Morin et al.: 1
March 31, 2004
United States
Department of
Agriculture
Forest Service
Mapping Host Species Abundance
of Selected Exotic Pests in the
Eastern United States
Northeastern Research Station
Research Paper NEDate 2004
Randall S. Morin, Andrew M. Liebhold, Eugene
R. Luzader, Andrew J. Lister, Kurt W.
Gottschalk, and Daniel B. Twardus
Morin et al.: 2
March 31, 2004
Abstract
Forests of the eastern United States seem to be particularly vulnerable to forest pest
invasions having suffered devastating effects by notorious forest pests such as chestnut
blight, gypsy moth, and beech bark disease over the last century. In this study we used
existing data to predict the geographical extent of future damage caused by three exotic
pests. The three alien pests considered here were beech bark disease, an insect-fungus
complex involving the beech scale insect (Cryptococcus fagisuga) and the exotic canker
fungus Nectria coccinea var. faginata or the native Nectria galligena, hemlock woolly
adelgid (Adelges tsugae), and gypsy moth (Lymantria dispar). The geographic
distributions of host species of the three pests were mapped in 1 km2 cells by
interpolating host basal area per ha from 93,611 forest inventory plots located throughout
the eastern U.S.A. These interpolated surfaces were adjusted for forest density by
multiplying values by an estimate of forest density derived from existing land cover map
data (30 m2 cells). These maps of host species abundance can be used in the future to
plan efforts to mitigate damages caused by these three exotic pests.
Acknowledgements
We would like to thank the USDA Forest Service Special Technology Development
Program for funding the project.
Morin et al.: 3
March 31, 2004
The Authors
RANDALL S. MORIN is research analyst, ANDREW M. LIEBHOLD is research
entomologist, EUGENE R. LUZADER is computer specialist, and KURT W.
GOTTSCHALK is research forester with the USDA Forest Service’s Northeastern
Research Station. Morin, Liebhold, Luzader and Gottschalk are located in Morgantown,
West Virginia.
ANDREW J. LISTER is research forester with the USDA Forest Service’s Northeastern
Research Station in Newtown Square, PA.
DANIEL B. TWARDUS is forest health specialist with Northeastern Area, State and
Private Forestry, in Morgantown, West Virginia.
Cover
The map on the cover is an estimated surface of percent forest cover.
Manuscript received for publication _____________.
USDA FOREST SERVICE
NEWTOWN SQUARE CORPORATE CAMPUS
11 CAMPUS BLVD., SUITE 200
NEWTOWN SQUARE, PA 19073
???, 2004
Morin et al.: 4
March 31, 2004
First-Grade Headings
Introduction
Methods
Results
Discussion
Literature Cited
Morin et al.: 5
March 31, 2004
Second-Grade Headings
Beech Bark Disease
Hemlock Woolly Adelgid
Gypsy Moth
Morin et al.: 6
March 31, 2004
Introduction
Invasions by exotic insects and diseases are one of the most important threats to the
stability and productivity of forest ecosystems around the world (Liebhold et al. 1995a,
Vitousak et al. 1996, Pimentel et al. 2000). The forests of the eastern U.S.A. seem to be
particularly vulnerable to these invasions; over the last century they have suffered
devastating effects by notorious forest pests such as chestnut blight, gypsy moth, and
beech bark disease (Mattson 1997) with cascading effects throughout invaded ecosystems
(Redman and Scriber 2000). One of the most important steps in the development of
effective strategies for management of alien species is to evaluate the risk of future
impacts from specific exotic organisms (Reichard and Hamilton 1997, Byers et al. 2002).
The generic activity, “risk assessment” is considered an important component to
management of exotics both before and after their arrival in new habitats (Liebhold et al.
1995a).
A risk assessment for an alien species that has not yet been established typically is
comprised of three components: 1) estimates of the probability of arrival and subsequent
establishment in the alien habitat; 2) estimates of the geographical extent that the species
is likely to achieve in it’s new habitat; and 3) estimation of the ecological, economical,
and sociological impact that is likely to be caused by the species. When an alien species
has already become established but is still expanding its range, step 1 is obviously not
necessary.
