Open Journal of Forestry
2012. Vol.2, No.1, 1-8
Published Online January 2012 in SciRes (http://www.SciRP.org/journal/ojf) http://dx.doi.org/10.4236/ojf.2012.21001
Carbon Stock Changes in Soil and Aboveground Biomass from
House Lot Development in King County, Washington, USA
Stephen Por der 1, Deborah Lipson2, Robert Harrison3
1Ecology and Evolutionary Biology, Brown University, Providence, USA
2Center for Environmental Stud ies , Brown University, Providence, US A
3College of Forest Resources, University of Washington, Seattle, USA
Received November 28th, 2011; revised December 28th, 2011; acc epted January 6th, 2012
Fossil fuel burning and deforestation have driven dramatic increases in atmospheric CO2 since the indus-
trial revolution. However, forests in the northern temperate region sequester a substantial (~0.6 Pg·yr–1)
amount of carbon (C), largely through the regrowth of secondary forests that were originally cleared for
timber over one hundred years ago. In the United States, however, some regions are approaching a maxi-
mum regrowth as forests are cleared again, this time for suburban and exurban development. Here we ex-
plore the effects of such development on C stocks in King County, WA, an area with high forest cover but
rapid suburban expansion. We measured soil and biomass C on 18 paired-house/forest lots, and found
house lots stored ~80 Mg·C·ha–1 less soil C, and between 130 and 280 Mg·C·ha–1 less above-ground bio-
mass C than adjacent forest lots. Combining soil C losses with estimates of C emissions from forest
products yields average C emissions of 130 - 280 Mg·C·ha–1, with the majority of losses occurring at the
time of lot conversion. As a comparison, suburban dwellers drive ~30% more than city residents, but this
increase in annual emissions from increased driving is 1% - 2.5% of the losses of C associated with con-
verting forests to house lots. If forestland conversion in the Seattle area continues apace, in the coming
decades C emissions each year from that land-use conversion will equal ~4% of King County’s 2008 C
emissions.
Keywords: Land-Use Change; Carbon; Urban Soils; Emissions; Urban Growth; Development
Introduction
Land-use change and the burning of fossil fuels have dra-
matically increased atmospheric carbon dioxide (CO2) concen-
trations since the industrial revolution, to a level not seen dur-
ing the past 650 thousand years (IPCC 2007). Carbon (C) se-
questration in regrowing forests, particularly in the northern
temperate forest of North America and Europe, has partially
offset these emissions (Pacala et al., 2001; Schimel et al., 2001;
Houghton, 2007). However, as reforestation in some areas reaches
a peak, and suburban and exurban development begins to re-
verse forest regrowth (Wienert 2006; Dwyer et al., 2000), rates
of C sequestration may slow. In the United States, most current
deforestation for suburban development occurs in forests that
were previously cleared, either for agriculture or silviculture.
Since urba n and exurban landscapes account for 1.5 million·km2,
or about 25% of the conterminous US, and have grown at an
average rate of 24,600 km2·yr–1 for the past fifty years (Brown
et al., 2005), the fate of C in secondary forested landscapes
undergoing conversion to housing bears closer examination.
Despite its potential importance, the effects of suburban de-
velopment on soil and biomass C has only recently been as-
sessed in several regions of the US (Pouyat et al., 2002). In two
northeastern temperate cities, Boston and Syracuse, urban soils
contained ~60% less C than was stored belowground pre-evel-
opment, while in Chicago and Oakland soil C was slightly
higher (4% - 6%) in urban soils (Pouyat et al., 2006). In more
arid regions, including Phoenix (Oleson et al., 2006) and the
Colorado Front Range (Ka ye et al., 2 005; Go lubiews ki, 2006 ; Qian
& Follett, 2002), C was higher in residential soils than native
grass or desert soils. Similar results from Larimer County, CO
suggest that surface soils in urban lawns can contain as much as
65% more C than shortgrass steppe soils (Kaye et al., 2005),
which is not surprising give n low organic C in semi-arid systems
and the increased growth on lawns in response to water and
fertilizer. In general, urban area s in arid regions have higher soil
C than surrounding ecosystems and urban areas in wet, forested
regions have lo wer soil C (Pataki et al., 2006). Whether changes
in soil carbon content translates to emissions to the atmosphere
is less well understo od. For example, in Kin g County, WA, large
developments (>4 houses) typically remove all topsoil and ship
it to a topsoil dump, whereas single house developments typi-
cally remove soil from a 60 cm foundation footprint and spread
it around the rest of the lot. The emission C from these soils to
the atmosphere may be quite different.
Similarly, quantifying development-driven C fluxes to the at-
mosphere for above ground biomass (AGB) depends on more
than documenting differences in C stocks pre and post devel-
opment, since the fate of forest products must be considered
when calculating atmospheric C emissions from that land use
change. The fate of woody biomass can vary (lumber, paper,
mulch, fuel, non-harvest residue), and with it the C released to
the atmosphere. Estimates of C losses to the atmosphere from
post-harvest biomass range between 30% - 77% over 90 years
(Harmon et al., 1996; Heath et al., 1996; Perez-Garcia et al.,
2006). Sc al i ng u p q ua n ti fi ed pl ot- l ev el C a ff ec ts is a l so di ff ic ul t

S. PORDER ET AL.
given the spatial patchiness and complexity of urban landscape
mosaics (Kaye et al., 2006).
