 

€2.15

4. .

:43
. . .h

an: E

{.

.
L

.1!

he.

. .—

3....

x;

. . .. raw.
Mg
‘

t:
$325...
.. 12x31.
Kin
it.

wag“. 2%
‘ .. %1 a

i

3519‘
WW»! Sun.» $
8.2.. SP

.. a

.3331».

‘0. III! '|

5;...1‘1 {.2 2
xiualciw.

i... {33.
III-‘3'Ia 1‘9:
€53.25);
32.4.. 3:!!!
1.... ..
.51

i3 .15. 3‘1
1 152.19% .101 .
3.. ‘3‘. .1!l\..lu>t:§:n
sixls‘lvt? a a K .123... I .
1:53.19. til . .5, t
3| 9.. n; V,
x1.

\I....!
.52: 7 .

 

 

 

I.-

' ' LIBRARY

" . ~' -' O§n§
lVllDlllgan \Plale
Umversuty

 

 

This is to certify that the

dissertation entitled

EXTENSIVE GREEN ROOFS: CARBON SEQUESTRATION

POTENTIAL AND SPECIES EVALUATIONS

presented by

KRISTIN LOUISE GETTER

has been accepted towards fulﬁllment

of the requirements for the

 

 

 

PhD degree in Horticulture
M'ajor Professor’s Signature
3 ~ 3 l —- o 7
Date

 

MSU is an Afﬁnnative Action/Equal Opportunity Employer

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K.IPrq/Acc&Pres/ClRCIDateDue.indd

EXTENSIVE GREEN ROOFS: CARBON SEQUESTRATION POTENTIAL AND
SPECIES EVALUATIONS

By

Kristin Louise Getter

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Horticulture

2009

 

ABSTRACT

EXTENSIVE GREEN ROOFS: CARBON SEQUESTRATION POTENTIAL AND
SPECIES EVALUATIONS

BY

Kristin Louise Getter
Green roofs, or vegetated roofs, are often adopted for energy savings and can
thus mitigate climate change by lowering demand for heating and air conditioning
use which results in less carbon dioxide emitted from power plants and furnaces.
The goal of this research was to quantify the intrinsic carbon storage potential of
extensive green roofs by plants and soils. Plots of four species of Sedum plus a
substrate only treatment were established on a roof with a 6.0 cm substrate
depth, replicated four times. Carbon analysis was performed by sampling above-
ground biomass, below-ground biomass (roots), and substrate carbon content,
harvested seven times across two growing seasons. After two growing seasons,
above-ground plant material stored an average of 168 g C-m‘2 and below-ground
biomass stored an average of 107 g C-m‘z. Substrate carbon content averaged
913 g C-m'z, sequestering 100 g C-rn'2 beyond what was in the initial substrate.
The entire extensive green roof system sequestered 375 g C-m'z. There was a

species effect on above- and below-ground carbon, but not on substrate carbon.

Proper plant selection for green roofs is of the utmost importance, for if the plants
fail, the energy savings and carbon mitigation will likely be reduced and future
green roof sales may be affected. Two additional studies evaluated the effect of

green roof substrate depth on plant community development over four years in a

Midwestern climate. The ﬁrst study evaluated plugs of 12 Sedum species bi-
weekly for absolute cover (AC) at 4.0 cm, 7.0 cm, and 10.0 cm substrate depths.
Most species exhibited greater growth and coverage at a substrate depth of 7.0
cm and 10.0 cm relative to 4.0 cm. Species exhibiting the greatest AC at all
substrate depths were 8. ﬂon'ferum, S. sexangulare, S. spun'um ‘John Creech’,
and S. stefco. In general, species that are less suitable at these substrate depths
are 8. ‘Angelina’, S. cauticola ‘Lidakense’, S. ewersii, S. ochroleucum, and S.
reﬂexum ‘Blue Spruce’. The second study quantiﬁed the effect of solar radiation
(full sun vs. full shade) on several US. native and non-native species for
potential use on extensive green roofs. Plugs of six native and three non-native
species and seed of six typical extensive green roof species of Sedum were
established at three different substrate depths (8.0 cm, 10.0 cm, and 12.0 cm)
both in sun and shade. AC was recorded as before. By week 174 (23 Sept
2008), most species exhibited different AC within a depth between sun and
shade. However, when all species were combined, overall AC did not differ
between sun and shade within a depth. This indicated that while species make-
up was changing among solar radiation Ievels, that overall coverage was not
signiﬁcantly different. The most abundant species were AIIium cemuum, 8. acre,
8. album, 8. kamtschaticum, S. spurium and Talinum calycinum. Less suitable
species included Carex ﬂacca, S. divergens, S. pulchellum, S. stenopetalum, and
Talinum parviﬂorum. With the exception of T. calycinum. native species were
less abundant than non-native species. AIIium cemuum, 8. acre, S.

kamtschaticum, and S. spun'um are suitable in the shade.

This dissertation is dedicated to the memory of my grandmother, Frances
Rossman, who was instrumental in my interest in higher education. Frances
received her masters from Michigan State University and nearly completed her
PhD as well. We miss you grandma.

ACKNOWLEDGMENTS

There are many people I would like to thank for helping to make this thesis
possible. To Dr. Bradley Rowe, thank you for allowing me to research projects
interesting to me and for giving me the freedom to succeed in my own way. To
my advising committee, Dr. Jeff Andresen, Dr. Bert Cregg, and Dr. Phil
Robertson, thank you for all of the help and insight you offered in implementing
this research. Most of all, I want to thank my wonderful husband, Rick, who
gladly supported me during these many years of schooling and whom without his

encouragement I would have never returned to graduate school.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................. vii
LIST OF FIGURES .............................................................................................. ix
LITERATURE REVIEW:

GREEN ROOFS AS ONE TOOL TO MITIGATE URBANIZATION ....................... 1
Problems with Urbanization ........................................................................ 2
Green Roofs as One Tool to Mitigate Urbanization .................................... 4
Importance of Plant Selection for Green Roofs ........................................ 11
Research Objectives ................................................................................ 14
Literature Cited ......................................................................................... 17

CHAPTER ONE:

SUBSTRATE DEPTH INFLUENCES SEDUM PLANT COMMUNITY ON A

GREEN ROOF .................................................................................................... 23
Abstract .................................................................................................... 24
Introduction .............................................................................................. 25
Materials and Methods ............................................................................. 28
Results and Discussion ............................................................................ 32
Conclusions .............................................................................................. 38
Literature Cited ......................................................................................... 40

CHAPTER TWO:

SOLAR RADIATION INTENSITY INFLUENCES EXTENSIVE GREEN ROOF

PLANT COMMUNITIES ...................................................................................... 56
Abstract .................................................................................................... 57
Introduction .............................................................................................. 58
Materials and Methods ............................................................................. 62
Results and Discussion ............................................................................ 67
Conclusions .............................................................................................. 75
Literature Cited ......................................................................................... 77

CHAPTER THREE:

CARBON SEQUESTRATION POTENTIAL OF EXTENSIVE GREEN ROOFS..97
Abstract .................................................................................................... 98
Introduction .............................................................................................. 99
Materials and Methods ........................................................................... 102
Results and Discussion .......................................................................... 105
Conclusions ............................................................................................ 112
Literature Cited ....................................................................................... 114

DISSERTATION CONCLUSION ....................................................................... 130

vi

Table

1.1

‘ 1.2

1.3

2.1
2.2

2.3

LIST OF TABLES

Page

Mean absolute cover :I: standard deviations of 12 Sedum species
cultivated at three substrate depths (4.0, 7.0, and 10.0 cm) at the end of
four growing seasons (2005 to 2008). Cover reported in weeks, where
week 15 = 23 September 2005; week 67= 19 September 2006; week 120
= 24 September 2007; week 172=25 September 2008. Absolute cover
was calculated for each species at each substrate depth as the total
number of point-frame contacts divided by the number of data collection
points. ...................................................................................................... 44

Mean absolute cover 1 standard deviations of 12 Sedum species
cultivated at three substrate depths (4.0, 7.0, and 10.0 cm) reported on
week 172 (25 September 2008). Absolute cover was calculated for each
species at each substrate depth as the total number of point-frame
contacts divided by the number of data collection points. ........................ 47

Mean absolute cover :1: standard deviations averaged over species
cultivated at three substrate depths (4.0, 7.0, and 10.0 cm) over the 2005
to 2008 growing seasons. Cover reported in weeks, where week 15 = 23
September 2005; week 67= 19 September 2006; week 120 = 24
September 2007; week 172=25 September 2008. Absolute cover was
calculated for each species at each substrate depth as the total number of
point-frame contacts divided by the number of data collection points. .....48

Initial physical and chemical properties of substrate. ............................... 81

Mean absolute cover +/- standard errors of nine species at two solar
radiation intensities (sun versus shade) cultivated at two substrate depths
(8.0 cm and 12.0 cm) reported on week 174 (23 September 2008). Six
species are native to the US. (denoted with (N) after the species name).
Absolute cover was calculated for each species as the total number of
point-frame contacts divided by the number of data collection points. ..... 82

Mean substrate volumetric moisture content (m3/m3) +/- standard errors of
two substrate depths (8.0 cm and 12.0 cm) at two solar radiation
intensities (sun versus shade) averaged over the 2007 and 2008 growing
seasons. .................................................................................................. 83

vii

2.4

3.1

3.2

3.3

3.4

3.5

Mean absolute cover +/- standard errors of 6 Sedum species at two solar
radiation intensities (sun versus shade) cultivated at a substrate depth of
10.0 cm reported on week 174 (23 September 2008). Absolute cover was
calculated for each species as the total number of point-frame contacts

divided by the number of data collection points. ..................................... 84
Location and description of sampled green roofs. Each location was

sampled in four random places on sampling date. ................................. 118
Initial Physical and Chemical Properties of Substrate. ........................... 119

Mean carbon (g0m'2) 1 standard errors for above-ground biomass on
twelve extensive green roofs. ................................................................. 120

ANOVA table for mean carbon (g-m'z) over two growing seasons (April
2007 to October 2008) for four species and a substrate only treatment
replicated four times. Grams of carbon per square meter of green roof is
the dependent variable. Date harvested“, species”, and partition" are
independent variables. ........................................................................... 121

Mean carbon per square meter and mean percent carbon t standard
errors at end of second growing season (13 October 2008). ................. 122

viii

Figure

1.1

1.2

1.3

1.4

2.1

2.2

2.3

LIST OF FIGURES

Page

Absolute cover 1: standard deviations of 12 Sedum spp. cultivated at three
substrate depths (4.0, 7.0, and 10.0 cm) over the 2005-2008 growing
seasons. Symbols represent absolute cover means with standard
deviations (n=3). For clarity of graphing, weeks 0 through 120 were
plotted at every other data point. .............................................................. 49

Monthly average maximum air temperatures (°C), monthly average
minimum air temperatures (°C), and monthly total precipitation (mm)
throughout the study (1 June 2005 to 30 September 2008). Data is from
the Michigan Automated Weather Network’s East Lansing weather station
(located adjacent to the research site). .................................................... 53

Substrate volumetric moisture content (m3/m3) :l: standard errors for three
substrate depths (4.0, 7.0, and 10.0 cm) averaged over two growing
seasons (2007 and 2008). Uppercase letters represent mean separation
by LSD (P5005) ....................................................................................... 54

Substrate volumetric moisture content (m3/m3) 1 standard errors for a
selected three day period following a 22.1 mm rain event at three
substrate depths (4.0, 7.0, and 10.0 cm). ................................................. 55

Representative data for maximum daily solar radiation intensities (kW/m2)
and maximum daily air temperatures (C) during the second year
(1 January 2006 to 31 December 2006) ................................................... 85

Absolute cover of 9 species cultivated at two substrate depths (8.0 cm and
12.0 cm) at two solar radiation intensities (sun versus shade) during the
four growing seasons (2005-2007). Six species are native to the US.
(denoted with (N) after species name). Symbols represent absolute cover
means with standard errors (n=9). For clarity of graphing, weeks 0
through 120 were plotted at every other data point. ................................. 86

Mean chlorophyll ﬂuorescence (Fv/Fm) with standard errors of 9 species
averaged over two substrate depths (8.0 cm and 12.0 cm) at two solar
radiation intensities (sun versus shade) over the last two growing seasons
(2007 - 2008). Lower case letters denote mean separation by LSD
(P5005). .................................................................................................. 91

2.4

2.5

2.6

3.1

3.2

3.3

3.4

Absolute cover of 6 Sedum species cultivated at a 10.0 cm substrate
depth at two solar radiation intensities (sun versus shade) over four
growing seasons (2005 - 2008). Symbols represent absolute cover
means with standard errors (n=3). For clarity of graphing, weeks 0
through 120 were plotted at every other data point. ................................. 92

Mean substrate volumetric moisture content (m3/m3) with standard errors
at two solar radiation intensities (sun versus shade) over the 2007 and
2008 growing seasons. .......................................................................... 95

Mean chlorophyll ﬂuorescence (FvlFm) with standard errors of 6 Sedum
species cultivated at 10.0 cm substrate depth at two solar radiation
intensities (sun versus shade) over the last two growing seasons (2007 -
2008) where SA=Sedum acre; SAL=S. album; SK=S. kamtschaticum;
SP=S. pulchellum; SPUR=S. spun'um ‘Coccineum’; SR=S. reﬂexum;
OVERALL=alI species combined. *=Signiﬁcant at P=0.1; **=Signiﬁcant at
P=.05; n.s.=not signiﬁcant. ....................................................................... 96

Simple linear regression of above-ground biomass on twelve extensive
green roofs, each sampled four times. ................................................... 123

Monthly average maximum air temperatures (°C), monthly average
minimum air temperatures (°C), and monthly total precipitation (mrn)
throughout the study (1 April 2007 to 31 October 2008). Data is from the
nearby Michigan Automated Weather Network’s East Lansing, MI weather
station ..................................................................................................... 124

Carbon content :t standard errors for above-ground biomass, root
biomass, and substrate across two growing seasons (April 2007 to
October 2008). ....................................................................................... 125

Carbon content (9 C-m'z) :t standard errors for above-ground biomass,
root biomass, and substrate across two growing seasons for four species
and substrate only treatment (April 2007 to October 2008) .................... 126

A LITERATURE REVIEW

Green Roofs as a Tool to Mitigate Urbanization

Problems with Urbanization

According to the United Nations Department of Economic and Social Affairs (UN
ESA), the world’s urban population is rapidly increasing. In 2005, 49% of the
world population resided in urban areas worldwide (UN ESA, 2007). This
number is expected to grow to 60% by 2030. As urban areas expand,
constructed surfaces will increase in the form of buildings, roads, and parking

lots.

In the continental United States, urban areas are covered by up to 95% in
impervious surfaces (Ferguson, 1998). As a result, an average of only 25% of
rainfall inﬁltrates into soil (Scholz-Barth, 2001) and the remaining 75% generates
urban runoff that may be highly contaminated with pollutants such as oil, heavy
metals, salts, pesticides, and animal wastes (Novotny and Chesters, 1981).
Furthermore, high volumes of stormwater runoff can also overwhelm municipal

sewer systems (US EPA, 2008b).

Since the quantity of water available for evapotranspiration is reduced due to
urbanization, a great deal of incoming solar energy that would have been used to
evaporate water is instead transformed into sensible heat, which in turn warms
the surface (Barnes et al., 2001). In addition, since the albedo of urban surfaces
is generally 10% lower than the albedo of rural surfaces (Oliver, 1973), urban
areas can have higher ambient air temperatures as compared to surrounding

suburbs or countryside, a phenomenon known as the Urban Heat Island Effect

 

(UHIE). According to the US. Environmental Protection Agency (US EPA)
(2009), urban air temperatures can be 10 to 15 °C (18 - 27 °F) warmer than the
surrounding countryside. In the summer, this translates into higher air
conditioning use. For every 0.6 °C (1 °F) increase in air temperature, peak

summer utility load may increase by 2% (US EPA, 2009).

In order to meet increased demands for energy due to the UHIE and population
growth, more than 100 new coal ﬁred plants are proposed in the US. by 2017
(US Department of Energy, 2008). Such fossil fuel use releases CD; as a by-
product of combustion. According to the Intergovernmental Panel on Climate
Change (IPCC, 2007), human activity related to the combustion of fossil fuels
has increased carbon dioxide (C02) concentrations in the atmosphere 32% since
1750. Because 002 is one of the atmospheric gases that reduce the amount of
outgoing terrestrial energy from escaping into space, increased levels of this gas
will ultimately lead to an increase in earth’s temperature. The National Academy
of Sciences reports that warming of the earth’s surface is already occurring as air
temperatures have warmed 0.6°C over the last 100 years (National Research

Council, 2001).

In 1997, the United Nations Framework Convention on Climate Change
(UNFCCC) protocol was adopted in Kyoto, Japan as an international effort to
reduce six greenhouse gases, especially C02. As of 05 February 2009, 184

countries ratiﬁed or accepted the Kyoto Protocol (UNFCCC, 2008), for an

average 5.2% reduction in 1990 greenhouse gases. While the protocol
mandates each nation to meet their own reductions by 2012, it also allows other
ﬂexible mechanisms to achieve these reductions, including emissions trading and

emission reduction/removal projects in other countries.

Green roofs as one tool to mitigate urbanization

Establishing green roofs, or vegetated roofs, is one way to mitigate some of the
many problems caused by population growth and urbanization. They are similar
to other cool roof technologies, since they have a high albedo - ranging from 0.7
to 0.85, depending on water availability (Gafﬁn et al., 2005). Other cool roof
technologies, such as white roofs, may start with an albedo of 0.8, but reﬂectivity
can decline up to 11% due to dust and debris accumulation. Conventional roof

surfaces have much lower albedos ranging from 0.05 to 0.25 (US EPA, 2005a).

Green roofs are categorized as ‘intensive’ or ‘extensive’ systems. Intensive
green roofs are designed to be similar to landscaping found at natural ground
level, and as such require substrate depths greater than 15.2 cm (6 in) and have
‘intense’ maintenance needs. By contrast, extensive green roofs use shallower
media depths (less than 15.2 cm (6 in)) and require minimal maintenance. Due
to building weight restrictions and costs, shallow substrate extensive green roofs
are much more common than deeper intensive roofs. Because of their shallower

media depth, plant species are limited to herbs, grasses, mosses, and drought

tolerant succulents such as Sedum. In addition, extensive green roots can be

built upon a sloped surface.

Green roofs in many areas of the world are typically implemented because of
their ability to reduce stormwater runoff. In a green roof system, much of the
precipitation is captured in the media or vegetation and will eventually evaporate
from the soil surface or will be released back into the atmosphere by
transpiration. Green roofs reduce runoff by 50% to 100% depending on the type
of green roof system (Carter and Jackson, 2007; DeNardo et al., 2005; Getter et
al., 2007; Hilten et al., 2008; Jarrett and Berghage, 2008; VanWoert et al.,
2005a). Water retention depends on design factors such as substrate depth,
composition, roof slope, and plant species, as well as weather factors such as

intensity, duration of rainfall, and anteoendent substrate moisture conditions.

However, in many areas of North America the primary reason for choosing a
green roof is due to the energy conservation they provide. Green roofs shade
and insulate buildings, which when combined with the evapotranspiration of the
substrate and plant material reduce summer air temperatures just above a green
roof as well as indoor temperatures underneath a green roof (Connelly and Liu,
2005; Santamouris et al., 2007; Sailor, 2008; Takebayashi and Moriyama, 2007).
Since buildings consume 39% of total energy use and 71% total electricity
consumption (US. Green Building Council, 2007), green roof implementation on

a wide scale could significantly impact the UHIE while providing economic

savings. Laberge (2003) estimated that for the Chicago city hall, energy savings
alone could result in $4,000 annually for heating and cooling combined. If all of
Chicago had green roofs, the savings could be $100,000,000 annually (Laberge,
2003)

Other beneﬁts of green roofs include extended roof lifespan, habitat for wildlife,
mitigation of noise and air pollution, and improved aesthetic value. Because
growing media and plant material protect rooﬁng membranes from solar
exposure and ultraviolet radiation that can damage the traditional bituminous roof
membrane, green roofs are estimated to last 45 years or longer in terms of
mechanical lifespan (Kosareo and Ries, 2007). This translates into a better
return on investment compared to traditional roofs (Clark et al., 2008). Since
most extensive green roofs are inaccessible to the public, they can provide
undisturbed habitat for microorganisms, insects, and birds (Baumann, 2006;
Brenneisen, 2006). When humans view green plants and nature, it has
beneﬁcial health effects (Ulrich and Simmons, 1986) as well as improved health
and worker productivity (Kaplan et al., 1988; Ulrich, 1984). Green roofs have
been also found to reduce noise pollution by 10 db (Van Renterghem and
Botteldooren, 2008) and air pollution (in the form of 03, N02, PM"), and $02) by

85 kg-rla'1-yr1 (Yang et al., 2008).

Green roofs have also been shown to achieve public health beneﬁts. Nitrogen

oxides (NOx) result from combustion of fossil fuels and can form ground level

ozone causing serious human respiratory problems, including premature death,
as well as a reduction in crop yields, all of which have economic impacts (US
EPA, 1998). Clark et al. (2008) found that green roofs yield an annual beneﬁt of
$0.45 to $1.70 per m2 ($0.04 to $0.16 per square foot) in terms of nitrogen oxide

(NO,) uptake per year.

Although green roofs are often adopted for energy savings and heat island
mitigation, rarely has this technology been promoted for its ability to mitigate
climate change, even though they do reduce carbon emissions. By lowering
electricity demand for air conditioning use, less carbon dioxide is released from
power plants. Sailor (2008) integrated green roof energy balance into Energy
Plus, a building energy simulation model supported by the US. Department of
Energy. His simulations suggested 2% reductions in electricity consumption and
9 to 11% reductions in natural gas consumption. Based on his model of a
generic building with a 2,000 m2 green roof, these annual savings ranged from
27.2 to 30.7 GJ of electricity saved and 9.5 to 38.6 GJ of natural gas saved,
depending on climate and green roof design. When considering the national
averages of CO; produced for generating electricity and burning natural gas (US
EPA, 2007; US EPA, 2008a), these ﬁgures translate to 2.3 to 2.6 kg C02 per
square meter of green roof in electricity and 0.24 to 0.97 kg C02 per square
meter of green roof in natural gas each year. Another 25% reduction in electricity

use may additionally occur due to indirect heat island reduction achieved from

large-scale green roof implementation throughout an urban area (Akbari and

Konopacki, 2005).

As an example of these potential emission savings using Sailor’s (2008) model,
the campus of Michigan State University in East Lansing, MI has 1.1 km2 of ﬂat
roof surface. If all of these roofs were greened, they could avoid 3,640,263 kg
CO2 emitted per year in electricity and natural gas consumption combined. This
is the equivalent of taking 661 vehicles off the road each year (US EPA, 2005b).
These ﬁgures depend on climate, green roof design, and source of fuel for

electricity and gas.

However, many components of a green roof have a C02 ‘cost’ in terms of the
manufacturing process. Embodied energy is a term used to describe total
energy consumed, or carbon released, of a product over its life cycle. Typical
components of a green roof include a root barrier installed on top of the normal
rooﬁng membrane (which protects the roof from root penetration damage), a
drainage layer above the root barrier (which allows excess water to ﬂow away
from the roof), and a growing substrate. Many life cycle analysis studies ignore
these unique components of a green roof by making the assumption that the root
barrier, drainage layers, and substrate will all have a carbon cost similar to the
traditional roofs’ gravel ballast (Alcazar and Bass, 2006; Kosareo and Ries,

2007). But, this assumption may not be valid.

