HomeMy WebLinkAbout20080868 Ver 2_Report_20080101NONRIVERINE WET HARDWOOD FORESTS IN NORTH CAROLINA
STATUS AND TRENDS
Michael P. Schafale, North Carolina Natural Heritage Program
January 2008
INTRODUCTION
Nonriverine Wet Hardwood Forests are among the most threatened of North Carolina's natural
communities, and in some ways among the least well known. Also called oak flats, they were
once widespread in the outer Coastal Plain of northeastern North Carolina, but were long ago
reduced to a small fraction of their presettlement abundance. Today few citizens of North
Carolina have seen and appreciated this part of the state's natural heritage.
Definition and Description
Nonriverine Wet Hardwood Forests, as defined in Schafale and Weakley (1990), are wetland
forests of poorly drained, mineral soils on broad interstream flats. They correspond to the
Quercus michauxii-Quercus pagoda/Clethra alnifolia-Leucothoe axillaris Forest and Quercus
laurifolia-Nyssa biflora Forest associations of the International Classification of Ecological
Communities (NatureServe 2007). They would be classified as type 91, Swamp Chestnut Oak-
Cherrybark Oak in the Society of American Foresters system, where they represent a small
minority amid the more common bottomland hardwoods along rivers (Eyre 1980).
Nonriverine Wet Hardwood Forests are naturally dominated by some of the same trees as
bottomland hardwood forests along large brownwater rivers: swamp chestnut oak (Quercus
michauxii), laurel oak (Quercus laurifolia), and cherrybark oak (Quercus pagoda). Water oak
(Quercus nigra), sweetgum (Liquidambar styraciflua), loblolly pine (Pinus taeda), red maple
(Acer rubrum), and tulip poplar (Liriodendron tulipifera) have increased with past logging and
are often abundant. Unlike the canopy, the understory, shrub, and herb layers consist primarily
of plants shared with pocosins and nonriverine swamp forests, with some shared with blackwater
river floodplains, but only the most widespread species also shared with brownwater rivers. The
most typical understory trees are red bay (Persea palustris), red maple (Acer rubrum), and
ironwood (Carpinus caroliniana). Common shrubs are sweet pepperbush (Clethra alnifolia),
evergreen dog hobble (Leucothoe axillaris), and cane (Arundinaria gigantea ssp. tecta). The
dominant herbs are netted chain fern (Woodwardia areolata), Virginia chain fern (Woodwardia
virginica), and royal fern (Osmunda regalis). Peat moss (Sphagnum spp.) is usually present in
small amounts. There is natural variation in composition clearly related to wetness, with swamp
black gum (Nyssa biflora) and laurel oak increasing in wetter sites and swamp chestnut oak and
cherrybark oak increasing in less wet sites. Variation in the amount of ironwood versus red bay,
and in sweet pepperbush versus fetterbush, may be related to soil base status or fertility, as may
be the presence of unusual species such as shagbark hickory (Carya ovata) and Shumard oak
(Quercus shumardii) in a few examples. Variation in amount of red maple, sweetgum, and pine
appears to relate primarily to logging history.
Animals include widespread species such as white-tailed deer, black bear, and gray squirrel. The
large oak component makes Nonriverine Wet Hardwood Forests excellent habitat for wild
turkeys. The multi-layered structure characteristic of mature Nonriverine Wet Hardwood Forests
supports high densities and diversities of neotropical migrant birds such as wood thrush,
ovenbird, Swainson's warbler, worm-eating warbler, prothonotary warbler, hooded warbler,
white-breasted nuthatch, and the Coastal Plain black-throated green warbler. In the outer Coastal
Plain, where large river floodplains with bottomland hardwoods are absent, the once-extensive
Nonriverine Wet Hardwood Forests may once have supported much larger populations of these
species than now occur in this region. Invertebrates of these communities have not been studied,
but it is likely that a suite of insects specialized to feed on oaks and a suite of soil organisms
adapted to the unique hydrological conditions are present.
In contrast to bottomland hardwood forests along rivers, wetland conditions in Nonriverine Wet
Hardwood Forests are caused by seasonal high water tables and limited runoff of rainfall, due to
flatness and natural absence of streams. Rheinhardt and Rheinhardt (1998a) found that soil
drainage class in this type did not correlate with soil texture as is common in many places, but
was more subject to topography and landscape position. However, the continued wetness of
small remnant sites despite the drainage in the surrounding landscape suggests that perching of
water by impermeable soils may be partly responsible for the wetness in some sites. The soil is
generally saturated or flooded with a few inches of water through most winters and well into the
early summer, and the lower soil probably remains moist through most summers. The water
never gets as deep as it may in river floodplains, but the soil undoubtedly stays saturated longer
than in bottomland hardwoods. Furthermore, no additional nutrients are brought in by flowing
water, and aquatic animals cannot move in from the river during flooded times. Thus, debris
processing and nutrient cycling are likely very different from floodplain communities.
A number of other natural community types occur on wet nonriverine flats and share some
characteristics with the Nonriverine Wet Hardwood Forest type. Most similar are Nonriverine
Swamp Forests, which are wetter and lack oaks, but share some of the shrubs and herbs. Mesic
Mixed Hardwood Forests contain some of the same tree species but are drier than Nonriverine
Wet Hardwood Forests and have beech (Fagus grandifolia) as a major component. Often the
centers of nonriverine flats are so wet that organic matter has accumulated, burying the mineral
soils. These peatlands support either pocosin communities (Low Pocosin, High Pocosin, or Pond
Pine Woodland, or Bay Forest), Nonriverine Swamp Forests, or Peatland Atlantic White Cedar
Forests (all names from Schafale and Weakley 1990). Fires were an important part of the natural
dynamics of the pocosin and white cedar communities. Fire is believed to have been much less
frequent in Nonriverine Wet Hardwood Forests, due to the limited flammability of the leaf litter
and lack of continuous live fuel layers, but they certainly would have burned with low intensity
surface fires at times. Fire has not been believed to be important to Nonriverine Wet Hardwood
Forests or Nonriverine Swamp Forests, but Rheinhardt and Rheinhardt (1998a) suggest it may
play a significant role in maintaining dominance of oaks over other hardwoods.
