . . . . . . . . . . . . . . . . 205
A.1.6 Claggett Farm (CF) . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
A.1.7 Chevy Chase Bank (CC) . . . . . . . . . . . . . . . . . . . . . . . . 208
A.1.8 Beltway Plaza Mall (BP) . . . . . . . . . . . . . . . . . . . . . . . . 210
A.1.9 Laurel Regional Hospital . . . . . . . . . . . . . . . . . . . . . . . . 212
B Soil Testing Procedures 213
B.1 Procedure for Hydrometer Test . . . . . . . . . . . . . . . . . . . . . . . . . 213
B.1.1 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
B.1.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
B.1.3 Hydrometer test data sheet . . . . . . . . . . . . . . . . . . . . . . . 215
B.2 Procedure for Sieve Testing of Soil Samples . . . . . . . . . . . . . . . . . . 219
B.2.1 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
B.2.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
C Taxonomic Key to the Earthworms of Maryland 220
vii
D Raw Field Data 224
D.1 University of Maryland (UMCP) . . . . . . . . . . . . . . . . . . . . . . . . 224
D.2 Washington Navy Yard (WNY) . . . . . . . . . . . . . . . . . . . . . . . . . 227
D.3 \Mother" Jones Elementary School (MJES) . . . . . . . . . . . . . . . . . . 230
D.4 Chesapeake Bay Foundation Headquarters (CBF) . . . . . . . . . . . . . . . 233
D.5 Northwestern High School (NWHS) . . . . . . . . . . . . . . . . . . . . . . 236
D.6 Inglewood Center III (PP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
D.7 Claggett Farm (CF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
D.8 Chevy Chase Bank (CC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
D.9 Beltway Plaza Mall (BP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
D.10 Laurel Regional Hospital (LRH) . . . . . . . . . . . . . . . . . . . . . . . . 251
E Microcosm Pilot Project 254
F Detailed Earthworm Survivorship Data for Microcosm Experiment 256
G RAINGARDEN Model Development 262
G.1 State Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
G.2 Forcing Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
G.3 Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
viii
List of Tables
1.1 Summary of bioretention research to date. . . . . . . . . . . . . . . . . . . . 21
2.1 Common and scienti c names of animal taxa. . . . . . . . . . . . . . . . . . 54
2.2 Summary of data collected at UMCP, averaged. . . . . . . . . . . . . . . . . 57
2.3 Summary of data collected at WNY, averaged. . . . . . . . . . . . . . . . . 61
2.4 Summary of data collected at MJES, averaged. . . . . . . . . . . . . . . . . 64
2.5 Summary of data collected at CBF, averaged. . . . . . . . . . . . . . . . . . 67
2.6 Summary of data collected at NWHS, averaged. . . . . . . . . . . . . . . . . 70
2.7 Summary of data collected at PP, averaged. . . . . . . . . . . . . . . . . . . 73
2.8 Summary of data collected at CF, averaged. . . . . . . . . . . . . . . . . . . 76
2.9 Summary of data collected at CC, averaged. . . . . . . . . . . . . . . . . . . 80
2.10 Summary of data collected at BP, averaged. . . . . . . . . . . . . . . . . . . 84
2.11 Summary of data collected at LRH, averaged. . . . . . . . . . . . . . . . . . 88
2.12 Results of in ltration testing at BP site. . . . . . . . . . . . . . . . . . . . . 92
2.13 Results of in ltration testing at LRH site. . . . . . . . . . . . . . . . . . . . 92
2.14 Results of in ltration testing at NWHS site. . . . . . . . . . . . . . . . . . . 92
2.15 The roles of major invertebrate taxa in the soil ecosystem (Gobat et al.,
2003; Hole, 1981) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2.16 Average soil texture for each of the research sites. . . . . . . . . . . . . . . . 102
2.17 Variation in total number of soil animals counted and number of di erent
animal taxa encountered at each sampling depth with age. . . . . . . . . . . 112
2.18 Land use within the watersheds of sampling locations. . . . . . . . . . . . . 122
3.1 Treatment assignments for microcosms showing strati ed randomized ex-
perimental design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.2 Particle size distribution for bioretention soil mix. . . . . . . . . . . . . . . 128
3.3 Particle Size Distribution of manufactured sediment. . . . . . . . . . . . . . 129
ix
3.4 30-year rainfall averages for Baltimore, MD. . . . . . . . . . . . . . . . . . . 132
3.5 Design storm frequencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.6 Design storm volumes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.7 Particle size distribution for manufactured suspended solids (SS). . . . . . . 135
3.8 Results of saturated hydraulic conductivity tests for Control group. . . . . . 138
3.9 Two-sided paired t-test of Control group change. . . . . . . . . . . . . . . . 139
3.10 Results of saturated hydraulic conductivity tests for Treatment 1 group. . . 140
3.11 Two-sided paired t-test of Treatment 1. . . . . . . . . . . . . . . . . . . . . 140
3.12 Results of saturated hydraulic conductivity tests for Treatment 2 group. . . 143
3.13 Two-sided paired t-test of Treatment 2. . . . . . . . . . . . . . . . . . . . . 143
3.14 Mean and variance of initial saturated hydraulic conductivity for each of
the treatment groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
3.15 Levine?s test for homogeneity of variance for initial saturated hydraulic
conductivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.16 Analysis of variance for initial saturated hydraulic conductivity. . . . . . . . 147
3.17 Mean and variance of  nal saturated hydraulic conductivity for each of the
treatment groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.18 Levine?s test for homogeneity of variance for  nal saturated hydraulic con-
ductivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
3.19 Analysis of variance for  nal saturated hydraulic conductivity. . . . . . . . . 148
3.20 Comparison of variance in initial and  nal saturated hydraulic conductivity
among treatment groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
3.21 Comparison of bioretention soil medium used in this study to Li and Davis
(2008b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.22 Comparison of simulated suspended solids used in this study to Li and
Davis (2008b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.23 Comparison of experimental results with Li and Davis (2008b). . . . . . . . 165
4.1 Numerical values for the inputs, state variables, and  ows in the RAIN-
GARDEN model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4.2 Values of coe cients in model equations. . . . . . . . . . . . . . . . . . . . . 176
x
A.1 UMCP Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . 196
A.2 UMCP Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
A.3 WNY Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . 198
A.4 WNY Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
A.5 MJES Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . 200
A.6 MJES Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
A.7 NWHS Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . 202
A.8 NWHS Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
A.9 PP Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . . . 205
A.10 PP Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
A.11 CF Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . . . 207
A.12 CF Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
A.13 CC Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . . . 208
A.14 CC Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
A.15 BP Bioretention Speci cation . . . . . . . . . . . . . . . . . . . . . . . . . . 210
A.16 BP Planting Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
D.1 UMCP Earthworms - number of individuals and sum of lengths . . . . . . . 224
D.2 UMCP Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . 224
D.3 UMCP Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . 224
D.4 UMCP Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
D.5 UMCP Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . 226
D.6 UMCP Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . 226
D.7 WNY Earthworms - number of individuals and sum of lengths . . . . . . . 227
D.8 WNY Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . 227
D.9 WNY Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . 227
D.10 WNY Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
xi
D.11 WNY Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . 229
D.12 WNY Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . 229
D.13 MJES Earthworms - number of individuals and sum of lengths . . . . . . . 230
D.14 MJES Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . 230
D.15 MJES Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . 230
D.16 MJES Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
D.17 MJES Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . 232
D.18 MJES Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . 232
D.19 CBF Earthworms - number of individuals and sum of lengths . . . . . . . . 233
D.20 CBF Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . 233
D.21 CBF Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . 233
D.22 CBF Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
D.23 CBF Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . . 235
D.24 CBF Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . 235
D.25 NWHS Earthworms - number of individuals and sum of lengths . . . . . . . 236
D.26 NWHS Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . 236
D.27 NWHS Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . 236
D.28 NWHS Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
D.29 NWHS Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . 238
D.30 NWHS Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . 238
D.31 PP Earthworms - number of individuals and sum of lengths . . . . . . . . . 239
D.32 PP Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
D.33 PP Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . . 239
D.34 PP Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
D.35 PP Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . . . 241
D.36 PP Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . . 241
D.37 CF Earthworms - number of individuals and sum of lengths . . . . . . . . . 242
xii
D.38 CF Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
D.39 CF Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . . 242
D.40 CF Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
D.41 CF Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . . . 244
D.42 CF Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . . 244
D.43 CC Earthworms - number of individuals and sum of lengths . . . . . . . . . 245
D.44 CC Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
D.45 CC Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . . 245
D.46 CC Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
D.47 CC Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . . . 247
D.48 CC Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . . 247
D.49 BP Earthworms - number of individuals and sum of lengths . . . . . . . . . 248
D.50 BP Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
D.51 BP Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . . 248
D.52 BP Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
D.53 BP Soil invertebrates total tally for entire site. . . . . . . . . . . . . . . . . 250
D.54 BP Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . . 250
D.55 LRH Earthworms - number of individuals and sum of lengths . . . . . . . . 251
D.56 LRH Root biomass, air-dried . . . . . . . . . . . . . . . . . . . . . . . . . . 251
D.57 LRH Soil organic matter content . . . . . . . . . . . . . . . . . . . . . . . . 251
D.58 LRH Soil texture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
D.59 LRH Soil invertebrates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
D.60 LRH Earthworm species identi ed. . . . . . . . . . . . . . . . . . . . . . . . 253
F.1 Column 2, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
F.2 Column 3, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
F.3 Column 4, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
xiii
F.4 Column 6, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
F.5 Column 8, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
F.6 Column 9, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
F.7 Column 10, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
F.8 Column 12, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
F.9 Column 13, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
F.10 Column 14, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
F.11 Column 17, Treatment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
F.12 Column 18, Treatment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
xiv
List of Figures
1.1 Schematic of a typical rain garden . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Initial stages of pedogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Factors in uencing the drawdown time of a rain garden. . . . . . . . . . . . 11
1.4 A basic terrestrial food web that may be present in rain gardens. . . . . . . 13
2.1 Locations of  eld sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2 Photo of UMCP site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 UMCP site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4 Photo of WNY site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.5 WNY site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.6 Photo of MJES site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.7 MJES site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.8 Photo of CBF site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.9 CBF site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.10 Photo of NWHS site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.11 NWHS site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.12 Photo of PP site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.13 PP site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.14 Photo of CF site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.15 CF site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.16 Photo of CC site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.17 CC site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.18 Photo of BP site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.19 BP site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.20 Photo of LRH site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
xv
2.21 LRH site planview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.22 UMCP particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . 58
2.23 UMCP animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.24 UMCP animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.25 WNY particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.26 WNY animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.27 WNY animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.28 MJES particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.29 MJES animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.30 MJES animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.31 CBF particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.32 CBF animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.33 CBF animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.34 NWHS particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . 70
2.35 NWHS animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.36 NWHS animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.37 PP particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.38 PP animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.39 PP animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.40 CF particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.41 CF animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.42 CF animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.43 CC particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.44 CC animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.45 CC animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.46 BP particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . 84
2.47 BP animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
xvi
2.48 BP animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.49 LRH particle size distribution. . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.50 LRH animal diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
2.51 LRH animal depth pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
2.52 Mean root density for all research sites. . . . . . . . . . . . . . . . . . . . . 94
2.53 Soil animals found at each of the research sites. . . . . . . . . . . . . . . . . 97
2.54 Mean earthworm density for each of the  eld sites sampled. . . . . . . . . . 99
2.55 Mean upper layer % SOM for all sites. . . . . . . . . . . . . . . . . . . . . . 102
2.56 Comparison of measured versus predicted in ltration rates for the sites
sampled. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.57 Soil organic matter pro le development over time. . . . . . . . . . . . . . . 105
2.58 Organic carbon pro le development in a mine spoil chronosequence (Leis-
man, 1957). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.59 Soil organic matter in uppermost 10 cm versus age. . . . . . . . . . . . . . . 107
2.60 Root biomass versus age. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
2.61 Root pro le development over time. . . . . . . . . . . . . . . . . . . . . . . 109
2.62 Earthworm abundance versus age. . . . . . . . . . . . . . . . . . . . . . . . 110
2.63 Earthworm pro le development over time. . . . . . . . . . . . . . . . . . . . 111
2.64 Relationship between earthworm abundance and sand content. . . . . . . . 114
2.65 E ect of month of sampling on earthworms and total number of soil animals
detected. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.66 E ect of ambient temperature on earthworms and total numbers of soil
animals detected. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.67 E ect of precipitation on numbers of earthworms and other soil animals
detected. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.68 Changes in sand content with rain garden age. . . . . . . . . . . . . . . . . 123
3.1 Diagram showing microcosm setup. . . . . . . . . . . . . . . . . . . . . . . . 128
3.2 Diagram of setup to establish constant head for in ltration testing. . . . . . 130
xvii
3.3 Average saturated hydraulic conductivity for Control group. . . . . . . . . . 138
3.4 Quantile-quantile plot of Control group change. . . . . . . . . . . . . . . . . 139
3.5 Average saturated hydraulic conductivity for Treatment 1 group. . . . . . . 141
3.6 Quantile-quantile plot of Treatment 1 change. . . . . . . . . . . . . . . . . . 142
3.7 Average saturated hydraulic conductivity for Treatment 2 group. . . . . . . 144
3.8 Quantile-quantile plot for Treatment 2. . . . . . . . . . . . . . . . . . . . . . 145
3.9 Changes in earthworm size distribution over time for Treatment 1 group. . 149
3.10 Changes in earthworm size distribution over time for Treatment 2 group. . 150
3.11 Comparison of initial and  nal Ks for all treatment groups. . . . . . . . . . 152
3.12 Results of initial and  nal in ltration tests for Control group showing de-
clining in ltration rates over the course of the tests. . . . . . . . . . . . . . 155
3.13 Results of initial and  nal in ltration tests for Treatment 1 group showing
declining in ltration rates over the course of the tests. . . . . . . . . . . . . 156
3.14 Results of initial and  nal in ltration tests for Treatment 2 group showing
decreasing in ltration rates over the course of the tests. . . . . . . . . . . . 157
3.15 Final Ks versus total number of earthworms. . . . . . . . . . . . . . . . . . 159
3.16 Final Ks versus number of surviving adult L. terrestris. . . . . . . . . . . . 160
3.17 Initial Ks of Control group versus sampling date. . . . . . . . . . . . . . . . 161
3.18 Initial Ks of Treatment 1 group versus sampling date. . . . . . . . . . . . . 161
3.19 Initial Ks of Treatment 2 group versus sampling date. . . . . . . . . . . . . 161
3.20 Final Ks of Control group versus sampling date. . . . . . . . . . . . . . . . 162
3.21 Final Ks of Treatment 1 group versus sampling date. . . . . . . . . . . . . . 162
3.22 Final Ks of Treatment 2 group versus sampling date. . . . . . . . . . . . . . 162
4.1 Symbols used in the energy systems language. . . . . . . . . . . . . . . . . . 169
4.2 Energy systems diagram of simulation model RAINGARDEN. . . . . . . . 170
4.3 Standard run: crust thickness (C) over 1500 days. . . . . . . . . . . . . . . 177
4.4 Standard run: void space (V) and pore water (H) over 1500 days. . . . . . . 178
xviii
4.5 Soil organic matter (O) in a system without earthworms over 1500 days. . . 179
4.6 Pore space (V) and pore water (H) in a system without earthworms over
1500 days. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4.7 Crust thickness (C) in a system without earthworms over 1500 days. . . . . 180
4.8 Loss of porosity (V) over 1500 days in a system where organic matter is
less cohesive. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
5.1 Factors in uencing the drawdown time of a rain garden, with emphasis on
those factors included in the simulation model. . . . . . . . . . . . . . . . . 187
A.1 UMCP construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
A.2 WNY construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
A.3 MJES construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
A.4 NWHS construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
A.5 PP construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
A.6 CC construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
A.7 BP construction detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
B.1 Graph to determine silt and clay content. . . . . . . . . . . . . . . . . . . . 216
B.2 USDA textural triangle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
D.1 UMCP soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
D.2 WNY soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
D.3 MJES soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
D.4 CBF soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
D.5 NWHS soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
D.6 PP soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
D.7 CF soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
D.8 CC soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
D.9 BP soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
xix
D.10 LRH soil pro le. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
E.1 Survivorship of earthworms in media of varying sand content. . . . . . . . . 255
xx
List of Abbreviations
BSM Bioretention Soil Medium
CEC Cation Exchange Capacity
LID Low Impact Development
SOM Soil Organic Matter
SS Suspended Solids
UMCP University of Maryland, College Park
WNY Washington Navy Yard
MJES \Mother" Jones Elementary School
NWHS Northwestern High School
CBF Chesapeake Bay Foundation
PP \Peppercorn Place" (Inglewood Center, III)
CF Claggett Farm
CC Chevy Chase Bank
BP Beltway Plaza
LRH Laurel Regional Hospital
EPA Environmental Protection Agency
NOAA National Oceanic and Atmospheric Administration
NCDC National Climatic Data Center
PG County Prince George?s County, Maryland
USDA United States Department of Agriculture
xxi
Chapter 1
Introduction and Literature Review
1.1 Background
Bioretention is a new best management practice for the treatment of stormwater
runo from impervious areas such as driveways and parking lots. Bioretention cells (com-
monly referred to as rain gardens) represent a low-tech treatment option that e ectively
promotes in ltration, reduces stormwater quantity, and removes heavy metals, nutrients,
and other pollutants contained in stormwater runo (Davis et al., 2001a, 2003). A rain
garden is a small planted depression  lled with approximately 0.5 to 1 m (2 to 3 ft) of a soil
medium engineered to promote in ltration while at the same time providing enough clay
and organic matter to promote the breakdown or sorption of pollutants. Runo  owing
into the rain garden is  ltered as it percolates through the soil mix. Figure 1.1 shows the
con guration of a typical rain garden.
The increasing popularity of rain gardens as stormwater control structures highlights
the need for a more thorough understanding of bioretention performance. Rain gardens
are essentially planted soil  lters, and their performance during a storm event depends on
the hydraulic conductivity and pollutant removal capacity of the soil. These soil properties
are in uenced by the physical e ects of storm events as well as the actions of the plants,
microbes and soil animals of the rain garden ecosystem. Over time, physical, chemical
and biological forces change the composition and structure of the soil, which a ects the
performance of the rain garden. Thus, the development of bioretention soils over time
must be examined if the long-term performance of rain gardens is to be understood.
1
Figure 1.1: Schematic of a typical rain garden, consisting of a 2.5-foot-deep
layer of bioretention soil medium (BSM), topped with a 3-inch layer of mulch,
planted with a variety of native plants, shrubs and trees. The system is de-
signed to allow a 6-inch ponding depth at the soil surface, and is often out tted
with and underdrain leading to a stormwater discharge.
Many environmental factors can be expected to act on the structures, and these factors
may impact their hydraulic performance. For example, suspended solids are a major
component of urban stormwater runo . The fate of suspended solids in rain gardens is
not yet well understood, but it is likely that these solids are trapped in the upper layers of
the soil (Li and Davis, 2008b,a). Over time, this may lead to clogging of the rain garden,
decreasing the in ltration rate.
2
1.1.1 Pedogenesis
The key to understanding how rain gardens change over time lies in understanding
the soil. Thus, it is necessary to examine how soils in general behave, and how they
develop over time. This knowledge can then be applied to bioretention soils.
Soil is composed of weathered rock of various sizes, combined with organic matter,
and organized into aggregates. Aggregates provide structure to the soil, permitting air and
water to enter, and providing habitat for a multitude of organisms. A soil begins as a loose
assemblage of weathered rock, which over time undergoes pedogenesis. The  rst step in
the process of pedogenesis is the formation of an A horizon at the soil surface (Figure 1.2).
Organic matter begins to accumulate in the upper part of the soil, changing the soil?s color
and physical properties. The soil develops a characteristic black color due to the presence of
dark-colored humic substances. Humus increases the water and nutrient holding capacity
of the soil. Sticky organic substances, such as polysaccharides, are produced by a variety of
soil organisms, including plant roots, mycorrhizae and bacteria. These organic substances
cause soil particles to clump together into spheroidal aggregates. Many of these aggregates
are water-stable, and create a stable, porous soil structure. Soil aggregates are also formed
in the guts of earthworms. Many earthworms ingest soil, mixing it with organic matter
before excreting the mixed matter back into the soil. Earthworm fecal pellets are often
water-stable, and in some soils constitute a sizeable fraction of the soil aggregates (Lee and
Foster, 1991). Plant roots and macroinvertebrates such as earthworms create macropores,
further increasing  ow through the soil. Thus, it is biological activity that creates and
maintains the permeability of the soil.
3
Jenny (1941, 1980) proposed that pedogenesis is governed by what he termed the
\CLORPT" equation:
s = f(cl;o;r;p;t) (1.1)
which states that soil properties (s) are a function of climate (cl), organisms (o), topogra-
phy (r), parent material (p), and time (t). This equation neatly encapsulates the major
processes involved in soil pedogenesis. Since it was originally proposed, the exact struc-
ture of the CLORPT equation has been the subject of intense debate among soil scientists
(Crocker, 1952; Johnson and Watson-Stegner, 1987; Stevens and Walker, 1970; Yaalon,
1975), but the equation?s utility as a conceptual model of the main in uences on soil
development has always been recognized.
Climate. The principal climatic factors in uencing soil development are temperature
and precipitation (Yaalon, 1983). At higher temperatures, organisms are more active,
and therefore have greater in uence on soil structure. High temperatures also increase
evaporation rates, decreasing the moisture content of the soil. Vegetation is in uenced
by climate. Even the reaction rates of chemical processes within the soil are increased
by increasing temperature. The soil moisture regime plays a major role in soil formation.
The balance between precipitation and evapotranspiration determines the  ux of water
through the soil, redistributing soil particles and leaching soluble materials. Plants and
soil organisms require adequate soil moisture in order to  ourish. Saturated conditions
can occur where there is excess precipitation, and if these saturated conditions persist,
oxygen can be depleted in the soil. This causes distinct and dramatic changes in the  oral
and faunal communities, as well as chemical changes in the soil.
Organisms. Organisms living within and on the surface of a soil produce pronounced
changes in the soil pro le (Gobat et al., 2003). Plants growing at the soil surface con-
4
tribute organic matter via litterfall and decay of plant roots. Plant roots exude oxygen
and simple sugars, creating a region known as the rhizosphere. The rhizosphere provides
critical habitat for bacteria and fungi. Soil animals and microorganisms work together to
break down plant litter and incorporate it into the soil matrix. This increase in organic
matter changes the structure of the soil. Soil organic matter decreases bulk density, in-
creases water holding capacity, and provides the glue that holds soil aggregates together
(Brady and Weil, 2002; Weil et al., 2004). Bioturbation by macroinvertebrates, such as
earthworms, as well as burrowing vertebrates, such as moles, aerates and mixes the soil
and creates large macropores.
Topography. Rain falling on a slope will tend to run downhill before it has the
opportunity to completely in ltrate into the soil. As rain runs o , it removes soil particles,
depositing them at the base of the slope. Thus, soils on slopes tend to be drier and have
shallower pro les than soils at the bases of slopes.
Parent material. The parent material of a soil acts is baseline from which the soil
develops. The parent material determines many of the basic characteristics of the soil.
A soil formed from silty loess deposits will be fundamentally di erent from a soil formed
from sandy marine deposits, even if these two soils form under identical climatic and
topographic conditions. Sandy soils have much higher hydraulic conductivities than silt
or clay soils (Ferguson, 1994), and will therefore leach more rapidly.
The mineral content of soils is also important in their development. For example, a
soil formed from limestone will tend to have a high pH, which will e ect the composition
of the plant community that develops, and will in uence the chemical reactions taking
place in the soil.
5
Time. All of the soil processes take place on di erent time scales, ranging from years
to hundreds or even thousands of years. The weathering of rock tends to take thousands of
years. Under favorable conditions (generally warm and moist climates), buildup of organic
matter can take place over a period of years, and downward migration of clays can occur
over several decades (Brady and Weil, 2002; Stevens and Walker, 1970; van Breemen and
Buurman, 2002).
Figure 1.2: Initial stages of pedogenesis, after Brady and Weil (2002). An
organic horizon (O) forms at the soil surface, and is gradually broken down
and incorporated into the uppermost soil layer, forming an A horizon.
6
1.1.2 Pedogenesis in rain gardens
The same soil forming factors act on bioretention soils, which begin as a uniform mix-
ture of sand, topsoil and organic matter. Speci cations for the bioretention soil medium
vary by municipality. One commonly used speci cation is 50% sand, 30% sandy loam
topsoil, and 20% organic matter in the form of mulch or compost by volume, uniformly
mixed (LIDC, 2003).
Climate. Rain gardens are often constructed in parking lots or other urban settings,
where paved surfaces create microclimates with elevated temperatures (Pickett et al.,
2001). These higher temperatures can be expected to accelerate soil formation by increas-
ing the activity of soil organisms. As rain gardens are designed to receive the precipitation
falling on an area much larger than the rain garden itself, the e ective precipitation on
the rain garden is much higher than that of the surrounding area. Bioretention soils are
highly permeable, allowing this precipitation to in ltrate quickly through the soil. Anoxic
conditions rarely develop. High e ective precipitation can be expected to accelerate the
leaching of soluble substances and the downward migration of clay particles.
Organisms. Rain gardens are largely devoid of organisms when they are  rst in-
stalled, though the topsoil and organic matter will contain microbes, fungi, seeds, and
some meso- and macro-invertebrates. After installation, the rain garden is planted. Over
time, invertebrates colonize the soil, and plant root systems develop into dense networks.
As the populations of organisms increase, the organic matter content of the upper soil
layer can be expected to increase, and a granular structure will develop. These changes
should begin almost immediately, and can be expected to have a major impact on the
rain garden throughout its operational life. The e ect of the rain garden ecosystem on
7
the bioretention soil is the major focus of this research, and will be discussed in greater
detail.
Topography. Rain gardens are located at the bases of largely impervious micro-
watersheds. Rain falling on an impervious surface will dislodge  ne particles. These
particles are carried o by the runo , and are ultimately deposited in the rain garden.
Over time, this deposition of  ne particles into the rain garden may clog the soil surface,
and may eventually alter the rain garden soil texture.
Parent material. Bioretention soils typically have a very high sand content, and
very high hydraulic conductivity, typically 40 { 50 cm/h (Li and Davis, 2008b). Rapid
 ow of water through the soil accelerates the leaching of soluble materials from the soil,
and also limits the soil?s water holding capacity, which prevents the development of anoxic
conditions during wet periods, but may stress plants during dry periods.
Time. With the high temperatures, high precipitation and high permeability de-
scribed above, bioretention soils may be expected to develop rapidly.
Over the  rst few years, an organic (O) horizon would be expected to form at the
soil surface, followed by a gradually thickening A horizon just below the soil surface (Fig-
ure 1.2). These changes in soil structure will a ect the hydraulic properties of the rain
garden. The in ltration rate of the bioretention soil is critically important to their proper
function. Figure 1.3 illustrates the major factors that in uence the drawdown time of a
rain garden. The drawdown time, or the time it takes ponded in ow to exit the system,
is the most important design parameter of a rain garden. A rain garden must be able to
process stormwater e ciently, while holding the stormwater for a su cient minimum time
to remove pollutants. The major external forces that in uence the drawdown time are
storm characteristics, maintenance procedures, and the initial planting plan. These exter-
8
nal forces in uence the rain garden?s internal physical and biological systems, impacting
such factors as the soil?s porosity, texture, in ltration rate, root biomass, organic matter
content, and invertebrate populations.
Storm characteristics, such as intensity, frequency and duration, determine the level
of suspended solids loading to the system, the amount of sodium loading from road salts,
and the toxic pollutant load that the system must handle. Suspended solids loading
directly impacts the in ltration rate of the system by clogging the soil surface. As these
solids are incorporated deeper into the soil mix, they decrease the soil?s porosity, and
gradually change the soil texture. Sodium used in road salts has a direct negative impact
on soil structure, decreasing the porosity and in ltration rate of the soil (Ramakrishna
and Viraraghavan, 2005). It is also highly toxic to plants and invertebrates, contributing
to the system?s toxic pollutant load. The porosity and the in ltration rate of the system
combine to determine the soil?s percolation rate, which, along with the presence or absence
of a liner, determines the drawdown time of the rain garden.
The initial planting plan plays an important role in how the biological system de-
velops, and therefore how it responds to storm events. The types of plants used and their
spacing determine the root biomass in the soil and the amount of litter falling on the soil
surface. Litterfall serves as food for earthworms (as well as other soil-dwelling inverte-
brates). The earthworm population is also sensitive to the soil texture, preferring loamy
soils without excessive sand or clay, and is negatively impacted by the toxic pollutant load.
Earthworms create aggregates and biopores, increasing the soil?s porosity. Roots create
biopores, which increase the soil?s porosity, and generate root exudates, which increase soil
aggregation. Ants create biopores. Soil organic matter is created through the metabolic
activities of earthworms, ants, and other soil organisms, as well as through root exudates.
9
This organic matter increases soil aggregation both directly and indirectly by serving as
a food source for fungal hyphae, which also increase the aggregation of the soil.
Maintenance procedures also impact the functioning of the system. Periodic removal
of trapped solids (cake) from the soil surface increases the soil?s in ltration rate and
porosity, and a ects the soil texture by removing  ne sand, silt and clay that might
otherwise be incorporated into the soil. Removal of plant detritus (litterfall), whether
intentional or as a byproduct of rubbish removal, limits the amount of food available to
the biological system.
The diagram reveals the links between a rain garden?s physical and biological sys-
tems. The biological system plays a fundamental role in the maintenance and regeneration
of the soil porosity after storm events. This biological system is highly dependent on good
initial design and enlightened maintenance to ensure an adequate food supply. Any action
that limits the amount of plant litter available limits the carrying capacity of the rain
garden. Toxic pollutants carried in stormwater runo can negatively a ect invertebrate
populations, but the high capacity of humus to adsorb heavy metals may limit exposure,
mitigating this risk.
10
Figure
1.3:
Factors
in uencing
the
dra
wdo
wn
time
of
arain
garden.
11
1.1.3 The role of biology in pedogenesis
Biological activity is the key to the creation and maintenance of aggregates and
macropores in the upper soil layers (Six et al., 2004). Organic material in the form of
plant litter, decaying roots and animal wastes is gradually broken down and incorporated
into the soil. This process is carried out by soil organisms. Soil contains a wide diversity
of organisms, ranging in size from mammals, such as prairie dogs and moles, down to
microscopic mites and nematodes, all the way to bacteria and viruses. These organisms
each inhabit specialized ecological niches, forming a complex soil ecosystem.
The soil fauna are responsible for the modi cation of soil structure. Macroinver-
tebrates (earthworms, isopods, millipedes, ants, termites) in uence soil structure both
through their feeding (incorporating, decomposing and mixing organic matter into the
soil) and their burrowing (aeration and mixing). Mesofauna (pot worms, springtails,
mites) feed on organic matter and fungal hyphae (Berry et al., 1994). This action breaks
down organic matter and mixes it into the soil matrix. Figure 1.4 illustrates a basic rain
garden food web, outlining the major organisms likely to be present and their roles in the
system.
The process of transforming plant tissue to humus is quite complex and involves the
entire soil ecosystem. Leaf litter is produced by plants. This litter is partially decomposed
on the soil surface by bacteria and fungi. The partly decayed plant matter is then con-
sumed by shredders, such as earthworms, which break the litter into smaller pieces, mix
it with bacteria in their digestive tracts, and deposit this mixture into the soil. Bacteria
and fungi then further attack the litter, gradually breaking complex organic substances
down into simpler ones. Most of the organic matter that is input into the soil system
is consumed by soil organisms. About  ve percent of the organic matter is converted to
12
simple compounds, such as polysaccharides, which are extremely important in the for-
mation of soil aggregates (Brady and Weil, 2002). About twenty percent of the organic
matter is transformed into the complex, heterogeneous substance known as humus. Humic
substances are highly resistant to degradation, colloidal, have a very high cation exchange
capacity (CEC), and can bind a large quantity of water (Weil et al., 2004). These traits
increase the ability of the soil to retain water and nutrients.
Figure 1.4: A basic terrestrial food web that may be present in rain gardens.
13
1.1.4 In uence of earthworms on soil structure
While a number of biological factors work together to improve soil structure, earth-
worms are among the most important (Lee and Foster, 1991; Kretzschmar and Edwards,
2004; Langmaack et al., 1999; Bouch e and Al-Addan, 1997; Coleman et al., 2004). In 1881,
Charles Darwin recognized that earthworms were responsible for the formation of topsoil
(Darwin, 1881). Since then, the e ect of earthworms on soil structure has been intensively
studied. Several chapters in earthworm ecology and soil science texts are dedicated to the
subject (Edwards and Arancon, 2004; van Vliet and Hendrix, 1999; Edwards and Bohlen,
1996; Brady and Weil, 2002). Earthworms can exert signi cant in uence on soil structure,
organic matter and nutrient cycling, and other soil organisms (van Vliet and Hendrix,
1999). This in uence is carried out in three principle ways:
1. The production of feces (casts) which harden to form stable aggregates.
2. The creation of burrows, which promote in ltration and soil aeration.
3. The incorporation of organic matter into the soil, through burrowing as well as
metabolic activities.
Earthworms have been divided into three important ecological groups: epigeic, endogeic
and anecic species (Bouch e et al., 1977). Epigeic species tend to live on the soil surface,
beneath the litter layer. They play a major role in the breakdown of surface litter. En-
dogeic species inhabit the upper soil horizons. They tend to form horizontal, meandering
burrows, and consume organic matter within the soil. Anecic species form long vertical
burrows in the soil. These burrows help to aerate the soil and increase in ltration (Lee,
1985). Through their feeding, earthworms incorporate organic matter into the soil. In-
gestion of plant litter and soil particles results in the mixing of the soil mineral fraction
with organic matter, glueing the two together as earthworm casts. Shaw and Pawluck
(1986) examined the e ects of di erent earthworm species on soil structure. Their  nd-
14
ings suggest that anecic species play a primary role in transporting organic matter from
the surface into the soil. Endogeic species are then responsible for mixing this introduced
organic matter throughout the soil.
The e ect of earthworms on aggregate stability, or the resistance of aggregates to
destruction when subjected to wetting or mechanical stresses, is complex (Six et al., 2004).
Some studies have found that earthworms increase aggregate stabilization (Shaw and
Pawluck, 1986), while others have found evidence that earthworms can actually destroy soil
structure (Schrader and Zhang, 1997). The ultimate e ect of earthworms on soil structure
seems to be dependent on a multitude of factors speci c to the soil and circumstances in
question.
The stability of fresh earthworm casts is generally lower than that of the surrounding
soil, but stability increases as the casts age over even a few days (Kay and Angers, 1999).
Cast stability is believed to derive from the colonization of the cast by fungal hyphae
present in the surrounding soil (Lee and Foster, 1991). Soils with high sand contents
lack su cient clay for abiotic aggregation to take place (Oades, 1993). In such soils,
aggregation depends on biotic forces, such as the ingestion of soil by macroinvertebrates
and the growth and decay of roots. Once created, this structure is stabilized by fungal
hyphae, colonies of microorganisms, and metabolic products from the decomposition of
higher plants (Oades, 1993). Soils with earthworms commonly have up to 70% pore space
(Lee, 1985).
1.1.5 Self-organization and ecological succession
Self-organization is the process by which a developing ecosystem self-selects the
assemblage of species best adapted to its particular environmental conditions, or energy
15
signature. At the most basic level, this process can be seen simply as the survival and
persistence of those species which do the best in the given environment. A basic strategy
of ecological engineering is to seed a site with as many species as possible, then allow the
site to self-organize. Those species which are well-adapted to the site will survive, and
those that are not will die out.
Succession is the process of community change in response to environmental changes.
