Open Journal of Forestry
2012. Vol.2, No.4, 213-218
Published Online October 2012 in SciRes (http://www.SciRP.org/journal/ojf) http://dx.doi.org/10.4236/ojf.2012.24026
Bloodroot (Sanguinaria canadensis L.) Extent and Sustainability
in Western North Carolina
Jill Furgurson1*, Fred Cubbage2, Erin Sills2, Peter Bates3
1Department of Geography, University of North Carolina, Chapel Hill, USA
2Department of Forestry and Environmental Resources, North Carolina State University, Raleigh, USA
3Department of Geosciences and Natural Resources, Western Carolina University, Cullowhee, USA
Received June 13th, 2012; revised July 24th, 2012; accepted August 8th, 2012
Bloodroot distribution and abundance were assessed in the Waynesville watershed in Western North
Carolina. This high quality site provides a benchmark for bloodroot populations in the region. Summary
data from an inventory of nine stands of bloodroot in the watershed are presented. Analysis of inventory
data reveals that both petiole height and petiole diameter are negatively associated with overstory tree
DBH, suggesting that there is an optimal overstory structure for bloodroot. In the Waynesville watershed,
seven out of nine stands have an average tree DBH between 27.38 cm and 36.17 cm. Allometric equations
relating belowground biomass to bloodroot petiole height and diameter have strong explanatory power,
indicating that harvesters could selectively harvest large rhizomes by targeting plants with larger petioles.
These results in combination with natural history, field observations and literature provide insights on the
sustainability of bloodroot harvest in Southern Appalachia. Wild bloodroot is likely becoming scarce due
to loss of favorable sites, such as rich cove forests, as well as harvest pressure.
Keywords: Bloodroot; Nontimber Forest Product; NTFP; Sustainability; Waynesville Watershed; Western
North Carolina
Introduction
The forest plant bloodroot (Sanguinaria canadensis L.) is
harvested from natural stands in the eastern United States for
both artistic and medicinal uses. In order to ensure this resource
is available for future generations, baseline information is re-
quired for establishing sustainable harvesting regimes. Density
and size-class structure are essential data for determining sus-
tainable harvest levels for nontimber forest products from natu-
ral forest ecosystems (Peters, 1994). Sustainable, as used in this
study, has both biological and managerial dimensions. Harvests
are sustainable if they allow the continuation of viable natural
populations of the resource and should incorporate management
practices that support indefinite yields. Additional goals include
maintenance of biodiversity within natural stands, preservation
of traditional knowledge, and generation of income for local
communities.
Bloodroot Uses and Sustainability
Bloodroot is an herbaceous perennial plant that grows in rich
well-drained forest soils (Predny & Chamberlain, 2005; Green-
field & Davis, 2004). Populations of the plant can be found
from the Atlantic to the Rocky Mountains, and from southern
Canada through the southern United States (Cech, 2002). In
Western North Carolina, bloodroot is harvested for both me-
dicinal and artistic uses and is currently used by the Eastern
Band of Cherokee Indians as a dye plant for hand woven bas-
kets, a tradition that goes back for hundreds of years. In recent
years the plant has been considered as an alternative to syn-
thetic antibiotics used in cattle feeds (Greenfield & Davis,
2004), and bloodroot alkaloids have been studied as a compo-
nent of cancer treatment medications (Nihal et al., 2000).
Bloodroot is considered to be a nontimber forest product
(NTFP) since nearly all harvesting is from natural stands, as
opposed to cultivated sources. NTFPs are biological resources,
such as medicinal plants, nuts, resins, dyes and ornamental
plants, which can be harvested from the forest to provide in-
come. Despite the widely touted conservation potential of
NTFPs, ecological assessments of harvesting implications are
rare (Ticktin, 2004). NTFPs offer local communities in South-
ern Appalachia a supplemental income from the forest and are
important to local culture, but over-harvesting of wild popula-
tions could pose a threat to long term population viability.
Over-collection is an important factor in population decline
of medicinal plants, and this is especially true of plants where
the root, bark or whole plant is harvested (Hayden, 2005). Since
the rhizome of bloodroot contains the desired medicinal com-
pounds, the plant is especially susceptible to mortality follow-
ing harvesting if careful collection practices are not followed.
Members of the Eastern Band of Cherokee Indians recommend
replanting a piece of the rhizome back into the ground while
harvesting bloodroot in order to maintain viable populations.