A typical natural landscape pattern on the largest nonriverine flats is a complex of peatland
communities in the center of the flat, with a fringe of Nonriverine Wet Hardwood Forest where
the peat gives way to wet mineral soils, then a band of upland communities on the gentle slopes
closer to the streams, then stream swamps and tidal swamps and marshes along the drainages.
On some of the smaller flats farther inland, no peat may be present, and Nonriverine Swamp
2
Forest and Nonriverine Wet Hardwood Forest on mineral soils may be in the center of the flat.
Because the easiest lands to drain and convert to other uses are the least wet and those close to
slopes, the Nonriverine Wet Hardwood Forests were generally among the first wetlands to be put
into agriculture and later intensive silviculture.
The primary range of Nonriverine Wet Hardwood Forests is northeastern North Carolina. They
range from Craven County north into the southeastern counties of Virginia. None are definitely
known south of North Carolina or north of Chesapeake Bay. Although one example was known
inland nearly to Tarboro, the vast majority of acreage was, and is, on the outermost terrace of the
Coastal Plain, east of New Bern, Washington, and Plymouth.
Composition and Quality of Nonriverine Wet Hardwood Forests
Most early and more recent qualitative descriptions of Nonriverine Wet Hardwood Forests
describe them as being dominated by oaks. The only extensive quantitative study of Nonriverine
Wet Hardwood Forest composition is that of Rheinhardt and Rheinhardt (1998b). They
measured canopy and understory basal area and density in most of the known remaining
examples in North Carolina. They noted that, in contrast to earlier qualitative descriptions, most
stands were dominated by sweetgum, red maple, or tulip poplar. Oaks were abundant, but only
in a few places were they codominant.
Sweetgum, red maple, and tulip poplar are, ecologically speaking, weedy species, producing
abundant, small, widely dispersed seeds, and able to take advantage of disturbance much more
readily than oaks. It is clear that, although these native species have always been present, they
have increased in absolute and relative abundance as a result of logging. Thus, although the
precise composition of the earlier natural forests is not well known, examples with more oak are
believed to be closer to natural composition. This belief is supported by the abundance of oak
saplings in examples that contain a strong minority of oak in the canopy, suggesting that over
time without severe disturbance oaks will increase in the forest. The presence and abundance of
oaks therefore serves both as an indicator that a community is a Nonriverine Wet Hardwood
Forest rather than a Nonriverine Swamp Forest and as an indicator of its natural condition. In the
best remnants known, Rheinhardt and Rheinhardt (1998a, 1998b) found oaks to be 1.2% to 50%
of basal area and 1.5% to 42.9% of canopy stem density. Given current conditions, examples
with oaks comprising more than 10% of the basal area or of the canopy cover should be
considered good examples. They have the best potential to recover to natural oak abundance in
time, and are most likely to retain species associated with oaks.
Observations of areas clearcut in recent years indicate that the tree regeneration is primarily
weedy hardwoods or loblolly pine, with little or no oak component. Given the abundance of
weedy tree species and the scarcity of oaks in the landscapes where they once were abundant, it
is unlikely that oaks will ever again become abundant on these sites. Any animal species which
are dependent on oaks are presumably eliminated. Although all remaining stands with oak have
been logged in the more distant past, many probably by clearcutting, it appears that these
communities no longer have the ability to recover readily from clearcutting. The reason is likely
some combination of altered seed rain, the cumulative impact of repeated logging events, and
perhaps subtle changes in hydrology or fire regime. Therefore, Nonriverine Wet Hardwood
Forests that are clearcut at present must be considered lost. Those that are selectively cut may be
expected to recover if a substantial amount of oak is left in the stand. However, oaks are
generally the most desirable species for removal.
Besides abundance of oaks in the canopy and understory, other indicators of good condition in
Nonriverine Wet Hardwood Forests are canopy maturity, canopy age structure, extent, and
connection to other natural communities. The most mature examples known have many trees
16-24 inches in diameter, with some exceeding 36 inches. However, given the scarcity of these
communities, examples with trees averaging 12 inches in diameter are considered significant
examples. Even those with trees averaging 8 to 10 inches in diameter are significant if the
canopy composition is good and the example is extensive. As with most North Carolina
hardwood forests, the natural canopy is believed to be uneven-aged, with trees reproducing
primarily in small to medium canopy gaps that formed periodically from storms and possibly
fires, and with old trees abundant. In no remaining examples is this structure well developed, but
it can be expected to develop over time in the oldest examples. Examples with some canopy
gaps containing oak saplings, or with large old trees that will form canopy gaps in the near
future, will have more of the characteristics of natural forests than those with uniform younger
canopies. Nonriverine Wet Hardwoods naturally occurred in large patches, and some aspects of
ecosystem function probably depend on large extent. Therefore, large examples are more likely
viable and are more significant than small examples.
CURRENT STATUS AND TRENDS IN NONRIVERINE WET HARDWOOD FORESTS
Methods
The North Carolina Natural Heritage Program first started recording occurrences of Nonriverine
Wet Hardwood Forest in the 1980s. The most concentrated survey work occurred as part of the
Albemarle/Pamlico Estuarine Study (Frost, LeGrand, and Schneider 1990; LeGrand, Frost, and
Fussell 1992), which covered all of the range of Nonriverine Wet Hardwood Forest in North
Carolina. Additional examples have been added continuously to the database as they were
discovered. To assess the current status of the community type, all known examples were
checked against current aerial photography (2006 NAIP DOQQs). While Nonriverine Wet
Hardwood Forests cannot be definitively identified on aerial photography, the loss of known
stands by clearcutting or conversion can usually be recognized with confidence. Acreages were
estimated and ratings (EO ranks) were updated. The EO ranks are based on a combination of
condition, size, and landscape context, using the criteria described above. Condition, primarily
based on stand maturity and composition, was assumed to be the same as initially described, but
size and landscape context were re-evaluated using the aerial photos.