One cradle-to-gate study (Hammond and Jones, 2008) analyzed many building
materials from the beginning (cradle) through the entire production process up
until the product leaves the factory (gate). Material consisting of low density
polyethylene (LDPE) similar to a root barrier was found to average 78.1 MJ-kg‘1
of embodied energy or 1.7 kg CO2-kg'1 of embodied carbon (Hammond and
Jones, 2008). Assuming 0.51 mm thickness and a density of 925 kg-m‘3 (Raven
Industries, Sioux Falls, SD), this translates to 0.8 kg CO2 per square meter of
green roof. Drainage layers are commonly made from polypropylene (Colbond
lnc., Enka, NC) which is found to contain 5.03 kg CO2 per kg of product
(Hammond and Jones, 2008). Assuming a weight of 0.39 kg-m'2 (Xero Flor
America LLC, Durham, NC), this translates into 1.9 kg CO2 per square meter of
green roof. The embodied energy for a substrate consisting of sand and
expanded slate is 0.005 and 0.44 kg CO2-kg", respectively (GBC, 2008;
Hammond and Jones, 2008). A 6.0 cm substrate depth consisting of half sand
and half expanded slate by volume with densities of 2,240 and 1,600 kg-m'3,
respectively (Expanded Shale, Clay and Slate Institute. 2008; Hammond and
Jones, 2008) calculates to 0.03 and 21.1 kg CO2 per square meter of green roof.
These green roof components add up to an embodied carbon content of 24.2 kg
CO2 per square meter of green roof. But, the embodied carbon of a traditional
roof’s gravel ballast is estimated at 0.017 kg CO2-kg'1 which equates to 50.0 kg
CO2-m'2 of roof, assuming a density of 1,800 kg-m'3 and a 2 cm depth
(Hammond and Jones, 2008). Subtracting out the unneeded gravel ballast, total

adjusted embodied carbon for this extensive green roof is then 23.6 kg CO2 per

square meter of green roof. Based on Sailor’s (2008) minimum energy savings
(above) of 2.54 kg CO2-m'2, it would take at most 9.3 years to ‘pay’ for the carbon

cost of the green roof materials.

In addition, green roofs may also be a tool for further reducing carbon footprints
due to the presence of plants and soils. The process of photosynthesis removes
carbon dioxide from the atmosphere and stores carbon in plant biomass,
commonly referred to as terrestrial carbon sequestration. Carbon is transferred
to the substrate via plant litter and exudates. The length of time that this carbon
remains in the soil before decomposition has yet to be quantiﬁed for green roofs.
But, immediately after initial green roof installation, net primary production should
exceed decomposition making this man-made ecosystem a carbon sink, at least
in the short-term. Once an equilibrium has been reached whereby carbon gain is
offset by respiratory losses and decomposition, this ecosystem will likely no
longer be sequestering new carbon, but will still be storing more carbon than the

original barren roof.

Due to both emission improvements and sequestration potential, green roofs
could be a new way for companies to develop carbon trading credits in a cap and
trade system. Market based carbon trading is a low-cost method for many
companies to manage and lower emissions. Chicago Climate Exchange (CCX)
is such a pilot trading program that is voluntary in the United States. This

program ”seeks to demonstrate that greenhouse gas trading can reduce

10

emissions across different business sectors” (CCX, 2007). It is similar to the
highly successful sulfur dioxide ($02) emissions trading program implemented in
the 19905 (Chestnut and Mills, 2005). Participation in CCX trading for
greenhouse gases could help companies meet their reduction targets by
purchasing credits from other companies who have exceeded their reduction
target, as well as provide ﬁnancial incentives for those companies who exceed

their reduction goals.

Importance of Plant Selection for Green Roofs

In order to capitalize on the beneﬁts of green roofs, plant selection is critical. If
the plants do not survive, albedo, stormwater retention, evaporative cooling, and
carbon sequestration may all be reduced. But selecting plants for green roofs is
often difﬁcult due to the harsh urban environment and because plants are also
more susceptible to extremes in temperature and drought due to their shallow

substrate and elevation above ground.

The genus Sedum is a popular choice among extensive green rooﬁng projects
due to its tolerance for drought (Durhman et al., 2006; Wolf and Lundholm,
2008), shallow substrate adaptability (Durhman et al., 2007; Emilsson, 2008),
persistence (Kohler, 2006; Monterusso et al., 2005; Rowe et al., 2006), and
ability to limit transpiration (Kluge, 1977; Lee and Kim, 1994) and store water
(Gravatt, 2003; Teeri et al., 1986). But even for such a well suited genus,

substrate depth can inﬂuence the rate of substrate coverage and subsequent

11

plant growth (Durhman et al., 2007; Getter and Rowe, 2008; Rowe et al., 2006).
Deeper substrates are beneﬁcial for both increased water holding capacity
(VanWoert et al, 2005a; VanWoert et al., 2005b) and as a buffer for
ovenrvintering survival as shallow substrates are more subject to ﬂuctuations in
temperature (Boivin et al., 2001). Despite the cultural limitations of shallow
substrate depths, they are often desirable because buildings must be structurally

able to support the added weight of the green roof.

Overall climate and roof microclimate are factors that will determine plant
success as well as the design intent for the green roof project. A good example
of this is the state of Florida’s ﬁrst pilot green roof installed in Naples, Florida in
2003. The plant species chosen for this project were selected because of their
green roof success in northern latitudes. But these species failed to survive in
Florida’s environment. The researchers concluded that the “type of media used

and the green roof proﬁle structure are secondary to the correct choice of plants

(Livingston et al., 2004).

Many studies for plant survival on green roofs collect data for one to two years
(Durhman et al., 2007; Emilsson and Rolf, 2005; Kircher, 2004; MacDonagh et
al., 2006; Nagase and Dunnett, 2008). Since green roofs exhibit such a long
lifespan (Kosareo and Ries, 2007), long-ten'n plant performance beyond the ﬁrst

few years’ growth is important. Plants that survive initially on a green roof may

12

not continue to exist there in the long term because of variability in climate and

other factors.

In addition, research involving plant selection thus far has typically been for roofs
exposed only to full sun (Monterusso et al., 2005; Durhman et al., 2007; Getter
and Rowe, 2008). However, as green roof implementation continues, it is likely
that many green roofs will be shaded by other structures. Incoming solar
radiation (insolation) directly and indirectly impacts plant growth. Depending on
the species, this impact may be severe or negligible. In Utah, it was observed
that high solar radiation zones differed from low solar radiation zones in terms of
species performance (Dewey et al., 2004). Another study in the Paciﬁc
Northwest looked at a sloped roof and found that in some cases the amount of

solar radiation inﬂuenced plant performance (Martin and Hinckley, 2007).

Often clients desire the use of native species because of their real and perceived
beneﬁts, such as their longevity without the use of pesticides, fertilizers, or
irrigation (US EPA, 2008c). But, plants could not actually be native to rooftops
since roofs are man-made artiﬁcial structures. However, many plant species are
adapted to these type of conditions. Because of client demand for native
species, researchers in Michigan evaluated 18 native taxa on unirrigated
extensive green roof platforms (Monterusso et al., 2005; Rowe et al., 2005).
After three years, only four of the species survived. The majority of the plants

tested were considered to be drought tolerant, but their survival in a native

13

environment relies on deep tap roots to obtain moisture. In a shallow extensive
roof, these roots can still grow sideways, but without supplemental irrigation

periods of drought resulted in death.

Research Objectives

The research presented here involves three studies on green roofs. The ﬁrst two
studies evaluate initial growth, survival, and persistence of species on extensive
green roofs. The third study focuses on how traditional green roof plants can
contribute to the terrestrial carbon sink in the artiﬁcial roof environment. The
three studies are complementary and build upon previous species research in
that the best suited species for green roof habitation should also be those that
sequester the most carbon simply because those species are more likely to

survive and persist in a roof ecosystem.

Because of the need to expand the plant palette of green roof species in local
climates, the objective of the ﬁrst study was to evaluate the effect of substrate
depth on long-term plant survival of 12 Sedum species, seven of which have
never been tested in a Midwestern climate. The expected outcome of this study
was that some species would be more suited for roof habitation than others and
that some would perform better at deeper or shallower substrate depths relative
to other species. The mechanism for this prediction is that soil depth inﬂuences

plant water availability and thus plant growth. Therefore, the research hypothesis

14

of this study was that the mean percentage of abundance of each species will

differ among substrate treatments and species.

There is also a need to evaluate the effect of solar radiation intensities on plant
community development, as well as evaluate the performance of native species.
Therefore, the objective of the second study was to evaluate long-terrn growth,
survival, and persistence of both native and typical green roof species as a
function of media depth and irradiation levels (full sun vs. full shade). The
expected outcome of this study was that some species would be more suited for
roof habitation than others and that some would perform better at deeper or
shallower substrate depths as well as different irradiation levels relative to other
species. The mechanism for this prediction was that soil depth and irradiation
levels inﬂuence plant water availability and photosynthetic rates, both of which
impact plant growth. Therefore, the research hypothesis of this study was that
the mean percentage of abundance of each species would differ between

substrate treatments, irradiance levels, and species.

Finally, there is an increasing momentum in the US. to offset carbon emissions.
Thus far, most studies on terrestrial carbon sequestration have focused on
forests and crops, but there have been no studies of green roofs. The objective
of the third study was to quantify the carbon storage potential of extensive green
roofs and to evaluate the effect that species has on carbon ﬂux. The research

hypothesis was twofold. The ﬁrst hypothesis was that species which exhibit

15

greater coverage and growth on extensive green roofs would affect the total
carbon sequestered both above-ground and below-ground due to greater plant
size and survival. The second hypothesis was that green roofs would have

greater carbon storage than a traditional roof.

16

Literature Cited

Akbari, H., and S. Konopacki. 2005. Calculating energy-saving potentials of heat
, island reduction strategies. Energy Policy 33(6):721-56.

Alcazar, SS. and B. Bass. 2006. Life cycle assessment of green roofs - case
Study of an eight-story residential building in madrid and implications for green
roof beneﬁts. In Proc. of 4"1 North American Green Roof Conference: Greening
Rooftops for Sustainable Communities, Boston, MA. 11-12 May 2006. The
Cardinal Group, Toronto.

Barnes, K., J. Morgan, and M. Roberge. 2001. lmpervious Surfaces and the
Quality of Natural Built Environments. Report prepared for Towson University’s
NASA/Raytheon/Synergy project.

Baumann, N. 2006. Ground-Nesting Birds on Green Roofs in Switzerland:
Preliminary Observations. Urban Habitats 4:37-50.

Boivin, M., M. Lamy, A. Gosselin, and B. Dansereau. 2001. Effect of artiﬁcial
substrate depth on freezing injury of six herbaceous perennials grown in a
green roof system. HortTechnology 11(3):409—412.

Brenneisen, S. 2006. Space for Urban Wildlife: Designing Green Roofs as
Habitats in Switzerland. Urban Habitats 4:27-36.

Carter, T., and CR. Jackson. 2007. Vegetated roofs for stormwater
management at multiple spatial scales. Landscape and Urban Planning 80:84-
94.

Chestnut, L.G., and OM. Mills. 2005. A Fresh Look at the Beneﬁts and Cost of
the US Acid Rain Program. Journal of Environmental Management. 77(3):252-
266.

Chicago Climate Exchange. 2007. Chicago Climate Exchange Overview. 16
December 2008. http://www.chicagoclimatex.com/index.jsf.

Clark, C., P. Adriaens, and F .B. Talbot. 2008. Green Roof Valuation: A
Probabilistic Economic Analysis of Environmental Beneﬁts. Environmental
Science and Technology 42:2155—2161.

Connelly, M. and K. Liu. 2005. Green roof research in British Columbia — an
overview. p. 416-432. In Proc. of 3rd North American Green Roof Conference:
Greening Rooftops for Sustainable Communities, Washington, DC. 4-6 May
2005. The Cardinal Group, Toronto.

17

DeNardo, J.C., A.R. Jarrett, H.B. Manbeck, D.J. Beattie, and RD. Berghage.
2005. Stormwater mitigation and surface temperature reduction by green roofs.
Transactions of ASAE 48(4): 1491 -1496

Dewey, D.W., P.G. Johnson, and R.K. Kjelgren. 2004. Species composition
change in a rooftop grass and wildﬂower meadow. Native Plants 2:56-65.

Durhman, A.K., D.B. Rowe, and CL. Rugh. 2006. Effect of watering regimen on
chlorophyll fluorescence and growth of selected green roof plant taxa.
HortScience 41:1623-1628.

Durhman, A.K., D.B. Rowe, and CL. Rugh. 2007. Effect of substrate depth on
initial coverage, and survival of 25 succulent green roof plant taxa. HortScience
42:588-595.

Emilsson, T. and K. Rolf. 2005. Comparison of establishment methods for
extensive green roofs in southern Sweden. Urban Forestry & Urban Greening
3: 1 03-1 1 1.

Emilsson, T. 2008. Vegetation development on extensive vegetated green roofs:
Inﬂuence of substrate composition, establishment method and species mix.
Ecological Engineering 33:265-277.

Expanded Shale, Clay and Slate Institute. 2008. Lightweight Expanded Shale,
Clay and Slate Aggregate for Geotechnical Applications lnforrnation Sheet 6001.
http://www.escsi.org/pdfdoc1/Geotech_lnfo_Sheet_6001 -J une_2008.pdf

Ferguson, B.K. 1998. Introduction to Stormwater: Concept, Purpose, Design.
John Wiley and Sons, Inc., New York.

Gafﬁn, S., C. Rosenzweig, L. Parshall, D. Beattie, R. Berghage, G. O’Keefe, and
D. Braman. 2005. Energy balance modeling applied applied to a comparison of
white and green roof cooling efﬁciency. p. 583-597. In Proc. of 3rd North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Washington, DC. 4-6 May 2005. The Cardinal Group, Toronto.

GBC. Unknown. The Green Building Challenge Handbook of the D-A-C-H clay
brick industry. 16 December 2008. http://www.gbc-ziegelhandbuch.orgldefault-
eng.htm.

Getter, K.L., D.B. Rowe, and J.A. Andresen. 2007. Quantifying the effect of
slope on extensive green roof stormwater retention. Ecological Engineering
31:225-231.

Getter, KL. and DB. Rowe. 2008. Media Depth Influences Sedum Green Roof
Establishment. Urban Ecosystems 11:361-372.

18

Gravatt, DA. 2003. Crassulacean acid metabolism and survival of asexual
propagules of Sedum wrightii. Photosynthetica 41 :449-452.

Hammond GP. and Cl. Jones. 2008. Embodied energy and carbon in
construction materials. Energy 161 (2):87 - 98.

Hilten, R.N., T.M. Lawrence, and E.W. Tollner. 2008. Modeling stormwater
runoff from green roofs with HYDRUS-1 D. Journal of Hydrology 358:288-293.

Intergovernmental Panel on Climate Change. 2007. Climate Change 2007: The
Physical Science Basis. Cambridge University Press. Cambridge, UK, New .
York.

Jarrett, AR. and RD. Berghage. 2008. Annual and individual green roof
stormwater response models. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto.

Kaplan, 8., JP Talbot, and R. Kaplan. 1988. Coping with daily hassles: The
impact of the nearby natural environment. Project Report. USDA Forest Service,
North Central Forest Experiment Station, Urban Forestry Unit Cooperative.
Agreement 23-85-08.

Kircher, W. 2004. Annuals and Sedum-cuttings in seed-mixtures for extensive
roof gardens. Acta Hort. 643:301-303.

Kluge, M. 1977. Is Sedum acre L. a CAM plant? Oceologia 29:77-83.

Kohler, M. 2006. Long-term vegetation research on two extensive green
roofs in Berlin. Urban Habitats 4(1):3-26.

Kosareo, L and R. Ries. 2007. Comparative environmental life cycle assessment
of green roofs. Building and Environment 42:2606—2613.

Laberge, K. M. 2003. Urban oasis: Chicago’s City Hall green roof. p. 194-203.
In Proc. of 1st North American Green Roof Conference: Greening Rooftops for
Sustainable Communities, Chicago. 29-30 May 2003. The Cardinal Group,
Toronto.

Lee, KS. and J. Kim. 1994. Changes in crassulacean acid metabolism (CAM)
of Sedum plants with special reference to soil moisture conditions. Journal of
Plant Biol. 37(1):9-15.

Livingston, E, C. Miller, and M. Lohr. 2004. Green Roof Design and
Implementation in Florida. In: Proc. of 2nd North American Green Roof

19

Conference: Greening Rooftops for Sustainable Communities, Portland, OR. 2-4
June 2004, The Cardinal Group, Toronto, pp. 353-364.

MacDonagh, L.P., N.M. Hallyn, and S. Rolph. 2006. Midwestern USA plant
communities + design =Bedrock bluff prairie greenroofs. In: Proc. of 4th North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Boston, MA. 11-12 May 2006. The Cardinal Group, Toronto.

Martin, MA. and TM. Hinckley. 2007. Native Plant Performance On A Seattle
Green Roof. In: Proc. Of 4th North American Green Roof Conference: Greening
Rooftops for Sustainable Communities, Minneapolis, MN, 29 April — 1 May 2007,
The Cardinal Group, Toronto.

Monterusso, M.A., D.B. Rowe, and CL. Rugh. 2005. Establishment and
persistence of Sedum spp. and native taxa for green roof applications.
HortScience 40(2):391-396.

Nagase, A. and N. Dunnett. 2008. Experiments in plant selection for extensive
green roofs: Performance of annual plant species and the amount of water runoff
from different vegetation types. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto.

National Research Council, Committee on the Science of Climate Change,
Division on Earth and Life Studies. 2001. Climate Change Science: An Analysis
of Some Key Questions. National Academy Press. Washington, DC.

Novotny, V. and G. Chesters. 1981. Handbook of Urban Nonpoint Pollution:
Sources and Management. New York: Van Nostrand Reinhold Company.

Oliver, J. E. 1973. Climate and Man’s Environment: An Introduction to Applied
Climatology. New York: John Wiley & Sons, Inc.

Rowe, B., M. Monterusso, and C. Rugh. 2005. Evaluation of Sedum species
and Michigan native taxa for green roof applications. p. 469-481. In Proc. of 3rd
North American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Washington, DC. 4-6 May 2005. The Cardinal Group, Toronto.

Rowe, D.B., C.L. Rugh, and AK. Durhman. 2006. Assessment of substrate
depth and composition on green roof plant performance. In Proc. of 4th North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Boston, MA. 11-12 May 2006. The Cardinal Group, Toronto.

Sailor, DJ. 2008. A green roof model for building energy simulation programs.
Energy and Buildings 40:1466-1478.

20

Santamouris, M., C. Pavlou, P. Doukas, G. Mihalakakou, A. Synnefa, A.
Hatzibiros, and P. Patargias. 2007. Investigating and analysing the energy and
environmental performance of an experimental green roof system installed in a
nursery school building in Athens, Greece. Energy 32:1781—1788.

Scholz-Barth, K. 2001. Green roofs: Stormwater management from the top
down. Environmental Design & Construction 4(1):63-70.

Takebayashi, H. and M. Moriyama. 2007. Surface heat budget on green roof
and high reﬂection roof for mitigation of urban heat island. Building and
Environment 42:2971—2979.

Teeri, J.A., M. Turner, and J. Gurevitch. 1986. The response of leaf water
potential and crassulacean acid metabolism to prolonged drought in Sedum
rubrotinctum. Plant Physiology 81 :678-680

Ulrich, RS. 1984. View through a window may inﬂuence recovery from surgery.
Science 224:420-421.

Ulrich, RS. and R. Simons. 1986. Recovery from stress during exposure to
everyday outdoor environments. In J. Wineman, R. Barnes, and C. Zimring, eds.
The Costs of Not Knowing, Proceedings of the 17th Annual Conference of the
Environmental Research Association. Washington, DC: Environmental
Research Association.

United Nations, Department of Economic and Social Affairs. 2007. Population
Division. Urban and Rural Areas 2007. 15 March 2009.
httpzllwww.un.org/esa/population/publications/wup2007/2007_urban_rural_chart.
pdf.

United Nations Framework Convention on Climate Change. 2008. Kyoto Protocol
Status of Ratiﬁcation. 02 December 2008.
http://unfccc.int/ﬁles/kyoto_protocol/status_of_ratiﬁcation/application/pdflkp_ratiﬁ
cation.pdf.

US. Department of Energy. National Energy Technology Laboratory. 2008.
Tracking New coal-ﬁred power plants. 2 December 2008.
http://www.netl.doe.gov/coal/refshelf/ncp.pdf.

US. Environmental Protection Agency. 1998. NOX: How nitrogen oxides affect
the way we live and breathe. Environmental Protection Agency Ofﬁce of Air
Quality Planning and Standards Research. EPA-456/F-98-005. Washington,
DC.

US. Environmental Protection Agency. 2005a. Cool roofs. 04 January 2006.
http://www.epa.gov/heatisland/strategies/coolroofs.html

21

US. Environmental Protection Agency. Ofﬁce of Transportation and Air Quality.
2005b. Emission Facts: Greenhouse Gas Emissions from a Typical Passenger
Vehicle. EPA420-F-05-004. Washington, DC.

US. Environmental Protection Agency. 2007. Inventory of US. Greenhouse Gas
Emissions and Sinks: Fast Facts 1990-2005. Conversion Factors to Energy Units
(Heat Equivalents) Heat Contents and Carbon Content Coefﬁcients of Various
Fuel Types. EPA-430-R-07-002. Washington, DC.

US. Environmental Protection Agency. 2008a. Climate Leaders Greenhouse
Gas Inventory Protocol Core Module Guidance: Indirect Emissions from
Purchases/Sales of Electricity and Steam. EPA -430-K-03-006. Washington, DC.

US. Environmental Protection Agency. 2008b. Combined Sewer Overﬂows:
Demographics. 30 January 2008. http://cfpub.epa.gov/npdeslcso/demo.cfm

US. Environmental Protection Agency. 2008c. Green Landscaping. 2 November,
2008. http://www.epa.gov/greenacresflndex.html#Beneﬁts.

US. Environmental Protection Agency. 2009. Reducing Urban Heat Islands:
Compendium of Strategies; Urban Heat Island Basics. 15 March 2009.
http://www.epa.gov/heatislandlresources/pdf/BasicsCompendium.pdf.

US. Green Building Council Research Committee. 2007. A National Green
Building Research Agenda. 11 December 2008.
httpzllwww.usgbc.org/ShowFile.aspx?DocumentlD=3402.

Van Renterghem, T., Botteldooren, D. 2008. Numerical evaluation of sound
propagating over green roofs. Journal of Sound and Vibration 317, 781—799.

VanWoert, N.D., D.B. Rowe, J.A. Andresen, C.L. Rugh, R.T. Fernandez, and L.
Xiao. 2005a. Green roof stormwater retention: Effects of roof surface, slope, and
media depth. J. Environ. Quality 34(3):1036-1044.

VanWoert, N.D., D.B. Rowe, J.A. Andresen, C.L. Rugh, and L. Xiao. 2005b.
Watering regime and green roof substrate design affect Sedum plant growth.
HortScience 40(3):659-664.

Wolf, D. and J.T. Lundholm. 2008. Water uptake in green roof microcosms:
Effects of plant species and water availability. Ecological Engineering 33:179-
186.

Yang, J., Yu, Q., Gong, P. 2008. Quantifying air pollution removal by green
roofs in Chicago. Atmospheric Environment 42, 7266—7273.

22

CHAPTER ONE

Getter, K.L., and DB. Rowe. 2009. Substrate Depth Inﬂuences Sedum Plant
Community on a Green Roof. HortScience. Accepted.