This practice is in accordance with recommended propagation
methods, which suggest breaking the rhizome into small pieces,
approximately one to two inches, for successful clonal propaga-
tion (Persons & Davis, 2005).
As Greenfield and Davis (2004) report, increasing demand
for bloodroot is putting pressure on naturally occurring popula-
tions. For centuries, the many uses of bloodroot have led to its
being intensively harvested from North American forests. Al-
*Corresponding author.

J. FURGURSON ET AL.
though small quantities of dried bloodroot from wild stock can
be found on the market, and some nurseries offer small vol-
umes of cultivated stock, large volumes are not available from
cultivated sources (Persons & Davis, 2005). Conventional cul-
tivation of forest medicinal plants can be cost prohibitive,
which is one factor contributing to diminishing wild popula-
tions (Naud et al., 2009). The USDA Forest Service Southern
Research Station (SRS) has recognized an increase in the har-
vesting of special forest products in the southeastern United
States and associated threats to the long-term social, ecological
and economic sustainability of these forest resources (Cham-
berlain, 2004).
The development of a management plan for a medicinal plant
should incorporate knowledge from various sources. The
Cherokee people utilized and influenced forest resources long
before the development of forest science. In order to most ef-
fectively manage forest resources such as bloodroot, collabora-
tion between local communities, government agencies and for-
est scientists is necessary. Understanding traditional knowledge
regarding forest resources is an integral part of developing sus-
tainable management plans. However, consumption patterns
today differ from historical trends, and knowledge of the pre-
sent state of the forest is also important for guiding manage-
ment plans. With a greater demand for forest products, it is
necessary to combine traditional knowledge of distribution and
growth patterns with monitoring efforts and inventory analysis.
Specifically, a forest inventory is necessary to ascertain the
current distribution and abundance of the populations of the
species under consideration (Peters, 1994).
Case Study Overview
We implemented a detailed bloodroot inventory in the
Waynesville city watershed in Haywood County in western
North Carolina. The site was chosen because of its mature for-
ests that provide high-quality sites for bloodroot and because
bloodroot harvest had not been allowed. The strategic man-
agement plan for the watershed has the primary objective of
maintaining a healthy and diverse forest to ensure a supply of
high quality drinking water for the town of Waynesville. The
preservation of unique plant species is a secondary objective,
and although selective timber harvests occasionally occur,
nearly all stands in the bloodroot growth areas contain trees of
approximately 80 - 85 years.
A single resource inventory at one point in time allows quan-
tification of the distribution and abundance of the resource and
estimation of relationships pertinent to sustainable harvesting.
In this case, inventory data on plant sizes and population densi-
ties from the Waynesville watershed in Western North Carolina
are used to establish a baseline on biological supply of blood-
root in natural stands not subject to harvesting. This provides a
reference point useful both for development of guidelines for
sustainable levels of extraction and for assessment of the poten-
tial for this resource to be sustainably utilized. Further, we
combine the data from this site with the literature on forest
types in the Appalachians to estimate the extent of bloodroot in
the region.
We also analyze the relationship between easily recognized
aboveground characteristics (petiole height and petiole diameter)
and belowground biomass of rhizomes. The ability to predict
belowground biomass without actually harvesting the plant is
critical for sustainable harvesting. Specifically, if harvesters can
practice more selective harvesting, plant mortality rates may
decline.
Methods
In the summer of 2006, we canvassed sites with both pro-
tected and harvested stands of bloodroot. We measured the
density of bloodroot and forest stand conditions in both types of
sites and then focused on the 8030 acre tract of mature second
growth forest that comprises the Waynesville watershed (Fig-
ure 1). The Waynesville watershed is a north-facing, rich forest
cove site with trees approximately 80 - 85 years old. It is a high
Figure 1.
Map of location of Waynesville watershed in Western North Carolina.
214

J. FURGURSON ET AL.
quality mature hardwood forest with good soils and drainage,
and provides high quality habitat for bloodroot growth. Situated
within the Great Balsam Mountains/Pisgah Ridge Natural
Heritage Megasite, the watershed is part of a natural heritage
area recognized by the NC Natural Heritage Program (Conser-
vation Trust for North Carolina, 2005).The watershed is man-
aged under a conservation easement that limits timber harvests.
Further, the bloodroot stands in the watershed are considered
protected and not subject to harvesting.