To assess trends, the current status was compared to two previous times when a reasonably
complete picture of the status of the community type was determined. A study by Rheinhardt
and Rheinhardt (1998b) reviewed the status of all known sites. They attempted to sample
vegetation at every known example, and examined aerial photos and consulted with foresters to
determine the status of examples they were unable to visit. In the few cases where Rheinhardt
and Rheinhardt's report did not give the condition of the site, I attempted to determine its status
4
by personal communication and by consulting aerial photos. Sites not confirmed to be destroyed
were assumed extant, possibly resulting in an overestimate of the amount remaining. The results
of this comprehensive picture were recorded at that time.
To obtain an earlier picture of the status of the community type, I reconstructed the status of
Nonriverine Wet Hardwood Forest around 1990 by estimating acreage for all the examples we
have evidence of having existed at that time. This reflected the Albemarle/Pamlico Estuarine
Study surveys, for which most field work was conducted in 1989 and 1990. More of the known
occurrences were found during these studies than at any other single time. The occurrences were
recorded in the Natural Heritage Program database when the studies were completed, and some
occurrences were visited and the records updated over the years. This time is labeled 1990, but it
really represents a range of last observation dates from 1990 through 1998, with information on
the state before the most recent data no longer readily accessible.
It is clear that not all of the Nonriverine Wet Hardwood Forests were found in the 1990s.
Several additional sites have been discovered. In addition, the extent of the community in some
previously known sites has become better known more recently. To make the analysis of trends
as accurate as possible, newly discovered sites and more accurate estimates of acreage were used
to adjust the earlier data. Thus, sites discovered since 1998 were added to the figures for 1990
and 1998, since they clearly existed at those times as well. Because these sites may have been
reduced in size between 1990 and 2007, the figures for acreage lost in that time period are an
underestimate. Some additional sites likely were destroyed in that time period without ever
having become known to the Natural Heritage Program, making the trend figures more of an
underestimate. Sites that were thought to be Nonriverine Wet Hardwood Forest at the time and
were later determined to be other community types were not included for any of the times.
To get an indication how much Nonriverine Wet Hardwood Forest may have once been present,
I analyzed digitized soil survey maps for two sample counties: Hyde and Currituck (USDA-
NRCS 1996 and 1997). Acreage of all soil series believed to support Nonriverine Wet
Hardwood Forest in the past was determined by GIS. The list of soils to include in the acreage
included those series mapped under known remnant Nonriverine Wet Hardwood Forest
occurrences, plus a few closely related series that are not known to support other natural
community types. Series were rated as high (most supporting this community in natural
remnants), medium (supporting this community but also substantial amounts of other natural
community types), and low (supported this community only in small patches that probably
represent inclusions). Series with high and medium potential are listed in Table 3. Acreage
totals were calculated for high rated soils alone and for high and medium rated soils together.
The former can be expected to be a serious underestimate of the amount of Nonriverine Wet
Hardwood Forest once present, the latter an overestimate.
Results
Table 1 lists the known remaining sites, with their estimated acreage of Nonriverine Wet
Hardwood Forest and current EO rank. Map 1 shows their distribution. A total of 25 separate
extant sites are known to the Natural Heritage Program. These occurrences total approximately
5576 acres. Of these, three sites have EO ranks of A, seven have rankings of B or possibly B
5
(BC), 14 have rankings of C, and one is too little known to rank. Only eight sites are 100 acres or
larger. Though patches of this community once covered thousands of acres, only two remnants
are as large as 500 acres. Of the 25 sites, seven have some kind of land status that may protect
them from future destruction. Only one of these, however, is of substantial size. About 600
acres total are protected.
In the study period, the total acreage of Nonriverine Wet Hardwood Forest went from 35 sites in
1990 to 25 sites in 2006. The area went from 13,885 acres in 1990 to 7907 acres in 1998 to 5576
acres in 2006. This represents an acreage decline of 43% in the first 8-year period and 42% in
the next eight years. Of the 35 occurrences known to have existed in 1990, nine (26%) are
completely destroyed and another 13 (38%) were reduced in acreage, most by more than half.
Six sites (17%) are known not to have declined. The other seven sites (20%) were newly
discovered since 1990, and may or may not have been larger in 1990. Some of the remaining
sites had selective cutting or other damage that reduced their EO rank. The loss has been
particularly heavy for large occurrences. Of six occurrences of 500 acres or more in 1990, only
two remain over 500 acres. There were 19 occurrences of at least 100 acres in 1990, and now
there are only six.
Based on soil mapping, it appears that Hyde County once had between 34961 and 50,586 acres
of Nonriverine Wet Hardwood Forest. Currituck County had 15,317 to 43,941 acres. The most
extensive of the medium-rated series, Roanoke, probably supported Nonriverine Wet Hardwood
Forest on a majority of its acreage, but is also known to support significant amounts of Mesic
Mixed Hardwood Forest. It is likely that the true estimate, at least for Currituck County, is
closer to the higher figure. This is more difficult to evaluate for Hyde County, where several
abundant soil series were not mapped in other counties.
Discussion
The results indicate a community type in serious decline. While there is high uncertainty in the
estimates of original acreage based on soils, this analysis shows the drastic loss of Nonriverine
Wet Hardwood Forest. Currituck County alone, one of the smallest counties in North Carolina,
once had several times, perhaps ten times, as much Nonriverine Wet Hardwood Forest as now
remains in the whole state. Much of the loss occurred long ago, as the most easily converted
lands were put into agriculture or repeatedly logged. However, recent losses are still
proportionally high. While comparable figures do not exist for other community types, this
appears to exceed that of virtually all other community types in North Carolina. In overall
portion lost from original extent, it is comparable to the losses of wet and mesic longleaf pine
savannas, and probably exceeds that of any other wetland community type. In percentage of
remaining examples unprotected and likely to be lost in the near future, it far exceeds longleaf
pine communities and virtually all other community types in the state.