Biotic communities change gradually in response to the physical environment. Eventually
the ecosystem reaches equilibrium with its energy signature. The community of plants and
animals at equilibrium is known as the climax community. In areas where bare ground
is generated due to earth-moving, abandonment of agricultural  elds, deforestation,  re,
or some other disturbance, the plant and animal communities go through a series of suc-
cessional stages, which follow a predictable general pattern. The area is  rst colonized
by pioneer species, which are typically fast-growing, annual plants and fast-reproducing,
opportunistic animal species. This community is gradually replaced by more mature
communities, which are characterized by slower-growing, perennial plants and slower- re-
producing, more specialized animal species. Eventually, the climax community is reached,
and the system is in equilibrium with its environment. The composition of this climax
community varies widely depending on the environment or biome; in the eastern United
States, the climax community is typically a hardwood forest. In the Midwestern United
States, the climax community may be tallgrass prairie. Climax ecosystems are character-
ized by low productivity and high stability, while immature ecosystems are characterized
by high productivity and low stability (Gutierrez and Fey, 1980). Both ecosystems may
be of use to engineers, depending on their needs.
16
1.1.6 Use of microcosms in ecological research
Microcosms are a common experimental unit in ecological research. Use of a micro-
cosm in a laboratory setting permits a level of experimental control that is often impossible
to achieve in the  eld. Most ecosystems of interest are much too large to be manipulated
experimentally, so they must either be observed or replicated at a smaller scale. In addi-
tion, manipulation of ecosystems in the  eld can alter the ecosystem in undesirable ways
that compromise its function. For example, manipulating the habitat of an endangered
species in order to  nd out what aspects of the habitat the species depends on for survival
would be an unacceptable method of conducting research. In such situations, it is some-
times possible to miniaturize the ecosystem and bring it into the controlled environment
of the laboratory. Microcosms o er several bene ts to ecological researchers:
1. They can be manipulated without fear of compromising the function of a live ecosys-
tem.
2. They can be replicated.
3. Conditions can be controlled.
4. The ecosystem of interest can be simpli ed to the desired level.
There are, however, limitations to the utility of microcosms, and the practice has come
under some criticism. The fundamental issue is whether or not a microcosm can be
constructed to accurately mimic the ecosystem of interest. Ecosystems consist of a web
of complex interactions, which are usually not fully understood. Microcosms intended
to simulate these ecosystems may be lacking essential components. In simplifying an
ecosystem to study a limited number of interactions, the system may again be altered to
the point where the behavior of the microcosm does not accurately re ect behavior in the
real world (Carpenter, 1996). Ecosystems which are very large can be very di cult to
simulate at a small scale, because they are dependent on very large processes. Ecosystems
17
do not exist independently from their surroundings. The  ows of energy and material in
and out of the system must be accounted for and accurately simulated.
1.2 Summary of Literature Findings and Research
1.2.1 The history of bioretention design
Bioretention was developed in Prince George?s County, Maryland in the early 1990s.
The design was conceived as a way to promote in ltration of storm water while simulta-
neously removing pollutants. As engineers have gained experience using rain gardens, the
design has evolved. In particular, as researchers have come to understand the importance
of the soil medium (Hsieh and Davis, 2005a), the speci cations for the soil medium have
become more tightly circumscribed. In early rain gardens, a wide range of soil textural
classes were deemed acceptable components. For example, at the Beltway Plaza Mall site,
constructed in 1997, the bioretention medium speci ed consisted of  ve parts topsoil to
one part wet loose peat moss or rotted manure (see Appendix A). Acceptable topsoil tex-
tures included: loam, sandy loam, clay loam, silt loam, sandy clay loam, or loamy sand.
By 2001, The Bioretention Manual, published by Prince Georges County?s Department of
Environmental Resources (PGCO, 2001), speci ed a medium consisting of: 50{60% sand,
20{30% leaf compost, and 20{30% topsoil by volume. The topsoil must be loamy sand,
sandy loam, or loam, with an in ltration rate higher than 0.5 in/hr. The Low Impact
Development Center, in cooperation with the US Environmental Protection Agency, has
developed a bioretention speci cation intended for nationwide use. This speci cation re-
quires that the Bioretention Soil Mixture (BSM) be composed of : 30% planting soil, 20%
shredded 2x hardwood mulch, and 50% sand by volume (LIDC, 2003). The planting soil
must be loamy sand or sandy loam. These speci cations are the result of hands-on ex-
18
perience and experimental optimization of the soil medium (Hsieh and Davis, 2005a). In
general, it seems that the sand/soil/organic ratio has been settled on, but the source of or-
ganic matter shifts between shredded hardwood mulch and leaf compost. This discussion
applies principally to variations in design speci c to Prince Georges County, Maryland and
its environs. Bioretention speci cations around the country do not necessarily follow this
formula, and can vary widely. For example, Emerson and Travers bioin ltration tra c
island was constructed using a 1:1 mixture of silt and sand, with no organic amendment
at all (Emerson and Traver, 2008).
Planting schemes are much less tightly controlled. The selection and placement
of plants is left to the landscape designer?s discretion, with the recommendation to se-
lect native plants that will tolerate variable hydrologic conditions, and will require little
maintenance (PGCO, 2001). In The Bioretention Manual (PGCO, 2001), Prince George?s
County provides a list of suggested species suitable for bioretention in the region.
1.2.2 Bioretention research
Research into the performance of bioretention as a stormwater treatment device
is still in its early days. Up until a few years ago, only a handful of papers had been
published. More recently, there has been an explosion of interest in the subject, and a
corresponding increase in the number of research studies. Table 1.1 is a compilation of the
research studies published to date. This research has focused primarily on the pollutant
removal performance of bioretention, testing di erent designs, soil media, and plantings in
the laboratory and in the  eld. A great deal of research has also concerned the hydrologic
performance of bioretention (e.g. peak  ow reduction).
19
Researchers have recently begun to turn their attention to the performance of biore-
tention systems over time. Li and Davis have studied the potential for bioretention systems
to become clogged with sediment (Li and Davis, 2008b,a). Emerson and Traver (2008)
recently published a four-year study of the hydrologic performance of a bioretention cell.
They found seasonal variations in in ltration rate, but no evidence of clogging over time.
The majority of bioretention research has focused on the soil medium. One study
(Hong et al., 2006) looked at the e ect of mulch on oil and grease removal. Only two
studies have looked at rain gardens as ecosystems. Culbertson and Hutchinson (2004)
looked at the e ect of plant types on performance. Greene et al. (2008), are currently
conducting a study that aims to quantify the e ect of vegetation (Tallgrass prairie grasses)
and earthworms (L. terrestris) on hydraulic performance and pollutant removal.
20
Table
1.1:
Summary
of
bioreten
tion
researc
hto
date.
Citation
Rep
orting
format
Factors
studied
(Da
vis
et
al.,
2001b)
Journal
article
Hea
vy
metals,
Nitrogen,
Phosphorus
(Da
vis
et
al,
2003)
Journal
article
Hea
vy
metals
(Kim
et
al.,
2003)
Journal
article
Nitrogen
(Culb
ertson
and
Hutc
hinson,
2004)
Conference
pap
er
Nitrogen
(Dietz
and
Clausen,
2005)
Journal
article
Phosphorus,
Nitrogen,
Hea
vy
metals,
Hydrologic
performance
(Hsieh
and
Da
vis,
2005a)
Journal
article
Hydrologic
performance,
TSS,
Oil
and
grease,
Phosphorus,
Nitrogen,
Hea
vy
me
tals
(Hsieh
and
Da
vis,
2005b)
Journal
article
TSS,
Oil
an
dgrease,
Hea
vy
metals,
Nitrogen,
Phosphorus
(Shark
ey
and
Hun
t,2005)
Conference
pap
er
Nitrogen,
Phosphorus,
Hydrologic
performance
(Da
vis
et
al.,
2006)
Journal
article
Nitrogen,
Phosphorus
(Dietz
and
Clausen,
2006)
Journal
article
Nitrogen,
Phosphorus,
Hea
vy
metals,
Hydrologic
performance
(Hong
et
al.,
2006)
Journal
article
Oil
and
gre
ase
(Hun
te
tal.,
2006)
Journal
article
TSS,
Nitrogen,
Phosphorus,
Hea
vy
metals,
Hydrologi
cp
erformance
(Zhang
et
al.,
2006)
Conference
pap
er
Phosphorus
(Da
vis,
2007)
Journal
article
TSS,
Hea
vy
metals,
Nitrogen,
Phosphorus
(Doughert
yet
al.,
2007)
Conference
pap
er
Nitrogen,
Phosphorus,
Hydrologic
performance
(Hsieh
et
al.,
2007a)
Journal
article
Phosphorus,
TSS
(Hsieh
et
al.,
2007b)
Journal
article
Nitrogen
(UNHSC,
2007)
Ann
ual
rep
ort
TSS,
Nitrogen,
Phosphorus,
Hea
vy
metals,
Hydrologi
cp
erformance,
Oil
and
greas
e
(Muthanna
et
al.,
2007a)
Journal
article
Hea
vy
metals,
TSS,
Chloride
(Muthanna
et
al.,
2007b)
Journal
article
Hea
vy
metals,
Hydrologic
performance
(Sun
and
Da
vis,
2007)
Journal
article
Hea
vy
metals
(Priv
ette
an
dW
eeb
er,
2007,
2008)
Conference
pap
ers
Hea
vy
metals,
Nitrogen,
Phosphorus,
Oil
and
grease,
Hydrologic
performance
(Rusciano
and
Obr
opta,
2007)
Journal
article
Fecal
coliform,
TSS
(Greene
et
al.,
2008)
Conference
pap
er
Earth
worms,
Vegetation,
Hydraulic
performance
(Hun
te
tal.,
2008)
Journal
article
TSS,
Nitrogen,
Phosphorus,
Hea
vy
metals,
BOD,
Fecal
coliform,
Hydrologic
performance
(Jones
and
Hun
t,2008)
Conference
pap
er
Temp
eratu
re
(Li
and
Da
vis,
2008a)
Journal
article
Hea
vy
metals
(Li
and
Da
vis,
2008b,c)
Journal
articles
TSS
(Thompson
et
al.,
2008)
Journal
article
Hydraulic
performance
(Da
vis,
2008)
Journal
article
Hydrologic
performance
(Emerson
and
Tra
ver,
2008)
Journal
article
Hydrologic
performance
21
1.2.3 Use and value of chronosequences
Succession and pedogenesis can be studied either by designing long-term experi-
ments in which a site is observed to undergo the successional changes of interest or by
using a space-for-time substitution (Pickett, 1989), in which multiple sites are selected
as representatives of the same ecosystem at various stages of development. These sites
form what is known as a chronosequence. Chronosequences are used where the process of
interest operates over a very long time period, where direct observation is impractical or
impossible.
Chronosequences have, for example, been used to study the recovery of ecosystems
in areas disturbed by mining. Abandoned open-pit mines and piles of mine spoils present
valuable opportunities to observe primary succession and the early stages of pedogenesis.
Leisman (1957) studied primary succession on spoil banks from iron mining in Minnesota
using a chronosequence. The spoil banks were composed of glacial till, and had been
deposited unamended and left to revegetate on their own. Banks were formed over time
as the mining operation proceeded, with the oldest bank 51 years old at the time of
sampling. The chronosequence consisted of banks at 10-year intervals in time of formation.
The plant community composition was surveyed, and soil pro les were studied down to a
15-inch depth. The depth of the A and Ao horizons were measured, and soil samples were
taken at multiple depths. These samples were analyzed for: bulk density, particle size
distribution, pH, organic carbon, and total Kjeldahl nitrogen. Young sites were found to
be dominated by herbaceous weed species, then, as the sites aged, this community tended
to be supplanted by a poplar woodland with a mixed herbaceous/grass understory. A
few sites, however, were never colonized by woody species, and retained their grassland
character. The chronosequence showed a uniform increase in the thickness of the Ao
22
horizon to 0.5 in after 51 years. Depth to the A horizon increased monotonically from zero
to 1-inch over the  rst 20 years, but thereafter increased more quickly on sites dominated
by herbaceous vegetation than on sites dominated by woody vegetation. No evidence
was found of B horizon development, which would have been evidenced by downward
migration of silt and clay particles. Organic matter content started out very low, and
gradually increased over time, from 0.08% at 2 years to 1.26% at 51 years at the 1-inch
depth. Total nitrogen also increased over time. The soil was very sandy (69% sand, on
average).
Frouz et al. (2001) compared pedogenesis and soil biota in two chronosequences of
coal mine spoils. One site was a orested with conifers, which produced little decomposable
litter, and one a orested with deciduous trees, which produced a much thicker and more
habitable litter layer. Both sites showed increasing organic matter over time. Both sites
also showed increasing density and species richness of invertebrates over time, though
there were di erences in patterns of development.
Kangas (1979) used a chronosequence to study succession on spoil mounds from
phosphate mining. The chronosequence consisted of eight sites, ranging from 1 to 50
years in age. Vegetation surveys were conducted at the sites. Soil organic matter was
measured, and ants were collected. Percent cover, tree height, litter weight, soil organic
matter, and ant nest density all increased with age.
Foster and Tilman (2000) studied succession in a chronosequence of 23 abandoned
 elds, then re-surveyed the plant communities at the sites 14 years later, in order to
test the validity of the chronosequence. They found that the chronosequence accurately
showed a deceleration of compositional change over time. Newly abandoned  elds were
dominated by fast-growing annuals, which were gradually displaced by slower-growing
23
perennials. The follow-up survey did not show the increase in species richness over time
suggested by the chronosequence, though the authors postulate that this may have been
due to weather-related perturbations rather than the inapplicability of the chronosequence
(implying that the chronosequence was actually more accurate than the long-term study
in measuring long-term successional trajectories).
It is important to be aware that the quality of the data generated using a chronose-
quence will only be as good as the sites which have been selected. Sites must be as similar
as possible. Ideally, time is the only di erence between sites. Practically, this is never the
case, so it is up to the researcher?s judgement to decide which di erences are signi cant
enough to in uence the site?s development. Pickett (1989) examines the use of chronose-
quences as an alternative to long-term ecological studies. He concludes that space-for-time
substitution is useful where general trends are sought. Space-for-time studies can reveal
averages, but may not re ect conditions at a particular site. Stevens and Walker (1970)
contend that all chronosequence studies fail in one way or another to control for all soil
forming factors apart from time, but assert that even  awed chronosequences can be useful
for making qualitative comparisons.
1.2.4 Pedogenesis in abandoned farmland and reclaimed mines
The e ects of ecological development on pedogenesis are most readily observed in the
regeneration of disturbed soils. Abandoned farmland and mine spoils are common places
where physically degraded soils are deserted and allowed to revert to natural habitat.
Scullion and Malik (2000) conducted a 9-year study of the in uence of earthworms on
soil development on coal mine spoils. They found that earthworm inputs increased the
creation of stable soil aggregates.
24
Hoogerkamp et al. (1983) inoculated abandoned pastures with earthworms, and
observed their e ect on the soil pro les over a decade. They found that after about three
years, earthworms had incorporated surface residues into the soil, and an A horizon had
begun to develop. After nine years, the A horizon had increased in thickness to 5-8 cm.
Roberts et al. (1988b,a) conducted experiments to compare pedogenesis on mine
spoils topped with topsoil to those where the soils were amended with organic material
(sawdust and sewage sludge). All plots were mulched and hydroseeded with tall fescue
(Festuca arundinacea). Soil samples were taken at various depths up to 30 cm every year
for three years. Soil pro les were analyzed to a depth of 1 m after 1 and 3 years. Within 1
year, A horizons were distinguishable in all plots. After three years, A horizons were more
developed, and AC and C1 horizons were observable. The authors note that soil structure
development was concentrated in root zones. A horizons were thicker in the topsoiled and
organically-amended plots, and thinner in the control plots (4 cm for control, 7 cm for
topsoil and sawdust, and 14 cm for sludge). Soil organic matter was observed to increase
over time in all plots, though the increase was signi cant only in the controls, due to the
higher background organic matter levels in the treatments.
Gonzalez-Sangregorio et al. (1991) studied pedogenesis in lignite mine spoils over
the  rst three years. The soils were plowed, limed, fertilized, hydroseeded with grasses
and legumes, and mulched at the beginning of the experiment. Organic carbon content
was observed to increase over time.
1.2.5 Earthworms and pedogenesis
Shaw and Pawluck (1986) demonstrated the ability of earthworms to produce well-
aggregated, granular soil structures from unstructured soils. Scullion and Malik (2000)
25
introduced earthworms to physically degraded coal mine spoils. They found that earth-
worms increased the formation of stable soil aggregates. Bouch e and Al-Addan (1997)
demonstrated a correlation between earthworm populations and soil in ltration rates.
Blanchart (1992) measured the ability of earthworms to aggregate sieved soils. He found
that M. anomala exerted a strong in uence on soil structure, forming 60% of the soil into
aggregates after 30 months.
The bene cial e ect of earthworms on soil permeability has been demonstrated
by a number of laboratory and  eld studies. Bouch e and Al-Addan (1997) conducted
 eld surveys at seventeen sites in France. The sites displayed a variety of soil types and
earthworm species. They found a correlation between in ltration rates and earthworm
biomass. Lachnicht et al. (1997) conducted a  eld experiment in which an agricultural
 eld was subdivided and subjected to three treatments: reduced earthworms, unaltered
earthworms, and supplemented earthworms. After several weeks, they found a signi cantly
greater number of macropores in the treatment with supplemental earthworms, but no
signi cant di erence in in ltration rates between the treatments.
Urb anek and Dole zal (1992) compared the extent of earthworm burrows to in ltra-
tion rates of agricultural soils. They found a positive relationship between the extent of
earthworm burrows and in ltration rate, and conclude that earthworm burrowing plays a
critical role in the maintenance of soil porosity in soils where subsurface drainage has been
employed. Joschko et al. (1992) conducted soil column experiments comparing the satu-
rated hydraulic conductivity of soil columns with and without earthworms. They found
that the columns with earthworms had higher in ltration rates than did the controls.
Microcosms have been used extensively to study the burrowing behavior of earth-
worms, and the e ect of earthworm burrowing on soil macroporosity and hydraulic con-
26
ductivity. Capowiez (Capowiez, 2000; Capowiez and Belzunces, 2001; Capowiez et al.,
2001) used microcosms to study the burrowing behavior of two earthworm species of dif-
ferent ecological types (anecic and endogeic). Gupta et al. (2001) used microcosms to
assess the e ect on macroporosity and potential for preferential  ow through the burrows
of three earthworm species.
Shaw and Pawluck (1986) conducted a microcosm study to assess the ability of
earthworms to create structure in di erent soil textures. After a year, the soil fabric of
the sandy loam control (no earthworms) consisted of individual mineral grains coated with
clays and organics. In the treatment containing endogeic species, the matrix at the soil
surface consisted of mineral grains coated with matrix material so that bridges formed
between the grains. In the treatments containing anecic species (L. terrestris), the sandy
loam soil showed few fecal deposits on the soil surface, in contrast to the other soil types
(silty clay loam and clay loam), which showed greater numbers of fecal deposits on the soil
surface. In the treatment containing both endogeic and anecic species, the surface soil of
the sandy loam consisted of units of densely packed matrix material fused into aggregates.
In other words, a more granular surface soil structure was present in the column containing
a diverse assembly of earthworm ecotypes. The authors hypothesize that this is due to
the synergistic e ect of the activities characteristic of the two ecotypes: anecic species
feed at the surface, adding organic matter to the soil matrix. Endogeic species then feed
within the soil, incorporating this organic matter subsidy into the soil matrix, increasing
the granularity of the soil structure.
Ponder et al. (2000) tested the ability of earthworms to reduce soil compaction in
microcosms. Using a loamy soil and the species Diplocardia ornata, they found that the
27
earthworms were able to signi cantly reduce the bulk density of compacted microcosms
after three months.
In a review of studies of earthworm populations on reclaimed lands, Curry and
Cotton (1983) found that A. caliginosa, A. chlorotica, and L. rubellus tend to be the
most successful early colonizers in European mine spoils. These are endogeic and epigeic
species, which inhabit the soil surface and upper layers of the soil, respectively. They tend
to have high reproductive rates. They found that anecic species, such as L. terrestris,
tend to take longer to colonize sites. This is due in part to their low reproductive rates.
1.3 Goals of This Study
The objective of this research is to investigate the importance of ecological pro-
cesses in the long-term performance of rain gardens. The research cataloged variations
between rain garden soils of di erent ages, with emphasis on soil ecosystem development.
The in uence of earthworms on in ltration rates was quanti ed, and the potential use of
earthworms to regenerate clogged soils was investigated. The knowledge gained by this
investigation will allow the development of more informed maintenance procedures for rain
gardens. This will allow the stormwater management community to better ensure reliable
water quality and quantity performance for the lifetime of a rain garden.
In order to develop new recommendations for rain garden maintenance, it is  rst
necessary to understand how rain garden soils will behave over time. Current engineering
design of rain gardens assumes that the soil mix installed will remain largely unchanged
for the lifetime of the cell. The only development anticipated is the accumulation of
a sediment layer on the soil surface, which may need to be removed periodically. The
principles of pedogenesis as understood by soil science suggest that these assumptions are
28
inadequate. Once installed, a rain garden should evolve in a manner similar to a degraded
soil in any other setting. That is, plants and soil animals will colonize the soil, and will
systematically change the soil structure. This research is intended to explore the rain
garden soil system. The research is divided into three components:
1. descriptive studies of pre-existing rain garden soil pro les,
2. laboratory experiments quantifying the e ect of earthworms on in ltration rates,
and
3. a simulation model describing the in uence of earthworms and soil organic matter
on soil porosity and surface crust thickness.
Field surveys of several di erent rain gardens of various ages provide the  rst detailed
descriptions of rain garden soil pro les. While di erences in design and life histories
between rain gardens will likely confound direct comparison, it is expected that general
patterns of rain garden pedogenesis will emerge from the data set.
A detailed investigation of the in uence of earthworms on soil in ltration rates was
be performed in the laboratory. Soil columns provided a controlled environment that
allowed the e ects of earthworms to be isolated from other factors which may also a ect
in ltration rates. The use of microcosms allowed multiple replicates to be constructed.
These microcosms were used to test the hypothesis that bioretention soils containing
earthworms have higher in ltration rates than bioretention soils lacking earthworms. This
provides a quantitative assessment of a direct impact of soil ecological development on
rain garden performance.
A simulation model of the action of earthworms on in ltration was developed. The
purpose of the model is to illustrate the physical processes taking place as a result of
earthworm activity. The model was developed using data from the  eld study and the
microcosm experiment. Emphasis was given to the potential role of crusting on the soil
surface and pore dynamics within the soil on the rain garden?s storage capacity.
29
This research is intended to provide a  rst glimpse into the biological processes
at work in rain garden soils. The three components, taken together, show that soil or-
ganisms are present in rain gardens, and explore their potential impact on bioretention
performance.
30
Chapter 2
Field Study
2.1 Introduction
Ten existing rain gardens of various ages were surveyed. The purpose of the surveys
was to assess the level of biological activity in the rain gardens, and to characterize their soil
pro les. The data collected include: earthworm quantity, species, and size, soil organic
matter, soil particle size distribution, root biomass, macroinvertebrate abundance and
species richness, and in ltration rate.
These data yield a chronosequence showing the evolution of biological activity and
soil pro les over time. While comparison between research sites is complicated by di er-
ences in design and history, these data provide a valuable glimpse into processes working
at a much larger time scale than are practical to study experimentally.
2.1.1 Background on  eld sites
Ten existing rain gardens were selected for assessment. They were selected to rep-
resent a wide range of rain garden ages and design styles, in order to get a sense of the
full spectrum of rain gardens currently in use. Rain gardens ranged in age from one year
to ten years. Figure 2.1 shows the locations of the research sites. All were located in the
Washington, DC metro area, most in Prince George?s County, Maryland. Greater detail
on the sites is included as Appendix A.
31
Figure
2.1:
Lo
cations
of
 eld
sites
(Quikmaps
-h
ttp
://ww
w.q
uikmaps.com).
32
2.1.1.1 University of Maryland (UMCP)
The UMCP site is located on the College Park campus of the University of Maryland,
at the end of parking lot PP1, on the south side of the Comcast Center. This rain garden
was constructed in 2004, and was sampled in the summer of 2005, making the rain garden
1 year old at the time of sampling. It is a component of a Low Impact Development (LID)
retro t of the Comcast Center parking areas, which included several rain gardens. The
cell is in the  oodplain of Campus Creek, and was overtopped during a storm event in
2004. The  ooding deposited a layer of  ne material on the cell surface, which caused the
death of many of the original plantings. At the time of sampling, the cell was sparsely
covered by herbaceous vegetation, as shown in Figure 2.2. Figure 2.3 shows the shape and
location of the site in context. The approximate sampling locations are indicated.
Figure 2.2: Photo of UMCP site.
33
Figure 2.3: UMCP site planview (source: UMCP Department of Facilities
Planning). The rain garden of interest is circled, and the sampling locations
(1,2, and 3) are indicated.
34
2.1.1.2 Washington Navy Yard (WNY)
The WNY site is located in the Washington Navy Yard in Washington, DC, in the
parking lot of Building 166, south of the O Street visitor center and east of the visitor
parking garage. This rain garden was constructed in 2002, and was sampled in the summer
of 2004, making the rain garden 2 years old at the time of sampling. It is one component
of an LID retro t of this parking lot, which included several rain gardens and a section of
permeable pavement. At the time of sampling, the cell was sparsely planted with small
shrubs (red chokeberry, Aronia arbutifolia), and the soil surface was covered with mulch,
as shown in Figure 2.4. The cell is lined with concrete on the bottom and sides, and
is equipped with an underdrain. Maintenance consists of annual removal of mulch and
trapped sediment from the soil surface. Figure 2.5 shows the shape and location of the
site in context. The approximate sampling locations are indicated.
Figure 2.4: Photo of WNY site.
35
Figure 2.5: WNY site planview (source: The Low Impact Development Center,
Inc.). The rain garden of interest is circled, and the sampling locations (1,2,
and 3) are indicated.
36
2.1.1.3 Mary Harris \Mother" Jones Elementary School (MJES)
The MJES site is located at Mary Harris \Mother" Jones Elementary School in
Adelphi, Maryland. The rain garden sampled is located directly in front of the school,
between the  rst and second rows of parking spaces. The rain garden was constructed in
2002, and was sampled in the summer of 2005, making the rain garden 3 years old at the
time of sampling. At the time of sampling, the cell was densely vegetated with a wide
variety of herbaceous plants, shrubs, and small trees, as shown in Figure 2.6. The soil
was covered with a layer of mulch. Figure 2.7 shows the shape and location of the site in
context. The approximate sampling locations are indicated.
Figure 2.6: Photo of MJES site.
37
Figure 2.7: MJES site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
38
2.1.1.4 Chesapeake Bay Foundation Headquarters (CBF)
The CBF site is located at the Philip Merrill Environmental Center in Annapolis,
MD, in the parking lot on the north side of the building entrance. The rain garden was
constructed in 2001 and was sampled in the summer of 2005, making the rain garden 4
years old at the time of sampling. The surrounding parking lot is paved with gravel rather
than asphalt. At the time of sampling, the cell was densely populated with small trees,
shrubs, and herbaceous vegetation, as shown in Figure 2.8. The soil was covered with a
thick litter layer. Figure 2.9 shows the shape and location of the site in context. The
approximate sampling locations are indicated.
Figure 2.8: Photo of CBF site.
39
Figure 2.9: CBF site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
40
2.1.1.5 Northwestern High School (NWHS)
The NWHS site is located at Northwestern High School in Hyattsville, MD, directly
in front of the school, between the driveway and the  rst row of parking spaces. This rain
garden was constructed in 1999, and was sampled in the summer of 2004, making the rain
garden 5 years old at the time of sampling. It is a component of an LID retro t of the
entire parking lot, which included several rain gardens. At the time of sampling, the cell
was densely vegetated with a wide variety of herbaceous plants, shrubs, and small trees,
as shown in Figure 2.10. The soil was covered with a thin litter layer. Figure 2.11 shows
the shape and location of the site in context. The approximate sampling locations are
indicated.
Figure 2.10: Photo of NWHS site.
41
Figure 2.11: NWHS site planview (source: Prince George?s County Public
Schools). The rain garden of interest is circled, and the sampling locations
(1,2, and 3) are indicated.
42
2.1.1.6 Inglewood Center III (PP)
The PP site is located at Inglewood Center III, 9400 Peppercorn Place, Upper
Marlboro, MD. The rain garden sampled is located on the north side of the building,
between the driveway and the parking lot. The rain garden was constructed in 1999, and
was sampled in the summer of 2004, making it 5 years old at the time of sampling. At
the time of sampling, the cell was densely populated with herbaceous vegetation, with the
exception of a bare area directly in front of the over ow inlet, as shown in Figure 2.12.
The soil surface was covered with a thick mulch layer. Figure 2.13 shows the shape and
location of the site in context. The approximate sampling locations are indicated.
Figure 2.12: Photo of PP site.
43
Figure 2.13: PP site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
44
2.1.1.7 Claggett Farm (CF)
The CF site is located on Claggett Farm in Upper Marlboro, MD. This rain garden
was constructed in 1999, and was sampled in the summer of 2005, making the rain garden
6 years old at the time of sampling. The rain garden at this site is unique in several
respects. It is located alongside a barn, whose roof constitutes the entire drainage area
of the rain garden. In addition, this rain garden was constructed using only unamended,
in-situ soil. The soil was aerated to a depth of one foot, then bordered by wooden planks.
At the time of sampling, the rain garden was dominated by a dense stand of goldenrod
(Solidago spp.) and an unidenti ed shrub, as shown in Figure 2.14. Figure 2.15 shows
the shape and location of the site in context. The approximate sampling locations are
indicated.
Figure 2.14: Photo of CF site.
45
Figure 2.15: CF site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
46
2.1.1.8 Chevy Chase Bank (CC)
The CC site is located at the Glenmont Branch of Chevy Chase Bank, on Randolph
Road in Silver Spring, MD. The rain garden sampled is located between the parking lot
and the bank drive-thru on the west side of the bank building. The rain garden was
constructed in 1998, and was sampled in the summer of 2005, making the rain garden
7 years old at the time of sampling. At the time of sampling, the cell was very densely
populated with ornamental grasses and shrubs, and the soil was covered with a thick
mulch layer, as shown in Figure 2.16. Figure 2.17 shows the shape and location of the site
in context. The approximate sampling locations are indicated.
Figure 2.16: Photo of CC site.
47
Figure 2.17: CC site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
48
2.1.1.9 Beltway Plaza Mall (BP)
The BP site is located behind the Beltway Plaza Mall in Greenbelt, MD. This rain
garden was constructed in 1997, and was sampled in the summer of 2004, making the rain
garden 7 years old at the time of sampling. A series of rain garden retro ts have been
constructed in this parking lot over the years, starting in 1993. The cell is planted with
small trees (mainly Acer rubrum), and is densely colonized by a variety of herbaceous
plants, as shown in Figure 2.18. The rain garden is mowed every fall, and then the plants
are allowed to regenerate over the following year. Figure 2.19 shows the shape and location
of the site in context. The approximate sampling locations are indicated.
Figure 2.18: Photo of BP site.
49
Figure 2.19: BP site planview (source: Beltway Plaza Developers). The rain
garden of interest is circled, and the sampling locations (1,2, and 3) are indi-
cated.
50
2.1.1.10 Laurel Regional Hospital (LRH)
The LRH site is located at Laurel Regional Hospital in Laurel, MD, to the north
of the Emergency entrance, between a service road and a small parking lot. It was con-
structed in 1994, and was sampled in the summer of 2004, making the rain garden 10
years old at the time of sampling. It is one of several rain gardens on the site, and one
of the oldest rain gardens in existence. At the time of sampling, the cell contained a
stand of well-established trees, with an understory dominated by large shrubs, as shown
in Figure 2.20. The soil surface was covered by a thin litter layer beneath the vegetation.
Figure 2.21 shows the shape and location of the site in context. The approximate sampling
locations are indicated.
Figure 2.20: Photo of LRH site.
51
Figure 2.21: LRH site planview. The rain garden of interest is circled, and the
sampling locations (1,2, and 3) are indicated.
52
2.2 Methodology
2.2.1 Field sampling
The sites were sampled during the summers of 2004{2006. Each site was sampled
once, at three locations. The sampling locations were spaced evenly along a transect
spanning the length of the rain garden. At each location, the litter layer was removed, and
plants were clipped at the soil surface. A 20 cm x 30 cm hole was dug, and the  rst 10 cm of
soil was removed. This material was hand sorted. Plant roots were collected. Earthworm
lengths were recorded. Representatives of di erent earthworm species were then sedated
in a weak ethanol solution, and preserved in 4% formalin solution for later identi cation.
Other macroscopic invertebrates were tallied and returned to the rain garden. Soil animals
were classi ed into easily identi able taxa (see Table 2.1). The volume of the removed
soil was recorded. Approximately 1 liter of soil was removed for laboratory analysis. The
procedure was repeated for the soil from 10{20 cm depth, and from 20{30 cm depth. The
major features of the soil pro le were recorded, including thicknesses, textures and colors
of identi able layers. Soil colors were described using a Munsell soil color chart. Texture
was estimated by feel. The residual soil was then returned to the pit.
At a later date, the in ltration rate of some of the rain gardens was measured using
a double ring in ltrometer (ASTM, 2003). The in ltration rate was measured at two to
three locations spaced evenly along the same transect used in the soil sampling. A level
site was selected that had not been disturbed by the initial soil sampling, and the soil
surface was cleared of mulch and plant litter. Plants were clipped at the soil surface, but
the roots were not disturbed. Two concentric rings, of 12 in and 24 in diameters, were
driven into the soil using a brace and a sledgehammer. When the progress of the rings
53
Table 2.1: Common and scienti c names of animal taxa.
COMMON NAME SCIENTIFIC NAME
Earthworms Oligochaeta: Lumbricidae
Beetles Coleoptera
Common white grubs Coleoptera: Scarabaeidae larvae
Snts Hymenoptera: Formicidae
Centipedes Myriapoda: Chilopoda
Millipedes Myriapoda: Diplopoda
Potworms Oligochaeta: Enchytraeidae
Pill bugs Crustacea: Isopoda
Spiders Arachnida: Araneae
Springtails Hexapoda: Collembola
Slugs and snails Mollusca: Gastropoda
Fly larvae Hexapoda: Diptera larvae
Mites Arachnida: Acari
Other insect larvae Insecta
was impeded by large roots, a machete was inserted into the ground alongside the rings to
sever the roots. Great care was taken to minimize disturbance of the soil. The outer ring
was driven to a depth of 6 inches, and the inner ring was driven to a depth of 2{4 inches.
The inner ring and annular space were  lled with water to a depth of 3 inches. Water
was added to both spaces as needed in order to maintain a constant head. The volume of
water added to the inner ring was measured at regular intervals. Measurements were taken
until the volume of water added per unit time reached equilibrium. The experiment was
continued for one hour at equilibrium. The incremental in ltration velocity was calculated
as follows:
I =  VA t (2.1)
where:
I = inner ring incremental in ltration velocity (cm/h)
 V = volume of liquid used to maintain constant head in inner ring (cm3)
54
A = internal area of inner ring (cm2), and
 t = timer interval (h).
The  nal in ltration rate of the soil was calculated by taking the average of the
incremental in ltration velocity of the  nal three measurements.
2.2.2 Laboratory analysis
Plant roots were washed, and live roots were separated from dead. Live roots were
characterized by  exibility, light-colored interior, with no signs of decay. Roots considered
dead were dark throughout, brittle, or decayed. The live roots were dried overnight in a
70 C oven, then weighed. To estimate the mass of very  ne roots, all visible root fragments
were removed from a 100 mL soil subsample. The roots were washed, oven dried at 70 C,
and weighed. The weight of this subsample was then multiplied by the measured volume of
removed soil to give the total mass of  ne roots. Soil organic matter, SOM, was measured
using the loss-on-ignition method (ASTM, 2000). The soil sample was oven dried at 70 C.