Bloodro ot a nd Forest Sampling
Bloodroot stands were defined as naturally isolated popula-
tions of individual plants geographically separated from other
bloodroot growth areas. Random sampling, which can be time
consuming and leave large sections of an area unsampled, is not
always appropriate for inventories and monitoring of NTFPs
(Kerns et al., 2002). Based on the literature (Predny & Cham-
berlain, 2005; Persons & Davis, 2005), we identified areas
within the watershed likely to support bloodroot growth and
then completely surveyed these areas. The identified areas were
based on forest type and all rich cove, mesic and northern
hardwood stands were sampled. All stands of bloodroot identi-
fied were inventoried to provide a census of bloodroot popula-
tions within the watershed.
The width and length of the identified stands were measured
and then a central transect was created that ran the entire length
of the stand. Petiole height (from the soil surface) and petiole
diameter at ground level were measured for all plants growing
within one half meter of either side of the transect. The total
number of plants measured was 2789. Data were also collected
on species and DBH of overstory trees, as well as stand eleva-
tion.
The analyses of above and belowground biomass relation-
ships were based on 174 plants harvested from the Waynesville
watershed between July 25 and August 4, 2006. Occasionally,
one bloodroot rhizome can support several stems. Since the
intent was to examine the relationship between above ground
leaf characteristics and rhizome mass, only rhizomes supporting
one petiole were measured.
The length and wet weight of rhizomes were measured. Rhi-
zomes were weighed wet so that they could be replanted di-
rectly following measurement. However, the rhizomes are usu-
ally marketed in a dried form.
Statistical Analyses
A one-way analysis of variance (ANOVA) test was per-
formed to detect differences among stands in density of blood-
root (using JMP from SAS). The relationships between the
aboveground plant characteristics of petiole height and diame-
ter and tree dbh were analyzed with ordinary least squares
(OLS) regression analysis, using a robust estimator for cluster-
ing by stand (using LIMDEP). We also estimated OLS regres-
sions of rhizome biomass as a function of petiole height and
petiole diameter to identify relationships between the above-
ground plant characteristics and below ground biomass. Fol-
lowing the methods of Ott and Longnecker (2001), all analyses
were conducted with and without an outlier stand that is excep-
tionally small and that has an exceptionally high density of
bloodroot. We included intercepts in all regressions, as recom-
mended by Bond-Lamberty et al. (2002).
Results
Bloodroot Extent
Bloodroot was found growing on approximately 34.8 acres
(141,445 m2) of the 433 acres (1,752,000 m2) of suitable habitat
cruised on the Waynesville watershed, or about 4.2 percent of
the watershed. The bloodroot was distributed across the water-
shed in nine stands, of which eight are located on northern
hardwood slopes or poplar coves. The ninth stand is located in a
mesic oak-hickory stand. The statistics for the nine bloodroot
stands are summarized in Table 1. The bloodroot stands vary in
dimensions, from 2.2 × 2.4 to 395 × 140 meters. Eight of the
nine stands have a mean area of 17,679 m2, while an outlier
stand has an area of just 5.28 m2. Average bloodroot densities
for the nine stands vary from 1.68 to 35.42 plants per square
meter. These values are based on the density measurements
obtained from sampling along the central transect and are used
for the ANOVA analyses. Eight out of nine of the stands have
average densities of 6.77 plants per square meter or lower. The
small stand with a density of 35.42 plants per square meter is
unusual when compared to all other stands sampled and can-
vassed for this study, and thus is treated as an outlier.
The petiole height of bloodroot plants in the Waynesville
watershed varies from 7.62 cm to 33.02 cm and petiole diame-
ter ranges from 0.08 cm to 0.75 cm. Average petiole heights at
the stand level vary from 13.94 cm to 26.39 cm and average
petiole diameters at the stand level vary from 0.23 cm to 0.40
cm.
The average tree diameter at breast height (DBH) of the
overstory trees measured in each stand varies from 15.27 cm to
31.83 cm. However, 7 out of 9 stands have an average tree dbh
between 27.38 cm and 36.17 cm. Tree species richness, defined
as the number of overstory tree species identified within the
stand, ranges from four to eight species. The most common
species are tulip poplar, red oak, hickory and white oak.
The results of the one way ANOVA revealed that there were
no significant statistical differences (α = 0.05) in mean densities
among the nine bloodroot stands in the Waynesville watershed
(F = 2.583; df = 1.7; p = 0.152), even including the outlier stand.
Excluding the outlier stand, the F statistic drops to 0.092, with a
p value of 0.772.