The goal of this study was to explore the effects of forest to
house lot conversion on soil and biomass C stocks in King County,
Washington, US, an area with high forest cover but rapid sub-
urban expansion. Like many local governments in the US, King
County is try ing to assess the so urces of its gree nhouse gas emis-
sions, and has set the aggressive emission reduction target of
bringing annual emissions 80% below 2007 levels by 2050
(http://your.ki ngcounty.gov/exec/news/2007/pdf/ClimatePlan.pdf).
However, while the county recognizes that deforestation in-
creases emissions and has taken some action to curb urban en-
croachment into forests, the county does not currently include
potential soil and biomass C losses associated with development
in its estimates of emissions. Thus we explored these C losses
from development represented a substantial fraction of regional
emissions given potential projected growth scenarios.
Methods
Site Description
Our 18 study sites were all in or around the City of Issaquah,
Washington, US which covers an area of about 22 km2 in east-
central King County to the east of Seattle and Lake Washington,
in the foothills of the Northern Cascades. The climate is temper-
ate, with an annual precipitation of 150 cm (Brown, 2008). The
region is in the western he mlock zone, and supports a humid co-
niferous forest (Franklin & Dyrness, 1973). The area is largely
forested, with a mixture of private forestland and state forest;
most of the forest was logged in the late 1800s and early 1900s,
and thus is ~100 year old secondary growth (Robbins, 1985).
The population of King County grew from 1.3 million to 1.7
million between 1980 and 2000 and is expected to grow to 2.2
million by 2030 (http://www.wsdot.wa.gov/planning/wtp/data-
library/population/PopGrowthCounty.htm). An estimated 18,000
ha of King County forestland were converted to urban and sub-
urban u se s betwe en 19 79 and 19 89 ( MacLean & B ol singer, 1997)
and 36,400 ha were converted between 1988 and 2004 (Erick-
son & Rogers, 2008). Similarly, some projections suggest that
21% of population growth in King County between 2000 and
2040 (362,000 people) will occur in unincorporated areas, many
of which are for ested (Pu get Sound R egional C ouncil, 2008) .
King County estimates that development of individual lots
for single family residences constituted ~50% of all residential
building permits for unincorporated areas (where much of the
construction on forestland occurs) between 1998-2008 (King
County Department of Development and Environmental Ser-
vices, pers. Comm.). Across development types, fossil-fuel inten-
sive lawnmowers, pesticides and fertilizers are sometimes used
for lawn maintenance, and while this certainly would affect a
lifecycle C analysis of development, we limit our analysis here
to soil and bio mass C loss. We fo cused our study on single ho me
developments, because it was not possible to assess the fate of
excavated topsoil that was transported off site (as is often done
in larger housing developments in the region).
The sites we sampled spanned urban (>386 people·km–2),
suburban (115 - 385 people·km–2), and exurban (19 - 114 peo-
ple·km–2) areas, though the majority were in suburban settings
(http://www.census.gov/population/censusdata/urdef.txt). We se-
lected only sites whose lots had been cleared for the develop-
ment of <5 homes (15 of 18 were cleared as single house lots),
which makes it p robable that soil was not re moved from the pr o-
pe rty during construction. Mo st of the houses wer e in hilly areas,
because we restri cted our analysis to forest-to-ho me conversions ,
and much of the flat land in the region had been previously
cleared for agriculture. The house lots ranged in elevation from
23 m - 330 m a.m.s.l. Th e underlying soi ls were Incepti sols and
Entisols (Alderwood, Everett, Beausite, & Neilton soil series;
Soil Survey Staff, 2008), which are typical of the glacial till-
rich soils of the region.
The house lots ranged from 1 - 88 years since development
(mea n 28 year s). Lot size (i nclu ding un cleare d for est areas) ra nge d
from 590 m2 - 28,000 m2, while the total area cleared ranged
from 400 m2 - 5300 m2, and the house footprint from 130 m2 -
340 m2. The majority of the cleared area at each site was lawn,
but the cleared area also included gardens, planted trees and
shrubs, and lone trees left standing during the clearing process.
Gardens accounted for ~6% of cleared area on the average house
lot. We drew house lot soil C samples from the lawn, and at two
sites we also dre w samples fro m gardens for comparis on. We as-
sumed that the intact forest left in place on the property did not
lose any C during development.
Soil Carbon Determination
At each house lot we took three soil samples from the lawn (at
1, 5, and 10 m perpendicular from the house) and one sample
from an adjacent forest site within 200 m of the lawn. There was
no difference in soil C between lawns and gardens (Appendix 1),
so these were combined in further analyses. We used a hammer
core to a depth of 25 cm to determine bulk density (Grossman
& Reinsch, 2002). Bulk density is notoriously difficult to mea-
sure in rocky soils (Vincent & Chadwick, 1994), and it was not
possible to get home owner permission for the preferred method
of excavating large areas quantitatively. Our use of a hammer
co re certainly introduces erro r to our bulk den sity measurements ,
but that error is equally distributed between our two land use types.
In the hole created by the hammer corer we used a twist auger
to collect three samples to a total depth of 75 cm (25 cm - 42
cm, 42 cm - 59 cm, and 59 cm - 75 cm). Because of the high rock
and gravel content of the soil we were unable to sample to 75 cm
at every location, but we collected at least one core per house,
and one per forest, to a depth of 42 cm and to a depth of 59 cm
for all but four houses and four forests. Since there were no dif-
ferences in either bulk density or soil C in house lot soils as a
function of distance from the house (Appendix 1), we averaged
our house lot samples and considered a paired house lot/forest
for each site.