A 6 g subsample was crushed and sieved using a #10 (2mm) mesh sieve. The subsample
was weighed, then ashed at 400 C for four hours. The ashed sample was weighed, and the
SOM was calculated using the formula:
% organic matter = mass (oven-dried) mass (ashed)mass (oven-dried)  100% (2.2)
Particle size distribution was measured using the hydrometer method and dry siev-
ing. Samples with more than 10% organic matter were  rst oxidized using a 6% hydrogen
peroxide solution. Details of the procedures followed are included in Appendix B. The
resulting particle size distribution curve was used to calculate the coe cient of uniformity
55
(d60/d10), where d60 and d10 are the soil particle diameters for which 60% and 10% of the
mass of a soil sample is  ner, respectively.
A taxonomic guide to earthworms commonly found in Maryland (Csuzdi and Sl avecz,
2003; Reynolds, 1974) was developed by compiling descriptions from a number of sources
(Baker and Barrett, 1994; Eaton, 1942; Fender and McKey-Fender, 1990; Gates, 1937,
1958, 1972a,b, 1973, 1974; James, 1990; Lee, 1959; Reynolds et al., 1974; Schwert, 1990;
Sims et al., 1999; Worm Watch Canada, 2002). This key is included as Appendix C. The
preserved earthworm samples were examined using a dissecting microscope. Earthworms
were identi ed as completely as possible by external features using the key. The extent to
which the earthworm samples were identi able depended upon their maturity, and on the
presence or absence of unique identifying features. Non-clitellate (immature) specimens
lack many identifying features, and can rarely be identi ed, even to genus.
2.3 Results
The data presented here are average values for the three locations sampled at each
of the sites. Results are reported as mean  standard deviation. Complete data for all
samples taken at each of the sites are included as Appendix D.
2.3.1 University of Maryland (UMCP)
A summary of the data collected at UMCP is presented in Table 2.2. An average
of 4.3  2.5 earthworms were collected from the uppermost soil layer. No earthworms
were found below 10 cm. Mean earthworm size was 2.6  0.7 cm. Root biomass was
highest in the upper and lower soil layers (0.71  0.20 g and 0.73  0.69 g, respectively,
averaging 1.74 0.89 g for the entire depth). Soil organic matter was highest at the surface
56
(2.91  0.33%), decreasing with depth to 1.44  0.03% at the 20{30 cm depth. The soil
texture was sand throughout, averaging 91.0 1.2% sand, 4.6 0.9% silt, and 3.8 0.3%
clay. The particle size distribution is presented graphically in Figure 2.22. A coe cient
of uniformity of 10 was measured throughout the site. A total of 5 animal taxa were
collected, all from the uppermost soil layer. The composition of the animal community
is presented in Figure 2.23. The UMCP animal community is dominated by earthworms
(Lumbricidae), followed by beetles (Coleoptera). Individual grubs, ants, millipedes and
spiders were also collected. Nine of the collected earthworms were identi ed as Diplocardia
singularis (see Appendix D). The distribution of major soil animal groups with depth is
shown in Figure 2.24. The major groups were found only in the uppermost soil layers.
Table 2.2: Summary of data collected at UMCP, averaged.  indicates a parameter for
which values are summed over the sampling depth rather than averaged. yLayers where
no earthworms were found are not included in the calculation of the average earthworm
size.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 4.3  2.5 0 0 4.3  2.5 
Earthworm Size (cm) 2.6  0.7 0 0 2.6  0.7y
Root Biomass (g) 0.71  0.20 0.30  0.15 0.73  0.69 1.74  0.89 
Soil Organic Matter (%) 2.91  0.33 1.53  0.05 1.44  0.03 2.05  0.25
Soil Textural
Classi cation
Sand Sand Sand Sand
Sand Content (%) 89.4  2.4 92.6  1.7 92.0  0.7 91.0  1.2
Silt Content (%) 6.1  2.9 4.6  0.8 4.4  0.9 4.6  0.9
Clay Content (%) 4.5  0.6 2.9  1.0 3.6  0.3 3.8  0.3
d60 (mm) 0.5 1 1 n/a
d10 (mm) 0.05 0.1 0.1 n/a
Coe cient of Uniformity
(d60/d10)
10 10 10 n/a
Soil Animals (#s,
includes earthworms)
7  4 0 0 7  4  
Total Number of Taxa
(not averaged)
5 0 0 5 
57
Figure 2.22: UMCP particle size distribution.
Figure 2.23: UMCP animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
58
Figure 2.24: UMCP animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
59
2.3.2 Washington Navy Yard (WNY)
A summary of the data collected at WNY is presented in Table 2.3. An average of
4.0 5.3 earthworms were collected from the uppermost soil layer. 1.7 2.1 earthworms
were collected from the 10{20 cm depth. No earthworms were found below 10 cm. Mean
earthworm size was 5.0  0.8 cm. Root biomass was lowest in the uppermost layer (2.74
 2.23 g), increasing in the middle and lower layers (3.50  3.42 g and 3.49  3.78 g,
respectively). The total root biomass over the entire sampling depth was 9.73 9.18 g. Soil
organic matter was highest at the surface (6.28  1.82%), decreasing with depth to 2.05
 0.19% at the 20{30 cm depth. The soil texture was sandy loam throughout, averaging
71.3  3.4% sand, 21.3  5.7% silt, and 7.4  2.8% clay. The particle size distribution
is presented graphically in Figure 2.25. A coe cient of uniformity of 10 was measured at
the 0{10 and 10{20 cm depths, and a coe cient of uniformity of 100 was measured at the
20{30 cm depth. A total of 2 animal taxa were collected. Animals were found at the upper
and middle soil layers. The composition of the animal community is presented in Figure
2.26. The WNY animal community is dominated by grubs (Coleoptera: Scarabaeidae
larvae), followed by earthworms (Lumbricidae). No other animals were found. None of
the collected earthworms were identi ed (see Appendix D). The distribution of major soil
animal groups with depth is shown in Figure 2.27. The major groups were found primarily
in the upper soil layer, and at lower densities in the middle layer. None were found in the
lower layer.
60
Table 2.3: Summary of data collected at WNY, averaged.  indicates a parameter for
which values are summed over the sampling depth rather than averaged. yLayers where
no earthworms were found are not included in the calculation of the average earthworm
size.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 4.0  5.3 1.7  2.1 0 5.7  7.4 
Earthworm Size (cm) 5.0  0.8 0 0 5.0  0.8y
Root Biomass (g) 2.74  2.23 3.50  3.42 3.49  3.78 9.73  9.18 
Soil Organic Matter (%) 6.28  1.82 2.34  0.16 2.05  0.19 3.56  0.62
Soil Textural
Classi cation
Sandy Loam Sandy Loam Sandy Loam Sandy Loam
Sand Content (%) 71.6  3.5 72.0  3.9 70.2  4.5 71.3  3.4
Silt Content (%) 23.0  2.9 21.2  5.0 19.8  11.1 21.3  5.7
Clay Content (%) 5.3  1.3 6.8  1.4 10.0  6.8 7.4  2.8
d60 (mm) 0.5 0.5 0.5 n/a
d10 (mm) 0.05 0.05 0.005 n/a
Coe cient of Uniformity
(d60/d10)
10 10 100 n/a
Soil Animals (#s,
includes earthworms)
11 4 2  2 0 14  5 
Total Number of Taxa
(not averaged)
2 2 0 2 
Figure 2.25: WNY particle size distribution.
61
Figure 2.26: WNY animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
Figure 2.27: WNY animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
62
2.3.3 \Mother" Jones Elementary School (MJES)
A summary of the data collected at MJES is presented in Table 2.4. An average of
45.2 65.9 earthworms were collected from the uppermost soil layer. 2.5 4.3 earthworms
were collected from the 10{20 cm depth. Only 0.7 1.2 earthworms were found below 20
cm. A very large number of small earthworms (121) were found in the upper soil layer at
one sampling location. Mean earthworm size was 3.1  1.1 cm. Root biomass was 1.62
 0.74 g in the upper and middle layers, increasing slightly to 2.08  2.97 g in the lower
layer. Total root biomass for the entire soil depth was 5.33  4.18 g. Soil organic matter
was highest at the surface (17.53  14.15%), decreasing with depth to 2.09  1.60% at
the 20{30 cm depth. The average soil texture was sandy loam at all depths, averaging
65.8  21.3% sand, 19.7  12.4% silt, and 14.5  8.9% clay. The particle size distribu-
tion is presented graphically in Figure 2.28, and shows considerable variability between
sampling locations (see Appendix D). A coe cient of uniformity of 50 was measured at
all depths. A total of 12 animal taxa were collected. Animals were found primarily in the
upper soil layer, but also occurred in the middle and lower soil layers. The composition
of the animal community is presented in Figure 2.29. The MJES animal community is
dominated by grubs (Coleoptera: Scarabaeidae larvae), followed by earthworms (Lumbri-
cidae). Potworms (Enchytraeidae) and springtails (Collembola) were found in moderate
numbers. Grubs, other beetle larvae, ants, centipedes, millipedes, spiders, slugs and snails
were collected in much lower numbers. Representatives of Allolobophora chlorotica, Bi-
mastos parvus, Pheretima spp., and Lumbricus spp. were identi ed among the collected
earthworms (see Appendix D). The distribution of major soil animal groups with depth
is shown in Figure 2.30. The major groups were found primarily in the upper soil layer.
Animal densities decreased with depth.
63
Table 2.4: Summary of data collected at MJES, averaged.  indicates a parameter for
which values are summed over the sampling depth rather than averaged. yLayers where
no earthworms were found are not included in the calculation of the average earthworm
size.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 45.2  65.9 2.5  4.3 0.7  1.2 48.3  71.4 
Earthworm Size (cm) 4.3  2.1 2.1 3.0 3.1  1.1 y
Root Biomass (g) 1.62  0.74 1.62  1.40 2.08  2.97 5.33  4.18 
Soil Organic Matter (%) 17.53  14.15 2.31  1.70 2.09  1.60 7.31  5.75
Soil Textural
Classi cation
Sandy Loam Sandy Loam Sandy Loam Sandy Loam
Sand Content (%) 57.2  18.9 71.3  21.7 68.8  23.8 65.8  21.3
Silt Content (%) 28.8  14.3 14.5  11.5 15.8  12.6 19.7  12.4
Clay Content (%) 14.0  5.3 14.2  10.3 15.4  11.5 14.5  8.9
d60 (mm) 0.25 0.25 0.25 n/a
d10 (mm) 0.005 0.005 0.005 n/a
Coe cient of
Uniformity
(d60/d10)
50 50 50 n/a
Soil Animals (#s,
includes earthworms)
73  84 4  6 1  1 78  90 
Total Number of Taxa
(not averaged)
11 4 1 12 
Figure 2.28: MJES particle size distribution.
64
Figure 2.29: MJES animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
Figure 2.30: MJES animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
65
2.3.4 Chesapeake Bay Foundation Headquarters (CBF)
A summary of the data collected at CBF is presented in Table 2.5. An average of
14.0 8.5 earthworms were collected from the uppermost soil layer. 1.3 1.2 earthworms
were collected from the 10{20 cm depth. Only 0.3 0.6 earthworms were found below 20
cm. Mean earthworm size was 2.7  1.9 cm. Root biomass was highest in the uppermost
soil layer (10.10 7.18 g), decreasing sharply in the middle and lower layers to 1.86 0.49
g and 0.89 0.61 g, respectively. Soil organic matter was very low throughout, and similar
at all depths (upper: 2.68  0.66%, middle: 2.69  0.33%, lower: 2.83  0.17%). The
average soil texture was sandy clay loam at all depths, averaging 61.0  0.9% sand, 16.5
 0.6% silt, and 22.5  1.3% clay. The particle size distribution is presented graphically
in Figure 2.31. A coe cient of uniformity of 50 was measured at all depths. A total of 11
animal taxa were collected. Animals were found primarily in the upper soil layer, but also
occurred in the middle and lower soil layers. The composition of the animal community
is presented in Figure 2.32. The CBF animal community is dominated by earthworms
(Lumbricidae) and pill bugs (Isopoda). Beetles (Coleoptera) were found in moderate
numbers. Beetle larvae, ants, centipedes, millipedes, potworms, spiders, and springtails
were collected in much lower numbers. The majority of the earthworms collected at
CBF were identi ed as Apporectodea caliginosa (see Appendix D). One representative of
Pheretima di ringens was also identi ed. The distribution of major soil animal groups
with depth is shown in Figure 2.33. Lumbricid density declined with depth. Isopods
and Coleoptera occurred primarily in the uppermost soil layer, though one Coleoptera
individual was found in the middle soil layer.
66
Table 2.5: Summary of data collected at CBF, averaged.  indicates a parameter for which
values are summed over the sampling depth rather than averaged. y Layers where no
earthworms were found are not included in the calculation of the average earthworm size.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 14.0  8.5 1.3  1.2 0.3  0.6 15.7  9.7 
Earthworm Size (cm) 2.3  0.3 4.8  3.9 1.0 2.7  1.9y
Root Biomass (g) 10.10 7.18 1.86  0.49 0.89  0.61 12.85  8.18 
Soil Organic Matter (%) 2.68  0.66 2.69  0.33 2.83  0.17 2.74  0.36
Soil Textural
Classi cation
Sandy Clay
Loam
Sandy Clay
Loam
Sandy Clay
Loam
Sandy Clay
Loam
Sand Content (%) 61.0  1.4 60.8  1.0 61.3  0.6 61.0  0.9
Silt Content (%) 15.1  1.6 17.1  0.6 17.2  1.0 16.5  0.6
Clay Content (%) 23.9  2.0 22.1  1.2 21.5  1.5 22.5  1.3
d60 (mm) 0.25 0.25 0.25 na
d10 (mm) 0.005 0.005 0.005 na
Coe cient of Uniformity
(d60/d10)
50 50 50 na
Soil Animals
(#s, includes
earthworms)
44  18 2  2 0  1 46  16 
Total Number of Taxa
(not averaged)
10 3 1 11 
Figure 2.31: CBF particle size distribution.
67
Figure 2.32: CBF animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
Figure 2.33: CBF animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
68
2.3.5 Northwestern High School (NWHS)
A summary of the data collected at NWHS is presented in Table 2.6. An average
of 18.0  27.8 earthworms were collected from the uppermost soil layer. 9.3  16.2
earthworms were collected from the 10{20 cm depth. 1.3  2.3 earthworms were found
below 20 cm. Mean earthworm size was 2.3  0.6 cm. Root biomass was highest in the
uppermost soil layer (7.58  5.41 g), decreasing sharply to 1.43  0.12 g in the middle
layer, and 1.22  0.59 g in the lower layer. Total root biomass for the entire sampling
depth was 10.22  5.09 g. Soil organic matter was highest in the upper layer (4.39  
1.11%), decreasing to 2.81 0.69% in the lower layer. Soil texture showed some variability
with depth, with the middle layer exhibiting the  nest texture. When averaged over the
entire depth, the soil had loamy texture, averaging 51.4  4.0% sand, 34.2  1.0% silt,
and 14.4  3.7% clay. The particle size distribution is presented graphically in Figure
2.34. A coe cient of uniformity of 50 was measured in the upper and lower layers, and a
coe cient of uniformity of 20 was measured in the middle layer. A total of 11 animal taxa
were collected. Animals were found primarily in the upper and middle soil layers, but also
occurred in the lower layer. The composition of the animal community is presented in
Figure 2.35. The NWHS animal community is dominated by earthworms (Lumbricidae).
Potworms (Enchytraeidae) were found in moderate numbers. Grubs, other beetle larvae,
ants, centipedes, millipedes, pill bugs, beetles, spiders, and springtails were collected in
much lower numbers. None of the collected earthworms could be identi ed (see Appendix
D). The distribution of major soil animal groups with depth is shown in Figure 2.36. The
densities of both Lumbricidae and Enchytraeidae were highest in the uppermost soil layer
and decreased with depth.
69
Table 2.6: Summary of data collected at NWHS, averaged.  indicates a parameter for
which values are summed over the sampling depth rather than averaged. yLayers where
no earthworms were found are not included in the calculation of the average earthworm
size.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 18.0  27.8 9.3  16.2 1.3  2.3 28.7  46.2 
Earthworm Size (cm) 2.9  1.2 1.9 2.0 2.3  0.6 y
Root Biomass (g) 7.58  5.41 1.43  0.12 1.22  0.59 10.22  5.09 
Soil Organic Matter (%) 4.39  1.11 3.59  0.41 2.81  0.69 3.60  0.25
Soil Textural
Classi cation
Sandy Loam Loam Sandy Loam Loam
Sand Content (%) 53.1  4.0 43.4  7.5 57.8  9.3 51.4  4.0
Silt Content (%) 33.0  1.5 40.0  8.1 29.6  5.1 34.2  1.0
Clay Content (%) 14.0  3.3 16.6  3.2 12.6  4.6 14.4  3.7
d60 (mm) 0.25 0.1 0.25 n/a
d10 (mm) 0.005 0.005 0.005 n/a
Coe cient of Uniformity
(d60/d10)
50 20 50 n/a
Soil Animals
(#s, includes
earthworms)
40  37 22  16 12  16 74  46 
Total Number of Taxa
(not averaged)
9 7 2 11 
Figure 2.34: NWHS particle size distribution.
70
Figure 2.35: NWHS animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
Figure 2.36: NWHS animal depth pro le, showing the variation in density
with depth of the most abundant macroinvertebrate taxa.
71
2.3.6 Inglewood Center III (PP)
A summary of the data collected at PP is presented in Table 2.7. An average of 4.2
 2.5 earthworms were collected from the uppermost soil layer. 0.7 0.6 earthworms were
collected from the 10{20 cm depth. 0.3 0.6 earthworms were found below 20 cm. Mean
earthworm size was 5.9 2.8 cm, increasing from 3.5 0.5 cm in the uppermost layer to
5.9  2.8 cm in the lower layer. Root biomass was highest in the upper soil layer (1.92
 1.30 g), decreasing to 1.02  0.53 g in the middle layer and 0.64  0.17 g in the lower
layer. The total root biomass for the entire sampling depth was 3.58 1.78 g. Soil organic
matter was very high in the upper layer (17.90  9.31%), decreasing to 3.75  1.69% in
the lower layer. Soil texture became sandier with depth, ranging from sandy loam in the
upper layer to sand in the lower layer. When averaged over the entire depth, the soil had
82.3  3.0% sand, 8.1  2.0% silt, and 9.6  1.6% clay. The particle size distribution
is presented graphically in Figure 2.37. A coe cient of uniformity of 50 was measured
in the upper layer, and a coe cient of uniformity of 10 was measured in the middle and
lower layers. A total of 9 animal taxa were collected. Animals were found primarily in the
upper soil layers, but also occurred in the middle and lower layers. The composition of the
animal community is presented in Figure 2.38. The PP animal community is dominated
by grubs (Scarabaeidae larvae). Earthworms (Lumbricidae) were found in moderate num-
bers. Other beetle larvae, ants, pill bugs, potworms, beetles, spiders, and springtails were
collected in much lower numbers. Representatives of Dendrodrilus rubidus, Apporectodea
caliginosa and Octolasion tyrtaeum were identi ed among the collected earthworms (see
Appendix D). The distribution of major soil animal groups with depth is shown in Fig-
ure 2.39. The densities of both Lumbricidae and Scarabaeidae larvae were highest in the
uppermost soil layer and decreased with depth.
72
Table 2.7: Summary of data collected at PP, averaged.  indicates a parameter for which
values are summed over the sampling depth rather than averaged. yLayers where no
earthworms were found are not included in the average.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 4.2  2.5 0.7  0.6 0.3  0.6 5.2  2.8 
Earthworm Size (cm) 3.5  0.5 5.3  2.5 9.0 5.9  2.8 y
Root Biomass (g) 1.92  1.30 1.02  0.53 0.64  0.17 3.58  1.78 
Soil Organic Matter (%) 17.90  9.31 8.27  4.69 3.75  1.69 9.98  5.16
Soil Textural
Classi cation
Sandy Loam Loamy Sand Sand Loamy Sand
Sand Content (%) 70.5  8.1 86.6  2.4 89.7  2.3 82.3  3.0
Silt Content (%) 13.0  4.6 6.8  4.4 4.4  3.0 8.1  2.0
Clay Content (%) 16.5  3.6 6.6  2.0 5.9  2.4 9.6  1.6
d60 (mm) 0.25 0.5 0.5 n/a
d10 (mm) 0.005 0.05 0.05 n/a
Coe cient of Uniformity
(d60/d10)
50 10 10 n/a
Soil Animals (#s,
including earthworms)
19  10 2  1 0  1 22  9 
Total Number of Taxa
(not averaged)
9 3 1 9 
Figure 2.37: PP particle size distribution.
73
Figure 2.38: PP animal diversity. Total number of individuals of major inver-
tebrate taxa found over entire site.
Figure 2.39: PP animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
74
2.3.7 Claggett Farm (CF)
A summary of the data collected at CF is presented in Table 2.8. No earthworms
were collected from the uppermost soil layer. 3.7  4.6 earthworms were collected from
the 10{20 cm depth. 1.7  2.9 earthworms were found below 20 cm. Sampling was
performed in late summer, after a prolonged dry period, and all earthworms collected
were aestivating. Mean earthworm size was 4.6 cm. Root biomass was highest in the
uppermost layer (21.97  12.97 g), decreasing sharply to 8.47  3.12 g in the middle
layer, and 2.25  1.30 g in the lower layer. Total root biomass for the entire sampling
depth was 32.69  16.68 g. Soil organic matter was highest in the upper layer (5.59  
0.28%), decreasing to 2.49  0.39% in the lower layer. Soil texture was sandy loam at all
depths, averaging 66.9 1.6% sand, 17.8 0.6% silt, and 15.4 1.0% clay. The particle
size distribution is presented graphically in Figure 2.40. A coe cient of uniformity of 50
was measured in the upper and middle layers, and a coe cient of uniformity of 20 was
measured in the lower layers. A total of 10 animal taxa were collected. Animals were
found primarily in the upper soil layers, but also occurred in the middle and lower layers.
The composition of the animal community is presented in Figure 2.41. The CF animal
community is dominated by ants (Formicidae). Earthworms (Lumbricidae) and beetles
(Coleoptera) were found in moderate numbers. Grubs, other beetle larvae, centipedes,
millipedes, pill bugs, spiders, and springtails were collected in much lower numbers. Four
Allolobophora chlorotica were identi ed among the collected earthworms (see Appendix
D). The distribution of major soil animal groups with depth is shown in Figure 2.42. A
large number of ants (Formicidae) were collected, with the highest density occurring in
the upper soil layer, and with decreasing density with depth. Coleoptera were found only
in the uppermost soil layer. Interestingly, earthworms (Lumbricidae) were found only in
75
the middle and lower soil layers. No earthworms were found in the upper soil layer. This
is consistent with the dry conditions encountered at this site, as earthworms tend to move
deeper into the soil to avoid dessication.
Table 2.8: Summary of data collected at CF, averaged.  indicates a parameter for which
values are summed over the sampling depth rather than averaged. yLayers where no
earthworms were found are not included in the average.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 0 3.7  4.6 1.7  2.9 5.3  4.0 
Earthworm Size (cm) 0 4.6  1.5 4.6  0 4.6  0
Root Biomass (g) 21.97  12.97 8.47  3.12 2.25  1.30 32.69  16.68 
Soil Organic Matter
(%)
5.59  0.28 3.75  0.33 2.49  0.39 3.94  0.33
Soil Textural
Classi cation
Sandy Loam Sandy Loam Sandy Loam Sandy Loam
Sand Content (%) 70.0  2.5 67.6  2.6 63.0  1.9 66.9  1.6
Silt Content (%) 17.5  0.8 18.4  1.5 17.4  0.6 17.8  0.6
Clay Content (%) 12.5  1.9 14.0  2.1 19.6  2.2 15.4  1.0
d60 (mm) 0.25 0.25 0.1 n/a
d10 (mm) 0.005 0.005 0.005 n/a
Coe cient of
Uniformity
(d60/d10)
50 50 20 n/a
Soil Animals (#s,
includes earthworms)
34  12 22  16 14  16 70  24 
Total Number of Taxa
(not averaged)
6 7 2 10 
76
Figure 2.40: CF particle size distribution.
Figure 2.41: CF animal diversity. Total number of individuals of major inver-
tebrate taxa found over entire site.
77
Figure 2.42: CF animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
78
2.3.8 Chevy Chase Bank (CC)
A summary of the data collected at CC is presented in Table 2.9. An average of
10.2 9.2 earthworms were collected from the uppermost soil layer, 10.0 10.4 from the
10{20 cm depth, and 7.7 8.0 below 20 cm. Mean earthworm size was 2.7 0.7 cm. Root
biomass was highest in the uppermost soil layer (26.04  14.65 g), decreasing to 19.30  
18.30 g in the middle layer and 11.15  15.32 g in the lower layer. Soil organic matter
was very high in the upper layer (30.13  11.42%), decreasing to 10.06  11.02% in the
lower layer. Soil texture was sandy loam at all depths, averaging 66.1 25.2% sand, 25.3
 16.8% silt, and 8.6  8.4% clay. The particle size distribution is presented graphically
in Figure 2.43. A coe cient of uniformity of 50 was measured in all layers. A total of 12
animal taxa were collected. Animals were found primarily in the upper and middle soil
layers, but also occurred in the lower layer. The composition of the animal community
is presented in Figure 2.44. The CC animal community is dominated by earthworms
(Lumbricidae). Ants (Formicidae) and centipedes (Chilopoda) were found in moderate
numbers. Grubs, other beetle larvae, pill bugs, potworms, spiders, springtails, snails and
slugs, and mites were collected in much lower numbers. The majority of the earthworms
collected were identi ed to be Pheretima spp. (see Appendix D). The distribution of major
soil animal groups with depth is shown in Figure 2.45. Earthworms (Lumbricidae) were
found at nearly equal densities in the upper and middle soil layers, and a lower density in
the lower layer. Formicidae were found only in the upper and middle layers. Chilopods
were found at all layers, with density decreasing with depth.
79
Table 2.9: Summary of data collected at CC, averaged.  indicates a parameter for which
values are summed over the sampling depth rather than averaged. yLayers where no
earthworms were found are not included in the average.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 10.2  9.2 10.0  10.4 7.7  8.0 27.8  27.0 
Earthworm Size
(cm)
2.7  1.4 1.9  0.8 3.4  1.3 2.7  0.7y
Root Biomass (g) 26.04  14.65 19.30  18.30 11.15  15.32 56.49  15.08 
Soil Organic Matter
(%)
30.13  11.42 16.37  14.19 10.06  11.02 18.85  11.89
Soil Textural
Classi cation
Sandy Loam Sandy Loam Sandy Loam Sandy Loam
Sand Content (%) 62.4  16.8 53.9  25.1 64.5  25.6 66.1  25.2
Silt Content (%) 31.0  11.4 36.1  20.4 24.6  18.5 25.3  16.8
Clay Content (%) 6.5  5.5 10.0  6.0 10.8  7.4 8.6  8.4
d60 (mm) 0.25 0.25 0.25 n/a
d10 (mm) 0.05 0.005 0.005 n/a
Coe cient of
Uniformity
(d60/d10)
5 50 50 n/a
Soil Animals (#s,
includes
earthworms)
23  22 18  17 11  15 52  54 
Total Number of
Taxa
(not averaged)
8 9 4 12 
80
Figure 2.43: CC particle size distribution.
Figure 2.44: CC animal diversity. Total number of individuals of major inver-
tebrate taxa found over entire site.
81
Figure 2.45: CC animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
82
2.3.9 Beltway Plaza Mall (BP)
A summary of the data collected at BP is presented in Table 2.10. An average of
19.5  12.3 earthworms were collected from the uppermost soil layer, 4.0  4.6 from the
10{20 cm depth, and 0.7  0.6 below 20 cm. Mean earthworm size was 3.8 cm. Root
biomass was highest in the uppermost soil layer (19.08  11.44 g), decreasing to 7.12
 2.16 g in the middle layer and 2.98  2.15 g in the lower layer. Total root biomass
for the entire sampling depth was 29.18  11.49 g. Soil organic matter was highest in
the upper layer (9.50  3.89%), decreasing to 2.92  0.08% and 3.03  1.18% in the
middle and lower layers, respectively. Soil texture was sandy clay loam in the upper and
middle layers and loam in the lower layer, averaging 53.0  6.6% sand, 23.5  2.8% silt,
and 23.4  4.3% clay over the entire sampling depth. The particle size distribution is
presented graphically in Figure 2.46. A coe cient of uniformity of 50 was measured in
the upper and middle layers, and a coe cient of uniformity of 20 was measured in the
lower layer. A total of 6 animal taxa were collected. Animals were found primarily in the
upper soil layer, but also occurred in the middle and lower layers. The composition of the
animal community is presented in Figure 2.47. The BP animal community is dominated
by earthworms (Lumbricidae). Ants, centipedes, millipedes, pill bugs, and beetles were
collected in much lower numbers. Representatives of Eisenia fetida, Lumbricus sp., and
Pheretima sp. were identi ed among the collected earthworms (see Appendix D). The
distribution of major soil animal groups with depth is shown in Figure 2.48. Earthworms
(Lumbricidae) were found at all depths, with density decreasing with depth. Formicidae
were found in the upper and middle layers, and at a higher density in the middle layer
than in the upper layer. Chilopoda and Coleoptera were found only in the upper soil layer.
83
Table 2.10: Summary of data collected at BP, averaged.  indicates a parameter for which
values are summed over the sampling depth rather than averaged. yestimated.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 19.5  12.3 4.0  4.6 0.7  0.6 24.2  15.1 
Earthworm Size (cm) 3.8  0.0 3.8  0.0 3.8  0.0 3.8  0.0 y
Root Biomass (g) 19.08  11.44 7.12  2.16 2.98  2.15 29.18  11.49 
Soil Organic Matter
(%)
9.50  3.89 2.92  0.08 3.03  1.18 5.15  0.96
Soil Textural
Classi cation
Sandy Clay
Loam
Sandy Clay
Loam
Loam Sandy Clay
Loam
Sand Content (%) 62.2  9.7 54.4  1.0 42.5  13.8 53.0  6.6
Silt Content (%) 16.2  5.0 22.8  1.5 31.5  9.2 23.5  2.8
Clay Content (%) 21.6  4.8 22.7  1.3 26.0  6.9 23.4  4.3
d60 (mm) 0.25 0.25 0.1 n/a
d10 (mm) 0.005 0.005 0.005 n/a
Coe cient of
Uniformity
(d60/d10)
50 50 20 n/a
Soil Animals (#s,
includes earthworms)
24  12 6  2 1  1 30  12 
Total Number of Taxa
(not averaged)
6 2 1 6 
Figure 2.46: BP particle size distribution.
84
Figure 2.47: BP animal diversity. Total number of individuals of major inver-
tebrate taxa found over entire site.
85
Figure 2.48: BP animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
86
2.3.10 Laurel Regional Hospital (LRH)
A summary of the data collected at LRH is presented in Table 2.11. An average of
13.7  0.7 earthworms were collected from the uppermost soil layer, 4.2  3.3 from the
10{20 cm depth, and 3.8 3.8 below 20 cm. Mean earthworm size was 4.0 0.3 cm. Root
biomass was highest in the uppermost layer (7.74 6.05 g), decreasing to 5.04 5.01 g in
the middle layer and 0.95 0.85 g in the lower layer. The total root biomass for the entire
sampling depth was 13.73  11.69 g. Soil organic matter was highest in the upper layer
(5.96  2.11%), decreasing to 2.13  0.20% in the lower layer. Soil texture was sandy
loam in the upper layer and loam in the middle and lower layers, averaging 46.4  14.4%
sand, 33.3 14.5% silt, and 20.3 4.1% clay over the entire sampling depth. The particle
size distribution is presented graphically in Figure 2.49. A coe cient of uniformity of 50
was measured in the upper layer, and a coe cient of uniformity of 20 was measured in
the middle and lower layers. A total of 10 animal taxa were collected. Animals were
found primarily in the upper soil layer, but also occurred in the middle and lower layers.
The composition of the animal community is presented in Figure 2.50. The LRH animal
community is dominated by earthworms (Lumbricidae). Grubs, other beetle larvae, ants,
centipedes, millipedes, pill bugs, beetles, slugs and snails, and  y larvae were collected
in much lower numbers. The majority of the earthworms collected were of indeterminate
Lumbricus sp. Representatives of Lumbricus rubellus, and Apporectodea caliginosa were
also identi ed (see Appendix D). The distribution of major soil animal groups with depth
is shown in Figure 2.51. Earthworms (Lumbricidae) were found at all depths, with density
decreasing with depth. Isopoda were found only in the upper soil layer.
87
Table 2.11: Summary of data collected at LRH, averaged.  indicates a parameter for
which values are summed over the sampling depth rather than averaged. yLayers where
no earthworms were found are not included in the average.
Depth
0{10 cm 10{20 cm 20{30 cm Entire Depth
Pro le
Earthworms (#s) 13.7  0.7 4.2  3.3 3.8  3.8 21.7  6.5 
Earthworm Size (cm) 4.2  0.8 3.6  0.8 4.1  0.9 4.0  0.3y
Root Biomass (g) 7.74  6.05 5.04  5.01 0.95  0.85 13.73  11.69 
Soil Organic Matter (%) 5.96  2.11 3.36  0.94 2.13  0.20 3.82  1.00
Soil Textural
Classi cation
Sandy Loam Loam Loam Loam
Sand Content (%) 52.7  14.2 45.7  14.6 40.8  18.4 46.4  14.4
Silt Content (%) 29.3  14.5 31.2  13.0 39.4  18.0 33.3  14.5
Clay Content (%) 18.0  1.9 23.1  14.1 19.8  0.4 20.3  4.1
d60 (mm) 0.25 0.1 0.1 n/a
d10 (mm) 0.005 0.005 0.005 n/a
Coe cient of Uniformity
(d60/d10)
50 20 20 n/a
Soil Animals (#s,
includes earthworms)
17  4 4  4 3  3 24  10 
Total Number of Taxa
(not averaged)
9 4 1 10 
Figure 2.49: LRH particle size distribution.
88
Figure 2.50: LRH animal diversity. Total number of individuals of major
invertebrate taxa found over entire site.
89
Figure 2.51: LRH animal depth pro le, showing the variation in density with
depth of the most abundant macroinvertebrate taxa.
90
2.3.11 Results of in ltration tests
In ltration tests were performed at three rain garden sites, using a double-ring
in ltrometer. Results are compared to the in ltration rate that would be expected given
the soil textures measured at each of the  eld sites (Appendix D). The expected in ltration
rate was estimated using published values for the saturated hydraulic conductivities of
the soil textures found at each of the sites (Rawls et al., 1982). These published values
were derived by taking the means of a large set of published and unpublished saturated
hydraulic conductivity measurements for each of the USDA soil textural classi cations.
Under saturated conditions, the in ltration rate is approximately equal to the saturated
hydraulic conductivity. For vertical  ow through a layered soil, the e ective saturated
hydraulic conductivity, Ke, is given by:
Ke =
Pn
i=1LiP
n
i=1
Li
Ki
(2.3)
where Li = layer height, and
Ki = layer saturated hydraulic conductivity (Hillel, 1998).
