e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e c o l e n g

Biomass production of hybrid poplar (Populus sp.) grown on
deep-trenched municipal biosolids

Erika Felix

a

, David R. Tilley

b

,

∗

, Gary Felton

b

, Eric Flamino

c

a

Department of Biological Resources Engineering, University of Maryland, College Park, United States

b

Department of Environmental Science and Technology, University of Maryland, College Park, United States

c

ERCO Inc., Brandywine, MD, United States

a r t i c l e

i n f o

Article history:
Received 22 September 2006
Received in revised form
1 October 2007
Accepted 12 October 2007

Keywords:
Biosolids
Deep row application
Hybrid poplar
Allometric models
Wood production
Waste recycling
Bioenergy

a b s t r a c t

Environmentally sustainable options for recycling municipal biosolids generated from

wastewater treatment plants are needed. We measured the wood biomass production of

a 50 ha hybrid poplar (Populus spp.) tree farm where biosolids were applied at a rate of

380 metric tonnes per hectare in deep trench rows (46 cm deep, 107 cm wide, and 20 cm below

the surface) of a clay spoil at a former sand and gravel mine located near Washington, DC.

Our aim was to quantify how much wood biomass was produced within a 6-year rotation

and to develop allometric models useful for easily estimating wood biomass for this type of

novel plantation. We randomly sampled trees from stands 2 to 6 years old to measure total

tree height, diameter at breast height (DBH), and total tree weight. The 6-year-old trees had

a mean height of 9.7 m and a mean biomass of 20.5 kg. DBH was the best allometric predictor

of biomass (r

2

= 0.98, P < 0.001), especially for trees with diameters greater than 4 cm. Tree

height was a significant, but less precise predictor of biomass (r

2

= 0.87, P < 0.001). Standing

wood biomass after 6 years was 22,100 kg/ha. Ecological recycling of municipal biosolids to

tree plantations can be an environmentally sustainable and energy conscientious means

for producing energy, while restoring degraded habitat.

© 2007 Elsevier B.V. All rights reserved.

1.

Introduction

Municipal wastewater treatment plants produce large
amounts of biosolids that must either be disposed in sanitary
landfills or recycled to land, like pastures and cropland. While
the former is unsustainable and wastes a valuable resource,
the latter can cause odor problems, especially when near
residential communities, or consume too much fuel during
long-distance hauling from urban to rural areas. For example,
in Washington, DC, biosolids are truck-hauled over 300 km
to rural southern Virginia, which consumes increasingly
expensive diesel fuel.

∗

Corresponding author. Tel.: [phone].

E-mail address:

[e-mail]

(D.R. Tilley).

URL: http://www.nrmt.umd.edu/tilley.htm (D.R. Tilley).

A treatment option that avoids these pitfalls is to bury

the biosolids in underground trenches on marginal lands to
support a plantation of fast growing trees. A sand and gravel
mining company (ERCO, Inc. of Brandywine, MD), located near
Washington, DC, operates such a plantation at one of their
former mines. They buried biosolids from the Blue Plains
wastewater treatment plant (Washington Sanitary Suburban
Commission, Washington, DC) in underground trenches of
a heavy clay spoil and planted rows of fast-growing hybrid
poplar (Populus spp.) trees on a 6 years rotation. The trees
were subsequently harvested and chipped for the local plant
nursery industry (

Felton et al., 2005; Buswell et al., 2006

). At

0925-8574/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:

10.1016/j.ecoleng.2007.10.009

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

9

today’s historically high energy prices there is also interest in
providing the produced wood to homeowners or businesses
with wood boilers. In the longer-term, the produced wood may
serve as a source of lignocellulose for manufacturing ethanol
as a transportation fuel (

Fleck, 2006

) or as feedstock for wood-

fired electric power plants. In addition, the close proximity
of the site to the wastewater treatment plant will help reduce
transportation costs. Finally, this recycling approach promotes
the tenets of ecological engineering by reducing the consump-
tion of non-renewable fuels, relying mostly on free renewable
energy and restoring ecosystem services on degraded land, all
while treating an urban waste problem.