We assumed, based on ten interviews with contractors, that
60 cm of soil was excavated from the house foundation and spread
ar ound the site (and therefore included in the soil C values for the
lawn). Since the mean house size was only 20% of the mean
cl eared lot size (Appendix 1 ), this assumption is unlikely to sub-
stantially influence our results. We conservatively assumed no
soil C loss from beneath the driveway.
All samples were air-dried at room temperature for a minimum
of 48 hours and sieved through a 2 mm screen. We determined
bulk density of the <2 mm fraction for the 0 cm - 25 cm core,
and assumed bulk density did not change below 25 cm. This as-
sumption underestimates soil losses from house lots, since sur-
face soils were significantly less dense in forests than in house
lo ts (0.77 vs 1.0 g· cm–3, respectively , p = .003; Appendix 1), and
this difference likely gets smaller with depth. A subsample of
each sample was ground in an agate mortar and pestle, dried at
65˚C for at least 48 hours, and analyzed for C concentration on
2

S. PORDER ET AL.
a Carlo Erba NC2100 model C/N analyzer. A second subsam-
ple was dried at 105˚C and all stocks are reported per 105˚C
oven-dried mass. Samples were run in duplicate, and 10% were
run in triplicate. Ninety-five percent of the standards run as un-
knowns were within 5% of their accepted value. All reported
values are mea ns ± 1 S.E. unless othe rwise noted. Statistical com-
parisons between forest and house lots were done in Matlab (Ver-
sion 7.4, Mathworks, Inc.) via a paired t-test or ANOVA after test-
ing that the data did not violate the assumptions of normality.
Carbon in Biomass
We did not measure biomass in forest lots adjacent to house
lots, but instead calculated changes in AGB by comparing AGB
on each house site with the range of forest AGB values reported
for the Pacific Northwest (140 - 290 Mg·C·ha–1; Adams et al.,
2005; Binkley et al., 1992; Hutyra et al., 2011), though more
recent estimates are somewhat lower. We chose to use a bio-
mass range, rather than to measure forest biomass directly for
two reasons: 1) allometric equations developed for tree species
in the region are problematic (Harrison et al., 2009); and 2)
biomass lost at the time of clearing would not be the same as
biomass in adjacent forests today, particularly for older home
sites. Rather than include a non-random error in our assessment
we chose to assess a reasonable range of biomasses that could
give us an idea of the importance of soil relative to biomass C
losses.
Estimates of how much AGB C ends up as atmospheric C
post-harvest also vary considerably, and we chose a range that
encompasses the majority of the literature for these estimates.
Heath et al. (1996) argued that 30% of biomass C from cleared
forests in the United States is emitted as CO2 over 90 years by
decomposition of wood products, and another 35% is emitted
from biomass burning for energy. Because the latter likely re-
places fossil fuels that would have been used anyway we do not
consider this an additional loss of C associated with develop-
ment per se, and use 30% over 90 years as a lower bound of
AGB loss. For an upper bound, we use an estimate of 77% loss
(Harmon et al., 1996). Thus we assume that 40 - 222 Mg·C·ha–1,
less the amount of biomass still on the cleared portion of the
house lot, are lost to the atmosphere in the 90 years after house
development. We use 90 years as a time frame because it is ap-
propriate for this study (the oldest house is ~90 years) and be-
cause both Heath et al. (1996) and Harmon et al. (1996) give
ag gregate emissions estimates based on C fate in forest pro ducts
over this time period, since the collection period of the harvest
data from the Forest Service was 1900-1990.
Biomass on each house lot was calculated by measuring di-
ameter at breast height (DBH) for all trees on the site and using
a generic allometric equat ion to determine biomass (Jenkins et al. ,
2004):


01
Above Ground Biom assExpBBlndbh (1)
where B0 = –2.4800 and B1 = 2.4835; these parameters are
mean values for mixed hardwoods (most house lots contained a
broad mix of native softwoods and non-native planted hardwoods).
While these parameters are meant specifically for trees growing
in a canopied forest, using other c ommon esti mate s for B0 and B1
in this type of system resulted in changes of <5 Mg·C·ha–1. While
some studies have shown that using allometric equations appro-
priate for trees in a canopied forest can overestimate biomass of
urban trees by ~20% (Nowak, 20 04), to our kn owledge this i ssue
has not been explored in the Pacific Northwest, and other author s
have argued that standard allometry may either over or under-
estimate actual urban biomass (McHale et al., 2009). Regard-
less, there was so little biomass (mean 8 Mg·C·ha–1) in house
lots relative to forests (or soils) that the choice of allometric
equation has very little influence on our results. We assumed
AGB was 50% C (Schlesinger, 1997), and omitted both grass
and shrubs under 1 m tall, since they account for <2% of total
biomass in urban and subur ban settings (Golubiewski, 2006; Jo &
McPherson, 1995). Finally, we conservatively assumed no
belowground biomass loss because there was insufficient data to
make a more nuanced estimate.