This rapid decline comes from a unique combination of vulnerabilities. The mineral soils on
which they occur are more easily drained than the organic soils that cover much of the
nonriverine wet flats, and the drained soils are productive for agriculture and growing of planted
pines. Their occurrence around the edges of nonriverine wet flats, or in smaller flats, also makes
them easier to drain than most wetlands. Finally, their vegetation is very vulnerable to loss with
6
logging. More extreme wetlands often have a limited range of tree species that can regenerate if
a forest is logged, and are more likely to regenerate in the characteristic species of the natural
community. Nonriverine Wet Hardwood Forests rarely regenerate to the characteristic oak
species and tend to become stands of weedy tree species that show little tendency ever to return
to an oak canopy. Although Nonriverine Wet Hardwood Forests are jurisdictional wetlands,
occurrences can be destroyed by common activities that are fully legal. In this situation, only
active land protection in the form of public acquisition or conservation easements is likely to
save any example in the long run.
The fate of land formerly supporting Nonriverine Wet Hardwood Forest varied. A few sites
appeared to have been developed or converted to cropland. A number were found to have been
converted to pine plantations. Many sites that had been recently clearcut at the time the aerial
photos were taken were likely converted to pine plantation later. Other sites presumably are
regenerating with weedy canopies. It is likely that lands in the last category could be restored,
and a few might recover spontaneously. Some pine plantations may also have good restoration
potential, but common practices of bedding, fertilization, and herbicide treatment mean that less
of the characteristic flora and local fauna can be expected to remain present.
Given the limitations of Natural Heritage inventories, some additional examples of Nonriverine
Wet Hardwood Forest probably remain undiscovered. During the eight years after 1990, four
new occurrences were found. Since 1998, six new examples have been discovered, and a couple
other examples have been found to be larger than realized. There is therefore some hope that
additional occurrences will be found, but these are unlikely to be anywhere as large individually
or collectively as the sites now known. Without substantial effort at protecting remaining
examples, the expected trend for this endangered community is continued rapid decline.
REFERENCES
Eyre, F.H., ed. 1980. Forest Cover Types of the United States and Canada. Soc. Am. For.
Washington, D.C. 148 pp.
Frost, C.C., H.E. LeGrand, Jr., R.E. Schneider. 1990. Regional Inventory for Critical Natural
Areas, Wetland Ecosystems, and Endangered Species Habitats of the Albemarle-Pamlico
Estuarine Region: Phase 1. Report to NC Natural Heritage Program. A/P Study Project
No. 90-01.
LeGrand, H.E., Jr., C.C. Frost, and J.O. Fussell III. 1992. Regional Inventory for Critical
Natural Areas, Wetland Ecosystems, and Endangered Species Habitats of the Albemarle-
Pamlico Estuarine Region: Phase 2. Report to NC Natural Heritage Program. A/P Study
Project No. 92-07.
NatureServe. 2007. International Ecological Classification Standard: Terrestrial Ecological
Classifications. NatureServe Central Databases. Arlington, VA. U.S.A. Data
current as of December 2007.
7
Rheinhardt, M.C., and R.D. Rheinhardt. 1998a. Canopy and woody subcanopy composition of
wet hardwood flats in eastern North Carolina and southeastern Virginia.
Manuscript submitted to Journal of the Torrey Botanical Society.
Rheinhardt, M.C., and R.D. Rheinhardt. 1998b. Canopy and Woody Subcanopy Composition of
Nonriverine Wet Hardwood Forests in Eastern North Carolina. Report to North
Carolina Natural Heritage Program.
Schafale, M.P., and A.S. Weakley. 1990. Classification of the Natural Communities of North
Carolina. Third Approximation. North Carolina Natural Heritage Program,
Division of Parks and Recreation.
U.S. Department of Agriculture-Natural Resources Conservation Service. 1996. Detailed
County Soils - Currituck County, North Carolina. Digital soil map. USDA-
NRCS, Raleigh, NC.
U.S. Department of Agriculture-Natural Resources Conservation Service. 1997. Detailed
County Soils -Hyde County, North Carolina. Digital soil map. USDA-NRCS,
Raleigh, NC.
Table 1. Remaining Nonriverine Wet Hardwood Forest sites in North Carolina. EO ranks are
from the December 2007 Natural Heritage Program database. A is excellent, B very good, C
fair, E uncertain, X destroyed. Size figures are in acres.
site name county EO
rank size protection status
Bonnerton Road Wet Hardwood Forest and Seeps Beaufort B 198
Drinkwater Creek Wet Hardwood Forest Beaufort C 130
Hell Swamp Beaufort C 44
Jackson Swamp Remnants Beaufort C 20
Sparrow Road Wet Hardwood Forest Beaufort B 40
Roquist Pocosin Bertie A 500 State Ecosystem Enhancement Program
wetland mitigation land
Whitehall Shores Hardwood Forest Camden C 30
Gum Swamp Bottomland Hardwoods Craven B 40 National Forest, Special Interest Area,
Registred
Sea Gate Woods Craven C 85 Land trust preserve
Buckskin Creek/Great Swamp Currituck C 73
Lower Tull Creek Woods and Marsh Currituck C 30
Indiantown Creek/North River Cypress Forest Currituck B 30 State Game Land, Dedicated
Gibbs Woods/Tull Bay Marshes Currituck C 135
Troublesome Point/Gibbs Point Forests and
Marshes Currituck C 10
Alligator River/Swan Lake Swamp Forest Dare C 15 National Wildlife Refuge
Great Dismal Swamp National Wildlife Refuge Gates CD 5 National Wildlife Refuge, Registered
Scranton Hardwood Forest Hyde A 3580
Gull Rock Game Land Hyde BC 20 State Game Land
South Prong Natural Area Pamlico B 60
Light Ground Pocosin Southeast Section Pamlico C 55
Little Flatty Creek Forests Pasquotank C 40
Big Flatty Creek Forests and Marshes Pasquotank E 50
Belvoir Carolina Bays and Flats Pitt C 30
Bethel/Grindle Hardwood Flats Pitt A 216
East Dismal Swamp Washington C 125 State Agricultural Experiment Staton
Palmetto-Peartree Swamp Forest Washington C 15
total 25 sites 12 counties 5576
9
Table 2. Change in Nonriverine Wet Hardwood Forest from 1990 to 1998. EO ranks are based on 1998
condition.