The results of in ltration testing at the BP site are presented in Table 2.12. The
in ltration rate was measured at three locations, and ranged from a high of 18.4 cm/h
to a low of 1.6 cm/h. The average in ltration rate was 8.2  8.9 cm/h. The e ective
saturated hydraulic conductivity that would be expected given the measured soil texture
is 0.5 cm/h. The results of in ltration testing at the LRH site are given in Table 2.13. The
in ltration rate was measured at three locations, and ranged from a high of 12.7 cm/h
to a low of 3.0 cm/h. The average in ltration rate was 8.3  4.9 cm/h. The e ective
saturated hydraulic conductivity that would be expected given the measured soil texture
is 1.4 cm/h. The in ltration rate was measured at two locations at the NWHS site. The
91
results of these tests are given in Table 2.14. The measured in ltration rate ranged from
a high of 10.6 cm/h to a low of 0.2 cm/h. The average measured in ltration rate was 5.4
 7.4 cm/h. The e ective saturated hydraulic conductivity that would be expected given
the measured soil texture is 1.7 cm/h.
Table 2.12: Results of in ltration testing at BP site. In ltration rate was measured at
three locations. Average in ltration rate is reported as mean  standard deviation, and
is compared to the predicted in ltration rate of the soil given it?s measured texture, using
values from Rawls et al. (1982).
Site In ltration Rate (cm/h)
1 18.4
2 1.6
3 4.6
Average In ltration Rate (cm/h) 8.2  9
Predicted In ltration Rate (cm/h) 0.5
Table 2.13: Results of in ltration testing at LRH site. In ltration rate was measured at
three locations. Average in ltration rate is reported as mean  standard deviation, and
is compared to the predicted in ltration rate of the soil given it?s measured texture, using
values from Rawls et al. (1982).
Site In ltration Rate (cm/h)
1 12.7
2 9.0
3 3.0
Average In ltration Rate (cm/h) 8.3  4.9
Predicted In ltration Rate (cm/h) 1.4
Table 2.14: Results of in ltration testing at NWHS site. In ltration rate was measured
at two locations. Average in ltration rate is reported as mean  standard deviation, and
is compared to the predicted in ltration rate of the soil given it?s measured texture, using
values from Rawls et al. (1982).
Site In ltration Rate (cm/h)
1 10.6
2 0.2
Average In ltration Rate (cm/h) 5.4  7.4
Predicted In ltration Rate (cm/h) 1.7
92
2.4 Discussion
The  eld surveys were intended to assess the extent of biological activity in the soil,
to collect baseline soil physical data, and to look for evidence of pedogenesis in existing
rain gardens.
2.4.1 Extent of biological activity
2.4.1.1 Root systems
The extent of the root systems present in these rain gardens is an important measure
of the level of biological activity in the soil. Roots are the backbone of the soil ecosystem,
playing vital structural and functional roles (Six et al., 2004). As they grow and die
back, roots create void space in the soil. Root exudates glue soil particles together into
aggregates. Root cells slough o and die back, providing a major source of organic matter
for micro- and mesofauna (Gobat et al., 2003). The area immediately surrounding each
root (about 2 mm thick), called the rhizosphere, provides an oxygen and nutrient rich
habitat for microorganisms, especially bacteria, which are responsible for many of the
important chemical reactions taking place in the soil, such as nitri cation. Roots form
mutualistic associations with mycorrhizal fungi. These fungi grow into large networks of
 ne hyphae extending deeply into the soil. This improves the plant?s nutrient absorption,
and provides structure and stability to the soil. Roots also act as organisms, consuming
oxygen, water and nutrients from the soil.
As young plants establish themselves, they put a great deal of energy into developing
their root systems (Brady and Weil, 2002). Root biomass can therefore be expected to
increase quickly with the age of the rain garden. Figure 2.52 shows the root biomass
measured at each of the  eld sites. The trend toward increasing root biomass with age is
93
not monotonic, but the data suggest that older rain gardens tend to have greater biomass
than younger rain gardens. This analysis is of course somewhat confounded by variations
in vegetation between sites.
Figure 2.52: Mean root density for all research sites. Error bars represent the
standard deviation between sampling locations. Numbers above each column
indicate the age of the rain garden in years at the time of sampling.
94
2.4.1.2 Soil animals
This study has revealed the basic structure of the soil community. A robust com-
munity of fauna was found. Each of the fauna found can be expected to play a role in
the development of the soil. These roles are described in Table 2.15. The  rst step of
pedogenesis, the development of an organically-enriched A horizon, is fundamentally a
biological process. Plant litter falls on the soil surface. Millipedes, pillbugs, ants and
earthworms fragment this litter, mix it with bacteria in their guts, and incorporate this
inoculated plant material into the soil matrix. Earthworms, potworms, springtails, and
mites consume this pre-digested plant material, further fragmenting it, and further boost-
ing bacterial and fungal activity. Bacteria and fungi complete the process of humi cation,
converting this decayed plant material into stable humus. Predators, such as centipedes,
spiders, and some ants, regulate invertebrate populations, preventing explosive growth of
a population, which would lead to collapse. Invertebrates act as earth-movers as well.
Burrowing animals, such as earthworms, ants, beetle larvae, mix soil layers, incorporate
organic matter into the soil, aerate the soil, improve in ltration, and create and destroy
aggregates. Earthworms, snails and slugs exude mucilages which act to glue soil particles
together. Figure 2.53 shows the relative numbers of taxa found at each site.
95
Table 2.15: The roles of major invertebrate taxa in the soil ecosystem (Gobat et al., 2003;
Hole, 1981)
Taxon Role
Earthworms (Lumbricidae)
Create voids
Mix soil layers
Create and destroy soil aggregates
Distribute bacteria throughout soil
Fragment organic matter
Mix surface residues into soil
Consume organic matter
Potworms (Enchytraeidae)
Create voids
Stimulate bacteria and fungi
Consume organic matter
Beetles (Coleoptera)
Larvae create voids
Larvae consume plant roots
Adults consume plants, animal wastes, carrion, other
invertebrates
Ants (Formicidae)
Create voids
Mix soil layers
Consume plants, microarthropods
Millipedes (Diplopoda) Create voidsFragment plant litter
Springtails (Collembola)
Consume fungi
Fragment and mix of organic matter
Regulate microbial populations
Slugs and Snails (Gastropoda) Exude mucus that helps glue soil particles togetherConsume decaying organic matter, fresh plant
material, and carrion
Fly larvae (Diptera)
Fragment plant litter
Consume carrion
Consume vertebrate waste
Pillbugs (Isopods)
Fragment plant litter
Mix soil
Bury organic matter
Mites (Acari)
Fragment organic matter
mix organic matter throughout soil
Regulate microbial populations
Centipedes (Chilopoda) Feed on other invertebrates - control populations
Spiders (Araneae) Feed on other invertebrates - control populations
96
Figure 2.53: Soil animals found at each of the research sites, reported as total
numbers of individuals tallied over the entire site. Numbers above each column
indicate the age of the rain garden in years at the time of sampling.
97
2.4.1.3 Earthworms
Earthworms were found at all ten sites. Even the youngest site, UMCP, which
had only been in operation for one year, and had the sandiest soil, had earthworms. In
fact, earthworms are the only soil animal that was observed at all sites. Figure 2.54
shows the numbers of earthworms counted at each of the sites, and compares these values
to reported earthworm densities for a comparable natural system, an old pasture (109{
646 individuals/m2)(Edwards and Bohlen, 1996). All earthworm densities were close to
this range, with four sites within the range, four sites slightly below the range, and one
site above it. These data show that earthworms are colonizing rain gardens, and reach
densities similar to other natural areas. Sites where earthworm populations are lower may
be showing the in uence of many factors, such as the age of the rain garden, sand content
of the soil mix, soil moisture regime, or proximity to existing earthworm populations.
These factors are discussed in more detail in Sections 2.4.3 and 2.4.4.
98
Figure 2.54: Mean earthworm density for each of the  eld sites sampled.
Error bars represent the maximum and minimum measured values. The
shaded area represents reported earthworm densities for old pasture (109{646
individuals/m2) (Edwards and Bohlen, 1996).
99
2.4.1.4 Soil organic matter enrichment
The average organic content of the upper soil layer (0{10 cm) of each of the sites
is shown in Figure 2.55. The expected organic matter content of a \typical" bioretention
soil medium using 50% sand, 20% topsoil, and 30% compost by volume was calculated
for comparison purposes. Assuming the following dry bulk densities: sand, 1.5{2.0 g/cm3
(Brady and Weil, 2002); sandy loam topsoil, 1.3{1.5 g/cm3 (Brady and Weil, 2002); com-
post, 0.3{0.4 g/cm3 (Glancey and Ho man, 1996); and BSM, 1.27{1.32 g/cm3 (Thompson
et al., 2008), the compost content of this mixture is 7{9% by weight. The organic mat-
ter content of compost can vary widely, from 30{65% on a dry weight basis (CSPMA,
1996). This would yield a soil mix with 2{6% organic matter. Six of the ten sites sampled
showed organic matter levels within this range. One site (BP) had a slightly elevated
organic matter content, and three sites (MJES, PP and CC) had much higher organic
matter content.
SOM was measured by the loss-on-ignition method (ASTM, 2000), which measures
the total organic matter fraction smaller than 2 mm in diameter. This method, there-
fore, does not include large, undecomposed organic matter fragments. Bioretention soil
media sometimes use shredded hardwood mulch as an organic matter source, and much
of this organic matter would be excluded from the measurement. This is appropriate, as
organic matter pieces this large do not play the same role as smaller, more decomposed
organic matter fractions. In addition, the method does not distinguish between small,
undecomposed particles and the more active, humi ed organic matter fractions, which
are of interest due to their prominent role in the maintenance of soil structure and their
strong a nity for cations (Brady and Weil, 2002; Weil et al., 2003). Measurement of this
active fraction falls to future researchers. Some sites did contain a great deal of hardwood
100
mulch, which had often degraded su ciently to create a lot of very small mulch pieces.
This tended to drive up the measured organic matter content. Indeed, the three sites
with elevated organic content, MJES, CC, and PP, were observed to have a lot of mulch
incorporated into the upper soil layers.
Di erent speci cations were used in the construction of many of these sites, but
these speci cations were not available for all sites. All available data on the research sites
is included as Appendix A. At Beltway Plaza, the speci cation called for  ve parts topsoil
to one part peat moss or rotted manure. Peat moss typically contains 80% organic matter
(CSPMA, 1996), which would yield a soil medium with 13% organic matter, a higher value
than the average measured content of 9.5%. Unfortunately, soil testing data for the soil
medium installed at this site is not available, so it is impossible to say whether the organic
matter content of the soils has increased or decreased over time.
2.4.2 Baseline physical data
2.4.2.1 Soil texture
Current speci cations for bioretention soil mix (PGCO, 2001; LIDC, 2003) specify
that the mix shall contain 50% sand, 20% sandy loam topsoil, and 30% compost or shred-
ded hardwood mulch. This would yield a soil texture of 83.7{96.7% sand, 0{14.3% silt,
and 0{5.7% clay. Of the ten sites sampled, only UMCP was within these speci cations,
though PP was very close. All other sites had lower sand and higher silt and clay contents.
The cause of this discrepancy could not be determined de nitively, because testing
data are not available for the soils used in the construction of the cells. Indeed, even
speci cations were unavailable for most of the sites. We do know that Prince George?s
County?s bioretention speci cations have evolved over time, requiring greater quantities of
101
Figure 2.55: Mean upper layer % SOM for all sites. Error bars show standard
deviations. The shaded area represents the expected organic matter content
of a typical BSM soil medium (2{6%).
Table 2.16: Average soil texture for each of the research sites.
Site Sand (%) Silt (%) Clay (%) Textural Classi cation
UMCP 91.0  1.2 4.6  0.9 3.8  0.3 Sand
WNY 71.3  3.4 21.3  5.7 7.4  2.8 Sandy Loam
MJES 65.8  21.3 19.7  12.4 14.5  8.9 Sandy Loam
CBF 61.0  0.9 16.5  0.6 22.5  1.3 Sandy Clay Loam
NWHS 51.4  4.0 34.2  1.0 14.4  3.7 Loam
PP 82.3  3.0 8.1  2.0 9.6  1.6 Loamy Sand
CF 66.9  1.6 17.8  0.6 15.4  1.0 Sandy Loam
CC 66.1  25.2 25.3  16.8 8.6  8.4 Sandy Loam
BP 53.0  6.6 23.5  2.8 23.4  4.3 Sandy Clay Loam
LRH 46.4  14.4 33.3  14.5 20.3  4.1 Loam
sand in the soil mix. Beltway Plaza Mall?s plans speci ed  ve parts topsoil to one part wet
loose peat moss or rotted manure. The topsoil could be: loam, sandy loam, clay loam, silt
loam, sandy clay loam, or loamy sand (see Appendix A). In contrast, the current Prince
102
George?s County bioretention speci cations, which were used in the construction of the
UMCP cell, specify a soil mix that is 50% sand, 30% sandy loam topsoil, and 20% organic
material in the form of compost or shredded hardwood mulch (see UMCP site details,
Appendix A). Di erences in the particle size distribution of the original soil mix could
account for the di erences in soil textures measured in this study. It is also possible that
we are seeing the e ect of years of enrichment of the system with  ne material carried in
by stormwater.
At several sites, large clods of soil with high clay content were found in the deeper
soil layers. The origin of these clods is unknown, though it is likely that they are unmixed
components of the original soil used in construction. Mottling and gleying were observed
in many of these clods, indicating the presence of reducing conditions in these microsites.
These clods could be providing a location for denitri cation and other chemical reactions
that require anaerobic conditions.
2.4.2.2 In ltration rate
At the three sites where the in ltration rate was measured, the measured in ltration
rate greatly exceeded the in ltration rate that would be expected given the particle size
distribution of the soil (Figure 2.56). This is an expected result of the presence of biological
actors in the soil, one of whose e ects is to increase soil porosity through burrowing and
the creation of aggregates. This measurement highlights the value of biological activity in
the successful operation of bioretention systems.
103
Figure 2.56: Comparison of measured versus predicted in ltration rates for the
sites sampled. The measured value is the mean, and the error bars represent
the maximum and minimum measured values. Predicted values represent the
e ective in ltration rate for the soil textures measured onsite, based on Rawls
et al. (1982).
104
2.4.2.3 Evidence of pedogenesis
When treated as a chronosequence, the ten  eld sites yield a picture of the devel-
opment of a rain garden over a decade. In these  rst ten years, we see the beginnings of
pedogenesis, with the gradual formation of an A horizon, enriched with organic matter.
Figure 2.57 shows the gradual development of an enriched organic layer in the uppermost
soil layer (0{10 cm) over time. The pro les show both a shift toward higher organic mat-
ter in the upper layer and a shift toward a steeper gradient of organic matter between
the upper and lower soil layers. A very similar pattern was seen by Leisman (1957) in a
chronosequence on abandoned mine spoil banks in Minnesota (Figure 2.58).
Figure 2.57: Soil organic matter pro le development over time. Data series
are plotted with organic matter increasing to the right, and depth increasing
downward, so that the soil surface is at the top of the chart. The 5-year pro le
was derived by averaging data from the CBF and NWHS sites. The 7-year
pro le was derived by averaging data from the CC and BP sites.
105
Figure 2.58: Organic carbon pro le development in a mine spoil chronose-
quence (Leisman, 1957).
Organic matter in the uppermost soil layer (0{10 cm) appears to increase slightly
over time, but the coe cient of correlation (R2= 0.0475), as shown in Figure 2.59, is not
high enough to declare that this relationship de nitively exists. It is worth noting that the
extremely high organic matter content of the CC site is likely an artifact of the presence
of a very large amount of mulch mixed into the soil, rather than to decomposed organic
matter that has been incorporated into the soil matrix.
2.4.2.4 Changes in root biomass over time
The sequence appears to show an increase in root biomass over time, though the co-
e cient of correlation (R2= 0.27), as shown in Figure 2.60, is not high enough to state that
106
Figure 2.59: Soil organic matter in uppermost 10 cm versus age.
such a correlation de nitively exists. Root biomass increases as plants become established,
therefore root biomass is expected to increase over the  rst years after planting.
Here again, Figure 2.61 reveals a shift to the right with the age of the rain garden.
The diagram shows the gradual formation of a gradient with depth, with the greatest root
biomass in the topsoil, and decreasing at deeper soil levels. This is an expected result,
corresponding to declining oxygen levels at greater soil depths.
107
Figure 2.60: Root biomass versus age.
108
Figure 2.61: Root pro le development over time. Data series are plotted with
root biomass increasing to the right, and depth increasing downward, so that
the soil surface is at the top of the chart. The 5-year pro le was derived by
averaging data from the CBF and NWHS sites. The 7-year pro le was derived
by averaging data from the CC and BP sites.
109
2.4.2.5 Changes in earthworm abundance over time
Figure 2.62 shows no correlation between earthworm abundance and rain garden
age (R2 = 0). Figure 2.63 suggests an increase in earthworm abundance over time at
all depths, with greater increases in the uppermost soil layer. This is consistent with the
gradual colonization of the sites by earthworms washed in from surrounding natural areas.
Note the MJES had many more earthworms than any of the other sites. The reason for
this is unknown.
Figure 2.62: Earthworm abundance versus age.
110
Figure 2.63: Earthworm pro le development over time. Data series are plotted
with earthworms increasing to the right, and depth increasing downward, so
that the soil surface is at the top of the chart. The 5-year pro le was derived
by averaging data from the CBF and NWHS sites. The 7-year pro le was
derived by averaging data from the CC and BP sites.
111
2.4.2.6 Colonization by soil animals
Table 2.17 summarizes how the depth at which taxa were found varied with the
age of the rain garden sampled. At one year, macroinvertebrates were con ned to the
uppermost soil layer. After two years, they had spread to the 10{20 cm depth. After
three years, they had spread through the entire 30 cm sampling depth. All sites showed
higher numbers of macroinvertebrates in the upper soil layers. Increasing species richness
and density are expected patterns of the gradual colonization of a newly disturbed site.
Earthworms had colonized all sites, including the youngest site, UMCP, which had only
been in operation for one year. This suggests that earthworms may be an important
pioneer species in the rain garden ecosystem.
Table 2.17: Variation in total number of soil animals counted and number of di erent
animal taxa encountered at each sampling depth with age. Where two rain gardens are
of the same age, the results are averaged. Partial animal counts are the result of severing
of earthworms during sampling.
Depth
Age Site(s) 0-10 cm 10-20 cm 20-30 cm
1 year UMCP 21 animals
5 taxa
{ {
2 years WNY 34 animals
2 taxa
7 animals
2 taxa
{
3 years MJES 219.5 animals
11 taxa
12.5 animals
4 taxa
3 animals
1 taxon
4 years NWHS, CBF 126 animals
9.5 taxa
36.5 animals
5 taxa
18 animals
1.5 taxa
5 years PP 57 animals
9 taxa
7 animals
3 taxa
1 animal
1 taxon
6 years CF 102 animals
6 taxa
66 animals
7 taxa
41 animals
2 taxa
7 years CC, BP 70 animals
7 taxa
35.5 animals
5.5 taxa
18 animals
2.5 taxa
10 years LRH 50 animals
9 taxa
13 animals
4 taxa
10 animals
1 taxon
112
2.4.3 Confounding factors
Variations in the histories of these sites make them an imperfect chronosequence.
They di er in many respects, some of which may in uence organic matter, root biomass,
and animal populations to a greater degree than the passage of time. The most important
of these factors are: variations in soil texture, di erences in plant cover, di erences in the
litter/mulch layer, and variations in the proximity to natural areas.
2.4.3.1 E ect of soil texture on earthworm populations
It is well known that earthworms dislike very sandy soils (Curry, 1998). This is
thought to be due in part to the abrasive e ects of sands on their soft bodies. Figure
2.64 suggests an inverse relationship exists between earthworm abundance and the sand
content of the soil, though the coe cient of correlation (R2 = 0.3) is not high enough
to make such a statement de nitively. This may be due to the presence of MJES as an
outlier.
113
Figure 2.64: Relationship between earthworm abundance and sand content.
114
2.4.3.2 E ect of plant cover and litter layer
The type and extent of plant cover can be expected to have a major in uence on
many aspects of the biology of the rain garden. Plant cover will determine to a large
extent the root biomass that can be expected to develop over time. It will also determine
the amount of plant litter than will accumulate on the rain garden surface, contributing
to the organic content of the soil. The type of plants present can a ect soil animals, as the
litter produced may be more or less easily broken down. For example, earthworms tend to
dislike leaves containing high levels of tannin, such as spruce, oak, and beech (Edwards and
Bohlen, 1996). Dominance by one of these species could depress earthworm populations.
The type and thickness of the litter layer can directly e ect soil animal populations. In
addition to serving as a food source, the litter layer provides shelter for surface-dwelling
organisms, and protects the soil surface from temperature extremes.
2.4.3.3 Proximity to natural habitat
Rain gardens are not inoculated with soil organisms when they are constructed, so
any invertebrates that establish themselves within the soil must have either been acci-
dentally introduced with the soil mix, or must have colonized from nearby natural areas.
Therefore, the proximity of a rain garden to such a natural area might be an important
factor in determining how quickly it will be colonized by soil invertebrates, and may also
determine which animals are likely to be introduced.
2.4.3.4 E ect of management
Various management schedules could have an in uence on biological development.
For example, mowing, sediment removal and mulch replacement, use of pesticides and
115
herbicides within the watershed, use of road salts within the watershed, would be ex-
pected to impact both plant and animal populations. Unfortunately, detailed data on the
management of these sites are unavailable.
2.4.4 Sources of error in animal sampling
Sampling of soil-dwelling macroinvertebrates is notoriously di cult to perform ac-
curately (Southwood and Henderson, 2000). A number of factors could have in uenced
the survey results. Environmental factors, such as time of year, temperature on the sam-
pling date, and the antecedent weather conditions, may have in uenced macroinvertebrate
populations at the time of sampling. In addition, the sampling method used, while ap-
propriate for assessing the earthworm population, does not give an accurate population
estimate for all taxa. Many of the very small soil animals, such as springtails, were prob-
ably missed during hand sorting. For this reason, animal numbers should be used only
for comparison between these sites and not for comparison with other studies. Many soil
animals are likely to have escaped during the sampling process.
2.4.4.1 Time of year
Soil macroinvertebrates can be expected to be at di erent stages of their life cycles
at di erent times of year. All samples were taken during the warmer months, between
May and October, but populations at the beginning of the season could di er substantially
from populations at the end of the season. Figure 2.65 compares the number of animals
found at each site with the month during which sampling took place. The data show a
tendency toward higher numbers of animals found at the end of the summer than at the
beginning of the summer. Earthworm populations seem to peak in midsummer.
116
Figure 2.65: E ect of month of sampling on earthworms and total number of
soil animals detected.
2.4.4.2 Outdoor temperature
High ambient temperatures could have caused variations in the number of animals
found. Animals could have retreated to cooler spots, either deeper in the soil, or elsewhere
in the rain garden (under trees, for example). This would result in a decrease in the number
of animals found with increasing temperature. Figure 2.66 reveals no such pattern on
declining animal numbers with increasing temperature. In fact, if anything, the data tend
toward higher numbers of animals detected at higher temperatures.
117
Figure 2.66: E ect of ambient temperature on earthworms and total numbers
of soil animals detected. Ambient temperature is the daily maximum temper-
ature recorded at the nearest NOAA weather station (Beltsville, MD).
118
2.4.4.3 Antecedent weather conditions
As soil invertebrates are dependent on water and oxygen for their survival, weather
conditions in the week prior to sampling might in uence the number of animals found.
During droughts, earthworms tend to aestivate, curling up into a tight ball, often inside
a soil aggregate. This makes their detection di cult. Many other animals will die o 
if no water is available. However, too much rain could also depress animal populations
by limiting the amount of available oxygen in the soil. Therefore, we would expect to
see lower numbers of earthworms and other animals after particularly wet or dry periods.
Interestingly, Figure 2.67 shows a di erent pattern, possibly even the inverse of what is
expected. The highest animal counts occurred after the driest periods. This may suggest
that animals are better able to survive lack of water than lack of oxygen, or it may be a
coincidence only, caused by some other factor.
Figure 2.67: E ect of precipitation on numbers of earthworms and other soil
animals detected. 7-day rainfall totals are those recorded at the nearest NOAA
weather station (Beltsville, MD).
119
2.4.5 In ltration tests
In ltration tests were conducted on only three of the ten research sites. In ltration
testing on all sites were not attempted due to time constraints and concerns about the
reliability of the data. The in ltration tests were performed using a double-ring in ltrom-
eter, which proved to be unsuitable to the soil conditions encountered in a rain garden.
Double-ring in ltrometers are best used in sites with little root biomass, and where the
rings can be driven into the ground using hydraulic equipment (ASTM, 2003). Biore-
tention cells are not accessible to hydraulic equipment, such as truck jacks, because the
surrounding soil must not be compacted. Therefore, the rings must be driven in by hand
using a sledgehammer. Bioretention cells are heavily vegetated, and many of the older
sites contain tree and shrub species. Roots more than 1 cm in diameter are common.
These cannot be severed by the rings as they are driven into the soil. Where the rings
encountered large roots, the roots were cut through by pushing a machete down alongside
the rings. This added to the disturbance of the surrounding soil, which could have lead
to inaccurate measurements.
Bioretention cells are constructed with a thick (commonly 3-inch) mulch layer at
the soil surface. Over time, this mulch layer breaks down and is incorporated into the
soil surface. This poses a particular di culty in using the double ring in ltrometer, as
the mulch surface must be removed in order to drive the rings in to the ground, but the
mulch layer does not end abruptly, but rather mulch pieces become incorporated into the
upper soil layers. The driving of the rings is easily halted by hitting a large mulch piece.
The data shows wide variability in in ltration rates between tests at the same site, but it
is impossible to say whether this is the result of spatial heterogeneity or inaccuracy due
to the inapplicability of the measurement technique to the site conditions.
120
2.4.6 Diversity among rain gardens
This study reveals the wide diversity in what we refer to as a rain garden; there
are huge di erences in the soils, vegetation, animals, designs and watersheds. Age was
expected to be the primary driver of di erences between the sites. Factors were expected
to change steadily over time, with higher organic matter, roots and animal populations at
older sites. This study found a great deal of variation between sites, suggesting that other
factors are at least as important as time in determining the level of biological activity at
a given site.
2.4.6.1 Di erences in history, location
Unfortunately, data on the original site designs were often unavailable. The oldest
site for which as-built plans were available was the Beltway Plaza Mall (BP) site (see
Appendix A). The rain garden that was sampled was constructed on this site in 1997.
The bioretention speci cation contained in these plans allows a wide range of soil textures
to be used in the bioretention mixture. No sand was to be added, and the mix was 5 parts
soil to 1 part organic matter, which is much less than the 3:2 mix currently speci ed.
Di erences in land use within watershed would a ect runo composition. The dom-
inant land use within the watersheds of the sampling sites are given in Table 2.18. The
unpaved parking lot at the Chesapeake Bay Foundation (CBF) might be expected to pro-
duce more sediment, but less runo overall due to increased in ltration over the parking
surface. The runo coming from the roof at Claggett Farm (CF) would carry pollutants
from atmospheric deposition only, and would not contain the same mix of pollutants re-
lated to automobiles, such as oil and grease, and heavy metals from tire wear. Reduced
121
heavy metal loading at Claggett Farm could be expected to provide better habitat for soil
animals.
Table 2.18: Land use within the watersheds of sampling locations.
Site Watershed Land Use
UMCP paved parking lot
WNY paved parking lot
MJES paved parking lot
CBF unpaved parking lot
NWHS paved parking lot
PP paved parking lot
CF roof
CC paved parking lot
BP paved parking lot
LRH paved parking lot
As shown in Figure 2.68, the sites show a signi cant decrease in sand content with
the age of the rain garden (R2 = 0.5). Unfortunately, we lack su cient historical data
to discern whether this trend is due to changes in the soil speci cations or to in ow of
additional silt and clay in the form of suspended solids. It is likely that both play a role,
though changes in soil speci cations are probably the primary driving force.
122
Figure 2.68: Changes in sand content with rain garden age.
123
2.5 Summary
1. Biological activity is ubiquitous in rain garden soils.
2. Rain gardens begin to undergo pedogenesis immediately after installation.
3. Rain gardens develop a characteristic soil pro le with exponentially decreasing bio-
logical activity with depth.
4. No evidence of clogging due to normal operation was found.
5. Earthworms were found at all  eld sites.
6. Earthworms occurred at densities similar to fallow pastures.
7. Representatives of thirteen major soil-dwelling invertebrate taxa were found.
8. Earthworms and other soil invertebrates rapidly colonize rain gardens in spite of
their physical isolation.
9. Some rain gardens showed higher than expected in ltration rates.
10. Substantial di erences in particle size distribution were found between rain gardens
of di erent ages, with newer sites containing more sand and less clay and silt.
11. Older rain gardens tend to have higher root biomass than younger rain gardens.
12. Many sites had lower soil organic matter than would be expected from the typical
bioretention soil medium.
13. Assessment of long-term rain garden development is hampered by the scarcity of
original design documents and soil testing data.
124
Chapter 3
Microcosm Study
3.1 Introduction
A microcosm study was performed to evaluate the e ect of earthworms on the
saturated hydraulic conductivity of soil columns subjected to simulated storm events.
Eighteen soil columns were divided into three treatment groups. The treatments were
designed to test the ability of earthworms to maintain the hydraulic conductivity of a soil
column subjected to sediment loading, and to improve the saturated hydraulic conductivity
of a soil column that is already clogged with sediment. The initial saturated hydraulic
conductivity of the columns was measured, then the columns were subjected to simulated
storm events for six months. The  nal saturated hydraulic conductivity was measured
at the end of the experiment, and earthworm survivorship was assessed. The change in
hydraulic conductivity over the course of the experiment within each treatment group was
evaluated, and the initial and  nal hydraulic conductivities of the treatment groups were
compared. An earlier study was performed to test the ability of earthworms to survive in
bioretention soil media (BSM). The results of this study are included as Appendix E.
3.2 Methods
Eighteen rain garden microcosms were constructed using 24-inch lengths of 4-inch
diameter PETG (polyethylene terepthalate glycol). These were divided into three treat-
ment groups of six columns each: the Control group, without earthworms; Treatment
1, with earthworms; and Treatment 2, with earthworms and an added sediment layer at
125
the surface. Treatments were assigned according to a strati ed randomized design, with
the columns arranged into groups of three (Table 3.1). This was intended to eliminate
potential sources of systematic error, such as di erences in temperature across the room
where the columns were set up.
Table 3.1: Treatment assignments for microcosms showing strati ed randomized experi-
mental design.
Column # Treatment Abbreviation
1 Control C
2 Worms T1
3 Worms + sediment T2
4 Worms + sediment T2
5 Control C
6 Worms T1
7 Control C
8 Worms + sediment T2
9 Worms T1
10 Worms + sediment T2
11 Control C
12 Worms T1
13 Worms + sediment T2
14 Worms T1
15 Control C
16 Control C
17 Worms + sediment T2
18 Worms T1
Earthworms were collected by hand digging from a rain garden in Beltway Plaza
Mall in Greenbelt, MD, and from a residential yard in Takoma Park, MD. Additional
adult Lumbricus terrestris were obtained by mail order (Carolina Biological Supply). The
collected earthworms were classi ed by size and coloration, but were all juveniles, and
were therefore not identi able. They were collected just below the soil surface, and were
likely endogeic species, which burrow horizontally through the topsoil. L. terrestris, an
anecic species, were added in order to increase the range of earthworm ecotypes present
in the microcosms.
126
The columns were constructed on November 4, 2005. Each column rested on a
funnel  lled with #57 washed stone gravel, as shown in Figure 3.1. The column was
 lled to a 45 cm depth with a variant of the standard bioretention soil medium (BSM):
50% sand, 30% sandy loam topsoil, 10% compost, and 10% shredded hardwood mulch by
volume (LIDC, 2003; PGCO, 2001). The BSM was tamped down by hand. Each of the
treatment columns was inoculated with:
 1 \rain garden" earthworm hatchling, unpigmented, length < 3 cm,
 1 medium pink \rain garden" earthworm, length 3{7 cm, and
 3{4 pink \Takoma Park" worms, with lengths ranging from 2{7 cm.
The earthworms were allocated in such a way as to balance out the total length of
earthworms added (i.e. one large worm = two small worms). The earthworms were placed
on the soil surface in the Treatment 1 and Treatment 2 columns. The Treatment 1 and
Control columns were then topped with a 3 cm depth of moistened shredded hardwood
mulch. Treatment 2 was topped with 150 g of manufactured sediment, creating a 1-cm-
deep sediment layer. This was then covered with 3 cm of moistened shredded hardwood
mulch. Sediment was created by passing BSM through a 0.6 mm (#30) mesh sieve, in
a manner similar to Hsieh and Davis (2005a). The sediment was intended to represent
trapped suspended solids.
Particle size analysis was performed using hydrometer and sieve tests as described
in A.1.9. The results are shown in Tables 3.2 and 3.4.
Three adult L. terrestris were added to each of the Treatment 1 and 2 columns on
December 15, 2005, following the completion of the initial tests, as they were expected to
quickly construct deep burrows, which would have a signi cant e ect on the soil saturated
hydraulic conductivity. These earthworms were placed on the soil surface, under the mulch
layer.
127
Figure 3.1: Diagram showing microcosm setup.
Table 3.2: Particle size distribution for bioretention soil mix.
Particle Size Percentage
sand 93.2%
silt 4.0%
clay 2.8%
d60 0.50 mm
d10 0.25 mm
Coe cient of uniformity, d60=d10 2
USDA textural classi cation sand
3.2.1 In ltration testing
The saturated hydraulic conductivity of the columns was measured using a con-
stant head in ltration setup (Figure 3.2). As equipment constraints allowed testing of
128
Table 3.3: Particle Size Distribution of manufactured sediment.
Particle Size Percentage
sand 88.4%
silt 6.5%
clay 5.1%
d60 0.50 mm
d10 0.05 mm
Coe cient of Uniformity, d60=d10 10
USDA Textural class loamy sand
only one column at a time, the columns were tested sequentially by column number. Ini-
tial tests were performed between 11/6/05 and 12/14/05. Final tests were performed
between 6/19/06 and 7/6/06. A constant water depth of 10 cm above the soil surface
was established, and the soil was allowed to saturate for one hour. Then, for the next
six hours, the volume of water exiting the base of the funnel was collected and measured
every hour to yield the  ow rate per hour. The tests were conducted for six hours in order
to ensure that the columns were completely saturated, and that the measured in ltration
rate would reach steady state. These hourly values were then averaged to determine  ow
rate of the column. Saturated hydraulic conductivity, Ks, was calculated by applying
Darcy?s Law in the following form (Hillel, 1998):
Ks = QLAh (3.1)
where Q = e uent  ow rate (cm/h)
L = soil height (cm)
A = column cross-sectional area (cm2), and
h = head (cm).
129
Figure 3.2: Diagram of setup to establish constant head for in ltration testing.
130
3.2.2 Earthworm feeding
The earthworms were fed a variety of seasonally available foods. During the win-
ter and early spring, the earthworms were fed assorted decaying leaves, primarily cherry
(Prunus sp.). These were supplemented with cornmeal in order to avoid any possibility
of malnutrition a ecting survivorship. Partially decayed leaves were collected outdoors,
soaked in tap water for three hours to soften, then roughly chopped. In late spring and
early summer, the earthworms were fed assorted fresh wild greens, supplemented with
cornmeal. Greens were collected outdoors, then roughly chopped. Greens included Ken-
tucky Bluegrass (Poa pratensis), white clover (Trifolium repens), dandelion (Taraxacum
o cinale) and plantain (Plantago major). Additional cornmeal and leaves were added
when they had been consumed in one or more of the columns. In total, the treatment
columns received 8.0 g of wet chopped decayed leaves, 3.0 g of chopped fresh greens, and
4.0 g of cornmeal. No foods were added to the Control columns, to avoid the possibility
of arti cially clogging the columns with unconsumed plant litter or mold growth.