23

Abstract

Since the waterprooﬁng membrane beneath green roofs is estimated to last at
least 45 years, long-tenn plant performance beyond initial establishment is
critical. Plants that survive initially on a green roof may not exist in the long term
because of variability in climate and other factors. This study evaluated the
effect of green roof substrate depth on substrate moisture, plant stress as
measured by chlorophyll ﬂuorescence, and plant community development and
survival of 12 Sedum species over four years in a Midwestern U.S. climate during
four years of growth. Plugs of 12 species of Sedum were planted on 8 June
2005 and evaluated bi-weekly for absolute cover (AC). Most species exhibited
greater growth and coverage at a substrate depth of 7.0 cm and 10.0 cm relative
to 4.0 cm. For the species evaluated, substrate depths of at least 7.0 cm are
highly recommended. AC of Sedum was signiﬁcantly greater at this substrate
depth than at 4.0 cm. Mean volumetric moisture content of the three substrate
depths followed the same pattern as AC. When averaged over time, the 4.0 cm
substrate depth held less moisture than depths of 7.0 or 10.0 cm, while the 7.0
and 10.0 cm substrate depths were statistically the same. Species exhibiting the
greatest AC at all substrate depths were S. ﬂon'ferum, S. sexangulare, S.
spun'um ‘John Creech’, and S. stefco. In general, species that are less suitable at
these substrate depths are 8. ‘Angelina’, S. cauticola ‘Lidakense’, S. ewersii, S.

ochroleucum, and S. reﬂexum “Blue Spruce’.

24

Introduction

In 2007, 223,666 m2 (2,407,525 ftz) of green roofs, or vegetated roofs, were
installed in North America representing a 30% increase over 2006 (Green Roofs
for Healthy Cities, 2008). By placing plants on rooftops, the vegetated footprint
that was previously destroyed during building construction is at least partially
replaced. Increased adoption of this rooﬁng technology may be due to the many
beneﬁts they provide, such as improved stormwater management (Carter and
Jackson, 2007; Getter et al., 2007; Hilten et al., 2008; Jarrett and Berghage,
2008), energy conservation (Sailor, 2008; Santamouris et al., 2007), mitigation of
the urban heat island effect (T akebayashi and Moriyama, 2007), increased
longevity of rooﬁng membranes (Kosareo and Ries, 2007), a better return on
investment than traditional roofs (Clark et al., 2008), reduced noise and air
pollution (Van Renterghem and Botteldooren, 2008; Yang et al., 2008), increased
urban biodiversity (Baumann, 2006; Brenneisen, 2006), as well as providing a
more aesthetically pleasing environment to experience (Getter and Rowe, 2006;

Obemdorfer et al., 2007).

In order for green roofs to be successful, as well as to meet client expectations,
plant selection is critical. Species selected must survive extremes in roof
microclimate. Green roofs are likely to experience drought and severe
ﬂuctuations in root zone temperatures due to shallow substrates, as well as high

temperatures, and windy conditions. These conditions combined with heat

25

radiating from the building will likely alter hardiness zones and soil moisture

content, thus impacting which species are able to survive.

Successful candidate species for extensive green roofs (i.e., green roofs with
substrate depths of 10.0 cm (3.93 in) or less) must exhibit characteristics such as
easy propagation, rapid establishment, and high groundcover density (Dunnett

and Kingsbury, 2004; Getter and Rowe, 2006; Snodgrass and Snodgrass, 2006).

 

Low growing plants that spread and cover the substrate in a short period of time
reduce potential erosion problems, inhibit weeds, and provide improved
aesthetics. Although rapid coverage is important, the ability of plant species to
be self-sustaining reduces the need for future replanting and maintenance.
Species that are long-lived, that reseed themselves, or spread vegetatively
should continue to provide ample coverage (60% or greater, as deﬁned by F LL

guidelines (F LL, 1995)) as long as environmental conditions are favorable.

The genus Sedum is a popular choice among extensive green rooﬁng projects
due to its tolerance for drought (Durhman et al., 2006; Wolf and Lundholm,
2008), shallow substrate adaptability (Durhman et al., 2007; Emilsson, 2008),
persistence (Kohler, 2006; Monterusso et al., 2005; Rowe et al., 2006), and
ability to limit transpiration (Kluge, 1977; Lee and Kim, 1994) and store water
(Gravatt, 2003; Teeri et al., 1986). But even for such a well suited genus,
substrate depth can inﬂuence the rate of substrate coverage and subsequent

plant growth (Durhman et al., 2007; Getter and Rowe, 2008; Rowe et al., 2006).

26

Deeper substrates are beneﬁcial for both increased water holding capacity
(VanWoert et al, 2005a; VanWoert et al., 2005b) and as a buffer in ﬂuctuating
winter temperatures (Boivin et al., 2001 ). Despite the cultural limitations of

shallow substrate depths, they are often desirable because of lighter roof loads.

Plant stress due to shallow substrate and other conditions on a roof can be
recorded by measuring chlorophyll ﬂuorescence. This technique is used to
quantify the efﬁciency of the photosynthetic apparatus (Maxwell and Johnson,
2000). Photon energy absorbed by a chlorophyll molecule can be used to fuel
photosynthesis, dissipated as heat, or re-emitted as ﬂuorescence. Measurement
of the latter is used to indicate how efﬁcient the former two processes are
proceeding. Fluorimeters are used to measure this value, usually reporting the
ratio (Fv/Fm) of variable ﬂuorescence (Fv) to maximum ﬂuorescence (Fm) that
typically ranges from 0.70 to 0.83, with values less than 0.60 indicating

photosynthetic stress (Ritchie, 2006).

Many studies for plant survival on green roofs collect data for one to two years
(Durhman et al., 2007; Emilsson and Rolf, 2005; Kircher, 2004; MacDonagh et
al., 2006; Nagase and Dunnett, 2008). Since the waterprooﬁng membrane
beneath green roofs is estimated to last 45 years or longer (Kosareo and Ries,
2007), long-term plant performance beyond the ﬁrst few years’ growth is
important. Plants that survive initially on a green roof may not exist there in the

long term because of variability in climate and other factors. Therefore, the

27

 

objective of this study was to evaluate the effect of substrate depth on substrate
moisture, plant stress as measured by chlorophyll ﬂuorescence, and plant
community development and survival of 12 Sedum species over a period of four

years.

Materials and Methods

Green Roof Platforms. Three roof platforms with dimensions of 2.44 m x 2.44 m
(8.0 ft x 8.0 ft) were utilized at the Michigan State University Horticulture
Teaching and Research Center (East Lansing, MI). Each platform was situated
at ground level and replicated a commercial extensive green roof, including
insulation, protective and waterprooﬁng membrane layers. Construction details

are outlined in VanWoert et al. (2005a).

The wood-framed platforms included sides that extend 20.3 cm (8.0 in) above the
platform deck. Each platform was divided into three equal sections measuring
0.77 m x 2.40 m (2.53 ft x 7.87 ft) using wood dividers. The platform sides and
dividers were also covered with waterprooﬁng membrane. Each platform was set
at a 2% slope and was placed with the low end of the slope facing south to

maximize sun exposure.

Drainage System and Vegetation Carrier. Each platform was constructed with a

Xero Flor XF 108 drainage mat (Wolfgang Behrens Systementwicklung, GmbH,

GroB lppener, Germany) installed over the waterprooﬁng system which allows

28

excess water to ﬂow off the roof. For additional water holding capacity, a 0.75
cm (0.26 in) thick moisture retention fabric (Xero Flor XF159) capable of retaining
5.92 kg-m'2 of water was placed over the drainage layer followed by the
vegetation carrier (Xero Flor XF301). Growing substrate was placed on the
vegetation carrier at three different depths (4.0 cm, 7.0 cm, or 10.0 cm (1.6, 2.8,
and 3.9 in)). Initially, the substrate consisted of 86% sand, 10% silt, and 4% clay
and had a bulk density of 1.37 g-cm'3 and a water holding capacity at 0.01 MPa
of 16.05%. Further details of the initial physical and chemical properties of the

substrate are detailed in Getter and Rowe (2008).

» Plant Species. The twelve species tested included Sedum ‘Angelina’ (crooked
stonecrop), Sedum cauticola 'Lidakense' (stonecrop), Sedum ewersii (stonecrop),
Sedum ﬂoriferum (kamtschatka stonecrop), Sedum hispanicum (Spanish
stonecrop), Sedum ochroleucum (European stonecrop), Sedum reﬂexum 'Blue
Spruce' (crooked stonecrop), Sedum sannentosum (stringy stonecrop), Sedum
sedifonne (pale stonecrop), Sedum sexangulare (tasteless stonecrop), Sedum
spun’um 'John Creech' (creeping sedum), and Sedum stefco (stonecrop). Plants
were obtained from Emory Knoll Farms (Street, MD) as plugs (120 cm3; 72/ﬂat)
that were established in a standard propagation mix of peat, perlite, and
vermiculite (Super-Fine Germination Media, Farfard, Inc., Agawam, MA). Plugs
were planted on 8 June, 2005 with four plants in twelve rows. Each species was
planted four times randomly in each section resulting in plugs spaced 17.0 cm

(6.7 in) apart from each other and platform walls. All plots were fertilized

29

(Nutricote controlled release fertilizer 18-6-8 type 120 (Agrivert, Webster, Tex.) at
100.0 g-m'z) and watered to ﬁeld capacity by hand on the day of planting. No

further irrigation was provided.

This planting arrangement resulted in a split plot design that was arranged in
randomized complete blocks with two factors replicated three times. The main
plot was substrate depth (4.0, 7.0, and 10.0 cm (1.6, 2.8, and 3.9 in)) and the
sub-plot factor was plant species which had 12 treatments, each replicated four

times within a sub-plot for a total of 48 plants per substrate depth per plot.

Data Collection and Analysis. A transect (a stainless steel point-frame) was used
every two weeks during the ﬁrst three growing seasons and monthly in the fourth
growing season to measure community composition and change (Waite, 2000).
The point-frame had internal measurements of 0.77 m x 1.2 m (2.5 ft x 3.9 ft) and
had eight strings (50 pound Berkley Gorilla Super Braid Fishing Line) vertically
and eight strings horizontally to create 64 measurement points. The point-frame
sat directly on top of the platform sides, secured by ﬁnishing nails that allowed
the frame to be positioned in the exact same spot every time. A stainless steel
skewer was placed vertically at each measuring point and each species the

skewer contacted was recorded, up to three canopy layers.

Substrate moisture and chlorophyll ﬂuorescence data were collected throughout

the last two growing seasons. Substrate moisture measurements were

30

monitored by inserting a theta probe (ML2x, Delta-T Devices, Ltd., Cambridge,
United Kingdom) with 6.0 cm (2.4 in) rods into the media until they were
completely buried. Measurements were collected at random times each week in
triplicate in each subplot. Chlorophyll ﬂuorescence measurements were
collected at six different times (21 May 2007, 18 June 2007, 2 August 2007, 17
September 2007, 29 May 2008, and 20 August 2008) using a Hansatech plant
efﬁciency analyzer (PEA; Hansatech Instruments, Ltd., Norfolk, England). These
measurement dates were selected in order to cover a wide range of
environmental conditions including active growth and drought stress during the
growing season. Leaves from each plant were dark adapted for 20 minutes prior
to measurement. Maximum quantum efﬁciency of photosystem II was recorded
(Fv/Fm). Three single leaf blades of each surviving species were randomly
selected in each subplot and excised from the plant to be dark adapted and
measured. This was necessary because the PEA clips were not secure on the

leaf of most species while still attached to the whole plant.

Data Analysis. Absolute cover (AC) was calculated for each species at each
substrate depth as the total number of contacts recorded divided by the number
of data collection points. Data was then analyzed as mean AC using repeated
measures. Although original means are presented, all AC values were
transformed prior to analysis using a log transformation to stabilize the variance
and normalize the data set (Underwood, 1998). Signiﬁcant differences between

treatments were determined using multiple comparisons (PROC MIXED, SAS

31

version 8.02, SAS Institute, Cary, NC). In addition, mean substrate volumetric
moisture content and mean chlorophyll ﬂuorescence data were analyzed by
PROC MIXED, least signiﬁcant differences (SAS version 8.02, SAS Institute,

Cary, NC)

Results and Discussion

Four points in time (weeks 15, 67, 120, and 172) were chosen for comparing
growth that represent the end of each growing season (before ﬁrst frost). At a
substrate depth of 4.0 cm four species exhibited no signiﬁcant growth between
the end of the ﬁrst and fourth growing seasons (S. ‘Angelina’, S. cauticola
‘Lidakense’, S. ewersii, and S. ochroleucum), while at 7.0 cm ﬁve species (S.
ewersii, S. ochroleucum, S. reﬂexum ‘Blue Spruce ’, and S. sedifonne) ﬁt this
category (Table 1.1 rows, Figure 1.1). At the 10.0 cm substrate depth there were
six such species (S. ‘Angelina’, S. cauticola ‘Lidakense’, S. ewersii, S.
ochroleucum, S. reﬂexum ‘Blue Spruce’, and S. sedifonne). As a result, at their
respective substrate depths, all of these species have zero or near zero AC by

week 172.

At each substrate depth there were also species that decreased in AC across the
four growing seasons, also resulting in zero AC by week 172 (Table 1.1 rows;
Figure 1.1). At substrate depths of 4.0 cm (S. sannentosum and S. sedifonne)

and 10.0 cm (S. hispanicum and S. sannentosum) there were two such species.

32

The 7.0 cm substrate depth had three species (8. cauticola ‘Lidakense’, S.

hispanicum, and S. sannentosum).

The remaining species at each substrate depth increased in AC across the four
growing seasons (Table 1.1; Figure 1.1). All substrate depths had the same four
species (8. ﬂon'ferum, S. sexangulare, S. spun'um ‘John Creech’, and S. stefco)
with the 4.0 cm substrate depth having an additional two species (8. hispanicum
and S. reﬂexum ‘Blue Spruce’). For all species, the increase in AC was only
signiﬁcant between the ﬁrst and second growing season. This perhaps indicates
that by the end of the second growing season, the plant community had reached

a mature or stable state.

Results for the ﬁrst growing season, as previously published, demonstrated that
by the end of the ﬁrst growing season (week 19) at all three substrate depths S.
sannentosum exhibited a much higher AC than all other species (Getter and
Rowe, 2008; Figure 1.1). However, subsequent growing seasons produced very
different results, highlighting the importance of long term studies. For the 4.0 cm
substrate depth, this species remained at near zero AC for the remaining growing
seasons. At the 7.0 cm and 10.0 cm substrate depths though, this species
represents more than haIf of the coverage by the end of the second growing
season, has a slow recovery in the third growing season to represent nearly 20%
of coverage, but then falls to near zero AC at the end of the fourth growing

season.

33

It is possible that the reason for S. sannentosum’s eventual failure at all three
substrate depths is due to incorrect hardiness zone classiﬁcation for this species.
Minimum air temperatures were lower during the second winter (-22.0°C) and
had a longer duration of extreme cold (33 days less than -10°C) as compared to
the ﬁrst winter (-19.9°C; 20 days less than -10°C) in which this species
successfully recovered (Figure 1.2). East Lansing, Mich. is classiﬁed as zone 5
on the USDA plant hardiness map, a value corresponding to an average
minimum temperature between -26 °C to -29 °C (-20 °F to -10 °F) (Cathey, 1990).
Plants that are classiﬁed with hardiness zones less than or equal to a
geographical area should in theory survive in that climate (assuming all other
plant requirements are met). Sedum sannentosum is categorized as a zone 5
species (Snodgrass and Snodgrass, 2006), but typically species are assigned a
hardiness zone based on observation only or based on how related species have
performed in the past. Hardiness zones are also meant for plants growing at
ground level. While initial work with this species indicated that it was able to
overwinter successfully when given enough establishment time before ﬁrst frost
(Getter and Rowe, 2007), this current study shows that S. sannentosum is likely
unsuitable for this climate. In climates with warmer winters, it is possible that this
species may cover a majority of the roof, as suggested by ﬁrst year data here.
However, this situation poses a threat in that an unusual cold spell may

completely wipe out S. sannentosum resulting in a many bare spots on the roof.

34

Differences in growth as measured by AC across the sampled substrate depths
and time may be partially explained by the habit of the species themselves.
Some species, such as S. ‘Angelina’, S. cauticola ‘Lidakense’, S. ochroleucom,
S. reﬂexum ‘Blue Spruce’, and S. sedifonne are more erect in nature and
therefore may be underrepresented in point-frame sampling techniques (Wilson,
1960). Furthermore, some species vegetatively reproduce easier than others,
allowing them to cover more area than simple growth would allow. Within this
genus, Stephenson (2002) reports that increasing degrees of succulence are
correlated with improved ability of a species to reproduce asexually (such as S.
stefco, S. sexangulare, and to some extent 8. sannentosum). Some species
also seem to allocate more energy to sexual reproduction (such as S.
hispanicum) while others allocate more energy to creeping growth (such as S.

sannentosum, which has sterile ﬂowers).

At all substrate depths, by the end of the fourth growing season (week 172) the
same four species (S. ﬂon'ferum, S. sexangulare, S. spurium ‘John Creech’, and
S. stefco) consistently exhibited the greatest AC (Table 1.2). However, relative
abundance differed between substrate depths. For example, 8. stefco was most
abundant at the 4.0 cm substrate depth, followed by S. ﬂon'femm, S. spurium
‘John Creech’, and S. sexangulare. By contrast, at 7.0 cm and 10.0 cm, 8.
ﬂon'ferum was most abundant, followed by S. spurium ‘John Creech’, S.

sexangulare, and S. stefco.

35

These same four species are also the only four species that were inﬂuenced by
substrate depth at the end of four growing seasons (week 172; Table 1.2).
Sedum ﬂon'ferum is the only species exhibiting signiﬁcant differences in AC at all
three substrate depths. The remaining three species primarily see the difference

between 4.0 cm and 7.0 cm, but not between 7.0 cm and 10.0 cm.

When AC is averaged over species, all depths demonstrated a statistically
signiﬁcant increase in AC across time (Table 1.3). AC for the 4.0 cm substrate
depth increased from 0.1406 at week 15 to 1.1953 at week 172, while AC for the
10.0 cm substrate depth increased from 0.4036 to 1.3411. With the exception of
week 15, the other times all showed signiﬁcant substrate depth effects within
sampled times, but only between 4.0 cm and 7.0 cm, but not between 7.0 cm and
10.0 cm. This indicates that for the surviving and most abundant species,
substrate depths greater than 7.0 cm gains no beneﬁt in terms of abundance as
measured with a point-frame. However, at deeper substrate depths, these plants
would likely be healthier, contain greater biomass, and be less susceptible to

adverse environmental conditions.

Mean volumetric moisture content of the three substrate depths follows the same
pattern as AC (Figure 1.3). While extremely variable, when averaged over time
the 4.0 cm substrate depth held less moisture then the 7.0 or 10.0 cm depths.
The 7.0 and 10.0 cm substrate depths were statistically the same. The

observation that deeper extensive green roof substrates consistently had higher

36

 

moisture content than shallower substrates is consistent with similar studies I
(Liesecke, 1998; VanWoert et al., 2005a; VanWoert et al., 2005b). In addition, it
appears that initially after a rain event the 4.0 cm substrate depth dries out faster

than either the 7.0 or the 10.0 depths (Figure 1.4).

This greater water availability at the deeper substrate depths may explain why
the 7.0 cm and 10.cm substrate depths had higher AC than the 4.0 cm depth.
Another factor may be due to substrate temperature differences. Boivin et al.
(2001) found that shallower extensive green roof substrates experienced much
more severe temperature ﬂuctuations than deeper substrates. During the
growing season, shallower substrates will likely experience higher soil
temperatures, which in turn will inﬂuence plant growth (Bouma et al. 1997;
Prasad et al. 2000). This is exacerbated by the fact that lower AC of the 4.0 cm
substrate depth exposes more substrate to direct sun resulting in higher
substrate temperatures. In addition, some species are more suited for these
temperature or water ﬂuctuations than others. Durhman et al. (2007) found that
S. album was the only species of 25 that exceeded 1.5 cm2 of growth per day at
a substrate depth of 2.5 cm (1.0 in). At a substrate depth of 5.0 cm (2.0 in) and

7.5 cm (3.0 in), this increased to 3 species and 8 species respectively.
Chlorophyll ﬂuorescence data did not follow the same pattern as AC and

substrate moisture content. Mean FVIFm values were not signiﬁcantly different at

0.795, 0.779, and 0.781 for substrate depths of 4.0 cm, 7.0 cm, and 10 cm,

37

respectively. There were also very few differences between species within the
same depth (data not shown). This may be due to the fact that measurements
for chlorophyll ﬂuorescence occurred during the third and fourth growing season
only, whereby three (Sedum cauticola ‘Lidakense', Sedum ochroleucum, and
Sedum sedifonne) of the twelve initial species had zero AC (i.e., no plants to take
measurements upon). Had chlorophyll ﬂuorescence data been taken during the
ﬁrst growing season, perhaps noticeable differences would have been detected
between species that ultimately survived the four year study and those that did
not. In addition, at individual measuring times which represent the driest portions
of the growing season (18 June 2007, 2 August 2007, and 20 August 2008),
mean chlorophyll ﬂuorescence values of individual species never fell below 0.60,
indicating very little, if any, stress to the photosynthetic system (Ritchie, 2006).
Other research has established the same trend for many species in this genus.
In a controlled greenhouse watering study, Durhman et al. (2006) found that
three Sedum species maintained active photosynthetic capacity for at least 88
days without water. This may at least partially explain why Sedum species are

such good candidates for extensive green roofs.

Conclusions

These results show the importance of substrate depth on plant performance, as
well as long-term evaluation of species. Of the substrate depths and species
evaluated in this paper, substrate depths of at least 7.0 cm are highly

recommended. AC was signiﬁcantly greater at this substrate depth relative to the

38

shallower depth of 4.0 cm. Species exhibiting the greatest AC at all substrate
depths were 8. ﬂon'ferum, S. sexangulare, S. spurium ‘John Creech’, and S.
stefco. In general, species that are less suitable are 8. ‘Angelina’, S. cauticola

‘Lidakense’, S. ewersii, S. ochroleucum, and S. reﬂexum ‘Blue Spruce’.

39

 

Literature Cited

Baumann, N. 2006. Ground-Nesting Birds on Green Roofs in Switzerland:
Preliminary Observations. Urban Habitats 4(1):37-50.

Boivin, M., M. Lamy, A. Gosselin, and B. Dansereau. 2001. Effect of artiﬁcial
substrate depth on freezing injury of six herbaceous perennials grown in a green
roof system. HortTechnology 11:409-412.

Bouma, T.J., K.L. Nielsen, D.M. Eissenstat, and JP. Lynch. 1997. Estimating
respiration of roots in soil: Interactions with soil CO2, soil temperature and soil
water content. Plant and Soil 195(2):221-232.

Brenneisen, S. 2006. Space for Urban Wildlife: Designing Green Roofs
as Habitats in Switzerland. Urban Habitats 4(1):27-36.

Cathey, HM. 1990. USDA plant hardiness zone map. USDA Agric. Res. Serv.
Misc. Publ. 1475.

Carter, T. and CR. Jackson. 2007. Vegetated roofs for stormwater
management at multiple spatial scales. Landscape and Urban Planning 80:84-
94.

Clark, C., P. Adriaens, and PB. Talbot. 2008. Green Roof Valuation: A
Probabilistic Economic Analysis of Environmental Beneﬁts. Environmental
Science and Technology 42: 2155—2161.

Dunnett, N. and N. Kingsbury. 2004. Planting Green Roofs and Living Walls.
Timber Press, Inc., Portland, OR.