Average petiole height is significantly negatively correlated
with average tree dbh (coefficientdbh = -1.2, p-value = 0.02, R2
= 0.67, df = 2788). Average petiole diameter is also negatively
correlated with average tree dbh (coefficientdbhl = -0.02, p-value
= 0.05, R2 = 0.34, df = 2788). In stands with larger overstory
trees, bloodroot plants tend to be smaller, whether measured by
petiole height or diameter.
Relationship between Aboveground Form and
Belowground Volume
Site-specific allometric equations were developed relating
the aboveground plant characteristics of petiole height and
petiole diameter to the belowground biomass of rhizomes.
Simple linear regressions of rhizome weight on each of these
characteristics were statistically significant for petiole height,
R2 = 0.65, pmodel < 0.0001 (Figure 2); and for petiole diameter,
R2 = 0.73; pmodel < 0.0001 (Figure 3). Thus, diameter has the
greatest explanatory power.
A single multiple regression was estimated after controlling
for petiole characteristics and confirms that the model continues

J. FURGURSON ET AL.
Table 1.
Bloodroot stand inventory data.
Stand number Average density (plants/m) Average petiole height (cm)Average petiole diameter (cm)Average tree dbh (cm) Count of tree species
1 35.42 26.39 0.40 15.27 5
2 5.99 16.18 0.26 36.17 4
3 2.09 19.61 0.34 27.99 8
4 2.10 18.14 0.29 28.55 5
5 1.68 18.13 0.29 27.38 7
6 6.77 16.33 0.25 27.94 5
7 1.82 16.93 0.26 29.64 8
8 3.31 15.58 0.23 30.12 5
9 5.79 13.94 0.29 31.83 6
0
5
10
15
20
25
0510 15 20 25 30 35 40
Rhi z om e wei g ht ( g)
Petiole height (cm)
Figure 2.
Scatter plot and fitted simple linear regression of bloodroot rhizome weight on stem height,
N = 174, Rhizome weight (g) = -5.1387 + 0.4319 (petiole height); R2 = 0.65; p < 0.0001.
Figure 3.
Scatter plot and fitted simple linear regression of bloodroot rhizome weight on stem diameter,
N = 174, Rhizome weight (g) = -2.9825 + 27.5459 (petiole diameter); R2 = 0.73; p < 0.0001.
to provide a good fit for the data (R2 = 0.77, pmodel < 0.0001),
and the coefficients on both petiole height and diameter re-
mained significant at the 1% level. The allometric equation for
predicting belowground rhizome biomass with only these sta-
tistically significant predictor variables of petiole height and
petiole diameter is:
Rhizome weight (g) = -4.6284 + 0.1802 (petiole height) +
18.7254 (petiole diameter).
Discussion
While nontimber forest products have received increasing
216

J. FURGURSON ET AL.
attention (Shackleton et al., 2011, Kerns et al., 2002, Ticktin,
2004), our research found relatively little literature relating to
their sustainable harvest (Wong, 2000). We did not find any
research recommending guidelines for harvest of bloodroot. In
order to address this gap in the scientific literature, this study
provides plant size and stand density information for nine
stands of protected bloodroot in the Waynesville watershed of
North Carolina. We also analyzed the relationship between
easily observable aboveground plant characteristics and below-
ground biomass, which could contribute to guidelines for sus-
tainable harvest of the resource. In this section, we complement
the analyses with a brief discussion of the distribution and long
term sustainability of bloodroot in western North Carolina to
provide context for comparisons between the protected stands
that we inventoried and nearby harvested stands.
Waynesville Case Study
In the Waynesville watershed, we found that most bloodroot
(eight of the nine stands identified) occurs on northern hard-
wood slopes or poplar coves. Within this habitat type, blood-
root stands have a clumped distribution. Eight stands had simi-
lar characteristics. While the mean density of the ninth stand
was not statistically different at the 15% level, it was much
higher than the other stands in absolute terms. This may be due
to unusual stand characteristics also reflected in its small size.
Bloodroot is known to respond to increased sunlight through
increased clonal growth (Marino et al., 1997), and the stand
was located in a patch of sunlight with little overstory cover,
possibly the result of a treefall gap.
Mean petiole diameter of the bloodroot stands is significantly
negatively correlated with average overstory tree dbh within the
range of conditions where bloodroot was found. It is important
to note that excluding the outlier population, all of the stands
are located within areas where average tree dbh is between
27.38 and 36.17 cm. Bloodroot is not found growing naturally
in large open fields or in very young stands with a low mean
tree dbh and lots of competitive herbaceous cover. Preliminary
literature review and canvassing of possible research sites indi-
cated that bloodroot is not found in young stands or dry sites.