Morin et al.: 7
March 31, 2004
Maps representing the geographic extent of estimated future disturbance caused by exotic
pests are a critically important part of risk assessment (Rouget et al. 2002). Prediction of
disturbance risk in forest ecosystems is often limited both by the lack of data and models
for predicting the geographic extent of impacts caused by specific pest species. In this
paper, we addressed this deficiency by developing statistical models that predict forest
pest host species abundance from interpolated forest inventory plots distributed across the
eastern United States. The three alien pests considered here were beech bark disease, an
insect-fungus complex involving the beech scale insect (Cryptococcus fagisuga) and the
exotic canker fungus Nectria coccinea var. faginata or the native Nectria galligena,
hemlock woolly adelgid (Adelges tsugae), and gypsy moth (Lymantria dispar). These
species have become established in North America and are currently expanding their
ranges. These estimates of host species abundance will be useful for both long-range
planning of population management activities and implementing cultural practices to
minimize impacts.
Beech bark disease, also known as beech scale Nectria canker, is an insect-fungus
complex involving the beech scale insect (Cryptococcus fagisuga) and the exotic canker
fungus Nectria coccinea var. faginata or the native Nectria galligena that kills or injures
American beech (Fagus grandifolia) (Houston 1994). The disease results when a Nectria
fungus infects the bark through feeding wounds caused by beech scale insects. Around
1900, the beech scale insect was accidentally introduced to Nova Scotia from Europe. It
has since spread southwestward into New England, New York, Pennsylvania, and West
Virginia (Manion 1991). Fungal invasion occurs about 3 to 5 years after the scale insects
Morin et al.: 8
March 31, 2004
appear. Extensive mortality often occurs in the year that Nectria becomes evident; some
trees may survive for several years.
The hemlock woolly adelgid is a destructive exotic insect pest of forest and ornamental
hemlock trees. It was first observed in the United States in the 1950s in Virginia. In the
eastern USA, hemlock woolly adelgid feeds only on the foliage of eastern hemlock
(Tsuga canadensis) and Carolina hemlock (Tsuga caroliniana), a less common species of
hemlock found in isolated areas of the southern Appalachians. Heavy infestations have
killed trees in as little as four years, but some trees have survived infestations for more
than 10 years (McClure et al. 2001).
The gypsy moth was accidentally introduced in 1868 or 1869 near Boston, Massachusetts,
and outbreaks began to occur in the area about 10 years later (Liebhold et al. 1992). The
range of the gypsy moth has been gradually expanding, and defoliating populations are
typically found 5-6 years following establishment in a new area (Liebhold et al. 1994,
Sharov and Liebhold 1998). The gypsy moth is a polyphagous insect. North American
populations feed on over 300 different shrub and tree species, and there is considerable
variation among tree species in their susceptibility to defoliation (Leibhold et al. 1995b).
The most preferred gypsy moth hosts are in the genera Quercus, Populus, and Larix.
Gypsy moth is an outbreak pest species that has been acknowledged as the most
destructive forest defoliating insect in the United States (Davidson et al. 1999, Sharov et
al. 2002).
Morin et al.: 9
March 31, 2004
Methods
The geographic distributions of suitable habitats for beech bark disease, hemlock woolly
adelgid, and the gypsy moth were mapped by interpolation of host species abundance
estimated from 93,611 forest inventory plots located throughout the eastern U.S. The
data were extracted from the Eastwide Database (Table 1), which at the time the project
was executed was the most recent available source of USDA Forest Service Forest
Inventory and Analysis data for the 37 states in the study. Host abundance for beech bark
disease was measured as basal area/ha of American beech, host abundance for hemlock
woolly adelgid was measured as basal area/ha of eastern and Carolina hemlock, and host
abundance for gypsy moth was measured as the total basal area/ha of tree species
preferred by gypsy moth. A host species was considered preferred if it had a high
preference rating (Liebhold et al. 1995b).
We used the ordinary kriging procedure (Deutsch and Journel 1998) to interpolate a
surface of basal area/ha for the host species of each disease or insect. Kriging is a
geostatistical method that provides estimates for unsampled locations by computing
weighted averages of sampled values from nearby locations (Isaaks and Srivistava 1989,
Cressie 1993). The weights are determined based on the semivariogram, a statistical
model of the relationship between spatial autocorrelation and distance between pairs of
sampled values. In this analysis, we generated maps from the plot data by calculating
kriged estimates on a grid of 1x1 km cells. Variography and kriging were performed
using the GSLIB software library (Deutsch and Journel 1998).