Durhman, A.K., D.B. Rowe, and CL. Rugh. 2006. Effect of watering regimen on
chlorophyll ﬂuorescence and growth of selected green roof plant taxa.
HortScience 41 (7): 1623-1628.

Durhman, A.K., D.B. Rowe, and CL. Rugh. 2007. Effect of substrate depth on
initial coverage, and survival of 25 succulent green roof plant taxa. HortScience
42(3):588-595.

Emilsson, T. and K. Rolf. 2005. Comparison of establishment methods for
extensive green roofs in southern Sweden. Urban Forestry & Urban Greening 3:
103—1 1 1.

Emilsson, T. 2008. Vegetation development on extensive vegetated green roofs:

Inﬂuence of substrate composition, establishment method and species mix.
Ecological Engineering 33:265—277.

'40

 

FLL (Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau). 1995.
Guidelines for the Planning, Execution and Upkeep of Green-Roof Sites.
Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau. Bonn,
Germany.

Getter, KL. and DB. Rowe. 2006. The role of extensive green roofs in
sustainable development. HortScience 41(5):1276-1285.

Getter, K.L., D.B. Rowe, and J.A. Andresen. 2007. Quantifying the effect of
slope on extensive green roof stormwater retention. Ecological Engineering
31:225-231.

Getter, KL. and DB. Rowe. 2008. Media depth inﬂuences Sedum green roof
establishment. Urban Ecosystems 11:361-372.

Gravatt, DA. 2003. Crassulacean acid metabolism and survival of asexual
propagules of Sedum wrightii. Photosynthetica 41:449-452.

Green Roofs for Healthy Cities. 2008. Final Report Green Roof Industry Survey
2007. Green Roofs for Healthy Cities. Toronto, Ontario.

Hilten, R.N., T.M. Lawrence, E.W. Tollner. 2008. Modeling stormwater runoff
from green roofs with HYDRUS-1 D. Journal of Hydrology 358:288-293.

Jarrett, AR. and RD. Berghage. 2008.Annual and individual green roof
stormwater response models. In: Proc. of 6"1 North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto

Kircher, W. 2004. Annuals and Sedum-cuttings in seed-mixtures for extensive
roof gardens. Acta Hort. 643:301-303.

Kluge, M. 1977. Is Sedum acre L. a CAM plant? Oceologia 29:77-83.

Kbhler, M. 2006. Long-term vegetation research on two extensive green
roofs in Berlin. Urban Habitats 4(1):3-26.

Kosareo, L. and R. Ries. 2007. Comparative environmental life cycle assessment
of green roofs. Building and Environment 42:2606—2613.

Lee. KS. and J. Kim. 1994. Changes in crassulacean acid metabolism (CAM) of
Sedum plants with special reference to soil moisture conditions. Journal of Plant
Biol 37(1):9-15.

Liesecke, H.J. 1998. Das retentionsvermbgen von dachbegrtinungen. (Water
retention capacity of vegetated roofs). Stadt und Griin 47(1):46-53.

41

 

MacDonagh, L.P., N.M. Hallyn, and S. Rolph. 2006. Midwestern usa giant
communities + design =Bedrock bluff prairie greenroofs. In: Proc. of 4 North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Boston, MA. 11-12 May 2006. The Cardinal Group, Toronto

Maxwell, K. and 6N Johnson. 2000. Chlorophyll ﬂuorescence - a practical
guide. Journal of Experimental Botany. 51(345):659-668.

Monterusso, M.A., D.B. Rowe, and CL. Rugh. 2005. Establishment and
persistence of Sedum spp. and native taxa for green roof applications.
HortScience 40(2):391-396.

Nagase, A. and N. Dunnett. 2008. Experiments in plant selection for extensive
green roofs: Performance of annual plant species and the amount of water runoff
from different vegetation types. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto

Obemdorfer, E., J. Lundholm, B. Bass, R.R. Coffman, H. Doshi, N. Dunnett,

S. Gafﬁn, M. Kohler, K.K.Y. Liu, and DB. Rowe. 2007. Green Roofs as Urban
Ecosystems: Ecological Structures, Functions, and Services. BioScience
57(10):823-833.

Prasad, P.V., P.W. Craufurd, and R.J. Summerﬁeld. 2000. Effect of high air and
soil temperature on dry matter production, pod yield and yield components of
groundnut. Plant and Soil 222:231-239.

Ritchie, G. 2006. Chlorophyll ﬂuorescence: what is it and what do the numbers
mean? In: USDA Forest Service Proceedings RMRS-P-43. Fort Collins, CO: US.
Department of Agriculture, Forest Service, Rocky Mountain Research Station. p.
34-43.

Rowe, D.B., C.L. Rugh, and AK. Durhman. 2006. Assessment of substrate
depth and composition on green roof plant performance. In Proc. of 4“1 North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Boston, MA. 11-12 May 2006. The Cardinal Group, Toronto.

Sailor, DJ 2008. A green roof model for building energy simulation programs.
Energy and Buildings 40:1466-1478.

Santamouris, M., C. Pavlou, P. Doukas, G. Mihalakakou, A. Synnefa, A.
Hatzibiros, and P. Patargias. 2007. Investigating and analysing the energy and
environmental performance of an experimental green roof system installed

in a nursery school building in Athens, Greece. Energy 32:1781—1788.

42

 

Snodgrass, E. and L. Snodgrass. 2006. Green Roof Plants: A Resource and
Planting Guide. Timber Press, Inc., Portland, OR.

Stephenson, R. 2002. Sedum: cultivated stonecrop. Timber Press, Inc.,
Portland, OR.

Takebayashi, H. and M. Moriyama. 2007. Surface heat budget on green roof
and high reﬂection roof for mitigation of urban heat island. Building and
Environment 42:2971-2979.

Teeri, J.A., M. Turner, and J. Gurevitch. 1986. The response of leaf water
potential and crassulacean acid metabolism to prolonged drought in Sedum
rubrotinctum. Plant Physiology 81:678-680.

Underwood, A.J. 1998. Experiments in Ecology: Their Logical Design and
Interpretation using Analysis of Variance. University Press, Cambridge.

Van Renterghem, T. and D. Botteldooren. 2008. Numerical evaluation of sound
propagating over green roofs. Journal of Sound and Vibration 317:781—799.

VanWoert, ND. D.B. Rowe, J.A. Andresen, C.L. Rugh, R.T Fernandez, and L.
Xiao. 2005a. Green roof stormwater retention: Effects of roof surface, slope, and
media depth. J Environ Quality 34(3):1036-1044.

VanWoert, N.D., D.B. Rowe, J.A. Andresen, C.L. Rugh, and L. Xiao. 2005b.
Watering regime and green roof substrate design affect Sedum plant growth.
HortScience 40(3):659-664.

Waite, S. 2000. Statistical Ecology in Practice: A guide to analyzing
Environmental and Ecological Field Data. Prentice Hall, Harlow, England.

Wilson, W.J. 1960. Inclined point quadrats. New Phytologist 59(1):1-8.

Wolf, D. and J.T. Lundholm. 2008. Water uptake in green roof microcosms:
Effects of plant species and water availability. Ecological Engineering 33:179-
186.

Yang, J., Q. Yu, and P. Gong. 2008. Quantifying air pollution removal by green
roofs in Chicago. Atmospheric Environment 42:7266—7273.

43

 

p-

202000000: 0:E:_oo 0:0 026: :00 .3th 0:0 030.0
H 00; 5:0 0:00:05 .Amwucv 020000 0:20 000:0:0E0 08:00 05:28 :_ 0002 0000000: gurus 006000 0:0 0:300 00000000 00305
55:5 08: :0>o 060000800 08:00 026: :_ 0000. 00000026.. 30de: am... E 000000 :000 :2 0:E:_oo 0:0 026: :_ :00000000 :005.

 

 

0 0 3000 H 0000.0 0 0 00000 H 0000.0 0 0 080 H 0000.0 0 0 00000 H 0200 000000 .0
0 0 00000 H 0.000 0 0 00000 H 22.0 00 0 03.00 H N: :0 00 0 00 0.0 H 0000 .5005 ::o_.. E00000 .0
0 0 0500 H 0000.0 0 0 2000 H 300.0 00 0 00000 H 09 :0 00 . 0 0H 00.0 H 0000.0 0030:0000 .m
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 00< 0 00000 H 0000.0 0:50.000 .0.
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 m< 0 0H 3.0 H 0900 0 0 00000 H 00.30 E0000:0::00 .0
m< 0 00000 H 30.0 0 0 0000 H 0000.0 00 0 0000 H 00000 < 0 0000.0 H 0000.0 00200 020. E0000: .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 2 0 0000 H 000.0 9.. 00 00000 H 0080 E88850 .0.
m 0 00000 H 5000 0 0 0000 H 38.0 00 0 00000 H 83.0 < 0 0000.0 H 0000.0 E00.E0Q0.E .m
0 0 00000 H 0800 0 0 0080 H 03:0 00 0 0200 H 000 00 0 00000 H 0200 E30000: .m
m< 0 0500 H 020.0 9.. 0 0300 H 020.0 9. 0 0080 H 0000 < 0 0000.0 H 0000.0 0000.0 .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 00:00.00? 060.0000 .0
0< 0 00000 H 0000.0 9.. 00 0H 00.0 H 0000 00 0 00000 H «20.0 00< 0 00000 H 0000.0 05.0094. .0.
«t 0.025 02 0.00; 5 0.00; 0. 0.025 020000

 

A80 0.35009 .060 033004.

 

00:60 5:00:00 000 .6 00:5: 05 >0 00220

0000050 0:003:50 .6 .0083: .002 05 00 5000 00050000 :000 00 000000 :000 Low 0000.30.00 003 :0>o0 0020000.
.moow :00E0000m 3an x003 umeow :00E0Ha0w 0N u ONF x002, umeow 00:00:00.0 0.. "00 x003 umeow. :00E0000m

mm u m? 0.003 000:; 00.003 :_ 00:80: .050 .308 9 wood 0:00000 05305 Soc. 00 0:0 05 00 E0 0.00 0:0

.oN .o.$ 0:300 000.0000 00.5 00 000020.30 000000 E000m NH .6 0:0:0300 0:00:00 H 050 056000 :005. .2 0.00...

44

 

202000000: 0:80.00 0:0 0:50: :0: 0000.00" 0:0 0000.0
H 002, :90 0:00:05 .Auwucv 000000 0:080 000:0:0E0 08:00 05:28 :_ 0:000. 0000:0002 .ANFHS 00:0000 0:0 05000 00000000 .0:0_>_0:_
5:02, 080 :0>0 05000950 00:00 030: :_ 0:000. 0000:0300 $00000 ow.‘ :5 000000 :000 :00 0:50.00 0:0 026: :_ :o_0.0:0000 c005.

 

 

0 0 00000 H 0000.0 0 0 00000 H 8000 um. 0 0000 H 6&0 00 0 H0000 H 00000 00:00 .0 M.
0 0 00000 H «08.0 0 0 00000 H 88.0 :0 0 3000 H 00.30 00 0 0H 00.0 H 0000.0 .5005 :00? E00000 .0
0 0 00000 H 0000.0 0 0 00000 H 0000.0 ”.0 0 88.0 H 0.50 00 0 00000 H 0080 0030:0000 .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 0 0 00000 H 00000 < 00 00000 H 0080 02000000 .0.
0 0 0080 H 0300 0 0 00000 H 0800 0 0 0020 H 0000.0 0 0 00000 H 30:0 E0000:0HE00 .0.
< 0 0000.0 H 0000.0 0 0 0000.0 H 0000.0 00 0 0200 H 0080 < 0 0080 H 0800 .00200 00.0. E9000: .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 000 0 00000 H 0800 < 0 00000 H 0000.0 58:00:50 .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 00 0 0800 H 0200 00 0 0200 H 0200 E00083 .0.
0 0 00000 H 0000.0 0 0 00000 H 00:00 :0 0 0200 H 000:0 00 0 00000 H 0300 530:5: .0
0 0 00 00.0 H 0900 m 00 00000 H 0000.0 00 0 8000 H 0000.0 00 00 0200 H 00.00 00000 .0
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 9.. 0 00000 H 0000.0 0009.003. 0:000:00 .0
0 0 00 00.0 H 0000.0 0 0 30.0 H 0000.0 00 0 00000 H 0000.0 m< 0 0000 H 0800 05.0050 .0
0: 0.00:5 000 0.025 00 0.00:5 00 0.02s 020000

 

as 0.050000 :38 30.0004:

0000008 E 0.000

 

202000000: 0:80.00 0:0 0:50: :00 03 3.0H 0:0 0000.0

H 003:0:0 0:00:05 .ANFuS 00.0000 0:0E0 000:0:0t.0 06:00 05:28 :. 0:000. 0000:0003 .Amvucv 00.0000 0:0 0.0000 0.000000 .0:0_>_0:.
5.0.3 080 :0>0 0:00.:00E00 20:00 0:50: :. 0:000. 0002030.. .Amodwn: 00.. .3 00.0000 :000 :00 05:28 0:0 030: :_ :000:0000 c005.

 

 

o 00 33.0 H 03.20 0 0 33.0 H 33.0 mo 00 303 H 033 we 0 003.0 H 300.0 00.0000 .m.
o 0 83.0 H 0000.0 0 0 030.0 H 0000.0 mo 0 030.0 H 33.0 o 0 0000.0 H 008.0 .5090 0000. 53.23000 .0.
o 0 300.0 H 303 o 0 003.0 H 88.0 no 0 030.0 H 003.0 a 0 03 3.0 H 0000.0 0003020080 .0,
< 0 88.0 H 88.0 < 0 0000.0 H 0000.0 < 0 030.0 H 800.0 9.. 00 030.0 H 080.0 030.0000 .m
m 0 003.0 H 80.0 o 0 003.0 H 003.0 m 0 0003 H 330.0 m 0 0000.0 H 03.3 0:002:0300 .m.
< 0 0000.0 H 080.0 m 0 0000.0 H 33.0 m 0 080.0 H 008.0 0.... 00 030.0 H 080.0 .0020w 02m. 530000: .m
< 0 0000.0 H 88.0 < 0 88.0 H 0000.0 .0. 0 88.0 H 0000.0 0.0. 0 030.0 H 080.0 23030300000 .m
< 0 0000.0 H 88.0 m 00 33.0 H 003.0 .0. 0 030.0 H 800.0 00 0 030.0 H 83.0 E00.EOQ0.E .m
o 0 800.0 H 0000.0 0 0 030.0 H 38.0 mo 00 003.0 H 033 m0 0 03 3.0 H 300.0 5:005:00 .w
< 0 003.0 H 030.0 m 0 0.. 3.0 H 003.0 00 0 03 3.0 H 0000.0 00 0 03 3.0 H 33.0 000030 .0.
< 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 < 0 0000.0 H 0000.0 00:00.00? 0.00.0000 .0
m 0 03 3.0 H 33.0 0.... 0 003.0 H 030.0 000 0 38.0 H 003.0 m< 0 0000.0 H 88.0 00:093.. .m.

«t x002. 02 0.00:5 :0 0.00.5 0. 0.00:5 00.0000

 

0:0 0.0."..300. .030 000.002

.830008 w... 0300.

46

 

.0300 H 8:0

0.00000 5.3 .mrucv 39 00.00 05 0.5.3 0:00:00E00 000.000 0000. 00000030.. 000.0000 00.. >0 00.0000 0000 .00 030. 0. 00.00.0000 000.).

 

 

 

47

0 :30 H 00000 0 00000 H 0000.0 0 00000 H 00000 000.000 .m.
a 003.0 H 88.0 a 38.0 H 88.0 m 88.0 H 053 0000.0 5.2.. 25.20% .0
0 300.0 H «0000 0 003.0 H 0000.0 0 0000.0 H 5000 0003020080 .m
0 0000.0 H 0000.0 0 0000.0 H 0000.0 0 0000.0 H 0000.0 020000.000 .m
0 0000.0 H P0000 0 00000 H 0300 0 0000.0 H 0000.0 E000~:0..E00 .w
0 0000.0 H 0000.0 0 0000.0 H 0000.0 0 0.00.0 H 3.00 .moaam 02m. 230800.. .m
0 0000.0 H 0000.0 0 0000.0 H 0000.0 0 0000.0 H 0000.0 230029.30 .m
0 0000.0 H 0000.0 0 0000.0 H 0000.0 0 0030 H 5000 E00..:0Q0.E .m
0 00000 H 0000.0 0 0000.0 H 0000.0 0 00000 H 53.0 E30555 .m
0 00 5.0 H 0000.0 0 0H 5.0 H 020.0 0 050.0 H N050 000030 .m
m 88.0 H 88.0 m 88.0 H 88.0 m 88.0 H 88.0 000000.... 0.8.300 .0
m 0H 3.0 H 83.0 m 2 8.0 H 35.0 a 0.8.0 H 008.0 0500.2. .0

Eu 0.0.. E0 0.x. E0 0.0 020000

at :83. .260 3282

 

0.0.00 00:00:00 0000 n.0 0006:: 05 >0 000.20 00000000 00.000.00.00 00 .0080: .002 05 00 5000 00000000
0000 00 00.0000 0000 .00 0000.00.00 002. .0>oo 05.0000. .3000 0000.000 08 «t 0.002, :0 000000.. .80 0.0? 000
.0. .0. 0.00 05000 00000000 00.5 00 00.05.30 00.0000 E0000, NF “.0 00000300 00000000 H .050 05.0000 0005. .0; 0.00 ..

.>.02.00000.
0050.00 000 02,0. .0. 0000.0H 000 .0000 H 003 .000 0.00006 .Amrucv 05000 0000.0 00000.0...0 0.0000 00.00.00 0. 0.000. 0000.000: guru...

05000 0.0..0000 .000.200. 0.5.; 00... .05 0000000080 0.0000 0.050. 0002026.. 000000. 00.. >0 5000 0000 .0. 0.50. 0. 02.0.0000 000s.

 

m 0 0036 H :3... m 0 0000.0 H 500.. m 0 0000.0 H 000.; < 0 0000.0 H 0000.0 50 0.00
m 0 0000.0 H #000... m 0 0000.0 H 003.9 m 00 mmnod H mmmmé < 0 0P 3.0 H mmmmd Eo ON
<0 rmvrdemmré (0000—..0HO0Noé < mmvvwdHnmnod <0N0N06H0036 Eu 0%

ND. 0.00.5 ON? x002. no x002. 2. 0.00.5 5000
.0>00 0.0.0004.

0.0.00 00:00:00 0.00 .0 .000.00

05 >0 000.20 0.00.000 0E0....0.00 .0 .00E00 .0.0. 05 00 5000 0.0..0000 0000 .0 00.0000 0000 .0. 00.0.0200 003
.0>00 0.0.000< .0000 5000.000 gum: x002, u>00N 0080.000 00 u om. v.00; .000N 0050.000 0. ">0 0.002. .0000
0000.000 0N u 0. x003 0.002, 6.003 0. 000000. .0>00 .0000000 00.2.0.0 000m 0. 000m 05 .0>0 000 0.0. 000 .0. >
.00. 05000 0.050000 00.5 .0 00.02.30 00.0000 .0>0 0000.0>0 000..0.>00 20000.0 H .0>00 0.0.0000 000.2 .0... 0.00 .r

Figure 1.1. Absolute cover :t standard deviations of 12 Sedum spp. cultivated at
three substrate depths (4.0, 7.0, and 10.0 cm) over the 2005-2008 growing
seasons. Symbols represent absolute cover means with standard deviations
(n=3). For clarity of graphing, weeks 0 through 120 were plotted at every other

data point.

49

 

00..0...0. >030 .000 0000.5

 

 

 

 

 

 

 

 

 

 

 

 

 

00—. 0a.. 00—.
I V
q
S
I «.0 m-
_, .0.
mm 3
00.0.0 .0 . .O M.
0 0000.0 000.... 05.500 .0 I I .0. I I M
I 0.0300880 .0 I IDI I I 0.0 h...
0.52.600 .0 I (r I V
5000.00.53 .0 IAVI IO
000.00 00.0. 53800.. .0 . I9. I
. _, . , .. 030303.500 .0 l IDI l
C . ,C, nw -C. I 530.5010... .0 I I 0.0
50.00.00: .0 I I .dr I I
L. 00.9.... .0 l I I
I LI 0 000930.... 0.00.0000 .0 I O. I
0.000 0.8.0000 .00 0... .. 00:00.2. .0 I
\ \ \ W \ \ m6
\ \ \ \ \ \

.0: 0.00.“.

50

 

 

 

5000 30.00000 .00 0.5
\ \

00.00.00. >030 .000 00.002.

0.00...

'99A9

00..

 

00

9‘

 

00.0.0 .0
2000.0 000... 0.00000 .0
0003000080 .0
00000000 .0
.0000.00.E00 .0
.000000 00.0. 000.000. .0
000000.203 .0
0.00.0000... .0
0.0000000 .0
00.000 .0
00000.02... 0.00.0000 .0
05.092. .0

K \

 

 

I 0.0

I 0.0

(0v) moo aInIosqv

I 0.0

 

 

\\

V\

E0

.2... 059...

51

 

 

 

 

0.000 0.00.0000 .00 0.0..
\ 0

our

00..0...0. >030 .000 0000.5
00. 00

.. .. I .WWﬁIWIw...

   

A

  

00.0.0 .0

0000.0 000.... 0.0.0000 .0

20.000000 .0

00000000 .0

5008000000 .0

000.00 00.0. 00..0...0.. .0

E0000.0.000 .0

500.0000... .0

0.0.0.000 .0

00.030 .0

0000000.... 0.00.0000 .0

05.0053 .0
\ N

 

I
‘1
0

I
‘0.
6
(0v) 1000 emlosqv

I 0.0

 

 

V\

Q\

0...
.0... 0.00.“.

52

 

 

 

 

 

 

 

 

 

 

 

 

4° 0 Maximum
0 Minimum
30 -
... . . ..
A O . .
g 0 O . . . .
2 20- .' ° . .-
3 o o O O O
"' 1o - ° 0 o 0 0
a O . o o o o o. 0
GE) . O o . o .0
I- o _ 9 O O 00
I- .0 O O o O O
._ o o
< o o o 0o
-10 - o
o
-20 I
200 -
E
g 150 -
C
.9 .
:‘2 100 l; :2
.9- $2;
5 iii 5: ' a
50 3‘ i.”
{‘3} 5 .‘i 1 . E
:9 :3 l ’ " "‘
iig Hg .. ‘0‘
o la :1 .a' 0‘1 “ .
Jun-2005 Dec-2005 Jun-2006 Dec-2006 Jun-2007 Dec-2007 Jun-2008

Date

Figure 1.2. Monthly average maximum air temperatures (°C), monthly average
minimum air temperatures (°C), and monthly total precipitation (mm) throughout
the study (1 June 2005 to 30 September 2008). Data is from the Michigan
Automated Weather Network’s East Lansing weather station (located adjacent to
the research site).

53

 

0.14

0.12 -

0.10 -

0.08 -

0.06 -

0.04 -

Volumetric Moisture Content (m3lm3)

 

 

 

4 7 1 0
Substrate Depth (cm)

Figure 1.3. Substrate volumetric moisture content (m3/m3) 1 standard errors for
three substrate depths (4.0, 7.0, and 10.0 cm) averaged over two growing
seasons (2007 and 2008). Uppercase letters represent mean separation by LSD
(P5005).

 

 

—O— 4cm
0.35 - —O— 7cm
__ —-v— 10cm

 

 

 

 

 

  
 

Volumetric Moisture Content (m3lm3)
O
8

 

 

 

0.20 -
0.15 ~
0.10 . . . . . . .
05/09/07 05/10/07 05/1 1/07
Date

Figure 1.4. Substrate volumetric moisture content (m3/m3) :I: standard errors for a
selected three day period following a 22.1 mm rain event at three substrate
depths (4.0, 7.0, and 10.0 cm).