Hybrid poplars were chosen for the plantation because they

are among the fastest-growing trees in North America pro-
ducing between 8000 and 22,000 kg-dry weight of wood per
hectare per year (kg/(ha year)) and achieving heights of 20 m
when grown under short-rotation silvicultural practices (

Van

Ham et al., 2000

).

Taylor (2002)

described the plant breeding,

physiology, biochemistry, and molecular biology of Populus and
suggested it as a model forestry tree for understanding the
unique processes of woody plants.

Le Roux et al. (2001)

eval-

uated 27 tree growth models, of which two were based on
Populus species. The models were mechanistic portrayals of
tree growth represented by differential or partial differential
equations that simulated several tree growth state variables,
such as tree structure, photosynthesis and photosynthate dis-
tribution, respiration, and energy reserve dynamics. The two
models developed based on Populus performed simulations on
an hourly or daily time step and were able to represent the tree
response to the environment. Tree models have been used
to study the likely response of forest communities to global
warming, specifically elevated carbon dioxide, nitrogen, and
temperature (

Hyvonen et al., 2007; Belote et al., 2003; Dreyer

et al., 2001

).

Ceulemans (1996)

stated that a benefit of process mod-

els over empirical models was that process models could be
used to make predictions outside the range of parameters
used for calibration.

Ceulemans (1996)

reviewed 10 models

that were predominantly process-based and categorized these
models as diagnostic, empirical, or mechanistic. Input, output,
and main assumptions were presented. He observed that the
below-ground compartment and below-ground interactions
were not well understood and poorly modeled. The poplar tree
farm that is the subject of our investigation violates two basic
assumptions common to most of these type of process models.
First, our site does not have a conventional soil profile where
the majority of organic matter and nutrients accumulate near
the surface. Rather, our soil profile is inverted because most of
the nutrients are buried in a trench, well below the surface
and out of reach of shallow roots. Second, our site experi-
ences heavy herbivory from deer periodically, which is not a
component of most models. Therefore, we pursued calibrating
simple, univariate empirical models to predict wood accumu-
lation and tree growth, rather than parameterizing a process
model that did not fit the necessary assumptions.

Ceulemans et al. (1995)

modeled height of Populus as a sig-

moid growth function during early development between days
145 and 300. A model by

Das and Chaturvedi (2005)

used above-

ground biomass as a predictor of diameter at breast height
(DBH). Essentially, non-linear regression was used to fit data to

several biomass parameters (bole, branch, twig, aboveground
biomass, etc.) as predictors of DBH (cm). Solving their equation
for aboveground biomass yielded:

Biomass

=

1

0

.120

2

.956 − ln

173

DBH

− 1

As a predictor of biomass (kg/tree), this equation predicts

a negative biomass for trees with a DBH less than 8 cm,
which would not be useful for our trees because our largest
DBH’s were 12 cm. Their model does demonstrate that a sin-
gle parameter is useful for predicting tree biomass. Their
observed data indicated that a 9-year rotation developed a
standing biomass of 90,600 kg/ha.

Norby et al. (2001)

stated that “young trees undergo a period

of exponential growth

. . ..”, but their temporal graph of basal

area was much closer to two straight-line segments with
the first much steeper than the second. The change in slope
occurred approximately at year four.

Norby et al. (2001)

pre-

sented models of aboveground biomass as variations of linear
relationships:

DM

= −2.24 + 0.355A

1

.3

where DM = mass of aboveground dry matter (kg) and,
A

1.3

= cross-sectional area at 1.3 m height (cm

2

) and

DM

= 30.25 + 0.174(A

1

.3

)

H − 56.81T

where DM = mass of aboveground dry matter (kg), A

1.3

= cross-

sectional area at 1.3 m height (cm

2

), H = tree height (m),

T = taper index = 1

− A

4

/A

1.3

and, A

4

= cross-sectional area at

4 m height (cm

2

).