3.2.3 Storm simulations
Storm simulations were designed to approximate a typical rainfall for May-October
in Prince George?s County, MD. This six-month period was selected because earthworms
are most active during late spring, summer, and early fall. The microcosms were intended
to represent unit elements of a rain garden. Ponding during simulated storms was intended
to re ect the ponding occurring in rain gardens during storm events. In order to obtain
representative values, 30-year rainfall averages published by the National Climatic Data
Center for the nearest weather station (Baltimore, MD) were used (NCDC, 2003).
131
Table 3.4: 30-year rainfall averages for Baltimore, MD (NCDC, 2003).
Month Precipitation (in) # of Rainy Days
May 3.89 11
June 3.43 10
July 3.85 9
August 3.74 9
September 3.98 8
October 3.16 8
It was assumed that the theoretical rain garden would drain a 1-acre paved parking
lot (CN = 98). The size of the theoretical rain garden was calculated following the sizing
procedure described in Low-Impact Development Hydrologic Analysis (PGCO, 2000). The
calculations are presented below.
Given:
Site area: 1 acre = 43,560 ft2
Existing CN: 98 (pavement)
Proposed CN: 60 (woods, fair condition, Hydrologic Soil Group B soils)
Design storm: Type II, 1-inch over 24 hours
Solution:
Using Appendix A in PGCO (2000), total storage volume required using retention storage
= 0.8 inches over the entire site.
Storage volume required = 43,560 ft2  0.8 in = 2,900 ft3
Depth of bioretention soil = 2.5 ft
Bioretention soil porosity (ideal) = 0.5
132
Ponding depth = 0.5 ft
E ective storage depth = soil depth  porosity + ponding depth = 1.75 ft
Size of bioretention cell required (rain garden area) = volume required  storage depth
= 1657.14 ft
Each column represented a unit volume within this rain garden.
The historic rainfall data contained in Table 3.5 was converted into a stormwater dosing
schedule for the columns by the following calculations. Assume equal rainfall for each
storm in a given month.
Rainfall = Total PrecipitationNumber of Storms (3.2)
Runo Volume = Rainfall Watershed Area (3.3)
Ponding Depth = Runo VolumeRain Garden Area (3.4)
Storm Volume = Ponding Depth Column Surface Area (3.5)
Revision to Storm Schedule. It was important to avoid waterlogging the columns to the
extent that anoxic conditions developed, which would have killed the earthworms. At
the end of month four, it was apparent that the soil columns were not drying su ciently
between storms. The storm volume used for months 5 and 6 was reduced to 2.00 L (see
Table 3.6), but the suspended solids (SS) added with each storm event was held steady at
133
Table 3.5: Design storm frequencies, based on 30-year rainfall averages for Baltimore, MD
(NCDC, 2003).
Real
Month
Real Dates Data Month Days in
Data
Month
# Storms Storm
Frequency
(days)
1 12/15/05-1/14/06 May 31 10 2.8
2 1/15/06-2/13/06 June 30 10 3
3 2/14/06-3/17/06 July 31 9 3
4 3/18/06-4/17/06 August 31 9 3.4
5 4/18/06-5/17/06 September 30 8 3.75
6 5/18/06-6/17/06 October 31 8 3.9
Table 3.6: Design storm volumes, based on 30-year rainfall averages for Baltimore, MD
(NCDC, 2003). Storm volumes for months 5 and 6 were reduced to 2.00 L per storm.
Real Month Real Dates Data Month Storm Volume
Scheduled (L)
Storm Volume
Actual (L)
1 12/15/05-1/14/06 May 1.91 1.91
2 1/15/06 2/13/06 June 1.85 1.85
3 2/14/06 3/17/06 July 2.31 2.31
4 3/18/06 4/17/06 August 2.25 2.25
5 4/18/06 5/17/06 September 2.69 2.00
6 5/18/06 6/17/06 October 2.14 2.00
3.0 g, as discussed below, in order to avoid changing the total SS input over the course of
the experiment.
3.2.3.1 Suspended solids loading
In order to maximize the potential for clogging over the 6-month experiment, suf-
 cient solids were added to potentially create a 1-cm sediment layer. 3.0 g of SS were
added to each column during every storm event, yielding a total of 165.0 g over the course
of the experiment. The SS concentration of the simulated runo ranged between 1,100
mg/L and 1,600 mg/L, an order of magnitude higher than typical values for urban runo ,
134
which are typically on the order of 150 mg/L (Hsieh and Davis, 2005a; Kadlec and Knight,
1996).
To ensure that the entire SS load was added with each storm event, the solids were
not mixed with the water, but rather sprinkled on the column surface prior to the addition
of the water. A thin plastic shield was placed over the surface after the solids were added
but prior to wetting in order to minimize the disturbance of the soil surface. This  lm was
removed as soon as all of the water had been added to the column, so that the natural
 ltration of the solids by the soil would not be impeded.
Solids were manufactured in two batches. The  rst batch was the same as the
sediment placed in the Treatment 2 columns during setup, created by passing BSM through
a 0.6 mm (#30) sieve. The second batch was manufactured at a later date by the same
method, using the BSM used in the soil columns, combined with waste soil collected from
the sampled rain gardens. The addition of waste soil was intended to increase the silt and
clay content of the sediment. The second batch was used from day 79 (3/3/06) through
the end of the experiment. The particle size distributions of the two batches are given in
Table 3.7.
Table 3.7: Particle size distribution for manufactured suspended solids (SS).
First Batch Second Batch
Sand 88.4% 78.5%
Silt 6.5% 13.1%
Clay 5.1% 8.4%
d60 0.5 mm 0.25 mm
d10 0.05 mm 0.05 mm
Coe cient of Uniformity, d60=d10 10 5
USDA textural classi cation loamy sand loamy sand
135
3.2.3.2 Statistical analyses
The mean, standard deviation and variance of the initial and  nal saturated hy-
draulic conductivity (Ks) were calculated for each of the treatment groups. A two-tailed
paired Student?s t-test was used to examine the change in Ks over time within each of the
treatment groups. In a paired t-test, the population of interest is the di erence between
a pair of measurements ( D = Ksf { Ksi). In this case, the paired measurements are
the initial and  nal Ks for each column. Student?s t-test requires that the population of
interest be normally distributed or approximately normally distributed, that samples are
randomly selected from the populations of interest, and that the pairs of observations are
independent.
Null hypothesis: mean initial Ks equals the mean  nal Ks (H0: D = 0, Ksi = Ksf).
Alternative hypothesis: mean initial Ks does not equal the mean  nal Ks (Ha: D 6= 0,
Ksi 6= Ksf).
If Student?s t-test yields a signi cant result, then the null hypothesis is rejected, and
it can be concluded with a high degree of con dence that the mean values of the popula-
tions being compared di er. Student?s t-test is sensitive to outliers and extreme skewness,
but is reasonably accurate for data sets that are approximately normally distributed (Ott
and Longnecker, 2001). Outliers and skewness were assessed visually by examination of
quantile comparison plots generated using the R statistical software package (R Develop-
ment Core Team, http:nnwww.r-project.org). A normally distributed data set will appear
as a straight line. If the data set deviates far from this line, either by forming a more
curved line (evidence of skewness) or by the presence of data points falling far from the
136
line (outliers), then this is evidence that the data are far from normally distributed, and
the t-test is not appropriate.
One-way analysis of variance (ANOVA) was employed to compare the saturated
hydraulic conductivities of the treatment groups at the beginning and end of the experi-
ment. ANOVA relies on the comparison of the variance between treatment groups to the
variance within treatment groups. The higher the ratio, the more likely that a signi cant
di erence exists between treatment groups.
Null hypothesis: Ks is the same for all treatments (H0: KsC = KsT1 = KsT2)
Alternative hypothesis: Ks di ers between one or more treatments.
ANOVA requires that: (1) each of the populations is approximately normally dis-
tributed, (2) each of the sets of measurements is an independent random sample from its
population, and (3) the populations must have equal variances. Normality of the popu-
lations was established prior to conducting Students t-tests. Homogeneity of variance is
assessed using Levine?s Test:
Null hypothesis: variances are equal (H0:  2C =  2T1 =  2T2)
Alternative hypothesis: variances are not all equal.
All tests were performed using the R statistical software program (R Development Core
Team, http:nnwww.r-project.org).
3.3 Results
3.3.1 Control
Table 3.8 and Figure 3.3 illustrate the test results for the control group. The average
hydraulic conductivity for the group decreased from 84  12 cm/h at the beginning of
137
the experiment to 60  11 cm/h at the end of the experiment. Hydraulic conductivity
decreased for all columns in the group.
Table 3.8: Results of saturated hydraulic conductivity tests for Control group. Average
values are presented as mean  standard deviation.
Column Initial Ks (cm/h) Final Ks (cm/h) Change (cm/h)
1 76 45 -31
5 83 66 -16
7 106 55 -51
11 74 55 -18
15 86 76 -10
16 82 61 -20
Average 84  12 60  11 -24  15
Figure 3.3: Average saturated hydraulic conductivity for Control group. Error
bars indicate the standard deviation from the mean. Average initial saturated
hydraulic conductivity was 84  12 cm/h. Average  nal saturated hydraulic
conductivity was 60  11 cm/h.
138
A paired Students t-test was performed to compare the initial saturated hydraulic
conductivity (Ks) to the  nal Ks of the Control group. Examination of the quantile-
quantile plot (Figure 3.4), shows that the data does not deviate far from a normal dis-
tribution, therefore, the t-test is applicable. Results of the t-test are presented in Table
3.9. The p-value <0.05, therefore there was a signi cant change in Ks over time for the
Control group.
a71
a71
a71
a71
a71
a71
?1.0 ?0.5 0.0 0.5 1.0
?50
?40
?30
?20
?10
Normal Q?Q Plot
Theoretical Quantiles
Sample Quantiles
Figure 3.4: Quantile-quantile plot of Control group change. The straight line
represents a normal distribution. Visual inspection of the data set suggests
that the data fall closely enough to the normal line to be treated as normally
distributed.
Table 3.9: Two-sided paired t-test of Control group change.
Null hypothesis degrees of freedom T p-value Con dencelevel Signi cant?
H0 : Ksfin = Ksinit 5 4.1 0.01 95% YES
139
3.3.2 Treatment 1
Table 3.10 and Figure 3.5 illustrate the test results for the Treatment 1 group.
The average hydraulic conductivity for the group decreased from 78  14 cm/h at the
beginning of the experiment to 69  28 cm/h at the end of the experiment. There was
a high degree of variability in the  nal saturated hydraulic conductivity, ranging from 34
cm/h for Column 6 to 114 cm/h for Column 2.
Table 3.10: Results of saturated hydraulic conductivity tests for Treatment 1 group. Av-
erage values are presented as mean  standard deviation.
Column Initial Ks (cm/h) Final Ks (cm/h) Change (cm/h)
2 80 114 34
6 64 34 -31
9 77 47 -31
12 80 81 1
14 103 65 -38
18 65 72 7
Average 78  14 69  28 -10  28
Examination of the quantile-quantile plot (Figure 3.6) shows that the data set does
not deviate far from a normal distribution. Therefore the t-test is applicable. Results of
the t-test are presented in Table 3.11. P-value >> 0.05, therefore the null hypothesis is
not rejected. The average saturated hydraulic conductivity of the T1 treatments did not
signi cantly decrease over time.
Table 3.11: Two-sided paired t-test of Treatment 1.
Null hypothesis degrees of freedom T p-value Con dencelevel Signi cant?
H0 : Ksfin = Ksinit 5 8.2 0.45 95% NO
140
Figure 3.5: Average saturated hydraulic conductivity for Treatment 1 group.
Error bars indicate the standard deviation from the mean. Average initial
saturated hydraulic conductivity was 78  14 cm/h. Average  nal saturated
hydraulic conductivity was 69  28 cm/h.
141
Figure 3.6: Quantile-quantile plot of Treatment 1 change. The straight line
represents a normal distribution. Visual inspection of the data set suggests
that the data Visual inspection of the data set suggests that the data fall
closely enough to the normal line to be treated as normally distributed.
142
3.3.3 Treatment 2
Table 3.12 and Figure3.7 illustrate the test results for the Treatment 2 group. The
average saturated hydraulic conductivity for the group decreased from 71  19 cm/h at
the beginning of the experiment to 61  12 cm/h at the end of the experiment. There
was a high degree of variability in the  nal hydraulic conductivity, though less than that
observed in the Treatment 1 group. Final saturated hydraulic conductivity ranged from
47 cm/h for Column 17 to 80 cm/h for Column 3. Saturated hydraulic conductivity
increased in three of the columns (Columns 3, 4, and 8), and decreased in three of the
columns (Columns 10, 13, and 17).
Table 3.12: Results of saturated hydraulic conductivity tests for Treatment 2 group. Av-
erage values are presented as mean  standard deviation.
Column Initial Ks (cm/h) Final Ks (cm/h) Change (cm/h)
3 50 80 31
4 54 56 2
8 58 63 6
10 82 64 -18
13 92 54 -39
17 87 47 -40
Average 71  19 61  12 -10  28
Examination of the quantile-quantile plot (Figure 3.8) shows that the data does not
deviate far from normality, so the t-test is applicable. Results of the t-test are presented
in Table 3.13. P-value >> 0.05, therefore there is no signi cant di erence between initial
and  nal average Ks for Treatment 2.
Table 3.13: Two-sided paired t-test of Treatment 2.
Null hypothesis degrees of freedom T p-value Con dencelevel Signi cant?
H0 : Ksfin = Ksinit 5 0.9 0.42 95% NO
143
Figure 3.7: Average saturated hydraulic conductivity for Treatment 2 group.
Error bars indicate the standard deviation from the mean. Average initial
saturated hydraulic conductivity was 71  19 cm/h. Average  nal saturated
hydraulic conductivity was 61  12 cm/h.
144
Figure 3.8: Quantile-quantile plot for Treatment 2. The straight line represents
a normal distribution. Visual inspection of the data set suggests that the data
Visual inspection of the data set suggests that the data fall closely enough to
the normal line to be treated as normally distributed.
145
3.3.4 E ect of treatments
The average initial and  nal hydraulic conductivities were compared between treat-
ment groups in order to discern what e ect if any the addition of earthworms may have
had on hydraulic conductivity. Table 3.14 shows the mean and variance of initial Ks for
the treatment groups. Mean initial Ks values were similar for the three groups (Con-
trol: 84 cm/h, Treatment 1: 78 cm/h, Treatment 2: 71 cm/h). One-way analysis of
variance (ANOVA) was performed to compare the initial values between treatments. As
ANOVA requires that the treatment groups have homogeneity variances, Levine?s test
was performed to ensure the homogeneity of variances, and therefore the applicability of
the ANOVA test. Results of Levine?s test are presented in Table 3.15. The dataset fails
Levine?s test, therefore the variances are homogeneous, and the ANOVA test is applicable.
Results of the ANOVA test are presented in Table 3.16. We fail to reject the
null hypothesis. Therefore, there is no signi cant di erence among the initial saturated
hydraulic conductivities of the three treatment groups. This suggests that the inclusion of
a 1 cm arti cial sediment layer in Treatment 2 had no measurable e ect on the saturated
hydraulic conductivity of the soil columns in this treatment group.
Table 3.14: Mean and variance of initial saturated hydraulic conductivity for each of the
treatment groups.
Treatment Mean Ks (cm/h) Variance (cm2/h2)
Control 84 130
Treatment 1 78 200
Treatment 2 71 350
Table 3.17 shows the mean and variance of  nal Ks for the treatment groups. Mean
 nal Ks values were similar for the three groups (Control: 60 cm/h, Treatment 1: 69
cm/h, Treatment 2: 61 cm/h). One-way analysis of variance (ANOVA) was performed
146
Table 3.15: Levine?s test for homogeneity of variance for initial saturated hydraulic con-
ductivity.
Nullhypothesis Degreesof freedom F-value p-value Con dencelevel Signi cant?
H0 :  C = T1 = T2 2 2.5 0.12 95% NO
Table 3.16: Analysis of variance for initial saturated hydraulic conductivity.
Nullhypothesis Sumof Squares F-value p-value Con dencelevel Signi cant?
H0 : KsC =KsT1 =KsT2 595.2 1.3 0.31 95% NO
to compare the initial values between treatments. Levine?s test was performed to ensure
the homogeneity of variances, and therefore the applicability of the ANOVA test. Results
of Levine?s test are presented in Table 3.18. Visual inspection of the variance in  nal
Ks (Table 3.17) suggest that  nal variances may not be homogeneous, but Levine?s test
yields a p-value of 0.14, which is greater than 0.05. The dataset fails Levine?s test, therefore
variances are homogeneous, and the ANOVA test is applicable.
Results of the ANOVA test are presented in Table 3.19. We fail to reject the null
hypothesis. There is no signi cant di erence in  nal saturated hydraulic conductivities
between the three treatment groups. Therefore, the presence of earthworms had no mea-
surable e ect on the saturated hydraulic conductivities of the soil columns.
Table 3.17: Mean and variance of  nal saturated hydraulic conductivity for each of the
treatment groups.
Treatment Mean Ks (cm/h) Variance (cm2/h2)
Control 60 120
Treatment 1 69 800
Treatment 2 61 130
147
Table 3.18: Levine?s test for homogeneity of variance for  nal saturated hydraulic conduc-
tivity.
Nullhypothesis Degreesof freedom F-value p-value Con dencelevel Signi cant?
H0 :  C = T1 = T2 2 2.3 0.14 95% NO
Table 3.19: Analysis of variance for  nal saturated hydraulic conductivity.
Nullhypothesis Sumof Squares F-value p-value Con dencelevel Signi cant?
H0 : KsC =KsT1 =KsT2 303.4 0.42 0.66 95% NO
3.3.5 Earthworm survivorship
After the completion of the  nal in ltration tests, the soil columns were disassem-
bled, and earthworms were removed from the soil and tallied. Large worms were removed
by sieving the soil sample through a 2 mm (#10) sieve. Hatchlings and cocoons were
removed by wet sieving through a 0.6 mm (#30) sieve. Detailed tallies of the number and
size of each of the earthworm types are included in Appendix F. Summaries are presented
in Figures 3.9 and 3.10.
148
Figure 3.9: Changes in earthworm size distribution over time for Treatment 1 group.
149
Figure 3.10: Changes in earthworm size distribution over time for Treatment 2 group.
150
3.4 Discussion
The only group to show a signi cant decrease in saturated hydraulic conductivity
over the course of the experiment was the Control group, which contained no earthworms.
In other words, the Control group clogged. On average, the columns with earthworms
did not clog, and were able to maintain their saturated hydraulic conductivities over the
course of the experiment. There was no signi cant di erence in  nal saturated hydraulic
conductivities between treatment groups, however. This may be explained in part by
the high degree of variability in saturated hydraulic conductivity between columns where
earthworms were present. Variance increased over the course of the experiment for the
Treatment 1 group, but decreased for the Treatment 2 group, and stayed approximately
the same for the Control group (Table 3.20). The initial and  nal Ks for all treatment
groups is shown in Figure 3.11.
Table 3.20: Comparison of variance in initial and  nal saturated hydraulic conductivity
among treatment groups.
Treatment Variance in initial Ks (cm2/h2) Variance in  nal Ks (cm2/h2)
Control 130 120
Treatment 1 200 800
Treatment 2 350 130
Some evidence of pore instability in the treated columns was found. The in ltra-
tion rate was observed to decrease over the course of the six-hour in ltration test for
all columns, whether or not they contained earthworms (see Figure 3.12, Figure 3.13,
and Figure 3.14), but the decreases were more pronounced in columns containing earth-
worms. The gradual decrease in in ltration rate observed in all columns has been noted
by other researchers (Thompson et al., 2008), who found that sandy engineered biore-
tention soils were susceptible to compaction as a result of wetting and the weight of the
151
Figure 3.11: Comparison of initial and  nal Ks for all treatment groups. The
error bars represent the standard deviation from the mean.
ponded water occurring during in ltration tests. The same e ects would likely be seen
during storm events, where the soil became saturated and ponded conditions developed.
Indeed, Dougherty et al. (2007) observed evidence of compaction over the  rst six months
of operation in two newly constructed rain gardens. Hillel (1998) notes that soil in ltra-
bility generally decreases monotonically as water  ows through a soil. In addition to the
expected decrease due to the development of saturated conditions (resulting in a decreas-
ing matric suction gradient), decreasing in ltrability is also sometimes attributable to the
gradual deterioration of the soil structure (Hillel, 1998).
Final saturated hydraulic conductivities for all columns in the Control group were
lower than initial saturated hydraulic conductivities. Final saturated hydraulic conductiv-
ities for treatment columns were more variable. Many soil columns containing earthworms
showed initially high in ltration rates, which decreased dramatically over the course of
the six-hour in ltration tests (Figure 3.13 and Figure 3.14). This is likely due to the grad-
ual disintegration of surface aggregates and the collapse of biopores. The destruction of
152
earthworm casts on the soil surface was observed during the simulated storms, but casts
quickly reappeared after storms, indicating the presence of active earthworms.
Biopores and aggregates created by earthworms may have been unstable due to
several factors. Three key factors have been identi ed:
1. low clay content of the BSM
2. lack of fungal hyphae (oversimpli cation of the ecosystem), and
3. persistent wet conditions.
The BSM contained 93.2% sand, 4% silt, and only 2.8% clay (Table 3.2). Clay plays
an important role in stabilizing pores and aggregates within the soil matrix. Chemical
bonds between clay and organic matter are critical to binding soil aggregates together
(Schrader and Zhang, 1997). According to Oades (1993), a soil must have at least 15%
clay in order to form aggregates, in the absence of signi cant biological activity. A soil
with so little clay will not be able to retain any structure created by earthworms.
In an e ort to isolate the e ect of earthworms on soil structure, all other biological
actors were excluded from the soil columns. In sandy soils, aggregates are stabilized
by plant roots, fungal hyphae, microbial colonies, and the metabolic products of plant
decay (Forster, 1990). The microcosms lacked these components. Fungal hyphae play
a particularly prominent role in aggregation of sand particles, which are only weakly
joined by microbial colonies or their metabolic products (Six et al., 2004). Hyphae grow
and form a sticky mesh around aggregates, protecting soil structure (Oades, 1993). This
highlights the importance of the entire soil ecosystem in the creation and persistence of
soil macropores. Earthworms can create macropores, but without the stabilizing e ect of
fungal hyphae, the macropores quickly collapse. Fungal hyphae are also responsible for
stabilizing earthworm casts, which are weak when freshly deposited, but strengthen after
153
a few days, as they are colonized by fungal hyphae (Lee and Foster, 1991). The columns
had relatively low organic matter, and lacked bacteria, fungal hyphae or plant roots, which
would have stabilized the soil structure created by the earthworms.
The stability of earthworm casts is increased by drying (Oades, 1993). Over the
course of the experiment, the columns were subjected to frequent storm events, with
little opportunity for drying between storms. Drying was further impeded by the small
surface area of the columns, combined with the arti cial moisture barrier of the columns
themselves. In a real rain garden, moisture would wick away from the rain garden into the
surrounding soil. Water uptake by plants and higher evaporation due to increased sunlight
and air movement would also increase drying in a  eld setting. Several studies have shown
that fresh, moist casts are less stable than uningested soil (Barois et al., 1993; Schrader
and Zhang, 1997). Casts need to dry and age in order to stabilize. In our arti cially wet
system, the aggregates never had the opportunity to dry out. Without clay, fungal hyphae,
roots, or the opportunity to dry out, the casts and burrows formed by the earthworms
were likely very unstable.
154
Column 1
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 7
0
20
40
60
80
100
120
140
160
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 5
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 11
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 15
0
20
40
60
80
100
120
140
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 16
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Figure 3.12: Results of initial and  nal in ltration tests for Control group
showing declining in ltration rates over the course of the tests.
155
Column 18
0
20
40
60
80
100
120
140
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 14
0
20
40
60
80
100
120
140
160
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 12
0
20
40
60
80
100
120
140
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 6
0
20
40
60
80
100
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 2
0
40
80
120
160
200
240
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 9
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Figure 3.13: Results of initial and  nal in ltration tests for Treatment 1 group
showing declining in ltration rates over the course of the tests.
156
Column 4
0
20
40
60
80
100
120
140
160
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 3
0
20
40
60
80
100
120
140
160
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 8
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 10
0
20
40
60
80
100
120
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 13
0
20
40
60
80
100
120
140
160
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Column 17
0
20
40
60
80
100
120
140
0 2 4 6
Hour
Infiltration Rate (cm/h)
Initial
Final
Figure 3.14: Results of initial and  nal in ltration tests for Treatment 2 group
showing decreasing in ltration rates over the course of the tests.
157
3.4.1 Sources of error
Three major sources of error have been identi ed: (1) earthworm mortality, (2)
error introduced by the initial saturated hydraulic conductivity testing schedule, and (3)
insu ciently large sample size. Earthworm mortality may have been a major factor in
the high variance observed between treatment columns. The  nal assessment of earth-
worm survivorship at the end of the experiment showed dramatic changes in the columns
earthworm populations over the course of the experiment. Figure 3.15 compares the num-
ber of surviving earthworms to the  nal Ks for each column, but  nds that there is no
correlation between the two (R2 = 0). The majority of the earthworms in the columns
were immature, and quite small in size. Very small earthworms would not be expected to
have any e ect on hydraulic conductivity, as immature earthworms have little e ect on
macroporosity (Francis and Fraser, 1998).
Figure 3.16 compares the number of surviving adult L. terrestris to the  nal Ks,
and  nds that there is an inverse relationship (R2 = 0.4). L. terrestris tend to construct
large, deep burrows, which may have proved to be unstable, as discussed above. Another
possibility is that the presence of these earthworms in their burrows impeded  ow through
these large biopores during storm events. These results suggest that under these experi-
mental conditions, earthworm burrowing was as often bad for in ltration as it was good.
The burrowing activity of the earthworms may have actually caused compaction in some
cases. Depending on a multitude of factors which are not yet completely understood,
earthworms can act as either compacting or decompacting agents in the soil system (Six
et al., 2004).
Due to equipment constraints, initial in ltration tests were performed on one column
at a time. The columns were constructed, and earthworms were added, on November 4,
158
Figure 3.15: Final Ks versus total number of earthworms.
2005. The initial in ltration tests were conducted between November 6 and December 14.
Final in ltration tests were performed on two columns at a time. The delay in sampling the
columns in the Control and Treatment 1 groups had no e ect on their measured saturated
hydraulic conductivity (Figure 3.17, Figure 3.18). However, there is a signi cant increase
in the saturated hydraulic conductivity of the Treatment 2 columns over time (R2 = 0.70,
as shown in Figure 3.19. This is likely due to the e ect of the introduced earthworms
on the sediment layer. Indeed, these  gures show that the  rst measurement taken from
the Treatment 2 group is much lower than the Control and Treatment 1 groups, but
that the  nal measurement taken from the Treatment 2 group is similar to the Control
and Treatment 1 groups. It appears that the earthworms began to substantially alter
the sediment layer in the Treatment 2 columns immediately after they were introduced,
even before the initial measurements were taken. Therefore, the average initial Ks for
159
Figure 3.16: Final Ks versus number of surviving adult L. terrestris.
the Treatment 2 group does not accurately re ect the Ks of the columns as they were
constructed, with a 1 cm sediment layer on top of the BSM. The e ect of delay on the
measured Ks for the  nal in ltration tests is harder to interpret. Figures 3.20 to 3.22 show
an increase in Ks over time for the Control group, no change over time for the Treatment
1 group, and decreasing Ks over time for the Treatment 2 group. No underlying cause for
these trends could be identi ed.
160
Figure 3.17: Initial Ks of Control group versus sampling date.
Figure 3.18: Initial Ks of Treatment 1 group versus sampling date.
Figure 3.19: Initial Ks of Treatment 2 group versus sampling date.
161
Figure 3.20: Final Ks of Control group versus sampling date.
Figure 3.21: Final Ks of Treatment 1 group versus sampling date.
Figure 3.22: Final Ks of Treatment 2 group versus sampling date.
162
The analysis is signi cantly limited by the small number of replicates in each of the
treatment groups. Given the measured variance, the sample size required to have 95%
con dence that the measured sample mean equals the true population mean plus or minus
5 cm/h is calculated by applying the formula:
n = (1:96)
2 2
52 (3.6)
(Ott and Longnecker, 2001), where  2 is the highest measured variance. The variance
in the measured  nal Ks for the Treatment 1 group was 800 cm2/h2. Each treatment
would need to have 123 replicates in order to compensate for this high variance. In this
experiment, each treatment had only six replicates. This complicates interpretation of
the results, because (a) there is signi cant potential for error in the prediction of the true
population mean, and (b) it is di cult to determine whether or not the data sets are truly
normally distributed, and thus which statistical tests are appropriate.
3.4.2 Comparison to Li and Davis (2008b)
Li and Davis (2008b) conducted a column study similar in many respects to this one,
but did not include earthworms in their experiment. The purpose of their study was to
determine what potential incoming sediment had to clog bioretention media, and whether
sediment would be deposited at the soil surface or deeper in the soil matrix. Li and Davis
used bioretention soil media of very similar composition to the medium used in this study,
except that their medium used mulch to provide organic matter, while this study used a
combination of mulch and compost. The particle size distributions of the media are nearly
identical (Table 3.21). Several storm events were simulated using a series of suspensions
of varying particle size distributions. The \sand" suspension was closest to the SS used
in this study (Table 3.22). The resulting changes in the columns? hydraulic conductivity
163
were compared. Filter cake thickness was measured as well. The researchers found that
incoming sediment was primarily deposited on the soil surface, forming a cake. This cake
signi cantly reduced the  ow rate of water through the soil column. The researchers
investigated the e ectiveness of removal of the cake and uppermost soil layer in restoring
the columns hydraulic conductivity,  nding that removal could restore some but not all
of the original hydraulic conductivity, and that the bene t of increasing the depth of
soil removal was marginal. Table 3.23 compares Li and Davis? results to the results of the
current study. Li and Davis observed a signi cant loss in hydraulic conductivity as a result
of solids loading, as did the Control group in this experiment. The columns containing
earthworms did not experience signi cant clogging. This suggests the potential utility of
earthworms in maintaining the hydraulic conductivity of bioretention cells.
Table 3.21: Comparison of bioretention soil medium used in this study to Li and Davis
(2008b).
Characteristic Ayers Li and Davis (2008b)
Sand (%) 93.2 96 - >98
Silt (%) 4.0 <2{4
Clay (%) 2.8 <2{4
Soil texture Sand Sand
Table 3.22: Comparison of simulated suspended solids used in this study to Li and Davis
(2008b).
Characteristic Ayers Li and Davis (2008b)
Sand (%) 78.5{88.4 >99
Silt (%) 6.5{13.1 <1
Clay (%) 5.1{8.4 <1
Soil texture Loamy Sand Sand
164
Table 3.23: Comparison of experimental results with Li and Davis (2008b). Average values
are reported as mean  standard deviation.
Control Treatment 1 Treatment 2 Li and Davis (2008b)
initial hydraulic
conductivity
(cm/h)
84  12 78  14 71  19 55
 nal hydraulic
conductivity
(cm/h)
60  11 69  28 61  12 7
total solids
loading (kg/m2)
20.2 20.2 38.6 6.4
TSS texture Loamy Sand Loamy Sand Loamy Sand Sand
TSS loading rate
(mg/L)
1,299{2,222 1,299{2,222 1,299{2,222 35{1,729
165
3.5 Summary
1. Soil columns without earthworms showed a signi cant decrease in saturated hy-
draulic conductivity over the course of the experiment.
2. Soil columns with earthworms did not show a signi cant decrease in saturated hy-
draulic conductivity.
3. Earthworm activity was not signi cantly impeded by the inclusion of an initial sed-
iment cake at the soil surface.
4. The saturated hydraulic conductivities of soil columns with earthworms had a high
variance, which may be attributable to pore instability.
5. High pore instability may be attributable to the absence of plant roots and fungal
hyphae in the soil columns.
6. High earthworm mortality was observed in the soil columns.
7. Earthworm populations showed a shift from adults to hatchlings over the course of
the experiment.
166
Chapter 4
Model
4.1 Introduction
The RAINGARDEN model was developed in order to study the in uence of earth-
worm growth and soil organic matter on the hydraulic performance of a rain garden.
Hydraulic performance is assessed in terms of soil porosity, which limits the stormwater
detention capacity of the soil, and by the thickness of a surface sediment layer, which
limits the available volume for detention on the rain garden surface. Factors in uencing
the creation, destruction, and stability of pore space are included.
Soil organic matter serves to improve the stability of pore space and as a food
source for earthworms. Earthworms play a dual role, increasing pore space and reducing
the thickness of the sediment layer. Storm events deposit sediment at the soil surface,
and destroy soil pores. Litterfall serves as an external source of organic matter. Emphasis
is given to earthworms because of their well-known roles in maintaining soil structure in
agricultural and other soil types. Earthworms increase soil porosity both by the creation
of macropores through their burrowing and by the aggregation of soil particles through
their metabolic processes (Edwards, 1998; Lee, 1985). Ultimately, soil animals such as
earthworms may be able to be used as elements of design for the ecological engineering of
bioretention cells. An earlier version of the RAINGARDEN model is described in Ayers
and Kangas (2004). The RAINGARDEN model is a conceptual, or synthetic model,
and is intended to begin the description of mechanisms within the soil a ecting system
performance. The model is developed using real values from the literature and other parts
167
of this research, and illustrates the dynamics of soil physical parameters in response to
biological activity. At this point, the model serves as a conceptual illustration of the
processes believed to be at work in rain gardens. It has not been subjected to sensitivity
analysis, calibration or validation.
4.1.1 The energy systems language
The energy systems language (also referred to as energy circuit language) was devel-
oped by H.T. Odum in the 1950s and 60s (Odum, 1971; Brown, 2004). It is a systematic
method of creating diagrammatic models of systems, which is commonly used in the  elds
of ecological engineering and systems ecology. The energy systems language is used to
visualize systems and to describe them mathematically. Every symbol used is mathemat-
ically de ned (Figure 4.1). Drawing a diagram is equivalent to writing a set of di erential
equations. A system is de ned, and the components of interest and the relationships
between them are drawn.
The principle advantage of the energy systems approach is that the act of draw-
ing the diagram yields a model based on fundamental physical, chemical and biological
relationships between the model components. Ecosystems tend to be very complex, and
thus quite di cult to study. The energy systems language forces the modeler to distill the
system to its essentials, choosing only a few key variables and interactions to focus on.
The  rst step in the simulation is to draw a detailed energy systems diagram. A
boundary is de ned, delineating the system of interest. The relevant inputs, state variables
and interactions are identi ed. The diagram then yields a set of di erential equations de-
scribing the system dynamics. These equations are generated using the method described
in Odum (1971).
168
Figure 4.1: Symbols used in the energy systems language.
4.1.2 RAINGARDEN model
Sources:
L = SOM inputs from litter fall (g)
I = water  owing into raingarden (L)
D = sediment carried in by water (g)
Storages:
S = soil mass (g)
O = soil organic matter mass (g)
W = number of earthworms (individuals)
V = void space (L)
H = pore water (L)
169
Figure 4.2: Energy systems diagram of simulation model RAINGARDEN.
170
C = surface crust thickness (cm)
Flows:
J1= k1  S  O  W = SOM consumed by earthworms (g/d)
J2 = k2  S  O  W = worm growth (individuals/d)
J3 = k6  W2 = worm death (individuals/d)
J4 = k3  S  O  W = creation of voids due to earthworm burrowing (L/d)
J5 = k4  HO = loss of void space due to wetting (L/d)
J6 = [IF V >H] I  (1-k  C) = water entering voids (L/d)
J7 = k5  H= water exiting voids (L/d)
J8 = k7  I  D = deposition to crust (cm/d)
J9 = k8  C  W = loss of crust due to earthworm burrowing (cm/d)
J10 = k9  O = loss of organic matter due to microbial degradation (g/d)
K through k8 are rate constants applied to the associated  ows.