Morin et al.: 10
March 31, 2004
Forest susceptibility maps were then adjusted for forest density. During the period of
data collection for this study, the FIA plots were in effect located randomly, and tree data
were in general only collected in forested areas. Therefore, it was necessary to adjust host
abundance estimates in each cell for the proportion of the cell area that was forested.
This was accomplished using a region-scale land cover dataset (National Land Cover
Data) that was acquired from the Multi-Resolution Land Characteristics Consortium as a
raster matrix of 30x30 m cells that were classified based on their land cover
(http://www.epa.gov/mrlc/nlcd.html) (Vogelmann et al. 2001). These data were
aggregated to a 1x1 km resolution to estimate percent forest cover for each cell (Figure
1). Next, the kriged host abundance maps were multiplied by the forest density map to
generate forest susceptibiltiy maps adjusted for percent forest cover.
When combined with current range maps, the host density maps allow the current range
and potential range of these three exotic pests to be estimated (Table 2). A 1 km2 pixel
was included in the range estimates if it was estimated to contain any host basal area.
Results
Beech Bark Disease
American beech basal area/ha was estimated (Figure 2) and then mulitiplied by the forest
density map (Figure 1) to create a beech abundance map adjusted for forest density
(Figure 3). The greatest concentration of beech is in the Adirondack Mountain region of
New York though smaller concentrations existed in Maine, New Hampshire, Vermont,
West Virginia, and northern Pennsylvania. Beech also exists at very low levels over a
Morin et al.: 11
March 31, 2004
large range extending through most forested regions of the eastern U.S.A. The current
range and historical spread of the beech scale insect is shown in Figure 4. Our estimates
indicate that beech bark disease currently occupies only about 27% of its potential range
(Table 2).
Hemlock Woolly Adelgid
Hemlock basal area/ha was estimated using ordinary kriging (Figure 5) and then
mulitiplied by the forest density map (Figure 1) to create a hemlock abundance map
adjusted for forest density (Figure 6). Hemlock is widely distributed throughout the
northeastern U.S. and the northern portion of Michigan and Wisconsin. Elsewhere, its
distribution is limited to the higher elevations of the Appalachian Mountains. The current
range and historical spread of hemlock woolly adelgid is shown in Figure 7. Our
estimates indicate that hemlock woolly adelgid currently occupies nearly 26% of its
potential range (Table 2).
Gypsy Moth
Basal area of species preferred by gypsy moth was estimated using ordinary kriging
(Figure 8) and then mulitiplied by the forest density map (Figure 1) to create a preferred
species abundance map adjusted for forest density (Figure 9). Of the three pest species,
gypsy moth hosts were the most widely distributed. In addition, maximum host densities
(>20 m2/ha) were much higher than hosts for the other two species. The current range
and historical spread of gypsy moth is shown in Figure 10. Our estimates indicate that
hemlock woolly adelgid currently occupies only 23% of its potential range (Table 2).
Morin et al.: 12
March 31, 2004
Discussion
The planning of forest pest management activities can be significantly improved with
reliable mapped estimates of future disturbances. The methods developed here and maps
of host abundance presented here should prove useful for future efforts of this nature.
Specifically, these maps may be used to help guide land managers make decisions about
where pest surveys and silivicultural manipulations should be located.
Beech bark disease has already invaded those areas with the greatest concentration of
beech but vast portions of the range of beech remain to be invaded in the future. In fact
only about 27% of the native range of beech has been invaded (Table 2). American beech
is most abundant in the Adirondacks and in the central and northern Appalachians.
Elsewhere it is widely distributed at low levels throughout the northeast and southeast
US. The fact that beech is still abundant in many regions where beech bark disease has
already invaded illustrates that while it can cause considerable mortality, beech bark
disease generally does not totally eliminate beech from stands (Houston 1994, Morin et
al. 2001). The estimated American beech abundance map (Figure 3) highlights eastern
Kentucky, northern Ohio, and northern Indiana as the highest risk areas that are currently
uninfested.
In contrast to beech bark disease, hemlock woolly adelgid has only recently begun to
move into areas with a large eastern hemlock component. Only about 26% of the native
range of eastern and Carolina hemlock have been invaded (Table 2). Estimated hemlock
Morin et al.: 13
March 31, 2004
abundance was highest in a large portion of New England, as well as New York and
northern Pennsylvania (Figure 6).