55

CHAPTER TWO

Solar Radiation Intensity Inﬂuences Extensive Green Roof Plant Communities

56

Abstract

Two studies were conducted on a third-story rooftop to quantify the effect of solar
radiation (full sun vs. full shade) on several US. native and non-native species
for potential use on extensive green roofs. In the ﬁrst study, plugs of six native
and three non-native species were planted in May 2005 at two different substrate
depths (8.0 cm and 12.0 cm) both in sun and shade. Absolute cover (AC) was
recorded using a point-frame transect during the growing season beginning in
June 2005 and every two weeks thereafter for a period of four years. By week
174 (23 Sept 2008), most species exhibited different AC within a depth between
sun and shade. However, when all species were combined, overall AC did not
differ between sun and Shade within a depth. This indicated that while species
make-up was changing among solar radiation levels, that overall coverage was
not signiﬁcantly different between sun and shade. For all substrate depths and
solar levels, the most abundant species were Sedum acre, AIIium cemuum,
Sedum album ‘Coral Carpet’, and Talinum calycinum. Less suitable species
included Talinum parviﬂorum, Carex ﬂacca, Sedum stenopetalum, and Sedum
divergens, which all exhibited 0 or near 0 AC regardless of depth or solar
radiation levels. With the exception of T. calycinum, native species were less

abundant than non-native species.

In the second study, six typical extensive green roof species of Sedum grown

from seed in 10.0 cm (3.9 in) of substrate were compared in both sun and shade

during May 2005. AC was evaluated as in the previous study. Solar radiation

57

did not affect AC, but over all species composition differed between sun and
shade levels. The most proliﬁc species in full sun were Sedum acre (0.57 AC)
and Sedum album ‘Coral Carpet’ (0.51 AC). Sedum kamtschaticum (0.57 AC)
and Sedum spun'um ‘Coccineum’ (0.35 AC) performed the best in the shade. For
both solar levels, the least abundant species at week 174 were Sedum

pulchellum (0.0 AC) and Sedum album ‘Coral Carpet” (0.1 AC).

Introduction

Green roofs, or vegetated roofs, are frequently installed in urban areas because
of their ability to improve stormwater management (Carter and Jackson, 2007;
Getter et al., 2007; Hilten et al., 2008; Jarrett and Berghage, 2008) and mitigate
the urban heat island effect (T akebayashi and Moriyama, 2007). Green roofs
provide many additional beneﬁts as compared to traditional roofs including
energy conservation (Santamouris et al., 2007; Sailor, 2008), increased longevity
of rooﬁng membranes (Kosareo and Ries, 2007), reduction in noise and air
pollution (Van Renterghem and Botteldooren, 2008; Yang et al., 2008), increased
urban biodiversity (Baumann, 2006; Brenneisen, 2006), as well as providing a
more aesthetically pleasing living environment (Getter and Rowe, 2006;
Obemdorfer et al., 2007). Recent research has suggested that green roofs are

also a better return on investment than traditional roofs (Clark et al., 2008).

Green roofs are categorized as ‘intensive’ or ‘extensive’ systems. Intensive

green roofs are designed to be similar to landscaping found at ground level, and

58

as such require substrate depths greater than 15.0 cm and have ’intense’
maintenance needs. In contrast, extensive green roofs use shallower substrate
depths (less than 15.0 cm) and usually require minimal maintenance. Due to
building weight restrictions and costs, shallow substrate extensive green roofs
are much more common than deeper intensive roofs. Therefore, the focus of this

paper is on extensive green roofs.

Plant selection for green roofs is often difﬁcult due to the harsh urban
environment and because plants are frequently subjected to extremes in
temperature and drought due to their shallow substrate and elevation above
ground. In addition, research involving plant selection thus far has typically been
for roofs exposed only to full sun (Monterusso et al., 2005; Durhmanet al., 2007;
Getter and Rowe, 2008). However, as green roof implementation continues, it is
likely that many green roofs will be shaded by other structures. Incoming solar
radiation (insolation) directly and indirectly impacts plant growth. Depending on
species, this impact may be severe or negligible. In Utah, USA, high solar
radiation zones differed from low solar radiation zones in terms of species
performance (Dewey et al., 2004). Another study in the US. Paciﬁc Northwest
examined sloped roofs and found that in some cases the amount of solar

radiation inﬂuenced plant performance (Martin and Hinckley, 2007).

Furthermore, overall climate and roof microclimate will determine plant success

as well as the design intent for the green roof. A good example of this is the

59

state of Florida’s ﬁrst pilot green roof installed in Naples, Florida, USA in 2003.
The plant species chosen for this project were selected because of their green
roof success in northern latitudes. But these species failed to survive in Florida’s
environment. The researchers concluded that the “type of media used and the
green roof proﬁle structure are secondary to the correct choice of plants”

(Livingston et al., 2004).

Often clients desire the use of native species because of their real and perceived
beneﬁts, such as their longevity without the use of pesticides, fertilizers, or
irrigation (US EPA, 2008). But, plants could not actually be native to rooftops
since roofs are man-made artiﬁcial structures. However, many plant species
have evolved in extreme environments and are adapted to green roof conditions.
Because of client demand for native species, researchers in Michigan, USA
evaluated 18 native taxa on unirrigated extensive green roof platforms
(Monterusso et al., 2005; Rowe et al., 2005). After three years, only four of the
species survived. The majority of the plants tested were considered to be
drought tolerant, but their survival in a native environment relies on deep tap
roots to obtain moisture. In a shallow extensive roof, these roots can still grow
sideways, but periods of drought resulted in death. These plants may have all

survived with deeper substrates or supplemental irrigation.

One technique used to assess plant stress in harsh conditions is measurement of

chlorophyll ﬂuorescence. This technique is used to quantify the efﬁciency of the

60

photosynthetic apparatus (Maxwell and Johnson, 2000). Photon energy
absorbed by a chlorophyll molecule can be used to fuel photosynthesis,
dissipated as heat, or re-emitted as ﬂuorescence. Measurement of the latter is
used to indicate how efﬁcient the former two processes are proceeding.
Fluorimeters are used to measure this value, usually reporting the ratio (FJFm) of
variable ﬂuorescence (Fv) to maximum ﬂuorescence (Fm) which typically range
from 0.70 to 0.83, with values less than 0.60 indicating photosynthetic stress

(Ritchie, 2006).

Many studies of plant survival on green roofs are based on one to two years data
(Kircher, 2004; Emilsson and Rolf, 2005; MacDonagh et al., 2006; Durhman et
al., 2007; Nagase and Dunnett, 2008). Since green roofs are estimated to last
45 years or longer in terms of mechanical lifespan (Kosareo and Ries, 2007),
long-term plant performance beyond the ﬁrst few years of establishment is
important. Plants that survive initially on a green roof may not continue to exist in
the long term because of variability in climate and other factors. Therefore, the
objective of this study was to evaluate the effect of substrate depth and solar
radiation intensities (full sun vs. full shade) on substrate moisture, plant stress as
measured by chlorophyll ﬂuorescence, and plant community development of both
US. native and typical non-native green roof species in a Midwestern climate
over a period of four years. Native species chosen were native to the US. and

were drought tolerant without the beneﬁt of deep rooted systems.

61

Materials and Methods

Green Roof Plots. Green roof plots were established on the Communication Arts
building on the campus of Michigan State University in East Lansing, Michigan,
USA in May 2005. The roof of this building is constructed such that it can easily
accommodate varying substrate depths and it is tiered in such a way that
approximately 10% of the roof is always in complete shade (deﬁned here as the
absence of direct solar radiation). Six wooden frames were built measuring 120
cm x 234 cm with sides of 15.2 cm in depth. Each frame was divided into three
subplots with internal dimensions of 115 cm x 75 cm. Three frames were
situated in full shade (rarely receiving direct sunlight) and three in full sun (rarely

receiving shade from natural or human structures).

The wood frames were placed directly on the roof over a root barrier membrane.
Drainage material (Xero Flor XF 108; XeroFlor America, Durham, North Carolina,
USA) was placed inside each subplot. For additional water holding capacity, a
0.75 cm thick moisture retention fabric (Xero Flor XF159) capable of retaining up
to 5.92 kg-m'2 of water was placed over the drainage layer. Above the retention
fabric was the vegetation carrier (Xero Flor XF301). Growing substrate (Table
2.1) was then placed in each subplot on top of the vegetation carrier to a depth of
8.0 cm, 10.0 cm, or 12.0 cm. Substrate treatments were blocked by arranging
each depth randomly in each plot, replicated three times. This resulted in an
experimental model that was a randomized complete block design with three

factors replicated three times.

62

Plant Establishment —- Study 1. In order to compare native with non-native
species, six species native to the US. (AIIium cemuum (wild nodding onion),
Carex ﬂacca (heath sedge), Sedum divergens (cascade stonecrop), Sedum
stenopetalum (narrow-petaled stonecrop), Talinum calycinum (largeﬂower
fameﬂower), and Talinum parviﬂorum (sunbright)), as well as three non-native
species (Sedum acre ‘Oktoberfest’ (biting stonecrop), Sedum album ‘Coral
Carpet’ (white stonecrop), and Sedum urvillei (stonecrop)), were acquired as
plugs (120 cm3; 72lﬂat; Emory Knoll Farms, Street, Maryland, USA) that were
established in a standard propagation mix of peat, perlite, and vermiculite
(Super-Fine Germination Media, Farfard, lnc., Agawam, Massachusetts, USA).
On May 25, 2005, plugs were planted in both the 8.0 cm and 12.0 cm substrate
depth plots with seven plants per row 15 cm apart resulting in four rows. Each
plant species was randomly planted three times in each subplot, resulting in a
total of 162 plants in the study. Since this arrangement results in 28 spots and
there were nine species planted three times each, one spot remained empty in

each sub-plot.

Plant Establishment - Study 2. Seeds (Jelitto Staudensamen, GmbH
(Schwarmstedt, Germany» of six typical extensive green roof species (Sedum
acre (biting stonecrop), Sedum album (white stonecrop), Sedum kamtschaticum
(stonecrop), Sedum pulchellum (bird’s claw sedum), Sedum reﬂexum (crooked

sedum), and Sedum spun'um ‘Coccineum’ (creeping sedum)) were planted in

63

each 10.0 cm subplot. On May 25, 2005, seeds were sown after weighing 1.13 g
of each species in a beaker and combining the seed with 28.0 grams of ﬁne
vermiculite. This mixture was then evenly distributed in the subplot. Shade cloth
was placed over the seeds for three weeks to aid in germination and

establishment.

For both studies, plots were irrigated three times daily for 20 minutes for the ﬁrst
four weeks and once daily for 20 minutes for the next six weeks. Irrigation was
then terminated for the remainder of the studies. Removal of weeds took place

every two weeks.

Data Collection and Analysis. For both studies, a stainless steel point-frame was
used every two weeks during the ﬁrst three growing seasons and monthly in the
fourth growing season to measure community composition and change (Waite,
2000). The point frame had internal measurements of 115 cm x 75 cm and had 7
strings (50 pound Berkley Gorilla Super Braid Fishing Line) vertically and 10
strings horizontally to create 70 measurement points. The point frame sat 14 cm
above the frame surface on top of wooden pegs which were secured into the
corners of each frame. The point-frame sat directly on top of the wooden plots,
secured by ﬁnishing nails that allowed the frame to be positioned in the exact
same spot every time. A stainless steel skewer (measuring 38.0 cm long) was
placed vertically at each measuring point in the point-frame and each species the

skewer contacted was recorded, up to three canopy layers. Absolute cover was

64

calculated for each species at each substrate depth as the total number of

contacts recorded divided by the number of data collection points (Waite, 2000).

Environmental conditions were continuously recorded on a CR10X Campbell
Scientiﬁc Datalogger (Campbell Scientiﬁc Inc., Logan, Utah, USA) throughout the
experiments. Two 105T-L thermocouples (Campbell Scientiﬁc Inc., Logan, Utah,
USA) were placed in radiation shields and mounted on the wooden plots, one in
the sun and one in the shade, and connected to the datalogger. In addition, two
Ll200X-L Ll-COR Silicon Pyranometers (Campbell Scientiﬁc Inc., Logan, Utah,
USA) quantiﬁed total incoming solar radiation at each location; one in the sun
and one in the shade. A weather station with a 03001-L R.M. Young Wind
Sentry Set (consisting of a 3-cup anemometer and a wind vane) (Campbell
Scientiﬁc Inc., Logan, Utah, USA) and a TE525WS-L Texas Electronics 8" Rain
Gage (Campbell Scientiﬁc Inc., Logan, Utah, USA) was also situated on the roof.
A MSX20 solar panel (Campbell Scientiﬁc Inc., Logan, Utah, USA) was installed

to power all electronic equipment.

Substrate moisture and chlorophyll ﬂuorescence data were collected throughout
the ﬁnal two growing seasons. Substrate volumetric moisture was monitored by
inserting a theta probe (ML2x, Delta-T Devices, Ltd., Cambridge, United
Kingdom) into the substrate until the 6 cm prongs were completely buried.
Measurements were collected at random times each week in triplicate in each

subplot. Chlorophyll ﬂuorescence measurements were collected at six different

65

times (21 May 2007, 18 June 2007, 2 August 2007, 17 September 2007, 29 May
2008, and 20 August 2008) using a Hansatech plant efﬁciency analyzer (PEA;
Hansatech Instruments, Ltd., Norfolk, England). These measurement dates were
selected in order to cover a wide range of environmental conditions including
active growth and drought stress during the growing season. Each plant was
dark adapted for 20 minutes prior to measurement and illuminated with a 50%
light level. Maximum quantum efﬁciency of photosystem II was recorded (FvlFm).
Three single leaf blades of each surviving species were randomly selected in
each subplot and excised from the plant to be dark adapted and measured. This
was necessary because the PEA clips were not secure on the leaf of most

species while still attached to the whole plant.

Data Analysis. Absolute cover (AC) was calculated for each species at each
solar radiation intensity and substrate depth as the total number of contacts
recorded divided by the number of data collection points. Data was then
analyzed as mean AC using repeated measures with a ﬁrst-order autoregressive
(AR(1)) covariance structure (Wolﬁnger, 1996). Signiﬁcant differences between
treatments were determined using multiple comparisons (PROC MIXED, SAS
version 9.1.3, SAS Institute, Cary, NC). In addition, mean substrate volumetric
moisture content and mean chlorophyll ﬂuorescence data were analyzed by
PROC MIXED, least signiﬁcant differences (SAS version 9.1.3, SAS Institute,

Cary, NC)

66

Results and Discussion

Throughout the study, maximum daily solar radiation and maximum daily air
temperatures varied in the sun and shade. Representative data for year two (1
January 2006 - 31 December 2006) are shown in Figure 2.1. Mean maximum
daily air temperature and mean maximum daily solar radiation ﬂux during July
2006, which represents the peak of the growing season, were 33.9 C and 0.71
kW-m'2 in the sun and 33.8 C and 0.44 kW-m'2 in the shade. More variability was
seen among maximum solar radiation than in temperature between radiation
intensities, likely because temperatures above the surface of the roof were well

mixed (Figure 2.1).

Study 1 — Plugs of U. S. native and non-native species

When analyzed across all growing seasons, solar exposure, substrate depth, and
species treatments were all signiﬁcant (P=0.05). There were also signiﬁcant
interactions between species and solar level, species and substrate depth, as
well as a three-way interaction between depth, solar exposure, and species.
These results conﬁrm that the choice of species depends on solar exposure and

substrate depth.

At the end of the ﬁrst growing season, C. ﬂacca appeared to be one of the most
abundant species for both substrate depths in the shade (Figure 2.2). However,
in subsequent years, it decreased in abundance during the driest portions of the

summer which likely impacted overall regeneration. By the end of the four years,

67

this species exhibited zero or near-zero AC. In contrast, at the end of the second
growing season, 8. acre ‘Oktoberfest’ had established itself as the most
abundant species for both substrate depths in the shade and exceeded an AC of
0.6 by the third growing season (Figure 2.2). For both substrate depths in the
shade, A. cemuum is the next most abundant species by the end of the fourth
growing season, followed by 8. album ‘Coral Carpet’ and T. calycinum. For all
depths and solar radiation levels, T. parviﬂorum (which had completely
disappeared after the ﬁrst growing season), 8. stenopetalum, and S. divergens

had the lowest AC (Figure 2.2).

In the sun, by the second growing season both substrate depths were dominated
by 8. album ‘Coral Carpet”, followed by T. calycinum and 8. acre ‘Oktoberfest’
(Figure 2.2). At 12 cm, A. cemuum closely follows as the fourth most abundant,
but this species was not nearly as abundant as it was in the shade at the same
depth. This is likely due to a severe phototrophic bending when grown in the
shade, which may overrepresent this species in point-frame sampling techniques
(Wilson, 1960), as it will fall under a sampling point more often than the erect
habit exhibited in the sun. While not measured in this study, A. cemuum
appeared to have greater biomass per given area due to its upright growth habit

than the AC measure suggests.

For speciﬁc point-in-time comparisons, week 174 (23 September 2008) of the

study was chosen to represent the end of the fourth growing season, where

68

maximum biomass should have been reached for the year for most species. At
the 8 cm substrate depth, AC only varied for four species between sun and
shade (A. cemuum, 8. acre ‘Oktoberfest’, 8. album ‘Coral Carpet’, and T.
calycinum) (Table 2.2). At 12 cm, the same four species, plus C. ﬂacca had
differing AC between sun and shade. The remaining four species all exhibited 0
or near 0 AC for both sun and shade, indicating that they are unsuitable species
for these substrate depths in this climate regardless of solar radiation intensities,

at least when supplemental irrigation is not available.

When all species are combined, overall AC did not differ between solar
exposures within each substrate depth (Table 2.2). This indicates that while
species make-up changed among solar radiation levels, overall coverage was
not signiﬁcantly different between sun and shade for this time. Species most
suited to the environmental conditions present will expand into open space. This
emphasizes the importance of planting numerous species to ensure plant
diversity in order to obtain and maintain full coverage on a roof. However, what
this study does not capture is the difference in above-ground biomass between
solar levels. Within a given substrate depth, the shade canopy was much taller

than the sun canopy (personal observation).
As a group, the native species chosen for this study did not grow as well as non-

natives. Talinum calycinum was the most abundant native species in the sun,

while A. cemuum had a higher AC in the shade. Two of the native species (8.

69

stenopetalum and S. divergens) are native to the western US. (USDA, 2008)
and thus may not have been suited for the hot humid summer conditions of the
Midwestern U.S. (Snodgrass and Snodgrass, 2006). By contrast, C. ﬂacca is
native to Michigan (USDA, 2008), but its poor performance is likely because it
requires a substrate depth of at least 15 cm (Snodgrass and Snodgrass, 2006).
The native Talinum species (T. calycinum and T. parviﬂorum) were outside of
their hardiness zone (zone 6; Snodgrass and Snodgrass, 2006) in this study
(zone 5; Cathey, 1990) and as such were self-sowing annuals, which may
explain why they did not perform as well as other (non-native) species. They are

proliﬁc seeders, but need bare soil in order to germinate.

Mean substrate volumetric moisture content did not differ between the two solar
radiation levels at the 8.0 cm substrate depth (Table 2.3). However, at the 12.0
cm substrate depth and when averaged across depths, soil moisture was higher
in the shade than in the sun. These results indicate that the full sun environment
likely had greater evaporation from the soil and plant transpiration losses leading
to less water held in the soil at any given measurement time. _ It is likely that
differences in AC between shade and sun for some species were due to lower
water availability in the sun compared to the shade. This conﬁrms what others
have found; that when all else is equal, shaded environments often have greater
substrate moisture than sunny environments (Williams et al., 1993; Kobayashi et

aL,1997)

7O

Within the same solar radiation level, mean volumetric moisture content was
statistically the same between the substrate depths. But, had a shallower
substrate depth been used, it is likely that there would be a difference in moisture
content. VanWoert et al. (2005) found substrate moisture differences between
2.0 cm and 6.0 cm substrate depths. Getter and Rowe (2009) found that 4.0 cm
substrate depths held less moisture then 7.0 or 10.0 cm depths, but that 7.0 and
10.0 cm substrate depths were statistically the same. The latter two are very

similar to the two substrate depth treatments (8 cm and 120m) used in this study.

Chlorophyll ﬂuorescence data exhibited mixed results. Within an individual
species, mean F,,IFm values were not signiﬁcantly different between solar
radiation or depth treatments (data not shown). This may be due to the fact that
measurements for chlorophyll ﬂuorescence occurred during the third and fourth
growing season only, whereby three (Sedum divergens, Sedum stenopetalum,
and Talinum parviﬂorum) of the nine initial species had zero AC (i.e., no plants to
take measurements upon). Had chlorophyll ﬂuorescence data been taken during
the ﬁrst growing season, perhaps noticeable differences would have been
detected between species that ultimately survived the four year study and those
that did not. Then again, chlorophyll ﬂuorescence is only a tool to quantify plant

stress and not necessarily a prediction of survival.

When averaged across all species, solar levels were signiﬁcant, while substrate

depth and the interaction between the two were not signiﬁcant. Mean Fv/Fm

71

values within the same substrate depth are higher in the shade than in the sun
(Figure 2.3). This indicates that perhaps the shaded environment induced less
photosynthetic stress than the sun, although neither of the average F,,IFm values
was considered stressful (Ritchie, 2006). It is possible that the number of
observations (N) was insufﬁcient for individual species to ﬁnd a difference in

Fv/Fm values, whereas combined data provided more degrees of freedom.

Study 2 - Seed of typical non-native green roof species

When analyzed across all growing seasons, solar exposure and species were
both signiﬁcant (P=0.05). There were also signiﬁcant interactions between solar
level and species, highlighting the importance of species choice for any given

solar exposure.

In the shade, S. pulchellum was initially the most abundant species during the
majority of the ﬁrst growing season and into the beginning of the second (Figure
2.4). But as the community developed, 8. acre clearly becomes more abundant
by the end of the second growing season and the ﬁrst half of the third. However,
the fourth growing season produced very different results, highlighting the
importance of long term studies. During the entire fourth growing season, 8.
pulchellum was the least abundant species and virtually absent. By the end of
that season, S. kamtschaticum was most abundant, followed by S. spurium
‘Coccineum’, 8. acre, and 8. album. While 8. reﬂexum was not highly abundant

at this time, it was still a respectable accent plant with an AC of 0.2.

72

By contrast, in the sun, S. acre was the most abundant species across all four
growing seasons (Figure 2.4), although by the end of the fourth season S. album
was nearly as abundant. The next most abundant species are S. kamtschaticum
and S. spurium ‘Coccineum’. Sedum pulchellum begins the study as the second
most abundant species in the sun, but then fails to reappear after the second
growing season. Sedum reﬂexum is also virtually non-existent at this solar load,

making it an unsuitable choice as well.

The initial prominence of S. pulchellum in the ﬁrst two growing seasons, followed
by its disappearance is likely due to its growth strategy. This species is a spring
ephemeral, whereby its seed germinates in very early spring or late winter,
ﬂowers, and sets seed before other species crowd it out in mid June. In the
second growing season, this species reached a maximum AC of 0.77 and 0.29 in
the shade and sun, respectively, before dieing off for the summer (Figure 2.4). In
year three, S. pulchellum only reached a maximum AC of 0.14 and 0.07 in the
shade and sun, respectively. This is probably due to fewer gaps in the canopy
for seed germination to occur because evergreen species like S. acre and S.

album had closed much of the canopy by that time.