Plant litter (L), is the only source of organic matter to the system. In the model,
plant litter adds directly to the organic content (O) of the soil (S). Organic matter is lost
through consumption by earthworms (W) and decay. The earthworm population increases
through consumption of organic matter, and decreases by death. The rate of organic
matter consumption by earthworms is dependent on the earthworm population, and is
therefore represented by a feedback loop. The rate of earthworm death is proportional
to the square of the earthworm population, producing logistic decay. This simpli cation
of the rain garden ecosystem is necessary in order to produce a model of manageable
complexity. In reality, plant litter (L) would create an organic layer at the soil surface,
which would then eventually be decomposed by soil animals and incorporated into the
171
soil matrix, gradually increasing organic matter (O). In addition, earthworms (W) are
assumed to use soil organic matter (O) as a food source. This is also a simpli cation of
a complex process, in which earthworms of di erent ecotypes feed on di erent substances
in di erent regions of the soil. For example, anecic species, such as L. terrestris, ingest
plant litter from the soil surface, but only digest the bacteria living on the litter, not the
plant material itself. Chapter 1 discusses these processes in greater detail.
The primary factor of interest in the model is the ability of the rain garden to store
runo (I) during storm events, and to discharge this stored runo into the groundwater
between storm events. This capacity is a function of the available pore space (V), and
of the thickness of the surface crust (C). In this model, the surface crust limits storage
capacity by decreasing the ponding volume available. The e ects of the surface crust and
changing soil porosity on the rate of stormwater in ltration into the rain garden are not
included in the model. The amount of water that can be held within the soil (H) is limited
to the available pore space, minus the amount of water ex ltrated from the system. The
available pore space is a function of organic matter and earthworm population, and is
limited to the soil volume (S). Pore space is lost as a result of wetting, but this loss is
retarded by the stabilizing e ect of the organic matter. Surface crust grows as a result of
the deposition of suspended solids (D), and is destroyed by bioturbation by earthworms.
The model does not capture all processes at work within the soil, but does illustrate some
key relationships, and highlights the critical role of the soil biology in the maintenance of
the systems hydraulic performance.
The energy systems diagram shown in Figure 4.2 was converted to a set of di erential
equations using the method described in Odum (1971):
172
dO
dt = L k1SOW k9O (4.1)
dW
dt = k2SOW k6W
2 (4.2)
dV
dt = k3SOW k4
H
O (4.3)
dH
dt = IF(I(1 kC) > (V  H +k5H))THEN(V  H +k5H)ELSE(I(1 kC)) (4.4)
dC
dt = k7ID k8CW (4.5)
Where O = soil organic matter content, expressed as total mass,
L = litterfall
S = soil volume,
W = total mass of earthworms in soil,
V = total volume of soil void space (porosity),
H = total volume of water held in soil pores,
C = thickness of surface crust (clogging layer),
I = in ow into raingarden (vol), and
D = suspended solids carried in in ow.
173
A computer model was constructed to solve this set of equations numerically us-
ing STELLATMsoftware (version 5.1.1, High Performance Systems, Inc., Hanover, New
Hampshire). Numerical values for the model inputs,  ows and storages were obtained
from a wide array of sources, including published papers and, where appropriate, values
obtained from the  eld study that is a part of this research. In particular, the soil organic
matter content and earthworm population values were derived directly from the average
values for all of the  eld sites, and the ex ltration rate was derived from the hydraulic
conductivity of the columns in the microcosm experiment. Detailed derivations of these
values are included in Appendix G. The values are given in Table 4.1. These values were
then used to calculate the coe cients in Equations 4.1 to 4.5. These are given in Table
4.2. For the purpose of the calculation of coe cients, an assumption was made that the
model was at a steady state in which all state variables remained constant. Each model
run was continued for 1500 days, with the model iterating once per day. The initial values
for the model runs were the values listed in Table 4.1 except where otherwise noted. A
runo hydrograph (I) was derived from precipitation data taken from a NOAA weather
station in Beltsville, MD (NCDC, 2008) using the SCS Method (McCuen, 2005). The
calculation is included in Appendix G.
174
Table 4.1: Numerical values for the inputs, state variables, and  ows in the RAINGAR-
DEN model. Detailed derivations appear in Appendix G.
Parameter Description Value Source
L SOM inputs due to
litter fall and root decay
2.05 g/m2/d steady state assumption
I water  owing into rain
garden
0.162 m3 (McCuen, 2005)
D sediment carried in by
water
8.9 x 10 3 cm/m3 (Li and Davis, 2008b)
S soil volume 1 m3 assumed
O soil organic matter mass 9.15 x 104 g  eld study
W number of earthworms 310.7 individuals  eld study
V void space 0.5 m3 (Brady and Weil, 2002)
H pore water 0.08 m3 (Brady and Weil, 2002)
C surface crust thickness 0.4 cm (Li, 2007)
J1 SOM consumed by
earthworms
2.05 g/d (Binet and Trehen,
1992; Satchell, 1967)
J2 worm growth rate 0.38 individuals/d (Lakhani and Satchell,
1970)
J3 worm death 0.38 individuals/d steady state assumption
J4 creation of voids due to
earthworm burrowing
3.915 x 10 4 m3/d (Bastardie et al., 2003;
Guild, 1955)
J5 loss of void space due to
wetting
3.915 x 10 4 m3/d steady state assumption
J6 water entering voids 0.158 m3/d assumed
J7 water exiting voids 2.45 m3/d microcosm experiment
J8 deposition to crust 3.46 x 10 4 cm/d (Li and Davis, 2008b)
J9 loss of crust due to
earthworm burrowing
3.46 x 10 4 cm/d steady state assumption
J10 rate of microbial decay
of OM
0 g/d assumed
175
Table 4.2: Values of coe cients in model equations.
Coe cient Value units
k 6.60 x 10 2 cm 1
k1 7.21 x 10 8 m 3ind 1d 1
k2 1.34 x 10 8 m 3g 1d 1
k3 1.38 x 10 11 g 1ind 1d 1
k4 447.78 gd 1
k5 30.625 d 1
k6 3.94 x 10 6 ind 1d 1
k7 0.24 m 3d 1
k8 2.78 x 10 6 ind 1d 1
k9 0 d 1
176
4.2 Results and Discussion
4.2.1 Model run 1: organic matter (O) and earthworms (W) held constant
Earthworm population and soil organic matter are held constant in the steady state
scenario. Crust thickness gradually decreases as a result of earthworm activity, despite
intermittent inputs of suspended solids (Figure 4.3). Figure 4.4 shows void space gradually
increasing over time, as earthworms burrow within the soil, and the soil organic matter
preserves this structure. During storm events in which the runo volume exceeds the
empty pore space in the soil, only a portion of the in owing water is stored as pore water
within the void space. The excess can be assumed to be discharged through an over ow.
Figure 4.3: Standard run: crust thickness (C) over 1500 days.
177
Figure 4.4: Standard run: void space (V) and pore water (H) over 1500 days.
178
4.2.2 Model run 2: no earthworms (W = 0)
A model run with the initial earthworm population set to zero illustrates the bene-
 cial e ect of earthworms within the system. Results are presented graphically in Figure
4.5, Figure 4.6, and Figure 4.7. Organic matter increases linearly, as it has a constant
input of plant litter, but no source of decay. Void space is gradually lost in spite of the
added stability provided by increased organic matter, since there are no earthworms to
regenerate lost void space. Crust thickness increases with each storm event. This simu-
lation makes clear the critical role of earthworms in the maintenance of a rain garden?s
hydraulic performance.
Figure 4.5: Soil organic matter (O) in a system without earthworms over 1500 days.
179
Figure 4.6: Pore space (V) and pore water (H) in a system without earthworms
over 1500 days.
Figure 4.7: Crust thickness (C) in a system without earthworms over 1500 days.
180
4.2.3 Model run 3: less cohesive organic matter (increased k4 from 448.35 to 5,000)
Organic matter also plays an important role in the maintenance of soil porosity.
The cohesiveness of organic matter, represented in the model by the inverse of the k4 con-
stant, is primarily responsible for the resistance of void space to deterioration by in owing
stormwater. When k4 is increased, void space rapidly deteriorates (Figure 4.8). This sit-
uation is similar to that observed in the microcosm study of Chapter 3, in which biopores
created by earthworms collapsed as a result of wetting during the  nal in ltration tests.
Figure 4.8: Loss of porosity (V) over 1500 days in a system where organic
matter is less cohesive.
181
4.3 Summary
These model runs illustrate the role that earthworms and soil organic matter may
play in the maintenance of the hydraulic function of the rain garden. In the model,
earthworms prevent the formation of a surface crust that would inhibit in ltration, and
burrow through the soil, creating void space, which is stabilized by soil organic matter.
In the future, the model will be rigorously tested and validated.
182
Chapter 5
Conclusions
This research is the  rst to describe the ecosystems that are present in rain gardens.
Because the designers of rain gardens usually conceive of them primarily as physical and
chemical systems for the retention, in ltration and removal of pollutants from stormwater,
little consideration is given to the biological system. Bacteria are assumed to be important
to the degradation of pollutants, but the role of higher plants and animals has not been
su ciently explored. The  eld study reveals a robust invertebrate community within the
soil. Earthworms were found at all sites, and representatives of twelve other invertebrate
taxa were observed.
Rain gardens were found to undergo pedogenesis, changing their soil structure and
composition in response to increasing biological activity. In the  rst years after construc-
tion, the uppermost soil layer begins to become enriched with organic matter, plant roots,
and soil organisms. The soil pro le develops a characteristic pattern of decreasing bio-
logical activity with depth. This pattern is an emergent property of rain garden soils,
resulting from the cumulative e ects of lower level processes, such as the metabolic activ-
ities of earthworms and other invertebrates. In this way, rain gardens self-organize from
a planted depression  lled with a simple mixture of topsoil, sand and mulch or compost
into a complex soil ecosystem.
The  eld sites surveyed showed no signs of clogging due to the trapping of suspended
solids carried in stormwater runo . Some evidence was found of higher than expected
183
in ltration rates at the  eld sites, which may be attributable to the e ects of bioturbation
and of the formation of a granular soil structure.
Earthworms act to regenerate soil porosity and bury deposited sediment between
storms. This \ecological service" is crucial to the sustained performance of a rain garden
over its operational life. It is critical, however, to recognize that earthworms do not and
cannot act in isolation. Many di erent components of the rain garden ecosystem work
together to sustain the earthworm population and to stabilize the biopores and aggregates
that earthworms create. The microcosm experiment found that earthworms were able
to maintain the in ltration rate of a soil column subjected to simulated storms over a
six-month period. Soil columns lacking earthworms showed decreased in ltration rates
after six months. However, the variance of the  nal in ltration rate measurement of
the treatment groups was much higher than that of the control group, suggesting that
the e ects of earthworms operating in isolation could be unpredictable. Without fungal
hyphae or plant roots to stabilize the structure created by the earthworms, pores were
vulnerable to collapse during storm events.
Shedding light on the complex web of relationships between a rain garden?s physical
and biological systems is the central theme of this study. Some, though by no means all,
of these relationships were identi ed in Figure 1.3. The simulation model explores the
relationships between  ve of these factors in detail (see Figure 5.1). Data taken from the
 eld and microcosm studies were used in conjunction with data from the literature to
create a simpli ed but realistic representation of the way in which earthworms and soil
organic matter work together to maintain a rain garden?s soil structure. In the model,
litterfall serves as the direct source of organic matter to the soil. Organic matter acts
both as a food source for earthworms and as a stabilizer, limiting the destructive e ect of
184
storms on soil porosity. Earthworms increase soil porosity and decrease the thickness of the
surface crust of solids deposited during storm events. Model runs illustrate the importance
of earthworms and organic matter in the prevention of clogging in rain gardens.
Li and Davis (2008b) conducted a  eld observation of a rain garden in Washington
DC. They measured the total suspended solids (TSS) loading to the rain garden over
several storm events, and calculated that the rain garden would likely clog after 1 to 2
years. However, while the researchers observed a small surface layer of sediment at the
entrance zone of the rain garden, they did not observe widespread clogging of the media
during their 1.5-year monitoring period. The authors speculate that the vegetation and
fauna present at the site may play a role in preventing the development of a sediment
layer at the soil surface. Similar conditions were observed at the rain gardens evaluated
in the current study. The only site observed to have a surface sediment layer was UMCP,
the youngest. This layer was not produced by sediment carried in by runo , but rather by
deposition during one very large storm event, in which a creek adjacent to the rain garden
overtopped the facility, depositing a large quantity of sediment. The soil surfaces of the
other nine sites did not have apparent surface sediment layers. The surfaces of most were
covered with a layer of mulch or plant litter. In the rare instances where the surface of the
soil was bare, there was no pronounced di erence in soil texture between the soil surface
and deeper soil layers. Thus, this study lends support to Li and Davis suggestion that
biological activity in rain gardens may play a critical role in the prevention of clogging by
incoming sediment.
In summary, this study  nds that:
1. Biological activity is ubiquitous in rain gardens.
185
2. Rain gardens develop a characteristic soil pro le of exponentially decreasing biolog-
ical activity with depth.
3. No evidence of clogging due to normal operation was found.
4. Earthworms may prevent clogging caused by trapped solids.
186
Figure
5.1:
Factors
in uencin
gthe
dra
wdo
wn
time
of
arain
garden,
with
emphasis
on
those
factors
included
in
the
sim
ulation
mo
del.
187
5.1 Implications for Rain Garden Design
This study has yielded an improved understanding of and appreciation for the living
organisms present in rain gardens. This new knowledge gives rise to several recommenda-
tions for rain garden design and maintenance. These recommendations will improve the
value of rain gardens as habitat for desirable organisms, and exploit the innate abilities of
the natural system to improve rain garden performance.
5.1.1 Design with self-organization in mind
Rain gardens are living systems, and their capacity to self-organize in response to
external forces must be taken into account in their design. Designers often fail to anticipate
the changes that will be brought about by the activities of the organisms that will colonize
the system. These changes include bioturbation and the enrichment of the soil with
organic matter. The trajectory of pedogenesis within rain gardens is of inherent interest
to engineers, who must anticipate how structures they design will perform throughout their
operational lives. Changes in the soil structure and composition may impact in ltration
rates and pollutant removal performance. For example, the gradual formation of a granular
soil structure in the topsoil may lead to increasing in ltration rates. The enrichment of the
soil with organic matter may improve pollutant removal. On the other hand, earthworm
burrowing and plant root growth may create undesirable preferential  ow paths that
hamper the pollutant removal performance of the rain garden.
5.1.2 Limit pesticide use in the rain garden and its watershed
Earthworms are very susceptible to pesticides, as are other soil animals. These an-
imals are critical to the soil development. Each soil organism plays a role in the complex
188
conversion of plant detritus into humus. Soil animals colonizing the rain garden are valu-
able, and care must be taken to avoid harming them. Pesticide use in the rain garden
and its watershed should be strictly controlled. Nonetheless, on occasion a particularly
destructive \pest" will invade the system, and may destroy more rain garden plants than
is desirable. In such a case, it would be preferable to use a targeted approach, such as
Integrated Pest Management, rather than a broad-spectrum insecticide, in order to limit
the impact on more desirable species.
5.1.3 Avoid disturbance of topsoil and mulch
The  eld surveys revealed that the majority of the biological activity occurring in the
rain garden occurs at the soil surface. This is where the soil organisms break down the litter
layer at the soil surface and incorporate it into the soil matrix. If a rain garden becomes
clogged, the temptation will be to remove the mulch and upper soil layers. Disturbance
of these layers will destroy the soil ecosystem, along with its bene cial e ects. All of the
soil structure that has developed, as well as all of the highly decomposed humic material
with which the soil has been enriched, and which has a high cation exchange capacity, and
therefore very high a nity for most pollutants found in stormwater, is removed. Therefore
care should be taken to avoid disturbance of the soil unless absolutely necessary. Mulch
may be added, but decayed mulch should not be removed. The same natural processes
that create and maintain soil porosity in natural soils are at work in rain garden soils.
With proper management, these processes can be exploited to maintain the in ltration
rate of the rain garden over the long term.
189
5.1.4 Limit the sand content of the Bioretention Soil Medium (BSM)
In designing bioretention cells, it is tempting to use a very high sand content, in
order to maximize the in ltration rate of the soil mix. The need to balance the intrinsic
permeability of sand with the superior pollutant removal capacity of clay and organic
matter has been recognized by other researchers (Hsieh and Davis, 2005a). Very sandy soils
create additional problems for rain garden biota. Sand does not hold water or nutrients
well, and can create an inhospitable environment for rain garden plants and soil animals.
In particular, earthworms can  nd sand to be a di cult medium in which to thrive. The
microcosm experiment revealed another potential pitfall of using too much sand: sand does
not hold a macropore structure very well, which limits the in ltration rate of the soil to
the permeability of the medium. Earthworms and plant roots may create pores, but in an
unstable soil, this structure will be destroyed every time it rains. In fact, the microcosm
experiment results suggest that burrowing of anecic species in very sandy soils might
actually promote compaction, as their burrows may be unstable if the soil lacks su cient
clay. There is a potential for these burrows to be stabilized by a well-developed network of
fungal hyphae, but further research will need to be done to support this hypothesis. The
microcosm experiment suggests that the sand content of the current speci cation may be
too high to take full advantage of the macropore-creating ability of resident earthworms.
It is interesting to note that all of the sites surveyed appeared to be in good working
order, though the in ltration rate was measured directly at only three of ten sites. The
soil texture at these sites varied greatly. If further investigation reveals that the older
sites, which contain much higher clay contents, function as well as the newer, sandier rain
gardens, then this would suggest that the soil structure created by the soil organisms may
indeed play a critical role in the promotion of in ltration in rain gardens. If a robust soil
190
ecosystem can create a stable, granular soil structure that in ltrates well, then this may
permit use of a soil with a higher clay content, which may enhance pollutant removal.
5.1.5 Consider seeding rain gardens with earthworms
Given the importance of earthworms to the maintenance of hydraulic conductivity
in rain gardens, it may be worthwhile to consider introducing earthworms into the system
during construction. However, the appropriate earthworm species are not easily available
commercially. Earthworms used in vermiculture, Eisenia fetida, are a surface dwelling,
endogeic species, and do not burrow into the soil. They are therefore not appropriate to
this application. Earthworms used as  shing bait are relatively easy to obtain in season,
but are of mixed species. Canadian night crawlers, Lumbricus terrestris, are harvested in
the wild. European night crawlers, Eisenia hortensis, and African night crawlers, Eudrilus
eugeniae, are farmed, and are becoming increasingly popular as  shing bait. All three are
non-native species, though Lumbricus terrestris is ubiquitous in many parts of the United
States. It is generally best to avoid introducing non-native species where possible to avoid
unintentionally introducing invasive pests.
A better approach might be to inoculate a newly constructed rain garden with a
small amount of topsoil collected from a nearby habitat. This soil would contain represen-
tatives of the local earthworm population, as well as other invertebrates, bacteria, fungi,
and the seeds of local plants, all of which are bene cial to the rain garden. The soil should
be collected from a habitat similar to a rain garden, such as a meadow or abandoned lot,
or from another rain garden in the vicinity. The soil should be transferred immediately
from the source habitat to the rain garden, to minimize invertebrate mortality due to
191
dessication, and should be installed intact. A small amount of soil, perhaps one to three
cubic feet, would likely be su cient to inoculate a typical rain garden.
5.1.6 Consider seeding rain gardens with mycorrhizal fungi
This study theorizes that mycorrhizal fungi play a critical role in the stabilization of
soil pores and aggregates. They form a  ne, sticky mesh that spreads throughout the soil,
holding soil particles in place. This role has not yet been observed or quanti ed in rain
gardens, but is well known in other soil systems. Mycorrhizal fungi inoculants are available
commercially from a number of mail-order suppliers, and are commonly used to improve
plant growth. Inoculating rain gardens with mycorrhizal fungi during construction may
improve soil permeability by stabilizing the soil structure, and improve plant establishment
at the same time.
5.1.7 Plant densely
A wide variety of planting schemes have been employed in rain gardens. Some are
planted with a dense cover of native, fast-growing annuals, and others are planted in a
more ornamental fashion, with isolated shrubs surrounded by mulch. Plant roots create
biopores, oxygenate the soil, provide a food source, exude sticky substances that bind soil
aggregates, and provide microsites for bacterial degradation of pollutants. Therefore, rain
gardens can be expected to perform best when densely rather than sparsely planted. In
addition, densely planted native annuals will ensure a steady supply of plant litter, the
primary food source for the soil ecosystem.
192
5.1.8 Mow to encourage high productivity
Over time, rain gardens can be expected to undergo succession, where fast-growing
annuals are gradually overtaken by slower-growing perennials. Fast-growing plants o er a
number of advantages to the rain garden designer. Their high productivity means higher
root growth, higher plant litter deposition, and faster take-up of nutrients carried into the
rain garden in stormwater runo . In order to maintain this early successional stage, rain
gardens should be mowed periodically. Mowing will kill slow growing perennials, and will
stimulate the growth of fast growing annuals.
5.2 Suggestions for future research
This research represents a major step in opening the "black box" which is the rain
garden ecosystem. Completing this picture, and gaining an understanding of precisely
how this ecosystem a ects rain garden performance are tasks that must be left to future
researchers. Biological activity can be expected to a ect both hydrologic function and
pollutant removal. The e ects of management strategies, such as mowing, pesticide ap-
plication, and mulch replacement, on the rain garden ecosystem should be explored. The
types of plants used in rain gardens may a ect their performance. Planting the rain garden
with highly productive early successional species may help to establish a highly produc-
tive ecosystem, in which take-up of nutrients and creation of biopores and soil organic
matter are maximized. Some of the rain gardens installed at the University of Maryland
are instrumented in such a way as to allow detailed analysis of hydraulic and pollutant
removal performance, and could be useful in assessing the a ects of biological activity.
Comprehensive measurements of the in ltration rates of a great number of estab-
lished rain gardens would be of great value in understanding their performance over the
193
long term. The instrument most commonly used to measure in ltration, the double ring
in ltrometer, is not well-suited for use at these sites. The double ring in ltrometer is
designed for use on agricultural  elds, where thick roots are rare. Rain gardens are full of
thick roots, especially when they are planted with shrubs and perennials. The use of an
alternative technique not requiring disturbance of the soil might prove more successful.
The role of fungal hyphae in the stabilization of soil structure should be explored.
The microcosm experiment could be repeated, this time including fungi.
Design experiments based on the variance measured in this study. When this re-
search was initiated, very little was known about the variability of di erent parameters
within or among rain gardens. Now that some of these values have been measured, it is
possible to calculate the number of samples and/or replicates that would be required to
obtain statistically signi cant measurements.
Examine the potential negative impacts of the use of sodium chloride deicers in
the rain garden watershed. Sodium chloride is known negative impacts on terrestrial and
freshwater ecosystems (Ramakrishna and Viraraghavan, 2005). The capture of sodium
ions in the rain garden soil may have destructive e ects, including the dispersal of soil
aggregates, and toxicity to plants and animals.
Most of the testing of rain gardens has been performed on the east coast of the
United States. Rain garden performance should be tested in other areas of the country
and the world. Di erences in soil properties, particularly the presence of expansive clays,
may require design modi cations.
A note of caution. When exploring the role of ecological factors in the performance
of a system, it is tempting to take a reductionist approach, attempting to measure the
impact of each component individually. Scientists and engineers seek linear, causal rela-
194
tionships that are quanti able and predictable. Unfortunately, this approach often fails
when applied to ecological systems, where components interact synergistically, and the
whole is something more than the sum of its parts. Even this study fell into this trap
in attempting to isolate the e ect of earthworms on in ltration rates in the microcosm
study described in Chapter 3. In the microcosm study, the earthworms created pores,
but lacked the support system they would have had in a natural ecosystem. In a natural
soil, the pores would have been stabilized by fungal hyphae and the soil would have been
glued together by organic matter. A more robust microcosm experiment would contain a
soil ecosystem of su cient complexity to re ect the behavior of the system. This is part
of a major question in ecology, namely how to capture the richness and complexity of an
ecosystem in a controlled laboratory setting.
Engineering a living system is a challenging task, requiring an understanding of both
the physical requirements to be met and the ways in which the system will behave and
evolve over time. This study is an attempt to integrate ecology and soil science with civil
engineering. Rain gardens are a new kind of ecosystem, and are not yet well understood.
Revealing the functioning of the rain garden ecosystem will necessarily be an iterative
process, with each new discovery raising new questions.
195
Appendix A
Additional Background on Field Sites
A.1 University of Maryland (UMCP)
Address: University of Maryland, College Park, MD 20742
Data source: UMCP Department of Facilities Planning
Table A.1: UMCP Bioretention Speci cation
Bioretention soil mix
50% sand
30% topsoil, sandy loam or  ner
20% shredded 2x hardwood mulch
mulch depth 3 inches
underdrains 6" perforated pipe
cell dimensions 3,000 sf
drainage area not available
ponding depth not available
soil depth not available
Table A.2: UMCP Planting Plan
Trees Betula nigra (river birch)
Shrubs Aronia arbutifolia (red chokeberry)
Herbaceous
Eupatorium  stulosum (joe pye weed)
Lobelia siphilitica ?blue selection? (blue lobelia)
Aster novae-angeliae (New England aster)
Solidago rugosa (goldenrod)
196
Figure A.1: UMCP construction detail (Source: UMCP Department of Facil-
ities Planning).
197
A.1.1 Washington Navy Yard (WNY)
Address: 1014 N Street, SE, Washington, DC 20374
Data Source: The Low Impact Development Center Inc.
Table A.3: WNY Bioretention Speci cation
Bioretention soil mix
50% sand, conforming to ASTM c144-97
30% topsoil conforming to ASTM D5268-92,
modi ed, clay content less than 10%, organic
content less than 20%
20% shredded 2x hardwood mulch
soil depth not available
ponding depth not available
mulch depth 3 inches
underdrains 4" perforated pipe
cell dimensions 225 sf
drainage area 0.32 ac
Table A.4: WNY Planting Plan
Trees not available
Shrubs not available
Herbaceous not available
198
Figure A.2: WNY construction detail (Source: The Low Impact Development
Center Inc.).
199
A.1.2 Mary Harris \Mother" Jones Elementary School (MJES)
Address: 2405 Tecumseh Street, Adelphi, MD 20783
Data Source: Prince George?s County Department of Environmental Resources
(DER) (PGCO, 2004b)
Table A.5: MJES Bioretention Speci cation
Bioretention soil mix
50% sand
25% topsoil
25% compost
soil depth 1.5 ft, with 1.5 ft stone retention area underneath
ponding depth 6 inches
mulch depth 3 inches
underdrains 6" perforated pipe
cell dimensions 800 sf
drainage area not available
Table A.6: MJES Planting Plan
Trees not available
Shrubs not available
Herbaceous not available
200
Figure A.3: MJES construction detail (PGCO, 2004b).
201
A.1.3 Chesapeake Bay Foundation Headquarters (CBF)
Address: Philip Merrill Environmental Center, 6 Herndon Avenue, Annapolis, MD 21403
No information available
A.1.4 Northwestern High School (NWHS)
Address: 7000 Adelphi Road, Hyattsville, MD 20782
Data Sources: Prince George?s County DER (PGCO, 2004c), construction plans
obtained from Prince George?s County Schools, Hyattsville, MD
Table A.7: NWHS Bioretention Speci cation
Bioretention soil mix
50% sand
25% topsoil (35{60% sand, 30{55% silt, 5{10%
clay)
25% compost
soil depth 2.6 ft
ponding depth 5 inches
mulch depth 3 inches
underdrains 6" perforated pipe
cell area 4040 sf
drainage area 3.17 acres, mostly paved
202
Table A.8: NWHS Planting Plan
Trees
Quercus palustris (pin oak)
Acer rubrum (red maple)
Koelreuteria paniculata (golden rain tree)
Shrubs
Ilex glabra ?Shamrock? (inkberry ?Shamrock?)
Fothergilla gardenii (Fothergilla)
Aesculus parvi ora (Bottlebrush buckeye)
Ilex verticillata ?Sparkleberry? (winterberry ?Sparkleberry?)
Herbaceous
Hemerocallis sp. (Hyperion daylily)
Iris kaempfera (Japanese bearded iris)
Liatris spicata (Blazing star)
Vernonia noveboracensis (New York ironweed)
Lobelia cardinalis (Cardinal  ower)
Saccharrum ravenne (Plume grass)
Narcissus sp. ?Mount Hood? (Mount Hood da odils)
Aster novae-angliae (New England aster)
Verbena hastata (Blue vervain)
Eupatorium dubium (Joe pye weed)
Rudbeckia ?Goldsurm? (Black-eyed susan)
Kosteletzkya virginica (Seashore mallow)
Pennisetum alopecuroides (Fountain grass)
203
Figure A.4: NWHS construction detail (Source: Prince George?s County Schools)
204
A.1.5 Inglewood Center III (PP)
Address: 9400 Peppercorn Place, Upper Marlboro, MD 20774
Data Sources: Prince George?s County DER (PGCO, 2004a)
Table A.9: PP Bioretention Speci cation
Bioretention soil mix 70% sand30% compost
soil depth 2 ft over 1 ft stone retention area
ponding depth 6 inches
mulch depth 3 inches
underdrains not available
cell area 324 sf
drainage area not available
Table A.10: PP Planting Plan
Trees not available
Shrubs not available
Herbaceous not available
205
Figure A.5: PP construction detail (PGCO, 2004a).
206
A.1.6 Claggett Farm (CF)
Address: 11904 Old Marlboro Pike, Upper Marlboro, MD 20772
Data Source: Chesapeake Bay Foundation (CBF, 2005)
Table A.11: CF Bioretention Speci cation
Bioretention soil mix not available
soil depth not available
ponding depth not available
mulch depth not available
underdrains not available
cell area 77 sf
drainage area 1125 sf
Table A.12: CF Planting Plan
Trees not available
Shrubs not available
Herbaceous not available
207
A.1.7 Chevy Chase Bank (CC)
Address: 2315 Randolph Road, Silver Spring, MD 20902
Data Source: Montgomery County Department of Environmental Protection (MCDEP,
2004)
Table A.13: CC Bioretention Speci cation
Bioretention soil mix not available
soil depth 22 inches
ponding depth not available
mulch depth not available
underdrains 6" perforated pipe
cell area 134 sf
drainage area 5,750 sf (4,850 sf impervious)
Table A.14: CC Planting Plan
Trees not available
Shrubs not available
Herbaceous not available
208
Figure A.6: CC construction detail (MCDEP, 2004)
209
A.1.8 Beltway Plaza Mall (BP)
Address: 6000 Greenbelt Road, Greenbelt, MD 20770
Data Source: Beltway Plaza Developers
Table A.15: BP Bioretention Speci cation
Bioretention soil mix 5 parts topsoil
1
1 part wet loose peat moss or rotted manure
soil depth 36 inches
ponding depth 12 inches
mulch depth 3 inches
underdrains 4" perforated pipe
cell area 1,300 sf
drainage area 0.5 acres, paved
Table A.16: BP Planting Plan
Trees Liquidambar styraci ua (sweet gum)Cornus stoloniferia (red twig dogwood)
Shrubs
Ilex glabra (inkberry)
Myrica pennsylvanica (Northern bayberry)
Juniperus conferta (shore juniper)
Herbaceous
Spartina pectinata (cord grass)
Scirpus cyperinus (woolgrass)
Lythrum salicaria (loosestrife)
Panicum virgatum ?Rotstrahlbusch? (red switch grass)
Hemerocallis fulva (wild daylily)
1Topsoil specs: loam, sandy loam, clay loam, silt loam, sandy clay loam, or loamy sand
210
Figure A.7: BP construction detail (Source: Beltway Plaza Developers)
211
A.1.9 Laurel Regional Hospital
Address: 7300 Van Dusen Road, Laurel, MD 20707
No information available
212
Appendix B
Soil Testing Procedures
B.1 Procedure for Hydrometer Test
Adapted from Weil (1998)
B.1.1 Equipment
 1 L cylinder with stopper
 1 hydrometer (ASTM 151H)
 #10, #200 sieves
 250 mL beaker
 thermometer
 wash bottle
 dropper
 scale
 electric blender with metal cup
 hot plate or Bunsen burner
 drying oven
 6% hydrogen peroxide
 10% Calgon solution, (NaPO3)6
 amyl alcohol
 distilled water
B.1.2 Procedure
1. Collect an air-dried soil sample
2. Pass through #10 sieve
3. On a separate sample, determine soil moisture content
4. Determine texture by feel
213
5. Weigh out soil into a 250 mL beaker. If coarser than sandy loam, use 100 g. If sandy
loam or  ner, use 50 g
6. If organic matter content is greater than 5%, oxidize sample:
(a) Add 100 mL of 6% hydrogen peroxide to beaker
(b) Mix well and allow to stand overnight
(c) Warm gently in a water bath until reaction subsides (control frothing by adding
2 drops of amyl alcohol)
(d) Cool for a few minutes
(e) After reaction has subsided, boil for 5 minutes (not to dryness)
(f) Cool completely
7. Add 100 mL distilled water and 25 mL 10% Calgon solution to the beaker
8. Stir well for 5 minutes. For clayey soils, allow mixture to stand overnight.
9. Using distilled water from a wash bottle, completely transfer the contents of the
beaker to a metal blender cup. Fill the cup no more than 1/3 full using distilled
water.
10. Place metal cup on blender. If coarse soil is used, blend for 5 minutes. If  ne soil is
used, blend for 10 minutes.
11. Transfer contents of blender cup completely to a Bouyoucos cylinder.
12. Fill cylinder partway with distilled water.
13. Insert hydrometer, and  ll cylinder to 600 mL mark with distilled water.
14. Remove hydrometer
15. Stopper the cylinder. Rotate end over end for about 1 minute (ten times).
16. Immediately stand the cylinder on the bench and start a stop watch. Quickly insert
the hydrometer. If a froth is present at the surface, add a few drops of amyl alcohol.
17. Read the hydrometer at 40, 50, 60, 70 and 80 seconds.
18. Plot readings on Figure B.1. Draw a straight line through the readings. Record the
value at which the line intersects the heavy reference line.
19. Remove the hydrometer and measure the temperature of the suspension.
20. After 7 hours, take another hydrometer reading and temperature measurement.
21. Pour contents of cylinder through a 200-mesh sieve. Wash sieve until water runs
clear.
22. Oven dry sand retained on sieve at 158 F (70 C) for 4 hours, and store for sieve
analysis.
23. Fill cylinder with 25 mL of 10% Calgon solution an enough distilled water to make
up 1 L.
24. Take hydrometer and temperature readings.
214
B.1.3 Hydrometer test data sheet
Soil ID: Date:
Texture by feel:
Air-dried mass, ma (g):
Moisture content: 1.5% (assumed)
Oven-dried mass, mo = ma  ma  0.015 (g) =
Organic Matter Content: >10% : yes / no
Oxidized: yes / no
Notes on oxidation:
Start Time:
Initial Temperature ( C):
Composite Correction Factor, ccf:
ccf = -0.1575  T + 6.57
Timer Hydrometer Reading Corrected Reading (Reading ccf) Mass in suspension (g)
40 sec
50 sec
60 sec
70 sec
80 sec
mass in suspension = corrected hydrometer reading
Plot values on chart in Figure B.1
End Time (+7 hours):
End Temperature ( C):
Composite Correction Factor, ccf:
Hydrometer Reading Corrected Reading (Reading ccf) Mass in suspension (g)
215
Figure B.1: Graph to determine silt and clay content. Plot 5 readings (40,
50, 60, 70 and 80 sec). Use value corresponding to the intersection of you plot
with heavy diagonal line (Weil, 1998).