Gypsy moth is currently occupying less than one-quarter of its favorable habitat. The
currently uninfested areas that have the highest densities of host species are western
Virginia, southwestern West Virginia, north-central North Carolina, and eastern
Minnesota (Figure 9).
A limitation that might affect the accuracy of this information is that host species
abundance is a major factor that determines susceptibility, but it is not the only factor that
determines vulnerability of a forest to a pest. For example, local site characteristics, such
as climate and soils (e.g. Houston and Valentine 1977), also affect forest susceptibility.
The methodologies employed thus represents a first approximation of host species
abundance and these methods can be elaborated in the future to generate improved host
species abundance maps for these and other pests.
Literature Cited
Andow, D.A.; Kareiva, P. M.; Levin, S.A.; Okubo, A. 1990. Spread of invading
organisms. Landscape Ecology 4: 177-188.
Bailey, R.G. 1995 Description of the ecoregions of the United States. USDA Forest
Service Misc. Pub. 1391, 2nd ed., 108 p.
Morin et al.: 14
March 31, 2004
Byers, J.E.; Reichard, S.; Randall, J.M.; Parker, I.M.; Smith, C.S.; Lonsdale, W.M.;
Atkinson, I.A.E.; Seastedt, T.R.; Williamson, M.; Chornesky, E.; Hayes, D. 2002.
Directing research to reduce the impacts of nonindigenous species. Conservation
Biology 16: 630-640.
Cressie, N.A.C. 1993. Statistics for Spatial Data. John Wiley & Sons, Inc., New York.
Davidson, C.B.; Gottschalk, K.W.; Johnson, J.E. 1999. Tree mortality following
defoliation by the European gypsy moth (Lymantria dispar L.) in the United States:
a review. Forest Science 45: 74-84.
Deutsch, C.V.; Journel, A.G. 1998. GSLIB: Geostatistical Software Library and
User’s Guide, Second Edition. Oxford University Press, New York.
Houston, D.R. 1994. Major new tree disease epidemics: beech bark disease. Annual
Review of Phytopathology 32: 75-87.
Houston, D.R.; Valentine, H.T. 1977. Comparing and predicting forest stand
susceptibility to gypsy moth. Canadian Journal of Forest Research 7: 447-461.
Isaaks, E.H.; Srivastava, R.M. 1989. An Introduction to Applied Geostatistics. Oxford
University Press, New York.
Morin et al.: 15
March 31, 2004
Levin, S.A. 1989. Analysis of risk for invasions and control programs. In: Drake, J.A.;
Mooney, H.A.; di Castri, F.; Groves, R.H.; Kruger, K.J.; Rejmánek, M.; Williamson, M.
(Eds.), Biological Invasions: A Global Perspective. Wiley, Chichester, pp. 425-435.
Liebhold, A.M.; Halverson, J.A.; Elmes, G.A. 1992. Gypsy moth invasion in North
America: a quantitative analysis. Journal of Biogeography 19: 513-520.
Liebhold, A.M.; Rossi, R.E.; Kemp, W.P. 1993. Geostatistics and geographic
information systems in applied insect ecology. Annual Review of Entomology. 38:
303-327.
Liebhold, A.M.; Elmes, G.A.; Halverson, J.A.; Quimby, J. 1994. Landscape
characterization of forest susceptibility to gypsy moth defoliation. Forest Science 40:
18-29.
Liebhold, A.M.; Macdonald, W.L.; Bergdahl, D.; Mastro, V.C. 1995a. Invasion by
exotic pests: a threat to forest ecosystems. Forest Science Monographs 30: 49 p.
Liebhold, A.M.; Gottschalk, K.W.; Muzika, R.; Montgomery, M.E.; Young, R.; O’Day,
K.; Kelley, B. 1995b. Suitability of North American tree species to the gypsy moth: a
summary of field and laboratory tests. USDA Forest Service Gen. Tech. Rep. NE-211,
34 p.
Morin et al.: 16
March 31, 2004
Liebhold, A.M.; Gottschalk, K.W.; Luzader, E.R.; Mason, D.A.; Bush, R.; Twardus, D.B.