Results
Soil Carbon
The mean surface (0 cm - 25 cm) soil C concentration was
significantly higher in forests (6.4% ± .78%) than on house lots
(3.8% ± .29%; p = .00005; Figure 1). Similarly, the 25 cm - 75
cm soils were significantly more C rich in forests than lawns
(3.6% ± .65% versus 1.9% ± .20%; p = .006). The mean bulk
densities for the forest and house lots also differed significantly
at .77 ± .24 and 1.0 ± .05 g·cm–3, respectively (p = .003, n = 18).
These data indicate that forest soils stored ~80 Mg more C ha-1
than house lot soils (240 ± 25 vs 160 ± 11 Mg·C·ha–1, respec-
tively, p = .002). There was no significant difference in C con-
centration between the three depths for the 25 cm - 75 cm sam-
ples (p = .71) nor in bulk density or the means of the C conc en-
tration among the three distances from the house (p = .55
and .18 respectively). Finally, there was no correlation between
house age, or lot size, and soil C concentration (p = .4 and .2,
respectively).
Carbon in Biomass
The aboveground C in biomass on the house lots ranged from 0
Mg - 44 Mg, with a mean of 8.0 ± 3.0 Mg·C·ha–1 (Figure 2). Na-
tive trees, left behind on clearing, made up ~40% of the this bio-
mass. There wa s n o co rr ela ti on betwee n age a n d bi omass o n hou se
sites (r2 = .05, p = .38). Forest AGB C estimates drawn from the
literature sugges t that there is 140 - 290 Mg ·C·h a –1 in AGB (A dams
et al., 2005; Binkley et al., 1992; Hutyra et al., 2011). Assuming
between 30% and 77% C emitted from this AGB over 90 years
(Heath et al., 1996; Harmon et al., 1996) after conversion from
forest to house lot, between 40 - 220 Mg·C·ha–1 are likely lost to
the atmosphere over this time. If loss rates are invariant over 90
years, this sugg ests em issio ns of .4 0 - 2.4 Mg ·C·h a–1·yr–1.
Figure 1.
Percent carbon in house lot soils and forest soils. The error bars repre-
sent 1 SE. p-values are from comparisons of forest and house soils at a
given depth.

S. PORDER ET AL.
Figure 2.
Mean carbon stores in soil and aboveground biomass in house and fo-
rest lots. The error bars for the house lots and the forest soil represent 1
SE. The error bar on forest AGB represent the range found in the lite-
rature for this region.
Discussion
Soil Carbon and Biomass Loss
Given the trajectory of suburban encroachment onto forests
or reforested land, our results are important for those calculat-
ing greenhouse gas budgets in the Pacific Northwest. Our data
suggest that a pulse of C is released from the soils of a house
lot during development that does not re-accumulate over time.
Furthermore, the soil C loss associated with lot development is
large (mean 80 Mg·ha–1), roughly 30% - 60% of the aboveground
C loss. This finding is consistent with those of Pouyat et al. (2006)
in cities with climates similar to that of this study, but contrasts
with those of Golubiewski (2006), Kaye (2005) and Qian and
Follett (2002), which found relatively higher C in urban soils in
drier and/or warmer climates. Similarly, the lack of correlation
between age of development and soil C contrasts with the re-
sults of those studies. We hypothesize that in wetter environ-
ments, such as the one in our study, development stimulates a
pulse of decomposition that tapers off as the new land use is in
place. In contrast , poor arid soils accumulate C over time as wa-
ter and nutrients are continually provided. These results support
the review by Pataki et al. (2006) and Pouyat et al. (2009), both
of which suggest that affect of development on soil C varies by
ecosystem. Given the low biomass on house lots, the vast ma-
jority (>90%) of C in house lots is stored in soils (Figure 2).
Similar C distributions have been observed in other urban areas
(Jo & McPherson, 1995).
Assumptions
While understanding the C footprint of development is critical
for planning how best to reduce emissions, there is considerable
uncertainty in determining losses that bear closer examination
here. Perhaps most importantly, we assume that soil C losses
from the site result in C emissions to the atmosphere in the form
of CO2 or that C originally in forest soils was decomposed and
respired upon conversion to house lots. However, some fraction
of house lot soil C that we measure as missing may actually have
been transported off of the site via erosion of topsoil. The fate of
such C i s unclear, and likely depends on processing in river s and
estuaries (Berhe et al., 2007). This may lead us to overestimate
emissions from soil C.
We also assume that bulk density was the same from 25 cm
to 75 cm as it was from 0 cm - 25 cm at each site, which propa-
gates the higher bulk density in house lots to depth. If compac-
tion did not affect the lower house lot horizons, and bulk den-
sity in the lower soil is similar between pairs of sites, we will
overestimate the amount of C in house lots by ~10%.
In addition, the fate of AGB removed from the house lot greatly
influences the magnitude of C lost. There is a dearth of data on
the fate of C in harvested wood products, particularly on how
long wood products used in const ruction take to decompose, and
how that varies with wood type, specific use, and region. In ad-
dition, the time scale of these losses is poorly constrained. We
took 90 y ears of losses as our ben chmark, based on the available
literature and the age of the homes in our study area. However,
if these losses occur more rapidly, or if losses decrease expo-
nentially with time, we may be considerably underestimating
losses from house lot clearing.