For speciﬁc point-in-time comparisons, week 174 (23 September 2008) of the

study was chosen to represent the end of the fourth growing season, where

maximum biomass should have been reached for the year for most species.

73

Within a given species, most exhibited similar AC between sun and shade (Table
2.4). The exception was 8. acre and 8. album, both of which were more
abundant in the sun than in the shade. This agrees with the general observation
that these two species tend to dominate shallow green roofs. They seem to be
better adapted to withstand full sun and limited substrate moisture. For all
species combined, there was no difference between total AC in the sun and
shade (Table 2.4). This indicates that while species make-up is changing
amongst solar radiation levels, that overall coverage is not signiﬁcantly different

between sun and shade for this time period.

Mean substrate volumetric moisture content for the two solar radiation levels at
this 10 cm substrate depth was 0.138 and 0.122 for shade and sun respectively
(data not shown). Over the entire growing season, it is evident that at almost
every measurement time, shade moisture content was equal to or greater than

sun moisture content (Figure 2.5).

Chlorophyll ﬂuorescence differed between solar radiation levels. Mean F,,/Fm
values for two species (8. album and S. spun’um ‘Coccineum’), as well as when
averaged across all species, had higher values in the shade than in the sun
(Figure 2.6). Mean chlorophyll ﬂuorescence values of individual species never
fell below 0.70, indicating very little, if any, stress to the photosynthetic system
(Ritchie, 2006). Other research has established the same trend for many

species in this genus. In a controlled greenhouse watering study, Durhman et al.

74

(2006) found that three Sedum (S. acre, 8. kamtschaticum, and S. reﬂexum)
species maintained active photosynthetic capacity for at least 88 days without
water. In addition, FV/Fm values for these species did not fall below 0.5 until at
least 40 days without water. This may at least partially explain why Sedum

species are such good candidates for extensive green roofs.

Conclusions

Results show the importance of long-term plant evaluations at different substrate
depths and solar radiation intensities. For shaded locations, S. acre and S.
kamtschaticum are both excellent choices for all of the depths tested. Other
suitable species are S. spun'um, S. album “Coral Carpet’, and A. cemuum.
Where native species are desired, T. calycinum is also a good choice. For sunny
locations, S. album ‘Coral Carpet’ and S. album are the best choices of the
species tested. Other good choices for the sun are 8. acre and the US. native
species T. calycinum. Species to avoid in Midwestern (or similar) climates
include T. parviﬂorum, S. stenopetalum, S. divergens, and possibly 8.
pulchellum. The latter would work well as a ﬂowering spring species, but may
disappear after several years. Plants that would make good accent species, but
will likely never dominate or ﬁll in gaps include S. reﬂexum, S. urvillei, and C.

ﬂacca.

Acknowledgements

Funding for these studies was provided by Ford Motor Company, Dearborn, MI,

75

ChristenDETROlT Rooﬁng Contractors, Detroit, MI; XeroFlor America LLC,
Durham, NC; the Michigan Agricultural Experiment Station; and Emory Knoll

Farms, Street, MD.

76

Literature Cited

Baumann, N. 2006. Ground-Nesting Birds on Green Roofs in Switzerland:
Preliminary Observations. Urban Habitats 4137—50.

Brenneisen, S. 2006. Space for Urban Wildlife: Designing Green Roofs
as Habitats in Switzerland. Urban Habitats 4:27-36.

Carter, T. and CR. Jackson. 2007. Vegetated roofs for stormwater management
at multiple spatial scales. Landscape and Urban Planning 80:84-94.

Cathey, HM. 1990. USDA plant hardiness zone map. USDA Agricultural
Research Service Miscellaneous Publication No. 1475.

Clark, C., P. Adriaens, and F.B. Talbot. 2008. Green Roof Valuation: A
Probabilistic Economic Analysis of Environmental Beneﬁts. Environmental
Science and Technology 42:2155—2161.

Dewey, D.W., P.G. Johnson, and R.K. Kjelgren. 2004. Species composition
change in a rooftop grass and wildﬂower meadow. Native Plants 2:56-65.

Durhman, A.K., D.B. Rowe, and C.L. Rugh. 2006. Effect of watering regimen on
chlorophyll ﬂuorescence and growth of selected green roof plant taxa.
HortScience 41 :1623-1628.

Durhman, A.K., D.B. Rowe, and C.L. Rugh. 2007. Effect of substrate depth on
initial coverage, and survival of 25 succulent green roof plant taxa. HortScience
42:588-595.

Emilsson, T. and K. Rolf. 2005. Comparison of establishment methods for
extensive green roofs in southern Sweden. Urban Forestry & Urban Greening
3: 1 03—1 1 1 .

Getter, KL. and DB. Rowe. 2006. The role of extensive green roofs in
sustainable development. HortScience 41:1276-1285.

Getter, K.L., D.B. Rowe, and J.A. Andresen. 2007. Quantifying the effect of slope
on extensive green roof stormwater retention. Ecological Engineering 31 :225-
231.

Getter, K.L. and DB. Rowe. 2008. Media Depth Inﬂuences Sedum Green Roof
Establishment. Urban Ecosystems 11:361-372.

Getter, KL. and DB. Rowe. 2009. Substrate depth inﬂuences Sedum plant
community on a Green Roof. HortScience. lnPress.

77

Hilten, R.N., T.M. Lawrence, and E.W. Tollner. 2008. Modeling stormwater runoff
from green roofs with HYDRUS-1 D. Journal of Hydrology 358:288-293.

Jarrett, AR. and RD. Berghage. 2008. Annual and individual green roof
stormwater response models. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto.

Kircher, W., 2004. Annuals and Sedum-cuttings in seed-mixtures for extensive
roof gardens. Acta Hort. 643:301-303.

Kobayashi, T., Y. Hori, and N. Nomoto. 1997. Effects of Trampling and
Vegetation Removal on Species Diversity and Micro-Environment under Different
Shade Conditions. Journal of Vegetation Science 81873-880.

Kosareo, L., and R. Ries. 2007. Comparative environmental life cycle
assessment of green roofs. Building and Environment 42:2606—2613.

Livingston, E, C. Miller, and M. Lohr. 2004. Green Roof Design and
Implementation in Florida. In: Proc. of 2"d North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Portland, OR. 2-4
June 2004, The Cardinal Group, Toronto, pp. 353-364.

MacDonagh, L.P., N.M. Hallyn, and S. Rolph. 2006. Midwestern USA plant
communities + design =Bedrock bluff prairie greenroofs. In: Proc. of 4th North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Boston, MA. 11-12 May 2006. The Cardinal Group, Toronto.

Martin, MA. and TM. Hinckley. 2007. Native Plant Performance On A Seattle
Green Roof. In: Proc. Of 4"1 North American Green Roof Conference: Greening
Rooftops for Sustainable Communities, Minneapolis, MN, 29 April — 1 May 2007,
The Cardinal Group, Toronto.

Maxwell, K. and GM Johnson. 2000. Chlorophyll ﬂuorescence - a practical
guide. Journal of Experimental Botany 51:659-668.

Monterusso, MA, 03. Rowe, and C.L. Rugh. 2005. Establishment and
persistence of Sedum spp. and native taxa for green roof applications.
HortScience 40:391-396.

Nagase, A. and N. Dunnett. 2008. Experiments in plant selection for extensive
green roofs: Performance of annual plant species and the amount of water runoff
from different vegetation types. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto.

78

Obemdorfer, E., J. Lundholm, B. Bass, R.R. Coffman, H. Doshi, N. Dunnett, S.
Gafﬁn, M. Kohler, K.K.Y. Liu, and DB. Rowe. 2007. Green Roofs as Urban
Ecosystems: Ecological Structures, Functions, and Services. BioScience
57:823-833.

Ritchie, G. 2006. Chlorophyll ﬂuorescence: what is it and what do the numbers
mean? In: USDA Forest Service Proceedings RMRS-P-43. Fort Collins, CO: US.
Department of Agriculture, Forest Service, Rocky Mountain Research Station.
pp. 34-43.

Rowe, D.B, M. Monterusso, and C.L. Rugh. 2005. Evaluation of Sedum species
and Michigan native taxa for green roof applications. In: Proc. of 3rd North
American Green Roof Conference: Greening Rooftops for Sustainable
Communities, Washington, DC, 4-6 May 2005, The Cardinal Group, Toronto, pp.
469-481.

Sailor, DJ. 2008. A green roof model for building energy simulation programs.
Energy and Buildings 40:1466-1478.

Santamouris, M., C. Pavlou, P. Doukas, G. Mihalakakou, A. Synnefa, A.
Hatzibiros, and P. Patargias. 2007. Investigating and analysing the energy and
environmental performance of an experimental green roof system installed in a
nursery school building in Athens, Greece. Energy 32:1781-1788.

Snodgrass, E. and L. Snodgrass. 2006. Green Roof Plants: A Resource and
Planting Guide. Timber Press, Inc., Portland, OR.

Takebayashi, H. and M. Moriyama. 2007. Surface heat budget on green roof and
high reﬂection roof for mitigation of urban heat island. Building and Environment
42:2971—2979.

USDA, NRCS. 2008. The PLANTS Database, National Plant Data Center, Baton
Rouge, LA USA. 25 November 2008. http://plants.usda.gov.

US. Environmental Protection Agency. 2008. Green Landscaping. 2 November
2008. http://www.epa.gov/greenacres/index.htmI#Beneﬁts.

Van Renterghem, T. and D. Botteldooren. 2008. Numerical evaluation of sound
propagating over green roofs. Journal of Sound and Vibration 317:781—799.

VanWoert, N.D., D.B. Rowe, J.A. Andresen, C.L. Rugh, and L. Xiao. 2005.
Watering regime and green roof substrate design affect Sedum plant growth.
HortScience 40:659-664.

Waite, S. 2000. Statistical Ecology in Practice: A guide to analyzing
Environmental and Ecological Field Data. Prentice Hall, Harlow, England.

79

Williams, K., M.M. Caldwell, and J.H. Richards. 1993. The inﬂuence of shade and
clouds on soil water potential: The buffered behavior of hydraulic lift. Plant and
Soil 157:83-95.

Wilson, W.J. 1960. Inclined point quadrates. New Phytologist 59:1-8.
Wolﬁnger, RD. 1996. Heterogeneous variance: covariance structures for
repeated measures. Journal of Agricultural, Biological, and Environmental
Statistics 1:205-230.

Yang, J., O. Yu, and P. Gong. 2008. Quantifying air pollution removal by green
roofs in Chicago. Atmospheric Environment 42:7266—7273.

80

Table 2.1. Initial Physical and Chemical Properties of Substrate.

 

 

Component Unit Method

Total Sand 86 % Gee and Bauder, 1986
Very Coarse Sand (1-2 mm) 2.62 % Gee and Bauder, 1986
Coarse Sand (0.5-1 mm) 20.08 % Gee and Bauder, 1986
Medium Sand (0.25-0.5 mm) 41.68 % Gee and Bauder, 1986
Fine Sand (0.10-0.25 mm) 19.92 % Gee and Bauder, 1986
Very Fine Sand (0.05-0.10 mm) 1.7 % Gee and Bauder, 1986

Silt 10 % Bouyoucos, 1962

Clay 4 % Bouyoucos, 1962

Soil Textural Class Loamy Sand Bouyoucos, 1962

Bulk Density 1.37 g-cm’3 Ferguson et al., 1960

Capillary Pore Space 22.05 % Ferguson et al., 1960

Non-Capillary Pore Space 10.30 % Ferguson et al., 1960

Water Holding Capacity at 0.01 MPa 16.05 % Ferguson et al., 1960

Inﬁltration Rate 13.472 in/hr Ferguson et al., 1960

pH 7.9 NCR-13, 1998

Conductivity (EC) 1.38 mmho-cm'1 NCR-13, 1998

Nitrate 47 ppm NCR-13, 1998

Phosphorus 3.6 ppm NOR-13, 1998

Potassium 23 ppm NCR-13, 1998

Calcium 388 ppm NOR-13, 1998

Magnesium 38 ppm NCR-13, 1998

Sodium 77 ppm NCR-13, 1998

Sulfur 73 ppm NOR-13, 1998

Boron .7 ppm NOR-13, 1998

Iron 8.2 ppm NOR-13, 1998

Manganese 2.6 ppm NCR-13, 1998

Zinc 6.4 ppm NCR-13, 1998

Copper 1.0 ppm NCR-13, 1998

 

Analysis per A&L Great Lakes Laboratories, Inc., Ft. Wayne, Indiana

81

.Aoucv 5:200 0800 05 55:5 0:00:00Eoo 06:00

0.050. 0000 .095 .8u5 Bo: 9:00 05 55:5 0:03:0an0 03:00 90:2 0002026.. .Amodwov om._ E 020000 :000 :8 026: :_ 530.0000 :00:

 

 

 

 

82

m otd H ﬁve 0 mmod H oomé 0 oood H mm; m nwmod H dome UmEnEoO 0060mm =<
< 0 oood H oood < m oood H oood < m oood H oood < m oood H oood AZV EEoEEmQ 232:3.
m m mmod H mmod o n omod H movd m m mvdd H vmmd m n mmod H omvd AZV 2:350me :5:th
m 0 mmod H ovod m 0 wood H omrd < 0 god H o5d < 0 too H mmod 8:33 Eobmm
< m oood H oood < 0 mood H mood < 0 mood H mood < m mood H mood AZV EES0Q020~0 5300M,
< m oood H oood < m mood H mood < m oood H oood < 0 mood H mood AZV 0.209926 Ezbmm
o m mood H mord m a mood H nmmd m m mood H Homo o n Sod H mid .5900 3.60. E380 Ezbmm
m a mood H 30d 0 m mmod H mood o a mood H Food m 0 Food H mmod metmnowxo. 0.60 Ezbmm
m n mmod H onod < m oood H oood < m oood H oood < m oood H oood AZV moomt .50ng
o n vood H vad m 0 wood H mo_.d m a mood H ovvd < 0 wood H Nmod AZV E33500 E3S<

0225 55 0095 cam 0200mm
:6 oﬁuﬁgoo :3 c.0238

 

 

:5 0.33. .260 8283

.350: 5:00:00 9.00 do 00:5: 0:13 .0220 089:8 0E9.E:_oa 00 09:5:

.39. 05 00 006000 :000 :8 0000.30.00 002, .0>00.0S_o0n< .A0E0: 006000 05 Eta g :22, 88:03 .wj 0:0. 2 050:
0.0 02030 x5 .Aooow LonEoﬁ0w mmv v: 0.00; :o 00:80: 3.5 oNF 0:0 80 odv 0500c 90.50930 05 mm 00002230
000:0 0:0:0> :s0v 00E0:0E_ coo—0:00.. 5.00 oz; 00 006000 0:_: mo 06:0 0:00:30 4+ .0>00 05.0000 :002 .N.N 030k

Table 2.3. Mean substrate volumetric moisture content +/- standard errors of two
substrate depths (8.0 cm and 12.0 cm) at two solar radiation intensities (sun
versus shade) averaged over the 2007 and 2008 growing seasons.

 

Volumetric Moisture Content (m3lm3)

 

 

 

 

Substrate Degh Sun Shade Combined

8 cm 0.131 00.011 aA 0.15010.011aA 0.14010.008A
12 cm 0.109:0.009aA 0.13310.009bA 0.12110.006A
Combined 0.120 1 0.007 a 0.141 1 0.007 b

Mean separation in rows for each species by LSD (P5005). Lowercase letters denote
comparisons within the same row (n=3). Upper case letters denote comparisons within the same

column (n=3).

83

Table 2.4. Mean absolute cover +/- standard errors of 6 Sedum species at two
solar radiation intensities (sun versus shade) cultivated at a substrate depth of
10.0 cm reported on week 174 (23 September 2008). Absolute cover was

calculated for each species as the total number of point-frame contacts divided

by the number of data collection points.

 

Absolute Cover (Week 174)

 

 

 

Species Sun Shade

Sedum acre ’Oktoberfest’ 0.576 1 0.243 b C 0.109 1 0.055 a AB
Sedum album 0.509 :I: 0.252 b C 0.095 1: 0.055 a A
Sedum kamschaticum 0.286 1 0.254 a B 0.571 1 0.254 a C
Sedum pulchellum 0.000 1 0.000 a A 0.000 t 0.000 a A
Sedum reﬂexum 0.033 1 0.022 a A 0.200 1 0.159 a AB
Sedum spun'um ’Coccineum’ 0.267 1 0.194 a B 0.348 1 0.085 a BC
All Species Combined 1.671 1 0.054 a 1.324 1 0.080 a

Mean separation in rows for each species by LSD (P3005). Lowercase letters denote
comparisons within the same row (n=3). Upper case letters denote comparisons within the same

column (n=3).

84

 

Maximum Daily Sola Radiation

A 08- O 639300 0

~ (66’ ‘9» 0 ‘6 o 0
g o o o 0 0°
3. 0C9 (3 O Q30 %
V 0.6- O 0 0 %.‘O: .009) 00
.. o 9 a. 0 00
E C‘30 ‘..‘ ‘.. (DO
1: o o o
g 04“ 080 9's; '0 C?

a o

'5

(D

 

 

 

 

 

 

 

 

Temperature (C)

 

 

 

 

'20 T l T T T l l 1 I l 1
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Date (2006)

Figure 2.1. Representative data for maximum daily solar radiation intensities
(kW/m2) and maximum daily air temperatures (C) during the second year (1
January 2006 to 31 December 2006).

85

Figure 2.2. Absolute cover of 9 species cultivated at two substrate depths (8.0
cm and 12.0 cm) at two solar radiation intensities (sun versus shade) during the
four growing seasons (2005-2007). Six species are native to the US. (denoted
with (N) after species name). Symbols represent absolute cover means with
standard errors (n=9). For clarity of graphing, weeks 0 through 120 were plotted
at every other data point

86

5:025 >030 10:0 0x025

mt. at. may as m2. c3. 2...

 

 

 

E0 N: 6.00.5

 

 

o: m2. on no em mm

. I .ll lW/sl 3....
<

‘\

IW\W “WNW 51110111200 E11513

sz E32538 E3210 .1.
1011.12: E11000.

sz E11101oqocm10 Sodom
2 0110901110 E168
.6900 .0100. 1.11310 Ezbmm
0.100 E11000.

g 00001.1 .6100

g 2111:2100 21:11:11

cm on me or m

 

 

 

d
V
a.
s
m.
- to m.
a
O
m
- 0d a
J
V
b
- 0d
d...

 

\\
\\

.0N.N 01:9...

87

05.0055 >030 00.00 00.0015

mt. o: m2. 02. mm_. on: m: 2... m2. on no co mm cm. .ou 9. 2. m

 

 

E0 o 009%

 

\\

 

\
\W\ \Wv EEoEqu E1510 .1.

EV E05058 E0510 .1.

*\ \W 1011.31: E1000m
sz E01010qo11010 E1000m

\W 2 020900.10 E1100m.
000100 _0100. E11010 E1600...

0.00 Eob0m

2 00000 0100

A20 E::E00 E01311

 

 

 

 

\\

\\\ \\\

0
- «.0 v
q
S
m.
. 0.0 m
o
O
A
. 0.0 0.
v
b
- 0.0
0..

00.0 050:1

88

:oz0EE >030 10.00 00.0015

mt. at. m3. cor m2. our 2... 2... m2. as no em mm on .ON 2. 2. m

r s .

 

 

E0 N_. .::w
\\

‘

rs

 

1.-- e
- sz 0111110111200 E02110 .0

EV E30300 E0510 .1.
1011.12: E11000.

sz E01010qo11010 E1100m
sz 0110901110 E1100m.
000100 .0100. E11010 E1100m
0100 E1100m.

sz 000011 0100

A20 E1111E00 E01511

 

 

o
T Nd V
a.
S
W
n
- 0.0 m
0
o
A
a
- 0.0 Ur
V
b
- dd
d...

 

 

.1. 1.

.0N.N 059m

89

00:03:: 30:00 10.00 00.0015

wt. a: new one mm? our 2... 3... m2. on no co mm on cu m: 2. m 9

71%|».

 

Illuuutﬂlntﬂx“: IIHl-mulll ..
01min? , rmlw

sz E20533 E31101.

sz E03038 E0310 .1.
1011.310 E11000.

A23 E01010000010 E000m
20.000.100.10 E0000,
000.100 .0100. E11010 E1100m
0100 E11000.

g 0000: 10100

g E1110E00 E15101

 

- 06

V
a.
s
m.
n
1.-
a
o
A
a
J
V

b

.. 0.0

- QO

 

 

 

.% \\

o...

00.0 050:

90

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

’g 0.84
E b
If 0'32 T
v ' . '
g 0.80 . a i}: a
g T T e ,.
o 0.78 -
h - , ' -
O 0.76 L 1 3".
2 "
u. 0.74
E 0.72
a.
O / ’;.;,1..,.+ 31.... 4. lav-1...”; $5.13; /
h 0.70 / /
/

2 // ’37 L ‘2'. '. 31,3 F711“ 3-'—' fit; [3w Eff-"£713; -0111“. /
5 O 00 “i1-;1:_:'. 1:"; .' a... '-.. r " '7.‘ .1' '1'1‘ 1;?" f.‘

Shade Sun Shade Sun

12 cm 12 cm 8 cm 8 cm

Category (insolation level, substrate depth)

Figure 2.3. Mean chlorophyll ﬂuorescence with standard errors of 9 species
averaged over two substrate depths (8.0 cm and 12.0 cm) at two solar radiation
intensities (sun versus shade) over the last two growing seasons (2007 - 2008).
Lower case letters denote mean separation by LSD (P5005).

91

Figure 2.4. Absolute cover of 6 Sedum species cultivated at a 10.0 cm substrate
depth at two solar radiation intensities (sun versus shade) over four growing
seasons (2005 - 2008). Symbols represent absolute cover means with standard
errors (n=3). For clarity of graphing, weeks 0 through 120 were plotted at every

other data point.

92

 

 

 

 

 

 

 

coumEﬁ >35 Lots @625

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mt. o2. m2. :2. mm? our 3... 2... M3 on mm 8 mm cm on m_. 3.
4 I4 I 4 4 {llil \DLQILQIG: m WlﬂmlrlqnlrUl Cd
m _* . \n -W
.- ll. 3 - \OI I? \W
2. \ f Nd
\ V
- q
x m.
a II? \ . m
+ W l .W.\\ - .v o m.
-- O
T? . 0
TL... a.
\ ed U;
/ \Eztaqm Ezbom. l IDI l V
nEaxotmT Ezbom. _ l raw
9. EEEESQ 258% I I .dr I I m. o
. anzmcoﬂme Ezbcm. | JDl l
ESE Sabom I OI l.
mtmcm meow 538% i
\ \ \\ 3

.mvN 059“.

93

cor—3:5 33m hots mxoo>>
mt. 02. m3. 69. m2. GNP m: 2... me. on mm co mm on cm 3. 3.

 

 

 

 

 

(ov) Jero ainlosqv

 

 

 

 

 

 

 

32353me Ezbmm | Ibl I
::w szotm: Ezbmm. ,. NI .. E=£m Embom. I OI II
EEBESQ Encom I I 4 I I Eco 258% O
\\ \\ \x. 3

new 2:9”.

94

 

 

 

 

 

 

 

 

 

 

 

,1:
”g 0.35 // o Shade
E 0.301 Q 0 8”"
.- g g
C
3 °'25I i g
5 i
o 0.20% Q éé % g 5
g 0.15- 5 Q 5
.2 5 c}
a I
E 0.10 - § §
0
g 0'05 d O Q Q g 0 g
g 0 00 806 // 00
3 05101107 07101107 09101107 05101108 07101108 09I01l08
>
Date

Figure 2.5. Mean substrate volumetric moisture content with standard errors at
two solar radiation intensities (sun versus shade) over the 2007 and 2008
growing seasons.