1997. Gypsy moth in the United States: an atlas. USDA Forest Service Gen. Technical
Rep. NE-233, 33 pp.
Manion, P.D. 1991. Tree Disease Concepts, Second Edition. Prentice Hall Career &
Technology, Upper Saddle River, NJ.
Mattson, W.J. 1997. Exotic insects in North American forests – ecological systems
forever altered. In: Proceedings of Exotic Pests of Eastern Forests. USDA Forest Service
& TN Exotic Pest Plant Council, Nashville, TN., pp. 187-193
McClure, M.S.; Salom, S.M.; Shields, K.M. 2001. Hemlock woolly adelgid. USDA
Forest Service FHTET-2001-03, 14 p.
Morin, R.S.; Liebhold, A.M.; Gottschalk, K.W.; Twardus, D.B.; Acciavatti, R.E.; White,
R.L.; Horsley, S.B.; Smith, W.D.; Luzader, E.R. 2001. Forest health conditions on the
Allegheny national forest (1989-1999): analysis of forest health monitoring surveys.
USDA Forest Service Northeastern Area Tech. Pap. NA-TP-04-01, 68 p.
Pimentel, D.; Lach, L.; Zuniga, R.; Morrison, D. 2000. Environmental and economic
costs of nonindigenous species in the United States. BioScience 50: 53-65.
Morin et al.: 17
March 31, 2004
Redman, A.M.; Scriber, J.M. 2000. Competition between the gypsy moth, Lymantria
dispar, and northern tiger swallowtail, Papilio canadensis: interactions mediated by
host plant chemistry, pathogens, and parasitoids. Oecologia. 125: 218-228.
Reichard, S.H.; Hamilton, C.W. 1997. Predicting ivasions of woody plants introduced
into North America. Conservation Biology 11: 193-203.
Rouget, M.; Richardson, D.M.; Nel, J.L.; Van Wilgen, B.W. 2002. Commercially
important trees as invasive aliens – towards spatially explicit risk assessment at a
national scale. Biological Invasions 4: 397-412.
Sharov, A.A.; Liebhold, A.M.; Roberts, E.A. 1997. Spatial variation among counts of
gypsy moths (Lepidoptera: Lymantriidae) in pheromone-baited traps at the
expanding front. Environmental Entomology 25: 1312-1320.
Sharov, A.A.; Liebhold, A.M. 1998. Model of slowing the spread of the gypsy moth
(Lepidoptera: Lymantriidae) with a barrier zone. Ecological Applications 8: 11701179.
Sharov, A.A.; Leonard, D.; Leibhold, A.M.; Roberts, E.A.; Dickerson, W. 2002. “Slow
the spread”: a national program to contain the gypsy moth. Journal of Forestry 100:
30-35.
Morin et al.: 18
March 31, 2004
Shigesada, N.; Kawasaki, K. 1997. Biological Invasions: Theory and Practice. Oxford
University Press, New York.
Shigesada, N.; Kawasaki, K.; Takeda, Y. 1995. Modeling stratified diffusion in
biological invasions. American Naturalist 146: 229-251.
Skellam, J.G. 1951. Random dispersal in random populations. Biometrika 38: 196218.
Souto, D.; Luther, T.; Chianese, B. 1996. Past and current status of HWA in eastern
and Carolina hemlock stands. In: Salom, S.M., Tignor, T.C., Reardon, R.C. (Eds.),
Proceedings of the First Hemlock Woolly Adelgid Review, USDA Forest Service,
Morgantown, WV, pp. 9-15.
Vitousek, P.M.; D’Antonio, C.M.; Loope, L.L.; Westbrooks, R. 1996. Biological
invasions as global environmental change. American Scientist 84: 468-478.
Vogelmann, J.E.; Howard, S.M.; Yang, L.; Larson, C.R.; Wylie, B.K.; Van Driel, N.
2001. Completion of the 1990s National Land Cover Data Set for the conterminous
United States from Landsat Thematic Mapper data and ancillary data sources.
Photogrammetric Engineering and Remote Sensing 67: 650-684.
Morin et al.: 19
March 31, 2004
LIBRARY CARD
Morin, Randall S.; Liebhold, A.M.; Luzader, E.R.; Lister, A.J.; Gottschalk, K.W.;
Twardus, D.B. 2004. Mapping Host Species Abundance of Selected Exotic Pests in the
Eastern United States. Res. Pap. NE-???. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northeastern Research Station, ??? p.