Water deficit is one of the most common external factors

that have an impact on biomass production. Hybrid poplars
are among the fastest growing trees in temperate latitudes
but this productivity is coupled to high water requirements
(

Monclus et al., 2006

). In their work, twenty-nine 3.5-year-old

Populus deltoides

× Populus nigra genotypes were subjected to

a water deficit for approximately 2.5 months. This moderate
water deficit lead to a significant decrease in biomass produc-
tion for most genotypes. The most productive genotypes were
sensitive to drought while less responsive genotypes exhib-
ited a wide range of drought sensitivity. In general, drought
reduced aboveground biomass production but also induced
biomass accumulation in the root systems (

Yanbao et al.,

2006

).

Our aim was to determine the net wood accumulation

and productivity of a humid temperate forest plantation that
received deep-trenched biosolids and to develop allometric
equations for estimating tree wood biomass specific to this
type of forest. To assess the potential for this type of eco-
agricultural forest plantation to provide wood as a feedstock
for an industry, knowledge of wood productivity and mod-
els for monitoring wood accumulation on the plantation are
needed.

2.

Materials and methods

2.1.

Site description

The study site was located within Prince Georges County,
Maryland, in the Washington, DC metropolitan area (

Fig. 1

).

10

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

Fig. 1 – Location of hybrid poplar tree farm (star) within the metropolitan Washington, DC, area.

About 6 m of sand and gravel was mined from the site between
1972 and 1983, which left behind an earth substrate consisting
predominantly of clay (

Kays, 2002

). Site morphology consisted

of a plateau where the hybrid poplar plantation was located.
Steep banks surrounding the plateau were characterized by
incised streams and unplanted mixed hardwood forest. The
edge of the plateau was bermed to divert runoff to one of four
detention ponds. Prior to biosolid application, the reclamation
site was representative of abandoned sand and gravel mines
in the metropolitan area. Surface hydrology was significantly
altered by the mining.

The spoil layer consisted mostly of clay with occasional

remnants of sand and gravel. The clay layer was 5.0–21.3 m
thick and overlaid the lower Miocene Calvert Formation, which
was a light to medium, olive gray to olive green, micaceous,
clayey silt formed from marine shelf deposition (

Wilson and

Fleck, 1990; Tompkins, 1983

).

2.2.

Site treatment

Prior to the tree planting rotation for our study the entire site
had undergone one complete cycle of deep-trenched biosolids
application, 6-year tree growth and complete clear cut. Land
was prepared by excavating trenches 76 cm deep and 107 cm
wide, which were spaced on 244 cm centers (

Fig. 2

). The

trenches were filled with biosolids at a rate of 383,000 kg-dry
weight per hectare. Filling occurred within 45 min of biosolids
delivery, trenches were covered with 20–30 cm of overburden
and the site was leveled using a low-ground pressure bulldozer
and disked, in preparation for tree planting. The application
rate used was similar to experimental trenching conducted by

Sikora et al. (1982)

on a well-drained, silt loam. The biosolids

contained approximately 24% solids, 1.15% total nitrogen, 72%
moisture and had a mean pH of 12 (

Buswell et al., 2006

). The

high pH was due to lime stabilization conducted at the treat-
ment plant.

The biosolids remained in a fairly stable anaerobic envi-

ronment for the months leading up to the spring planting. All
cuttings were from hybrid poplar clones (

Felton et al., 2005

).

Tree plantation management was ‘low-intensity’; i.e., there

was no irrigation, no supplemental fertilization, no pesticides,
and no herbicide used. However, the understory was mowed
for the first 2 years of the rotation to suppress plant competi-
tion. The applied biosolids were the main source of nutrients
for the trees.

2.3.