Lost Carbon Seq ue stration Pote nt ial
Many of the forestlands into which Seattle is expanding were
last logged in the late 1800s and early 1900s, meaning that most
of these stands are currently around 100 years old (Robbins,
1985). S ta n ds of No rt h we st e ve rg re e n s, su c h a s Douglas-fir and
Western Hemlock, take up to 250 years to mature after being
logged, and some conifers can live up to 700 years (Spies &
Franklin, 1996). Smithwick et al. (2002) found that on average
an additional 338 Mg·C·ha–1 would be stored in the biomass
and soils of coastal Washington and Oregon forests if second
growth stands were allowed to return to their maximum C hold-
ing capacity, and that the upper bound C potential for just the
biomass of these old growth forest is 380 Mg·C·ha–1.
For C losses, and loss rates, we have used a range of between
140 Mg·C·ha–1 and 290 Mg·C·ha–1 to estimate preclearing bio-
mass (Adams et al., 2005; Binkley et al., 1992; Hutrya et al.,
2011). However, the development of housing on regrowing fo-
restland also represents a lost C uptake in the future. Assuming
that the forests reach their full C holding potential after an addi-
tional 150 years (given an average stand age of ~100 years;
Smithwick et al., 2002; Franklin et al., 1986), the lost seques-
tration potential (LSP) is the difference between the biomass at
the time of removal and the assumed biomass of a fully grown
forest. Lost sequestration potential at our sites ranges from 92 -
245 Mg·C·ha–1 developed, or .6 - 1.6 Mg·C·ha–1·yr–1 for the next
150 years. From the perspective of King County’s emissions,
this LSP represents a C sink that is lost by development that
otherwise would have helped King County to reach its net C
emissions reduction goals. Instead of this forested land acting as
a C sink as the forest matures, when it is developed it acts as a
C source as the biomass and soil C is emitted.
Carbon Emissions in Context
Although there are considerable uncertainties in the estimate,
the C loss due to development is substantial, even when LSP or
fossil fuel intensive inputs such as fertilizer and pesticides are
le ft out of the equation. Assumin g lower bound fo rest C and loss
ra tes, 120 ± 31 Mg·C· ha–1 is li berated from soils and AGB within
90 years of conversion, while upper bound estimates suggest a
300 ± 31 Mg·C·ha–1 loss (Figure 2). This loss is comparable to
other major C costs of suburban development (Figure 3). For
example, for every house that is built on a forest lot rather than
4

S. PORDER ET AL.
Figure 3.
Estimates for emissions from house lots over a time period of 90 years
assuming a mean houselot size of .16 ha. Soil, above ground biomass
low and high calculations as described in the text. Additional emissions
from driving calculated assuming suburban households drive 31% more
than urban households.
dense urban infill, there is on average a 31% increase in the num-
ber of miles that that particular household drives (Kahn, 2000).
If the average person in the Puget Sound region drives ~13,500
km·yr–1 (Overby, 2008), a household with two drivers travels an
additional 8000 km yr–1 by building in suburbia compared to a
similar household living in a denser urban area. Assuming an
average car gets the 2004 CAFÉ standard of 11.7 km·l–1 (27.5
miles per gallon) (EPA, 2003) and that emissions are 633
g·C·l–1 (EPA, 2003), given a 99% combustion efficiency, then the
transportation-based C impact of building a single family home
in the forest instead of as urban infill is .43 Mg·C·yr–1 (Figure
3). Thus it would take 30 years for the emissions from in-
creased driving to equal just the soil C emissions from clearing
0.16 ha of forest, the mean area cleared at our sites (Figure 3,
Appendix 1 ). It would take 43 - 107 years for increased tailpipe
emissions to equal emissions from soil and biomass combined
over 90 years, and even longer for driving emissions to equal
soil and biomass emissions if LSP was incl uded.
Implications for King County and Future Directions
Although the recent economic downturn and a new national
awareness of environmental issues could slow Seattle’s growth
into forestlan d, it is likely that suburban encroac hment into fore st
will still occur in the coming decades. Assuming that growth will
ro ughly continue at the 1988- 2004 pace, ~36,000 ha will be con-
verted from forestland to suburbs in the next 15 years in King
County (Erickson & Rogers, 2008). This means that emissions
from soil and biomass C will likely be between 210,000 - 240,000
Mg·C·yr–1 , or ~4% King County’s annual emissions, based on
the county’s 2008 emissions.
Such extrapolation is subject to considerable uncertainty. Not
all of the land which is converted from forest to suburbs will be
deforested and built as single house lots; according to estimates
from the King County Department of Development and Envi-
ronmental Services, single house lots likely represent just over
half of total homes developed in the past decade around Issa-
quah and similar suburb s. In large subdivisions or shopping cen -
ters the treatment of biomass and especially of soils is different,
and therefore the C emissions may vary . Different soil types may
respond to development differently, and because the sites in this
study were located primarily in highland area s, riparian and ot her
particularly soil C-rich areas were excluded. Despite this varia-
tion in both initial conditions and owner decision-making, these
data demonstrate that emissions from house lot development can
be a measureable fraction of total emissions in some regions.
Given the rapid rate of suburban expansion in this country, we
suggest the fate of biomass and soil C from different settings be
a priority of future research across ecosystems.
Acknowledgements
We are grateful to Steven Hamburg and one anonymous re-
viewer for their helpful comments on a previous version of this
manuscript. This work was supported by a Royce Foundation
grant to D.L.