Forests of the eastern United States seem to be particularly vulnerable to forest pest
invasions having suffered devastating effects by notorious forest pests such as chestnut
blight, gypsy moth, and beech bark disease over the last century. In this study we used
existing data to predict the geographical extent of future damage caused by three exotic
pests through 2025. The three alien pests considered here were beech bark disease, an
insect-fungus complex involving the beech scale insect (Cryptococcus fagisuga) and the
exotic canker fungus Nectria coccinea var. faginata or the native Nectria galligena,
hemlock woolly adelgid (Adelges tsugae), and gypsy moth (Lymantria dispar). The
geographic distributions of host species of the three pests were mapped in 1 km2 cells by
interpolating host basal area per ha from 93,611 forest inventory plots located throughout
the eastern U.S.A. These interpolated surfaces were adjusted for forest density by
multiplying values by an estimate of forest density derived from existing land cover map
data (30 m2 cells). These maps of host species abundance can be used in the future to
plan efforts to mitigate damages caused by these three exotic pests.
Keywords: beech bark disease; hemlock woolly adelgid; gypsy moth
Morin et al.: 20
March 31, 2004
Table 1. Year and number of forest inventory plots in each eastern state
State
Year # of Plots
Alabama
1990 3575
Arkansas
1995 3090
Connecticut
1985
283
Delaware
1986
134
Florida
1995 3822
Georgia
1997 5789
Illinois
1998 1586
Indiana
1986 1976
Iowa
1990
636
Kansas
1994 1548
Kentucky
1988 1921
Louisiana
1991 2233
Maine
1995 2587
Maryland
1986
672
Massachussets 1985
371
Michigan
1993 9933
Minnesota
1990 11553
Mississippi
1994 2960
Missouri
1989 4664
Nebraska
1994
428
New Hampshire 1983
586
New Jersey
1987
251
New York
1993 2850
North Carolina 1990 4558
North Dakota 1995
265
Ohio
1991 1587
Oklahoma
1993
862
Pennsylvania 1989 2964
Rhode Island
1985
117
South Carolina 1993 3358
South Dakota 1980
46
Tennessee
1999 2348
Texas
1992 1925
Vermont
1989
619
Virginia
1992 3600
West Virginia 1989 2491
Wisconsin
1996 5423
Morin et al.: 21
March 31, 2004
Table 2. Current and potential ranges of beech bark disease, hemlock woolly adelgid, and
gyspy moth (areas in sq. km.)
Potential area
Current area of
%
Pest
of host habitat infested host habitat infested
beech bark disease
1,589,744 sq. km.
434,548 sq. km.
27.3
hemlock woolly adelgid 714,426 sq. km.
182,942 sq. km.
25.6
gypsy moth
3,637,478 sq. km.
851,753 sq. km.
23.4
Morin et al.: 22
March 31, 2004
FIGURE LEGENDS
FIGURE 1.--Map of percent forest density from NLCD data
FIGURE 2.--Kriged map of estimated American beech basal area
FIGURE 3.--Kriged map of estimated American beech basal area adjusted for forest
density
FIGURE 4.--Map of beech bark disease range (2003)
FIGURE 5.--Kriged map of estimated hemlock basal area
FIGURE 6.--Kriged map of estimated hemlock basal area adjusted for forest density
FIGURE 7.--Map of hemlock woolly adelgid range (2003)
FIGURE 8.--Kriged basal area map of species preferred by gypsy moth
FIGURE 9.--Kriged basal area map of species preferred by gypsy moth adjusted for forest
density
FIGURE 10.--Map of gypsy moth range (2003)
Morin et al.: 23
March 31, 2004
Figure 1
Morin et al.: 24
March 31, 2004
Figure 2
Morin et al.: 25
March 31, 2004
Figure 3
Morin et al.: 26
March 31, 2004
Figure 4
Morin et al.: 27
March 31, 2004
Figure 5
Morin et al.: 28
March 31, 2004
Figure 6
Morin et al.: 29
March 31, 2004
Figure 7
Morin et al.: 30
March 31, 2004
Figure 8
Morin et al.: 31
March 31, 2004
Figure 9
Morin et al.: 32
March 31, 2004
Figure 10