Experimental design

Tree stands ranged in age from 2 to 6 years. In the fall of
2005, samples were randomly selected from stands planted
in years 1999, 2000, 2001, 2002 and 2003. At each stand,
five trees were randomly selected for sampling, except for
the 2001 stand where a total of nine samples were selected
due to the larger area it covered and only three samples
were collected from the 2003 stand. Trees were randomly
selected by generating random numbers for both its row
and column location in the plantation grid. If the randomly
selected tree was missing or dead, it was excluded from mea-
surements and a new replacement sample was randomly
identified.

Fig. 2 – Excavated trench (foreground) in preparation for
biosolids (background).

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

11

2.4.

Data collection

The diameter at breast height (DBH) of each sample tree
was measured with a standard DBH tape at the end of the
growing season in September 2005. Each sample tree was
felled by cutting it at the ground surface with a chainsaw in
October 2005. The height (HT) of felled trees was measured
using a 50-m tape immediately following cutting. In addi-
tion, the green weight of the entire tree (GW

tree

) was obtained

by suspending the felled tree from a load cell (National
Scale Technology, Huntsville, AL), which was suspended from
the bucket of the bulldozer. Two- to three-centimeter-thick
disks were then cut from the sample tree every 1.6 m from
the bottom, inclusive of the bottom. A minimum of three
disks per tree were collected. The green weight (GW) of
each disk was taken at the site with a scale. The disks
were transported to a lab where they were oven dried at
70

◦

C until reaching a constant weight. Drying required from

3 to 5 days. The percent moisture (mass basis) content of
each disk (MC

disk

) was calculated using the following equa-

tion:

MC

disk

=

GW

disk

− DW

disk

DW

disk

× 100

(1)

where GW

disk

= green weight of tree sample disk (grams)

and

DW

disk

= oven-dried

weight

of

tree

sample

disk

(grams).

Sample tree moisture content (MC

tree

) was determined by

taking the average of MC

disk

, which was then used in Eq.

(2)

to

determine a tree’s aboveground biomass (BA

tree

).

BA

tree

=

GW

tree

1

+ MC

tree

/100

(2)

where BA

tree

was in grams dry weight, GW

tree

was in grams

green weight of sample tree, and MC

tree

is the estimated per-

cent moisture content of sample tree.

Standing aboveground biomass of each stand-age (BA

stand

)

was determined from the following equation

(3)

:

BA

stand

= s × BA

tree

−mean

(3)

where s was planted stem density (stems per hectare), which
was known from planting arrangement (1075 trees/ha), and
BA

tree-mean

was mean aboveground biomass of sample trees

per stand (grams per tree).

2.5.

Data analysis

The mean net production of aboveground tree biomass was
found by fitting a straight line to BA

stand

as a function of

stand-age. The best-fit line was found using simple linear least
squares regression.

Allometric relationships predictive of BA

tree

were deter-

mined for DBH and HT using simple linear regression for
each metric. Stepwise linear regression was used to explore
whether an allometric equation that combined both DBH
and HT was a better predictor. However, strong correlation
between these independent variables (i.e., DBH and HT) was
found, which precluded us from using them in combination

Fig. 3 – Hybrid poplar standing biomass after 6 years of
growth on trenched municipal biosolids.

as independent variables to avoid multicollinearity problems.
SPSS for Windows 10.0 (Chicago, IL) was used for all statistical
analysis.

3.

Results and discussion

3.1.

Biomass

After 6 years of growth on trenched biosolids the hybrid poplar
plantation appeared healthy with a developing canopy and full
understory (

Fig. 3

).

As shown in

Table 1

the mean tree dry weight increased

with stand age, except for the 2001 planting, which suffered
from heavy deer grazing and drought conditions during its first
year of growth. Mean tree height and DBH also increased with
stand age, with the 2001 planting again the exception. After
6 years of growth, trees weighed an average of 20.5 kg, were
974 cm tall and had diameters of 11.7 cm (

Table 1

).