Rather, typical sites for bloodroot are medium-sized hardwood
stands on northern aspects. As the mean dbh of mature stands
increases, bloodroot growth may decline. One explanation for
this is that an older mature overstory of large hardwood trees
allows more light to reach the forest floor and increases the
competition from other understory plants. It is also possible that
increased tree growth results in limited nutrients, which may in
turn lead to a decline in bloodroot growth rates.
Allometric Volume Equations
The relationship between easily recognized aboveground
characteristics (petiole height and petiole diameter) and below-
ground biomass of rhizomes was analyzed to assess the feasibil-
ity of restricting harvest to the largest rhizomes. Regression
analyses based on the 174 harvested rhizomes and their corre-
sponding aboveground plant parts reveal that rhizome biomass
can be predicted fairly accurately with two variables: petiole
height (cm) and petiole diameter (cm) (R2 = 0.77; p < 0.0001).
The regression equations demonstrate that harvesters can
estimate belowground biomass based on the observable above
ground plant characteristics of petiole height and petiole diameter.
Regional Bloodroot Sustainability
In a preliminary reconnaissance made when selecting the
Waynesville study site, four additional forests in Western North
Carolina were visited, including a national forest research site
and private and community forests. This confirmed that rich
cove sites provide excellent habitat for bloodroot.
Field observations and a few sample plots were taken in one
of the sites that had been open to bloodroot harvests and subject
to periodic timber harvests. Bloodroot on these lands was
scarce, even on rich cove sites comparable to those sampled in
the Waynesville watershed. The plants had comparable mean
petiole heights to those on the Waynesville watershed, but stand
densities were notably reduced.
Thus, our inventory may provide an upper bound on blood-
root distribution and abundance, since the Waynesville site is
managed under a conservation easement and not subject to
bloodroot harvest. Based on our canvass of potential sites, the
mean density of harvested stands was less than one plant per
square meter compared to 7.2 plants per square meter in the
Waynesville watershed. If the Waynesville high density stand
outlier is removed, the mean stand density drops to 3.7 plants
per square meter, still significantly larger than the harvested
stands. Due to the similarity in forest types and bloodroot plant
sizes, it is likely that this discrepancy in stand density is due to
harvesting pressure. This in turn suggests that some regulation
of harvest (whether by law or by community norms) and/or
active management such as enrichment plantings will be re-
quired to prevent further decline of bloodroot populations.
According to work on the ecological zones of the Southern
Appalachian Mountains by Simon et al. (2005), 12 percent
(695,000 acres) of the Southern Appalachian Mountains are in
Northern Hardwood or Rich Cove ecological zones. If the
Waynesville watershed can be considered representative of
Southern Appalachia, then we can extrapolate from the occur-
rence of bloodroot in 4.2% of the suitable habitat in Waynes-
ville to estimate that there are 56,000 acres of bloodroot in the
Southern Appalachian Mountains. This represents less than one
percent of the total forest area. If the 1,772,000 acres of Mesic
Oak-Hickory (Simon et al., 2005) are added to the total area of
suitable habitat, the total potential bloodroot area would be
about 197,000 or 3.4 percent of total forest area, although this is
likely an over-estimate, because we found only one bloodroot
stand in this forest type.
Furthermore, anecdotal evidence indicates that bloodroot
may be more common in the Waynesville watershed than in the
region as a whole, suggesting that the total area occupied by
bloodroot in Southern Appalachia is even less than projected
above. Sampling more stands under a variety of ownership and
management types would shed further light on whether the
Waynesville stands are representative and whether diminished
stand density is typical of sites that are open to harvest. Last we
offer a few observations about the natural history of bloodroot
and forest management practices. Bloodroot clearly favors deep
rich soils on north and east facing slopes and rich cove sites, as
well as soils that are not too acidic, which leads to domination
by rhododendron. Most of the Waynesville watershed has this
combination of aspect and soils. Furthermore, though not well
documented, our results suggest that bloodroot thrives under a
certain age, species, and density of forest overstory. Most of the
forest in the Waynesville watershed is about 80 - 85 years old
with a general oak-hickory-poplar forest composition, and a

J. FURGURSON ET AL.
218
moderate stand density. The few additional areas where we
observed bloodroot were mature stands with trees 60 to 100
years old, which prevented lush vegetation on the forest floor,
thus allowing the less competitive bloodroot to prosper. Recent
clearcuts and young stands with trees 30 - 50 years old had no
observable bloodroot at all. The dense vegetation of pioneer
species and young saplings resulting in nearly complete forest
floor cover may prevent bloodroot survival. This implies that
older stands of oak-hickory-poplar on good sites are required to
nurture bloodroot, which in turn suggests that landowners face
a trade-off between timber and non-timber production: tracts of
valuable old growth timber should be left unharvested where
conservation of bloodroot is a high priority.