REFERENCES
Adams, A., Harriso n, R., Sletten, R., Strahm, B., Tumblo m, E., & Jen-
sen, C. (2005). Nitrogen-fertilization impacts on carbon sequestration
and flux in managed coastal Douglas-Fir stands of the Pacific North-
west. Forest Ecology and Management, 220, 313-325.
doi:10.1016/j.foreco.2005.08.018
Berhe, A., Harte, J., Harden, J., & Torn, M. (2007). The significance of
the erosion-induced terrestrial carbon sink. Bioscience, 57, 337-346.
doi:10.1641/B570408
Binkley, D., Sollins, P., Bell, R., Sachs, D., & Myrold, D. (1992). Bio-
geochemistry of adjacent Conifer and Alder-Conifer stands. Ecology,
73, 2022-2033. doi:10.2307/1941452
Brown, D. G., Johnson, K. M., Loveland, T. R., & Theobald, D. M.
(2005). Rural land-use trends in the conterminous United States,
1950-2000. Ecological Applications, 15, 1851-1863.
doi:10.1890/03-5220
Brown, T. (2008). Period of record monthly climate summary. In:
NOAA Washington Climate Summary. URL (last checked 20 De-
cember 2008). http://www.wrcc.dri.edu/cgi-bin/cliMAIN.pl?wa7468
Dwyer, J. F., Nowak, D. J., Noble, M. H., & Sisinna, S. (2000). Con-
necting people with ecosystems in the 21st Century: An assessment
of our nation’s urban forests. United States Department of Agricul-
ture Forest Service, PNW-GTR-4990.
Erickson, A., & Rogers, L. (2008) Western Washington land use
change. Rural Technology Initiative.
http://www.ruraltech.org/projects/wwaluc/
EPA (2003). US inventory of greenhouse gas emissions and sinks
1990-2001. Washington DC: Office of Atmospheric Programs, US
Environmental Protection Agency, EPA 430-R-03-004.
Franklin, J. F., & Dyrness, C. T. (1973). Natural vegetation of Oregon
and Washington. United States Department of Agriculture General
Technical Report PNW-8. Portland, OR.
Franklin, J., Hall, F., Laudenslayer, W., Maser, C., Nunan, J., Poppino,
J., Ralph, C. J., & Spies, T. (1986). Old growth definition task group
report: Interim definitions for old-growth Douglas-Fir and Mixed-
Conifer forests in the Pacific Northwest and California. USDA For-
est Service, Pacific Northwest Research Station Research Note
PNW-447.
Golubiewski, N. (2006). Urbanization increases grassland carbon pools:
Effects of landscaping in Colorado’s front range. Ecological Appli-
cations, 16, 555-571.
doi:10.1890/1051-0761(2006)016[0555:UIGCPE]2.0.CO;2
Grossman, R., & Reinsch, T. (2002) Bulk density and linear extension
In J. H. Dane, & G. C. Topp (Eds.), Methods of soil analysis part IV:
Physical methods (p. 1692). Madison, WI: Soil Science Society of
America.

S. PORDER ET AL.
6 Copyright © 2012 SciRes.
Harmon, M., Harmon, J., Ferrell, W., & Brooks, D. (1996). Modeling
carbon stores in Oregon and Washington forest products: 1900-1992.
Climatic Change, 33.
Harrison, R., et al. (2009). Biomass and stand characteristics of highly
productive mixed Douglas Fir and Western Hemlock Plantation in
Coastal Washington. Western Journal of Applied Forestry, 24,
180-286. doi:10.1146/annurev.earth.35.031306.140057
Heath, L., Birdsey, R., Clark, R., & Plantinga, J. (1996) Carbon pools
and flux in US forest products. Forest ecosystems, forest manage-
ment and global carbon cycle (pp. 271-278). New York: Springer-
Verlag.
Houghton, R. (2007). Balancing the global carbon budget. Annual
Review of Earth and Planetary Sciences, 35, 313-347.
Hutyra, L. R., Yoon, B., & Albert, M. (2011). Terrestrial carbon stocks
across a gradient of urbanization: A study of the Seattle, WA region.
Global Change Biology, 17, 783-797.
doi:10.1111/j.1365-2486.2010.02238.x
IPCC (2007). Climate Change 2007: The physical science basis. Con-
tribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge: Cam-
bridge University Press.
Jenkins, J., Chojnacky, D., Heath, L., & Birdsey, R. (2004). Compre-
hensive database of diameter-based biomass regressions for North
American tree species. United States Department of Agriculture
General Technical Report NE-319.
Jo, H., & McPherson, E. (1995). Carbon storage and flux in urban
residential green space. Journal of Environmental Management, 45,
109-133. doi:10.1006/jema.1995.0062
Kahn, M. (2000). The environmental impact of suburbanization. Jour-
nal of Policy Analysis and Management, 17, 569-586.
doi:10.1002/1520-6688(200023)19:4<569::AID-PAM3>3.0.CO;2-P
Kaye, J. P., McCulley, R. L., & Burke, I. C. (2005). Carbon fluxes,
nitrogen cycling, and soil microbial communities in adjacent urban,
native and agricultural ecosystems. Global Change Biology, 11,
575-587. doi:10.1111/j.1365-2486.2005.00921.x
Kaye, J. P., Groffman, P., Grimm, N. B., Baker, L., & Pouyat, R.
(2006). A distinct urban biogeochemistry? Trends in Ecology and
Evolution, 21, 192-199. doi:10.1016/j.tree.2005.12.006
MacLean, C., & Bolsinger, C. L. (1997). Urban expansion in the for-
ests of the Puget Sound Region. Resource Bulletin PNW-RB-225.