3.2.

Allometric models

Field inventory of the wood biomass of a hybrid poplar planta-
tion, such as the one studied here, requires allometric models
that are precise with easy to measure field variables. We tested
the efficacy of models that were based either on tree height or
DBH. Diameter at breast height (r

2

= 0.974, P < 0.01) and height

(r

2

= 0.932, P < 0.01) were strongly predictive of above ground

tree biomass (

Fig. 4

).

From our regression analysis we found tree height (HT) to

be a strong estimator of above ground tree biomass (WB

tree

)

using the following equation:

WB

tree

= 0.0274 × HT − 8.1

(4)

where WB

tree

was in kg and HT was tree height in cm.

Eq.

(5)

estimated WB

tree

as a function of DBH:

WB

tree

= 2.0 × DBH − 4.64

(5)

where DBH was diameter at breast height in cm.

12

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

Table 1 – Weight, height and diameter of hybrid poplar trees grown on trenched municipal biosolids in Maryland (USA)

Year of planting

Age (years)

Dry weight (kg)

Height (cm)

DBH (cm)

Samples

Mean

S.D.

Mean

S.D.

Mean

S.D.

n

2003

2

0.3

0.2

225

47

1.3

0.7

4

2002

3

4.6

4.5

494

188

5.0

2.7

10

2001

4

3.0

1.9

442

106

4.2

1.5

9

2000

5

14.9

6.7

844

91

9.5

2.5

5

1999

6

20.5

2.9

974

66

11.7

1.0

5

From

Fig. 4

we noticed that the relationship between tree

biomass and DBH was not quite linear across the range of
0–14 cm for DBH. There was an apparent inflection point at
a DBH of 4 cm (

Fig. 4

). Likely, young trees did not penetrate the

20 cm of clayey spoil to reach the fertile biosolids until they
were greater than 4 cm in diameter, which would explain the
inflection at 4 cm.

To improve the precision of the DBH allometric model, we

divided the data into two sets based on whether DBH was
greater or less than 4 cm. This division resulted in different

Fig. 4 – Dry weight of hybrid poplar trees as a function of (a)
tree height and (b) diameter at breast height (DBH) during
their first 6 years of growth on trenched municipal
biosolids near Washington, DC.

model coefficients for the two data sets as shown in Eqs.

(6)

and (7)

:

WB

tree

= 2.6 × DBH − 9.64

(6)

when DBH was greater than 4 cm.

WB

tree

= 0.5 × DBH − 0.35

(7)

when DBH was less than 4 cm.

Eq.

(6)

provided slightly better predictability of tree biomass

(r

2

= 0.98; P < 0.001) than Eq.

(5)

. If a plantation owner were

more interested in estimating the biomass of large rather than
small trees, then Eq.

(6)

would be more appropriate than Eq.

(5)

and it would provide a slightly better estimate. Eq.

(7)

, on the

other hand, did not offer as much precision (r

2

= 0.78; P < 0.001)

as Eq.

(5)

, and it would not be of use for estimating the biomass

of large trees.

While it was tempting to build an allometric model

that combined DBH and height to estimate tree biomass,
strong correlation between these two metrics (r

2

= 0.970,

P < 0.01) indicated they were not independent of each
other. The stepwise regression selected DBH and excluded
HT because it was highly correlated to DBH. Combin-
ing these types of collinear variables in a regression
model typically overestimates fitness and builds an unstable
model.

3.3.

Net wood productivity

To estimate the net wood productivity of the hybrid poplar
plantation, we evaluated how well three types of curves (i.e.,
linear, exponential and power) fit the plot of standing tree
biomass as a function of age (

Fig. 5

). As explained above,

the 2001-age class had less standing wood than the 2002-
age class due to a summer drought and heavy deer grazing
during its planting year. Since there were only five years
of data, this single year strongly influenced the regression
analysis.