site name county EO rank 1990
size 1998
size 2006
size status change since discovery
Bonnerton Road Wet Hardwood Forest and Seeps Beaufort B 198+ 198 198 newly discovered in 2000s
Drinkwater Creek Wet Hardwood Forest Beaufort C 130 130 130 newly discovered in 2000s
Hell Swamp Beaufort C 44 44 44 newly discovered in 2000s
Jackson Swamp Remnants Beaufort C 165 165 20 part destroyed in 2000s
Sparrow Road Wet Hardwood Forest Beaufort B 40 40 40 newly discovered in 2000s
Roquist Pocosin Bertie A 500+ 500+ 500 part destroyed in 1990s, part
newly discovered in 2000s
Whitehall Shores Hardwood Forest Camden C 100 100 30 part destroyed in 2000s
Forest Wet Hardwood Forest Craven x 138 0 0 destroyed in 1990s
Gum Swamp Bottomland Hardwoods Craven B 40 40 40
Sea Gate Woods Craven C 280 85 85 part destroyed in 1990s
Buckskin Creek/Great Swamp Currituck C 420 100 73 part destroyed in 1990s and
2000s
Lower Tull Creek Woods and Marsh Currituck C 120 30 30 part destroyed in 1990s
Indiantown Creek/North River Cypress Forest Currituck B 30 30 30 affected by rising sea level
Gibbs Woods/Tull Bay Marshes Currituck C 135 135 135 selectively cut in 1990s
Maple Swamp Gordonia Forest Currituck x 40 0 0 destroyed in 1990s
Northwest Backwoods Currituck x 900 0 0 destroyed in 1990s
Troublesome Point/Gibbs Point Forests and Marshes Currituck C 40 40 10 part destroyed in 2000s, part
selectively cut in 1990s
Alligator River/South Lake Swamp Forest Dare C 15 15 15 newly discovered in 1990s
Mildred Wet Hardwood Flat Edgecombe x 40 0 0 destroyed in 1990s, restoration
occurring
Black Mingle Pocosin Gates x 150 0 0 destroyed in 1990s
Great Dismal Swamp National Wildlife Refuge Gates CD 5 5 5 heavy blow down has degraded
Scranton Hardwood Forest Hyde A 5700 4100 3580 part destroyed in 1990s and
2000s
Gull Rock Game Land Hyde BC 20 20 20
South Prong Natural Area Pamlico B 100 70 60 part destroyed in 1990s and
2000s
Light Ground Pocosin Southeast Section Pamlico C 60 60 55 part destroyed in 1980s and in
2000s
Merritt Hardwoods Pamlico x 1400 900 0 last remnant destroyed in 2000s
Little Flatty Creek Forests Pasquotank C 185 40 40 part destroyed in 1990s
10
Big Flatty Creek Forests and Marshes Pasquotank E 1500 300 50 part destroyed in 1990s and
2000s
Menzies Pond Perquimans X 20 20 0 destroyed in 2000s
Belvoir Carolina Bays and Flats Pitt B 85 30 30 part destroyed in 1990s
Bethel/Grindle Hardwood Flats Pitt AB 1080 530 216 part destroyed in 1990s and
2000s
Lewis Point Swamp Forest Tyrrell X 15+ 15 0 destroyed in 2000s
Highway 99 Nonriverine Hardwood Flat Washington X 50 25 0 part destroyed in 1990s, rest in
2000s
East Dismal Swamp Washington C 125 125 125 part damaged by logging
Palmetto-Peartree Swamp Forest Washington C 15 15 15 newly discovered in 2000s, may
be larger
total 35 sites 15 counties, 12
remaining 13885 7907 5576
Table 3. Soil series of Nonriverine Wet Hardwood Forest occurrences and additional series likely to
have supported Nonriverine Wet Hardwood Forest. Rating is an estimate of the fraction of the soil
series acreage that would have supported Nonriverine Wet Hardwood Forest.
series taxonomy sites rating
Acredale Typic Endoaqualf Scranton high
Arapahoe Typic Humaquept Merritt Hardwoods low
Argent Typic Endoaqualf Light Ground Pocosin Southeast, South Prong, Scranton Hardwoods high
Brookman Typic Umbraqualf Scranton med.
Cape Fear Typic Umbraquult Highway 99 Hardwood Flats, Belvoir Carolina Bays high
11
Deloss Typic Umbraquult Sea Gate Woods low
Hydeland Typic Umbraqualf Scranton med.
Leaf Typic Albaquult Roquist Pocosin, Gum Swamp low
Pantego Umbric Paleaquult Gum Swamp low
Pasquotank Typic Endoaqualf high
Portsmouth Typic Umbraquult Jackson Swamp Remnants low
Roanoke Typic Endoaquult Whitehall Shores, Sea Gate Woods, Gibbs Woods/Tull Bay,
Troublesome Point, Lower Tull Creek, Buckskin Creek/Great Swamp,
Big Flatty Creek, Little Flatty Creek, Menzies Pond, Bethel/Grindle,
East Dismal Swamp med.
Tomotley Typic Endoaquult Indiantown Creek, Lewis Creek high
Yonges Typic Endoaqualf high
12
Table 4. Amount of soils likely to have supported Nonriverine Wet Hardwood Forest in Hyde and
Currituck counties. Rating is an estimate of the fraction of the soil series acreage that would have
supported Nonriverine Wet Hardwood Forest.
series Hyde Currituck rating
Argent 9467 high
Acredale 12425 high
Brookman 7777 medium
Cape Fear 6104 medium
Hydeland 17848 medium
Pasquotank 590 1509 high
Roanoke 28624 medium
Tomotley 6104 high
Yonges 2479 high
13
total 50586 43941
total of high probability only 24961 15317
14
Map 1. Locations of Nonriverine Wet Hardwood Forest community occurrences remaining in North
Carolina in 2006.