216
Initial Y-intercept (silt and clay) (g):
% silt and clay = corrected reading / O.D. sample mass (g) * 100 =
% sand = 100 - % silt and clay =
 nal mass in suspension (clay) (g):
% clay = corrected reading / O.D. sample mass (g) * 100 =
% silt = % silt and clay - % clay =
% sand:
% silt:
% clay:
Textural designation:
217
Figure B.2: USDA textural triangle (Weil, 1998).
218
B.2 Procedure for Sieve Testing of Soil Samples
B.2.1 Equipment
 stackable sieves: #4, #10, #20, #40, #60, #100, #140, #200
 cap and bottom pan for sieves
 mechanical shaker
 drying oven
 scale
 250 mL beaker
 wash bottle
 distilled water
B.2.2 Procedure
1. Measure out enough soil so that about 500 g of will remain after removing silt and
clay fractions. Record mass.
2. Wash soil on a #200 sieve (with tap water) until water runs clear.
3. Wash soil retained on #200 sieve into an evaporating dish using distilled water.
Oven-dry dish at 110 C.
4. After oven-drying, weigh soil and subtract mass of evaporating dish.
5. Assemble stack of dry and clean sieves, with a bottom pan.
6. Pour soil into top sieve, then shake mechanically for about 15 minutes.
7. Weigh the soil retained on each of the sieves.
219
Appendix C
Taxonomic Key to the Earthworms of Maryland
1. Setae closely paired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
 Setae widely paired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
 Setae equidistant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
 Setae perichaetine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2 (1). Prostomium epilobic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
 Prostomium tanylobic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3 (2). Clitellum annular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
 Clitellum saddle-shaped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4 (3). Dorsal color unpigmented . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . Bimastos heimburgeri, Microscolex dubius, Bimastos palustris
 Dorsal color white . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color blue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color grey . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea, Bimastos palustris
 Dorsal color pink . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea, Bimastos palustris
 Dorsal color reddish . . . Bimastos palustris, Bimastos heimburgeri, Bimastos tumidus
 Dorsal color reddish violet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color violet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color slate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color dark reddish brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri, Bimastos tumidus, Bimastos palustris
 Dorsal color red-brown in transverse bands . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color brown . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri, Bimastos palustris
 Dorsal color yellowish brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color yellowish green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color dark green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
 Dorsal color iridescent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos palustris
5 (3). Dorsal color unpigmented . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
 Dorsal color white . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
 Dorsal color blue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Octolasion tyrtaeum
 Dorsal color grey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
 Dorsal color pink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
 Dorsal color reddish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
 Dorsal color reddish violet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
 Dorsal color violet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenoides lonnbergi
 Dorsal color slate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenoides lonnbergi
 Dorsal color dark reddish brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
 Dorsal color red-brown in transverse bands . . . . . . . . . . . . . . . . . . . . . . Eisenia fetida
 Dorsal color brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
 Dorsal color yellowish brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aporrectodea longa
 Dorsal color yellowish green . . . . . . . . . . . . . . . . . . . . . . . . . . Allolobophora chlorotica
 Dorsal color dark green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6 (5). Tuberculata pubertatis indistinct . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Tuberculata pubertatis sucker-like papillae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
220
 Tuberculata pubertatis ridge-like . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . Aporrectodea caliginosa turgida, Aporrectodea caliginosa trapezoides
 Tuberculata pubertatis bipartite . . . . . . . . . . . . . . . . Aporrectodea caliginosa turgida
 Tuberculata pubertatis tripartite . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
 Tuberculata pubertatis triangular . . . . . . . . . . . . Aporrectodea calignosa tuberculata
 Tuberculata pubertatis bilboate . . . . . . . . . . . . . . . . . Aporrectodea caliginosa turgida
 Tuberculata pubertatis crescentic . . . . . . . . . . . . . . . . Aporrectodea caliginosa turgida
7 (6). Setae (aa:ab:bc:cd:dd) 4.5:1.5:3:1:15; tuberculata pubertatis two pairs . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
 Setae (aa:ab:bc:cd:dd) 13:1.5:6:1:25; tuberculata pubertatis three pairs . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allolobophora chlorotica
8 (5). Spermathecal pores above setal line c; tuberculata pubertatis narrow strip; setae
(aa:ab:bc:cd:dd) 3.3: 1.6: 1.3: 1.0: 7.3 . . . . . . . . . . . . . . . . . . . . Octolasion tyrtaeum
 Spermathecal pores between setal lines cd; tuberculata pubertatis triangular; setae
(aa:ab:bc:cd:dd) 4.5:1.5:3:1:15 . . . . . . . . . . . . . . . . Aporrectodea calignosa tuberculata
9 (5). Spermathecal pores above setal line c . . . . . . . . . . . . . . . . . . Octolasion tyrtaeum
 Spermathecal pores between setal lines cd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
10 (9). Tuberculata pubertatis indistinct . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis sucker-like papillae . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis elliptical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis oval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis saucer-like . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis narrow strip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis broad, rectangular . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis ridge-like . . . Eisenia rosea, Aporrectodea caliginosa turgida
 Tuberculata pubertatis bipartite . . . Aporrectodea caliginosa turgida, Eisenia rosea
 Tuberculata pubertatis tripartite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 Tuberculata pubertatis triangular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea, Aporrectodea calignosa tuberculata
 Tuberculata pubertatis bilboate . . . . Aporrectodea caliginosa turgida, Eisenia rosea
 Tuberculata pubertatis crescentic . . . Eisenia rosea, Aporrectodea caliginosa turgida
11 (5). First dorsal pore 3/4 . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia fetida, Eisenia rosea
 First dorsal pore 4/5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea, Eisenia fetida
 First dorsal pore 5/6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 6/7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 7/8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 8/9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 9/10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 10/11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 11/12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
 First dorsal pore 12/13 . . . . . . . . . . . . Aporrectodea caliginosa turgida, Eisenia rosea
 First dorsal pore 13/14 . . . . . . . . . . . . Aporrectodea caliginosa turgida, Eisenia rosea
 First dorsal pore 14/15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia rosea
12 (5). Spermathecal pores absent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
 Spermathecal pores two pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
 Spermathecal pores three pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
13 (12). Tuberculata pubertatis indistinct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
 Tuberculata pubertatis sucker-like papillae . . . . Aporrectodea caliginosa trapezoides
221
 Tuberculata pubertatis ridge-like . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
 Tuberculata pubertatis tripartite . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
14 (13). Male pores inconspicuous . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Male pores with small tumescences . . . . . . . Bimastos heimburgeri, Bimastos parvus
 Male pores with moderate tumescences . . . Bimastos tumidus, Bimastos heimburgeri
 Male pores with large tumescences . . . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Male pores con ned to a single segment . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Male pores large . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Male pores paired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
15 (12). Setae (aa:ab:bc:cd:dd) 4: 1: 4: 1: 16 . . . . . . . . . . . . . . . . . . . . . . . . Eisenia fetida
 Setae (aa:ab:bc:cd:dd) 4.5:1.5:3:1:15 . . . . . . . . . . Aporrectodea caliginosa trapezoides
16 (12). Spermathecal pores between setal lines cd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
 Spermathecal pores setal line d . . . . . . . . . . . . . . . . . . . . . . . . Eisenoides carolinensis
17 (5). Ventral color unpigmented; spermathecal pores two pairs; tuberculata pubertatis
narrow strip; male pores paired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eisenia fetida
 Ventral color yellowish; spermathecal pores absent; tuberculata pubertatis indistinct;
male pores with small tumescences . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos parvus
18 (5). Tuberculata pubertatis indistinct . . . Bimastos heimburgeri, Bimastos tumidus
 Tuberculata pubertatis sucker-like papillae . . . . Aporrectodea caliginosa trapezoides
 Tuberculata pubertatis oval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aporrectodea longa
 Tuberculata pubertatis ridge-like . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
 Tuberculata pubertatis tripartite . . . . . . . . . . . . . Aporrectodea caliginosa trapezoides
19 (5). Spermathecal pores absent; tuberculata pubertatis indistinct . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bimastos heimburgeri
 Spermathecal pores two pairs; tuberculata pubertatis narrow strip . . . . . . . . . . . . 20
20 (19). Posterior square; setae (aa:ab:bc:cd:dd) 3: 1: 3: 1: 6 . . . . Eiseniella tetraedra
 Posterior trapezoidal; setae (aa:ab:bc:cd:dd) 4: 1: 4: 1: 16 . . . . . . . . . Eisenia fetida
21 (5). Posterior depressed dorso-ventrally; spermathecal pores three pairs; tuberculata
pubertatis sucker-like papillae; setae (aa:ab:bc:cd:dd) 13:1.5:6:1:25 . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allolobophora chlorotica
 Posterior square; spermathecal pores two pairs; tuberculata pubertatis narrow strip;
setae (aa:ab:bc:cd:dd) 3: 1: 3: 1: 6 . . . . . . . . . . . . . . . . . . . . . . . . Eiseniella tetraedra
22 (2). Tuberculata pubertatis indistinct . . . Bimastos heimburgeri, Bimastos palustris
 Tuberculata pubertatis saucer-like . . . . . . . . . . . . . . . . . . . . . . . . . . . Lumbricus friendi
 Tuberculata pubertatis narrow strip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
 Tuberculata pubertatis broad, rectangular . . . . . . . . . . . . . . . . . . . Lumbricus rubellus
 Tuberculata pubertatis ridge-like . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
23 (22). Spermathecal pores setal line c; setae (aa:ab:bc:cd:dd) 5: 1.2: 5: 1: 20 . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lumbricus castaneus
 Spermathecal pores setal line d; setae (aa:ab:bc:cd:dd) 7: 1.5: 6: 1: 22 . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lumbricus friendi
24 (22). Spermathecal pores between setal lines cd; setae (aa:ab:bc:cd:dd) 6: 1.3: 5: 1:
22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Lumbricus terrestris
 Spermathecal pores setal line d; setae (aa:ab:bc:cd:dd) 7: 1.5: 6: 1: 22 . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lumbricus friendi
25 (1). Clitellum annular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Diplocardia singularis, Microscolex dubius, Bimastos heimburgeri, Bimastos palustris
222
 Clitellum saddle-shaped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
26 (25). Tuberculata pubertatis indistinct . . . . Bimastos heimburgeri, Bimastos parvus
 Tuberculata pubertatis narrow strip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
 Tuberculata pubertatis broad, rectangular . . . . . . . . . . . . . . . . . Dendrodrilus rubidus
27 (26). Setae (aa:ab:bc:cd:dd) 2: 1: 2: 1: 12; posterior cylindrical . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dendrodrilus rubidus
 Setae (aa:ab:bc:cd:dd) 3.3: 1.6: 1.3: 1.0: 7.3; posterior octagonal . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Octolasion tyrtaeum
28 (1). Clitellum annular; ventral color unpigmented . . . . . . . . . . . . . Microscolex dubius
 Clitellum saddle-shaped; ventral color reddish . . . . . . . . . . . . . Dendrobaena octaedra
29 (1). Spermathecal pores absent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
 Spermathecal pores two pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
 Spermathecal pores three pairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
30 (29). Unusual features fuzzy . . . . . . . . . . . . Pheretima agrestis, Pheretima di ringens
 Unusual features white  ecks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
 Unusual features brown  ecks . . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima di ringens
 Unusual features bulbous head . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima di ringens
31 (30). Ventral color unpigmented . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microscolex dubius
 Ventral color reddish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima di ringens
 Ventral color same as dorsal color . . . . . . . . . . . . . . . . . . . . . . . Pheretima di ringens
32 (29). Unusual features fuzzy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima agrestis
 Unusual features white  ecks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microscolex dubius
33 (29). Unusual features fuzzy . . . . . . . . . . . . Pheretima hupeiensis, Pheretima agrestis
 Unusual features white  ecks . . . . . . . . . . Microscolex dubius, Pheretima hupeiensis
 Unusual features brown  ecks . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima hupeiensis
 Unusual features bulbous head . . . . . . . . . . . . . . . . . . . . . . . . . . . Pheretima hupeiensis
This key was generated using DELTA (DEscriptive Language for TAxonomy, by
IntKey). The taxonomic descriptions included in this key were compiled from a number
of sources (Baker and Barrett, 1994; Eaton, 1942; Fender and McKey-Fender, 1990; Gates,
1937, 1958, 1972a,b, 1973, 1974; James, 1990; Lee, 1959; Reynolds et al., 1974; Schwert,
1990; Sims et al., 1999; Worm Watch Canada, 2002). The earthworms included in this
key are those identi ed by Reynolds (1974) and Csuzdi and Sl avecz (2003) as occurring
in Maryland. The key includes only external features. In some cases, the data included
in the key are insu cient to distinguish between closely related species.
223
Appendix D
Raw Field Data
D.1 University of Maryland (UMCP)
Sampling Date: July 29, 2005
Age at sampling: 1 year
Table D.1: UMCP Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 7 worms, 22.5 cm 2 worms, 5.5 cm 4 worms, 7.0 cm
10-20 cm 0 worms 0 worms 0 worms
20-30 cm 0 worms 0 worms 0 worms
Table D.2: UMCP Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 0.49 g 0.88 g 0.77 g
10-20 cm 0.18 g 0.46 g 0.26 g
20-30 cm 0.75 g 1.42 g 0.03 g
Table D.3: UMCP Soil organic matter content
Location
Depth 1 2 3
0-10 cm 3.19% 2.98% 2.55%
10-20 cm sample was lost prior to testing 1.50% 1.57%
20-30 cm 1.46% 1.44% 1.40%
224
Figure D.1: UMCP soil pro le.
225
Table D.4: UMCP Soil texture
Location
Depth 1 2 3
0-10 cm
Sand Sand Sand
88.3% sand 87.7% sand 92.1% sand
7.3% silt 8.2% silt 2.8% silt
4.3% clay 4.0% clay 5.1% clay
10-20 cm
sample was lost prior to testing Sand Sand
91.4% sand 93.8% sand
5.1% silt 4.0% silt
3.6% clay 2.2% clay
20-30 cm
Sand Sand Sand
92.0% sand 92.6% sand 91.3% sand
4.1% silt 3.8% silt 5.4% silt
3.9% clay 3.6% clay 3.3% clay
Table D.5: UMCP Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 4
beetle larvae Coleoptera larvae -
common white grubs Coleoptera: Scarabaeidae larvae 1
ants Hymenoptera: Formicidae 1
centipedes Myriapoda: Chilopoda -
millipedes Myriapoda: Diplopoda 1
potworms Oligochaeta: Enchytraeidae -
pill bugs Crustacea: Isopoda -
spiders Arachnida: Araneae 1
springtails Hexapoda: Collembola -
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.6: UMCP Earthworm species identi ed.
Earthworm species Number identi ed
Diplocardia singularis 9
226
D.2 Washington Navy Yard (WNY)
Sampling Date: May 13, 2004
Age at sampling: 2 years
Table D.7: WNY Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 10 worms, 55.8 cm 0 worms 2 worms, 8.9 cm
10-20 cm 4 worms, 19.1 cm 0 worms 1 worm, 3.8 cm
20-30 cm 0 worms 0 worms 0 worms
Table D.8: WNY Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 3.88 g 4.18 g 0.17 g
10-20 cm 3.38 g 6.97 g 0.14 g
20-30 cm 2.96 g 7.51 g 0 g
Table D.9: WNY Soil organic matter content
Location
Depth 1 2 3
0-10 cm 8.26% 4.67% 5.90%
10-20 cm 2.35% 2.49% 2.18%
20-30 cm 2.16% 2.16% 1.83%
227
Figure D.2: WNY soil pro le.
228
Table D.10: WNY Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy loam Sandy loam Loamy sand
68.1% sand 71.8% sand 75.0% sand
26.4% silt 21.7% silt 21.0% silt
5.4% clay 6.5% clay 3.9% clay
10-20 cm
Sandy loam Sandy loam Sandy loam
67.9% sand 75.6% sand 72.4% sand
26.8% silt 17.1% silt 19.7% silt
5.2% clay 7.3% clay 7.9% clay
20-30 cm
Sandy loam Sandy loam Sandy loam
66.5% sand 75.2% sand 68.9% sand
26.8% silt 7.0% silt 25.6% silt
6.7% clay 17.8% clay 5.5% clay
Table D.11: WNY Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera -
beetle larvae Coleoptera larvae -
common white grubs Coleoptera: Scarabaeidae larvae 24
ants Hymenoptera: Formicidae -
centipedes Myriapoda: Chilopoda -
millipedes Myriapoda: Diplopoda -
potworms Oligochaeta: Enchytraeidae -
pill bugs Crustacea: Isopoda -
spiders Arachnida: Araneae -
springtails Hexapoda: Collembola -
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.12: WNY Earthworm species identi ed.
Earthworm species Number identi ed
none identi ed
229
D.3 \Mother" Jones Elementary School (MJES)
Sampling Date: June 16, 2005
Age at sampling: 3 years
Table D.13: MJES Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 121 worms, 353.5 cm 12.5 worms, 83.0 cm 2 worms, 6.5 cm
10-20 cm 7.5 worms, 15.5 cm 0 worms 0 worms
20-30 cm 2 worms, 6.0 cm 0 worms 0 worms
Table D.14: MJES Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 1.35 g 1.06 g 2.46 g
10-20 cm 2.77 g 0.07 g 2.03 g
20-30 cm 5.51 g 0.16 g 0.58 g
Table D.15: MJES Soil organic matter content
Location
Depth 1 2 3
0-10 cm 30.87% 2.69% 19.02%
10-20 cm 3.52% 0.37% 3.04%
20-30 cm 3.00% 0.24% 3.03%
230
Figure D.3: MJES soil pro le.
231
Table D.16: MJES Soil texture
Location
Depth 1 2 3
0-10 cm
Loam Loamy sand Loam
43.1% sand 78.7% sand 49.9% sand
41.3% silt 13.2% silt 31.8% silt
15.6% clay 8.1% clay 18.3% clay
10-20 cm
Sandy Clay Loam Sand Sandy Loam
57.2% sand 96.3% sand 60.4% sand
21.2% silt 1.2% silt 21.0% silt
21.6% clay 2.5% clay 18.6% clay
20-30 cm
Sandy Clay Loam Sand Sandy Clay Loam
57.2% sand 96.2% sand 53.1% sand
19.9% silt 1.6% silt 25.8% silt
22.9% clay 2.2% clay 21.1% clay
Table D.17: MJES Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 13
beetle larvae Coleoptera larvae 4
common white grubs Coleoptera: Scarabaeidae larvae 9
ants Hymenoptera: Formicidae 8
centipedes Myriapoda: Chilopoda 1
millipedes Myriapoda: Diplopoda 3
potworms Oligochaeta: Enchytraeidae 26
pill bugs Crustacea: Isopoda -
spiders Arachnida: Araneae 2
springtails Hexapoda: Collembola 23
slugs and snails Mollusca: Gastropoda 1
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.18: MJES Earthworm species identi ed.
Earthworm species Number identi ed
Allolobophora chlorotica 2
Bimastos parvus 1
Pheretima spp. 14
232
D.4 Chesapeake Bay Foundation Headquarters (CBF)
Sampling Date: September 14, 2005
Age at sampling:4 years
Table D.19: CBF Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 22 worms, 53.0 cm 5 worms, 13.0 cm 15 worms, 29.5 cm
10-20 cm 2 worms, 15.0 cm 0 worms 2 worms, 4.0 cm
20-30 cm 0 worms 0 worms 1 worm, 1.0 cm
Table D.20: CBF Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 8.05 g 4.17 g 18.09 g
10-20 cm 2.01 g 1.31 g 2.25 g
20-30 cm 0.50 g 0.58 g 1.59 g
Table D.21: CBF Soil organic matter content
Location
Depth 1 2 3
0-10 cm .41% 2.52% 2.12%
10-20 cm 3.07% 2.45% 2.56%
20-30 cm 2.95% 2.91% 2.65%
233
Figure D.4: CBF soil pro le.
234
Table D.22: CBF Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy Clay Loam Sandy Clay Loam Sandy Clay Loam
60.6% sand 59.8% sand 62.6% sand
16.9% silt 14.0% silt 14.5% silt
22.5% clay 26.2% clay 22.9% clay
10-20 cm
Sandy Clay Loam Sandy Clay Loam Sandy Clay Loam
61.8% sand 59.8% sand 60.8% sand
17.5% silt 17.4% silt 16.5% silt
20.7% clay 22.8% clay 22.7% clay
20-30 cm
Sandy Clay Loam Sandy Clay Loam Sandy Loam
61.4% sand 60.6% sand 61.8% sand
16.9% silt 16.5% silt 18.3% silt
21.7% clay 22.9% clay 19.9% clay
Table D.23: CBF Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 23
beetle larvae Coleoptera larvae 1
common white grubs Coleoptera: Scarabaeidae larvae -
ants Hymenoptera: Formicidae 5
centipedes Myriapoda: Chilopoda 3
millipedes Myriapoda: Diplopoda 1
potworms Oligochaeta: Enchytraeidae 3
pill bugs Crustacea: Isopoda 47
spiders Arachnida: Araneae 6
springtails Hexapoda: Collembola 3
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
insect larvae Insecta 1
Table D.24: CBF Earthworm species identi ed.
Earthworm species Number identi ed
Apporectodea caliginosa 46
Pheretima di ringens 1
235
D.5 Northwestern High School (NWHS)
Sampling Date: July 7, 2004
Age at sampling: 5 years
Table D.25: NWHS Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 50 worms, 102.5 cm 0 worms 4 worms, 15.0 cm
10-20 cm 28 worms, 52.5 cm 0 worms 0 worms
20-30 cm 4 worms, 4.0 cm 0 worms 0 worms
Table D.26: NWHS Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 4.81 g 13.82 g 4.11 g
10-20 cm 1.35 g 1.56 g 1.38 g
20-30 cm 1.05 g 0.73 g 1.87 g
Table D.27: NWHS Soil organic matter content
Location
Depth 1 2 3
0-10 cm 5.4 % 4.56% 3.20%
10-20 cm 4.03% 3.20% 3.54%
20-30 cm 2.01% 3.21% 3.21%
236
Figure D.5: NWHS soil pro le.
237
Table D.28: NWHS Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy loam Loam Loam
57.6% sand 51.3% sand 50.3% sand
32.1% silt 32.1% silt 34.7% silt
10.3% clay 16.6% clay 15.0% clay
10-20 cm
Loam Loam Loam
40.1% sand 38.1% sand 51.9% sand
46.7% silt 42.4% silt 31.0% silt
13.2% clay 19.5% clay 17.1% clay
20-30 cm
Sandy loam Sandy loam Sandy loam
68.5% sand 52.7% sand 52.1% sand
24.0% silt 30.9% silt 33.9% silt
7.5% clay 16.4% clay 14.0% clay
Table D.29: NWHS Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 8
beetle larvae Coleoptera larvae 11
common white grubs Coleoptera: Scarabaeidae larvae 1
ants Hymenoptera: Formicidae 10
centipedes Myriapoda: Chilopoda 1
millipedes Myriapoda: Diplopoda 2
potworms Oligochaeta: Enchytraeidae 34
pill bugs Crustacea: Isopoda 3
spiders Arachnida: Araneae 4
springtails Hexapoda: Collembola 9
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.30: NWHS Earthworm species identi ed.
Earthworm species Number identi ed
none
238
D.6 Inglewood Center III (PP)
Sampling Date: July 29, 2004
Age at sampling: 5 years
Table D.31: PP Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 7 worms, 26.5 cm 3 worms, 11.5 cm 2.5 worms, 7.5 cm
10-20 cm 1 worm, 3.5 cm 1 worm, 7.0 cm 0 worms
20-30 cm 0 worms 1 worm, 9.0 cm 0 worms
Table D.32: PP Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 3.15 g 0.56 g 2.06 g
10-20 cm 1.57 g 0.51 g 0.98 g
20-30 cm 0.69 g 0.79 g 0.45 g
Table D.33: PP Soil organic matter content
Location
Depth 1 2 3
0-10 cm 16.27% 9.52% 27.92%
10-20 cm 5.89% 5.25% 13.68%
20-30 cm 2.69% 2.86% 5.70%
239
Figure D.6: PP soil pro le.
240
Table D.34: PP Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy Loam Sandy loam Sandy loam
62.9% sand 79.1% sand 69.6% sand
17. 5% silt 8.3% silt 13.2% silt
19.6% clay 12.6% clay 17.2% clay
10-20 cm
Loamy sand Loamy sand Loamy sand
83.8% sand 88.1% sand 87.8% sand
11.9% silt 4.3% silt 4.2% silt
4.3% clay 7.6% clay 7.9% clay
20-30 cm
Sand Sand Loamy sand
92.2% sand 89.2% sand 87.8% sand
1.5% silt 7.5% silt 4.2% silt
6.3% clay 3.3% clay 8.0% clay
Table D.35: PP Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 1
beetle larvae Coleoptera larvae 4
common white grubs Coleoptera: Scarabaeidae larvae 33
ants Hymenoptera: Formicidae 4
centipedes Myriapoda: Chilopoda -
millipedes Myriapoda: Diplopoda -
potworms Oligochaeta: Enchytraeidae 3
pill bugs Crustacea: Isopoda 2
spiders Arachnida: Araneae 1
springtails Hexapoda: Collembola 3
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.36: PP Earthworm species identi ed.
Earthworm species Number identi ed
Dendrodrilus rubidus 1
Apporectodea caliginosa 1
241
D.7 Claggett Farm (CF)
Sampling Date: September 16, 2005
Age at sampling: 6 years
Table D.37: CF Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 0 worms 0 worms 0 worms
10-20 cm 1 worm, 5.0 cm 9 worms, 53.5 cm 1 worm, 3.0 cm
20-30 cm 0 worms 0 worms 5 worms, 23.0 cm
Table D.38: CF Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 26.52 g 7.35 g 32.06 g
10-20 cm 11.37 g 5.16 g 8.88 g
20-30 cm 3.55 g 0.95 g 2.25 g
Table D.39: CF Soil organic matter content
Location
Depth 1 2 3
0-10 cm 5.90% 5.36% 5.51%
10-20 cm 4.09% 3.44% 3.71%
20-30 cm 2.93% 2.17% 2.38%
242
Figure D.7: CF soil pro le.
243
Table D.40: CF Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy Loam Sandy Loam Sandy Loam
72.6% sand 67.5% sand 69.9% sand
16.6% silt 18.0% silt 17.9% silt
10.8% clay 14.5% clay 12.1% clay
10-20 cm
Sandy Loam Sandy Loam Sandy Loam
70.2% sand 67.5% sand 65.1% sand
16.8% silt 19.8% silt 18.5% silt
13.0% clay 12.7% clay 16.4% clay
20-30 cm
Sandy Loam Sandy Clay Loam Sandy Loam
62.8% sand 61.2% sand 64.9% sand
18.0% silt 16.8% silt 17.4% silt
19.2% clay 22.0% clay 17.7% clay
Table D.41: CF Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 16
beetle larvae Coleoptera larvae 9
common white grubs Coleoptera: Scarabaeidae larvae -
ants Hymenoptera: Formicidae 148
centipedes Myriapoda: Chilopoda -
millipedes Myriapoda: Diplopoda 7
potworms Oligochaeta: Enchytraeidae -
pill bugs Crustacea: Isopoda 1
spiders Arachnida: Araneae 4
springtails Hexapoda: Collembola 1
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
insect larvae Insecta 2
Table D.42: CF Earthworm species identi ed.
Earthworm species Number identi ed
Allolobophora chlorotica 4
244
D.8 Chevy Chase Bank (CC)
Sampling Date: July 7, 2005
Age at sampling: 7 years
Table D.43: CC Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 20.5 worms, 46.5 cm 7 worms, 11.0 cm 3 worms, 13.0 cm
10-20 cm 22 worms, 33.0 cm 4 worms, 5.5 cm 4 worms, 13.0 cm
20-30 cm 16.5 worms, 31.0 cm 1 worm, 4.0 cm 5.5 worms, 24.0 cm
Table D.44: CC Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 9.25 g 32.65 g 36.22 g
10-20 cm 14.18 g 39.62 g 4.10 g
20-30 cm 28.77 g 0.99 g 3.69 g
Table D.45: CC Soil organic matter content
Location
Depth 1 2 3
0-10 cm 42.54% 27.80% 20.05%
10-20 cm 32.74% 7.53% 8.83%
20-30 cm 22.46% 1.36% 6.35%
245
Figure D.8: CC soil pro le.
246
Table D.46: CC Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy Loam Loamy Sand Loam
67.0% sand 76.4% sand 43.8% sand
28.3% silt 21.3% silt 43.5% silt
4.6% clay 2.3% clay 12.7% clay
10-20 cm
Silt Loam Loamy Sand Loam
32.3% sand 81.4% sand 48.0% sand
56.1% silt 15.3% silt 37.0% silt
11.6% clay 3.3% clay 15.0% clay
20-30 cm
Loam Sand Sandy Loam
46.7% sand 93.8% sand 53.1% sand
39.1% silt 3.8% silt 31.0% silt
14.2% clay 2.4% clay 15.9% clay
Table D.47: CC Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 6
beetle larvae Coleoptera larvae 1
common white grubs Coleoptera: Scarabaeidae larvae 5
ants Hymenoptera: Formicidae 22
centipedes Myriapoda: Chilopoda 29
millipedes Myriapoda: Diplopoda -
potworms Oligochaeta: Enchytraeidae 3
pill bugs Crustacea: Isopoda 3
spiders Arachnida: Araneae 2
springtails Hexapoda: Collembola 1
slugs and snails Mollusca: Gastropoda 1
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari 1
Table D.48: CC Earthworm species identi ed.
Earthworm species Number identi ed
Pheretima sp. 29
247
D.9 Beltway Plaza Mall (BP)
Sampling Date: July 28, 2004
Age at sampling: 7 years
Table D.49: BP Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 20 worms, 75.9 cm 31.5 worms, 119.6 cm 7 worms, 26.0 cm
10-20 cm 9 worms, 34.2 cm 3 worms, 11.4 cm 0 worms
20-30 cm 1 worm, 3.8 cm 1 worm, 3.8 cm 0 worms
Table D.50: BP Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 13.44 g 32.24 g 11.55 g
10-20 cm 9.60 g 5.63 g 6.12 g
20-30 cm 4.93 g 3.35 g 0.67 g
Table D.51: BP Soil organic matter content
Location
Depth 1 2 3
0-10 cm 12.62% 10.75% 5.14%
10-20 cm 3.02% 2.87% 2.88%
20-30 cm 2.52% 2.19% 4.38%
248
Figure D.9: BP soil pro le.
249
Table D.52: BP Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy clay loam Sandy loam Sandy clay loam
60.0% sand 72.8% sand 53.7% sand
17.9% silt 10.6% silt 20.1% silt
22.1% clay 16.6% clay 26.2% clay
10-20 cm
Sandy clay loam Sandy clay loam Sandy clay loam
53.3% sand 55.1% sand 54.9% sand
24.0% silt 23.4% silt 21.1% silt
22.7% clay 21.5% clay 24.0% clay
20-30 cm
Sandy clay loam Loam Clay loam
55.1% sand 44.8% sand 27.7% sand
21.1% silt 34.9% silt 38.6% silt
23.8% clay 20.4% clay 33.7% clay
Table D.53: BP Soil invertebrates total tally for entire site.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 3
beetle larvae Coleoptera larvae -
common white grubs Coleoptera: Scarabaeidae larvae -
ants Hymenoptera: Formicidae 4
centipedes Myriapoda: Chilopoda 3
millipedes Myriapoda: Diplopoda 1
potworms Oligochaeta: Enchytraeidae -
pill bugs Crustacea: Isopoda 4
spiders Arachnida: Araneae -
springtails Hexapoda: Collembola -
slugs and snails Mollusca: Gastropoda -
 y larvae Hexapoda: Diptera larvae -
mites Arachnida: Acari -
Table D.54: BP Earthworm species identi ed.
Earthworm species Number identi ed
Eisenia fetida 4
Lumbricus sp. 5
Pheretima sp. 3
250
D.10 Laurel Regional Hospital (LRH)
Sampling Date: July 20, 2004
Age at sampling: 10 years
Table D.55: LRH Earthworms - number of individuals and sum of lengths
Location
Depth 1 2 3
0-10 cm 13.5 worms, 67.5 cm 13 worms, 44.0 cm 14.5 worms, 63.0 cm
10-20 cm 8 worms, 34.5 cm 2.5 worms, 7.0 cm worms, 7.5 cm
20-30 cm 7.5 worms, 35.5 cm 4 worms, 14.0 cm 0 worms
Table D.56: LRH Root biomass, air-dried
Location
Depth 1 2 3
0-10 cm 12.46 g 0.92 g 9.84 g
10-20 cm 10.24 g 0.25 g 4.64 g
20-30 cm 1.86 g 0.17 g 0.82 g
Table D.57: LRH Soil organic matter content
Location
Depth 1 2 3
0-10 cm 3.86% 8.08% 5.94%
10-20 cm 3.18% 4.38% 2.52%
20-30 cm 2.02% 2.36% 2.02%
251
Figure D.10: LRH soil pro le.
252
Table D.58: LRH Soil texture
Location
Depth 1 2 3
0-10 cm
Sandy loam Sandy loam Loam
59.2% sand 62.4% sand 36.4% sand
20.7% silt 21.2% silt 46.0% silt
20.1% clay 16.4% clay 17.5% clay
10-20 cm
Sandy loam Clay loam Loam
62.4% sand 39.0% sand 35.6% sand
25.4% silt 22.0% silt 46.1% silt
12.1% clay 39.0% clay 18.2% clay
20-30 cm
Sandy loam Loam Silt loam
59.4% sand 40.3% sand 22.6% sand
21.1% silt 40.1% silt 57.1% silt
19.5% clay 19.6% clay 20.3% clay
Table D.59: LRH Soil invertebrates.
Common name Scienti c name Number of individuals
adult beetles Coleoptera 1
beetle larvae Coleoptera larvae 3
common white grubs Coleoptera: Scarabaeidae larvae 1
ants Hymenoptera: Formicidae 2
centipedes Myriapoda: Chilopoda 1
millipedes Myriapoda: Diplopoda 3
potworms Oligochaeta: Enchytraeidae -
pill bugs Crustacea: Isopoda 3
spiders Arachnida: Araneae -
springtails Hexapoda: Collembola -
slugs and snails Mollusca: Gastropoda 1
 y larvae Hexapoda: Diptera larvae 1
mites Arachnida: Acari -
Table D.60: LRH Earthworm species identi ed.
Earthworm species Number identi ed
Lumbricus rubellus 2
other Lumbricus spp. 25
Aporrectodea caliginosa 3
253
Appendix E
Microcosm Pilot Project
Short-term (1-month) microcosm experiments were conducted in Fall 2004 in order
to assess earthworm survivorship in bioretention soil media with variable sand content.
Earthworms survived in all three treatments. The experiment was conducted with the
assistance of students enrolled in NRMT 470 (Heidi McMillen, Nicholas Regalia, Laura
Senkowsky, Katie Struder, and Kyle Wagner). The results of this study were presented in
poster form at the 2005 meeting of the American Ecological Engineering Society (AEES)
in Columbus, Ohio.
Eighteen soil columns were constructed from 4-inch diameter clear plastic tubing.
Each soil column was about 24 inches tall. A funnel was placed over the base of the tubing
to allow drainage but prevent the soil from falling out. Three treatments were used:
Treatment 1: 50% sand, 30% sandy loam topsoil, 10% shredded hardwood mulch, 10%
compost.
Treatment 2: 25% sand, 55% sandy loam topsoil, 10% shredded hardwood mulch, 10%
compost.
Treatment 3: 80% sandy loam topsoil, 10% shredded hardwood mulch, 10% compost
Earthworms were collected from a previously sampled rain garden at Beltway Plaza
Mall. Three soil columns for each treatment were inoculated with three large and two
small earthworms, which were not identi ed. 100 mL of composted manure was added
to each column to provide food for the worms. Each column was topped with 2{3 cm
254
of hardwood mulch. The columns were kept moist throughout the experiment. After
4 weeks, the columns were disassembled, and the surviving earthworms were retrieved.