When we included the 2001-age class in the regression

analysis, all three models explained more than 83% of the
variation in standing wood biomass (all P < 0.05). When we
excluded the 2001-age class from the regression, the linear
model was much superior (r

2

= 0.998; P < 0.001) to the expo-

nential (r

2

= 0.84; not significant) and power (r

2

= 0.922; P < 0.05)

models.

There is anecdotal evidence from the tree farm that trees

that are stunted due to drought and deer browse eventually

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

13

Fig. 5 – Standing wood biomass of hybrid poplar grown on
trenched municipal biosolids for stand ages from 2 to 6
years with exponential, linear and power models fitted to
observed data.

catch up. Similarly,

Sidhu and Dhillon (2007)

found that differ-

ent planting sizes exhibited no differences in height or DBH
by the fifth year of growth. Hence, for our allometric model,
excluding the single year outlier seems justified.

Our estimate of net wood productivity, taken as the slope

of the linear model (

Fig. 5

), was 5450 kg/(ha year) whether the

2001-age class was included or excluded in the regression;
only the y-intercept and coefficient of determination were
changed. By comparison the models of

Das and Chaturvedi

(2005)

and

Norby et al. (2001)

were almost linear in the same

DBH range (4–14 cm) as our poplar study.

4.

Conclusions

A healthy hybrid poplar plantation was grown on the clay spoil
of a former sand and gravel mine where municipal biosolids
were applied in deep trench rows. The oldest tree stand,
which was 6 years old, had a mean standing wood biomass
of 22,100 kg/ha. The most precise allometric model for esti-
mating above ground tree biomass used DBH. Tree height also
served as a strong estimator of tree biomass. The mean net
wood productivity of the hybrid poplar plantation during its
first 6 years of growth was estimated to be 5450 kg/(ha year).

Acknowledgements

Partial financial support was provided by the Washington Sub-
urban Sanitary Commission (Steve Gerwin project manager)
and the University of Maryland College of Agriculture and Nat-
ural Resources, College Park. The Washington, DC Water and
Sewer Authority (Chris Peot project manager) assisted with
project planning and implementation. Logistical field and lab
support was provided by Carrie Buswell, Tommy Griffith, Bob
Dixon, Peter Sharpe and Scott Flamino.

r e f e r e n c e s

Belote, R.T., Weltzin, J.F., Norby, R.J., 2003. Response of an

understory plant community to elevated [CO

2

], depends on

differential responses of dominant invasive species and is
mediated by soil water availability. New Phytol. 161, 827–835.

Buswell, C.U., Felton, G.K., Kays, J.S., Flamino, E.J., 2006. Water

Quality of Deep Row Biosolids Incorporation on a Tree Farm.
ASABE Paper: 06-4048 Presented at 2006 ASABE Annual
Meeting, Portland, OR. 29 p., July 9–July 12, 2006.

Ceulemans, R., 1996. An inventory of tree stand growth models

with potential application in short-rotation forestry. Biomass
Bioenerg. 11 (2/3), 95–107.

Ceulemans, R., Jiang, X.N., Shao, B.Y., 1995. Growth and

physiology of one-year old Poplar (Poplus) under elevated
atmospheric CO

2

levels. Ann. Bot. 75, 609–617.

Das, D.K., Chaturvedi, O.P., 2005. Structure and function of

Populus deltoides agroforestry systems in eastern India. 1. Dry
matter dynamics. Agroforest. Syst. 65, 215–221.

Dreyer, E., Le Roux, X., Montpied, P., Daudet, F.A., Masson, F., 2001.

Temperature response of leaf photosynthetic capacity in
seedlings from seven temperate tree species. Tree Physiol. 21,
223–232.

Felton, G.K., Kays, J.S., Buswell, C.U., Flamino, E., 2005.