15
SITE SURVEY REPORT FORM
NC Natural Heritage Program
SITE NAME: Bonnerton Road Wet Hardwood Forest and Seep
DATES VISITED: April 15, 2005, January 26, 2005
INVESTIGATORS: Mike Schafale, David Lekson (US Army Corps of Engineers), Ed Schwartzman
(Unique Wetlands Program), Julia Berger (CZR), Tom Steffan (CZR) (April)
Mike Schafale with David Lekson, John Dorney (DWQ), LeiLani Paugh (DOT), Ed Schwartzman,
Sandy Smith (EcoScience), Matt Cusak (EcoScience), and other members of the Wetland Functional
Assessment Team
REPORT AUTHOR: Mike Schafale
OWNER: PCS Phosphate
OWNER CONTACT: Contact made through CZR staff and David Lekson.
COUNTY: Beaufort QUAD: Aurora
SIZE: 203.08 acres total: 194.22 acres primary, 8.86 acres secondary.
HOW DETERMINED: GIS
LOCATION: On the east side of SR 1936 (Bonnerton Road), on the north side of 1958, extending
from the top of the Suffolk Scarp to the headwaters of Porter Creek. About 5 air miles northwest of
Aurora.
PROVINCE: Coastal Plain, Embayed Region WATERSHED: Pamlico River
GENERAL DESCRIPTION: This site includes a segment of the Suffolk Scarp and the adjacent wet
flats at its base. The flats have shallow standing water through much of the growing season, with a
narrow slough that is somewhat wetter. They support a Nonriverine Wet Hardwood Forest
community in very good condition. Both subtypes are present. The scarp is saturated by seepage all
the way to the top. It supports an unusual seepage wetland community dominated by water oak, red
maple, and loblolly bay, with a dense layer of evergreen shrubs. Porter Creek forms in the eastern
side of the site, with a small area of Cypress-Gum Swamp in the headwater wetland.
SIGNIFICANCE OF SITE: State. The extensive Nonriverine Wet Hardwood Forest communitiy in
excellent condition is one of the best remaining examples in the state. It is the fifth largest example
remaining in the state, and is in very good condition. The unclassified seepage community may also
potentially be one of the best examples in the state.
PHYSICAL DESCRIPTION
ASPECT: East on the scarp, flat below.
SLOPE: Moderate slope, probably about 10%.
ELEVATION: 15-30 feet.
TOPOGRAPHY: Open, undissected slope on the face of the scarp. Flat with a couple gentle linear
1
swales below.
HYDROLOGY AND MOISTURE: Scarp face is semipermanently to permanently saturated. Flat is
seasonally saturated to shallowly flooded. The headwaters of Porters Creek are seasonally flooded.
The swales may drain to it, but appear to be primarily nonriverine.
GEOLOGY: Unconsolidated Holocene sediments
SOIL (from USSCS soil map):
Tomotley (Fine-loamy, mixed, thermic Typic Ochraquult): most of the wet flats.
Portsmouth (Fine-loamy over sandy, mixed, thermic Typic Umbraquult ): western fringe of the wet
flat.
Leon (Sandy, siliceous, thermic Aeric Haplaquod) on the seepy Suffolk Scarp face.
COMMENTS ON PHYSICAL DESCRIPTION:
NATURAL COMMUNITY DESCRIPTION
Nonriverine Wet Hardwood Forest (Oak Flat Subtype): This is the most extensive community within
the site, occurring on the wet flats. Shallow water was standing on a minority of the surface at this
visit. The canopy is dominated by a mix of Quercus laurifolia and Quercus michauxii, with some
Quercus phellos, Quercus nigra, Liriodendron tulipifera, Acer rubrum, Liquidambar styraciflua,
Ulmus americana, and Pinus taeda. The understory is dominated by Acer rubrum and oaks in parts,
Carpinus caroliniana elsewhere, with some Persea palustris and Ilex opaca. The shrub layer is
moderate in density. Leucothoe axillaris dominates some patches. Most parts have no strong
dominant. Species include Persea palustris, Symplocos tinctoria, Ilex coriacea, Vaccinium
formosum, Gaylussacia frondosa, and Sabal minor. Woody vines are fairly abundant. The herb layer
is generally moderate to sparse, with no strong dominants. The canopy is quite mature in large parts,
with trees averaging 12" dbh, 14" in some places. Many tree reach 18" dbh and some reach 24". The
northwest portions have canopy trees averaging 8-10" dbh, but with sparse old trees up to 24" dbh
present. Lonicera japonica is present in scattered places, but is mostly confined to the drier edges of
the community.
Nonriverine Wet Hardwood Forest (Oak-Gum Slough Subtype): Occurs as a couple of narrow bands
in low swales that are incipient drainages. These areas were flooded over 100% of the ground, and
more deeply than in the Oak Flat. The canopy is dominated by Nyssa biflora, Quercus laurifolia,
Quercus nigra, and Liquidambar styraciflua. The understory is sparse and consists mostly of canopy
species. Shrubs are sparse, though a few Vaccinium formosum are present. The herb layer is less
diverse and includes abundant Osmunda regalis. The canopy is quite mature in this area, with trees
averaging 12" dbh, trees up to 18" dbh common, and some up to 36" dbh.
Suffolk Scarp seepage community: The scarp here is a moderate slope with sandy soil with a muck
surface all the way to the top. There is an open canopy dominated by Quercus nigra and Acer
rubrum, with abundant Gordonia lasianthus. Oxydendrum arboreum occurs in the undestory along
with canopy species. Local Nyssa biflora and Chamaecyparis thyoides are present. The shrub layer
is dense, with Leucothoe axillaris and Ilex coariacea dominant. Symplocos tinctoria and Persea
palustris are common. Smilax laurifolia is abundant. Herbs are extremely sparse. Community
condition is difficult to judge. Logging maybe responsible for the open canopy, and pines may once
have been present. Lack of fire may also have affected the community, but its location adjacent to
the non-flammable Nonriverine Wet Hardwood Forest and downhill of former longleaf pine
2
communities would have limited fire frequency and intensity. The appropriate classification for this
community is unclear. It probably represents an undescribed type of seepage wetland. For now, it
will be classified as a Low Elevation Seep, a very broad category.