66.7% of the earthworms in Treatment 1 and Treatment 2 survived (Figure E.1). 60% of
the earthworms in Treatment 3 survived. The lower survivorship in the 0% sand columns
may be explained by the development of low-oxygen conditions within the soil.
Figure E.1: Survivorship of earthworms in media of varying sand content.
255
Appendix F
Detailed Earthworm Survivorship Data for Microcosm Experiment
Table F.1: Column 2, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 8
medium unpigmented juveniles, 3{7 cm 27
Table F.2: Column 3, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
medium unpigmented juveniles, 3{7 cm 1
medium unpigmented adult, 3{7 cm 1
256
Table F.3: Column 4, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 57
medium unpigmented juveniles, 3{7 cm 16
medium unpigmented adult, 3{7 cm 1
medium adult violet/greenish, 3{7 cm 2
large adult L. terrestris, > 10 cm 1
cocoons 2
Table F.4: Column 6, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
medium unpigmented adult, 3{7 cm 1
small L. terrestris juveniles, < 3 cm 8
medium L. terrestris juveniles, 3{7 cm 9
large adult L. terrestris, > 10 cm 2
257
Table F.5: Column 8, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
medium unpigmented juveniles, 3{7 cm 11
large unpigmented adult, 7{10 cm 1
medium green adult 1
medium L. terrestris juveniles, 3{7 cm 2
large adult L. terrestris, 7{10 cm 1
cocoons 1
Table F.6: Column 9, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 25
medium unpigmented juveniles, 3{7 cm 16
medium unpigmented adults, 3{7 cm 2
small L. terrestris juveniles, < 3 cm 13
medium L. terrestris juveniles, 3{7 cm 5
258
Table F.7: Column 10, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small red juveniles, < 3 cm 1
medium red juveniles, 3{7 cm 2
medium unpigmented juveniles, 3{7 cm 3
medium unpigmented adults, 3{7 cm 1
cocoons 2
Table F.8: Column 12, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 73
medium unpigmented juveniles, 3{7 cm 1
small juvenile L. terrestris, < 3 cm 3
medium juvenile L. terrestris, 3{7 cm 1
cocoons 14
Table F.9: Column 13, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium L. terrestris juveniles, 3{7 cm 1
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 2
medium L. terrestris juveniles, 3{7 cm 1
cocoons 1
259
Table F.10: Column 14, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 188
medium unpigmented juveniles, 3{7 cm 1
small juvenile L. terrestris, < 3 cm 8
medium juvenile L. terrestris, 3{7 cm 2
medium grey adult, 3{7 cm 1
cocoons 5
Table F.11: Column 17, Treatment 2
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 14
medium unpigmented juveniles, 3{7 cm 17
large grey adult, 7{10 cm 1
small juvenile L. terrestris, < 3 cm 3
medium juvenile L. terrestris, 3{7 cm 9
medium adult Lumbricus sp., 3{7 cm 1
large adult L. terrestris, > 10 cm 1
cocoons 8
260
Table F.12: Column 18, Treatment 1
Initial earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 1
medium unpigmented juveniles, 3{7 cm 4
large adult L. terrestris, > 10 cm 3
Final earthworm size distribution
Category Number
small unpigmented juveniles, < 3 cm 15
medium unpigmented juveniles, 3{7 cm 13
medium L. terrestris juveniles, 3{7 cm 6
large adult L. terrestris, > 10 cm 14
261
Appendix G
RAINGARDEN Model Development
G.1 State Variables
S: soil volume
S = 1 m3
Derivation: assumed unit volume
O: soil organic matter
O = 9.15 x 104 g
Derivation: average value for all  eld samples from  eld survey
Average SOM = 6.% (by dry weight)
Convert to mass:
Bulk density of BSM = 1.5 g/cm3 (typical value for a cultivated sand or sandy loam
(Brady and Weil, 2002))
Total soil mass = soil volume  bulk density = 1.5 g/cm3  1 m3  106 cm3/m3
= 1.5 x 106 g
Mass of organic matter = total soil mass % SOM = 1.5 x 106 g 0.061 = 9.15 x 104 g
W: earthworms (number of individuals)
W = 310.7 individuals
Derivation: average value for all  eld samples from  eld survey
Average number of worms = 18.64 individuals for the entire sampled depth (30 cm) of
262
each 20 x 30 cm sampling area. Assuming that there are no earthworms deeper than
30 cm, convert to number per unit volume:
Number of worms in unit volume = 18.64  600 cm2  10,000 cm2/m2  1 m3
= 310.7 individuals
V: void space
V = 0.5 m3
Derivation: pore volume for an ideal soil = 50% of total soil volume (Brady and Weil,
2002)
Void space = 50%  total soil volume, S = 50%  1 m3 = 0.5 m3
H: pore water
H = 0.08 m3/d
Derivation:
For a sand/loamy sand at  eld capacity, water content = 8% of soil volume (Brady and
Weil, 2002)
Volume of water held in pores at  eld capacity = 8%  total soil volume, S = 8% x 1
m3 = 0.08 m3
C: crust thickness
C = 0.4 cm
Derivation: average thickness for column tests subjected to continuous loading using
kaolin as suspended solids in Table 5-1 of (Li, 2007)
263
G.2 Forcing Functions
I: In ow Volume
I = 0.162 m3 Derivation: based on the SCS Method of runo estimation (McCuen,
2005)
Rainfall depth, P = 1/2"
Watershed area: raingarden area = 20:1
Watershed area, A = 20 m2
CN = 98 (assume a paved surface)
Retention, S = (1000) / (CN 10) = 0.20
Runo depth, Q = (P { 0.2S)2 / (P + 0.8S) = 0.32" = 0.0081 m
Runo volume, I = Q  A = 20  0.0081 = 0.162 m3
D: TSS
D = 0.13 g/L (Li and Davis, 2008b)
expressed as incoming thickness in cm per L of in owing runo :
D = TSS / bulk density / area
Assuming TSS is kaolin as in Li and Davis (2008b), bulk density of kaolin = 1.46 g/cm3
(average value of oven-dried surface-weathered and sedimentary kaolin samples, Table
1-A, Baumann and Keller (1975))
D = 0.13 g/L  (103 L/m3) / (1.46 g/cm3) / 104 cm2
= 8.9 x 10 3 cm/m3 of runo 
L: litterfall
L = 2.05 g/d
264
Derivation: at steady state, L = J1 - J10
G.3 Flows
J1 = k1SOW
J1: earthworm assimilation rate
J1 = 2.05 g/d
Derivation:
Earthworm biomass increases by 10% over 85 days (Binet and Trehen, 1992)
Growth rate = rate of assimilation of organic matter (feeding rate is not important, as
most is egested back into the soil)
Growth rate = 0.1 g/g  85 days = 1.18 x 10 3g/g/d
Assimilation rate = growth rate  number of earthworms  average size of adult L.
terrestris)
Assimilation rate = 1.18e 3g/g/d 5.6 g/ ind (Satchell, 1967) (310.7 ind) = 2.05 g/d
J2 = k2SOW
J2: rate of earthworm reproduction
J2 = 0.38 ind/d
Derivation:
Earthworm growth rate: .33{.56 g new worms  g existing worms/yr (Lakhani and
Satchell, 1970)
average = .45g/g/yr  310.7 ind  365 d = 0.38 ind/d
265
J3 = k6WW
J3: rate of earthworm death
J3 = 0.38 ind/d
Derivation: J2 = J3 at steady state
J4 = k3SOW
J4: rate of void creation by earthworm burrowing
J4 = 3.915 x 10 4 m3/d
Derivation: 2.5 cm/d/ind (average length of burrow created per day, (Bastardie et al.,
2003)   (0.8 cm  2)2 (average burrow diameter, Guild (1955))  310.7 ind
= 391.5 cm3/d  1 m3/106 cm3 = 3.915 x 10 4 m3/d
J5 = k4VH/O
J5: rate of void collapse due to wetting
J5 = 3.915 x 10 4 m3/d
Derivation: at steady state, J5 = J4
J6 = I(1-kC)
J6: in ltration rate
J6 = 0.158 m3/d
Derivation: at kCmax = 1
Cmax = 6" = 15.24 cm (assumes entire ponding depth is  lled)
k = 1/ Cmax = 0.066 cm 1
266
J6 = (0.162 m3)  (1-(0.066 cm 1)  (0.4 cm)) = 0.158 m3/d
J7 = k5H
J7: ex ltration rate
J7 = 2.45 m3/d
Derivation: based on results of microcosm experiment. Average initial ex ltration rate,
control group = 10.22 cm/h 1/100 m/cm 24 h/d 1 m2 (surface area) = 2.45 m3/d
J8 = k7DI
J8: rate of crust growth (cm/d)
J8 = 3.46 x 10 4 cm/d
24% of TSS is captured at the soil surface (Li and Davis, 2008b)
Crust growth rate = 0.24  D  I = 0.24  (8.9 x 10 3 cm/m3)  (0.162 m3)
= 3.46 x 10 4 cm
J9 = k8CW
J9: rate of crust removal by earthworms
J9 = 3.46 x 10 4 cm
Derivation: At steady state, assume J8 = J9
J10 = k9O
J10: rate of microbial degradation of organic matter
J10 = 0
Derivation: assumed for simplicity
267
Bibliography
ASTM. 2000. Standard test methods for moisture, ash, and organic matter of peat and
other organic soils. Tech. Rep. ASTM D2974-00, American Society for Testing and
Materials, West Conshohocken, PA.
ASTM. 2003. Standard test method for in ltration rate of soils in  eld using
Double-Ring in ltrometer. Tech. Rep. ASTM D 3385-03, American Society for Testing
and Materials, West Conshohocken, PA.
Ayers, E. and P. Kangas. 2004. A hydro-ecological model of in ltration in bioretention
soils. In Putting the LID on Stormwater Management, September 21-23, 2004, College
Park, MD: Metropolitan Washington Council of Governments.
Baker, G. and V. Barrett. 1994. Earthworm Identi er. Melbourne, Australia: CSIRO.
Barois, I., G. Villemin, P. Lavelle and F. Toutain. 1993. Transformation of the soil
structure through Pontoscolex corethrurus (Oligochaeta) intestinal tract. Geoderma
56:57{66.
Bastardie, F., Y. Capowiez, J. de Dreuzy and D. Cluzeau. 2003. X-ray tomographic and
hydraulic characterization of burrowing by three earthworm species in repacked soil
cores. Applied Soil Ecology 24:3{16.
Baumann, D. and W. D. Keller. 1975. Bulk densities of selected dried natural and  red
kaolin clays. Clays and Clay Minerals 23:424{427.
Berry, E., J. Hat eld and B. Stewart. 1994. Earthworms and other fauna in the soil. In
Soil Biology: E ects on Soil Quality, 61{90, Boca Raton, FL: CRC Press.
Binet, F. and P. Trehen. 1992. Experimental microcosm study of the role of Lumbricus
terrestris (Oligochaeta:Lumbricidae) on nitrogen dynamics in cultivated soils. Soil
Biology and Biochemistry 24(12):1501{1506.
Blanchart, E. 1992. Restoration by earthworms (megascolecidae) of the macroaggregate
structure of a destructured savanna soil under  eld conditions. Soil Biology and
Biochemistry 24(12):1587{1594.
Bouch e, M. and F. Al-Addan. 1997. Earthworms, water in ltration and soil stability:
some new assessments. Soil Biology and Biochemistry 29(3/4):441{452.
Bouch e, M., U. Lohm and T. Persson. 1977. Strat egies lombriciennes, Ecological
Bulletins, vol. 25, 122{132. Stockholm: Swedish Natural Reseach Council, NFR.
Brady, N. and R. Weil. 2002. The Nature and Properties of Soils, 13th ed. Upper Saddle
River, N.J.: Prentice Hall.
Brown, M. T. 2004. A picture is worth a thousand words: energy systems language and
simulation. Ecological Modelling 178:83{100.
Capowiez, Y. 2000. Di erences in burrowing behaviour and spatial interaction between
the two earthworm species Aporrectodea nocturna and Allolobophora chlorotica.
Biology and Fertility of Soils 30:341{346.
268
Capowiez, Y. and Belzunces. 2001. Dynamic study of the burrowing behaviour of
Aporrectodea nocturna and Allolobophora chlorotica: interactions between earthworms
and spatial avoidance of burrows. Biology and Fertility of Soils 33:310{316.
Capowiez, Y., P. Renault and L. Belzunces. 2001. Three-dimensional trajectories of
60Co-labelled earthworms in arti cial cores of soil. European Journal of Soil Science
52:365{375.
Carpenter, S. R. 1996. Microcosm experiments have limited relevance for community and
ecosystem ecology. Ecology 77(3):677{680.
CBF. 2005. Build Your Own Rain Garden. Annapolis, MD: Chesapeake Bay Foundation.
Coleman, D., D. Crossley and P. Hendrix. 2004. Fundamentals of Soil Ecology, 2nd ed.
Amsterdam: Elsevier.
Crocker, R. 1952. Soil genesis and the pedogenic factors. The Quarterly Review of
Biology 27(2):139{168.
CSPMA. 1996. Canadian sphagnum peat moss - a horticultural teaching plan.
http://www.peatmoss.com/hortprog1.php.
Csuzdi, C. and K. Sl avecz. 2003. Lumbricus friendi Cognetti, 1904, a new exotic
earthworm in North America. Northeastern Naturalist 10(1):77{82.
Culbertson, T. L. and S. L. Hutchinson. 2004. Assessing bioretention cell function in a
midwest continental climate. paper number 047051. In 2004 ASAE/CSAE Annual
International Meeting, 1-4 August 2004, Ottawa, Ontario, Canada: American Society
of Agricultural Engineers.
Curry, J. 1998. Factors a ecting earthworm abundance in soils. In C. Edwards, ed.,
Earthworm Ecology, 37{64, Boca Raton, FL: St. Lucie Press.
Curry, J. and D. Cotton. 1983. Earthworms and land reclamation. In J. Satchell, ed.,
Earthworm Ecology: From Darwin to Vermiculture, 215{228, London: Chapman and
Hall, Ltd.
Darwin, C. 1881. The Formation of Vegetable Mould, through the Action of Worms,
with Observations on Their Habits. Republished in 1985. Chicago, IL: University of
Chicago Press.
Davis, A., M. Shokouhian and S. Ni. 2001a. Loading estimates of lead, copper, cadmium
and zinc in urban runo from speci c sources. Chemosphere 44(5):997{1009.
Davis, A., M. Shokouian, H. Sharma and C. Minami. 2001b. Laboratory study of
biological retention (Bioretention) for urban storm water management. Water
Environment Research 73(1):5{14.
Davis, A., M. Shokouian, H. Sharma, C. Minami and D. Winogrado . 2003. Water
quality improvement through bioretention: lead, copper and zinc. Water Environment
Research 75(1):73{82.
Davis, A. P. 2007. Field performance of bioretention: water quality. Environmental
Engineering Science 24(8):1048{1063.
269
Davis, A. P. 2008. Field performance of bioretention: hydrology impacts. Journal of
Hydrologic Engineering 13(2):90{95.
Davis, A. P., M. Shokouhian, H. Sharma and C. Minami. 2006. Water quality
improvement through bioretention media: nitrogen and phosphorus removal. Water
Environment Research 78(3):284{293.
Dietz, M. E. and J. C. Clausen. 2005. A  eld evaluation of rain garden  ow and
pollutant treatment. Water 167(1-4):123{138.
Dietz, M. E. and J. C. Clausen. 2006. Saturation to improve pollutant retention in a rain
garden. Environmental Science and Technology 40(4):1335{1340.
Dougherty, M., C. LeBleu, E. Brantley and C. Francis. 2007. Evaluation of bioretention
nutrient removal in a rain garden with an internal water storage (IWS) layer. paper
number: 077085. In 2007 ASABE Annual International Meeting, June 17-20, 2007,
Minneapolis Convention Center, Minneapolis, Minnesota: ASABE.
Eaton, T. 1942. Earthworms of the northeastern United States: a key, with distribution
records. Journal of the Washington Academy of Sciences 32(8):242{249.
Edwards, C. 1998. Earthworm Ecology. Boca Raton, FL: St. Lucie Press.
Edwards, C. and N. Arancon. 2004. Interactions among organic matter, earthworms, and
microorganisms in promoting plant growth. In F. Magdo and R. Weil, eds., Soil
Organic Matter in Sustainable Agriculture, 327{376, Boca Raton, FL: CRC Press.
Edwards, C. and P. Bohlen. 1996. Biology and Ecology of Earthworms. London:
Chapman & Hall.
Emerson, C. H. and R. G. Traver. 2008. Multiyear and seasonal variation of in ltration
for storm-water best management practices. Journal of Irrigation and Drainage
Engineering 134(5):598{605.
Fender, W. M. and D. McKey-Fender. 1990. Oligochaeta: Megascolecidae and other
earthworms from western North America. In D. L. Dindal, ed., Soil Biology Guide,
357{377, New York: John Wiley & Sons.
Ferguson, B. 1994. Stormwater In ltration. Boca Raton: CRC Press.
Forster, S. M. 1990. The role of microorganisms in aggregate formation and soil
stabilization: types of aggregation. Arid Soil Research and Rehabilitation 4:85{98.
Foster, B. L. and D. Tilman. 2000. Dynamic and static views of succession: Testing the
descriptive power of the chronosequence approach. Plant Ecology 146:1{10.
Francis, G. S. and P. M. Fraser. 1998. The e ects of three earthworm species on soil
macroporosity and hydraulic conductivity. Applied Soil Ecology 10:11{19.
Frouz, J., B. Keplin, V. Pizl, K. Tajovsky, J. Stary, A. Lukesov, A. Novkov, V. Balk,
L. Hnel, J. Materna, C. Dker, J. Chalupsky, J. Rusek and T. Heinkele. 2001. Soil biota
and upper soil layers development in two contrasting post-mining chronosequences.
Ecological Engineering 17:275{284.
270
Gates, G. 1937. The genus Pheretima in North America. Bulletin of the Museum of
Comparative Zoology, Harvard University 80:339{373.
Gates, G. 1958. On some species of the oriental earthworm genus Pheretima Kinberg,
1867, with key to species reported from the Americas. American Museum Novitates
1888:1{33.
Gates, G. 1972a. Contributions to North American Earthworms (Annelida), No. 4: On
American earthworm genera. I. Eisenoides (Lumbricidae). Bulletin of Tall Timbers
Research Station 3(13):1{17.
Gates, G. 1972b. Contributions to North American Earthworms (Annelida), No. 5: On
variation in another anthropochorous species of the oriental earthworm genus
Pheretima Kinberg 1866 (Megascolecidae). Bulletin of Tall Timbers Research Station
3(13):18{44.
Gates, G. 1973. Contributions to North American Earthworms (Annelida), No. 8: The
Earthworm Genus Octolasion in America. Bulletin of Tall Timbers Research Station
4(14):29{50.
Gates, G. 1974. Contributions to North American Earthworms (Annelida), No. 10:
Contributions to a revision of the Lumbricidae. X., Dendrobaena octaedra (Savigny)
1826, with special reference to the importance of its parthenogenetic polymorphism for
the classi cation of earthworms. Bulletin of Tall Timbers Research Station
5(15):15{57.
Glancey, J. L. and S. C. Ho man. 1996. Physical properties of solid waste materials.
Applied Engineering in Agriculture 12(4):441{446.
Gobat, J., M. Aragno and W. Matthey. 2003. The Living Soil: Fundamentals of Soil
Science and Soil Biology. En eld, NH: Science Publishers.
Gonzalez-Sangregorio, M., M. Trasar-Cepeda, M. Leiros, F. Gil-Sotres and
F. Guitian-Ojea. 1991. Early stages of lignite mine soil genesis: changes in biochemical
properties. Soil Biology and Biochemistry 23(6):589{595.
Greene, A., S. L. Hutchinson, R. Christianson and T. L. Culbertson. 2008. Impacts of
ecological diversity on bioretention cell performance during establishment in the
midwest. paper number: 084758. In 2008 ASABE Annual International Meeting, June
29 - July 2, 2008, Rhode Island Convention Center, Providence, Rhode Island:
ASABE.
Guild, W. J. M. 1955. Earthworms and soil structure. In D. K. M. Kevan, ed., Soil
Zoology. Proceedings of the University of Nottingham Second Easter School in
Agricultural Science, 1955., 83{98, London: Butterworths Scienti c Publications.
Gupta, S. C., A. Bhattacharjee, J. F. Moncrief, E. C. Berry and J. E. Zachmann. 2001.
In uence of earthworm species and depth of residue placement on macropore
characteristics and preferential transport. In D. D. Bosch and K. W. King, eds.,
Preferential Flow, Water Movement and Chemical Transport in the Environment,
Proc. 2nd Int. Symp. (3-5 January 2001, Honolulu, Hawaii, USA), 153{156, St.
Joseph, MI: ASAE.
271
Gutierrez, L. T. and W. R. Fey. 1980. Ecosystem Succession: A General Hypothesis and
a Test Model of a Grassland. Cambridge, MA: MIT Press.
Hillel, D. 1998. Environmental Soil Physics. San Diego: Academic Publishers.
Hole, F. 1981. E ects of animals on the soil. Geoderma 25:75{112.
Hong, E., E. A. Seagren and A. P. Davis. 2006. Sustainable oil and grease removal from
synthetic stormwater runo using bench-scale bioretention studies. Water
Environment Research 78(2):141{155.
Hoogerkamp, M., H. Rogaar and H. Eijsackers. 1983. E ect of earthworms on grassland
on recently reclaimed polder soils in the netherlands. In J. Satchell, ed., Earthworm
Ecology: From Darwin to Vermiculture, 85{105, London: Chapman and Hall, Ltd.
Hsieh, C. H. and A. Davis. 2005a. Evaluation and optimization of bioretention media for
treatment of urban storm water runo . Journal of Environmental Engineering, ASCE
131(11):1521{1531.
Hsieh, C. H. and A. P. Davis. 2005b. Multiple-event study of bioretention for treatment
of urban storm water runo . Water Science and Technology 51(3-4):177{181.
Hsieh, C. H., A. P. Davis and B. A. Needelman. 2007a. Bioretention column studies of
phosphorus removal from urban stormwater runo . Water Environment Research
79(2):177{184.
Hsieh, C. H., A. P. Davis and B. A. Needelman. 2007b. Nitrogen removal from urban
stormwater runo through layered bioretention columns. Water Environment Research
79(12):2404{2411.
Hunt, W. F., A. R. Jarrett, J. T. Smith and L. J. Sharkey. 2006. Evaluating bioretention
hydrology and nutrient removal at three  eld sites in North Carolina. Journal of
Irrigation and Drainage Engineering 132(6):600{608.
Hunt, W. F., J. T. Smith, S. J. Jadlocki, J. M. Hathaway and P. R. Eubanks. 2008.
Pollutant removal and peak  ow mitigation by a bioretention cell in urban Charlotte,
N.C. Journal of Environmental Engineering 134(5):403{408.
James, W. M. 1990. Oligochaeta: Megascolecidae and other earthworms from southern
and midwestern North America. In D. L. Dindal, ed., Soil Biology Guide, 379{386,
New York: John Wiley & Sons.
Jenny, H. 1941. Factors of Soil Formation. New York: McGraw-Hill.
Jenny, H. 1980. The Soil Resource. New York: Springer-Verlag.
Johnson, D. and D. Watson-Stegner. 1987. Evolution model of pedogenesis. Soil Science
143(5):349{366.
Jones, M. P. and W. F. Hunt. 2008. The e ect of bioretention on runo temperature in
trout sensitive waters. paper number: 084072. In 2008 ASABE Annual International
Meeting, June 29 - July 2, 2008, 7, Rhode Island Convention Center, Providence,
Rhode Island: ASABE.
272
Joschko, M., W. Sochtig and O. Larink. 1992. Functional relationship between
earthworm burrows and soil water movement in column experiments. Soil Biology and
Biochemistry 24(12):1545{1547.
Kadlec, R. and R. Knight. 1996. Treatment Wetlands. Boca Raton, FL: CRC Press.
Kangas, P. C. 1979. Succession as an alternative for reclaiming phosphate spoil mounds.
Kay, B. and D. Angers. 1999. Soil structure. In Handbook of Soil Science, A{229{A{275,
Boca Raton, FL: CRC Press.
Kim, H., E. Seagren and A. Davis. 2003. Engineered bioretention for removal of nitrate
from stormwater runo . Water Environment Research 75(4):355{367.
Kretzschmar, A. and C. Edwards. 2004. E ects of Earthworms on Soil Organization,
201{209. Boca Raton, FL: CRC Press.
Lachnicht, S., R. Parmelee, D. McCartney and M. Allen. 1997. Characteristics of
macroporosity in a reduced tillage agroecosystem with manipulated earthworm
populations: implications for in ltration and nutrient transport. Soil Biology and
Biochemistry 29(3/4):493{498.
Lakhani, K. and J. Satchell. 1970. Production by Lumbricus terrestris (L.). Journal of
Animal Ecology 39:473{492.
Langmaack, M., S. Schrader and U. Rapp-Bernhardt. 1999. Quantitative analysis of
earthworm burrow systems with respect to biological soil-structure regeration after soil
compaction. Biology and Fertility of Soils 28:219{229.
Lee, K. 1959. The Earthworm Fauna of New Zealand. Wellington, New Zealand: New
Zealand Department of Scienti c and Industrial Research.
Lee, K. 1985. Earthworms: Their Ecology and Relationships with Soils and Land Use.
Orlando, FL: Academic Press.
Lee, K. and R. Foster. 1991. Soil fauna and soil structure. Australian Journal of Soil
Research 29:745{775.
Leisman, G. A. 1957. A vegetation and soil chronosequence on the Mesabi Iron Range
Spoil Banks, Minnesota. Ecological Monographs 27(3):221{245.
Li, H. 2007. Urban particle and pollutant capture via stormwater  lter facilities and the
concomittant water quality and hydrological bene ts. Ph.D. thesis, University of
Maryland.
Li, H. and A. P. Davis. 2008a. Heavy metal capture and accumulation in bioretention
media. Environmental Science and Technology 42(14):5247{5253.
Li, H. and A. P. Davis. 2008b. Urban particle capture in bioretention media. i:
Laboratory and  eld studies. Journal of Environmental Engineering 134(6):409{418.
Li, H. and A. P. Davis. 2008c. Urban particle capture in bioretention media. II: theory
and model development. Journal of Environmental Engineering 134(6):419{432.
273
LIDC. 2003. Low impact development center - bioretention speci cation.
http://www.lowimpactdevelopment.org/epa03/biospec.htm.
McCuen, R. H. 2005. Hydrologic Analysis and Design, 3rd ed. Upper Saddle River, N.J.:
Prentice Hall.
MCDEP. 2004. Fact Sheet: Chevy Chase Bank Low Impact Stormwater Project.
Rockville, MD: Montgomery County Department of Environmental Protection.
Muthanna, T. M., M. Viklander, G. Blecken and S. T. Thorolfsson. 2007a. Snowmelt
pollutant removal in bioretention areas. Water Research 41(18):4061{4072.
Muthanna, T. M., M. Viklander and S. T. Thorolfsson. 2007b. An evaluation of applying
existing bioretention sizing methods to cold climates with snow storage conditions.
Water Science and Technology 56(10):73{81.
NCDC. 2003. Climate inventories.
http://www.ncdc.noaa.gov/oa/climate/climateinventories.html.
NCDC. 2008. Climate inventories.
http://www.ncdc.noaa.gov/oa/climate/climateinventories.html.
Oades, J. 1993. The role of biology in the formation, stabilization and degradation of soil
structure. Geoderma 56:377{400.
Odum, H. T. 1971. Environment Power and Society. New York: John Wiley & Sons.
Ott, R. L. and M. Longnecker. 2001. An Introduction to Statistical Methods and Data
Analysis, 5th ed. Paci c Grove, CA: Duxbury.
PGCO. 2000. Low-Impact Development Hydrologic Analysis. Largo, MD: Programs and
Planning Division, Department of Environmental Resources, Prince George?s County,
MD.
PGCO. 2001. The Bioretention Manual. Largo, MD: Programs and Planning Division,
Department of Environmental Resources, Prince George?s County, MD.
PGCO. 2004a. Inglewood Center III Low Impact Development Project. Factsheet.
Largo, MD: Programs and Planning Division, Department of Environmental
Resources, Prince George?s County, MD.
PGCO. 2004b. "Mother" Jones Elementary School Low Impact Development Retro t
Project. Factsheet. Largo, MD: Programs and Planning Division, Department of
Environmental Resources, Prince George?s County, MD.
PGCO. 2004c. Northwestern High School Low Impact Development Retro t Project.
Factsheet. Largo, MD: Programs and Planning Division, Department of
Environmental Resources, Prince George?s County, MD.
Pickett, S., M. Cadenasso, J. Grove, C. Nilon, R. Pouyat, W. Zipperer and R. Costanza.
2001. Urban ecological systems: Linking terrestrial, ecological, physical, and
socioeconomic components of metropolitan areas. Annual Review of Ecology and
Systematics 32:127{157.
274
Pickett, S. T. 1989. Space-for-time substitution as an alternative to long-term studies. In
Long-Term Studies in Ecology: Approaches and Alternatives, 110{135, New York:
Springer-Verlag.
Ponder, F., F. Li and D. Jordan. 2000. Assessing the impact of Diplocardia ornata on
physical and chemical properties of compacted forest soil in microcosms. Biology and
Fertility of Soils 32:166{172.
Privette, C. V. and B. L. Weeber. 2007. Evaluation of an anaerobically enhanced
bioretention cell?s treatment of  rst  ush from highway runo . paper number: 072030.
In 2007 ASABE Annual International Meeting, June 17-20, 2007, Minneapolis
Convention Center, Minneapolis, Minnesota: ASABE.
Privette, C. V. and B. L. Weeber. 2008. Nutrient and metal removal e ciency of an
anaerobically enhanced bioretention cell. paper number: 084209. In 2008 ASABE
Annual International Meeting, June 29 - July 2, 2008, Rhode Island Convention
Center, Providence, Rhode Island: ASABE.
Ramakrishna, D. M. and T. Viraraghavan. 2005. Environmental impact of chemical
deicers a review. Water, Air, and Soil Pollution 166(1):49{63.
Rawls, W. J., D. L. Brakensiek and K. E. Saxton. 1982. Estimation of soil water
properties. Transactions of the ASAE 25(5):1316{1320.
Reynolds, J. 1974. The earthworms of Maryland (Oligochaeta: Acanthodrilidae,
Lumbricidae, Megasolecidae and Sparganophilidae). Megadrilogica 1(11):1{12.
Reynolds, J. W., E. E. C. Clebsch and W. M. Reynolds. 1974. Contributions to North
American Earthworms (Oligochaeta), No. 13: The Earthworms of Tennessee
(Oligochaeta). I. Lumbricidae. Bulletin of Tall Timbers Research Station 3(17):1{133.
Roberts, J., W. Daniels, J. Bell and J. Burger. 1988a. Early stages of mine soil genesis as
a ected by topsoiling and organic amendments. Soil Science Society of America
Journal 52:730{738.
Roberts, J., W. Daniels, J. Bell and J. Burger. 1988b. Early stages of mine soil genesis in
a southwest Virginia spoil lithosequence. Soil Science Society of America Journal
52:716{723.
Rusciano, G. M. and C. C. Obropta. 2007. Bioretention column study: fecal coliform and
total suspended soids reductions. Transactions of the ASABE 50(4):1261{1269.
Satchell, J. 1967. Lumbricidae. In A. Burges and F. Raw, eds., Soil Biology, 259{322,
New York: Academic Press.
Schrader, S. and H. Zhang. 1997. Earthworm casting: stabilization or destabilization of
soil structure? Soil Biology and Biochemistry 29(3/4):469{475.
Schwert, D. P. 1990. Oligochaeta: Lumbricidae. In D. L. Dindal, ed., Soil Biology Guide,
341{356, New York: John Wiley & Sons.
Scullion, J. and A. Malik. 2000. Earthworm activity a ecting organic matter,
aggregation and microbial activity in soils restored after opencast mining for coal. Soil
Biology and Biochemistry 32(1):119{126.
275
Sharkey, L. J. and W. F. Hunt. 2005. Design implications on bioretention performance as
a stormwater BMP: water quality and quantity. paper number: 052201. In 2005 ASAE
Annual International Meeting, July 17-20, 2005, Tampa Convention Center, Tampa,
Florida: ASAE.
Shaw, C. and S. Pawluck. 1986. The development of soil structure by Octolasion
tyrtaeum, Aporrectodea turgida and Lumbricus terrestris in parent materials belonging
to di erent textural classes. Pedobiologia 29:327{339.
Sims, R. W., B. M. Gerard, R. S. K. Barnes and J. H. Crothers. 1999. Earthworms.
Synopses of the British Fauna, Shrewsbury, UK: Linnean Society of London.
Six, J., J. Bossuyt, S. Degryze and K. Denef. 2004. A history of research on the link
between (micro)aggregates, soil biota, and soil organic matter dynamics. Soil and
Tillage Research 79:7{31.
Southwood, T. and P. Henderson. 2000. Ecological Methods, 3rd ed. Malden, MA:
Blackwell Science Ltd.
Stevens, P. and T. Walker. 1970. The chronosequence concept and soil formation. The
Quarterly Review of Biology 45(4):333{350.
Sun, X. and A. P. Davis. 2007. Heavy metal fates in laboratory bioretention systems.
Chemosphere 66(9):1601{1609.
Thompson, A. M., A. C. Paul and N. J. Balster. 2008. Physical and hydraulic properties
of engineered soil media for bioretention basins. Transactions of the ASAE
51(2):499{514.
UNHSC. 2007. 2007 annual report.
http://www.unh.edu/erg/cstev/2007 stormwater annual report.pdf.
Urb anek, J. and F. Dole zal. 1992. Review of some case studies on the abundance and on
the hydraulic e ciency of earthworm channels in Czechoslovak soils, with reference to
the subsurface pipe drainage. Soil Biology and Biochemistry 24(12):1563{1571.
van Breemen, N. and P. Buurman. 2002. Soil Formation, 2nd ed. Dordrecht, The
Netherlands: Kluwer Academic Publishers.
van Vliet, P. and P. Hendrix. 1999. Enchytraeids. In Handbook of Soil Science,
C{70{C{85, CRC Press.
Weil, R., F. Magdo , F. Magdo and R. Weil. 2004. Signi cance of soil organic matter
to soil quality and health. In Soil Organic Matter in Sustainable Agriculture, 1{43,
Boca Raton, FL: CRC Press.
Weil, R. R. 1998. Laboratory Manual for Introductory Soils. Dubuqe, IA: Kendall/Hunt
Publishing Co., 6th ed.
Weil, R. R., K. R. Islam, M. A. Stine, J. B. Gruver and S. E. Samson-Liebig. 2003.
Estimating active carbon for soil quality assessment: A simpli ed method for
laboratory and  eld use. American Journal of Alternative Agriculture 18(1):3{17.
276
Worm Watch Canada. 2002. Field guide to earthworms.
http://www.naturewatch.ca/english/wormwatch/resources/guide/index.html.
Yaalon, D. 1975. Conceptual models in pedogenesis: can soil-forming functions be
solved? Geoderma 14:189{205.
Yaalon, D. 1983. Climate, time and soil development. In L. Wilding, N. Smeck and
G. Hall, eds., Pedogenesis and Soil Taxonomy. I. Concepts and Interactions.,
Developments in Soil Science, vol. 11A, 233{251, Amsterdam: Elsevier.
Zhang, W., G. O. Brown and D. E. Storm. 2006. Enhancement of phosphorus removal in
bioretention cells by soil amendment. paper number: 062193. In 2006 ASABE Annual
International Meeting, July 9-12, 2006, Oregon Convention Center, Portland, Oregon:
ASABE.
277