Portland: USDA Forest Service.
McHale, M., Burke, I., Lefsky, M., Peper, P., & McPherson, E. (2009).
Urban forest biomass estimates: Is it important to use allometric rela-
tionships developed specifically for urban trees? Urban Ecosystems,
12, 95-113. doi:10.1007/s11252-009-0081-3
Nowak, D. J. (1994). Atmospheric carbon dioxide reduction by Chi-
cago’s urban forest. In: E. G. McPherson, D. J. Nowak, & R. A.
Rowntree (Eds.), Chicago’s urban forest ecosystem: Results of the
Chicago Urban Forest Climate Project. General Technical Report
NE-186. US Department of Agriculture, Forest Service: 83-94.
Oleson, J., Hope, D., Gries, C., & Kaye, J. P. (2006). A Baysian ap-
proach to estimating regression coefficients for soil properties in
land-use patches with varying degrees of spatial variation. Environ-
metrics, 17, 517-525. doi:10.1002/env.789
Overby, K. (2008). Puget sound trends: Trends in vehicle mi l es traveled.
URL (last checked 31 January 2009).
www.psrc.org/publications/pubs/trends/t2sep08.pdf.
Perez-Garcia, J., Lippke, B., Comnick, J., & Manriquez, C. (2006). An
assesment of carbon pools, storage and wood products market sub-
stitution using life-cycle analysis results. Wood Fiber Science, 37,
140-148.
Pacala, S., et al. (2001). Consistent land- and atmosphere-based US
carbon sink estimates. Science, 292, 2316-2320.
doi:10.1126/science.1057320
Pataki, D., Alig, R., Fung, A., Goliubski, E., Kennedy, C., McPherson,
E., Nowak, K., Pouyat, R., & Pomero Lankao, P. (2006). Urban eco-
systems and the North American carbon cycle. Global Change Biol-
ogy, 12, 2092-2102. doi:10.1111/j.1365-2486.2006.01242.x
Pouyat, R., Groffman, P., Yesilonis, I., & Hernandez, L. (2002). Soil
carbon pools and fluxes in urban ecosystems. Environmental Pollu-
tion, 116, 107-1 18. doi:10.1016/S0269-7491(01)00263-9
Pouyat, R., Yesilonis, I., & Nowak, D. (2006). Carbon storage by urban
soils in the United States. Journal of Environmental Quality, 35,
1566-1575. doi:10.2134/jeq2005.0215
Pouyat, R., Yesilonis, I., & Golubiewski, N. (2009). A comparison of
soil organic carbon stocks between residential turf grass and native
soil. Urban Ecosystems, 12, 45-62. doi:10.1007/s11252-008-0059-6
Puget Sound Regional Council (2008). Vision 2040. In: Puget Sound
Regional Council Documents. URL (last checked 18 September
2008).
http://psrc.org/projects/vision/pubs/vision2040/index.htm
Qian, Y., & Follet, R. F. (2002). Assessing soil carbon sequestration in
turfgrass systems using long-term soil testing data. Agronomy, 94,
930-935. doi:10.2134/agronj2002.0930
Robbins, W. (1985). The social context of forestry: The Pacific North-
west in the 20th Century. The Western History Quarterly, 16,
413-427. doi:10.2307/968606
Schimel, D. S., House, J. I., Hibbard, K. A., Bousquet, P., Ciais, P.,
Peylin, P., Braswell, B. H., Apps, M. J., Baker, D., Bondeau, A.,
Canadell, J., Churkina, G., Cramer, W., Denning, A. S., Field, C. B.,
Friedlingstein, P., Goodale, C., Heimann, M., Houghton, R. A.,
Melillo, J. M., Moore III, B., Murdiyarso, D., Noble, I., Pacala, S. W.,
Prentice, I. C., Raupach, M. R., Rayner, P. J., Scholes, R. J., Steffen,
W. K., & Wirth, C. (2001). Recent patterns and mechanisms of car-
bon exchange by terre st ri al ec osys te ms. Nature, 414, 169-172.
doi:10.1038/35102500
Schlesinger, W. (1997). Biogeochemistry: An analysis of global change
(2nd ed.). San Diego, CA: Academic Press.
Smithwick, E., Harmon, M., Remillard, S., Acker, S., & Franklin, J.
(2002) Potential upper bounds of carbon stores in forests of the Pa-
cific Northwest. Journal of Appli ed Ecology, 12, 1303-1317.
doi:10.1890/1051-0761(2002)012[1303:PUBOCS]2.0.CO;2
Soil Survey Staff (2008). Soil survey of King County, Washington. In:
Natural Resources Conservation Service, United States Department
of Agriculture Soil Sur vey. URL (last checked 10 December 2008).
http://soildatamart.nrcs.usda.gov/Survey.aspx?State=WA
Spies, T. A., & Franklin, J. F. (1996). The diversity and maintenance of
old-growth forests. In R. C. Szaro, & D. W. Johnson (Eds.), Biodi-
versity in managed landscapes: Theory and practice (pp. 296-314).
Oxford, New York.
Vincent, K., & Chadwick, O. (1994). Synthesizing bulk density for
soils with abundant rock fragments. Soil Science Society of America
Journal, 58, 455-4 64.
doi:10.2136/sssaj1994.03615995005800020030x
Wienert, A. (2006). From forestland to house lot: Carbon stock
changes and greenhouse gas emissions from exurban land develop-
ment in central New Hampshire. Masters Thesis, Providence: Brown
University.