OTHER NATURAL COMMUNITIES PRESENT: Several acres of Mesic Mixed Hardwood Forest
is present on the south edge. It has some large Fagus grandifolia and Quercus nigra, up to 18-22"
dbh, is in excellent condition, but is too small to be highly significant for this community type. It is
heavily invaded by Lonicera japonica in parts.
About 10 acres of Cypress-Gum Swamp occur in the headwaters of Porter Creek. It is transitional to
the Oak-Gum Slough subtype, with only weak stream flow. The canopy is dominated by Nyssa
biflora, with some Acer rubrum and a few Quercus laurifolia.
ANIMAL HABITAT FACTORS
HABITAT HETEROGENEITY: Medium. Closed and open-canopy forest, with and without dense
shrub thickets. Artificial open areas are nearby.
AMPHIBIAN BREEDING SITES: Extensive standing water in the early growing season.
DENNING SITES: None noted.
BIG TREES/LARGE CAVITIES: Trees up to 20-22" dbh are present.
SNAGS AND LOGS: Moderate amounts, probably a bit less than natural levels.
MAST PRODUCING SPECIES: Oaks are dominant throughout. Some grapes and berry-bearing
shrubs. Black gums are common in parts.
NECTAR SOURCES: Ericaceous shrubs.
PRESENCE OF WATER: Abundant water except in dry seasons. A marginal creek is present in the
east side of the site.
AQUATIC HABITAT FACTORS: The headwaters of Porter Creek within the site have only a
marginally developed channel.
SPECIAL STATUS SPECIES PRESENT: None noted.
POTENTIAL FOR OTHER SPECIAL STATUS SPECIES: A Bald Eagle site is present within 2
miles to the west, and they may occasionally use the site. Wayne's black-throated green warbler
could potentially occur.
OTHER NOTEWORTHY SPECIES OR FEATURES PRESENT:
SITE ECOSYSTEM INTEGRITY: Fairly high. The canopy is mature, but the area is fairly small,
and small amounts of invasive plants are present. There is some drainage on adjacent lands.
AVERAGE DBH OF CANOPY TREES: 12" over most. 10" in some parts, 14" locally.
MAXIMUM DBH OF CANOPY TREES: 18-24".
DISTURBANCE-SENSITIVE SPECIES: None noted. Unknown if present.
FIRE REGIME: No sign of fire. The Nonriverine Wet Hardwood Forests are not very flammable
and probably seldom if ever burned. The vegetation on the scarp probably burned occasionally, but
3
its location adjacent to non-flammable vegetation except uphill probably limited fire frequency.
OTHER DISTURBANCES OR IMPACTS
LOGGING: The more mature oak flat areas have not been logged for many decades. The area on the
north side may have been selectively cut some years ago. Young successional stands developed after
clearcutting border the site. The scarp seepage community may have had selective logging not too
many years ago, but this is unclear.
FARMING: Not in site, but fields occur nearby. There is an abandoned farm house on the south
edge of the site.
DITCHES: Several small ditches cross the site. Larger ditches are present on some of the edges of
the site. The agricultural lands to the south are intensively ditched. A mining pit is present just a
short distance east of the site, but it is downstream and does not cross the outflow creek.
ROADS: Public roads border the site on two sides. A few old overgrown logging roads are present.
ALTERED FLOOD REGIME: Somewhat. The ditch on the south edge of the eastern unit appears to
be intercepting sheet flow into the site from the south.
EXOTIC/WEEDY SPECIES: Lonicera japonica is present in several places in the Nonriverine Wet
Hardwood Forest, and is abundant in the small Mesic Mixed Hardwood Forest.
UNDERSTORY CLEARING: No.
DIRECT HUMAN INTRUSION: Probably low. No sign of visitation.
LANDSCAPE FACTORS
BOUNDARY INTEGRITY/SHAPE:
ADJACENT LAND USE/OFFSITE STRESSES: Surrounded by pine plantations and logged forests,
with a few fields.
RELATION/CONNECTION TO OTHER SITES: Drinkwater Creek, with poorer Nonriverine Wet
Hardwood Forest, lies 4 miles to the east. Suffolk Scarp Bogs, with similar seepage wetlands, lies
7.5 miles to the south. Other small seepage wetlands occur along the Suffolk Scarp closer to this
site.
DEGREE OF THREAT/POTENTIAL FOR CHANGE: High. Without protection, the mature forest
is likely to be logged soon. The area is underlain by phosphate deposits, and may be mined in the
future.
BOUNDARY JUSTIFICATION: The boundary was drawn based on field survey and 1998 and 2003
DOQQs to include the intact wet hardwood and seep communities. The boundary is marked on all
sides by pine plantations and successional forests. The small area of young hardwoods that connects
the two primary areas is included as a secondary area.
RECOMMENDATIONS FOR PROTECTION: The area would be worthy of dedication, registry, or
acquisition for conservation.
MANAGEMENT RECOMMENDATIONS: No urgent management needs are apparent.
Honeysuckle should be controlled before invasion becomes more extensive. The secondary area, and
ideally some of the surrounding young stands, should be managed to restore Nonriverine Wet
Hardwood Forest.
NEED FOR FURTHER STUDY: Need more investigation into the character of the seepage
community. Need to study effect of adjacent ditches and drainage on adjacent lands.