Determination of Optimum Tree Density, Biosolid Application
Rate, Water Quality Impacts, and Tree Growth Effects Using
the Deep Row Biosolids Incorporation Method. Washington
Suburban Sanitary Commission, Laurel, MD, 127 pp.

Fleck, J. Road to new fuels: from corn to algae other sources for

powering a car are studied. Albuquerque Journal (New
Mexico), June 18, 2006, p. A1.

Hyvonen, R., Agren, G.I., Linder, S., Persson, T., Cotrufo, M.F.,

Ekblad, A., Freeman, M., Grelle, A., Janssens, I.A., Jarvis, P.G.,
Kellomaki, S., Lindroth, A., loustau, D., Lundmark, T., Norby,
R.J., Oren, R., Pilegaard, K., Ryan, M.G., Sigurdsson, B.D.,
Stromgren, M., van Oijen, M., Wallin, G., 2007. The likely
impact of elevated [CO

2

], nitrogen deposition, increased

temperature and management on carbon sequestration in
temperate and boreal forest ecosystems: a literature review.
New Phytol. 173, 463–480.

Kays, J.S., 2002. Results of hybrid poplar clone trial after two years

at the ERCO Tree Farm in Prince George’s County, Maryland.
Extension Research Note RNS3. Maryland Cooperative
Extension, Western Maryland Research and Education Center,
Keedysville, MD. Available online:

www.naturalresources.umd.edu/Pages/Clone FS/Clone FS.htm

.

Le Roux, X., Lacointe, A., Escobar-Gutierrez, A., Le Dizes, S., 2001.

Carbon-based models of individual tree growth: a critical
appraisal. Ann. For. Sci. 58, 469–506.

Monclus, R., Dreyer, E., Villar, M., Delmotte, F.M., Delay, D., Petit,

J.-M., Barbaroux, C., Le Thiec, D., Brechet, C., Brignolas, F.,
2006. Impact of drought on productivity and water use
efficiency in 29 genotypes of Poplus deltoides

× Poplus nigra.

New Phytol. 169, 765–777.

Norby, R.J., Todd, D.E., Fultz, J., Johnson, D.W., 2001. Allometric

determination of tree growth in a CO

2

-enriched sweetgum

stand. New Phytol. 150, 477–487.

Sidhu, D.S., Dhillon, G.P.S., 2007. Field performance of ten clones

and two sizes of planting stock of Populus deltoids on the
Indo-gangetic plains of India. New Forests 34,115–122.

Sikora, L.J., Burge, W.D., Jones, J.E., 1982. Monitoring of a

municipal sludge entrenchment site. J. Environ. Qual. 11 (2),
321–326.

Taylor, G., 2002. Poplus: Arabidopsis for forestry. Do we need a

model tree? Ann. Bot. 90, 681–689.

Tompkins, M.D., 1983. Prince Georges County Ground-Water

Information: Well Records, Chemical-Quality Data, Pumpage,

14

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 8–14

Appropriation Data, Observation Well Records, and Well Logs.
Maryland Geological Survey Water Resources Basic Data
Report No. 13. 160 pp.

Van Ham, M., Lee, L., McLean, B., 2000. Pit to park: gravel mine

reclamation using biosolids. In: Planning for End Uses in Mine
Reclamation. Proceedings of the Twenty-Fourth Annual
British Columbia Mine Reclamation Symposium, Williams
Lake, BC. GVRD, Vancouver, BC, 38–51. Available online at:

www.nutrifor.com/publications.html

.

Wilson, J.M., Fleck, W.B., 1990. Geology and Hydrologic

Assessment of Coastal Plain Aquifers in the Waldorf Area,
Charles County, Maryland. Maryland Geological Survey,
Report of Investigation No. 53. Baltimore, MD. 138 pp.

Yanbao, L., Yin, C., Li, C., 2006. Differences in some

morphological, physiological, and biochemical responses to
drought stress in two contrasting populations of Populus
przewalskii. Physiol. Plantarum 127, 182–191.