Acknowledgements
We would like to thank the City of Waynesville for allowing
us to conduct our research on their watershed. We would also
like to thank the USDA Forest Service Health Monitoring Pro-
gram. We are grateful to the Revitalization of Traditional
Cherokee Artisan Resources (RTCAR) and the Community
Forestry & Environmental Research Partnerships for funding.
We thank Seth Holling for assistance in the field and Drs.
Jeanine Davis and David Danehower for valuable discussion
and insight.
REFERENCES
Bond-Lamberty, B., Wang, C., & Gower, S. T. (2002). Aboveground
and belowground biomass and sapwood area allometric equations for
six boreal tree species of northern Manitoba. Canadian Journal of
Forest Research, 32, 1441-1450. doi:10.1139/x02-063
Cech, R. (2002). Growing at-risk medicinal herbs: Cultivation, con-
servation and ecology. Williams, OR: Horizon Herbs.
Chamberlain, J. (2004). Special forest products: A southern strategy for
research and technology transfer. Washington DC: USDA Forest
Service Publication.
Conservation Trust for North Carolin (2005). Waynesville’s clean
drinking water protected. Raleigh, NC: Conservation Trust for North
Carolina.
Greenfield, J., & Davis, J. M. (2004). Medicinal herb production guide:
Bloodroot (Sanguinaria canadensis L.). Raleigh, NC: North Carolina
Consortorium on Natural Medicines and Public Health.
Hayden, L. (2005). The role of herbalism in the loss of native plants of
the northeast. Rhode Is la nd Na tur alist, 12, 1-3.
Kerns, B. K., Liegel, L., Pilz, D., & Alexander, S. J. (2002). Biological
inventory and monitoring. In E. T. Jones, R. J. McClain, & J. Wei-
gand (Eds.), Non-timber forest products in the United States. Law-
rence, KS: University Press of Kansas.
Marino, P. C., Eisenberg, R. M., & Cornell, H. V. (1997). Influence of
sunlight and soil nutrients on clonal growth and sexual reproduction
of the understory perennial herb Sanguinaria canadensis L. Journal
of the Torrey Botanical Society, 124, 219-227. doi:10.2307/2996609
Naud, J., Olivier, A., Belanger, A., & Lapointe, L. (2009). Medicinal
understory herbaceous species cultivated under different light and
soil conditions in maple forests in Southern Quebec. Agroforestry
Systems, 79, 303-326.
Nihal, A., Gupta, S., & Husain, M. M. (2000). Differential antiprolif-
erative response of sanguinarine for cancer cells versus normal cells.
Clinical Cancer Research, 6, 1524-1528.
Ott, R. L. & Longnecker, M. (2001). An introduction to statistical
methods and data analysis (5th ed.). Duxbury: Thomson Learning.
Persons, W. S., & Davis, J. M. (2005). Growing and marketing ginseng,
goldenseal, and other woodland medicinals. Fairview, NC: Bright
Mountain Books.
Peters, C. M. (1994). Sustainable harvest of non-timber plant resources
in tropical moist forest: An ecological primer. Biodiversity Support
Group.
Predny, M. L., & Chamberlain, J. L. (2005). Bloodroot (Sanguinaria
canadensis): An annotated bibliography. US Department of Agri-
culture, Forest Service, Southern Research Station.
Shackleton, S., Shackleton, C., & Shanley, P. (Eds.) (2011). Non-tim-
ber forest products in the global context. Tropical Forestry, 7, 3-21.
doi:10.1007/978-3-642-17983-9_1
Simon, S. A., Collins, T. K., Kauffman, G. L., McNab, W. H., & Ulrey,
C. J. (2005). Ecological zones in the Southern Appalachians: First
Approximation. USDA Forest Service Southern Research Station.
Ticktin, T. (2004). The ecological implications of harvesting non-tim-
ber forest products. Journal of Applied Ecol o g y, 41, 11-21.
doi:10.1111/j.1365-2664.2004.00859.x