4
PLANT SPECIES OBSERVED
Thoroughness of list: moderate
Wo = Nonriverine Wet Hardwood Forest (Oak Flat Subtype)
Ws = Nonriverine Wet Hardwood Forest (Oak-Gum Slough Subtype)
S = seepage community on scarp
M = Mesic Mixed Hardwood Forest
C = Cypress-Gum Swamp (Acid Blackwater Subtype)
canopy
Acer rubrum S, Wo
Chamaecyparis thyoides S
Fagus grandifolia M, (S, WO)
Fraxinus pensylvanica (Ws)
Gordonia lasianthus S
Liquidambar styraciflua Wo, Ws, M, S
Liriodendron tulipifera S, Wo
Nyssa biflora B, Ws
Pinus taeda Wo, (Ws)
Quercus laurifolia Wo, Ws, (C)
Quercus michauxii Wo, Ws
Quercus nigra S, Wo, M
Quercus phellos Wo
Ulmus americana Wo
understory
Acer rubrum Wo
Amelanchier Wo
Carpinus caroliniana Wo
Ilex opaca S, Wo, Ws, M
Lyonia lucida S, (Wo)
Nyssa biflora S
Oxydendrum arboreum S
Persea palustris Wo, M
Quercus michauxii Wo
shrub layer
Arundinaria gigantea Wo
Cyrilla racemiflora Wo
Euonymus americana Wo
Gaylussacia frondosa Wo
Hypericum hypericoides Wo
Ilex coriacea S, Wo
c
c
c
c
c
c
c
c
c
c
c
c
c
c
u
u
u
u
u
u
u
u
u
s
s
s
s
s
s
5
Leucothoe axillaris S, Wo s
Morella cerifera S s
Persea palustris S s
Sabal minor Wo s
Symplocos tinctoria S, Wo s
Vaccinium formosum Ws s
vines
Anisostichus capreolata Wo v
Berchemia scandens Wo, M v
Decumaria barbara Wo v
Gelsemium sempervirens Wo v
Lonicera japonica Wo, M v
Smilax laurifolia S v
Smilax walteri? Wo v
herb layer
Athyrium asplenioides Wo h
Atrichum sp. Wo h
Botrychium sp. Wo h
Carex spp. Wo h
Chasmanthium laxum Wo, M h
Climacium americanum Wo h
Hexastylis arifolia Wo h
Listera australis Wo h
Mitchella repens Wo h
Mnium sp. Wo h
Osmunda cinnamomea Wo h
Osmunda cinnamomea S h
Osmunda regalis Ws, Wo h
Polypodium Wo h
Sphagnum sp. Ws h
Tipularia discolor Wo h
Viola primulifolia? Wo h
Viola sp. Wo h
Vitis rotundifolia Wo h
Woodwardia areolata Wo h
ANIMAL SPECIES OBSERVED
Thoroughness of list: casual
Prothonotary warbler
Ovenbird
6
White-eyed vireo
Northern parula
Carolina wren
Stt' S Yi..' " '? {? r e!MI+ r. , .? f 7Z1, ,'? flW'["` ll F M4.
-Ike
'dp
1l,, x Y" a`? ??`? ,ra ...?. ?' q ? nr. }'?. .x
lis Fes; k•a", i 'S' x 1 ?!• kr't ?,o, + p a f w . ?' * J r:
.. -4'^?. al A•'Ir/n,?.,.?; ?.. ?A r .•M• 4 ;A r+V k ? p tP,.?? f w :`4
? ati,Ft}C ?.x, .a*i 'a ?w x ` yx,, . ? p ;' ,?•?vR?*x?? "`'a• .,' ?' ,?:= r.-' yam' I
? .: ,l R ?? ti ?,,,q. t`"t?e ,?? ?? t1 y ?f ? `"? - r , ya 'it.
$'
?? t i v i7x'? K? wy a r y K.q ?• w r Y` ? w4 d ?y,:"M'?Yb ±.t '
(" ,.+?, !" R. ,y? ??I i '? W V +Wb• ':, C .?.4y 'fir '.#r i .?.J'r5 . ?•7'I
d'-.? + /-"•'ii ...y „r./. ?11?.- •M'? C FS!?[n.?l' q.{ r'",1??,•? y ?• -?^-
Yrj rjf t. ?• ? 'k li, " ?k' _ r^ _ r }}y4U Ff f++?........
'f? ?p11 sy?4`> w ? •lb ? ? `:5. a av?? y?i? ? ? F \? ? ?y.F +u?N?r..
?,ys„ytli ,??w a'*+v.•? q'? A?4 ;.,.. ? ^i` -?:. - ?'?-.?yp,? rk ? sp• ti .,.,.,
?`. '?4p' . - ?,y•" `? -,?r?"G-t??xr+G ...+? X54.. ? H ? ,..?1./? ?
,?"' :: "?". ??. nt.• b? " uci `n {rte tk * ?? r
411
y
ee ` Li. - F h Y { s{?i?h?
Y+fi F ? , w? l ? T ,r. ? \ ! t rtiti'' ? r? .. 5 ,, ? sSa".`?3 f A, i - _.,?y,?
r"?•r', y ? a. l ` n *S?.`LEJ? h ? ? f r fiS? ;v T q ?-. i. 2'} ?v. 'z` '
•^'ty. ?"`i ?`W \ '?-j''?gll.-` t it`b.?' q+Y'.. '4 ?1 rya
-NI
-WO
.?AIr . -fit r,i 3.4'4 f S kVS- dy ?l ? ?? /
.. ti`s ?' f..k ? ' +?. 4 i? r .." "°` ,•e"p; fig, -
ev-
16
strr.w
S, ???? ?;,x'4iy7 k?"':ti a ,x''?a?"?t 7?ryf?i ,._ •l, r '+?Cr_
ra - D o.
d ? ? f'? ?4r , x ri ? *•m 1 ?'rqr"?•:.• r q, ,a yry"?.,.iv ?;??,,. ??` , ff +? M +e 'w.a".,s* ? +, ? «? `fir ?,
WIR
\
..'•?r?+?- .art _r y,k'; "`&,;,' ??* :v+y r ?? ? s' ?• ?..
A V,
PCS Bonnerton area.
S ? S •c ? oY. ?, AM,., d?c`?[a.-S ?o ?• C O h r?e c-?-?
?,eo,,,,.?