95

 

 

0-35 ‘— _ Shade
".5.

0.84 - sun ** n .8.

 

 

 

 

0.82 ‘
0.80 -
0.78 -

0.76 -

 

 

 

   

 

 

Chlorophyll Fluorescence (Fv/Fm)

  

; 2.
mg“

SA SAL SK SP SPUR SR Overall

 

 

Species
Figure 2.6. Mean chlorophyll ﬂuorescence with standard errors of 6 Sedum
species cultivated at 10.0 cm substrate depth at two solar radiation intensities
(sun versus shade) over the last two growing seasons (2007 - 2008) where
SA=Sedum acre; SAL=S. album; SK=S. kamtschaticum; SP=S. pulchellum;
SPUR=S. spurium ‘Coccineum’; SR=S. reﬂexum; OVERALL=aII species
combined. *=Signiﬁcant at P=0.1; **=Signiﬁcant at P=.05; n.s.=not significant.

96

CHAPTER THREE

Carbon Sequestration Potential of Extensive Green Roofs

97

Abstract

The objective of this study was to quantify the carbon storage potential of
extensive green roofs and to evaluate the effect of species selection on carbon
ﬂux. Two studies were conducted. The ﬁrst was performed on eight roofs in
Michigan and four in Maryland, ranging from one year to six years in age. All
twelve Sedum based extensive green roofs ranged from 2.5 cm to 12.7 cm in
substrate depth. Above-ground plant material was harvested in the fall of 2006.
On average, these roofs stored 162 g C-m'2 in above—ground biomass These
ﬁgures are based on the end of the growing season (after ﬂowering) and
therefore should be the maximal biomass of most Sedum species. The second
study was conducted on the roof of the Plant and Soil Sciences Building on the
campus of Michigan State University in East Lansing, MI. Twenty plots were
established on 21 April 2007 with a substrate depth of 6.0 cm. In addition to a
substrate only control treatment, the other plots were sown with a single species
of Sedum (8. acre, 8. album, 8. kamtshaticum, or S. spun’um). Species and
substrate depth represent typical extensive green roofs in the United States.
Carbon analysis was performed by sampling above-ground biomass, below-
ground biomass (roots), and substrate carbon content. Plant material and
substrate were harvested seven times across two growing seasons. Results at
the end of the second year show that above-ground plant material storage varied
by species, with an average of 168 g C-m'2 across species, ranging from 64 to
239 g C-m'z. Below-ground biomass averaged 107 g C-m'z, but ranged from 37

9 Com'2 (8. acre) to 185 g Cwm‘2 (S. kamtschaticum). Substrate carbon content

98

averaged 913 g C-m‘z, with no species effect, sequestering 100 g (3-m'2 beyond
what was in the initial substrate. The entire extensive green roof system

sequestered 375 g C-m'z.

Introduction

Establishing green roofs, or vegetated roofs, mn improve stormwater
management (Carter and Jackson, 2007; Getter et al., 2007; Hilten et al., 2008;
Jarrett and Berghage, 2008), conserve energy (Santamouris et al., 2007; Sailor,
2008), mitigate urban heat island effects (Takebayashi and Moriyama, 2007),
increase longevity of rooﬂng membranes (Kosareo and Ries, 2007), improve
return on investment compared to traditional roofs (Clark et al., 2008), reduce
noise and air pollution (Van Renterghem and Botteldooren, 2008; Yang et al.,
2008), increase urban biodiversity (Baumann, 2006; Brenneisen, 2006), and
provide a more aesthetically pleasing environment (Getter and Rowe, 2006;
Obemdorfer et al., 2007). Green roofs are either ‘intensive’ or ‘extensive’.
Intensive green roofs may include shrubs and trees and appear similar to
landscaping found at natural ground level. As such, they require substrate
depths greater than 15 cm and have ‘intense’ maintenance needs. By contrast,
extensive green roofs consist of herbaceous perennials or annuals, use
shallower media depths (less than 15 cm), and require minimal maintenance.

Due to building weight restrictions and costs, shallow substrate extensive green

99

roofs are more common than deeper intensive roofs and will be the focus of this

study.

Although green roofs are often adopted for energy savings and heat island
mitigation, rarely has this technology been promoted for its ability to mitigate
climate change. By lowering demand for heating and air conditioning use, less
carbon dioxide is released from power plants and furnaces. Sailor (2008)
integrated green roof energy balance into Energy Plus, a building energy
simulation model supported by the US. Department of Energy. His simulations
found 2% reduction in electricity consumption and 9 to 11% in natural gas
consumption. Based on his model of a generic building with a 2000 m2 green
roof, these annual savings ranged from 27.2 to 30.7 GJ of electricity saved and
9.5 to 38.6 GJ of natural gas saved, depending on climate and green roof design.
When considering the national averages of C02 produced for generating
electricity and burning natural gas (US. EPA, 2007; US. EPA, 2008), these
ﬁgures translate to 637 to 719 g C per square meter of green roof in electricity
and 65 to 266 g C per square meter of green roof in natural gas each year.
Another 25% reduction in electricity use may additionally occur due to indirect
heat island reduction achieved from large-scale green roof implementation

throughout an urban area (Akbari and Konopacki, 2005).

Green roofs may also sequester carbon in plants and soils. Photosynthesis

removes carbon dioxide from the atmosphere and stores carbon in plant

100

biomass, a process commonly referred to as terrestrial carbon sequestration.
Carbon is transferred to the substrate via plant litter and exudates. The length of
time that this carbon remains in the soil before decomposition has yet to be
quantiﬁed for green roofs. But, immediately after initial green roof installation, if
net primary production exceeds decomposition, this man-made ecosystem will be

a net carbon sink, at least in the short-term.

However, this ecosystem will not likely sequester large amounts of carbon due
the types of species used and shallow substrate. Many species used on
extensive green roofs exhibit some form of Crassulacean acid metabolism (CAM;
Getter and Rowe, 2006). CAM photosynthesis operates by opening stomata
during the night to uptake CO; and storing it in the form of an organic acid in the
cells vacuoles. During the following daylight period, stomata remain closed while
stored organic acid is decarboxylated back into 002 as the source for the normal
photosynthetic carbon reduction cycle (Cushman, 2001). When operating in
CAM mode, rates for daily carbon assimiliation are 1/2 to 1/3 that of non-CAM

species (Hopkins and Huner, 2004).

The goal of this research was to evaluate the intrinsic carbon storage potential of
extensive green roofs and the effect of species selection on carbon
accumulation. Two studies were conducted in the United States to meet these

objectives.

101

Materials and Methods

Study 1. Above-ground biomass was determined on eight Sedum based
extensive green roofs in Michigan and four in Maryland (Table 3.1). Roofs
ranged from one year to six years in age and from 2.5 cm to 12.7 cm in substrate
depth. Above-ground biomass was sampled in quadruplicate on each roof with a
13.0 cm ring during the Fall of 2006 (see Table 3.1 for speciﬁc dates). Any
above-ground biomass that was within the ring was clipped at substrate level,
placed in paper bags, and dried in an oven at 70 °C for one week. Samples were
then weighed and ground through a 60-mesh stainless steel screen using a
Wiley mill. The material was stored in glass vials in a desiccator prior to carbon
analysis to prevent moisture uptake. Total carbon concentration was determined
by using a Carlo Erba NA1500 Series 2 N/C/S analyzer (CE Instruments, Milan,
Italy). Carbon accumulation was determined by multiplying dry matter weight by
total C concentration. Regression analysis was performed with location of the
roof (Michigan or Maryland) as a categorical independent variable and age of
roof (in months) and substrate depth of the roof (in cm) as independent variables
against grams of carbon per square meter of green roof as the dependent

variable (PROC REG, SAS version 9.1, SAS Institute, Cary, NC).

Study 2. The second study was performed on the roof of the Plant and Soil
Sciences Building on the campus of Michigan State University in East Lansing,
MI. An existing extensive green roof was expanded on 21 April 2007 to include a

study area measuring 2.84 m x 4.6 m with 20 plots, each measuring 0.71 m x

102

0.92 m. The study area was covered with a Xero Flor XF108 drainage layer
(Xero Flor America LLC, Durham, NC) installed over the waterprooﬁng system.
Above the drainage layer, substrate (T able 3.2) was installed to a uniform depth
of 6.0 cm, which represents the physical limitations of this building and many

other roofs in the US.

Each plot was covered with one of four species of Sedum typically used on US.
green roofs; Sedum acre L. (biting stonecrop), Sedum album L. (white
stonecrop), Sedum kamtschaticum var ellacombianum Fisch. (stonecrOp), and
Sedum spurium Bieb. ‘Summer Glory’ (creeping sedum), or a ﬁfth “substrate only’
treatment. Planted plots were sown with 0.65 g of seeds of the treatment
species (Jelitto Staudensamen, GmbH, Schwarmstedt, Germany) mixed with
20.0 g of ﬁne vermiculite (T herm-O-Rock, Inc., New Eagle, PA) and distributed
evenly. The ‘substrate only’ treatment had 20.0 g of ﬁne vermiculite distributed
evenly across it. This arrangement resulted in a randomized complete block
design, with four replicates and ﬁve treatments. To improve seed germination,
the entire study area was covered with shade cloth until 27 June 2007. In
addition, the roof was irrigated for the ﬁrst three months, three times daily for 20

minutes, with 2.8 mm applied at each irrigation cycle. Weeding occurred on a

monthly basis, when needed.

Carbon analysis was performed by sampling above-ground biomass, below-

ground biomass (roots), and substrate carbon content over two growing seasons.

103

Sampling occurred every second month (30 June 2007, 23 August 2007, 17
October 2007, 15 April 2008, 12 June 2008, 15 August 2008, 13 October 2008)
in order to capture the full variability of the green roof ecosystem, especially
since different species exhibit varying growth rates and timing of peak biomass.
To keep track of which portions of a plot had been sampled, a transect was used
to split each plot into 12 equal portions. The portion that was sampled was

randomized and recorded so that no portion was sampled twice.

At each collection time, above-ground biomass was sampled and analyzed as in
Study 1. Below—ground substrate carbon content and below-ground biomass
were also determined at each sampling time. All substrate and below-ground
biomass were removed from the 13.0 cm ring into a plastic bag. The entire bag
was weighed and the substrate was passed through a 4.0 mm sieve. Gravel that
was retained on the sieve was saved and weighed. Roots were removed from
the retained and sieved matter with forceps. Roots were then cleaned with a
phosphate-free dilute detergent followed by a 0.01 mol-L’1 NaEDTA solution for 5
minutes, each cleaning followed by a rinse with deionized water. The cleaned
roots were placed in paper bags and dried for 2 days at 65 °C. Dried biomass
was ground and analyzed for carbon as previously described. Remaining sieved
substrate was mixed and a portion (25.0 g) was removed and oven dried at 105
°C in a small paper bag. Dried substrate and bag weight were subtracted from

original weight to determine moisture content. All substrate material was ground

104

with a roller mill until it was a completely pulverized powder and then analyzed

for carbon as above.

Mean percent carbon and grams of carbon per square meter were analyzed
using an ANOVA model with species as a ﬁxed effect. Signiﬁcant differences
between treatments were determined using multiple comparisons by LSD (PROC

MIXED, SAS version 9.1, SAS Institute, Cary, NC).

Results and Discussion

Study 1 - Aboveground harvest of 12 green roofs. Average above-ground
carbon stored at the time of sampling for the twelve roofs was 162.3 g C-m'2
(Table 3.3). These ﬁgures are based on the end of the growing season (after
ﬂowering) and therefore should be the maximal biomass of most Sedum species.
However, there was a high degree of variability. Carbon sequestered ranged
from 73.2 to 276.5 g C-m'z, inﬂuenced by such factors as age of the green roof

and substrate depth.

Regression analysis (Figure 3.1) showed that there was a weak association
(r2=0.32) between age of the green roof and grams of carbon sequestered,
indicating that older roofs tend to store more above-ground carbon than younger
roofs. However, substrate depth may be a confounding factor, because
substrate depth has been shown to inﬂuence plant growth on a green roof

(Durhman et al., 2007; Getter and Rowe, 2008). This appears to be the case in

105

Table 3.3 where three roofs have a mean substrate depth of 2.5 cm. These roofs
increase in carbon with respect to age. The 15, 39, and 48 month aged roofs

contain 96.8, 143.5, and 196.3 g C-m'z, respectively.

There was no signiﬁcant location (Ml or MD) effect, suggesting that other
variables besides substrate depth and roof age may be impacting carbon
storage. For example, management techniques would likely impact how much
carbon can be sequestered in above-ground material at any given time. Fertilizer
applications or the use of supplemental irrigation may increase plant biomass
since nitrogen and water are often the limiting component to primary production
in many ecosystems (Vitousek and Howarth, 1991). Design choices may also
inﬂuence carbon storage. For example, substrate composition has been shown
to impact plant growth on an extensive green roof (Hunt et al., 2006; Rowe et al.,
2006), which in turn impacts total above-ground carbon. Species selection may
also impact above-ground carbon stores, as species vary in their net primary

production and allocation to biomass (Naeem et al, 1996).

Study 2 — Plant and Substrate Harvested Seven Times over Two Growing
Seasons. Monthly average maximum air temperatures (°C), monthly average
minimum air temperatures (°C), and monthly total precipitation (mm) (not
including initial irrigation), are shown in Figure 3.2. East Lansing, MI is in the
midwestern US. and is characterized as a temperate climate with four well-

defined seasons. Thirty-year average mean low temperatures in January and

106

high temperatures in July are -10.2°C and 275°C, respectively. The average
number of days per year with precipitation greater than 2.54 mm is 70.3, which
are generally well distributed throughout the year (Michigan State Climatologist’s

Ofﬁce, 2008).

Harvest date, species, and the interaction affected total C for above- and below-
ground plant biomass, but not for substrate (Table 3.4). Across all species,
mean carbon on an area basis in above-ground biomass was comparable to the
previous study, averaging 168 g C-m'2 at the end of the second year (Table 3.5).
Sedum album held the greatest amount of carbon, followed by S. kamtschaticum,
S. spurium, and S. acre. Carbon contained in root biomass averaged 107 9 Com
2 at the end of the second growing season (Table 3.5). Root biomass values for
S. kamtschaticum were highest, while values for 8. acre were lowest. The
relatively low allocation to roots may help explain why S. acre seems to be the

least heat and drought tolerant species among those tested.

At the end of the second growing season, substrate carbon content averaged
913 9 Corn’2 (Table 3.5). Unlike the plant material, there were no signiﬁcant
differences among species treatments for substrate carbon content. At
establishment, substrate consisted of 810 g C~rn'2 (Table 3.2). As such, there

was 100 9 Corn'2 sequestered in the soils after two growing seasons.

107

Percent carbon averaged 42.1% C, 41.4% C, and 4.6% C for above-ground
biomass, root biomass, and substrate, respectively (Table 3.5). The percentage
of carbon stored in plant tissues is slightly lower than the expected 45-50% C in
most other vascular plants (Ovington, 1957; Pettersen, 1984; Schlesinger, 1997),
but is similar to other succulent species (Lindsay and French, 2004). Substrate
values correspond well to other ecosystems (Jones et al., 2005; Harte et al.,
2006). Averaged across species, above-ground and root carbon (g-m’z)
increased the ﬁrst year as plants established themselves and then remained
steady throughout the second growing season (Figure 3.3). Substrate carbon

also remained steady over the two growing seasons.

For individual species, ﬁrst year above ground carbon for 8. acre was similar to
the other species, but by the end of the second season 8. acre held the lowest
amount of carbon (Figure 3.4). This may be because the second growing
season did not have supplemental irrigation and S. acne is known to die back in
hot growing conditions (Snodgrass and Snodgrass, 2006). The other three
species also declined through the hottest portion of the second growing season
without this irrigation (July and August 2008), but then rebounded by October
2008. Two species above-ground carbon decreases slightly between the end of
the ﬁrst and the beginning of the second growing season (Figure 3.4). This likely
occurred because these species are either semi-evergreen (S. spun'um) or
deciduous (S. kamtschaticum) and full spring leaf-out had not yet occurred on the

ﬁrst sampling of 2008. In contrast, above-ground carbon values for S. acre and

108

S. album, both evergreen species, are very similar before and after the winter

donnantseason.

Over time, S. kamtschaticum contains the greatest carbon in root biomass
(Figure 3.4). This is likely due to the woody nature of this species’ roots
(personal observation). However, all species generally increased in root carbon
across both growing seasons. Sedum acre allocated the least amount of carbon
to roots. For substrate carbon over time, all species treatments remained steady

throughout the study.

Application. This entire extensive green roof system sequestered 375 g C-m'z,
(168 g C-m'2 in above-ground plant biomass, 107 9 Corn'2 in below-ground plant
biomass, and 100 9 Corn'2 in substrate carbon) beyond what was stored originally
in the initial substrate (Table 3.2; Table 3.5). However, many components of a
green roof have a carbon ‘cost’ in terms of the manufacturing process.
Embodied energy is a term used to describe total energy consumed, or carbon
released, by a product over its life cycle. Typical components of a green roof
include a root barrier installed on top of the normal rooﬁng membrane (which
protects the roof from root penetration damage), a drainage layer above the root
barrier (which allows excess water to ﬂow away from the roof), and a growing
substrate. Many life cycle analysis studies ignore these unique components of a
green roof by making the assumption that the root barrier, drainage layers,

substrate, and plant material will all have a carbon cost similar to the traditional

109

roofs’ gravel ballast (Alcazar and Bass, 2006; Kosareo and Ries, 2007). But, this

assumption may not be valid.

HammOnd and Jones (2008) analyzed building materials through the entire
production process. Material consisting of low density polyethylene (LDPE)
similar to a root barrier was found to average 78.1 MJ-kg‘1 of embodied energy or
1.7 kg COz-kg'1 of embodied carbon. Assuming 0.51 mm thickness and density
of 925 kg-m‘3 (Raven Industries, Sioux Falls, SD), this translates to 219 g C per
square meter of green roof. Drainage layers are commonly made from
polypropylene (Colbond Inc., Enka, NC) which contains 5.0 kg CO; per kg of
product (Hammond and Jones, 2008). Assuming a weight of 0.39 kg-m'2 (Xero
Flor America LLC, Durham, NC), this translates into 535 g C per square meter of
green roof. The embodied energy for a substrate consisting of sand and
expanded slate is 0.005 and 0.44 kg COykg", respectively (GBC, 2008;
Hammond and Jones, 2008). A 6.0 cm substrate depth consisting of half sand
and half expanded slate by volume with densities of 2,240 and 1,600 kg-m'3,
respectively (Expanded Shale, Clay and Slate Institute. 2008; Hammond and
Jones, 2008) translates to 92 and 5769 g C per square meter of green roof.
These green roof components add up to an embodied carbon content of 6615 g

C per square meter of green roof.

But, a traditional roof may also have a gravel ballast, which is no longer needed

on a green roof. Assuming a density of 1800 kg-m'3 and a 2 cm depth, a gravel

110

ballast has an embodied energy content of 0.017 kg COzwkg‘1 or 167 g C per
square meter of roof (Hammond and Jones, 2008). Subtracting out the
unneeded ballast, total adjusted embodied carbon for the extensive green roof
components above and beyond a traditional roof is then 6,448 g C per square

meter of green roof.

While the embodied energy in the initial green roof system is greater than what is
stored in the substrate and plant biomass at any given time, the emissions
avoided due to energy savings will ‘pay’ for those costs in time. Sailor (2008)
integrated green roof energy balance into Energy Plus, a building energy
simulation model supported by the US. Department of Energy.“ His simulations
found 2% reduction in electricity consumption and 9 to 11% in natural gas
consumption. Based on his model of a generic building with a 2,000 m2 green
roof, the minimum annual savings were 27.2 GJ of electricity and 9.5 GJ of
natural gas saved. When considering the green house potential of generating
electricity and burning natural gas (US. EPA, 2007; US. EPA, 2008), these
ﬁgures translate to 702 g C per square meter of green roof in electricity and
natural gas savings combined per year. Since the embodied cost is

6,448 g C-m'2 (from above), it will take nine years to ‘pay’ for the carbon cost of
the green roof materials. After this time, the emissions avoided would simply add
on to the sequestration potential of the roof. The carbon sequestered by growing
biomass (375 g C-m'2 in this study) will shorten the carbon payback period in this

scenario by two years.

111

As typically unused spaces, roofs provide a unique opportunity to sequester
carbon. For example, in the Detroit Metropolitan area, land area of rooftops is
estimated to be 6,335 and 8,399 hectares of commercial and industrial land use,
respectively (Clark et al., 2005). If all of these roofs were covered with
vegetation in a design similar to this study and thus were able to sequester 375 g
C'rn'2 of green roof, 55,252 metric tons of carbon could be sequestered in the
plants and substrates alone (not including avoided emissions). This is similar to
taking over 10,000 mid-sized SUV or trucks off the road for a year (US. EPA,
2005). While these ﬁgures depend on climate and green roof design, they
nonetheless represent a signiﬁcant, if minimal, potential for sequestering carbon

in urban environments.

Conclusions

One time sampling from twelve roofs ranging from one year to six years in age
and from 2.5 cm to 12.7 cm in substrate depth found above ground biomass to
store 162.3 g C-m’z. This corresponds well with two-year multiple sampling
results that averaged 168 g C-m'2 at the end of two growing seasons. In the
latter study, carbon storage varied by species (Sedum album held the greatest
amount of carbon, followed by S. kamtschaticum, S. spun'um, and 8. acre).
Carbon contained in root biomass averaged 107 g C-m'2 and at the end of the
second growing season, with a species effect (S. kamtschaticum were highest,

while S. acre was lowest). Substrate carbon content averaged 913 g Com'z, with

112

no species effect, sequestering 100 g C-m’2 beyond what was in the initial

substrate. The entire extensive green roof system sequestered 375 g C-m'z.

These conclusions are applicable to the midwestern United States and other
geographical areas with similar climates. In addition, management techniques
(such as fertilizer applications or the use of supplemental irrigation) and design
choices (like substrate composition, substrate depth, or species selection) may
inﬂuence these results as well. Future work needs to be performed to quantify

carbon ﬂuxes in the green roof system.

Acknowledgements
Funding for these studies was provided by Ford Motor Company, Dearborn, MI,
XeroFlor America LLC, Durham, NC; Michigan Nursery and Landscape

Association, Okemos, MI; and the Michigan Agricultural Experiment Station.

113

Literature Cited

Akbari, H., and S. Konopacki. 2005. Calculating energy—saving potentials of heat
island reduction strategies. Energy Policy 33(6):721-56.

Alcazar, S. S. and B. Bass. 2006. Life cycle assessment of green roofs - case
Study of an eight-story residential building in madrid and implications for green
roof beneﬁts. In Proc. of 4th North American Green Roof Conference: Greening
Rooftops for Sustainable Communities, Boston, MA. 1142 May 2006. The
Cardinal Group, Toronto.

Baumann, N. 2006. Ground-Nesting Birds on Green Roofs in Switzerland:
Preliminary Observations. Urban Habitats 4:37—50.

Bouyoucos, G. J. 1962. Hydrometer method improved for making particle size
analysis of soils. Agronomy Journal 54:464-465.

Brenneisen, S. 2006. Space for Urban Wildlife: Designing Green Roofs as
Habitats in Switzerland. Urban Habitats 4:27-36.

Carter, T., and C. R. Jackson. 2007. Vegetated roofs for stormwater
management at multiple spatial scales. Landscape and Urban Planning 80:84-
94.