S. PORDER ET AL.
Appendix 1. Soil Carbon, Bulk Density, Above Ground Biomass (AGB) and Selected Site Attribute
Measurements.
HOUSE
Site Distance from
house m BD g·cm–3 wt%·C 0 - 25 cm wt%·C 25 - 75 cmLawn Soil C 0 - 7 5 cm Mg·ha–1 Cleared Area ha*House Area ha Total Soil C Mg·ha–1 AGB-C Mg·ha–1
1 1 1.0 1.4 1.2 100
1 5 .88 3.3 1.7 150
1 10 .70 2.7 1.0 83
Site 1 mean .87 2.5 1.3 110 .053 .028 83 6
2 1 1.4 1.9 .79 120
2 5 1.1 .87 .74 63
2 10 1.1 4.2 1.4 180
Site 2 mean 1.7 2.3 1.0 190 .082 .013 170 20
3 1 1.0 2.1 2.6 190
3 5 1.2 2.9
3 10 .71 1.7 1.4 78
Site 3 mean 1.5 2.2 2.0 230 .042 .025 150 .5
4 1 .90 3.0 2.8 190
4 5 .93 3.3
4 10 1.0 3.0 2.6 210
Site 4 mean 1.0 3.1 2.7 200 .20 .070 160 .09
5 1 1.0 2.7 .84 110
5 5 1.0 2.9 1.0 120
Site 5 mean 1.0 2.8 1.5 140 .061 .018 120 10
6 1 1.3 2.9
6 5 1.0 2.8 .34 86
6 10 .83 2.8 .37 73
Site 6 mean 1.0 2.8 .36 91 .021 .022 59 30
7 1 1.0 6.1 1.4 230
7 5 1.0 5.6 1.3 220
7 10 1.0 7.2 2.2 300
Site 7 mean 1.0 6.3 1.7 250 .018 .018 170 .3
8 1 .77 3.5 .63 92
8 5 1.4 4.6
8 10 .84 5.4 1.3 170
Site 8 mean 1.0 4.5 1.0 160 .035 .022 121 40
9 1 1.3 3.9
9 5 1.3
9 10 1.0 3.4 1.4 160
Site 9 mean 1.2 3.7 1.4 190 .20 .053 170 8**
10 1 1.0 5.0 3.1 290
10 5 1.3 5.4
10 10 .89 4.5
Site 10 mean 1.1 5.0 3.1 300 .070 .018 240 8**
11 1 .92 3.8
11 5 1.0 3.2 1.6 150
11 10 .95 5.6 6.4 440
Site 11 mean .95 4.2 4.0 290 .073 .021 240 5
12 1 .67 6.8 3.5 230
12 5 .58 7.6 2.1 170
12 10 .67 4.8 2.3 160
Site 12 mean .64 6.4 2.6 190 .021 .015 140 .1
13 1 .92 2.8 1.8 140
13 5 .80 4.5

S. PORDER ET AL.
Continued
13 10 1.0 4.4
Site 13 mean .90 3.9 1.8 170 .20 .019 150 .1
14 1 1.1 3.8 2.4 240
14 5 1.0 3.2
14 10 1.3 4.2
Site 14 mean 1.1 3.7 2.4 250 .10 .016 220 1
15 1 1.0 2.6 2.0 170
15 5 .93 3.0 3.3 230
15 10 .67 2.9 1.3 91
Site 15 mean .88 2.8 2.2 160 .037 .029 110 4
16 1 1.0 2.8 1.5 140
16 5 1.1 3.1
16 10 1.0 4.2 1.4 170
Site 16 mean 1.0 3.4 1.5 160 .34 .034 150 10
17 1 1.0 4.3 1.8 400
17 5 1.4 3.8
17 10 .87 4.0 1.0 130
Site 17 mean 1.1 4.1 1.4 190 .34 .030 180 1
18 1 1.0 3.6 .79 130
18 5 1.1 4.2
18 10 1.0 3.6 2.1 2.0*102
Site 18 mean 1.0 3.8 1.4 170 .48 .021 170 .8
MEAN 1.0 3.8 1.9 190 .13 .026 160 8
S.E. .050 .29 .20 13 .03 .003 11 3.0
*Cleared area values do not include the house footprint; they show total cleared area less the house footprint. Total cleared area = cleared area + house footprint. **No
data-value a ssumed to be the mean of the dataset.
FOREST
Site Bulk density g·cm–3 wt% C 0 - 25 cm wt% C 25 - 75 cm Total Soil C Mg·ha–1
1 .63 3.8 2.5 140
2 .48 8.8 2.8 170
3 1.0 2.0 .86 98
4 .70 5.4 1.4 140
5 1.0 3.0 2.5 2.0*102
6 .46 5.3 3.5 140
7 .77 4.5 7.9 390
8 .67 11 3.0 290
9 .53 11 5.5 290
10 .39 12 2.5 160
11 .47 11 12.4 420
12 1.0 9.6 3.7 420
13 .93 5.1 2.3 230
14 1.1 1.5 .94 92
15 .86 5.6 2.6 230
16 1.1 3.7 3.8 320
17 .91 8.5 3.2 340
18 .91 5.0 2.7 240
MEAN .77 6.4 3.6 240
Standard Error .24 .78 .65 25
8