Clark, C., B. Talbot, J. Bulkley, and P. Adriaens. 2005. Optimization of green
roofs for air pollution mitigation. p. 482-494. In Proc. of 3ml North American
Green Roof Conference: Greening Rooftops for Sustainable Communities,
Washington, DC. 4-6 May 2005. The Cardinal Group, Toronto.

Clark, C., P. Adriaens, and F. B. Talbot. 2008. Green Roof Valuation: A
Probabilistic Economic Analysis of Environmental Beneﬁts. Environmental
Science and Technology 42:2155—2161.

Cushman, J.C. 2001. Crassulacean acid metabolism: A plastic photosynthetic
adaptation to arid environments. Plant Physiol. 127:1439-1448.

Durhman, A. K., D. B. Rowe, and C. L. Rugh. 2007. Effect of substrate depth on
initial coverage, and survival of 25 succulent green roof plant taxa. HortScience
42(3):588-595.

Expanded Shale, Clay and Slate Institute. 2008. Lightweight Expanded Shale,

Clay and Slate Aggregate for Geotechnical Applications Information Sheet 6001.
http://www.escsi.org/pdfdoc1/Geotech_lnfo_Sheet_6001 -June_2008.pdf

114

Ferguson, M. H., L. Howard, and M. E. Bloodworth. 1960. Laboratory methods
for evaluation of putting green soil mixtures. USGA Journal and Turf
Management 13(5):30-32.

Gee, G. W., and J. W. Bauder. 1986. Particle-size Analysis. p. 383-411. In: A.
Klute (ed.) Methods of soil analysis: Part 1. Physical and mineralogical methods.
ASA Monograph Number 9.

Getter, K. L., and D. B. Rowe. 2006. The role of extensive green roofs in
sustainable development. HortScience 41(5):1276-1285.

Getter, K. L., D. B. Rowe, and J. A. Andresen. 2007. Quantifying the effect of
slope on extensive green roof stormwater retention. Ecological Engineering
31:225-231.

Getter, K. L., and D. B. Rowe. 2008. Media Depth Inﬂuences Sedum Green
Roof Establishment. Urban Ecosystems 11:361-372.

GBC. Unknown. The Green Building Challenge Handbook of the D-A-C-H clay
brick industry. 16 December 2008. http://www.gbc-ziegelhandbuch.org/default—
eng.htm

Hammond, G. P., and C. I. Jones. 2008. Embodied energy and carbon in
construction materials. Proceedings of Institution of Civil Engineers: Energy
161(2):87 — 98.

Harte, J., S. Saleska, and T. Shih. 2006. Shifts in plant dominance control
carbon-cycle responses to experimental warming and widespread drought.
Environmental Research Letters. 1:014001 (4pp)

Hilten, R. N., T. M. Lawrence, and E. W. Tollner. 2008. Modeling stormwater
runoff from green roofs with HYDRUS-1D. Journal of Hydrology 358:288-293.

Hopkins, W.G., and N.P.A. Hijner. 2004. Introduction to Plant Physiology. 3rd ed.
John Wiley & Sons, New York.

Hunt, W. F., A. M. Hathaway, J. T. Smith, and J. Calabria. 2006. Choosing the
right green roof media for water quality. In Proc. of 4th North American Green
Roof Conference: Greening Rooftops for Sustainable Communities, Boston, MA.
11-12 May 2006. The Cardinal Group, Toronto.

Jarrett, A. R., and R. D. Berghage. 2008. Annual and individual green roof
stormwater response models. In: Proc. of 6th North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Baltimore, MD. 30
April-2 May 2008. The Cardinal Group, Toronto.

115

Jolivet, C., D. Arrouays, and M. Bernoux. 1998. Comparison between analytical
methods for organic carbon and organic matter determination in sandy spodosols
of France. Communications in Soil Science and Plant Analysis. 29:2227-2233.

Jones, R. J. A., R. Hiederer, E. Rusco, and L. Montanarella. 2005. Estimating
organic carbon in the soils of Europe for policy support. European Journal of Soil
Science 56:655-671.

Kosareo, L., and R. Ries. 2007. Comparative environmental life cycle
assessment of green roofs. Building and Environment 42:2606—2613.

Lindsay, EA, and K. French. 2004. Chrysanthemoides monilifera ssp. rotundata
invasion alters decomposition rates in coastal areas of south-eastern Australia.
Forest Ecology and Management. 198:387-399.

Michigan State Climatologist's Ofﬁce. 2008. Michigan Climatological Resources
Program. 2008. 06 June 2008. http://climate.geo.msu.edul.

Naeem, S., K. Hakansson, J. H. Lawton, M. J. Crawley, and L. J. Thompson.
1996. Biodiversity and plant productivity in a model assemblage of plant species.
Oikos 76(2):259-264.

NCR-13. 1998. Recommended chemical soil test procedures for the North
Central Region. North Central Regional Publ. 221. Missouri Agric. Exp. Stn.,
Columbia.

Obemdorfer, E., J. Lundholm, B. Bass, R. R. Coffman, H. Doshi, N. Dunnett,

S. Gafﬁn, M. Kohler, K. K. Y. Liu, and DB. Rowe. 2007. Green Roofs as Urban
Ecosystems: Ecological Structures, Functions, and Services. BioScience
57(10):823-833. '

Ovington, J. D. 1957. The volatile matter, organic carbon and nitrogen contents
of tree species grown in close stands. The New Physiologist 56:1-11.

Pettersen, R. C. 1984. The Chemical composition of wood. In Rowell, Roger
M., ed. The chemistry of solid wood. Advances in chemistry series 207.
Washington, DC: American Chemical Society.

Rowe, D. B., M. A. Monterusso, and C. L. Rugh. 2006. Assessment of heat-
expanded slate and fertility requirements in green roof substrates.
HortTechnology 16(3):471-477.

Sailor, D. J. 2008. A green roof model for building energy simulation programs.
Energy and Buildings 40:1466-1478.

116

Santamouris, M., C. Pavlou, P. Doukas, G. Mihalakakou, A. Synnefa, A.
Hatzibiros, and P. Patargias. 2007. Investigating and analysing the energy and
environmental performance of an experimental green roof system installed in a
nursery school building in Athens, Greece. Energy 32:1781—1788.

Schlesinger, W. H. 1997. Biogeochemistry: An Analysis of Global Change.
Academic Press, San Diego, CA.

Snodgrass, E. and L. Snodgrass. 2006. Green Roof Plants: A Resource and
Planting Guide. Timber Press, Inc., Portland, OR.

Takebayashi, H. and M. Moriyama. 2007. Surface heat budget on green roof
and high reﬂection roof for mitigation of urban heat island. Building and
Environment 42:2971-2979.

US Environmental Protection Agency. Ofﬁce of Transportation and Air Quality.
2005. Emission Facts: Greenhouse Gas Emissions from a Typical Passenger
Vehicle. EPA420-F-05-004. Washington, DC.

US. Environmental Protection Agency. 2007. Inventory of US. Greenhouse Gas
Emissions and Sinks: Fast Facts 1990-2005. Conversion Factors to Energy Units
(Heat Equivalents) Heat Contents and Carbon Content Coefﬁcients of Various
Fuel Types. EPA-430-R-07-002. Washington, DC.

US. Environmental Protection Agency. 2008. Climate Leaders Greenhouse Gas
Inventory Protocol Core Module Guidance: Indirect Emissions from
Purchases/Sales of Electricity and Steam. EPA ~430-K-03-006. Washington, DC.

Van Renterghem, T., and D. Botteldooren. 2008. Numerical evaluation of sound
propagating over green roofs. Journal of Sound and Vibration 317:781-799.

Vitousek, P. M., and R. W. Howarth. 1991. Nitrogen limitation on land and in the
sea: How can it occur? Biogeochemistry 13287-115.

Yang, J., O. Yu, and P. Gong. 2008. Quantifying air pollution removal by green
roofs in Chicago. Atmospheric Environment 42:7266—7273.

117

.Eataqm .m Espamcanmx .m. ESE .w

as. 8:825.

 

”Buses 828% SEE 38. 62m 3288 NE 62 89.238: .5... 352mm 528 88.5 88 >82 >mm
.Eataqm
.m .EmSucmxmw .m .EamoEmEmm .m .Eaxmumc .m .Eaem .m .daw as. cmﬁBmmum. $8.00
Essa. “8.328. 898% begun. .8. Sea 3.225 NE 8? 52m 58m .8 .950 8889201 88:5 Bow :8... cm
.EDEDQW .W .E3xmke .W .EDOBGCQQme .W .E:£m .m _2 .mecwn— «www
.96.... .m H.2232. 868% tween. .32 :85 02228 NE mum .325m 8228 :8 new .52.. :ms. 88:5 38 as. mmme
.Eatzqm .m .Em.=m:mxcw .m .EacmctobcmbbE
.m €22.33.me .m .anqumE .m ..&m mgoqonQ
.UOUEOE mm_omam EmEtn. goo; amen $29563 NE NMN DS— dammmﬁ 59:22 mooth NOON 9.th ”=2
dam ESSE...
.teEccoe .w .EESQm .m .EacmELobzonbE .m, Espamﬁame .m
.Eaﬁm .m ”Own—.35 «3.0QO metn. $00.. :36 $22568 NE th DS— dBEBOO .cozmhoeoo ODS. OOONFE meow “Qmw mum—2
.Eatzqmw .szmtm: Eabcm .EacmELobccbnE .w €39.28.me
588 Ease .m ”Buses 8288 beach. Essa “mom .2 5583 sew .05: :92 88.9.. 88 we; 38E.
Eat: Wm
.52.... .m .98 .m 6828. 888% bases gouge Loom :2 .0523 saw .05: 8.2 Boga 88 as. 85:
.EESQmw
.Eaxocm: Eabcm .EEmELobcebbE .m €38.28..me 5.8%
Ease .m Sum .m Hcauses $.0QO mecn. 5.86... “.03". .5. 6595.. .mmm. dab... 322 885m moom 9.2. womb...
Satan Eznmm. .EacmEgobcmbbE Esncm .Eanﬁ =2 .mEmcm. .mmm goat... .950
256$ .88 253% Bongo... $.0on bests .Eotma “.8”. 29881 new @5289» 93.32.81 :22 mooQQm moom 95... 085.:
.EacmutobzonbE .m. €39.38..me .m EBB .m =2 .Eoﬁmmn.
.Eom .m Hnausea. 8.0on metn. coo. :35 9229.3 NE :56 .255 cmzom .EmquO .222 Eon. moomEm Noom __mn. 010...
93363.8 .m .EEmEBbEbEE .m €3.35».me .m. EBB .m =2 .9653
”causes. $6on begun. goo. :o 2826 cocoa 30... 558cm. 33 .mEESm wt< cozmoEzEEoo Ems. 886$ meow >22 «.20
. 652m
.8353 .m .Emsgmxcm .m .E:=m§ob=obb.§ .m .ESQE .m .mmm. 65.2.3 3?. cozmﬁcaEEomV.
6328. 8.28% tween. .82 8 >896 883 see 5.88m 522. 299...: 93m 595.2 8865 88 as. 096
3:98:30 5330.. Sun. 635... “.00”.
0.9.55 2.2.93
3on—

oﬁu mEaEmm
co meow... Eoncm. .52 :. 838mm mm? 5.500. comm. £02 :36 umaEmm .0 5:98.36 new c0280.. .3” 933

118

Table 3.2. Initial Physical and Chemical Properties of Substrate.

 

 

Component Unit Method
Total Sand 0.79 99'1 Gee and Bauder, 1986
Very Coarse Sand (1-2 mm) 0.13 g-g'1 Gee and Bauder, 1986
Coarse Sand (0.5-1 mm) 0.25 g-g‘1 Gee and Bauder, 1986
Medium Sand (0.2505 mm) 0.29 99" Gee and Bauder, 1986
Fine Sand (0.10-0.25 mm) 0.11 g-g'1 Gee and Bauder, 1986
Very Fine Sand (0.05-0.10 mm) 0.01 g-g'1 Gee and Bauder, 1986
Silt 0.15 g-g'1 Bouyoucos, 1962
Clay 0.06 g-g'1 Bouyoucos, 1962
Bulk Density1 1.17 g-cm’3 Ferguson et al., 1960
Capillary Pore Space 0.34 cms-cm'3 Ferguson et al., 1960
Non-Capillary Pore Space 009 cm3-cm'3 Ferguson et al., 1960
Water Holding Capacity at 0.01 MP3 0.29 cm3-cm'3 Ferguson et al., 1960
pH 8.2 NOR-13, 1998
Conductivity (EC) 10.36 mmno-cm'1 NOR-13, 1993
Organic Matter by LOI @ 360 02 5.6 g-kg" NCR-13, 1998
Organic Carbon3 810 gm'2 Jolivet et al., 1998
Nitrate 2.0 rig-g“ NCR-13, 1993
Phosphorus 53.9 ug-g'1 NOR-13, 1998
Potassium 2174.0 rig-g" NCR-13, 1998
Calcium 203.0 ug-g'1 NOR-13, 1993
Magnesium 57.0 p.g-g'1 NOR-13, 1998
Sodium 439.0 ug-g'1 NOR-13, 1998
Sulfur 434.0 ,ug-g'1 NOR-13, 1993
Boron 1.5 099‘1 NOR-13, 1998
Iron 74.0 ug-g" NOR-13, 1998
Manganese 25.4 ug-g'1 NOR-13, 1998
Zinc 24.5 ug-g‘1 NOR-13, 1998
Copper 3.0 pig-1 NOR-13, 1998

 

 

Analysis per A&L Great Lakes Laboratories, Inc., Ft. Wayne, Indiana
1Including gravel. 2Gravel free. 3Calculated from organic matter LOl per Jolivet et al., 1998.

119

Table 3.3. Mean carbon (g'm'z) :I: standard errors for above-ground biomass on
twelve extensive green roofs.

 

 

Mean Age at Plant
Roof" Substrate Sampling Aboveground

Depth (cm) (months) Carbon mm“)
CA2.5 2.5 15 96.8 :1: 27.9
CA3.2 3.2 15 126.8 :1: 19.0
FORD 2.5 48 196.3 1 64.8
HTRCZ.5 2.5 39 143.5 i 16.0
HTRC5 5.0 39 158.7 t 32.4
HTRC6 6.0 52 224.2 :1: 52.6
HTRC7.5 7.5 39 201.7 :I: 11.1
MDG 7.0 12 73.2 :I: 16.0
MF 7.1 53 188.9 :I: 33.5
PSSB 4.0 28 148.8 1: 26.7
RC 6.4 48 276.5 :t 28.0
SEV 10.8 4 111.8: 30.1

162.3 :1: 11.7

‘See table 1 for descriptions.

120

2:08.00: .25 0.080030. 0 0:0 Satan anom 8:020:00..me 280% 820.0 8380.0 .0000 2.380% 00020:. 00.08% a

.88 cocooo

2 new .88 .803. B .88 82. NF .88 __8< 2 .88 .828 t .88 .882 8 .88 83.. 8 So; 8.828 8.8 ..
9.8.8 8.8 8. 88.9. 88 8 :88 8.8 8 8.0080 x 00.820193

 

wt rd 08 v wooodv Nada m 2.86 5...“ m 300.0025
Emod 3.? m Fooodv 0...? o rooodv ﬁww 0 «00.0020: 0.0m.
ovomd mu; m 33.0 Fwd m 0290 58 m 2005
“In. n. no unn— n. no unn— ..._ ".0 550...; .o 00.3w
80.53:...“ ﬂoom u::o..mo>oa<

 

0030.09 809.0000... 0.0 A.00.00n.0 0cm 300.0020; 0.0m. 030...?
8020000 05 0. .oo. c005 .6 0.0.: 0.0300 .00 5900 8o 0:85 .008: So. 00.00.30. 80.500: >.:o 0.020030 0 0:0
00.00% 52 cos. 80cm 000800 o. Bow ....o<. 0.80000 05265 o3. .0>o .988. c0900 :00... .3 050. <>Oz< .1. 0.00...

121

.N 0.00.... N8.0 m 0.0 .0 00.0050 0.08.0080 ..:08:0._00.00 .< 8:08.008. .80 0.08.0080 08020:. .02..
3.8. 00.0000 0:080 00000.0...0 0.0:00 0:80.00 :. 0.000. 0000.000: .8000. am. .5 0:88.00 :_ 500.0000 :005.

 

 

 

8.8 a 8... 8.8 a 0.8 .03, a 8.8 0.8 a 8.... 8.... a 8. 8.8 a .88 0.8 u 8. 800.2

< 0.8 a N... < 0.80 0 .8 8:0 29.0880

< 8.8.. a 88.. < 0.8 a 8... < .8. n. .8 0 8.8 u. 8.8.. 8 0.8 a 8. < 8.8 « 0.... 9. .8 a 8. 83.88% .0

< .8. a 08.. < 8.8 u 0... < .8 a 88 m 8.8 a 8.8 0 8.8. u 08. < .8 a 0.8.. 8 0.88 H 88 588.8880.me .0

< 8.8 a 88.. < «.8 a 3 < 8.8 u 88 m< .8 u. 8.8.. m< 0. .. a 8 < 8.8 a 8.. m 0.8 n. 88 .550 .0

< 8.8 a 88 < 8.8 a 8... < 8.8. 0 808 < 0.. a 8.8 < «.8 n. 8 < 8.8 a 0.8 < 0.0 u .8 0.88 .0

a. N.:..0 3.. 0 .0. N8.0 3.. 0 .8. N.:_.0 3.. 0 4035.0 3.0080
.80... 0.8.0880 0.03. 05.20 962

 

.__2 .0580. .00m. :_ 800m 500.00 0.. :00000 05306 0:0000
.0 0:0 .0 0.00: :0080 00000 80000. .0 0:0..0..:00:00 :0800 0:0 :0800 .0.0. .0 0.080 0.00:0.0 n." :00... .0...” 0.00...

122

 

500
y = 71.5 + (2.72 x Age in months)

r2=0.32
400 -
.

Carbon (g . m")

 

 

 

 

o l I I I l
0 10 20 30 40 50 60

Age of Green Roof (months)

Figure 3.1. Simple linear regression of above-ground biomass on twelve
extensive green roofs, each sampled four times.

123

 

 

 

 

 

 

 

30 0 Maximum
. o . 0 Minimum . o o
a - ‘ '
I 20 ‘ . O
'5 o ’ o o
E . o o o o
o 10 d O o o
a. 0 o
E o
'2 ’ o o
h o _O O O
._ O O
<
O
O o
-10 l l l l I

 

 

Precipitation (mm)

 

 

Date

Figure 3.2. Monthly average maximum air temperatures (°C), monthly average
minimum air temperatures (°C), and monthly total precipitation (mm) throughout
the study (1 April 2007 to 31 October 2008). Data is from the nearby Michigan
Automated Weather Network’s East Lansing, MI weather station.

124

 

1 200

 

 

 

 

 

 

 

 

 

 

 

 

E\\\\ //,,,f i
NA 800 - ‘i ’
E
9 —O— Above-ground biomass
5 60° ‘ — -O — Root biomass
'E — —v— — Substrate
m
0 400 -
.___—————-G —o——-—“O”‘””G—ﬂ—‘Q
o “—0— ﬂ ’- 7. ﬂ 0 I *r ///l I r T
50 //
/ A} W
40 - /——'-A \ \ / /
aw \ 0
Q, .. , _..
§ 30 —
c
E
«I 20 ~
0
10 ~
v— —————— v— —————— v —V-—-'—-V———“*——~—v
o I I f ﬁ //F T I I
July Aug Sep Oct Apr Jun Aug Oct
2007 2007 2007 2007 2008 2008 2008 2008
Date

Figure 3.3. Carbon content :t standard errors for above-ground biomass, root

biomass,
2008)

and substrate across two growing seasons (April 2007 to October

125

 

Figure 3.4. Carbon content :I: standard errors for above-ground biomass, root
biomass, and substrate across two growing seasons for four species and
substrate only treatment (April 2007 to October 2008).

126

mann—

 

 

 

 

 

 

 

 

mE—ch 82.28 8:28 82.23 3:23. 3?ch 32.28
_ _ _ L\\ _ _ c
- cm
- 2: m
J
a.
- one m
m
- com 3
w
- omw (Z
- can
5.0090300 Illdrll
8.8 m> 8.8 ||>il l\\ c:=o..0o>on< own
mm 60 m> on 60 I O: I
5.00 w>om_oo IIIOIII

 

 

 

.mv.m 2:9“.

127

mumm—
wEFEOF ”233$ @2356 $2235 5233—,

he EEO

 

 

 

_ _ _ _\\

 

030098.00 IILGrll
$100 m> mm 60 I JPl |
Nv .00 m> mm 60 I: O. I

 

 

 

 

CV 60 m> mm 60 l- I \\

 

 

 

5:28
a
- on
$2 m
m.
-8; m
no;
58 m
s w
emu Iv.
, 8..”
$00K
or,»

him 9:9“.

128

8860 88 92 8c

 

 

_

 

 

memnnm

25.23% Eatom
Ezommcoﬂme 259%
E: 2m Ezbmm.

ohm Eznmm

 

Baa
N :33 wch 3< scam “00

L \\

\\

 

 

 

 

 

38 92 Bow :2.
_ _ o
\\
\;\
, com 3
B
J
m
. com u
«m
3
e 02:. W
R
ﬂ8N.
92395
cows

 

 

dim 9:9“.

129

DISSERTATION CONCLUSION

The results presented in this thesis add to the data disseminated from the Green
Roof Research Program at Michigan State University. These chapters
demonstrated the importance of selecting proper species for green roofs, as well
as the impact that species have on carbon sequestration. If the plants fail, either
because of the speciﬁc geographical climate or because of extremes in

microclimate conditions on the rooftop, carbon sequestered will be reduced.

Results from the Sedum plant community development study show the
importance of substrate depth on plant performance, as well as long-term
evaluation of species. Of the substrate depths and species evaluated, substrate
depths of at least 7.0 cm are highly recommended. Species exhibiting the
greatest coverage at all substrate depths were 8. ﬂon‘ferum, S. sexangulare, S.
spurium ‘John Creech’, and S. stefco. In general, species that are less suitable

are S. ‘Angelina’, S. cauticola ‘Lidakense’, S. ewersii, and S. ochroleucum.

The solar radiation and substrate depth study demonstrated the impact of
insolation on roof-top plant growth. Species make-up differed between sun and
shade. For shaded locations, AIIium cemuum, Sedum acre, 8. album, 8.
kamtschaticum, and S. spun'um are excellent choices for all of the depths tested.
Where native species are desired, T. calycinum is also a good choice. For sunny
locations, Sedum acre, 8. album, and T. calycinum are suitable choices.

Species to avoid in Midwestern (or similar) climates include T. parviﬂorum, S.

130

stenopetalum, and S. divergens. Plants that would make good accent species,
but will likely never dominate or ﬁll in gaps include C. ﬂacca, S. reﬂexum, and S.

urvillei.

Results from the carbon study showed how extensive green roofs are able to
sequester carbon. While the average 375 g C-m'2 is not much compared to an
old-growth forest in terms of carbon storage, it still is more than a traditional
barren roof. The substrate stores the majority of the carbon, which is similar to
other ecosystems. However, there was a species effect on both above-ground

and below-ground plant material carbon results.

These conclusions are applicable to the midwestern United States and other
geographical areas with similar climates. Of course, plant growth and carbon
sequestration on any roof depends on climatic factors, such as rainfall
distribution and ambient air temperatures. In addition, management techniques
(such as fertilizer applications-or the use of supplemental irrigation) and design
choices (like substrate composition, substrate depth, or species selection) may

inﬂuence these results as well.

131